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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
277942d1-427a-4c50-b5b5-574bd0315946 | gravitational-laws-of-focus-of-attention | null | null | https://ieeexplore.ieee.org/document/8730418 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8730418 | Gravitational Laws of Focus of Attention | The understanding of the mechanisms behind focus of attention in a visual scene is a problem of great interest in visual perception and computer vision. In this paper, we describe a model of scanpath as a dynamic process which can be interpreted as a variational law somehow related to mechanics, where the focus of attention is subject to a gravitational field. The distributed virtual mass that drives eye movements is associated with the presence of details and motion in the video. Unlike most current models, the proposed approach does not estimate directly the saliency map, but the prediction of eye movements allows us to integrate over time the positions of interest. The process of inhibition-of-return is also supported in the same dynamic model with the purpose of simulating fixations and saccades. The differential equations of motion of the proposed model are numerically integrated to simulate scanpaths on both images and videos. Experimental results for the tasks of saliency and scanpath prediction on a wide collection of datasets are presented to support the theory. Top level performances are achieved especially in the prediction of scanpaths, which is the primary purpose of the proposed model. | ['Stefano; Gori', 'Dario; Melacci', 'Zanca', 'Marco'] | 2019-06-04 | null | null | null | ieee-transactions-on-pattern-analysis-and-7 | ['scanpath-prediction'] | ['computer-vision'] | [ 1.66750520e-01 1.36351645e-01 -2.55843848e-02 -7.70468339e-02
5.00894845e-01 -1.28467217e-01 4.59781885e-01 -1.94181189e-01
-4.82292950e-01 4.69252020e-01 2.34446060e-02 -8.15764815e-02
-2.74202675e-01 -2.99733013e-01 -8.69281232e-01 -7.70423830e-01
1.49684504e-01 9.18332860e-02 8.58964741e-01 -3.86171579e-01
9.88107264e-01 2.92848200e-01 -2.11967349e+00 -1.27259701e-01
7.85901308e-01 6.37151659e-01 9.67062116e-01 7.15819895e-01
4.68258224e-02 6.52199924e-01 -9.73732919e-02 -7.25142434e-02
1.26785085e-01 -7.14506745e-01 -6.65583551e-01 1.36543348e-01
3.00136745e-01 5.33331856e-02 -2.27021202e-01 1.29868245e+00
3.04197192e-01 4.33077812e-01 6.28894389e-01 -1.07132483e+00
-6.05016291e-01 -2.07230691e-02 -8.88230443e-01 9.65361059e-01
3.02173734e-01 1.13201022e-01 8.11458170e-01 -9.39159393e-01
8.81764174e-01 1.14961338e+00 1.56210838e-02 5.08367360e-01
-9.32498455e-01 1.07090168e-01 2.33533397e-01 7.79122114e-01
-1.25266445e+00 -1.91915706e-01 7.11713552e-01 -5.99569917e-01
6.25898123e-01 3.15447181e-01 1.08193922e+00 5.13761222e-01
9.61649418e-01 7.52546608e-01 8.08827043e-01 -5.61355710e-01
3.48648906e-01 4.69007313e-01 2.27192640e-01 5.13784528e-01
2.80838519e-01 1.25161901e-01 -7.94418216e-01 2.72572488e-01
8.93072546e-01 -1.38554052e-01 -4.80347782e-01 -6.11448407e-01
-9.07157123e-01 6.71190739e-01 4.70315188e-01 2.87172943e-01
-5.42145967e-01 -8.31203535e-02 -2.31374621e-01 -2.05812827e-01
4.88881320e-01 2.02063203e-01 4.24625762e-02 1.38432994e-01
-1.06458819e+00 3.83304060e-01 4.68177378e-01 8.14516783e-01
4.79807258e-01 3.41257378e-02 -1.49979949e-01 2.22531572e-01
5.84954500e-01 6.99845970e-01 7.30070353e-01 -8.78030598e-01
6.71852529e-02 4.34598565e-01 4.02811468e-01 -1.15667307e+00
-3.57514888e-01 -5.76530755e-01 -2.64782935e-01 5.97541094e-01
4.27841872e-01 1.17669143e-01 -6.16720378e-01 1.63174772e+00
4.05956864e-01 3.64456117e-01 -3.56972903e-01 1.36837590e+00
4.64618981e-01 7.18723476e-01 -1.57556877e-01 -5.58038414e-01
1.23006988e+00 -8.27167928e-01 -1.08122444e+00 -1.25899881e-01
4.93095145e-02 -6.91933215e-01 8.41998160e-01 2.48249158e-01
-1.53462529e+00 -6.88826323e-01 -6.75886095e-01 -5.64878546e-02
-9.76810902e-02 -1.20644227e-01 2.13907048e-01 4.85129841e-02
-1.49336505e+00 3.87322843e-01 -8.01088095e-01 -7.47830331e-01
5.96747175e-02 1.90789938e-01 1.41137704e-01 5.76523602e-01
-8.81004453e-01 1.29164720e+00 2.14939386e-01 1.49883494e-01
-6.56075180e-01 -5.00532091e-01 -4.60151136e-01 2.29161233e-01
8.61577224e-04 -6.95245326e-01 1.05480027e+00 -1.36244822e+00
-1.07770813e+00 8.40670526e-01 -8.23834002e-01 -5.92406988e-01
4.05568272e-01 -3.07261318e-01 -9.28930715e-02 3.32607627e-01
1.03063583e-01 7.40643799e-01 1.14460003e+00 -1.04932606e+00
-8.33388507e-01 -2.83160895e-01 -1.70955077e-01 6.83353424e-01
8.00626725e-02 1.07576646e-01 -3.40474099e-01 -1.88419506e-01
1.78212747e-01 -9.59103525e-01 -8.02515447e-02 3.63213159e-02
-9.00226980e-02 -1.52302742e-01 6.13707542e-01 -5.38958013e-01
1.32205892e+00 -1.93711007e+00 6.19742751e-01 -1.84667021e-01
2.78034538e-01 1.17739275e-01 2.18588740e-01 1.63186625e-01
-2.26520032e-01 -2.72357434e-01 1.44832104e-01 -3.07993814e-02
-4.66594249e-01 -2.20287338e-01 -5.39726138e-01 6.25247061e-01
1.27467766e-01 8.08208942e-01 -6.19856536e-01 -6.24891520e-01
2.35714912e-01 4.47613448e-01 -6.64192855e-01 -1.97311565e-02
-2.23111123e-01 5.93236446e-01 -5.57938933e-01 2.00601742e-01
7.59533823e-01 -3.01465780e-01 -2.75378942e-01 -2.53278594e-02
-8.00958097e-01 -2.11795084e-02 -1.05675912e+00 1.27551019e+00
2.03917488e-01 1.08130395e+00 -2.41191179e-01 -8.82784903e-01
5.85911870e-01 -3.48830409e-02 2.05307290e-01 -7.53248394e-01
3.30300808e-01 2.79788510e-03 4.03495103e-01 -8.07509005e-01
7.48479843e-01 9.28735435e-02 5.69757581e-01 3.42308998e-01
-1.95398927e-02 1.64691582e-02 3.09517920e-01 2.12095395e-01
4.09591615e-01 4.27558303e-01 2.40437299e-01 -6.12457335e-01
6.73567116e-01 9.24805999e-02 2.35973254e-01 5.35725653e-01
-3.30769151e-01 4.35397208e-01 3.98376197e-01 -3.72857362e-01
-1.04581916e+00 -7.49584317e-01 -2.10087299e-01 9.07725453e-01
9.02107298e-01 1.67658299e-01 -1.11738324e+00 2.05086306e-01
-3.32039058e-01 9.41521883e-01 -8.91206741e-01 -2.04004854e-01
-4.00461555e-01 -6.86677814e-01 -3.52817476e-01 2.66538002e-02
3.57866108e-01 -1.43148482e+00 -1.58048415e+00 -5.18277138e-02
-3.34398687e-01 -8.29450369e-01 -3.18392336e-01 -3.86746824e-01
-9.81962740e-01 -1.21619916e+00 -7.52374649e-01 -7.43186474e-01
6.51390791e-01 5.00639200e-01 8.68110120e-01 1.70440301e-01
-3.96400899e-01 5.45975626e-01 -5.35163805e-02 -8.16513300e-01
-1.07176393e-01 -3.79458368e-01 8.98993835e-02 4.68800962e-01
4.41361576e-01 -3.32731903e-01 -9.29421186e-01 3.33824694e-01
-6.81527615e-01 2.87482888e-01 3.36486161e-01 3.56014937e-01
6.04115605e-01 -1.10416375e-01 1.47405759e-01 -2.01137781e-01
3.70457828e-01 -5.25091946e-01 -1.09783745e+00 1.03508040e-01
-4.42359686e-01 1.50084086e-02 -4.85214032e-03 -3.90867680e-01
-1.24609280e+00 -1.87490761e-01 3.55479747e-01 -4.38546002e-01
2.25870358e-03 1.13520086e-01 2.87754864e-01 -2.72723824e-01
3.98297101e-01 6.15474880e-01 2.31246874e-01 -2.83856839e-01
-9.20107067e-02 1.14006490e-01 3.22060019e-01 2.05788419e-01
5.02096593e-01 7.19655633e-01 4.25360352e-01 -1.01073956e+00
-8.00154150e-01 -4.76525635e-01 -7.76672363e-01 -7.55542099e-01
1.10134804e+00 -5.87963164e-01 -9.90456104e-01 5.04712641e-01
-1.27509212e+00 -1.10805839e-01 -2.19158545e-01 8.09604526e-01
-8.83486211e-01 2.57739902e-01 -5.65998331e-02 -1.16944122e+00
-8.00952986e-02 -1.15914690e+00 7.65244424e-01 8.37993681e-01
1.16168819e-01 -1.23513329e+00 1.44863650e-01 -7.96397123e-03
4.18283820e-01 -7.61587992e-02 5.99231064e-01 -1.86597854e-01
-1.12184000e+00 2.08608970e-01 -7.22937658e-02 1.80938037e-03
-2.67565191e-01 3.05344105e-01 -7.77281761e-01 1.58153456e-02
6.87247217e-01 4.52826887e-01 7.01862931e-01 1.27434063e+00
6.88537896e-01 2.65397251e-01 -5.12782216e-01 5.48390865e-01
1.63335347e+00 4.11248922e-01 7.33588517e-01 2.95268089e-01
3.56549203e-01 1.00647271e+00 1.13565195e+00 4.35078740e-01
1.45720184e-01 8.25799108e-01 9.82863605e-01 8.30514580e-02
-1.68056954e-02 -4.28087190e-02 1.41906962e-01 4.97155130e-01
-5.83094358e-01 -3.41034830e-01 -7.24289417e-01 7.20526278e-01
-1.88706517e+00 -1.21316826e+00 -7.48700619e-01 2.31271553e+00
6.05787076e-02 3.28952000e-02 4.62812930e-02 -1.11122616e-01
1.11147773e+00 -1.39842689e-01 -6.53969407e-01 -3.34028482e-01
-1.56825572e-01 -1.40046284e-01 4.09796655e-01 7.80458748e-01
-6.31968200e-01 9.25252199e-01 6.83714294e+00 3.27281922e-01
-1.36207354e+00 2.04480976e-01 2.29803741e-01 -3.38505894e-01
-1.50414538e-02 1.63195565e-01 -1.06583047e+00 8.43494773e-01
7.50394702e-01 -4.94776905e-01 3.00291508e-01 5.28923392e-01
7.96228766e-01 -8.69297028e-01 -7.00546205e-01 8.00202250e-01
3.49383414e-01 -1.10660601e+00 -5.51228859e-02 2.20707357e-01
5.65900087e-01 -5.42174140e-03 5.18542290e-01 -3.50907266e-01
-5.69769621e-01 -6.44005299e-01 1.07970035e+00 1.05043626e+00
3.28903556e-01 -4.82304603e-01 4.77133602e-01 6.41799033e-01
-7.67889142e-01 -1.50341958e-01 -7.45807111e-01 -2.73559839e-01
4.83565390e-01 2.29233295e-01 -6.00443900e-01 9.16468650e-02
7.09060550e-01 7.46092558e-01 -7.42796957e-01 1.54667687e+00
-6.78113941e-03 4.33434844e-01 -1.02944903e-01 -4.34487432e-01
8.93890187e-02 -4.88539696e-01 1.07658660e+00 6.85364544e-01
5.02703786e-01 7.20442012e-02 -6.54224992e-01 1.25700843e+00
6.36898637e-01 2.11814895e-01 -5.77578843e-01 4.27011102e-01
-6.47426695e-02 1.03234136e+00 -7.92319059e-01 -6.86408728e-02
-2.51411140e-01 7.85126030e-01 3.01632453e-02 5.02289712e-01
-1.08637357e+00 -3.12334374e-02 3.98522943e-01 8.19345772e-01
4.37661201e-01 2.68506780e-02 -6.01919517e-02 -1.07940316e+00
-7.42532238e-02 -4.23462391e-02 -1.22023508e-01 -1.32682157e+00
-3.56329083e-01 5.37644506e-01 2.61898518e-01 -1.29038799e+00
-1.15709029e-01 -4.95358318e-01 -8.99882257e-01 1.21523643e+00
-1.58572769e+00 -5.80812097e-01 -4.18521494e-01 7.63801038e-01
7.88808823e-01 -4.57661785e-02 1.30523026e-01 -1.78352565e-01
-3.53531837e-01 -2.04350024e-01 -3.58501561e-02 -6.85788691e-01
4.31005269e-01 -1.06031942e+00 1.90515116e-01 1.04051769e+00
-8.85564908e-02 4.93519545e-01 1.38703537e+00 -7.45006084e-01
-1.08116329e+00 -5.49802780e-01 1.02657008e+00 -5.87668478e-01
6.75158083e-01 -1.00593947e-01 -1.05454850e+00 4.98190671e-01
3.41178864e-01 -1.05996951e-01 1.61640823e-01 -5.18870294e-01
7.03386843e-01 2.90463448e-01 -7.32630551e-01 4.43905950e-01
9.10987735e-01 1.89608056e-02 -8.51588607e-01 2.74036944e-01
5.11667669e-01 -4.10656631e-01 5.33461273e-02 -3.43084522e-02
4.31287318e-01 -1.40974963e+00 7.03701019e-01 -5.28495133e-01
2.20920369e-01 -3.74971211e-01 3.43919307e-01 -1.02008975e+00
-5.01208842e-01 -4.84914392e-01 -1.09743543e-01 6.44731283e-01
4.07710206e-03 -5.72187960e-01 5.60130477e-01 2.32151225e-01
6.13936875e-03 -4.50957686e-01 -8.87579501e-01 -3.45937580e-01
-4.31073725e-01 -1.27159804e-01 -5.52452952e-02 3.09950262e-01
-1.17262155e-01 2.20707595e-01 -2.79495120e-01 1.99522093e-01
8.26234162e-01 1.53055694e-02 3.65352541e-01 -1.44776726e+00
-1.69932902e-01 -3.95978540e-01 -6.22383833e-01 -1.00124562e+00
4.19337749e-02 -4.28266495e-01 -2.59472355e-02 -1.43289089e+00
1.97987139e-01 2.46633798e-01 -2.70057648e-01 -4.66715813e-01
-1.00693099e-01 -5.13401106e-02 2.52474010e-01 5.19934297e-01
-4.87640530e-01 4.87619817e-01 1.49601841e+00 5.09859383e-01
-3.71008545e-01 3.36462677e-01 -3.66555959e-01 8.57143939e-01
6.06659174e-01 -3.21777970e-01 -8.58035445e-01 -2.82012522e-01
4.35119212e-01 2.05598488e-01 6.94908500e-01 -8.98893297e-01
4.81142849e-01 -2.24155650e-01 4.91888486e-02 -8.51915240e-01
4.78150576e-01 -6.69348478e-01 3.32921147e-02 5.92749774e-01
-2.37719446e-01 1.45218968e-01 1.56636328e-01 6.98145449e-01
-1.91523239e-01 -5.55606842e-01 9.76063669e-01 1.23897009e-02
-7.97309935e-01 1.99814364e-02 -4.74457979e-01 -5.56528307e-02
1.28113520e+00 -4.44173008e-01 -6.72393665e-02 -5.47266781e-01
-9.15383339e-01 -1.43284574e-01 4.01641697e-01 4.69483107e-01
6.99678898e-01 -9.09018099e-01 -4.47576582e-01 3.13535303e-01
-3.68858099e-01 -3.10628682e-01 4.71933097e-01 1.33598936e+00
-6.37908220e-01 7.22649336e-01 -6.22828722e-01 -8.97530496e-01
-1.28312314e+00 9.31853175e-01 4.86408263e-01 3.05921257e-01
-2.85917819e-01 7.28663981e-01 7.83768952e-01 5.73890924e-01
2.01137308e-02 -2.70139992e-01 -9.15113747e-01 4.80370410e-02
6.76596165e-01 4.49968010e-01 -4.86406028e-01 -1.08504128e+00
-2.23228529e-01 1.01261950e+00 1.78556398e-01 -1.97499737e-01
1.15419710e+00 -7.95368075e-01 -2.67083913e-01 6.98103189e-01
4.71406132e-01 -3.23649123e-02 -1.37304986e+00 4.88965362e-02
-1.42664537e-01 -5.31352580e-01 8.98667946e-02 -3.81134450e-01
-8.13065350e-01 1.08318424e+00 6.93134964e-01 2.70011812e-01
1.05714595e+00 3.93587984e-02 1.43557802e-01 -1.92004412e-01
4.15652275e-01 -1.03748214e+00 1.38653755e-01 2.60098040e-01
9.76590931e-01 -1.07004213e+00 -1.65597871e-01 -5.06976724e-01
-7.86782265e-01 8.22576642e-01 7.43113101e-01 -4.60233063e-01
7.22723842e-01 -1.27487361e-01 -3.69444340e-01 -2.60465890e-01
-9.28691208e-01 -3.35982800e-01 5.53724527e-01 5.92192769e-01
2.23934785e-01 -4.29787457e-01 -6.44584775e-01 1.22747287e-01
6.32936805e-02 1.25548512e-01 1.03254879e+00 6.89738512e-01
-9.54894900e-01 -3.11260253e-01 -4.91599590e-01 -3.83075476e-02
-3.05989236e-01 -1.61430061e-01 8.08544308e-02 7.83632338e-01
1.94075480e-01 7.28095531e-01 4.99546021e-01 1.44117951e-01
1.55720294e-01 -1.11896046e-01 5.45780361e-01 -4.26032901e-01
-5.16253524e-02 2.65707433e-01 -7.12905169e-01 -5.65653801e-01
-7.81016886e-01 -9.19632494e-01 -1.19096172e+00 7.96796381e-02
-4.39421654e-01 2.10259274e-01 9.45902646e-01 7.87132204e-01
2.64667660e-01 4.62492794e-01 3.50970000e-01 -8.57486665e-01
-1.77703232e-01 -9.23738122e-01 -8.64567459e-01 1.50893420e-01
3.39227974e-01 -1.02496886e+00 -5.47077537e-01 2.87664860e-01] | [9.99693775177002, 1.5428884029388428] |
99710274-741f-4cc6-a092-02ef66c6e532 | lstm-based-ecg-classification-for-continuous | 1812.04818 | null | https://arxiv.org/abs/1812.04818v3 | https://arxiv.org/pdf/1812.04818v3.pdf | LSTM-Based ECG Classification for Continuous Monitoring on Personal Wearable Devices | Objective: A novel ECG classification algorithm is proposed for continuous cardiac monitoring on wearable devices with limited processing capacity. Methods: The proposed solution employs a novel architecture consisting of wavelet transform and multiple LSTM recurrent neural networks. Results: Experimental evaluations show superior ECG classification performance compared to previous works. Measurements on different hardware platforms show the proposed algorithm meets timing requirements for continuous and real-time execution on wearable devices. Conclusion: In contrast to many compute-intensive deep-learning based approaches, the proposed algorithm is lightweight, and therefore, brings continuous monitoring with accurate LSTM-based ECG classification to wearable devices. Significance: The proposed algorithm is both accurate and lightweight. The source code is available online [1]. | ['Saeed Saadatnejad', 'Mohammadhosein Oveisi', 'Matin Hashemi'] | 2018-12-12 | null | null | null | null | ['ecg-classification'] | ['medical'] | [ 4.88671064e-01 -4.28223908e-01 -1.79217368e-01 -4.16654646e-01
-6.83201551e-01 4.69356738e-02 -4.97419596e-01 4.85920459e-01
-5.23214698e-01 6.33086503e-01 -1.27358556e-01 -5.05055845e-01
-1.93497077e-01 -4.50619966e-01 -6.50829822e-02 -5.28096199e-01
-3.81503552e-01 5.61051667e-02 -6.32085130e-02 7.20618367e-02
3.03567853e-02 5.77085197e-01 -9.87933636e-01 4.71491158e-01
4.87816989e-01 1.43162942e+00 -2.28893787e-01 1.03849506e+00
3.92007858e-01 9.30671215e-01 -9.66237068e-01 -8.85948725e-03
-1.69602893e-02 -7.32266545e-01 -5.13250768e-01 -3.85187030e-01
9.10620689e-02 -3.94600660e-01 9.55525339e-02 4.54119414e-01
1.34100401e+00 -1.68535933e-02 1.24196760e-01 -5.63843668e-01
-1.14943974e-01 4.99412596e-01 -4.54171032e-01 9.61151540e-01
3.34517509e-01 -2.31959403e-01 2.49113157e-01 -6.02779329e-01
-3.91604565e-02 3.07693481e-01 1.62870038e+00 3.91607434e-01
-9.18064833e-01 -6.35702550e-01 -3.48029017e-01 2.94383556e-01
-1.42523873e+00 -4.37632829e-01 8.75469804e-01 -3.37028429e-02
1.45476508e+00 3.65075141e-01 1.15820801e+00 1.14008737e+00
9.13054764e-01 2.30788261e-01 9.42572176e-01 -4.36461598e-01
1.76402315e-01 -2.29236603e-01 4.74672347e-01 8.20039451e-01
4.56002295e-01 3.99824455e-02 -6.42219543e-01 -6.39425457e-01
7.56790102e-01 5.37229419e-01 5.01424447e-02 4.95294660e-01
-1.26481450e+00 3.47563833e-01 5.87027110e-02 6.64208353e-01
-9.30029273e-01 4.34112966e-01 1.18903232e+00 5.64599574e-01
6.09766126e-01 1.45054534e-01 -4.68713969e-01 -6.57426417e-01
-1.30457687e+00 -1.63874954e-01 7.02790916e-01 7.72813380e-01
-2.63759166e-01 8.14828873e-01 -7.81267881e-02 5.73008299e-01
-4.44192626e-02 3.85519743e-01 1.08045864e+00 -4.81692165e-01
1.42299518e-01 3.01825821e-01 -1.83309719e-01 -1.07744479e+00
-9.84816551e-01 -8.98000598e-01 -1.49284315e+00 -1.93940669e-01
-1.97552964e-01 -5.45780182e-01 -5.69584608e-01 9.69627678e-01
-5.38172433e-03 5.68900585e-01 4.95648682e-02 6.23690128e-01
1.06425834e+00 3.44840199e-01 4.14145321e-01 -5.67796767e-01
1.44102156e+00 -4.53391880e-01 -1.10813606e+00 2.61643827e-01
5.22909939e-01 -5.33830225e-01 6.20987773e-01 6.37567699e-01
-1.41887867e+00 -7.95255065e-01 -1.30705380e+00 1.95341915e-01
9.46128294e-02 6.40141070e-01 5.61215401e-01 1.28910029e+00
-1.16042042e+00 9.12053108e-01 -1.32356012e+00 -4.88520145e-01
2.56547987e-01 6.03277624e-01 4.23103333e-01 6.00727737e-01
-1.11548889e+00 8.18597972e-01 3.30557823e-01 6.76295936e-01
-2.49318033e-01 -4.19831187e-01 -5.08973062e-01 -5.28782867e-02
-4.17807728e-01 -1.05722129e+00 1.28802705e+00 -8.46438169e-01
-1.58578038e+00 6.24025464e-01 -4.13739741e-01 -1.03829932e+00
3.33030492e-01 -2.50013471e-01 -7.69734025e-01 3.68468761e-01
-3.87652904e-01 -2.70399988e-01 9.89488184e-01 -2.46099189e-01
-5.33893883e-01 -4.06673014e-01 -6.49013102e-01 -1.21590137e-01
-6.27940178e-01 2.94130921e-01 4.10203606e-01 -8.81378591e-01
2.24431813e-01 -7.31677175e-01 -2.31287017e-01 -9.99224484e-02
1.66909218e-01 6.84073120e-02 6.65120304e-01 -9.76333737e-01
1.56016552e+00 -2.00024891e+00 -3.88232619e-01 2.11396083e-01
4.67894316e-01 4.59824353e-01 5.49833357e-01 2.48023227e-01
-1.26616165e-01 -2.05382407e-01 -4.66910824e-02 -1.80323526e-01
-4.70138103e-01 1.43131211e-01 -2.65846521e-01 7.03650355e-01
-2.39114583e-01 1.06207407e+00 -4.65944231e-01 -5.35508931e-01
4.20757800e-01 8.92540753e-01 9.02116951e-03 -1.38938818e-02
6.14849746e-01 4.49626416e-01 -4.51967835e-01 6.24931514e-01
2.34790072e-01 -1.19726196e-01 3.90359282e-01 -4.47066516e-01
-8.24434217e-03 1.58102781e-01 -9.47332084e-01 1.78623390e+00
-5.12482047e-01 3.86016607e-01 -5.71273327e-01 -1.15366888e+00
1.28462923e+00 1.01119936e+00 7.92763293e-01 -6.16594732e-01
4.57966894e-01 4.23442394e-01 -5.52378520e-02 -7.85590410e-01
9.62929130e-02 -1.55769020e-01 2.48889625e-01 4.47117358e-01
-8.51929411e-02 5.70005000e-01 -2.57570863e-01 -4.42410201e-01
1.27781188e+00 2.66676873e-01 5.46032846e-01 -3.48206758e-01
1.84698761e-01 -3.76436263e-01 7.58107245e-01 1.00521898e+00
-3.02109063e-01 2.59898722e-01 -3.20983142e-01 -1.33639848e+00
-6.77407384e-01 -9.38270211e-01 -9.23634320e-02 8.89113486e-01
-4.32525456e-01 -4.04256731e-01 -6.08476758e-01 1.01248696e-01
-3.84346664e-01 -1.02874627e-02 -5.14433920e-01 -7.46240765e-02
-1.01233459e+00 -9.02057171e-01 1.37438238e+00 9.54635501e-01
6.11884773e-01 -1.12105489e+00 -2.01306200e+00 7.54944026e-01
-2.79287547e-01 -1.02998495e+00 -2.31986567e-01 1.35765836e-01
-1.81616127e+00 -6.09231710e-01 -7.96130478e-01 -9.31454360e-01
1.57303661e-01 -2.84936011e-01 1.18959796e+00 2.02337146e-01
-8.34085286e-01 4.56205338e-01 -3.03428620e-01 -8.88214290e-01
1.09267309e-01 8.49852338e-02 3.40135753e-01 -7.62423351e-02
3.79893035e-01 -8.89131069e-01 -1.04879034e+00 -3.91185313e-01
-3.71079415e-01 -2.86212832e-01 6.23693168e-01 8.31012785e-01
4.73267049e-01 -8.92409161e-02 8.60621810e-01 -7.38419235e-01
1.09797275e+00 -7.52081126e-02 -8.02829042e-02 1.10444061e-01
-1.01964283e+00 -4.53075886e-01 6.93874180e-01 -5.31649411e-01
-5.65934837e-01 1.29636541e-01 -4.14748609e-01 -4.19143230e-01
7.52489865e-02 5.49282074e-01 6.60689294e-01 -1.21009164e-01
6.96355283e-01 4.65035200e-01 -1.25877997e-02 -4.80847359e-01
-5.66893756e-01 5.86765349e-01 5.29287696e-01 -3.63105565e-01
-4.87670638e-02 2.80513912e-01 7.54349902e-02 -9.60920632e-01
-3.01943481e-01 -2.54400492e-01 -6.74724221e-01 -3.52978200e-01
7.76783288e-01 -7.94792175e-01 -1.13041139e+00 3.86383563e-01
-1.21521723e+00 8.92079249e-02 -3.50621670e-01 6.86427593e-01
-3.60547721e-01 2.51677483e-01 -1.17001426e+00 -1.08388066e+00
-1.61941028e+00 -5.08412600e-01 9.32142198e-01 -2.18417540e-01
-8.42708170e-01 -1.23900843e+00 -3.53129730e-02 -1.56544030e-01
8.32197428e-01 9.09479260e-01 5.72156250e-01 -4.41811740e-01
3.20172310e-01 -6.15365624e-01 2.77346611e-01 1.00900553e-01
3.50551337e-01 -2.49539256e-01 -9.33947325e-01 -6.20343566e-01
8.60283911e-01 6.39499500e-02 4.90539670e-01 7.39824474e-01
1.23016965e+00 -2.59318143e-01 -2.96840847e-01 5.57190716e-01
1.30501413e+00 5.56038618e-01 5.33403456e-01 8.59507397e-02
3.54725778e-01 -1.28140301e-01 1.97215199e-01 7.25728571e-01
5.29061593e-02 2.16098383e-01 -3.80789042e-01 -7.85891116e-01
1.49569407e-01 3.90817612e-01 2.23693162e-01 1.46577525e+00
-6.63562357e-01 4.74628001e-01 -9.88611102e-01 1.92851737e-01
-1.83934855e+00 -1.02064455e+00 -3.47727716e-01 2.15759158e+00
6.02640450e-01 1.67136967e-01 4.43411678e-01 8.24171007e-01
5.05844891e-01 -2.46448651e-01 -6.46106839e-01 -9.44775283e-01
2.19286710e-01 9.85720158e-01 3.48083109e-01 2.22391263e-02
-9.49927449e-01 4.26157862e-02 7.35561609e+00 6.15865551e-02
-1.51940119e+00 7.94137061e-01 4.46433395e-01 -3.08176935e-01
7.01352119e-01 -7.34539390e-01 -2.08255202e-01 3.95730466e-01
1.63417161e+00 -2.14125991e-01 -3.27055603e-01 5.11168957e-01
4.68109131e-01 3.21065545e-01 -8.43728781e-01 1.53480649e+00
2.29619995e-01 -1.28829718e+00 -1.91036984e-01 -4.02575821e-01
-6.59592748e-02 -2.08429545e-02 -2.87616197e-02 2.72269044e-02
-6.67141318e-01 -7.57420838e-01 3.70927274e-01 6.77874386e-01
1.15571845e+00 -9.09059584e-01 1.16579664e+00 1.21776666e-02
-1.55752861e+00 -3.06776941e-01 -1.95949674e-01 -5.78894734e-01
-3.32945101e-02 7.90437400e-01 -7.05776572e-01 6.10012174e-01
1.06523204e+00 9.12767649e-01 -4.04447407e-01 8.18875253e-01
3.79423022e-01 1.02191496e+00 -9.17934254e-02 -1.52974248e-01
-2.87576884e-01 2.10638508e-01 2.58514255e-01 1.60416162e+00
5.18480837e-01 4.42662120e-01 3.24986607e-01 2.76108116e-01
3.50736350e-01 1.43563539e-01 -6.03957951e-01 3.06474566e-01
1.80205330e-01 1.01319015e+00 -9.36503172e-01 -7.27122426e-01
-1.28106728e-01 1.04163134e+00 -2.82662362e-01 1.89120889e-01
-9.14548635e-01 -8.81099105e-01 -6.86086491e-02 1.17124408e-01
-1.60217747e-01 -1.89309359e-01 -9.10618901e-01 -7.60641694e-01
4.17588651e-01 -6.30022943e-01 6.98349237e-01 -3.26764911e-01
-9.80989695e-01 1.09435880e+00 -3.98719639e-01 -1.57097101e+00
-2.77678043e-01 -1.69636831e-01 -5.46113491e-01 7.51064122e-01
-1.01705265e+00 -1.08850873e+00 -3.04293543e-01 7.19571412e-01
5.29753983e-01 -2.18350247e-01 1.65022695e+00 6.39921904e-01
-5.50454855e-01 8.22828114e-01 -1.50577977e-01 1.73671152e-02
2.51039445e-01 -8.30639184e-01 4.81852740e-01 7.96086311e-01
-1.86038911e-01 1.00901043e+00 4.50944513e-01 -6.50373220e-01
-1.51297903e+00 -1.14188981e+00 9.41887200e-01 -2.23810539e-01
2.22791433e-01 -4.64581028e-02 -7.79918194e-01 4.73316163e-01
4.09801662e-01 9.76183116e-02 1.23737955e+00 1.49353355e-01
1.52416945e-01 -5.88750482e-01 -1.21735704e+00 9.03506279e-02
6.22903645e-01 -5.88186502e-01 -6.30617380e-01 3.05527985e-01
1.89085424e-01 -6.71157300e-01 -1.43215740e+00 5.68210483e-01
1.09355295e+00 -6.10000908e-01 8.55313778e-01 -3.79547834e-01
-1.01470247e-01 -6.83969120e-03 4.16530848e-01 -7.10653663e-01
-3.01288366e-01 -1.04398286e+00 -5.51340580e-01 5.13705432e-01
2.70031482e-01 -1.01807594e+00 5.85001111e-01 3.68372262e-01
-1.22431330e-01 -1.05769777e+00 -1.09393704e+00 -6.80337429e-01
-6.34951532e-01 -3.18023205e-01 1.78179353e-01 9.31180239e-01
2.81738281e-01 3.47481370e-01 -6.25436127e-01 6.07982511e-03
7.03755677e-01 1.08114325e-01 -2.41733357e-01 -1.33446658e+00
-3.45524997e-01 4.33964729e-02 -7.75480747e-01 -1.92055508e-01
-4.57696348e-01 -7.61416197e-01 -4.78466064e-01 -1.47929025e+00
-1.09727398e-01 -1.99472055e-01 -1.03687036e+00 7.55150974e-01
1.58161134e-01 8.31209064e-01 -2.64857829e-01 1.25314504e-01
-4.52643454e-01 -1.75032578e-02 5.17647684e-01 -7.90834278e-02
-4.96011078e-01 3.21677834e-01 -2.84308702e-01 7.15178013e-01
1.08577526e+00 -6.61441088e-01 -3.57830584e-01 -3.56887013e-01
1.39469773e-01 4.75192249e-01 2.75767535e-01 -1.79033816e+00
5.35282075e-01 6.32934988e-01 9.92580354e-01 -4.32794273e-01
3.73048604e-01 -7.47046232e-01 4.37914163e-01 1.48810101e+00
-3.14158559e-01 8.91595721e-01 4.21846747e-01 2.35072419e-01
-1.36940619e-02 3.01944464e-01 5.45497358e-01 -2.41676047e-02
-1.22953318e-01 1.13301948e-01 -8.37071300e-01 -3.03665996e-01
8.07392061e-01 -6.86103642e-01 3.25698584e-01 -7.36230910e-02
-1.10490668e+00 -4.18526679e-01 -4.24177170e-01 1.72563061e-01
1.19726241e+00 -1.23330629e+00 -7.58903205e-01 8.81164819e-02
-2.48990268e-01 -5.44128060e-01 2.21754447e-01 1.25018442e+00
-8.58310044e-01 3.57446671e-01 -5.68688035e-01 -7.37713456e-01
-1.71978211e+00 2.33701929e-01 6.34541512e-01 -1.40725076e-02
-1.43725288e+00 4.15942043e-01 -1.02158225e+00 2.52219707e-01
4.24181283e-01 -6.78970397e-01 -7.13199079e-02 -9.28472951e-02
7.45017886e-01 6.07728899e-01 6.73339725e-01 6.40757196e-03
-5.31449914e-01 5.91823399e-01 2.02943757e-01 1.94811478e-01
1.27469933e+00 5.59841134e-02 -2.43656859e-01 1.10640788e+00
9.48818326e-01 -5.75793862e-01 -6.83310926e-02 -8.54236260e-02
1.67719200e-01 -7.88657144e-02 1.15229171e-02 -6.97462678e-01
-1.05395496e+00 1.05835474e+00 1.64743972e+00 -5.89467585e-02
1.75711274e+00 -1.03097594e+00 1.29102588e+00 6.49478555e-01
4.41059411e-01 -1.16592026e+00 -1.68327719e-01 1.52478755e-01
5.18294096e-01 -5.45414329e-01 2.47924536e-01 2.26791114e-01
-3.03106427e-01 1.32453620e+00 6.12830520e-02 -3.38887691e-01
8.30168188e-01 5.47000229e-01 3.33759457e-01 -2.79059768e-01
-6.09992623e-01 4.58369046e-01 -3.47394235e-02 4.35368866e-01
9.28513348e-01 3.85243520e-02 -6.41868651e-01 5.60176551e-01
-3.13005596e-02 6.72242641e-01 2.39031583e-01 1.31566429e+00
-3.91995981e-02 -8.53254199e-01 -1.92292064e-01 8.65153253e-01
-1.03750265e+00 -1.69216946e-01 3.39069739e-02 1.65153727e-01
5.26535138e-02 1.01060796e+00 -2.30097532e-01 -3.88800502e-01
3.98111552e-01 2.15984121e-01 5.52080274e-01 -2.47572795e-01
-1.47599030e+00 9.26482975e-02 1.39088323e-02 -5.21210909e-01
-4.87901092e-01 -3.46352994e-01 -1.24536920e+00 -5.86288795e-03
5.61883766e-03 1.38589084e-01 7.87170827e-01 8.02692413e-01
7.97130942e-01 1.29648030e+00 4.66417253e-01 -7.09363341e-01
-4.77084965e-01 -1.10031068e+00 -5.74669063e-01 -2.69227065e-02
5.88033140e-01 -8.70851725e-02 3.15033674e-01 3.10309440e-01] | [14.234594345092773, 3.2509312629699707] |
ea152f72-1772-4c75-a904-d7bedabf41a0 | cuni-system-for-the-wmt18-multimodal-1 | null | null | https://aclanthology.org/W18-6441 | https://aclanthology.org/W18-6441.pdf | CUNI System for the WMT18 Multimodal Translation Task | We present our submission to the WMT18 Multimodal Translation Task. The main feature of our submission is applying a self-attentive network instead of a recurrent neural network. We evaluate two methods of incorporating the visual features in the model: first, we include the image representation as another input to the network; second, we train the model to predict the visual features and use it as an auxiliary objective. For our submission, we acquired both textual and multimodal additional data. Both of the proposed methods yield significant improvements over recurrent networks and self-attentive textual baselines. | ['Du{\\v{s}}an Vari{\\v{s}}', "Jind{\\v{r}}ich Libovick{\\'y}", 'Jind{\\v{r}}ich Helcl'] | 2018-10-01 | null | null | null | ws-2018-10 | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 4.65477318e-01 2.26833895e-01 -2.38607496e-01 -4.31014955e-01
-9.72627580e-01 -6.20031595e-01 1.07792473e+00 -1.13322876e-01
-6.24827862e-01 5.96156597e-01 5.55068552e-01 -4.08162862e-01
6.72681153e-01 -4.14295524e-01 -9.55527008e-01 -3.88467222e-01
3.14651549e-01 5.36346614e-01 1.99150527e-03 -4.55252260e-01
1.69332162e-01 2.29687765e-01 -1.01140416e+00 9.29139316e-01
7.67004788e-01 9.89188969e-01 2.56163746e-01 7.60463655e-01
4.87683527e-03 8.72185111e-01 -4.46628332e-01 -4.97301936e-01
9.41260066e-03 -5.65771282e-01 -1.06215692e+00 3.39513049e-02
5.86159885e-01 -3.14381301e-01 -6.53971255e-01 6.39182866e-01
5.93680918e-01 3.25905532e-01 8.24118733e-01 -1.07898533e+00
-1.00439262e+00 7.41119206e-01 -5.05206823e-01 1.58876061e-01
5.03577411e-01 3.28342855e-01 9.44666207e-01 -1.32091224e+00
9.26671624e-01 1.44684792e+00 3.52430880e-01 6.79230273e-01
-1.37228680e+00 -3.92724246e-01 2.38353491e-01 4.77807432e-01
-8.63330603e-01 -8.32598984e-01 9.07416165e-01 -1.98688701e-01
1.15629542e+00 1.59644559e-01 5.81396222e-01 1.68351758e+00
1.06676400e-01 1.20541883e+00 1.16140759e+00 -5.54924726e-01
-2.36581057e-01 3.78154188e-01 -1.20707102e-01 7.11748898e-01
-7.20632553e-01 1.94680899e-01 -5.70036232e-01 1.53974921e-01
5.21380603e-01 -5.06566130e-02 -7.22357184e-02 -1.43262014e-01
-1.45222247e+00 7.87872255e-01 7.51825988e-01 2.12602571e-01
-2.50820726e-01 1.73334226e-01 4.14466858e-01 4.93530959e-01
5.78718305e-01 4.46523577e-01 -7.54039958e-02 -1.31005108e-01
-8.65150511e-01 -1.54348135e-01 2.38189429e-01 8.15689504e-01
8.06164145e-01 7.23068602e-04 -8.96799088e-01 1.07083750e+00
3.50335002e-01 5.59020579e-01 6.70254767e-01 -7.24550605e-01
9.09811795e-01 4.89683390e-01 -7.40592852e-02 -7.95641005e-01
-3.56071144e-01 -1.88114792e-01 -6.60422206e-01 -2.46777594e-01
-2.05148906e-02 -2.39516854e-01 -1.26166141e+00 1.74452114e+00
-2.43424147e-01 -3.14632893e-01 2.48295069e-01 9.38782513e-01
1.30248702e+00 9.53949153e-01 2.01699510e-01 -1.19691879e-01
1.06284821e+00 -1.61825120e+00 -1.08832681e+00 -1.55763984e-01
6.44102097e-01 -9.87402260e-01 1.19263387e+00 -6.27548695e-02
-1.33618736e+00 -5.60818195e-01 -9.85365748e-01 -3.04859966e-01
-5.87979615e-01 4.77598548e-01 2.59807974e-01 1.23620309e-01
-1.47964239e+00 4.54187155e-01 -5.27308464e-01 -5.92544556e-01
1.92313671e-01 2.50632823e-01 -3.93581480e-01 9.53990221e-02
-1.24298096e+00 1.36804223e+00 1.74941316e-01 2.43513212e-01
-9.16864693e-01 -1.20536357e-01 -1.15476191e+00 -1.41001210e-01
8.87267664e-02 -7.95725167e-01 1.46987760e+00 -1.33157098e+00
-1.82104039e+00 1.03554285e+00 -6.83061123e-01 -2.04215214e-01
5.73755205e-01 -1.28790215e-01 -2.77401954e-01 4.00494546e-01
-8.16319808e-02 1.23114753e+00 9.26897526e-01 -1.49635530e+00
-3.16391200e-01 -6.67128991e-03 1.35065481e-01 5.59107721e-01
-4.38336104e-01 -2.04927418e-02 -5.67671418e-01 -7.68599033e-01
-2.54746437e-01 -1.09656930e+00 -3.37270200e-02 -2.50919878e-01
-6.51212335e-01 -1.62749901e-01 1.08161151e+00 -6.86669052e-01
7.94791102e-01 -2.07857275e+00 6.35938048e-01 -1.91452816e-01
1.53694093e-01 2.79488359e-02 -8.53956282e-01 5.75376987e-01
-2.30348736e-01 1.76543370e-01 -9.57171917e-02 -8.67717206e-01
2.11581793e-02 4.38169949e-02 -4.40836370e-01 1.08522430e-01
6.29654646e-01 1.68263018e+00 -6.89102709e-01 -5.47240973e-01
2.79011101e-01 4.49845344e-01 -1.87678069e-01 6.15060151e-01
-4.18710679e-01 5.13690889e-01 -4.14407551e-02 5.29792249e-01
4.63806629e-01 -2.40637138e-01 4.72051054e-02 -4.15056914e-01
-1.33021951e-01 5.26398778e-01 -2.50917107e-01 1.74210441e+00
-4.65432286e-01 8.75533819e-01 -3.57081234e-01 -8.42479825e-01
6.12634003e-01 5.18214643e-01 1.08667873e-01 -1.18512785e+00
1.89715281e-01 -2.00003982e-01 -2.81491607e-01 -5.33339441e-01
6.66114748e-01 -1.21483900e-01 1.54025189e-03 6.30555630e-01
2.57428467e-01 5.15336320e-02 9.16498303e-02 4.11051601e-01
7.22432673e-01 4.28693622e-01 -1.16112076e-01 3.13882649e-01
3.61757815e-01 4.48910519e-02 8.01907033e-02 6.80369020e-01
-8.42590164e-03 8.69653344e-01 5.83645403e-01 -3.57169569e-01
-1.10721552e+00 -9.22102928e-01 3.69213998e-01 1.43860900e+00
-3.32282158e-04 -2.85085797e-01 -3.81047487e-01 -9.77718174e-01
-4.54926789e-01 7.40174890e-01 -1.03564644e+00 -3.11024845e-01
-5.11570692e-01 -5.21682858e-01 4.28146034e-01 7.51317799e-01
4.88860816e-01 -1.18791234e+00 -2.80756712e-01 -1.31583840e-01
-6.17253006e-01 -1.18145835e+00 -7.60825396e-01 2.55383909e-01
-7.62524545e-01 -4.05041784e-01 -9.26847637e-01 -8.84921074e-01
7.03337610e-01 3.63282830e-01 1.13789713e+00 5.65211289e-02
1.19288385e-01 4.17833000e-01 -4.28602636e-01 -1.43966660e-01
-3.35731626e-01 3.34106475e-01 -4.01514471e-01 -4.97048534e-02
-3.98412272e-02 -1.99623585e-01 -5.71843326e-01 1.69097960e-01
-6.71354771e-01 5.00979960e-01 7.38232434e-01 1.00943255e+00
3.09344381e-01 -9.87762094e-01 3.78300607e-01 -7.44980574e-01
8.29452634e-01 -1.89614341e-01 -1.42793268e-01 4.84735906e-01
-2.82648236e-01 4.68435735e-02 3.47983599e-01 -6.32158458e-01
-1.12203979e+00 2.54669070e-01 -2.26574615e-01 -4.66046065e-01
5.14976345e-02 6.39186561e-01 -4.69966140e-03 -3.11605781e-02
1.82110652e-01 2.25058585e-01 8.81258771e-02 -2.31381178e-01
7.66466618e-01 5.68446338e-01 3.64511907e-01 -5.63777089e-01
7.75203168e-01 6.80447817e-02 -2.72169083e-01 -6.94614053e-01
-7.72836864e-01 -1.44187376e-01 -8.37087333e-01 -2.83910900e-01
8.88867438e-01 -8.87652218e-01 -5.77844262e-01 2.59331167e-01
-1.52279317e+00 -6.19817674e-01 -5.33976257e-02 3.27572644e-01
-6.54784083e-01 2.04907358e-01 -7.56547987e-01 -6.15178704e-01
-3.90609413e-01 -1.32135046e+00 1.18818414e+00 7.04852166e-03
-2.05527902e-01 -1.09571612e+00 1.85690746e-01 4.67172176e-01
5.66920578e-01 7.88056552e-02 1.12332940e+00 -6.44874215e-01
-4.08304036e-01 -9.11226049e-02 -6.47175550e-01 -1.21400565e-01
1.89917237e-02 4.16463092e-02 -1.17502701e+00 -3.15253556e-01
-1.95603728e-01 -9.05537605e-01 1.21244621e+00 1.71779886e-01
8.84078026e-01 -4.66972411e-01 -2.64990211e-01 6.07906699e-01
1.04336894e+00 5.54324612e-02 7.61010349e-01 2.79252231e-01
8.21919024e-01 5.99085212e-01 3.62675011e-01 -1.05909809e-01
7.36931026e-01 9.69745338e-01 4.84917223e-01 -6.62696242e-01
-3.32989603e-01 -3.92747551e-01 6.49377823e-01 1.07809877e+00
-2.00494021e-01 -4.11641449e-01 -8.45543087e-01 4.34073567e-01
-2.25181842e+00 -8.67775321e-01 1.19329616e-01 1.85490632e+00
8.58223677e-01 -2.63787746e-01 2.92943686e-01 -5.49747288e-01
5.44666827e-01 1.51909575e-01 -3.84180635e-01 -6.98566198e-01
-4.62894201e-01 -1.17844425e-01 1.23956636e-01 5.91220200e-01
-1.23563600e+00 1.25857151e+00 7.81677246e+00 2.75394529e-01
-1.45049071e+00 2.34520525e-01 5.64703763e-01 -2.33610332e-01
-3.61541212e-01 -2.16561064e-01 -4.15690333e-01 -8.31070635e-03
1.04722536e+00 1.34561434e-01 4.17881966e-01 3.05748701e-01
2.27989495e-01 4.76721935e-02 -1.21216500e+00 7.17352748e-01
5.38563251e-01 -1.25826132e+00 5.35965741e-01 -3.96581888e-02
5.85808814e-01 -1.50032798e-02 5.67758024e-01 5.31779170e-01
2.45135590e-01 -1.37800944e+00 7.61629164e-01 6.63502812e-01
8.24616492e-01 -6.84511840e-01 8.86211634e-01 3.36002372e-02
-8.96916151e-01 1.88111320e-01 -1.28451854e-01 1.84119062e-03
1.70055732e-01 -2.34896630e-01 -8.28590751e-01 5.57752430e-01
3.59106928e-01 9.44321632e-01 -9.77787435e-01 7.42646575e-01
-3.22403312e-01 3.58084649e-01 1.68437079e-01 2.59481650e-02
4.56627488e-01 5.62216304e-02 5.65345645e-01 1.47778213e+00
-5.37320524e-02 -2.16217801e-01 2.98826545e-01 7.97923386e-01
-3.93950552e-01 2.31790334e-01 -8.05416048e-01 -3.70979965e-01
8.17701742e-02 1.26137340e+00 -3.80034417e-01 -4.31843758e-01
-5.51664174e-01 1.26782382e+00 8.00743699e-01 7.79931128e-01
-7.35417247e-01 -3.28121841e-01 1.27008781e-01 -2.41522312e-01
3.49137902e-01 -1.51914552e-01 -3.49536330e-01 -1.34276307e+00
-2.21735671e-01 -9.39057589e-01 2.98970491e-01 -1.22857475e+00
-1.21959841e+00 8.79508376e-01 -1.18431799e-01 -9.97106135e-01
-4.76069987e-01 -6.93310916e-01 -5.72600007e-01 9.77824450e-01
-1.62809789e+00 -1.65312469e+00 -6.35953397e-02 6.17608905e-01
6.65187657e-01 -2.14374259e-01 9.86943007e-01 1.22455224e-01
-6.87579215e-01 9.18540239e-01 -2.13587016e-01 2.42320776e-01
1.17656052e+00 -1.20864785e+00 2.46997565e-01 6.62203610e-01
1.44136846e-01 7.10700572e-01 6.24482393e-01 -5.41649342e-01
-1.61443985e+00 -1.02451539e+00 1.01675379e+00 -5.38659930e-01
7.35579789e-01 -5.12284040e-01 -8.51803482e-01 1.12730289e+00
1.26105475e+00 -3.61846060e-01 5.08420825e-01 1.60225928e-01
-4.01296675e-01 2.19653472e-01 -8.29761207e-01 6.74490571e-01
6.67885482e-01 -6.52425468e-01 -8.90358686e-01 3.31994653e-01
1.01355314e+00 -4.53666300e-01 -6.22070670e-01 4.06885773e-01
5.40451944e-01 -3.86554152e-01 8.74940693e-01 -7.80629456e-01
9.71589506e-01 4.73843291e-02 -1.79385871e-01 -1.64207911e+00
-2.86182582e-01 -6.23088360e-01 -1.15347795e-01 1.09446704e+00
1.01766777e+00 -3.97204071e-01 3.30021083e-01 3.30981821e-01
-1.83815435e-01 -6.46385372e-01 -7.80968547e-01 -4.52287704e-01
2.79222846e-01 5.52649284e-03 4.45838682e-02 8.73714745e-01
3.13293934e-01 9.36066270e-01 -7.18123019e-01 -2.98437506e-01
3.48107368e-01 1.13354884e-01 7.95150697e-01 -6.56486869e-01
4.84605804e-02 -5.34367085e-01 6.35415167e-02 -1.16029716e+00
2.38418385e-01 -1.03864431e+00 2.55843401e-01 -1.79803395e+00
6.96570754e-01 2.69859701e-01 -3.05384010e-01 9.24921036e-01
-1.83942020e-01 6.47715449e-01 3.93579990e-01 1.66243181e-01
-8.90356183e-01 9.43253934e-01 1.30420709e+00 -5.65766752e-01
-1.62787035e-01 -2.42576838e-01 -4.88619268e-01 3.15056175e-01
5.83222091e-01 -7.65185505e-02 -1.96000889e-01 -8.03084373e-01
2.48507425e-01 2.09111292e-02 4.92732584e-01 -2.99216300e-01
1.92649573e-01 -1.18380405e-01 7.48778641e-01 -9.63908434e-01
6.99066579e-01 -5.57919681e-01 -4.02699649e-01 2.20974162e-01
-8.35807562e-01 4.84557003e-01 4.88160312e-01 2.64772505e-01
-3.14402193e-01 6.65141121e-02 6.56343520e-01 -8.24825019e-02
-2.87914306e-01 5.15127517e-02 -6.55405998e-01 -1.95933014e-01
4.53721434e-01 -1.07503548e-01 -7.13551521e-01 -7.43271291e-01
-9.39933956e-01 5.36128879e-01 4.94571656e-01 7.50636578e-01
9.82356191e-01 -1.66816127e+00 -7.14918673e-01 -1.97902590e-01
3.88765901e-01 -5.94355047e-01 5.55124786e-03 1.14136481e+00
-1.37874612e-03 5.82201481e-01 -4.53027964e-01 -7.12193251e-01
-1.33399642e+00 7.44908631e-01 1.93285674e-01 -1.46106184e-01
-5.48396528e-01 4.44611579e-01 2.10693717e-01 -6.19411409e-01
4.39936489e-01 2.19832182e-01 -5.13439059e-01 2.29357615e-01
3.52610439e-01 -7.06742033e-02 -1.01388656e-01 -8.78994823e-01
-2.66949803e-01 4.96837944e-01 -3.53407085e-01 -6.68981254e-01
1.45203590e+00 -3.91415864e-01 -2.17622161e-01 9.35204804e-01
1.29483187e+00 -2.94246465e-01 -9.96246338e-01 -5.32334447e-01
-1.85185358e-01 -6.91692159e-02 3.57324369e-02 -1.18922174e+00
-9.17322636e-01 1.21949887e+00 6.26965225e-01 -1.13674849e-01
1.04938614e+00 9.25880894e-02 5.56251526e-01 5.52500248e-01
-2.58092552e-01 -1.05994487e+00 4.66278553e-01 9.69572902e-01
1.26657581e+00 -1.56544292e+00 -2.11180404e-01 -3.93774472e-02
-9.78811800e-01 1.14884388e+00 7.04100370e-01 1.53108478e-01
-5.00498451e-02 -8.19767118e-02 4.63866919e-01 -1.16818115e-01
-1.21346653e+00 -2.90773779e-01 8.55734527e-01 5.33018470e-01
8.50877762e-01 -3.03990871e-01 -2.06282973e-01 2.47596830e-01
-6.55134022e-02 -2.98706353e-01 3.43132108e-01 8.42742383e-01
-1.27789631e-01 -9.07967448e-01 -1.19644582e-01 2.24376824e-02
-9.72130001e-02 -3.16975921e-01 -1.00806618e+00 5.68038404e-01
-3.96047980e-01 8.78223538e-01 -4.12388742e-02 -5.98787963e-01
4.19818521e-01 2.37694547e-01 6.58192098e-01 -6.10293210e-01
-7.71261394e-01 2.89760232e-01 1.85317323e-01 -4.49226767e-01
-4.91892040e-01 -5.59669733e-01 -8.84538114e-01 -2.26360317e-02
-8.41589868e-02 -6.16026968e-02 7.95390248e-01 1.12528253e+00
2.63051033e-01 5.92203498e-01 7.79212594e-01 -1.21284127e+00
-9.81725529e-02 -1.38074660e+00 3.68228048e-01 4.43699062e-01
6.02409840e-01 -4.59018528e-01 -1.27034321e-01 1.19109534e-01] | [11.408249855041504, 1.5511177778244019] |
3de853e4-d598-4d60-a249-645bb204a404 | ttvos-lightweight-video-object-segmentation | 2011.04445 | null | https://arxiv.org/abs/2011.04445v3 | https://arxiv.org/pdf/2011.04445v3.pdf | TTVOS: Lightweight Video Object Segmentation with Adaptive Template Attention Module and Temporal Consistency Loss | Semi-supervised video object segmentation (semi-VOS) is widely used in many applications. This task is tracking class-agnostic objects from a given target mask. For doing this, various approaches have been developed based on online-learning, memory networks, and optical flow. These methods show high accuracy but are hard to be utilized in real-world applications due to slow inference time and tremendous complexity. To resolve this problem, template matching methods are devised for fast processing speed but sacrificing lots of performance in previous models. We introduce a novel semi-VOS model based on a template matching method and a temporal consistency loss to reduce the performance gap from heavy models while expediting inference time a lot. Our template matching method consists of short-term and long-term matching. The short-term matching enhances target object localization, while long-term matching improves fine details and handles object shape-changing through the newly proposed adaptive template attention module. However, the long-term matching causes error-propagation due to the inflow of the past estimated results when updating the template. To mitigate this problem, we also propose a temporal consistency loss for better temporal coherence between neighboring frames by adopting the concept of a transition matrix. Our model obtains 79.5% J&F score at the speed of 73.8 FPS on the DAVIS16 benchmark. The code is available in https://github.com/HYOJINPARK/TTVOS. | ['Nojun Kwak', 'Ganesh Venkatesh', 'Hyojin Park'] | 2020-11-09 | null | null | null | null | ['template-matching'] | ['computer-vision'] | [ 3.45987640e-02 -3.31335694e-01 -2.65155226e-01 -3.05154234e-01
-3.88371170e-01 -1.18979082e-01 3.42763305e-01 -1.81613550e-01
-5.88344097e-01 5.06000102e-01 -2.39465833e-01 1.25484720e-01
5.41892573e-02 -6.65295780e-01 -7.12183535e-01 -7.14823246e-01
1.53599203e-01 1.50140867e-01 1.16652656e+00 5.87114990e-02
2.71708578e-01 3.54281753e-01 -1.49002647e+00 1.75066501e-01
1.07776403e+00 1.23466396e+00 6.74326122e-01 3.36259037e-01
-4.12572682e-01 6.83106840e-01 -3.92846495e-01 -2.68339306e-01
4.32022363e-01 -2.54137099e-01 -6.40227258e-01 1.66765049e-01
6.30631566e-01 -4.47191477e-01 -5.35271525e-01 1.28965342e+00
4.86859202e-01 2.75955230e-01 2.63528109e-01 -1.36043906e+00
-4.30776387e-01 3.27968985e-01 -7.66631961e-01 5.66547692e-01
-1.46419341e-02 5.25537074e-01 4.84212607e-01 -7.68943846e-01
5.63830614e-01 1.21666741e+00 6.00750268e-01 5.78724444e-01
-8.77833009e-01 -8.61130476e-01 4.16166693e-01 5.39955616e-01
-1.37881887e+00 -4.85451877e-01 7.55279243e-01 -5.00370979e-01
6.30733550e-01 6.66825520e-03 6.91990495e-01 7.88123667e-01
3.33476156e-01 9.89164472e-01 9.27830219e-01 3.10219191e-02
7.09456299e-03 9.46553349e-02 1.25066161e-01 8.60569894e-01
1.65754423e-01 1.61313340e-01 -4.31368619e-01 3.81641001e-01
9.05717015e-01 2.59250402e-01 -3.95140827e-01 -3.68177325e-01
-1.07371831e+00 4.65299278e-01 6.11732721e-01 2.95221627e-01
-2.95776635e-01 1.56387746e-01 4.30300653e-01 -4.72556651e-02
3.67643654e-01 -3.12043931e-02 -4.61408079e-01 -9.02708098e-02
-1.11189616e+00 1.43755469e-02 3.53820592e-01 1.00326991e+00
6.85116529e-01 2.40137875e-01 -4.56790954e-01 7.42233276e-01
3.94045115e-01 5.46914876e-01 6.27848029e-01 -9.76445019e-01
4.90985185e-01 5.88022888e-01 2.06854835e-01 -1.23968875e+00
-3.54976445e-01 -6.36034667e-01 -8.34633470e-01 2.53626317e-01
6.13631845e-01 3.77238654e-02 -1.08497989e+00 1.69809306e+00
7.02992082e-01 6.53533101e-01 -3.18746120e-01 1.17421103e+00
7.70980418e-01 7.57955015e-01 1.59922332e-01 -3.35513651e-01
1.13910413e+00 -1.43872774e+00 -8.34033549e-01 -1.60655156e-01
3.79412770e-01 -8.61216545e-01 8.60921264e-01 1.77991420e-01
-1.13519025e+00 -9.53429759e-01 -9.19396937e-01 6.53936788e-02
-2.27432206e-01 3.40855084e-02 5.21416903e-01 4.45948780e-01
-8.31282437e-01 7.32340455e-01 -1.01812637e+00 -1.10941641e-01
7.44644523e-01 3.20463985e-01 -7.03627244e-02 -3.97910327e-02
-9.89062011e-01 6.58741713e-01 4.34112966e-01 2.88311481e-01
-7.59042263e-01 -6.46608293e-01 -7.04919517e-01 -3.67505997e-02
6.40097320e-01 -7.19463050e-01 1.00183332e+00 -1.10240245e+00
-1.54266334e+00 5.37068903e-01 -2.67070293e-01 -4.29344505e-01
7.34520316e-01 -2.90747583e-01 -4.07153755e-01 1.20195068e-01
2.22455516e-01 7.37342834e-01 1.10004914e+00 -8.22902083e-01
-8.28848779e-01 -3.52246612e-01 -1.64938912e-01 2.13989958e-01
-3.38480473e-01 -1.18080869e-01 -8.71166527e-01 -7.70010650e-01
2.38264903e-01 -8.62510443e-01 -2.88109004e-01 3.61050099e-01
-1.84362486e-01 -2.32627586e-01 1.06352055e+00 -5.38604677e-01
1.24502826e+00 -2.11109376e+00 -1.47798121e-01 -2.53824294e-01
1.73016384e-01 8.16943288e-01 -1.83321625e-01 -2.56780952e-01
1.69448331e-01 -8.45410600e-02 -2.39001483e-01 -2.50002652e-01
-3.94192904e-01 8.62195995e-03 -7.07772970e-02 4.05991077e-01
2.51771092e-01 8.31436515e-01 -9.10776496e-01 -8.12676013e-01
4.73013610e-01 4.59729046e-01 -3.97593737e-01 3.09622675e-01
-2.41216123e-01 4.82276887e-01 -4.16651160e-01 6.01967275e-01
9.71655190e-01 -3.00541252e-01 -3.87899607e-01 -4.06288356e-01
-2.82807052e-01 -9.00947750e-02 -1.44694185e+00 1.80991447e+00
-1.63987100e-01 4.01540041e-01 5.30720577e-02 -8.83713663e-01
6.60339236e-01 2.96597630e-02 6.35228932e-01 -7.95882523e-01
3.74567747e-01 1.67799309e-01 7.91275203e-02 -6.70130193e-01
3.18528175e-01 3.20701212e-01 6.34019256e-01 9.21109915e-02
-6.97277933e-02 2.61827737e-01 4.22498733e-01 8.09324384e-02
8.73149753e-01 4.92410541e-01 -8.51120651e-02 -1.52885780e-01
7.70591021e-01 3.69372033e-02 1.10394061e+00 5.90846062e-01
-6.96557522e-01 6.01755738e-01 -1.43233016e-01 -6.27498150e-01
-7.35180199e-01 -7.77709782e-01 -8.49889070e-02 7.46444583e-01
7.65874684e-01 -1.96820617e-01 -6.38181686e-01 -7.15996563e-01
8.89204443e-03 3.07599992e-01 -3.77330363e-01 -2.64264643e-01
-6.47244632e-01 -6.15939915e-01 2.99763709e-01 5.93325436e-01
9.87714529e-01 -1.14113533e+00 -7.48577356e-01 4.77483332e-01
-2.30161145e-01 -1.31286645e+00 -9.41834867e-01 -3.67504984e-01
-1.11759055e+00 -1.06346381e+00 -9.26165283e-01 -6.62498891e-01
6.05930448e-01 5.60469806e-01 7.87161052e-01 2.16106325e-01
-4.16853368e-01 6.41848817e-02 -2.14874953e-01 -2.74882585e-01
-5.76702692e-02 -9.94854197e-02 1.69384256e-02 3.01319420e-01
1.53538272e-01 -4.09659594e-01 -8.98965180e-01 5.77783942e-01
-8.78807068e-01 2.62012839e-01 6.23533070e-01 7.55646050e-01
7.29082346e-01 -1.11539900e-01 4.59089875e-01 -6.65419519e-01
-4.66539450e-02 -3.03202987e-01 -9.12505269e-01 2.30364963e-01
-6.87409461e-01 -2.73175035e-02 5.16540706e-01 -7.10827589e-01
-1.11233282e+00 1.77939400e-01 -9.15201232e-02 -7.51901209e-01
2.03863736e-02 -9.96485278e-02 -1.55519217e-01 -3.52093995e-01
1.46854565e-01 3.27178836e-01 7.72570148e-02 -3.76961559e-01
7.84390494e-02 4.74694937e-01 4.41745847e-01 -2.85282791e-01
8.19767892e-01 5.49475431e-01 -1.06540091e-01 -5.15593052e-01
-9.14388299e-01 -6.24393523e-01 -5.09190142e-01 -5.23149133e-01
9.30974245e-01 -8.95122349e-01 -7.37295091e-01 8.26206684e-01
-1.25854599e+00 -3.89784753e-01 -1.42662272e-01 6.09301984e-01
-3.32732439e-01 5.52717626e-01 -5.69332838e-01 -6.80007696e-01
-4.86884326e-01 -1.26707125e+00 8.41140091e-01 7.16475427e-01
4.23377275e-01 -7.14215875e-01 -2.55356669e-01 3.30464065e-01
6.27826691e-01 -1.56114763e-02 1.46973327e-01 -1.82013854e-01
-9.90436852e-01 1.29878819e-01 -6.40159249e-01 3.31671119e-01
1.06313214e-01 -4.16103192e-02 -8.02184463e-01 -3.35054725e-01
1.14694931e-01 7.98501670e-02 1.01432705e+00 5.20653069e-01
1.21247315e+00 -1.08761042e-01 -3.67056310e-01 8.77890944e-01
1.47229052e+00 5.71836412e-01 5.61428308e-01 3.30973864e-01
9.84789550e-01 3.34627420e-01 9.67575014e-01 3.33093435e-01
2.78539121e-01 7.66926885e-01 3.99931878e-01 -2.14796271e-02
-5.41882813e-01 -3.89414839e-02 3.70055914e-01 7.62850642e-01
-1.75779518e-02 -1.00192718e-01 -7.39904106e-01 5.78505993e-01
-2.08339190e+00 -9.34729457e-01 -1.98717058e-01 2.32804179e+00
7.78413177e-01 4.70015854e-01 1.08890217e-02 -1.86649725e-01
1.02277398e+00 1.80876702e-01 -8.02827716e-01 8.83575678e-02
-3.97702446e-03 -1.55070186e-01 5.65425754e-01 3.82610172e-01
-1.17202604e+00 1.10432601e+00 4.75090551e+00 1.15520132e+00
-1.21909893e+00 3.01896960e-01 6.86855674e-01 9.38370526e-02
2.10028842e-01 2.72095501e-02 -1.02517605e+00 8.61839235e-01
5.23092210e-01 1.35809451e-01 1.58432081e-01 8.00814867e-01
3.75480086e-01 -2.99355060e-01 -7.95390666e-01 1.19567585e+00
-3.12254447e-02 -1.40812838e+00 -2.77221249e-03 -3.26035231e-01
6.23810470e-01 -2.14677043e-02 -1.41418085e-01 1.84299111e-01
-3.30345899e-01 -5.04741192e-01 7.41899490e-01 6.15633249e-01
5.78041673e-01 -4.03500915e-01 6.34077966e-01 3.12971413e-01
-1.53832006e+00 -2.15061754e-02 -3.97835165e-01 8.04796219e-02
2.58580863e-01 6.61997080e-01 -2.98210651e-01 5.32691181e-01
7.97916830e-01 8.21517706e-01 -6.45926416e-01 1.50839245e+00
1.75799262e-02 4.82761830e-01 -4.53677863e-01 1.54619634e-01
2.57678717e-01 -3.40455443e-01 7.67811000e-01 1.17652845e+00
1.16645321e-01 8.61836225e-02 4.70952570e-01 7.74928391e-01
1.55404225e-01 7.37406611e-02 -2.33693376e-01 2.60280579e-01
4.01236057e-01 1.27224696e+00 -9.59871650e-01 -5.73976815e-01
-4.03595805e-01 1.05083728e+00 1.07563309e-01 2.50998527e-01
-1.14708698e+00 -4.71318483e-01 3.10684204e-01 2.23806798e-01
3.33440363e-01 -2.66570061e-01 -1.77992761e-01 -1.17239761e+00
1.54471055e-01 -4.65207189e-01 3.20291430e-01 -6.45323277e-01
-1.01581299e+00 5.64193666e-01 -2.17602983e-01 -1.41010320e+00
1.82184637e-01 -2.75166690e-01 -5.37157893e-01 5.26731312e-01
-1.73867118e+00 -9.09671724e-01 -6.15453243e-01 6.29676640e-01
9.15020108e-01 -2.39412822e-02 1.50727153e-01 7.68171847e-01
-8.84939849e-01 6.18253112e-01 -1.65538713e-01 2.14031860e-01
7.84395397e-01 -9.62846696e-01 2.34717563e-01 1.03464711e+00
2.22062469e-02 3.46400589e-01 5.30634880e-01 -7.36339450e-01
-1.14253354e+00 -1.35722184e+00 6.02767706e-01 -4.87570912e-02
3.99152428e-01 1.26879374e-02 -1.19980597e+00 3.34832817e-01
4.65723276e-02 5.72761953e-01 -1.04572296e-01 -5.26773214e-01
-2.51231402e-01 -4.22784477e-01 -1.14439964e+00 4.76784140e-01
1.22282732e+00 -9.18331817e-02 -3.63095999e-01 2.83959806e-01
8.21346164e-01 -5.22427678e-01 -6.27896070e-01 5.26148021e-01
5.43107092e-01 -1.08894551e+00 8.11208427e-01 -1.92802846e-01
1.03011660e-01 -8.09764445e-01 3.15549195e-01 -7.83439040e-01
-3.89881641e-01 -6.16878033e-01 -2.20816761e-01 1.29697168e+00
-2.39463337e-02 -6.39885783e-01 9.04754639e-01 6.16557598e-01
-2.37720966e-01 -9.06771600e-01 -8.51274431e-01 -1.04131031e+00
-3.33611906e-01 -3.01603884e-01 2.96955109e-01 7.72017837e-01
-6.24357700e-01 1.17184743e-01 -3.06804866e-01 1.58361688e-01
8.57872844e-01 2.62988031e-01 6.54950976e-01 -9.47823584e-01
-1.84719518e-01 -4.97184038e-01 -5.37458539e-01 -1.27487183e+00
-8.91515315e-02 -6.89282715e-01 6.85276762e-02 -1.35387576e+00
2.58380711e-01 -5.81322372e-01 -3.58159095e-01 4.17971194e-01
-3.57639343e-01 3.15466702e-01 4.52545732e-01 3.04275215e-01
-9.36982334e-01 6.75618112e-01 1.53530180e+00 -1.72763035e-01
-3.20423633e-01 1.26829490e-01 -1.70574170e-02 8.14699233e-01
8.92629147e-01 -6.74291193e-01 -3.38939786e-01 -5.62402964e-01
-3.90518695e-01 6.78142682e-02 4.35350627e-01 -1.32629013e+00
6.54071093e-01 -2.33876854e-01 4.47417080e-01 -6.82967365e-01
2.83456594e-01 -7.87783444e-01 1.29085526e-01 7.29965627e-01
1.61647908e-02 8.07850286e-02 2.27782950e-01 7.28649795e-01
-2.72596180e-01 -1.56820238e-01 1.14689410e+00 -5.67011796e-02
-1.13618004e+00 7.52287149e-01 -5.90637922e-02 1.52933866e-01
1.12264943e+00 -3.84707719e-01 -1.34701520e-01 -4.07960452e-02
-6.13746405e-01 4.67732668e-01 2.81282365e-01 5.64074516e-01
7.24599361e-01 -1.24090970e+00 -5.54910421e-01 1.16429977e-01
-1.36330113e-01 1.70073718e-01 6.22695744e-01 1.09378564e+00
-4.81443167e-01 2.46899813e-01 -3.77546966e-01 -8.76741588e-01
-1.14703512e+00 6.27697527e-01 4.42833126e-01 -1.74040556e-01
-6.85557306e-01 9.10508513e-01 3.62607330e-01 1.16352744e-01
4.35253531e-01 -1.74330309e-01 -1.87295750e-01 -8.11423510e-02
6.50265574e-01 5.02619445e-01 -1.36421040e-01 -6.89671934e-01
-4.10667807e-01 9.60446179e-01 -1.70095801e-01 1.72490865e-01
9.73472416e-01 -3.35428804e-01 9.31307524e-02 3.05172980e-01
1.03101683e+00 -3.10358226e-01 -1.66197813e+00 -4.40389544e-01
2.28498466e-02 -7.97839463e-01 1.33388475e-01 -5.24427712e-01
-1.60360301e+00 7.62050509e-01 1.05744112e+00 -1.61197796e-01
1.06247914e+00 -3.51697624e-01 1.23845661e+00 6.44306764e-02
3.45546842e-01 -1.19950831e+00 1.74700350e-01 4.31761920e-01
5.55170894e-01 -1.53522730e+00 -7.60423243e-02 -6.05439484e-01
-4.97274190e-01 1.03011763e+00 1.01022720e+00 -2.68201888e-01
6.15210295e-01 1.66490778e-01 3.42375934e-02 6.12811930e-02
-5.21839917e-01 -3.82796884e-01 3.93722832e-01 4.56604302e-01
2.19310336e-02 -3.37470293e-01 -4.13905591e-01 4.46778387e-01
4.40221906e-01 1.37767911e-01 2.64353424e-01 7.46991098e-01
-4.37952876e-01 -8.65258336e-01 -2.91118532e-01 3.67984354e-01
-4.62978840e-01 2.97802705e-02 2.76950836e-01 6.35904670e-01
3.70879740e-01 8.21564734e-01 1.11239120e-01 -3.56565833e-01
2.44033843e-01 -2.17494428e-01 4.19885308e-01 -2.91409403e-01
-4.18949068e-01 2.52508610e-01 -3.87172282e-01 -9.34598327e-01
-8.07965636e-01 -6.47807956e-01 -1.35768461e+00 -1.20151497e-01
-5.35222471e-01 -6.84490576e-02 5.28376043e-01 8.48138154e-01
4.95762229e-01 6.35804892e-01 5.67272007e-01 -8.76564920e-01
-3.51178586e-01 -7.06086397e-01 -2.79867232e-01 5.26058078e-01
1.95166275e-01 -8.67235661e-01 -2.20247805e-01 5.03621809e-02] | [9.200998306274414, -0.26618725061416626] |
de897c57-c5dc-4e0c-8e10-aa7bb27acf61 | learnable-path-in-neural-controlled | 2301.04333 | null | https://arxiv.org/abs/2301.04333v1 | https://arxiv.org/pdf/2301.04333v1.pdf | Learnable Path in Neural Controlled Differential Equations | Neural controlled differential equations (NCDEs), which are continuous analogues to recurrent neural networks (RNNs), are a specialized model in (irregular) time-series processing. In comparison with similar models, e.g., neural ordinary differential equations (NODEs), the key distinctive characteristics of NCDEs are i) the adoption of the continuous path created by an interpolation algorithm from each raw discrete time-series sample and ii) the adoption of the Riemann--Stieltjes integral. It is the continuous path which makes NCDEs be analogues to continuous RNNs. However, NCDEs use existing interpolation algorithms to create the path, which is unclear whether they can create an optimal path. To this end, we present a method to generate another latent path (rather than relying on existing interpolation algorithms), which is identical to learning an appropriate interpolation method. We design an encoder-decoder module based on NCDEs and NODEs, and a special training method for it. Our method shows the best performance in both time-series classification and forecasting. | ['Sunhwan Lim', 'Sungpil Woo', 'Noseong Park', 'Seungji Kook', 'Minju Jo', 'Sheo Yon Jhin'] | 2023-01-11 | null | null | null | null | ['irregular-time-series'] | ['time-series'] | [ 1.32554278e-01 7.31873959e-02 -9.59026814e-02 -1.12929523e-01
-9.94461626e-02 -1.49173573e-01 6.42914057e-01 -3.31027925e-01
-2.26238832e-01 8.34288895e-01 2.33885460e-03 -6.01400733e-01
-1.84813980e-02 -7.40746260e-01 -7.67927170e-01 -5.33306301e-01
-3.89372706e-01 -9.90127474e-02 1.05734877e-01 -4.12880629e-01
2.25537583e-01 6.88257873e-01 -1.63647127e+00 -1.02691345e-01
1.03139913e+00 1.12406158e+00 2.37898212e-02 6.73307002e-01
-4.85396087e-01 1.14525568e+00 -3.76295984e-01 -6.19646022e-03
1.35620683e-01 -7.79575288e-01 -5.45372903e-01 -4.60611343e-01
-2.29088828e-01 -4.16658401e-01 -3.34153682e-01 7.97315538e-01
2.33537033e-01 2.69000262e-01 9.38859344e-01 -1.27397311e+00
-1.02974522e+00 4.55481440e-01 -2.00326890e-01 1.99428484e-01
2.44140655e-01 -3.05356383e-01 5.52160680e-01 -1.00522947e+00
6.54036701e-01 9.44187105e-01 1.36652040e+00 6.08898163e-01
-1.22479594e+00 -2.74534792e-01 -3.70549895e-02 -1.53755113e-01
-1.41101539e+00 -4.69366133e-01 9.48228180e-01 -3.28135341e-01
9.27506566e-01 1.63207278e-01 7.10355401e-01 9.72406745e-01
5.95176995e-01 7.76525974e-01 7.63325930e-01 -5.75882912e-01
5.00339210e-01 3.97728421e-02 -1.96556710e-02 2.93158770e-01
-3.89586151e-01 2.04388291e-01 -1.63451791e-01 -1.41990811e-01
1.35647190e+00 2.52374798e-01 -3.10294271e-01 -3.12040187e-02
-1.04117513e+00 7.56913185e-01 3.21622521e-01 5.02291560e-01
-8.26386929e-01 2.89903373e-01 4.71928269e-01 7.44531929e-01
5.95035911e-01 1.79520056e-01 -3.69383633e-01 -1.70921162e-01
-1.14449668e+00 3.52190673e-01 1.05681026e+00 1.13887143e+00
6.36959136e-01 7.45793045e-01 -1.83685988e-01 7.15051472e-01
6.74068406e-02 2.49162376e-01 8.51929307e-01 -9.27377939e-01
-5.81572428e-02 2.95473754e-01 8.83849636e-02 -8.62014174e-01
-3.90817434e-01 -5.20282269e-01 -1.44961953e+00 8.16636458e-02
3.48529696e-01 -4.84628409e-01 -6.35487914e-01 1.69949496e+00
-7.75780110e-03 5.70084870e-01 1.18830904e-01 4.41502601e-01
2.65174896e-01 1.04603052e+00 -2.90628999e-01 -6.46936059e-01
6.45143509e-01 -8.39989126e-01 -1.12109816e+00 4.65947986e-01
6.53557956e-01 -3.86063129e-01 7.92931139e-01 2.38999113e-01
-1.12993097e+00 -7.22446084e-01 -9.13361669e-01 -1.90987647e-01
-5.41205883e-01 1.22149803e-01 4.76950407e-01 1.40598103e-01
-1.39318430e+00 1.16583884e+00 -8.31751525e-01 -7.87027180e-02
-2.78717935e-01 2.48946622e-01 1.38323009e-01 7.15567708e-01
-1.57803094e+00 8.90568733e-01 1.35509044e-01 4.02142346e-01
-3.16365391e-01 -1.00296807e+00 -7.24660158e-01 -1.15804255e-01
-8.29026103e-02 -3.20182413e-01 1.48244858e+00 -1.40907478e+00
-1.94384813e+00 1.75161511e-01 -4.66648400e-01 -9.05804455e-01
4.60919321e-01 8.91482607e-02 -7.14949489e-01 -1.32435575e-01
-1.72355220e-01 5.25146186e-01 9.08952594e-01 -7.54094481e-01
-3.17510754e-01 2.05295071e-01 -3.78626525e-01 -3.15471351e-01
-2.21097469e-01 -4.47784066e-02 3.04796305e-02 -1.09574068e+00
2.93750405e-01 -7.64667451e-01 -5.78842223e-01 2.04479441e-01
-1.28514290e-01 -4.68144357e-01 9.17253673e-01 -9.52541173e-01
1.62116706e+00 -1.97151566e+00 -5.96376918e-02 2.36227334e-01
-9.43227019e-03 3.34497064e-01 1.40280217e-01 7.38901854e-01
-3.60653996e-01 5.44678792e-02 -3.87761146e-01 -3.65748614e-01
-1.21291772e-01 2.24634647e-01 -7.00905323e-01 3.02875608e-01
4.48963225e-01 8.92494261e-01 -8.68884921e-01 -2.45968208e-01
-4.78448868e-02 6.87267721e-01 -3.33989292e-01 -7.01965988e-02
-4.46725428e-01 5.66803098e-01 -3.48234266e-01 1.94945082e-01
5.40347695e-01 -6.33031949e-02 -3.19434881e-01 3.83423232e-02
-8.68185043e-01 4.62884337e-01 -1.18476295e+00 1.44930351e+00
-5.57228863e-01 8.67879331e-01 -2.66218930e-01 -1.00590158e+00
1.35187221e+00 7.78277636e-01 4.91062462e-01 -6.56585932e-01
-5.15993983e-02 5.89783192e-01 -4.17982429e-01 -3.99532467e-01
5.23005962e-01 -4.07510139e-02 3.32495898e-01 4.84795511e-01
-5.26236594e-02 1.60172898e-02 4.99286018e-02 -1.10624537e-01
7.63399363e-01 3.48342210e-01 2.33821854e-01 -3.14696968e-01
6.47028506e-01 -1.25859007e-01 7.38657594e-01 6.08626962e-01
1.29514173e-01 7.09214687e-01 6.39387667e-01 -6.55836046e-01
-1.40612328e+00 -8.16336811e-01 -2.22384483e-01 4.84689057e-01
-3.49869221e-01 -2.18372926e-01 -6.86843812e-01 1.66093960e-01
-2.84695715e-01 8.71815860e-01 -4.78956312e-01 2.71303523e-02
-6.96474254e-01 -2.31891990e-01 8.81713569e-01 6.69243097e-01
5.39862096e-01 -1.23589265e+00 -3.68059397e-01 6.70745134e-01
-2.72290464e-02 -7.16234803e-01 -6.28273487e-01 2.91966736e-01
-1.15135455e+00 -2.94734299e-01 -1.20321071e+00 -8.88147950e-01
5.53499937e-01 -9.35727134e-02 8.86512339e-01 -1.01773910e-01
2.36989528e-01 7.73304701e-02 -2.84581065e-01 -4.81087536e-01
-6.48729801e-01 -1.61260844e-03 2.89723247e-01 1.26327857e-01
2.21718237e-01 -1.08105552e+00 -2.42575616e-01 7.67332688e-02
-1.00115180e+00 2.62777686e-01 2.16836572e-01 5.80602825e-01
7.96950936e-01 1.07820936e-01 9.51956093e-01 -6.20581865e-01
1.09353244e+00 -6.36534512e-01 -8.15269649e-01 2.49568850e-01
-7.16013610e-01 2.26755768e-01 1.24164438e+00 -7.89156675e-01
-8.18519652e-01 6.23106621e-02 -4.63476509e-01 -8.16681921e-01
1.22513756e-01 7.69937932e-01 5.60477793e-01 -1.04273837e-02
6.20889843e-01 5.35491705e-01 2.99189568e-01 -5.88249981e-01
1.54809415e-01 7.48182774e-01 5.53372085e-01 -3.26508760e-01
3.69272530e-01 2.62620211e-01 -4.62032929e-02 -1.02947712e+00
-1.45198792e-01 -1.17537297e-01 -6.59106433e-01 -2.03251556e-01
6.26093864e-01 -6.53855979e-01 -6.59914434e-01 6.91186905e-01
-1.71281517e+00 -3.69159907e-01 -6.59466207e-01 6.13847196e-01
-6.77072942e-01 9.74892676e-02 -9.43713844e-01 -1.23008311e+00
-2.76902556e-01 -6.70812070e-01 6.81271315e-01 2.26130947e-01
-3.03039938e-01 -1.29136097e+00 2.34747678e-01 -9.86180604e-01
8.01426947e-01 3.85955095e-01 6.79620266e-01 -3.22763294e-01
-1.83520749e-01 -1.81544229e-01 2.60413829e-02 7.79298842e-01
4.18950953e-02 2.99090177e-01 -9.15788054e-01 1.36585563e-01
6.55470014e-01 2.72175997e-01 5.17882586e-01 6.37174070e-01
1.24017692e+00 -5.32535553e-01 -1.45480037e-01 7.54631460e-01
1.40013552e+00 6.63591921e-01 9.83779013e-01 -1.28027529e-01
3.76771241e-01 4.83834863e-01 -4.25680727e-02 3.80780250e-01
2.78442025e-01 4.65365022e-01 -7.79063851e-02 -9.76940021e-02
1.68039471e-01 -4.77155566e-01 7.54201174e-01 1.44709790e+00
-4.26910251e-01 1.71836361e-01 -7.77523518e-01 4.49284703e-01
-1.87012863e+00 -1.10730600e+00 -5.75354993e-01 2.10760355e+00
9.12499547e-01 7.84138665e-02 1.22538961e-01 4.50449079e-01
6.76358461e-01 2.20363792e-02 -7.23221540e-01 -8.58818352e-01
-1.53976589e-01 2.95868069e-01 5.08142173e-01 3.71063530e-01
-7.60846019e-01 5.62402606e-01 7.00594711e+00 7.66522765e-01
-1.42295802e+00 -1.40570059e-01 3.41719866e-01 4.55597132e-01
-3.71757358e-01 -1.69135898e-01 -6.88449323e-01 6.30339980e-01
1.48608983e+00 -2.98294425e-01 5.47234118e-01 6.93140745e-01
5.96565723e-01 3.51333588e-01 -1.16842926e+00 7.06970632e-01
-3.10398757e-01 -1.48562956e+00 -1.13687381e-01 -2.15806678e-01
6.99687839e-01 -1.73666641e-01 -1.22399695e-01 2.63687164e-01
1.30083457e-01 -9.22432542e-01 7.87790716e-01 1.14379156e+00
8.39210391e-01 -5.82054853e-01 5.46782792e-01 6.43070936e-01
-1.29186654e+00 1.23380445e-01 -4.33232605e-01 -3.71147096e-01
3.26616079e-01 8.62573147e-01 -3.53017569e-01 5.92759252e-01
3.98380190e-01 9.35697854e-01 5.13208658e-02 9.46174800e-01
-6.53107390e-02 7.85253167e-01 -4.19089794e-01 -2.36309111e-01
2.82298148e-01 -5.14708400e-01 5.35147667e-01 1.04390621e+00
7.11621761e-01 -5.87869482e-03 -3.64724755e-01 1.35578728e+00
3.04642856e-01 -9.99339391e-04 -7.31990576e-01 -2.03497216e-01
4.77842987e-01 5.90647817e-01 -4.43627268e-01 -1.75828502e-01
-5.08070648e-01 9.71976638e-01 -8.44945982e-02 7.27956951e-01
-7.92626560e-01 -8.00898850e-01 3.15922946e-01 2.95611769e-02
2.68451720e-01 -4.51850057e-01 -3.04740101e-01 -1.02524066e+00
2.16181502e-01 -6.23626590e-01 -1.14891782e-01 -8.72058213e-01
-1.33857298e+00 9.40201879e-01 -2.26466388e-01 -1.71425009e+00
-8.30715060e-01 -4.12206292e-01 -8.30229998e-01 1.19743526e+00
-1.43307316e+00 -6.75270677e-01 2.46507242e-01 7.69756317e-01
2.63196945e-01 7.13359714e-02 9.43529189e-01 3.94077063e-01
-4.40305144e-01 3.74985695e-01 4.62011218e-01 1.42978653e-01
8.49179700e-02 -1.03502011e+00 6.66058898e-01 7.54415929e-01
-3.02099735e-01 6.48357093e-01 7.75343120e-01 -5.85979104e-01
-1.30937099e+00 -1.10103345e+00 1.32745945e+00 1.15121759e-01
7.51663923e-01 -3.34199905e-01 -1.32403994e+00 7.84749806e-01
7.70551665e-03 -1.71333060e-01 2.20879108e-01 -2.88356721e-01
1.82976667e-02 -7.37026781e-02 -8.83273602e-01 7.40325212e-01
9.48637605e-01 -5.81923008e-01 -5.51310241e-01 9.17497277e-02
1.00694692e+00 -4.30767983e-01 -8.67966771e-01 3.28995138e-01
6.15831733e-01 -8.49704683e-01 8.20273161e-01 -4.14292753e-01
6.34023786e-01 -2.93763727e-01 5.69747761e-02 -1.38561869e+00
-3.19443971e-01 -1.07632720e+00 -3.94577235e-01 1.18895447e+00
4.24305320e-01 -1.01826048e+00 3.58687818e-01 5.52537262e-01
-3.76379877e-01 -9.69870329e-01 -7.93941259e-01 -1.09922671e+00
2.44998157e-01 -5.48489273e-01 8.00516129e-01 1.05911446e+00
-1.50915697e-01 -1.29331663e-01 -5.28950691e-01 -3.11588850e-02
2.41109699e-01 -2.03076661e-01 3.06362122e-01 -1.39923882e+00
-6.79205060e-02 -4.56205517e-01 -6.60836548e-02 -1.51151478e+00
1.58498824e-01 -7.17007756e-01 6.41231090e-02 -1.38216662e+00
-8.97344947e-01 -7.83801258e-01 -2.70237595e-01 1.77387476e-01
4.92957026e-01 -2.93102503e-01 -1.91933304e-01 4.57816392e-01
1.32647172e-01 7.55724549e-01 1.20723283e+00 4.45264250e-01
-7.86687791e-01 3.74642134e-01 -2.34875038e-01 6.35909259e-01
8.18739712e-01 -2.68525034e-01 -4.90621775e-01 -2.32975602e-01
3.94443244e-01 6.07208550e-01 3.98262143e-01 -1.08474374e+00
6.74103141e-01 -1.43060029e-01 6.20559417e-02 -7.87857413e-01
2.23739475e-01 -7.13093281e-01 5.85343659e-01 4.71268892e-01
-3.91574085e-01 3.73449981e-01 9.67391878e-02 3.09304327e-01
-4.89631802e-01 -3.67653519e-01 4.59358513e-01 -6.37781844e-02
-5.67296088e-01 1.87545627e-01 -7.48861313e-01 -2.30430201e-01
6.19859755e-01 -4.61517960e-01 1.18075505e-01 -7.12189615e-01
-5.88631451e-01 -2.79676970e-02 1.02363728e-01 1.67535305e-01
5.74814141e-01 -1.58928835e+00 -5.48730671e-01 6.24810994e-01
-5.43596148e-01 1.21606663e-01 -9.89684165e-02 9.65936601e-01
-5.57896554e-01 7.03756928e-01 -1.17480665e-01 -3.03814083e-01
-2.88948596e-01 5.49940705e-01 6.26333654e-01 -1.28673866e-01
-7.40705729e-01 4.35996026e-01 -3.70775193e-01 -4.59563911e-01
2.74116099e-01 -8.10841978e-01 -2.14459494e-01 8.40014815e-02
5.72119474e-01 5.08544505e-01 2.90372148e-02 -3.97383600e-01
6.96593449e-02 4.67346847e-01 3.47046465e-01 -1.61269918e-01
1.49143016e+00 3.09149623e-02 -4.95533198e-01 1.06685865e+00
1.44840586e+00 -1.99187577e-01 -1.21062791e+00 -3.33999515e-01
9.95814949e-02 3.20670962e-01 -7.30474666e-02 -1.63084134e-01
-9.88778591e-01 8.48581493e-01 1.77861556e-01 8.83234262e-01
1.32316411e+00 -6.40517950e-01 1.03507400e+00 1.59200519e-01
2.79576302e-01 -1.06742334e+00 -4.61964637e-01 1.02544975e+00
1.00376439e+00 -4.76742029e-01 -5.02010345e-01 -2.30765387e-01
-3.00055981e-01 1.60339570e+00 1.22944243e-01 -5.33140898e-01
1.17583144e+00 5.87120354e-01 -2.48894487e-02 2.72691518e-01
-9.88680482e-01 1.69618756e-01 1.68607861e-01 4.26243573e-01
6.56165957e-01 -2.02501878e-01 -5.28341293e-01 6.29970014e-01
-2.77223825e-01 7.26716101e-01 4.91530359e-01 8.57123733e-01
-4.07769978e-02 -8.83292556e-01 -2.88728356e-01 3.87442827e-01
-2.85199523e-01 -9.54680443e-02 7.68266618e-02 5.23921430e-01
-2.79522166e-02 6.99515104e-01 4.29998606e-01 -3.99878055e-01
3.06413442e-01 3.52478474e-01 -4.39958200e-02 -7.54188225e-02
-4.76868689e-01 -1.69779778e-01 -2.27367133e-01 -2.92129040e-01
-4.83354181e-01 -5.98414361e-01 -1.39759052e+00 -4.95513469e-01
-2.49445289e-01 2.33409941e-01 8.42021108e-01 9.12335813e-01
3.52436692e-01 5.81649184e-01 8.00265431e-01 -9.47868109e-01
-6.34608269e-01 -7.98451185e-01 -7.02888370e-01 -7.34795928e-02
7.77496219e-01 -2.93096989e-01 -5.82211792e-01 2.45334938e-01] | [6.997147560119629, 3.2323763370513916] |
61e42bea-e36b-4c2d-906c-e3c250be9de2 | technical-report-auxiliary-tuning-and-its | 2006.16823 | null | https://arxiv.org/abs/2006.16823v1 | https://arxiv.org/pdf/2006.16823v1.pdf | Technical Report: Auxiliary Tuning and its Application to Conditional Text Generation | We introduce a simple and efficient method, called Auxiliary Tuning, for adapting a pre-trained Language Model to a novel task; we demonstrate this approach on the task of conditional text generation. Our approach supplements the original pre-trained model with an auxiliary model that shifts the output distribution according to the target task. The auxiliary model is trained by adding its logits to the pre-trained model logits and maximizing the likelihood of the target task output. Our method imposes no constraints on the auxiliary architecture. In particular, the auxiliary model can ingest additional input relevant to the target task, independently from the pre-trained model's input. Furthermore, mixing the models at the logits level provides a natural probabilistic interpretation of the method. Our method achieved similar results to training from scratch for several different tasks, while using significantly fewer resources for training; we share a specific example of text generation conditioned on keywords. | ['Or Sharir', 'Yoel Zeldes', 'Dan Padnos', 'Barak Peleg'] | 2020-06-30 | null | null | null | null | ['conditional-text-generation'] | ['natural-language-processing'] | [ 5.57532847e-01 7.53894746e-01 -1.92225397e-01 -4.57801223e-01
-9.75255132e-01 -6.16189420e-01 1.06441927e+00 -1.45606160e-01
-6.17119908e-01 1.05932903e+00 3.41464818e-01 -4.71138418e-01
4.09086704e-01 -7.76301444e-01 -1.02328849e+00 -5.38931012e-01
4.92179990e-01 9.66988802e-01 9.71668065e-02 -1.43541917e-01
3.02553892e-01 1.19154461e-01 -1.15101540e+00 5.12895465e-01
8.38001966e-01 5.06499827e-01 7.80779898e-01 8.91666114e-01
-2.43268460e-01 4.69884187e-01 -8.91100705e-01 -4.60890144e-01
-1.67410560e-02 -5.10554254e-01 -9.84359086e-01 -2.17424557e-01
2.37970203e-02 -7.72788823e-02 -5.96438497e-02 6.30761623e-01
2.13659480e-01 2.58808881e-01 1.26336122e+00 -1.13782740e+00
-7.65754521e-01 1.06623232e+00 -1.81667060e-01 -5.53655624e-03
2.05836520e-01 7.37602338e-02 1.00731981e+00 -9.68480349e-01
4.54684079e-01 1.48961294e+00 3.37711066e-01 8.90323520e-01
-1.83336484e+00 -5.63033521e-01 3.90038133e-01 -3.82728159e-01
-9.52315629e-01 -3.90729517e-01 4.96881962e-01 -5.22952974e-01
9.02184963e-01 -3.48865949e-02 1.80952549e-01 1.59173214e+00
1.67020753e-01 1.01401949e+00 1.26101220e+00 -8.21051776e-01
2.40564615e-01 8.78569186e-01 5.90109602e-02 3.23942602e-01
3.85886431e-02 1.44055456e-01 -6.20620012e-01 -2.88734198e-01
4.43223268e-01 -4.84515458e-01 3.74383992e-04 -7.92666301e-02
-1.41811109e+00 9.03315127e-01 1.02007084e-01 2.69445211e-01
-3.05467457e-01 3.90869826e-01 2.54747849e-02 5.13920225e-02
5.90472281e-01 5.54381073e-01 -1.00217342e+00 5.64086214e-02
-9.75270689e-01 4.20254976e-01 9.73910391e-01 1.20455801e+00
9.45399106e-01 5.78793287e-02 -7.67519116e-01 6.50716782e-01
4.85779494e-01 5.87751567e-01 5.93945205e-01 -7.31652081e-01
6.77494347e-01 7.88916722e-02 2.32390836e-01 -3.88046056e-02
-1.10790849e-01 -4.18522269e-01 -6.12629950e-01 6.73680454e-02
5.34391820e-01 -5.76463640e-01 -1.21959817e+00 2.06848621e+00
-3.04710846e-02 -3.24107945e-01 1.86285287e-01 1.46304429e-01
3.45889866e-01 9.71685410e-01 4.34607774e-01 -1.94351207e-02
1.13604832e+00 -1.00179505e+00 -5.84389508e-01 -5.88107586e-01
5.90674579e-01 -8.39877605e-01 1.43852282e+00 2.38384455e-01
-1.14473403e+00 -7.28638351e-01 -6.72722936e-01 -1.26367554e-01
-5.34409463e-01 5.31880498e-01 6.39318347e-01 6.40119553e-01
-1.23893023e+00 3.79468739e-01 -5.82910299e-01 -1.24689072e-01
5.29358201e-02 4.21931088e-01 -8.06660578e-02 2.63173580e-01
-1.43014228e+00 1.05828810e+00 7.76100159e-01 -5.13291836e-01
-1.03376114e+00 -6.90432966e-01 -8.45520377e-01 2.71139741e-01
2.54516006e-01 -1.07551694e+00 1.84830046e+00 -8.84134829e-01
-1.83622730e+00 5.80318809e-01 -6.42560422e-01 -2.79981464e-01
4.12879229e-01 -3.07041883e-01 4.46853340e-02 -3.12586844e-01
3.21335882e-01 1.17405200e+00 1.09346068e+00 -1.39444828e+00
-6.83642924e-01 1.18737459e-01 2.58077160e-02 2.46725559e-01
-2.16897130e-01 -1.08880177e-01 -3.67473990e-01 -1.02479708e+00
-5.69104075e-01 -8.02980125e-01 -3.14164430e-01 -5.00300646e-01
-6.38640940e-01 -6.21547878e-01 3.77273560e-01 -3.61338258e-01
1.13320482e+00 -1.94282210e+00 3.75496417e-01 2.51324438e-02
-5.90483062e-02 -1.10317640e-01 -3.02422285e-01 3.30708057e-01
-1.44528449e-01 4.29480344e-01 -4.54641730e-01 -7.66202629e-01
2.81729281e-01 8.86030197e-02 -6.97544575e-01 -3.64611208e-01
5.53360343e-01 1.06335759e+00 -8.65107298e-01 -5.88993013e-01
-9.71942320e-02 1.72342420e-01 -7.31040537e-01 4.97933328e-01
-7.70419717e-01 4.10057813e-01 -3.25294048e-01 2.26950552e-03
2.37572640e-01 -2.81312853e-01 -3.99623588e-02 2.54227132e-01
1.18711576e-01 6.97851837e-01 -9.18014526e-01 1.53115070e+00
-7.85182834e-01 3.93501312e-01 -1.11386478e-01 -8.58156979e-01
7.36315370e-01 3.83609235e-01 -6.88239709e-02 -2.09345259e-02
-1.97311655e-01 8.47882405e-02 -1.40518276e-03 -1.85387999e-01
5.62975764e-01 -4.16557640e-01 -3.77264827e-01 9.62340474e-01
5.10420442e-01 -5.13108075e-01 3.17202687e-01 5.41292191e-01
5.74022889e-01 5.28073668e-01 2.85098225e-01 -2.14317530e-01
2.53566861e-01 -1.77454814e-01 2.39313152e-02 1.08298171e+00
4.76058662e-01 2.68039614e-01 6.15260839e-01 9.18563604e-02
-8.26331854e-01 -1.36520994e+00 -2.44040024e-02 1.57001805e+00
-5.93809068e-01 -3.96411449e-01 -7.93227315e-01 -1.06844354e+00
-4.39323820e-02 1.40948820e+00 -8.12736452e-01 -9.59827472e-03
-4.40914273e-01 -7.35161424e-01 5.22549033e-01 5.35663843e-01
9.14706588e-02 -1.37639976e+00 -2.06709981e-01 1.92920312e-01
-4.65156794e-01 -7.89087415e-01 -7.10846603e-01 7.15122581e-01
-8.40873301e-01 -5.28417647e-01 -8.63132834e-01 -8.94519329e-01
9.45382178e-01 -2.59600103e-01 1.38779128e+00 -2.88018733e-01
1.73861846e-01 1.87474072e-01 5.21948040e-02 -6.91925824e-01
-9.34219182e-01 5.63191473e-01 -2.47372359e-01 -2.05366403e-01
2.90226310e-01 -2.62451023e-01 -6.71105906e-02 -5.08594885e-02
-1.04661608e+00 3.91658604e-01 9.12715137e-01 9.61946249e-01
2.70030439e-01 -2.57980138e-01 7.86874771e-01 -1.16357183e+00
9.57484961e-01 -6.53323412e-01 -4.78468418e-01 2.19435930e-01
-5.49296021e-01 7.10846305e-01 7.95784235e-01 -6.09685898e-01
-1.46738505e+00 3.02638471e-01 -2.86469515e-02 2.02771276e-01
-3.80796671e-01 4.92302179e-01 -1.77613646e-01 6.38680518e-01
7.20835984e-01 2.42866859e-01 -2.39946708e-01 -5.38525999e-01
6.81634605e-01 5.40485084e-01 4.19266850e-01 -8.99866819e-01
9.32952046e-01 -2.56896075e-02 -1.45715147e-01 -3.12160939e-01
-9.55798328e-01 -5.74508905e-02 -7.58092284e-01 2.38859370e-01
7.33371615e-01 -8.27795684e-01 -4.31001276e-01 3.17063540e-01
-1.55520129e+00 -9.08359647e-01 -4.84938979e-01 5.15556991e-01
-7.58578598e-01 1.16771668e-01 -3.51318836e-01 -8.37732494e-01
-7.46512115e-02 -1.03109646e+00 1.23622453e+00 -8.16289261e-02
-5.61522245e-01 -1.43018079e+00 1.02128208e-01 -5.60271293e-02
4.88906831e-01 -3.75996083e-01 1.41157687e+00 -8.90781164e-01
-6.12318039e-01 -9.68025774e-02 -3.02233435e-02 4.82197016e-01
1.77689254e-01 -1.97960719e-01 -1.09244716e+00 1.24144323e-01
1.60751149e-01 -5.04694700e-01 1.20479226e+00 4.09222454e-01
1.18081760e+00 -4.19905096e-01 -5.17237008e-01 2.65147954e-01
9.56194341e-01 -3.19011733e-02 2.70101160e-01 7.15740621e-02
4.36112791e-01 7.89299250e-01 3.72337043e-01 1.79460108e-01
4.26374406e-01 5.72551072e-01 -8.31094906e-02 -1.62950188e-01
-8.49067867e-02 -6.62017286e-01 7.94584930e-01 3.84631574e-01
9.51040313e-02 -5.31761229e-01 -6.55961454e-01 5.67308605e-01
-1.68777061e+00 -8.29652786e-01 1.31981224e-01 2.18497515e+00
1.36955082e+00 4.60464299e-01 -8.52244124e-02 -1.51473895e-01
5.08662701e-01 -6.96882829e-02 -3.62566143e-01 -4.34526116e-01
3.15780863e-02 6.66502774e-01 3.49942833e-01 9.41763341e-01
-9.39613938e-01 1.24899018e+00 7.94105864e+00 8.98337722e-01
-8.02209377e-01 1.29652351e-01 6.77015960e-01 -1.78969018e-02
-7.64974058e-01 1.70185000e-01 -1.26142406e+00 4.24194247e-01
9.58521247e-01 -4.49425876e-01 4.84362453e-01 8.31316113e-01
1.22167282e-01 -2.55579174e-01 -1.67040849e+00 2.70170748e-01
6.77813515e-02 -1.09693813e+00 4.94374454e-01 1.88679501e-01
7.23382950e-01 -3.00047100e-01 2.82092452e-01 8.73832762e-01
9.23767269e-01 -1.10796988e+00 7.84620404e-01 4.91100073e-01
8.70924652e-01 -4.74914730e-01 3.73770446e-01 7.75660574e-01
-6.74101233e-01 1.71287984e-01 -8.59587342e-02 -8.02584291e-02
2.27895483e-01 4.30972546e-01 -1.29637218e+00 1.82905093e-01
2.59488016e-01 2.56337017e-01 -6.05971813e-01 3.39850843e-01
-7.19514430e-01 8.07153165e-01 -2.56690830e-01 -1.53161600e-01
9.59172696e-02 5.59343807e-02 5.03512144e-01 1.49903250e+00
2.87755847e-01 -3.30932438e-01 2.59939641e-01 1.21726251e+00
-2.86753833e-01 1.58374328e-02 -7.33716846e-01 6.08154126e-02
2.50447869e-01 1.19084835e+00 -3.50271821e-01 -7.68492818e-01
-1.86467722e-01 8.62038732e-01 4.56379324e-01 7.86556363e-01
-7.13678956e-01 -5.40780127e-01 9.03130844e-02 2.50714682e-02
3.23006809e-01 -8.05025920e-02 -5.22262931e-01 -1.02419806e+00
-1.42796442e-01 -6.82753026e-01 2.79321641e-01 -9.45264935e-01
-1.25902641e+00 5.94918311e-01 4.38268483e-01 -6.28794014e-01
-9.62916613e-01 -6.81374848e-01 -6.72660053e-01 1.69342136e+00
-1.53948522e+00 -1.07412207e+00 2.75118381e-01 5.74957192e-01
7.68286049e-01 -3.82005811e-01 1.08355701e+00 -3.83471698e-01
-1.45147935e-01 7.06183553e-01 -1.63889125e-01 -2.36436985e-02
9.95188713e-01 -1.77939141e+00 8.00831556e-01 7.60816813e-01
7.59905577e-02 9.51023161e-01 6.64073288e-01 -7.23674417e-01
-7.34544039e-01 -1.20423412e+00 1.43106318e+00 -9.37662661e-01
6.00388288e-01 -7.76610076e-01 -5.65072656e-01 1.20672309e+00
4.92197156e-01 -7.55725741e-01 7.11288393e-01 2.84988165e-01
-3.01686138e-01 3.07625383e-01 -9.07468021e-01 8.31997573e-01
6.20961607e-01 -5.64120471e-01 -9.88073230e-01 6.05094671e-01
1.07708168e+00 -2.62216300e-01 -3.68444592e-01 1.23543590e-01
2.72008866e-01 -3.22515428e-01 8.42501521e-01 -8.98850858e-01
6.62378788e-01 -1.42173663e-01 -1.84677802e-02 -1.74755073e+00
-5.14754176e-01 -6.74980938e-01 -2.57919550e-01 1.20298266e+00
1.06558549e+00 -6.60586238e-01 6.08308077e-01 5.28184652e-01
-1.48441300e-01 -4.81296152e-01 -4.95513171e-01 -5.97974777e-01
5.33051312e-01 -4.70726520e-01 4.14162010e-01 3.49221826e-01
-1.48489013e-01 8.91530156e-01 -2.58476883e-01 -1.24776945e-01
4.87097591e-01 -2.07479537e-01 9.50050294e-01 -1.13996363e+00
-7.13721573e-01 -4.23496962e-01 6.45944178e-01 -1.49288476e+00
4.86322761e-01 -1.17381275e+00 5.03591180e-01 -1.60312104e+00
2.28896111e-01 -5.10298014e-01 -2.67174810e-01 7.06872225e-01
-6.55504644e-01 5.76622551e-03 2.88894713e-01 7.08767846e-02
-1.12531550e-01 5.19649327e-01 1.23134911e+00 -2.27961406e-01
-3.10886025e-01 5.78266382e-01 -1.16546583e+00 6.66390359e-01
7.94200540e-01 -7.36301959e-01 -6.20895803e-01 -4.87753391e-01
2.33980596e-01 -1.45590380e-01 2.26375237e-01 -5.37021875e-01
5.09402715e-02 -2.36659706e-01 6.00486755e-01 -5.37483275e-01
3.26443523e-01 -3.72015774e-01 -1.78928599e-01 2.53360212e-01
-1.03514969e+00 -1.00905828e-01 3.58243763e-01 5.66640556e-01
4.38756905e-02 -6.73502982e-01 5.84921837e-01 -2.33234823e-01
-8.49717036e-02 -3.61507162e-02 -8.62802684e-01 1.81669176e-01
6.89021707e-01 2.26997677e-02 -7.93871954e-02 -4.83956397e-01
-1.00417507e+00 1.26920924e-01 2.40421459e-01 5.05959809e-01
3.38098645e-01 -1.04463208e+00 -8.21132123e-01 2.28909627e-01
-1.87896658e-02 7.00366646e-02 -3.59173119e-01 3.24844390e-01
2.90894926e-01 7.63022125e-01 9.88697857e-02 -4.86217409e-01
-8.10713530e-01 6.54545546e-01 4.01219912e-02 -8.81830156e-01
3.29678990e-02 8.00258279e-01 7.62878895e-01 -6.59722626e-01
2.07193583e-01 -4.92709309e-01 -6.97748959e-02 -1.01936698e-01
2.96320915e-01 -6.98656663e-02 -1.36800289e-01 -1.34520158e-01
3.35359424e-02 1.42694935e-01 -3.35435152e-01 -9.11899626e-01
9.49632585e-01 -3.46657995e-04 9.54155251e-03 8.68816614e-01
7.66125917e-01 2.34081104e-01 -1.10495412e+00 -3.32609177e-01
2.66917739e-02 -5.75889163e-02 -3.12707543e-01 -1.21813214e+00
-3.09605032e-01 1.12603140e+00 -2.21117213e-01 8.04346651e-02
7.46822417e-01 1.69245034e-01 2.65596867e-01 5.97616196e-01
2.42425933e-01 -1.10550725e+00 3.25837433e-01 9.02531207e-01
9.70846474e-01 -1.03607500e+00 -2.14091673e-01 -2.25027457e-01
-8.60016048e-01 8.95263255e-01 7.08023906e-01 4.36796658e-02
6.53702617e-01 2.85506397e-01 5.40204272e-02 2.75079429e-01
-1.18659532e+00 5.81641169e-03 3.72321188e-01 8.18652332e-01
7.37203836e-01 5.12376316e-02 -5.76446764e-02 6.84227228e-01
-5.59015274e-01 1.73077565e-02 4.54200566e-01 6.24825954e-01
-2.27412209e-01 -1.38659823e+00 -3.73420298e-01 5.64580619e-01
-5.01461744e-01 -6.38054729e-01 -5.29737175e-01 6.97724581e-01
3.88676450e-02 8.39390159e-01 -8.00437760e-03 6.05938472e-02
8.80433992e-02 7.50100970e-01 3.97220492e-01 -1.32145965e+00
-6.48024619e-01 9.42952111e-02 -3.73856612e-02 5.51452935e-02
-8.24768841e-02 -4.92945611e-01 -9.08886671e-01 2.31928021e-01
-2.87793428e-01 4.16703522e-01 7.58493841e-01 1.04832017e+00
2.38969013e-01 5.57632208e-01 5.95981658e-01 -1.00304127e+00
-8.50003600e-01 -1.35565221e+00 -1.70843303e-01 2.63196707e-01
2.71091431e-01 -3.93303722e-01 -3.90836686e-01 5.94083667e-01] | [11.534234046936035, 8.983741760253906] |
820d09e0-81ff-408b-a0c2-a65247341ae7 | visibility-inspired-models-of-touch-sensors | 2203.04751 | null | https://arxiv.org/abs/2203.04751v2 | https://arxiv.org/pdf/2203.04751v2.pdf | Visibility-Inspired Models of Touch Sensors for Navigation | This paper introduces mathematical models of \sensors\ for mobile robots based on visibility. Serving a purpose similar to the pinhole camera model for computer vision, the introduced models are expected to provide a useful, idealized characterization of task-relevant information that can be inferred from their outputs or observations. Possible tasks include navigation, localization and mapping when a mobile robot is deployed in an unknown environment. These models allow direct comparisons to be made between traditional depth sensors, highlighting cases in which touch sensing may be interchangeable with time of flight or vision sensors, and characterizing unique advantages provided by touch sensing. The models include contact detection, compression, load bearing, and deflection. The results could serve as a basic building block for innovative touch sensor designs for mobile robot sensor fusion systems. | ['Steven M. LaValle', 'Manivannan M.', 'Prasanna Routray', 'Basak Sakcak', 'Kshitij Tiwari'] | 2022-03-04 | null | null | null | null | ['contact-detection'] | ['robots'] | [ 5.01596093e-01 -3.31518091e-02 -1.26690164e-01 -2.49625623e-01
-3.37457247e-02 -6.65512025e-01 4.97447938e-01 9.94556397e-02
-4.53962356e-01 4.99411315e-01 -6.41547024e-01 -3.10437828e-01
-2.93127149e-01 -7.09750295e-01 -4.17548597e-01 -8.36348593e-01
1.39759749e-01 4.02051657e-01 3.44616324e-01 -3.24330091e-01
6.29456580e-01 7.62535512e-01 -1.99825680e+00 -3.59749496e-01
4.18674290e-01 1.48159337e+00 9.88640189e-01 6.08301222e-01
7.60113355e-03 6.59747515e-03 -3.04759979e-01 2.49534324e-02
2.73529977e-01 2.93871731e-01 -9.44664106e-02 -1.15846820e-01
-2.01872841e-01 -2.99713552e-01 -1.74151976e-02 1.00940430e+00
2.99809009e-01 -3.89410615e-01 1.05500221e+00 -1.10815263e+00
-2.25433454e-01 9.77235287e-02 -5.34481049e-01 -1.88763946e-01
1.03104746e+00 -1.50918081e-01 5.79598725e-01 -1.15037096e+00
5.52337229e-01 1.16092598e+00 8.40456665e-01 4.20872748e-01
-8.85908842e-01 -9.02882665e-02 -2.44243652e-01 3.56521085e-02
-1.15264225e+00 -5.39945602e-01 8.90249848e-01 -4.74919677e-01
8.48158896e-01 4.48567897e-01 2.26331919e-01 1.04663897e+00
9.79878485e-01 2.30074823e-01 9.72609758e-01 -5.81864178e-01
2.72942513e-01 4.14841741e-01 2.21815646e-01 7.62316048e-01
9.22641456e-01 2.01431051e-01 -2.95465589e-01 1.84556562e-02
6.70998216e-01 1.44997030e-01 -3.97421271e-01 -7.67965972e-01
-1.13944435e+00 4.75891411e-01 3.23888719e-01 8.77460614e-02
-4.00140047e-01 3.96817029e-01 1.21848792e-01 3.17311347e-01
-3.11299980e-01 4.36088949e-01 -1.46217570e-01 -9.87432152e-02
-1.31789520e-01 4.01747376e-02 9.69452679e-01 1.15627658e+00
7.69219279e-01 -9.88909230e-03 5.43499529e-01 5.60711563e-01
8.53278875e-01 1.39588094e+00 3.18541497e-01 -1.05346775e+00
4.26187932e-01 3.98612469e-01 6.40758753e-01 -1.00334477e+00
-7.26887822e-01 1.95405498e-01 -4.16267455e-01 8.51077080e-01
3.29223201e-02 -4.77321118e-01 -5.70460439e-01 9.18097317e-01
-5.89263961e-02 -5.96967518e-01 3.64881307e-01 7.55067408e-01
7.94534504e-01 2.67378330e-01 -6.76814377e-01 -3.41821432e-01
1.26251638e+00 -2.62820393e-01 -8.14093530e-01 -4.01893586e-01
2.08747610e-01 -6.86345160e-01 8.15138698e-01 6.27892971e-01
-6.84223413e-01 -5.32239139e-01 -1.44751859e+00 1.74744114e-01
-6.87221408e-01 1.84734836e-01 4.77060050e-01 7.03763187e-01
-9.82731998e-01 1.61419898e-01 -9.02410865e-01 -7.10522830e-01
-5.11076272e-01 5.03507137e-01 1.90982362e-03 1.85034573e-01
-6.24332428e-01 1.25250721e+00 -1.04790322e-01 5.26424885e-01
-4.45166200e-01 3.12569171e-01 -6.90969825e-01 -3.93340379e-01
3.23132515e-01 -3.96968752e-01 1.19167888e+00 7.14717731e-02
-1.51465487e+00 7.24663198e-01 -2.63036579e-01 -4.16133076e-01
2.72153139e-01 -4.19088453e-01 -2.44305342e-01 3.28083664e-01
7.89464638e-03 3.50665122e-01 8.72995079e-01 -1.77563226e+00
-5.59402406e-01 -5.04357100e-01 -1.95061266e-02 3.56422216e-01
1.26113549e-01 -3.30300689e-01 -8.80866647e-02 8.27078596e-02
5.20434380e-01 -9.99838829e-01 -3.85121435e-01 2.54667431e-01
-4.43245858e-01 -1.84239522e-02 1.33992195e+00 1.36555180e-01
6.20550156e-01 -1.69899213e+00 -1.50597960e-01 5.25060534e-01
-2.16652732e-02 -5.05332649e-02 1.95342749e-01 1.05453396e+00
5.94202638e-01 -2.57919103e-01 -1.71921745e-01 -2.99415410e-01
-3.92022096e-02 3.16422611e-01 -3.80542278e-01 5.94045818e-01
-1.34844244e-01 7.17305541e-01 -5.40827990e-01 -3.15948546e-01
6.79324925e-01 2.29685485e-01 3.13523531e-01 -1.36842519e-01
5.40304184e-02 2.64823228e-01 -7.09260046e-01 1.05107856e+00
3.18857491e-01 4.10940588e-01 -1.94395810e-01 -2.46200338e-02
-3.86768281e-01 -1.53610751e-01 -1.18241620e+00 1.25433946e+00
-4.11241233e-01 8.37315500e-01 6.20226145e-01 -1.05398035e+00
1.68971455e+00 2.69288599e-01 4.33271945e-01 -6.08952880e-01
4.68559623e-01 4.59908128e-01 -6.79157197e-01 -9.90793943e-01
7.62306929e-01 3.06992494e-02 -4.16535705e-01 2.72210464e-02
-4.96362686e-01 -6.07157350e-01 -2.90878087e-01 -4.94013280e-01
8.79095733e-01 1.53118402e-01 2.27850631e-01 -3.80736470e-01
4.15660292e-01 2.54669785e-01 1.90656502e-02 8.65377903e-01
3.73950265e-02 1.98703021e-01 -1.26139134e-01 1.57093644e-01
-6.08197033e-01 -1.20455658e+00 -3.68927002e-01 6.05451167e-01
1.09468520e+00 2.85372138e-01 -2.94121623e-01 2.07692742e-01
2.64221549e-01 1.64472967e-01 -3.26352894e-01 -3.94425988e-02
-3.62909913e-01 -2.46955380e-01 4.88100022e-01 6.41102552e-01
4.28601086e-01 -8.69759738e-01 -1.55295420e+00 1.25935405e-01
6.95029572e-02 -1.14830935e+00 5.31951189e-01 8.19492459e-01
-1.01365912e+00 -1.27976871e+00 -5.74661672e-01 -8.40631485e-01
5.48047841e-01 9.79532659e-01 5.95354021e-01 -1.46787539e-01
-1.23595960e-01 1.27047253e+00 -5.86556196e-01 -8.73733222e-01
-3.38352546e-02 -2.80859530e-01 6.41226292e-01 -3.24502915e-01
4.41919386e-01 -3.50980490e-01 -5.74686825e-01 6.58276856e-01
-5.26580393e-01 -7.31065810e-01 5.31481445e-01 5.58047116e-01
5.54428458e-01 -9.55936238e-02 3.34041685e-01 -3.34468901e-01
8.92144382e-01 -3.34314108e-01 -7.45314538e-01 6.39992356e-02
-9.41227257e-01 -2.12536529e-01 4.42411751e-01 -2.08860725e-01
-1.09387863e+00 1.51431575e-01 1.14159293e-01 -1.01343423e-01
-3.09960902e-01 6.29243016e-01 -8.31954330e-02 -4.88914788e-01
7.67992496e-01 -1.47767588e-02 3.45236838e-01 -5.17231762e-01
4.75394651e-02 1.08956504e+00 8.28101575e-01 -4.30366725e-01
7.06423223e-01 8.63447011e-01 5.91501057e-01 -1.48550713e+00
2.98930913e-01 -8.57233465e-01 -8.87619913e-01 -3.39995414e-01
6.92910969e-01 -6.16238117e-01 -9.21789348e-01 3.78655612e-01
-1.09211528e+00 -1.06119834e-01 -3.23958516e-01 6.98691785e-01
-8.40392053e-01 4.48508859e-01 -3.19981575e-01 -1.50462997e+00
3.40683223e-03 -1.20899940e+00 1.20240557e+00 3.31382185e-01
-3.45518231e-01 -1.20329392e+00 -2.43352950e-02 2.70491261e-02
1.87861487e-01 2.28179961e-01 3.88360709e-01 -3.68968248e-02
-4.12698507e-01 -6.14878178e-01 1.00257002e-01 1.10266164e-01
1.15143292e-01 2.61948034e-02 -1.08049488e+00 -2.81561702e-01
2.83283591e-01 -8.70449394e-02 8.52269411e-01 4.01412219e-01
2.51351118e-01 2.11513326e-01 -1.02105355e+00 3.30170661e-01
1.74493802e+00 7.75094986e-01 3.21493030e-01 2.00683028e-01
2.91156918e-01 7.33218372e-01 1.10263312e+00 4.60930109e-01
5.35618253e-02 7.00489879e-01 1.07479632e+00 1.93281353e-01
4.02184099e-01 5.01189008e-02 4.10713315e-01 5.64945400e-01
-3.35847974e-01 -2.80065805e-01 -8.19552243e-01 3.13097447e-01
-1.64444244e+00 -6.52900279e-01 -5.23711264e-01 2.20046139e+00
-3.09467942e-01 1.71988562e-01 -2.69886196e-01 4.62202311e-01
7.12451756e-01 -7.62936175e-02 -7.39264071e-01 -6.43028975e-01
-1.59052107e-02 -1.47001088e-01 9.58074927e-01 6.74550533e-01
-7.82816648e-01 5.22960365e-01 7.44503784e+00 1.31155893e-01
-1.17677677e+00 -3.64728928e-01 -4.94312108e-01 5.50969660e-01
-3.12056780e-01 -2.12159634e-01 -9.26204085e-01 1.09542467e-01
5.24067938e-01 1.96638778e-01 2.90325321e-02 7.26693809e-01
2.70023316e-01 -9.05111194e-01 -1.15237558e+00 1.14690304e+00
8.57760832e-02 -5.56380332e-01 -2.64237285e-01 2.95031101e-01
-1.78482179e-02 -7.47849233e-03 2.65984923e-01 -2.37447143e-01
-2.21458957e-01 -6.63590074e-01 6.39556706e-01 9.15962636e-01
1.03086138e+00 -1.58787519e-01 8.40874851e-01 6.91757143e-01
-1.42232120e+00 -3.63172233e-01 -5.80552876e-01 -6.00600719e-01
2.78794259e-01 3.40576440e-01 -1.07841897e+00 4.79245007e-01
5.99627316e-01 4.06328410e-01 -1.37836576e-01 8.89078081e-01
7.88923576e-02 -7.44477436e-02 -7.87629843e-01 -7.38705814e-01
-1.35046924e-02 -3.95621181e-01 7.87703991e-01 1.08727443e+00
6.54124975e-01 -9.35789347e-02 -5.65729216e-02 6.90544426e-01
8.03142548e-01 -3.22618902e-01 -1.47628331e+00 2.71783412e-01
6.88210547e-01 1.15503263e+00 -9.67767358e-01 1.26585707e-01
-5.19218087e-01 8.71564627e-01 -5.03788710e-01 4.40213710e-01
-4.99690294e-01 -8.44874203e-01 4.80670363e-01 1.21894218e-01
2.98278391e-01 -8.00529242e-01 -7.53600061e-01 -6.82452738e-01
2.68390447e-01 1.75520048e-01 -5.23578227e-01 -1.03339052e+00
-8.89364421e-01 1.61752433e-01 2.23205104e-01 -1.68507731e+00
-5.12630165e-01 -1.26350844e+00 -4.79206860e-01 5.95721245e-01
-1.12022424e+00 -6.66495740e-01 -4.98222888e-01 4.77993488e-01
3.98389071e-01 -1.45890266e-01 9.63981569e-01 -4.66778457e-01
8.42929110e-02 1.30680278e-01 4.04137582e-01 -3.36662501e-01
3.97245646e-01 -1.11941838e+00 -8.15330520e-02 5.40997565e-01
-2.48238981e-01 5.98679900e-01 1.02936459e+00 -6.84360206e-01
-2.13359523e+00 -4.32237923e-01 4.00951058e-01 -6.82212472e-01
4.98378336e-01 -1.66079760e-01 -3.22597206e-01 4.32332188e-01
-8.55305642e-02 -5.36722243e-01 3.26963425e-01 -1.92112476e-01
1.77197874e-01 -3.72582465e-01 -1.22888386e+00 4.55678225e-01
9.09042060e-01 -5.57341099e-01 -8.37894559e-01 -2.20224246e-01
4.69662964e-01 -3.73761058e-01 -6.28081560e-01 4.43814129e-01
1.09943414e+00 -9.99692738e-01 9.40346420e-01 4.54796225e-01
-2.36148715e-01 -1.91709429e-01 -5.44427991e-01 -1.06892586e+00
-2.89603695e-02 -5.30339003e-01 1.66747868e-01 7.38785267e-01
4.94093955e-01 -1.03670049e+00 7.45286047e-01 3.59856009e-01
-3.94528806e-01 -6.23089075e-01 -9.85140324e-01 -9.98248279e-01
-4.67046082e-01 -6.51326001e-01 9.84428898e-02 4.91931349e-01
5.59708118e-01 4.70922947e-01 -7.25482628e-02 2.99974531e-01
7.30952978e-01 -1.32794127e-01 6.95901275e-01 -1.85873747e+00
4.67585288e-02 -2.81163901e-01 -5.23116887e-01 -1.53324687e+00
-2.62289822e-01 -1.24913245e-01 5.74448228e-01 -1.84361196e+00
-4.29280251e-01 -4.28704828e-01 8.30984265e-02 -8.08051601e-02
4.97571826e-01 3.04713190e-01 -5.47920242e-02 4.81131166e-01
-1.65605947e-01 1.00136951e-01 9.95052040e-01 8.23352262e-02
-3.42325896e-01 5.43937325e-01 -3.69475871e-01 1.01236618e+00
8.59136879e-01 1.17127895e-01 -5.63460290e-01 -2.65255988e-01
4.29611146e-01 4.41715300e-01 4.28752095e-01 -1.33232617e+00
5.52084744e-01 -1.25688300e-01 8.49537402e-02 -8.62799048e-01
1.00914598e+00 -1.61184335e+00 7.05970675e-02 6.68446183e-01
1.12190396e-01 2.04963952e-01 -2.59216070e-01 9.12015975e-01
9.08245053e-03 -6.72614217e-01 3.11784387e-01 -1.68691576e-01
-9.45289850e-01 -3.38115871e-01 -9.34303701e-01 -8.64929199e-01
1.13238966e+00 -1.17298198e+00 -3.72805864e-01 -5.32850325e-01
-6.10417604e-01 1.15666375e-01 6.60076499e-01 2.81121522e-01
1.11743867e+00 -9.01682794e-01 1.53260827e-01 2.77949005e-01
2.69175828e-01 9.07699093e-02 -3.70130241e-01 7.16509283e-01
-5.81939757e-01 7.98259795e-01 -3.47685695e-01 -9.42252278e-01
-1.18087065e+00 6.14663064e-01 2.11851090e-01 4.66092497e-01
-2.64512926e-01 5.55146635e-01 -9.10601690e-02 -2.22226098e-01
3.86814296e-01 -7.01719224e-01 -2.81642348e-01 -7.80295730e-02
-1.45064630e-02 6.71410620e-01 -8.63327309e-02 -6.52487814e-01
-5.19301772e-01 1.43685186e+00 5.58786869e-01 -5.65542400e-01
6.66136324e-01 -8.51626217e-01 2.54273176e-01 1.26000035e+00
7.24675357e-01 3.11457179e-02 -1.17888153e+00 -4.97769900e-02
6.49295598e-02 1.13797024e-01 -3.65635544e-01 -5.64930499e-01
-3.61822695e-01 1.00388193e+00 7.46948481e-01 5.79424739e-01
1.02147520e+00 1.74397767e-01 3.98045123e-01 1.22177577e+00
1.03731930e+00 -1.34099376e+00 -3.99066880e-02 5.38332582e-01
9.95299101e-01 -1.12327135e+00 -4.00179252e-02 -6.98241651e-01
-3.60069633e-01 1.41871727e+00 4.33223397e-01 -2.48368859e-01
7.73538709e-01 8.19480181e-01 2.66773999e-01 -2.41444841e-01
-3.84095997e-01 -3.90404135e-01 -1.48480773e-01 1.22669256e+00
1.79764293e-02 1.51360765e-01 -3.32592040e-01 2.07131833e-01
-1.00942671e-01 -2.34599158e-01 7.36803710e-01 1.37440503e+00
-1.34023404e+00 -6.98375583e-01 -8.61855805e-01 4.06857520e-01
3.22490185e-02 4.29712594e-01 -4.65740144e-01 9.11606550e-01
-5.03511727e-03 1.36712277e+00 -1.00527918e-02 -8.97374511e-01
5.28634071e-01 -9.48893651e-02 5.09951472e-01 -3.42980623e-01
2.17261329e-01 5.00678569e-02 2.21607387e-01 -4.86294866e-01
-4.92235333e-01 -6.24734342e-01 -1.36376131e+00 2.69551843e-01
-4.88090873e-01 1.43300623e-01 1.16445100e+00 9.26449478e-01
3.82819884e-02 8.53286833e-02 5.98533511e-01 -1.26646399e+00
-5.35931766e-01 -9.77363825e-01 -9.23214614e-01 -4.05506939e-01
5.57875037e-01 -1.18067098e+00 -4.20842230e-01 4.98196594e-02] | [7.288428783416748, -2.0604984760284424] |
38af3628-764a-41f3-9f56-a9ad586aee84 | a-semantic-backdoor-attack-against-graph | 2302.14353 | null | https://arxiv.org/abs/2302.14353v1 | https://arxiv.org/pdf/2302.14353v1.pdf | A semantic backdoor attack against Graph Convolutional Networks | Graph Convolutional Networks (GCNs) have been very effective in addressing the issue of various graph-structured related tasks, such as node classification and graph classification. However, extensive research has shown that GCNs are vulnerable to adversarial attacks. One of the security threats facing GCNs is the backdoor attack, which hides incorrect classification rules in models and activates only when the model encounters specific inputs containing special features (e.g., fixed patterns like subgraphs, called triggers), thus outputting incorrect classification results, while the model behaves normally on benign samples. The semantic backdoor attack is a type of the backdoor attack where the trigger is a semantic part of the sample; i.e., the trigger exists naturally in the original dataset and the attacker can pick a naturally occurring feature as the backdoor trigger, which causes the model to misclassify even unmodified inputs. Meanwhile, it is difficult to detect even if the attacker modifies the input samples in the inference phase as they do not have any anomaly compared to normal samples. Thus, semantic backdoor attacks are more imperceptible than non-semantic ones. However, existed research on semantic backdoor attacks has only focused on image and text domains, which have not been well explored against GCNs. In this work, we propose a black-box Semantic Backdoor Attack (SBA) against GCNs. We assign the trigger as a certain class of nodes in the dataset and our trigger is semantic. Through evaluation on several real-world benchmark graph datasets, the experimental results demonstrate that our proposed SBA can achieve almost 100% attack success rate under the poisoning rate less than 5% while having no impact on normal predictive accuracy. | ['Zhipeng Xiong', 'Jiazhu Dai'] | 2023-02-28 | null | null | null | null | ['graph-classification'] | ['graphs'] | [ 4.71787661e-01 2.87478447e-01 -1.77452117e-01 3.07024326e-02
1.14016213e-01 -1.04417956e+00 5.27115107e-01 4.08820599e-01
6.94046244e-02 3.85497600e-01 -5.04210174e-01 -6.59968376e-01
2.12643251e-01 -1.47413802e+00 -1.06896138e+00 -7.05837786e-01
-1.86199561e-01 3.82192037e-03 7.06921101e-01 -2.27135971e-01
1.18248209e-01 5.02259314e-01 -1.08801019e+00 3.52349728e-01
6.53083920e-01 7.30231404e-01 -2.81579137e-01 5.27076602e-01
-1.95456788e-01 5.63908517e-01 -1.08004260e+00 -7.25235105e-01
4.46833819e-01 -4.51216847e-01 -6.44196570e-01 -3.21189642e-01
3.26981902e-01 -1.63292289e-01 -6.51165009e-01 1.82009649e+00
4.75875475e-02 -2.72696674e-01 2.92022049e-01 -1.73659647e+00
-5.85989714e-01 6.63987994e-01 -2.90651083e-01 1.89407781e-01
3.35066199e-01 3.58967036e-01 8.41188133e-01 -2.87401408e-01
4.44485098e-01 1.39247990e+00 4.09884661e-01 6.78520679e-01
-1.01854599e+00 -1.10041380e+00 2.85520941e-01 8.93379971e-02
-1.37496471e+00 1.49958059e-01 8.68764818e-01 6.17687292e-02
5.22121966e-01 7.74763584e-01 4.97925699e-01 1.43116736e+00
6.26929581e-01 4.48580891e-01 8.59059513e-01 -8.79646242e-02
2.73049414e-01 2.52099633e-01 1.73195347e-01 6.96632147e-01
7.17882872e-01 3.78490239e-01 -4.44970326e-03 -5.24831176e-01
4.23885286e-01 1.89199746e-01 -3.83733809e-01 -8.63282606e-02
-6.95777833e-01 8.38095784e-01 8.89128745e-01 1.76867142e-01
-1.79533605e-02 3.11251402e-01 4.48810726e-01 3.83641630e-01
3.12747918e-02 2.00109154e-01 -3.87199342e-01 5.25673926e-01
-2.67729491e-01 1.21615909e-01 8.79994571e-01 7.30534136e-01
6.94767535e-01 1.90782890e-01 -1.41628250e-03 1.05756670e-01
2.42210969e-01 7.15102792e-01 3.90720785e-01 -2.16413476e-02
3.53177279e-01 9.83792841e-01 -3.66169542e-01 -1.84725320e+00
-1.57828569e-01 -7.08459318e-01 -7.95764446e-01 -4.93117841e-03
5.02543628e-01 -5.31266667e-02 -1.09497702e+00 1.96381307e+00
4.94972557e-01 7.77929246e-01 9.74295363e-02 7.93275177e-01
9.26561058e-01 6.39738977e-01 2.03378081e-01 9.27486271e-02
1.44286716e+00 -4.69406724e-01 -5.80852449e-01 -3.17794621e-01
7.14100122e-01 -3.82555753e-01 1.19181383e+00 2.83339024e-01
-2.53413826e-01 -1.87895417e-01 -1.42217517e+00 6.24114394e-01
-8.80661130e-01 -5.66816986e-01 7.02403426e-01 1.16843700e+00
-5.81718922e-01 4.96529132e-01 -5.64838171e-01 -2.82323986e-01
3.85193676e-01 2.97374308e-01 -4.71526057e-01 -1.92429498e-01
-1.70641434e+00 2.80500382e-01 6.69414520e-01 4.67115082e-02
-1.29233515e+00 -5.27467430e-01 -7.21527696e-01 9.79720801e-02
8.00686836e-01 -2.35356435e-01 6.10119760e-01 -1.17355847e+00
-8.91830564e-01 7.12988615e-01 2.02079907e-01 -6.75844431e-01
5.65371394e-01 2.75608897e-01 -1.07998216e+00 9.66022089e-02
-8.00632872e-03 1.68748219e-02 1.15906084e+00 -1.17389083e+00
-1.97812498e-01 -4.39748466e-01 5.17885089e-01 -3.52741092e-01
-6.89449549e-01 -1.37757793e-01 -3.70036401e-02 -6.59501851e-01
7.84098506e-02 -9.13235247e-01 -1.67163536e-01 -2.21229509e-01
-1.07995474e+00 -5.16749769e-02 1.39047599e+00 -2.32796371e-01
1.41220927e+00 -2.31968594e+00 -4.69165772e-01 7.33243704e-01
3.36598098e-01 6.00950778e-01 -6.07976690e-02 3.90517384e-01
-3.61939132e-01 6.72713041e-01 -2.63060898e-01 4.69220191e-01
-3.06446135e-01 2.28764445e-01 -7.60856330e-01 5.11090815e-01
-9.73960664e-03 8.48373592e-01 -9.27578092e-01 1.13137513e-02
2.21432224e-02 4.34040725e-02 -4.19037193e-01 3.37097235e-02
-4.52472180e-01 7.36834034e-02 -6.72581673e-01 5.52885771e-01
9.08497989e-01 -2.46022642e-01 2.22201675e-01 -1.72657222e-01
6.06909871e-01 5.39356284e-02 -1.26061881e+00 6.99556410e-01
-1.07483737e-01 2.07758009e-01 -1.64862216e-01 -8.48948479e-01
9.52782273e-01 2.26184070e-01 -2.80905038e-01 -2.63867617e-01
2.10345924e-01 2.56551176e-01 3.35576653e-01 -1.52110010e-01
1.28249809e-01 -5.57227526e-03 -1.96453884e-01 3.92244190e-01
-3.01595002e-01 4.48763967e-01 -1.75243065e-01 5.31803429e-01
1.47300184e+00 -5.19678593e-01 2.58716106e-01 -2.83605188e-01
9.45546091e-01 6.38807416e-02 6.84711933e-01 1.00503826e+00
-1.05370484e-01 1.78963378e-01 8.62247109e-01 -5.08776426e-01
-4.51185495e-01 -1.25496638e+00 1.63453877e-01 4.02856350e-01
6.44966424e-01 -6.15004063e-01 -8.78829956e-01 -1.41113496e+00
2.27195606e-01 6.42935097e-01 -6.39266849e-01 -1.04541147e+00
-3.43839556e-01 -5.62733233e-01 1.06144834e+00 1.93442628e-01
8.90979111e-01 -1.07309580e+00 -5.91575243e-02 -2.41803564e-02
1.75421059e-01 -1.20238769e+00 -4.31314498e-01 -2.54177153e-01
-6.16921604e-01 -1.71207869e+00 1.93921551e-01 -5.42156994e-01
9.86243069e-01 3.70598197e-01 7.40981162e-01 7.06348956e-01
-2.09336504e-01 -1.13912500e-01 -2.47884691e-01 -4.14515615e-01
-7.88176417e-01 -4.00343835e-02 4.59841714e-04 3.39932650e-01
4.22998101e-01 -3.61344486e-01 -4.09044802e-01 3.88549864e-01
-1.25649035e+00 -4.35979366e-01 2.99553722e-01 5.12469411e-01
2.64376789e-01 6.24614716e-01 7.08372831e-01 -1.53511143e+00
6.66422963e-01 -7.04177022e-01 -6.01386011e-01 2.71426350e-01
-4.35141921e-01 -1.49140283e-01 1.45653749e+00 -7.84268558e-01
-3.83818269e-01 -3.29032809e-01 -4.78031971e-02 -5.85735500e-01
-1.93409488e-01 3.81392181e-01 -7.16072440e-01 -3.22923094e-01
8.08286965e-01 3.67216468e-01 -1.22052684e-01 -3.32783498e-02
-3.26326117e-02 3.10055882e-01 3.28755766e-01 -3.89248580e-01
1.43931305e+00 6.41008615e-01 3.98828894e-01 -7.47104108e-01
-4.46910918e-01 1.68601692e-01 2.54712135e-01 -1.57676294e-01
6.34061694e-01 -3.95807207e-01 -6.42684042e-01 7.23635137e-01
-8.97785068e-01 1.26354665e-01 1.11101106e-01 7.80271664e-02
2.42315665e-01 5.67956090e-01 -4.28399175e-01 -5.59486985e-01
-3.67133409e-01 -1.22915018e+00 3.60563785e-01 3.58289868e-01
-1.02736577e-02 -1.09072804e+00 -5.25843322e-01 -5.26175536e-02
1.09140478e-01 6.46286368e-01 1.37865543e+00 -1.24620855e+00
-7.10253716e-01 -6.83922172e-01 6.00347221e-02 2.59259790e-01
3.25781971e-01 1.71878263e-01 -9.34987783e-01 -5.25102079e-01
-4.97229863e-03 1.25017837e-01 5.41200817e-01 -1.86824635e-01
1.41155660e+00 -6.63666308e-01 -6.32310271e-01 6.95079446e-01
1.38427269e+00 3.27386737e-01 7.15613008e-01 2.48964757e-01
8.87502432e-01 2.77774334e-01 4.48621660e-01 -9.65666994e-02
-2.64023900e-01 2.82909513e-01 1.00591028e+00 -1.66598737e-01
2.66952217e-01 -8.02811325e-01 3.49847078e-01 -5.17147966e-02
4.66754436e-01 -6.17858768e-01 -7.66663432e-01 1.41553685e-01
-1.45091677e+00 -6.93015099e-01 -3.23861986e-01 2.35014606e+00
4.28183347e-01 5.91728449e-01 -2.05620825e-01 4.54015285e-01
1.11131024e+00 4.24962014e-01 -6.07433975e-01 -5.35035193e-01
-1.09715424e-01 3.05030882e-01 7.33437777e-01 2.97376633e-01
-1.06228578e+00 1.14933574e+00 4.44426155e+00 1.01073730e+00
-1.19089508e+00 -8.77306238e-02 6.35065913e-01 4.16360766e-01
-6.58374786e-01 3.77050877e-01 -5.82611024e-01 6.16809309e-01
6.36171877e-01 -2.78950810e-01 4.14707333e-01 9.00826812e-01
-8.70375708e-02 3.87986571e-01 -9.25722182e-01 4.08535302e-01
-5.05594499e-02 -1.15000248e+00 5.41361392e-01 1.85555205e-01
4.29264039e-01 -4.62194055e-01 1.33162260e-01 2.95349061e-01
2.66120106e-01 -1.13662410e+00 5.01481175e-01 -1.42346203e-01
6.63510740e-01 -1.11262429e+00 7.14040816e-01 5.65534532e-01
-1.23837984e+00 -2.08476454e-01 -3.61941040e-01 1.48419395e-01
-5.22908509e-01 5.81964731e-01 -7.22945631e-01 7.21278727e-01
4.92087901e-01 4.97403175e-01 -6.75503194e-01 5.54634869e-01
-5.60453057e-01 7.49476314e-01 -3.92055631e-01 -1.29453108e-01
4.11887795e-01 3.49689461e-02 9.88449097e-01 7.94682622e-01
-4.33926024e-02 -2.33476572e-02 2.74045497e-01 8.68973613e-01
-4.04831827e-01 6.53410107e-02 -1.08775139e+00 -2.31985107e-01
6.60158277e-01 1.10688972e+00 -1.00149536e+00 -9.02308151e-02
-1.78547293e-01 8.03246140e-01 5.18437102e-02 4.29353803e-01
-9.20242786e-01 -5.75274467e-01 6.76391482e-01 1.92529976e-01
-9.45149511e-02 2.01863468e-01 -1.03472307e-01 -1.03920412e+00
4.52872999e-02 -1.07861495e+00 7.02485323e-01 -3.17420304e-01
-1.23056197e+00 4.71264899e-01 -3.05446863e-01 -1.08178639e+00
2.08092824e-01 -4.89334732e-01 -1.13412428e+00 7.39549518e-01
-9.54554915e-01 -9.54353809e-01 -3.33770782e-01 9.55529630e-01
-4.06743623e-02 -1.54495165e-01 8.14625680e-01 1.33103833e-01
-7.02903211e-01 1.12074792e+00 -4.28777575e-01 5.57041109e-01
2.82839835e-01 -8.23689580e-01 5.60547233e-01 1.37096298e+00
2.38590166e-01 8.46843839e-01 7.42458403e-01 -1.09261584e+00
-1.47487879e+00 -1.35241425e+00 4.46559012e-01 -1.08787648e-01
7.36142576e-01 -6.75238371e-01 -1.12430203e+00 8.00417840e-01
-4.32581842e-01 3.29947978e-01 3.58358264e-01 -3.78779173e-01
-6.80286288e-01 -1.66589752e-01 -1.65787804e+00 8.38234603e-01
9.95281696e-01 -5.00478387e-01 -1.65955871e-01 3.57227802e-01
1.04964614e+00 -2.92643815e-01 -3.45813602e-01 5.63690543e-01
2.01588079e-01 -7.76343882e-01 9.14637983e-01 -9.94224429e-01
2.06920460e-01 -5.49876213e-01 -1.00755850e-02 -1.06876779e+00
2.19302960e-02 -5.20508230e-01 -3.04523587e-01 1.15295088e+00
3.61011922e-01 -1.32577348e+00 8.48659158e-01 1.00176640e-01
2.29759619e-01 -6.02356672e-01 -8.83284628e-01 -9.66939509e-01
-1.03513435e-01 -3.71327341e-01 1.16948938e+00 1.22468519e+00
-4.04344648e-01 -7.82200992e-02 -1.30484745e-01 8.37630749e-01
7.70862103e-01 -8.75843782e-03 8.10051918e-01 -1.12077606e+00
-1.57967940e-01 -2.33672783e-01 -8.19089293e-01 -4.35601145e-01
2.32384518e-01 -1.19191813e+00 -4.90432024e-01 -9.03311312e-01
-1.72733918e-01 -4.73916590e-01 -4.31654245e-01 5.15042663e-01
-3.41714710e-01 2.23763824e-01 4.27037515e-02 4.73074429e-02
4.44725063e-03 2.19291821e-01 1.11101735e+00 -3.34189862e-01
-1.92304924e-02 3.27721149e-01 -7.69938350e-01 6.82669401e-01
9.48572397e-01 -8.01199257e-01 -7.99350381e-01 1.04748785e-01
2.11477250e-01 -1.09415993e-01 7.00109541e-01 -7.99176812e-01
-3.98499742e-02 -2.13275298e-01 2.08606422e-01 -2.72139817e-01
-1.98968202e-01 -1.09745026e+00 3.99785519e-01 1.03886056e+00
7.06424797e-03 3.06521654e-02 1.66228876e-01 1.09501386e+00
7.21960217e-02 -1.46822050e-01 6.62861049e-01 -1.07361630e-01
-6.04910016e-01 5.48300862e-01 -1.12855606e-01 9.62687582e-02
1.34995759e+00 -3.16850513e-01 -7.06660867e-01 -2.43770763e-01
-5.54897845e-01 1.64262280e-01 4.75312263e-01 5.92081308e-01
6.97726488e-01 -1.14214623e+00 -4.75246847e-01 7.03921437e-01
2.71717161e-01 -1.81814253e-01 6.35716170e-02 2.22800434e-01
-5.37222028e-01 1.42907321e-01 9.37537327e-02 -4.35077727e-01
-1.33549643e+00 7.63436556e-01 6.29524648e-01 -1.77981332e-01
-7.53142297e-01 5.74494064e-01 4.12844330e-01 -4.59295064e-01
5.50796352e-02 -2.01084986e-02 -1.65566847e-01 -4.82233167e-01
4.16230232e-01 1.24640696e-01 2.02277675e-02 -4.17077512e-01
-6.21467113e-01 7.87484646e-02 -2.13531762e-01 5.07613778e-01
6.15098894e-01 5.53679585e-01 -3.17972571e-01 -1.38042029e-02
1.16029286e+00 1.94844559e-01 -6.29685700e-01 -1.81900784e-01
-1.84850708e-01 -7.63218999e-01 -2.48154819e-01 -6.62393034e-01
-1.33036768e+00 6.54444635e-01 4.57949698e-01 6.94252074e-01
9.40578759e-01 -1.78206131e-01 9.28191066e-01 2.75010079e-01
5.68321526e-01 -5.27997553e-01 -4.93826494e-02 2.90794015e-01
4.08198893e-01 -7.78851569e-01 -2.63360053e-01 -8.64182532e-01
-2.20963389e-01 9.02677357e-01 9.44864690e-01 -4.34957653e-01
6.66889369e-01 5.04903570e-02 -1.47122100e-01 -2.59522051e-01
-5.10113776e-01 4.38027024e-01 -4.38729674e-02 5.72977483e-01
-3.61229986e-01 3.37961346e-01 -2.11069882e-01 7.41791964e-01
-3.44570488e-01 -4.58164155e-01 6.54553413e-01 6.88879550e-01
-2.46966884e-01 -9.66321230e-01 -3.78549218e-01 5.78617096e-01
-7.62069404e-01 -8.51941332e-02 -6.90458596e-01 9.05451298e-01
2.17938703e-02 9.74223137e-01 -1.29310995e-01 -8.55875850e-01
2.30482399e-01 -7.96884969e-02 -3.64974253e-02 -6.84495449e-01
-8.53392899e-01 -5.94424605e-01 -6.26974255e-02 -6.34033501e-01
3.23782682e-01 -7.75563493e-02 -1.53919971e+00 -5.87939322e-01
-4.37793970e-01 3.54876220e-01 3.70391160e-01 8.56772244e-01
1.48165122e-01 5.76012313e-01 1.04112220e+00 3.15275788e-01
-6.79831445e-01 -5.38705230e-01 -5.72552800e-01 5.70872724e-01
1.36583120e-01 -6.46923184e-01 -8.80142331e-01 -5.95392227e-01] | [6.127984046936035, 7.381504535675049] |
9321ff68-8d1f-4853-91de-e60600664477 | gene-teams-are-on-the-field-evaluation-of | 2301.11763 | null | https://arxiv.org/abs/2301.11763v1 | https://arxiv.org/pdf/2301.11763v1.pdf | Gene Teams are on the Field: Evaluation of Variants in Gene-Networks Using High Dimensional Modelling | In medical genetics, each genetic variant is evaluated as an independent entity regarding its clinical importance. However, in most complex diseases, variant combinations in specific gene networks, rather than the presence of a particular single variant, predominates. In the case of complex diseases, disease status can be evaluated by considering the success level of a team of specific variants. We propose a high dimensional modelling based method to analyse all the variants in a gene network together. To evaluate our method, we selected two gene networks, mTOR and TGF-Beta. For each pathway, we generated 400 control and 400 patient group samples. mTOR and TGF-? pathways contain 31 and 93 genes of varying sizes, respectively. We produced Chaos Game Representation images for each gene sequence to obtain 2-D binary patterns. These patterns were arranged in succession, and a 3-D tensor structure was achieved for each gene network. Features for each data sample were acquired by exploiting Enhanced Multivariance Products Representation to 3-D data. Features were split as training and testing vectors. Training vectors were employed to train a Support Vector Machines classification model. We achieved more than 96% and 99% classification accuracies for mTOR and TGF-Beta networks, respectively, using a limited amount of training samples. | ['Yelda Tarkan Arguden', 'Ayse Cirakoglu', 'Emrah Yucesan', 'Cagri Gulec', 'Suha Tuna'] | 2023-01-27 | null | null | null | null | ['medical-genetics'] | ['miscellaneous'] | [ 2.35872343e-01 5.33141568e-02 -1.10601433e-01 -1.67805597e-01
-2.71095991e-01 -4.19829160e-01 3.02010030e-01 2.08911613e-01
-2.39959508e-01 8.38678062e-01 1.06602542e-01 -1.58756331e-01
-6.28329933e-01 -8.25336516e-01 -3.97997141e-01 -1.02023971e+00
-4.42182660e-01 7.38343179e-01 -1.23173602e-01 -1.19070165e-01
-1.18195176e-01 5.64704180e-01 -1.41534817e+00 3.15360993e-01
1.00527596e+00 7.73754060e-01 -1.69109792e-01 7.97697365e-01
-7.37394914e-02 2.91186363e-01 -6.16429687e-01 -4.10030961e-01
1.00528598e-01 -4.47692245e-01 -3.07126999e-01 3.03130299e-01
6.33480027e-02 4.90442038e-01 -9.09434855e-02 8.68338943e-01
5.62845230e-01 -1.88422635e-01 9.49301243e-01 -1.12088776e+00
-3.02217305e-01 4.67429608e-01 -4.58622187e-01 -9.71220806e-02
2.48086363e-01 2.15295911e-01 1.14936697e+00 -8.84475648e-01
9.36634719e-01 1.09765553e+00 5.68538070e-01 3.44833940e-01
-1.58968222e+00 -3.03801477e-01 -1.63158372e-01 2.17264429e-01
-1.60443878e+00 -1.39403224e-01 6.37722790e-01 -7.96719074e-01
9.23693061e-01 4.76854742e-01 1.27172542e+00 8.74699116e-01
4.75181878e-01 3.86785001e-01 8.82918417e-01 -2.03564689e-01
1.33984670e-01 -1.67746246e-01 5.60022332e-02 8.90432417e-01
4.31980789e-01 4.56600785e-02 -3.75729501e-01 -5.44341683e-01
6.66781187e-01 -1.14545785e-01 -2.32822716e-01 -2.74874151e-01
-1.49708128e+00 6.09445989e-01 7.82359764e-02 6.26310945e-01
-4.63320225e-01 -2.91371018e-01 2.36322552e-01 2.53407925e-01
3.07287067e-01 3.68905485e-01 -3.44487756e-01 1.22960083e-01
-7.06488252e-01 3.02202087e-02 8.00545573e-01 4.81379181e-01
6.68513775e-01 -1.97838619e-01 -1.10342413e-01 9.20409203e-01
3.10259223e-01 7.73610622e-02 6.57897472e-01 -5.12644231e-01
9.14136618e-02 1.13485765e+00 -3.76900792e-01 -1.39557183e+00
-8.31620216e-01 -8.03742707e-01 -1.50621676e+00 -1.19949937e-01
3.13501745e-01 -1.98990986e-01 -7.34796107e-01 1.68536389e+00
5.19586444e-01 4.44095105e-01 8.75878409e-02 7.46924639e-01
5.80309331e-01 3.28300446e-01 -2.79755116e-01 -3.73397946e-01
1.47121930e+00 -3.36342126e-01 -5.12963653e-01 4.61916208e-01
8.19162607e-01 -6.32432580e-01 4.57131565e-01 6.01456881e-01
-1.07066035e+00 -3.20238173e-01 -1.02531397e+00 3.19551557e-01
-3.79166752e-01 2.95721263e-01 6.94367707e-01 5.25498629e-01
-1.03332269e+00 6.60138488e-01 -7.23326266e-01 -2.73387611e-01
3.17294419e-01 5.50867558e-01 -6.08882427e-01 2.90780813e-02
-1.34407163e+00 6.59874797e-01 2.91835010e-01 3.55107307e-01
-7.01543212e-01 -6.49703979e-01 -5.69754541e-01 -1.22400254e-01
9.77938846e-02 -1.02574301e+00 1.84907138e-01 -5.40614128e-01
-1.14475858e+00 8.43519092e-01 -7.90838376e-02 7.28177233e-03
2.34221503e-01 4.35472846e-01 -8.55242193e-01 1.64688751e-01
-4.65561189e-02 1.96158960e-01 6.62812591e-01 -8.11394989e-01
-4.62841213e-01 -5.45995951e-01 -2.85916388e-01 6.41514137e-02
-2.38215715e-01 -2.05871254e-01 -2.96187133e-01 -6.71919763e-01
4.31458294e-01 -9.74734128e-01 -4.81918663e-01 1.08517811e-01
-7.79366493e-01 1.33000512e-03 3.39219838e-01 -5.03674805e-01
1.23284447e+00 -2.16525388e+00 8.18811893e-01 6.19588315e-01
7.14456499e-01 1.74321920e-01 -1.89484954e-01 2.80062407e-01
-4.46959108e-01 4.10679936e-01 -2.70947009e-01 1.71720952e-01
-3.44031036e-01 2.13322371e-01 4.91338402e-01 5.87092638e-01
4.94990885e-01 7.42712080e-01 -7.12036431e-01 -4.94737387e-01
5.98167302e-04 5.81653237e-01 -6.70308828e-01 6.81360066e-02
-2.07816646e-01 2.24378347e-01 -6.41285598e-01 1.02435148e+00
5.26540518e-01 -3.74366343e-01 6.18349552e-01 -4.39986616e-01
1.45682931e-01 -3.88645351e-01 -1.23584425e+00 1.42090285e+00
-1.19988769e-01 3.06539506e-01 7.71482736e-02 -1.29882300e+00
1.05049431e+00 6.21634543e-01 7.28407621e-01 -1.42907903e-01
7.02882931e-02 3.30383092e-01 5.63438058e-01 -6.88020051e-01
-2.41850689e-02 -1.30651668e-01 -5.13818376e-02 2.29378700e-01
8.87461007e-02 7.94273168e-02 6.13538504e-01 2.16483809e-02
1.34834933e+00 -3.19575369e-01 3.44259024e-01 -1.79680273e-01
6.44736826e-01 1.70529619e-01 6.77731395e-01 3.17771584e-01
-3.74410325e-03 5.79904795e-01 1.15070367e+00 -4.97868329e-01
-8.84707153e-01 -9.38916087e-01 -2.97052830e-01 4.03710723e-01
-2.75790006e-01 -4.00652975e-01 -3.22839946e-01 -5.19771934e-01
2.09989652e-01 1.25078812e-01 -7.54150987e-01 -2.46616811e-01
-3.23717892e-01 -1.44292057e+00 7.26027906e-01 -2.36506283e-01
1.73873857e-01 -3.95447701e-01 3.99857573e-02 3.80848050e-01
1.70855254e-01 -4.94859248e-01 -8.75103697e-02 1.36289611e-01
-8.98804128e-01 -1.56495500e+00 -7.03139603e-01 -7.55015314e-01
8.89167368e-01 -1.85202703e-01 7.55732179e-01 1.51268899e-01
-7.65533864e-01 6.93911361e-03 -2.33382568e-01 4.99974936e-02
-4.55302835e-01 -1.47071227e-01 4.21805888e-01 3.44372094e-01
-1.95357972e-03 -9.03069377e-01 -4.74212706e-01 4.67494339e-01
-9.09935296e-01 -2.77852081e-02 9.29352105e-01 1.13736820e+00
6.07366681e-01 1.90131769e-01 2.99018145e-01 -5.96116662e-01
7.03120530e-01 -6.26173675e-01 -2.76661605e-01 3.05964023e-01
-3.21084768e-01 -1.20647475e-01 5.25070906e-01 -5.62538862e-01
-5.84401548e-01 1.34430632e-01 -4.04180784e-04 -3.50685090e-01
-1.15699828e-01 9.87701178e-01 -1.56077415e-01 9.20925289e-02
5.95051467e-01 2.01884970e-01 5.04052937e-01 -4.03939188e-01
1.46285892e-01 3.94325167e-01 6.80745989e-02 -3.36474389e-01
4.64181572e-01 1.94324479e-01 5.59129834e-01 -1.05913353e+00
-1.93766594e-01 -2.33060390e-01 -7.79977798e-01 -3.14920276e-01
8.45933318e-01 -6.60045564e-01 -8.57317507e-01 5.33644676e-01
-8.15430105e-01 3.52340519e-01 2.83358432e-02 7.49275029e-01
-3.07691157e-01 1.58449411e-01 -5.27964890e-01 -2.37712011e-01
-5.70783131e-02 -1.18406999e+00 5.58410764e-01 -1.20147765e-01
-2.83477366e-01 -1.09601426e+00 3.99591386e-01 3.49881360e-03
2.30910122e-01 5.73799729e-01 1.47991431e+00 -8.74773741e-01
-2.48529553e-01 -4.57488090e-01 1.17775455e-01 1.89508036e-01
3.84305239e-01 4.51659083e-01 -4.45865542e-01 -1.45226838e-02
-2.11297110e-01 3.53141651e-02 7.07959592e-01 3.01753730e-01
8.38233888e-01 -2.44741425e-01 -6.75686359e-01 5.08432090e-01
1.22662079e+00 2.78074205e-01 4.36985999e-01 -9.70064551e-02
7.49233663e-01 6.76790535e-01 3.73044014e-01 3.44886512e-01
-1.04929566e-01 7.60690570e-01 3.00343722e-01 -2.28062555e-01
1.22369893e-01 1.25904962e-01 1.24120429e-01 8.27226758e-01
-4.89158660e-01 -2.95951992e-01 -9.30540025e-01 2.56131619e-01
-1.60553765e+00 -7.41217196e-01 -4.34805691e-01 1.99895489e+00
6.59232318e-01 -5.58627099e-02 1.17220730e-01 1.94208592e-01
8.79807413e-01 -7.09257647e-02 -5.33877075e-01 -2.86962446e-02
-6.71914816e-01 -3.92828137e-02 -3.18578035e-02 1.40380457e-01
-7.53189802e-01 3.61659944e-01 6.85748196e+00 8.27026963e-01
-1.21029735e+00 -2.60928899e-01 8.04146647e-01 -3.56565922e-01
-3.52330238e-01 -2.54957646e-01 -4.54919875e-01 2.96638399e-01
8.77469242e-01 -2.53948539e-01 8.22042525e-02 3.01922143e-01
3.55784208e-01 3.44853364e-02 -9.13034320e-01 8.40596914e-01
-2.86030054e-01 -1.50466597e+00 1.97693571e-01 3.41455758e-01
5.53639472e-01 -1.06169812e-01 -3.38512696e-02 -6.56695589e-02
2.09634468e-01 -1.14483452e+00 5.54664694e-02 8.50008428e-01
6.57470822e-01 -6.98584974e-01 6.85847998e-01 1.40098482e-01
-1.02361643e+00 1.43243507e-01 -4.07323807e-01 1.47591248e-01
2.48850450e-01 1.25106049e+00 -9.85456347e-01 8.51905465e-01
2.68706858e-01 9.43917632e-01 -4.15480345e-01 7.79936075e-01
3.91378067e-02 5.26179314e-01 -3.21807027e-01 -2.10131958e-01
-1.18137747e-02 -5.64640284e-01 7.85028636e-01 8.52644563e-01
5.81243336e-01 2.16085657e-01 -6.57048672e-02 8.01636398e-01
2.59730160e-01 2.45344475e-01 -4.20597881e-01 -4.50309187e-01
1.63318619e-01 1.58837116e+00 -7.20332146e-01 -3.46936166e-01
-3.69944930e-01 6.88629687e-01 2.75582582e-01 3.58087629e-01
-5.04842281e-01 -3.94524902e-01 1.03996408e+00 -3.59731093e-02
1.43499985e-01 -9.76327434e-02 -4.09974623e-03 -1.17296457e+00
-1.12023398e-01 -8.85885954e-01 5.06046712e-01 -3.73261720e-01
-1.25674355e+00 5.77619314e-01 -4.16727901e-01 -1.43755972e+00
-3.27867508e-01 -8.41608882e-01 -5.10359049e-01 9.26664889e-01
-8.53293061e-01 -7.90238559e-01 4.23965603e-02 4.11100119e-01
-1.88029170e-01 -5.19610524e-01 1.03171062e+00 3.09706360e-01
-1.05883622e+00 2.49201700e-01 2.90052444e-01 -5.27186319e-02
5.43034196e-01 -9.30477381e-01 -7.21505582e-02 3.59018177e-01
-2.21249104e-01 7.93259442e-01 5.10349512e-01 -8.07185292e-01
-1.69091570e+00 -1.10994339e+00 8.03514957e-01 3.69045995e-02
7.94888198e-01 -2.37522244e-01 -7.84024537e-01 4.52462524e-01
-3.30083668e-01 1.18002944e-01 1.17207098e+00 1.38395667e-01
-3.97172123e-02 1.06337346e-01 -1.30842710e+00 8.04698050e-01
1.08757949e+00 -1.09038875e-01 -2.34740764e-01 3.44834894e-01
5.01883984e-01 -9.58448723e-02 -1.55476701e+00 4.33341771e-01
6.76330447e-01 -1.08702946e+00 1.02257597e+00 -7.49882042e-01
4.35042083e-01 -5.90016305e-01 -1.70408502e-01 -1.51590323e+00
-5.11714756e-01 -2.83510149e-01 1.65472291e-02 8.37310433e-01
7.17759371e-01 -7.56481469e-01 7.84233510e-01 2.64497310e-01
-1.09683260e-01 -1.13999224e+00 -1.17846835e+00 -5.18003345e-01
-2.53437877e-01 -3.31589788e-01 6.64757848e-01 1.08439183e+00
2.24236548e-01 2.13779375e-01 -1.43603504e-01 1.65481433e-01
4.11963165e-01 -2.26927456e-02 4.32306498e-01 -1.20245063e+00
-4.01087403e-01 -6.20457530e-01 -9.44572031e-01 -2.98124999e-01
-4.95121852e-02 -1.32251930e+00 -5.98608553e-01 -1.30856586e+00
1.54087946e-01 -6.01659715e-01 -2.86227226e-01 3.68861228e-01
-1.03930049e-01 3.54694426e-01 -3.81587178e-01 5.93127720e-02
8.95388722e-02 3.67103904e-01 1.24369621e+00 -2.02425316e-01
-2.72362202e-01 7.49990493e-02 -6.51368856e-01 5.11461377e-01
6.40517056e-01 -1.22631617e-01 -1.75000638e-01 8.17586258e-02
3.34082633e-01 3.19357455e-01 2.30634049e-01 -7.91145265e-01
2.74114963e-02 -2.20402807e-01 7.45999396e-01 -3.90369952e-01
5.66922128e-01 -5.83677053e-01 8.35063040e-01 8.98236752e-01
-1.08613268e-01 -1.68813303e-01 -2.68204734e-02 3.42182100e-01
-3.53854597e-01 1.59639306e-02 4.40757781e-01 -1.77524552e-01
-2.80259699e-01 3.50024939e-01 -5.17331302e-01 -4.28409278e-01
1.13604569e+00 -2.15873688e-01 -5.68728931e-02 2.81554312e-02
-1.27135563e+00 6.27049208e-02 3.22646856e-01 1.08708240e-01
7.95893669e-01 -1.45843852e+00 -9.77872908e-01 4.29036647e-01
1.49420425e-01 -1.28682509e-01 6.58474207e-01 1.17894208e+00
-4.20153677e-01 3.05077285e-01 -3.16530049e-01 -9.26180184e-01
-1.44489527e+00 5.88944137e-01 5.15741050e-01 -2.26003110e-01
-2.23208934e-01 9.02157605e-01 1.42749846e-01 -4.83225942e-01
-2.52377242e-01 -1.41529888e-01 -6.14948452e-01 5.13705254e-01
4.83711004e-01 4.68277723e-01 1.02962293e-01 -7.22746313e-01
-3.38416040e-01 8.19278955e-01 3.81566025e-02 3.68841514e-02
1.46889484e+00 2.14965701e-01 -6.12046301e-01 4.73187417e-01
1.35498178e+00 -6.78213313e-02 -3.49793196e-01 -1.04594298e-01
-1.68378189e-01 -2.09302828e-01 -2.42615119e-01 -6.72760129e-01
-1.18475568e+00 3.46842021e-01 4.29713190e-01 1.23793378e-01
1.16990232e+00 5.50719649e-02 -1.53337298e-02 1.00237034e-01
2.31708333e-01 -5.09657323e-01 -2.61407822e-01 2.56432772e-01
9.18133855e-01 -7.17534065e-01 1.96011662e-02 -6.94064975e-01
-4.39001888e-01 1.21700668e+00 2.59963423e-01 -4.43587452e-02
7.02849984e-01 4.51635234e-02 -2.10579447e-02 -4.20135081e-01
-1.13430381e+00 1.03423409e-01 5.13528824e-01 5.80229759e-01
5.55088401e-01 3.29137444e-01 -5.20920575e-01 7.08115041e-01
-1.99396953e-01 -5.09559736e-02 5.04650354e-01 7.99865723e-01
-9.58741605e-02 -1.26108027e+00 -5.66166580e-01 9.96500134e-01
-2.75578558e-01 5.96483164e-02 -3.89852375e-01 7.78081536e-01
3.13275009e-01 6.82455540e-01 2.07508504e-01 -8.14992130e-01
3.74520302e-01 2.82984190e-02 3.23513478e-01 -5.56469738e-01
-5.08836985e-01 3.79570961e-01 3.83116096e-01 -4.96203750e-01
-3.02035093e-01 -8.29003036e-01 -1.07633412e+00 -2.67555624e-01
-5.74641936e-02 2.34721780e-01 6.79663241e-01 7.37314641e-01
6.51620507e-01 8.14991891e-01 6.16545141e-01 -6.04829729e-01
-1.89560607e-01 -8.38626146e-01 -1.13189113e+00 1.86265126e-01
1.80963293e-01 -8.46175969e-01 -6.13475621e-01 -1.72855824e-01] | [6.406226634979248, 5.468254089355469] |
84c61512-c4ee-4ba6-9713-ce4af4e22601 | learning-causal-effects-via-weighted | null | null | http://proceedings.neurips.cc/paper/2020/hash/95a6fc111fa11c3ab209a0ed1b9abeb6-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/95a6fc111fa11c3ab209a0ed1b9abeb6-Paper.pdf | Learning Causal Effects via Weighted Empirical Risk Minimization | Learning causal effects from data is a fundamental problem across the sciences. Determining the identifiability of a target effect from a combination of the observational distribution and the causal graph underlying a phenomenon is well-understood in theory. However, in practice, it remains a challenge to apply the identification theory to estimate the identified causal functionals from finite samples. Although a plethora of effective estimators have been developed under the setting known as the back-door (also called conditional ignorability), there exists still no systematic way of estimating arbitrary causal functionals that are both computationally and statistically attractive. This paper aims to bridge this gap, from causal identification to causal estimation. We note that estimating functionals from limited samples based on the empirical risk minimization (ERM) principle has been pervasive in the machine learning literature, and these methods have been extended to causal inference under the back-door setting. In this paper, we develop a learning framework that marries two families of methods, benefiting from the generality of the causal identification theory and the effectiveness of the estimators produced based on the principle of ERM. Specifically, we develop a sound and complete algorithm that generates causal functionals in the form of weighted distributions that are amenable to the ERM optimization. We then provide a practical procedure for learning causal effects from finite samples and a causal graph. Finally, experimental results support the effectiveness of our approach. | ['Elias Bareinboim', 'Jin Tian', 'Yonghan Jung'] | 2020-12-01 | null | null | null | neurips-2020-12 | ['causal-identification'] | ['reasoning'] | [ 6.13223195e-01 1.48205563e-01 -6.08931303e-01 -4.29945469e-01
-7.64382005e-01 -4.14863735e-01 5.72402000e-01 4.76081185e-02
-1.77276284e-02 1.09289300e+00 3.01719248e-01 -3.48998129e-01
-1.01048338e+00 -8.58507097e-01 -1.07293069e+00 -7.64098585e-01
-3.02138776e-01 1.17468707e-01 -2.62544066e-01 3.22351009e-01
2.51595497e-01 1.58234805e-01 -1.41526759e+00 -2.08572298e-01
1.20583987e+00 5.67842603e-01 -6.75909687e-03 3.11310112e-01
2.78247714e-01 6.27120554e-01 -9.05208811e-02 -2.92565852e-01
-8.15142691e-03 -6.85863674e-01 -8.44907463e-01 -1.56277478e-01
3.29108387e-01 -1.96717903e-01 -1.48037627e-01 1.03722131e+00
2.92852849e-01 -1.67404518e-01 8.86444926e-01 -1.33111620e+00
-7.89618492e-01 9.53824162e-01 -6.57642484e-01 -1.72452003e-01
3.62222373e-01 -2.40981519e-01 1.40307236e+00 -6.44677877e-01
3.25508416e-01 1.55932450e+00 6.21288061e-01 2.08374843e-01
-1.39549959e+00 -7.09105313e-01 9.52883288e-02 6.82137161e-02
-1.09195745e+00 -2.87991047e-01 6.99341953e-01 -7.65981138e-01
8.22947919e-02 2.42715567e-01 4.22609806e-01 1.26082456e+00
2.13543966e-01 6.97517931e-01 1.37059975e+00 -6.90746903e-01
3.51859659e-01 -4.87225167e-02 3.03547382e-01 6.02825522e-01
5.76810360e-01 6.13308251e-01 -6.05060995e-01 -3.90710324e-01
8.39384317e-01 -1.38774559e-01 -4.27378148e-01 -6.07372880e-01
-1.11940932e+00 1.28495252e+00 1.97206914e-01 1.27496570e-01
-4.37255204e-01 2.82852560e-01 2.40373090e-01 1.27048090e-01
5.84756374e-01 3.97670954e-01 -3.23124439e-01 3.37344140e-01
-6.85848951e-01 2.68617898e-01 9.32044625e-01 6.80120289e-01
4.76020545e-01 -1.83479875e-01 -9.92828906e-02 5.88021755e-01
5.01857162e-01 6.15455091e-01 -1.67982817e-01 -8.35352659e-01
-4.48139235e-02 4.39016998e-01 2.00091794e-01 -1.04916096e+00
-1.75650939e-01 -1.58480436e-01 -9.59355891e-01 -4.88904491e-02
7.77285874e-01 -3.48587841e-01 -3.52643669e-01 2.29040360e+00
5.83888590e-01 5.13982773e-01 -2.85088837e-01 8.79552007e-01
3.53464574e-01 3.99470419e-01 2.50484437e-01 -6.74442768e-01
1.05071294e+00 -2.78096408e-01 -8.51319909e-01 5.80290481e-02
4.28516954e-01 -4.59813923e-01 9.76893783e-01 3.00407887e-01
-7.92780876e-01 -3.05849046e-01 -8.51620913e-01 2.21958205e-01
-1.17979478e-02 -9.63620469e-02 1.02285981e+00 8.09536636e-01
-7.63629198e-01 7.03519583e-01 -5.33489645e-01 -4.20472890e-01
3.83773029e-01 4.13119197e-01 -2.04818442e-01 -7.09392726e-02
-1.47715271e+00 6.67501450e-01 3.35898936e-01 6.08034097e-02
-1.02021563e+00 -1.04719782e+00 -6.62547290e-01 5.49458042e-02
8.54610860e-01 -9.94797826e-01 1.05472970e+00 -1.15053415e+00
-1.29436028e+00 3.89757007e-01 -9.97215211e-02 -2.61997908e-01
5.59695899e-01 -2.87416756e-01 -9.85714644e-02 -7.13650733e-02
1.57570004e-01 -5.40671870e-02 9.28099036e-01 -1.22126520e+00
-6.04963779e-01 -4.94982421e-01 1.96367368e-01 -1.64838955e-01
-3.61958981e-01 1.83217581e-02 1.62919998e-01 -4.86403346e-01
-2.09341034e-01 -7.68552482e-01 -2.36430332e-01 -2.31155172e-01
-6.99008286e-01 -4.07825023e-01 4.35702860e-01 -4.19440120e-01
1.38265657e+00 -1.98612976e+00 2.55610287e-01 2.24197030e-01
1.37063459e-01 -3.41222614e-01 3.18180844e-02 5.25308669e-01
-5.15986204e-01 2.85423279e-01 -6.84315562e-01 2.14344859e-01
4.26503830e-02 -1.84656847e-02 -5.62590063e-01 8.76334250e-01
2.33207300e-01 7.39361525e-01 -1.19748747e+00 -3.99852544e-01
1.62766173e-01 3.24092835e-01 -6.99194968e-01 3.29019487e-01
-9.28888395e-02 5.27712822e-01 -8.02671134e-01 3.60173434e-01
5.51589668e-01 -3.34785134e-01 5.35272837e-01 -6.10220246e-02
-3.70643705e-01 2.25475758e-01 -1.46954203e+00 1.28842127e+00
-2.62833267e-01 1.08319446e-01 -3.78592461e-02 -1.71217251e+00
5.23005009e-01 4.45293099e-01 7.26636767e-01 -4.35763039e-02
9.15328935e-02 2.52446920e-01 -6.61014244e-02 -6.41236603e-01
-3.60477343e-02 -5.78245342e-01 -2.20922381e-01 4.11071718e-01
3.86286937e-02 2.47956514e-01 1.13384306e-01 -1.92808453e-02
1.07665396e+00 2.57074416e-01 8.58717740e-01 -5.88713348e-01
3.42747062e-01 -3.93458642e-02 5.29578149e-01 9.91746187e-01
1.95168972e-01 1.63428143e-01 6.82370007e-01 -8.00871029e-02
-8.42515826e-01 -1.22756720e+00 -5.83835542e-01 7.50298381e-01
-1.42412767e-01 -2.29890775e-02 -7.34130323e-01 -7.30419457e-01
3.03708911e-01 6.79593205e-01 -8.43162119e-01 -2.26539776e-01
-3.05438697e-01 -9.96890962e-01 3.27594578e-01 3.11428308e-01
7.85334110e-02 -7.71226227e-01 -2.96960652e-01 7.17905462e-02
-1.68210268e-01 -5.21691561e-01 -3.53008956e-01 9.94073227e-02
-9.15390372e-01 -1.41895306e+00 -3.60115707e-01 -2.53439248e-01
5.07525325e-01 4.62279245e-02 1.10099280e+00 -2.91327000e-01
-8.42244774e-02 5.38807809e-01 -2.02733666e-01 -5.51739633e-01
-5.28430045e-01 -1.61471233e-01 2.99742937e-01 4.36109811e-01
1.58108816e-01 -7.10719049e-01 -3.65877271e-01 1.63223580e-01
-9.04596806e-01 -1.49195135e-01 7.30412841e-01 1.02132308e+00
4.65489089e-01 2.52274245e-01 1.07348001e+00 -1.20555913e+00
4.46123987e-01 -9.10778940e-01 -8.92697453e-01 3.93203855e-01
-7.82751799e-01 2.79616416e-01 5.65081179e-01 -4.08227623e-01
-1.28717208e+00 6.08595461e-02 2.21183017e-01 -1.29720047e-01
-1.33916438e-01 9.00304019e-01 -5.38635433e-01 2.44267628e-01
4.95244205e-01 -2.82417148e-01 -1.44761950e-01 -4.01126176e-01
6.58100724e-01 5.84884465e-01 3.20511878e-01 -8.12368989e-01
7.24371433e-01 5.92541754e-01 3.02963287e-01 -7.29782701e-01
-1.22745788e+00 -3.46754402e-01 -5.99211216e-01 -2.27640092e-01
9.63404775e-01 -5.85726500e-01 -1.01792681e+00 3.48275490e-02
-1.03741682e+00 -1.62371799e-01 -3.29329252e-01 7.67010510e-01
-8.16008449e-01 2.57082582e-01 -1.05925076e-01 -1.21512341e+00
7.21116364e-02 -8.97561550e-01 9.17767942e-01 1.00955199e-02
-3.46544534e-01 -1.42576361e+00 3.58163029e-01 9.08789486e-02
-3.09348077e-01 3.87086958e-01 1.17131162e+00 -2.68918335e-01
-5.91418624e-01 -1.00822732e-01 -2.49019295e-01 7.17574954e-02
4.18673962e-01 3.82401049e-02 -9.97997046e-01 -9.94013995e-02
1.36677653e-01 -1.53086528e-01 8.81458521e-01 1.09786296e+00
1.07526815e+00 -4.28684533e-01 -5.69940746e-01 3.33395571e-01
1.61192584e+00 5.34092635e-02 3.65755379e-01 -1.99503288e-01
7.11673498e-01 1.07735765e+00 4.76638973e-01 4.76375580e-01
2.25173965e-01 5.92111945e-01 4.24346715e-01 7.71014541e-02
2.23543465e-01 -5.67828119e-01 2.52897710e-01 6.02405787e-01
-1.50964662e-01 -1.98748514e-01 -6.61864817e-01 5.14230609e-01
-2.03773642e+00 -1.23162210e+00 -5.98904967e-01 2.70056152e+00
1.16153467e+00 -3.21459383e-01 2.37813950e-01 3.36080678e-02
9.89855111e-01 -3.00450206e-01 -5.62082708e-01 -8.44983011e-03
-9.68194976e-02 9.37642753e-02 6.09065235e-01 5.96924603e-01
-1.15031028e+00 3.73818815e-01 6.96480560e+00 7.73801148e-01
-5.23875415e-01 1.34326413e-01 6.03479385e-01 3.67046982e-01
-4.87057537e-01 4.35202032e-01 -7.41541207e-01 4.03550565e-01
1.04873645e+00 -3.79011810e-01 2.63592511e-01 6.43587708e-01
7.52533793e-01 -1.84522182e-01 -1.53873098e+00 6.10809147e-01
-3.97204846e-01 -1.03407145e+00 -8.11427534e-02 4.59517181e-01
9.59224761e-01 -6.73957467e-01 -5.52707836e-02 -3.60088125e-02
6.82932556e-01 -1.12458563e+00 5.84237278e-01 7.92808473e-01
8.79479706e-01 -6.41260624e-01 4.67729986e-01 4.14865136e-01
-7.81020164e-01 -3.23791891e-01 -3.45168233e-01 -2.95714021e-01
1.42095849e-01 1.33768868e+00 -5.34662306e-01 8.77814651e-01
4.17758405e-01 9.72801328e-01 -1.23432027e-02 9.86632645e-01
-3.48289907e-01 1.03327417e+00 -6.40352145e-02 4.79446873e-02
-2.96231747e-01 -5.53037643e-01 4.85640287e-01 9.32137311e-01
4.95049685e-01 6.30237162e-02 -7.25086853e-02 1.14602268e+00
-1.23772450e-01 9.08283144e-02 -8.35960567e-01 -1.73895061e-01
4.75776255e-01 9.22628641e-01 -3.21946204e-01 -8.04153457e-02
-5.69960952e-01 3.30733508e-01 2.78947741e-01 2.91187942e-01
-8.88254404e-01 2.47214139e-01 4.63515848e-01 1.62115112e-01
-1.45189300e-01 2.98222810e-01 -5.20159304e-01 -9.74511802e-01
-1.90075368e-01 -8.97001326e-01 6.42034650e-01 -3.09633315e-01
-1.81243014e+00 -4.67795581e-01 4.77463216e-01 -9.46239829e-01
-2.19282851e-01 -5.52837610e-01 -4.18025821e-01 8.94700587e-01
-1.25012088e+00 -8.84534538e-01 7.52203688e-02 3.71587485e-01
2.78580397e-01 4.04256046e-01 7.60895669e-01 1.96062848e-01
-6.77755535e-01 2.16378704e-01 2.03546748e-01 -1.94521531e-01
8.33039582e-01 -1.55175519e+00 -3.12391311e-01 7.53608525e-01
-8.19721892e-02 8.67850661e-01 7.80579090e-01 -8.91958594e-01
-1.61578524e+00 -9.37878311e-01 9.41484809e-01 -5.36928236e-01
1.09484541e+00 -2.90918112e-01 -7.87399650e-01 6.98677659e-01
6.65657595e-02 -3.74531388e-01 5.73974729e-01 6.80984318e-01
-2.15911031e-01 -1.69227511e-01 -7.89349616e-01 6.44455552e-01
1.14434040e+00 -2.66573757e-01 -6.19661987e-01 3.92851979e-01
4.90747213e-01 1.98111132e-01 -9.27616119e-01 5.22035301e-01
6.16460919e-01 -7.60166109e-01 9.99956191e-01 -7.72253454e-01
7.88032949e-01 -1.28687501e-01 -4.25814604e-03 -1.40097582e+00
-3.06338847e-01 -6.26203120e-01 1.10704638e-02 1.39584887e+00
2.70296335e-01 -6.31050646e-01 2.58767247e-01 4.19332951e-01
1.01452187e-01 -4.13767248e-01 -8.69942725e-01 -7.06608891e-01
3.80755901e-01 -5.42654574e-01 2.42805913e-01 1.28043342e+00
1.14775687e-01 6.24755740e-01 -6.43942595e-01 3.69930774e-01
1.13520634e+00 2.36757353e-01 6.12654924e-01 -1.69003880e+00
-5.74564517e-01 -3.83709460e-01 1.08701147e-01 -7.59928524e-01
4.25303280e-01 -8.31833720e-01 1.83318064e-01 -1.32003701e+00
7.83491075e-01 -4.30694044e-01 -1.32573396e-01 2.43439689e-01
-5.51017880e-01 -2.92013735e-01 -2.75689363e-01 1.43844321e-01
-5.54282404e-02 4.72104609e-01 1.13555610e+00 -2.40583923e-02
-4.11150269e-02 1.89253956e-01 -8.39024544e-01 8.90648484e-01
5.66185355e-01 -7.29051769e-01 -6.93959415e-01 1.11002676e-01
5.47534406e-01 4.14329261e-01 8.33511770e-01 -2.75666296e-01
-8.80257487e-02 -4.76616383e-01 -1.84207559e-02 -1.50202245e-01
-3.25682968e-01 -6.72722936e-01 3.15950871e-01 2.91171640e-01
-5.75082242e-01 -4.04339015e-01 -1.80353135e-01 9.33487892e-01
-1.36053413e-02 -3.53335172e-01 7.28588223e-01 7.95955211e-02
-3.30470026e-01 1.23630948e-01 -4.46676910e-02 2.85717368e-01
7.78438389e-01 1.26260996e-01 -1.44988433e-01 -3.46240640e-01
-4.75288630e-01 1.00787051e-01 1.33708334e-02 1.25054196e-01
3.73784214e-01 -1.30708086e+00 -9.48729992e-01 -1.83937222e-01
-6.64658844e-02 -5.01594663e-01 3.08288693e-01 1.19228125e+00
3.78521651e-01 5.51854193e-01 2.71693975e-01 -5.26206791e-01
-7.77979374e-01 8.02132905e-01 1.36725813e-01 -2.50346601e-01
-3.35815609e-01 5.12891650e-01 9.32572246e-01 -3.42167139e-01
-8.19184184e-02 -3.07730250e-02 -1.12256944e-01 8.70777369e-02
4.17197019e-01 6.38086677e-01 -4.10515100e-01 -2.35122427e-01
-2.39446700e-01 3.90147120e-01 5.07785022e-01 -8.12955275e-02
1.33212626e+00 -1.49095908e-01 -4.25933659e-01 7.60691583e-01
1.00632405e+00 1.57717779e-01 -1.35039020e+00 -8.06548223e-02
2.92823553e-01 -3.26593965e-01 4.45122980e-02 -7.16942668e-01
-6.80936396e-01 6.03995442e-01 3.36720318e-01 5.66718459e-01
1.09293067e+00 1.16867185e-01 -5.30287772e-02 -3.44598405e-02
2.80928642e-01 -7.54648864e-01 -1.76896811e-01 3.24669331e-02
8.75734508e-01 -1.27016091e+00 -1.14475749e-02 -7.62671471e-01
-2.36787759e-02 9.13530707e-01 -2.60822512e-02 -2.16235384e-01
7.89408624e-01 3.05437624e-01 -4.90572125e-01 -2.00526148e-01
-5.88144779e-01 -2.03991801e-01 4.52493489e-01 5.80966234e-01
7.13558853e-01 3.25758845e-01 -7.77931333e-01 6.50990188e-01
1.16123697e-02 2.48794153e-01 5.43448389e-01 3.13142926e-01
-2.70783782e-01 -1.12890100e+00 -5.36645651e-01 4.52371150e-01
-5.00014663e-01 -1.34843424e-01 -3.73107910e-01 1.07058907e+00
7.37282857e-02 1.24191737e+00 -3.19654316e-01 5.40322363e-02
3.12356621e-01 1.10626139e-01 4.74589318e-01 -5.86981535e-01
8.19162950e-02 1.72006935e-01 5.60963824e-02 -4.61159050e-01
-9.74226952e-01 -1.09420538e+00 -8.38967264e-01 -3.25180799e-01
-5.02763867e-01 1.69473693e-01 2.37704366e-01 1.23286951e+00
-1.17042065e-01 5.29329419e-01 7.46577919e-01 -3.29515547e-01
-8.84321988e-01 -8.83492410e-01 -8.82246077e-01 1.94731876e-01
2.51974970e-01 -9.60092545e-01 -5.95145226e-01 2.90333867e-01] | [7.833114147186279, 5.296751499176025] |
5f489f91-7460-41af-8886-3dffd31e108e | real-time-cardiovascular-mr-with-spatio | 1803.05192 | null | http://arxiv.org/abs/1803.05192v3 | http://arxiv.org/pdf/1803.05192v3.pdf | Real-time Cardiovascular MR with Spatio-temporal Artifact Suppression using Deep Learning - Proof of Concept in Congenital Heart Disease | PURPOSE: Real-time assessment of ventricular volumes requires high
acceleration factors. Residual convolutional neural networks (CNN) have shown
potential for removing artifacts caused by data undersampling. In this study we
investigated the effect of different radial sampling patterns on the accuracy
of a CNN. We also acquired actual real-time undersampled radial data in
patients with congenital heart disease (CHD), and compare CNN reconstruction to
Compressed Sensing (CS).
METHODS: A 3D (2D plus time) CNN architecture was developed, and trained
using 2276 gold-standard paired 3D data sets, with 14x radial undersampling.
Four sampling schemes were tested, using 169 previously unseen 3D 'synthetic'
test data sets. Actual real-time tiny Golden Angle (tGA) radial SSFP data was
acquired in 10 new patients (122 3D data sets), and reconstructed using the 3D
CNN as well as a CS algorithm; GRASP.
RESULTS: Sampling pattern was shown to be important for image quality, and
accurate visualisation of cardiac structures. For actual real-time data,
overall reconstruction time with CNN (including creation of aliased images) was
shown to be more than 5x faster than GRASP. Additionally, CNN image quality and
accuracy of biventricular volumes was observed to be superior to GRASP for the
same raw data.
CONCLUSION: This paper has demonstrated the potential for the use of a 3D CNN
for deep de-aliasing of real-time radial data, within the clinical setting.
Clinical measures of ventricular volumes using real-time data with CNN
reconstruction are not statistically significantly different from the
gold-standard, cardiac gated, BH techniques. | ['Vivek Muthurangu', 'Simon Arridge', 'Felix Lucka', 'Andreas Hauptmann', 'Jennifer A. Steeden'] | 2018-03-14 | null | null | null | null | ['de-aliasing'] | ['computer-vision'] | [ 1.61714792e-01 2.11544454e-01 4.40587312e-01 -2.55412340e-01
-6.95988953e-01 -6.09858930e-01 -2.33084057e-02 -2.27536280e-02
-4.20381397e-01 6.45266294e-01 2.35671014e-01 -7.08645165e-01
-1.63296461e-01 -5.53932071e-01 -6.10962212e-01 -3.80714893e-01
-8.32303584e-01 8.01604271e-01 -8.10376778e-02 8.06903616e-02
-1.66401103e-01 8.90675426e-01 -8.26699138e-01 3.34667355e-01
3.08011532e-01 1.06966364e+00 1.74604550e-01 1.12066817e+00
4.36575443e-01 6.74142122e-01 -6.67594492e-01 3.67503941e-01
6.67866290e-01 -5.80237389e-01 -6.92681372e-01 -1.39271125e-01
7.83707976e-01 -1.07781589e+00 -9.19559896e-02 1.19165957e-01
1.46410608e+00 -4.68251854e-01 1.03471637e-01 -3.98440987e-01
-2.23963037e-01 5.63468099e-01 -2.33073607e-02 7.11068511e-01
1.18830418e-02 5.37594795e-01 1.96409598e-01 -1.03324461e+00
6.95908248e-01 4.52741116e-01 1.69326746e+00 4.63746756e-01
-1.50061512e+00 -5.65557361e-01 -1.27284098e+00 -6.18671954e-01
-9.13752496e-01 -2.75732845e-01 6.73232019e-01 -6.49533629e-01
1.05859542e+00 4.26425576e-01 1.67883790e+00 3.94923002e-01
2.82937199e-01 9.13116857e-02 1.02678323e+00 -4.53964502e-01
7.60729015e-02 -3.16139817e-01 -3.86091501e-01 4.80793923e-01
5.57169735e-01 6.14294410e-01 4.26242240e-02 -3.24786961e-01
1.47198176e+00 7.79871568e-02 -8.86661232e-01 -5.46559095e-01
-1.60848475e+00 6.07147694e-01 2.12748125e-01 6.07002795e-01
-4.77632672e-01 3.58429193e-01 9.50120151e-01 3.22261542e-01
4.33676660e-01 8.06149721e-01 -7.31188893e-01 -3.96940172e-01
-1.32216954e+00 2.61193156e-01 4.99408066e-01 7.27206528e-01
-6.69989362e-02 5.45334160e-01 -2.94668138e-01 7.24021077e-01
2.26430967e-01 6.51090503e-01 7.54099667e-01 -1.20253313e+00
3.91587287e-01 8.12097639e-02 -1.64711997e-01 -1.00744236e+00
-8.88220251e-01 -8.29642892e-01 -1.23939145e+00 4.89858478e-01
6.14944279e-01 -4.32387739e-01 -9.31956112e-01 1.24550354e+00
1.51391447e-01 5.22241145e-02 9.53548178e-02 1.23562014e+00
1.16094947e+00 1.45779103e-02 -3.44399184e-01 -4.07159448e-01
9.67791677e-01 -1.96843982e-01 -3.76037121e-01 2.57268637e-01
1.00548327e+00 -5.55449307e-01 7.80300915e-01 3.38539809e-01
-1.77594137e+00 -5.29937029e-01 -1.06868613e+00 2.24779993e-01
4.01825517e-01 2.48044595e-01 5.80929577e-01 8.54277432e-01
-1.48667502e+00 1.33271384e+00 -9.79849815e-01 -1.36685027e-02
1.00366342e+00 3.36646378e-01 -4.09973651e-01 -1.35280266e-01
-9.99934673e-01 9.38037992e-01 1.40462294e-01 3.37094933e-01
-8.12786579e-01 -1.58464050e+00 -6.91264153e-01 -2.53829490e-02
-3.61471951e-01 -7.40595579e-01 9.60502088e-01 -5.30479610e-01
-1.10991466e+00 1.28306901e+00 7.50597775e-01 -6.74030423e-01
9.07296360e-01 -5.92105091e-02 -3.51160616e-02 4.43694592e-01
-4.19422910e-02 2.35036775e-01 4.72803056e-01 -8.63591850e-01
2.65386939e-01 -5.35007596e-01 -4.99654531e-01 -2.24450886e-01
4.16593045e-01 -8.41869488e-02 1.90728799e-01 -8.01171064e-01
6.40210330e-01 -8.16252947e-01 -1.82564437e-01 4.28721458e-01
-7.91726187e-02 8.40790451e-01 5.49797058e-01 -1.15890896e+00
7.52660155e-01 -1.91415143e+00 -4.15917009e-01 2.38220587e-01
1.02200496e+00 6.78410292e-01 2.94968814e-01 -1.02343678e-01
-6.74611509e-01 3.06624383e-01 -5.52131951e-01 -7.69463405e-02
-8.15951586e-01 1.12468824e-01 2.43386611e-01 8.09881747e-01
-2.15694979e-01 1.14314127e+00 -7.45186567e-01 -5.71575463e-01
5.56492686e-01 7.07964242e-01 -6.72785223e-01 2.15305626e-01
5.78857124e-01 7.66804218e-01 1.13126174e-01 3.00468713e-01
9.06940937e-01 -2.98243254e-01 4.05424178e-01 -4.21804518e-01
-1.50562942e-01 6.89180791e-02 -7.81358600e-01 1.86340702e+00
-3.83795798e-01 8.80052269e-01 1.29438818e-01 -9.07682657e-01
9.93358314e-01 8.95368457e-01 9.42500412e-01 -8.13522339e-01
2.32693121e-01 7.45247304e-01 2.97743529e-01 -5.56449592e-01
-2.64228284e-01 -6.63429439e-01 7.22724378e-01 3.53021353e-01
1.09585971e-02 -7.12023973e-01 -4.51530278e-01 -7.60140121e-02
1.24430120e+00 2.50103008e-02 7.84435403e-03 -6.55997753e-01
-5.03452159e-02 1.02674313e-01 4.65601027e-01 7.89679348e-01
-5.11096954e-01 1.66728544e+00 5.20722628e-01 -1.15597737e+00
-1.76879847e+00 -8.84913743e-01 -4.96115804e-01 -2.36874118e-01
-5.79758704e-01 -1.21067233e-01 -4.05862212e-01 -1.32745758e-01
1.02562979e-02 1.42207950e-01 -6.67978406e-01 1.21293299e-01
-1.08222556e+00 -5.95028996e-01 9.91527259e-01 8.55643570e-01
2.44673222e-01 -1.17661560e+00 -1.65271664e+00 7.08630979e-01
1.86662993e-03 -7.97085285e-01 -1.13861598e-01 2.63671368e-01
-1.65330374e+00 -1.15957355e+00 -1.18561018e+00 -5.21371543e-01
3.65998536e-01 -3.57959479e-01 1.54132020e+00 5.56928098e-01
-9.64687228e-01 1.51972920e-01 -2.72531033e-01 -4.14093763e-01
-5.54704130e-01 -4.52843398e-01 -1.04290463e-01 -8.58471036e-01
-5.89341819e-01 -8.58108819e-01 -1.10015404e+00 -1.40435353e-01
-6.93279624e-01 1.94308028e-01 2.80219674e-01 9.99952078e-01
5.76328695e-01 -6.39356315e-01 4.15435106e-01 -7.22953498e-01
5.63530922e-01 -5.49102277e-02 -4.67683971e-01 -3.19939315e-01
-7.77060926e-01 -4.07402545e-01 5.81096530e-01 -1.11581758e-01
-2.74389774e-01 -2.07421049e-01 -2.88881987e-01 -1.06691813e+00
7.00547025e-02 3.42458278e-01 7.00768113e-01 -3.18305790e-02
1.03444242e+00 1.07151546e-01 9.09456789e-01 -1.70512319e-01
-3.06943595e-01 9.68043730e-02 5.02048612e-01 -4.40421700e-01
2.50252038e-01 5.76909959e-01 4.21657294e-01 -4.79578197e-01
-1.31133333e-01 -5.43127060e-02 -8.67146015e-01 -4.41758215e-01
8.73917758e-01 -6.96111560e-01 -4.27168816e-01 5.63923299e-01
-1.06050301e+00 -7.80131996e-01 -1.16365612e+00 8.50486040e-01
-7.89443016e-01 2.36801520e-01 -7.55856693e-01 -3.48462552e-01
-1.02068770e+00 -1.17826295e+00 7.50256300e-01 -3.96347851e-01
-2.81163394e-01 -9.84439552e-01 1.95944071e-01 1.31469652e-01
1.14137602e+00 1.01026046e+00 8.07261884e-01 -5.89525342e-01
-9.44199339e-02 -2.26146653e-01 -2.05805063e-01 6.78336799e-01
-4.67064269e-02 -2.50204891e-01 -8.39446485e-01 -5.73496759e-01
2.28046820e-01 -3.18597138e-01 3.49254340e-01 1.06437218e+00
1.14817441e+00 1.18389390e-02 1.26995131e-01 8.69749844e-01
1.57132697e+00 3.21675837e-01 9.12240028e-01 -1.18349366e-01
6.58094347e-01 1.03052363e-01 2.54824340e-01 7.26088643e-01
-3.71083856e-01 2.63551474e-01 5.13738871e-01 -6.45170748e-01
-4.51105446e-01 1.71095550e-01 -4.91972119e-01 8.77346158e-01
-5.37006736e-01 3.76407802e-01 -1.33889139e+00 5.87196827e-01
-1.05221879e+00 -8.23293090e-01 -4.53806251e-01 2.31261230e+00
5.68965971e-01 5.14484383e-02 -9.17389616e-03 4.06011581e-01
5.51948965e-01 -1.31738037e-01 -4.08093065e-01 -3.78555655e-01
-1.00770645e-01 7.89461434e-01 4.91168082e-01 1.35190383e-01
-6.39864504e-01 -5.56191355e-02 6.53763056e+00 -9.45474729e-02
-1.50513208e+00 3.07660520e-01 9.61657465e-01 -2.61999607e-01
-1.23107418e-01 -2.04622731e-01 1.08920902e-01 2.96117425e-01
9.07831907e-01 2.63300300e-01 1.86445147e-01 6.00373507e-01
4.70547885e-01 1.80569991e-01 -1.00182939e+00 1.25444019e+00
-9.98423770e-02 -2.07445860e+00 -5.72529972e-01 9.69307050e-02
3.47263336e-01 4.99245614e-01 -4.45298165e-01 5.42695448e-02
-4.61656243e-01 -1.28459227e+00 5.35397351e-01 4.88525867e-01
1.83907890e+00 -3.81657332e-01 1.01448417e+00 1.70297161e-01
-8.54074061e-01 2.91599631e-01 -9.29010585e-02 8.83478820e-02
3.44070457e-02 9.42939460e-01 -1.40202451e+00 2.50940531e-01
8.55460167e-01 6.25861049e-01 -2.83610970e-01 9.56182897e-01
6.30771995e-01 7.68778384e-01 -5.18537641e-01 4.14083987e-01
-2.02446386e-01 1.44217595e-01 6.47288322e-01 9.67827857e-01
3.56756151e-01 4.73992050e-01 -3.49328846e-01 1.15408945e+00
2.83356935e-01 3.26049253e-02 -8.32841575e-01 2.23207235e-01
1.61383286e-01 9.58355486e-01 -7.29461789e-01 -5.49728513e-01
-1.36855900e-01 3.94166052e-01 -2.71798074e-01 -1.02058843e-01
-5.43806672e-01 -1.78508893e-01 -2.75781937e-02 9.93100524e-01
1.26789868e-01 2.03438792e-02 -7.25601673e-01 -8.12012851e-01
1.08243711e-01 -7.62298167e-01 1.64682105e-01 -1.23466659e+00
-8.10643613e-01 6.38108373e-01 -2.27372989e-01 -1.38867295e+00
-2.15339482e-01 -3.89247835e-01 -5.81816137e-01 1.24840271e+00
-9.80824471e-01 -6.07651711e-01 -4.66553837e-01 3.65759701e-01
2.74899334e-01 -5.78334630e-02 1.10720098e+00 5.61382771e-01
4.06542927e-01 1.52463838e-01 -1.43009961e-01 4.51392472e-01
2.65963286e-01 -1.26408184e+00 2.09403664e-01 4.63622957e-01
-5.94568133e-01 5.50083697e-01 4.84182507e-01 -8.61190259e-01
-1.13154149e+00 -1.03615236e+00 4.86010671e-01 -6.32297024e-02
-2.43468940e-01 2.90441543e-01 -7.26514697e-01 5.80099881e-01
-7.82930776e-02 9.88773167e-01 4.48483169e-01 -5.24734855e-01
2.28238598e-01 7.72065595e-02 -1.79802811e+00 1.33337295e-02
8.38714600e-01 -2.32280523e-01 -4.13033843e-01 1.84963152e-01
5.10073960e-01 -9.98184502e-01 -1.49016249e+00 9.59385216e-01
8.23446870e-01 -1.35365868e+00 1.12539995e+00 -4.30851966e-01
8.51215184e-01 2.84662358e-02 6.69443086e-02 -1.10751176e+00
-8.50487128e-02 -4.01971489e-01 1.01847485e-01 3.85159969e-01
1.43481091e-01 -8.85178626e-01 1.03740764e+00 6.90384746e-01
-7.55982816e-01 -1.21533990e+00 -1.20619559e+00 -3.56696308e-01
3.80152732e-01 -6.82775855e-01 3.54787230e-01 1.10214043e+00
-2.93252945e-01 -3.30546319e-01 5.24251014e-02 -1.13587402e-01
5.07096827e-01 -8.40696841e-02 2.04361752e-01 -1.25948858e+00
-3.54366243e-01 -2.34535262e-01 -7.01267898e-01 -2.92837977e-01
-7.14317024e-01 -1.01205993e+00 -2.72188395e-01 -1.50342786e+00
-1.33190930e-01 -1.01047349e+00 1.27454950e-02 1.64803848e-01
3.86953086e-01 4.26963389e-01 2.24654123e-01 2.09037855e-01
3.65416497e-01 -9.02075991e-02 1.92429757e+00 3.41730386e-01
-8.83377939e-02 -2.32343767e-02 -9.83360782e-02 4.25447881e-01
8.70315194e-01 -3.74909669e-01 -3.48948449e-01 -3.27745885e-01
1.64466128e-01 1.10124278e+00 8.93603802e-01 -1.47540784e+00
-2.91364998e-01 6.94936633e-01 8.95760536e-01 -8.61645937e-01
7.65089616e-02 -8.76574516e-01 6.12951040e-01 9.79520202e-01
-1.98750526e-01 4.29514438e-01 4.27958012e-01 -3.00312489e-01
6.04268983e-02 -7.38563910e-02 1.11923707e+00 -7.04218984e-01
2.89449751e-01 4.58653182e-01 -4.96101193e-02 6.63618207e-01
6.78535938e-01 -7.23364830e-01 2.36873910e-01 -1.69812948e-01
-1.31357753e+00 -3.93860281e-01 1.71263829e-01 -3.42763156e-01
1.09789073e+00 -1.16026247e+00 -9.85853970e-01 5.67013860e-01
-4.23854560e-01 7.11032867e-01 6.29675686e-01 1.39718223e+00
-1.67680430e+00 3.26594412e-01 -3.27157319e-01 -1.20505166e+00
-1.06324828e+00 3.30325395e-01 1.10634112e+00 -2.59959280e-01
-1.33643401e+00 8.16057324e-01 -4.59189445e-01 -6.16842568e-01
-2.92033218e-02 -8.76305103e-01 1.66712344e-01 -3.24302077e-01
1.72676623e-01 2.28740603e-01 5.28143823e-01 -2.51141842e-02
-1.96009964e-01 5.79236031e-01 6.10278189e-01 6.12618178e-02
1.49515390e+00 1.39804333e-01 7.38108903e-02 3.39616656e-01
1.26902878e+00 -2.47523054e-01 -9.94215250e-01 1.51277572e-01
-7.02435553e-01 -5.83741784e-01 1.55485317e-01 -9.57310557e-01
-1.75779927e+00 1.10044658e+00 1.30584347e+00 2.78283179e-01
9.96388614e-01 -2.69420743e-01 8.86376143e-01 -4.77479190e-01
3.54561388e-01 -6.42691433e-01 -2.22803116e-01 6.81780837e-03
1.15037560e+00 -1.09403121e+00 2.14344442e-01 -1.42026201e-01
-3.32324713e-01 1.49517930e+00 2.66324103e-01 -3.04587990e-01
7.91817665e-01 5.34624457e-01 2.88603604e-01 -6.51183128e-01
-1.88030422e-01 6.47083819e-01 -2.65879720e-01 6.39834166e-01
6.83151364e-01 1.02199599e-01 -1.42057940e-01 6.65704608e-02
-1.20214999e-01 4.97789890e-01 6.29982352e-01 1.10694218e+00
5.56247570e-02 -3.08123320e-01 -3.62584710e-01 8.94768059e-01
-8.53224993e-01 -4.21465337e-01 5.00485957e-01 8.57594013e-01
4.77258354e-01 1.31841913e-01 9.30934846e-02 1.64556727e-01
4.73045617e-01 -2.46074889e-02 6.90499485e-01 -5.06613135e-01
-1.05826652e+00 -7.49628469e-02 1.38448417e-01 -5.31457007e-01
-2.54268318e-01 -8.44576001e-01 -1.22642815e+00 -2.30340511e-01
-2.25299329e-01 -2.87851900e-01 8.36375177e-01 5.08226633e-01
4.01390404e-01 9.16488945e-01 3.96713912e-01 -1.17050433e+00
-4.84193653e-01 -9.34536099e-01 -7.82191217e-01 3.91235590e-01
5.85046947e-01 -3.22915107e-01 -5.14638662e-01 6.27867132e-03] | [14.086670875549316, -2.4767367839813232] |
e6b769d6-d6f7-4164-83c3-c9747bc32b66 | rapid-learning-of-spatial-representations-for | 2206.02249 | null | https://arxiv.org/abs/2206.02249v3 | https://arxiv.org/pdf/2206.02249v3.pdf | Rapid Learning of Spatial Representations for Goal-Directed Navigation Based on a Novel Model of Hippocampal Place Fields | The discovery of place cells and other spatially modulated neurons in the hippocampal complex of rodents has been crucial to elucidating the neural basis of spatial cognition. More recently, the replay of neural sequences encoding previously experienced trajectories has been observed during consummatory behavior potentially with implications for quick memory consolidation and behavioral planning. Several promising models for robotic navigation and reinforcement learning have been proposed based on these and previous findings. Most of these models, however, use carefully engineered neural networks and are tested in simple environments. In this paper, we develop a self-organized model incorporating place cells and replay, and demonstrate its utility for rapid one-shot learning in non-trivial environments with obstacles. | ['Ali Minai', 'Dieter Vanderelst', 'Adedapo Alabi'] | 2022-06-05 | null | null | null | null | ['one-shot-learning'] | ['methodology'] | [-3.17444763e-04 -2.79536724e-01 9.04436707e-02 6.23583281e-03
9.35386792e-02 -6.95799768e-01 4.65964347e-01 3.89982730e-01
-9.30524409e-01 1.21915698e+00 -7.16279894e-02 -7.47524053e-02
-5.16281962e-01 -7.94701338e-01 -8.29495490e-01 -7.94145942e-01
-1.20292497e+00 1.74930245e-01 5.80870926e-01 -5.14937937e-01
6.23358428e-01 6.47398472e-01 -1.49418938e+00 6.05895482e-02
6.09509945e-01 4.62453246e-01 8.21876943e-01 5.10936558e-01
4.63463157e-01 8.17532420e-01 3.13628325e-03 3.48286420e-01
2.15994567e-01 -4.17087018e-01 -6.65504515e-01 -4.64597702e-01
-2.48814598e-01 -3.50868225e-01 -6.12554193e-01 6.34567916e-01
5.46249270e-01 7.35936761e-01 4.86129194e-01 -8.88520002e-01
-5.39936066e-01 3.21657151e-01 3.02065797e-02 7.76902556e-01
2.52025634e-01 3.32152098e-01 5.67528069e-01 -7.77375102e-01
8.22966993e-01 7.83121049e-01 6.59972310e-01 5.06682396e-01
-1.27205646e+00 -5.26455998e-01 2.01128181e-02 5.13451219e-01
-1.26635313e+00 -5.15011728e-01 3.99388701e-01 -2.66028732e-01
1.42220485e+00 -3.37011725e-01 1.42964005e+00 9.91040289e-01
9.72925663e-01 5.46694577e-01 1.28783011e+00 -1.56363830e-01
8.84549916e-01 -6.57028615e-01 -2.65725464e-01 6.90148890e-01
2.25543886e-01 5.81261158e-01 -8.06095004e-01 4.47083823e-02
1.02647007e+00 1.62063509e-01 -1.58988416e-01 -6.01719022e-01
-1.19789815e+00 5.67460060e-01 6.68358147e-01 4.86977011e-01
-4.94476438e-01 5.29395938e-01 5.90389296e-02 1.04773231e-02
-3.22712183e-01 8.06418836e-01 -6.20574392e-02 -1.52488306e-01
-8.14967215e-01 6.87674522e-01 2.90302485e-01 7.78483272e-01
6.83266282e-01 3.51649821e-01 2.55077064e-01 5.71238041e-01
3.31641585e-01 3.11284453e-01 8.32380116e-01 -1.32386971e+00
1.81447640e-02 2.16766551e-01 2.49535531e-01 -9.19672191e-01
-8.86814296e-01 -2.74968773e-01 -3.75680894e-01 4.87865359e-01
3.35989207e-01 6.16080165e-02 -6.80782437e-01 1.71957111e+00
5.66635514e-03 1.86014831e-01 8.23399499e-02 9.20970917e-01
1.61683425e-01 4.11326855e-01 4.20780778e-01 9.86700952e-02
7.16082275e-01 -7.40651906e-01 -3.16159338e-01 -5.52996814e-01
7.76846766e-01 1.51098177e-01 5.85236609e-01 2.70593137e-01
-9.85432982e-01 -6.90084398e-02 -1.35422468e+00 1.37586296e-01
-7.41856456e-01 -5.72198570e-01 6.79542542e-01 4.28037524e-01
-1.46309996e+00 1.02157998e+00 -1.00907803e+00 -8.96156728e-01
8.43010485e-01 3.96361291e-01 -5.03875732e-01 -2.11372986e-01
-1.06732059e+00 1.14274740e+00 3.58731270e-01 8.55079666e-03
-1.27125251e+00 -2.35510722e-01 -7.00607836e-01 -3.93686164e-03
-2.48456433e-01 -5.68226695e-01 1.25474179e+00 -6.17341027e-02
-1.58772039e+00 8.20186973e-01 -7.49111250e-02 -7.16937304e-01
-2.12055847e-01 -4.70245518e-02 1.79261877e-03 2.94299692e-01
7.72451013e-02 1.04863071e+00 -7.90808350e-02 -9.38251257e-01
-3.00396323e-01 -5.20217657e-01 -8.10323358e-02 3.61714840e-01
2.82136261e-01 -2.47209027e-01 4.21786904e-01 -5.67378879e-01
2.44007140e-01 -1.16339767e+00 -6.91536307e-01 3.07002068e-01
4.35316712e-01 1.64416894e-01 2.12402433e-01 -2.69678891e-01
6.13496423e-01 -1.98038316e+00 5.90213798e-02 9.61999816e-04
-9.40905362e-02 1.36048749e-01 -3.15142334e-01 7.51501501e-01
2.21550494e-01 -1.29162118e-01 -4.12744671e-01 4.21120226e-01
-2.73613483e-01 3.51494372e-01 -5.49790680e-01 5.92932463e-01
-8.13989528e-03 1.21673799e+00 -1.32133734e+00 -1.30910724e-01
-3.14649381e-02 1.02091737e-01 -7.89418340e-01 -7.38734752e-02
-2.58260578e-01 7.13558018e-01 -1.99730143e-01 5.68653703e-01
3.88996184e-01 1.10782824e-01 4.85732704e-01 6.23815358e-01
-7.15965092e-01 2.52515376e-01 -7.18197584e-01 1.94065285e+00
-1.63205087e-01 7.21391022e-01 -1.43465459e-01 -1.00990283e+00
6.71647847e-01 1.96201906e-01 4.64308798e-01 -1.39743841e+00
2.01970890e-01 2.71200806e-01 3.84412855e-01 -3.97833079e-01
3.28404903e-01 -2.49364585e-01 5.58754578e-02 5.85968673e-01
4.55257148e-01 -3.77768204e-02 3.30332935e-01 -1.24991886e-01
1.32516587e+00 5.17015874e-01 5.34799159e-01 -4.95023519e-01
-2.21271113e-01 2.90116459e-01 3.27232182e-01 9.88154173e-01
-7.04248130e-01 5.23144722e-01 1.32681489e-01 -2.52633065e-01
-1.01610446e+00 -1.42498016e+00 -1.43224290e-02 1.24488091e+00
3.76713216e-01 -1.37884334e-01 -5.19912601e-01 3.41585159e-01
-3.11938345e-01 8.75241816e-01 -5.90980589e-01 -5.21312654e-01
-8.66592348e-01 -7.04928637e-01 7.20287800e-01 5.99899769e-01
6.16123974e-01 -1.46389508e+00 -1.46732116e+00 8.23574483e-01
1.89949274e-01 -6.77190423e-01 1.61131307e-01 7.99473047e-01
-1.16123331e+00 -1.07901800e+00 -6.96582317e-01 -1.25665450e+00
5.37073016e-01 6.77007020e-01 5.01810908e-01 2.43501514e-01
-4.39380825e-01 4.88954931e-01 -1.59168407e-01 -1.44734323e-01
4.08087939e-01 -1.11615092e-01 3.04146886e-01 -7.56402135e-01
-2.67241821e-02 -1.12630534e+00 -8.77933443e-01 2.60641724e-01
-8.98601711e-01 -3.69668528e-02 4.75703716e-01 8.25721025e-01
6.45728588e-01 -3.39609444e-01 8.11492741e-01 -4.45104092e-01
6.58626676e-01 -8.76368344e-01 -4.45557356e-01 -1.40947521e-01
-4.74934042e-01 1.74451515e-01 5.97935617e-01 -3.28414023e-01
-9.56350625e-01 1.88857839e-02 -1.90878063e-01 3.41554314e-01
-3.44558448e-01 6.27533257e-01 3.00136417e-01 -3.88897240e-01
8.66522491e-01 7.98931122e-01 -1.18653819e-01 1.37822613e-01
1.91901967e-01 -2.84711659e-01 3.23300034e-01 -6.80763900e-01
3.44177186e-01 5.47451556e-01 1.69712350e-01 -9.07246828e-01
-2.11020768e-01 -1.85109749e-01 -8.65215480e-01 -4.50737655e-01
7.12122440e-01 -6.67672038e-01 -8.85486007e-01 6.67983830e-01
-8.05249274e-01 -1.18151450e+00 -1.21960506e-01 5.14615476e-01
-1.30541408e+00 -1.26316369e-01 -6.74950838e-01 -6.21726096e-01
3.50471675e-01 -6.96396232e-01 3.71285379e-01 5.34511626e-01
-4.66449201e-01 -8.05157721e-01 8.17332625e-01 -5.59447944e-01
6.34473920e-01 -9.55864936e-02 1.02057993e+00 -2.39414662e-01
-8.55021775e-01 1.27954513e-01 1.75625652e-01 -7.27199912e-01
-3.30468953e-01 -5.13543665e-01 -6.53540492e-01 -9.63598788e-02
-1.94247022e-01 -6.20682418e-01 8.73474538e-01 2.60889918e-01
8.89236629e-01 -2.74016619e-01 -5.79610407e-01 3.90866965e-01
1.28049052e+00 7.48130977e-01 8.54449928e-01 8.24123979e-01
-2.39730671e-01 6.35757565e-01 4.17746782e-01 3.31478357e-01
3.26812595e-01 2.14378640e-01 4.40959722e-01 7.00830698e-01
2.08842427e-01 -3.41921359e-01 1.26710996e-01 4.40266699e-01
-3.05977821e-01 1.85375419e-02 -1.02721298e+00 8.69631231e-01
-1.92364943e+00 -1.55414712e+00 4.25129205e-01 1.95689404e+00
6.90066218e-01 6.71360940e-02 -1.10384561e-02 -1.74634099e-01
2.54602015e-01 2.01568350e-01 -7.76728868e-01 -9.78126228e-02
-3.25121790e-01 4.04481888e-01 3.39030147e-01 3.37948263e-01
-7.05031335e-01 1.07405519e+00 8.04055977e+00 5.07232025e-02
-1.14612126e+00 9.60020497e-02 2.25287616e-01 -3.95630598e-01
-9.15502955e-04 6.34161942e-03 -5.99894583e-01 1.53027549e-01
1.12916934e+00 -3.92321020e-01 1.04800415e+00 7.06231713e-01
1.87009856e-01 -8.06672931e-01 -9.76088703e-01 6.52129710e-01
-2.19517633e-01 -1.58445692e+00 -4.58980471e-01 -9.07255858e-02
6.96518958e-01 2.47946024e-01 5.63758850e-01 5.43024659e-01
4.33962911e-01 -1.21152067e+00 7.91098595e-01 6.97904229e-01
2.46868104e-01 -4.90811199e-01 1.30478293e-02 6.63278818e-01
-8.74009728e-01 -4.41830069e-01 -4.72840011e-01 -5.81441522e-01
2.08124220e-01 -4.42076325e-02 -7.47535050e-01 -4.10636276e-01
6.87203646e-01 5.58998883e-01 -4.28721994e-01 1.76601613e+00
-7.23384693e-03 3.39651138e-01 -2.78284490e-01 -6.11537516e-01
4.21788394e-01 -2.99625881e-02 3.74839723e-01 8.21353376e-01
7.00234473e-01 5.35321474e-01 -2.58854330e-01 9.81860757e-01
3.13772947e-01 -1.21409088e-01 -1.06893837e+00 4.91126925e-02
6.86578095e-01 1.00668132e+00 -1.18584085e+00 1.67923465e-01
5.56966811e-02 8.51414680e-01 8.18748057e-01 5.63739479e-01
-6.07962191e-01 -2.81064153e-01 5.62293112e-01 2.93994278e-01
3.41439307e-01 -1.15629876e+00 -3.35569054e-01 -5.98904371e-01
-3.76289427e-01 -1.89357370e-01 -1.42641604e-01 -9.21804965e-01
-4.98797327e-01 2.85193622e-01 -7.38222823e-02 -9.28542972e-01
-3.52124065e-01 -7.17319012e-01 -6.49771154e-01 5.03004968e-01
-1.46431720e+00 -4.50399876e-01 -9.04254317e-02 5.23801982e-01
3.30226630e-01 -1.44362509e-01 1.21607363e+00 -7.26282522e-02
-2.81521380e-01 6.27769306e-02 4.10465211e-01 -1.60459802e-01
4.43639636e-01 -8.75900328e-01 3.28624815e-01 4.73067582e-01
8.93194303e-02 1.11218596e+00 7.17108965e-01 -7.44710982e-01
-1.25875270e+00 -1.03647995e+00 4.64473873e-01 -3.01196277e-01
4.16504443e-01 -5.40046275e-01 -8.45829964e-01 7.31890380e-01
1.24373764e-01 -9.07916762e-03 8.67340505e-01 -2.41420463e-01
-6.08558021e-02 2.74602950e-01 -1.17262352e+00 1.14216518e+00
1.43830502e+00 -4.87335712e-01 -6.02745116e-01 4.30901088e-02
1.06411912e-01 -2.11608335e-01 -3.45861167e-01 -6.90015927e-02
7.36508787e-01 -1.05419195e+00 9.16046858e-01 -6.71376407e-01
2.59010255e-01 -3.27040941e-01 -2.56016165e-01 -1.64387703e+00
-9.28531170e-01 -9.15034264e-02 6.13491945e-02 4.66085821e-01
2.73864418e-01 -5.36952913e-01 7.14026451e-01 3.21039230e-01
-3.71523470e-01 -5.75729787e-01 -1.31797290e+00 -9.37397957e-01
3.33539546e-01 -8.01561326e-02 -9.55530163e-03 5.13457119e-01
7.13134646e-01 -1.61253467e-01 -3.13538723e-02 -8.32694024e-02
3.20079088e-01 -2.08551943e-01 1.97061881e-01 -8.31954598e-01
5.53794131e-02 -4.09147680e-01 -3.20219040e-01 -7.80647218e-01
2.20238119e-01 -1.11538625e+00 5.99106073e-01 -1.57751465e+00
1.97607540e-02 -3.15680593e-01 -3.03726792e-01 4.65613991e-01
3.69853616e-01 2.65645310e-02 3.91067974e-02 2.00941175e-01
-9.15324628e-01 6.76372647e-01 1.06994224e+00 7.95090869e-02
-1.76904187e-01 -4.41404969e-01 -4.70070601e-01 6.64965749e-01
1.31207752e+00 -5.88617444e-01 -3.51220965e-01 -2.69614369e-01
4.09372389e-01 2.37410497e-02 6.96967185e-01 -1.70218563e+00
6.55240119e-01 -4.53359514e-01 6.98974073e-01 -3.69433165e-01
4.16635245e-01 -6.13554001e-01 -1.17431544e-02 9.28147137e-01
-4.62089479e-01 2.27484107e-01 4.05091286e-01 9.32434201e-01
2.14140534e-01 -3.47087145e-01 8.78869295e-01 -6.71454132e-01
-1.33098900e+00 1.48358524e-01 -1.67502701e+00 1.08097121e-01
1.35656333e+00 -6.09349549e-01 -5.41018784e-01 -7.15867132e-02
-9.21651781e-01 6.49824739e-02 3.91181052e-01 2.26112351e-01
9.14809942e-01 -1.46626318e+00 -8.33883882e-02 1.51500195e-01
-4.65271994e-02 -3.54318321e-01 5.40923834e-01 7.21417069e-01
-8.73522699e-01 7.69795001e-01 -1.20456123e+00 -1.80749044e-01
-3.28707457e-01 5.93174160e-01 4.85542685e-01 1.04363933e-01
-3.50852042e-01 6.34058237e-01 -5.99173866e-02 -4.79310095e-01
-1.40448347e-01 -7.68676475e-02 -3.72754544e-01 -1.57649040e-01
5.14337778e-01 4.91046757e-01 -1.76390618e-01 -3.14079493e-01
-3.63215804e-01 5.81911653e-02 1.44441873e-01 -5.33335865e-01
1.68252861e+00 -2.61955172e-01 -1.91314936e-01 8.03931832e-01
6.17431283e-01 -4.03969318e-01 -1.55858755e+00 5.46611175e-02
1.94470569e-01 -1.44327596e-01 -4.41912532e-01 -5.28420508e-01
-5.20500302e-01 1.05736434e+00 5.70593178e-01 -3.17577243e-01
6.63814545e-01 -1.74673215e-01 6.46178961e-01 1.05595565e+00
1.12334764e+00 -1.29609191e+00 6.19167566e-01 1.08322680e+00
9.36706007e-01 -6.70014262e-01 -3.32881868e-01 3.28607887e-01
-3.17236841e-01 1.10517120e+00 8.27725410e-01 -7.36332238e-01
7.16894686e-01 1.63194284e-01 -4.16110605e-01 -4.45744097e-02
-7.79654622e-01 -1.37694040e-02 -3.87266397e-01 1.24891639e+00
3.41490567e-01 -1.41387656e-01 -1.75066307e-01 7.31265426e-01
-2.31044501e-01 1.83928028e-01 7.44813859e-01 1.54488778e+00
-1.27875197e+00 -5.97597182e-01 -2.94075996e-01 2.20524207e-01
-3.87736666e-03 -1.30132154e-01 -2.77698897e-02 4.80034769e-01
-5.81315644e-02 5.45454502e-01 -3.73044237e-02 -1.72081187e-01
-1.16511017e-01 3.45424861e-01 1.00378907e+00 -6.90766454e-01
-2.45683372e-01 -3.59733909e-01 -2.18127593e-01 -6.43536985e-01
-7.81927630e-02 -1.05486298e+00 -1.74053872e+00 -2.39100158e-01
3.83865625e-01 -2.43337035e-01 2.83653617e-01 7.93656826e-01
4.19654012e-01 3.70804578e-01 4.61448766e-02 -1.23350978e+00
-1.01914257e-01 -5.41748881e-01 -9.09607410e-01 -2.06478402e-01
2.60706544e-01 -8.94762218e-01 2.07652509e-01 -5.75312600e-02] | [4.382242202758789, 1.1518956422805786] |
edc84333-fb86-4559-842d-42954ae2bf60 | stmt-a-spatial-temporal-mesh-transformer-for | 2303.18177 | null | https://arxiv.org/abs/2303.18177v1 | https://arxiv.org/pdf/2303.18177v1.pdf | STMT: A Spatial-Temporal Mesh Transformer for MoCap-Based Action Recognition | We study the problem of human action recognition using motion capture (MoCap) sequences. Unlike existing techniques that take multiple manual steps to derive standardized skeleton representations as model input, we propose a novel Spatial-Temporal Mesh Transformer (STMT) to directly model the mesh sequences. The model uses a hierarchical transformer with intra-frame off-set attention and inter-frame self-attention. The attention mechanism allows the model to freely attend between any two vertex patches to learn non-local relationships in the spatial-temporal domain. Masked vertex modeling and future frame prediction are used as two self-supervised tasks to fully activate the bi-directional and auto-regressive attention in our hierarchical transformer. The proposed method achieves state-of-the-art performance compared to skeleton-based and point-cloud-based models on common MoCap benchmarks. Code is available at https://github.com/zgzxy001/STMT. | ['Alexander Hauptmann', 'Celso M. de Melo', 'Junwei Liang', 'Po-Yao Huang', 'Xiaoyu Zhu'] | 2023-03-31 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Zhu_STMT_A_Spatial-Temporal_Mesh_Transformer_for_MoCap-Based_Action_Recognition_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Zhu_STMT_A_Spatial-Temporal_Mesh_Transformer_for_MoCap-Based_Action_Recognition_CVPR_2023_paper.pdf | cvpr-2023-1 | ['action-recognition-in-videos'] | ['computer-vision'] | [ 2.37091765e-01 1.30163446e-01 -2.49677882e-01 -4.32767928e-01
-8.24397326e-01 -4.26031575e-02 6.20278299e-01 -3.35159630e-01
-2.90203512e-01 4.18082923e-01 3.63814920e-01 1.40827835e-01
2.63075769e-01 -7.08266675e-01 -9.98418987e-01 -3.17575902e-01
5.98207042e-02 4.05504465e-01 7.32164502e-01 5.86865768e-02
1.15093291e-01 3.76360536e-01 -1.40087879e+00 8.49926710e-01
4.47958618e-01 1.13084364e+00 2.39811346e-01 8.89939368e-01
-1.28024042e-01 1.12039387e+00 -1.33191511e-01 -3.39023918e-01
3.68130118e-01 -3.41370910e-01 -9.77979064e-01 3.68070841e-01
8.08860242e-01 -2.54730880e-01 -6.98872983e-01 6.33802295e-01
3.18010658e-01 3.45961779e-01 4.26416159e-01 -1.23310471e+00
-3.82739395e-01 1.91244200e-01 -9.03112769e-01 4.92433965e-01
3.67104501e-01 5.74430525e-01 8.58194411e-01 -1.08689022e+00
7.15869367e-01 1.33037055e+00 7.55221248e-01 6.15981162e-01
-1.30495667e+00 -5.71403980e-01 4.61416185e-01 6.06017292e-01
-1.28668249e+00 -4.66989040e-01 9.99063969e-01 -7.06071317e-01
1.40413523e+00 4.51467186e-02 9.77591693e-01 1.28542757e+00
4.12845939e-01 9.25302863e-01 6.24669433e-01 -2.49138623e-01
-6.73852861e-02 -6.80814266e-01 -1.14336282e-01 9.40248966e-01
-4.01443630e-01 -1.52883225e-03 -7.96377718e-01 -6.98257983e-02
1.42297649e+00 1.00795448e-01 -1.59724832e-01 -5.51907897e-01
-1.36252129e+00 5.86628318e-01 5.06199419e-01 7.51192272e-02
-7.40459263e-01 8.77216339e-01 2.61486322e-01 -1.97907209e-01
4.88306642e-01 -1.09715037e-01 -4.87304598e-01 -2.79758185e-01
-1.11893272e+00 2.37034142e-01 1.21507145e-01 8.48728657e-01
6.66630208e-01 1.70144469e-01 -3.18956435e-01 6.10386491e-01
3.66349190e-01 2.87197739e-01 4.99909580e-01 -1.36991298e+00
6.14551425e-01 5.75572848e-01 5.43765165e-02 -8.03342700e-01
-2.08019659e-01 -2.52482027e-01 -5.12153447e-01 5.39153755e-01
4.97262418e-01 1.42866537e-01 -1.18336010e+00 1.42356133e+00
4.67108667e-01 7.58333027e-01 -4.09406930e-01 9.71754551e-01
7.81548858e-01 5.67030966e-01 4.95200992e-01 2.44213879e-01
1.16434228e+00 -1.55460727e+00 -4.15760309e-01 -4.09355849e-01
4.21009630e-01 -4.64916319e-01 1.16877437e+00 1.23847105e-01
-1.47850442e+00 -8.81717443e-01 -7.37550259e-01 -5.24639249e-01
-4.31532692e-03 4.49185912e-03 4.02638972e-01 -1.51573375e-01
-8.22375417e-01 7.16915905e-01 -1.48913097e+00 -2.27933958e-01
7.80754209e-01 3.61972779e-01 -4.99812752e-01 2.45379247e-02
-8.07343543e-01 5.72266102e-01 -4.71340902e-02 -1.38327060e-02
-1.15364206e+00 -6.63471043e-01 -9.15344536e-01 -3.27912010e-02
1.95811763e-01 -1.07508492e+00 1.23165739e+00 -1.27142906e+00
-1.50979984e+00 9.96355414e-01 -4.71410334e-01 -5.48960686e-01
7.59562552e-01 -4.35706586e-01 1.04328021e-02 4.07890886e-01
2.70079136e-01 1.08021820e+00 1.00508416e+00 -1.00124896e+00
-6.48643255e-01 -1.95646554e-01 -2.14881506e-02 2.42307454e-01
1.44562185e-01 1.81763336e-01 -8.50992262e-01 -1.04596853e+00
1.19979009e-01 -7.28551745e-01 -3.04818004e-01 4.28987712e-01
-1.63056418e-01 -9.17685553e-02 7.53632605e-01 -7.90800571e-01
1.10918093e+00 -1.84996402e+00 4.65123028e-01 -2.06181824e-01
1.09296806e-01 1.81872606e-01 -2.05551609e-01 8.04709643e-02
-2.91548669e-01 -2.25062862e-01 -4.20865536e-01 -7.71679938e-01
-2.66563863e-01 3.37371796e-01 -2.76199952e-02 4.06193912e-01
5.01355529e-01 1.21891630e+00 -8.88819695e-01 -8.01024735e-01
5.94542682e-01 7.48355806e-01 -5.50855875e-01 2.15031788e-01
-4.44616228e-01 7.10066020e-01 -4.77940500e-01 7.48103440e-01
1.50353000e-01 -5.65544307e-01 -2.25785613e-01 -1.12869523e-01
1.31299406e-01 1.33817688e-01 -8.89467895e-01 2.19866538e+00
-2.53661007e-01 5.52956283e-01 -1.92384690e-01 -8.67957652e-01
6.15185022e-01 3.90663505e-01 8.01165581e-01 -5.18274903e-01
1.27845690e-01 -3.00460726e-01 -3.46267104e-01 -5.92454016e-01
1.48568884e-01 -5.69459703e-03 2.84569353e-01 2.09310308e-01
1.82834998e-01 4.04933989e-01 -9.94623918e-03 -3.39729786e-02
1.32028484e+00 9.37887371e-01 5.51911741e-02 -6.52151834e-03
6.99259341e-01 8.76764134e-02 8.47179294e-01 2.69223511e-01
-2.59076953e-01 1.14572287e+00 3.29891503e-01 -7.92353749e-01
-9.77284551e-01 -1.06899023e+00 2.27682024e-01 1.16013348e+00
-2.92239391e-04 -4.90643233e-01 -6.52517200e-01 -7.42138922e-01
-6.70156255e-02 5.83017826e-01 -9.00402308e-01 9.43252221e-02
-9.23436046e-01 -1.53896451e-01 3.20572913e-01 1.06773150e+00
5.31385481e-01 -1.41495419e+00 -1.05389416e+00 3.40458483e-01
-3.70465428e-01 -1.02947700e+00 -6.78521454e-01 -9.89782289e-02
-1.07943451e+00 -1.11169910e+00 -8.95838380e-01 -5.30089915e-01
6.70120537e-01 -5.70673309e-02 1.17290199e+00 1.30050331e-01
-2.53796458e-01 6.00793004e-01 -1.83804482e-01 -1.07271113e-01
1.21654801e-01 2.71406174e-02 -1.93202198e-01 4.35135007e-01
4.11536336e-01 -7.42555976e-01 -7.46379018e-01 2.85030991e-01
-6.61510527e-01 5.11265755e-01 3.58021677e-01 6.44906342e-01
9.79974806e-01 -5.39576530e-01 9.28539857e-02 -5.79171300e-01
-1.18222140e-01 -4.51808989e-01 -3.76537383e-01 6.78579062e-02
-7.18031749e-02 -6.53044283e-02 1.50080413e-01 -5.68738699e-01
-9.27345395e-01 6.89455390e-01 -1.06809907e-01 -1.19094396e+00
-2.42874861e-01 1.66919872e-01 -9.03598890e-02 7.58230500e-03
5.42971969e-01 1.64945394e-01 -1.14284776e-01 -6.90418363e-01
1.93489939e-01 7.21925218e-03 8.29129815e-01 -5.21007538e-01
6.41997039e-01 7.30561733e-01 -1.46561041e-01 -6.26146674e-01
-7.29758441e-01 -3.10846508e-01 -1.26959455e+00 -5.14228642e-01
1.42346334e+00 -1.00565195e+00 -1.77095577e-01 4.66383487e-01
-1.33945930e+00 -9.84287024e-01 -3.41975302e-01 3.96885097e-01
-9.35984731e-01 3.95957202e-01 -6.01157665e-01 -6.96779072e-01
-3.15272003e-01 -1.00837111e+00 1.36707985e+00 -3.64365764e-02
-4.47664976e-01 -9.61759388e-01 1.34020656e-01 5.73139310e-01
-4.33847234e-02 7.06646681e-01 4.46034849e-01 -1.36007994e-01
-8.28784943e-01 2.44179126e-02 6.26330823e-02 9.95667577e-02
-2.37351116e-02 -3.30252051e-02 -8.76646459e-01 -8.25081207e-03
-3.06992203e-01 -1.84277236e-01 9.62143004e-01 7.17738688e-01
1.30767059e+00 -1.25286922e-01 -4.86627281e-01 9.77817178e-01
1.14601135e+00 -1.26524400e-02 9.09513652e-01 4.36178356e-01
1.18274534e+00 4.34577465e-01 4.47139710e-01 3.76094878e-01
4.95663881e-01 9.77818370e-01 4.01976228e-01 -1.96605489e-01
-3.48187983e-01 -3.40376139e-01 3.82702082e-01 2.97055095e-01
-5.87696493e-01 1.49068490e-01 -1.25590312e+00 6.88809454e-01
-2.20965791e+00 -1.20544958e+00 -3.82687539e-01 2.05370045e+00
6.00575626e-01 3.38506103e-01 2.83213764e-01 -6.75727203e-02
6.56482100e-01 2.54137725e-01 -5.06070912e-01 -4.30998094e-02
7.52335861e-02 3.42254311e-01 3.51635665e-01 7.12222338e-01
-1.26598418e+00 1.15350473e+00 5.64899349e+00 5.55270553e-01
-1.10609269e+00 3.43938500e-01 6.77976370e-01 -5.16776681e-01
1.83227465e-01 -7.84886554e-02 -4.49872792e-01 4.94726539e-01
8.11228335e-01 1.53120935e-01 2.36406863e-01 8.79534125e-01
3.40692103e-01 8.83029923e-02 -1.15365314e+00 1.03712308e+00
-5.27703054e-02 -1.47384489e+00 -3.77829373e-02 2.43542865e-02
6.88677728e-01 3.41050804e-01 -2.40722418e-01 2.78053060e-02
1.24382228e-01 -1.11558914e+00 1.04080892e+00 9.67623115e-01
8.44285548e-01 -4.95287716e-01 2.64746636e-01 2.04439566e-01
-1.58122277e+00 -4.15720269e-02 -2.39846274e-01 -1.40227377e-01
4.56081092e-01 4.60519046e-02 -1.69796973e-01 3.62981975e-01
9.29582953e-01 1.02800262e+00 -6.88614368e-01 9.80919957e-01
-1.82519600e-01 5.59248507e-01 -2.00425297e-01 6.37175322e-01
2.58777708e-01 9.35950875e-02 6.30433500e-01 1.13229871e+00
1.00388475e-01 2.29664728e-01 9.65458751e-02 7.96822727e-01
2.46665388e-01 -2.25427836e-01 -3.38810861e-01 2.81375378e-01
-1.61473118e-02 8.96718025e-01 -7.04165339e-01 -6.33182764e-01
-6.53510749e-01 1.15449297e+00 4.47976828e-01 4.02132750e-01
-1.16959381e+00 7.04493821e-02 7.69603491e-01 7.28070974e-01
7.56288350e-01 -3.91965210e-01 -4.31746721e-01 -1.25904119e+00
3.47585455e-02 -6.26914799e-01 5.29840112e-01 -1.09038830e+00
-1.06608224e+00 5.58228135e-01 1.39689088e-01 -1.32012665e+00
-1.93244621e-01 -4.77350056e-01 -8.83755803e-01 7.97587276e-01
-1.19607067e+00 -1.43748903e+00 -5.40234447e-01 8.61880541e-01
1.03127754e+00 -1.38202114e-02 6.50402248e-01 2.75263935e-01
-4.10425603e-01 2.43340984e-01 -5.20950735e-01 3.77802461e-01
3.91604453e-01 -9.81121659e-01 8.08073819e-01 8.56142402e-01
2.88262099e-01 1.42300919e-01 3.83636951e-01 -9.99939263e-01
-8.41787875e-01 -1.35896850e+00 8.47180963e-01 -1.03808403e+00
4.95482266e-01 -3.08272421e-01 -1.10073197e+00 1.07785070e+00
9.19819102e-02 6.39463603e-01 3.00152898e-01 -4.15210515e-01
-1.73437223e-01 1.90836996e-01 -6.93501651e-01 4.46918905e-01
1.39385927e+00 -4.06734973e-01 -7.28531599e-01 2.64882267e-01
4.14797813e-01 -5.84562123e-01 -8.13702345e-01 3.85576457e-01
6.34875953e-01 -9.86456931e-01 1.19440484e+00 -8.79973352e-01
7.92021275e-01 -4.23137009e-01 3.43537666e-02 -6.84432268e-01
-6.97113097e-01 -5.83027482e-01 -5.40311337e-01 8.15023601e-01
2.42706820e-01 -1.10833995e-01 1.19570792e+00 6.36779130e-01
-2.35922143e-01 -1.07245529e+00 -1.07036066e+00 -5.20177484e-01
1.82181969e-02 -6.22557819e-01 4.01434928e-01 8.30171108e-01
-2.63327241e-01 2.47332990e-01 -4.01494980e-01 1.33282363e-01
6.37532711e-01 -1.84168875e-01 8.95129800e-01 -1.05256999e+00
-5.46683848e-01 -4.23620015e-01 -6.26206160e-01 -1.14365041e+00
1.91980585e-01 -6.91422224e-01 -1.64539173e-01 -1.77125049e+00
-1.27601504e-01 5.48361540e-02 -3.22329462e-01 7.46054769e-01
-1.77229300e-01 4.92467821e-01 2.11390316e-01 3.08886170e-01
-7.01891601e-01 6.68700814e-01 1.23284590e+00 -1.25747621e-01
-1.34413838e-01 -2.91296933e-02 4.62017059e-02 9.17197824e-01
6.78952932e-01 -6.24013186e-01 -3.83260667e-01 -6.84438169e-01
-3.56822222e-01 2.44692713e-01 7.90763795e-01 -1.40079772e+00
1.20840728e-01 -3.83103222e-01 7.06557810e-01 -7.10236549e-01
8.00607324e-01 -7.30146289e-01 4.50654268e-01 4.68553483e-01
-3.47294480e-01 2.31425837e-01 9.83913913e-02 7.03622222e-01
5.79049205e-03 1.27178416e-01 6.69538021e-01 -4.57547158e-01
-7.65467763e-01 6.39242351e-01 -2.60461718e-01 6.06305664e-03
1.15838957e+00 -5.01144588e-01 7.78141245e-02 -2.92972416e-01
-1.12875044e+00 2.62227416e-01 6.40854418e-01 5.42998970e-01
7.54744947e-01 -1.51726019e+00 -6.13015115e-01 1.18321911e-01
3.74732204e-02 2.46797249e-01 3.44421655e-01 9.51397061e-01
-7.70595849e-01 2.67637968e-01 -5.81050277e-01 -8.25833142e-01
-1.26102889e+00 5.32257080e-01 5.13372302e-01 -2.52165794e-01
-1.12793076e+00 1.04059923e+00 4.90684301e-01 1.29186446e-02
2.82003790e-01 -3.82490516e-01 5.21244295e-02 -4.61049259e-01
4.50358123e-01 3.28312606e-01 -2.64519870e-01 -1.12061143e+00
-3.58498991e-01 9.09797847e-01 3.51592660e-01 -2.60297239e-01
1.32714486e+00 1.89254191e-02 2.45614335e-01 5.68259180e-01
1.04245973e+00 -4.26410526e-01 -1.97333932e+00 -2.18600512e-01
-7.45686293e-02 -7.62028217e-01 9.07502323e-02 -4.37041640e-01
-1.20630503e+00 1.15835559e+00 5.43696702e-01 -3.97847027e-01
8.92777979e-01 -1.74251944e-02 9.37254369e-01 -7.12006390e-02
1.54994860e-01 -1.01163292e+00 3.29142094e-01 3.17935079e-01
1.23015022e+00 -1.09734368e+00 -1.27656860e-02 -3.86812449e-01
-8.03634405e-01 9.80302036e-01 1.04125237e+00 -3.61880451e-01
6.26278281e-01 1.73442010e-02 5.70339188e-02 -3.71935993e-01
-9.00540113e-01 -2.10034475e-01 4.99112904e-01 7.06440210e-01
4.73295122e-01 -2.32874349e-01 4.22110222e-02 5.19590139e-01
7.15227285e-03 2.09095627e-01 7.65030086e-02 1.05531490e+00
-1.63762361e-01 -1.06162810e+00 -4.10784423e-01 3.80251855e-01
-4.02184665e-01 9.25856605e-02 -2.57203788e-01 6.38571203e-01
2.12523460e-01 4.85638648e-01 4.30857033e-01 -4.02916789e-01
3.16844940e-01 3.78761679e-01 5.95444441e-01 -7.07754672e-01
-6.03077412e-01 3.07057321e-01 6.32442459e-02 -1.05161548e+00
-5.81987321e-01 -8.63265991e-01 -1.32710338e+00 6.25328869e-02
2.30215684e-01 -3.89496535e-01 1.37379229e-01 7.91014135e-01
5.81691504e-01 7.28900373e-01 9.05615650e-03 -1.40643179e+00
5.24186976e-02 -1.00063658e+00 7.22421007e-03 4.86085147e-01
2.88059324e-01 -1.01252806e+00 6.43476322e-02 6.34286821e-01] | [8.447073936462402, 0.3512386977672577] |
85d7bec4-bf34-4642-80d5-61be20ed73ee | monocular-expressive-body-regression-through | 2008.09062 | null | https://arxiv.org/abs/2008.09062v1 | https://arxiv.org/pdf/2008.09062v1.pdf | Monocular Expressive Body Regression through Body-Driven Attention | To understand how people look, interact, or perform tasks, we need to quickly and accurately capture their 3D body, face, and hands together from an RGB image. Most existing methods focus only on parts of the body. A few recent approaches reconstruct full expressive 3D humans from images using 3D body models that include the face and hands. These methods are optimization-based and thus slow, prone to local optima, and require 2D keypoints as input. We address these limitations by introducing ExPose (EXpressive POse and Shape rEgression), which directly regresses the body, face, and hands, in SMPL-X format, from an RGB image. This is a hard problem due to the high dimensionality of the body and the lack of expressive training data. Additionally, hands and faces are much smaller than the body, occupying very few image pixels. This makes hand and face estimation hard when body images are downscaled for neural networks. We make three main contributions. First, we account for the lack of training data by curating a dataset of SMPL-X fits on in-the-wild images. Second, we observe that body estimation localizes the face and hands reasonably well. We introduce body-driven attention for face and hand regions in the original image to extract higher-resolution crops that are fed to dedicated refinement modules. Third, these modules exploit part-specific knowledge from existing face- and hand-only datasets. ExPose estimates expressive 3D humans more accurately than existing optimization methods at a small fraction of the computational cost. Our data, model and code are available for research at https://expose.is.tue.mpg.de . | ['Michael J. Black', 'Dimitrios Tzionas', 'Vasileios Choutas', 'Georgios Pavlakos', 'Timo Bolkart'] | 2020-08-20 | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/983_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123550018.pdf | eccv-2020-8 | ['3d-human-reconstruction', '3d-multi-person-mesh-recovery'] | ['computer-vision', 'computer-vision'] | [-2.12933812e-02 1.44040391e-01 -1.15605414e-01 -3.42085719e-01
-4.09670800e-01 -4.64044273e-01 3.41977865e-01 -5.37447691e-01
-3.26148570e-01 4.27376062e-01 3.09546739e-01 4.58934337e-01
3.42363775e-01 -4.58634049e-01 -6.86961949e-01 -3.66304904e-01
2.40985692e-01 8.83362889e-01 -1.60694957e-01 -1.31391600e-01
-4.53875601e-01 6.15583539e-01 -1.58654857e+00 6.60180673e-02
3.17220420e-01 1.19489980e+00 -3.07818413e-01 8.58669519e-01
2.35432699e-01 3.99723500e-01 -2.44388714e-01 -5.21164060e-01
6.04706287e-01 -3.68283957e-01 -6.30204678e-01 3.81081223e-01
9.68117833e-01 -9.24686670e-01 -4.08458203e-01 4.91922528e-01
6.62148178e-01 1.83159843e-01 5.24335265e-01 -1.24125075e+00
-1.76320866e-01 1.66935213e-02 -8.97133946e-01 -4.68607098e-01
5.72758138e-01 4.21296954e-01 6.73685133e-01 -8.72738481e-01
7.10535526e-01 1.45679629e+00 1.00917840e+00 1.05920136e+00
-1.26810038e+00 -7.54211426e-01 2.35591352e-01 -3.65367442e-01
-1.54326642e+00 -7.97239780e-01 6.83490634e-01 -5.27624846e-01
9.47600961e-01 3.89083683e-01 1.26160014e+00 1.11652708e+00
-2.63029337e-01 6.79153979e-01 1.09424841e+00 -2.86804050e-01
-2.50651896e-01 -7.92789161e-02 -1.74385726e-01 1.17732251e+00
2.51223426e-02 1.34291127e-02 -7.41988301e-01 -7.39260437e-03
1.16091609e+00 1.30861804e-01 -2.42291749e-01 -4.87721711e-01
-9.89807844e-01 4.38479781e-01 5.06533980e-01 -2.37847179e-01
-3.78490508e-01 4.18125331e-01 -5.15863486e-02 -2.08287224e-01
4.05642599e-01 -4.15866002e-02 -6.93362176e-01 -1.71669617e-01
-1.18688285e+00 5.61669648e-01 7.78868377e-01 8.34513128e-01
6.72530949e-01 -2.16270268e-01 1.52339519e-03 5.87968290e-01
3.56789500e-01 9.11330998e-01 4.13480587e-02 -1.27811515e+00
3.86646330e-01 6.30461037e-01 5.39686307e-02 -7.46303678e-01
-7.71772921e-01 -9.96416658e-02 -7.26024330e-01 4.95380461e-01
7.71498561e-01 -2.14107648e-01 -1.05992925e+00 1.71940923e+00
8.17425609e-01 -3.01811546e-01 -5.37477553e-01 1.24096322e+00
9.89430606e-01 1.96187273e-01 7.28688296e-03 1.29600391e-01
1.47409666e+00 -9.59242344e-01 -4.21849430e-01 -5.09651423e-01
6.80111498e-02 -4.43077266e-01 1.13712049e+00 3.91008705e-01
-1.49822891e+00 -5.77688932e-01 -5.98020077e-01 -5.15144646e-01
-2.01565370e-01 3.44692171e-01 7.86943078e-01 6.15051925e-01
-9.46973383e-01 6.52403593e-01 -1.03696382e+00 -5.22127867e-01
5.73256552e-01 8.51807296e-01 -7.04460680e-01 8.30142871e-02
-5.83516777e-01 8.35951686e-01 -6.74476177e-02 1.43749297e-01
-5.75114191e-01 -6.33413434e-01 -1.11835504e+00 -1.70104638e-01
4.18891817e-01 -9.90155220e-01 1.27451253e+00 -1.18980742e+00
-1.65037298e+00 1.27122235e+00 -1.97454885e-01 3.25110443e-02
8.38112175e-01 -5.04564524e-01 1.82637542e-01 1.83232427e-01
-3.31370592e-01 1.03639698e+00 1.13901424e+00 -1.09960210e+00
-6.98785633e-02 -9.38466847e-01 -4.31376584e-02 4.41431344e-01
-1.25141025e-01 1.44844027e-02 -1.07130289e+00 -4.21276689e-01
4.35173931e-03 -1.08450818e+00 1.33896127e-01 7.53516614e-01
-4.29347932e-01 2.04977781e-01 5.46478689e-01 -1.07606053e+00
7.95670927e-01 -1.87333536e+00 5.91370344e-01 2.46908754e-01
5.71777046e-01 9.53467861e-02 -5.54127172e-02 -9.73117724e-02
6.66227117e-02 -9.18836221e-02 -1.36048883e-01 -1.08416831e+00
1.75417408e-01 3.11433971e-01 2.23204911e-01 7.68627942e-01
6.60045594e-02 1.17258513e+00 -3.37532640e-01 -7.47930825e-01
3.10803086e-01 8.81451249e-01 -7.01467693e-01 2.72856832e-01
-2.22521797e-01 7.36248672e-01 -1.96707025e-01 1.09566665e+00
6.50788724e-01 -2.11347267e-01 1.49377465e-01 -3.67389470e-01
7.14138001e-02 2.13709474e-02 -1.17061698e+00 1.84955764e+00
-3.90495747e-01 2.72316128e-01 6.70674562e-01 -4.48996365e-01
3.25980604e-01 2.54543573e-01 6.73785865e-01 -4.40447599e-01
4.19056892e-01 -2.49746777e-02 -2.86085516e-01 -2.46075153e-01
-3.92709151e-02 -3.69291008e-01 2.42232382e-02 4.75021511e-01
1.71863988e-01 -4.03943151e-01 -4.27798107e-02 -1.48205489e-01
7.69552529e-01 7.33313084e-01 2.67150193e-01 2.32532710e-01
9.08513963e-02 -2.50319064e-01 4.28966582e-01 2.82672197e-01
-5.91941513e-02 8.78121078e-01 3.18714559e-01 -5.62799096e-01
-9.75341141e-01 -9.62882102e-01 6.22598641e-02 1.27938557e+00
-1.47460803e-01 -6.19332612e-01 -1.07839549e+00 -4.88017082e-01
3.60956967e-01 -2.19604354e-02 -8.60878289e-01 1.38598010e-01
-7.09006429e-01 -3.37762505e-01 4.61144388e-01 7.49645829e-01
4.33059931e-01 -9.20389891e-01 -8.58518183e-01 -2.28721395e-01
-1.89234257e-01 -1.06528389e+00 -6.70845151e-01 -6.81075975e-02
-7.42635369e-01 -1.20151460e+00 -9.34755862e-01 -1.37369975e-01
8.34986448e-01 -3.42527777e-01 1.28626990e+00 3.42198521e-01
-6.57486975e-01 7.20587611e-01 9.22556221e-02 -3.94758910e-01
1.91232517e-01 -4.33851629e-02 3.44947547e-01 -1.46419749e-01
1.02000237e-01 -5.82786560e-01 -6.03350580e-01 2.28413522e-01
-3.04833531e-01 2.84820884e-01 4.05575573e-01 5.39710283e-01
7.98201680e-01 -3.53759617e-01 -3.95085871e-01 -4.35000956e-01
8.91008601e-02 -9.01297957e-04 -6.04998767e-01 -5.33881262e-02
-1.18207686e-01 -2.88567040e-02 2.84102798e-01 -5.32174468e-01
-9.49520588e-01 6.60223722e-01 -2.35884860e-01 -6.14892423e-01
-2.35331774e-01 -2.09139928e-01 -3.41623813e-01 -2.51067758e-01
5.19203842e-01 -1.82190076e-01 3.49745750e-01 -7.37208068e-01
3.09589803e-01 4.35078442e-01 5.76696813e-01 -6.98388755e-01
7.39679158e-01 7.44999290e-01 1.55322909e-01 -8.27236593e-01
-9.83642042e-01 -2.99737632e-01 -1.11367643e+00 -4.00767326e-01
8.63503695e-01 -9.43706930e-01 -1.25582314e+00 5.29463887e-01
-9.40986156e-01 -8.01769376e-01 -4.69130784e-01 3.71391952e-01
-6.73176587e-01 1.31728187e-01 -5.95631778e-01 -9.19127047e-01
-6.77372098e-01 -8.18923235e-01 1.66254091e+00 1.65965319e-01
-7.42175996e-01 -6.11953080e-01 -1.45814465e-02 7.39838660e-01
2.02710871e-02 5.56978762e-01 2.42979482e-01 1.18161373e-01
-3.18857223e-01 -2.14009359e-01 -7.47600198e-02 1.28783435e-01
-1.87433869e-01 -2.73909532e-02 -1.02242362e+00 -3.04578096e-01
-3.60699832e-01 -7.63587296e-01 6.68882906e-01 4.61323321e-01
1.12893987e+00 -4.52091724e-01 -3.29350471e-01 1.27261639e+00
9.51815248e-01 -5.95905483e-01 2.18408465e-01 -7.79283494e-02
9.89015818e-01 8.20079505e-01 2.20367387e-01 6.79273069e-01
5.30720174e-01 9.34398174e-01 3.47873330e-01 -3.25557977e-01
-5.13586521e-01 -3.75152320e-01 3.82484376e-01 3.15180659e-01
-7.14505136e-01 1.89801067e-01 -9.65804160e-01 1.28487304e-01
-1.57729352e+00 -7.25021243e-01 4.41142581e-02 2.07113671e+00
9.94921505e-01 -2.27305919e-01 6.89350903e-01 1.98811144e-02
1.20573752e-01 -2.69942641e-01 -7.27632165e-01 3.70817818e-02
-1.34623824e-02 5.68776786e-01 4.15627331e-01 5.97143710e-01
-1.14935803e+00 9.18447673e-01 5.90537691e+00 2.07187831e-01
-1.05765152e+00 9.89027023e-02 4.59019661e-01 -8.53622139e-01
1.72725379e-01 -3.42442155e-01 -8.68895233e-01 2.39860043e-01
5.03710389e-01 4.71669197e-01 9.39186573e-01 7.44184315e-01
-2.38711890e-02 -7.57733509e-02 -1.27311385e+00 1.45584786e+00
1.53733060e-01 -6.15287602e-01 -4.46324646e-01 2.02100843e-01
4.57697600e-01 -5.29931672e-02 1.55480891e-01 1.17606498e-01
1.14443317e-01 -1.42021978e+00 1.10713434e+00 7.85281003e-01
1.23658991e+00 -5.93576074e-01 3.07603866e-01 4.42295611e-01
-1.13589692e+00 -2.60689631e-02 -1.53125241e-01 -2.52038956e-01
1.01213500e-01 2.48907700e-01 -3.16802830e-01 -6.84521422e-02
1.04004657e+00 5.42116284e-01 -5.25855184e-01 3.42205822e-01
-3.68735373e-01 2.15518013e-01 -9.33954060e-01 3.41675133e-01
-3.05440485e-01 -8.24339688e-02 2.83175260e-01 9.86212850e-01
1.26022995e-01 3.75347167e-01 1.47960618e-01 1.13861406e+00
-2.96040446e-01 -7.21861869e-02 -3.09155911e-01 3.65317799e-02
-1.41040921e-01 1.43694878e+00 -5.83207011e-01 -2.19800994e-01
-2.84400880e-01 1.29463649e+00 4.71334159e-01 2.19340563e-01
-8.20946217e-01 1.44427642e-01 7.97946513e-01 7.48232126e-01
1.63895443e-01 -2.97765523e-01 -2.31665269e-01 -1.33439350e+00
2.08318666e-01 -1.09318864e+00 1.29774213e-01 -8.75193536e-01
-8.60628366e-01 2.54318595e-01 8.21139216e-02 -5.74473679e-01
-3.38406056e-01 -7.91398525e-01 -9.15263593e-02 8.88399363e-01
-9.84899163e-01 -1.79987538e+00 -7.39711940e-01 8.76604974e-01
2.36159205e-01 3.55007529e-01 9.20508265e-01 1.64064109e-01
-6.09598160e-01 6.27434969e-01 -4.40404296e-01 3.07452917e-01
6.78728878e-01 -1.12006795e+00 4.74868685e-01 3.23946685e-01
9.76762399e-02 7.46757388e-01 4.35739636e-01 -7.02056050e-01
-1.61270750e+00 -6.53301835e-01 5.39510667e-01 -1.09043896e+00
-7.62382299e-02 -8.49730432e-01 -4.25157607e-01 9.22250569e-01
-3.07696879e-01 4.81324583e-01 6.36484563e-01 3.33871365e-01
-2.62349278e-01 -6.62113875e-02 -1.27260458e+00 4.98726666e-01
1.38660729e+00 -4.58578974e-01 -3.37185860e-01 2.40179315e-01
5.43471565e-03 -8.33359063e-01 -8.14271748e-01 1.50416359e-01
1.32024002e+00 -1.10784733e+00 1.29694295e+00 -4.12040979e-01
2.28347957e-01 -8.10167864e-02 -2.72124354e-02 -9.12385404e-01
-3.81561108e-02 -6.14879191e-01 -6.71051145e-01 1.05505598e+00
-9.74537134e-02 -2.34486252e-01 1.07658732e+00 1.12859213e+00
5.35270572e-01 -7.69343317e-01 -8.03764641e-01 -3.91334206e-01
-5.42966500e-02 -2.98304826e-01 6.10297561e-01 4.76080716e-01
-1.92501724e-01 2.33947575e-01 -5.74325621e-01 -2.20371068e-01
7.89519191e-01 9.89495143e-02 1.24571574e+00 -1.35804999e+00
-5.07208765e-01 -2.38030732e-01 -1.07831933e-01 -9.43748176e-01
2.16786146e-01 -5.65494180e-01 -2.56215818e-02 -1.28070378e+00
3.23568285e-01 -2.72221684e-01 3.67574275e-01 1.08120537e+00
-8.07798356e-02 7.29926050e-01 4.43830520e-01 1.96212992e-01
-4.51842397e-01 4.07434672e-01 1.25410390e+00 6.15630150e-02
-1.09793998e-01 5.52737787e-02 -4.08083379e-01 1.22780955e+00
7.01217473e-01 -1.97523072e-01 2.69761961e-02 -4.60848510e-01
1.52693644e-01 2.79308874e-02 8.08219433e-01 -9.25604939e-01
-9.86149758e-02 -5.36555704e-03 1.25503671e+00 -4.59906220e-01
1.08653426e+00 -9.31885481e-01 3.87670994e-01 3.99331391e-01
6.26401901e-02 -1.68848217e-01 2.78102607e-01 3.21387500e-02
4.25148815e-01 2.51540244e-01 9.43982184e-01 -5.15424013e-01
-3.16522479e-01 5.80803990e-01 1.76042795e-01 6.71941554e-03
7.95955777e-01 -3.84474099e-01 3.61387759e-01 -6.02185488e-01
-8.57205272e-01 -1.21994941e-02 9.31993365e-01 1.57801464e-01
4.89465207e-01 -1.15367877e+00 -5.45723617e-01 5.70098877e-01
-2.63075858e-01 3.31662178e-01 2.03479290e-01 9.60053146e-01
-6.41063750e-01 1.36683106e-01 -3.16279739e-01 -4.19487238e-01
-1.75862193e+00 4.18605208e-01 6.14300072e-01 -3.12900469e-02
-6.54053032e-01 1.08963442e+00 2.25910097e-01 -7.78767109e-01
3.38797987e-01 -3.73401314e-01 2.95055211e-01 -1.46245420e-01
5.31288683e-01 5.00522912e-01 -3.59435707e-01 -1.17161167e+00
-5.57697415e-01 1.23077810e+00 4.90569621e-01 -2.55525261e-01
1.31286168e+00 1.18695451e-02 -2.00629339e-01 1.53143359e-02
1.07575345e+00 2.50355124e-01 -1.65726912e+00 1.24463094e-02
-7.27222741e-01 -4.76181835e-01 5.23262359e-02 -8.23237002e-01
-1.15144622e+00 1.20130181e+00 5.02286911e-01 -5.27092814e-01
1.19089925e+00 3.21099102e-01 7.67918289e-01 8.67179781e-02
1.85222283e-01 -1.24125624e+00 3.14434655e-02 2.84475267e-01
1.15121293e+00 -1.14777040e+00 4.82018948e-01 -3.73485506e-01
-4.43159312e-01 1.05395055e+00 8.44616354e-01 2.32445702e-01
4.38656807e-01 5.59086144e-01 -1.01802602e-01 -2.73952782e-01
-2.39592940e-01 -4.34148550e-01 6.54757202e-01 7.31244206e-01
3.98730963e-01 4.18350548e-02 4.65757787e-01 7.31152356e-01
-6.00437820e-01 9.94472057e-02 -3.22707474e-01 8.57652366e-01
-5.82001507e-02 -1.00014842e+00 -7.96430349e-01 2.92902589e-01
-4.87300545e-01 7.66157731e-02 -8.54459941e-01 7.99218595e-01
5.98403990e-01 4.15571451e-01 9.60105658e-02 -1.38194785e-01
4.84237224e-01 1.41096726e-01 1.12301660e+00 -5.60178399e-01
-4.87766027e-01 1.51904345e-01 -1.39478207e-01 -1.06484032e+00
-2.90163070e-01 -7.06926644e-01 -1.28909659e+00 -5.68017840e-01
-1.21117860e-01 -5.79948127e-01 6.25375390e-01 7.96120167e-01
2.41462126e-01 2.21182704e-01 -1.73861519e-01 -1.65437949e+00
-4.32829648e-01 -8.43148649e-01 -6.43899500e-01 4.69414890e-01
4.94118899e-01 -8.34538162e-01 -1.47454426e-01 2.20189482e-01] | [7.123443603515625, -1.070980191230774] |
81ed42e5-444e-496f-869e-8b61f3e4ac79 | event-triggered-hybrid-energy-aware | 2302.00812 | null | https://arxiv.org/abs/2302.00812v1 | https://arxiv.org/pdf/2302.00812v1.pdf | Event-triggered Hybrid Energy-aware Scheduling in Manufacturing Systems | Incorporating renewable energy sources (RESs) into manufacturing systems has been an active research area in order to address many challenges originating from the unpredictable nature of RESs such as photovoltaics.In the energy-aware scheduling for manufacturing systems, the traditional off-line scheduling techniques cannot always work well due to their lack of robustness with respect to uncertainties coming from imprecise models or unexpected situations. On the other hand, on-line scheduling or rescheduling, which can improve the robustness by using the model and the latest measurements simultaneously, suffer from a high computational cost. This work proposes a hybrid scheduling framework, which combines the advantages of both off-line scheduling and on-line scheduling, to provide a balanced solution between robustness and computational cost. A novel concept of partially-dispatchable state is introduced. It can be treated as a constant in scheduling when the model works well. When the model does not work well, it is triggered as the variable to tune to improve the performance. Such an event-triggered structure can reduce the number of rescheduling and computational costs while achieving a reasonable performance and enhancing system robustness. Moreover, the choice of partially-dispatchable state also provides an extra design freedom in achieving green manufacturing. Simulation examples on a manufacturing system, of which consists a 100-kW solar photovoltaic system, a 10-machine flow shop production line, a 50-kWh energy storage system, a 100-kW gas turbine, and the grid for power supply, demonstrating the validity and applicability of this event-triggered hybrid scheduling (ETHS) framework. | ['Ying Tan', 'Wen Li', 'Zhean Shao'] | 2023-02-02 | null | null | null | null | ['manufacturing-simulation'] | ['knowledge-base'] | [ 1.39960676e-01 -1.73750803e-01 5.42844925e-03 6.84594437e-02
2.96682660e-02 -5.58591783e-01 4.60901618e-01 2.77186692e-01
1.59835711e-01 1.02166176e+00 -5.30491710e-01 -8.35763589e-02
-8.35426807e-01 -1.07797515e+00 -4.78441626e-01 -1.09043252e+00
1.19723871e-01 5.24006009e-01 5.17042838e-02 -1.76378563e-01
1.06116712e-01 6.11419559e-01 -1.75485790e+00 -3.62619132e-01
1.24980700e+00 8.97833407e-01 7.57596731e-01 1.95185989e-01
-2.34541763e-02 2.59871292e-03 -8.15545499e-01 5.32695711e-01
3.43962044e-01 -4.93325531e-01 7.65006710e-03 1.03309251e-01
-8.28917861e-01 9.83191654e-02 2.46959627e-01 9.62457716e-01
5.33491433e-01 2.98840404e-01 2.57145256e-01 -1.81296444e+00
-2.85555959e-01 3.05135995e-01 -4.17399049e-01 -3.58969122e-02
8.65519121e-02 2.62525529e-01 2.69043058e-01 -3.99275661e-01
4.08651829e-01 8.13394785e-01 -1.85844600e-02 2.21363097e-01
-1.07430005e+00 -3.64003837e-01 6.67542741e-02 1.48254007e-01
-1.28750491e+00 -8.60581174e-02 8.73917639e-01 -2.95869261e-01
1.01985931e+00 5.75212181e-01 9.00674224e-01 3.53881806e-01
8.50292921e-01 9.14243981e-02 1.39343679e+00 -5.31841516e-01
7.67703712e-01 -1.13126703e-01 -2.93942034e-01 -8.40464979e-02
5.87060392e-01 3.55847120e-01 -1.07668526e-01 -5.71004860e-02
5.15843451e-01 2.24557772e-01 -3.20974827e-01 -3.98680389e-01
-9.77414966e-01 3.74493152e-01 1.82035848e-01 5.84809601e-01
-6.04502499e-01 -1.28612876e-01 1.29883394e-01 3.12644452e-01
1.74055666e-01 5.87104559e-01 -5.14307976e-01 -1.64567903e-02
-7.77337313e-01 -1.78531200e-01 6.89509690e-01 9.36469495e-01
2.79056013e-01 5.39782822e-01 -2.10805655e-01 3.07508826e-01
1.16049089e-01 6.56428397e-01 4.20212805e-01 -7.06456661e-01
-3.22880484e-02 6.24534667e-01 6.39984906e-01 -3.14471126e-01
-7.06247807e-01 -5.92771053e-01 -9.38331962e-01 5.65239072e-01
6.80060452e-03 -3.37574512e-01 -8.99062872e-01 1.58210003e+00
4.84415472e-01 -2.46953994e-01 -1.08544447e-01 7.52328932e-01
-1.37415603e-01 9.44782376e-01 -3.31028312e-01 -1.20857990e+00
1.23945773e+00 -7.37132907e-01 -1.14479280e+00 2.95606524e-01
1.20639205e-01 -9.17027295e-01 2.82727093e-01 4.48834002e-01
-1.20635748e+00 -6.14256084e-01 -1.14484835e+00 7.32836246e-01
-6.28932893e-01 2.95750704e-02 1.20381732e-03 4.29345310e-01
-8.85571361e-01 5.91880023e-01 -8.10794413e-01 -2.27384672e-01
-4.49372172e-01 4.01559979e-01 1.19746774e-01 2.13135064e-01
-1.13264930e+00 1.30554473e+00 7.20832407e-01 5.23466289e-01
-3.34924400e-01 -5.23469031e-01 -4.84517843e-01 4.93614316e-01
7.91256309e-01 -9.32416916e-01 1.12627661e+00 -7.26827979e-01
-1.85399783e+00 -4.73449081e-01 -3.47231291e-02 -1.95305154e-01
5.05182683e-01 3.29502493e-01 -8.00695300e-01 -5.85843213e-02
-3.07314634e-01 -2.45533392e-01 6.76868796e-01 -1.21907854e+00
-6.61128998e-01 -2.54379869e-01 -1.89876214e-01 2.06374511e-01
4.01130654e-02 -4.11072254e-01 4.71975952e-01 -3.41962337e-01
-5.88437542e-02 -1.09730995e+00 -5.20894825e-01 -7.18207419e-01
-1.49578869e-01 -2.97196776e-01 9.64067459e-01 -2.15121821e-01
1.16810966e+00 -1.84539664e+00 9.17414576e-02 4.22179490e-01
-6.90371752e-01 3.19355160e-01 2.59777755e-01 1.24688590e+00
-3.61398071e-01 -9.30057690e-02 7.52424896e-02 2.09299982e-01
2.40823343e-01 5.99015772e-01 7.84845352e-02 1.40841648e-01
1.17621593e-01 4.69520062e-01 -9.84266400e-01 1.10371888e-01
8.35897744e-01 5.14398329e-02 3.45036745e-01 6.22750167e-03
-4.25225854e-01 5.59614956e-01 -4.38467115e-01 5.39161921e-01
7.18640089e-01 6.93065897e-02 3.84119421e-01 7.45043531e-02
-6.01523936e-01 -3.91404808e-01 -1.59380639e+00 1.31687403e+00
-8.67358446e-01 -9.39597562e-02 5.73280871e-01 -8.39383185e-01
1.04378200e+00 4.74844545e-01 8.01711619e-01 -9.00146306e-01
-9.76349786e-02 4.53176677e-01 1.62985459e-01 -2.46184573e-01
4.92369086e-01 -7.07014278e-02 1.99441493e-01 1.25241160e-01
-3.85039508e-01 -3.48910034e-01 6.58746600e-01 -3.41431648e-01
8.47070456e-01 2.41660967e-01 4.87330735e-01 -6.23861134e-01
7.46898890e-01 -1.40799776e-01 1.09775603e+00 1.11838490e-01
2.99612552e-01 -4.43989597e-02 1.63115084e-01 -3.26876715e-02
-8.01866114e-01 -7.36848116e-01 -1.00537404e-01 2.01851383e-01
5.38057685e-01 -7.58486539e-02 -1.78401545e-01 -2.68206477e-01
1.44900173e-01 1.40314901e+00 -2.21563373e-02 -1.74901918e-01
-3.88445526e-01 -6.89980984e-01 -7.12630033e-01 3.12310368e-01
1.77020460e-01 -7.47883916e-01 -1.08176172e+00 5.61045468e-01
4.04630184e-01 -7.99243152e-01 -2.54025012e-01 5.57745993e-01
-6.29490793e-01 -1.01818120e+00 -3.60867858e-01 -3.16570729e-01
1.09624040e+00 3.37671995e-01 6.84120953e-01 1.02970742e-01
-1.05489954e-01 1.57922477e-01 -1.61579698e-01 -6.56677723e-01
-5.36720812e-01 -2.20445499e-01 2.13719934e-01 -2.54096657e-01
-3.51306468e-01 -4.55242693e-01 -4.79520082e-01 5.92601299e-01
-1.10981822e+00 1.27910048e-01 4.86137867e-01 1.04425657e+00
7.36382067e-01 1.12647831e+00 1.37970650e+00 -6.22614205e-01
4.51335967e-01 -4.26315993e-01 -1.24363053e+00 5.91501713e-01
-1.10606062e+00 -1.89796716e-01 1.20833504e+00 -1.32937089e-01
-1.30137146e+00 2.46145129e-01 2.96068728e-01 -4.31609809e-01
-3.50866839e-02 5.29617071e-01 -3.26670408e-01 4.10346612e-02
-2.32376680e-01 2.22690821e-01 3.09285242e-02 -2.43067339e-01
7.36193582e-02 3.70121032e-01 1.94769338e-01 -3.69952828e-01
1.16400349e+00 -9.26816743e-03 8.32921922e-01 -3.09071064e-01
2.21921392e-02 -3.44539911e-01 -2.01745719e-01 -4.05373484e-01
3.64621311e-01 -5.04305005e-01 -1.00224423e+00 5.06019950e-01
-7.62336433e-01 -1.19927242e-01 -4.89126980e-01 5.34996450e-01
-7.07823038e-01 1.83340520e-01 -7.95956701e-02 -1.27879834e+00
-3.85604769e-01 -1.17100573e+00 6.08389080e-01 8.11965525e-01
2.14956701e-01 -7.88387716e-01 -2.38710731e-01 -8.26885700e-02
6.01557791e-01 8.97381485e-01 9.59805131e-01 -3.63726228e-01
-7.62254357e-01 -6.23003440e-03 4.44069117e-01 2.91913956e-01
6.08478248e-01 3.66327465e-01 -4.04025167e-01 -8.87655914e-01
2.09978759e-01 4.65764403e-01 -6.82146996e-02 3.49848241e-01
8.92492712e-01 -1.77507997e-01 -4.32763338e-01 -1.16878875e-01
2.04612136e+00 9.13633704e-01 3.16496789e-01 2.61801779e-01
-3.59999307e-04 4.13613856e-01 1.25304627e+00 7.13255644e-01
1.00884169e-01 8.31873477e-01 8.15633297e-01 1.77162271e-02
2.97807902e-01 1.27075002e-01 2.29046658e-01 8.38079572e-01
-6.58428445e-02 -5.18660426e-01 -5.46296835e-01 4.64870960e-01
-2.09088898e+00 -8.83634925e-01 -3.00470531e-01 2.74212694e+00
5.28623581e-01 2.26324886e-01 -2.56968379e-01 4.17226434e-01
8.90296638e-01 -2.06589520e-01 -5.40630579e-01 -8.85991991e-01
1.74154863e-01 -1.44901797e-02 6.49403751e-01 1.60171643e-01
-3.51313680e-01 -1.07894801e-01 4.19588518e+00 8.33563745e-01
-1.05903125e+00 -1.24412291e-01 1.75423905e-01 -1.50068074e-01
-3.33163977e-01 2.87564188e-01 -4.71170872e-01 1.24118662e+00
1.23906147e+00 -7.64564276e-01 7.91793048e-01 6.93662822e-01
8.80397439e-01 -7.64711738e-01 -1.08911097e+00 5.45705080e-01
-3.21231693e-01 -1.14145339e+00 -4.65760738e-01 2.15163350e-01
1.16094112e+00 -4.29498553e-01 -5.13599932e-01 4.34088670e-02
-4.37643826e-02 -5.07136226e-01 6.69903815e-01 7.87539959e-01
2.24306926e-01 -1.28759754e+00 1.01647794e+00 7.96308339e-01
-1.29814899e+00 -4.40174431e-01 1.07057756e-02 -1.40432492e-01
6.79651022e-01 1.06849122e+00 -7.13687837e-01 1.61153054e+00
5.59748769e-01 -1.15390621e-01 9.30754393e-02 1.15258181e+00
-1.80663377e-01 -9.87023190e-02 -5.32853663e-01 -3.35355774e-02
-2.13174090e-01 -6.09114647e-01 7.06867576e-01 4.56348449e-01
6.73865855e-01 1.07534595e-01 5.21690786e-01 7.57280588e-01
5.38780630e-01 -2.89510310e-01 -5.72526515e-01 4.17528674e-02
8.76367509e-01 1.49937880e+00 -9.01236713e-01 -1.78269506e-01
-1.13709450e-01 4.03821081e-01 -6.50805712e-01 2.81131297e-01
-8.79149020e-01 -5.03896177e-01 3.40297699e-01 -3.58961746e-02
2.02112511e-01 -6.20159686e-01 -4.01038110e-01 -5.22243619e-01
2.20055118e-01 -2.37570539e-01 6.74472302e-02 -9.10768449e-01
-1.28675663e+00 1.55061737e-01 1.35439560e-01 -1.47656572e+00
-7.03386962e-01 -1.02985881e-01 -6.84327662e-01 1.06178010e+00
-1.89697719e+00 -7.56702781e-01 -2.24192470e-01 2.76372880e-01
7.69604564e-01 2.28197634e-01 5.93665004e-01 1.38985381e-01
-8.24434221e-01 -2.49585241e-01 7.50884175e-01 -9.08442080e-01
4.32603866e-01 -1.34107351e+00 -2.71752089e-01 1.08122551e+00
-5.99833608e-01 3.24858636e-01 8.46598268e-01 -7.39674628e-01
-1.90650511e+00 -9.63916421e-01 7.52205491e-01 3.65455389e-01
5.75338304e-01 -2.13126671e-02 -7.55863309e-01 1.25428632e-01
6.97084963e-01 -2.14801416e-01 1.51316240e-01 -4.76972669e-01
7.47655332e-01 -4.72793907e-01 -1.43058622e+00 4.31186765e-01
6.53620422e-01 1.33912370e-01 -4.04948384e-01 4.81884062e-01
3.55153143e-01 -6.02826178e-01 -1.17793965e+00 6.82061851e-01
3.00243236e-02 -5.63778281e-01 4.04307425e-01 2.35505566e-01
-1.85325548e-01 -7.76962399e-01 3.25165927e-01 -2.09454560e+00
-3.42950195e-01 -9.61503565e-01 -2.22660884e-01 1.62689185e+00
1.84874237e-02 -1.29036558e+00 -6.30481318e-02 7.08042741e-01
-4.87267643e-01 -7.59877563e-01 -1.40958762e+00 -1.39030576e+00
-5.10290444e-01 3.29355419e-01 1.05756009e+00 8.19504797e-01
1.05350822e-01 3.24687026e-02 -5.28082065e-03 4.75679189e-01
5.25540590e-01 5.62024236e-01 4.99219209e-01 -1.24352801e+00
-1.50919080e-01 -2.34707654e-01 6.89584985e-02 3.19553651e-02
-1.58880293e-01 -3.69896859e-01 1.46425113e-01 -2.03620291e+00
-3.20334911e-01 -4.16567624e-01 -5.38658440e-01 4.75406438e-01
1.90836772e-01 -4.67619181e-01 5.50217390e-01 2.57801451e-02
7.75764734e-02 8.65801334e-01 1.24760485e+00 1.32557958e-01
-3.03927541e-01 3.11559647e-01 1.35130242e-01 3.91644061e-01
9.32719171e-01 -3.30131173e-01 -9.14768875e-01 -4.50677536e-02
1.44654959e-01 7.23167181e-01 8.46185605e-04 -9.57100868e-01
4.02036369e-01 -7.06845999e-01 1.49493411e-01 -7.73674071e-01
2.18398556e-01 -1.71144271e+00 1.27137983e+00 8.56078148e-01
3.34878057e-01 3.80540699e-01 1.16730347e-01 7.52804756e-01
-6.57654628e-02 -3.20306957e-01 5.75406611e-01 4.23576757e-02
-5.77577710e-01 -2.14380130e-01 -1.44053087e-01 -8.00233662e-01
1.82855630e+00 -3.84248197e-01 -7.33354390e-01 7.83617198e-02
-6.57054722e-01 8.87359917e-01 5.50729871e-01 6.01566255e-01
2.77026426e-02 -1.09009707e+00 -3.35982203e-01 8.09919611e-02
-4.23030138e-01 -7.49518350e-02 4.51480985e-01 9.03868377e-01
5.67274541e-03 4.64108735e-01 -4.58043784e-01 -5.51122963e-01
-9.55846846e-01 8.99575889e-01 1.21004805e-01 -3.52588058e-01
-2.28302926e-01 -2.75494576e-01 -4.08686668e-01 1.53563440e-01
-5.07866144e-01 -4.45370525e-01 9.92611796e-02 3.35859060e-01
8.71426165e-02 9.11845922e-01 6.52956367e-01 -4.21597734e-02
-4.18321013e-01 3.52816612e-01 7.04471946e-01 2.33294740e-01
1.24832320e+00 -4.49621707e-01 -2.69627213e-01 5.75660348e-01
3.65188807e-01 -3.93380970e-03 -1.01740766e+00 2.69058049e-01
-1.08433634e-01 -4.77361768e-01 1.22328252e-01 -1.21470690e+00
-1.16812050e+00 -7.08202198e-02 3.56171250e-01 9.82892036e-01
1.56882453e+00 -7.93808997e-01 5.48671246e-01 -4.35865857e-02
1.00043392e+00 -1.66423035e+00 -5.96479535e-01 2.22873181e-01
9.56994593e-01 -5.87621331e-01 4.72534373e-02 -4.25176173e-01
-3.17969710e-01 1.35723817e+00 5.12564778e-01 -7.42290691e-02
4.13834691e-01 4.51555341e-01 -4.75635082e-01 3.08311552e-01
-9.47529078e-01 -1.61691397e-01 -2.11467981e-01 2.45122060e-01
-1.42435819e-01 3.65909189e-01 -1.07703829e+00 2.70876944e-01
1.81716144e-01 4.58602756e-01 7.06636250e-01 1.21032941e+00
-1.82288677e-01 -1.26954341e+00 -5.77619255e-01 2.61038333e-01
9.78643373e-02 5.97435117e-01 4.59309489e-01 9.53429759e-01
4.44190383e-01 1.29820120e+00 2.99386848e-02 7.41261393e-02
7.37710178e-01 8.12854692e-02 2.42182851e-01 -6.31871164e-01
-7.61268795e-01 2.36015484e-01 2.95915067e-01 -2.69420773e-01
-3.89472067e-01 -6.35650992e-01 -1.61898780e+00 -9.47008133e-02
-8.52444708e-01 4.34413642e-01 1.15498793e+00 9.65861738e-01
5.70307255e-01 1.10570300e+00 1.35091114e+00 -5.97753823e-01
-9.75010216e-01 -5.79988539e-01 -8.33335698e-01 -2.13313773e-01
-9.40758809e-02 -8.64780605e-01 -3.60202581e-01 -4.21069354e-01] | [5.674221992492676, 2.5276670455932617] |
9c330423-544e-461c-81e0-2bb717928493 | optimising-the-input-image-to-improve-visual | 1903.11029 | null | http://arxiv.org/abs/1903.11029v1 | http://arxiv.org/pdf/1903.11029v1.pdf | Optimising the Input Image to Improve Visual Relationship Detection | Visual Relationship Detection is defined as, given an image composed of a
subject and an object, the correct relation is predicted. To improve the visual
part of this difficult problem, ten preprocessing methods were tested to
determine whether the widely used Union method yields the optimal results.
Therefore, focusing solely on predicate prediction, no object detection and
linguistic knowledge were used to prevent them from affecting the comparison
results. Once fine-tuned, the Visual Geometry Group models were evaluated using
Recall@1, per-predicate recall, activation maximisations, class activation
maps, and error analysis. From this research it was found that using
preprocessing methods such as the Union-Without-Background-and-with-Binary-mask
(Union-WB-and-B) method yields significantly better results than the widely
used Union method since, as designed, it enables the Convolutional Neural
Network to also identify the subject and object in the convolutional layers
instead of solely in the fully-connected layers. | ['Noel Mizzi', 'Adrian Muscat'] | 2019-03-26 | null | null | null | null | ['visual-relationship-detection'] | ['computer-vision'] | [ 3.11403304e-01 5.11651576e-01 2.34332159e-02 -3.50178331e-01
-1.36068448e-01 -5.26952207e-01 7.51027226e-01 5.72185159e-01
-4.08246011e-01 5.22696912e-01 -2.82319725e-01 -4.44773972e-01
-1.44438714e-01 -9.22756791e-01 -8.10160220e-01 -4.40862447e-01
8.80530924e-02 2.52106339e-01 6.06178343e-01 1.64988145e-01
4.15278077e-01 6.47457004e-01 -1.92772186e+00 6.58836782e-01
4.62721497e-01 1.18288660e+00 2.24832803e-01 3.41804713e-01
-1.27752438e-01 8.06638241e-01 -6.58673644e-01 -5.67450583e-01
2.73422927e-01 -3.64966780e-01 -9.15574610e-01 -3.41647901e-02
6.73370540e-01 -4.81055751e-02 2.25103483e-01 9.37503099e-01
2.65713960e-01 2.11520329e-01 8.15878034e-01 -8.92155528e-01
-7.21715987e-01 2.94204384e-01 -5.77011585e-01 3.30809712e-01
5.83390296e-01 1.37807012e-01 1.13044751e+00 -8.00175846e-01
8.04224968e-01 1.36183810e+00 4.44313079e-01 -5.37261851e-02
-1.50214314e+00 -5.61116934e-01 3.41354251e-01 2.02604160e-01
-1.59423482e+00 -6.25607520e-02 5.99275291e-01 -5.26822567e-01
1.39081240e+00 4.27050650e-01 7.06716061e-01 5.43156743e-01
1.09008931e-01 4.96785879e-01 1.17993629e+00 -8.48859668e-01
8.61782879e-02 6.32160902e-01 1.47644445e-01 7.41209507e-01
6.32933140e-01 1.97506294e-01 -3.39871079e-01 2.27854982e-01
5.76980352e-01 -5.31996191e-01 -2.78791219e-01 -3.16235870e-01
-9.80239809e-01 5.78208566e-01 6.48936391e-01 5.73165119e-01
-3.46005619e-01 -1.31254673e-01 2.31918499e-01 8.35903659e-02
3.44984800e-01 6.21324956e-01 -3.64832938e-01 4.30884868e-01
-1.01691806e+00 1.42892376e-01 6.00261450e-01 6.66593850e-01
8.37149918e-01 -1.93478987e-01 -4.59291399e-01 3.88804168e-01
3.26536298e-01 1.37972817e-01 1.64663404e-01 -5.62843800e-01
3.65261286e-01 1.07433569e+00 4.43365686e-02 -1.46747804e+00
-4.56428021e-01 -4.79887635e-01 -2.66812742e-01 5.46836436e-01
6.33226871e-01 1.87571183e-01 -1.18180883e+00 1.49164724e+00
3.98061782e-01 -2.37784773e-01 7.77419336e-06 1.06512487e+00
1.17994487e+00 3.42890710e-01 5.22879779e-01 -2.67942011e-01
1.57160830e+00 -6.01735592e-01 -6.66481555e-01 -5.53342462e-01
6.49382889e-01 -8.93707991e-01 1.15307581e+00 2.43190989e-01
-8.59093487e-01 -9.28770363e-01 -1.36233199e+00 -9.11118537e-02
-8.61633837e-01 2.82698691e-01 6.05771065e-01 6.93825126e-01
-9.85603988e-01 3.80818844e-01 -4.33997840e-01 -5.93777597e-01
4.14774537e-01 3.81225437e-01 -4.22614127e-01 1.98850259e-01
-9.85067010e-01 1.48380458e+00 8.44045758e-01 3.24221589e-02
-5.50871193e-01 -4.16860163e-01 -8.21230352e-01 1.17426425e-01
2.71980405e-01 -4.08135533e-01 8.48402262e-01 -1.36211300e+00
-9.09428120e-01 1.28302050e+00 -1.07755363e-01 -4.03929114e-01
3.96895170e-01 -1.66138232e-01 -1.89794108e-01 2.47199491e-01
8.35996680e-03 8.87257516e-01 6.74059331e-01 -1.58373892e+00
-7.96174288e-01 -2.23096043e-01 2.51812905e-01 2.99769074e-01
1.36463389e-01 1.62403822e-01 -5.01121759e-01 -4.54943120e-01
3.34442466e-01 -6.15877867e-01 1.81958944e-01 -7.02479184e-02
-3.83008540e-01 -2.73394167e-01 8.07421207e-01 -7.19604790e-01
1.05582964e+00 -2.38862419e+00 -3.82394135e-01 5.02545655e-01
7.47482032e-02 2.86346227e-01 2.01740608e-01 -6.80021942e-02
-4.92401093e-01 2.12332696e-01 1.52441218e-01 4.57648821e-02
-2.33719751e-01 2.06883997e-02 6.07484803e-02 5.13047159e-01
3.60808253e-01 6.76078260e-01 -6.41543210e-01 -8.95018041e-01
5.24161160e-01 4.45781797e-01 -2.73739129e-01 3.32804918e-02
-2.86688626e-01 1.04525521e-01 -9.44590271e-02 4.85351413e-01
6.31993592e-01 -2.61599183e-01 3.15225691e-01 -5.47531903e-01
-1.55980393e-01 7.84802586e-02 -1.15882301e+00 9.17356312e-01
-1.95429340e-01 9.24376905e-01 -2.35992506e-01 -8.48108888e-01
1.12970722e+00 3.04382145e-01 1.01626031e-01 -1.18477750e+00
2.75606215e-01 1.80404231e-01 2.57556349e-01 -5.93334258e-01
3.28350693e-01 -8.00347999e-02 3.45856041e-01 -1.19034596e-01
-7.61932880e-02 3.28836471e-01 3.11372399e-01 -9.72569436e-02
6.26506448e-01 4.39621359e-01 5.52661598e-01 -4.04227644e-01
6.38002515e-01 2.86559731e-01 3.27202648e-01 6.56439722e-01
-1.27071917e-01 5.08520424e-01 7.42088139e-01 -2.08170146e-01
-7.10928857e-01 -9.89842653e-01 -1.37645692e-01 9.66499388e-01
4.29505914e-01 -1.83933258e-01 -6.56897247e-01 -7.21760869e-01
-2.26799231e-02 1.00719833e+00 -8.00497830e-01 -5.62282577e-02
-4.62933838e-01 -4.95268583e-01 1.94685176e-01 4.48648989e-01
5.44990361e-01 -1.22774291e+00 -1.17112565e+00 -1.03792250e-01
7.65852928e-02 -1.06590426e+00 1.20058671e-01 3.08765262e-01
-7.12081969e-01 -1.39663553e+00 -4.38481748e-01 -1.00115931e+00
7.92914271e-01 -8.62067565e-02 1.05646980e+00 2.47790873e-01
-2.75274307e-01 9.86491293e-02 -3.15946281e-01 -4.38802838e-01
-2.07988456e-01 -2.79180110e-01 -5.58049142e-01 6.70064986e-02
6.40002668e-01 -6.29584640e-02 -5.19414663e-01 1.15187682e-01
-6.13678455e-01 1.32804692e-01 9.22806919e-01 3.31919998e-01
7.19403267e-01 8.99682045e-02 5.98987564e-02 -7.15994716e-01
5.09637058e-01 -9.49035361e-02 -5.48585176e-01 6.63080096e-01
-7.13650048e-01 5.63019440e-02 2.07182586e-01 -5.56795716e-01
-9.32349384e-01 1.71676382e-01 2.36721322e-01 -3.01262289e-01
-5.51598549e-01 2.44571522e-01 -3.41227263e-01 -9.03702974e-02
7.91171312e-01 -2.70246640e-02 -1.96776107e-01 -1.45434648e-01
3.08053970e-01 3.58596563e-01 3.76096487e-01 -1.93820819e-01
3.98862988e-01 2.93627381e-01 6.14694990e-02 -6.10820055e-01
-7.70864904e-01 -3.54066610e-01 -7.68158495e-01 -5.64371407e-01
1.26415944e+00 -6.59158289e-01 -6.70522571e-01 -2.34493688e-02
-1.33603191e+00 -8.82230848e-02 -1.89500213e-01 4.97734070e-01
-2.07264289e-01 5.49033098e-02 4.86908853e-02 -9.91530955e-01
-8.05701613e-02 -1.14007163e+00 7.07962751e-01 4.51497376e-01
-4.82280701e-01 -7.80536771e-01 -5.72069168e-01 1.84940204e-01
1.85906976e-01 4.48645949e-01 1.26848316e+00 -1.15755141e+00
-4.49670851e-01 -7.56468028e-02 -6.98876798e-01 3.12984288e-01
1.42922580e-01 1.98322296e-01 -1.07361889e+00 4.55269404e-02
-3.20098877e-01 1.13696471e-01 5.93806803e-01 3.25870544e-01
8.70612323e-01 -2.45422527e-01 -5.88743210e-01 3.51862431e-01
1.62012708e+00 4.67104435e-01 6.65141165e-01 5.77091098e-01
6.05591953e-01 8.79537344e-01 6.62991822e-01 -1.06378971e-02
2.14753643e-01 6.69794798e-01 5.74033499e-01 -3.31638098e-01
-3.33548307e-01 1.84745211e-02 -3.03797256e-02 -2.58201867e-01
-2.67050296e-01 -2.47086853e-01 -9.09561932e-01 5.81838191e-01
-1.53200698e+00 -8.71671855e-01 -3.53415936e-01 2.35252118e+00
6.35338247e-01 6.09128058e-01 2.06668963e-04 3.17398816e-01
8.41717362e-01 -2.46563498e-02 -3.66421044e-02 -6.77401721e-01
-1.57110870e-01 3.55660975e-01 5.10323346e-01 4.66285855e-01
-1.25580788e+00 9.45847273e-01 6.24495125e+00 7.39208281e-01
-1.15820217e+00 -7.16039687e-02 7.18443394e-01 1.15053684e-01
3.80229354e-02 1.08354218e-01 -9.22693193e-01 1.49105951e-01
4.71234709e-01 3.65122944e-01 9.01368409e-02 7.56564796e-01
1.94953412e-01 -7.77454734e-01 -1.15389204e+00 8.39816093e-01
8.11328366e-02 -9.94057536e-01 1.69907957e-01 -1.36423215e-01
2.99536496e-01 -6.95805907e-01 -8.46185908e-02 2.37426415e-01
-2.02425539e-01 -1.07278168e+00 1.18015480e+00 5.38135648e-01
6.28613532e-01 -6.08114779e-01 8.95814240e-01 1.01119868e-01
-1.10031605e+00 8.56705010e-02 3.04822829e-02 -1.95015520e-01
-2.03198254e-01 3.82013083e-01 -1.02809048e+00 5.31581879e-01
9.58343506e-01 7.35349730e-02 -9.24428999e-01 1.08194149e+00
-3.28782052e-01 4.63578790e-01 -3.96466672e-01 -2.45558694e-01
6.31431490e-02 1.30782977e-01 4.20783281e-01 1.52189529e+00
-8.77526775e-02 -1.24252439e-01 1.13156185e-01 1.04424143e+00
4.32552248e-01 3.56079489e-01 -4.48361933e-01 2.18234211e-01
2.19732836e-01 1.17996013e+00 -1.44979358e+00 -1.79692447e-01
-4.25035238e-01 6.49825513e-01 2.57231385e-01 5.17463505e-01
-9.87483799e-01 -3.54237348e-01 1.82691008e-01 5.76354325e-01
6.00689650e-01 5.55625372e-02 -7.53951132e-01 -1.35978013e-01
1.82302788e-01 -5.90488434e-01 4.13213909e-01 -1.04702592e+00
-8.40466380e-01 4.85814631e-01 4.36583281e-01 -7.81998932e-01
1.88945923e-02 -9.43100393e-01 -3.36671084e-01 1.15870512e+00
-1.13542736e+00 -1.29330158e+00 -3.11868995e-01 3.03626567e-01
1.59348607e-01 1.33027673e-01 6.32668793e-01 1.41184896e-01
-3.44970316e-01 5.18620610e-01 -7.00559914e-01 1.60422519e-01
4.41129446e-01 -1.11339581e+00 -2.80439168e-01 1.04568219e+00
1.81841463e-01 6.55608892e-01 1.11610949e+00 -8.65056276e-01
-5.26140153e-01 -6.99094594e-01 9.83148634e-01 -2.79283971e-01
2.93439329e-01 -2.23375246e-01 -8.84486020e-01 4.67823654e-01
2.88871080e-01 -4.45810296e-02 3.66235793e-01 -1.35431183e-03
-2.64448404e-01 -5.82087003e-02 -1.23300779e+00 5.09641290e-01
7.34204173e-01 -3.77149493e-01 -8.36067080e-01 2.15460613e-01
4.74742979e-01 -4.85163063e-01 -6.78957164e-01 6.67481899e-01
3.86147857e-01 -1.22287655e+00 9.42829549e-01 -4.22954053e-01
2.35390991e-01 -5.31239808e-01 -1.78325728e-01 -6.13105834e-01
-3.90148342e-01 1.30630478e-01 -4.03257012e-02 1.26976633e+00
8.95785272e-01 -4.80772972e-01 6.67003989e-01 4.26066458e-01
3.27365883e-02 -8.17398787e-01 -7.73214698e-01 -6.12697124e-01
-3.54951322e-01 -2.83306092e-01 3.51653695e-01 7.09420264e-01
-3.55789632e-01 2.59913027e-01 2.73383439e-01 3.28983337e-01
1.65556610e-01 1.59584619e-02 4.43757772e-01 -1.24875784e+00
-1.51430085e-01 -6.14365399e-01 -7.17918694e-01 -5.19349039e-01
-8.29044208e-02 -7.56523013e-01 1.04075950e-03 -1.84733605e+00
8.39784648e-03 -1.83257416e-01 -1.97289407e-01 7.12318301e-01
-3.49485040e-01 3.46241266e-01 2.20830381e-01 7.67405480e-02
-3.64365011e-01 -1.53386995e-01 1.24734032e+00 -6.35379413e-03
-1.70075342e-01 2.83151306e-02 -6.22331321e-01 6.47763908e-01
7.15747476e-01 -3.58652949e-01 -4.78982776e-01 -9.15448964e-02
3.57715636e-01 -3.53265136e-01 7.32169092e-01 -1.07834947e+00
7.80059025e-02 -9.63805169e-02 8.00482869e-01 -7.68838525e-01
2.57114381e-01 -1.00906336e+00 2.17422768e-01 4.57951725e-01
-2.33558774e-01 8.64115730e-02 5.64754069e-01 2.26616859e-01
-7.62787983e-02 -4.10310209e-01 7.35287189e-01 -1.58411145e-01
-1.19320023e+00 -4.98129398e-01 -1.38816372e-01 -1.81013599e-01
1.40506113e+00 -8.35394740e-01 -3.81707877e-01 8.98416154e-03
-8.93587947e-01 -1.17456108e-01 3.53108823e-01 3.09156656e-01
4.58559841e-01 -9.30232406e-01 -3.18462878e-01 1.03120133e-01
2.21810758e-01 -1.67753175e-02 2.07455754e-02 8.17856073e-01
-8.17614496e-01 4.76555496e-01 -1.71946406e-01 -6.08940065e-01
-1.56925380e+00 8.67198229e-01 5.17280936e-01 -1.96640581e-01
-5.54907739e-01 9.12648618e-01 2.74072498e-01 6.63406700e-02
4.14537787e-01 -4.11953479e-01 -6.40429020e-01 2.42566809e-01
2.31892928e-01 8.54733773e-03 2.75902599e-01 -8.24687719e-01
-5.97766459e-01 3.85948300e-01 1.83080822e-01 -6.64836243e-02
9.13527906e-01 9.18760672e-02 -1.30507685e-02 4.86474246e-01
9.52222109e-01 -4.16521840e-02 -9.30813551e-01 1.46306843e-01
2.45262519e-01 -3.75319868e-01 1.02038980e-01 -1.04617679e+00
-8.52151811e-01 6.12039387e-01 9.10233259e-01 5.28921783e-01
1.09467602e+00 7.59116039e-02 -2.98593253e-01 2.78592721e-04
2.17511933e-02 -1.08186007e+00 -1.78006351e-01 2.43958056e-01
8.19862664e-01 -1.23412132e+00 3.26114208e-01 -6.57489061e-01
-6.76167667e-01 9.37270820e-01 9.47616160e-01 1.15504541e-01
5.73177159e-01 1.97750628e-01 -3.19714062e-02 -5.12030780e-01
-3.43739212e-01 -6.66562855e-01 8.49597573e-01 7.36655772e-01
7.39647985e-01 1.43832562e-03 -5.04322469e-01 2.86855817e-01
-2.30832368e-01 -2.53196448e-01 -1.29648134e-01 8.57882023e-01
-4.78756398e-01 -6.01072550e-01 -5.56961060e-01 6.30866885e-01
-5.25484025e-01 -5.94759695e-02 -7.19759226e-01 1.46166670e+00
7.15494514e-01 9.36472178e-01 4.27014619e-01 -3.88824314e-01
4.59006190e-01 1.08702198e-01 5.44724047e-01 -5.98837912e-01
-9.18203533e-01 -1.74507007e-01 2.71685064e-01 -4.11166072e-01
-6.80846512e-01 -6.40884221e-01 -1.49446750e+00 5.57905249e-02
-6.04233503e-01 -1.25950560e-01 6.22482240e-01 1.17867374e+00
4.12842445e-02 7.22590446e-01 -8.15256611e-02 -5.49471438e-01
-2.71882936e-02 -1.14024997e+00 -2.46433243e-01 4.48627293e-01
-7.11486265e-02 -9.73708570e-01 -2.48542473e-01 1.32968843e-01] | [10.244171142578125, 1.664555549621582] |
e7aa7a80-80ba-40ac-bdd6-c4add9cfbd1e | how-many-labeled-license-plates-are-needed | 1808.08410 | null | http://arxiv.org/abs/1808.08410v1 | http://arxiv.org/pdf/1808.08410v1.pdf | How many labeled license plates are needed? | Training a good deep learning model often requires a lot of annotated data.
As a large amount of labeled data is typically difficult to collect and even
more difficult to annotate, data augmentation and data generation are widely
used in the process of training deep neural networks. However, there is no
clear common understanding on how much labeled data is needed to get
satisfactory performance. In this paper, we try to address such a question
using vehicle license plate character recognition as an example application. We
apply computer graphic scripts and Generative Adversarial Networks to generate
and augment a large number of annotated, synthesized license plate images with
realistic colors, fonts, and character composition from a small number of real,
manually labeled license plate images. Generated and augmented data are mixed
and used as training data for the license plate recognition network modified
from DenseNet. The experimental results show that the model trained from the
generated mixed training data has good generalization ability, and the proposed
approach achieves a new state-of-the-art accuracy on Dataset-1 and AOLP, even
with a very limited number of original real license plates. In addition, the
accuracy improvement caused by data generation becomes more significant when
the number of labeled images is reduced. Data augmentation also plays a more
significant role when the number of labeled images is increased. | ['Shugong Xu', 'Changhao Wu', 'Shunqing Zhang', 'Guocong Song'] | 2018-08-25 | null | null | null | null | ['license-plate-recognition'] | ['computer-vision'] | [ 6.00174777e-02 -1.79556578e-01 2.09918544e-01 -4.84053791e-01
-5.66977620e-01 -7.52488613e-01 5.01122475e-01 -5.44772029e-01
-4.61686432e-01 9.02626336e-01 -4.06298518e-01 -8.62401575e-02
5.98603666e-01 -9.48039114e-01 -1.00604784e+00 -7.09831715e-01
5.20548761e-01 6.77822709e-01 2.00437561e-01 -2.73931414e-01
9.20816511e-02 5.37783504e-01 -1.36282837e+00 1.65430948e-01
1.29079199e+00 8.01229894e-01 2.25676328e-01 2.70445585e-01
-3.96827877e-01 9.59543228e-01 -8.60025942e-01 -7.21654177e-01
6.69296443e-01 -3.88914645e-01 -3.91408771e-01 6.93113387e-01
3.69588703e-01 -3.92314166e-01 -2.98463762e-01 1.36916316e+00
2.18056440e-01 1.18330844e-01 6.21818483e-01 -1.36433470e+00
-9.83880043e-01 3.97731394e-01 -3.49926591e-01 -1.07213177e-01
-4.10838962e-01 4.74186093e-01 2.54692823e-01 -1.02681100e+00
4.65017408e-01 7.06900775e-01 6.65872633e-01 7.86929250e-01
-9.78809595e-01 -8.98777664e-01 -2.09414348e-01 8.88085458e-03
-1.44204319e+00 -2.57571816e-01 8.21146548e-01 -5.31763613e-01
2.18034834e-01 2.01122612e-02 6.26308680e-01 9.28301036e-01
-2.57966578e-01 6.83077276e-01 1.18421495e+00 -4.24169183e-01
1.48647442e-01 7.32496440e-01 1.11696526e-01 5.57937145e-01
3.72917920e-01 -1.07018657e-01 2.65484422e-01 3.27540308e-01
1.03856349e+00 2.22314447e-01 -2.83759743e-01 -1.42159417e-01
-8.71739984e-01 8.34483206e-01 3.19956571e-01 2.06364617e-01
-2.30024457e-01 -1.88234806e-01 2.85674810e-01 7.87183866e-02
3.65590721e-01 6.78029597e-01 -1.65121228e-01 2.75429301e-02
-9.50888574e-01 1.52055025e-01 5.92894018e-01 1.04173160e+00
8.90380025e-01 7.05114007e-01 2.37896904e-01 1.29642951e+00
1.05201259e-01 7.19274163e-01 7.58650601e-01 -4.45003480e-01
8.00489783e-01 8.12279582e-01 1.81862950e-01 -9.31595564e-01
-9.60981920e-02 -4.09459323e-01 -9.48850393e-01 6.40299380e-01
5.67359507e-01 -3.38257343e-01 -1.27916551e+00 1.43943906e+00
-2.74116039e-01 1.49755627e-01 3.60927582e-01 1.03225601e+00
7.89517343e-01 1.11115813e+00 -1.35438535e-02 8.77500027e-02
1.12290955e+00 -1.28211129e+00 -7.64201581e-01 -6.09225154e-01
4.06382084e-01 -7.62992561e-01 1.14878261e+00 1.64757535e-01
-8.51806700e-01 -9.11545217e-01 -1.16632760e+00 1.39848709e-01
-4.96767998e-01 5.29050827e-01 4.04264450e-01 7.40434408e-01
-8.26216757e-01 2.75610626e-01 -4.29615349e-01 5.35773709e-02
6.30555391e-01 4.73018199e-01 -5.96231818e-01 -4.54968691e-01
-1.15802193e+00 8.70895803e-01 6.04279637e-01 2.23987147e-01
-8.42260659e-01 -3.95694494e-01 -7.12887526e-01 -4.66932468e-02
1.27675161e-01 -2.14603655e-02 9.08296585e-01 -1.45457864e+00
-1.32728934e+00 8.67243052e-01 6.37229860e-01 -2.73838371e-01
6.74396157e-01 7.85176978e-02 -8.23232114e-01 -1.63100541e-01
-1.49621367e-01 9.55383420e-01 7.68532336e-01 -1.45843196e+00
-5.14211655e-01 -2.87560821e-01 -1.09740831e-01 1.56269491e-01
-4.59757477e-01 -1.47962779e-01 -6.37369215e-01 -6.56067371e-01
-2.88332999e-01 -1.15208650e+00 -1.71251208e-01 -1.57480076e-01
-4.91500795e-01 1.37629077e-01 9.17174459e-01 -7.41045296e-01
5.14999866e-01 -2.16524935e+00 -4.33250695e-01 2.36564681e-01
8.39647576e-02 9.95901942e-01 -3.09899688e-01 -3.93697172e-02
-1.80732876e-01 3.57870221e-01 -4.60989684e-01 -1.72261149e-01
-1.42585248e-01 8.09494779e-02 -1.48791924e-01 2.06400335e-01
5.31736016e-01 7.29535162e-01 -4.54413325e-01 -2.42525443e-01
2.63721853e-01 3.00854266e-01 -2.61004031e-01 3.86068225e-01
-1.39350086e-01 3.84961963e-01 -3.55121374e-01 6.11565113e-01
1.07777226e+00 -1.01483166e-01 -2.40013182e-01 -1.20173454e-01
3.61698517e-03 -4.19609070e-01 -1.05189443e+00 9.92306411e-01
-3.29110652e-01 1.09163988e+00 -2.79773533e-01 -8.71064663e-01
1.43120122e+00 8.25103149e-02 1.40263394e-01 -6.39942586e-01
3.37791383e-01 2.35733137e-01 2.13194892e-01 -5.86833060e-01
5.66180468e-01 -2.25074902e-01 -1.42918631e-01 2.52979547e-01
-9.95497778e-02 -2.59452432e-01 4.15044576e-01 -1.87274471e-01
7.75289834e-01 -2.07963914e-01 -1.83339104e-01 9.62472185e-02
6.31389499e-01 4.82030153e-01 6.25367165e-01 2.93157548e-01
1.00379381e-02 6.98428810e-01 3.23395610e-01 -6.17318571e-01
-1.76223397e+00 -6.23975456e-01 -3.65272462e-02 5.73775172e-01
8.47432613e-02 4.39625949e-01 -1.04709733e+00 -6.22818172e-01
-3.05143148e-01 7.38567173e-01 -4.89333868e-01 -1.29296184e-01
-5.74489594e-01 -9.92581666e-01 8.38642240e-01 6.31093740e-01
1.18212605e+00 -1.30206442e+00 3.25164795e-01 1.32332012e-01
-1.27072502e-02 -1.41160750e+00 -2.98641175e-01 -1.50029883e-01
-7.86199629e-01 -9.86451328e-01 -9.37799454e-01 -1.26929903e+00
1.25433445e+00 8.69190767e-02 8.67886364e-01 3.38466167e-02
1.52486071e-01 -3.25248271e-01 -3.62430751e-01 -5.44007003e-01
-1.06071067e+00 -1.25263706e-01 -2.53974833e-02 3.54869545e-01
3.30491364e-01 -1.66802466e-01 -1.87527239e-01 5.08963943e-01
-1.30647087e+00 3.15850407e-01 8.49019110e-01 7.71998048e-01
2.76421100e-01 2.23424777e-01 6.02579534e-01 -9.78745341e-01
6.67591512e-01 -2.90144980e-01 -1.00232244e+00 1.02986023e-01
-2.68162668e-01 -1.09075621e-01 1.26448333e+00 -6.37727261e-01
-1.16514361e+00 2.19377160e-01 -3.92115384e-01 -6.04800582e-01
-4.11862284e-01 3.16386849e-01 -3.92561078e-01 -1.92905992e-01
7.14039981e-01 4.48121786e-01 1.66601792e-01 -3.26527387e-01
2.60322630e-01 7.41171181e-01 7.33796597e-01 -1.67481363e-01
1.18569756e+00 1.74410790e-01 -4.29738969e-01 -6.58799887e-01
-2.76716918e-01 -1.60521641e-02 -5.65083563e-01 -3.69365841e-01
8.48222256e-01 -7.66146421e-01 -3.21891099e-01 1.06678319e+00
-1.00112891e+00 -4.94579524e-01 -2.33923852e-01 4.75579381e-01
-1.35651186e-01 1.65525556e-01 -5.17703533e-01 -4.75763708e-01
-1.33277699e-01 -1.34786832e+00 5.03825247e-01 4.33777571e-01
4.88263756e-01 -8.28784168e-01 -1.14689678e-01 6.23925269e-01
2.85740286e-01 3.02620202e-01 8.27551842e-01 -1.11814439e+00
-8.62762570e-01 -8.24097991e-01 -3.27098846e-01 1.10902262e+00
1.57425523e-01 9.91461352e-02 -8.21907341e-01 1.12990946e-01
-1.08460054e-01 -6.18530750e-01 4.46110755e-01 5.04698185e-03
1.05135012e+00 -4.09534723e-01 1.51905462e-01 4.57149237e-01
1.58231175e+00 4.88050252e-01 1.00048506e+00 2.90627271e-01
1.01384234e+00 3.61223578e-01 5.06346166e-01 1.48590118e-01
2.51806295e-03 4.52512324e-01 3.94213438e-01 -4.10877556e-01
-6.85294718e-02 -3.92446332e-02 2.88297594e-01 9.15641487e-01
-1.85680106e-01 -5.06912291e-01 -1.20934415e+00 3.01008850e-01
-1.28520072e+00 -9.92535532e-01 -2.78026640e-01 2.10057330e+00
9.44082916e-01 2.39490464e-01 -6.16670288e-02 2.41133735e-01
1.17170763e+00 -1.73930690e-01 -5.34315288e-01 -9.64076445e-02
-1.52066007e-01 -2.43385866e-01 6.37147605e-01 9.82179716e-02
-1.07673383e+00 1.12034845e+00 5.50455856e+00 7.93927252e-01
-1.53001118e+00 -1.01409346e-01 1.02730978e+00 2.77645975e-01
5.27384356e-02 -4.77867305e-01 -8.19522500e-01 8.78546238e-01
8.86608422e-01 1.51923984e-01 3.80605608e-01 1.08901310e+00
1.21137239e-01 2.14227915e-01 -9.59766209e-01 1.04846585e+00
2.79828429e-01 -1.57787824e+00 8.96685123e-02 2.21972778e-01
1.05256343e+00 6.94485661e-03 1.50013685e-01 5.36865056e-01
3.53101432e-01 -9.80835080e-01 5.90413451e-01 5.25767028e-01
9.92574573e-01 -8.18402588e-01 1.13792729e+00 6.00459337e-01
-5.88033319e-01 -4.98470431e-03 -5.38155377e-01 2.91642785e-01
-9.34889093e-02 2.58856595e-01 -1.11850810e+00 -1.70652885e-02
1.49146140e-01 2.82120377e-01 -7.83458591e-01 1.16456234e+00
-1.49833977e-01 7.48085022e-01 -1.25363275e-01 -1.69605002e-01
4.63523775e-01 -3.69898826e-01 5.05239666e-02 9.29793060e-01
2.47270957e-01 1.15372710e-01 1.45939970e-02 9.85419929e-01
-6.25468910e-01 5.46826720e-02 -5.75246811e-01 -2.20400795e-01
4.18423355e-01 1.32267964e+00 -8.63401651e-01 -5.65691829e-01
-5.21694541e-01 8.65116417e-01 4.17485572e-02 3.51464629e-01
-1.23039079e+00 -3.04465026e-01 2.08272099e-01 1.60144821e-01
4.38216478e-02 -1.45481387e-02 -2.21637130e-01 -1.15982521e+00
-8.59940648e-02 -1.06179273e+00 -2.19452843e-01 -9.93667424e-01
-1.22350514e+00 9.71414506e-01 -4.17033941e-01 -1.81995392e+00
-3.46373737e-01 -9.39789355e-01 -6.81667686e-01 9.07113492e-01
-1.19307244e+00 -9.26659048e-01 -7.71910787e-01 6.19493902e-01
8.09623063e-01 -8.53953540e-01 6.93129838e-01 5.86872041e-01
-8.11266601e-01 5.88469386e-01 4.88060355e-01 9.80211437e-01
4.72926110e-01 -9.23547506e-01 5.05570352e-01 1.03635788e+00
-5.16461171e-02 1.85498729e-01 4.23995614e-01 -5.27370036e-01
-1.10559154e+00 -1.53600979e+00 9.43585113e-02 -1.12395823e-01
2.72856295e-01 -3.80772978e-01 -1.19144118e+00 5.32272398e-01
1.58112675e-01 3.66456717e-01 8.02829146e-01 -6.65790021e-01
1.74635071e-02 -2.97606409e-01 -1.40169418e+00 5.52606761e-01
2.82174140e-01 -6.56912178e-02 -6.13249779e-01 3.73718977e-01
4.14529800e-01 -3.66583973e-01 -6.02531135e-01 1.75550699e-01
1.65574312e-01 -5.48233688e-01 7.18329787e-01 -4.73383725e-01
6.98105335e-01 -4.58659619e-01 4.32781726e-02 -1.45254004e+00
-2.09365964e-01 9.52119380e-02 6.33141518e-01 1.51604605e+00
6.37180209e-01 -5.05486190e-01 1.08796704e+00 1.08023894e+00
-3.38422477e-01 -3.44660580e-01 -3.67584646e-01 -8.07405412e-01
-1.89676974e-02 -1.75019413e-01 7.16559410e-01 1.32532191e+00
-5.76209366e-01 2.79431075e-01 -5.04582286e-01 1.41592667e-01
4.80169743e-01 -2.41280138e-01 1.07967532e+00 -1.20715666e+00
5.01430407e-02 -2.12108716e-01 -6.63235426e-01 -5.81628919e-01
1.56838268e-01 -7.26947129e-01 2.30637237e-01 -1.42027152e+00
-4.34566382e-03 -7.31017292e-01 6.42710999e-02 5.32195330e-01
-7.36522228e-02 8.21418345e-01 4.84314263e-01 4.44479465e-01
-2.95552969e-01 5.47409177e-01 1.45826828e+00 -3.65965962e-01
-1.66221619e-01 2.56603718e-01 -4.60716933e-01 9.12371933e-01
1.01105976e+00 -3.15760076e-01 -4.03066605e-01 -4.61268991e-01
-1.16099805e-01 3.39470804e-02 2.54817277e-01 -1.23706305e+00
1.61509082e-01 9.49694123e-03 8.27116251e-01 -4.78519946e-01
4.17885751e-01 -9.94445622e-01 8.88082609e-02 1.48763120e-01
-1.74058542e-01 -3.37227881e-02 4.62031037e-01 2.42450431e-01
-4.27199781e-01 -7.58753002e-01 1.08688200e+00 -2.83915192e-01
-1.18195438e+00 3.88782531e-01 -4.29330498e-01 5.16349487e-02
1.09666634e+00 -4.78833795e-01 -3.58004123e-01 -2.55430579e-01
-6.12018704e-01 -2.00641617e-01 5.85287154e-01 5.69388568e-01
5.86227715e-01 -1.64848471e+00 -8.21154654e-01 3.85694742e-01
4.77219969e-02 1.02087475e-01 2.66736388e-01 1.07525080e-01
-1.08899808e+00 2.49836490e-01 -7.73631394e-01 -4.01687860e-01
-1.05360711e+00 5.35200536e-01 3.70528638e-01 -1.27352163e-01
-3.35211366e-01 6.08698964e-01 8.92343298e-02 -5.62117875e-01
-4.56510251e-03 -2.10670173e-01 -2.52704561e-01 -1.37946546e-01
5.31521082e-01 2.19688430e-01 6.46793842e-02 -7.61725664e-01
5.83805330e-02 3.32800120e-01 -1.62216648e-01 3.84185016e-02
1.38457906e+00 4.52047229e-01 7.28990510e-02 -1.65460091e-02
1.13977492e+00 4.38482985e-02 -1.60384405e+00 -1.12366490e-01
-4.17546779e-01 -5.37822425e-01 -2.24991292e-01 -8.12370241e-01
-1.54847193e+00 8.42461705e-01 6.62539124e-01 2.03175485e-01
1.03525746e+00 -4.09318447e-01 6.36164367e-01 5.89442790e-01
2.29903311e-01 -1.12995505e+00 1.57630473e-01 3.44840825e-01
9.30188239e-01 -1.43047237e+00 -3.16541731e-01 -3.16858947e-01
-1.04897892e+00 1.10651088e+00 1.08976817e+00 -3.44282746e-01
9.41039771e-02 4.54561710e-01 2.44004503e-01 2.63892084e-01
-3.99014056e-02 1.43536881e-01 3.83377448e-02 6.67454243e-01
1.02169991e-01 -8.03786963e-02 1.58598498e-01 7.23390698e-01
-1.59955934e-01 -8.40503722e-03 9.50649917e-01 5.30929387e-01
-4.52744156e-01 -1.12702703e+00 -8.49746287e-01 4.94191200e-01
-3.92363667e-01 -1.45176411e-01 -2.82621503e-01 1.22011721e+00
4.34176266e-01 6.10280037e-01 2.71807194e-01 -4.15171176e-01
2.68640578e-01 1.68630421e-01 1.69232879e-02 -5.82248151e-01
-2.01808080e-01 -2.14327186e-01 -8.39344561e-02 2.71385521e-01
-2.45399117e-01 -2.92882562e-01 -1.22218359e+00 -3.45842957e-01
-4.61370319e-01 1.39499471e-01 8.64653349e-01 8.00721526e-01
1.03761479e-02 5.72110176e-01 8.38972926e-01 -8.12522173e-01
-4.22020495e-01 -1.12160051e+00 -6.30519450e-01 6.05693579e-01
-1.45589225e-02 -3.31668198e-01 -8.22840482e-02 5.48521578e-01] | [9.935583114624023, -4.642817497253418] |
e4c5d5b0-f61e-46cb-b3a9-f2e756e46b1d | the-visual-centrifuge-model-free-layered | 1812.01461 | null | http://arxiv.org/abs/1812.01461v2 | http://arxiv.org/pdf/1812.01461v2.pdf | The Visual Centrifuge: Model-Free Layered Video Representations | True video understanding requires making sense of non-lambertian scenes where
the color of light arriving at the camera sensor encodes information about not
just the last object it collided with, but about multiple mediums -- colored
windows, dirty mirrors, smoke or rain. Layered video representations have the
potential of accurately modelling realistic scenes but have so far required
stringent assumptions on motion, lighting and shape. Here we propose a
learning-based approach for multi-layered video representation: we introduce
novel uncertainty-capturing 3D convolutional architectures and train them to
separate blended videos. We show that these models then generalize to single
videos, where they exhibit interesting abilities: color constancy, factoring
out shadows and separating reflections. We present quantitative and qualitative
results on real world videos. | ['João Carreira', 'Jean-Baptiste Alayrac', 'Andrew Zisserman'] | 2018-12-04 | the-visual-centrifuge-model-free-layered-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Alayrac_The_Visual_Centrifuge_Model-Free_Layered_Video_Representations_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Alayrac_The_Visual_Centrifuge_Model-Free_Layered_Video_Representations_CVPR_2019_paper.pdf | cvpr-2019-6 | ['color-constancy'] | ['computer-vision'] | [ 2.20405012e-01 -1.83541164e-01 2.73284972e-01 -2.96648234e-01
-4.91925776e-01 -7.11937606e-01 6.61155164e-01 -4.23200101e-01
1.85702380e-03 5.40007949e-01 1.90571055e-01 -9.91461575e-02
7.31002688e-02 -4.55997288e-01 -1.01653326e+00 -7.45058239e-01
-5.86469829e-01 2.92105436e-01 4.65170771e-01 1.00996040e-01
-3.97237241e-02 5.91737092e-01 -1.72359443e+00 7.51471162e-01
4.96470369e-02 1.01979721e+00 1.83788538e-01 1.44127059e+00
-3.28347571e-02 1.47300887e+00 -4.85381722e-01 -1.21239863e-01
2.55228758e-01 -2.40330696e-01 -3.24490607e-01 6.17811978e-01
8.69131446e-01 -9.57467437e-01 -6.90931916e-01 7.26031423e-01
-3.18045050e-01 3.91966999e-02 6.29845142e-01 -1.28519571e+00
-7.30998456e-01 1.97105512e-01 -6.63663626e-01 1.48364589e-01
6.09088421e-01 5.51688112e-02 7.96746850e-01 -8.99274468e-01
5.76841474e-01 1.40916252e+00 8.48776221e-01 6.55970931e-01
-9.24671292e-01 -1.74353227e-01 6.76752090e-01 1.23177685e-01
-1.00503576e+00 -7.06771553e-01 5.62941015e-01 -5.06943941e-01
7.45829105e-01 4.09967184e-01 7.17953861e-01 1.26877367e+00
4.03475583e-01 9.31093395e-01 1.12444198e+00 -2.81637758e-01
1.91138953e-01 1.18951708e-01 3.41894105e-02 9.04123008e-01
4.03037876e-01 6.97221830e-02 -5.46637595e-01 -1.40530378e-01
1.00891471e+00 2.08820716e-01 -5.11648238e-01 -5.88089108e-01
-1.10698938e+00 2.36865953e-01 1.56354681e-01 1.84667557e-02
-1.48712337e-01 7.65216827e-01 -4.42398340e-02 2.25171950e-02
4.25552875e-01 -1.28290340e-01 -3.66693169e-01 5.99251427e-02
-9.84480619e-01 1.38961986e-01 8.31294596e-01 1.14122176e+00
7.76739657e-01 3.51852298e-01 2.42221311e-01 2.67579854e-01
5.64328671e-01 7.25797832e-01 -1.97089165e-01 -1.62225795e+00
2.53810078e-01 -9.14847255e-02 6.33472860e-01 -8.04738343e-01
-3.21410298e-01 2.16011539e-01 -5.55540025e-01 8.03545833e-01
3.38765949e-01 -2.17269674e-01 -1.30193329e+00 1.42656660e+00
3.44858095e-02 8.16355646e-01 2.19775289e-01 1.15279675e+00
9.16629791e-01 7.92724133e-01 -3.03431779e-01 -1.70064181e-01
1.14406896e+00 -6.23903513e-01 -6.46534503e-01 -3.08546275e-01
-7.09802508e-02 -6.59756601e-01 5.29973567e-01 8.08464408e-01
-1.29985130e+00 -2.50536114e-01 -1.10453510e+00 -4.53569181e-02
-7.55829960e-02 -1.56247422e-01 7.39993453e-01 7.23071873e-01
-1.29004502e+00 4.10166889e-01 -9.23632145e-01 -1.21413879e-01
2.86923617e-01 -9.55052897e-02 -4.44598526e-01 -6.69233561e-01
-6.58411682e-01 6.54018402e-01 -1.18754260e-01 3.28965217e-01
-1.46506071e+00 -6.08710587e-01 -9.38922524e-01 -9.41924099e-03
3.18025738e-01 -7.11855710e-01 1.22258401e+00 -1.28646398e+00
-1.37372923e+00 6.11523271e-01 -3.20552051e-01 -1.22761019e-01
3.56457531e-01 -5.29053271e-01 -4.36762273e-01 4.80991691e-01
-4.20585275e-01 2.45704830e-01 1.10372734e+00 -2.03959727e+00
-5.09185970e-01 -1.90829694e-01 7.64583349e-01 2.43369415e-01
3.41472805e-01 -7.95961246e-02 -5.19493163e-01 -3.02699804e-01
1.12354107e-01 -9.37167168e-01 -2.27282807e-01 5.44665813e-01
-1.83969527e-01 6.48343623e-01 1.06249571e+00 -4.97116596e-01
4.67806846e-01 -2.03974915e+00 7.12395236e-02 -3.04702241e-02
3.27010363e-01 -4.58760671e-02 -2.45132595e-01 2.67089605e-01
2.08534837e-01 -5.74125210e-03 -8.51825327e-02 -4.24324930e-01
-1.50155067e-01 5.93905151e-01 -3.92195880e-01 6.59423590e-01
3.42762649e-01 4.39910680e-01 -9.50361907e-01 -3.45379829e-01
5.40539443e-01 1.02506673e+00 -5.42733729e-01 2.19367385e-01
-5.27754962e-01 2.05541924e-01 -2.91545540e-01 8.52000594e-01
1.14563072e+00 -1.08067743e-01 1.24446005e-01 -3.28362174e-02
-4.37341928e-02 -1.20058663e-01 -1.46692693e+00 1.56859946e+00
-4.16666329e-01 1.14296758e+00 7.75357902e-01 -3.90003920e-01
1.95961148e-01 4.28548306e-01 5.59840858e-01 -3.64138067e-01
-5.71636185e-02 -2.57119387e-01 -5.10656893e-01 -6.88762426e-01
5.78653634e-01 -3.20698977e-01 2.81153530e-01 4.25762117e-01
-9.28466395e-02 -4.12620693e-01 -8.68596807e-02 4.62412834e-01
1.10729051e+00 6.04881823e-01 -3.85026395e-01 -4.27155793e-02
7.76222870e-02 -2.20436618e-01 5.09237051e-01 8.19278896e-01
-4.36939336e-02 1.14414895e+00 2.10549638e-01 -4.59111422e-01
-8.19178939e-01 -1.53669763e+00 1.07077934e-01 9.47776377e-01
4.80224609e-01 -2.79569030e-01 -4.58586872e-01 -2.25195110e-01
4.25374843e-02 6.89391673e-01 -6.85807526e-01 1.27829909e-01
-4.75668579e-01 -4.43905354e-01 3.00286710e-01 5.09579003e-01
7.81107098e-02 -4.30168092e-01 -9.41533804e-01 -8.78866110e-03
-2.29175791e-01 -1.49290538e+00 -5.23656458e-02 2.30127499e-01
-7.31591284e-01 -1.41154110e+00 -4.86294806e-01 -1.70634851e-01
3.83362710e-01 9.80648398e-01 1.38076472e+00 2.38269284e-01
-5.67086101e-01 1.40452039e+00 -3.36205631e-01 -4.00542468e-01
-2.14592859e-01 -1.32355165e+00 8.59697759e-02 2.88355261e-01
9.06592049e-03 -3.73546094e-01 -6.88593149e-01 2.28814464e-02
-1.08901668e+00 2.83870976e-02 2.79762805e-01 3.09614897e-01
1.70257270e-01 -9.86015052e-02 -6.16527915e-01 -6.75992489e-01
-1.95773616e-01 -4.67192411e-01 -5.03212929e-01 3.21389854e-01
2.67438442e-01 -4.86565717e-02 2.89335549e-01 -3.75118703e-01
-1.50948763e+00 7.45616034e-02 3.41360539e-01 -9.51027870e-01
-3.40082467e-01 -1.24538876e-01 -2.44842358e-02 -1.14550047e-01
2.71549702e-01 -4.36924770e-02 -3.38063687e-01 -1.94245979e-01
6.31571412e-01 1.68531209e-01 5.70542157e-01 -6.74602091e-01
8.14593434e-01 1.22817910e+00 6.09664284e-02 -1.23029411e+00
-7.84406066e-01 -2.58306950e-01 -5.68457186e-01 -5.45456707e-01
8.62890720e-01 -1.18524206e+00 -6.74203873e-01 4.64595586e-01
-1.45967472e+00 -3.88193935e-01 -8.97125527e-02 6.24260187e-01
-6.04743481e-01 4.16963845e-01 -8.27975452e-01 -1.29650235e+00
5.56826711e-01 -8.62747312e-01 1.20570564e+00 1.62212715e-01
1.96338788e-01 -1.16902399e+00 1.75724342e-01 1.98659539e-01
1.98770240e-01 4.43790406e-01 6.54664099e-01 2.60797113e-01
-1.28821182e+00 1.41079620e-01 -3.18780750e-01 3.88508648e-01
-7.16712922e-02 5.97991765e-01 -1.51815188e+00 -1.80054665e-01
3.84550840e-02 -3.30063373e-01 1.19594324e+00 5.78117490e-01
1.02203619e+00 -1.16005950e-01 -2.32054725e-01 8.50641370e-01
1.66610384e+00 -1.81660731e-03 6.44030094e-01 -6.39087409e-02
9.33608532e-01 4.26942110e-01 2.83674508e-01 6.14736199e-01
3.32479030e-01 2.86200702e-01 1.02931750e+00 -6.67689815e-02
-3.10780972e-01 1.88553989e-01 6.65854871e-01 2.93794394e-01
-5.25808096e-01 -7.06342697e-01 -4.57665712e-01 3.19948018e-01
-1.70385897e+00 -1.32985377e+00 -3.52257669e-01 2.12500596e+00
1.11178406e-01 6.50022551e-02 -3.93329203e-01 -1.76328033e-01
3.07090968e-01 4.62473720e-01 -5.24787366e-01 -3.33113641e-01
-3.69778097e-01 -1.71923473e-01 5.96383452e-01 8.34073365e-01
-1.00188005e+00 4.95747566e-01 7.69043064e+00 -1.13140441e-01
-7.63773143e-01 -6.19747955e-03 3.98356050e-01 -4.39134270e-01
-7.68437564e-01 8.81939456e-02 -4.20627415e-01 -1.60070807e-02
7.53697872e-01 5.15576184e-01 6.12051189e-01 3.39798301e-01
1.63082987e-01 -4.88987833e-01 -1.42203665e+00 9.87312376e-01
4.49632227e-01 -1.31528699e+00 1.31734237e-01 -1.56711400e-01
7.47964501e-01 2.79148161e-01 9.29763541e-02 -8.75765681e-02
8.11605811e-01 -9.24859107e-01 1.07752955e+00 1.14371884e+00
5.03018320e-01 -1.91360220e-01 1.26490563e-01 1.14747226e-01
-1.18972981e+00 -2.27451399e-01 -3.39732856e-01 -2.35018089e-01
2.84768045e-01 5.39075017e-01 -3.86499763e-01 3.85391772e-01
9.81656373e-01 7.68767715e-01 -2.43302748e-01 9.19741213e-01
6.78316876e-02 3.74641508e-01 -4.89888191e-01 4.32223499e-01
3.66198421e-01 -1.80871561e-01 6.69451654e-01 1.41961527e+00
4.36803371e-01 3.53252888e-01 4.52856943e-02 8.30620825e-01
1.96302280e-01 -8.70270967e-01 -8.30701172e-01 1.17802419e-01
-7.58023560e-02 1.19357467e+00 -5.27013302e-01 -1.83079377e-01
-9.60817814e-01 1.14840531e+00 -6.82070330e-02 1.07055843e+00
-8.32096696e-01 1.10764705e-01 1.13399303e+00 1.88831866e-01
4.91013557e-01 -6.09066546e-01 1.88079566e-01 -1.60589397e+00
1.30691286e-02 -3.61296684e-01 1.48751333e-01 -1.49200261e+00
-1.16898835e+00 2.78859586e-01 1.07178809e-02 -1.17966044e+00
-1.11345097e-01 -1.20907605e+00 -6.45594954e-01 4.22896981e-01
-1.71090817e+00 -1.22982132e+00 -5.57989478e-01 7.27128565e-01
5.46634972e-01 2.35041291e-01 7.09317088e-01 7.58558437e-02
-9.92910340e-02 -2.60892510e-01 4.83088672e-01 8.97962973e-03
5.14573157e-01 -1.24408948e+00 9.09777954e-02 9.03546751e-01
4.63793516e-01 4.11622494e-01 1.01327860e+00 -2.03589514e-01
-2.02535677e+00 -7.74513662e-01 4.38395962e-02 -7.95306325e-01
6.21364653e-01 -5.72744548e-01 -7.79833913e-01 1.07449579e+00
3.49241346e-01 3.96693975e-01 4.10108924e-01 -4.21280637e-02
-8.62396657e-01 -9.43202004e-02 -9.12460923e-01 3.06463957e-01
8.36250067e-01 -6.85199678e-01 -3.37304413e-01 2.10942104e-01
5.73320925e-01 -5.29290617e-01 -3.91329318e-01 1.17842227e-01
8.67933512e-01 -1.65210450e+00 1.28987229e+00 -5.95863163e-01
2.87347615e-01 -2.66084969e-01 -6.51916027e-01 -1.22003627e+00
-2.46066138e-01 -4.12701696e-01 -4.00071412e-01 7.50437677e-01
-2.29242779e-02 -1.81359544e-01 6.96662784e-01 8.89500380e-01
-2.49948114e-01 -3.99452567e-01 -7.70842791e-01 -5.03074586e-01
-1.15186885e-01 -8.93976808e-01 1.83645099e-01 6.12790585e-01
-3.89666736e-01 -2.74285167e-01 -6.55713737e-01 7.38714397e-01
9.85158920e-01 3.85326259e-02 4.83375400e-01 -1.27243865e+00
-4.19300407e-01 -1.40369534e-01 -4.00894105e-01 -1.05685818e+00
2.82537006e-02 -6.82279989e-02 3.17590028e-01 -1.73385322e+00
3.51909459e-01 -3.11985224e-01 -3.44586611e-01 -8.65183771e-02
1.46972850e-01 3.20229352e-01 3.41902912e-01 -9.75876004e-02
-1.00869727e+00 2.60262907e-01 9.66809690e-01 -2.22227648e-01
2.43782222e-01 -7.06264302e-02 -3.68504137e-01 1.12298286e+00
3.26275021e-01 -1.26369804e-01 -5.61678886e-01 -1.09112287e+00
5.13828278e-01 2.06108272e-01 9.25820529e-01 -1.00002444e+00
3.56228873e-02 -4.22495395e-01 8.56610894e-01 -5.65454960e-01
1.07918608e+00 -1.28679264e+00 2.80229539e-01 1.86054427e-02
-1.31341934e-01 -5.47513179e-02 3.01595956e-01 1.08969450e+00
1.13255776e-01 -1.15771286e-01 5.31165063e-01 -6.23298764e-01
-9.05388236e-01 2.52218336e-01 -8.45855892e-01 -3.78080979e-02
1.02944136e+00 -3.46614420e-01 -4.20107335e-01 -6.89427733e-01
-8.39208782e-01 1.23273171e-01 6.84019923e-01 4.58596259e-01
1.02324986e+00 -1.05123198e+00 -6.26192927e-01 2.04358369e-01
3.92288528e-02 -1.26046166e-01 4.14743990e-01 3.91724020e-01
-7.36910760e-01 1.59889713e-01 2.60955170e-02 -8.51210475e-01
-1.33213782e+00 4.69312847e-01 5.08096635e-01 4.40021574e-01
-6.96135461e-01 9.34308231e-01 5.51869869e-01 -1.99488048e-02
3.81946385e-01 -6.92589164e-01 2.02710792e-01 -2.02659607e-01
8.21110606e-01 3.30012083e-01 -3.20218176e-01 -7.06632674e-01
-4.86920714e-01 7.47227013e-01 1.77865133e-01 -2.01353982e-01
1.27213001e+00 -5.91272473e-01 -3.79043072e-02 9.99871254e-01
1.02634013e+00 6.87461644e-02 -2.02563167e+00 -1.14377148e-01
-4.07039791e-01 -7.52586603e-01 1.76169351e-01 -6.09209239e-01
-1.04061294e+00 1.00852132e+00 4.08751935e-01 1.74690858e-01
9.44800019e-01 6.58561885e-02 4.50741768e-01 4.86299187e-01
5.51227450e-01 -7.01813281e-01 2.56957769e-01 5.83001792e-01
5.34848928e-01 -1.18574667e+00 2.78688788e-01 -1.69076100e-01
-5.93273520e-01 1.50270593e+00 4.39907551e-01 -1.13608219e-01
8.46536160e-01 7.70278513e-01 2.87045091e-01 -1.34885058e-01
-1.02332318e+00 -1.29803732e-01 1.20781265e-01 9.09378052e-01
3.31195623e-01 -3.27247195e-02 9.83067870e-01 -3.31313200e-02
5.58921099e-01 -2.12506652e-01 1.06086230e+00 9.11983073e-01
-6.54125988e-01 -3.82711947e-01 -7.19512939e-01 9.95773301e-02
-3.25821221e-01 -6.76365802e-03 -2.85050660e-01 7.62710989e-01
1.73648492e-01 1.06195998e+00 3.22905719e-01 -1.72284156e-01
-3.69688161e-02 -1.79630533e-01 1.12928748e+00 -3.77972513e-01
3.49081717e-02 1.62074581e-01 1.18321687e-01 -9.66847181e-01
-9.34425294e-01 -7.54176915e-01 -1.24753857e+00 -1.82498693e-01
-1.30947053e-01 -1.37358814e-01 6.47460043e-01 9.22946453e-01
5.00236936e-02 4.78632629e-01 5.55800736e-01 -1.55622911e+00
-1.34130137e-03 -4.42020237e-01 -8.62721086e-01 2.62263477e-01
1.33015001e+00 -6.98684096e-01 -6.17118716e-01 4.08868074e-01] | [9.482486724853516, -2.884411334991455] |
cb765d48-89d9-476a-b341-08a003800c1b | sparse-illumination-learning-and-transfer-for | 1402.1879 | null | https://arxiv.org/abs/1402.1879v1 | https://arxiv.org/pdf/1402.1879v1.pdf | Sparse Illumination Learning and Transfer for Single-Sample Face Recognition with Image Corruption and Misalignment | Single-sample face recognition is one of the most challenging problems in face recognition. We propose a novel algorithm to address this problem based on a sparse representation based classification (SRC) framework. The new algorithm is robust to image misalignment and pixel corruption, and is able to reduce required gallery images to one sample per class. To compensate for the missing illumination information traditionally provided by multiple gallery images, a sparse illumination learning and transfer (SILT) technique is introduced. The illumination in SILT is learned by fitting illumination examples of auxiliary face images from one or more additional subjects with a sparsely-used illumination dictionary. By enforcing a sparse representation of the query image in the illumination dictionary, the SILT can effectively recover and transfer the illumination and pose information from the alignment stage to the recognition stage. Our extensive experiments have demonstrated that the new algorithms significantly outperform the state of the art in the single-sample regime and with less restrictions. In particular, the single-sample face alignment accuracy is comparable to that of the well-known Deformable SRC algorithm using multiple gallery images per class. Furthermore, the face recognition accuracy exceeds those of the SRC and Extended SRC algorithms using hand labeled alignment initialization. | ['S. Shankar Sastry', 'Tsung-Han Chan', 'Allen Y. Yang', 'Yi Ma', 'Liansheng Zhuang'] | 2014-02-08 | null | null | null | null | ['sparse-representation-based-classification'] | ['computer-vision'] | [ 7.49292195e-01 -1.59538642e-01 -2.42371783e-02 -5.86356759e-01
-9.90014315e-01 -4.77121413e-01 4.96395707e-01 -8.63527358e-01
-1.06920287e-01 6.82569265e-01 -1.51338562e-01 4.23983812e-01
-1.23387706e-02 -3.24447721e-01 -7.30014324e-01 -1.25225079e+00
5.66450238e-01 5.96061885e-01 -4.62649375e-01 -1.98930521e-02
1.44005880e-01 8.79163861e-01 -1.91818631e+00 6.49116561e-02
6.60587251e-01 1.16348624e+00 -1.22861529e-03 1.64169863e-01
2.78918035e-02 1.93082958e-01 -5.70726752e-01 -2.10630655e-01
5.95652521e-01 -6.15766048e-01 -3.85894656e-01 6.83007717e-01
1.31030226e+00 -1.23186283e-01 -1.47639915e-01 1.12677872e+00
5.57338417e-01 1.97358057e-01 4.70976889e-01 -1.29770410e+00
-5.69336832e-01 -2.11369187e-01 -6.96475208e-01 -1.07204266e-01
5.25072336e-01 -1.11070178e-01 2.96468496e-01 -1.39222443e+00
6.34130180e-01 1.41245186e+00 4.70639139e-01 7.33206332e-01
-1.49475753e+00 -7.91604638e-01 -1.48924915e-02 9.91348848e-02
-1.58538187e+00 -1.13930011e+00 1.02665651e+00 -3.23043495e-01
4.65444088e-01 3.86442572e-01 4.03898448e-01 1.08705199e+00
-2.46276438e-01 3.57504010e-01 1.51990116e+00 -7.68404663e-01
1.85775816e-01 2.10048735e-01 7.22061992e-02 9.49468434e-01
4.21311408e-01 1.25877604e-01 -7.54502356e-01 -2.21216306e-01
9.35426235e-01 2.37856090e-01 -3.51691842e-01 -6.45461023e-01
-9.02346313e-01 6.44702852e-01 1.10791832e-01 3.34484577e-01
-2.93170094e-01 -2.36253366e-01 -6.98225275e-02 4.00440961e-01
4.59739208e-01 1.01935141e-01 -1.00541532e-01 4.29714292e-01
-1.22211778e+00 -2.65038580e-01 9.03621793e-01 8.26212227e-01
1.16911995e+00 4.83874559e-01 -1.23514019e-01 1.00507677e+00
2.94086397e-01 1.12697780e+00 4.70011234e-01 -9.12601948e-01
3.42991024e-01 4.37258571e-01 -5.48702516e-02 -1.03727639e+00
-8.43437482e-03 -2.70935059e-01 -9.44993734e-01 2.71004558e-01
4.17205244e-01 2.75708497e-01 -9.95169640e-01 1.60276532e+00
4.96391624e-01 3.24007004e-01 1.64882123e-01 9.15048122e-01
6.84855700e-01 3.32661927e-01 -4.78631884e-01 -6.32386982e-01
1.31190228e+00 -8.73683035e-01 -8.04647923e-01 -2.82254994e-01
3.89499478e-02 -1.09377968e+00 7.04241574e-01 4.55952376e-01
-8.34235728e-01 -7.24352241e-01 -9.18163061e-01 1.79975163e-02
-8.14844072e-02 5.77900410e-01 2.98701555e-01 1.01020801e+00
-8.98789108e-01 2.77977765e-01 -6.48920357e-01 -4.10372138e-01
5.54499865e-01 6.24498427e-01 -8.43181491e-01 -7.33853042e-01
-4.67434436e-01 7.50638723e-01 -2.09058166e-01 3.84273082e-01
-9.50895309e-01 -4.07148123e-01 -1.05848122e+00 -2.01758265e-01
2.61057168e-01 -2.89309919e-01 5.45850515e-01 -1.29860187e+00
-1.83771312e+00 1.10574853e+00 -7.10187912e-01 2.20524281e-01
8.64231884e-02 -1.18364573e-01 -3.75094175e-01 3.28343272e-01
1.13795869e-01 3.76591653e-01 1.64398468e+00 -1.43261623e+00
3.89941037e-02 -7.51978457e-01 -5.22744536e-01 9.93446931e-02
-3.12161088e-01 1.43564433e-01 -5.53027511e-01 -7.23178089e-01
2.91295916e-01 -1.18070209e+00 1.53434917e-01 1.83426917e-01
-1.78029999e-01 2.98368216e-01 1.26986265e+00 -7.49087632e-01
4.86001074e-01 -2.33426881e+00 5.11598051e-01 4.67965335e-01
-1.64011180e-01 4.01669830e-01 -5.00001311e-01 6.01364262e-02
-3.76037210e-01 -7.13098288e-01 -4.38732386e-01 -6.54683769e-01
-3.14814389e-01 6.41048908e-01 -2.34767377e-01 9.28324223e-01
6.93120733e-02 6.50506854e-01 -4.40862894e-01 -3.84817719e-01
2.85847008e-01 9.02131379e-01 -4.18055356e-01 4.35142308e-01
3.08482438e-01 9.87247944e-01 -9.38062146e-02 1.00830615e+00
9.71168995e-01 -9.93072242e-02 1.23490117e-01 -5.50939620e-01
1.31649300e-01 -4.16893393e-01 -1.50551939e+00 1.64152908e+00
-4.46582109e-01 3.55407089e-01 5.77215075e-01 -1.33264387e+00
1.14119947e+00 5.10516286e-01 6.41387045e-01 -4.52740133e-01
-2.18471885e-02 4.09396470e-01 -2.82462388e-01 -3.59833747e-01
-1.35545209e-01 -1.34414643e-01 3.44219506e-01 4.84519750e-01
3.67499709e-01 -1.52566344e-01 -1.02810085e-01 -3.37672591e-01
4.77271616e-01 1.03053987e-01 2.80441731e-01 -2.68801153e-01
1.09876859e+00 -6.43702209e-01 6.99376106e-01 3.07372987e-01
8.04592893e-02 5.85457027e-01 -1.36280715e-01 -6.38804078e-01
-7.40772545e-01 -6.73692286e-01 -2.92427242e-01 8.80474985e-01
-5.67737930e-02 1.22051820e-01 -9.25300777e-01 -6.88148737e-01
2.28221044e-02 1.56469494e-01 -6.00193381e-01 4.32429351e-02
-7.11525917e-01 -7.23156273e-01 1.61965787e-01 2.11940974e-01
6.16299152e-01 -7.83916116e-01 -1.11251123e-01 -1.18092135e-01
-1.92276224e-01 -1.31545985e+00 -7.51623988e-01 -1.59644231e-01
-7.61010289e-01 -1.20867682e+00 -8.39626074e-01 -9.64401484e-01
1.44341254e+00 5.98243833e-01 6.20463490e-01 6.01796852e-03
-6.33543789e-01 7.88166940e-01 -1.73895106e-01 4.04393449e-02
-2.79343784e-01 -3.74686360e-01 4.15853262e-01 9.07577932e-01
1.96033567e-01 -5.71597397e-01 -4.05528724e-01 7.15265989e-01
-7.04151988e-01 -3.71188909e-01 6.42230809e-01 1.18325782e+00
7.99991965e-01 -1.83930144e-01 3.09492469e-01 -7.79686093e-01
-1.81961507e-01 -2.48721391e-02 -8.17765415e-01 2.90176213e-01
-5.42143703e-01 6.69081509e-02 5.23630440e-01 -5.67305863e-01
-1.12657130e+00 6.87419534e-01 1.48579806e-01 -7.83038497e-01
-2.85329342e-01 -2.17384279e-01 -5.65515161e-01 -1.12793422e+00
5.11581898e-01 5.86355925e-01 4.49391395e-01 -7.02598751e-01
2.46324912e-01 4.64516163e-01 6.55359626e-01 -6.37039959e-01
1.27686858e+00 7.19464302e-01 3.41901809e-01 -1.14312983e+00
-7.89540887e-01 -4.80172783e-01 -9.31030333e-01 -1.51078209e-01
4.88351434e-01 -8.05656254e-01 -4.42262888e-01 5.93240917e-01
-9.08858597e-01 1.32623076e-01 -5.16980648e-01 4.86492038e-01
-6.00114465e-01 6.73479855e-01 -1.96936995e-01 -7.01923549e-01
-3.12986583e-01 -1.37589633e+00 1.42816567e+00 1.78111970e-01
3.47090155e-01 -7.91445076e-01 -1.19953677e-01 7.84722567e-01
4.63090450e-01 1.82900533e-01 4.77222472e-01 -4.74464118e-01
-5.63505828e-01 -4.99343038e-01 -8.55829287e-03 6.16621852e-01
5.26271641e-01 -4.50461745e-01 -1.16905200e+00 -8.88341844e-01
4.54592198e-01 -3.22250247e-01 6.61220074e-01 1.24854617e-01
1.00144386e+00 -4.94156122e-01 -1.80168331e-01 7.65164852e-01
1.43103528e+00 1.75256319e-02 5.91278076e-01 -1.91787243e-01
8.99500787e-01 6.28133714e-01 3.67374390e-01 2.39655510e-01
-4.29366343e-03 1.14353096e+00 2.10521176e-01 -2.85719216e-01
-4.26066399e-01 1.31708726e-01 5.17088890e-01 6.84171081e-01
-1.87811211e-01 2.80149043e-01 -3.17914009e-01 7.92750418e-02
-1.56997776e+00 -1.13632834e+00 2.85295874e-01 2.56406665e+00
7.01046884e-01 -7.26915896e-01 -1.29195169e-01 3.90236974e-01
6.84587717e-01 9.68195125e-02 -4.43646133e-01 -4.57414240e-02
-3.55093896e-01 5.07015646e-01 2.19863102e-01 7.14201868e-01
-1.01740634e+00 8.18887055e-01 6.62883282e+00 6.57801151e-01
-1.13833916e+00 2.63639539e-01 3.87789458e-01 1.70770898e-01
1.03383072e-01 -1.09217040e-01 -1.10346544e+00 3.37176621e-01
5.42302966e-01 1.89471573e-01 9.20128882e-01 7.44334340e-01
-1.35654107e-01 4.41507436e-02 -1.26105869e+00 1.48025918e+00
1.02143049e+00 -8.65173459e-01 2.21906394e-01 2.81586468e-01
9.45183158e-01 -5.31863451e-01 2.53329873e-01 -1.13945313e-01
-3.28515828e-01 -1.02973676e+00 1.86848655e-01 6.62443042e-01
1.14002049e+00 -3.86215717e-01 4.91558790e-01 3.15988883e-02
-1.21634221e+00 -1.08790167e-01 -2.96712816e-01 2.71722525e-01
-1.60064280e-01 5.24416566e-01 -4.20273334e-01 5.46203434e-01
3.50776136e-01 7.51415789e-01 -5.48905790e-01 6.67110503e-01
-1.62222162e-01 3.29986125e-01 -4.74063665e-01 7.20266163e-01
-2.49601468e-01 -5.46922684e-01 6.52074754e-01 7.40700126e-01
3.99026841e-01 7.91686997e-02 4.30422395e-01 5.55970728e-01
-1.64516866e-01 1.03738800e-01 -6.83134675e-01 3.21172982e-01
2.75814921e-01 1.41673267e+00 -4.37746912e-01 -2.24239662e-01
-5.20213485e-01 1.28570831e+00 6.80896193e-02 5.17870545e-01
-4.05384034e-01 -3.05819865e-02 5.71315944e-01 -1.46416083e-01
5.87554634e-01 -9.78657454e-02 1.80436507e-01 -1.31164622e+00
1.92110211e-01 -1.10768378e+00 2.50829667e-01 -4.22399670e-01
-1.22570443e+00 5.65578103e-01 -9.26272944e-02 -1.20456803e+00
-1.32777795e-01 -7.16480911e-01 -3.84605587e-01 9.51877177e-01
-1.73937714e+00 -1.44144630e+00 -5.50154746e-01 1.13937604e+00
5.67922294e-01 -6.27186537e-01 1.14623940e+00 5.16040742e-01
-7.85464525e-01 9.83869553e-01 1.94273055e-01 4.54484373e-02
9.13864076e-01 -7.21156299e-01 -2.70369709e-01 9.32353199e-01
3.68706584e-01 5.77866435e-01 4.32408690e-01 -2.21098691e-01
-2.00365520e+00 -1.01488459e+00 5.58443069e-01 -2.87510812e-01
-4.48629744e-02 -2.47191653e-01 -8.40912223e-01 7.55315840e-01
-4.17988151e-02 5.52830458e-01 9.65389669e-01 -2.53909111e-01
-6.48639083e-01 -5.47655225e-01 -1.42719162e+00 -2.89231110e-02
8.40871751e-01 -5.79939902e-01 -4.90134507e-01 6.15935922e-01
-5.23862131e-02 -4.01400894e-01 -8.52341712e-01 2.76362091e-01
5.72111845e-01 -5.51203370e-01 1.12028229e+00 -1.60027534e-01
-2.14953721e-01 -3.91816795e-01 -5.10834157e-01 -1.17596614e+00
-2.22613305e-01 -8.28330338e-01 -2.05393896e-01 1.20255268e+00
-7.98528194e-02 -9.10136223e-01 7.92227805e-01 5.28827846e-01
1.01839587e-01 -1.60573795e-01 -1.32022774e+00 -1.04302895e+00
-3.63894492e-01 3.71098787e-01 4.31012928e-01 9.80922043e-01
-4.74384964e-01 1.77795738e-01 -7.26342559e-01 3.78953606e-01
1.32976282e+00 5.02105296e-01 8.58908176e-01 -1.38325644e+00
-1.88216403e-01 6.41393289e-02 -5.46195269e-01 -8.33289027e-01
5.86697102e-01 -9.49149668e-01 -1.80621352e-02 -8.30866516e-01
2.51216769e-01 -3.14307094e-01 -4.64601964e-02 7.64887452e-01
8.53212550e-02 8.47699761e-01 1.14423245e-01 4.92238641e-01
-2.32217833e-01 5.54846168e-01 1.31235361e+00 -1.31698400e-01
2.79514249e-02 3.26431287e-03 -4.90372777e-01 4.90529031e-01
4.14643794e-01 -3.95319670e-01 -1.07374236e-01 -3.41814727e-01
-4.26230550e-01 -3.00282687e-02 4.39000249e-01 -9.57444012e-01
2.51466244e-01 -9.37663168e-02 6.22355282e-01 -2.52005875e-01
7.32230902e-01 -1.13161576e+00 3.78751338e-01 3.23481798e-01
4.01726132e-03 -4.36736733e-01 4.54657264e-02 5.27667701e-01
-2.38966733e-01 -1.22282043e-01 1.19360530e+00 5.81112429e-02
-3.29544812e-01 3.89294535e-01 -1.86591428e-02 -2.88114637e-01
9.60726142e-01 -4.89749551e-01 -5.50756939e-02 -1.04930200e-01
-7.77385354e-01 -3.16176862e-01 4.51780498e-01 2.49274388e-01
7.14500189e-01 -1.58618367e+00 -6.75088704e-01 1.01901531e+00
4.61351499e-02 -2.51369655e-01 1.43215239e-01 1.01660836e+00
-5.31770699e-02 3.85084242e-01 -4.22903717e-01 -6.50690377e-01
-1.70629406e+00 6.14903033e-01 3.85471255e-01 9.83673334e-02
-6.71924889e-01 6.36913955e-01 1.70908317e-01 -3.77823383e-01
1.42214194e-01 3.26255739e-01 -9.15974304e-02 1.87507663e-02
7.09064424e-01 4.42971587e-01 2.32948214e-01 -1.35500419e+00
-3.74502838e-01 1.31901407e+00 -4.32528332e-02 8.34636390e-02
1.32502007e+00 -8.98742527e-02 -3.51929098e-01 1.61876325e-02
1.46056497e+00 1.22492455e-01 -1.18068922e+00 -5.22801697e-01
-5.33473790e-01 -9.78911042e-01 -1.37834176e-02 -4.97579753e-01
-1.52413344e+00 7.03812361e-01 8.45884085e-01 -3.87006581e-01
1.32757592e+00 -2.13957831e-01 5.36377192e-01 4.59042460e-01
4.94706959e-01 -7.60579109e-01 2.91656703e-01 2.04223558e-01
1.24073040e+00 -1.32155645e+00 2.24152759e-01 -7.40075469e-01
-2.33032793e-01 1.16059864e+00 4.42540079e-01 -1.19384058e-01
5.82556367e-01 1.54927760e-01 -1.35926427e-02 -1.59685493e-01
-1.24658883e-01 -1.01835276e-05 5.02894521e-01 7.53130496e-01
6.02636486e-02 -2.73661792e-01 1.83708191e-01 5.84470853e-02
1.53914005e-01 -1.76597089e-01 5.96455820e-02 7.74368048e-01
-2.15286598e-01 -1.35989499e+00 -1.01541376e+00 7.88850039e-02
-1.36308596e-01 2.44869128e-01 -3.04746151e-01 3.45022082e-01
1.33518642e-03 8.25691640e-01 -1.69819236e-01 7.54802749e-02
2.70634890e-01 2.87630826e-01 1.00392079e+00 -6.31772876e-01
-2.43127301e-01 2.31181622e-01 -3.94589901e-01 -8.78232539e-01
-8.52849603e-01 -8.40952933e-01 -6.29987895e-01 2.66086534e-02
-4.44123417e-01 9.22043994e-02 6.30954862e-01 9.37337101e-01
3.63696933e-01 -4.40068021e-02 1.19582725e+00 -1.42952585e+00
-5.40023804e-01 -6.69637144e-01 -6.97795510e-01 5.42597294e-01
4.49522048e-01 -9.67962623e-01 -6.92089379e-01 4.20253038e-01] | [12.922341346740723, 0.3428160548210144] |
0cc73e79-46d0-4aa6-a71a-e466d5a366f8 | stereo-event-lifetime-and-disparity | 1907.07518 | null | https://arxiv.org/abs/1907.07518v1 | https://arxiv.org/pdf/1907.07518v1.pdf | Stereo Event Lifetime and Disparity Estimation for Dynamic Vision Sensors | Event-based cameras are biologically inspired sensors that output asynchronous pixel-wise brightness changes in the scene called events. They have a high dynamic range and temporal resolution of a microsecond, opposed to standard cameras that output frames at fixed frame rates and suffer from motion blur. Forming stereo pairs of such cameras can open novel application possibilities, since for each event depth can be readily estimated; however, to fully exploit asynchronous nature of the sensor and avoid fixed time interval event accumulation, stereo event lifetime estimation should be employed. In this paper, we propose a novel method for event lifetime estimation of stereo event-cameras, allowing generation of sharp gradient images of events that serve as input to disparity estimation methods. Since a single brightness change triggers events in both event-camera sensors, we propose a method for single shot event lifetime and disparity estimation, with association via stereo matching. The proposed method is approximately twice as fast and more accurate than if lifetimes were estimated separately for each sensor and then stereo matched. Results are validated on real-world data through multiple stereo event-camera experiments. | ['Ivan Petrović', 'Ivan Marković', 'Antea Hadviger'] | 2019-07-17 | null | null | null | null | ['stereo-matching'] | ['computer-vision'] | [ 9.17074859e-01 -2.97289878e-01 5.10481060e-01 -3.04196775e-01
-5.01932263e-01 -5.99503577e-01 5.60785592e-01 2.74382234e-01
-9.57837224e-01 1.01425362e+00 -1.64027795e-01 2.35014558e-01
2.41025612e-01 -6.82181299e-01 -9.38655555e-01 -8.76208961e-01
-3.39272358e-02 -1.70220882e-02 9.68191087e-01 4.78828132e-01
5.43322980e-01 5.71236253e-01 -2.30594110e+00 2.91950077e-01
4.02187020e-01 1.09799039e+00 5.99515021e-01 1.28998733e+00
6.26553223e-02 9.72065806e-01 -7.03497708e-01 -2.10435521e-02
2.01627061e-01 -4.34923947e-01 -1.74734756e-01 2.01012716e-01
5.99215090e-01 -7.75015116e-01 -3.06872815e-01 8.12621653e-01
5.31572640e-01 1.19996533e-01 3.52787226e-01 -1.22481811e+00
3.84691596e-01 5.91667369e-02 -6.35718822e-01 6.29107237e-01
8.22389841e-01 2.52323568e-01 3.36803555e-01 -6.88088834e-01
8.45174432e-01 9.00907278e-01 5.02590835e-01 6.85826838e-01
-1.27062392e+00 -5.70513070e-01 -1.67979166e-01 2.53479540e-01
-1.13606131e+00 -8.22304904e-01 5.67848802e-01 -2.90676326e-01
1.19867218e+00 1.03061952e-01 9.21410859e-01 8.86062682e-01
6.75048947e-01 4.33789521e-01 1.25095856e+00 -3.47544134e-01
5.58517158e-01 -3.00204605e-01 -1.43625557e-01 5.41265428e-01
1.93899274e-01 3.93512994e-01 -1.11138713e+00 -8.08868557e-02
1.16321182e+00 2.59610713e-01 -4.96345848e-01 2.00703312e-02
-1.52181351e+00 1.14343651e-01 -1.10449202e-01 -8.73787999e-02
-4.82777417e-01 6.10653102e-01 2.38155305e-01 8.64261240e-02
3.70331667e-02 -9.51674879e-02 -1.92633867e-02 -4.90528494e-01
-9.09413099e-01 1.80257689e-02 6.95781171e-01 8.81518841e-01
8.50203812e-01 -2.15130731e-01 -3.14965695e-02 1.78556740e-01
-8.96829739e-03 5.44243097e-01 5.93842924e-01 -1.30128992e+00
1.28552821e-02 3.85629922e-01 3.07244301e-01 -5.80156803e-01
-7.29608163e-02 6.07056797e-01 -5.66337466e-01 7.72706807e-01
7.60415137e-01 -1.18833996e-01 -8.52184176e-01 1.52483141e+00
3.07517529e-01 8.32242608e-01 2.09409386e-01 7.35454738e-01
3.55290443e-01 9.35779572e-01 -1.65880218e-01 -7.14650750e-01
1.39839232e+00 -2.30853796e-01 -5.63569963e-01 -1.35290027e-01
-1.81087151e-01 -7.56404877e-01 5.63598871e-01 6.11547709e-01
-1.41944134e+00 -4.36181396e-01 -9.45669949e-01 -1.01735108e-01
-1.56939834e-01 -3.07888240e-01 6.06852949e-01 5.09552717e-01
-1.18505979e+00 4.43510681e-01 -9.75600541e-01 -5.11684060e-01
4.86691669e-02 5.87660015e-01 -3.35212320e-01 1.79543436e-01
-7.24393904e-01 6.13055468e-01 3.72575760e-01 -3.09026212e-01
-1.11123240e+00 -5.69614410e-01 -6.47540331e-01 -1.83701485e-01
-1.68380171e-01 -7.35946238e-01 1.34983158e+00 -1.02661633e+00
-1.87093544e+00 1.14736772e+00 -6.74831033e-01 -7.51962304e-01
4.15912360e-01 8.93037650e-04 -2.16641999e-03 5.22737622e-01
-1.18618384e-01 9.39304709e-01 8.13765109e-01 -8.65857959e-01
-1.22117221e+00 -2.22201943e-01 1.72905967e-01 1.90133438e-01
-1.76325828e-01 1.43641904e-01 -4.56440121e-01 -7.91286379e-02
-4.26808782e-02 -5.19659400e-01 -2.25176707e-01 4.71597522e-01
1.89882413e-01 1.59774095e-01 6.84793949e-01 -4.49329279e-02
8.34293187e-01 -2.10614371e+00 -1.94669381e-01 -4.21206683e-01
4.68648821e-02 -1.39737785e-01 1.02256261e-01 3.82479042e-01
2.66843021e-01 -6.19130671e-01 -2.43131146e-01 -3.57340962e-01
-8.09179962e-01 1.95370004e-01 -4.52924132e-01 6.14620388e-01
6.18728101e-02 2.04386801e-01 -1.05966914e+00 -8.14853013e-01
8.65641415e-01 5.92633486e-01 -2.56591767e-01 3.68232906e-01
-6.90327436e-02 5.19525886e-01 -1.44618794e-01 6.04294956e-01
7.17312694e-01 -8.30925629e-02 -6.74042553e-02 -1.86053932e-01
-7.00317502e-01 -1.54787451e-01 -1.27007771e+00 1.55305183e+00
-4.90160733e-01 1.22038579e+00 -1.86367914e-01 -5.20073593e-01
8.78071666e-01 5.36472797e-01 6.58464611e-01 -7.29146481e-01
2.41803691e-01 1.63188636e-01 -5.08457303e-01 -2.70546615e-01
7.73765743e-01 -4.91218753e-02 1.05811276e-01 2.87291497e-01
-1.41980378e-02 -3.35454553e-01 4.80369419e-01 9.16364342e-02
1.46682262e+00 1.10013463e-01 4.32167083e-01 5.09977490e-02
5.71334064e-01 -2.29987383e-01 7.13272154e-01 6.79314256e-01
-4.23855841e-01 8.42399180e-01 7.26390332e-02 -5.21651208e-01
-1.24417818e+00 -1.39767385e+00 -2.23284230e-01 5.08488774e-01
9.13864970e-01 -1.61518753e-01 -6.80439711e-01 1.57150552e-01
-2.81747222e-01 3.43212157e-01 -2.86678821e-01 1.91778317e-01
-6.43684566e-01 -6.56278491e-01 3.81132662e-01 4.97691184e-01
4.61304367e-01 -1.03588808e+00 -1.77007484e+00 7.27705538e-01
-4.15614136e-02 -1.48660946e+00 -1.40428334e-01 3.45993221e-01
-1.18351495e+00 -1.16465509e+00 -6.78572536e-01 -5.64029694e-01
6.84336782e-01 4.21808094e-01 1.03823626e+00 -3.21740508e-01
-7.10124910e-01 8.26414108e-01 -5.80711178e-02 -4.41987157e-01
-1.64994642e-01 -8.07814896e-01 -1.81051530e-02 1.97927937e-01
3.01623851e-01 -8.74629796e-01 -1.01396775e+00 2.28639185e-01
-1.15309548e+00 3.83063972e-01 2.31093541e-01 5.23843706e-01
8.27928841e-01 -2.41399378e-01 7.96388648e-03 -3.66560489e-01
3.88626568e-02 -7.76899830e-02 -1.26705027e+00 2.62137763e-02
-8.97605419e-02 -2.05408245e-01 6.69215679e-01 -6.01338685e-01
-1.46266055e+00 6.36686921e-01 4.26073164e-01 -3.27726841e-01
-3.27585638e-01 -2.86823243e-01 3.68679345e-01 -9.16697159e-02
5.61949492e-01 3.09241354e-01 -2.92799711e-01 7.75950700e-02
-7.83253461e-02 5.14484763e-01 1.04726100e+00 -3.25975984e-01
1.48284256e-01 1.31285846e+00 9.22609493e-02 -9.56334233e-01
-3.20141166e-02 -6.61902487e-01 -2.87104011e-01 -7.84926593e-01
8.75078797e-01 -1.09804916e+00 -9.06245768e-01 1.11910999e+00
-1.63680327e+00 -2.98670143e-01 -3.73310238e-01 5.94624758e-01
-9.24013615e-01 2.81342149e-01 -8.58262062e-01 -1.09357738e+00
-1.82924241e-01 -9.29018557e-01 1.39417994e+00 9.66862500e-01
-1.61193572e-02 -9.09500599e-01 2.04225421e-01 -8.59409869e-02
1.28099188e-01 3.78380150e-01 -5.93020916e-02 4.23088104e-01
-1.19882405e+00 -1.99662909e-01 -1.89737484e-01 -1.01280905e-01
1.43312722e-01 3.76714975e-01 -1.33516324e+00 -1.40844777e-01
1.93171561e-01 -5.68965673e-02 7.22950876e-01 8.28314960e-01
1.08362663e+00 4.03237581e-01 -5.60166001e-01 5.86988688e-01
1.96105039e+00 5.26776731e-01 1.11351919e+00 2.31598228e-01
6.63484111e-02 3.81348252e-01 6.29413486e-01 9.44194257e-01
3.54330167e-02 5.22027731e-01 5.03369391e-01 2.20445488e-02
-1.82374194e-01 1.52895497e-02 7.77454019e-01 3.43325347e-01
-3.42698961e-01 -6.60739541e-01 -4.72652107e-01 7.74854541e-01
-1.73209667e+00 -1.34660006e+00 -4.06547785e-01 2.75981331e+00
8.02340508e-01 1.61064900e-02 -1.65610805e-01 1.43746749e-01
1.15562963e+00 -4.34793569e-02 -6.38977110e-01 -3.02555561e-01
-4.90619779e-01 3.41374993e-01 8.78425002e-01 3.71782184e-01
-7.54635096e-01 7.15267837e-01 6.61842728e+00 3.04771721e-01
-1.21677327e+00 -2.79329587e-02 5.28914988e-01 -3.72289389e-01
-1.09697171e-01 2.94833541e-01 -1.00468397e+00 7.87842095e-01
1.09509516e+00 -4.06211108e-01 2.74090022e-01 2.48773009e-01
4.75044072e-01 -1.13809288e+00 -1.34760702e+00 1.44652629e+00
-1.83131203e-01 -1.29757845e+00 -1.40593946e-01 1.71609074e-02
7.82465100e-01 -7.81552568e-02 -3.58186096e-01 -7.31131136e-01
2.35257283e-01 -4.30992007e-01 7.45390177e-01 7.40107417e-01
8.37791622e-01 -4.09174770e-01 3.30300212e-01 2.45974004e-01
-1.35996771e+00 -8.72418433e-02 -4.78350997e-01 -3.95375878e-01
7.07590640e-01 8.94845724e-01 -6.14006221e-01 -4.66575883e-02
8.63484263e-01 6.78379416e-01 -1.45282045e-01 1.39564645e+00
8.88515785e-02 1.78858146e-01 -7.22159147e-01 4.59109992e-02
-2.79025108e-01 -9.78157073e-02 6.45387828e-01 1.11259794e+00
7.67101288e-01 3.15704048e-01 -3.83879006e-01 6.82489455e-01
2.01094419e-01 -4.87540603e-01 -7.90369451e-01 1.85140267e-01
8.14584792e-01 1.13682234e+00 -1.00858939e+00 -5.69128275e-01
-6.06505692e-01 1.59822392e+00 -2.29725949e-02 2.61879653e-01
-9.52521801e-01 -5.31829417e-01 5.12379527e-01 6.06692806e-02
1.92184463e-01 -1.78570926e-01 4.06799391e-02 -1.10453331e+00
6.76328391e-02 -2.06317216e-01 2.35283613e-01 -1.13839388e+00
-9.32513297e-01 2.01976001e-01 -6.78825751e-02 -1.47177649e+00
-4.03130442e-01 -6.15609169e-01 -6.92986727e-01 5.06165922e-01
-1.46510255e+00 -6.77717447e-01 -6.70027733e-01 9.32057977e-01
8.10260892e-01 3.80437940e-01 5.12670279e-01 1.68265924e-01
-1.59235984e-01 -5.50156794e-02 2.82126546e-01 -3.47905189e-01
7.43634284e-01 -1.12369263e+00 1.91181019e-01 1.19314945e+00
-6.75846860e-02 1.00586116e-01 7.83211708e-01 -4.96001959e-01
-1.64127350e+00 -6.29182994e-01 7.34870434e-01 -2.97747016e-01
4.43874627e-01 -1.75984085e-01 -5.34203827e-01 4.32012528e-01
2.85702974e-01 9.88173038e-02 2.25151703e-01 -8.70686829e-01
3.23239535e-01 -4.44099575e-01 -1.16540158e+00 4.52045858e-01
1.00792408e+00 -6.79364979e-01 -2.55714864e-01 2.80123204e-02
2.58446276e-01 -5.28971016e-01 -5.47420382e-01 2.01552629e-01
7.72616506e-01 -1.69308639e+00 8.82906199e-01 4.70022023e-01
2.77939349e-01 -6.59739673e-01 1.88822392e-02 -5.69592059e-01
4.08081055e-01 -8.63110602e-01 -2.19946671e-02 1.20031118e+00
-1.52172074e-01 -7.02398837e-01 7.71633446e-01 6.82026565e-01
1.21218590e-02 9.02303774e-03 -1.28942490e+00 -8.10554206e-01
-8.99550915e-01 -3.18998396e-01 1.27484694e-01 3.08240205e-01
-3.28384116e-02 -1.21911585e-01 -1.65398702e-01 2.69675016e-01
1.07988071e+00 3.43206450e-02 6.59435868e-01 -1.18014550e+00
-1.39001042e-01 -5.83174527e-02 -9.16202188e-01 -1.23458278e+00
-2.51678705e-01 9.42259803e-02 4.94313329e-01 -1.07160830e+00
4.42146510e-01 1.95393920e-01 -1.96492836e-01 -2.72215366e-01
-1.07077584e-01 4.52116072e-01 -1.72667339e-01 4.81463857e-02
-4.88740325e-01 2.87241712e-02 7.06511736e-01 3.07477802e-01
-1.67424694e-01 -2.84036309e-01 4.24509138e-01 8.16615522e-01
3.40313703e-01 -2.63161451e-01 -4.55557644e-01 -4.79432166e-01
2.01961905e-01 6.26154661e-01 7.07250476e-01 -1.58041394e+00
8.36127043e-01 -8.27201977e-02 4.43933904e-01 -5.60136080e-01
6.69127703e-01 -8.01239610e-01 6.83073819e-01 3.13076168e-01
-6.18295297e-02 2.51476943e-01 2.79095799e-01 8.96340013e-01
-2.80383497e-01 -3.47569078e-01 1.09497273e+00 -3.27451319e-01
-1.25250089e+00 1.83821499e-01 -9.96038735e-01 -3.04370761e-01
1.38918412e+00 -1.02946675e+00 -2.37416789e-01 -3.31470102e-01
-2.95341969e-01 -2.74248660e-01 1.06427729e+00 -1.15264110e-01
1.01327717e+00 -8.54659379e-01 -1.79546878e-01 1.89861864e-01
-4.07008827e-02 -5.11648320e-02 3.50934952e-01 6.42890155e-01
-8.56173038e-01 -3.83514427e-02 -5.47366440e-01 -9.79311883e-01
-1.49314427e+00 2.69979119e-01 3.50219905e-01 2.59940326e-01
-5.41460037e-01 9.76731718e-01 4.36961204e-01 6.69422269e-01
1.57316491e-01 -4.09552455e-01 1.08530059e-01 -1.52452692e-01
8.43092084e-01 4.33297306e-01 -2.96723753e-01 -2.59036988e-01
-3.22175443e-01 9.54809606e-01 2.25501150e-01 -6.13892317e-01
9.73384559e-01 -6.24071300e-01 -6.84557930e-02 6.28215671e-01
9.04241025e-01 -7.74138346e-02 -1.90151107e+00 1.37608841e-01
-2.81768411e-01 -7.16968656e-01 1.33699164e-01 -1.61960691e-01
-7.56968439e-01 9.23427284e-01 1.03614128e+00 3.66332829e-02
1.64721107e+00 -1.82763472e-01 1.05755377e+00 9.55035985e-02
6.42865062e-01 -1.08134878e+00 9.51057598e-02 6.26443774e-02
1.76025763e-01 -1.13303137e+00 -1.34852111e-01 -4.19451356e-01
-6.20894693e-02 1.45195186e+00 5.89334190e-01 -2.18187317e-01
6.08166456e-02 8.13238204e-01 -1.16163708e-01 5.94339855e-02
-1.07803285e+00 -2.55354285e-01 -4.33987886e-01 6.76207364e-01
3.81104797e-01 -3.07657510e-01 -3.47184598e-01 -3.29553813e-01
4.24771458e-01 5.77077687e-01 9.56496954e-01 1.12499774e+00
-6.96644366e-01 -9.54975963e-01 -5.08157969e-01 6.32224679e-02
-3.38821471e-01 -1.20202698e-01 -3.60038951e-02 3.71493727e-01
4.41806875e-02 9.15638328e-01 7.57535279e-01 -5.75506638e-05
1.01206668e-01 -2.09455207e-01 9.83841181e-01 -2.07233414e-01
-2.19115466e-01 -1.53633505e-01 -6.66473657e-02 -9.16435122e-01
-1.02604711e+00 -8.38677287e-01 -1.56319451e+00 -2.80211180e-01
-3.06275696e-01 -2.25776851e-01 7.28399456e-01 5.21227837e-01
3.58417660e-01 1.01657674e-01 6.25185728e-01 -1.13444066e+00
2.87744492e-01 -3.37281853e-01 -5.57541251e-01 4.44736600e-01
4.93292004e-01 -3.66493344e-01 -6.71708524e-01 1.01646149e+00] | [8.683606147766113, -1.3478641510009766] |
bca0da62-f8fb-480c-8b9c-bd008c157cbc | deep-reinforcement-learning-based-vehicle | 2304.02832 | null | https://arxiv.org/abs/2304.02832v1 | https://arxiv.org/pdf/2304.02832v1.pdf | Deep Reinforcement Learning Based Vehicle Selection for Asynchronous Federated Learning Enabled Vehicular Edge Computing | In the traditional vehicular network, computing tasks generated by the vehicles are usually uploaded to the cloud for processing. However, since task offloading toward the cloud will cause a large delay, vehicular edge computing (VEC) is introduced to avoid such a problem and improve the whole system performance, where a roadside unit (RSU) with certain computing capability is used to process the data of vehicles as an edge entity. Owing to the privacy and security issues, vehicles are reluctant to upload local data directly to the RSU, and thus federated learning (FL) becomes a promising technology for some machine learning tasks in VEC, where vehicles only need to upload the local model hyperparameters instead of transferring their local data to the nearby RSU. Furthermore, as vehicles have different local training time due to various sizes of local data and their different computing capabilities, asynchronous federated learning (AFL) is employed to facilitate the RSU to update the global model immediately after receiving a local model to reduce the aggregation delay. However, in AFL of VEC, different vehicles may have different impact on the global model updating because of their various local training delay, transmission delay and local data sizes. Also, if there are bad nodes among the vehicles, it will affect the global aggregation quality at the RSU. To solve the above problem, we shall propose a deep reinforcement learning (DRL) based vehicle selection scheme to improve the accuracy of the global model in AFL of vehicular network. In the scheme, we present the model including the state, action and reward in the DRL based to the specific problem. Simulation results demonstrate our scheme can effectively remove the bad nodes and improve the aggregation accuracy of the global model. | ['Qiang Fan', 'Pingyi Fan', 'Siyuan Wang', 'Qiong Wu'] | 2023-04-06 | null | null | null | null | ['edge-computing'] | ['time-series'] | [-7.30473697e-01 -6.73302338e-02 -2.66757816e-01 -1.85275421e-01
-1.43305466e-01 -2.70524234e-01 7.37651885e-02 -1.39648125e-01
-3.07417423e-01 7.76277423e-01 -4.04530436e-01 -4.21993315e-01
-1.00300103e-01 -1.03492880e+00 -6.59971356e-01 -9.01789844e-01
-3.36125523e-01 2.57041723e-01 4.45910990e-01 -7.50507042e-02
-3.05070966e-01 3.91642690e-01 -1.27443171e+00 -1.11692593e-01
1.27975345e+00 1.23237586e+00 5.24938703e-01 1.72990099e-01
-3.42776895e-01 6.32881343e-01 -5.04633844e-01 -1.02289826e-01
3.22549164e-01 8.20671842e-02 -3.02481115e-01 -7.62259364e-02
-4.36321050e-01 -5.72367966e-01 -6.83041692e-01 9.90130365e-01
4.06951338e-01 5.55759249e-03 2.26194471e-01 -2.09543681e+00
-2.64568925e-01 3.31482470e-01 -5.12506902e-01 8.62405747e-02
-2.22898796e-01 1.70560881e-01 2.21346557e-01 -7.47001290e-01
5.79133034e-01 8.85312438e-01 4.96042401e-01 4.60114270e-01
-4.98515159e-01 -1.04537392e+00 6.57568514e-01 7.35594928e-01
-1.43915451e+00 -5.18079400e-01 7.17336416e-01 -1.42282352e-01
4.53709781e-01 5.21337427e-02 7.85613298e-01 3.92322123e-01
3.28017086e-01 9.26768541e-01 7.49077916e-01 2.21138328e-01
4.97500658e-01 2.63630122e-01 -1.78839579e-01 2.73264259e-01
3.04372847e-01 7.50521868e-02 -2.29586467e-01 1.15964767e-02
4.90898430e-01 2.11414576e-01 -1.80094048e-01 -3.73167664e-01
-7.41449773e-01 4.90250200e-01 7.23573923e-01 -1.59556285e-01
-7.78869390e-01 4.38567817e-01 8.06683242e-01 4.55731988e-01
5.74935973e-01 -5.81497192e-01 -6.36409879e-01 1.66774727e-02
-5.51235676e-01 -8.75338241e-02 5.33632457e-01 1.21516943e+00
1.19662094e+00 3.97092283e-01 -8.88459682e-02 3.24018806e-01
2.80319870e-01 5.26547551e-01 1.70772567e-01 -9.48037326e-01
6.35785818e-01 5.96087933e-01 2.43813425e-01 -1.05771840e+00
-1.98696762e-01 -3.55674267e-01 -1.04328096e+00 1.57344282e-01
-2.37107009e-01 -8.53814602e-01 -7.10932612e-01 1.60222304e+00
7.73607850e-01 8.92928421e-01 1.71148017e-01 1.21836984e+00
6.30028486e-01 1.03264010e+00 3.52067411e-01 -6.20976746e-01
8.34714770e-01 -7.51376748e-01 -8.13906729e-01 5.83626982e-03
8.92277181e-01 -5.70917547e-01 2.68038511e-01 -9.00719687e-02
-8.80895793e-01 -4.50232685e-01 -8.11340511e-01 4.57238495e-01
-2.98216343e-01 2.23407865e-01 7.26573884e-01 4.08066899e-01
-1.12347460e+00 5.58734387e-02 -6.12186909e-01 -1.01512499e-01
5.62653065e-01 6.69144809e-01 -6.18587770e-02 -3.77499342e-01
-1.53568327e+00 5.13908207e-01 2.29478806e-01 4.29098815e-01
-1.04707897e+00 -8.19933891e-01 -5.51297069e-01 9.28519443e-02
7.17676163e-01 -4.85320419e-01 9.20616329e-01 -9.95337248e-01
-1.17175794e+00 -6.35302216e-02 -9.40157771e-02 -3.09083670e-01
6.62813783e-01 3.91264200e-01 -7.27858186e-01 -1.42510131e-01
9.25739668e-03 3.76163423e-01 6.62922382e-01 -1.17250919e+00
-1.01738441e+00 -2.79437870e-01 1.50711387e-01 3.28291535e-01
-3.86585504e-01 -2.73270369e-01 -6.88110173e-01 -5.92976213e-02
-2.10135147e-01 -9.37979460e-01 -4.13369119e-01 -1.61963224e-01
1.30192965e-01 -5.73141515e-01 1.70917618e+00 -4.76592094e-01
1.15384877e+00 -2.22491932e+00 -2.11352557e-01 3.13815475e-01
3.88008267e-01 5.15946507e-01 -1.92959994e-01 2.91610420e-01
3.63041013e-01 1.29154786e-01 3.26717198e-01 -1.34497033e-02
-2.91749984e-01 4.89931822e-01 3.50194126e-02 3.45998794e-01
-2.08427280e-01 7.22055018e-01 -9.20588732e-01 -6.87853932e-01
3.15808445e-01 4.08688694e-01 -2.27534443e-01 1.47825494e-01
-1.49764623e-02 2.82874882e-01 -1.03928435e+00 3.69342625e-01
1.28980565e+00 -3.69326919e-02 1.46490216e-01 -2.24869266e-01
-1.78774655e-01 -3.27396333e-01 -1.31748676e+00 1.05285907e+00
-8.09859216e-01 3.81163567e-01 8.19414556e-01 -1.18761003e+00
7.85152256e-01 5.57357371e-01 8.84761453e-01 -8.12805355e-01
2.17128262e-01 1.81183085e-01 -1.06434673e-01 -6.58405244e-01
2.81785816e-01 5.09929955e-01 8.97405669e-02 3.90429169e-01
-3.75254750e-01 6.78472102e-01 -4.22253869e-02 5.28977394e-01
1.01449049e+00 -1.80166498e-01 -5.60778320e-01 1.94043107e-02
7.05515623e-01 -6.92207590e-02 1.21670294e+00 1.31514966e-01
-4.48982835e-01 -3.09332371e-01 2.48267666e-01 -5.09042144e-01
-8.73759985e-01 -7.38282382e-01 2.67313182e-01 8.57627809e-01
8.41505766e-01 -1.90729514e-01 -6.08669758e-01 -5.84399641e-01
1.31916717e-01 3.52278113e-01 -1.70542061e-01 -6.39078796e-01
-4.87359494e-01 -5.11743367e-01 8.61183926e-02 3.20301086e-01
9.19499397e-01 -8.38877439e-01 -1.35321438e-01 4.89080727e-01
-5.85151352e-02 -1.20687580e+00 -4.98282701e-01 -3.51145506e-01
-6.04028106e-01 -9.37752604e-01 -2.86501974e-01 -6.66377783e-01
8.75943661e-01 9.43107247e-01 3.64025205e-01 5.43418169e-01
3.33901137e-01 1.28382862e-01 -3.79698455e-01 -4.82355833e-01
-2.38141045e-01 1.36428699e-01 2.60748923e-01 4.59597021e-01
2.63294727e-01 -2.93178260e-01 -7.42089152e-01 4.92624819e-01
-8.38688791e-01 4.43020426e-02 2.41737425e-01 5.20493984e-01
4.56284106e-01 5.37548721e-01 1.21849048e+00 -6.23581529e-01
5.90936959e-01 -1.08674991e+00 -8.31053078e-01 2.38568291e-01
-7.14759588e-01 -5.75527310e-01 1.10377073e+00 -3.17327410e-01
-9.80826795e-01 1.06910020e-01 4.36189771e-01 -7.87604809e-01
3.94339651e-01 5.77655315e-01 -4.68355715e-01 -2.78556138e-01
-8.64346921e-02 1.04518868e-01 2.60282487e-01 -5.50986975e-02
8.28852430e-02 9.03670609e-01 -5.38536683e-02 -1.49234831e-01
8.19306314e-01 2.13983551e-01 1.30395770e-01 -3.56389165e-01
1.71169862e-01 -1.21014684e-01 8.36316869e-02 -6.94168568e-01
5.73334157e-01 -1.45531833e+00 -1.17304409e+00 5.75093746e-01
-1.20716417e+00 -3.21141154e-01 3.09413403e-01 7.05360472e-01
-1.19809046e-01 1.00389622e-01 -5.21011591e-01 -9.18121338e-01
-3.41431946e-01 -1.39417565e+00 5.70997417e-01 6.51654124e-01
6.59415662e-01 -8.67877126e-01 -6.16850495e-01 1.08116604e-02
6.81737244e-01 -1.19415678e-01 6.96619928e-01 -7.30756223e-02
-1.12112474e+00 -1.93633676e-01 -4.65083718e-01 1.33060321e-01
2.43065767e-02 1.63987488e-01 -5.55998623e-01 -5.99815905e-01
-3.78002763e-01 7.78866857e-02 3.44315946e-01 3.92299473e-01
1.34703243e+00 -3.09024096e-01 -7.23677754e-01 5.56135178e-01
1.48227119e+00 2.82893628e-01 5.64573586e-01 1.15969226e-01
8.68300736e-01 4.09017891e-01 1.00650179e+00 4.61229950e-01
9.37459767e-01 4.22565758e-01 8.93042624e-01 -1.17537111e-01
9.32709966e-03 -2.79202580e-01 4.95617181e-01 8.34746540e-01
-6.34468570e-02 -3.40111732e-01 -4.62881774e-01 5.48361778e-01
-2.23101926e+00 -8.20927262e-01 -3.20400715e-01 2.34579182e+00
2.77061075e-01 -1.19474769e-01 -1.14312910e-01 -2.27097347e-01
1.03733647e+00 1.55427894e-02 -7.70972908e-01 -4.77839023e-01
2.17168570e-01 -5.57169139e-01 9.26955938e-01 1.91135257e-01
-5.29614806e-01 9.46347713e-01 4.78679705e+00 1.19555068e+00
-1.49526811e+00 3.78885508e-01 7.21690416e-01 -4.00144756e-02
-3.26933265e-01 1.95649520e-01 -5.85839093e-01 1.11356127e+00
9.15550709e-01 -5.31207800e-01 7.50065267e-01 1.09943449e+00
9.19345915e-01 -8.13817233e-02 -4.56453711e-01 9.71417487e-01
-6.55483902e-01 -1.44724739e+00 -2.39775941e-01 1.98093057e-01
9.58580315e-01 5.35507441e-01 -2.79152036e-01 4.84984159e-01
2.63119519e-01 -6.09432399e-01 2.86286771e-01 4.57732886e-01
8.57746243e-01 -1.19558454e+00 7.76661396e-01 6.49120748e-01
-1.63986814e+00 -3.83851677e-01 -7.48192012e-01 6.98923394e-02
2.27981701e-01 7.58262932e-01 -4.64878261e-01 7.53759682e-01
7.27445602e-01 5.81767976e-01 -1.84231207e-01 9.26902890e-01
2.10058257e-01 3.89065504e-01 -3.80066633e-01 -1.05140090e-01
1.48750097e-01 -4.94068533e-01 5.80507457e-01 3.83032322e-01
4.15723175e-01 2.26657897e-01 5.87508738e-01 4.46842104e-01
-4.30805862e-01 2.18335181e-01 -5.07916152e-01 3.77751261e-01
1.05988181e+00 1.64366078e+00 -2.97021627e-01 -4.80289191e-01
-4.98064339e-01 5.69880307e-01 2.75368452e-01 7.68066943e-01
-8.91037107e-01 -4.64169443e-01 8.81894052e-01 1.16880424e-01
1.00579239e-01 -2.28177011e-01 -7.22105578e-02 -8.26796889e-01
2.08454460e-01 -3.27017486e-01 4.82869480e-04 -7.67111242e-01
-9.32942152e-01 5.14002681e-01 -4.58767086e-01 -1.48158300e+00
2.05411524e-01 1.67761326e-01 -1.12989819e+00 8.71605754e-01
-1.88659036e+00 -1.00449789e+00 -6.07792616e-01 9.51423287e-01
2.77489930e-01 -3.06564867e-01 1.33167014e-01 8.01245451e-01
-8.46323907e-01 5.84915757e-01 3.02278042e-01 -1.28261700e-01
5.11533141e-01 -4.79900688e-01 -1.40381575e-01 8.09009492e-01
-7.52733111e-01 9.63085219e-02 3.41150880e-01 -8.18468273e-01
-1.84867108e+00 -1.71736729e+00 5.23000181e-01 2.71188289e-01
3.84264559e-01 -1.11564271e-01 -7.61921048e-01 5.30752182e-01
-3.40754017e-02 6.97402775e-01 1.96355447e-01 -3.63954455e-01
4.73410159e-01 -8.17681968e-01 -1.18542981e+00 4.62771177e-01
7.88108766e-01 -2.58309662e-01 5.38009584e-01 4.37544495e-01
8.64445865e-01 -2.66281635e-01 -6.96609259e-01 3.53228569e-01
7.85612762e-02 -4.09288228e-01 3.42033118e-01 -5.82635581e-01
-2.32596040e-01 -7.45256305e-01 1.71352535e-01 -1.72915709e+00
-2.45602772e-01 -5.61479270e-01 -6.84161857e-02 1.30464947e+00
2.06389502e-01 -1.02050352e+00 8.92993152e-01 8.34361732e-01
-4.22381282e-01 -7.38047302e-01 -1.25099742e+00 -5.95442593e-01
-2.96055794e-01 -4.50288266e-01 1.26622105e+00 7.79398322e-01
-1.85825512e-01 1.79635227e-01 -4.88745779e-01 5.50010085e-01
5.12327492e-01 1.92628857e-02 1.04948461e+00 -1.14925647e+00
4.16346937e-01 1.41007289e-01 -2.86827236e-01 -1.02386618e+00
3.72566462e-01 -9.34556484e-01 -1.38195246e-01 -1.72303486e+00
-2.46401012e-01 -1.37252569e+00 -4.00178671e-01 3.75261664e-01
-4.18262109e-02 -1.59613773e-01 3.39913994e-01 3.25031877e-01
-9.45090294e-01 6.45566165e-01 1.51978409e+00 -6.98808208e-02
-3.21574330e-01 2.64990240e-01 -3.63709033e-01 3.39774609e-01
9.52836812e-01 -4.39040124e-01 -7.44200408e-01 -7.28713155e-01
1.47615477e-01 6.29589558e-01 2.85162389e-01 -8.31830800e-01
6.99052453e-01 -3.86442065e-01 3.07200700e-01 -6.73483789e-01
9.50824916e-02 -1.44891894e+00 4.50700849e-01 4.91729021e-01
4.01550919e-01 1.56498794e-02 -1.15468882e-01 7.52835751e-01
-2.01096348e-02 3.12358737e-01 3.33355278e-01 2.57330239e-01
-9.96739686e-01 1.17890763e+00 -5.88357866e-01 -4.04672652e-01
1.71353900e+00 -1.90550342e-01 -3.58885050e-01 -5.16139507e-01
-4.87655371e-01 1.36100054e+00 3.71635318e-01 4.83461231e-01
5.51477730e-01 -1.51263571e+00 -5.86804807e-01 2.43956149e-01
-1.98251829e-01 4.88781594e-02 9.00036991e-01 1.12299287e+00
-1.78591236e-01 1.38802677e-01 -1.43956900e-01 -4.60594207e-01
-1.06767440e+00 7.35995352e-01 4.09491688e-01 1.86270401e-01
-1.37441233e-01 2.41651252e-01 -2.14669406e-02 -3.30036432e-01
-1.16705820e-01 2.46032804e-01 -1.57104433e-01 -9.85124707e-02
2.35162005e-01 7.34280527e-01 1.35120362e-01 -7.93834329e-01
-2.84977138e-01 1.14768706e-01 1.67231467e-02 6.04810357e-01
9.44806099e-01 -7.32652366e-01 -1.78921714e-01 -2.30376020e-01
1.30910826e+00 -2.37247080e-01 -1.33310843e+00 -3.24758768e-01
-7.69955158e-01 -5.62370777e-01 5.51289797e-01 -2.05893412e-01
-1.95275283e+00 5.07413268e-01 5.18760443e-01 2.43192762e-01
1.12378466e+00 -3.33741128e-01 1.26896620e+00 -3.87612283e-02
8.55640292e-01 -1.60604632e+00 -5.09192526e-01 3.63217771e-01
1.31428435e-01 -1.13702273e+00 -1.57877371e-01 -5.90164185e-01
-7.24022925e-01 8.78640056e-01 1.16724157e+00 -7.65206814e-02
9.41254616e-01 1.56696215e-01 1.32746557e-02 -1.33071840e-02
-9.08513069e-01 4.69045676e-02 -6.30149007e-01 7.99207628e-01
-4.44215447e-01 2.56968170e-01 -5.63816547e-01 5.11584580e-01
2.91556835e-01 2.58813918e-01 7.44417250e-01 7.17802405e-01
-3.76543611e-01 -1.10197866e+00 -8.65064189e-02 7.17603326e-01
-1.16135530e-01 4.07168925e-01 4.48287845e-01 3.97007406e-01
3.93025279e-01 1.19719326e+00 2.06988961e-01 -5.09330750e-01
-4.35233722e-03 -3.95820200e-01 -2.45945215e-01 -8.80386084e-02
-8.12590644e-02 -1.05916010e-02 -9.23067927e-02 -4.72850680e-01
-2.48625711e-01 -2.66749501e-01 -1.83272195e+00 -9.04809773e-01
-4.81730402e-01 5.40559351e-01 8.12733233e-01 9.32844698e-01
8.35103989e-01 4.79084402e-01 1.66628098e+00 -5.98831236e-01
-8.98320526e-02 -4.94592816e-01 -7.22700417e-01 -5.22540994e-02
2.10973322e-01 -6.39076948e-01 -1.58314914e-01 -4.90698159e-01] | [5.664848327636719, 1.487156867980957] |
e4b9a90d-cbe7-4f76-bc46-ac9f3e16b5d4 | productive-reproducible-workflows-for-dnns-a | 2206.09359 | null | https://arxiv.org/abs/2206.09359v1 | https://arxiv.org/pdf/2206.09359v1.pdf | Productive Reproducible Workflows for DNNs: A Case Study for Industrial Defect Detection | As Deep Neural Networks (DNNs) have become an increasingly ubiquitous workload, the range of libraries and tooling available to aid in their development and deployment has grown significantly. Scalable, production quality tools are freely available under permissive licenses, and are accessible enough to enable even small teams to be very productive. However within the research community, awareness and usage of said tools is not necessarily widespread, and researchers may be missing out on potential productivity gains from exploiting the latest tools and workflows. This paper presents a case study where we discuss our recent experience producing an end-to-end artificial intelligence application for industrial defect detection. We detail the high level deep learning libraries, containerized workflows, continuous integration/deployment pipelines, and open source code templates we leveraged to produce a competitive result, matching the performance of other ranked solutions to our three target datasets. We highlight the value that exploiting such systems can bring, even for research, and detail our solution and present our best results in terms of accuracy and inference time on a server class GPU, as well as inference times on a server class CPU, and a Raspberry Pi 4. | ['José Cano', 'Perry Gibson'] | 2022-06-19 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [-2.13466778e-01 -1.81363106e-01 2.90885717e-01 -4.67540950e-01
-4.41594422e-01 -5.77625930e-01 3.84471156e-02 6.88718911e-03
-3.76309305e-01 4.82069284e-01 -3.65080118e-01 -4.59080786e-01
-3.05083632e-01 -8.16347361e-01 -6.52497411e-01 -5.31784534e-01
5.71949892e-02 8.08654130e-01 2.36100569e-01 4.06239182e-02
1.98953807e-01 7.13522553e-01 -1.95974553e+00 2.56825954e-01
4.42049354e-01 1.19160080e+00 3.20445895e-01 6.56270444e-01
-1.00496158e-01 9.46123302e-01 -8.62313151e-01 -3.11686605e-01
3.46190810e-01 1.92387715e-01 -6.74216568e-01 -1.94028497e-01
5.32764375e-01 -6.22195125e-01 6.15189560e-02 7.47438252e-01
8.96516919e-01 -1.71164513e-01 -1.01070650e-01 -1.38241816e+00
-3.24733526e-01 1.88926667e-01 -1.56619534e-01 3.25578958e-01
1.75279424e-01 4.34997559e-01 5.67215204e-01 -4.66930717e-01
5.29160976e-01 7.78419077e-01 1.03705502e+00 3.47599417e-01
-1.01179540e+00 -7.67128885e-01 -1.89783484e-01 8.15858021e-02
-9.82372701e-01 -6.08900905e-01 4.12009209e-01 -5.18438995e-01
1.63254130e+00 1.11962138e-02 8.55074525e-01 1.23883772e+00
1.20336138e-01 6.72838390e-01 6.48216963e-01 -2.69558012e-01
4.28839475e-01 9.87303331e-02 3.15666735e-01 8.61310840e-01
5.06995082e-01 -1.37272716e-01 -6.02569520e-01 -2.12734923e-01
1.00370228e+00 2.47102961e-01 1.05818421e-01 -2.15796292e-01
-1.21455348e+00 5.01692593e-01 6.27361014e-02 2.88768351e-01
-5.21381974e-01 1.93564072e-01 7.83054769e-01 3.21314245e-01
4.87926483e-01 4.68943775e-01 -9.72259760e-01 -7.30399311e-01
-7.80247927e-01 5.51012695e-01 1.17745042e+00 1.08809090e+00
5.37971079e-01 7.04666078e-02 2.54617661e-01 8.93017352e-01
8.83940831e-02 -1.47256479e-02 4.69592273e-01 -1.48701024e+00
2.35511921e-02 6.90855265e-01 2.20388658e-02 -8.32800388e-01
-5.76197863e-01 -5.49401999e-01 -5.84034145e-01 6.46267295e-01
4.69810367e-01 -4.98129815e-01 -5.52673459e-01 1.09864271e+00
2.48731941e-01 -3.20833661e-02 -2.95164555e-01 8.05702984e-01
7.97177196e-01 2.51753807e-01 -1.38841748e-01 4.04553771e-01
1.24202836e+00 -9.25458968e-01 -4.57275927e-01 -2.50165671e-01
7.03719020e-01 -7.25405574e-01 1.00297892e+00 9.48117077e-01
-1.16970038e+00 -6.31662309e-01 -1.12990201e+00 -2.88415194e-01
-5.89477003e-01 -1.07747443e-01 1.08860183e+00 8.13139200e-01
-1.21799111e+00 1.07229459e+00 -1.45032644e+00 -5.63604772e-01
6.82343066e-01 4.98800248e-01 -3.16093743e-01 -1.12171680e-01
-5.43453753e-01 9.12800312e-01 2.51208603e-01 8.01878721e-02
-6.27499104e-01 -8.52346480e-01 -2.72024602e-01 -8.09129328e-02
2.56753802e-01 -8.79510403e-01 1.54145634e+00 -7.55429864e-01
-1.36064827e+00 9.48109984e-01 2.71369517e-01 -2.87148863e-01
4.53929573e-01 -1.16990089e-01 -2.59038627e-01 -1.69640079e-01
-9.20039136e-03 6.46713912e-01 3.51639003e-01 -5.63307643e-01
-1.08254075e+00 -5.90078533e-01 6.16130978e-02 -2.81422943e-01
-4.54176366e-01 3.25111389e-01 -2.99213767e-01 -1.32133722e-01
-1.84361234e-01 -6.12873137e-01 -9.95217040e-02 2.96108633e-01
4.07017730e-02 -1.77458584e-01 9.60041702e-01 -7.75373578e-01
7.32301891e-01 -2.14630294e+00 -4.06260312e-01 -1.03472643e-01
2.11662948e-01 4.46615756e-01 2.42327362e-01 2.89264232e-01
5.04638106e-02 -1.73780233e-01 -5.72405122e-02 -6.56410977e-02
3.84121597e-01 4.21931922e-01 1.15856640e-01 2.02635035e-01
3.35546359e-02 4.43635881e-01 -7.15136588e-01 -1.29110605e-01
3.01877052e-01 5.38044572e-01 -5.55526197e-01 1.52301431e-01
-1.66677982e-02 1.29038215e-01 -1.70693040e-01 9.44878221e-01
5.21358013e-01 -3.16087097e-01 1.23699412e-01 -1.58051215e-02
-2.65894979e-01 2.96287090e-01 -1.37669885e+00 1.77887762e+00
-5.70159793e-01 8.65482926e-01 6.19514585e-01 -9.67244089e-01
9.87500012e-01 3.95045817e-01 3.23750198e-01 -4.93863523e-01
9.59367305e-02 3.39689940e-01 4.39008698e-02 -7.00997770e-01
2.92399764e-01 4.16325480e-01 3.43308359e-01 4.06759143e-01
3.06832701e-01 -6.37096614e-02 2.85061419e-01 -1.61122024e-01
1.54096222e+00 4.12789106e-01 -2.59864293e-02 -3.45284373e-01
-2.31198341e-01 2.00799212e-01 3.90377223e-01 6.86373174e-01
-3.13737482e-01 4.85850960e-01 6.73474431e-01 -1.01614046e+00
-1.36140966e+00 -8.05891454e-01 -1.73399135e-01 1.40949619e+00
-6.92963362e-01 -2.18679309e-01 -8.41001034e-01 -2.87137210e-01
6.67114928e-02 3.80332857e-01 -1.16363183e-01 2.58645356e-01
-3.19949389e-01 -6.79924250e-01 7.52498507e-01 9.59645331e-01
6.03130996e-01 -1.18376303e+00 -1.27557898e+00 5.63139439e-01
3.74686152e-01 -1.12251151e+00 4.72192168e-01 5.87274492e-01
-1.01553822e+00 -8.65890682e-01 -4.76649612e-01 -8.15450490e-01
2.23264068e-01 -5.63594885e-02 1.35336852e+00 1.11547731e-01
-9.23870504e-01 2.81080276e-01 -5.71997650e-02 -8.48625839e-01
-6.17806800e-02 7.94140548e-02 1.12466495e-02 -8.65451574e-01
6.99291766e-01 -8.30983341e-01 -5.48009813e-01 2.41851099e-02
-6.87782824e-01 1.39821842e-01 3.91561419e-01 5.46691120e-01
1.14532307e-01 1.58998501e-02 4.43447232e-01 -6.77450120e-01
6.03364587e-01 -4.10291582e-01 -9.88124013e-01 -2.13166833e-01
-7.46577024e-01 -2.60380536e-01 4.32576418e-01 -5.02445959e-02
-8.43896687e-01 2.02091515e-01 -7.29871690e-01 -2.91731328e-01
-6.62148595e-01 3.29990119e-01 2.45110653e-02 2.33649854e-02
6.98239923e-01 -2.56255746e-01 1.34520754e-01 -6.59088194e-01
4.74963104e-03 8.41346204e-01 6.02882266e-01 -7.62980282e-01
-5.63944727e-02 2.56497979e-01 -1.40176550e-01 -7.12786376e-01
-4.26513016e-01 -2.83766329e-01 -5.53509951e-01 -1.56657845e-01
6.37240469e-01 -8.38034809e-01 -1.16993797e+00 7.45744526e-01
-1.35968125e+00 -4.28224832e-01 -2.14990720e-01 2.62639403e-01
-2.92882919e-01 -7.33254850e-02 -8.01689148e-01 -6.34307623e-01
-4.38851804e-01 -1.32081246e+00 9.34996188e-01 1.75197497e-01
-3.78002375e-01 -9.35017407e-01 -1.15798950e-01 5.07553756e-01
7.15807140e-01 3.48409653e-01 7.15604067e-01 -7.62701571e-01
-5.98044455e-01 -4.87560213e-01 -4.39576805e-01 7.62141287e-01
-1.29080802e-01 2.54938930e-01 -1.33950388e+00 -1.85927320e-02
-9.80363190e-02 -3.54881942e-01 3.48411828e-01 3.64357769e-01
1.46594846e+00 -9.64092091e-02 -3.59294653e-01 5.47475457e-01
1.29380691e+00 1.85190156e-01 5.05396962e-01 6.23597264e-01
7.25802243e-01 5.39795339e-01 1.64857298e-01 5.43640316e-01
1.17878832e-01 4.54764605e-01 5.53009808e-01 -9.14611220e-02
2.24697948e-01 4.21378434e-01 1.70982137e-01 7.38251150e-01
-5.20778477e-01 1.94523871e-01 -1.31926119e+00 3.80372226e-01
-1.83358681e+00 -7.90591419e-01 -3.56569558e-01 1.99015582e+00
6.51193142e-01 3.54243040e-01 1.24272808e-01 1.64970338e-01
3.72276217e-01 -3.54260474e-01 -7.54063368e-01 -6.38409078e-01
3.80571812e-01 5.79085648e-01 4.40771222e-01 -7.41230398e-02
-8.19684565e-01 5.33258557e-01 7.04107857e+00 6.03664637e-01
-9.93363500e-01 3.30042928e-01 5.17796814e-01 -4.55185592e-01
5.06165326e-01 -4.41637307e-01 -8.07041943e-01 4.53584433e-01
1.41396964e+00 5.67536913e-02 5.82737744e-01 1.63232315e+00
-3.16577889e-02 -1.53244480e-01 -1.34907830e+00 1.03784955e+00
-2.34860748e-01 -1.68236268e+00 -6.53075516e-01 1.97905689e-01
4.32369560e-01 6.15943015e-01 -2.88415790e-01 3.50259483e-01
6.49743438e-01 -8.23741317e-01 6.38052106e-01 3.52895230e-01
2.75313675e-01 -9.83157635e-01 1.08851469e+00 3.83975029e-01
-5.61406374e-01 -1.03693157e-01 -4.26730990e-01 -6.38941884e-01
-2.55242229e-01 7.59204984e-01 -8.60534549e-01 3.29288930e-01
1.29932523e+00 3.72015446e-01 -4.09461766e-01 1.07324803e+00
2.02569321e-01 5.36758125e-01 -4.49309349e-01 -1.72530502e-01
-2.00040787e-02 6.61841854e-02 -9.04200424e-04 1.51024044e+00
3.18241119e-01 -2.32531220e-01 -1.27403378e-01 7.57507443e-01
4.15232452e-03 -3.58326137e-01 -6.20863438e-01 -4.75121029e-02
3.97483051e-01 1.62931716e+00 -9.05849576e-01 -3.56305778e-01
-4.82192397e-01 8.61753821e-01 2.94239491e-01 -1.28311366e-01
-7.51149476e-01 -6.84182405e-01 1.16232336e+00 -2.12757345e-02
1.70322418e-01 -3.19454402e-01 -5.65743685e-01 -7.31727719e-01
1.51993737e-01 -1.10560226e+00 1.25590622e-01 -9.26022112e-01
-1.35253930e+00 4.28226084e-01 -2.92013377e-01 -6.00649416e-01
-1.98966473e-01 -1.18290889e+00 -3.45425099e-01 7.46772528e-01
-9.48058665e-01 -8.43915403e-01 -5.81427634e-01 1.41296655e-01
4.12399739e-01 -2.41928399e-01 1.12276602e+00 8.21193814e-01
-7.62638211e-01 2.49443218e-01 2.09297702e-01 3.05780824e-02
3.97133648e-01 -1.10019910e+00 5.54120600e-01 5.27811170e-01
-6.35217875e-02 8.23465168e-01 6.06702924e-01 -2.82857627e-01
-1.71392465e+00 -8.99586797e-01 3.86367202e-01 -7.49174595e-01
6.57798946e-01 -5.38052142e-01 -8.35083246e-01 1.02071929e+00
2.12278560e-01 -8.33105668e-03 6.41521454e-01 4.48308349e-01
-1.14304483e-01 -3.77489805e-01 -1.21947014e+00 2.22921833e-01
1.10150170e+00 -3.33896369e-01 -2.77297318e-01 6.32212460e-01
4.26480860e-01 -8.01817954e-01 -9.41136777e-01 2.93279916e-01
8.30891073e-01 -1.34825480e+00 8.66261423e-01 -2.95006752e-01
5.69457114e-01 -9.83503908e-02 1.38940485e-02 -9.89251494e-01
-3.19289535e-01 -3.92873675e-01 -6.24680668e-02 1.25201249e+00
1.08222269e-01 -7.42946148e-01 8.74662757e-01 1.00451589e+00
-5.40306509e-01 -7.77714550e-01 -7.82649338e-01 -7.02623844e-01
-2.46524185e-01 -9.99999583e-01 8.15194070e-01 8.84901643e-01
-1.29931405e-01 4.77185771e-02 1.96746454e-01 1.68359205e-02
3.91434580e-01 -2.32947513e-01 8.58385742e-01 -1.42629194e+00
-4.74281251e-01 -5.09067178e-01 -7.28375673e-01 -5.19129395e-01
-2.31812179e-01 -7.06564367e-01 -6.04877621e-02 -1.73487175e+00
-4.36705165e-03 -3.04580629e-01 -1.43159375e-01 9.08013940e-01
3.23215932e-01 2.88330197e-01 -6.32173344e-02 8.59998539e-03
-4.99158919e-01 -3.61030191e-01 6.96049511e-01 2.22347751e-01
1.16334684e-01 -1.04399934e-01 -7.88823485e-01 9.94082391e-01
7.44475007e-01 -4.13710833e-01 -1.31639600e-01 -8.28071713e-01
3.08628201e-01 -5.33315778e-01 6.31129563e-01 -1.68364060e+00
4.40594435e-01 3.21560353e-01 7.93465436e-01 -3.47463280e-01
4.09182429e-01 -9.20959234e-01 3.64911824e-01 4.03686076e-01
2.81208530e-02 1.63362324e-01 6.12392664e-01 -7.89436027e-02
8.04851651e-02 -2.62974173e-01 6.06558859e-01 -3.76423001e-01
-8.59920621e-01 5.29530868e-02 -3.46101850e-01 -2.57095426e-01
1.03642070e+00 -3.54189932e-01 -4.20538396e-01 1.06485859e-01
-5.77173352e-01 -8.26782584e-02 4.83095914e-01 3.23565215e-01
1.34515911e-01 -8.38263273e-01 -4.04556066e-01 4.41260904e-01
-9.71393883e-02 1.71486557e-01 1.93921432e-01 5.33233285e-01
-1.00925493e+00 3.70044768e-01 -7.32828498e-01 -6.40116572e-01
-1.18973160e+00 3.32617968e-01 2.29065716e-01 6.16516508e-02
-8.27113450e-01 1.06672299e+00 -5.34597218e-01 -7.50085771e-01
5.93695581e-01 -5.68440378e-01 4.25992727e-01 -2.37702783e-02
6.84354305e-01 8.50814462e-01 8.57155681e-01 3.57490867e-01
-3.75605285e-01 -1.27824545e-02 1.20738931e-01 9.57961604e-02
1.78605187e+00 3.92592847e-01 -3.21426302e-01 4.69121337e-01
9.90495503e-01 -5.10924101e-01 -1.22150147e+00 4.00191635e-01
2.49897651e-02 -2.16570556e-01 3.04317415e-01 -9.24578071e-01
-1.21242154e+00 9.14547145e-01 8.58548522e-01 3.91772628e-01
1.01855171e+00 -1.50331602e-01 9.68522429e-01 7.17014909e-01
6.51834846e-01 -1.27175868e+00 -3.15562487e-01 5.30889452e-01
5.50397813e-01 -1.09355402e+00 -2.12845355e-02 -5.98123968e-02
-2.74762452e-01 1.42795277e+00 6.73353016e-01 -1.94452003e-01
5.88363111e-01 1.03535354e+00 2.18357384e-01 -2.93201059e-01
-8.48822773e-01 1.34031758e-01 -6.34662390e-01 9.14016604e-01
7.52181351e-01 -9.37576517e-02 2.31126845e-01 5.04571140e-01
-2.42664143e-01 6.43085241e-01 1.99357793e-01 1.33946908e+00
-3.09087247e-01 -9.27580059e-01 -4.53982532e-01 7.94403136e-01
-7.36030519e-01 3.92089114e-02 6.72331778e-03 6.87622964e-01
4.36577797e-01 8.26945066e-01 3.32233340e-01 -2.09708050e-01
5.80916643e-01 1.76326931e-01 5.46705842e-01 -5.87685823e-01
-9.41276789e-01 -2.99855769e-01 4.33815837e-01 -7.69406438e-01
1.10552736e-01 -5.82393050e-01 -1.08104205e+00 -8.11120927e-01
-8.56530443e-02 -3.45388800e-01 1.40202224e+00 7.07649231e-01
8.32327068e-01 8.56405497e-01 -2.06123874e-01 -1.20817304e+00
-3.00412744e-01 -8.73122454e-01 -6.51145875e-01 -2.01524302e-01
-2.08186526e-02 -4.20631617e-01 -1.22541033e-01 2.63523668e-01] | [8.423568725585938, 2.68607234954834] |
189506c6-4500-435e-9964-246b9b39cfcc | tailoring-machine-learning-for-process-mining | 2306.10341 | null | https://arxiv.org/abs/2306.10341v1 | https://arxiv.org/pdf/2306.10341v1.pdf | Tailoring Machine Learning for Process Mining | Machine learning models are routinely integrated into process mining pipelines to carry out tasks like data transformation, noise reduction, anomaly detection, classification, and prediction. Often, the design of such models is based on some ad-hoc assumptions about the corresponding data distributions, which are not necessarily in accordance with the non-parametric distributions typically observed with process data. Moreover, the learning procedure they follow ignores the constraints concurrency imposes to process data. Data encoding is a key element to smooth the mismatch between these assumptions but its potential is poorly exploited. In this paper, we argue that a deeper insight into the issues raised by training machine learning models with process data is crucial to ground a sound integration of process mining and machine learning. Our analysis of such issues is aimed at laying the foundation for a methodology aimed at correctly aligning machine learning with process mining requirements and stimulating the research to elaborate in this direction. | ['Wil van der Aalst', 'Ernesto Damiani', 'Sylvio Barbon Junior', 'Paolo Ceravolo'] | 2023-06-17 | null | null | null | null | ['anomaly-detection'] | ['methodology'] | [ 6.13394618e-01 3.32045019e-01 -1.32434592e-01 -3.88180465e-01
-2.46525817e-02 -5.31426847e-01 9.38537955e-01 8.60949397e-01
-1.77317888e-01 2.58212864e-01 5.35320444e-03 -6.46950245e-01
-4.97065872e-01 -1.08377087e+00 -3.02714974e-01 -3.62202376e-01
-2.58147903e-02 6.62339807e-01 1.04547374e-01 3.06639105e-01
4.99285817e-01 6.37690842e-01 -1.76924539e+00 2.66539842e-01
5.87551355e-01 7.63955176e-01 -3.19794536e-01 6.39752865e-01
-6.54393077e-01 1.03863907e+00 -3.45373988e-01 -3.83091897e-01
6.76621422e-02 -5.34835756e-01 -7.55458832e-01 2.91367978e-01
-1.68435365e-01 2.60583401e-01 3.30392867e-01 1.02869618e+00
-2.09096104e-01 -1.89015582e-01 8.17711234e-01 -1.42975497e+00
-2.61464983e-01 4.42765862e-01 -2.58362710e-01 6.76851645e-02
3.24511200e-01 2.45252743e-01 8.65529776e-01 -5.16655326e-01
3.62720162e-01 1.08727658e+00 6.18475854e-01 5.66898346e-01
-1.43191731e+00 -3.49803627e-01 8.01590160e-02 -5.10480516e-02
-1.07863724e+00 -5.36965132e-01 5.50119221e-01 -7.33293295e-01
6.17376387e-01 5.30738711e-01 7.43829191e-01 9.68196452e-01
2.22184286e-01 7.21949279e-01 1.03365195e+00 -6.33012772e-01
6.52150810e-01 4.84525084e-01 2.40914956e-01 2.46544838e-01
7.12674141e-01 1.07284352e-01 -5.71670771e-01 -2.25857183e-01
8.60565603e-01 8.57321620e-02 8.22842568e-02 -7.21382499e-01
-8.56473088e-01 7.17072189e-01 -5.39726257e-01 6.69907272e-01
-4.31946069e-01 -5.04373163e-02 5.52935362e-01 6.64386094e-01
1.98495045e-01 7.26565540e-01 -7.08786547e-01 -5.27440965e-01
-8.89227271e-01 1.94721684e-01 1.26848972e+00 9.03931499e-01
7.91127026e-01 -5.71857132e-02 6.98573068e-02 2.16359839e-01
5.67617714e-01 1.00272506e-01 3.08886200e-01 -8.26022863e-01
1.18706748e-01 1.09656656e+00 1.35224134e-01 -5.27666569e-01
-1.61338389e-01 -9.08389166e-02 -5.96092284e-01 1.81735694e-01
7.04935730e-01 -3.29483524e-02 -6.07732654e-01 1.21100140e+00
1.30950391e-01 -1.87963650e-01 1.44613102e-01 3.10338676e-01
7.40657672e-02 2.64084876e-01 4.17373389e-01 -5.63132703e-01
1.22490394e+00 -3.01649690e-01 -9.01464701e-01 -2.14041054e-01
7.29650140e-01 -6.05703890e-01 1.21036923e+00 4.72258240e-01
-1.00718224e+00 -4.65570122e-01 -7.81526506e-01 3.45891207e-01
-1.22879975e-01 -4.49333787e-01 1.11886811e+00 1.10303390e+00
-6.76013231e-01 8.04707348e-01 -1.14385343e+00 -6.95709169e-01
3.73347759e-01 2.81634748e-01 -2.87374705e-01 2.34375000e-01
-6.30685210e-01 8.07713807e-01 4.28358614e-01 4.01885882e-02
-4.99034405e-01 -7.28680313e-01 -8.48897040e-01 2.02150509e-01
6.87786281e-01 -6.01575792e-01 1.33945096e+00 -1.05677199e+00
-1.29977429e+00 7.93094456e-01 -1.72161713e-01 -5.69508195e-01
8.17731023e-01 -3.01119745e-01 -5.49301505e-01 -3.82810563e-01
-2.22209677e-01 -2.24149853e-01 8.16795707e-01 -1.22895181e+00
-9.13394332e-01 -6.18320405e-01 -3.85303199e-01 -2.26126984e-01
8.88491236e-03 2.51394540e-01 -7.15875924e-02 -2.07741842e-01
1.85299262e-01 -5.40692210e-01 -4.35597181e-01 -2.33559847e-01
-3.63983482e-01 -1.89716488e-01 8.40746582e-01 -7.91379437e-02
1.20650637e+00 -2.20611548e+00 -2.22167075e-01 5.67564905e-01
-1.89711433e-02 1.15720339e-01 3.25939506e-01 9.20890987e-01
-7.84655064e-02 2.81700641e-01 -2.01517999e-01 -2.13181853e-01
2.16216534e-01 2.19354317e-01 -4.71811324e-01 3.84668648e-01
6.10391378e-01 6.58370912e-01 -7.88928926e-01 -2.66458988e-01
3.66540909e-01 1.46372646e-01 -2.45749921e-01 3.40015829e-01
-3.70103627e-01 6.02950215e-01 -6.22992814e-01 5.35007894e-01
1.90183595e-01 1.60433464e-02 4.79253054e-01 2.75319308e-01
-4.36053038e-01 3.16254139e-01 -1.48665226e+00 1.10390663e+00
-2.03525081e-01 3.62716496e-01 -6.05172198e-03 -9.94914949e-01
1.16236424e+00 2.65944630e-01 6.92669988e-01 -4.01528537e-01
8.87553766e-02 2.84632575e-02 2.39783302e-01 -6.35076404e-01
3.37002099e-01 -4.22278076e-01 9.97242630e-02 7.79447854e-01
-1.87330082e-01 -2.25249544e-01 3.61432821e-01 -4.96805072e-01
1.11341774e+00 2.13712782e-01 5.63871741e-01 -3.55371058e-01
7.27591872e-01 6.38234913e-02 6.85862303e-01 7.16408551e-01
4.07450311e-02 2.08201423e-01 9.86594558e-01 -6.44170284e-01
-1.09782088e+00 -9.62904215e-01 -9.96736139e-02 6.59995675e-01
-3.38847607e-01 -7.46100307e-01 -6.40932918e-01 -6.59227431e-01
3.00799850e-02 9.56060290e-01 -5.88021636e-01 -1.39455408e-01
-2.35882819e-01 -7.08212316e-01 4.83462006e-01 4.84608978e-01
2.93743946e-02 -1.02393007e+00 -8.45227838e-01 4.23166454e-01
5.48418820e-01 -8.29944551e-01 3.27115983e-01 5.39672852e-01
-1.26229715e+00 -1.51235890e+00 2.80420363e-01 -3.20659250e-01
5.82032502e-01 -3.00132334e-01 1.00536919e+00 9.29029658e-02
1.58153344e-02 4.07648206e-01 -4.40055966e-01 -1.04591179e+00
-9.64881480e-01 8.41219500e-02 -1.43811211e-01 1.07846204e-02
1.10231006e+00 -6.54750168e-01 -4.64898162e-02 1.81722343e-01
-1.21631813e+00 -1.45399645e-01 7.74792552e-01 3.39861691e-01
3.33590806e-01 6.07116878e-01 3.77298981e-01 -1.15911746e+00
8.37438464e-01 -2.68775344e-01 -5.01172245e-01 3.03934813e-01
-1.00724638e+00 2.87864715e-01 5.29267728e-01 -3.65233570e-01
-9.61358905e-01 1.07804582e-01 1.07495807e-01 -3.16304937e-02
-8.31813037e-01 5.03951848e-01 -5.70591748e-01 5.28735161e-01
4.17451829e-01 2.45877400e-01 3.13630402e-01 -3.83938640e-01
1.50233328e-01 3.53421837e-01 1.22984760e-01 -6.73587859e-01
9.47030306e-01 5.35141587e-01 1.21037632e-01 -8.21987391e-01
-3.42602462e-01 -4.39606398e-01 -7.76533663e-01 4.38873880e-02
6.76987767e-01 -3.00669491e-01 -6.40703440e-01 2.13517800e-01
-6.46421075e-01 -5.12294531e-01 -8.56939256e-01 2.99409688e-01
-7.95255363e-01 2.91868985e-01 -4.78596658e-01 -1.26250470e+00
-1.50745735e-01 -8.98666978e-01 6.32149458e-01 5.58076985e-03
-1.00457072e+00 -1.19852114e+00 -1.37698157e-02 1.64049059e-01
3.23170483e-01 1.36585549e-01 1.21666372e+00 -1.36561036e+00
-4.67367560e-01 -4.89073306e-01 2.13029221e-01 3.40779990e-01
4.66869265e-01 5.64940870e-01 -9.75238204e-01 9.68696401e-02
4.10231262e-01 1.97817549e-01 2.16537863e-01 1.19522162e-01
1.21763396e+00 -4.58957031e-02 -1.50318444e-01 5.47812246e-02
1.28061485e+00 2.66863734e-01 7.00080335e-01 5.17922699e-01
4.38038796e-01 1.30327845e+00 6.47199333e-01 4.90368187e-01
2.18763813e-01 2.23758906e-01 2.02392712e-01 1.89236999e-01
5.22318363e-01 -3.07211906e-01 2.22745210e-01 2.93510169e-01
-1.47122785e-01 2.56779253e-01 -1.20416486e+00 3.98169756e-01
-1.94702470e+00 -1.24433386e+00 -7.35134721e-01 2.36052489e+00
6.65990770e-01 5.26527166e-01 5.26649989e-02 7.86273539e-01
3.78209174e-01 -2.47562155e-01 -7.68913031e-02 -7.71354675e-01
4.37022984e-01 2.83083946e-01 3.51020575e-01 2.81479508e-01
-1.20184374e+00 5.73549092e-01 6.13957930e+00 2.68717617e-01
-9.13674593e-01 -3.79106790e-01 3.02424848e-01 1.70630246e-01
-4.48658764e-01 2.77322382e-01 -5.94883919e-01 3.06691885e-01
1.14123368e+00 -2.65284181e-01 2.40227208e-01 9.49486792e-01
6.75689161e-01 -2.32372195e-01 -1.90785456e+00 6.07486129e-01
-3.42167020e-01 -9.98921812e-01 -2.92994902e-02 4.21568900e-01
3.61700356e-01 -6.45212889e-01 -1.63231909e-01 3.12601924e-01
4.45653141e-01 -1.38482499e+00 6.57021642e-01 5.91511726e-01
4.28944975e-02 -8.46801937e-01 7.49319434e-01 4.57001120e-01
-7.45498598e-01 -2.47898057e-01 1.67248733e-02 -5.33296645e-01
6.42709881e-02 8.66547883e-01 -9.77682769e-01 6.43611491e-01
4.19257700e-01 4.49599504e-01 -3.86446208e-01 9.56592619e-01
-1.65728480e-01 6.91668928e-01 -7.58189335e-02 2.53458947e-01
6.55469950e-03 -3.94368917e-01 4.12763208e-01 1.10992682e+00
1.73360795e-01 -5.32264113e-01 9.24213731e-04 9.50797856e-01
5.96438348e-01 3.03382277e-01 -7.51437783e-01 -3.07627439e-01
3.69436741e-01 1.04165018e+00 -9.26719904e-01 -7.48289824e-02
-6.57306671e-01 5.78898370e-01 -1.45780653e-01 6.48445264e-02
-3.36311162e-01 1.75892487e-01 9.32481706e-01 5.62007248e-01
-4.22928333e-02 -3.83396268e-01 -1.05695844e+00 -8.76946509e-01
1.21875241e-01 -1.11160994e+00 3.84003371e-01 -8.00528750e-02
-1.40980816e+00 1.40036628e-01 4.45004459e-03 -9.13621366e-01
-4.79233325e-01 -4.97082084e-01 -9.04691339e-01 8.45604479e-01
-1.31625676e+00 -1.13021278e+00 -7.53757805e-02 4.53425169e-01
3.39514434e-01 -8.77970271e-03 1.01459002e+00 -1.68283246e-02
-5.27515769e-01 -1.04763262e-01 -3.35647404e-01 1.90535914e-02
5.87780774e-01 -1.43074644e+00 4.34033811e-01 8.13475013e-01
2.30833218e-01 7.96561182e-01 1.09945309e+00 -7.74741530e-01
-1.60697627e+00 -1.05496490e+00 1.01149094e+00 -7.84918368e-01
1.00347424e+00 -1.42992988e-01 -1.30698073e+00 8.38569582e-01
-4.27487046e-02 -4.46957827e-01 1.19985247e+00 4.00169998e-01
-5.26345847e-03 1.63016487e-02 -9.98087227e-01 3.86160165e-01
5.73367894e-01 -4.46505666e-01 -8.46307874e-01 -2.81371444e-01
1.11491859e-01 9.86390188e-02 -1.10928404e+00 3.68117988e-01
2.38809273e-01 -9.75963175e-01 5.20499229e-01 -1.04597485e+00
2.83955723e-01 -3.46895605e-01 5.62155135e-02 -8.08225691e-01
2.80204881e-02 -5.89058697e-01 -2.06793129e-01 1.61264157e+00
4.26049471e-01 -4.80054736e-01 1.14914823e+00 1.28796494e+00
2.32278422e-01 -4.51337337e-01 -5.09357750e-01 -6.05525315e-01
-2.02353001e-02 -1.06387019e+00 7.11605132e-01 1.19484663e+00
1.10974992e-02 1.81456253e-01 1.73442975e-01 8.80267620e-02
5.25907576e-01 -9.43711996e-02 1.09325099e+00 -1.92996049e+00
-1.69648707e-01 -5.80998302e-01 -5.12941241e-01 -2.87581503e-01
-2.45424383e-03 -4.39754993e-01 -2.70666718e-01 -1.58495116e+00
-2.23735999e-02 -6.57869577e-01 2.07511373e-02 3.13281715e-01
4.28219885e-02 -4.42069590e-01 -5.22481427e-02 3.16836387e-01
-2.71901250e-01 2.73950040e-01 9.49549615e-01 3.30617100e-01
-5.59073985e-01 4.83852893e-01 -9.16704416e-01 1.06681120e+00
9.94607985e-01 -4.74678755e-01 -7.14462280e-01 1.34891376e-01
4.47767079e-01 -2.86815763e-01 2.50327259e-01 -8.43595207e-01
4.16755587e-01 -6.16177142e-01 2.87460595e-01 -2.39222020e-01
-1.56128153e-01 -1.39561033e+00 4.22315508e-01 4.84108746e-01
-4.74984050e-01 6.12337627e-02 -5.92566282e-02 5.13856709e-01
-2.91575164e-01 -5.05712390e-01 4.39320982e-01 -3.48529130e-01
-7.89116502e-01 1.12099990e-01 -7.27559388e-01 -1.92097828e-01
1.42271936e+00 -5.84052920e-01 1.99021026e-01 -1.83380172e-01
-8.40515137e-01 8.15045908e-02 7.91494906e-01 4.82913464e-01
2.12736070e-01 -7.82131612e-01 -4.44302708e-01 7.17453361e-01
3.15925598e-01 2.41001919e-01 -3.06849748e-01 8.37444901e-01
-3.27581286e-01 2.08813816e-01 -6.04494251e-02 -4.35993284e-01
-1.24344850e+00 6.57299638e-01 1.73434332e-01 -4.93274987e-01
-5.65134823e-01 6.19402565e-02 -2.35440090e-01 -3.12754005e-01
1.11476853e-01 -4.59622025e-01 -1.96395084e-01 1.25619277e-01
6.13003433e-01 3.21249902e-01 1.63053647e-01 5.28215505e-02
-1.76410392e-01 -4.56611775e-02 3.46764252e-02 -7.28741363e-02
1.34794927e+00 2.08536834e-02 -4.16014791e-01 8.68748367e-01
2.74021327e-01 1.54466882e-01 -1.19598103e+00 -3.42654772e-02
1.12656319e+00 -4.58546817e-01 -3.47927868e-01 -4.84272718e-01
-5.35136878e-01 7.77125716e-01 1.12113208e-01 8.32127690e-01
8.06769907e-01 4.07417826e-02 -3.65237854e-02 2.79356122e-01
2.51224637e-01 -1.29451036e+00 -1.18010819e-01 3.78610909e-01
5.37935495e-01 -9.87173378e-01 -8.43134746e-02 -6.45581365e-01
-6.64975464e-01 1.37044418e+00 5.32547653e-01 1.25343189e-01
6.20697916e-01 7.19592631e-01 1.80495717e-02 -2.94158608e-01
-8.16706002e-01 -3.01009715e-01 -1.10044524e-01 8.27262819e-01
6.13823712e-01 1.44578572e-02 -1.76409498e-01 5.72967708e-01
-1.55567378e-01 3.06913108e-01 5.60003817e-01 1.29815745e+00
-5.19648552e-01 -1.67477846e+00 -5.47213137e-01 5.70738792e-01
-7.66131341e-01 1.15100503e-01 -4.61244792e-01 1.06456542e+00
1.06415704e-01 1.07357156e+00 3.75208795e-01 -1.48846313e-01
6.18095398e-01 4.96143430e-01 3.42914611e-01 -8.81357849e-01
-6.15474463e-01 -4.69189100e-02 3.11558008e-01 -3.54654521e-01
-2.94018954e-01 -1.03541589e+00 -1.03221762e+00 -4.11255956e-01
7.35377520e-02 2.06799060e-01 8.07258427e-01 1.09927213e+00
-3.93322064e-03 5.49621224e-01 1.39454231e-01 -1.34502918e-01
-7.34796286e-01 -7.72471130e-01 -5.36837101e-01 4.71283048e-01
-2.39497542e-01 -2.67933309e-01 -5.49733378e-02 2.74072230e-01] | [8.662229537963867, 6.004100322723389] |
fc50b77d-efaa-4a2d-aa2f-c3e16ca57a7b | repurposing-existing-deep-networks-for | 2201.02280 | null | https://arxiv.org/abs/2201.02280v1 | https://arxiv.org/pdf/2201.02280v1.pdf | Repurposing Existing Deep Networks for Caption and Aesthetic-Guided Image Cropping | We propose a novel optimization framework that crops a given image based on user description and aesthetics. Unlike existing image cropping methods, where one typically trains a deep network to regress to crop parameters or cropping actions, we propose to directly optimize for the cropping parameters by repurposing pre-trained networks on image captioning and aesthetic tasks, without any fine-tuning, thereby avoiding training a separate network. Specifically, we search for the best crop parameters that minimize a combined loss of the initial objectives of these networks. To make the optimization table, we propose three strategies: (i) multi-scale bilinear sampling, (ii) annealing the scale of the crop region, therefore effectively reducing the parameter space, (iii) aggregation of multiple optimization results. Through various quantitative and qualitative evaluations, we show that our framework can produce crops that are well-aligned to intended user descriptions and aesthetically pleasing. | ['Hyung Jin Chang', 'Ales Leonardis', 'Abhishake Kumar Bojja', 'Kwang Moo Yi', 'Kedi Xia', 'Nora Horanyi'] | 2022-01-07 | null | null | null | null | ['image-cropping'] | ['computer-vision'] | [ 6.35564685e-01 3.13471347e-01 6.86106756e-02 -4.42934960e-01
-6.65168941e-01 -6.30824208e-01 3.27036947e-01 4.60749455e-02
-3.22698683e-01 4.37390327e-01 2.25262225e-01 -2.59968638e-01
2.26991862e-01 -8.48298371e-01 -9.15825307e-01 -5.93858540e-01
3.42101067e-01 -6.80010393e-02 -2.07031682e-01 -1.72698408e-01
2.68869370e-01 3.55467826e-01 -1.38642383e+00 1.48440138e-01
1.16009557e+00 1.25275815e+00 2.63214082e-01 6.15966678e-01
8.25526416e-02 4.71418798e-01 -4.63125944e-01 -7.24482000e-01
2.64268577e-01 -6.02394879e-01 -6.64422452e-01 6.54765069e-01
2.91748554e-01 -3.08180839e-01 3.06970567e-01 1.21413052e+00
3.66272718e-01 -1.32416878e-02 6.00861251e-01 -1.34420669e+00
-1.21123922e+00 5.25483072e-01 -8.42131436e-01 -6.78502977e-01
1.39303058e-01 4.00308102e-01 1.13658392e+00 -9.07908380e-01
5.66770196e-01 1.37275970e+00 4.96557653e-01 3.99117440e-01
-1.30405307e+00 -4.38800573e-01 2.28862077e-01 -1.35795221e-01
-1.24506605e+00 -2.61048824e-01 9.05106008e-01 -2.82609850e-01
4.14764702e-01 3.59235674e-01 7.72460938e-01 9.38020289e-01
-1.03418037e-01 7.54615784e-01 1.05334568e+00 -5.84961712e-01
2.65052825e-01 3.15016747e-01 -3.87668163e-01 6.39186025e-01
1.70988768e-01 -1.06346421e-01 -2.76202381e-01 9.79423076e-02
9.53778207e-01 -2.74922192e-01 -1.39818788e-01 -6.97366118e-01
-1.17459905e+00 9.12611246e-01 7.91520000e-01 -6.66088089e-02
-5.60851991e-01 4.69559133e-01 1.54280156e-01 5.42911664e-02
4.56772894e-01 9.43394184e-01 -3.77074629e-01 3.02917361e-01
-7.99685955e-01 2.84094691e-01 3.21021318e-01 8.37196946e-01
1.04660702e+00 -1.06945701e-01 -3.17959249e-01 1.05732679e+00
4.76750791e-01 4.96377110e-01 1.47946268e-01 -1.03977740e+00
3.78027707e-01 5.75651526e-01 4.42345589e-01 -9.66140330e-01
-1.29582673e-01 -2.92332530e-01 -7.01951146e-01 4.46379781e-01
2.57953435e-01 -4.41604435e-01 -1.17858803e+00 1.77885640e+00
1.00955404e-01 -5.20557463e-01 1.32807996e-02 1.04950464e+00
5.90408325e-01 6.64049208e-01 4.03666168e-01 7.37431273e-02
1.35536754e+00 -1.29432678e+00 -7.07433045e-01 -5.24497986e-01
5.41648805e-01 -9.46439147e-01 1.62179339e+00 1.61402434e-01
-1.35217130e+00 -4.01308268e-01 -1.16322172e+00 -1.29299074e-01
-3.37532252e-01 4.33091611e-01 5.59980810e-01 5.79603314e-01
-1.31156874e+00 6.10828996e-01 -3.38374555e-01 -2.52689779e-01
4.48342144e-01 2.70090818e-01 -2.06860185e-01 1.33824706e-01
-1.04509473e+00 9.30516422e-01 4.47685808e-01 2.23360345e-01
-6.26773298e-01 -4.84632581e-01 -1.03972995e+00 2.48262152e-01
3.46759379e-01 -7.40614176e-01 1.02571607e+00 -1.72511518e+00
-1.63800943e+00 8.86943758e-01 7.27358907e-02 -1.15910113e-01
4.33591425e-01 -2.82259107e-01 1.41887859e-01 -6.56157956e-02
-4.86913249e-02 1.47463787e+00 9.06159759e-01 -1.74653995e+00
-3.37244689e-01 8.89679566e-02 4.34706569e-01 5.89941740e-01
-6.14878118e-01 -9.13574174e-02 -5.89868963e-01 -7.08547533e-01
-2.51225978e-02 -9.72467005e-01 -5.37883878e-01 4.52827364e-01
-6.63642228e-01 2.26240784e-01 5.39769351e-01 -7.44409144e-01
1.11795056e+00 -2.04496956e+00 1.90092549e-01 3.29546839e-01
-1.60883199e-02 6.60424605e-02 -5.69540620e-01 4.46845852e-02
-1.10312231e-01 5.01882970e-01 -4.01992053e-01 -4.94379789e-01
2.14032196e-02 -4.71765473e-02 -8.29057246e-02 6.17887974e-02
7.06848562e-01 1.29662395e+00 -8.58263493e-01 -5.78698277e-01
3.12288553e-01 5.21163344e-01 -7.69830108e-01 3.71534526e-01
-4.21547383e-01 2.55491346e-01 -3.05897325e-01 6.79160416e-01
7.73662984e-01 -3.47629964e-01 9.63755473e-02 -4.85475570e-01
8.66425708e-02 -4.57600206e-02 -1.01023066e+00 1.59571838e+00
-7.22338498e-01 5.97701192e-01 2.02782210e-02 -6.51754498e-01
1.06923389e+00 3.43523659e-02 3.81590635e-01 -6.98651671e-01
2.85013914e-01 9.84850228e-02 -3.80035788e-01 -4.79015052e-01
5.98295391e-01 -1.05657332e-01 -5.59977517e-02 4.73414510e-01
-1.23604797e-01 -4.14983213e-01 1.40301317e-01 -6.73790872e-02
5.34238756e-01 4.51272130e-01 1.89494178e-01 -9.06422064e-02
4.18434322e-01 -1.35193631e-01 1.55634701e-01 3.66480201e-01
-4.17195410e-02 1.01562977e+00 7.39886403e-01 -2.72259176e-01
-1.45712066e+00 -8.84766638e-01 2.66348451e-01 1.25516725e+00
2.70988077e-01 8.45806524e-02 -1.24255955e+00 -5.74317455e-01
-2.49775201e-01 8.22503150e-01 -9.24761415e-01 -1.22052334e-01
-4.76552784e-01 -8.10497284e-01 1.80216268e-01 3.37505430e-01
5.56553125e-01 -1.57033837e+00 -6.39718473e-01 -7.94761255e-02
-4.13016021e-01 -9.24730122e-01 -9.17835057e-01 2.44633779e-01
-5.56473672e-01 -7.26999342e-01 -1.07482862e+00 -9.08636689e-01
1.16826010e+00 3.29776973e-01 1.17070127e+00 3.97601992e-01
-1.07148318e-02 1.70346983e-02 -3.61225426e-01 -3.75556439e-01
-3.98144633e-01 2.74002254e-01 -5.35925806e-01 -1.79395080e-02
-1.93001553e-01 -3.60202074e-01 -7.52962530e-01 2.09114403e-01
-1.06697011e+00 5.09712577e-01 9.03841078e-01 7.46884942e-01
5.22709727e-01 -2.99465358e-01 4.61862385e-01 -5.41540205e-01
8.66924763e-01 -1.20498680e-01 -4.95174199e-01 6.57351971e-01
-6.71118855e-01 1.69199899e-01 3.86789978e-01 -6.67790890e-01
-9.13916409e-01 3.01901907e-01 4.68915291e-02 -3.68220091e-01
8.35317224e-02 3.23820978e-01 -3.38345259e-01 -3.29138249e-01
7.15347469e-01 -3.81004177e-02 6.54391898e-03 -1.87379960e-03
8.95344615e-01 3.94292325e-01 4.46276784e-01 -5.36290526e-01
9.83354747e-01 1.79367289e-01 -3.50727737e-01 -2.79032528e-01
-8.71813059e-01 1.65038481e-01 -6.02501035e-01 -4.98626053e-01
1.12764966e+00 -7.45111048e-01 -6.34198129e-01 3.76322627e-01
-1.20413280e+00 -4.13729638e-01 -3.07887048e-01 8.62260312e-02
-6.42172337e-01 2.17861980e-01 -2.11779237e-01 -8.82782340e-01
-5.15108049e-01 -1.31168711e+00 1.24167681e+00 4.48260099e-01
-3.18133831e-01 -8.04078460e-01 -2.74527073e-01 2.73172408e-01
6.13843858e-01 5.35555482e-01 9.51964498e-01 4.88795936e-02
-5.83109677e-01 -2.92502120e-02 -8.32534134e-01 3.43550712e-01
1.42046347e-01 1.84996128e-01 -9.20364380e-01 -1.29703507e-01
-4.38132316e-01 -6.18440807e-01 5.55647075e-01 5.92436850e-01
1.33415949e+00 -5.74148476e-01 4.22115158e-03 7.88014352e-01
1.59166598e+00 1.98592648e-01 7.64607072e-01 6.28934324e-01
8.00529659e-01 8.10725808e-01 5.26401877e-01 3.11138362e-01
3.64902794e-01 3.98262560e-01 7.65821218e-01 -7.09134281e-01
-1.90303139e-02 -4.49326724e-01 2.99002111e-01 8.86861160e-02
5.90869151e-02 -3.92732054e-01 -4.90998566e-01 6.65754497e-01
-1.81546187e+00 -5.34629047e-01 4.41790462e-01 2.04363680e+00
8.85758460e-01 9.28270519e-02 1.58470005e-01 -3.26306731e-01
8.42455924e-01 2.89552659e-01 -8.04379642e-01 -6.80184066e-01
-1.79528356e-01 9.32860449e-02 6.21449172e-01 4.97693241e-01
-1.05062282e+00 1.19469035e+00 6.93110943e+00 6.28716886e-01
-1.18289948e+00 -2.38955289e-01 1.19280422e+00 -6.27205446e-02
-8.51816833e-01 7.16285780e-02 -1.22387677e-01 1.72568604e-01
3.44716609e-01 -6.24639876e-02 8.35608304e-01 7.57519364e-01
3.99969935e-01 -2.21406370e-02 -7.69593835e-01 6.71347082e-01
-1.77304465e-02 -1.18166006e+00 2.59653658e-01 -3.29148062e-02
7.50514746e-01 -5.52547514e-01 2.40698621e-01 3.85292955e-02
4.76949126e-01 -1.26615047e+00 1.04352605e+00 3.31340790e-01
7.53214777e-01 -6.65869892e-01 3.80666554e-01 -1.03549354e-01
-9.25860465e-01 -7.13991374e-02 -6.67411238e-02 3.33338857e-01
3.43630552e-01 4.21046346e-01 -5.39787650e-01 1.56825483e-01
6.09968543e-01 3.31810802e-01 -5.96078753e-01 8.79478693e-01
-2.68027872e-01 1.65398359e-01 -1.13265194e-01 -1.50634453e-01
3.71347219e-01 -4.39994842e-01 1.99178606e-01 1.04513288e+00
4.69925344e-01 -1.97534412e-01 -4.54401821e-02 1.24119520e+00
-1.68161586e-01 3.85279119e-01 -2.87044704e-01 -2.13445440e-01
3.03123593e-01 1.46247470e+00 -8.61281872e-01 -1.12638757e-01
-1.75336778e-01 1.13949561e+00 3.17849785e-01 4.88448560e-01
-1.06714594e+00 -3.32799882e-01 4.04422671e-01 7.92221054e-02
3.97193074e-01 1.08147562e-02 -7.76937425e-01 -8.63216758e-01
1.31304786e-01 -7.85324574e-01 -1.40102699e-01 -1.32026470e+00
-9.84067321e-01 7.57365942e-01 -1.55548602e-01 -1.15895867e+00
-6.20300286e-02 -3.63387674e-01 -6.45974874e-01 9.07629073e-01
-1.53860867e+00 -1.44485271e+00 -4.92328852e-01 8.86784866e-02
3.31282854e-01 2.72247285e-01 5.76986909e-01 1.32796094e-01
-4.58362013e-01 6.42736197e-01 -3.42586815e-01 -1.01036228e-01
7.41408944e-01 -1.22693121e+00 5.70407927e-01 6.85261786e-01
-2.96236962e-01 4.17880982e-01 9.66369271e-01 -3.63173038e-01
-7.61106908e-01 -1.04964268e+00 7.88444102e-01 5.25015369e-02
4.76042986e-01 -4.18773562e-01 -7.39680469e-01 3.38357151e-01
4.68271494e-01 -3.46795261e-01 3.06888252e-01 -4.42888737e-02
-2.60866612e-01 1.23840747e-02 -1.22015667e+00 9.38974619e-01
8.34660351e-01 -1.78247765e-01 5.40185682e-02 4.34694409e-01
9.50266659e-01 -3.74340773e-01 -7.32691050e-01 2.96066672e-01
6.70445442e-01 -8.32440257e-01 1.01008797e+00 -4.38108146e-01
9.26134527e-01 -2.08472297e-01 1.03602149e-02 -1.47395229e+00
-4.03725803e-01 -5.69977403e-01 3.24976742e-01 1.13591123e+00
8.04036319e-01 -2.22350106e-01 8.26683640e-01 9.55401540e-01
1.13767937e-01 -1.02921629e+00 -2.96934962e-01 -3.19333375e-01
5.48952958e-04 3.66623439e-02 7.66746283e-01 7.04208732e-01
-2.87170172e-01 2.70355314e-01 -7.49529660e-01 -5.75556643e-02
5.21484554e-01 7.87971988e-02 6.79543018e-01 -7.66859829e-01
-8.43907222e-02 -6.57962024e-01 2.10951433e-01 -8.64252329e-01
-1.60570875e-01 -3.74649465e-01 4.98383552e-01 -1.56345630e+00
3.50686610e-01 -5.02288401e-01 -1.75505728e-01 8.11482966e-01
-5.19917965e-01 4.35273707e-01 3.40262532e-01 1.24961343e-02
-4.89846200e-01 6.39949322e-01 1.75616157e+00 -7.40510449e-02
-3.99038941e-01 -3.39036435e-01 -1.33157635e+00 4.89457011e-01
7.42663562e-01 -2.10626185e-01 -5.08683264e-01 -4.55456018e-01
5.05001128e-01 -1.24262832e-01 3.35754156e-01 -6.90139294e-01
-3.09138358e-01 -6.53454483e-01 5.42792797e-01 -3.28075200e-01
3.17855090e-01 -7.32859313e-01 1.95848085e-02 2.97571301e-01
-7.04743862e-01 3.52828018e-02 1.32312670e-01 2.45088622e-01
3.48615013e-02 -4.21614230e-01 1.07875454e+00 -1.01992659e-01
-4.99848336e-01 2.25345358e-01 -8.48541409e-02 -2.85435468e-01
1.03734183e+00 -2.07202122e-01 -9.44194198e-02 -6.64794028e-01
-5.88051438e-01 4.20050740e-01 8.16319644e-01 6.30162418e-01
3.30233455e-01 -1.49922085e+00 -5.75829983e-01 -3.89104672e-02
9.95484442e-02 2.55769845e-02 1.72552139e-01 3.81656677e-01
-6.57220066e-01 -1.27271609e-02 -4.06529874e-01 -3.88663262e-01
-1.11953402e+00 6.11463547e-01 4.00856584e-01 -3.86283368e-01
-1.81346565e-01 7.94906676e-01 4.79577512e-01 -5.67321181e-01
1.84439167e-01 -2.38583684e-01 -3.63999814e-01 8.90582576e-02
8.47688019e-02 -1.56565428e-01 -2.18992069e-01 -4.96109933e-01
-1.59195065e-01 5.42342484e-01 3.29355225e-02 -3.23229402e-01
1.28503323e+00 -4.00092095e-01 -1.89459734e-02 -1.59199253e-01
1.20551014e+00 -4.06247020e-01 -1.72436726e+00 -1.17573403e-02
-3.44675153e-01 -3.78317386e-01 8.59135315e-02 -1.02309918e+00
-1.26308203e+00 7.33531773e-01 6.52111650e-01 2.52226919e-01
1.54498720e+00 -3.35707545e-01 8.32419872e-01 9.85240340e-02
-1.12051673e-01 -1.22507346e+00 4.43364352e-01 1.99346215e-01
1.19312096e+00 -1.26570547e+00 -1.87350903e-02 -4.55331177e-01
-1.10878301e+00 1.08244288e+00 8.08413804e-01 -2.35352799e-01
3.29769343e-01 1.31838918e-01 3.45773935e-01 -2.14417167e-02
-4.81470644e-01 -2.08669752e-01 4.52566981e-01 5.24387777e-01
5.03111303e-01 5.18100485e-02 -2.45672554e-01 3.63920927e-01
-2.56480187e-01 7.30427727e-02 3.57090980e-01 6.70956254e-01
-6.04969323e-01 -1.04292667e+00 -4.24583763e-01 2.74489641e-01
-1.59513682e-01 -2.68133372e-01 -5.62813878e-01 5.54172456e-01
1.65758803e-01 6.86960042e-01 -3.74041945e-02 -3.45567793e-01
4.41651911e-01 -2.50204831e-01 3.15727621e-01 -2.68728435e-01
-5.83140075e-01 1.82291314e-01 -1.93721484e-02 -4.16090190e-01
-3.28358978e-01 -4.73863423e-01 -8.11413288e-01 -9.22933891e-02
-4.45677459e-01 -2.88895458e-01 8.46705139e-01 9.02797878e-01
2.16726184e-01 5.17434180e-01 8.44631374e-01 -1.13623428e+00
-3.71805370e-01 -8.57614458e-01 -3.15672100e-01 4.70861197e-01
1.47031322e-01 -4.79725868e-01 -3.18225831e-01 2.94898450e-01] | [11.482085227966309, -0.955616295337677] |
a558749a-35a5-4fec-8295-6e00bf0e79e5 | robot-intent-recognition-method-based-on | null | null | https://openreview.net/forum?id=xREjEGUoY4c | https://openreview.net/pdf?id=xREjEGUoY4c | Robot Intent Recognition Method Based on State Grid Business Office | Artificial intelligence is currently in an era of change, not only changing the artificial intelligence technology itself, but also changing human society. It has become more and more common to use artificial intelligence as the core human-computer interaction technology to replace manpower. Intention recognition is an important part of the human-machine dialogue system, and deep learning technology is gradually being applied to the task of intent recognition. However, intent recognition based on deep learning often has problems such as low recognition accuracy and slow recognition speed. In response to these problems, this paper designs a BERT fine-tuning to improve the network structure based on the pre-training model and proposes new continuous pre-training goals. To improve the accuracy of intent recognition, a method based on multi-teacher model compression is proposed to compress the pre-training model, which reduces the time consumption of model inference. | ['Hanchao Liu', 'Zhao Pu Hu', 'Lanfang Dong'] | 2021-09-29 | null | null | null | null | ['intent-recognition'] | ['natural-language-processing'] | [ 3.01773876e-01 9.91652757e-02 -3.69665235e-01 -6.02099061e-01
1.22081749e-01 1.10354178e-01 4.70978230e-01 -1.98769584e-01
-6.94221258e-01 3.58528316e-01 1.91424131e-01 -3.77965897e-01
6.74181506e-02 -7.98754334e-01 3.12655754e-02 -3.71451020e-01
4.65758741e-01 6.07169688e-01 -1.43849134e-01 -1.79471731e-01
5.13959050e-01 2.30251893e-01 -1.36213875e+00 1.87790811e-01
1.02417767e+00 1.05256820e+00 3.01897705e-01 5.29971838e-01
-4.03401673e-01 1.23249567e+00 -5.90409935e-01 -2.85322875e-01
-6.05804957e-02 -5.45193136e-01 -9.46889102e-01 7.84110464e-03
-3.98916602e-01 -6.99539363e-01 -5.83711147e-01 9.97508228e-01
3.13761234e-01 1.40293792e-01 5.25811374e-01 -1.15338635e+00
-4.27268863e-01 4.32810277e-01 -3.34943801e-01 3.47805396e-02
2.66889215e-01 9.11533087e-02 6.97956502e-01 -4.22506720e-01
-1.20692566e-01 1.15616381e+00 5.91317117e-01 7.45349705e-01
-5.73883593e-01 -5.49184680e-01 -1.27769575e-01 7.78929532e-01
-1.17420053e+00 -4.31829929e-01 8.88997436e-01 -3.03796120e-02
1.22973263e+00 3.24272066e-01 8.68349910e-01 8.22374105e-01
2.56683677e-01 1.11320925e+00 6.64372385e-01 -7.55637407e-01
3.17997426e-01 1.19143188e-01 5.20785987e-01 7.52339363e-01
1.40068904e-01 -1.13896079e-01 -1.66511253e-01 -5.01280203e-02
8.00595582e-01 5.07108569e-01 -8.65535438e-02 1.70784414e-01
-8.30990374e-01 1.06446815e+00 2.55410790e-01 5.38775444e-01
-5.55482924e-01 -1.01640478e-01 6.64516389e-01 2.06204683e-01
3.22252840e-01 4.39262211e-01 -5.99376202e-01 -8.79559338e-01
-5.51798046e-01 -1.38373598e-01 1.20593584e+00 5.76565444e-01
4.54909474e-01 2.57101625e-01 1.66846737e-01 1.17178094e+00
4.24593002e-01 1.97265729e-01 1.10925376e+00 -8.00735712e-01
7.41608739e-02 1.26034200e+00 -3.17092419e-01 -1.08168316e+00
-4.72996831e-01 -3.29286188e-01 -1.39171100e+00 -2.24579126e-01
4.41253409e-02 -2.56006747e-01 -9.17968154e-01 1.34963918e+00
1.40690684e-01 4.15658094e-02 1.29621834e-01 7.93237388e-01
7.15231001e-01 9.29154336e-01 -4.65375893e-02 -3.56629133e-01
1.38243139e+00 -1.14776814e+00 -9.65373397e-01 -4.66269821e-01
9.04681087e-01 -4.29798871e-01 9.47772920e-01 5.69817841e-01
-6.74957991e-01 -5.87311149e-01 -9.82514322e-01 -1.65272862e-01
-4.47060019e-01 9.97858047e-02 1.07779980e+00 7.59790897e-01
-5.40527105e-01 2.72166669e-01 -6.57956600e-01 -2.99985200e-01
2.59116650e-01 6.52538300e-01 -5.84016182e-02 -1.53489932e-01
-1.34501326e+00 1.12850845e+00 8.45940173e-01 3.02997231e-01
-1.27959281e-01 -2.06692845e-01 -9.08107698e-01 2.84164816e-01
4.08280313e-01 -5.27765512e-01 1.40980601e+00 -1.16848826e+00
-1.78863811e+00 4.69886780e-01 -1.23268634e-01 -6.15715206e-01
8.45452100e-02 -1.72026739e-01 -5.74414968e-01 -2.85897374e-01
-5.51863194e-01 6.00831866e-01 8.52812409e-01 -6.28222942e-01
-8.38734269e-01 -3.59144777e-01 -8.54687244e-02 5.08178651e-01
-7.51779974e-01 -5.68858199e-02 -5.28895557e-01 -3.11722964e-01
5.78838736e-02 -7.63142705e-01 -3.08054775e-01 -4.51123267e-01
-2.70114187e-02 -6.48061812e-01 1.19527507e+00 -7.28478193e-01
1.37639570e+00 -2.16462064e+00 -9.64922234e-02 6.67980537e-02
3.15489262e-01 8.17747056e-01 5.78310825e-02 1.82366356e-01
2.09118947e-01 -1.98235005e-01 -1.65993795e-01 -1.41270980e-01
4.41360250e-02 6.66400492e-01 -1.31298542e-01 -4.61713783e-03
-9.90449563e-02 9.56180155e-01 -6.23451173e-01 -4.22041595e-01
4.52075869e-01 3.34805548e-01 -4.54983473e-01 5.33157349e-01
-1.26627192e-01 9.87831056e-02 -4.31763798e-01 3.13130319e-01
3.22842091e-01 -4.75248694e-01 1.40340820e-01 -7.64221847e-02
8.96894634e-02 4.00310725e-01 -7.36074686e-01 1.36835694e+00
-5.70042431e-01 4.77325052e-01 -9.12624449e-02 -1.49359179e+00
1.13445425e+00 4.33702856e-01 4.57650483e-01 -9.06275570e-01
4.46099609e-01 1.47413984e-01 4.35416907e-01 -7.21546888e-01
5.36156535e-01 -4.87283096e-02 -1.57216948e-03 5.65576136e-01
-1.29574850e-01 1.31720621e-02 3.40536307e-03 -1.68919846e-01
8.00477505e-01 -4.07977879e-01 5.16306818e-01 1.61027357e-01
7.12995052e-01 9.98872519e-02 6.27375841e-01 3.79646748e-01
-1.97367772e-01 1.42782748e-01 1.39585033e-01 -1.02835166e+00
-9.50114310e-01 -1.85764372e-01 6.50178194e-02 8.96401405e-01
8.53034556e-02 -2.28095040e-01 -8.03492904e-01 -6.77424490e-01
-2.63764948e-01 8.28680456e-01 -2.56837368e-01 -7.07134187e-01
-6.07685328e-01 -7.38147736e-01 4.94510114e-01 3.81522477e-01
1.25915265e+00 -1.47129631e+00 -7.55074263e-01 3.45750779e-01
-2.13336170e-01 -1.00150061e+00 -2.44432867e-01 3.08330923e-01
-9.25809205e-01 -8.04031372e-01 -4.83880252e-01 -1.03080738e+00
6.96767747e-01 2.03546479e-01 8.45684946e-01 6.50272906e-01
-3.53447050e-02 5.58221005e-02 -4.74542171e-01 -5.48622906e-01
-5.92642248e-01 4.15109932e-01 2.48027310e-01 -9.01165530e-02
1.04767907e+00 -2.43014932e-01 -3.30052704e-01 1.76291570e-01
-8.88803720e-01 6.64452493e-01 9.30610478e-01 1.06567931e+00
-6.45169541e-02 5.84536731e-01 6.24432504e-01 -6.37804985e-01
7.63379633e-01 -2.65325963e-01 -2.28094995e-01 2.23906979e-01
-9.17285681e-01 -1.41333202e-02 9.51439798e-01 -4.87274349e-01
-1.18075681e+00 -1.25512227e-01 -4.52244818e-01 -1.21207856e-01
-3.77603620e-01 9.70215201e-01 -2.30525002e-01 1.16248131e-01
2.77518276e-02 4.43649113e-01 4.41925317e-01 -5.00033975e-01
-8.41125920e-02 1.49098027e+00 3.22411865e-01 -7.53229856e-03
1.67067587e-01 -1.71970636e-01 -3.15663725e-01 -1.13918126e+00
-7.78685272e-01 -3.81405950e-01 -4.46290165e-01 -2.05894202e-01
6.74466074e-01 -5.59733927e-01 -8.59026730e-01 8.57993186e-01
-1.35919809e+00 -3.11450422e-01 -3.91868725e-02 7.06379235e-01
-2.82641888e-01 4.55653429e-01 -7.08225191e-01 -7.56682277e-01
-7.25116849e-01 -1.11858559e+00 4.60711092e-01 5.29385746e-01
-2.75698274e-01 -1.12527966e+00 -6.42380863e-02 6.38597786e-01
5.54439664e-01 -7.23019004e-01 1.00431418e+00 -1.08060718e+00
-2.52540380e-01 -6.00294709e-01 -4.05821949e-01 7.30807364e-01
3.25462729e-01 -4.96615797e-01 -7.60457695e-01 -7.53962249e-02
6.36964381e-01 -2.89453089e-01 4.86082733e-01 1.62644625e-01
1.21695840e+00 -7.51976907e-01 -7.88675472e-02 3.56179535e-01
9.60798323e-01 8.01015019e-01 8.81684959e-01 2.63465464e-01
7.27005184e-01 5.08165121e-01 5.16930044e-01 3.42275292e-01
5.40179253e-01 6.13104105e-01 1.02127522e-01 -3.28110643e-02
2.40672797e-01 -1.15214616e-01 3.14292192e-01 1.43473387e+00
5.99581487e-02 2.19417904e-02 -1.06552744e+00 1.32247105e-01
-2.03315163e+00 -9.51214612e-01 2.60049105e-01 1.87393486e+00
9.58465219e-01 1.62456319e-01 -1.42901868e-01 3.55102539e-01
5.07070839e-01 -6.75160885e-02 -5.70123911e-01 -7.02403963e-01
5.25580347e-01 -1.69818342e-01 1.64685592e-01 3.89403850e-01
-8.86088073e-01 9.00331557e-01 6.02009916e+00 9.07875478e-01
-1.53089833e+00 -1.64141804e-01 7.70671904e-01 5.80988169e-01
6.91553354e-02 -2.00212762e-01 -7.27959752e-01 5.99669814e-01
9.56350088e-01 -9.76809114e-02 5.79486430e-01 1.14286423e+00
7.47391060e-02 -1.48541227e-01 -8.92466187e-01 1.23263884e+00
2.60278732e-01 -1.24844205e+00 3.17061655e-02 2.17227846e-01
2.83399940e-01 -2.81733572e-01 -1.69992536e-01 7.61008680e-01
3.70230004e-02 -1.02575481e+00 -7.90744126e-02 4.27858829e-01
4.19463485e-01 -8.15326452e-01 1.19337535e+00 9.01932538e-01
-8.02736878e-01 -3.34257483e-01 -4.14717346e-01 -5.49310088e-01
-6.43233731e-02 4.55845505e-01 -1.07054996e+00 1.48150355e-01
4.11682487e-01 7.07638621e-01 -3.38787049e-01 8.45211983e-01
4.30859402e-02 7.49866545e-01 -3.84777963e-01 -5.13854384e-01
1.48204952e-01 -4.14346695e-01 1.02593563e-01 1.00880253e+00
3.60959284e-02 2.17230335e-01 3.18651080e-01 5.20848811e-01
-3.62136215e-02 -3.00543439e-02 -7.07651794e-01 -4.07174408e-01
3.99416268e-01 1.10060036e+00 -5.76032698e-01 -6.02374077e-01
-5.29779494e-01 1.00611913e+00 1.37359947e-01 -5.37046939e-02
-6.81452155e-01 -4.51045364e-01 3.53456616e-01 -4.00921226e-01
-8.74846056e-02 -3.70018572e-01 -6.02164745e-01 -8.90505075e-01
-2.32414380e-01 -1.17967832e+00 1.81710348e-01 -4.24386144e-01
-8.57946157e-01 3.22957546e-01 -2.09298328e-01 -9.31309462e-01
-6.48725629e-01 -4.83510315e-01 -6.72611237e-01 5.93302548e-01
-1.10793149e+00 -1.01420808e+00 -2.93655932e-01 2.76447803e-01
8.80644977e-01 -4.79974926e-01 1.06755948e+00 2.72690356e-01
-6.83001399e-01 5.31675160e-01 -1.10674739e-01 3.55202258e-01
2.06238732e-01 -7.75280774e-01 3.25082988e-01 6.07783914e-01
1.33520946e-01 6.17929220e-01 3.49377483e-01 -4.50392336e-01
-1.70601881e+00 -7.78801680e-01 9.99606669e-01 1.32197231e-01
3.16738665e-01 -2.22051382e-01 -1.05958450e+00 5.07591844e-01
2.25465491e-01 -6.40345156e-01 6.47646189e-01 1.37840509e-01
1.72811449e-01 -1.50998533e-01 -1.05632031e+00 7.96419024e-01
5.28943896e-01 -5.37702560e-01 -7.98692942e-01 7.02496767e-02
8.57727766e-01 -2.71123171e-01 -7.20078766e-01 5.95660627e-01
4.40547854e-01 -6.82991743e-01 6.58131242e-01 -3.36102784e-01
1.52086899e-01 6.96375445e-02 1.60675675e-01 -1.00702524e+00
-3.10357809e-01 -4.74896193e-01 -5.09342492e-01 8.81194532e-01
1.44944191e-01 -6.54875338e-01 1.05230927e+00 9.43881750e-01
-1.44762024e-01 -9.18336213e-01 -8.02775562e-01 -4.54780936e-01
-3.19878876e-01 -4.23404485e-01 5.46138227e-01 9.92258549e-01
3.76985759e-01 8.85454595e-01 -5.00522316e-01 -3.67890239e-01
1.55720875e-01 8.42008460e-03 7.66826391e-01 -1.35395014e+00
-2.30936810e-01 -5.50064683e-01 -3.81476879e-01 -1.50652301e+00
3.19668651e-02 -4.15972054e-01 1.86842009e-01 -1.63898826e+00
1.54243603e-01 -3.28559846e-01 -1.16523467e-01 7.14152694e-01
-5.40500358e-02 -1.09550998e-01 1.65044442e-02 3.07973832e-01
-6.76959753e-01 7.59389222e-01 1.20326042e+00 -2.20932677e-01
-3.65065098e-01 1.85471952e-01 -5.70100725e-01 9.28316534e-01
1.06293988e+00 -7.46853873e-02 -4.63148028e-01 -2.17999443e-01
6.73737079e-02 1.06988356e-01 -1.57941595e-01 -1.21557581e+00
5.24264693e-01 -2.26705417e-01 2.43723318e-01 -5.03889143e-01
3.88116479e-01 -1.18847215e+00 -1.02178708e-01 9.05308664e-01
-3.73887926e-01 -1.04498841e-01 2.76863705e-02 1.78416356e-01
-3.39674056e-01 -6.32847905e-01 6.43672347e-01 1.41775981e-02
-1.12628090e+00 1.58392057e-01 -6.43139124e-01 -3.05408746e-01
7.68973410e-01 -3.73366058e-01 -1.73896804e-01 -5.44045627e-01
-8.10440853e-02 2.33852237e-01 2.52707690e-01 5.56627333e-01
9.14964378e-01 -1.05612791e+00 -2.61711597e-01 6.62777305e-01
-1.26458988e-01 1.51943669e-01 1.89783946e-01 6.95878923e-01
-5.66501260e-01 7.59585023e-01 -1.15663141e-01 -3.43785048e-01
-1.29949653e+00 3.37092876e-01 1.50776491e-01 -4.51775759e-01
-6.94919169e-01 5.19561708e-01 -1.13028593e-01 -5.76704443e-01
7.73206294e-01 -9.89394486e-02 -5.62490046e-01 -4.25380588e-01
7.66852021e-01 2.94509798e-01 -1.33017614e-01 -3.42411876e-01
-5.98942861e-02 1.03351146e-01 -4.75778461e-01 1.19546145e-01
1.10126925e+00 4.07745019e-02 -3.51220995e-01 4.00019079e-01
1.09873986e+00 -7.20984995e-01 -6.57117188e-01 -1.21117413e-01
8.87194946e-02 -9.20077339e-02 4.15391922e-01 -8.29267740e-01
-9.39651132e-01 8.84028614e-01 5.39702952e-01 4.99357134e-01
1.04587281e+00 -5.87963581e-01 1.29210269e+00 1.02707267e+00
4.48948711e-01 -1.41999018e+00 1.80603296e-01 1.10942471e+00
5.68571031e-01 -1.42588747e+00 -1.22526716e-02 1.22503571e-01
-7.30161667e-01 1.10192227e+00 7.79727697e-01 2.80839175e-01
7.09552228e-01 2.04622075e-01 8.80056694e-02 -4.36405241e-02
-7.33073711e-01 1.90154433e-01 1.71144024e-01 3.74745667e-01
4.92075086e-01 -1.18309885e-01 -4.27467078e-01 6.36999965e-01
-8.35127085e-02 3.61535192e-01 1.92642525e-01 1.01337993e+00
-7.22964287e-01 -1.25097156e+00 -1.27108470e-01 9.70986545e-01
-2.55502671e-01 -9.80825499e-02 -4.25594777e-01 3.61634225e-01
-2.71253530e-02 1.07257402e+00 2.34934911e-01 -8.52099776e-01
3.98764461e-02 2.42404491e-01 1.13339625e-01 -5.58178902e-01
-4.57204163e-01 -2.73921520e-01 2.20260806e-02 -2.69350171e-01
-3.37403506e-01 1.85342021e-02 -1.53693581e+00 -7.14198291e-01
-5.72032213e-01 3.87052715e-01 9.44112301e-01 1.32680368e+00
3.59373778e-01 3.93957436e-01 6.35679722e-01 -6.06865108e-01
-8.18420649e-01 -1.23358727e+00 -3.08848798e-01 1.74823463e-01
3.27051282e-02 -3.35645884e-01 -1.75952002e-01 -9.96663943e-02] | [9.41183853149414, 6.135878086090088] |
92c95bfb-ce69-4555-b199-c22c3aa87cae | poisoning-web-scale-training-datasets-is | 2302.10149 | null | https://arxiv.org/abs/2302.10149v1 | https://arxiv.org/pdf/2302.10149v1.pdf | Poisoning Web-Scale Training Datasets is Practical | Deep learning models are often trained on distributed, webscale datasets crawled from the internet. In this paper, we introduce two new dataset poisoning attacks that intentionally introduce malicious examples to a model's performance. Our attacks are immediately practical and could, today, poison 10 popular datasets. Our first attack, split-view poisoning, exploits the mutable nature of internet content to ensure a dataset annotator's initial view of the dataset differs from the view downloaded by subsequent clients. By exploiting specific invalid trust assumptions, we show how we could have poisoned 0.01% of the LAION-400M or COYO-700M datasets for just $60 USD. Our second attack, frontrunning poisoning, targets web-scale datasets that periodically snapshot crowd-sourced content -- such as Wikipedia -- where an attacker only needs a time-limited window to inject malicious examples. In light of both attacks, we notify the maintainers of each affected dataset and recommended several low-overhead defenses. | ['Florian Tramèr', 'Kurt Thomas', 'Andreas Terzis', 'Hyrum Anderson', 'Will Pearce', 'Daniel Paleka', 'Christopher A. Choquette-Choo', 'Matthew Jagielski', 'Nicholas Carlini'] | 2023-02-20 | null | null | null | null | ['data-poisoning'] | ['adversarial'] | [-5.31303287e-01 1.03150658e-01 -1.58035353e-01 -1.18656285e-01
-7.35016882e-01 -1.27504587e+00 6.94028020e-01 1.52033508e-01
-6.76332712e-01 7.08963990e-01 -9.27837864e-02 -4.56883937e-01
3.29037637e-01 -7.19430923e-01 -1.11660576e+00 -3.62140119e-01
-3.44133019e-01 5.75273812e-01 6.78569734e-01 -6.31567324e-04
3.13285887e-01 4.42674518e-01 -7.59980381e-01 5.02210975e-01
3.84393394e-01 6.53303802e-01 -4.13135409e-01 6.21681094e-01
-6.62131459e-02 1.27162933e+00 -1.09003782e+00 -1.13684428e+00
7.21787333e-01 9.37619284e-02 -8.91220748e-01 -2.21233085e-01
4.65204686e-01 -1.08873081e+00 -6.68332577e-01 1.18621850e+00
3.86966914e-01 -6.02092743e-01 2.06525981e-01 -2.16502428e+00
-5.44441938e-01 1.17993104e+00 -9.00370598e-01 2.92658180e-01
2.79796608e-02 6.33743405e-01 7.04810262e-01 -1.89147919e-01
9.52297926e-01 9.45860803e-01 7.23481178e-01 6.30617380e-01
-9.99881327e-01 -9.11188483e-01 4.19887081e-02 -1.02386214e-01
-8.79366994e-01 -4.49911863e-01 5.68250775e-01 -4.18724954e-01
6.43091857e-01 2.37407133e-01 1.65785700e-01 1.90084755e+00
2.65001088e-01 5.87899268e-01 9.82411742e-01 9.45458040e-02
3.88482928e-01 4.58143055e-01 3.00992876e-01 5.89875817e-01
1.02225471e+00 -1.42316148e-01 -5.95982850e-01 -1.19009042e+00
3.19935113e-01 -2.06986547e-01 -2.71826595e-01 -5.83597004e-01
-1.10029805e+00 1.20384824e+00 3.59308600e-01 -1.36956140e-01
-3.32652479e-01 5.43714404e-01 9.32150662e-01 4.12296981e-01
3.03287894e-01 3.97926748e-01 -6.76235318e-01 -2.67148018e-01
-7.04428554e-01 7.16623366e-02 1.41202509e+00 1.00823247e+00
4.65074986e-01 -2.60916781e-02 6.31821990e-01 6.70598671e-02
3.32931459e-01 5.25122941e-01 5.14028430e-01 -1.09554803e+00
6.69734359e-01 3.53534013e-01 4.20620471e-01 -9.68276739e-01
-1.15965776e-01 1.33134261e-01 -2.30427325e-01 2.41533905e-01
8.05199444e-01 -5.12466192e-01 -3.37851137e-01 1.54206061e+00
5.62028825e-01 -1.61288291e-01 -1.41427591e-01 6.74581707e-01
2.09999979e-01 -3.08681075e-02 8.45801532e-02 7.11953565e-02
1.06752920e+00 -6.03172064e-01 -2.30898812e-01 2.29089692e-01
5.16055167e-01 -4.33495641e-01 1.02205348e+00 4.99323130e-01
-9.86898124e-01 2.80158877e-01 -9.59878683e-01 2.53350407e-01
-7.67090559e-01 -6.40915513e-01 4.47786033e-01 7.56830037e-01
-7.67170548e-01 4.76226598e-01 -8.83869529e-01 -5.72263479e-01
8.04918766e-01 1.73292741e-01 -5.20865977e-01 2.72324324e-01
-1.15572095e+00 7.09279418e-01 -3.92934345e-02 -7.37860501e-01
-1.59856939e+00 -6.15742981e-01 -1.04757540e-01 -1.39855385e-01
3.57257366e-01 -2.23846570e-01 1.42225373e+00 -6.03790641e-01
-7.25781620e-01 8.99017334e-01 5.93984842e-01 -9.94371235e-01
1.11654770e+00 -3.61215681e-01 -2.33370930e-01 5.84417045e-01
2.62314618e-01 1.04101025e-01 1.01583874e+00 -1.82214010e+00
-3.20958912e-01 -3.71599525e-01 2.75101125e-01 -3.44726413e-01
-9.46425736e-01 3.07775170e-01 -2.93853809e-04 -2.74064839e-01
-7.01450169e-01 -8.07681501e-01 -5.44757694e-02 2.13253155e-01
-9.28467512e-01 2.95848757e-01 1.24061239e+00 -4.98519629e-01
8.83921862e-01 -1.89833212e+00 -4.44211036e-01 3.09265345e-01
6.99736357e-01 4.47066665e-01 -2.61005491e-01 8.25541675e-01
2.61168808e-01 8.63060713e-01 9.60394442e-02 -1.47514775e-01
1.82548970e-01 1.09241799e-01 -5.93397975e-01 7.48642623e-01
-4.16641861e-01 6.41595602e-01 -8.47027123e-01 -3.91194314e-01
-3.42924029e-01 2.06486925e-01 -4.58809763e-01 2.24814400e-01
-6.21050894e-01 1.05488542e-02 -4.86364156e-01 5.09747088e-01
7.21929789e-01 -4.90366280e-01 3.12213659e-01 -1.28612950e-01
2.00972065e-01 2.23987982e-01 -7.64810622e-01 1.18418360e+00
-1.66286215e-01 5.34430385e-01 4.90207344e-01 -1.36863187e-01
3.82490575e-01 2.86646336e-01 5.58475375e-01 -2.96643198e-01
1.28251091e-01 1.73479542e-01 -2.25350022e-01 -6.34914696e-01
3.52048501e-02 4.64287281e-01 2.33430080e-02 1.33321834e+00
-3.16524446e-01 4.30155426e-01 -5.25161512e-02 8.81006539e-01
1.77838314e+00 -3.14420611e-01 1.51136611e-02 -7.26639479e-03
2.10874025e-02 3.88323814e-01 1.65585548e-01 1.22648656e+00
-7.53542542e-01 9.83272269e-02 1.06047106e+00 -7.55253315e-01
-1.32973683e+00 -1.18471885e+00 2.78640777e-01 1.08760428e+00
5.53170219e-02 -5.33003032e-01 -1.07814229e+00 -1.77528429e+00
4.95302111e-01 8.91551375e-01 -5.61659932e-01 -7.46216848e-02
-5.56103170e-01 -7.24599838e-01 1.27655625e+00 1.68032318e-01
7.63089538e-01 -8.80989492e-01 -7.88597822e-01 9.17075202e-02
-2.36773446e-01 -9.19973910e-01 -4.78270918e-01 2.61141330e-01
-3.22124124e-01 -1.75422418e+00 -9.42663327e-02 -9.37137231e-02
4.33920056e-01 4.03159320e-01 1.15696669e+00 4.43476021e-01
-3.22765082e-01 6.93543136e-01 -3.70298326e-01 -5.20764887e-01
-8.74217093e-01 2.29494125e-01 2.60245174e-01 -1.58835933e-01
7.76821554e-01 -7.35845447e-01 -3.33158851e-01 3.20012003e-01
-1.11702871e+00 -7.04600513e-01 4.07136977e-02 2.49003485e-01
-2.28914589e-01 6.85047731e-02 4.76355940e-01 -1.34586573e+00
7.93244302e-01 -1.07679927e+00 -6.24591351e-01 2.28125528e-01
-6.01754963e-01 -2.26253673e-01 7.73196161e-01 -8.93252015e-01
-6.27925694e-01 -2.46899500e-01 3.61177534e-01 -6.70419812e-01
-2.61385173e-01 8.82069692e-02 -1.32463813e-01 -2.19995171e-01
1.26402676e+00 8.32127109e-02 2.23808438e-01 -6.45278394e-01
4.91764098e-01 7.51508236e-01 3.58994603e-01 -7.45397031e-01
1.39215219e+00 7.73606360e-01 -6.82961345e-01 -5.13120532e-01
-4.16669935e-01 -8.80878232e-03 -3.06908667e-01 -9.33807120e-02
6.19561672e-01 -6.86580181e-01 -1.01990175e+00 8.38161349e-01
-1.44256175e+00 -6.59837723e-01 -1.14022903e-01 -1.09558940e-01
-1.72936410e-01 6.80251718e-01 -1.16151249e+00 -5.20529330e-01
-3.76292914e-01 -9.14895236e-01 3.85375530e-01 -2.43529201e-01
-1.32727668e-01 -8.46785665e-01 1.14366710e-01 4.24382120e-01
7.16254950e-01 3.08353812e-01 7.79512882e-01 -1.50344110e+00
-5.51579773e-01 -4.85910207e-01 -3.29461008e-01 1.94228977e-01
-5.63005432e-02 4.14425820e-01 -9.80693638e-01 -5.04756451e-01
1.54810563e-01 -6.47595525e-01 2.66168386e-01 -4.57041234e-01
1.10982466e+00 -1.10757065e+00 -2.77387649e-01 3.10564339e-01
1.45947039e+00 -9.23903286e-03 4.29716676e-01 7.80716181e-01
7.22770274e-01 3.28544587e-01 4.51437384e-03 7.94716001e-01
1.85966313e-01 1.47567689e-01 1.02218091e+00 2.08137840e-01
1.26699537e-01 -4.42793936e-01 4.62556630e-01 1.07886970e-01
3.99419188e-01 -5.38062274e-01 -7.83774793e-01 4.05596405e-01
-1.52834320e+00 -1.11925197e+00 -3.35860103e-01 2.14999413e+00
8.81079853e-01 4.62710798e-01 5.57257295e-01 -3.03933203e-01
8.28369141e-01 2.99726248e-01 -1.05320990e+00 -2.21537635e-01
2.41244867e-01 -3.96381944e-01 1.14340997e+00 3.03633630e-01
-1.03072977e+00 7.09235489e-01 6.34244299e+00 3.01450849e-01
-8.12091529e-01 4.98910815e-01 4.54952806e-01 -5.92047989e-01
-2.76309401e-01 -1.38820618e-01 -7.58831799e-01 9.57385540e-01
1.41381478e+00 -2.77823776e-01 6.43279731e-01 1.36084533e+00
1.47850057e-02 5.77960312e-02 -9.78859484e-01 4.26358253e-01
1.11903206e-01 -1.27643204e+00 5.67065999e-02 6.53434277e-01
4.40244138e-01 6.99937344e-01 6.50766343e-02 9.70017239e-02
1.36557889e+00 -5.84317565e-01 6.33913755e-01 8.96972343e-02
5.20107985e-01 -8.05647552e-01 4.82004702e-01 5.44735730e-01
-4.76401210e-01 -2.04248846e-01 -4.23841417e-01 3.54784846e-01
-2.03820784e-03 6.62541687e-01 -6.00239277e-01 -5.12504168e-02
9.15735185e-01 2.24901527e-01 -8.95075202e-01 6.46046996e-01
-8.70615169e-02 8.97502482e-01 -4.97794956e-01 1.17098719e-01
2.24878997e-01 3.66418451e-01 5.81558287e-01 1.02356637e+00
-3.16875398e-01 -3.10195535e-01 -8.84712040e-02 8.37990940e-01
-6.77859545e-01 -4.29935515e-01 -1.09457421e+00 -1.38780624e-01
8.83193314e-01 1.31747961e+00 -3.34923506e-01 -5.00796258e-01
-4.17971641e-01 1.09824800e+00 5.09149015e-01 2.87860811e-01
-1.26446939e+00 -3.10149401e-01 1.01395977e+00 4.44716483e-01
1.70807734e-01 -1.10982746e-01 -2.96750739e-02 -1.46685874e+00
3.82301994e-02 -1.54729664e+00 4.68873739e-01 -6.35122001e-01
-1.89029777e+00 5.13059199e-01 -6.61666244e-02 -6.89332187e-01
1.23338632e-01 -3.33578080e-01 -5.91588318e-01 3.45231146e-01
-1.23150420e+00 -1.17959344e+00 -1.53842226e-01 9.01348710e-01
1.23114631e-01 -4.51223910e-01 6.32211924e-01 1.01091012e-01
-6.52784228e-01 7.72215545e-01 2.76077896e-01 6.79934382e-01
8.49528968e-01 -1.09055066e+00 8.24145079e-01 8.23126674e-01
2.36837804e-01 8.66701901e-01 6.28670752e-01 -1.01139021e+00
-1.31082225e+00 -1.08121789e+00 5.95990717e-01 -9.87490237e-01
1.23873425e+00 -5.51733971e-01 -9.54018772e-01 1.19893873e+00
3.03372800e-01 1.43289953e-01 7.34428585e-01 -3.98011357e-01
-1.28752947e+00 -9.15478468e-02 -1.63210452e+00 5.00329435e-01
7.98805058e-01 -6.69889152e-01 -2.51558065e-01 8.88925791e-01
7.37721860e-01 3.52999754e-02 -8.05026889e-01 -3.49818379e-01
5.24154782e-01 -1.26633334e+00 7.11426497e-01 -1.40818858e+00
2.97487110e-01 2.60752663e-02 -2.90534705e-01 -9.73150015e-01
1.34037793e-01 -8.56187344e-01 -6.70479298e-01 1.29947257e+00
2.58181900e-01 -8.88204873e-01 9.62496638e-01 1.08300006e+00
7.05263317e-01 -1.32411078e-01 -7.24133611e-01 -9.45050359e-01
3.53799582e-01 -1.20981164e-01 7.37536311e-01 1.23585045e+00
-1.16769748e-03 -1.70742422e-01 -4.37582463e-01 4.08972681e-01
1.19227862e+00 -5.39363146e-01 1.23389912e+00 -1.04443359e+00
-3.40237588e-01 5.29685477e-03 4.75960746e-02 -4.65318233e-01
-1.65101200e-01 -4.79064673e-01 -4.32984591e-01 -9.86036360e-01
4.44801748e-01 -1.83676898e-01 -1.73331171e-01 8.86362791e-01
3.58825535e-01 2.08865494e-01 3.63792598e-01 6.87906384e-01
-7.56528854e-01 -1.03331573e-01 4.36893523e-01 -1.41500294e-01
3.18164945e-01 -1.96922049e-01 -8.08074713e-01 8.15053523e-01
9.91728485e-01 -1.03414774e+00 -4.53998536e-01 -4.70694333e-01
2.63172626e-01 -2.77770847e-01 7.02223718e-01 -9.20269489e-01
2.88362056e-01 -6.78222924e-02 1.81627780e-01 -3.54240835e-01
-1.38354659e-01 -1.09690011e+00 1.70103267e-01 7.88441718e-01
-6.27149463e-01 3.74126613e-01 -1.72153845e-01 8.31706107e-01
5.14906347e-01 -3.34190041e-01 9.85701263e-01 -7.17287958e-01
-1.26812890e-01 3.46483320e-01 -5.70320487e-01 5.02440989e-01
1.34003103e+00 3.24759811e-01 -1.27242959e+00 -5.41950047e-01
-4.52323407e-01 8.51820782e-02 1.09144485e+00 2.57236123e-01
2.01024398e-01 -8.16513896e-01 -6.21749461e-01 -1.83470007e-02
-1.00440241e-01 -5.36477745e-01 -2.16071650e-01 4.40980852e-01
-7.61721849e-01 -7.89416954e-02 -2.26012960e-01 -1.25170901e-01
-1.18401635e+00 1.19935405e+00 3.89355421e-01 -1.70987502e-01
-6.82871461e-01 6.59392595e-01 -2.90844828e-01 -2.41264045e-01
5.17377615e-01 3.33147883e-01 4.80154306e-01 -4.60977778e-02
8.30422878e-01 5.39373934e-01 -1.03347979e-01 -1.96194440e-01
-4.03687418e-01 -3.90333176e-01 -5.85578442e-01 -1.22863144e-01
1.32164037e+00 -2.53556781e-02 -2.15706199e-01 2.36167327e-01
1.45859385e+00 3.08909088e-01 -1.37792909e+00 -3.62664461e-02
-3.32189016e-02 -6.19862378e-01 -5.02831638e-01 -1.19911027e+00
-1.23729646e+00 5.69888175e-01 6.43070191e-02 8.45289648e-01
4.67747271e-01 -1.20436348e-01 1.19573545e+00 7.25611746e-01
8.59629869e-01 -8.58249068e-01 3.82142574e-01 -3.46495621e-02
5.93913138e-01 -1.22003615e+00 7.68926069e-02 1.94693636e-02
-6.57608867e-01 9.88131166e-01 1.01538825e+00 1.83836892e-02
6.73246980e-01 5.24027228e-01 3.82285804e-01 -4.56668317e-01
-1.04802501e+00 6.17507875e-01 -9.14753437e-01 8.43223333e-01
-2.70389646e-01 -2.11967945e-01 6.13757744e-02 3.37467015e-01
1.41912580e-01 -2.64561415e-01 1.14780033e+00 9.58699167e-01
-5.38888931e-01 -8.07472467e-01 -3.77158612e-01 5.73178411e-01
-7.00493813e-01 -4.44209576e-03 -8.39407921e-01 1.24704456e+00
-1.91796720e-01 1.00230038e+00 -1.51232600e-01 -4.35965121e-01
-1.66802090e-02 1.24256060e-01 -2.13348910e-01 -3.00444454e-01
-9.71424222e-01 -3.32592249e-01 2.75624692e-01 -9.10874665e-01
1.16495103e-01 -4.91997480e-01 -8.76378357e-01 -1.06054151e+00
-1.14814796e-01 2.80192107e-01 8.62511456e-01 3.13651145e-01
5.16444623e-01 -4.35556740e-01 9.49710727e-01 -3.42793584e-01
-1.30022812e+00 -6.69018686e-01 -6.78465486e-01 6.91656113e-01
1.06087886e-01 -3.60148132e-01 -1.04954135e+00 1.27953812e-01] | [5.82712459564209, 7.6086931228637695] |
70edc54b-b63a-4924-b24f-6a2dd5bea838 | a-compact-embedding-for-facial-expression | 1811.11283 | null | http://arxiv.org/abs/1811.11283v2 | http://arxiv.org/pdf/1811.11283v2.pdf | A Compact Embedding for Facial Expression Similarity | Most of the existing work on automatic facial expression analysis focuses on
discrete emotion recognition, or facial action unit detection. However, facial
expressions do not always fall neatly into pre-defined semantic categories.
Also, the similarity between expressions measured in the action unit space need
not correspond to how humans perceive expression similarity. Different from
previous work, our goal is to describe facial expressions in a continuous
fashion using a compact embedding space that mimics human visual preferences.
To achieve this goal, we collect a large-scale faces-in-the-wild dataset with
human annotations in the form: Expressions A and B are visually more similar
when compared to expression C, and use this dataset to train a neural network
that produces a compact (16-dimensional) expression embedding. We
experimentally demonstrate that the learned embedding can be successfully used
for various applications such as expression retrieval, photo album
summarization, and emotion recognition. We also show that the embedding learned
using the proposed dataset performs better than several other embeddings
learned using existing emotion or action unit datasets. | ['Aseem Agarwala', 'Raviteja Vemulapalli'] | 2018-11-27 | a-compact-embedding-for-facial-expression-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Vemulapalli_A_Compact_Embedding_for_Facial_Expression_Similarity_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Vemulapalli_A_Compact_Embedding_for_Facial_Expression_Similarity_CVPR_2019_paper.pdf | cvpr-2019-6 | ['action-unit-detection'] | ['computer-vision'] | [ 3.45049441e-01 -3.82340439e-02 -3.05865824e-01 -9.29163456e-01
-3.17873061e-01 -3.53470802e-01 4.70831186e-01 -6.64679753e-03
-2.93969989e-01 4.03893709e-01 5.57222545e-01 3.80597204e-01
3.07842195e-01 -3.95382047e-01 -2.19686449e-01 -7.77874112e-01
-8.22030976e-02 -6.39613420e-02 -6.00507796e-01 -1.35355547e-01
-2.44072042e-02 8.36400568e-01 -1.72196472e+00 3.09546620e-01
2.88752645e-01 1.56481457e+00 -5.90310097e-01 4.35244232e-01
-2.13344648e-01 9.96803284e-01 -6.52400792e-01 -5.20296991e-01
-1.64785783e-03 -6.53069675e-01 -7.61508703e-01 4.23931897e-01
7.12509036e-01 -3.84577841e-01 -1.16097800e-01 1.19828689e+00
4.40826714e-01 1.83995962e-01 1.03168237e+00 -1.65656626e+00
-8.62075090e-01 -1.81313828e-01 -6.11488223e-01 -2.80155510e-01
5.61742187e-01 -3.62564594e-01 1.05758536e+00 -9.67604697e-01
8.31436574e-01 1.41018915e+00 5.16360879e-01 9.33915675e-01
-1.02950799e+00 -7.45630205e-01 1.04806952e-01 8.95282254e-02
-1.35541022e+00 -5.47072113e-01 1.01709700e+00 -3.08569461e-01
8.39348733e-01 3.04116935e-01 7.81523347e-01 1.24114656e+00
-2.55612880e-01 1.02001214e+00 9.17038083e-01 -3.90950620e-01
2.33990982e-01 2.06043214e-01 -6.68042079e-02 7.54494011e-01
-4.10632581e-01 -4.78938252e-01 -6.31837547e-01 -2.52005011e-01
4.91306007e-01 1.89027146e-01 -3.26238543e-01 -4.16351348e-01
-7.66436398e-01 9.23991799e-01 5.13037980e-01 3.96065116e-01
-5.21775782e-01 4.17906374e-01 7.91980207e-01 4.30879712e-01
7.28903472e-01 3.61312211e-01 -1.34854436e-01 -4.48917598e-01
-7.52878010e-01 1.25778571e-01 5.41252792e-01 7.10616946e-01
9.74198699e-01 2.59713262e-01 -8.86478946e-02 1.16182458e+00
2.22793609e-01 2.76654303e-01 5.53558588e-01 -1.20780599e+00
-2.86089808e-01 7.73919284e-01 -3.81355211e-02 -1.45152605e+00
-2.38779992e-01 5.20755529e-01 -7.92013347e-01 3.31297398e-01
7.36884624e-02 -2.27939203e-01 -5.78028023e-01 2.06223130e+00
2.40275517e-01 -3.76364365e-02 2.80490369e-01 1.07523167e+00
8.94338787e-01 7.53436029e-01 1.97069600e-01 -2.46442422e-01
1.38233101e+00 -9.36030626e-01 -1.12151515e+00 -3.30577642e-02
9.78875399e-01 -5.66240191e-01 9.35232520e-01 1.10911332e-01
-7.37818718e-01 -3.73746783e-01 -1.01113033e+00 -2.65853375e-01
-5.48229337e-01 3.13794225e-01 9.46414411e-01 5.16011477e-01
-1.04753172e+00 2.48358160e-01 -3.42265517e-01 -6.56319201e-01
6.10050559e-01 1.83187842e-01 -1.04419231e+00 1.35368764e-01
-1.07103920e+00 8.79594028e-01 3.16680036e-02 1.62413478e-01
-3.79515946e-01 -3.53687406e-01 -1.42518759e+00 4.28442992e-02
-1.12314522e-01 -3.23247582e-01 1.17639756e+00 -1.97368085e+00
-1.66590893e+00 1.32522690e+00 -3.69908959e-01 -5.07729314e-02
-8.90979485e-04 -2.21249148e-01 -5.58432877e-01 4.63439226e-01
-2.37299249e-01 9.53359187e-01 9.95485306e-01 -1.07679176e+00
-1.65615931e-01 -8.24253738e-01 8.33378509e-02 2.18354493e-01
-9.51174676e-01 5.56223392e-01 -3.18593591e-01 -5.84689915e-01
-3.00890446e-01 -8.53608668e-01 7.13972077e-02 7.96050608e-01
-4.27207211e-03 -5.13214648e-01 1.12591469e+00 -2.31268778e-01
1.21041155e+00 -2.32817793e+00 2.68002599e-01 1.39425218e-01
2.83578247e-01 1.39654696e-01 -2.83424944e-01 3.28168243e-01
-4.11199838e-01 8.43486786e-02 -1.86383292e-01 -5.76923847e-01
3.36603045e-01 3.62937063e-01 -2.05624476e-01 5.93407989e-01
5.07723272e-01 9.49749351e-01 -8.96752238e-01 -6.21167421e-01
-1.12746712e-02 5.73061526e-01 -5.19988477e-01 3.49657685e-01
1.91274330e-01 3.56812850e-02 -4.42107141e-01 9.57320631e-01
5.13176262e-01 4.14920971e-02 2.16966987e-01 -4.33516413e-01
2.27217838e-01 -4.44313765e-01 -7.50415385e-01 1.64858890e+00
-5.11456788e-01 1.18180978e+00 1.33692101e-01 -1.23004019e+00
1.16934133e+00 3.90343755e-01 7.25933433e-01 -6.55374050e-01
3.83552998e-01 -4.79704700e-02 -3.27801973e-01 -6.41060829e-01
5.62163889e-01 -3.78343672e-01 -1.94381744e-01 6.79535627e-01
3.31018984e-01 -4.52946685e-02 -4.52049309e-03 4.26822826e-02
7.73210227e-01 -7.87215084e-02 5.05726755e-01 1.82743091e-02
4.45701450e-01 -3.48405361e-01 6.05816901e-01 4.70052660e-02
-5.73989451e-01 4.14539963e-01 9.22270119e-01 -5.65227926e-01
-7.42299378e-01 -8.20697308e-01 -1.59723431e-01 1.44371760e+00
-6.31005019e-02 -7.02226460e-01 -7.36412346e-01 -7.17645943e-01
1.33378670e-01 3.13498795e-01 -1.19939184e+00 -2.58241773e-01
-1.04542300e-01 -2.38279060e-01 6.99644387e-01 8.13753426e-01
3.17780107e-01 -1.31256247e+00 -5.59999883e-01 -1.99512109e-01
1.61139086e-01 -1.20328188e+00 -5.04868150e-01 -9.38672870e-02
-2.65353143e-01 -9.59406018e-01 -7.47543216e-01 -9.09506023e-01
9.68313396e-01 7.09036179e-03 9.70573843e-01 -2.16532592e-02
-5.94730318e-01 8.95708084e-01 -6.26939774e-01 -6.34130359e-01
-2.16912940e-01 -8.17797184e-01 2.34475181e-01 6.95387185e-01
8.86452556e-01 -3.38112622e-01 -4.82359022e-01 9.00030136e-02
-1.09706461e+00 -3.48307341e-01 2.46800870e-01 8.78047585e-01
4.46457565e-01 -4.80115056e-01 5.47065914e-01 -5.42649508e-01
8.84293497e-01 -1.97008640e-01 1.94095410e-02 4.40086663e-01
-1.56901240e-01 -1.55199245e-01 3.42067689e-01 -6.85846984e-01
-8.24244916e-01 2.56526262e-01 9.84099414e-03 -8.48764658e-01
-1.46405667e-01 2.78101593e-01 -2.13102117e-01 -2.58335084e-01
3.94834697e-01 6.35257289e-02 5.21228135e-01 -6.07325733e-02
7.21994042e-01 9.03632462e-01 3.30987394e-01 -4.26825106e-01
1.63883492e-01 6.29621625e-01 -7.70106390e-02 -1.16171098e+00
-6.46753728e-01 -5.40008783e-01 -7.26765335e-01 -4.22498375e-01
1.18929482e+00 -7.57281721e-01 -8.98197114e-01 2.94737726e-01
-1.30414355e+00 1.37716150e-02 -1.93177715e-01 2.84398288e-01
-8.68431747e-01 4.58685130e-01 -5.56136727e-01 -1.01867568e+00
-3.61609310e-01 -9.50062931e-01 1.46337593e+00 1.79299399e-01
-7.47082174e-01 -1.02459359e+00 1.00463130e-01 -6.97040185e-02
3.16000044e-01 7.24003494e-01 7.71825969e-01 -3.31321359e-01
3.65867168e-01 -4.51950192e-01 -2.17050552e-01 7.14954197e-01
3.82655770e-01 4.54578221e-01 -1.08938396e+00 3.17943934e-03
-1.63750768e-01 -1.16852868e+00 7.70196259e-01 1.19291842e-02
1.49721491e+00 -4.11158800e-01 -5.20361327e-02 6.46660447e-01
1.02657914e+00 -6.25785142e-02 7.20062852e-01 -3.64661738e-02
4.89070147e-01 9.55043912e-01 5.52978516e-01 6.87141061e-01
4.05025296e-02 8.55971992e-01 3.26169848e-01 -3.77156168e-01
4.03131962e-01 -1.82546929e-01 5.55606067e-01 4.56107438e-01
-1.71765443e-02 8.44493806e-02 -5.16634524e-01 3.65194291e-01
-1.91138291e+00 -1.21545625e+00 5.53009450e-01 1.61872256e+00
8.05632055e-01 -6.00108147e-01 6.24499843e-02 -7.77713507e-02
4.05405074e-01 5.20409763e-01 -3.99390161e-01 -1.19326246e+00
-2.37103447e-01 2.20725402e-01 -1.40678197e-01 1.02966830e-01
-1.14282715e+00 1.09280252e+00 6.44747782e+00 5.26934505e-01
-1.55030203e+00 -5.58804646e-02 5.47198176e-01 -1.43954858e-01
-7.33254105e-02 -4.73302960e-01 -2.77006060e-01 6.07153326e-02
6.18801177e-01 -3.30769211e-01 -3.77950668e-02 1.17731977e+00
-2.18760967e-02 1.96357176e-01 -1.34244800e+00 1.59581435e+00
6.27545416e-01 -1.09001887e+00 3.53604138e-01 -1.85393840e-01
4.79584187e-01 -7.27170289e-01 1.83096617e-01 4.81000155e-01
-1.64070532e-01 -1.45809948e+00 4.12492335e-01 4.95498240e-01
9.98981357e-01 -8.55384290e-01 5.47121108e-01 -1.33492827e-01
-9.97532964e-01 -2.25065891e-02 -6.01420701e-01 -1.56103268e-01
1.87217623e-01 -2.95912586e-02 -5.14811695e-01 1.62810534e-01
6.16921723e-01 1.07600880e+00 -3.13032329e-01 3.00171137e-01
-1.05802692e-01 2.06086874e-01 -5.13903499e-02 -3.45715344e-01
3.66334170e-01 -3.18621725e-01 1.51763290e-01 1.48545539e+00
2.42467210e-01 2.49367669e-01 -4.92402129e-02 7.38839746e-01
-4.00792688e-01 4.71415192e-01 -1.10233128e+00 -6.43673062e-01
5.88893145e-02 1.56185484e+00 -1.57883912e-01 -4.24571127e-01
-6.61467016e-01 1.50417423e+00 5.06372154e-01 4.77506727e-01
-7.18637705e-01 -4.06034231e-01 1.53464699e+00 -2.93745130e-01
1.63834214e-01 6.39636591e-02 4.94874984e-01 -1.10409129e+00
-1.18717588e-02 -8.43595088e-01 2.92615265e-01 -1.00739849e+00
-1.46737611e+00 8.13964128e-01 -1.72983557e-01 -1.18806434e+00
-5.70628583e-01 -1.11613584e+00 -6.17284238e-01 5.79272628e-01
-1.27965510e+00 -1.21976185e+00 -5.30407310e-01 7.94967771e-01
2.90114969e-01 -7.96477646e-02 1.57823741e+00 1.56261370e-01
-5.51359892e-01 9.05988693e-01 -1.11460693e-01 4.02413458e-01
9.51909244e-01 -1.04326427e+00 -4.05877560e-01 1.52921155e-01
3.65011454e-01 6.69982612e-01 5.10558486e-01 9.73327979e-02
-1.33994091e+00 -9.41199064e-01 6.86070442e-01 -2.14728117e-01
7.57207334e-01 -4.47015792e-01 -7.51991510e-01 6.06001914e-01
3.89567256e-01 4.90409344e-01 1.33116436e+00 1.92576036e-01
-7.31972098e-01 -2.41206273e-01 -1.15838385e+00 6.08778536e-01
9.32499349e-01 -9.43074048e-01 -3.84167939e-01 2.22899884e-01
2.80812949e-01 -5.36307096e-02 -9.58195686e-01 4.44003642e-01
9.35940027e-01 -9.01175141e-01 6.99941218e-01 -1.38915634e+00
6.38152778e-01 5.50949536e-02 -5.33700824e-01 -1.29400635e+00
-6.19414710e-02 -5.24863780e-01 -2.69847121e-02 1.05361426e+00
-1.43435374e-02 -3.72769982e-01 6.91678703e-01 6.99921072e-01
2.78036624e-01 -9.44556236e-01 -8.63005698e-01 -4.98076797e-01
-1.36157453e-01 -2.12458953e-01 4.93610382e-01 1.35124290e+00
3.86069208e-01 3.39519322e-01 -4.49959427e-01 -2.72605598e-01
2.33062431e-01 3.60798180e-01 9.90783691e-01 -1.02094781e+00
2.11435288e-01 -7.25531578e-01 -1.28176928e+00 -9.40648198e-01
9.95667219e-01 -8.60609353e-01 -1.94088832e-01 -1.16769004e+00
2.11167306e-01 2.06777513e-01 -4.43080157e-01 6.86498344e-01
1.15275703e-01 5.11476099e-01 1.94078133e-01 -1.39273509e-01
-7.49377787e-01 1.10435534e+00 1.01134527e+00 -3.20985705e-01
2.52301127e-01 -5.88094234e-01 -7.52347171e-01 8.36572647e-01
5.12984812e-01 1.31578348e-03 -3.13231975e-01 -1.46987408e-01
1.98440086e-02 -1.74929276e-01 2.24447787e-01 -4.85338151e-01
-1.01729445e-01 -1.94125220e-01 5.77288330e-01 -6.16734438e-02
8.57577741e-01 -9.59319651e-01 -2.63971895e-01 -1.59735262e-01
-5.03125489e-01 1.42924473e-01 2.61941999e-01 3.76421452e-01
-7.90430307e-01 -1.07074201e-01 7.60718882e-01 2.02183545e-01
-1.16378653e+00 3.99886876e-01 -3.22893173e-01 -2.40811363e-01
1.23833215e+00 -4.94825125e-01 -2.01423895e-02 -8.19145858e-01
-7.67324209e-01 9.67180207e-02 5.22474587e-01 7.65461445e-01
8.75172317e-01 -1.92304707e+00 -5.44844449e-01 1.44910678e-01
8.83847117e-01 -6.31361485e-01 1.86083272e-01 6.69166028e-01
-2.55823731e-01 4.11046185e-02 -6.87467635e-01 -5.23854077e-01
-1.70733643e+00 3.89428109e-01 4.18494612e-01 2.61484355e-01
-2.25477368e-01 9.24312830e-01 3.47860247e-01 -4.44666475e-01
2.92094082e-01 2.92070806e-02 -3.82591188e-01 5.00030398e-01
9.57951963e-01 -1.25005707e-01 -4.98904556e-01 -1.25612223e+00
-4.03301269e-01 7.00715303e-01 1.06856905e-01 1.35213718e-01
1.33309925e+00 1.24049507e-01 -4.24462765e-01 6.37649298e-01
2.04345441e+00 -9.84724686e-02 -8.95303249e-01 -2.54004359e-01
-1.24358930e-01 -6.04750037e-01 -1.98408186e-01 -2.39867717e-01
-1.07662523e+00 1.17829108e+00 6.44414783e-01 -1.51251638e-02
1.30746603e+00 1.18077293e-01 4.65552241e-01 5.55739760e-01
4.24921252e-02 -1.31830442e+00 5.91175497e-01 3.39427799e-01
1.34988427e+00 -1.32295394e+00 -1.58314750e-01 -1.23170614e-01
-1.11763752e+00 1.38247073e+00 6.99227333e-01 -1.65591352e-02
6.61315084e-01 5.24491630e-02 4.60482478e-01 -4.54836458e-01
-6.63393199e-01 -2.96654612e-01 2.63094366e-01 6.12619400e-01
7.41290212e-01 1.06407002e-01 -1.02774136e-01 4.98955131e-01
-8.94165561e-02 -1.52755454e-01 2.81294078e-01 8.35968614e-01
-4.02242422e-01 -8.77784848e-01 4.68447320e-02 2.30191648e-01
-4.55514401e-01 2.44631320e-01 -1.05835152e+00 6.82458520e-01
-2.25479826e-01 6.65704250e-01 3.69591773e-01 -2.73269713e-01
3.29189867e-01 3.34236741e-01 7.03012347e-01 -4.99136358e-01
-8.42304379e-02 -3.02517951e-01 1.15318760e-01 -9.54936504e-01
-7.99913883e-01 -2.88242310e-01 -1.31431890e+00 -1.96504951e-01
1.52982160e-01 1.70342922e-02 6.08279467e-01 6.75302029e-01
3.83628547e-01 -6.58809347e-03 7.25355744e-01 -9.60780621e-01
-3.44381988e-01 -9.87088919e-01 -8.23251545e-01 1.12584662e+00
2.88921922e-01 -9.14245188e-01 -2.78828859e-01 9.76978242e-02] | [13.582874298095703, 1.795093297958374] |
e5f7ea30-8538-46f7-8e59-b32cae40aa76 | fishdreamer-towards-fisheye-semantic | 2303.13842 | null | https://arxiv.org/abs/2303.13842v2 | https://arxiv.org/pdf/2303.13842v2.pdf | FishDreamer: Towards Fisheye Semantic Completion via Unified Image Outpainting and Segmentation | This paper raises the new task of Fisheye Semantic Completion (FSC), where dense texture, structure, and semantics of a fisheye image are inferred even beyond the sensor field-of-view (FoV). Fisheye cameras have larger FoV than ordinary pinhole cameras, yet its unique special imaging model naturally leads to a blind area at the edge of the image plane. This is suboptimal for safety-critical applications since important perception tasks, such as semantic segmentation, become very challenging within the blind zone. Previous works considered the out-FoV outpainting and in-FoV segmentation separately. However, we observe that these two tasks are actually closely coupled. To jointly estimate the tightly intertwined complete fisheye image and scene semantics, we introduce the new FishDreamer which relies on successful ViTs enhanced with a novel Polar-aware Cross Attention module (PCA) to leverage dense context and guide semantically-consistent content generation while considering different polar distributions. In addition to the contribution of the novel task and architecture, we also derive Cityscapes-BF and KITTI360-BF datasets to facilitate training and evaluation of this new track. Our experiments demonstrate that the proposed FishDreamer outperforms methods solving each task in isolation and surpasses alternative approaches on the Fisheye Semantic Completion. Code and datasets are publicly available at https://github.com/MasterHow/FishDreamer. | ['Rainer Stiefelhagen', 'Kaiwei Wang', 'Huajian Ni', 'Yaozu Ye', 'Alina Roitberg', 'Kunyu Peng', 'Jiaming Zhang', 'Kailun Yang', 'Yu Li', 'Hao Shi'] | 2023-03-24 | null | null | null | null | ['image-outpainting'] | ['computer-vision'] | [ 2.29068339e-01 8.98836106e-02 2.47102365e-01 -4.43376154e-01
-7.54670262e-01 -8.57197523e-01 5.36861897e-01 -6.13341987e-01
-2.98926771e-01 3.37037981e-01 5.41368961e-01 -7.06548914e-02
-2.17525840e-01 -3.90530258e-01 -9.56781149e-01 -7.21932113e-01
3.53909492e-01 -9.28019732e-02 3.93378228e-01 -6.82420731e-02
2.42902756e-01 2.22917750e-01 -1.62864506e+00 1.68362096e-01
9.05381918e-01 9.17693138e-01 7.43981659e-01 6.08805060e-01
4.41249907e-01 6.59818530e-01 -1.44615144e-01 -4.89911228e-01
6.14477694e-01 -1.83586076e-01 -4.16982353e-01 4.24186915e-01
9.41241860e-01 -6.70040548e-01 -4.53861684e-01 1.20955396e+00
2.49771148e-01 3.55881035e-01 2.78419852e-01 -1.04597628e+00
-5.42673051e-01 3.49227302e-02 -5.42003989e-01 -2.25020438e-01
6.26210928e-01 2.44271815e-01 8.28557909e-01 -9.38595235e-01
5.45285404e-01 1.12111986e+00 5.82152367e-01 4.06397551e-01
-5.16339600e-01 -6.00952983e-01 5.45946777e-01 1.81650430e-01
-1.16796136e+00 -7.11573184e-01 7.52921820e-01 -3.34143043e-01
5.22576988e-01 -2.75351088e-02 6.76988780e-01 1.16900849e+00
-3.02626882e-02 8.08935165e-01 1.19047868e+00 -1.99763283e-01
1.74588323e-01 1.05026029e-02 -2.08309397e-01 8.03744197e-01
7.25069493e-02 2.97431231e-01 -5.17736256e-01 3.14067185e-01
8.97865415e-01 3.78657103e-01 -8.97228718e-01 -6.38011217e-01
-1.23944390e+00 6.87409580e-01 5.04121602e-01 -2.85389662e-01
7.91108534e-02 1.92984015e-01 -3.23001176e-01 -1.89145178e-01
3.40189755e-01 3.72402430e-01 -3.81668091e-01 9.33702737e-02
-8.56397808e-01 3.33334565e-01 6.83575392e-01 1.39120781e+00
6.70730352e-01 -2.12330952e-01 1.94549501e-01 7.33197927e-01
5.00462651e-01 7.97164261e-01 -1.58443555e-01 -1.38789511e+00
5.77163935e-01 2.86986500e-01 4.78020698e-01 -8.34932446e-01
-2.65837878e-01 -2.85834193e-01 -5.81313372e-01 3.11901838e-01
6.61886811e-01 -2.23279178e-01 -1.19748342e+00 1.47756183e+00
4.02126044e-01 3.85842204e-01 -7.25962892e-02 1.45156419e+00
7.47467160e-01 5.24829030e-01 -1.60065994e-01 1.89366132e-01
1.36466753e+00 -1.37748265e+00 -6.20583594e-01 -6.74498737e-01
2.09228814e-01 -8.12783718e-01 1.06127918e+00 5.42163789e-01
-9.09299433e-01 -1.69552356e-01 -9.40009058e-01 -4.64607805e-01
-2.65658408e-01 2.80794442e-01 6.38368726e-01 4.57525820e-01
-1.15369487e+00 -1.22040333e-02 -5.65730333e-01 -3.58849496e-01
4.17111278e-01 1.25490809e-02 -5.88028073e-01 -7.68100262e-01
-7.64139116e-01 7.12305248e-01 1.09149240e-01 3.47476840e-01
-1.00101209e+00 -6.90826654e-01 -1.13003278e+00 -1.23454846e-01
8.30225050e-01 -8.43902290e-01 1.13090682e+00 -7.50149727e-01
-1.51786208e+00 6.27754688e-01 -2.10366860e-01 -2.36535370e-01
4.05381888e-01 -6.41185999e-01 -4.14910987e-02 4.45558935e-01
1.45931154e-01 8.71646345e-01 1.17581975e+00 -1.49667025e+00
-7.56800771e-01 -5.00382483e-01 4.29823369e-01 6.39225423e-01
7.99952075e-02 -5.31092733e-02 -1.07923985e+00 -7.23630130e-01
1.70147866e-01 -9.48816180e-01 -9.03078169e-02 2.96386361e-01
-5.58238626e-01 2.25476727e-01 7.21698642e-01 -9.32783484e-01
7.79991031e-01 -2.17524004e+00 2.83237875e-01 -1.50797829e-01
3.96087974e-01 -7.19426423e-02 2.93364376e-02 1.88742131e-01
2.20340028e-01 -3.64216924e-01 -5.98849237e-01 -6.47996962e-01
-1.53401956e-01 1.67155847e-01 -4.56846923e-01 7.62423337e-01
-3.27277966e-02 7.34622121e-01 -9.76856411e-01 -1.89666077e-01
6.82378471e-01 4.15650904e-01 -7.94639111e-01 4.01152462e-01
-1.97636783e-01 6.58667922e-01 -3.50018710e-01 1.01515889e+00
9.84458148e-01 -1.54084608e-01 -5.47455847e-02 -1.84581786e-01
-1.71113506e-01 -2.10677788e-01 -1.05793083e+00 2.27184796e+00
-6.79386437e-01 5.35411894e-01 5.03390729e-01 -7.51888692e-01
5.23685813e-01 -1.32865295e-01 1.50407299e-01 -6.31802559e-01
5.45371734e-02 -8.34148377e-03 -6.38611913e-01 -6.39267206e-01
6.46380365e-01 1.85151510e-02 -2.74707049e-01 5.78569621e-02
7.48575479e-02 -4.49710339e-01 -3.07011575e-01 2.64322162e-01
8.79085302e-01 3.90763581e-01 2.54393532e-03 -4.25669961e-02
3.32429886e-01 -1.06227227e-01 5.66135287e-01 5.96664250e-01
-1.99383810e-01 1.23419976e+00 2.80006170e-01 -1.98355064e-01
-1.02693284e+00 -1.16507888e+00 -1.09846629e-01 7.75507748e-01
9.47224259e-01 -3.41071904e-01 -9.30335283e-01 -7.61525929e-01
-1.28807247e-01 6.25568867e-01 -6.00714326e-01 1.98171943e-01
-3.35735172e-01 -3.20131749e-01 1.50520116e-01 2.64309824e-01
7.55449772e-01 -6.26930058e-01 -7.76278675e-01 -3.22305888e-01
-4.81059104e-01 -1.72809124e+00 -7.85121679e-01 -1.45606011e-01
-4.66938406e-01 -1.37330329e+00 -8.27206731e-01 -6.36347353e-01
5.98430037e-01 8.98225904e-01 8.95889044e-01 -4.32423264e-01
-1.15024909e-01 7.59521484e-01 -5.19438446e-01 -2.66529411e-01
3.16839546e-01 -2.81703115e-01 -1.68426126e-01 1.30030721e-01
3.34471837e-02 -4.55767810e-01 -1.03788495e+00 4.13506359e-01
-9.06632781e-01 4.03508216e-01 6.35421515e-01 7.68286824e-01
5.29922128e-01 -3.54131125e-02 1.11269010e-02 -6.34640336e-01
-1.20915443e-01 -6.84784234e-01 -8.78357351e-01 2.08173871e-01
-2.86267966e-01 -1.56500489e-01 4.72364515e-01 -4.95769009e-02
-1.31234896e+00 4.49640781e-01 2.76688635e-02 -7.82874346e-01
-3.03078890e-01 -1.88669890e-01 -6.33124292e-01 -1.25802130e-01
7.37017393e-02 1.68318450e-01 -3.84666547e-02 -5.25105655e-01
5.81479967e-01 5.82495391e-01 7.47060299e-01 -2.59106666e-01
6.95805967e-01 9.49867189e-01 -3.56695056e-01 -8.84709775e-01
-1.13988674e+00 -7.65763521e-01 -3.01958323e-01 -2.80464828e-01
1.07256806e+00 -1.28296900e+00 -6.79979622e-01 7.43820965e-01
-1.16625595e+00 -4.83715147e-01 -7.24149169e-03 5.19572616e-01
-5.03513336e-01 6.01497650e-01 7.73323774e-02 -7.26900220e-01
3.64648849e-02 -1.37506187e+00 1.68699408e+00 2.72651136e-01
3.35142791e-01 -8.55533004e-01 -2.57779986e-01 9.25722539e-01
2.21135151e-02 1.11207373e-01 3.24043036e-01 -3.15639041e-02
-1.11675513e+00 1.84579402e-01 -5.83787739e-01 4.61488634e-01
-2.97371238e-01 -4.64695692e-01 -1.26243389e+00 -1.72378808e-01
1.06883727e-01 -2.69881725e-01 1.09848607e+00 6.24517858e-01
1.17287838e+00 -1.91785812e-01 -1.21784426e-01 1.35962200e+00
1.49417639e+00 -1.52182495e-02 7.14277029e-01 3.22147280e-01
8.50525618e-01 9.34994578e-01 8.11446369e-01 4.90340710e-01
8.33389044e-01 8.20924759e-01 9.72674131e-01 -2.77223196e-02
-4.45930690e-01 -3.98126483e-01 5.30023992e-01 2.47532487e-01
-1.06775342e-02 -5.73992372e-01 -6.82229757e-01 6.42414391e-01
-1.68226719e+00 -6.67266071e-01 -1.04138635e-01 2.15302515e+00
4.66139406e-01 -4.39881474e-01 -2.77170628e-01 -3.02701890e-01
4.99322981e-01 3.31053674e-01 -5.49630821e-01 2.62677401e-01
-3.30054581e-01 -1.25860557e-01 8.32567513e-01 8.42224181e-01
-1.18691468e+00 1.14604878e+00 5.22402048e+00 7.37752199e-01
-7.91840971e-01 2.38539219e-01 4.04243112e-01 -3.23499590e-01
-3.80139351e-01 1.74880579e-01 -9.83436823e-01 5.62447071e-01
3.81915122e-01 7.22496331e-01 6.92128360e-01 5.84686279e-01
1.96012795e-01 -6.69710577e-01 -9.53777254e-01 1.29435468e+00
3.36432099e-01 -1.13327968e+00 -2.10190535e-01 1.23657919e-01
7.85487771e-01 2.08981305e-01 2.30707049e-01 -1.72667578e-01
3.32269162e-01 -9.42534149e-01 9.82609630e-01 7.53577471e-01
1.06054688e+00 -5.00062287e-01 4.73375499e-01 5.29513061e-02
-9.80033398e-01 -4.83015537e-01 -2.70478070e-01 2.60392159e-01
2.67700404e-01 6.20296001e-01 -5.56259334e-01 5.96627414e-01
1.07516468e+00 1.19644475e+00 -5.35113394e-01 9.58023846e-01
-4.31485921e-01 1.24599606e-01 -3.09687763e-01 5.06128788e-01
3.83111477e-01 -3.62309575e-01 7.68281996e-01 8.22237551e-01
5.75981975e-01 2.28717446e-01 -1.44567400e-01 9.12981331e-01
1.37690725e-02 -3.55006427e-01 -7.29912937e-01 3.32792073e-01
4.86646533e-01 1.12579393e+00 -4.54615355e-01 1.16652213e-02
-7.52725184e-01 1.30985510e+00 6.87594013e-03 6.66951358e-01
-9.17791188e-01 -3.10877889e-01 8.85908782e-01 2.67232955e-01
5.72692037e-01 -2.04137579e-01 -2.68205792e-01 -1.53017068e+00
2.32484564e-01 -6.21673346e-01 -4.52727824e-02 -1.28717756e+00
-8.63971353e-01 3.91046345e-01 6.20094314e-03 -1.46091020e+00
5.93940318e-02 -6.39554262e-01 -2.71024585e-01 8.06081414e-01
-1.86334550e+00 -1.53188562e+00 -8.60818446e-01 1.10489821e+00
8.31358135e-01 -8.47790837e-02 4.82571691e-01 9.62253436e-02
-5.27942419e-01 4.50346828e-01 2.25601777e-01 -1.66955829e-01
9.83209074e-01 -1.09508872e+00 6.25763983e-02 1.34498799e+00
-1.78220376e-01 4.35240418e-01 5.32179892e-01 -5.59632540e-01
-1.65805936e+00 -1.32475638e+00 4.21968877e-01 -7.81299949e-01
4.86768544e-01 -5.49774647e-01 -4.14360046e-01 6.23836815e-01
5.07318527e-02 1.64765865e-01 1.79391116e-01 -3.21957022e-01
-2.98347652e-01 -7.92373046e-02 -8.88762176e-01 5.83568931e-01
1.39182174e+00 -5.41333556e-01 -3.28141630e-01 3.35049242e-01
8.67598891e-01 -5.54948747e-01 -5.64029396e-01 2.49887764e-01
5.65875232e-01 -1.41085827e+00 1.15240932e+00 1.26971364e-01
5.93705475e-01 -3.85305554e-01 -5.31670868e-01 -1.12136257e+00
3.16010416e-01 -7.36491978e-01 1.97309554e-02 9.87385094e-01
2.09401846e-02 -5.65935314e-01 6.30365551e-01 5.11771977e-01
-6.59784734e-01 -6.10030234e-01 -6.98779941e-01 -3.99203032e-01
-3.28044146e-01 -6.17988586e-01 4.06655222e-01 7.98630238e-01
-3.54376614e-01 9.47931334e-02 -8.08406055e-01 6.21618867e-01
9.11491573e-01 1.24109223e-01 7.82501757e-01 -9.25306141e-01
-2.89864421e-01 1.49598271e-02 -2.05488458e-01 -1.51270962e+00
1.39948884e-02 -5.76205075e-01 4.14100677e-01 -1.53684080e+00
1.52831078e-02 -4.48727816e-01 2.44649157e-01 2.82962412e-01
-9.14411321e-02 2.99159318e-01 1.65499985e-01 2.95953244e-01
-7.48089373e-01 6.80160046e-01 1.47878098e+00 5.59943058e-02
-1.52431682e-01 -2.91416682e-02 -9.18182790e-01 8.53243291e-01
4.48248506e-01 -1.70254052e-01 -8.49837184e-01 -9.13126349e-01
2.35546470e-01 3.48542511e-01 9.12900150e-01 -8.37993741e-01
3.66201162e-01 4.14919518e-02 2.17191830e-01 -4.30646420e-01
6.96808100e-01 -9.89283979e-01 -7.07526654e-02 -1.98946640e-01
-3.43537517e-02 -2.98433840e-01 -5.86828291e-02 1.09021890e+00
-3.07236195e-01 -5.80569729e-02 7.02732623e-01 -1.51501596e-01
-1.15885019e+00 4.06826258e-01 -1.11157008e-01 2.29205996e-01
1.04472315e+00 -4.22967225e-01 -4.34653133e-01 -5.03776073e-01
-5.80376685e-01 2.78002709e-01 8.91585767e-01 4.58040297e-01
1.12508702e+00 -8.16597223e-01 -5.59434235e-01 4.75679725e-01
2.96471298e-01 6.20522976e-01 7.22364128e-01 1.04252112e+00
-5.78090787e-01 4.20720875e-01 1.78491359e-03 -5.24987400e-01
-9.56878662e-01 4.26600248e-01 2.98733801e-01 1.80366695e-01
-1.04559433e+00 1.12263095e+00 9.65403914e-01 -4.46999460e-01
4.12668914e-01 -3.72548819e-01 -7.17260092e-02 -2.58679777e-01
4.69642341e-01 2.38651082e-01 -8.52670297e-02 -6.07763529e-01
-2.43958637e-01 7.93310106e-01 1.04878858e-01 -1.68201327e-01
1.32741976e+00 -6.70161009e-01 4.80673537e-02 -2.41118327e-01
9.36575353e-01 1.51835725e-01 -1.98854089e+00 -2.03287676e-01
-5.55503368e-01 -8.43603611e-01 2.63002813e-01 -8.85847032e-01
-1.06157553e+00 1.01363695e+00 4.93401229e-01 -4.99783307e-01
1.45022333e+00 1.77044347e-01 9.40862358e-01 1.57212272e-01
3.62239301e-01 -9.67252195e-01 4.06182196e-04 4.08836305e-01
1.09481740e+00 -1.40591156e+00 -8.72855708e-02 -7.14057088e-01
-8.31752479e-01 9.42787707e-01 5.91773033e-01 -1.03270024e-01
5.76060355e-01 1.63200974e-01 -6.25139028e-02 -1.57352269e-01
-3.88246685e-01 -2.05701068e-01 3.46770078e-01 9.47290063e-01
-2.94297636e-01 -1.66664943e-01 5.56468606e-01 7.79312074e-01
-1.97775528e-01 -2.72127211e-01 6.78472638e-01 5.93207359e-01
-2.64603078e-01 -6.04189098e-01 -5.31344652e-01 1.20585427e-01
-1.30042583e-01 -4.00185198e-01 -2.44191244e-01 8.48265469e-01
3.13044280e-01 1.08601296e+00 7.28078783e-02 -3.98509502e-01
1.09625213e-01 -4.02462244e-01 4.89335388e-01 -5.19725442e-01
-1.26044424e-02 2.88578749e-01 1.31123751e-01 -1.03445876e+00
-5.40656090e-01 -5.95729828e-01 -9.53026235e-01 -5.90501130e-02
-1.04189828e-01 -2.20825672e-01 7.11221755e-01 8.32876623e-01
3.47388119e-01 2.81758159e-01 6.39765382e-01 -1.14497352e+00
-1.69267163e-01 -7.45309234e-01 -7.95911849e-01 5.81832193e-02
6.57778323e-01 -8.90320659e-01 -4.53706861e-01 2.56198883e-01] | [8.843080520629883, -2.5760562419891357] |
fd74cdc3-0b4b-4f6d-a0f2-2f2f9195245c | a-speech-corpus-for-chronic-kidney-disease | 2211.01705 | null | https://arxiv.org/abs/2211.01705v1 | https://arxiv.org/pdf/2211.01705v1.pdf | A speech corpus for chronic kidney disease | In this study, we present a speech corpus of patients with chronic kidney disease (CKD) that will be used for research on pathological voice analysis, automatic illness identification, and severity prediction. This paper introduces the steps involved in creating this corpus, including the choice of speech-related parameters and speech lists as well as the recording technique. The speakers in this corpus, 289 CKD patients with varying degrees of severity who were categorized based on estimated glomerular filtration rate (eGFR), delivered sustained vowels, sentence, and paragraph stimuli. This study compared and analyzed the voice characteristics of CKD patients with those of the control group; the results revealed differences in voice quality, phoneme-level pronunciation, prosody, glottal source, and aerodynamic parameters. | ['Minhwa Chung', 'Sejoong Kim', 'Jiwon Ryu', 'Myeong Ju Kim', 'Sunhee Kim', 'Jihyun Mun'] | 2022-11-03 | null | null | null | null | ['severity-prediction'] | ['computer-vision'] | [-2.25098059e-01 -2.88185924e-01 8.35773498e-02 -2.49451160e-01
-4.96746033e-01 -1.10172197e-01 6.97893053e-02 5.02107859e-01
-1.92267969e-01 4.78871971e-01 9.84317601e-01 -4.50599104e-01
-1.79222003e-01 -4.69238281e-01 3.45008850e-01 -4.18113619e-01
-2.86006361e-01 5.04402816e-01 -4.78348076e-01 1.79316089e-01
1.61748901e-01 4.72376049e-01 -1.16592467e+00 2.60205418e-01
1.00960767e+00 3.56941432e-01 5.47984362e-01 1.24268591e+00
-5.28441548e-01 9.77744699e-01 -7.10550249e-01 -1.12128556e-02
-4.70606357e-01 -9.12693262e-01 -7.16714323e-01 1.21949680e-01
2.58798718e-01 -3.14064503e-01 -3.89056742e-01 6.08663797e-01
1.28084576e+00 -1.25069931e-01 7.54433215e-01 -2.38226712e-01
-6.08977854e-01 8.24813604e-01 5.06924391e-01 6.63876772e-01
3.39866251e-01 1.91176504e-01 4.26778615e-01 -6.01696014e-01
2.86453336e-01 1.25134254e+00 6.86348498e-01 1.05804074e+00
-1.27950609e+00 -3.72603416e-01 -4.34834301e-01 2.31430784e-01
-1.21445358e+00 -7.14767873e-01 6.58898234e-01 -1.02174795e+00
9.39677060e-01 2.96997339e-01 9.27587330e-01 9.31690931e-01
1.06931798e-01 6.86812401e-02 5.95398545e-01 -6.01580381e-01
2.16785327e-01 2.29226202e-01 5.20807207e-01 3.01620841e-01
-1.68202743e-01 5.82096912e-02 -1.14813291e-01 -4.07305032e-01
6.87575877e-01 -2.87459791e-01 -7.33521819e-01 -1.43484073e-02
-1.31215501e+00 5.86237311e-01 -3.50700170e-01 7.86229968e-01
-7.25800395e-01 3.78687382e-02 6.60002172e-01 4.09316421e-01
6.49551079e-02 9.99529753e-03 -4.81425047e-01 -5.49539328e-01
-2.47208700e-01 -5.83733320e-02 7.83588231e-01 4.34509784e-01
-4.19155538e-01 2.73499906e-01 -5.32868087e-01 1.47600424e+00
5.25162041e-01 9.01104629e-01 6.94056213e-01 -1.03626847e+00
4.27711338e-01 2.73492336e-01 -9.16812941e-02 -5.94169319e-01
-5.37904918e-01 -1.13680542e-01 -6.63536608e-01 -1.46027952e-01
2.74606884e-01 -1.04430154e-01 -8.99474680e-01 1.45845664e+00
-1.58179700e-01 -3.21493804e-01 5.74808061e-01 6.39580429e-01
7.81311631e-01 2.57123768e-01 4.89108115e-01 -7.52927363e-01
1.55172229e+00 -5.77832282e-01 -1.09689796e+00 5.73495887e-02
1.57447889e-01 -8.24725211e-01 1.32695019e+00 1.83142155e-01
-1.07479274e+00 -5.29059708e-01 -3.93094391e-01 2.27323264e-01
3.34507406e-01 1.47511646e-01 3.16144861e-02 1.39554298e+00
-8.53818476e-01 7.08176732e-01 -9.21142697e-01 -4.94434774e-01
-1.86117750e-03 -7.35530555e-02 6.24951944e-02 4.61752385e-01
-1.06823909e+00 1.03869402e+00 -3.25963467e-01 -3.03055663e-02
-4.16445911e-01 -8.67400408e-01 -8.21325064e-01 1.17757723e-01
-6.56846225e-01 -1.00031376e+00 1.46547174e+00 -2.38739893e-01
-1.58082974e+00 7.83274949e-01 -4.53117132e-01 -2.05866992e-01
3.49385053e-01 -2.00807363e-01 -7.72978544e-01 5.18902481e-01
-1.39732465e-01 -3.69501382e-01 3.79140377e-01 -8.26655865e-01
-2.42744625e-01 -5.12369812e-01 -8.86238635e-01 2.22178251e-01
1.39454842e-01 2.51808792e-01 2.80201137e-01 -6.12593412e-01
2.55755961e-01 -3.80220234e-01 -1.64514586e-01 -1.59953028e-01
-2.58633584e-01 -1.78337783e-01 1.24814093e-01 -1.05977631e+00
1.41159487e+00 -2.65460515e+00 -1.92247987e-01 1.44032583e-01
2.27470100e-01 2.81586260e-01 3.50384146e-01 6.37399137e-01
-1.34936884e-01 5.86083591e-01 -3.89714330e-01 -3.07373315e-01
4.36534956e-02 4.12667155e-01 3.30724977e-02 4.98082250e-01
-1.65944889e-01 3.36904436e-01 -7.46877432e-01 -5.48762381e-01
9.19369400e-01 5.45652449e-01 -2.36521259e-01 2.93249160e-01
4.88928884e-01 3.88341397e-01 -2.29090005e-01 6.15816772e-01
3.17117125e-01 3.30625176e-01 6.58192411e-02 -1.50175756e-02
-3.82267922e-01 6.57844484e-01 -7.48393059e-01 9.02657032e-01
-3.54350716e-01 5.94151318e-01 2.74145901e-01 -6.47809982e-01
9.08093870e-01 8.94322693e-01 7.07341015e-01 -4.25443888e-01
3.56579423e-02 4.47314650e-01 3.60924691e-01 -1.30965602e+00
-2.02411994e-01 -6.56568825e-01 3.09052587e-01 -6.16264157e-02
-3.22721243e-01 -5.83106466e-02 1.70583293e-01 -1.64885640e-01
7.90380597e-01 -9.85919952e-01 4.37037885e-01 -4.35181975e-01
1.03937864e+00 -3.70506018e-01 2.72479177e-01 5.44959843e-01
-8.02570879e-01 6.77174032e-01 4.12756175e-01 1.86861549e-02
-7.67508984e-01 -1.50899291e+00 -7.10660338e-01 3.38118583e-01
-4.36860353e-01 -2.33503327e-01 -6.32385790e-01 1.13966659e-01
3.83717567e-01 7.74059772e-01 -1.88479833e-02 5.23959892e-03
-6.74641013e-01 -3.33083659e-01 9.66287971e-01 4.61306453e-01
-1.34431213e-01 -1.42140162e+00 -1.14786059e-01 4.60978538e-01
-3.69319081e-01 -6.39392138e-01 -6.15400076e-01 -1.29086403e-02
-1.04046726e+00 -1.29570365e+00 -8.08767080e-01 -1.10281229e+00
1.63631037e-01 -1.91943184e-01 8.00767422e-01 -1.33086145e-01
-5.75845361e-01 6.83039963e-01 -1.33667067e-01 -3.71379286e-01
-1.00655115e+00 -4.10208195e-01 2.56568611e-01 -5.19247234e-01
2.79712260e-01 -7.73864746e-01 -5.64429283e-01 -1.31470159e-01
-2.08521411e-01 -5.95090270e-01 3.21552157e-01 2.96888977e-01
4.53662395e-01 -1.05258040e-01 8.49364817e-01 -4.14863884e-01
1.18691087e+00 -2.27692693e-01 3.04840505e-01 -9.45002288e-02
-8.51381600e-01 -3.01600605e-01 4.62671489e-01 -2.34698668e-01
-6.54963493e-01 -3.21277440e-01 -5.14047146e-01 -1.80372477e-01
-5.14102817e-01 3.57035041e-01 -7.80592412e-02 4.19731081e-01
8.90895009e-01 5.33002555e-01 6.51023328e-01 -8.14420164e-01
5.26592880e-02 1.36278737e+00 1.06738186e+00 -2.90681690e-01
1.97570875e-01 -7.35306293e-02 -5.99875987e-01 -1.44049418e+00
1.83746442e-01 -6.97868526e-01 -1.87714979e-01 -2.42013797e-01
1.15145695e+00 -6.69910431e-01 -7.56327808e-01 9.86040115e-01
-8.10980082e-01 -2.16342852e-01 -4.51254547e-01 1.28815722e+00
-5.38789690e-01 7.41088390e-01 -1.10424602e+00 -1.33109224e+00
-1.00929737e+00 -8.73509109e-01 5.13407826e-01 7.74948895e-02
-5.63088715e-01 -1.06781244e+00 3.77918512e-01 6.05834246e-01
7.34059811e-01 1.42651558e-01 1.74607539e+00 -4.42344666e-01
3.39411557e-01 2.32519343e-01 2.25697502e-01 7.42474258e-01
6.61504388e-01 6.45138174e-02 -4.36711639e-01 1.85038477e-01
4.63374346e-01 2.11791292e-01 4.23918456e-01 8.43491852e-01
6.91266179e-01 -1.16555728e-01 2.10217297e-01 1.87142372e-01
1.25249672e+00 6.17742956e-01 6.93466365e-01 -2.52470672e-01
3.46280068e-01 6.89884126e-01 -2.71906614e-01 3.45674723e-01
3.18524957e-01 5.22246882e-02 -7.23821223e-02 4.12688643e-01
-4.39339250e-01 1.75402001e-01 5.21674633e-01 1.39628732e+00
-7.03546777e-02 -2.08080932e-01 -1.13273406e+00 8.58106494e-01
-8.44424188e-01 -1.10350609e+00 -8.67756009e-01 1.87641478e+00
7.79487133e-01 -3.33054721e-01 2.55578190e-01 4.35893834e-01
1.17820644e+00 -6.69158399e-02 -1.34902343e-01 -6.40131652e-01
2.26764649e-01 4.22634870e-01 2.51482785e-01 8.21826816e-01
-6.53336167e-01 4.98698279e-02 7.75197554e+00 -1.07491523e-01
-1.19107318e+00 -3.56767595e-01 -7.94523507e-02 1.68090895e-01
-2.41884142e-01 -6.00879192e-01 2.04089843e-02 5.46783328e-01
1.29287350e+00 -3.45414579e-01 4.37413007e-01 5.91353059e-01
1.02224970e+00 -1.20786682e-01 -7.97689378e-01 9.09891367e-01
-6.80082589e-02 -8.61096442e-01 -2.12187737e-01 -1.50343433e-01
-3.30796838e-01 2.08444461e-01 -4.07489151e-01 1.04659416e-01
-7.42188811e-01 -9.62528348e-01 5.99568427e-01 7.85873473e-01
7.64568806e-01 -2.40552127e-01 8.04313958e-01 9.04550478e-02
-1.03189838e+00 -1.29308313e-01 -3.32661867e-02 -1.11536793e-01
4.77941334e-01 6.96360588e-01 -8.90062630e-01 2.20303416e-01
3.17354888e-01 2.13884845e-01 -3.71168070e-02 1.16960883e+00
7.03203678e-02 9.76776421e-01 -1.64611176e-01 -1.64441794e-01
-7.26497352e-01 -1.62893623e-01 7.66430497e-01 1.16770053e+00
3.92521977e-01 2.54842937e-01 -1.37548059e-01 8.65938187e-01
4.61730748e-01 6.09583497e-01 -1.01651527e-01 -4.48261887e-01
6.93216503e-01 4.08455163e-01 -2.99689531e-01 -2.73999840e-01
-2.54082054e-01 5.63714921e-01 -5.23414433e-01 1.83624968e-01
-1.51678145e-01 -3.35343480e-01 1.01956213e+00 2.38968477e-01
-1.97532505e-01 -2.64786184e-02 -7.59224415e-01 -7.63038695e-01
1.55301526e-01 -7.95179427e-01 3.68340701e-01 -4.62392449e-01
-1.47723818e+00 1.06765799e-01 -2.27999508e-01 -6.51966810e-01
-1.97161108e-01 -3.21497381e-01 -7.00406790e-01 1.52936816e+00
-1.04525077e+00 -2.86236197e-01 -2.93066531e-01 3.75146478e-01
6.35937512e-01 -3.24537642e-02 1.31329727e+00 8.30309987e-01
-5.10148644e-01 2.35361438e-02 -4.85974066e-02 3.35953951e-01
3.12808424e-01 -1.56960833e+00 8.70389491e-02 1.17153674e-01
-6.80112302e-01 8.92297804e-01 7.68535674e-01 -7.98847914e-01
-9.59494174e-01 -7.89778173e-01 1.61647832e+00 -1.50712430e-01
3.50798994e-01 2.15892419e-01 -7.66018569e-01 -5.87340407e-02
-2.03662701e-02 -5.70908964e-01 1.07596374e+00 -2.10896805e-01
2.94846892e-01 1.13177001e-02 -1.42820525e+00 3.01119715e-01
8.23600292e-01 -9.91414666e-01 -9.54476058e-01 9.79170725e-02
5.39291680e-01 -3.13894004e-01 -1.34716797e+00 2.69066602e-01
6.78049147e-01 -6.63406610e-01 9.70439792e-01 -5.84877491e-01
1.75649509e-01 -6.23379759e-02 -1.82561338e-01 -1.12242067e+00
-4.89679605e-01 -4.08595771e-01 -9.39042401e-03 1.15550923e+00
2.81611264e-01 -6.21903062e-01 4.20561105e-01 6.67849660e-01
-6.49457872e-01 -3.71145129e-01 -6.46145880e-01 -5.34311533e-01
2.55442739e-01 -5.15125632e-01 1.83911212e-02 6.02833152e-01
5.67828476e-01 4.58993584e-01 2.26083584e-02 1.11804254e-01
3.63945574e-01 -3.93206149e-01 7.97039792e-02 -1.29649556e+00
5.79849780e-02 -6.86983407e-01 -4.60684806e-01 -1.12075076e-01
-4.85206187e-01 -7.19785452e-01 -5.48139922e-02 -2.16142106e+00
-1.78164035e-01 -1.41895667e-01 1.58058442e-02 6.00822531e-02
-3.65849704e-01 -5.72512567e-01 1.79343849e-01 2.01893687e-01
8.67321312e-01 3.07155669e-01 1.00491154e+00 2.70426031e-02
-8.85449767e-01 5.46068311e-01 -6.24903619e-01 3.80553126e-01
1.01474440e+00 -3.98609579e-01 -6.87460899e-01 2.57922381e-01
-6.10117733e-01 5.66863179e-01 1.41431928e-01 -9.15853620e-01
-2.62067437e-01 -2.34085873e-01 3.03873103e-02 -5.94312847e-01
-1.79299340e-03 -6.48452878e-01 5.13055563e-01 1.16937375e+00
-5.76620638e-01 -5.25104403e-02 -2.27311492e-01 3.99172902e-01
-1.76696095e-03 -5.53973317e-01 9.30275798e-01 -2.06705436e-01
-1.34280613e-02 -1.34083092e-01 -1.30124271e+00 2.44831085e-01
6.93890274e-01 -5.74344732e-02 -4.59016293e-01 -3.96146715e-01
-1.45239723e+00 2.24918593e-02 1.41979113e-01 2.78502464e-01
6.82601631e-01 -1.05016708e+00 -9.24179792e-01 1.86637208e-01
-6.01059124e-02 -3.40814233e-01 4.72633302e-01 1.37415516e+00
-1.11286938e+00 5.04553795e-01 7.36162066e-02 -5.11183202e-01
-1.46668732e+00 2.71764278e-01 9.13663149e-01 2.27473855e-01
-1.09047163e+00 3.07615101e-01 -2.91464388e-01 -2.43897542e-01
5.00417590e-01 -7.38485336e-01 -5.44945419e-01 -8.95425025e-03
6.14167094e-01 6.13685489e-01 -3.37870196e-02 -3.46058398e-01
-4.21778232e-01 2.23981336e-01 6.44820571e-01 -4.57200892e-02
7.94709980e-01 -6.21808648e-01 -2.68278152e-01 8.11032832e-01
1.05658257e+00 4.66296524e-01 -6.23655431e-02 1.56175524e-01
3.51048224e-02 -2.05152452e-01 -2.41777226e-01 -6.91699624e-01
-7.97738969e-01 6.13133192e-01 1.21665180e+00 6.19855165e-01
8.39979172e-01 1.08673625e-01 9.32411015e-01 1.04206704e-01
-1.40624523e-01 -9.32474017e-01 -2.54697233e-01 3.22825313e-01
8.23369503e-01 -3.55115116e-01 -6.77305281e-01 -6.26482248e-01
-6.86631560e-01 1.21639085e+00 1.06332235e-01 6.84089065e-02
8.17859530e-01 1.16703082e-02 9.44876969e-01 -1.43565863e-01
-4.00960505e-01 -3.34347337e-01 -2.10921615e-01 1.16801691e+00
7.66673803e-01 3.59361261e-01 -1.03013670e+00 7.10709870e-01
-5.69542766e-01 -6.10132031e-02 7.47301400e-01 5.04189849e-01
-1.03702116e+00 -9.42347169e-01 -3.50679994e-01 9.11715448e-01
-8.76012146e-01 -8.78139511e-02 -3.85152668e-01 3.17744911e-01
-1.77243222e-02 1.03382778e+00 -1.22836813e-01 -1.55996859e-01
9.42917168e-01 7.36003220e-01 2.51103491e-01 -6.53421044e-01
-6.18443787e-01 5.84383719e-02 6.57363176e-01 1.32308668e-02
-4.34035331e-01 -1.00683534e+00 -1.78780055e+00 -1.89583808e-01
-6.15046844e-02 3.91541988e-01 7.33147800e-01 8.90499592e-01
8.98224243e-04 9.52912986e-01 4.99686956e-01 -5.19011617e-02
-7.59726942e-01 -1.43030894e+00 -9.03119206e-01 2.02230826e-01
8.31327319e-01 -7.88031593e-02 -7.47509658e-01 1.58314392e-01] | [14.059332847595215, 5.459327697753906] |
d3385574-307b-4778-91c6-c31990429cea | hope-hierarchical-object-prototype-encoding | null | null | http://openaccess.thecvf.com/content_cvpr_2017/html/Yu_HOPE_Hierarchical_Object_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Yu_HOPE_Hierarchical_Object_CVPR_2017_paper.pdf | HOPE: Hierarchical Object Prototype Encoding for Efficient Object Instance Search in Videos | This paper tackles the problem of efficient and effective object instance search in videos. To effectively capture the relevance between a query and video frames and precisely localize the particular object, we leverage the object proposals to improve the quality of object instance search in videos. However, hundreds of object proposals obtained from each frame could result in unaffordable memory and computational cost. To this end, we present a simple yet effective hierarchical object prototype encoding (HOPE) model to accelerate the object instance search without sacrificing accuracy, which exploits both the spatial and temporal self-similarity property existing in object proposals generated from video frames. We design two types of sphere k-means methods, i.e., spatially-constrained sphere k-means and temporally-constrained sphere k-means to learn frame-level object prototypes and dataset-level object prototypes, respectively. In this way, the object instance search problem is cast to the sparse matrix-vector multiplication problem. Thanks to the sparsity of the codes, both the memory and computational cost are significantly reduced. Experimental results on two video datasets demonstrate that our approach significantly improves the performance of video object instance search over other state-of-the-art fast search schemes.
| ['Junsong Yuan', 'Yuwei Wu', 'Tan Yu'] | 2017-07-01 | null | null | null | cvpr-2017-7 | ['instance-search'] | ['computer-vision'] | [-1.37535238e-03 -4.89456326e-01 -3.34864199e-01 -2.80292183e-01
-9.09028113e-01 -3.06946188e-01 3.07716757e-01 1.21980406e-01
-2.93434650e-01 2.33322233e-01 2.74543054e-02 4.44421142e-01
-2.88898647e-01 -4.11723405e-01 -7.50398099e-01 -7.04427600e-01
-9.23352614e-02 1.76213905e-01 7.06175864e-01 3.01350027e-01
4.63150859e-01 4.52898413e-01 -1.89419484e+00 1.94795713e-01
6.62951887e-01 1.56411171e+00 7.02715456e-01 3.57124925e-01
-3.85459438e-02 8.14709246e-01 -3.29836071e-01 -1.27690583e-01
3.58395010e-01 -1.24733217e-01 -6.20669901e-01 5.19336283e-01
6.78328514e-01 -5.17168581e-01 -7.68829048e-01 1.04248106e+00
3.93527865e-01 5.66200793e-01 3.21412474e-01 -1.17399049e+00
-4.66753840e-01 1.72137722e-01 -5.64753950e-01 5.31143248e-01
4.28739727e-01 1.02298148e-01 7.81752765e-01 -1.52442241e+00
5.00038266e-01 1.27413058e+00 3.36196810e-01 2.18392715e-01
-8.46205592e-01 -6.20012939e-01 2.44526133e-01 7.58034050e-01
-1.89390588e+00 -6.40986502e-01 7.71440387e-01 -4.03509170e-01
6.54655695e-01 4.30292398e-01 9.91301298e-01 3.42481107e-01
-3.90981734e-01 1.01043856e+00 4.94668812e-01 -3.24906898e-04
3.57145965e-01 -1.32052517e-02 -5.01687638e-02 7.01943219e-01
1.79451242e-01 -1.50789181e-02 -8.49375129e-01 -3.39507282e-01
1.00911593e+00 4.01284307e-01 -5.14335692e-01 -6.10195577e-01
-1.56200588e+00 7.10411668e-01 4.76890951e-01 1.23665571e-01
-5.99069297e-01 2.80413449e-01 4.14061099e-01 -2.26334497e-01
1.20908171e-01 1.20573491e-01 -2.43786424e-01 -1.08972341e-01
-1.19917727e+00 2.05405176e-01 4.58693653e-01 1.25784731e+00
8.53277445e-01 -1.32668033e-01 -4.57006276e-01 7.05196142e-01
5.32609582e-01 6.49364412e-01 3.65457982e-01 -1.02919888e+00
3.19649190e-01 6.14866138e-01 1.47062749e-01 -1.45978904e+00
3.98786925e-02 -1.90342948e-01 -6.02719545e-01 -4.08196449e-01
9.44258720e-02 5.27702451e-01 -5.88681519e-01 1.12600911e+00
9.42429125e-01 9.59424555e-01 -6.33975044e-02 1.31849205e+00
1.08052421e+00 8.65783513e-01 -1.42610997e-01 -4.76583749e-01
1.49088848e+00 -1.29523838e+00 -6.31445050e-01 9.22264680e-02
3.87113899e-01 -8.38533759e-01 6.80711746e-01 8.76861587e-02
-1.50069833e+00 -7.52516985e-01 -7.39461362e-01 1.87971834e-02
1.75884262e-01 2.97393799e-01 6.33061290e-01 1.68028742e-01
-6.51216447e-01 3.21898878e-01 -9.26397681e-01 4.21282388e-02
6.30414844e-01 4.17124629e-01 -2.30593875e-01 -3.20773184e-01
-7.96041250e-01 2.29958728e-01 6.64283097e-01 6.92761689e-02
-9.97323573e-01 -8.52490485e-01 -7.16871798e-01 2.53529042e-01
6.82773232e-01 -3.67828876e-01 1.04979444e+00 -7.76110947e-01
-1.11241257e+00 5.69519162e-01 -4.72684443e-01 -2.64145464e-01
6.80272281e-02 -2.04555139e-01 -1.99005514e-01 6.42096341e-01
1.68915659e-01 6.79692507e-01 1.14350271e+00 -1.09335911e+00
-9.42393482e-01 -2.23366991e-01 -2.25525293e-02 2.93237746e-01
-5.18354475e-01 2.25540891e-01 -1.31260800e+00 -8.40334475e-01
4.44394618e-01 -8.90694797e-01 -2.42431700e-01 4.52145964e-01
1.93003919e-02 -4.28990543e-01 9.60511148e-01 -5.41669846e-01
1.42296016e+00 -2.28371000e+00 1.97584942e-01 2.87033468e-01
1.23890720e-01 3.89800906e-01 -1.13388874e-01 6.73883110e-02
2.12585002e-01 -4.66316789e-01 1.01894261e-02 -4.72241156e-02
-2.69468129e-01 2.09531978e-01 -2.71892458e-01 7.23988473e-01
5.95085844e-02 1.02175677e+00 -1.04856884e+00 -9.49648142e-01
2.49955654e-01 5.85504591e-01 -8.01181793e-01 3.77353400e-01
4.09547891e-03 2.17423543e-01 -7.31529593e-01 8.28460038e-01
6.61421835e-01 -6.18948877e-01 -1.47659525e-01 -6.73991382e-01
5.70733137e-02 -6.71394542e-02 -1.50146902e+00 2.00015521e+00
6.70687947e-03 3.60006601e-01 4.69264481e-03 -1.11793637e+00
6.83316171e-01 1.59644380e-01 7.11683154e-01 -6.79524601e-01
-1.82780236e-01 2.62814343e-01 -4.11555469e-01 -7.03004956e-01
5.53540170e-01 4.34961259e-01 3.12725961e-01 1.21821731e-01
-1.16153955e-01 1.12539448e-01 3.22565705e-01 1.98005974e-01
7.02558875e-01 -2.70411130e-02 2.74676740e-01 -1.85189083e-01
8.66204441e-01 -2.41884985e-03 5.58903754e-01 5.57454288e-01
-1.81085646e-01 5.57936370e-01 -1.81744084e-01 -5.69813490e-01
-7.94746518e-01 -9.07307863e-01 -1.63408220e-01 1.02951074e+00
9.11072671e-01 -5.96491814e-01 -6.38851106e-01 -4.81076628e-01
6.16330132e-02 5.35224855e-04 -2.69710690e-01 -1.05135694e-01
-9.16366696e-01 -2.89732307e-01 -5.43357395e-02 5.32172561e-01
4.26244825e-01 -8.17022085e-01 -7.73835361e-01 1.88581198e-01
-2.94161350e-01 -1.17622447e+00 -1.06417441e+00 -5.66025436e-01
-9.50273335e-01 -1.08369720e+00 -9.90486085e-01 -9.85372245e-01
7.89858639e-01 1.02238739e+00 9.45691526e-01 4.37612742e-01
-5.10300636e-01 7.27089643e-01 -5.97982645e-01 1.69142187e-01
1.34692639e-01 -3.23694319e-01 1.47488281e-01 2.51328290e-01
3.17278415e-01 -2.02164903e-01 -1.21545184e+00 6.76619470e-01
-9.89733577e-01 -1.33975431e-01 5.74761569e-01 7.18995988e-01
1.00301003e+00 2.60662064e-02 2.95753896e-01 1.00764394e-01
3.52993272e-02 -4.51297373e-01 -7.26208866e-01 3.97509485e-01
-3.43398482e-01 -1.10736251e-01 5.60237058e-02 -7.69221425e-01
-5.70670128e-01 2.87247807e-01 3.59113067e-01 -1.03272903e+00
2.92113632e-01 3.80001605e-01 1.12349145e-01 -3.99612099e-01
1.96819216e-01 8.16630960e-01 -9.72224846e-02 -3.29323232e-01
2.64693499e-01 5.90086579e-01 5.97279549e-01 -5.23215592e-01
7.53184617e-01 6.84651077e-01 -6.55388907e-02 -8.23258817e-01
-7.05973864e-01 -1.06654024e+00 -3.92200857e-01 -3.45865637e-01
8.00567448e-01 -1.23350000e+00 -8.05705845e-01 1.21844821e-01
-1.17564058e+00 1.11598648e-01 -3.70277345e-01 7.07713723e-01
-5.95230639e-01 7.69295573e-01 -3.69160146e-01 -6.18504345e-01
-3.57511789e-01 -1.37210643e+00 1.48610139e+00 2.08803207e-01
2.35947236e-01 -5.60467422e-01 -2.91295797e-01 3.36919725e-01
1.45222664e-01 -1.76990345e-01 2.19939500e-01 -3.51162404e-01
-1.32368147e+00 -2.63286144e-01 -5.33465564e-01 -5.79242175e-03
-2.13847272e-02 -3.77702445e-01 -4.00737375e-01 -5.36046267e-01
7.79117048e-02 -2.19045833e-01 6.57600939e-01 3.63498807e-01
1.50385439e+00 -6.26703203e-01 -4.77435946e-01 6.05487287e-01
1.31692255e+00 1.13947392e-01 5.22591352e-01 1.57000750e-01
7.62495995e-01 3.78678054e-01 1.18445742e+00 9.45739388e-01
4.34066087e-01 1.25238883e+00 3.82993102e-01 3.06905568e-01
-1.72841236e-01 -1.08976379e-01 2.10658640e-01 9.04376030e-01
-6.17584325e-02 5.89208007e-02 -5.98633468e-01 6.14645302e-01
-2.12729025e+00 -1.10431826e+00 -3.59945372e-02 2.23590040e+00
7.21493423e-01 -4.32552278e-01 1.58014193e-01 6.19141087e-02
8.20416451e-01 1.39060214e-01 -6.07474148e-01 6.02576017e-01
-6.72250688e-02 -2.70691365e-01 2.01126739e-01 6.57546669e-02
-1.13321936e+00 7.90156066e-01 5.58999634e+00 1.39413476e+00
-8.63726437e-01 1.94078907e-01 3.18076879e-01 -5.03635824e-01
1.00909442e-01 -6.27098009e-02 -1.00244892e+00 7.05651700e-01
4.27777737e-01 -3.09124887e-01 4.25105602e-01 1.16915596e+00
1.81679860e-01 -8.43081325e-02 -1.11904466e+00 1.74078393e+00
4.04060930e-01 -1.68775117e+00 1.17877744e-01 -9.09801573e-02
7.54571617e-01 -1.95484981e-01 -2.41940003e-02 7.36056492e-02
-5.34290552e-01 -6.36767030e-01 1.03256571e+00 4.46058124e-01
6.24822617e-01 -5.55761039e-01 2.55911440e-01 1.39170215e-01
-1.84303212e+00 -2.65308022e-01 -6.87801898e-01 3.26061755e-01
3.69943678e-01 4.48134691e-01 -4.18532997e-01 1.88865826e-01
1.05057764e+00 8.81281972e-01 -3.25716317e-01 1.52124989e+00
3.22760761e-01 2.98426718e-01 -4.96970445e-01 5.24872728e-02
3.40086401e-01 -9.26859081e-02 6.70120895e-01 1.06513631e+00
5.42571366e-01 6.54749572e-01 6.22634351e-01 5.68983436e-01
1.60502821e-01 2.75831431e-01 -8.21168125e-02 1.29393905e-01
7.90981293e-01 1.24899614e+00 -7.29158163e-01 -6.27569854e-01
-5.57198584e-01 1.06324267e+00 1.45477459e-01 2.65040606e-01
-1.03220701e+00 -1.25263780e-01 7.60223091e-01 2.43071318e-01
9.12292242e-01 -2.88595378e-01 3.72549444e-01 -1.23693848e+00
3.34249765e-01 -7.74046600e-01 3.89867991e-01 -7.18631506e-01
-1.00224257e+00 3.13695222e-01 -2.12574285e-02 -1.56239593e+00
-1.60668197e-03 -2.15096757e-01 -1.93501726e-01 3.56125951e-01
-1.51323175e+00 -1.00753105e+00 -6.25019073e-01 1.01286089e+00
8.96997750e-01 -5.19311503e-02 2.69196361e-01 4.47223991e-01
-3.63766700e-01 6.05319619e-01 1.44339725e-01 7.83528313e-02
4.16727066e-01 -5.16920805e-01 3.64923067e-02 5.47156036e-01
3.37896466e-01 6.43096149e-01 3.21648091e-01 -6.25433505e-01
-2.01427579e+00 -1.22115481e+00 4.33608860e-01 -2.34564066e-01
5.83518207e-01 -4.70109060e-02 -9.72622573e-01 1.30560786e-01
-4.01969969e-01 6.96030438e-01 5.23162186e-01 -5.73554218e-01
-3.09558243e-01 -1.72570318e-01 -9.31671739e-01 4.07283306e-01
1.24666953e+00 -4.78431433e-01 -4.60911959e-01 6.97425902e-01
8.74475360e-01 -5.32107949e-01 -1.04988921e+00 4.11369503e-01
5.28655708e-01 -7.98728704e-01 1.54177320e+00 -4.23920900e-01
1.21184841e-01 -6.27827942e-01 -5.83923995e-01 -5.75481415e-01
-4.55546081e-01 -5.34970880e-01 -8.86753976e-01 9.57255661e-01
-2.06782490e-01 -1.09432764e-01 8.30701590e-01 6.76074624e-01
-5.19660376e-02 -1.15543485e+00 -1.14631760e+00 -8.93975973e-01
-7.11536646e-01 -3.41898859e-01 5.70156157e-01 6.70941293e-01
-5.07239699e-02 -1.32182881e-01 -3.60010028e-01 4.14115071e-01
8.87458801e-01 5.37510455e-01 7.51561821e-01 -9.38721359e-01
-3.25683981e-01 -3.35582793e-01 -9.28653419e-01 -1.77385986e+00
5.15746698e-02 -7.74516523e-01 2.35992353e-02 -1.21087921e+00
5.22716880e-01 -7.72347510e-01 -2.06983283e-01 1.10081181e-01
-4.37269270e-01 5.21316111e-01 4.17748392e-01 8.16733241e-01
-1.32624865e+00 7.44054437e-01 1.15651429e+00 -8.40757713e-02
-1.77761704e-01 -1.93153098e-01 -1.96819410e-01 5.82533062e-01
1.62216634e-01 -4.14830685e-01 -3.74436706e-01 -5.20017803e-01
-8.09482038e-02 2.46881112e-01 6.12331271e-01 -1.09286118e+00
6.17426455e-01 -2.32254222e-01 3.20482314e-01 -8.82802427e-01
7.00746894e-01 -9.67549205e-01 1.74089089e-01 4.41396266e-01
-2.10335329e-01 -1.51142642e-01 -7.75085539e-02 9.86226261e-01
-4.26812172e-01 -1.59748644e-01 6.96992517e-01 -1.09952301e-01
-1.07308578e+00 8.52450073e-01 1.32684276e-01 2.86554620e-02
1.44190931e+00 -5.73879957e-01 1.08549565e-01 -1.23788163e-01
-5.13038695e-01 4.11347479e-01 3.64540875e-01 5.48932195e-01
1.15069795e+00 -1.63977849e+00 -5.78215897e-01 1.51541710e-01
3.26451004e-01 2.19355717e-01 5.20322561e-01 1.07807267e+00
-2.19867289e-01 5.21817327e-01 2.37389162e-01 -1.11199880e+00
-1.42926753e+00 1.00781584e+00 -5.42221852e-02 2.40771830e-01
-6.72717810e-01 1.15400398e+00 4.90718722e-01 1.69010103e-01
4.39097792e-01 -2.42466219e-02 -1.52238414e-01 2.04367563e-02
1.00440335e+00 6.77934408e-01 -2.85839409e-01 -1.07311761e+00
-4.75681573e-01 1.06693697e+00 -8.35337937e-02 3.46739262e-01
1.16648161e+00 -4.31230664e-01 -7.72429034e-02 4.39049192e-02
1.37044251e+00 -3.63434970e-01 -1.43880129e+00 -6.00553155e-01
-1.99535504e-01 -1.07507706e+00 1.59661546e-01 6.65254379e-03
-1.28759587e+00 5.12272000e-01 7.19448328e-01 -1.26663670e-01
1.14451551e+00 3.79784137e-01 1.00692201e+00 5.78113317e-01
6.75990283e-01 -1.01360285e+00 4.16514426e-01 1.21396378e-01
8.55300128e-01 -1.25974762e+00 2.56289154e-01 -7.00680852e-01
-4.72856104e-01 1.03780997e+00 5.28975189e-01 -1.41391471e-01
7.08301425e-01 -1.66867450e-01 -4.91687953e-01 -2.85641670e-01
-6.22442901e-01 -1.07688315e-01 6.68510199e-01 3.68629575e-01
-5.00024632e-02 -3.14076543e-01 -1.19198181e-01 4.61402267e-01
3.57834250e-01 1.60588935e-01 -1.27067734e-02 8.96962941e-01
-7.26538122e-01 -5.48391223e-01 -6.01875663e-01 5.62820852e-01
-2.03800142e-01 -1.60138741e-01 2.83310324e-01 3.25982034e-01
-5.39642498e-02 8.32369506e-01 1.73211262e-01 -1.66261405e-01
2.72870272e-01 -5.03183544e-01 4.38593090e-01 -3.76630425e-01
-1.52003378e-01 2.32943982e-01 -3.88816953e-01 -1.03592515e+00
-7.80059814e-01 -6.83014035e-01 -1.31793189e+00 -3.70543636e-02
-7.13730395e-01 2.65723795e-01 5.26381910e-01 6.85946822e-01
7.20459998e-01 -1.50696393e-02 7.01284647e-01 -1.35293329e+00
-7.10554898e-01 -5.09661973e-01 -4.32491183e-01 4.05833066e-01
1.86896324e-01 -8.93545449e-01 -1.66831389e-01 3.09256762e-01] | [9.911163330078125, 0.4989762008190155] |
482e37c3-31c7-46e7-9175-776afcabe099 | single-image-deraining-from-model-based-to | 1912.07150 | null | https://arxiv.org/abs/1912.07150v2 | https://arxiv.org/pdf/1912.07150v2.pdf | Single Image Deraining: From Model-Based to Data-Driven and Beyond | The goal of single-image deraining is to restore the rain-free background scenes of an image degraded by rain streaks and rain accumulation. The early single-image deraining methods employ a cost function, where various priors are developed to represent the properties of rain and background layers. Since 2017, single-image deraining methods step into a deep-learning era, and exploit various types of networks, i.e. convolutional neural networks, recurrent neural networks, generative adversarial networks, etc., demonstrating impressive performance. Given the current rapid development, in this paper, we provide a comprehensive survey of deraining methods over the last decade. We summarize the rain appearance models, and discuss two categories of deraining approaches: model-based and data-driven approaches. For the former, we organize the literature based on their basic models and priors. For the latter, we discuss developed ideas related to architectures, constraints, loss functions, and training datasets. We present milestones of single-image deraining methods, review a broad selection of previous works in different categories, and provide insights on the historical development route from the model-based to data-driven methods. We also summarize performance comparisons quantitatively and qualitatively. Beyond discussing the technicality of deraining methods, we also discuss the future directions. | ['Yuming Fang', 'Shiqi Wang', 'Jiaying Liu', 'Robby T. Tan', 'Wenhan Yang'] | 2019-12-16 | null | null | null | null | ['single-image-deraining'] | ['computer-vision'] | [ 2.94600248e-01 -1.73931509e-01 3.10530931e-01 -4.46231812e-01
-6.98354006e-01 -2.21673682e-01 2.71786571e-01 -7.24537015e-01
2.92523503e-02 8.45625043e-01 -1.72546115e-02 -2.07382083e-01
2.31300041e-01 -6.64924800e-01 -8.34232450e-01 -1.14977121e+00
-8.81451964e-02 -1.00321166e-01 1.32325245e-02 -2.68106699e-01
-2.45526731e-01 7.52152801e-01 -1.68703926e+00 1.24712966e-01
1.29015410e+00 8.27398241e-01 1.79436579e-01 1.10973346e+00
-2.21799444e-02 1.20852733e+00 -8.92231047e-01 -2.76689857e-01
1.93920895e-01 -6.75134122e-01 -2.30145633e-01 3.33531618e-01
9.28272188e-01 -7.34791875e-01 -8.96164477e-01 9.13244009e-01
6.01948023e-01 -7.97655508e-02 6.05261922e-01 -1.11132371e+00
-9.05553460e-01 7.43684694e-02 -5.61789215e-01 4.13911939e-01
-1.55063897e-01 8.52345452e-02 6.52467549e-01 -1.27264595e+00
4.14743930e-01 1.20962775e+00 8.05060387e-01 6.37680113e-01
-9.96354759e-01 -4.45273668e-01 4.70556587e-01 2.32204169e-01
-1.13338947e+00 -3.32343429e-01 6.15334570e-01 -3.25350732e-01
6.55137599e-01 4.41722900e-01 7.84377754e-01 1.11680937e+00
2.03599796e-01 1.10528159e+00 1.35133791e+00 -4.75822181e-01
4.89814803e-02 -2.18548596e-01 2.59145737e-01 6.03626251e-01
5.39527953e-01 3.64824444e-01 -9.00838822e-02 4.68804091e-02
7.67122388e-01 1.79921702e-01 -6.11968875e-01 -2.20240667e-01
-5.00964582e-01 8.17479491e-01 6.82007194e-01 -1.95308570e-02
-3.55422735e-01 7.51399063e-03 -8.91854987e-02 4.43796754e-01
9.18381453e-01 1.65609151e-01 -2.42990777e-01 6.02530301e-01
-1.40153122e+00 5.38953483e-01 6.51143551e-01 7.95648336e-01
8.65490735e-01 8.35260332e-01 -3.21464717e-01 8.74757528e-01
2.67592341e-01 1.24729717e+00 -1.19627565e-01 -8.59126687e-01
2.94236273e-01 -1.18942946e-01 3.38001668e-01 -7.75038004e-01
-3.80100399e-01 -4.20055777e-01 -1.21261358e+00 6.64108336e-01
3.47937606e-02 -3.71395141e-01 -1.67051387e+00 1.38294065e+00
-6.99722022e-02 2.49716088e-01 2.03276530e-01 9.66052353e-01
1.13229132e+00 8.25168729e-01 -1.41476423e-01 -3.26555103e-01
1.03555763e+00 -1.20960402e+00 -8.23263407e-01 -5.35850585e-01
-1.25194401e-01 -7.75310159e-01 8.16076398e-01 2.79167712e-01
-1.07131326e+00 -6.33847654e-01 -1.10268974e+00 3.52873169e-02
-2.68741459e-01 2.73387343e-01 4.26773071e-01 5.99316001e-01
-1.13464403e+00 6.63467646e-01 -7.50960350e-01 -4.92164671e-01
5.24327159e-01 -7.61067197e-02 1.32695869e-01 -5.33222973e-01
-1.17873144e+00 1.06295991e+00 -2.57466018e-01 6.94764256e-01
-1.39993489e+00 -6.02811694e-01 -7.78180599e-01 -2.42965326e-01
-1.70710728e-01 -9.87481177e-01 1.16570210e+00 -1.23448920e+00
-1.36533844e+00 9.92971063e-01 -2.86156207e-01 -6.08263135e-01
2.83331841e-01 -8.70876610e-01 -6.42149746e-01 2.48195499e-01
-4.41804260e-01 2.95533329e-01 1.41337228e+00 -1.87099278e+00
-6.46010101e-01 -1.28558472e-01 9.87463593e-02 1.27223924e-01
3.19827080e-01 -6.15480430e-02 -4.86943185e-01 -9.33853149e-01
-1.57249436e-01 -8.82144868e-01 -4.11612004e-01 1.49507657e-01
-4.40043181e-01 6.05131328e-01 1.12555790e+00 -9.41771805e-01
1.13800979e+00 -2.12209821e+00 9.13317502e-02 -2.58974612e-01
3.65121931e-01 6.85786903e-01 -3.04400414e-01 3.67402166e-01
-1.36363909e-01 -2.45150357e-01 -8.46655965e-01 -6.16819561e-01
-2.92837083e-01 6.75877035e-01 -7.45630383e-01 6.19479537e-01
3.97176415e-01 8.59454393e-01 -5.77664137e-01 -1.07612453e-01
3.27314645e-01 8.12089920e-01 -1.32849246e-01 8.72310996e-01
-2.07020000e-01 3.88758123e-01 -1.85360745e-01 1.08758521e+00
1.28532159e+00 5.53760789e-02 -2.51001418e-01 -3.68878931e-01
-3.82739827e-02 -1.65413618e-01 -6.47690475e-01 9.23242450e-01
-5.98809838e-01 7.85007060e-01 5.74849367e-01 -7.66231239e-01
9.22179103e-01 1.36567950e-01 3.19408886e-02 -5.83614409e-01
-2.94698745e-01 2.88283795e-01 -3.00989628e-01 -6.90290272e-01
4.82129693e-01 -2.67256826e-01 5.69805562e-01 -9.40721110e-02
-1.50086060e-01 -3.33343208e-01 -9.70878080e-02 1.51958898e-01
9.48976815e-01 3.81029725e-01 5.71614169e-02 9.76563767e-02
1.99804336e-01 -4.05029505e-02 5.90227902e-01 1.06644249e+00
-1.97682083e-01 1.30719364e+00 2.76349727e-02 -6.51591480e-01
-1.11457241e+00 -1.25946140e+00 -2.26166904e-01 8.17751467e-01
1.89860344e-01 2.07622319e-01 -6.24234676e-01 -5.02679169e-01
9.69574600e-02 5.69568276e-01 -9.30902064e-01 5.42834587e-02
-6.75143063e-01 -1.56967652e+00 4.97962326e-01 5.46616077e-01
7.58922160e-01 -1.10208416e+00 -3.42238665e-01 3.77069712e-02
-1.76262170e-01 -1.11546326e+00 8.72135609e-02 2.36119598e-01
-1.11072350e+00 -9.98618782e-01 -1.13804591e+00 -6.41933501e-01
6.80742800e-01 6.96245193e-01 1.65664482e+00 2.21483156e-01
-5.45079947e-01 4.57864434e-01 -6.10353529e-01 -4.58179504e-01
3.46047133e-02 -5.39127350e-01 -2.78314263e-01 -7.02315941e-02
1.99707458e-03 -8.57230544e-01 -9.57787216e-01 -1.13212369e-01
-1.16449726e+00 -2.55490661e-01 8.64401102e-01 8.66755962e-01
5.13404667e-01 -3.90006274e-01 8.23755637e-02 -9.72839653e-01
1.52055815e-01 -4.32490796e-01 -5.17771006e-01 1.89286113e-01
-4.36849177e-01 -2.55958855e-01 3.67933542e-01 6.12433031e-02
-1.32899320e+00 -6.45916387e-02 -2.16700032e-01 -7.05392241e-01
-1.27917565e-02 2.77142227e-01 -2.24710107e-01 -2.85315961e-01
6.71284795e-01 3.96519721e-01 -1.80024594e-01 -6.46409214e-01
6.70497775e-01 3.03611517e-01 7.35012650e-01 -3.47217113e-01
1.22696173e+00 8.68139148e-01 -3.16606253e-01 -9.75669682e-01
-1.35238302e+00 -2.47504085e-01 -3.90250295e-01 -2.93930233e-01
7.73960769e-01 -1.21162963e+00 1.34388134e-01 9.25027251e-01
-1.23935187e+00 -7.27582037e-01 -3.98706585e-01 2.75259614e-01
-3.94547731e-01 3.70524526e-01 -8.60009491e-01 -9.80558097e-01
-6.29704714e-01 -8.81508768e-01 1.01629758e+00 5.57244360e-01
6.35692775e-01 -9.12079871e-01 2.88989305e-01 2.00699061e-01
8.46935809e-01 5.21272719e-01 5.83719611e-01 1.95381805e-01
-8.52999091e-01 -5.75477146e-02 -4.64168161e-01 7.87661791e-01
1.42939314e-01 3.84361506e-01 -1.34545326e+00 -5.44645488e-01
1.26846105e-01 -1.81349948e-01 1.41767943e+00 8.86947036e-01
9.50593412e-01 -2.23456249e-01 -2.45300874e-01 1.08811390e+00
1.70894146e+00 3.91607061e-02 1.23099172e+00 4.44686621e-01
7.08324134e-01 4.09137756e-01 5.84623158e-01 3.41716051e-01
2.32292488e-01 3.13148797e-01 9.46572244e-01 -8.38184476e-01
-4.73584980e-01 2.11954072e-01 5.82453370e-01 7.02984393e-01
-5.12170672e-01 -6.19738817e-01 -3.93159449e-01 6.13973737e-01
-1.63454604e+00 -1.05824029e+00 -1.43543676e-01 1.96499801e+00
4.86212492e-01 -2.80945271e-01 -2.32730687e-01 -3.63482594e-01
7.52752602e-01 6.89591467e-01 -6.18666232e-01 -1.40720606e-01
-7.81306684e-01 4.12690580e-01 5.98270178e-01 7.31528699e-01
-1.37940872e+00 1.00076818e+00 7.58482885e+00 4.21374857e-01
-9.60443795e-01 6.95571750e-02 6.20379925e-01 -1.43026307e-01
-1.11599691e-01 -3.20905149e-02 -8.06205451e-01 3.26784551e-01
6.69940948e-01 2.77607739e-01 3.34783077e-01 7.62815058e-01
3.61531764e-01 -1.76659971e-01 -4.49173599e-01 6.62903070e-01
3.28367531e-01 -1.13652289e+00 1.29635766e-01 -3.29239756e-01
9.35032010e-01 6.57569289e-01 1.18351787e-01 4.34627235e-01
4.75051999e-01 -1.20735848e+00 4.25310016e-01 9.76536751e-01
7.31661201e-01 -4.23355520e-01 8.90796483e-01 -1.96262807e-01
-1.03903854e+00 1.26939625e-01 -6.17576599e-01 -2.51973365e-02
2.19693631e-01 1.24632204e+00 1.24065310e-01 9.01140749e-01
1.14945924e+00 8.16958725e-01 -3.37052315e-01 1.08831882e+00
-5.83799183e-01 8.54188502e-01 -2.24503368e-01 7.36622632e-01
1.26962572e-01 -4.91918325e-01 5.86992919e-01 1.59309280e+00
3.03292483e-01 2.66101211e-01 6.56401925e-03 7.01392293e-01
3.77015211e-02 -5.38167477e-01 -6.20436132e-01 3.06558192e-01
1.98698372e-01 1.23890162e+00 -1.16094865e-01 -3.50393385e-01
-6.98408306e-01 1.11499739e+00 -1.22277904e-02 9.50995147e-01
-1.04119527e+00 -5.80861330e-01 1.03824759e+00 4.50334363e-02
4.99363154e-01 -3.42056155e-01 -6.33456856e-02 -1.31501436e+00
7.67521467e-03 -1.06339157e+00 3.78093049e-02 -1.09827638e+00
-1.57900047e+00 9.46169496e-01 -7.35548511e-02 -1.23309577e+00
3.09541881e-01 -6.43700898e-01 -9.54396427e-01 8.97897840e-01
-2.23059964e+00 -1.17081356e+00 -9.12424624e-01 3.54425907e-01
5.32923460e-01 -1.65429916e-02 6.29143775e-01 3.79895419e-01
-7.80929685e-01 2.07229376e-01 6.69260025e-01 1.25052899e-01
8.96491826e-01 -1.20215261e+00 5.71328223e-01 1.30451238e+00
-4.59486851e-03 3.13902140e-01 1.02056599e+00 -4.49632883e-01
-1.09880316e+00 -1.45356607e+00 5.38890123e-01 -7.63782784e-02
3.60114753e-01 -1.97726011e-01 -1.17061853e+00 8.68119001e-01
3.08516055e-01 4.85565722e-01 2.81264216e-01 -5.11353575e-02
-4.78416353e-01 -3.87661994e-01 -1.04048514e+00 5.59066474e-01
9.32288349e-01 -2.88882464e-01 -4.23444599e-01 4.62297797e-01
5.92089355e-01 -6.68762803e-01 -4.19976652e-01 6.46870434e-01
3.67321372e-01 -1.47559917e+00 1.05885410e+00 -4.12069649e-01
5.77294469e-01 -4.47792560e-01 -3.54517311e-01 -1.43630302e+00
-4.52682078e-01 -6.03736222e-01 -5.06163299e-01 8.83717716e-01
9.73436534e-02 -5.23250997e-01 6.75206423e-01 1.09513476e-01
-3.88587058e-01 -7.09287643e-01 -6.08462036e-01 -7.35466063e-01
1.32182047e-01 -8.45580772e-02 2.42882490e-01 5.94611943e-01
-1.02699018e+00 1.05518766e-01 -9.93775785e-01 6.30209446e-01
9.41931903e-01 5.79437137e-01 7.66824841e-01 -9.25139308e-01
-1.53968468e-01 -1.03141479e-01 -4.42578681e-02 -1.37159956e+00
-1.99119210e-01 -3.43442559e-01 3.85930330e-01 -1.84722757e+00
2.46073127e-01 -9.46636349e-02 -2.24984676e-01 4.99387890e-01
-4.62240219e-01 4.63781923e-01 2.55111337e-01 3.98717165e-01
-3.23920161e-01 9.23322380e-01 1.27594829e+00 -3.00393611e-01
-4.29627672e-02 1.46627709e-01 -5.14705956e-01 8.86164248e-01
9.61995482e-01 -4.33588088e-01 -6.38516620e-02 -8.65576446e-01
-9.35480967e-02 -3.13340634e-01 6.36506379e-01 -9.76941884e-01
-2.20417455e-01 -2.78132632e-02 5.40045857e-01 -6.16408467e-01
5.60029864e-01 -6.08322084e-01 -1.08985581e-01 3.48508030e-01
2.66383618e-01 -9.31822360e-02 3.50081921e-01 7.01290190e-01
-5.54007173e-01 7.28105158e-02 1.39617777e+00 -1.13981098e-01
-6.34988129e-01 4.75080431e-01 -6.64238513e-01 -1.74028911e-02
6.19068623e-01 -1.36457399e-01 -5.39929092e-01 -6.57981098e-01
-1.08411300e+00 1.24007307e-01 2.78177381e-01 1.91981927e-01
6.76667333e-01 -8.60286057e-01 -1.07652533e+00 1.37532959e-02
-2.30411619e-01 -1.00344896e-01 5.98903537e-01 8.42365265e-01
-8.24856520e-01 5.25935814e-02 -2.28683919e-01 -4.05909657e-01
-1.23537648e+00 2.75436342e-01 6.91363692e-01 -3.16251695e-01
-9.68024433e-01 8.42928708e-01 6.97556973e-01 -1.38846457e-01
2.45610297e-01 -2.05962926e-01 -1.49468081e-02 -3.11787665e-01
6.04012311e-01 1.96477801e-01 1.69867650e-01 -2.53815472e-01
-2.73272514e-01 6.38385117e-01 -1.40347511e-01 2.64728397e-01
1.55840683e+00 -2.43638024e-01 -1.71868682e-01 2.67700195e-01
6.74668729e-01 -8.24626982e-02 -1.63702643e+00 -2.66478091e-01
-5.35300493e-01 -4.16478217e-01 1.29612952e-01 -9.06980336e-01
-1.72580612e+00 1.04953170e+00 7.96726346e-01 -1.27709312e-02
1.65024316e+00 -2.29828015e-01 6.33618772e-01 3.52496862e-01
2.35208478e-02 -5.64233720e-01 -1.67288759e-03 8.48446429e-01
1.06843817e+00 -1.11682391e+00 3.45939577e-01 -3.57136607e-01
-6.19992614e-01 1.03848124e+00 6.36747301e-01 -4.92809504e-01
9.48719621e-01 5.39047778e-01 5.12444794e-01 -2.10569426e-01
-4.50779378e-01 -5.46282709e-01 -7.75478333e-02 9.53915775e-01
4.17655587e-01 -1.89768896e-01 4.15662751e-02 3.14660579e-01
2.42782831e-01 -1.19792588e-01 5.09687126e-01 9.67316628e-01
-8.66097510e-01 -9.66046393e-01 -7.21910954e-01 3.21713984e-01
-2.81536818e-01 -4.70461011e-01 -1.85161740e-01 8.93108428e-01
1.62650540e-01 9.36175525e-01 -4.02909778e-02 -3.90336141e-02
4.80164319e-01 -2.63307869e-01 6.67054236e-01 -3.16315532e-01
-2.76540637e-01 -7.90883303e-02 -7.10558593e-02 -4.34866160e-01
-8.63361061e-01 -3.81102413e-01 -5.47889769e-01 -3.28330100e-01
-2.04429522e-01 -1.04052515e-03 2.78663576e-01 8.38101923e-01
3.27259898e-01 6.02714479e-01 7.22248077e-01 -1.36383021e+00
-1.39968857e-01 -9.93719459e-01 -9.67910349e-01 -3.03953737e-02
8.54709089e-01 -4.67518359e-01 -6.82219207e-01 3.27693611e-01] | [10.920541763305664, -3.2454254627227783] |
5ffc7c12-ce8e-4f4b-b9d3-cd325a2a9727 | two-stage-pipeline-for-multilingual-dialect | 2303.03487 | null | https://arxiv.org/abs/2303.03487v2 | https://arxiv.org/pdf/2303.03487v2.pdf | Two-stage Pipeline for Multilingual Dialect Detection | Dialect Identification is a crucial task for localizing various Large Language Models. This paper outlines our approach to the VarDial 2023 shared task. Here we have to identify three or two dialects from three languages each which results in a 9-way classification for Track-1 and 6-way classification for Track-2 respectively. Our proposed approach consists of a two-stage system and outperforms other participants' systems and previous works in this domain. We achieve a score of 58.54% for Track-1 and 85.61% for Track-2. Our codebase is available publicly (https://github.com/ankit-vaidya19/EACL_VarDial2023). | ['Aditya Kane', 'Ankit Vaidya'] | 2023-03-06 | null | null | null | null | ['dialect-identification'] | ['natural-language-processing'] | [-7.02063918e-01 -4.36538905e-01 -2.43488982e-01 -6.09178364e-01
-1.52863824e+00 -1.06619072e+00 8.75600219e-01 4.77786511e-02
-5.18897593e-01 8.74455750e-01 4.20026422e-01 -6.25597179e-01
2.53821760e-01 -4.98327732e-01 -3.36949170e-01 -4.94466573e-01
1.65463597e-01 8.82062972e-01 2.11113900e-01 -3.54105085e-01
3.58811289e-01 2.33831927e-01 -9.72526431e-01 4.85484779e-01
8.20723534e-01 2.97431141e-01 -5.61705977e-02 1.00874436e+00
-1.60745159e-01 3.48964423e-01 -4.06524599e-01 -4.44076687e-01
7.34952688e-02 -4.50298160e-01 -1.19003320e+00 -6.06455207e-01
9.77689385e-01 8.07049125e-03 -5.33873849e-02 1.08552659e+00
8.79429162e-01 -1.05124749e-01 7.22777724e-01 -1.03238714e+00
-6.20776355e-01 1.19863248e+00 -6.73793375e-01 4.79911655e-01
6.07908010e-01 -2.32104301e-01 9.71891880e-01 -1.09552503e+00
5.86765766e-01 1.40744555e+00 8.54448915e-01 6.20261669e-01
-1.14029336e+00 -1.09366798e+00 -5.68348961e-03 1.23329677e-01
-1.94043159e+00 -9.33451295e-01 5.19101918e-01 -7.58373439e-01
9.68298852e-01 4.46824968e-01 1.41448781e-01 1.07758868e+00
-2.47143105e-01 7.83514500e-01 1.58147502e+00 -3.92114639e-01
-2.83355236e-01 5.99795043e-01 8.56443882e-01 4.14164275e-01
9.34823975e-03 -2.36781821e-01 -6.41211331e-01 -2.43027508e-01
4.16832596e-01 -7.84213424e-01 8.92436132e-02 3.20203960e-01
-1.02724576e+00 8.12390804e-01 5.88017032e-02 6.19673073e-01
-1.22589678e-01 -4.35046941e-01 5.61708212e-01 6.31704509e-01
4.80553597e-01 -2.48318627e-01 -4.01317358e-01 -2.62365103e-01
-1.06924021e+00 5.96835494e-01 6.20425880e-01 8.78402531e-01
4.60799336e-01 -7.62568265e-02 2.64126388e-03 1.33607137e+00
5.87272763e-01 6.82272315e-01 6.10141754e-01 -2.75003344e-01
7.67748535e-01 3.76098841e-01 4.57006916e-02 -4.40844983e-01
-3.88956338e-01 -1.98237017e-01 -7.47018874e-01 -9.92202535e-02
6.61421299e-01 -3.81297410e-01 -6.89843893e-01 2.03380346e+00
3.63427073e-01 -6.06208714e-03 -7.44059756e-02 5.62338889e-01
8.32819760e-01 6.45420372e-01 2.18833629e-02 1.13126442e-01
1.57312548e+00 -8.81702542e-01 -6.15337431e-01 -8.59407708e-02
7.59724259e-01 -1.32284236e+00 9.86896932e-01 3.45983744e-01
-1.20196986e+00 -6.55722558e-01 -7.64675736e-01 -3.12955268e-02
-6.89158142e-01 3.23809862e-01 2.61467874e-01 1.19642115e+00
-1.47838247e+00 -3.37513685e-01 -2.18579784e-01 -9.10268426e-01
8.49425346e-02 2.72440940e-01 -5.23969471e-01 6.23444095e-02
-1.27502251e+00 7.58184493e-01 9.59668607e-02 7.76661634e-02
-6.93562746e-01 -5.48178554e-01 -4.78343397e-01 -5.11263430e-01
-5.12098074e-01 -1.60191476e-01 1.16744471e+00 -4.18404013e-01
-1.33773553e+00 1.67488396e+00 -5.38441300e-01 -2.12658584e-01
7.78239250e-01 -3.28329206e-01 -8.76183569e-01 -4.64716941e-01
4.85253572e-01 3.54540706e-01 3.58378366e-02 -8.97101343e-01
-7.88119972e-01 -5.08623183e-01 -3.33179832e-01 2.73873329e-01
-6.80169165e-02 9.14642930e-01 -5.35312474e-01 -6.62574351e-01
-1.52947858e-03 -1.05045688e+00 1.53741613e-01 -6.65804207e-01
-7.17914045e-01 -3.98162723e-01 3.56287211e-01 -1.16399264e+00
1.59265769e+00 -1.76484084e+00 -2.01027095e-01 2.92744547e-01
-7.27822930e-02 4.10161883e-01 8.73641074e-02 6.74890459e-01
-2.15424612e-01 1.54655248e-01 -6.35826513e-02 -4.45068568e-01
4.65776771e-02 -5.41703820e-01 -3.20060849e-02 6.91137493e-01
-3.48612010e-01 5.02370358e-01 -3.98645401e-01 -3.86018366e-01
7.61595517e-02 5.03754199e-01 -4.79922622e-01 1.44908577e-01
3.57132643e-01 4.99443978e-01 -7.55181611e-02 6.06585383e-01
1.27788627e+00 3.44233990e-01 3.17868739e-01 2.92370409e-01
-6.99004710e-01 6.84343398e-01 -1.50412023e+00 1.50367677e+00
-3.42006385e-01 6.46658659e-01 2.69478887e-01 -6.75439954e-01
1.01064014e+00 4.77459252e-01 -8.65579844e-02 -4.46931630e-01
8.98022875e-02 4.68254596e-01 -6.33237930e-03 -2.95053929e-01
3.74153793e-01 -2.79462598e-02 -5.42860389e-01 5.68940878e-01
2.14403737e-02 6.19180024e-01 1.11708514e-01 2.97945052e-01
6.94377601e-01 -1.18255101e-01 3.06518406e-01 -7.82451212e-01
1.03208268e+00 -1.75521225e-01 4.56192851e-01 9.94761109e-01
-6.45747185e-01 6.02132857e-01 1.40708894e-01 -9.71922874e-02
-7.66945660e-01 -1.37782407e+00 -3.78781140e-01 1.34254432e+00
-5.79152107e-01 -3.23417574e-01 -9.69295084e-01 -4.07477796e-01
-2.04340056e-01 5.46500862e-01 -4.24894005e-01 3.65255088e-01
-8.60739708e-01 -7.24806786e-01 1.30772769e+00 1.74388364e-01
5.45954645e-01 -8.52155864e-01 5.38937569e-01 5.19171022e-02
-5.59632301e-01 -8.81619036e-01 -8.57511759e-01 -1.76243752e-01
-2.99117178e-01 -8.91399801e-01 -9.74654317e-01 -1.40504062e+00
2.96196342e-01 2.00197957e-02 9.60951924e-01 -2.55397558e-01
2.98718270e-03 -1.17281787e-01 -2.33961511e-02 -1.60718217e-01
-2.87127286e-01 4.93097931e-01 3.48714948e-01 9.34575945e-02
8.72245073e-01 -3.70297551e-01 -2.38395840e-01 2.08702698e-01
-1.02989122e-01 -1.68975919e-01 5.32104373e-02 3.04576159e-01
7.72266136e-03 -4.25424129e-01 6.95031941e-01 -1.24439847e+00
8.70179176e-01 -7.24796593e-01 -6.93308830e-01 4.36793774e-01
-1.96981654e-01 -3.34659070e-01 3.50043774e-01 -2.78759360e-01
-1.07291806e+00 1.32802753e-02 -6.54828966e-01 3.98178101e-01
-5.49541831e-01 2.65473038e-01 -3.25168312e-01 1.71012446e-01
5.32627344e-01 2.98825979e-01 -3.82483780e-01 -9.35517609e-01
3.21794778e-01 1.40476060e+00 4.87833440e-01 -6.49271786e-01
6.96888387e-01 3.98203991e-02 -1.05178356e+00 -1.04902112e+00
-4.64517087e-01 -9.16549087e-01 -9.92689729e-01 -2.45799527e-01
8.95764410e-01 -1.41822553e+00 -7.00754762e-01 1.24166560e+00
-1.02744007e+00 -1.60400674e-01 6.45504832e-01 4.68269765e-01
-4.13413048e-02 3.48782800e-02 -8.88015747e-01 -8.62489402e-01
-5.94213247e-01 -9.35439646e-01 6.82573080e-01 2.03565344e-01
-6.77990377e-01 -1.18272007e+00 7.59816051e-01 5.40835619e-01
4.83171165e-01 -2.47825205e-01 5.03848851e-01 -9.88800645e-01
-6.44006506e-02 -4.36662808e-02 -1.58976525e-01 -6.02265038e-02
1.32895932e-01 -9.75677744e-02 -1.27489531e+00 -3.46178114e-01
-4.45701808e-01 -4.48669851e-01 5.98814666e-01 3.81435335e-01
3.86215538e-01 -1.77346781e-01 -1.30837098e-01 4.57904041e-01
1.42321360e+00 1.18590303e-01 3.39560330e-01 3.14313591e-01
7.39513516e-01 6.33144975e-01 2.08504260e-01 2.30906054e-01
1.02392280e+00 8.91359806e-01 -6.19190574e-01 2.47489419e-02
-2.38572225e-01 -3.22729588e-01 7.10494816e-01 1.34539294e+00
3.37561786e-01 -1.29992694e-01 -1.60547352e+00 1.03789115e+00
-1.65838969e+00 -1.08638656e+00 -6.31516755e-01 2.28212214e+00
1.07090151e+00 -8.47619176e-02 8.12612057e-01 -9.86086950e-03
1.10432506e+00 1.92190241e-02 -7.56463557e-02 -7.16197670e-01
-4.89458174e-01 2.26924747e-01 2.79418737e-01 1.30291438e+00
-1.05613458e+00 1.48094440e+00 6.33726311e+00 7.46875167e-01
-1.03269315e+00 3.11933041e-01 4.64757293e-01 1.26004741e-02
-4.16919924e-02 2.87370495e-02 -1.54971123e+00 6.02158606e-01
1.31269121e+00 -3.63549411e-01 2.73180366e-01 1.81730345e-01
2.25511521e-01 5.69367670e-02 -7.41772234e-01 1.03693533e+00
1.58869609e-01 -1.07284546e+00 1.12710826e-01 -1.41666261e-02
4.32849318e-01 5.46625912e-01 1.59558002e-02 3.97309482e-01
7.36714244e-01 -1.00832415e+00 7.70582438e-01 2.35656276e-01
7.86291122e-01 -7.63943315e-01 4.19179022e-01 3.56648475e-01
-1.50666618e+00 3.91509384e-01 -1.10548533e-01 1.37886897e-01
3.04928333e-01 1.74370497e-01 -7.76250422e-01 4.36517000e-01
8.45928848e-01 3.37516904e-01 -6.78879619e-01 1.03078878e+00
-2.43156427e-03 1.01866889e+00 -6.20654464e-01 7.78420419e-02
2.77454376e-01 -2.14816406e-01 4.35021430e-01 1.71858895e+00
1.66944653e-01 -3.20701689e-01 1.10310011e-01 2.62156188e-01
-7.58433864e-02 5.56367815e-01 -6.41676009e-01 2.31761053e-01
8.09629917e-01 9.83563185e-01 -4.06350583e-01 -4.43530113e-01
-3.35258424e-01 9.73477423e-01 6.57117784e-01 2.65707165e-01
-7.21042633e-01 -5.34107029e-01 8.13650310e-01 3.87245119e-02
9.34850350e-02 -9.73853022e-02 -1.82213068e-01 -1.19722843e+00
-1.20687336e-02 -1.12493837e+00 9.19342637e-01 -3.57601196e-01
-1.31391907e+00 7.24943221e-01 -1.57033339e-01 -7.13437200e-01
-2.27016762e-01 -5.20178497e-01 -7.93576598e-01 1.40049803e+00
-1.10216188e+00 -1.69598198e+00 1.68768957e-01 9.06034410e-01
5.30019522e-01 -5.25022686e-01 1.02371216e+00 9.92095232e-01
-7.55447090e-01 1.27351522e+00 3.28465581e-01 5.50726116e-01
1.24542892e+00 -1.15533209e+00 6.32119060e-01 9.34250712e-01
9.94122494e-03 9.16125953e-01 5.78694701e-01 -5.79236805e-01
-8.56099486e-01 -9.69371915e-01 1.84896231e+00 -7.30850995e-01
8.12828124e-01 -1.01528645e+00 -6.23347580e-01 1.06796527e+00
6.30381823e-01 -4.42482919e-01 9.14916813e-01 6.76696360e-01
-5.90177715e-01 -6.09077699e-03 -1.22809827e+00 6.18686497e-01
9.38445210e-01 -1.02171493e+00 -3.33464950e-01 3.10799509e-01
1.10857800e-01 -8.27394426e-02 -8.71871769e-01 -3.27427417e-01
6.62343919e-01 -8.97670031e-01 8.51847708e-01 -5.81670821e-01
-2.52179623e-01 -7.00633153e-02 -4.27132726e-01 -1.11524963e+00
-3.02108526e-01 -7.56165981e-01 5.69931865e-01 1.96844935e+00
7.22937822e-01 -8.98331940e-01 4.14580941e-01 1.03944935e-01
3.18825692e-01 1.50683090e-01 -1.01639104e+00 -6.35238469e-01
8.30271065e-01 -4.61770386e-01 3.17813784e-01 1.43965423e+00
-1.24294758e-01 6.81612611e-01 -5.03237009e-01 3.67038995e-01
7.56062984e-01 -5.02675436e-02 8.31076145e-01 -9.65275109e-01
1.70337819e-02 -5.63074052e-01 -9.16889682e-02 -8.36430132e-01
1.30528539e-01 -1.29192090e+00 -3.14757317e-01 -1.14661372e+00
2.73301601e-01 -7.85508692e-01 -3.39986473e-01 4.15255755e-01
-6.99399486e-02 5.18591702e-01 7.57845566e-02 6.40010908e-02
-4.52653021e-01 -1.77961141e-02 2.35404298e-01 7.35733286e-02
-2.66988277e-01 1.86298758e-01 -8.96992683e-01 4.77243841e-01
1.59185505e+00 -5.96906483e-01 -1.22322507e-01 -5.36698163e-01
2.33329274e-02 -2.87203550e-01 -1.79398015e-01 -8.78967881e-01
3.20440024e-01 1.41460851e-01 2.38494635e-01 -9.61984158e-01
2.21766546e-01 -3.34501714e-02 1.74935013e-01 4.61569905e-01
-2.67295688e-01 6.00198865e-01 3.81055921e-01 -2.09314585e-01
-1.77066058e-01 -5.15904464e-03 9.04167056e-01 -1.18584320e-01
-5.99528491e-01 5.96334040e-02 -8.40408623e-01 4.40112144e-01
7.66610146e-01 1.10389955e-01 -3.46336305e-01 -1.63127512e-01
-7.30265260e-01 3.46426308e-01 2.48758703e-01 6.67970598e-01
2.03392833e-01 -1.54193687e+00 -1.36866033e+00 4.37967390e-01
2.01447725e-01 -8.22534919e-01 3.79306227e-01 7.46712446e-01
-6.69268131e-01 6.94141924e-01 -3.81097883e-01 -4.17182267e-01
-1.86222267e+00 -1.51909336e-01 3.51510316e-01 -3.49281728e-01
-1.34199187e-01 1.19840467e+00 4.43208665e-02 -1.34944856e+00
1.82899401e-01 4.89502460e-01 -5.35713077e-01 3.54586691e-01
7.45338142e-01 2.98257411e-01 -5.42404540e-02 -1.27017212e+00
-9.03843224e-01 7.82834232e-01 -6.62024856e-01 -6.51319742e-01
1.00335455e+00 -3.35828871e-01 -2.77764976e-01 8.33088875e-01
1.32612014e+00 6.71454430e-01 -1.00008175e-01 -3.45117569e-01
2.67632872e-01 -5.10495864e-02 -1.93297714e-01 -9.72463131e-01
-6.14674747e-01 7.36927629e-01 9.42869544e-01 -4.99240048e-02
6.28658175e-01 2.09183339e-02 5.69878042e-01 -1.16259702e-01
1.88151076e-01 -1.14875138e+00 -7.86991477e-01 9.42717791e-01
6.57190740e-01 -1.32052135e+00 -3.64771605e-01 -2.11586297e-01
-4.67245042e-01 4.51020271e-01 7.11722314e-01 6.25610948e-02
9.86580491e-01 3.18328083e-01 7.66882241e-01 -1.41807860e-02
-6.69583142e-01 -1.06972769e-01 1.94157675e-01 5.74565649e-01
1.05661654e+00 3.64582151e-01 -4.73985732e-01 5.60509324e-01
-7.03335226e-01 -3.53887230e-01 1.76009506e-01 7.37058461e-01
-2.63664871e-01 -1.37164724e+00 -4.86370713e-01 1.64802615e-02
-7.07418799e-01 -3.84878427e-01 -8.51320326e-01 8.60632896e-01
-2.40622163e-02 9.49134648e-01 3.07392236e-02 -4.13339853e-01
9.46851969e-02 4.64808077e-01 3.28915209e-01 -4.57960784e-01
-1.18441963e+00 1.79256409e-01 6.22670710e-01 -1.67732984e-01
-1.19000757e-02 -9.17783558e-01 -1.00350606e+00 -9.20503557e-01
6.46045133e-02 3.06236893e-01 5.77674031e-01 4.40260589e-01
1.32547736e-01 -1.40838906e-01 5.37914991e-01 -1.37428805e-01
-2.07776815e-01 -1.15362740e+00 -5.08634388e-01 2.12050721e-01
3.81322771e-01 -2.00650722e-01 -1.46979034e-01 -3.97959277e-02] | [10.152641296386719, 10.739961624145508] |
a30aa448-2fb6-4311-a410-b73af93706f1 | robust-discovery-of-positive-and-negative | null | null | http://www.eurecom.fr/fr/publication/5469/detail/robust-discovery-of-positive-and-negative-rules-in-knowledge-bases-1 | https://www.dropbox.com/s/4hgcli75ccqe20t/Rudik_CR_ICDE.pdf?dl=0 | Robust Discovery of Positive and Negative Rules in Knowledge-Bases | We present RUDIK, a system for the discovery of declarative rules over knowledge-bases (KBs). RUDIK discovers rules that express positive relationships between entities, such as “if two persons have the same parent, they are siblings”, and negative rules, i.e., patterns that identify contradictions in the data, such as “if two persons are married, one cannot be the child of the other”. While the former class infers new facts in the KB, the latter class is crucial for other tasks, such as detecting erroneous triples in data cleaning, or the creation of negative examples to bootstrap learning algorithms. The system is designed to: (i) enlarge the expressive power of the rule language to obtain complex rules and wide coverage of the facts in the KB, (ii) discover approximate rules (soft constraints) to be robust to errors and incompleteness in the KB, (iii) use disk-based algorithms, effectively enabling rule mining in commodity machines. In contrast with traditional ranking of all rules based on a measure of support, we propose an approach to identify the subset of useful rules to be exposed to the user. We model the mining process as an incremental graph exploration problem and prove that our search strategy has guarantees on the optimality of the results. We have conducted extensive experiments using real-world KBs to show that RUDIK outperforms previous proposals in terms of efficiency and that it discovers more effective rules for the application at hand. | ['Paolo Papotti', 'Stefano Ortona', 'Venkata Vamsikrishna Meduri'] | 2018-04-16 | null | null | null | icde-2018-4 | ['knowledge-graphs-data-curation'] | ['knowledge-base'] | [-6.35629706e-03 6.70998454e-01 -4.58594471e-01 -2.52310902e-01
1.09942965e-01 -5.26283145e-01 7.41707906e-02 5.84295630e-01
-4.30214852e-02 1.26000071e+00 -2.84145683e-01 -6.73512459e-01
-6.75422251e-01 -1.49672365e+00 -8.64724338e-01 -2.34632701e-01
-5.37661254e-01 1.00000405e+00 7.35857844e-01 -9.23928246e-02
-7.52465278e-02 5.82975745e-01 -1.78063738e+00 2.57052273e-01
9.34643090e-01 9.62178111e-01 -1.85931712e-01 1.01024903e-01
-9.81502086e-02 1.00028574e+00 -2.80895144e-01 -6.17859304e-01
2.97677577e-01 -3.79090048e-02 -1.21378970e+00 -3.36434953e-02
-3.41062509e-02 -6.44638343e-03 4.05415939e-03 1.16437507e+00
-2.11136773e-01 -5.86549379e-02 3.24661493e-01 -1.41551363e+00
-2.79941708e-01 8.13400984e-01 -4.53764677e-01 7.41856918e-02
6.74645185e-01 -3.25722009e-01 1.02086425e+00 -5.04377246e-01
1.08118975e+00 9.43968475e-01 4.80494738e-01 3.75572979e-01
-1.12095606e+00 -4.38339949e-01 1.91733599e-01 5.01743019e-01
-1.49123704e+00 -5.90671420e-01 2.84925252e-01 -3.48508433e-02
1.16907108e+00 9.61359501e-01 6.81446135e-01 3.02105576e-01
-1.20118268e-01 4.40898031e-01 5.41270375e-01 -8.52728009e-01
4.83177632e-01 3.94233346e-01 3.97715390e-01 1.11114562e+00
1.15995193e+00 -1.03825770e-01 -7.36691654e-01 -5.70738614e-01
2.50228465e-01 -3.84437531e-01 -1.16155647e-01 -7.01707482e-01
-7.78799951e-01 7.34210670e-01 -4.51981947e-02 3.55924964e-01
-3.70163649e-01 -2.10138649e-01 2.58818507e-01 5.14397025e-01
4.31724042e-02 6.04967833e-01 -8.09253991e-01 1.62954777e-01
-5.38549960e-01 4.21699643e-01 1.28710783e+00 9.79847729e-01
8.70736599e-01 -4.57843602e-01 2.72266448e-01 4.76964712e-01
1.22921772e-01 2.52468288e-01 9.15036127e-02 -8.48259926e-01
4.88129675e-01 1.18482745e+00 3.95540476e-01 -9.51239824e-01
-3.61922950e-01 -9.33217704e-02 -5.86132050e-01 -7.13690147e-02
3.68835837e-01 1.63803843e-03 -7.40598202e-01 1.70032251e+00
6.14417613e-01 -2.09662452e-01 8.39958712e-02 4.28070992e-01
5.96064389e-01 3.93937111e-01 -1.27358645e-01 -8.53497565e-01
1.31702566e+00 -2.64803201e-01 -5.62476635e-01 -1.39402941e-01
7.97951281e-01 -8.86507779e-02 5.16912282e-01 6.09060943e-01
-1.20745122e+00 -4.41006683e-02 -9.35398102e-01 1.83685765e-01
-5.24108052e-01 -2.57410884e-01 1.05876374e+00 6.28729820e-01
-8.20194840e-01 4.44607764e-01 -5.98640919e-01 -3.87922078e-01
1.44333944e-01 5.89415967e-01 -5.47434449e-01 -3.80487680e-01
-1.33585250e+00 9.11576927e-01 9.06897366e-01 -1.20918274e-01
-1.41217440e-01 -6.35406494e-01 -9.99423146e-01 2.73139626e-01
1.08231449e+00 -7.70695627e-01 7.22560048e-01 -9.58443701e-01
-6.21090770e-01 8.74105155e-01 -3.78811419e-01 -6.03236973e-01
3.18588287e-01 1.97266191e-01 -7.29128778e-01 4.85310405e-02
2.47760862e-01 4.76355255e-02 2.18389928e-01 -1.22552037e+00
-1.04835284e+00 -6.62372708e-01 3.54219198e-01 -7.17176348e-02
-3.16187650e-01 -6.41660690e-02 -6.35546267e-01 -3.74971293e-02
3.28598380e-01 -7.81233013e-01 -1.51446685e-01 -4.52531815e-01
-6.53335035e-01 -3.54098499e-01 6.08907819e-01 -4.67228383e-01
1.50209439e+00 -1.85861790e+00 -1.42131001e-01 1.18595028e+00
2.41958931e-01 2.46071354e-01 4.83289003e-01 3.96072865e-01
-5.86123616e-02 3.27636898e-01 -1.12953395e-01 4.50800210e-01
-2.65175283e-01 7.91121364e-01 -3.53751212e-01 1.10892236e-01
-3.87731120e-02 5.75802267e-01 -7.85915077e-01 -6.43761218e-01
-2.27365777e-01 -4.22328830e-01 -4.23023462e-01 -1.61256760e-01
-4.26562279e-01 -4.31247175e-01 -3.65751386e-01 6.52817726e-01
5.28670192e-01 -2.78692424e-01 7.87665248e-01 -4.94825467e-02
2.16895305e-02 2.72454768e-01 -1.73891747e+00 6.27939224e-01
5.40681630e-02 -1.90755025e-01 4.27029133e-02 -1.04806185e+00
7.74418712e-01 8.57878178e-02 4.22328949e-01 -5.76925993e-01
-4.50864643e-01 2.17243075e-01 -1.66263863e-01 -7.09920228e-01
3.59530866e-01 1.24437362e-02 4.64694723e-02 3.93893093e-01
-2.02603355e-01 3.54870349e-01 9.12974238e-01 4.79463309e-01
1.38583589e+00 -2.97923684e-01 6.55854583e-01 -3.78979504e-01
5.91907203e-01 3.26079488e-01 9.90384996e-01 7.87939608e-01
5.18577933e-01 -3.70889425e-01 7.86120534e-01 -8.34833503e-01
-5.24268568e-01 -9.12766695e-01 1.13118999e-02 7.27880001e-01
3.04062545e-01 -6.98366106e-01 -2.97076255e-01 -9.23840880e-01
4.38675404e-01 8.43581080e-01 -5.94572902e-01 1.37078320e-03
-3.73874903e-01 -6.02221429e-01 3.85450155e-01 2.92140931e-01
1.38378322e-01 -9.28909242e-01 -7.56900132e-01 2.50692844e-01
-2.61125863e-01 -1.12415314e+00 1.84672490e-01 4.21080798e-01
-7.88278520e-01 -1.83219981e+00 6.48283303e-01 -5.36337972e-01
8.93445432e-01 -1.49574518e-01 1.26055002e+00 5.71223021e-01
-2.03153223e-01 1.54819086e-01 -4.46231782e-01 -5.43257356e-01
-3.52236927e-01 -8.31994116e-02 2.66800046e-01 -3.71607065e-01
4.93190885e-01 -6.33037686e-01 7.76487812e-02 3.74920011e-01
-8.48622203e-01 -9.40273777e-02 3.78403723e-01 5.96119404e-01
6.83276474e-01 7.94839084e-01 6.65036917e-01 -1.45064533e+00
4.65299368e-01 -4.76532400e-01 -7.01019704e-01 9.32612956e-01
-1.17703462e+00 2.15007514e-01 6.05469286e-01 -1.47603795e-01
-9.45563853e-01 1.48017779e-01 2.96084374e-01 5.55190518e-02
-3.22462507e-02 9.48813677e-01 -3.62581849e-01 6.56803176e-02
6.32552147e-01 5.89704029e-02 -3.02218765e-01 -1.32980332e-01
1.48722559e-01 1.73596591e-01 7.05886781e-01 -9.12538588e-01
7.30873823e-01 4.77865726e-01 3.48423272e-01 -7.08505332e-01
-5.55303335e-01 -4.29610372e-01 -4.07168686e-01 4.42124233e-02
3.95322219e-02 -4.68644381e-01 -9.87051487e-01 -2.75688618e-01
-7.76408672e-01 -5.28027937e-02 -4.81087416e-01 1.06727168e-01
-3.70281756e-01 5.76738596e-01 -4.59882677e-01 -1.01428103e+00
-3.15058440e-01 -4.92601603e-01 3.09902966e-01 -9.84492619e-03
-5.62943041e-01 -6.06163859e-01 -2.16185115e-03 2.22442299e-01
-2.72707522e-01 3.14750135e-01 1.46266913e+00 -1.01086855e+00
-5.16943097e-01 -1.88544974e-01 6.03821501e-02 -1.16177775e-01
1.72937468e-01 1.54874161e-01 -3.42351437e-01 -1.14583485e-01
-4.89767224e-01 -1.13534383e-01 3.97998840e-01 -3.35093066e-02
1.16283739e+00 -6.78968668e-01 -7.78809905e-01 1.93295181e-01
1.24851704e+00 3.02003175e-01 7.76371479e-01 3.88498127e-01
1.18946552e-01 6.61607325e-01 7.40297854e-01 4.87321079e-01
4.80579883e-01 6.29192650e-01 3.55133682e-01 1.80280775e-01
5.57578504e-01 -2.67913908e-01 -6.70262054e-02 1.93797320e-01
-4.30707276e-01 5.45910485e-02 -8.13713431e-01 7.28420377e-01
-2.14899445e+00 -1.18062520e+00 -1.56167150e-01 2.41808724e+00
1.12403941e+00 4.01240945e-01 2.47885272e-01 5.58317304e-01
5.14290154e-01 -5.43593824e-01 -3.91043663e-01 -6.49577320e-01
-4.49332558e-02 2.73110479e-01 4.59422201e-01 5.30272007e-01
-6.63804889e-01 7.24611044e-01 5.69677830e+00 4.11399096e-01
-5.47249556e-01 -2.48740554e-01 3.64270180e-01 -1.30031720e-01
-5.41823387e-01 3.44010890e-01 -8.01191270e-01 2.17454955e-01
5.92094660e-01 -3.89658570e-01 4.66393679e-01 9.30078804e-01
-1.33122593e-01 -4.44960564e-01 -1.31592035e+00 5.07239461e-01
-2.91860193e-01 -1.42431092e+00 3.43075804e-02 3.05314034e-01
5.07755041e-01 -3.91215771e-01 -6.90550804e-01 2.46873647e-01
5.94726980e-01 -8.43265235e-01 7.08100200e-01 4.64701504e-01
6.28248453e-01 -1.19887781e+00 6.94069684e-01 6.51710629e-01
-9.68782902e-01 -3.94824415e-01 -3.54728878e-01 -8.78063813e-02
-8.72887149e-02 1.17604136e+00 -1.06606448e+00 1.02981019e+00
8.90304804e-01 3.60038579e-02 -3.55778247e-01 8.02067935e-01
-1.89341351e-01 3.39873612e-01 -7.06554472e-01 1.25724688e-01
-2.52757072e-01 -1.54815689e-01 3.35462779e-01 1.04743195e+00
2.21431747e-01 4.81827766e-01 5.75554445e-02 5.86767614e-01
-9.86344218e-02 1.89026166e-02 -8.69854748e-01 2.18104661e-01
8.48493457e-01 9.14277971e-01 -6.09038472e-01 -5.08891881e-01
-2.05685854e-01 3.33051026e-01 5.40942550e-01 2.71126330e-01
-3.84131044e-01 -3.36902976e-01 4.01419312e-01 5.06652117e-01
3.90902847e-01 1.07091606e-01 -2.76491016e-01 -9.45259452e-01
5.50825357e-01 -8.72400999e-01 1.04023099e+00 -4.77428794e-01
-8.18590701e-01 2.40998134e-01 1.51383474e-01 -4.16720450e-01
-5.82288086e-01 -4.33223724e-01 -3.50140959e-01 5.18856347e-01
-1.26120973e+00 -8.34732234e-01 3.03619821e-02 8.05409312e-01
-1.77336007e-01 1.68946892e-01 9.08197165e-01 3.07670422e-02
-2.12724015e-01 3.96300554e-01 -4.50287163e-01 -1.22941501e-01
2.19896749e-01 -1.03769326e+00 -1.64055288e-01 1.02780139e+00
2.01607481e-01 9.62176085e-01 7.68034756e-01 -9.54332650e-01
-1.48425245e+00 -8.54766309e-01 1.37684417e+00 -1.34549171e-01
3.07185382e-01 -1.35950735e-02 -8.42727423e-01 9.22305405e-01
-4.53922570e-01 1.24701358e-01 5.96068978e-01 5.89412808e-01
-4.05513793e-01 -4.77614164e-01 -1.36595488e+00 3.12113047e-01
1.27563834e+00 -2.27678977e-02 -5.58849037e-01 2.50263304e-01
2.77267933e-01 -1.63686603e-01 -8.89952779e-01 8.28894198e-01
6.80747271e-01 -1.02360845e+00 7.71342456e-01 -1.10817587e+00
-1.97949000e-02 -6.20226204e-01 4.42706682e-02 -9.52793956e-01
-4.05192316e-01 -4.07131821e-01 -8.86322320e-01 1.01361060e+00
7.78356433e-01 -5.67404032e-01 7.50662804e-01 1.01378357e+00
1.72941908e-01 -9.64552641e-01 -1.15736794e+00 -8.92405927e-01
-6.19119704e-01 -2.41720542e-01 8.82689178e-01 1.23466492e+00
6.54082716e-01 2.46783733e-01 -2.77288586e-01 3.96904081e-01
3.80821586e-01 6.14931226e-01 6.49807513e-01 -1.83815205e+00
-3.09066921e-01 -5.92872016e-02 -2.57810146e-01 -2.28502989e-01
7.85042867e-02 -7.09710360e-01 -2.24751890e-01 -1.61324465e+00
2.40358040e-01 -8.17753673e-01 -1.11621001e-03 1.02463579e+00
2.10469097e-01 -2.43669763e-01 -2.59946585e-01 -7.79932961e-02
-7.77310729e-01 -2.53715426e-01 6.10361397e-01 -4.66906168e-02
-4.02548164e-01 2.88894400e-02 -9.82365966e-01 8.71448994e-01
4.30171579e-01 -3.93299192e-01 -3.73348117e-01 4.23058383e-02
1.08199334e+00 3.28890443e-01 5.33551574e-02 -5.21880627e-01
3.71035457e-01 -6.26576722e-01 2.82022446e-01 -2.54665852e-01
-1.94235399e-01 -1.05122828e+00 7.44890928e-01 7.94142306e-01
-1.13600172e-01 -8.80149975e-02 -1.22169433e-02 3.90395910e-01
-5.96272647e-02 -5.10289192e-01 3.22939515e-01 -9.68091711e-02
-7.36366451e-01 1.45825431e-01 -4.01143879e-02 -1.83627844e-01
1.20282483e+00 -2.87254274e-01 -3.81236166e-01 -1.70499504e-01
-7.60980189e-01 5.01964450e-01 5.21773100e-01 6.36544526e-02
6.44801557e-01 -9.67006385e-01 -5.07434487e-01 1.23885311e-01
4.45113569e-01 3.49129915e-01 -2.95960307e-01 7.11595714e-01
-2.99178928e-01 2.48989969e-01 8.09545442e-02 4.77682054e-02
-1.63045681e+00 8.66834641e-01 5.79757057e-02 -5.58219194e-01
-4.39272523e-01 4.32788312e-01 -2.46338397e-01 -1.56579494e-01
1.67193010e-01 -2.42493242e-01 -1.18275650e-01 -1.91461220e-02
6.19587302e-01 3.76614541e-01 5.08248448e-01 -9.68994871e-02
-7.93525338e-01 -1.37612939e-01 -1.43851072e-01 4.29281116e-01
1.49649405e+00 4.17788997e-02 -6.05229676e-01 4.85395677e-02
1.71092361e-01 4.74721730e-01 -2.57040620e-01 -3.06294203e-01
5.29854417e-01 -4.29519832e-01 -4.12359983e-01 -1.06175458e+00
-8.80942583e-01 -1.05506338e-01 -1.83220655e-01 6.89325988e-01
1.15182877e+00 3.21576297e-01 3.98530275e-01 8.89419079e-01
7.82565713e-01 -1.21365237e+00 -6.82384431e-01 3.35903943e-01
5.87043583e-01 -6.93910837e-01 5.04737020e-01 -8.91017914e-01
-5.54011464e-01 1.20204020e+00 5.60547113e-01 3.87910128e-01
3.47117215e-01 5.35497844e-01 -6.79934084e-01 -3.44979852e-01
-1.12852216e+00 -2.57480085e-01 -1.65671052e-03 5.41079819e-01
-1.19562015e-01 2.28399232e-01 -6.06845617e-01 8.44376147e-01
-1.62743509e-01 5.12902796e-01 3.82064670e-01 9.39876080e-01
-6.57233417e-01 -1.11764491e+00 -3.26271027e-01 8.32696676e-01
-3.72589707e-01 1.11004956e-01 -6.32479191e-01 9.09570336e-01
5.73990226e-01 1.05244410e+00 -1.42828658e-01 -1.71914995e-01
5.32666206e-01 1.84807494e-01 6.65096283e-01 -5.04380226e-01
-9.56611410e-02 -5.78928590e-01 9.45260406e-01 -3.57548743e-01
-2.42429733e-01 -5.61927378e-01 -1.36774135e+00 -4.94691908e-01
-5.30910492e-01 6.61643982e-01 1.92076653e-01 1.09735739e+00
4.15297210e-01 5.97607121e-02 3.82749140e-01 3.50488633e-01
-3.50292444e-01 -5.23983419e-01 -7.88933337e-01 2.22073331e-01
-1.59871668e-01 -6.51785493e-01 3.48611223e-03 1.12871103e-01] | [8.867708206176758, 7.161899089813232] |
652ba512-6dee-41fa-acce-2155f7d0c436 | learning-to-personalize-for-web-search | 2009.08206 | null | https://arxiv.org/abs/2009.08206v1 | https://arxiv.org/pdf/2009.08206v1.pdf | Learning to Personalize for Web Search Sessions | The task of session search focuses on using interaction data to improve relevance for the user's next query at the session level. In this paper, we formulate session search as a personalization task under the framework of learning to rank. Personalization approaches re-rank results to match a user model. Such user models are usually accumulated over time based on the user's browsing behaviour. We use a pre-computed and transparent set of user models based on concepts from the social science literature. Interaction data are used to map each session to these user models. Novel features are then estimated based on such models as well as sessions' interaction data. Extensive experiments on test collections from the TREC session track show statistically significant improvements over current session search algorithms. | ['Stephen Clark', 'Saad Aloteibi'] | 2020-09-17 | null | null | null | null | ['session-search'] | ['natural-language-processing'] | [ 3.55325431e-01 -5.88080585e-02 -5.23008704e-01 -7.38608420e-01
-8.25743556e-01 -5.54431677e-01 8.79065275e-01 5.66336036e-01
-8.94047022e-01 5.81561208e-01 6.79158330e-01 -1.08785771e-01
-7.28377998e-01 -3.55510592e-01 -2.03363672e-01 -3.33817378e-02
-4.41976875e-01 7.05508590e-01 5.15539765e-01 -3.59715641e-01
1.03284156e+00 8.79645571e-02 -2.02082968e+00 8.30949783e-01
1.07202268e+00 6.40125811e-01 4.56522018e-01 1.02726436e+00
-1.71259686e-01 -2.64954180e-01 -3.95165443e-01 -1.09400012e-01
4.25317734e-02 -3.30404788e-01 -1.27028823e+00 -2.85430193e-01
2.49106243e-01 -6.56350553e-02 -1.58458844e-01 7.20839798e-01
3.08612734e-01 8.14031482e-01 7.11849570e-01 -8.36429060e-01
-2.28433043e-01 5.95208108e-01 -2.56906860e-02 5.56136131e-01
9.60716069e-01 -6.30455911e-01 1.33748758e+00 -6.58105791e-01
5.67131579e-01 1.16723514e+00 2.19220206e-01 6.23832762e-01
-1.49212325e+00 -4.30471301e-01 2.94205427e-01 7.05481946e-01
-1.32696259e+00 -2.81558335e-01 4.65783596e-01 -3.87921035e-01
8.87500226e-01 7.87003100e-01 4.56497103e-01 6.50912046e-01
-9.91667137e-02 7.68904448e-01 7.47438073e-01 -7.48029470e-01
1.92815661e-01 6.89031720e-01 6.71545923e-01 1.53197423e-01
-4.01563644e-02 -8.21255744e-02 -7.81056643e-01 -7.86831379e-01
1.71994895e-01 5.49279079e-02 -3.44748318e-01 -4.86282796e-01
-8.19074273e-01 8.21677327e-01 4.31015104e-01 5.90901554e-01
-5.72199702e-01 -3.77824038e-01 1.74915612e-01 4.70771283e-01
4.11981761e-01 9.61377561e-01 -6.05587900e-01 -1.26703233e-01
-7.33173311e-01 4.29349661e-01 1.11132777e+00 8.87109995e-01
9.98751700e-01 -1.24765766e+00 -6.25783920e-01 1.14905059e+00
4.45022523e-01 -3.21529508e-02 8.39354634e-01 -9.24445391e-01
9.09734219e-02 6.01776183e-01 4.05093223e-01 -6.26345932e-01
-1.59409210e-01 -4.15427238e-01 -2.06016213e-01 -4.05570060e-01
-4.13400307e-02 3.68926346e-01 -1.00785029e+00 1.51649928e+00
1.30091161e-01 8.60309079e-02 -5.75971231e-02 3.32757175e-01
5.27881861e-01 4.61195379e-01 4.97812539e-01 -6.41879022e-01
1.17349505e+00 -6.38961852e-01 -4.61659342e-01 1.23377152e-01
8.89131188e-01 -8.73225212e-01 1.06658483e+00 3.60994399e-01
-9.26286399e-01 -5.65054297e-01 -5.79583824e-01 2.73694307e-01
-4.08352196e-01 -4.49878395e-01 4.65669036e-01 7.31929958e-01
-1.29366791e+00 8.63190591e-01 -4.99156147e-01 -1.04457760e+00
-2.14459896e-01 6.00987852e-01 2.70502865e-01 2.29079381e-01
-1.44623840e+00 7.68881619e-01 6.43302023e-01 -4.87047732e-01
-2.20861867e-01 -7.65853524e-01 -3.53592575e-01 2.31273293e-01
9.38421860e-02 -8.33857596e-01 1.54610157e+00 -4.53056842e-01
-1.29088402e+00 7.67333567e-01 -6.83764875e-01 -3.73633325e-01
6.66874647e-02 -1.30346730e-01 -4.92082268e-01 -1.50546536e-01
6.39747009e-02 3.44159335e-01 4.90352064e-01 -1.34835684e+00
-1.56908429e+00 -3.00072551e-01 2.43732095e-01 6.86439633e-01
-7.09439874e-01 7.25314170e-02 -1.07604575e+00 -3.00242491e-02
1.50725886e-01 -1.24511814e+00 -4.83912915e-01 -7.69951761e-01
9.08208638e-02 -9.53300476e-01 6.76732004e-01 -5.56795776e-01
2.03996062e+00 -1.61370623e+00 2.38105774e-01 1.00442421e+00
1.07624613e-01 1.75521314e-01 -5.63049838e-02 6.45308912e-01
4.67781276e-02 3.99354041e-01 5.26470423e-01 -1.98281065e-01
-5.84159791e-02 -2.05124721e-01 2.94853393e-02 -2.57012844e-01
-9.94353652e-01 6.92357123e-01 -9.64925110e-01 -7.13084757e-01
3.81282810e-03 2.41768025e-02 -9.19810772e-01 3.99214029e-01
-1.60728604e-01 4.14141744e-01 -7.92712867e-01 1.26944005e-01
1.21142603e-01 -3.49672705e-01 3.08268607e-01 -1.24889068e-01
-4.85917972e-03 4.21923935e-01 -7.92023242e-01 1.84858012e+00
-6.72210574e-01 5.51352389e-02 -4.55622107e-01 -8.05577517e-01
3.78379226e-01 2.82214969e-01 7.63163626e-01 -8.34268630e-01
-2.92675823e-01 1.32143930e-01 -1.31178632e-01 -6.25200331e-01
7.68631577e-01 5.29408097e-01 -1.64222836e-01 8.22343767e-01
-8.33241642e-02 2.64229476e-01 6.14788830e-01 7.30265021e-01
1.08518612e+00 -2.57313579e-01 4.17266428e-01 -4.40606982e-01
6.87695920e-01 -9.42923203e-02 5.06165661e-02 1.39129305e+00
8.44326690e-02 -1.05063962e-02 -2.43324235e-01 3.55329365e-02
-7.76413500e-01 -9.23095107e-01 -2.05164775e-01 1.92809820e+00
1.03863753e-01 -1.04394484e+00 -5.07746279e-01 -7.57293582e-01
1.15119228e-02 8.61191332e-01 -7.40728259e-01 -3.30334753e-01
-3.59457195e-01 -3.41514647e-01 -2.21720934e-01 1.94859996e-01
1.68233499e-01 -1.03109872e+00 -1.17934391e-01 2.42537022e-01
-3.33660275e-01 -3.89250129e-01 -7.77490139e-01 -7.38480389e-02
-8.01402986e-01 -7.84174204e-01 -7.28703022e-01 -9.11094964e-01
5.73552907e-01 4.42334801e-01 1.09653425e+00 4.47922766e-01
-1.18663073e-01 1.09287441e+00 -6.49720490e-01 -2.02053726e-01
-2.64283687e-01 4.96527404e-01 2.56267637e-01 -1.60344645e-01
6.59533918e-01 -5.64972937e-01 -8.83411884e-01 5.50047219e-01
-8.58629346e-01 -1.32416487e-01 3.47190797e-01 7.39866436e-01
1.78638443e-01 2.41280168e-01 3.34127694e-01 -1.18099272e+00
1.19795024e+00 -5.56716383e-01 -1.90595374e-01 9.40016389e-01
-1.37233841e+00 2.17509583e-01 -1.61298722e-01 -4.10500765e-01
-1.48360300e+00 6.09819079e-03 3.51076871e-01 2.88662523e-01
-4.51520421e-02 9.74808872e-01 5.88381775e-02 -1.94506600e-01
9.71955240e-01 6.83235675e-02 -6.18884146e-01 -6.55626714e-01
2.77113199e-01 9.46795166e-01 2.95950979e-01 -5.53049922e-01
7.43455887e-01 -1.27018020e-01 -3.98140639e-01 -7.53751755e-01
-9.06172514e-01 -1.52626693e+00 -8.21184754e-01 -3.55835795e-01
4.57527876e-01 -3.24569374e-01 -9.15275872e-01 -5.18017747e-02
-7.71766186e-01 -1.07182756e-01 9.89400223e-02 5.10327280e-01
-4.40106958e-01 5.94325185e-01 -1.91070959e-01 -7.78273106e-01
-3.71188104e-01 -4.81514573e-01 7.47274101e-01 7.01277912e-01
-6.52278423e-01 -1.36454630e+00 5.27706683e-01 4.65329349e-01
4.35933948e-01 -7.27268338e-01 1.16273928e+00 -1.10676217e+00
-5.67389250e-01 -6.93098664e-01 3.37199606e-02 -1.59107849e-01
3.38723958e-01 -4.87088561e-01 -7.24228799e-01 -5.03238380e-01
-4.67249960e-01 1.48243085e-01 9.34924543e-01 1.34962842e-01
1.46809268e+00 -2.35784456e-01 -1.03285408e+00 5.83941154e-02
1.06779492e+00 4.97526437e-01 4.47366953e-01 4.70079601e-01
2.58553416e-01 6.69179916e-01 9.00013030e-01 5.65726399e-01
2.13001087e-01 9.62921560e-01 -2.53962010e-01 4.62932438e-01
3.75007778e-01 -3.62970412e-01 -2.29118884e-01 1.88220114e-01
-2.99629331e-01 -4.77098644e-01 -8.24688435e-01 4.00961965e-01
-2.03332281e+00 -1.09552455e+00 1.70939758e-01 2.71243644e+00
9.07772541e-01 1.96233734e-01 2.64334559e-01 -1.71076506e-01
7.53717124e-01 -3.51746112e-01 -3.79594028e-01 -2.44129255e-01
6.28042817e-01 3.99646729e-01 3.23884606e-01 8.33027661e-01
-8.74521077e-01 8.80826831e-01 6.68030691e+00 6.22931540e-01
-3.90556842e-01 -2.18242645e-01 1.69119030e-01 9.14072320e-02
-5.00771880e-01 2.52939045e-01 -1.05631316e+00 3.69950473e-01
1.08418858e+00 -1.27973104e+00 4.73953843e-01 5.98922670e-01
2.03565821e-01 -3.68305981e-01 -1.54997027e+00 9.46888626e-01
1.62290469e-01 -9.81930315e-01 2.32282460e-01 2.89098829e-01
8.15244436e-01 -3.77267450e-01 1.43521696e-01 7.64162779e-01
5.95138609e-01 -4.94366467e-01 -4.00312185e-01 8.21042180e-01
3.58644634e-01 -5.08806884e-01 3.57379198e-01 4.95720923e-01
-1.10124803e+00 -2.22640529e-01 -7.91580006e-02 4.13328618e-01
2.07857639e-01 1.02786601e-01 -1.54051614e+00 3.77094895e-01
8.97696614e-01 7.15085149e-01 -1.05894578e+00 1.67552876e+00
2.31336772e-01 6.13925159e-01 -4.14815456e-01 -1.84017047e-01
3.02460836e-03 1.73022941e-01 5.56863010e-01 1.08951688e+00
3.42346162e-01 2.34925836e-01 3.19318682e-01 8.75821188e-02
2.52437145e-01 5.44413567e-01 -2.20516846e-01 1.73015296e-01
4.27401483e-01 1.05830562e+00 -4.78103936e-01 -7.88858175e-01
-1.21163264e-01 1.25440490e+00 -2.81046685e-02 4.92679656e-01
-1.21701308e-01 -4.48328555e-01 3.15525472e-01 2.51772970e-01
-1.54060677e-01 1.44890547e-01 4.79902625e-02 -1.09529698e+00
-1.30752593e-01 -4.29330856e-01 1.17415309e+00 -4.70464110e-01
-1.12518466e+00 4.79258835e-01 6.30997300e-01 -1.13037777e+00
-8.10589492e-01 -6.02810942e-02 -7.38132417e-01 1.03006661e+00
-1.00523758e+00 -7.70415068e-01 -4.90292370e-01 5.61470032e-01
7.45100319e-01 -1.78478345e-01 1.11859047e+00 1.80845946e-01
2.98583563e-02 7.79245377e-01 3.90034646e-01 -6.45209491e-01
1.00737917e+00 -1.31267834e+00 2.62569934e-01 2.65541226e-01
7.68263713e-02 1.15418577e+00 1.05913007e+00 -9.58460867e-01
-5.97688377e-01 -5.43264329e-01 1.44049883e+00 -7.50239789e-01
5.91003656e-01 4.70582321e-02 -1.12316418e+00 5.28600156e-01
1.63424969e-01 -8.37567031e-01 1.11229336e+00 1.05836177e+00
-7.33685717e-02 1.45404087e-02 -7.63659537e-01 6.57208800e-01
1.26028061e+00 -6.17397845e-01 -9.15483117e-01 5.60656369e-01
2.47992322e-01 -1.62004698e-02 -8.02900195e-01 3.35192442e-01
7.99680352e-01 -5.00700355e-01 1.19767249e+00 -8.96736860e-01
-4.58671659e-01 1.46419138e-01 2.36921366e-02 -1.52133477e+00
-7.32356369e-01 -7.12267935e-01 -7.62223303e-02 1.08927345e+00
6.45895600e-01 -2.48928353e-01 1.14866984e+00 1.27706456e+00
1.73740700e-01 -2.76848108e-01 -4.66054946e-01 -5.81160665e-01
-4.37557071e-01 4.90385890e-02 3.87005091e-01 5.67700922e-01
5.76160729e-01 6.28095627e-01 -2.02616215e-01 7.05820546e-02
7.92989552e-01 4.43280078e-02 7.20149934e-01 -1.95617509e+00
-2.67031550e-01 -4.05226976e-01 -9.89492387e-02 -1.25137377e+00
1.27163291e-01 -1.03040695e+00 1.13529891e-01 -1.56164563e+00
6.08851612e-01 -4.43653166e-01 -1.05326414e+00 -7.89109338e-03
-3.93007934e-01 -2.98385620e-02 -2.34633490e-01 4.90756541e-01
-1.25919271e+00 1.44256592e-01 5.30266404e-01 1.47054240e-01
-1.05111325e+00 8.95770133e-01 -7.67239869e-01 5.07464230e-01
6.39637470e-01 -3.14936012e-01 -7.09192216e-01 2.54773553e-02
2.49197140e-01 1.26525909e-01 -1.78828299e-01 -7.39345610e-01
7.73142219e-01 -4.64063615e-01 2.40138292e-01 -7.87494123e-01
3.34974289e-01 -7.10899472e-01 -1.46716544e-02 3.57537478e-01
-1.18175948e+00 -2.99861908e-01 -1.96167797e-01 9.48079348e-01
-1.94796547e-02 -4.70367700e-01 4.10877377e-01 -1.42762840e-01
-1.15796328e+00 3.55444908e-01 -4.46697295e-01 -3.90145302e-01
8.00129294e-01 -2.54178643e-01 4.23272192e-01 -8.16817045e-01
-1.27883768e+00 6.54886007e-01 1.98843479e-01 7.81601131e-01
3.41327399e-01 -1.26560569e+00 -3.39545965e-01 -1.76192801e-02
7.52428353e-01 -7.51466274e-01 3.88109922e-01 2.20025659e-01
2.21475989e-01 1.02464676e+00 4.38200720e-02 -5.34232199e-01
-1.91425598e+00 3.09991598e-01 -5.06777838e-02 -5.05844951e-01
2.48238258e-02 8.28935623e-01 2.95783579e-01 -8.81341323e-02
6.36381447e-01 3.39500487e-01 -8.85735691e-01 2.19741464e-01
7.09279537e-01 2.87452549e-01 1.39964923e-01 -1.92734703e-01
-2.07779631e-01 2.70644546e-01 -8.74095142e-01 -4.97828484e-01
1.10972774e+00 -7.38651633e-01 1.48594126e-01 3.86943042e-01
1.21678329e+00 -1.97452620e-01 -4.16518569e-01 -1.01393092e+00
7.93834329e-01 -7.90960908e-01 -4.10097130e-02 -9.72927272e-01
-1.85796261e-01 1.75485581e-01 9.47298765e-01 4.14945513e-01
1.00470471e+00 3.18628699e-01 4.84530389e-01 9.41755772e-01
4.35799152e-01 -1.51962590e+00 2.39245400e-01 5.88906527e-01
7.20839024e-01 -1.15672886e+00 1.36039957e-01 -2.51840264e-01
-4.51997876e-01 7.73164392e-01 6.18751287e-01 4.68104959e-01
1.10147297e+00 -9.33139741e-01 -2.02938095e-01 5.72217396e-03
-9.25554812e-01 -4.10056263e-01 9.58644211e-01 3.52783561e-01
6.26212537e-01 -1.33063033e-01 -9.22431231e-01 3.81070673e-01
-1.38082445e-01 3.13616754e-03 9.11787376e-02 1.10583639e+00
-9.37269688e-01 -1.63664210e+00 -5.47247492e-02 9.30488467e-01
-2.06697211e-01 -7.06554577e-02 -4.30886835e-01 1.42793491e-01
-4.45075065e-01 1.05545831e+00 -1.92310125e-01 -5.84008098e-01
4.20939505e-01 3.78389835e-01 2.79564172e-01 -1.13312364e+00
-4.85082030e-01 1.77535385e-01 1.44900516e-01 -5.81980824e-01
-3.87150735e-01 -9.15996730e-01 -6.84132516e-01 -1.74893066e-01
-4.81536359e-01 1.14460409e+00 7.70043254e-01 7.48638511e-01
4.33576494e-01 -4.74813059e-02 8.47379446e-01 -7.29955316e-01
-4.25070584e-01 -1.10052407e+00 -4.60546821e-01 9.26934183e-01
-9.40799564e-02 -5.88678002e-01 -3.41491997e-01 2.48885170e-01] | [11.757423400878906, 7.607293128967285] |
3b6b4295-3de3-443a-b2f2-c16d87498fdf | pivotal-prior-driven-supervision-for-weakly | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Rizve_PivoTAL_Prior-Driven_Supervision_for_Weakly-Supervised_Temporal_Action_Localization_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Rizve_PivoTAL_Prior-Driven_Supervision_for_Weakly-Supervised_Temporal_Action_Localization_CVPR_2023_paper.pdf | PivoTAL: Prior-Driven Supervision for Weakly-Supervised Temporal Action Localization | Weakly-supervised Temporal Action Localization (WTAL) attempts to localize the actions in untrimmed videos using only video-level supervision. Most recent works approach WTAL from a localization-by-classification perspective where these methods try to classify each video frame followed by a manually-designed post-processing pipeline to aggregate these per-frame action predictions into action snippets. Due to this perspective, the model lacks any explicit understanding of action boundaries and tends to focus only on the most discriminative parts of the video resulting in incomplete action localization. To address this, we present PivoTAL, Prior-driven Supervision for Weakly-supervised Temporal Action Localization, to approach WTAL from a localization-by-localization perspective by learning to localize the action snippets directly. To this end, PivoTAL leverages the underlying spatio-temporal regularities in videos in the form of action-specific scene prior, action snippet generation prior, and learnable Gaussian prior to supervise the localization-based training. PivoTAL shows significant improvement (of at least 3% avg mAP) over all existing methods on the benchmark datasets, THUMOS-14 and ActivitNet-v1.3. | ['Mei Chen', 'Mubarak Shah', 'Sandra Sajeev', 'Matthew Hall', 'Ye Yu', 'Gaurav Mittal', 'Mamshad Nayeem Rizve'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['weakly-supervised-action-localization', 'action-localization', 'action-recognition'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 5.43294966e-01 -8.21950089e-04 -7.66666293e-01 -1.91511959e-01
-9.44489360e-01 -5.49788058e-01 7.50446796e-01 -1.78937525e-01
-3.03550094e-01 4.67235953e-01 5.85857689e-01 6.49056509e-02
1.81119725e-01 -1.55758023e-01 -9.83038187e-01 -7.05052912e-01
-1.79023281e-01 5.18947542e-02 7.29405522e-01 2.16944039e-01
2.32765958e-01 1.33214489e-01 -1.22023344e+00 7.69001067e-01
3.82850438e-01 1.08407688e+00 2.16749802e-01 6.26327276e-01
3.28549653e-01 1.68393612e+00 -2.48539582e-01 8.99835229e-02
2.49850616e-01 -5.79553246e-01 -8.51068139e-01 4.29373592e-01
6.56916082e-01 -5.23384511e-01 -5.30524194e-01 6.55979574e-01
8.84954035e-02 1.86248809e-01 4.96883780e-01 -1.40421271e+00
-2.02529877e-01 4.69163388e-01 -6.50884748e-01 3.72789383e-01
5.02308190e-01 4.12851751e-01 1.07572043e+00 -8.93118918e-01
6.98743641e-01 1.12794137e+00 6.79327488e-01 4.05851603e-01
-1.18403792e+00 -2.91359216e-01 4.30203885e-01 4.91023511e-01
-1.35497069e+00 -5.45662403e-01 6.48759723e-01 -6.46677196e-01
1.03554702e+00 -1.91573113e-01 5.12898803e-01 1.27341437e+00
2.29228288e-01 1.28661299e+00 8.13592970e-01 -1.92641765e-01
3.66706848e-01 -4.12186414e-01 -2.91518122e-01 6.90219939e-01
-3.53592962e-01 -8.58528540e-02 -9.65724230e-01 1.94681302e-01
8.73804152e-01 2.35608026e-01 -6.42597824e-02 -5.06079674e-01
-1.46563900e+00 3.83016616e-01 3.53890061e-01 2.06699342e-01
-6.52268291e-01 5.09301543e-01 6.46742284e-01 -1.83475628e-01
5.26425600e-01 1.23272598e-01 -6.95258856e-01 -4.87337410e-01
-1.25753939e+00 5.43615408e-03 2.34861314e-01 8.86954486e-01
7.30509222e-01 -1.02646779e-02 -6.59021556e-01 3.93103182e-01
2.05276817e-01 1.70720950e-01 3.20764154e-01 -1.27645612e+00
6.88239813e-01 6.06034577e-01 2.26692766e-01 -6.73424542e-01
-1.05187811e-01 -2.17825845e-01 -2.81293839e-01 8.47886875e-02
4.23633128e-01 -2.76673157e-02 -1.08807492e+00 1.63932955e+00
3.86759907e-01 7.32917488e-01 -5.72045334e-02 9.34947670e-01
2.59753495e-01 5.07560372e-01 3.69889528e-01 8.88046809e-03
1.15054023e+00 -1.58215344e+00 -6.18301272e-01 -3.78211588e-01
8.22645664e-01 -4.80869681e-01 1.14360535e+00 3.25462610e-01
-8.90350580e-01 -6.84180796e-01 -8.30262423e-01 1.74593879e-03
2.79116724e-02 6.93164229e-01 2.86655366e-01 7.44439289e-02
-9.29895222e-01 5.45626938e-01 -1.43374598e+00 -4.37954336e-01
8.28960359e-01 -3.00251767e-02 -5.77510893e-01 -1.34739175e-01
-7.28340209e-01 5.78114808e-01 4.78577048e-01 -3.90317403e-02
-1.48469901e+00 -7.02158928e-01 -1.11172378e+00 -1.12405404e-01
8.83897662e-01 -3.56690526e-01 1.31612849e+00 -1.36777925e+00
-1.43060422e+00 5.04119515e-01 -4.12120193e-01 -8.40803206e-01
5.72065473e-01 -6.06216133e-01 -5.65670058e-02 7.12404013e-01
4.57667917e-01 7.78790474e-01 8.85459900e-01 -9.25068617e-01
-1.05137885e+00 -3.50946337e-02 2.63682991e-01 1.94057137e-01
-1.60518974e-01 1.15727544e-01 -7.59234786e-01 -7.33033478e-01
-1.61451811e-03 -9.03289735e-01 -1.44939497e-01 6.64092302e-02
-3.08225155e-01 -3.60703588e-01 1.07905233e+00 -7.39104509e-01
1.27334416e+00 -2.26666212e+00 1.53128669e-01 -3.41238827e-01
1.17685609e-01 1.21569425e-01 -2.42876217e-01 4.78195190e-01
-1.20783873e-01 -2.28211805e-01 1.13683492e-01 -7.46474802e-01
-9.25388038e-02 2.88960397e-01 -3.16116035e-01 6.79886341e-01
2.88758874e-01 1.04896605e+00 -1.24629307e+00 -6.65913224e-01
5.82908332e-01 3.69432032e-01 -6.59495831e-01 2.77370572e-01
-5.03797770e-01 6.80795372e-01 -4.91858333e-01 9.85260904e-01
9.10804570e-02 -2.83067733e-01 6.63718283e-02 -4.03649479e-01
-1.61821038e-01 1.93409652e-01 -8.53540003e-01 1.99968052e+00
-2.61757791e-01 7.39479184e-01 -9.30926725e-02 -1.24547541e+00
1.46504521e-01 5.02552330e-01 1.07745028e+00 -4.32081521e-01
-1.74469516e-01 -4.97330129e-02 -3.57472628e-01 -6.26398563e-01
1.48586661e-01 1.39126122e-01 -3.42691056e-02 3.66799116e-01
4.36691254e-01 5.59664965e-01 3.24575692e-01 3.99099052e-01
1.62956262e+00 1.00393653e+00 2.46304363e-01 1.71387851e-01
4.54527259e-01 1.31245658e-01 7.93020368e-01 7.03930378e-01
-5.22184670e-01 7.94514477e-01 5.70523798e-01 -3.79995495e-01
-8.23575914e-01 -9.92534816e-01 2.97826678e-01 1.50858605e+00
8.47487822e-02 -7.67541945e-01 -8.95911098e-01 -1.26822960e+00
-3.18586767e-01 4.66211945e-01 -7.75609434e-01 -1.94338620e-01
-6.58374071e-01 8.32250267e-02 4.73223358e-01 9.23829854e-01
7.08690047e-01 -1.17382550e+00 -7.58367598e-01 2.25592270e-01
-3.16046685e-01 -1.64936280e+00 -7.86254585e-01 2.24906921e-01
-7.00581670e-01 -1.22676063e+00 -6.89139426e-01 -5.45549095e-01
7.27423310e-01 2.32139975e-01 8.95369828e-01 -3.12602729e-01
-5.77120371e-02 5.86332142e-01 -7.20679998e-01 1.49748579e-01
-5.31589566e-03 -1.80120543e-01 3.87688056e-02 4.81308430e-01
3.98686081e-01 -4.92336869e-01 -7.94893563e-01 5.55932045e-01
-6.65959656e-01 9.58995670e-02 8.21576655e-01 6.41356111e-01
7.10819781e-01 5.95347881e-02 1.48020968e-01 -4.20961529e-01
-2.77262360e-01 -5.44586957e-01 -3.36852282e-01 3.12744856e-01
-9.10982192e-02 -2.19861820e-01 5.84519207e-01 -6.08653069e-01
-9.26864445e-01 6.88617468e-01 2.46839598e-01 -9.34623241e-01
-4.03936505e-01 3.97350520e-01 -2.14921176e-01 1.69064328e-01
5.11020362e-01 5.71524799e-01 -3.12049091e-01 -3.96784037e-01
3.03976476e-01 3.04077923e-01 6.99727952e-01 -4.29718077e-01
6.15338624e-01 8.14375818e-01 -1.78658053e-01 -5.56797922e-01
-1.32487822e+00 -7.28005111e-01 -1.05659223e+00 -4.91859168e-01
1.34578300e+00 -1.19593346e+00 -4.74806875e-01 3.77076298e-01
-9.81653094e-01 -8.21167231e-01 -3.80122155e-01 7.08086193e-01
-9.52391148e-01 3.78080457e-01 -5.39528012e-01 -7.13905096e-01
1.57387391e-01 -1.02335382e+00 1.51043272e+00 -2.08290130e-01
-3.31528366e-01 -8.97819042e-01 -7.04847276e-02 6.74331784e-01
-2.32469309e-02 3.27578008e-01 3.07424903e-01 -5.37138879e-01
-8.20252657e-01 -2.85275280e-01 -5.09374999e-02 6.94325030e-01
2.50025809e-01 -1.77627906e-01 -7.65631258e-01 -1.35331899e-01
-1.86068073e-01 -5.17363131e-01 9.34123933e-01 5.53017080e-01
1.03960896e+00 -3.32376629e-01 -4.08214897e-01 5.27443171e-01
1.10269368e+00 1.94023345e-02 5.02085984e-01 1.52518928e-01
9.07121241e-01 3.75678688e-01 1.03434157e+00 5.02365589e-01
3.17537606e-01 9.23061073e-01 4.79618847e-01 -4.38506678e-02
-3.12244058e-01 -6.58709884e-01 9.21339333e-01 4.97928187e-02
-1.76287264e-01 -2.16124386e-01 -7.08607435e-01 6.32273018e-01
-2.31945205e+00 -1.24514878e+00 1.57189682e-01 1.96082163e+00
6.79980338e-01 1.62169889e-01 3.83257121e-01 -6.97244555e-02
6.00669861e-01 4.40741986e-01 -6.00914598e-01 4.15525526e-01
9.12839994e-02 -1.90486714e-01 5.84471345e-01 2.97607869e-01
-1.55886841e+00 1.29707658e+00 5.63299894e+00 9.60482478e-01
-1.00799978e+00 3.96523446e-01 4.88491774e-01 -3.29298049e-01
4.05638367e-01 2.95001835e-01 -6.97035015e-01 5.47167242e-01
7.27682054e-01 4.12521958e-01 7.85121843e-02 9.39975798e-01
8.09887171e-01 -4.98921663e-01 -1.41769958e+00 8.19979131e-01
1.04307443e-01 -1.32599151e+00 -1.77892998e-01 -2.08177906e-03
7.54131973e-01 1.81710556e-01 -1.02070235e-01 4.35456902e-01
2.10133314e-01 -8.60419333e-01 1.08170116e+00 5.35928965e-01
4.70711082e-01 -3.70991081e-01 5.83311617e-01 4.35194314e-01
-1.47169936e+00 -2.77848572e-01 4.94615771e-02 -2.36962736e-01
5.34949958e-01 1.86615825e-01 -7.15389371e-01 3.54298383e-01
6.38179243e-01 1.50608325e+00 -4.15073693e-01 7.39533484e-01
-6.24658346e-01 1.07310367e+00 -1.44157439e-01 5.72576225e-01
6.22400522e-01 5.73589504e-02 3.01945835e-01 1.07526839e+00
1.43429756e-01 5.03759533e-02 8.21677506e-01 5.50311089e-01
1.92046344e-01 -2.32947037e-01 -2.80766606e-01 -2.18363896e-01
1.38806939e-01 1.08345878e+00 -7.73629546e-01 -4.47335005e-01
-6.84524000e-01 1.36901903e+00 9.22132060e-02 5.70376933e-01
-1.19087899e+00 1.73988178e-01 5.00123799e-01 4.74017978e-01
5.87862670e-01 -4.18861508e-01 8.29676017e-02 -1.14185202e+00
1.98471770e-01 -6.64602876e-01 4.89690095e-01 -8.96112204e-01
-7.78820217e-01 1.60114259e-01 1.52268901e-01 -1.50635219e+00
-2.15809509e-01 -3.84421080e-01 -5.89036584e-01 4.03310299e-01
-1.28192091e+00 -1.50549400e+00 -1.30414248e-01 7.10416675e-01
1.11703169e+00 -4.11735959e-02 2.48994648e-01 2.91518897e-01
-6.39207006e-01 2.83080667e-01 -2.21593782e-01 3.95620376e-01
8.04672956e-01 -1.24201107e+00 1.50785327e-01 1.07328522e+00
4.63155866e-01 2.21756399e-01 5.23312390e-01 -7.62559414e-01
-1.20964217e+00 -1.38841915e+00 7.20372677e-01 -9.13332105e-01
8.93870771e-01 -4.31090087e-01 -5.20291090e-01 9.04432595e-01
1.26418084e-01 5.46104848e-01 3.44663918e-01 -1.86369300e-01
-1.69872969e-01 -7.08249062e-02 -6.74016416e-01 4.87849355e-01
1.30091906e+00 -6.56432986e-01 -4.58150655e-01 6.54986322e-01
5.82095265e-01 -2.22255841e-01 -6.19282067e-01 4.04872239e-01
4.47875440e-01 -9.75825310e-01 9.63867903e-01 -5.32140255e-01
7.21951008e-01 -6.19016945e-01 -2.08869949e-01 -6.61726058e-01
-1.97162613e-01 -7.35038340e-01 -4.24450368e-01 1.05548406e+00
1.34312257e-01 5.59252081e-03 9.55418229e-01 1.80500284e-01
-2.94453830e-01 -8.36719573e-01 -8.64990830e-01 -8.42641294e-01
-5.74169993e-01 -6.93989635e-01 -8.59355405e-02 6.57403469e-01
7.65442699e-02 2.15695754e-01 -7.26346791e-01 2.50147820e-01
4.73905146e-01 -2.26905793e-01 8.69670451e-01 -4.49298799e-01
-4.41198647e-01 -1.26495302e-01 -6.56800151e-01 -1.51060748e+00
3.79925281e-01 -5.27186036e-01 4.06902403e-01 -1.45085561e+00
2.30815962e-01 4.24001999e-02 -5.38567960e-01 9.98921156e-01
-4.11349945e-02 5.19543111e-01 2.59032678e-02 3.93654108e-01
-1.41816878e+00 5.59628785e-01 1.00208223e+00 -2.63636205e-02
-9.22468752e-02 -3.88567448e-02 -1.46079600e-01 9.33828950e-01
4.08252448e-01 -4.32960510e-01 -7.00847149e-01 -2.92684913e-01
-2.41922542e-01 9.50016081e-02 7.26347685e-01 -1.25726986e+00
3.33996564e-01 -3.39921474e-01 2.24483520e-01 -6.93461001e-01
3.97129953e-01 -7.15264916e-01 -2.54645288e-01 2.39081785e-01
-5.18269598e-01 -2.72067726e-01 -2.08283976e-01 8.22983027e-01
-3.72389048e-01 1.45639196e-01 6.22080863e-01 -6.10062405e-02
-1.15501916e+00 5.12786567e-01 -4.85915005e-01 1.12859666e-01
1.38888812e+00 -4.28953260e-01 5.13645336e-02 -4.18629080e-01
-9.91081476e-01 1.89772114e-01 3.49866241e-01 3.11979681e-01
3.81266505e-01 -1.23282933e+00 -5.60599566e-01 1.27307801e-02
2.64596552e-01 -1.32699043e-01 3.36766422e-01 1.37491405e+00
-2.27683499e-01 4.36262429e-01 9.38320979e-02 -1.01113451e+00
-1.26064265e+00 5.94168961e-01 2.71805704e-01 -1.60254478e-01
-7.82799959e-01 8.80647600e-01 4.71032321e-01 1.62780493e-01
4.40474987e-01 -4.18681979e-01 -2.63803769e-02 -1.59686103e-01
4.68245536e-01 2.31294587e-01 -3.34349006e-01 -9.57788348e-01
-5.17835200e-01 4.37289566e-01 8.82236660e-02 -1.46779060e-01
1.18850756e+00 -1.14193514e-01 2.56750971e-01 3.71378630e-01
1.19745970e+00 -1.91590935e-01 -2.31521988e+00 -3.95741969e-01
2.51013607e-01 -5.29270172e-01 3.57671157e-02 -7.33103335e-01
-1.07069099e+00 7.04895735e-01 2.60710239e-01 -3.35594118e-01
1.10206079e+00 3.57041687e-01 6.64176524e-01 2.20442310e-01
3.91745299e-01 -1.28452265e+00 7.16153145e-01 4.87919629e-01
6.50793433e-01 -1.23075676e+00 -7.47668836e-03 -2.80850887e-01
-8.86164248e-01 8.08262646e-01 7.63206303e-01 -1.54733956e-01
4.69987869e-01 1.66574821e-01 -7.80683681e-02 -7.11757690e-02
-8.75416100e-01 -3.52411628e-01 4.66342986e-01 4.62546974e-01
2.76665092e-01 -3.74500066e-01 4.68603298e-02 4.07725841e-01
5.63628614e-01 1.85614154e-01 6.50881752e-02 1.20003295e+00
-3.61151755e-01 -8.79858375e-01 -3.55115086e-02 1.30990103e-01
-4.49368119e-01 -5.71224000e-03 -2.87165374e-01 5.47514200e-01
4.02093261e-01 9.48660374e-01 -7.25422651e-02 -4.22372222e-01
1.49413288e-01 2.93060001e-02 4.02364045e-01 -7.78035283e-01
-1.33165345e-01 4.62914497e-01 5.48306629e-02 -1.33388972e+00
-8.12988043e-01 -1.01456475e+00 -1.27407622e+00 3.60348016e-01
-2.76054833e-02 -4.87147234e-02 2.03039885e-01 1.31310320e+00
4.95408088e-01 5.61572552e-01 4.16167825e-01 -1.12491953e+00
-3.23599398e-01 -9.49403048e-01 -3.87641162e-01 4.65372920e-01
4.12391156e-01 -7.01586843e-01 -3.13291311e-01 6.00046575e-01] | [8.456406593322754, 0.5996009111404419] |
1087b9fc-55b6-477f-95cf-e086094d3d37 | faircop-facial-image-retrieval-using | 2205.15870 | null | https://arxiv.org/abs/2205.15870v1 | https://arxiv.org/pdf/2205.15870v1.pdf | FaIRCoP: Facial Image Retrieval using Contrastive Personalization | Retrieving facial images from attributes plays a vital role in various systems such as face recognition and suspect identification. Compared to other image retrieval tasks, facial image retrieval is more challenging due to the high subjectivity involved in describing a person's facial features. Existing methods do so by comparing specific characteristics from the user's mental image against the suggested images via high-level supervision such as using natural language. In contrast, we propose a method that uses a relatively simpler form of binary supervision by utilizing the user's feedback to label images as either similar or dissimilar to the target image. Such supervision enables us to exploit the contrastive learning paradigm for encapsulating each user's personalized notion of similarity. For this, we propose a novel loss function optimized online via user feedback. We validate the efficacy of our proposed approach using a carefully designed testbed to simulate user feedback and a large-scale user study. Our experiments demonstrate that our method iteratively improves personalization, leading to faster convergence and enhanced recommendation relevance, thereby, improving user satisfaction. Our proposed framework is also equipped with a user-friendly web interface with a real-time experience for facial image retrieval. | ['Rajiv Ratn Shah', 'Ponnurangam Kumaraguru', 'Rishi Raj Jain', 'Shagun Uppal', 'Sarthak Bhagat', 'Drishti Bhasin', 'Aditya Saini', 'Devansh Gupta'] | 2022-05-28 | null | null | null | null | ['face-image-retrieval'] | ['computer-vision'] | [ 4.26063150e-01 -1.64067224e-01 -4.23525661e-01 -8.72182667e-01
-5.89485765e-01 -3.81371886e-01 5.28445780e-01 3.87553200e-02
-6.10646546e-01 3.40956956e-01 -5.06244553e-03 -6.29421696e-02
-1.63252428e-01 -4.04930383e-01 -3.94557118e-01 -5.77235341e-01
-1.80291943e-02 1.14461601e-01 -1.00678019e-01 -1.42791942e-01
3.33103210e-01 4.35171902e-01 -1.98733759e+00 1.57144755e-01
7.10680187e-01 1.36569798e+00 1.92990333e-01 1.93693757e-01
2.27083102e-01 1.49214134e-01 -3.82567108e-01 -6.33693457e-01
3.70555907e-01 -1.72257438e-01 -6.17760301e-01 5.95441401e-01
7.89668739e-01 -4.72631633e-01 -7.76845589e-02 1.18670070e+00
5.70550442e-01 2.51833320e-01 4.97824997e-01 -1.31699514e+00
-6.74454689e-01 -8.57378766e-02 -6.76004589e-01 1.18952431e-01
6.35689020e-01 8.88874084e-02 1.03675783e+00 -8.56944025e-01
4.36055332e-01 1.38935339e+00 1.26850158e-01 7.96820760e-01
-1.11456692e+00 -8.64436626e-01 3.80169332e-01 3.16286772e-01
-1.66674995e+00 -7.25303411e-01 9.50618982e-01 -9.47051570e-02
2.64580607e-01 3.83570254e-01 1.78641304e-01 7.72752225e-01
-2.84312367e-01 8.97542298e-01 1.09982383e+00 -3.11582714e-01
1.01598710e-01 7.77602792e-01 -8.25302955e-03 8.35801959e-01
-1.46518081e-01 -1.39520928e-01 -6.27041757e-01 -4.24488544e-01
5.42665362e-01 4.06923771e-01 -3.97206575e-01 -3.54061007e-01
-6.85770035e-01 6.82175279e-01 2.65794337e-01 -1.15511656e-01
-4.10185874e-01 -1.58233672e-01 2.18030453e-01 3.38445574e-01
4.67475235e-01 2.90248613e-03 -1.17797703e-01 1.38455495e-01
-7.60958254e-01 1.22948154e-03 5.46601057e-01 8.53455901e-01
8.91726971e-01 -3.35912824e-01 -3.36741775e-01 1.07165909e+00
5.51511824e-01 5.25081694e-01 6.17073894e-01 -9.84417677e-01
9.68400761e-02 3.41746300e-01 3.27323973e-01 -1.26954782e+00
1.42734557e-01 -3.37988347e-01 -6.28508091e-01 1.45366222e-01
1.09589733e-01 1.00146271e-01 -6.79858208e-01 2.02240944e+00
3.85485709e-01 2.78440863e-01 -2.24503741e-01 1.21934092e+00
5.44385314e-01 4.01848942e-01 2.15196580e-01 -4.49841887e-01
1.59043050e+00 -6.17507279e-01 -5.91375470e-01 2.11436287e-01
2.11868480e-01 -6.89050496e-01 1.38130629e+00 4.87379789e-01
-9.13169265e-01 -5.18706262e-01 -7.76809812e-01 1.91389695e-01
-6.65210709e-02 3.31633925e-01 7.16098428e-01 1.06878710e+00
-1.08479130e+00 4.29229975e-01 -3.58688623e-01 -6.90842330e-01
3.96070451e-01 5.56552947e-01 -4.37431723e-01 -3.26715000e-02
-1.13242376e+00 5.12204289e-01 -1.07111171e-01 3.15911621e-02
-6.36553288e-01 -4.06877190e-01 -7.87998080e-01 2.34810799e-01
4.89495248e-01 -5.68794429e-01 1.20935488e+00 -1.38925660e+00
-1.66954517e+00 1.07624626e+00 -4.28474694e-01 -1.31183535e-01
2.33794197e-01 -2.61855811e-01 -4.06342804e-01 3.12903106e-01
5.51815983e-03 6.44240081e-01 1.28448904e+00 -1.29326439e+00
-6.62434340e-01 -6.51498973e-01 3.34037244e-01 3.10643673e-01
-9.36197579e-01 3.08320671e-01 -7.20107257e-01 -5.09053946e-01
-1.48874283e-01 -9.02305007e-01 -7.16057569e-02 3.58553141e-01
5.04330508e-02 -5.16957045e-01 1.05365944e+00 -2.62590945e-01
1.09550667e+00 -2.10589552e+00 -2.50974596e-01 6.44338250e-01
7.56944045e-02 3.54399115e-01 -2.35525534e-01 1.11859396e-01
-5.31728156e-02 5.88998199e-02 9.90186781e-02 -6.02266371e-01
-6.33907467e-02 -2.62630936e-02 -9.83583182e-02 4.17299092e-01
2.06477657e-01 4.36335236e-01 -9.64908659e-01 -5.21877646e-01
-1.52925819e-01 4.86743361e-01 -7.01173484e-01 3.10479552e-01
1.46879375e-01 2.35623479e-01 -6.05366468e-01 7.59268701e-01
6.24986470e-01 -3.88316542e-01 2.42326051e-01 -3.53164285e-01
4.04725313e-01 -1.18407592e-01 -1.20553005e+00 1.56432271e+00
-5.58253288e-01 2.94709772e-01 3.23140144e-01 -9.97223020e-01
9.99928832e-01 3.25004101e-01 4.32377100e-01 -7.80253410e-01
1.21294565e-01 -1.83810756e-01 -4.31763470e-01 -6.20262861e-01
5.64809740e-01 -1.12112919e-02 2.84387171e-01 9.26885843e-01
-2.79332846e-02 3.33746493e-01 -6.97170151e-03 4.01041508e-01
7.62847841e-01 3.20831500e-02 2.13741601e-01 -9.75640193e-02
8.72008443e-01 -8.09123933e-01 5.13523221e-01 5.67813039e-01
-3.56933057e-01 3.75201553e-01 1.63774043e-01 -1.14604682e-01
-4.88439411e-01 -8.93620431e-01 -1.52512804e-01 1.40790761e+00
3.34850401e-01 -5.18565476e-01 -6.55018151e-01 -7.57082403e-01
5.85212037e-02 2.80814081e-01 -6.00931644e-01 -2.23785639e-01
-7.81909674e-02 -4.64311033e-01 2.34609321e-01 1.10968657e-01
5.08393526e-01 -1.03005648e+00 -3.50877792e-01 -3.57155442e-01
-2.76058447e-02 -8.48908007e-01 -8.95020843e-01 -6.12775266e-01
-6.51407778e-01 -7.93131411e-01 -8.68627787e-01 -7.13437438e-01
1.09090054e+00 6.83476985e-01 9.69789326e-01 2.94023544e-01
-2.77480900e-01 9.41250801e-01 -3.62634271e-01 -1.46451041e-01
3.76527049e-02 -2.23991886e-01 4.22997087e-01 7.48783410e-01
4.48874027e-01 -5.00202656e-01 -1.00647795e+00 5.57444572e-01
-1.24600828e+00 -3.03599775e-01 6.03064775e-01 9.33941245e-01
3.75528961e-01 3.00954562e-02 6.31058335e-01 -1.08326292e+00
1.00251257e+00 -5.88344276e-01 -4.36510116e-01 4.77389961e-01
-8.22362304e-01 -1.11081414e-01 4.11620289e-01 -5.13548195e-01
-1.11853111e+00 1.42497763e-01 8.43819678e-02 -5.05613685e-01
-1.45122543e-01 3.40635359e-01 -1.76428035e-01 -2.44012445e-01
4.63469744e-01 1.49683639e-01 2.58614421e-01 -2.85318047e-01
2.79902041e-01 1.02617049e+00 4.91083771e-01 -7.70741463e-01
7.84945309e-01 4.56960469e-01 -2.14363128e-01 -6.68149710e-01
-7.15481043e-01 -6.82477653e-01 -3.50258797e-01 -4.65699106e-01
3.44134450e-01 -7.40761518e-01 -1.32528496e+00 1.37791455e-01
-8.76248956e-01 3.47480983e-01 3.24155718e-01 3.49840313e-01
-3.21139157e-01 6.53220117e-01 -3.42247546e-01 -9.93113220e-01
-5.14634311e-01 -1.14027131e+00 1.22828209e+00 4.35139805e-01
-8.78527462e-02 -7.47204959e-01 -2.90719032e-01 5.25873005e-01
3.78506124e-01 -2.87319958e-01 5.96900761e-01 -3.60947311e-01
-4.18113053e-01 -3.67564946e-01 -4.11816150e-01 3.79769236e-01
5.44886708e-01 -1.26268432e-01 -1.18549192e+00 -4.74439085e-01
-6.39340058e-02 -5.02668977e-01 5.68977296e-01 -7.94539377e-02
1.56441414e+00 -3.31899136e-01 -2.27765873e-01 3.85519654e-01
1.19574368e+00 3.38695161e-02 4.73244876e-01 1.09233066e-01
4.20874208e-01 8.39797854e-01 9.43853319e-01 7.76791751e-01
3.34707946e-01 8.52056086e-01 3.25286984e-01 -1.88876435e-01
4.35669541e-01 -2.29811490e-01 3.17616582e-01 1.44624352e-01
-1.48932680e-01 -1.27413407e-01 -4.74882096e-01 2.79384047e-01
-1.81778622e+00 -9.44737971e-01 6.39492273e-01 2.62142444e+00
8.59566271e-01 -2.07021564e-01 1.97045267e-01 -1.19864553e-01
8.48311961e-01 -6.71099275e-02 -6.48859560e-01 -2.49461293e-01
3.06555361e-01 2.14724496e-01 2.20335886e-01 3.66336107e-01
-9.96699333e-01 8.95350993e-01 5.69382238e+00 9.20839012e-01
-1.36756754e+00 2.48030829e-03 9.78330851e-01 -2.05044687e-01
-4.16611105e-01 -2.93139309e-01 -6.50681853e-01 4.58504647e-01
6.32317245e-01 -3.05947036e-01 4.35435176e-01 7.51655638e-01
4.98964876e-01 -5.14680073e-02 -1.15789723e+00 1.39003980e+00
4.64065492e-01 -9.98082042e-01 2.86990851e-01 -2.79609598e-02
4.17595267e-01 -6.84504688e-01 6.01468742e-01 7.83786476e-02
-1.65473223e-01 -8.62536728e-01 3.93212825e-01 4.76275444e-01
9.08965409e-01 -8.12782466e-01 4.53857481e-01 2.35842466e-01
-8.56796622e-01 -1.95945837e-02 -3.08193922e-01 6.55530915e-02
-1.51899531e-01 2.51694202e-01 -9.57969725e-01 2.91682690e-01
6.57758892e-01 6.18438959e-01 -6.42693758e-01 7.20684826e-01
2.63099913e-02 4.47341383e-01 -1.67288437e-01 3.39020528e-02
-3.92702930e-02 -3.47594112e-01 2.12909833e-01 1.03342140e+00
2.54066825e-01 2.88643748e-01 2.59686649e-01 4.74900544e-01
-2.92085111e-01 6.03199959e-01 -5.56474686e-01 1.19966052e-01
3.34578216e-01 1.51517630e+00 -4.36291516e-01 -2.92952895e-01
-4.29883510e-01 1.17484927e+00 2.08325952e-01 4.58228678e-01
-6.40674233e-01 -2.94468820e-01 7.76715815e-01 1.90015927e-01
3.36571373e-02 2.09688634e-01 2.43199587e-01 -1.06761062e+00
1.10461794e-01 -9.31056678e-01 4.45333928e-01 -7.48923779e-01
-1.33925891e+00 6.19341493e-01 -1.36869878e-01 -1.39282548e+00
-2.02421293e-01 -4.47870791e-01 -4.95563507e-01 8.88777077e-01
-1.61619794e+00 -9.84071791e-01 -3.98964852e-01 9.70883250e-01
3.62520427e-01 -2.91255593e-01 8.01346898e-01 4.93451625e-01
-4.51002270e-01 1.14813542e+00 -2.22105473e-01 -8.42272192e-02
1.30250049e+00 -8.76815438e-01 -2.10686848e-01 5.01856804e-01
1.40085798e-02 1.05161619e+00 6.32124662e-01 -3.04864824e-01
-1.42346549e+00 -8.82599533e-01 6.29329741e-01 -3.84423099e-02
3.96678418e-01 -1.56381577e-01 -7.40532279e-01 3.45794618e-01
1.65448889e-01 1.75132364e-01 1.09203207e+00 2.04702958e-01
-4.95625973e-01 -6.03562772e-01 -1.50517428e+00 7.04902053e-01
1.03515148e+00 -7.06860244e-01 -1.11575253e-01 3.13651979e-01
2.27657810e-01 6.97388873e-02 -7.80520737e-01 3.23188066e-01
7.94648528e-01 -8.19021463e-01 7.45658278e-01 -6.81919038e-01
1.58405602e-01 -3.69154036e-01 -3.29357721e-02 -1.00263536e+00
-2.89703429e-01 -8.48442018e-01 3.13309819e-01 1.35333943e+00
8.25586766e-02 -5.60532987e-01 1.05586851e+00 1.06908000e+00
3.79620731e-01 -6.84381366e-01 -5.24247944e-01 -4.36966985e-01
-6.82246804e-01 -1.67674452e-01 5.65187335e-01 8.51356685e-01
1.71175584e-01 3.29191864e-01 -6.11168623e-01 3.81421953e-01
7.50225782e-01 1.61022529e-01 7.93754756e-01 -1.16762245e+00
-3.83971840e-01 -2.89655864e-01 -5.22296488e-01 -1.22962248e+00
4.32479531e-01 -7.29168713e-01 -4.41866405e-02 -6.98951125e-01
5.96421599e-01 -4.84202772e-01 -7.66869068e-01 3.44567478e-01
-2.39471301e-01 6.16302133e-01 9.88953710e-02 1.78959236e-01
-1.03550112e+00 5.48561633e-01 1.13394797e+00 -2.17657968e-01
-5.16167842e-02 2.53100067e-01 -9.55606878e-01 6.47665083e-01
6.79917395e-01 -1.63741797e-01 -7.74923682e-01 -2.93236047e-01
9.31620318e-03 1.01205729e-01 2.28165507e-01 -5.68310499e-01
2.71067381e-01 -1.65922493e-01 1.81207344e-01 -1.15381092e-01
4.71009165e-01 -8.94760311e-01 -3.82831186e-01 3.75477038e-02
-5.26505470e-01 -4.19963663e-03 -4.41725850e-02 6.63196504e-01
-3.99845064e-01 -1.64813459e-01 6.38355613e-01 -5.40026575e-02
-8.02021086e-01 6.39672995e-01 -1.94015458e-01 -4.65043575e-01
8.83154273e-01 -2.34067366e-01 -5.31161111e-03 -9.97097790e-01
-7.37237573e-01 2.91110188e-01 4.57852036e-01 4.66704279e-01
8.92963588e-01 -1.35287631e+00 -5.17361701e-01 3.20740521e-01
4.69346821e-01 -6.91926003e-01 1.20690294e-01 5.78507423e-01
1.05456948e-01 2.74674833e-01 -2.26536438e-01 -6.76862061e-01
-1.89091039e+00 4.50665236e-01 6.22694530e-02 9.13474634e-02
-7.41736293e-02 7.59464562e-01 4.10738140e-01 -1.14558645e-01
4.07435298e-01 3.56434047e-01 -4.19539005e-01 2.42450424e-02
7.72068739e-01 -8.89389515e-02 -5.45221418e-02 -7.59345412e-01
-2.38010541e-01 4.86164510e-01 -4.11957562e-01 -4.09005843e-02
1.03455389e+00 -4.38163280e-01 1.93046108e-02 1.78360213e-02
1.25331724e+00 1.19144924e-01 -1.07951152e+00 -6.18835092e-01
-9.28378999e-02 -9.36707497e-01 1.27492677e-02 -5.03069043e-01
-1.19431973e+00 5.78972280e-01 1.09188831e+00 -7.81133771e-02
1.49089742e+00 -1.37371525e-01 6.32470608e-01 6.28441572e-01
6.38766587e-01 -1.08099163e+00 2.83164591e-01 -7.53654167e-02
6.71140015e-01 -1.72453094e+00 5.29985549e-03 -4.70247537e-01
-7.61713564e-01 7.61885941e-01 5.62734365e-01 4.12151590e-02
7.72490501e-01 -3.74556094e-01 3.38131368e-01 -3.31981741e-02
-8.62363756e-01 -3.05783153e-01 4.73737985e-01 4.18026924e-01
5.93394816e-01 -7.25582093e-02 -3.77561480e-01 4.37103212e-01
1.61459178e-01 4.00900207e-02 1.75172761e-01 7.47639239e-01
-3.15327644e-01 -1.35400259e+00 -4.17114735e-01 4.84777480e-01
-5.23494363e-01 8.50492641e-02 -1.28912330e-01 1.89156845e-01
-1.85119897e-01 1.09674525e+00 -7.98744615e-03 -4.17281061e-01
1.10862598e-01 -7.73507506e-02 4.77158308e-01 -7.74592400e-01
-2.90883213e-01 6.68713748e-02 -2.51716733e-01 -6.89226210e-01
-6.77621841e-01 -6.28028631e-01 -7.59175301e-01 -1.30033821e-01
-1.61176816e-01 3.18252623e-01 6.51627600e-01 7.99279451e-01
4.69375432e-01 -1.51790723e-01 1.20461953e+00 -8.06878686e-01
-5.38351119e-01 -7.05772996e-01 -6.16133451e-01 1.03176427e+00
9.34497267e-02 -7.92493999e-01 -1.75395429e-01 2.68709660e-01] | [13.2274169921875, 0.7681884169578552] |
cb0219f0-4510-4916-b8fa-b8dd391bc716 | document-level-entity-based-extraction-as | 2109.04901 | null | https://arxiv.org/abs/2109.04901v1 | https://arxiv.org/pdf/2109.04901v1.pdf | Document-level Entity-based Extraction as Template Generation | Document-level entity-based extraction (EE), aiming at extracting entity-centric information such as entity roles and entity relations, is key to automatic knowledge acquisition from text corpora for various domains. Most document-level EE systems build extractive models, which struggle to model long-term dependencies among entities at the document level. To address this issue, we propose a generative framework for two document-level EE tasks: role-filler entity extraction (REE) and relation extraction (RE). We first formulate them as a template generation problem, allowing models to efficiently capture cross-entity dependencies, exploit label semantics, and avoid the exponential computation complexity of identifying N-ary relations. A novel cross-attention guided copy mechanism, TopK Copy, is incorporated into a pre-trained sequence-to-sequence model to enhance the capabilities of identifying key information in the input document. Experiments done on the MUC-4 and SciREX dataset show new state-of-the-art results on REE (+3.26%), binary RE (+4.8%), and 4-ary RE (+2.7%) in F1 score. | ['Nanyun Peng', 'Sam Tang', 'Kung-Hsiang Huang'] | 2021-09-10 | null | https://aclanthology.org/2021.emnlp-main.426 | https://aclanthology.org/2021.emnlp-main.426.pdf | emnlp-2021-11 | ['role-filler-entity-extraction', '4-ary-relation-extraction', 'binary-relation-extraction'] | ['natural-language-processing', 'natural-language-processing', 'natural-language-processing'] | [ 5.13342798e-01 4.05218303e-01 -4.80250478e-01 -2.71164805e-01
-1.04893196e+00 -8.68025243e-01 6.77262604e-01 3.99436802e-01
-6.77957654e-01 8.70643675e-01 2.54918993e-01 -3.55676115e-01
-5.14362752e-02 -7.53476918e-01 -8.08987260e-01 -2.00636268e-01
-6.78247586e-02 6.78763390e-01 3.12957197e-01 -2.08615944e-01
3.07968594e-02 2.94127375e-01 -1.25670278e+00 5.91601908e-01
1.12822270e+00 7.39899576e-01 1.29410192e-01 7.42208898e-01
-3.84739697e-01 8.79308879e-01 -8.01105559e-01 -8.65936279e-01
-2.48682946e-01 -2.68335342e-01 -1.09903347e+00 -1.00913160e-01
-6.11415766e-02 9.69659835e-02 -1.38270840e-01 7.53841817e-01
3.09063822e-01 7.69561306e-02 8.79189491e-01 -1.05548155e+00
-6.84367657e-01 1.07876658e+00 -5.38351238e-01 1.60507485e-01
2.63017803e-01 -4.80218291e-01 1.52096868e+00 -9.90715683e-01
1.12703931e+00 1.02531898e+00 1.63029000e-01 5.19126296e-01
-1.07270193e+00 -5.89200795e-01 2.29773194e-01 3.55638772e-01
-1.41308939e+00 -4.30029750e-01 4.72850591e-01 -3.27647150e-01
1.70820248e+00 4.10090327e-01 1.66675419e-01 1.04833007e+00
-2.56763548e-02 1.18907940e+00 7.19457567e-01 -9.21972692e-01
-2.62763172e-01 1.13721378e-01 3.49706262e-01 4.45394516e-01
3.67543161e-01 -2.68064618e-01 -6.40259266e-01 1.05818912e-01
3.66292626e-01 -5.35237551e-01 -1.64826915e-01 1.89449981e-01
-9.21131611e-01 5.83490252e-01 1.25947729e-01 5.78222454e-01
-4.59207356e-01 -2.56117493e-01 4.32645261e-01 7.42881419e-03
4.22543019e-01 7.69334197e-01 -9.03579354e-01 -9.16864425e-02
-7.36035645e-01 1.43391311e-01 9.14159834e-01 1.46932375e+00
4.49295431e-01 -3.22537541e-01 -6.42102957e-01 1.01765692e+00
1.64353713e-01 4.04263079e-01 5.29550493e-01 -2.96473503e-01
8.08379471e-01 8.85209262e-01 -5.36622033e-02 -6.36180639e-01
-3.25785160e-01 -8.76473963e-01 -6.39176369e-01 -7.05380976e-01
-1.82794407e-01 -2.52231508e-01 -1.14617479e+00 1.74843919e+00
4.09069091e-01 -4.40397561e-02 4.11426336e-01 2.50777245e-01
9.88849521e-01 8.49827111e-01 4.49258834e-01 -3.09055895e-01
1.70249450e+00 -1.20893419e+00 -9.36371446e-01 -5.41887939e-01
9.56593096e-01 -6.90901101e-01 6.77447379e-01 -8.09930041e-02
-1.07673049e+00 -5.19964755e-01 -1.00763559e+00 -3.60067755e-01
-7.16282308e-01 6.88192129e-01 4.72350627e-01 3.40076625e-01
-5.54408491e-01 9.62153152e-02 -6.44273579e-01 -7.91087076e-02
1.10140055e-01 4.03142035e-01 -4.48534667e-01 -1.96541939e-02
-1.57422495e+00 8.71944487e-01 8.88690531e-01 1.62191942e-01
-5.04926324e-01 -8.14780235e-01 -1.05296910e+00 3.44623446e-01
9.15907800e-01 -6.11078143e-01 1.24145794e+00 -3.53171915e-01
-1.01744485e+00 8.31345856e-01 -4.70361501e-01 -4.03334051e-01
-4.29445989e-02 -6.91375256e-01 -8.22670579e-01 -3.63924690e-02
2.81513900e-01 5.27681768e-01 4.43189889e-01 -1.30168962e+00
-8.86515319e-01 -2.96134591e-01 -4.43222709e-02 3.30954045e-01
-2.75806457e-01 4.48437750e-01 -8.35286558e-01 -8.76253009e-01
-2.01324150e-01 -1.05907369e+00 8.15477297e-02 -9.98767614e-01
-7.99126685e-01 -6.30673289e-01 6.09370172e-01 -1.01381052e+00
1.70982409e+00 -1.98637629e+00 2.17484549e-01 1.75246716e-01
3.94710675e-02 4.90888953e-01 -1.17873639e-01 5.47988594e-01
-1.63260773e-01 4.41336393e-01 -3.03347915e-01 -3.70782137e-01
9.54867154e-02 2.15414882e-01 -1.83787107e-01 -4.91802990e-01
7.61924028e-01 1.25738704e+00 -7.95839965e-01 -5.54473639e-01
-3.90040517e-01 2.55144119e-01 -2.82243758e-01 4.28021908e-01
-4.54182744e-01 3.70617323e-02 -4.14158374e-01 6.69127822e-01
4.59659904e-01 -2.98500180e-01 6.95227981e-01 -2.76355267e-01
1.35636732e-01 7.82988727e-01 -1.13477564e+00 1.36989617e+00
-7.32692182e-01 2.87408859e-01 -1.63346991e-01 -5.33032238e-01
6.55009091e-01 5.17395854e-01 1.97945580e-01 -6.64703965e-01
1.85275469e-02 2.80166119e-01 6.87860548e-02 -4.27938432e-01
9.73382235e-01 9.27766711e-02 -4.31490928e-01 2.11431801e-01
4.15700763e-01 3.20285320e-01 8.05990338e-01 5.00714481e-01
1.18831074e+00 8.35634097e-02 5.65426290e-01 1.34239823e-01
7.11967230e-01 5.77623099e-02 5.53711176e-01 4.99443740e-01
7.13548005e-01 2.30762213e-01 5.51284015e-01 1.85852200e-01
-7.36468732e-01 -5.52771866e-01 6.47508800e-02 1.22993028e+00
-4.29033935e-02 -9.89633322e-01 -5.45425713e-01 -1.24140429e+00
-1.06697425e-01 1.00313866e+00 -5.86306036e-01 -2.53517777e-01
-1.02410293e+00 -7.29074597e-01 6.39983356e-01 8.63659680e-01
2.51444608e-01 -1.14106202e+00 -1.38993591e-01 5.92967868e-01
-4.60537910e-01 -1.66144443e+00 -4.09967661e-01 6.32068515e-01
-4.92082208e-01 -9.40909386e-01 -4.75389183e-01 -7.98600316e-01
6.69151723e-01 -3.66323471e-01 1.47887039e+00 -7.65121263e-03
4.27573733e-02 -3.78455631e-02 -8.14477563e-01 -3.78777504e-01
-4.30074602e-01 7.78423846e-01 -3.84246647e-01 -2.22578734e-01
6.11330628e-01 -2.60117501e-01 -1.81977917e-02 1.79070771e-01
-9.58021700e-01 6.87552840e-02 1.16929245e+00 9.82830465e-01
8.85918379e-01 1.21559553e-01 5.21984041e-01 -1.49088645e+00
6.73835218e-01 -4.09160107e-01 -3.09497952e-01 7.65186727e-01
-7.12634921e-01 4.14313644e-01 6.51402652e-01 -4.21183884e-01
-1.41721499e+00 6.87189177e-02 -2.49237299e-01 6.66465908e-02
-1.56168431e-01 9.09451127e-01 -5.54268837e-01 6.43703938e-01
4.05005991e-01 3.81781399e-01 -8.72247994e-01 -5.72904170e-01
6.38826132e-01 7.88056552e-01 7.78761744e-01 -8.85496259e-01
7.11376488e-01 -3.03853989e-01 -1.66988909e-01 -5.99403739e-01
-1.26881826e+00 -7.60040998e-01 -7.79228508e-01 3.93871933e-01
9.03107405e-01 -1.07832623e+00 -2.75091588e-01 2.60287672e-01
-1.42024255e+00 -6.51788041e-02 -3.13777477e-01 2.66216874e-01
-4.58642989e-02 1.67300757e-02 -7.42867410e-01 -6.16483271e-01
-6.30052865e-01 -7.57523060e-01 1.30567002e+00 3.06938976e-01
-3.52237880e-01 -6.85963035e-01 1.15571348e-02 4.02599216e-01
-4.38011736e-02 -7.67665058e-02 1.44073272e+00 -1.10151446e+00
-6.01366103e-01 -1.72367856e-01 -2.69456267e-01 2.59979367e-01
4.25650589e-02 -1.83169141e-01 -7.06409752e-01 6.20450415e-02
-4.51801896e-01 -2.66811728e-01 8.10498834e-01 -1.51579872e-01
8.54992330e-01 -4.60096151e-01 -6.98050380e-01 1.85084671e-01
1.06359839e+00 4.60559487e-01 5.74427366e-01 1.96554974e-01
8.66378903e-01 6.04606390e-01 8.00716817e-01 1.68353200e-01
6.02265716e-01 8.79995704e-01 -6.03815950e-02 -2.34770849e-02
-4.61613536e-01 -5.69123149e-01 1.64769053e-01 9.71102417e-01
8.76404420e-02 -7.03565180e-01 -8.52238894e-01 6.92436278e-01
-1.74014223e+00 -5.81869006e-01 -1.75814152e-01 1.67203200e+00
1.38806307e+00 3.64427269e-01 -1.33405760e-01 1.59609079e-01
7.21573174e-01 -1.14565179e-01 -2.91525543e-01 -2.23679423e-01
-1.17830507e-01 4.12379086e-01 3.61822277e-01 3.42594206e-01
-1.26176727e+00 1.41553974e+00 4.96819401e+00 1.07577491e+00
-7.22380698e-01 -6.70259492e-03 3.70671153e-01 3.44068795e-01
-3.81875336e-01 8.13082885e-03 -1.67196310e+00 3.01460832e-01
1.14262295e+00 -1.55017838e-01 -2.56430477e-01 8.02759767e-01
-4.72300828e-01 5.90375029e-02 -1.24028111e+00 6.31580651e-01
3.00756954e-02 -1.17402899e+00 2.63490200e-01 4.05759998e-02
6.71376407e-01 -2.47795343e-01 -3.32452118e-01 8.14375997e-01
3.28015089e-01 -9.32471037e-01 6.94361627e-01 2.20282361e-01
9.96958911e-01 -9.04633164e-01 1.01116920e+00 3.45816106e-01
-1.48359764e+00 2.36351207e-01 -1.57275405e-02 4.92714047e-01
4.27475303e-01 6.87743485e-01 -1.02424634e+00 1.05121434e+00
3.72838825e-01 5.80092609e-01 -5.68249583e-01 6.25479281e-01
-8.60937059e-01 5.92272520e-01 -2.41661936e-01 -1.97974280e-01
1.91968083e-01 2.82132685e-01 5.27775884e-01 1.76726687e+00
6.71940297e-02 2.39685833e-01 -9.88582000e-02 5.54637790e-01
-6.88224554e-01 2.55936086e-01 -9.26963538e-02 -4.14259642e-01
6.56702876e-01 1.35182858e+00 -7.05370963e-01 -4.70127314e-01
-3.65538120e-01 1.04824936e+00 5.49462855e-01 1.70256510e-01
-6.61284506e-01 -8.34691048e-01 2.73882091e-01 -1.73431143e-01
7.24299729e-01 -2.79416949e-01 1.10039534e-03 -1.18265069e+00
3.36783648e-01 -1.04399216e+00 6.48752511e-01 -5.38524508e-01
-1.02439332e+00 8.59146297e-01 1.03015758e-01 -8.43383789e-01
-6.41617358e-01 -5.27701259e-01 -5.33995666e-02 7.73680151e-01
-1.63848901e+00 -1.46837282e+00 1.49104342e-01 2.49855205e-01
6.81304157e-01 9.17894840e-02 9.27751124e-01 6.15874887e-01
-6.92577362e-01 9.54012632e-01 -1.29698843e-01 5.66408575e-01
6.02458060e-01 -1.45422912e+00 5.43227136e-01 1.14121985e+00
7.00994134e-01 8.13638926e-01 2.30980590e-01 -9.00420249e-01
-1.13008690e+00 -1.25574374e+00 1.70692670e+00 -5.01257956e-01
4.87821430e-01 -5.63344240e-01 -9.30455327e-01 8.62228334e-01
1.12605728e-01 -7.79452100e-02 8.53047550e-01 6.52066529e-01
-5.17932713e-01 2.02343777e-01 -7.73276389e-01 6.05322778e-01
1.18788660e+00 -5.94622254e-01 -7.48067081e-01 4.78309393e-02
1.02222347e+00 -6.44307077e-01 -1.04052794e+00 6.99772358e-01
4.20356393e-01 -1.83602303e-01 8.52644503e-01 -8.10768485e-01
5.60332775e-01 -1.87981829e-01 1.93180016e-03 -1.23672736e+00
-2.85536915e-01 -6.41637266e-01 -8.44842970e-01 1.77628148e+00
1.23757350e+00 -2.35705711e-02 5.42391777e-01 5.18996835e-01
-1.98717877e-01 -9.12380278e-01 -5.29324830e-01 -9.79327083e-01
-2.28569448e-01 -3.51292819e-01 5.77068627e-01 9.31368828e-01
-2.15418171e-02 1.24525726e+00 -1.87176868e-01 3.09011519e-01
3.79418358e-02 2.76627064e-01 4.76805508e-01 -1.04877317e+00
-4.49621737e-01 -1.72430083e-01 5.59930364e-03 -1.22009730e+00
4.70644146e-01 -1.11429989e+00 7.41956383e-02 -1.66254461e+00
2.64582962e-01 -4.81086701e-01 -3.20589662e-01 7.58078933e-01
-5.60243547e-01 -2.29804814e-01 1.23022966e-01 6.71214983e-02
-6.61687434e-01 5.95845520e-01 9.44134891e-01 -1.04307510e-01
-3.37190598e-01 -5.61090633e-02 -8.62913728e-01 4.41759944e-01
4.09533590e-01 -8.91345084e-01 -4.06465411e-01 -5.49005926e-01
2.86070436e-01 4.46468475e-04 -3.30128431e-01 -5.94062567e-01
3.51119280e-01 2.06490159e-02 3.12379569e-01 -7.58957744e-01
2.20107898e-01 -7.40930855e-01 -1.43097863e-01 1.00416504e-01
-4.72956687e-01 4.44524288e-02 2.88477302e-01 3.24200809e-01
-4.84845459e-01 -5.46642303e-01 2.59077489e-01 2.19624992e-02
-7.95728862e-01 1.08747587e-01 -6.13701344e-02 5.24569273e-01
8.52097273e-01 2.50180960e-01 -4.63349998e-01 -2.38670874e-02
-7.22150624e-01 3.31408232e-01 -3.35877508e-01 6.61726952e-01
4.08954203e-01 -1.05908597e+00 -8.51436555e-01 1.59900680e-01
4.51761246e-01 2.80747920e-01 1.02493152e-01 4.24082577e-01
2.35218313e-02 6.94472790e-01 3.32585692e-01 -1.13376610e-01
-1.54852629e+00 3.77148330e-01 -1.62852839e-01 -1.22619629e+00
-2.84109771e-01 1.24609208e+00 -1.03066310e-01 -3.76764476e-01
2.86930911e-02 -2.57015675e-01 -4.33008760e-01 2.68237025e-01
4.47118580e-01 7.45062307e-02 5.17639041e-01 -5.87905228e-01
-5.00945270e-01 2.18526974e-01 -7.60303080e-01 -1.92164525e-01
1.32234931e+00 3.01372167e-02 -1.46855235e-01 -1.15541339e-01
1.03924584e+00 4.11642373e-01 -6.48392797e-01 -3.80672663e-01
7.60297954e-01 3.75414491e-02 -1.48712188e-01 -1.28514230e+00
-6.64881468e-01 6.06413245e-01 -1.62142977e-01 1.11510657e-01
1.03687823e+00 4.55684870e-01 1.09189427e+00 4.77432936e-01
3.16293761e-02 -1.04136229e+00 -1.03223979e-01 8.81207168e-01
8.45616996e-01 -9.28300440e-01 -1.20892920e-01 -9.96092498e-01
-7.49694824e-01 7.13330626e-01 8.02095294e-01 4.44063544e-01
4.68011230e-01 5.39953172e-01 -3.07542890e-01 -5.73236905e-02
-9.19683695e-01 -5.47363520e-01 7.63608396e-01 3.52627248e-01
7.65874684e-01 1.09150102e-02 -5.54384112e-01 1.13005209e+00
-1.19034626e-01 -3.03000987e-01 2.09765896e-01 1.17200327e+00
-1.28368944e-01 -1.72004449e+00 3.23660046e-01 3.57506603e-01
-8.64578426e-01 -6.15111709e-01 -6.06951475e-01 8.14672410e-01
2.13824496e-01 8.73605609e-01 -1.79059640e-01 -3.28158289e-01
5.85658729e-01 4.29196656e-01 4.27475095e-01 -1.05797029e+00
-6.18590832e-01 1.70247972e-01 8.50977540e-01 -1.50792465e-01
-4.46638465e-01 -3.72738928e-01 -1.35284495e+00 3.89583319e-01
-6.52233183e-01 5.72148561e-01 5.21614790e-01 1.03369057e+00
6.08029783e-01 8.98462296e-01 3.22152108e-01 -5.92836440e-02
-1.93244517e-01 -1.10752416e+00 -3.61603409e-01 3.73052031e-01
-1.07673727e-01 -5.35913765e-01 4.44177948e-02 2.00088769e-01] | [9.363066673278809, 8.75308895111084] |
99ca78db-a4de-4c4a-93b3-f7e31c3e8543 | from-community-to-role-based-graph-embeddings | 1908.08572 | null | https://arxiv.org/abs/1908.08572v2 | https://arxiv.org/pdf/1908.08572v2.pdf | On Proximity and Structural Role-based Embeddings in Networks: Misconceptions, Techniques, and Applications | Structural roles define sets of structurally similar nodes that are more similar to nodes inside the set than outside, whereas communities define sets of nodes with more connections inside the set than outside. Roles based on structural similarity and communities based on proximity are fundamentally different but important complementary notions. Recently, the notion of structural roles has become increasingly important and has gained a lot of attention due to the proliferation of work on learning representations (node/edge embeddings) from graphs that preserve the notion of roles. Unfortunately, recent work has sometimes confused the notion of structural roles and communities (based on proximity) leading to misleading or incorrect claims about the capabilities of network embedding methods. As such, this paper seeks to clarify the misconceptions and key differences between structural roles and communities, and formalize the general mechanisms (e.g., random walks, feature diffusion) that give rise to community or role-based structural embeddings. We theoretically prove that embedding methods based on these mechanisms result in either community or role-based structural embeddings. These mechanisms are typically easy to identify and can help researchers quickly determine whether a method preserves community or role-based embeddings. Furthermore, they also serve as a basis for developing new and improved methods for community or role-based structural embeddings. Finally, we analyze and discuss applications and data characteristics where community or role-based embeddings are most appropriate. | ['Sungchul Kim', 'Ryan A. Rossi', 'John Boaz Lee', 'Danai Koutra', 'Nesreen K. Ahmed', 'Di Jin'] | 2019-08-22 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [ 4.45531309e-02 2.96959370e-01 -2.62986988e-01 -1.14737175e-01
2.91614830e-01 -1.03322041e+00 9.15932298e-01 1.05722761e+00
-1.20484725e-01 3.71597618e-01 6.43380404e-01 -4.43821758e-01
-5.48809171e-01 -1.22144794e+00 -1.81829631e-01 -7.64171660e-01
-5.64592898e-01 3.70363116e-01 2.28830129e-01 -3.03212076e-01
3.25338483e-01 4.99915361e-01 -1.48107588e+00 -2.10803390e-01
5.26270270e-01 2.73545772e-01 -2.41255417e-01 4.83068258e-01
-2.63033539e-01 8.23133051e-01 -4.92650211e-01 -2.19396845e-01
1.03030793e-01 -4.56476957e-01 -7.77833164e-01 -1.73746288e-01
3.77131663e-02 1.32516056e-01 -5.51716268e-01 9.46453154e-01
2.42069900e-01 6.21126480e-02 6.67175710e-01 -1.77651381e+00
-1.40104425e+00 8.35244238e-01 -5.02492785e-01 2.73525804e-01
5.05591452e-01 -4.70924079e-01 1.61404073e+00 -6.72514498e-01
6.21501327e-01 1.37436080e+00 9.45155740e-01 4.76039827e-01
-1.48880494e+00 -3.28354061e-01 2.50015795e-01 1.61664814e-01
-1.31007874e+00 -1.27823189e-01 7.93072701e-01 -6.22381091e-01
5.38809121e-01 3.43934953e-01 8.50698650e-01 7.43419409e-01
-5.46203852e-02 3.83702338e-01 8.67685258e-01 -5.13270438e-01
6.93619922e-02 3.29887480e-01 4.04394895e-01 5.90084314e-01
1.01972747e+00 -2.02190831e-01 -1.25643179e-01 -7.55375504e-01
9.27190006e-01 3.79143327e-01 -4.40022498e-01 -9.00902510e-01
-1.42344272e+00 1.28357470e+00 6.50701642e-01 9.31502342e-01
3.81705090e-02 4.32748020e-01 3.90307248e-01 5.60002983e-01
5.48979938e-01 8.82902801e-01 -1.34980515e-01 2.75259763e-02
-4.42350894e-01 1.42709659e-02 9.07447100e-01 4.14832383e-01
7.71198094e-01 -2.87998527e-01 4.98501033e-01 6.42754793e-01
4.95832711e-01 -6.86793327e-02 2.53068268e-01 -7.10975230e-01
-1.44460246e-01 1.01762855e+00 -3.35661732e-02 -1.69547331e+00
-1.62366822e-01 -1.56603768e-01 -9.26211059e-01 -1.20304555e-01
2.99390227e-01 3.19186568e-01 -2.67522186e-01 1.85283720e+00
2.27686659e-01 6.54306337e-02 -1.66094005e-01 5.55708230e-01
6.39527023e-01 4.04266626e-01 -2.01246843e-01 2.18631942e-02
1.19244170e+00 -6.86514914e-01 -5.52966833e-01 2.45598793e-01
6.97909236e-01 -4.17939603e-01 8.13098013e-01 -4.94746804e-01
-7.88133144e-01 -3.10933590e-01 -1.06419182e+00 1.73384726e-01
-7.32211292e-01 -6.44697726e-01 8.72036219e-01 8.67968202e-01
-1.67129183e+00 7.30617940e-01 -6.37409508e-01 -8.39002430e-01
1.58449516e-01 2.56581247e-01 -6.16079986e-01 3.21146660e-03
-1.22694421e+00 6.75350189e-01 1.56495273e-02 -2.31037006e-01
-4.75980759e-01 -6.48746610e-01 -1.05911112e+00 2.27301091e-01
1.35331288e-01 -6.97470844e-01 3.43890548e-01 -8.81302059e-01
-4.18972135e-01 9.08589721e-01 -6.02287985e-02 -3.89233649e-01
4.97425087e-02 4.07132804e-01 -3.37951392e-01 3.51009846e-01
4.00581658e-01 4.69969928e-01 5.44698000e-01 -1.38563836e+00
-2.62222737e-01 -3.79701406e-01 4.54735160e-01 8.42811093e-02
-9.89017069e-01 1.11266933e-01 1.90290466e-01 -6.74557030e-01
3.37101012e-01 -9.40197468e-01 -2.16631979e-01 4.19823140e-01
-2.16762379e-01 -6.88220501e-01 7.76596487e-01 -1.34916916e-01
1.41121709e+00 -2.21434498e+00 1.85428828e-01 4.03676808e-01
1.13508022e+00 -2.25573584e-01 -2.68454403e-01 1.08858562e+00
-4.24801648e-01 1.20000362e+00 -2.47800574e-01 4.28119078e-02
-3.46305817e-02 3.48669559e-01 1.78728849e-02 8.32388699e-01
1.11126058e-01 7.21086681e-01 -1.25878704e+00 -3.06419492e-01
7.15191709e-03 6.17257893e-01 -4.94651139e-01 -2.40352735e-01
5.56538045e-01 1.13298841e-01 -2.97432423e-01 5.07314146e-01
3.26514184e-01 -5.38019300e-01 5.40519476e-01 2.11589247e-01
-6.07801080e-02 2.75537133e-01 -1.09474599e+00 8.13095689e-01
-3.19698639e-02 8.87181818e-01 3.25625718e-01 -1.25240779e+00
8.77547503e-01 3.77486885e-01 8.88436377e-01 5.21090180e-02
-2.16927946e-01 1.32222315e-02 4.38082546e-01 3.41197625e-02
5.01139045e-01 -2.60713011e-01 1.20914821e-02 1.19875777e+00
-3.96238297e-01 2.36828402e-01 1.31337509e-01 6.10278010e-01
1.59892333e+00 -7.14719713e-01 5.94999194e-01 -5.14108956e-01
4.08204913e-01 -1.38202801e-01 3.99877518e-01 5.95433712e-01
-5.26607931e-01 4.90637422e-01 8.32956016e-01 -3.62634420e-01
-1.23619366e+00 -1.27269995e+00 -3.02350800e-02 1.09413683e+00
4.55889821e-01 -7.82176614e-01 -1.72277257e-01 -5.90387881e-01
5.67720532e-01 -4.31379229e-02 -1.13478208e+00 -3.16041827e-01
-4.51336592e-01 -5.23133397e-01 5.88321567e-01 7.01123357e-01
-1.16907269e-01 -7.55919755e-01 -1.14646025e-01 8.83224234e-02
2.88168117e-02 -5.28856993e-01 -4.76436824e-01 4.06089798e-02
-1.08881545e+00 -1.48795581e+00 -5.18601596e-01 -8.98455381e-01
1.07263994e+00 8.36602449e-01 1.15829349e+00 6.31175160e-01
-1.21494561e-01 8.05362463e-01 -7.87841856e-01 7.35557526e-02
-3.48499805e-01 -8.06605592e-02 4.05780792e-01 9.82841179e-02
1.69080198e-01 -9.94726419e-01 -5.30671835e-01 4.37433898e-01
-1.03458941e+00 -7.18921125e-01 3.33182633e-01 7.74527490e-01
1.40950143e-01 2.40905121e-01 5.73221684e-01 -8.38446796e-01
1.19944274e+00 -9.34371531e-01 1.06848925e-01 1.89956278e-01
-9.37868476e-01 1.35109365e-01 3.78510863e-01 -4.68188584e-01
-3.72426987e-01 -6.70658886e-01 3.26888144e-01 -2.72526115e-01
2.20710933e-01 7.32470274e-01 -6.10022880e-02 -1.56033620e-01
6.71176195e-01 1.80327326e-01 3.08773696e-01 -3.74069273e-01
5.92947423e-01 3.85326058e-01 -2.37275571e-01 -5.65702438e-01
1.14921224e+00 8.57066214e-01 -1.83443442e-01 -9.38833475e-01
-1.79303437e-01 -6.85930073e-01 -7.13499844e-01 1.23735264e-01
6.09189332e-01 -6.40546143e-01 -3.30866247e-01 -1.18448548e-01
-9.72244561e-01 2.75306910e-01 -4.53692108e-01 1.62373647e-01
-1.00648560e-01 7.48159170e-01 -5.55135787e-01 -7.60611236e-01
4.31098603e-02 -5.12683272e-01 4.27358031e-01 -3.50633003e-02
-7.06218600e-01 -1.66535187e+00 3.10476333e-01 9.74862650e-02
5.23244381e-01 5.27665794e-01 1.31070232e+00 -9.32195902e-01
-3.94903600e-01 -2.84878165e-01 -3.42019230e-01 1.77370280e-01
4.89220768e-01 1.45773962e-01 -5.06366670e-01 -5.11847258e-01
-3.44542176e-01 1.28070056e-01 7.16641009e-01 2.92810231e-01
6.20754302e-01 -4.00791228e-01 -6.82476401e-01 1.35959432e-01
1.40083516e+00 1.07131295e-01 4.56218749e-01 1.60906166e-01
7.52225280e-01 9.28417921e-01 3.05790175e-02 2.50428438e-01
3.61449927e-01 2.36172721e-01 4.00133073e-01 1.00420684e-01
1.23660393e-01 -4.74688321e-01 2.93245316e-01 9.65357721e-01
-1.79658189e-01 -6.63385838e-02 -9.10756946e-01 1.06859541e+00
-1.67639387e+00 -1.21035647e+00 -4.53443825e-01 2.11708307e+00
6.88846052e-01 7.17701018e-02 3.86987001e-01 4.15778846e-01
1.17978930e+00 4.91184473e-01 -1.75438359e-01 -2.74741262e-01
-1.68759227e-01 9.01214033e-02 3.00081879e-01 4.53781068e-01
-8.71386707e-01 5.82698047e-01 6.75813341e+00 3.09596598e-01
-7.42408276e-01 2.82466710e-01 3.65392029e-01 3.29440773e-01
-9.03489709e-01 3.47904652e-01 -2.61443257e-01 1.72190383e-01
6.10658824e-01 -6.17647707e-01 7.75710121e-02 9.53601778e-01
8.71997923e-02 1.82358384e-01 -1.38237453e+00 7.73810923e-01
-2.50034872e-02 -1.47116911e+00 -4.58705035e-05 5.29404879e-01
6.18671358e-01 -1.88133523e-01 -3.13659571e-02 -5.14893904e-02
6.65235877e-01 -1.31119096e+00 5.07001042e-01 2.02724244e-02
7.97256291e-01 -5.06621957e-01 6.98865354e-01 3.84444967e-02
-1.69920254e+00 -1.50709122e-01 -5.01756072e-01 -6.66905046e-01
-1.13332644e-01 7.15567589e-01 -7.78077483e-01 3.60298872e-01
6.24724329e-01 1.22005916e+00 -5.55194855e-01 8.64804745e-01
-3.13243300e-01 6.46369100e-01 -6.05024360e-02 -1.97093979e-01
1.53124794e-01 -3.06410611e-01 6.92342997e-01 6.70665324e-01
1.10134132e-01 -2.67234892e-01 6.98521510e-02 1.17126191e+00
-6.42554611e-02 -1.27286121e-01 -1.22952592e+00 -8.11273038e-01
1.11739695e+00 1.28264034e+00 -1.16817868e+00 -2.85224281e-02
-4.03485894e-01 7.23715544e-01 1.29544124e-01 1.35584146e-01
-6.01060152e-01 -4.99619484e-01 1.23059857e+00 9.17757511e-01
-3.55610140e-02 -4.70302075e-01 -2.05463842e-01 -9.92384791e-01
-1.27174720e-01 -5.06709099e-01 3.54247421e-01 -2.68127680e-01
-1.62720037e+00 3.52309674e-01 -4.87617813e-02 -9.78754222e-01
1.95708215e-01 -3.69684607e-01 -8.82466018e-01 6.36824846e-01
-1.16937160e+00 -8.99682403e-01 -1.49976850e-01 3.97746474e-01
-7.18812943e-02 1.79369211e-01 9.82469201e-01 5.36566675e-02
-1.40653580e-01 4.10885692e-01 2.79282540e-01 5.18431365e-01
4.52086896e-01 -1.51381087e+00 3.76462042e-01 4.94186014e-01
3.10828209e-01 1.27222288e+00 4.00208086e-01 -5.14097214e-01
-1.07675886e+00 -9.14258540e-01 1.29218769e+00 -8.85193765e-01
1.11677527e+00 -5.35090029e-01 -8.30209196e-01 6.23686910e-01
1.67601809e-01 1.52948216e-01 9.88392293e-01 5.02468407e-01
-6.92092478e-01 7.39716589e-02 -1.11655331e+00 4.89464849e-01
1.44122410e+00 -8.55007529e-01 -6.58618331e-01 7.53351972e-02
8.45201373e-01 9.42341208e-01 -1.01263404e+00 1.81789771e-01
4.81345326e-01 -1.10135078e+00 1.11725688e+00 -5.02280772e-01
2.12681085e-01 -3.26443911e-01 -1.27203599e-01 -1.41037989e+00
-8.84833336e-01 -5.65908313e-01 2.18328774e-01 1.12103593e+00
2.78893739e-01 -1.09886587e+00 7.72380292e-01 1.29293069e-01
2.88263619e-01 -7.49963999e-01 -7.58769214e-01 -8.23214889e-01
2.27729023e-01 1.82447776e-01 5.43348014e-01 1.72529852e+00
2.02787846e-01 5.62510908e-01 2.50984997e-01 -1.73867688e-01
4.26502168e-01 5.97729497e-02 4.70270783e-01 -2.04718661e+00
-2.48247222e-03 -7.76305676e-01 -8.59441578e-01 -6.26347840e-01
1.67619899e-01 -9.18238461e-01 -6.28533959e-01 -1.94425714e+00
3.94501358e-01 -9.56278980e-01 -2.68736005e-01 5.21103859e-01
2.24665552e-02 1.27294838e-01 7.21710846e-02 5.85518897e-01
-4.43386495e-01 4.16645676e-01 9.79081511e-01 -2.43143573e-01
-1.19765922e-01 -2.27146804e-01 -1.37391734e+00 6.64762318e-01
6.10117912e-01 -5.37538826e-01 -6.72880232e-01 -7.40055367e-02
7.80040860e-01 -3.84012252e-01 3.63193363e-01 -6.55227065e-01
3.86901647e-02 -8.42907466e-03 -7.85316154e-03 5.23339100e-02
3.91395316e-02 -7.45984375e-01 1.25995561e-01 6.34023845e-01
-3.80073249e-01 4.74204749e-01 -3.79035264e-01 1.08727801e+00
-3.59897852e-01 -1.66722715e-01 4.61742789e-01 -3.02162975e-01
-5.67488194e-01 4.22323383e-02 -5.09925187e-01 4.55972962e-02
1.18457210e+00 -7.73851752e-01 -2.29811937e-01 -7.09278882e-01
-9.12226379e-01 8.18075836e-02 9.25494671e-01 5.62503040e-01
7.23573148e-01 -1.56465983e+00 -7.90923834e-01 -7.73247406e-02
3.49373400e-01 -4.52577263e-01 -3.02312255e-01 9.18706715e-01
-4.85452801e-01 1.64140493e-01 -6.94284439e-02 -3.70019943e-01
-1.62503362e+00 4.65228796e-01 1.24955125e-01 -2.80175537e-01
-2.94127673e-01 9.68013287e-01 2.36299753e-01 -7.74226010e-01
-5.55192642e-02 -2.36880645e-01 -3.11250955e-01 5.23051381e-01
3.13293457e-01 4.65527743e-01 -4.65648532e-01 -6.73631072e-01
-7.81077802e-01 3.75331491e-01 1.11558713e-01 5.36891334e-02
1.44376791e+00 -9.87870898e-03 -8.53920817e-01 5.69377661e-01
1.26879001e+00 3.40096444e-01 -5.44424236e-01 -3.24143469e-02
2.52026767e-01 -6.05740488e-01 -2.68330812e-01 -2.18890160e-01
-8.15880954e-01 8.93943429e-01 1.23530991e-01 1.08596516e+00
5.68174958e-01 4.77114677e-01 3.29113841e-01 1.04476228e-01
3.49947810e-01 -6.11320019e-01 5.19224763e-01 3.60154778e-01
7.38622010e-01 -8.55251014e-01 2.47005336e-02 -6.94050491e-01
-2.06315577e-01 8.80783319e-01 2.79418647e-01 -2.97126830e-01
1.05690348e+00 -1.00605704e-01 -4.68991458e-01 -4.01586860e-01
-5.69716394e-01 -2.05611452e-01 5.37008159e-02 1.12460744e+00
6.48991764e-01 2.67354369e-01 -5.24724245e-01 4.28192019e-01
-4.47003655e-02 -7.22163796e-01 7.86284864e-01 1.01558816e+00
-5.74662447e-01 -1.39802551e+00 -4.05313522e-01 4.98176783e-01
-1.32265434e-01 -2.42968962e-01 -1.16203916e+00 1.02012205e+00
1.99751511e-01 1.07530022e+00 3.24161708e-01 -6.81022942e-01
-1.70666739e-01 -1.99342985e-02 4.18229789e-01 -1.08087862e+00
-4.31580365e-01 -3.87491792e-01 8.82389620e-02 -1.18487425e-01
-5.33694923e-01 -5.85412383e-01 -1.09007561e+00 -8.59780192e-01
-3.95698935e-01 6.28003836e-01 1.57061845e-01 4.57849622e-01
5.76319635e-01 1.19746640e-01 7.35963404e-01 -3.97145331e-01
-4.73033488e-01 -9.31504369e-01 -8.67036641e-01 5.88499844e-01
1.45120680e-01 -7.14546502e-01 -7.84275174e-01 -2.39356488e-01] | [7.199984073638916, 6.16398811340332] |
7331bcca-b205-4195-9698-f5a795a8a9f9 | blind-image-quality-assessment-via | 2305.09353 | null | https://arxiv.org/abs/2305.09353v1 | https://arxiv.org/pdf/2305.09353v1.pdf | Blind Image Quality Assessment via Transformer Predicted Error Map and Perceptual Quality Token | Image quality assessment is a fundamental problem in the field of image processing, and due to the lack of reference images in most practical scenarios, no-reference image quality assessment (NR-IQA), has gained increasing attention recently. With the development of deep learning technology, many deep neural network-based NR-IQA methods have been developed, which try to learn the image quality based on the understanding of database information. Currently, Transformer has achieved remarkable progress in various vision tasks. Since the characteristics of the attention mechanism in Transformer fit the global perceptual impact of artifacts perceived by a human, Transformer is thus well suited for image quality assessment tasks. In this paper, we propose a Transformer based NR-IQA model using a predicted objective error map and perceptual quality token. Specifically, we firstly generate the predicted error map by pre-training one model consisting of a Transformer encoder and decoder, in which the objective difference between the distorted and the reference images is used as supervision. Then, we freeze the parameters of the pre-trained model and design another branch using the vision Transformer to extract the perceptual quality token for feature fusion with the predicted error map. Finally, the fused features are regressed to the final image quality score. Extensive experiments have shown that our proposed method outperforms the current state-of-the-art in both authentic and synthetic image databases. Moreover, the attentional map extracted by the perceptual quality token also does conform to the characteristics of the human visual system. | ['Aljosa Smolic', 'Pan Gao', 'Jinsong Shi'] | 2023-05-16 | null | null | null | null | ['blind-image-quality-assessment', 'image-quality-assessment', 'no-reference-image-quality-assessment'] | ['computer-vision', 'computer-vision', 'computer-vision'] | [ 3.51278543e-01 -3.77157390e-01 2.84246773e-01 -2.97027677e-01
-6.61030233e-01 1.03804715e-01 3.75597894e-01 4.99325152e-03
-3.01615030e-01 4.53418225e-01 -5.29714152e-02 8.16036537e-02
-1.59626991e-01 -8.88257623e-01 -6.47077262e-01 -8.07175517e-01
2.62204498e-01 -2.00221673e-01 1.83505490e-01 -1.79038048e-01
4.17170763e-01 2.91770369e-01 -1.66620982e+00 1.68419942e-01
1.21234298e+00 1.67310810e+00 5.19119143e-01 4.93544221e-01
1.80531427e-01 9.23674941e-01 -7.35077918e-01 -6.06521428e-01
3.31775278e-01 -5.91882825e-01 -4.82646137e-01 2.14186132e-01
2.64448553e-01 -5.08780897e-01 -5.88654280e-01 1.38701463e+00
8.03211153e-01 5.08717373e-02 4.20648903e-01 -1.13910866e+00
-7.78409004e-01 6.42898083e-02 -3.69067192e-01 3.28182876e-01
2.72804350e-01 5.40955007e-01 6.98006868e-01 -9.71715629e-01
2.96265274e-01 1.09576237e+00 3.41500580e-01 1.94533437e-01
-9.58428442e-01 -5.29326022e-01 -2.50986874e-01 8.06737363e-01
-1.22196794e+00 -3.67898494e-01 9.86769199e-01 -2.89935529e-01
7.38167405e-01 5.02790995e-02 7.84808099e-01 7.73292363e-01
5.82327724e-01 6.03512108e-01 1.24955213e+00 -3.54356706e-01
2.44361565e-01 1.10080354e-01 -2.87133038e-01 5.80498993e-01
-3.61015312e-02 3.31772864e-01 -4.46969271e-01 4.06954080e-01
7.56826997e-01 -1.00559399e-01 -5.25644720e-01 -3.27890456e-01
-1.07578135e+00 4.04794484e-01 7.75608599e-01 4.21751231e-01
-5.98784149e-01 -1.04309566e-01 2.78306156e-01 3.59149098e-01
1.72541276e-01 2.77251005e-01 -7.93185756e-02 7.43790045e-02
-9.64296103e-01 -1.53451875e-01 1.95313588e-01 4.70644385e-01
6.61741316e-01 3.14635307e-01 -4.26544875e-01 8.70585501e-01
2.97604829e-01 7.53304958e-01 6.70822620e-01 -9.80728686e-01
3.44270647e-01 6.66979194e-01 2.83053219e-01 -1.36485410e+00
-4.81680557e-02 -6.88027084e-01 -1.12719834e+00 6.19325161e-01
2.86100596e-01 3.35830182e-01 -7.82476187e-01 1.53082490e+00
6.63982704e-02 7.49976560e-03 1.49930835e-01 1.09259593e+00
5.54821432e-01 8.03086400e-01 -1.53682142e-01 -4.22051996e-01
1.17350483e+00 -8.22083116e-01 -9.67560947e-01 2.57057082e-02
-2.57512834e-02 -8.44649613e-01 1.25576448e+00 7.84799814e-01
-1.31114876e+00 -1.34925973e+00 -1.40605688e+00 1.84552446e-02
-1.32006824e-01 2.93133289e-01 6.54665083e-02 5.90264440e-01
-1.19069755e+00 6.93037689e-01 -6.29160643e-01 -2.37130071e-03
4.40203816e-01 1.99833676e-01 -1.52717739e-01 -2.69912332e-01
-1.37277997e+00 1.13715041e+00 3.81280124e-01 4.18240964e-01
-1.07483315e+00 -3.30770314e-01 -7.26059079e-01 2.58753836e-01
2.28683546e-01 -6.55988634e-01 9.15295064e-01 -1.34895408e+00
-1.70874214e+00 6.24085069e-01 2.49288511e-02 -3.05798799e-01
4.62930292e-01 -1.12098485e-01 -7.08664596e-01 2.80490249e-01
-5.30261733e-02 5.03519297e-01 1.17291999e+00 -1.18639100e+00
-6.71601892e-01 -3.15636516e-01 1.45214066e-01 2.23446131e-01
-2.42824629e-01 -6.79409504e-02 -5.38683116e-01 -4.56842035e-01
-7.06688091e-02 -3.62490028e-01 7.57663026e-02 1.64065868e-01
-1.29984751e-01 -1.34182781e-01 5.74091673e-01 -8.22568834e-01
1.26357424e+00 -2.21809149e+00 -1.85047630e-02 1.61315069e-01
1.40827850e-01 5.02523720e-01 -2.90540099e-01 5.32890074e-02
-1.48159727e-01 -1.81144044e-01 -1.95671126e-01 -6.05476089e-02
-1.53912812e-01 -1.85667858e-01 3.16429250e-02 3.87268126e-01
3.48838180e-01 6.94654703e-01 -8.41232717e-01 -6.15222752e-01
5.51802576e-01 4.68130708e-01 -3.25118512e-01 6.82282984e-01
2.51004640e-02 4.16493386e-01 -2.48082191e-01 3.97005975e-01
8.52046371e-01 -2.34812990e-01 -1.28473043e-01 -7.49285161e-01
-1.00042410e-01 4.67755571e-02 -1.03000271e+00 1.66434717e+00
-5.31502128e-01 6.11878693e-01 -1.36992916e-01 -1.00913405e+00
1.10671902e+00 3.38901430e-01 4.24136221e-01 -1.46696830e+00
2.65283734e-01 2.67372459e-01 2.13211343e-01 -7.08702385e-01
2.09296629e-01 -1.59520805e-01 3.60542297e-01 6.26408756e-02
2.97747791e-01 -1.34692624e-01 9.76111442e-02 -1.10447809e-01
8.79211426e-01 1.85605749e-01 3.08145016e-01 1.31478310e-01
9.11739826e-01 -4.83078420e-01 5.52391231e-01 4.05399710e-01
-4.08694893e-01 7.04021335e-01 1.67330727e-01 -4.73564506e-01
-1.30375588e+00 -1.20938528e+00 -1.10006668e-01 5.15692532e-01
5.58918417e-01 -5.42252548e-02 -8.70914161e-01 -4.29974496e-01
-5.42985141e-01 5.13122201e-01 -4.16905254e-01 -6.11538947e-01
-3.78771693e-01 -5.14310122e-01 2.05155268e-01 2.10483417e-01
1.18932486e+00 -1.41112673e+00 -6.86374605e-01 2.84957439e-01
-5.41629612e-01 -9.85716343e-01 -3.19039434e-01 -1.93764374e-01
-5.81681550e-01 -1.00703549e+00 -7.44570255e-01 -5.86055696e-01
4.45187151e-01 2.44989514e-01 1.01803398e+00 1.99828163e-01
-1.08703405e-01 6.13786876e-02 -2.47213557e-01 -1.11168839e-01
-4.10188884e-01 -4.72280174e-01 -1.83308125e-01 5.06158412e-01
3.18049490e-02 -4.60903257e-01 -9.50635970e-01 3.30285639e-01
-1.01507819e+00 9.97215286e-02 9.52042222e-01 9.12221789e-01
6.04711533e-01 5.06514966e-01 5.39642274e-01 -3.54355611e-02
6.08294725e-01 2.51994338e-02 -6.66400850e-01 3.74894708e-01
-7.25321770e-01 4.24065329e-02 7.18743205e-01 -2.16747329e-01
-1.28805661e+00 -3.58073056e-01 -3.10326815e-01 -5.83783805e-01
-1.19872382e-02 4.42531973e-01 -7.10883498e-01 -1.32541228e-02
4.81681406e-01 5.17203927e-01 -3.25855389e-02 -1.97180584e-01
7.33327642e-02 7.24853635e-01 7.15118885e-01 -9.90820900e-02
7.52725482e-01 1.88403338e-01 -4.59390245e-02 -4.23772693e-01
-6.64773107e-01 -1.27439678e-01 -3.89199167e-01 -5.88450313e-01
9.68298256e-01 -7.54386723e-01 -8.27042162e-01 1.00510728e+00
-1.23941350e+00 -1.00076586e-01 -2.16248825e-01 5.77609420e-01
-6.91508293e-01 4.98542011e-01 -4.31482285e-01 -7.01166868e-01
-5.77996910e-01 -1.47648120e+00 8.07667911e-01 4.01921391e-01
3.64456236e-01 -6.14778697e-01 -1.67370826e-01 3.37428838e-01
5.98856688e-01 2.71925237e-02 1.00536823e+00 -7.63100311e-02
-7.35757649e-01 9.36739668e-02 -5.47919869e-01 9.12438631e-01
2.32085511e-01 -2.86143988e-01 -9.54717815e-01 -2.20415965e-01
4.92300808e-01 -2.69452929e-01 5.84057987e-01 4.49248224e-01
1.25600946e+00 -1.39434338e-01 2.47734904e-01 5.97602904e-01
1.64039969e+00 5.74113786e-01 1.19927752e+00 3.49630952e-01
4.05035377e-01 3.63729686e-01 6.92922890e-01 3.13712716e-01
1.68313667e-01 7.48055518e-01 6.39204204e-01 -3.22609246e-01
-3.30647320e-01 -1.36280313e-01 4.16261733e-01 8.64569783e-01
-1.08482622e-01 -3.18868965e-01 -5.14210999e-01 3.34205925e-01
-1.52015758e+00 -1.07697093e+00 2.41492495e-01 2.32101250e+00
7.42139101e-01 4.14989233e-01 -2.68410146e-01 6.06097400e-01
6.81439757e-01 -2.08162721e-02 -7.72857428e-01 -3.10953707e-01
-1.98736981e-01 2.34271556e-01 1.44835055e-01 2.14955494e-01
-7.99718201e-01 5.10436893e-01 5.36909437e+00 1.01656270e+00
-1.33643544e+00 3.22714224e-02 8.06790590e-01 3.26662987e-01
-8.58091637e-02 -2.56601006e-01 -8.10120702e-02 6.42620564e-01
8.79253387e-01 -7.64606372e-02 5.92907727e-01 5.07342875e-01
5.85993111e-01 -2.75005966e-01 -1.02017295e+00 1.33657217e+00
2.23035276e-01 -8.07893097e-01 2.44050056e-01 -1.08329959e-01
4.83727813e-01 -4.06209141e-01 4.51749206e-01 1.19071312e-01
-2.28024095e-01 -1.05512059e+00 8.42751205e-01 8.77610624e-01
9.01651621e-01 -7.21820235e-01 1.06069207e+00 2.96861500e-01
-1.06588805e+00 -2.08605573e-01 -4.97381598e-01 7.40519911e-02
1.61732901e-02 7.10482776e-01 -3.14062387e-01 8.49274814e-01
8.19612622e-01 7.04555154e-01 -8.16964746e-01 1.20236814e+00
-2.03411758e-01 4.29544568e-01 1.48771957e-01 3.72641325e-01
-1.53069012e-02 -3.04338336e-01 3.11920583e-01 6.45686865e-01
6.59798324e-01 6.56592995e-02 -2.31841341e-01 1.00287056e+00
-2.68335994e-02 4.14393991e-02 -2.77622372e-01 2.09407672e-01
1.43312275e-01 1.13763046e+00 -3.44250381e-01 -4.83277410e-01
-3.52426082e-01 1.07748008e+00 -1.48665747e-02 3.28143537e-01
-8.64493906e-01 -4.79693413e-01 2.66997457e-01 -7.41199180e-02
3.45095575e-01 1.87257484e-01 -1.51871279e-01 -1.04591167e+00
2.76102871e-01 -1.13802874e+00 -7.97112286e-02 -1.42127669e+00
-9.90316987e-01 8.81196082e-01 -3.62380683e-01 -1.62573600e+00
-1.23175845e-01 -5.22211730e-01 -5.03943205e-01 1.05211639e+00
-1.64289510e+00 -8.80838215e-01 -6.60778821e-01 7.30756998e-01
5.34774244e-01 -6.99458867e-02 5.57138026e-01 4.70844656e-01
-4.23570752e-01 5.64383388e-01 -2.46645603e-02 1.38840511e-01
7.02635527e-01 -1.04070115e+00 3.19544226e-02 1.06445086e+00
-1.33696541e-01 2.33484805e-01 7.34062552e-01 -3.59962970e-01
-1.25543749e+00 -9.69386220e-01 5.29593408e-01 7.03737512e-02
2.44642660e-01 2.83053201e-02 -1.10394871e+00 -4.03513275e-02
5.28481722e-01 2.25827560e-01 1.04879603e-01 -6.13110006e-01
-2.16517180e-01 -7.63726294e-01 -1.25150013e+00 2.93747455e-01
7.60771811e-01 -5.41356206e-01 -4.96426433e-01 -1.92452103e-01
6.49300694e-01 -1.50984019e-01 -8.90505731e-01 5.17837644e-01
5.21929085e-01 -1.42191648e+00 8.68113279e-01 1.93085104e-01
6.41737759e-01 -5.60925245e-01 -1.27623215e-01 -1.54770851e+00
-3.09363693e-01 -4.74775955e-02 1.45568803e-01 1.18888366e+00
1.28305806e-02 -2.86429793e-01 3.31732780e-01 1.54762551e-01
-1.97299302e-01 -7.15746403e-01 -8.64247143e-01 -5.56646109e-01
-3.37555438e-01 -2.49320984e-01 6.29769325e-01 4.61118311e-01
-3.72589916e-01 2.24792942e-01 -4.94275212e-01 1.60976127e-01
7.59873033e-01 -2.87748035e-02 4.83118474e-01 -9.62119401e-01
-3.49721670e-01 -4.00382161e-01 -6.70159996e-01 -9.39140499e-01
-2.48466387e-01 -5.06691754e-01 1.20237544e-01 -1.63199103e+00
2.41303056e-01 -1.13192357e-01 -6.68705046e-01 2.61886101e-02
-3.51396173e-01 2.71243066e-01 2.85651863e-01 1.48005560e-01
-7.51320481e-01 8.53699982e-01 1.59713864e+00 -5.34931958e-01
-4.92221937e-02 -1.10557556e-01 -4.22186226e-01 4.69855577e-01
6.12324119e-01 -2.33663291e-01 -4.81104106e-01 -4.59723085e-01
1.75666288e-01 4.12571043e-01 4.13996428e-01 -1.50939119e+00
1.87134728e-01 -5.19286580e-02 7.58818448e-01 -5.78016698e-01
3.40937763e-01 -9.86469209e-01 1.90483227e-01 5.19501209e-01
-2.70041376e-01 8.50715861e-02 -1.48964778e-01 3.55210543e-01
-6.74700499e-01 -1.03155263e-01 1.09641588e+00 -5.50188869e-02
-8.10488760e-01 2.82034069e-01 -1.71290651e-01 -2.03119934e-01
8.01132917e-01 -5.07262170e-01 -1.80411905e-01 -3.25938493e-01
-4.59794760e-01 -5.64633720e-02 3.82026047e-01 2.02874318e-01
1.18398798e+00 -1.43521237e+00 -7.48714030e-01 2.35721618e-01
1.74468786e-01 -3.19816887e-01 5.80745518e-01 7.75244594e-01
-2.15341389e-01 1.43692404e-01 -6.91778123e-01 -6.70479476e-01
-9.56437111e-01 9.05153215e-01 6.34967208e-01 -3.91376585e-01
-3.47524732e-01 2.32118309e-01 2.72436559e-01 9.54854414e-02
1.65639475e-01 -2.96893269e-01 -4.57152128e-01 -3.00247818e-01
7.22933173e-01 3.62033814e-01 3.75050277e-01 -8.07555854e-01
-5.56037724e-02 7.11487114e-01 3.15091193e-01 -9.22622457e-02
1.07778561e+00 -4.23631012e-01 3.56779993e-02 2.11972162e-01
1.26365924e+00 -2.57396638e-01 -1.21509814e+00 -3.15738887e-01
-2.51241714e-01 -6.21654332e-01 2.91324615e-01 -1.12363863e+00
-1.40448976e+00 1.15642595e+00 1.30415308e+00 6.14399426e-02
1.90292525e+00 -4.55523163e-01 7.30600893e-01 3.48806456e-02
3.87690306e-01 -9.50327158e-01 5.29887557e-01 7.10688084e-02
9.54367280e-01 -1.32151747e+00 -2.66067743e-01 -5.89404069e-03
-4.87855405e-01 9.70042944e-01 6.78777874e-01 3.59050301e-03
4.22616184e-01 -7.22179189e-02 1.94114193e-01 4.72542383e-02
-4.83114183e-01 -1.69352472e-01 4.14182752e-01 7.49843180e-01
1.18506946e-01 -2.88764745e-01 -1.91265583e-01 4.55446213e-01
-7.74392113e-02 3.75499547e-01 4.44175571e-01 2.93782085e-01
-4.55101311e-01 -8.76271486e-01 -4.41632420e-01 3.09361815e-01
-4.18908864e-01 -3.16667669e-02 1.93893880e-01 4.13398415e-01
5.99675417e-01 1.23227692e+00 -8.02327842e-02 -5.48047483e-01
5.58593273e-01 -3.06086361e-01 5.05597770e-01 -4.13162410e-02
-4.54122335e-01 4.21301462e-02 -3.63497943e-01 -7.51132905e-01
-5.97036779e-01 -1.48514912e-01 -1.03671598e+00 -1.12608619e-01
-3.72128636e-01 6.20939024e-02 6.87596142e-01 8.71369660e-01
1.51330829e-01 7.67576456e-01 9.36384976e-01 -7.53463149e-01
-4.74332005e-01 -1.09666634e+00 -5.07038593e-01 6.55648053e-01
3.41971308e-01 -7.43948817e-01 -2.73403078e-01 1.46507844e-01] | [11.805098533630371, -1.8970340490341187] |
f6d91657-583b-41ef-8221-4d8e8b1fe965 | designing-biological-circuits-from-principles | 2111.04508 | null | https://arxiv.org/abs/2111.04508v1 | https://arxiv.org/pdf/2111.04508v1.pdf | Designing biological circuits: from principles to applications | Genetic circuit design is a well-studied problem in synthetic biology. Ever since the first genetic circuits -- the repressilator and the toggle switch -- were designed and implemented, many advances have been made in this area of research. The current review systematically organizes a number of key works in this domain by employing the versatile framework of generalized morphological analysis. Literature in the area has been mapped based on (a) the design methodologies used, ranging from brute-force searches to control-theoretic approaches, (b) the modelling techniques employed, (c) various circuit functionalities implemented, (d) key design characteristics, and (e) the strategies used for the robust design of genetic circuits. We conclude our review with an outlook on multiple exciting areas for future research, based on the systematic assessment of key research gaps that have been readily unravelled by our analysis framework. | ['Karthik Raman', 'Raghunathan Rengaswamy', 'Debomita Chakraborty'] | 2021-11-05 | null | null | null | null | ['robust-design'] | ['miscellaneous'] | [ 9.19135392e-01 6.41577542e-02 -1.19302854e-01 1.81478545e-01
2.15110645e-01 -1.07695842e+00 4.28468317e-01 2.87999749e-01
-2.03240126e-01 1.00632465e+00 -2.00205162e-01 -4.87756938e-01
-5.44857264e-01 -7.17303634e-01 -6.52252436e-01 -1.11501026e+00
-2.15387106e-01 2.90216625e-01 2.20999107e-01 -7.57350266e-01
6.82767212e-01 6.45720661e-01 -1.97825658e+00 2.03803219e-02
6.32242322e-01 7.72692084e-01 2.72583008e-01 8.54821265e-01
1.20079637e-01 2.24612698e-01 -5.73002100e-01 -3.75618607e-01
-6.17762394e-02 -1.34481692e+00 -2.76025891e-01 -2.48499930e-01
-3.95770758e-01 5.37615538e-01 -5.64626791e-02 5.27755439e-01
7.86871850e-01 -3.34067702e-01 6.77169621e-01 -8.04430127e-01
-6.65658176e-01 5.50760806e-01 2.06282157e-02 -7.69681185e-02
3.32447737e-01 2.44389892e-01 7.04261363e-01 -5.29001236e-01
1.04345489e+00 1.00205135e+00 4.20216441e-01 8.75170231e-01
-1.30674469e+00 -2.49968082e-01 -5.59540950e-02 -2.97786951e-01
-1.44612527e+00 -5.94667912e-01 4.37539279e-01 -3.57243717e-01
1.03501201e+00 4.80729252e-01 1.09517634e+00 5.37246287e-01
6.75210774e-01 2.41371840e-01 1.13601923e+00 -7.37411320e-01
4.47017074e-01 1.27800889e-02 -4.48691428e-01 7.62105882e-01
6.87914014e-01 3.21402967e-01 -5.47672510e-01 -8.83475393e-02
9.18099046e-01 -7.31091976e-01 -1.60315111e-01 -5.73547304e-01
-8.80376756e-01 5.86638153e-01 -2.01178357e-01 8.45333457e-01
6.61957413e-02 3.24732751e-01 2.86656827e-01 3.56237829e-01
-1.45000190e-01 8.63742590e-01 -5.50437093e-01 -1.36601031e-01
-6.85343504e-01 4.52685982e-01 1.10385907e+00 9.49682474e-01
2.37738863e-01 2.89374292e-01 1.23782583e-01 6.34086013e-01
2.94096202e-01 6.07148325e-03 2.55778879e-01 -7.55949259e-01
-3.04335147e-01 6.30044580e-01 -7.07553476e-02 -7.83982694e-01
-5.13399541e-01 -2.28550211e-01 -5.21923125e-01 8.65879729e-02
5.81164956e-01 -5.81112742e-01 -5.59554994e-01 1.70798421e+00
1.99767306e-01 -3.07897300e-01 -8.99999589e-02 2.71551818e-01
6.50975108e-01 3.31518382e-01 -1.02615356e-01 -5.43373346e-01
1.37596083e+00 -2.85377920e-01 -8.34252238e-01 1.91424608e-01
4.34042513e-01 -6.33861363e-01 4.51461166e-01 3.11490208e-01
-1.34357584e+00 -3.53949368e-02 -1.36940837e+00 1.59537107e-01
-7.86748409e-01 7.53797293e-02 8.20281744e-01 1.38885605e+00
-1.20274138e+00 6.06420815e-01 -7.97058046e-01 -9.87062812e-01
2.36215144e-01 7.16035724e-01 -1.36659831e-01 2.72382915e-01
-8.01490247e-01 9.79349434e-01 4.36741561e-01 1.42281234e-01
-9.67852771e-01 -6.10098183e-01 -5.52035511e-01 -7.91753680e-02
4.59050745e-01 -9.20951605e-01 6.76043808e-01 -4.81741667e-01
-2.20059609e+00 1.17325306e+00 9.64109004e-02 -1.68789342e-01
2.71587998e-01 4.81689960e-01 -8.82162079e-02 3.26282606e-02
-5.49181223e-01 7.34589636e-01 1.01299966e-02 -1.19210041e+00
-4.34220850e-01 -3.01754862e-01 -4.66628186e-02 -1.96707159e-01
-1.26505848e-02 4.03833598e-01 -1.09820664e-01 -9.08467948e-01
1.37419060e-01 -1.10446310e+00 -4.45745856e-01 8.21320117e-02
-4.23219204e-01 3.62508558e-03 3.96235049e-01 1.04650341e-01
1.42727327e+00 -1.77414656e+00 6.47614717e-01 2.53858995e-02
-2.08626553e-01 4.50201452e-01 -7.71431439e-03 1.26138616e+00
-2.04880968e-01 4.05174881e-01 -5.35009146e-01 3.75985414e-01
-1.83791906e-01 3.11744418e-02 -1.75320983e-01 6.93302989e-01
1.84882566e-01 1.04553938e+00 -7.20455945e-01 1.75231509e-02
2.82455325e-01 3.59985173e-01 -5.32650292e-01 -1.60650596e-01
-2.60846823e-01 5.08013427e-01 -4.52823639e-01 9.99020576e-01
2.12864563e-01 2.74776280e-01 6.38469040e-01 2.78094232e-01
-7.80754268e-01 6.39688298e-02 -8.03435624e-01 1.69569838e+00
2.32961118e-01 4.50598747e-01 2.60997057e-01 -1.08853590e+00
1.22136796e+00 4.68009621e-01 4.33455229e-01 -2.38582775e-01
4.58241016e-01 4.70856547e-01 4.51494902e-01 -4.26881284e-01
-5.07763326e-02 -2.79010296e-01 -1.03350662e-01 2.04560474e-01
2.17739761e-01 -6.45182729e-01 6.32409453e-01 -3.15632790e-01
1.22529280e+00 6.19105995e-01 6.52539968e-01 -7.11915612e-01
6.77761853e-01 2.33836591e-01 6.08137488e-01 7.15549886e-01
-2.19643205e-01 3.88109028e-01 9.12272573e-01 -2.81720638e-01
-1.08978832e+00 -4.98459369e-01 -3.46988499e-01 6.36888564e-01
6.35457337e-02 -5.30526817e-01 -1.22829485e+00 7.48562738e-02
3.04301493e-02 3.42809051e-01 -7.82846510e-01 -3.52618456e-01
-6.95159376e-01 -1.32510388e+00 1.23880458e+00 1.70874730e-01
2.36183740e-02 -8.69259477e-01 -9.25092638e-01 4.33447510e-01
5.01995504e-01 -5.18203378e-01 4.25284505e-01 5.58294356e-01
-1.02107835e+00 -1.03277099e+00 -4.87094998e-01 -1.05052400e+00
6.92672491e-01 1.17952116e-01 6.28311157e-01 3.49103332e-01
-7.21341014e-01 1.60291180e-01 -2.71555930e-01 -7.42102027e-01
-2.98677415e-01 -1.62340477e-01 -7.32099041e-02 -5.84092617e-01
6.20229775e-03 -7.89043844e-01 -3.21710795e-01 5.42980194e-01
-1.11594260e+00 -3.10301960e-01 4.70309436e-01 1.00770855e+00
5.74959338e-01 1.73410758e-01 9.90752995e-01 -7.49564588e-01
5.86465597e-01 -1.60909280e-01 -8.92046809e-01 5.00931621e-01
-4.85724479e-01 -1.38894737e-01 4.88661289e-01 2.14884952e-02
-8.83205831e-01 1.59948617e-01 -2.64494658e-01 7.07186937e-01
-1.42860666e-01 4.56667870e-01 -5.13243318e-01 -7.85712540e-01
6.06900394e-01 2.75612414e-01 1.51576415e-01 -1.61811218e-01
4.01595294e-01 1.64717108e-01 1.75359666e-01 -8.40917051e-01
4.60118115e-01 5.98250270e-01 6.52388334e-01 -1.22594261e+00
1.36723161e-01 1.27238885e-01 -6.29424751e-01 -2.35084668e-01
6.16117120e-01 -2.62644470e-01 -9.46561515e-01 6.72594607e-01
-7.11007416e-01 -2.08569393e-01 -1.12142131e-01 1.38874546e-01
-9.94746745e-01 7.55520537e-02 -6.33729577e-01 -9.60238159e-01
1.00036412e-02 -1.37719941e+00 7.38003552e-01 3.87341648e-01
-3.39154929e-01 -9.93788481e-01 3.37657720e-01 -8.90862793e-02
3.75644773e-01 7.75009751e-01 1.50216877e+00 -1.74274683e-01
-5.15366435e-01 -6.07738979e-02 3.64033520e-01 -3.10329556e-01
4.47914079e-02 7.59721696e-01 -8.27251673e-01 -2.14994222e-01
5.17373458e-02 -1.63731515e-01 6.12513363e-01 6.28206015e-01
6.85540318e-01 3.04163307e-01 -8.10376644e-01 5.76227903e-01
1.60737002e+00 8.90360355e-01 9.01156962e-01 4.48132843e-01
-2.53724996e-02 1.03873682e+00 6.32907271e-01 3.38686377e-01
-1.75685614e-01 6.56852365e-01 2.51969755e-01 1.73486933e-01
-1.04408875e-01 -2.15020344e-01 9.08459947e-02 5.40325880e-01
-4.09185112e-01 -7.76223898e-01 -5.88786423e-01 3.60237628e-01
-1.72080350e+00 -6.63335919e-01 7.49990493e-02 2.05411911e+00
6.98898256e-01 -1.09164771e-02 2.33068213e-01 3.94394189e-01
7.64602840e-01 -3.15735400e-01 -4.62617993e-01 -5.45046508e-01
-6.32268071e-01 4.42077935e-01 4.88440037e-01 8.33821893e-02
-7.22263515e-01 8.72553170e-01 8.32095051e+00 7.51527727e-01
-1.16725039e+00 -5.62742293e-01 5.45486689e-01 -2.06493326e-02
-3.01429927e-01 4.79534984e-01 -8.84564877e-01 2.88602799e-01
8.58108342e-01 -3.42309415e-01 4.07743931e-01 2.63624340e-01
3.18519175e-01 -2.68706083e-01 -1.01205385e+00 4.52492952e-01
1.14181440e-03 -1.58338475e+00 -9.51302424e-02 9.71426219e-02
8.09962034e-01 -6.42098665e-01 1.01934515e-01 -1.74965352e-01
1.11120194e-01 -1.00667787e+00 8.49769592e-01 3.50299448e-01
7.90298462e-01 -5.92652142e-01 2.56651610e-01 1.28464058e-01
-9.28212464e-01 -3.46689016e-01 -5.34705937e-01 -3.67556900e-01
2.63139188e-01 5.83598912e-01 -4.28311586e-01 6.69471323e-01
4.38522756e-01 4.74498659e-01 -2.51720160e-01 1.27301073e+00
-4.47859108e-01 4.62970138e-01 -2.45116472e-01 -5.43276787e-01
-8.62497184e-03 -5.44758558e-01 6.48941576e-01 1.13277364e+00
4.36998546e-01 3.06173801e-01 -5.27968943e-01 9.84377444e-01
3.62184227e-01 -3.30986306e-02 -5.81515431e-01 -6.49378479e-01
4.76194412e-01 1.07914305e+00 -1.42031348e+00 9.92909297e-02
-7.47270957e-02 3.19762498e-01 -1.38425514e-01 1.58565745e-01
-5.26057959e-01 -9.58704710e-01 7.77731061e-01 1.14208728e-01
3.06639552e-01 -2.97901839e-01 -4.11760509e-01 -5.98108530e-01
-5.26524663e-01 -8.55427980e-01 2.95118600e-01 -5.20118713e-01
-5.24006128e-01 1.50014758e-01 -2.96956878e-02 -6.03726387e-01
-7.99387619e-02 -8.82085502e-01 -3.79402161e-01 5.54937005e-01
-9.22546148e-01 -7.74110317e-01 4.94324356e-01 -3.85981739e-01
3.06539237e-01 -9.37452987e-02 9.92103636e-01 1.77830428e-01
-8.98809671e-01 3.82572889e-01 4.90499020e-01 -5.70223749e-01
3.67965698e-01 -7.31138289e-01 3.35187882e-01 5.00618875e-01
-4.71656889e-01 1.13241005e+00 8.25617135e-01 -6.75381184e-01
-2.04294658e+00 -5.85149467e-01 8.75866592e-01 -3.41620773e-01
4.62951779e-01 -5.99797130e-01 -3.88585508e-01 3.33803385e-01
1.20124839e-01 -6.68056726e-01 8.36841524e-01 -5.11449158e-01
3.82544965e-01 6.77740350e-02 -1.27481902e+00 9.61069643e-01
1.13347554e+00 -1.17534906e-01 -6.98088109e-02 -9.53822061e-02
3.63569230e-01 -3.50422770e-01 -8.92847955e-01 4.38210607e-01
1.03832090e+00 -8.37432325e-01 8.39042306e-01 -7.24237740e-01
3.31350237e-01 -6.31040215e-01 2.31104389e-01 -1.06357741e+00
-5.14370322e-01 -1.12505972e+00 3.85556638e-01 1.30955458e+00
4.63825226e-01 -5.41442692e-01 5.61332226e-01 2.79658556e-01
-4.74178225e-01 -1.14955676e+00 -8.55666459e-01 -5.67619085e-01
1.52375638e-01 3.59249920e-01 4.72805798e-01 6.54812813e-01
5.99993050e-01 1.26455322e-01 -5.85922913e-04 -5.78810394e-01
-5.70474155e-02 -1.16343096e-01 5.40311813e-01 -8.25626910e-01
-9.91359428e-02 -7.52881527e-01 -7.78414488e-01 -6.69931293e-01
-3.47080201e-01 -6.83059096e-01 6.37412965e-02 -1.23333895e+00
1.86909419e-02 -1.02997422e-01 3.19284871e-02 7.87185282e-02
1.64615825e-01 1.50258258e-01 -1.68195784e-01 -3.11728299e-01
-1.52243882e-01 1.80799544e-01 1.18229842e+00 2.91418672e-01
-2.50374049e-01 -2.53279716e-01 -1.30102336e+00 5.55186033e-01
7.57210135e-01 -5.17797470e-01 -4.48496163e-01 -1.75150618e-01
6.16156757e-01 -8.99759936e-04 -2.87565559e-01 -8.69494438e-01
1.82938218e-01 -2.24827498e-01 2.01414049e-01 -1.85448289e-01
1.60482869e-01 -5.29090405e-01 5.56013465e-01 9.40026760e-01
-4.07507628e-01 2.64892995e-01 4.24656719e-01 5.44815958e-01
2.11299598e-01 -4.64772284e-01 9.50908899e-01 -3.70334744e-01
-3.56285125e-01 -3.35130960e-01 -1.18021023e+00 -4.97241795e-01
1.43293631e+00 -6.65640712e-01 -4.70011473e-01 5.52416518e-02
-5.84270835e-01 -1.00358620e-01 8.74361992e-01 5.21842651e-02
3.25539500e-01 -7.98898399e-01 -4.45342809e-01 -1.07347723e-02
1.84590202e-02 -6.58362627e-01 8.92468318e-02 7.68499911e-01
-1.01795328e+00 8.91633749e-01 -5.93503892e-01 -3.72589946e-01
-1.07917547e+00 7.34240055e-01 4.07579124e-01 2.14977816e-01
1.66280419e-01 9.11343038e-01 3.27585846e-01 -2.83844799e-01
-1.14815924e-02 -1.06694520e-01 -3.09986442e-01 -1.20498233e-01
2.25208774e-01 5.87535143e-01 1.45143464e-01 -4.85105604e-01
-3.41904581e-01 8.94469440e-01 5.09440124e-01 -4.43841144e-02
1.37627900e+00 -2.55415648e-01 -7.51033962e-01 2.54048973e-01
5.03026843e-01 -2.45147318e-01 -7.01130807e-01 6.33814931e-01
-4.35532890e-02 -8.98200274e-02 -2.67532051e-01 -9.48254704e-01
-7.07642555e-01 6.17170215e-01 2.61398584e-01 3.44806686e-02
1.12106776e+00 -3.26330483e-01 6.94012363e-03 2.61873037e-01
5.03457189e-01 -1.05298281e+00 -2.59846449e-01 3.74415576e-01
7.94006765e-01 -1.81150988e-01 6.78259507e-02 -8.96448851e-01
-7.89016336e-02 1.08822358e+00 2.12543070e-01 -1.34791210e-01
6.30401671e-01 6.86658323e-01 -2.99613237e-01 -2.30741888e-01
-1.01410246e+00 -1.19967714e-01 -2.91489780e-01 8.23023140e-01
9.92942870e-01 -2.36245543e-01 -1.21922052e+00 5.68024218e-01
-1.07886158e-01 1.96092099e-01 5.54420769e-01 1.51498306e+00
-5.80585778e-01 -1.63469911e+00 -3.97592008e-01 1.90013170e-01
-5.27964771e-01 2.55477309e-01 -1.20459580e+00 6.84261620e-01
3.34146500e-01 1.28520334e+00 -4.29039896e-01 -2.07976490e-01
4.68408197e-01 -1.38700232e-01 9.62701857e-01 -4.31277245e-01
-8.42478931e-01 2.49041498e-01 2.65242100e-01 -1.03292041e-01
-3.35227489e-01 -7.80451357e-01 -9.82164323e-01 -3.24298352e-01
-6.72155857e-01 3.29515129e-01 8.72835457e-01 8.20805550e-01
4.53278780e-01 8.20661902e-01 1.45063400e-01 -6.87997580e-01
2.22531423e-01 -3.47911805e-01 -8.48249733e-01 -6.74098432e-01
-2.84284472e-01 -6.74176693e-01 -1.75628264e-03 2.25738123e-01] | [5.634951114654541, 4.203885555267334] |
21913acf-d982-46d0-aadf-c276af34b35c | pdformer-propagation-delay-aware-dynamic-long | 2301.07945 | null | https://arxiv.org/abs/2301.07945v2 | https://arxiv.org/pdf/2301.07945v2.pdf | PDFormer: Propagation Delay-Aware Dynamic Long-Range Transformer for Traffic Flow Prediction | As a core technology of Intelligent Transportation System, traffic flow prediction has a wide range of applications. The fundamental challenge in traffic flow prediction is to effectively model the complex spatial-temporal dependencies in traffic data. Spatial-temporal Graph Neural Network (GNN) models have emerged as one of the most promising methods to solve this problem. However, GNN-based models have three major limitations for traffic prediction: i) Most methods model spatial dependencies in a static manner, which limits the ability to learn dynamic urban traffic patterns; ii) Most methods only consider short-range spatial information and are unable to capture long-range spatial dependencies; iii) These methods ignore the fact that the propagation of traffic conditions between locations has a time delay in traffic systems. To this end, we propose a novel Propagation Delay-aware dynamic long-range transFormer, namely PDFormer, for accurate traffic flow prediction. Specifically, we design a spatial self-attention module to capture the dynamic spatial dependencies. Then, two graph masking matrices are introduced to highlight spatial dependencies from short- and long-range views. Moreover, a traffic delay-aware feature transformation module is proposed to empower PDFormer with the capability of explicitly modeling the time delay of spatial information propagation. Extensive experimental results on six real-world public traffic datasets show that our method can not only achieve state-of-the-art performance but also exhibit competitive computational efficiency. Moreover, we visualize the learned spatial-temporal attention map to make our model highly interpretable. | ['Jingyuan Wang', 'Wayne Xin Zhao', 'Chengkai Han', 'Jiawei Jiang'] | 2023-01-19 | null | null | null | null | ['time-series-prediction'] | ['time-series'] | [-1.10560037e-01 -5.17146230e-01 -5.34075558e-01 -3.97728801e-01
4.26733587e-03 -2.85093606e-01 5.32341361e-01 -1.51275918e-01
-8.19526240e-02 4.36887622e-01 1.71125144e-01 -8.95888805e-01
-5.39540112e-01 -1.16686583e+00 -5.85683703e-01 -3.93373162e-01
-8.34234357e-02 2.06794411e-01 7.46655226e-01 -5.21239519e-01
-2.59054750e-02 7.31033802e-01 -1.36002207e+00 -9.10280570e-02
1.20501351e+00 9.72583234e-01 2.12861732e-01 4.41883087e-01
-3.21106076e-01 9.14272189e-01 -2.94888645e-01 -2.29994968e-01
2.23306343e-01 -1.26481161e-01 -3.59498024e-01 -1.49196759e-01
2.51003176e-01 -3.19602042e-01 -1.29633689e+00 8.32455337e-01
1.58528626e-01 3.29803050e-01 3.18932772e-01 -1.58790660e+00
-8.04208517e-01 3.10353011e-01 -6.18904769e-01 7.91824162e-01
-5.94151281e-02 6.07459009e-01 8.69710207e-01 -3.88652503e-01
1.70063332e-01 1.32627058e+00 2.96307981e-01 2.48214573e-01
-9.99144912e-01 -7.58601606e-01 7.64486730e-01 7.98212528e-01
-1.38636184e+00 -2.10827559e-01 1.01904833e+00 -3.91920984e-01
8.85062873e-01 3.66376281e-01 7.27177739e-01 6.67325735e-01
2.95760393e-01 6.67014301e-01 5.75848222e-01 1.89713076e-01
-2.04486758e-01 -2.63780653e-01 2.40485758e-01 6.00094438e-01
-2.27898620e-02 2.98276633e-01 -3.26276451e-01 3.81324291e-01
6.87573671e-01 4.06463385e-01 -1.04510836e-01 -3.21835354e-02
-9.15848732e-01 5.39645970e-01 1.14941716e+00 1.99419558e-01
-3.33053976e-01 4.38322693e-01 2.79310018e-01 4.76101153e-02
4.37660694e-01 -2.70213842e-01 -9.53920931e-02 -1.81301743e-01
-4.39936429e-01 -1.07851531e-02 1.22712880e-01 1.06020796e+00
9.66531038e-01 2.53486335e-01 -2.66518176e-01 5.93333781e-01
2.65932977e-01 8.34488094e-01 3.30496579e-02 -6.09335423e-01
1.03788447e+00 8.85112345e-01 -2.42015645e-01 -1.68488336e+00
-6.07061863e-01 -5.00020683e-01 -1.00298727e+00 -1.55470386e-01
3.63124669e-01 1.19579667e-02 -8.64114523e-01 1.73311114e+00
1.63454652e-01 6.80331707e-01 -4.26636904e-01 8.97732139e-01
6.00217521e-01 9.51268673e-01 2.59412885e-01 2.87693720e-02
1.02623117e+00 -1.00458288e+00 -8.66912961e-01 -4.15538490e-01
5.64209163e-01 -2.45208025e-01 1.02499449e+00 -4.53253001e-01
-7.65791178e-01 -6.37670815e-01 -6.60213828e-01 4.67942841e-02
-6.30537391e-01 -2.07802609e-01 6.07955933e-01 4.35232967e-01
-6.97682202e-01 1.00723952e-01 -7.95742631e-01 -8.93993825e-02
4.55076009e-01 2.81268865e-01 1.13216192e-02 -3.03335756e-01
-1.59301174e+00 6.99508190e-01 1.32696316e-01 4.89936858e-01
-4.23270136e-01 -8.93520832e-01 -8.97137284e-01 3.32987040e-01
6.09084725e-01 -4.41816956e-01 7.04946578e-01 -5.40503800e-01
-1.08989155e+00 5.21913469e-02 -5.05222619e-01 -3.66128951e-01
5.09586096e-01 1.57912701e-01 -1.23931205e+00 -7.63486922e-02
1.83405370e-01 2.99193799e-01 3.57124746e-01 -9.92165983e-01
-8.72256637e-01 -1.78130791e-01 2.46415704e-01 -1.59949455e-02
-3.33117068e-01 -2.75310367e-01 -9.02710974e-01 -6.75242782e-01
-1.30423047e-02 -8.21932673e-01 -2.97017157e-01 -5.39411642e-02
-5.08426666e-01 -1.34354293e-01 1.18583274e+00 -5.46875179e-01
1.92724490e+00 -2.08159518e+00 -3.71216893e-01 5.60401201e-01
5.12022913e-01 5.77193260e-01 -2.73044914e-01 4.63843375e-01
-1.85713787e-02 1.39696270e-01 -6.61370978e-02 1.27777472e-01
1.57407727e-02 4.06018972e-01 -3.76889199e-01 1.48065045e-01
1.98891923e-01 1.31681359e+00 -1.02569783e+00 -2.60755152e-01
7.62592494e-01 5.84846139e-01 -5.08971632e-01 -8.45262632e-02
-1.42845005e-01 6.23851836e-01 -7.27243543e-01 1.49273261e-01
8.22111309e-01 -3.67178768e-01 -1.15860872e-01 -2.67824322e-01
-3.03396136e-01 2.83575833e-01 -9.33159292e-01 1.03067052e+00
-6.95062637e-01 8.97530079e-01 -4.13938284e-01 -9.10324335e-01
8.08325708e-01 -1.09902639e-02 4.90934342e-01 -1.47637105e+00
-8.07378292e-02 -9.46515724e-02 1.59029484e-01 -6.42639041e-01
3.08469951e-01 2.46070385e-01 5.21515235e-02 2.67919719e-01
-5.50301075e-01 7.53612518e-01 4.76417124e-01 2.86386579e-01
1.20837617e+00 -4.38994437e-01 -2.84168571e-01 9.09604430e-02
8.28301907e-01 -2.34274730e-01 8.27247679e-01 3.05389136e-01
-4.24568534e-01 3.25975031e-01 6.66801274e-01 -7.80065179e-01
-4.61346656e-01 -1.04222250e+00 1.97027162e-01 9.97182012e-01
5.50585985e-01 -2.64710307e-01 -4.26612973e-01 -7.78475344e-01
1.44618064e-01 6.75018609e-01 -7.09891379e-01 -4.30778921e-01
-9.49689806e-01 -5.05555630e-01 2.58407027e-01 6.55308545e-01
7.90833533e-01 -6.49045408e-01 -6.64515197e-02 3.51589531e-01
-3.64469588e-01 -1.47382724e+00 -8.91015947e-01 -6.29724979e-01
-4.26579177e-01 -9.96814013e-01 -3.87338996e-01 -5.05174220e-01
6.05529189e-01 8.83572578e-01 7.61558175e-01 3.49239826e-01
-2.78860014e-02 3.02859973e-02 -1.04780927e-01 -2.04666108e-02
4.75752279e-02 2.47817084e-01 -2.50162363e-01 5.68702221e-01
5.51444829e-01 -8.12022686e-01 -8.54506791e-01 8.08522284e-01
-7.23686159e-01 2.00028956e-01 5.73114514e-01 3.88359547e-01
4.54614788e-01 4.16725397e-01 7.61815190e-01 -6.83751225e-01
4.55200613e-01 -7.20789611e-01 -6.54650509e-01 2.30687141e-01
-5.57853520e-01 -1.44558907e-01 9.80818510e-01 -2.68822104e-01
-9.90914285e-01 -3.98933589e-01 -1.99638233e-01 -4.39249754e-01
-1.23350404e-01 6.82655215e-01 -5.46251774e-01 8.11333731e-02
1.44832566e-01 4.03681368e-01 -1.68910086e-01 -2.46345773e-01
3.08989257e-01 3.12189162e-01 3.40915412e-01 -2.47235149e-01
1.03646755e+00 5.03726363e-01 3.49855930e-01 -8.11178923e-01
-5.83510518e-01 -3.50906789e-01 -6.65492594e-01 -5.55405498e-01
7.50468552e-01 -6.69159472e-01 -1.05397165e+00 3.96228939e-01
-9.71577704e-01 -4.14591432e-01 4.20150906e-02 3.97156686e-01
-2.28023663e-01 2.35133559e-01 -3.71259928e-01 -5.55693030e-01
2.38178551e-01 -1.09188676e+00 6.53075635e-01 3.38685244e-01
4.47629213e-01 -1.34045219e+00 -2.06520155e-01 3.71278644e-01
7.11949229e-01 7.26946443e-02 1.18823040e+00 -1.37877420e-01
-1.14315295e+00 -1.35104895e-01 -8.86124015e-01 -5.26719019e-02
4.18789923e-01 -5.32484986e-03 -4.94540691e-01 1.00064121e-01
-7.90269494e-01 6.55512214e-01 9.05427635e-01 5.16629577e-01
1.38590801e+00 -4.12851840e-01 -5.99367559e-01 6.73719823e-01
1.06527495e+00 2.51556069e-01 7.45885015e-01 3.30779254e-02
1.23454881e+00 7.67567456e-01 4.47238415e-01 1.32219866e-01
9.95576084e-01 8.64481807e-01 5.14061272e-01 -2.11268663e-01
-4.47010219e-01 -6.35402858e-01 7.66697200e-03 7.79171884e-01
-6.01418652e-02 -6.11753941e-01 -1.18518245e+00 4.97674972e-01
-2.18594956e+00 -1.23770106e+00 -5.96918046e-01 1.99770963e+00
-1.32350507e-03 4.22810823e-01 3.37174326e-01 2.37934049e-02
8.19890201e-01 5.16673386e-01 -6.60369992e-01 1.69342346e-02
-5.83891496e-02 -3.26275766e-01 6.86898768e-01 6.36778176e-01
-8.14098656e-01 1.10456920e+00 5.18623638e+00 8.96978438e-01
-1.30173159e+00 4.57688747e-03 5.15186310e-01 -2.73777153e-02
-4.62378174e-01 -2.25092694e-01 -5.97235084e-01 1.05124617e+00
1.00340748e+00 -2.50963748e-01 5.23083687e-01 3.78841400e-01
7.89563298e-01 2.44200945e-01 -6.14716768e-01 8.58670175e-01
-3.93693388e-01 -1.41530228e+00 2.61092275e-01 2.32323200e-01
4.28988725e-01 8.98035094e-02 3.36726964e-01 3.70797455e-01
1.77022249e-01 -9.72826600e-01 3.69875789e-01 5.93130589e-01
7.05586553e-01 -1.00755990e+00 3.07858139e-01 2.40400985e-01
-1.93471313e+00 -2.99201757e-01 -1.13760404e-01 -1.40716076e-01
7.66945302e-01 6.01564825e-01 -2.40378857e-01 6.27608001e-01
4.39556748e-01 1.22483146e+00 -8.35410595e-01 1.18625557e+00
-2.15575457e-01 7.41716743e-01 -2.14905933e-01 8.54705200e-02
4.79237199e-01 -2.82703876e-01 4.64850605e-01 1.00886893e+00
1.91732109e-01 1.47448868e-01 1.57990977e-01 8.68888974e-01
1.31706223e-01 -1.91616222e-01 -6.77743673e-01 2.27784187e-01
6.56490862e-01 8.70083928e-01 -5.06984770e-01 -2.57598516e-03
-8.00366938e-01 4.46409315e-01 2.56641716e-01 8.81907344e-01
-1.26501870e+00 -4.30566221e-01 1.03160834e+00 6.00292742e-01
3.58579874e-01 -5.29139340e-01 -1.79209292e-01 -9.42518055e-01
1.67786583e-01 -2.43674144e-01 2.99098730e-01 -5.76961994e-01
-1.15247798e+00 4.98228163e-01 -2.46657841e-02 -1.42343843e+00
1.79433495e-01 -4.65440303e-01 -9.16998684e-01 7.92884886e-01
-1.95664310e+00 -1.42414367e+00 -5.44540465e-01 7.98181593e-01
4.04858857e-01 -1.60481147e-02 -2.30206014e-03 8.94004166e-01
-1.11752367e+00 6.25391126e-01 -1.52246490e-01 4.06118304e-01
1.95245892e-01 -7.71909416e-01 7.13904500e-01 1.04994476e+00
-1.00005597e-01 4.23648208e-01 2.87424952e-01 -5.96092224e-01
-1.29393840e+00 -1.64637673e+00 8.83921266e-01 -3.56485754e-01
7.97306776e-01 -2.45399013e-01 -1.07205045e+00 7.95551777e-01
-2.53162891e-01 4.65076834e-01 4.81523693e-01 -4.89287684e-03
-3.82798672e-01 -6.68856502e-01 -5.53990364e-01 8.04696083e-01
1.39600849e+00 -5.09241283e-01 5.85435629e-02 1.99116930e-01
7.52794147e-01 -1.72134191e-01 -4.18791443e-01 3.34317088e-01
4.18675065e-01 -8.61440241e-01 9.98499870e-01 -4.62131500e-01
1.05684705e-01 -5.90064585e-01 4.60269190e-02 -1.21556020e+00
-6.36224389e-01 -4.33683425e-01 -3.86608332e-01 1.25319052e+00
4.86246973e-01 -1.16631782e+00 6.83887661e-01 7.68813133e-01
-1.78340644e-01 -7.79628217e-01 -9.35729682e-01 -8.34679008e-01
-1.59075052e-01 -7.79146791e-01 1.13277316e+00 7.32904375e-01
-3.89344364e-01 4.82345521e-01 -5.03493130e-01 3.67497891e-01
3.61080974e-01 1.22157812e-01 8.31834316e-01 -1.16026783e+00
2.65522838e-01 -8.04043770e-01 -7.26768434e-01 -1.61156535e+00
3.42076480e-01 -7.24995196e-01 -3.39121521e-01 -1.64675260e+00
-7.73826912e-02 -7.03881562e-01 -4.37981993e-01 2.01070577e-01
-3.00383449e-01 -1.13126986e-01 9.60452333e-02 4.13094535e-02
-6.30028367e-01 8.41094792e-01 1.42259991e+00 -2.32115731e-01
-1.83050498e-01 3.21990877e-01 -5.75323462e-01 3.79299164e-01
6.99178576e-01 -2.17337221e-01 -8.05560887e-01 -8.07788908e-01
1.39793754e-01 4.85594347e-02 5.74995160e-01 -1.05135417e+00
5.28584957e-01 -5.99552989e-01 -5.66385575e-02 -8.64195168e-01
1.99326366e-01 -1.12646306e+00 2.01891810e-02 2.28502288e-01
-6.92588165e-02 3.07707518e-01 2.35701069e-01 1.05802095e+00
-2.85873771e-01 7.02248633e-01 3.65699619e-01 3.83325666e-01
-1.08966660e+00 1.12426436e+00 -5.60491920e-01 3.94634977e-02
1.15302908e+00 -2.62358010e-01 -5.91132104e-01 -5.39261699e-01
-3.83154005e-01 8.31633091e-01 1.60717890e-01 8.66369188e-01
5.50806999e-01 -1.67217827e+00 -4.38956112e-01 4.11355764e-01
2.18441829e-01 -1.48615271e-01 9.25190032e-01 9.56432641e-01
-4.26947951e-01 8.18148613e-01 -5.77718988e-02 -6.37290180e-01
-7.94464231e-01 8.27000618e-01 3.82101625e-01 -2.84388751e-01
-6.24659121e-01 3.66774648e-01 4.80745584e-01 -4.44700778e-01
-5.57780899e-02 -3.73360217e-01 -2.78264046e-01 -2.50528038e-01
6.52206719e-01 6.28734589e-01 -8.60960633e-02 -1.18082690e+00
-4.49749112e-01 7.62476623e-01 2.36295372e-01 2.90907085e-01
1.05002975e+00 -7.14488626e-01 1.95921257e-01 1.72967732e-01
1.21876431e+00 -1.37223274e-01 -1.43606889e+00 -6.09583974e-01
-2.70504624e-01 -8.45784545e-01 2.11648092e-01 -5.74508429e-01
-1.70196855e+00 1.29020679e+00 3.52406979e-01 3.94015640e-01
1.23619914e+00 -3.48861337e-01 1.28314066e+00 1.96585000e-01
1.63694307e-01 -8.24665904e-01 -1.09739944e-01 7.85967827e-01
5.78995109e-01 -1.20460582e+00 -4.47236568e-01 -7.44451284e-01
-4.49765444e-01 9.24524069e-01 9.78042424e-01 2.10459813e-01
1.05351686e+00 -1.19215980e-01 -1.41761601e-01 -2.17150837e-01
-7.55026817e-01 -6.16266608e-01 6.41881108e-01 6.54602170e-01
5.04095592e-02 6.77694455e-02 1.84050962e-01 3.03278714e-01
1.05032593e-01 -1.83032319e-01 4.70171086e-02 5.51831245e-01
-2.57581681e-01 -8.20329964e-01 1.09026149e-01 4.10315245e-01
1.55128703e-01 7.62765408e-02 5.15140221e-02 7.95836687e-01
9.66443215e-03 1.16484320e+00 2.61578351e-01 -8.01635921e-01
6.72846913e-01 -4.25555408e-01 -9.64279845e-02 -1.99485898e-01
-1.01230383e-01 -3.02580625e-01 -1.58107877e-01 -7.72676170e-01
-5.66408224e-02 -3.78179073e-01 -1.26572096e+00 -9.43820059e-01
-5.74357174e-02 -5.29516414e-02 2.32604474e-01 1.06263864e+00
8.69993508e-01 1.00653136e+00 9.19727802e-01 -5.87218106e-01
3.83970201e-01 -4.46112156e-01 -4.03181523e-01 4.35290456e-01
6.65350914e-01 -8.80799353e-01 1.78194530e-02 -2.69577384e-01] | [6.471134662628174, 2.050156354904175] |
b4b90d5c-316c-4b03-afa9-488bf48b5a68 | wasserstein-kelly-portfolios-a-robust-data | 2302.13979 | null | https://arxiv.org/abs/2302.13979v1 | https://arxiv.org/pdf/2302.13979v1.pdf | Wasserstein-Kelly Portfolios: A Robust Data-Driven Solution to Optimize Portfolio Growth | We introduce a robust variant of the Kelly portfolio optimization model, called the Wasserstein-Kelly portfolio optimization. Our model, taking a Wasserstein distributionally robust optimization (DRO) formulation, addresses the fundamental issue of estimation error in Kelly portfolio optimization by defining a ``ball" of distributions close to the empirical return distribution using the Wasserstein metric and seeking a robust log-optimal portfolio against the worst-case distribution from the Wasserstein ball. Enhancing the Kelly portfolio using Wasserstein DRO is a natural step to take, given many successful applications of the latter in areas such as machine learning for generating robust data-driven solutions. However, naive application of Wasserstein DRO to the growth-optimal portfolio problem can lead to several issues, which we resolve through careful modelling. Our proposed model is both practically motivated and efficiently solvable as a convex program. Using empirical financial data, our numerical study demonstrates that the Wasserstein-Kelly portfolio can outperform the Kelly portfolio in out-of-sample testing across multiple performance metrics and exhibits greater stability. | ['Jonathan Yu-Meng Li'] | 2023-02-27 | null | null | null | null | ['portfolio-optimization'] | ['time-series'] | [ 9.15316399e-03 8.94894302e-02 5.25469072e-02 -2.63108075e-01
-1.26239192e+00 -6.46336734e-01 1.04726955e-01 -5.58369868e-02
-3.69426101e-01 1.01555586e+00 -1.01661764e-01 -7.36198008e-01
-1.03517783e+00 -7.16352642e-01 -7.94363976e-01 -9.26566243e-01
-1.87614664e-01 2.96575159e-01 -1.88757345e-01 -5.85705936e-02
6.18614793e-01 4.80760932e-01 -1.10966992e+00 -4.22817409e-01
9.72174764e-01 1.37275207e+00 -1.45615175e-01 7.13444412e-01
1.96738496e-01 4.38006461e-01 -5.86003959e-01 -5.80129623e-01
8.12012911e-01 -2.62301981e-01 -2.36730486e-01 5.39262742e-02
1.52027622e-01 -2.19395012e-01 2.75858313e-01 1.32080197e+00
5.85737228e-01 4.36153293e-01 9.17780459e-01 -1.34491348e+00
-8.65402520e-01 3.66980940e-01 -8.78955483e-01 3.29701185e-01
-9.49982181e-02 -5.13815172e-02 1.28981328e+00 -9.88175750e-01
5.37344888e-02 1.14323294e+00 8.64620745e-01 1.14204012e-01
-1.10437953e+00 -4.28506613e-01 -1.33678690e-01 -3.64424795e-01
-1.12154305e+00 -4.12173383e-03 5.34753442e-01 -7.03942955e-01
5.71545959e-01 5.99649668e-01 3.24818730e-01 6.17088854e-01
4.67773646e-01 7.18910515e-01 1.31045389e+00 -4.38272834e-01
3.32775593e-01 2.93995261e-01 3.49118002e-02 4.56942350e-01
6.97562158e-01 4.12095189e-01 -2.91819960e-01 -6.47875428e-01
8.29606414e-01 -1.32528096e-01 -3.03734481e-01 -5.38270891e-01
-7.51917422e-01 1.01757598e+00 -2.21880123e-01 -1.38903633e-01
-3.36146832e-01 4.42754358e-01 -6.74490258e-02 5.58768690e-01
9.52558994e-01 7.30011582e-01 -3.76754910e-01 -1.87510177e-01
-9.90035772e-01 8.38784397e-01 9.56635118e-01 8.52939069e-01
3.12314749e-01 1.82196289e-01 -5.04489601e-01 5.23920238e-01
9.45244372e-01 8.18585873e-01 1.67424038e-01 -9.25454795e-01
8.55232000e-01 2.64576465e-01 6.99966013e-01 -8.29847455e-01
-4.24532108e-02 -5.34783185e-01 -3.41533691e-01 4.05781269e-01
7.60394692e-01 -4.73814130e-01 -9.77703705e-02 1.48274040e+00
4.12133217e-01 2.11071029e-01 -3.15415524e-02 5.59343278e-01
-3.39618206e-01 4.36318070e-01 -4.21490997e-01 -5.48640430e-01
8.29093695e-01 -6.75245702e-01 -3.62503767e-01 8.42301846e-02
4.29888308e-01 -7.14240372e-01 9.83253777e-01 3.32831383e-01
-1.09340703e+00 5.68985790e-02 -1.17694092e+00 6.39418542e-01
3.71181555e-02 -4.17706043e-01 2.20881596e-01 1.13888168e+00
-1.06971991e+00 1.16034603e+00 -3.72992754e-01 1.05846301e-01
2.77293295e-01 2.18226343e-01 2.01764777e-02 1.76343530e-01
-8.21918249e-01 8.24638009e-01 2.96845257e-01 2.78786689e-01
-7.25113571e-01 -9.55536783e-01 -7.74894476e-01 9.47926193e-02
5.25455058e-01 -3.71146113e-01 1.20773661e+00 -5.46958506e-01
-1.59186804e+00 3.56284499e-01 4.00485367e-01 -5.76507270e-01
1.06305420e+00 -4.03312951e-01 7.07323179e-02 -1.90878451e-01
-5.44595979e-02 -3.70572329e-01 1.08510101e+00 -6.90111578e-01
-4.18938279e-01 -2.31867701e-01 -2.03034043e-01 -1.64946109e-01
-3.33824456e-01 1.54600829e-01 4.14653748e-01 -1.16285241e+00
-2.85371274e-01 -7.55688190e-01 -3.85635823e-01 -1.47748321e-01
-3.10065359e-01 -2.58655071e-01 2.30307132e-01 -3.68706703e-01
1.30818892e+00 -1.94054067e+00 -1.52573094e-01 6.24717057e-01
-5.09934649e-02 -1.89599901e-01 -6.76369444e-02 4.65994895e-01
-2.08950624e-01 3.59873623e-01 -3.57989520e-01 -2.17982471e-01
5.63024580e-01 -1.62404865e-01 -7.09513187e-01 9.56533551e-01
2.91608155e-01 8.83286119e-01 -1.05788374e+00 -1.23806164e-01
-1.53377131e-01 -2.66254023e-02 -6.61912739e-01 3.82854342e-01
-8.79865512e-02 -7.19010010e-02 -6.92691028e-01 6.22829258e-01
6.06074333e-01 -1.98050871e-01 -2.22807094e-01 3.59477729e-01
-1.03587456e-01 -3.71000797e-01 -1.45174253e+00 6.59516990e-01
-1.51962608e-01 5.36329448e-01 -1.37625709e-01 -1.17704344e+00
1.07603455e+00 8.41463208e-02 6.50409281e-01 -1.54820278e-01
2.42375195e-01 5.75176477e-01 -3.84851187e-01 -2.99026459e-01
5.99999249e-01 -9.29402053e-01 -1.29652724e-01 9.88083005e-01
-1.61969125e-01 -7.87981749e-02 1.65253311e-01 -4.37464714e-01
9.89774406e-01 2.18643084e-01 2.87047416e-01 -9.55893040e-01
1.14592046e-01 -3.46018940e-01 5.46681881e-01 9.96888399e-01
-3.57674509e-01 6.04128242e-01 9.26212966e-01 2.48738546e-02
-9.65591550e-01 -1.50569832e+00 -4.51588839e-01 7.72370875e-01
-2.29479373e-01 2.17301890e-01 -5.06289303e-01 -6.88377678e-01
6.27660215e-01 7.28321135e-01 -7.94418037e-01 -1.35908589e-01
-1.79388642e-01 -1.14924300e+00 4.05539960e-01 5.15083969e-01
2.41854072e-01 -6.18672907e-01 -6.15072072e-01 4.82839972e-01
4.53661054e-01 -3.00833702e-01 -9.25293088e-01 2.62365341e-02
-8.24225724e-01 -1.39731443e+00 -1.44767332e+00 -4.22902070e-02
4.54207271e-01 -1.40347764e-01 1.04374015e+00 -4.10683244e-01
-2.12005019e-01 8.42405736e-01 -1.98816851e-01 -8.50359738e-01
-1.53484046e-01 -5.83704412e-01 1.92354143e-01 4.50196326e-01
1.47239402e-01 -1.96305692e-01 -6.41556978e-01 5.08141756e-01
-7.95216739e-01 -8.51996481e-01 3.44785273e-01 1.03382707e+00
6.69267952e-01 2.04161674e-01 1.05031025e+00 -8.00769687e-01
1.17073810e+00 -7.97989547e-01 -1.08094168e+00 4.98987943e-01
-9.33326304e-01 3.65992427e-01 1.51576206e-01 -5.67738533e-01
-7.78258860e-01 -7.39335299e-01 4.03210610e-01 -4.06487852e-01
8.06995153e-01 5.41091383e-01 1.72326297e-01 -2.36401588e-01
4.20650899e-01 -7.31577575e-02 1.52001962e-01 -4.48005468e-01
1.05399288e-01 6.04720891e-01 2.36226201e-01 -1.21825624e+00
9.70570564e-01 1.86854690e-01 2.00804219e-01 -5.83776414e-01
-1.17361808e+00 -4.36897695e-01 -1.28851950e-01 -2.67256945e-01
5.57518423e-01 -2.19902515e-01 -7.84877002e-01 2.47445658e-01
-6.32522523e-01 -4.78295922e-01 -8.63795877e-01 5.05645275e-01
-1.14457011e+00 1.98407039e-01 -1.97106734e-01 -1.49930227e+00
-3.47011775e-01 -1.00827920e+00 9.44155216e-01 1.60675645e-01
-1.52669683e-01 -1.60606098e+00 5.92459381e-01 1.02250874e-01
4.66863841e-01 9.66649890e-01 7.90569067e-01 -6.29288912e-01
-3.44100058e-01 -5.19306660e-01 -1.22972824e-01 8.24632823e-01
-1.61863834e-01 7.62853324e-02 -7.37919092e-01 -5.49203396e-01
6.08170688e-01 -1.45381227e-01 4.32370961e-01 5.28801799e-01
9.52847540e-01 -4.49561626e-01 1.81498513e-01 6.35816276e-01
1.57235610e+00 9.88219902e-02 1.91109061e-01 6.96009219e-01
1.71433508e-01 7.02219844e-01 9.64079082e-01 8.00810516e-01
5.10573648e-02 7.07839727e-02 2.86469251e-01 5.82233191e-01
8.98475587e-01 -6.78392723e-02 6.46081269e-01 5.35048604e-01
1.69754893e-01 -8.46342966e-02 -8.03890169e-01 4.76262867e-01
-1.97780466e+00 -1.01075053e+00 1.21418774e-01 2.61737299e+00
7.24102259e-01 1.99810803e-01 2.89552182e-01 -4.31004167e-02
6.19043827e-01 -4.57256474e-02 -7.27449656e-01 -5.79887033e-01
-3.04148477e-02 3.33547562e-01 9.50862110e-01 5.35049975e-01
-8.74561191e-01 -1.87986270e-02 7.14798832e+00 9.27779436e-01
-6.39348567e-01 4.16498557e-02 8.39788914e-01 -3.11213821e-01
-6.18756175e-01 -2.07511619e-01 -8.58962476e-01 5.83395660e-01
9.69138443e-01 -1.05667806e+00 2.03245074e-01 1.05012250e+00
3.37579519e-01 -1.61700917e-03 -1.19796813e+00 7.88941383e-01
-1.86531290e-01 -1.21632981e+00 -4.52469468e-01 5.61234772e-01
1.06000245e+00 -2.34138459e-01 5.60234189e-01 3.58160287e-01
4.76385027e-01 -1.28146243e+00 9.68844056e-01 9.09516633e-01
6.03478789e-01 -1.09156907e+00 9.18175995e-01 3.26159537e-01
-9.80851114e-01 -1.21159784e-01 -6.92477047e-01 1.54591426e-01
3.61096188e-02 9.29873407e-01 -5.74554741e-01 4.40288097e-01
6.38197660e-01 4.74194586e-01 -2.12565288e-01 1.39288354e+00
3.56710941e-01 5.30868828e-01 -4.19546694e-01 -1.53001428e-01
4.09896851e-01 -7.13442624e-01 7.51651883e-01 8.42177987e-01
8.15346956e-01 -2.25418881e-01 -8.94616172e-02 9.71202374e-01
3.62384170e-02 2.83138573e-01 -6.33583426e-01 -2.44890198e-01
2.28018567e-01 8.33777249e-01 -4.59689468e-01 6.03170209e-02
-1.49287060e-01 2.68638879e-01 2.39189431e-01 4.56508219e-01
-6.66325331e-01 -6.19891286e-01 7.40285218e-01 -5.67819960e-02
2.56390870e-01 -1.00045323e-01 -4.24430788e-01 -8.46186638e-01
3.69494677e-01 -6.04352891e-01 5.57334781e-01 -4.23017561e-01
-1.65582740e+00 1.39594704e-01 3.62514764e-01 -1.22593391e+00
-3.16167951e-01 -1.04423201e+00 -8.79368961e-01 1.12508261e+00
-1.50455987e+00 -3.52860004e-01 3.50773126e-01 1.99546948e-01
4.12781537e-02 -3.38825345e-01 2.93152511e-01 -1.59262065e-02
-6.51236951e-01 7.61858284e-01 9.69674587e-01 -4.09890920e-01
6.48940384e-01 -1.89455163e+00 -1.08430013e-02 7.97294080e-01
-1.96531400e-01 5.57969451e-01 9.93767679e-01 -5.97840846e-01
-1.35338783e+00 -1.17855823e+00 3.82389843e-01 -7.77776182e-01
1.39735425e+00 1.02655105e-01 -7.13142931e-01 6.36969090e-01
-1.31093308e-01 -8.21168423e-02 9.20746267e-01 -1.59324974e-01
3.66456658e-02 -1.07339643e-01 -1.49395669e+00 7.05161452e-01
5.21521330e-01 -2.72743553e-01 -6.17877901e-01 5.74700594e-01
5.09088099e-01 -9.27919447e-02 -1.39343917e+00 2.33406156e-01
5.74955523e-01 -8.59206200e-01 7.64770091e-01 -7.37511039e-01
9.60629284e-02 7.52489567e-02 -3.24409276e-01 -1.38346422e+00
8.30802768e-02 -1.43126512e+00 -1.83758408e-01 1.25849342e+00
2.99307823e-01 -1.15688002e+00 7.49622762e-01 6.63249493e-01
2.77782947e-01 -1.38323534e+00 -1.12119985e+00 -1.33118212e+00
6.14513636e-01 -7.66382337e-01 6.79060459e-01 5.03700018e-01
-9.57995504e-02 -6.34602368e-01 -4.01714951e-01 2.24280190e-02
1.34095883e+00 5.23507744e-02 5.67168474e-01 -9.82636213e-01
-7.75689900e-01 -7.63738990e-01 -1.61969379e-01 -7.97246635e-01
8.61764029e-02 -8.21950674e-01 2.20363945e-01 -8.93613636e-01
1.05227027e-02 -4.73684639e-01 -5.63031256e-01 -1.50668085e-01
-3.74354899e-01 6.72305971e-02 1.18197009e-01 -6.54888824e-02
-3.11793268e-01 7.57645249e-01 1.17409420e+00 6.86736032e-02
3.19248922e-02 4.61476386e-01 -1.12731981e+00 7.33294725e-01
6.74707890e-01 -4.78918046e-01 -4.64210629e-01 2.38251109e-02
5.38224220e-01 1.65549144e-01 1.54060885e-01 -3.50268364e-01
-1.86262894e-02 -6.27449274e-01 4.01410274e-02 -5.04274905e-01
-6.04295209e-02 -5.09801269e-01 -2.18577520e-03 2.62289315e-01
-1.65570006e-01 4.61347193e-01 -5.13933711e-02 9.83233750e-01
2.31014360e-02 -7.68059134e-01 9.23167229e-01 1.46843702e-01
-4.70980862e-03 3.86606902e-01 7.34904315e-03 6.60522342e-01
1.31721067e+00 -2.33904511e-01 -1.63674548e-01 -3.78333539e-01
-5.19597232e-01 5.40632844e-01 2.24553168e-01 -2.75192820e-02
5.73946834e-01 -1.48243344e+00 -8.43242824e-01 7.13224709e-02
-1.44116059e-01 -1.49298683e-01 -1.89533934e-01 1.04564011e+00
-5.56858718e-01 3.77486140e-01 3.68218899e-01 -3.66614074e-01
-5.54009438e-01 1.76114202e-01 6.52177215e-01 -6.40905380e-01
-3.58426362e-01 1.04096508e+00 6.09119087e-02 -2.58018017e-01
1.53847635e-01 -4.86500353e-01 2.24675819e-01 9.09892619e-02
6.98858142e-01 9.78032291e-01 -1.74770221e-01 -1.04036100e-01
-4.95170280e-02 5.31168878e-01 2.77873844e-01 -3.48892152e-01
1.45880663e+00 4.75511551e-02 -9.38139036e-02 5.72762549e-01
1.19837892e+00 6.58830926e-02 -1.67849815e+00 -7.63404146e-02
7.53912032e-01 -7.74299502e-01 -1.97438270e-01 -2.33121261e-01
-7.20790744e-01 5.06472468e-01 2.85512447e-01 5.19633532e-01
6.93945050e-01 -3.37168276e-01 4.47410792e-01 2.86432058e-01
3.81397426e-01 -1.19178045e+00 1.42077759e-01 3.54578197e-01
1.18821692e+00 -1.07675946e+00 4.24359664e-02 2.05615714e-01
-4.92600143e-01 1.13121367e+00 1.21697031e-01 -8.24120700e-01
1.23163271e+00 3.17189634e-01 -1.12703331e-01 4.74005528e-02
-6.18900716e-01 2.18611091e-01 5.91137588e-01 3.97038281e-01
6.41652718e-02 1.46715045e-01 -2.86311179e-01 7.07480192e-01
-4.49077338e-01 -3.87141258e-01 5.68238556e-01 9.63602483e-01
-5.46203494e-01 -8.55498314e-01 -6.56083405e-01 7.56139636e-01
-8.96266222e-01 5.89831583e-02 1.07601777e-01 8.69164765e-01
-5.62703907e-01 8.43477428e-01 1.70916736e-01 1.85217232e-01
5.26789844e-01 1.06774248e-01 4.56672519e-01 -6.42044723e-01
-2.88935810e-01 1.76761463e-01 -2.95321405e-01 -3.87046754e-01
-1.74417809e-01 -1.02871609e+00 -3.97433758e-01 -4.71290082e-01
-5.11491239e-01 3.87145698e-01 3.63265395e-01 1.17284524e+00
-2.41204783e-01 2.93337882e-01 1.03504133e+00 -7.12526798e-01
-1.93710899e+00 -8.28734457e-01 -1.20484078e+00 -2.00122632e-02
3.82569551e-01 -6.90154135e-01 -8.26508164e-01 -5.53435266e-01] | [5.113101959228516, 3.8997154235839844] |
68498880-8440-424a-aeb4-4effa533c64b | zero-shot-text-to-speech-synthesis | 2304.11976 | null | https://arxiv.org/abs/2304.11976v1 | https://arxiv.org/pdf/2304.11976v1.pdf | Zero-shot text-to-speech synthesis conditioned using self-supervised speech representation model | This paper proposes a zero-shot text-to-speech (TTS) conditioned by a self-supervised speech-representation model acquired through self-supervised learning (SSL). Conventional methods with embedding vectors from x-vector or global style tokens still have a gap in reproducing the speaker characteristics of unseen speakers. A novel point of the proposed method is the direct use of the SSL model to obtain embedding vectors from speech representations trained with a large amount of data. We also introduce the separate conditioning of acoustic features and a phoneme duration predictor to obtain the disentangled embeddings between rhythm-based speaker characteristics and acoustic-feature-based ones. The disentangled embeddings will enable us to achieve better reproduction performance for unseen speakers and rhythm transfer conditioned by different speeches. Objective and subjective evaluations showed that the proposed method can synthesize speech with improved similarity and achieve speech-rhythm transfer. | ['Yusuke Ijima', 'Takafumi Moriya', 'Hiroki Kanagawa', 'Takanori Ashihara', 'Kenichi Fujita'] | 2023-04-24 | null | null | null | null | ['text-to-speech-synthesis', 'speech-synthesis'] | ['speech', 'speech'] | [ 1.58304766e-01 3.45815881e-03 -1.83764711e-01 -4.22117084e-01
-8.05094659e-01 -1.92281738e-01 6.69532478e-01 -2.20562547e-01
2.25314070e-02 6.30964100e-01 6.37244344e-01 2.03637648e-02
1.39099583e-01 -5.19709826e-01 -3.07044387e-01 -9.59835470e-01
-9.70624238e-02 1.31984517e-01 -2.24750668e-01 -2.54389137e-01
-9.40808877e-02 3.36253941e-01 -1.93986416e+00 -5.89426905e-02
8.33386242e-01 6.94273353e-01 4.95954663e-01 9.72503006e-01
-1.61802351e-01 1.69742256e-01 -8.18856239e-01 -7.02928081e-02
3.17106843e-02 -7.08964825e-01 3.68796401e-02 4.82991673e-02
6.07210305e-03 -9.77281947e-03 -6.15054369e-01 6.32630825e-01
1.01910889e+00 2.59411961e-01 9.89666998e-01 -1.13047838e+00
-1.08480155e+00 6.12794638e-01 -4.33393605e-02 7.19901621e-02
4.00699437e-01 1.16560152e-02 9.37107682e-01 -9.18462753e-01
1.34342909e-01 1.32977819e+00 3.21484506e-01 7.67687738e-01
-1.41290355e+00 -8.83556783e-01 -2.62049675e-01 2.86397994e-01
-1.32204258e+00 -1.09724081e+00 1.15354943e+00 -2.14722946e-01
8.06960702e-01 5.75703979e-01 5.59335053e-01 1.63537872e+00
7.06035144e-06 6.20806456e-01 1.05045462e+00 -7.91886568e-01
1.84531406e-01 7.65385568e-01 -2.80502409e-01 4.44166273e-01
-3.51154245e-02 6.52443349e-01 -8.10708880e-01 9.74621400e-02
6.04457438e-01 -1.75687909e-01 -3.79610777e-01 -3.09829563e-01
-1.36960232e+00 7.54942894e-01 -9.33136269e-02 6.68820143e-01
-5.91570176e-02 -8.26066136e-02 4.65011001e-01 4.94852304e-01
5.37877321e-01 3.24578226e-01 -2.67626226e-01 -2.39789337e-01
-9.87746954e-01 -2.92892992e-01 9.75307226e-01 9.77224946e-01
6.08581126e-01 9.90969658e-01 -2.52568692e-01 1.05726755e+00
5.46908796e-01 1.11609113e+00 1.12964153e+00 -3.23117137e-01
3.13558221e-01 -1.64791629e-01 -1.50853530e-01 -8.24506462e-01
1.31239533e-01 -5.30003548e-01 -7.37127542e-01 2.50342548e-01
-1.43872827e-01 -1.44161135e-01 -7.79782951e-01 1.72993934e+00
-4.75481525e-02 4.51584041e-01 6.47190690e-01 6.11354947e-01
7.62201488e-01 1.15511322e+00 -1.77184314e-01 -4.94601578e-01
1.16672254e+00 -9.11404610e-01 -1.36424434e+00 2.71380562e-02
3.31089884e-01 -9.45638239e-01 1.12530804e+00 1.02926373e-01
-7.98263967e-01 -1.07149065e+00 -1.48147368e+00 3.57583135e-01
-4.15171891e-01 3.84934992e-01 8.37442204e-02 1.17751288e+00
-7.62178183e-01 5.01660407e-01 -4.37314361e-01 -1.05709732e-01
-1.79253489e-01 1.71570852e-01 -3.78756851e-01 4.07459080e-01
-1.50962603e+00 9.24898088e-01 1.64959699e-01 2.91884821e-02
-1.04783010e+00 -5.88392973e-01 -1.16153896e+00 3.15520138e-01
-1.41419813e-01 -3.51781487e-01 1.03526163e+00 -6.71417475e-01
-2.21446776e+00 5.29138982e-01 -3.86012673e-01 -2.43832618e-01
1.16223870e-02 -8.00032690e-02 -1.13271821e+00 3.38294357e-02
-1.27445668e-01 1.38826162e-01 1.45406616e+00 -1.37382662e+00
-1.50621980e-01 -1.99952051e-01 -9.86318827e-01 3.93065393e-01
-8.94062698e-01 -2.80433357e-01 1.70362577e-01 -9.86765683e-01
1.09252118e-01 -6.58888757e-01 2.67992318e-01 -3.29028815e-01
-2.88754255e-01 -2.68881261e-01 1.00345004e+00 -6.73636556e-01
1.14159966e+00 -2.40516257e+00 2.59261787e-01 -1.81711525e-01
-3.10950279e-02 5.15464842e-01 -3.51185739e-01 8.80957007e-01
-3.72282684e-01 -1.12615868e-01 8.64783973e-02 -8.85086834e-01
2.50239164e-01 1.57879084e-01 -6.68827653e-01 3.94900411e-01
2.06806332e-01 6.40425324e-01 -8.72119486e-01 -5.28179884e-01
3.48575920e-01 6.83685005e-01 -3.52049768e-01 7.02312410e-01
1.40372068e-01 5.01027465e-01 -1.41672373e-01 2.05605567e-01
4.62332785e-01 6.04600251e-01 -5.62787764e-02 -1.74065009e-01
-2.28300616e-02 2.04371631e-01 -1.01050019e+00 1.74959099e+00
-9.24133539e-01 8.74053478e-01 -1.00087263e-01 -9.85357940e-01
1.60142982e+00 1.00155067e+00 1.72130540e-01 -6.08437836e-01
8.54152590e-02 2.24160522e-01 -1.92756020e-03 -8.06209087e-01
2.31237739e-01 -5.78640401e-01 4.54506427e-02 3.56426358e-01
6.27899826e-01 -3.43277395e-01 -4.56336170e-01 -1.49655700e-01
5.24500191e-01 -5.22886366e-02 1.75261676e-01 -1.13226496e-01
6.73050880e-01 -8.47491801e-01 4.78813291e-01 3.52688670e-01
-9.28630531e-02 5.31875730e-01 -9.90655571e-02 2.03329772e-01
-1.07399642e+00 -1.55940020e+00 -3.10606599e-01 9.85390961e-01
-1.99981809e-01 -1.79809272e-01 -4.21795160e-01 -9.52454880e-02
-1.71405837e-01 1.12875855e+00 -4.04389739e-01 -4.84850466e-01
-2.74911404e-01 -2.03274086e-01 7.13057399e-01 4.00267303e-01
2.06796657e-02 -1.03214073e+00 2.02382714e-01 2.66323656e-01
-2.61669420e-02 -9.98260975e-01 -7.35753775e-01 2.53150851e-01
-6.83358014e-01 -1.06564045e-01 -1.11709583e+00 -1.18990684e+00
3.95406663e-01 1.15255438e-01 5.20571053e-01 -6.69637263e-01
-2.27438256e-01 2.50429988e-01 -5.72956920e-01 -4.76755679e-01
-8.97100151e-01 -1.30619898e-01 7.48633623e-01 4.68811870e-01
1.66348353e-01 -9.96894300e-01 -2.95597970e-01 3.27426881e-01
-6.12575948e-01 -1.32877693e-01 5.38189113e-01 9.28725719e-01
7.70654017e-03 -7.42083043e-02 1.12080967e+00 -2.13895127e-01
8.31435502e-01 -3.22851688e-01 -1.91784501e-01 1.89262688e-01
-6.65560186e-01 3.06515813e-01 7.73379683e-01 -7.36272335e-01
-1.30175817e+00 -2.54408896e-01 -1.55791998e-01 -6.29478812e-01
-3.65891844e-01 9.77846095e-04 -5.45949817e-01 4.58948046e-01
7.21613944e-01 8.44442606e-01 4.38121974e-01 -4.40408766e-01
6.75617099e-01 1.43989146e+00 5.39346278e-01 -3.99174631e-01
9.77371275e-01 4.12854850e-02 -4.76168901e-01 -1.32734656e+00
-1.69539690e-01 -4.05080050e-01 -4.35746759e-01 -5.24711609e-02
7.54574418e-01 -1.12299466e+00 -5.92206717e-01 3.45636427e-01
-1.14474988e+00 1.15054838e-01 -6.30798101e-01 1.14183116e+00
-7.81953514e-01 4.35511798e-01 -4.70915407e-01 -1.36305916e+00
-2.52660513e-01 -1.02091002e+00 1.02841783e+00 1.47648454e-01
-1.27481282e-01 -1.01235509e+00 5.73719978e-01 1.98555306e-01
6.51350200e-01 -2.56007493e-01 7.56651998e-01 -8.99203897e-01
-1.90584064e-01 -3.29581261e-01 3.86327714e-01 8.11510563e-01
6.54046953e-01 -1.89812720e-01 -1.41416228e+00 -3.85982573e-01
3.96367311e-01 -1.94159836e-01 5.94926298e-01 1.54252157e-01
6.42563105e-01 -2.94287086e-01 -8.65584388e-02 4.86530542e-01
1.13009560e+00 3.71938884e-01 5.80397785e-01 -3.68444860e-01
4.08675045e-01 7.46957123e-01 2.04261899e-01 4.31953311e-01
-1.54811114e-01 7.27706671e-01 -1.41695395e-01 -9.35316533e-02
-4.12746102e-01 -6.74690485e-01 7.67515898e-01 1.96331549e+00
2.26160139e-01 -3.27439398e-01 -2.58938283e-01 5.96738517e-01
-1.28095710e+00 -1.07854295e+00 3.68318141e-01 2.42120004e+00
8.24294627e-01 -2.07130760e-02 -7.59509206e-02 6.51074171e-01
8.84971499e-01 2.95751631e-01 -3.77829701e-01 -6.19505942e-01
-1.30430058e-01 4.19798374e-01 -3.64050008e-02 6.00326061e-01
-5.62449813e-01 7.46421695e-01 6.30516577e+00 1.16706085e+00
-1.17227387e+00 2.30353978e-02 5.69243245e-02 7.45903403e-02
-5.14429808e-01 -2.16569409e-01 -6.94184780e-01 3.84329528e-01
1.42586720e+00 -4.96649712e-01 4.78834033e-01 6.31156981e-01
3.66829336e-01 4.36126947e-01 -1.23555243e+00 1.22485924e+00
5.93991995e-01 -9.19709086e-01 1.16067678e-01 5.74213527e-02
4.65276927e-01 -4.53390598e-01 4.95805532e-01 4.51329172e-01
-2.50741601e-01 -1.18284655e+00 4.24689710e-01 5.55382431e-01
1.21055210e+00 -5.28586805e-01 4.00406897e-01 4.22979653e-01
-1.14935672e+00 1.48459719e-02 -3.80094588e-01 1.21278375e-01
1.44700646e-01 3.85506988e-01 -1.24349618e+00 7.21113503e-01
4.30475585e-02 5.75206757e-01 -5.83058894e-02 6.81402564e-01
-3.15722190e-02 8.71594667e-01 1.01144865e-01 -4.74712133e-01
-3.09962749e-01 -3.16965908e-01 8.41018438e-01 1.25854647e+00
5.55889130e-01 -3.43725502e-01 -4.53265518e-01 9.26443100e-01
2.13020995e-01 2.66665936e-01 -9.44343030e-01 -5.62708914e-01
6.48273587e-01 1.00324094e+00 -1.50907904e-01 -2.83386528e-01
-2.42234021e-01 1.12031174e+00 -1.67372853e-01 5.49931169e-01
-6.77484393e-01 -9.35434580e-01 6.54017389e-01 -7.94609115e-02
3.38297844e-01 -5.17058611e-01 -9.35632661e-02 -1.16645718e+00
-2.12331235e-01 -6.80286169e-01 -4.93400455e-01 -8.38977218e-01
-1.45197141e+00 9.89785314e-01 -1.28549501e-01 -1.68798423e+00
-5.96027851e-01 -4.05758888e-01 -7.67028451e-01 1.29852307e+00
-1.22660041e+00 -1.02172279e+00 5.81244789e-02 5.16027987e-01
7.71116793e-01 -9.55727935e-01 1.42847872e+00 1.47773847e-01
-3.69409889e-01 9.05962050e-01 5.40967464e-01 7.34659936e-03
7.48515189e-01 -1.17833567e+00 2.71223992e-01 4.32729989e-01
5.66943109e-01 5.62447309e-01 9.32795048e-01 -1.49350896e-01
-1.30591905e+00 -7.27137625e-01 1.11500525e+00 -2.80495435e-01
4.22349721e-01 -7.93941677e-01 -7.79707432e-01 2.33610287e-01
6.45699143e-01 -2.03229368e-01 1.18301189e+00 6.54802248e-02
-2.77239472e-01 -5.52945614e-01 -8.43277931e-01 5.86309612e-01
7.61180758e-01 -9.71945584e-01 -1.20766103e+00 9.00096968e-02
1.10067356e+00 1.39202967e-01 -7.86496162e-01 4.50414494e-02
6.01048827e-01 -6.42202020e-01 9.93434191e-01 -5.19642293e-01
8.65856111e-02 -1.31859064e-01 -4.18243945e-01 -1.83032560e+00
-1.84649542e-01 -8.97979736e-01 5.53718954e-02 1.56659138e+00
4.28421050e-01 -7.87016988e-01 3.54489833e-01 -1.63039461e-01
-2.27594927e-01 -3.16197693e-01 -9.88063157e-01 -1.21110475e+00
-6.50506541e-02 -2.85360605e-01 5.57134151e-01 8.51417303e-01
4.20478612e-01 7.01200008e-01 -6.76669240e-01 2.90952832e-01
6.25576198e-01 -1.24924570e-01 6.99260890e-01 -1.03362548e+00
-4.94512707e-01 -1.98754687e-02 -5.28777301e-01 -1.00021613e+00
3.91218394e-01 -1.09606278e+00 1.93329096e-01 -1.17191482e+00
-3.51887405e-01 -2.23761201e-01 -3.88003349e-01 -1.10843331e-01
9.42620113e-02 -7.77341500e-02 1.77792698e-01 -2.58483421e-02
1.14758126e-01 1.45997405e+00 1.33224869e+00 1.57213006e-02
-3.63092452e-01 2.18281567e-01 -4.37634468e-01 1.75455183e-01
7.32661784e-01 -4.76291209e-01 -7.44681180e-01 -1.28446557e-02
-5.55675507e-01 5.40172935e-01 -2.15865802e-02 -1.21554267e+00
9.00985450e-02 2.09510058e-01 2.95625925e-01 -4.83424067e-01
8.45579207e-01 -7.32566118e-01 -1.17102288e-01 4.39696014e-01
-5.08951783e-01 -2.76939720e-01 1.60723627e-01 9.08857644e-01
-4.43650126e-01 -1.98487908e-01 5.44304609e-01 2.88355827e-01
-1.49508402e-01 -3.70460115e-02 -5.44153154e-01 -2.53154457e-01
9.12929416e-01 -3.52187783e-01 1.18051648e-01 -6.70842946e-01
-1.04801393e+00 -3.79700452e-01 -2.76277274e-01 7.98672915e-01
8.77674937e-01 -1.66770172e+00 -9.20913935e-01 8.68117690e-01
2.41672054e-01 -7.77700186e-01 3.08771521e-01 2.71227807e-01
5.74509203e-02 6.80787861e-01 -2.24532574e-01 -5.11271954e-01
-1.13698745e+00 6.02051914e-01 1.53496331e-02 3.01597774e-01
-3.72079641e-01 6.42751396e-01 1.04206949e-01 -6.40282214e-01
3.44447613e-01 -1.75926089e-01 -2.06480697e-01 2.26223283e-03
3.80332738e-01 2.18741566e-01 -6.53841719e-02 -7.91152000e-01
-1.46208867e-01 4.87117887e-01 2.92325318e-01 -7.67564535e-01
1.21306932e+00 -2.49711558e-01 4.79578584e-01 1.09011936e+00
1.72023594e+00 5.74860275e-01 -9.12313581e-01 -4.44435149e-01
-4.43648666e-01 -5.38612485e-01 1.38165783e-02 -4.87494081e-01
-6.77029073e-01 1.26554680e+00 9.98124003e-01 2.52691209e-01
7.55434215e-01 -7.90869370e-02 8.26007485e-01 1.27288073e-01
6.38616160e-02 -9.72009003e-01 4.55021590e-01 9.60078984e-02
1.12829149e+00 -1.00265634e+00 -4.86812741e-01 -1.50575086e-01
-7.89833784e-01 1.19383168e+00 3.32086205e-01 5.17259957e-03
8.09944630e-01 1.49945870e-01 2.84045905e-01 3.65030199e-01
-8.06717515e-01 -2.73458540e-01 4.31089699e-01 1.00220954e+00
4.44598258e-01 2.96346009e-01 -1.22686714e-01 5.96733034e-01
-5.36197245e-01 -4.14693177e-01 3.20400238e-01 3.97477955e-01
-4.78415042e-01 -1.30503893e+00 -4.97091055e-01 2.44821310e-02
1.64293841e-01 -8.95261690e-02 7.72666335e-02 2.38151491e-01
-2.07551107e-01 1.23645973e+00 1.19917877e-01 -8.20782006e-01
3.39598447e-01 5.39025247e-01 4.10553485e-01 -7.93031454e-01
-6.46360219e-02 3.68736476e-01 -9.82629880e-02 1.10865869e-01
-1.98867142e-01 -4.48493332e-01 -9.40099239e-01 7.42997825e-02
-6.46623194e-01 4.62861866e-01 1.11845338e+00 6.80294096e-01
2.67888665e-01 7.66502857e-01 1.40943801e+00 -8.47478569e-01
-1.04448187e+00 -1.29962313e+00 -1.07806063e+00 3.65356266e-01
5.95497251e-01 -4.10696507e-01 -8.21434796e-01 1.80310503e-01] | [14.911758422851562, 6.5262651443481445] |
458bc163-b030-4b3d-9dd5-a79dcd1d90e9 | emospeech-guiding-fastspeech2-towards | 2307.00024 | null | https://arxiv.org/abs/2307.00024v1 | https://arxiv.org/pdf/2307.00024v1.pdf | EmoSpeech: Guiding FastSpeech2 Towards Emotional Text to Speech | State-of-the-art speech synthesis models try to get as close as possible to the human voice. Hence, modelling emotions is an essential part of Text-To-Speech (TTS) research. In our work, we selected FastSpeech2 as the starting point and proposed a series of modifications for synthesizing emotional speech. According to automatic and human evaluation, our model, EmoSpeech, surpasses existing models regarding both MOS score and emotion recognition accuracy in generated speech. We provided a detailed ablation study for every extension to FastSpeech2 architecture that forms EmoSpeech. The uneven distribution of emotions in the text is crucial for better, synthesized speech and intonation perception. Our model includes a conditioning mechanism that effectively handles this issue by allowing emotions to contribute to each phone with varying intensity levels. The human assessment indicates that proposed modifications generate audio with higher MOS and emotional expressiveness. | ['Vitaly Shutov', 'Daria Diatlova'] | 2023-06-28 | null | null | null | null | ['emotion-recognition', 'speech-synthesis'] | ['computer-vision', 'speech'] | [-1.73222750e-01 5.21202147e-01 2.52654254e-01 -2.99215734e-01
-3.53581309e-01 -4.03358728e-01 5.55687606e-01 -1.18160516e-01
-5.28172702e-02 7.26267099e-01 4.57475364e-01 -1.43103644e-01
3.34396452e-01 -5.17077446e-01 -1.97216719e-01 -3.72124851e-01
2.71324784e-01 1.88930929e-01 1.02256671e-01 -6.31423175e-01
-1.26356378e-01 5.18532336e-01 -2.02216458e+00 6.26108825e-01
8.01067591e-01 9.61800277e-01 2.68438071e-01 1.06383586e+00
-5.48954368e-01 9.76810396e-01 -1.18221509e+00 -6.59468591e-01
-3.48092347e-01 -6.33895040e-01 -5.85604846e-01 3.71211991e-02
-3.25820953e-01 2.79773399e-02 -5.25194854e-02 7.64277220e-01
9.08114433e-01 2.70919025e-01 7.62393355e-01 -1.29703021e+00
-4.10425723e-01 7.68145084e-01 2.68049926e-01 -2.47902215e-01
5.55797279e-01 -1.75904483e-01 8.54187548e-01 -9.29954290e-01
4.97954488e-01 1.14289951e+00 3.72247070e-01 9.38913822e-01
-8.57148051e-01 -4.59967077e-01 -1.33977771e-01 1.68788344e-01
-1.12773180e+00 -8.91772211e-01 7.57044315e-01 -7.47128874e-02
1.28092945e+00 9.28681016e-01 6.80303812e-01 1.39960480e+00
-7.56461546e-02 6.99537158e-01 1.03274679e+00 -8.04481745e-01
2.96595037e-01 8.69793713e-01 -2.73827970e-01 -4.36288156e-02
-6.79246068e-01 1.41175181e-01 -6.12013578e-01 1.23956360e-01
1.94414407e-01 -7.40473747e-01 -1.73979521e-01 4.50082064e-01
-6.97501481e-01 5.66321969e-01 -3.90061855e-01 8.55542839e-01
-7.53711879e-01 -9.88597348e-02 6.85140967e-01 5.38769960e-01
6.24385774e-01 3.97726595e-01 -3.16351861e-01 -8.22957158e-01
-1.16330111e+00 -1.66322589e-01 1.06843460e+00 9.02411044e-01
1.91689894e-01 7.02426732e-01 -2.64963776e-01 1.26245904e+00
2.73647606e-01 6.96872175e-01 4.72126514e-01 -8.66310537e-01
-1.52723456e-03 3.03716570e-01 4.33621416e-03 -8.59639823e-01
-1.24065265e-01 -5.38530111e-01 -7.05424905e-01 2.28874117e-01
-3.18636864e-01 -5.44676840e-01 -6.24583483e-01 1.55684543e+00
1.88869387e-01 -2.20998496e-01 5.22859275e-01 6.81028843e-01
1.02721405e+00 1.20555866e+00 2.44095102e-01 -6.03354871e-01
1.25923121e+00 -8.83569479e-01 -1.60040045e+00 -5.14312126e-02
4.15824711e-01 -1.22063708e+00 1.24593675e+00 6.39353275e-01
-1.47894037e+00 -7.22724974e-01 -9.95418787e-01 2.25445703e-01
-4.68254358e-01 4.11488950e-01 3.70626420e-01 1.39144146e+00
-1.20406651e+00 3.51151586e-01 -3.39852273e-01 -2.67896086e-01
-4.67282623e-01 2.53605545e-01 -6.76231682e-02 8.36379409e-01
-1.53086960e+00 1.07665217e+00 1.55623600e-01 -1.44362357e-02
-1.99170500e-01 -2.96928704e-01 -7.65950680e-01 3.51473629e-01
4.11686525e-02 -3.84732425e-01 1.45305753e+00 -1.32008839e+00
-2.44068241e+00 5.91973126e-01 -3.04232210e-01 -3.83753121e-01
3.12488824e-01 -1.36907890e-01 -1.05531144e+00 3.65769058e-01
-6.33430660e-01 7.47007430e-01 8.88504982e-01 -1.57065976e+00
-5.13411999e-01 3.25933874e-01 -4.75558162e-01 1.99573904e-01
-4.68869448e-01 5.83245337e-01 -2.47150004e-01 -8.81183624e-01
-4.09684598e-01 -6.56956375e-01 1.21439263e-01 -5.67521513e-01
-1.65561393e-01 -6.41232356e-02 7.73972392e-01 -7.57797241e-01
1.74267626e+00 -2.08846784e+00 7.66446143e-02 2.18202785e-01
-2.99046844e-01 6.77592576e-01 7.13326335e-02 7.63172805e-01
-3.84536624e-01 2.69031107e-01 1.51595831e-01 -8.41302156e-01
4.08662498e-01 2.63836324e-01 -4.05867934e-01 -2.29412720e-01
2.54210562e-01 6.57661498e-01 -5.33762276e-01 -5.66941321e-01
6.85578942e-01 9.10314441e-01 -4.61906344e-01 4.91287351e-01
-1.09223709e-01 2.39270866e-01 -1.02602214e-01 3.08737755e-01
4.81069356e-01 8.06046724e-01 -9.65155065e-02 -1.68219879e-01
-3.98642063e-01 2.59621590e-01 -1.16779137e+00 1.10687816e+00
-1.01977670e+00 5.91480255e-01 3.96643966e-01 -5.74773312e-01
1.35476089e+00 1.13696182e+00 3.70192796e-01 -7.54254818e-01
5.85784078e-01 3.46246779e-01 8.64164457e-02 -7.65193701e-01
8.01926851e-01 -4.92719978e-01 8.94594181e-04 -1.99018531e-02
1.52701661e-01 -7.75844157e-01 3.61388959e-02 4.56917100e-03
4.23752934e-01 -1.76881239e-01 2.52633661e-01 -4.04302813e-02
8.44236135e-01 -5.41624546e-01 3.74947414e-02 2.58995265e-01
-1.93916827e-01 3.66517723e-01 4.95709062e-01 3.93315256e-01
-9.49744105e-01 -7.25646913e-01 2.39402175e-01 9.71292257e-01
-4.59296167e-01 -4.66832280e-01 -1.16976583e+00 -1.50301114e-01
-7.92823255e-01 1.42201233e+00 -2.84050405e-01 -1.22958757e-01
-7.79099017e-02 -7.75528625e-02 9.11218762e-01 8.61642882e-02
-1.90543812e-02 -1.55844796e+00 -4.45589393e-01 4.20618057e-01
-4.70094144e-01 -1.19538713e+00 -2.69307733e-01 3.60073864e-01
-4.33693618e-01 -3.32982630e-01 -8.00079405e-01 -7.09178746e-01
8.09017047e-02 -3.57412666e-01 9.96371448e-01 -8.50727707e-02
1.43811062e-01 3.82593781e-01 -9.33717191e-01 -7.87363887e-01
-1.26722789e+00 -1.86830070e-02 -5.74987084e-02 1.35111764e-01
5.83361723e-02 -6.30875170e-01 -3.21622677e-02 2.59790033e-01
-1.00550890e+00 -9.66089591e-02 3.56503129e-01 5.05373120e-01
8.40210095e-02 2.19623223e-01 8.26590240e-01 -3.04053843e-01
1.15647018e+00 -1.05460763e-01 1.01342224e-01 7.44038075e-02
-1.35990292e-01 -1.86183915e-01 9.08055723e-01 -4.37796116e-01
-1.46611893e+00 -1.67995945e-01 -1.11200178e+00 -2.85364509e-01
-5.23846269e-01 2.57050484e-01 -5.33068359e-01 1.85079530e-01
4.42470968e-01 1.20255679e-01 -8.14286172e-02 -1.26061991e-01
3.51074904e-01 1.35585594e+00 3.02681983e-01 -4.66905385e-01
3.31051469e-01 -1.58970892e-01 -3.25336337e-01 -1.42357755e+00
-6.71458766e-02 -2.59769112e-01 -1.78978279e-01 -6.43104136e-01
7.69348323e-01 -6.68607354e-01 -9.30122018e-01 4.65186298e-01
-1.33955717e+00 -3.54022592e-01 -4.37092811e-01 6.35720313e-01
-5.13717473e-01 3.18837285e-01 -6.21954262e-01 -1.75206625e+00
-5.30647695e-01 -1.04805493e+00 1.09948647e+00 5.16004972e-02
-7.13982642e-01 -8.97220433e-01 -8.52283239e-02 2.21585259e-01
7.53900886e-01 -1.30300313e-01 6.50543451e-01 -5.91758907e-01
4.67670113e-01 -9.41358805e-02 2.99800545e-01 8.13838720e-01
1.28060102e-01 6.05573714e-01 -1.26770997e+00 4.47181433e-01
2.91760474e-01 -3.29104602e-01 3.37208688e-01 1.81582853e-01
7.44562984e-01 -1.93223700e-01 3.82460952e-01 6.60984889e-02
8.23772192e-01 7.31092334e-01 1.10780466e+00 -1.42510101e-01
-5.08748069e-02 1.01319814e+00 8.20164979e-01 6.87774599e-01
4.46382426e-02 5.86237848e-01 3.79444391e-01 -3.39688212e-01
-3.31119716e-01 -6.79058284e-02 6.46278560e-01 1.52444053e+00
1.19023882e-01 -8.32090259e-01 -4.31948096e-01 4.09179688e-01
-1.32887733e+00 -1.02539468e+00 -3.52646977e-01 1.82256973e+00
7.88539112e-01 2.01540247e-01 1.29111424e-01 7.21740127e-01
7.18047082e-01 6.64712414e-02 2.22094342e-01 -1.42199099e+00
-4.27937388e-01 6.37013316e-01 -8.67851079e-02 9.62187529e-01
-6.61143899e-01 1.09112430e+00 6.47493505e+00 1.20375609e+00
-1.36952734e+00 7.57668726e-03 4.87262547e-01 -5.12462817e-02
-5.52733719e-01 -4.75199521e-01 -3.75145048e-01 3.77936304e-01
1.51655912e+00 -5.67983510e-03 4.65914994e-01 5.99269807e-01
8.18187296e-01 6.50920197e-02 -7.14064002e-01 9.24440026e-01
1.23511031e-01 -7.45826960e-01 8.32234621e-02 -2.90286034e-01
3.00026476e-01 -5.65273464e-01 -1.79078933e-02 4.84497339e-01
-2.97087610e-01 -1.05604124e+00 1.02556562e+00 4.20154750e-01
8.38578820e-01 -9.88564193e-01 8.00843656e-01 2.17809170e-01
-1.16413546e+00 2.24005297e-01 1.34458423e-01 -1.33062854e-01
5.79636395e-01 4.01653975e-01 -9.31044757e-01 6.41851664e-01
2.04593211e-01 -7.55036697e-02 5.29317278e-03 5.12768090e-01
-3.82086605e-01 9.01496530e-01 -2.55326897e-01 -6.19541168e-01
1.39861003e-01 1.30592557e-02 6.32356107e-01 1.58894587e+00
7.99317360e-01 1.77038327e-01 -3.36636662e-01 6.10379934e-01
2.94428349e-01 8.84920597e-01 -2.87135571e-01 -3.94470036e-01
5.22347510e-01 1.17873502e+00 -5.57144105e-01 -4.08630878e-01
5.29243350e-02 1.27681398e+00 -3.63860488e-01 1.89093485e-01
-1.14864230e+00 -7.24706173e-01 5.37169278e-01 -1.96200162e-01
1.58334449e-01 5.09935282e-02 -2.92054504e-01 -6.80119932e-01
-1.48538753e-01 -1.21702325e+00 -3.02701026e-01 -1.26649964e+00
-8.22788656e-01 1.19139254e+00 -3.00393730e-01 -1.00164688e+00
-4.05966192e-01 -5.52999318e-01 -6.07484460e-01 9.27726924e-01
-1.21470439e+00 -1.10407555e+00 4.66652252e-02 3.61276835e-01
7.57042527e-01 -4.40827720e-02 1.19710302e+00 6.28794611e-01
-3.53109896e-01 6.14251852e-01 -3.54546994e-01 -4.60093766e-01
7.33922005e-01 -1.08292282e+00 1.51746288e-01 6.45481825e-01
-9.36837196e-02 3.22946459e-01 1.37761295e+00 -3.51280212e-01
-9.36669946e-01 -6.64316416e-01 1.29012728e+00 1.33101150e-01
5.53215742e-01 -4.37919915e-01 -7.06871867e-01 3.20929149e-03
8.13214779e-01 -8.22352290e-01 8.79568756e-01 -1.26155153e-01
1.80002764e-01 7.14884475e-02 -1.20590401e+00 8.91948760e-01
6.14852905e-01 -5.77225745e-01 -5.52262306e-01 -2.95634478e-01
7.84677148e-01 -1.65416554e-01 -8.79487157e-01 7.81505480e-02
4.45542902e-01 -1.29080212e+00 5.96085668e-01 -1.54350266e-01
3.13316703e-01 1.80393655e-03 -2.77898818e-01 -1.74990535e+00
1.84142962e-01 -1.19010496e+00 -1.26528582e-02 1.67593694e+00
6.35849237e-01 -5.94772458e-01 4.57060486e-01 4.94750649e-01
-4.88488078e-01 -5.06875694e-01 -7.82667518e-01 -6.25258744e-01
-4.17437553e-02 -9.07909751e-01 5.92427790e-01 5.71785033e-01
5.57332039e-01 5.56473851e-01 -7.56426692e-01 5.96164390e-02
-4.07311507e-02 -4.98870820e-01 6.29774511e-01 -8.79717588e-01
-3.33898455e-01 -7.38236010e-01 -6.57415837e-02 -7.73763537e-01
3.01680624e-01 -3.78308982e-01 2.04374015e-01 -1.40844035e+00
-5.80408394e-01 -5.56107201e-02 1.17952831e-01 2.99497783e-01
7.75341541e-02 1.01915844e-01 4.33334738e-01 -7.07530141e-01
4.12415043e-02 9.48685706e-01 1.17090809e+00 2.74233550e-01
-4.27154720e-01 1.32512115e-03 -4.29535329e-01 5.45749307e-01
9.97926831e-01 -3.19898337e-01 -5.07751822e-01 1.33291662e-01
2.35550895e-01 3.54018569e-01 -1.33500963e-01 -1.11671937e+00
6.38308823e-02 1.68155394e-02 -1.00065529e-01 -6.26364052e-01
9.57705200e-01 -1.16018236e+00 3.68064404e-01 2.61826515e-01
-3.77040476e-01 -2.72186488e-01 4.50411946e-01 -2.65563875e-01
-5.31929910e-01 -4.53177392e-01 7.93961942e-01 2.79069930e-01
-2.01406240e-01 -3.12190533e-01 -1.09540117e+00 -2.49042720e-01
8.01279545e-01 -4.56317067e-01 7.51633290e-03 -1.02358496e+00
-9.86311018e-01 -1.42354727e-01 5.96777238e-02 4.58957583e-01
5.62510669e-01 -9.53563094e-01 -5.96580565e-01 1.40721276e-01
-1.30590856e-01 -9.32856083e-01 4.95097101e-01 7.24646688e-01
-2.72150874e-01 5.36832452e-01 -1.12281628e-01 7.59047782e-03
-1.85464954e+00 3.17086399e-01 4.03767914e-01 -7.24894702e-02
2.45788265e-02 7.09938288e-01 -2.81703949e-01 -3.98292720e-01
5.01009345e-01 -3.91375482e-01 -4.52603966e-01 2.87255108e-01
2.69123852e-01 4.69662040e-01 3.07865351e-01 -9.31402266e-01
-2.45576113e-01 1.13014348e-01 5.58543086e-01 -7.18254447e-01
1.01668906e+00 -3.57723206e-01 1.12170046e-02 6.63356125e-01
9.59054589e-01 6.85786247e-01 -5.36061525e-01 5.52706778e-01
-3.01613271e-01 -2.71666870e-02 1.58544779e-01 -1.16658151e+00
-7.28279769e-01 1.12045085e+00 3.16935211e-01 5.62790453e-01
1.35413325e+00 -3.93651843e-01 1.04088485e+00 2.55415529e-01
-7.26028010e-02 -1.64461529e+00 5.98922446e-02 7.98110127e-01
1.04192996e+00 -6.90642476e-01 -6.50539398e-01 -6.83049679e-01
-1.20033014e+00 1.16095459e+00 3.22348922e-01 5.06780207e-01
6.13081217e-01 8.37658346e-01 6.04263902e-01 1.04935288e-01
-9.95147049e-01 -2.57381052e-01 1.49492979e-01 6.09616578e-01
8.20520282e-01 1.73835427e-01 -6.03841066e-01 1.02888274e+00
-6.15590274e-01 -9.07721221e-02 4.91729110e-01 5.32170773e-01
-6.03809237e-01 -1.42881298e+00 -5.63054383e-01 -2.32185021e-01
-7.90633202e-01 -2.18876094e-01 -7.58555412e-01 4.99971688e-01
1.51656643e-01 1.69998932e+00 -1.48929760e-01 -6.07470393e-01
6.50846779e-01 5.21058261e-01 2.07027182e-01 -2.98102707e-01
-1.14394248e+00 5.65552413e-01 7.58198142e-01 -1.98586851e-01
-4.59360331e-01 -3.01513642e-01 -1.51003349e+00 -1.33063480e-01
-3.14120650e-01 6.08273625e-01 1.28642106e+00 7.62936652e-01
4.09525633e-01 1.09141910e+00 9.51247990e-01 -8.40817928e-01
-2.40879506e-01 -1.15608978e+00 -7.00096846e-01 2.22171366e-01
4.02102284e-02 -2.54683763e-01 -5.36984265e-01 1.56800114e-02] | [14.754486083984375, 6.5207414627075195] |
f88b9956-af96-4a54-b9f3-20b96d0a7afb | proselflc-progressive-self-label-correction | 2005.03788 | null | https://arxiv.org/abs/2005.03788v6 | https://arxiv.org/pdf/2005.03788v6.pdf | ProSelfLC: Progressive Self Label Correction for Training Robust Deep Neural Networks | To train robust deep neural networks (DNNs), we systematically study several target modification approaches, which include output regularisation, self and non-self label correction (LC). Two key issues are discovered: (1) Self LC is the most appealing as it exploits its own knowledge and requires no extra models. However, how to automatically decide the trust degree of a learner as training goes is not well answered in the literature? (2) Some methods penalise while the others reward low-entropy predictions, prompting us to ask which one is better? To resolve the first issue, taking two well-accepted propositions--deep neural networks learn meaningful patterns before fitting noise [3] and minimum entropy regularisation principle [10]--we propose a novel end-to-end method named ProSelfLC, which is designed according to learning time and entropy. Specifically, given a data point, we progressively increase trust in its predicted label distribution versus its annotated one if a model has been trained for enough time and the prediction is of low entropy (high confidence). For the second issue, according to ProSelfLC, we empirically prove that it is better to redefine a meaningful low-entropy status and optimise the learner toward it. This serves as a defence of entropy minimisation. We demonstrate the effectiveness of ProSelfLC through extensive experiments in both clean and noisy settings. The source code is available at https://github.com/XinshaoAmosWang/ProSelfLC-CVPR2021. Keywords: entropy minimisation, maximum entropy, confidence penalty, self knowledge distillation, label correction, label noise, semi-supervised learning, output regularisation | ['David A. Clifton', 'Yang Hua', 'Xinshao Wang', 'Neil M. Robertson', 'Elyor Kodirov'] | 2020-05-07 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Wang_ProSelfLC_Progressive_Self_Label_Correction_for_Training_Robust_Deep_Neural_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Wang_ProSelfLC_Progressive_Self_Label_Correction_for_Training_Robust_Deep_Neural_CVPR_2021_paper.pdf | cvpr-2021-1 | ['self-knowledge-distillation'] | ['computer-vision'] | [ 4.61890459e-01 7.58395731e-01 -2.97494233e-01 -5.75927258e-01
-5.18222272e-01 -5.42311847e-01 4.22373652e-01 3.18803310e-01
-8.16339076e-01 9.96444762e-01 9.24352407e-02 -2.09568202e-01
-4.39646691e-01 -4.64305580e-01 -7.88607419e-01 -9.42277968e-01
2.34379664e-01 1.77724257e-01 1.30874008e-01 5.49781695e-03
9.10773799e-02 3.28506976e-01 -1.43471420e+00 -1.00493431e-01
9.52999115e-01 1.05749917e+00 -1.74000472e-01 5.09168386e-01
4.40046102e-01 9.72075701e-01 -3.63005877e-01 -8.57373059e-01
3.34666282e-01 -5.13634443e-01 -1.05015039e+00 -3.98475111e-01
2.26184353e-01 1.57081544e-01 1.06420048e-01 1.41222298e+00
6.21746480e-01 2.94978261e-01 7.37224936e-01 -1.19241858e+00
-5.97408175e-01 9.93593812e-01 -1.28133833e-01 1.09275952e-01
-2.05603316e-01 1.52043089e-01 1.11791039e+00 -4.85879600e-01
4.61967826e-01 8.92724156e-01 8.92404974e-01 9.56796527e-01
-1.29497063e+00 -6.18010044e-01 -2.22365186e-02 6.61277324e-02
-1.13976800e+00 -6.03987873e-01 6.04553163e-01 -2.66028523e-01
7.32964814e-01 4.83508140e-01 3.91087294e-01 1.35326338e+00
1.27129838e-01 7.69127071e-01 1.28305185e+00 -6.04442418e-01
4.33199465e-01 7.70490050e-01 2.38859758e-01 6.51490033e-01
1.50738776e-01 5.19853771e-01 -3.83891702e-01 1.84745844e-02
1.39728799e-01 -3.78301322e-01 -3.39058280e-01 -2.38096789e-01
-8.94129932e-01 7.50425279e-01 2.99183160e-01 3.75762939e-01
-1.82220593e-01 4.34449734e-03 6.80470109e-01 4.96665686e-01
6.15742207e-01 7.30825365e-01 -8.15645099e-01 -3.28790694e-01
-9.30534899e-01 3.91970985e-02 8.40280771e-01 6.38928950e-01
7.26936162e-01 -6.34478852e-02 -1.97562590e-01 8.48324060e-01
4.08050567e-01 -5.61078219e-03 8.48488569e-01 -9.55796957e-01
-4.57231048e-03 5.86625338e-01 -1.84838951e-01 -5.95203400e-01
-5.12297630e-01 -7.10520804e-01 -1.08522677e+00 3.13237846e-01
2.08194017e-01 -4.66267645e-01 -7.21879363e-01 2.10020661e+00
2.35730633e-01 1.54601574e-01 4.12819356e-01 7.63495862e-01
6.58836126e-01 2.73352772e-01 1.81407124e-01 -5.82441390e-01
9.96423960e-01 -9.58820522e-01 -6.75826490e-01 -2.11733088e-01
9.74293113e-01 -3.65633368e-01 8.70667458e-01 5.68593740e-01
-8.70984793e-01 -5.14640510e-01 -1.07518768e+00 4.63439636e-02
-3.84447902e-01 4.59666476e-02 1.76382437e-01 7.89713144e-01
-1.10617292e+00 1.10992396e+00 -6.58646584e-01 -1.63200587e-01
5.02656341e-01 3.64259630e-01 -3.82146299e-01 4.14688915e-01
-1.52792442e+00 1.13473690e+00 1.02463746e+00 2.23934148e-02
-5.18774629e-01 -5.80026925e-01 -6.23682857e-01 8.37091822e-03
4.40744489e-01 -4.53100652e-01 1.46094799e+00 -1.42533922e+00
-1.66593254e+00 8.66860151e-01 2.18015909e-01 -7.84871817e-01
8.10372114e-01 -3.84277314e-01 -2.78601259e-01 -3.37391645e-01
-2.13183731e-01 7.56211579e-01 7.99528003e-01 -1.20793581e+00
-6.17175519e-01 -9.46034640e-02 -2.14614823e-01 2.13575006e-01
-5.81851721e-01 -1.74194634e-01 6.29933476e-02 -4.77340788e-01
-2.28377521e-01 -9.48737323e-01 -1.90553308e-01 -3.53950977e-01
-6.31536245e-01 -4.19550061e-01 5.80479920e-01 -3.97902638e-01
1.45111728e+00 -2.11439657e+00 -3.15598063e-02 3.44645768e-01
2.15631843e-01 6.83817744e-01 9.61934701e-02 -1.13948047e-01
-4.06538308e-01 4.32486832e-01 -4.37257290e-01 -3.35540146e-01
3.23314190e-01 3.07046086e-01 -1.69673428e-01 3.75536144e-01
1.65579811e-01 9.04259086e-01 -9.51633036e-01 -3.12379867e-01
-3.97209786e-02 6.26969159e-01 -4.62467730e-01 1.85749739e-01
-4.22392450e-02 1.75177589e-01 -4.40307707e-03 2.37240508e-01
4.50212508e-01 -1.10406674e-01 5.76117858e-02 -3.82459819e-01
4.02661636e-02 2.61812091e-01 -1.23797715e+00 1.08159900e+00
-2.00360790e-01 5.79878926e-01 -2.64004081e-01 -1.05303121e+00
1.00361598e+00 4.35962319e-01 3.56572032e-01 -4.14135247e-01
3.09997469e-01 3.02437991e-01 1.20981212e-03 -5.19064188e-01
3.26494962e-01 -3.30925107e-01 7.55148008e-02 4.32756782e-01
2.63653278e-01 2.45627940e-01 1.33073404e-01 4.53526005e-02
1.17568970e+00 2.24586070e-01 4.96720284e-01 -2.48547614e-01
4.50572699e-01 -4.30733472e-01 6.91171944e-01 7.53129303e-01
-6.97604001e-01 3.75457287e-01 5.89253485e-01 -1.37593329e-01
-8.90725255e-01 -6.89463794e-01 -2.76376754e-01 1.32143688e+00
-1.10221818e-01 -3.45305234e-01 -9.66041982e-01 -1.21190095e+00
-1.31609246e-01 9.71391559e-01 -8.27246189e-01 -7.20974743e-01
-2.18988463e-01 -7.95884132e-01 7.36529708e-01 3.95554572e-01
5.85501432e-01 -1.27178013e+00 -8.22235107e-01 1.34076104e-01
-4.47517522e-02 -6.16420388e-01 -2.66119838e-01 9.69831944e-01
-7.37157047e-01 -8.26088369e-01 -3.96127909e-01 -4.74728644e-01
6.81297243e-01 -4.54084426e-01 1.04419172e+00 1.64190620e-01
3.10804874e-01 1.48160085e-01 -4.85547274e-01 -4.09426838e-01
-7.97334373e-01 2.55855262e-01 3.59394372e-01 -1.10502839e-01
5.54846227e-01 -4.06372160e-01 -3.48023027e-01 2.27248818e-01
-9.37210262e-01 -5.85346073e-02 8.50885272e-01 8.82283926e-01
5.71350098e-01 2.95318037e-01 7.22532749e-01 -1.34870827e+00
8.70360613e-01 -3.78247142e-01 -3.00573379e-01 3.21715832e-01
-1.36416209e+00 3.57383281e-01 7.28639841e-01 -4.67788786e-01
-9.88486648e-01 1.44274786e-01 -4.72002387e-01 -3.63202661e-01
-5.14175653e-01 4.23813879e-01 -1.95226535e-01 7.59628490e-02
1.15850186e+00 7.21485540e-02 -2.76537165e-02 -3.73017281e-01
4.07696664e-01 6.22200966e-01 4.68639910e-01 -4.31042165e-01
6.80263162e-01 -2.51900200e-02 -1.79981902e-01 -2.79837668e-01
-1.42229581e+00 -2.86205709e-01 -8.88504744e-01 -2.39052460e-01
6.52721107e-01 -5.55993378e-01 -7.59375274e-01 2.24386707e-01
-8.21307361e-01 -5.41953623e-01 -6.89903080e-01 2.54622877e-01
-5.44287980e-01 3.23300779e-01 -2.18055859e-01 -9.67209399e-01
-6.62291229e-01 -1.03505659e+00 4.33148056e-01 4.23138499e-01
-4.64007974e-01 -1.29261458e+00 1.27919778e-01 1.89711615e-01
3.17332864e-01 3.90167028e-01 6.30376220e-01 -1.27935934e+00
1.62941650e-01 -1.09625392e-01 -3.95857394e-02 1.04109824e+00
-4.34497073e-02 8.41048136e-02 -1.24110675e+00 -3.43920052e-01
2.19188154e-01 -6.35565460e-01 9.55589294e-01 1.75323591e-01
1.11187196e+00 -8.52115571e-01 1.15726449e-01 4.92418170e-01
1.30808008e+00 -4.57627475e-02 4.04376686e-01 6.21665716e-01
6.01941109e-01 6.43469274e-01 5.37354469e-01 3.05523187e-01
1.22726329e-01 3.57837290e-01 4.65549558e-01 3.35162021e-02
2.04742342e-01 -1.09890342e-01 6.89941347e-01 8.30654681e-01
-5.23237288e-02 -2.47127131e-01 -7.50302434e-01 2.55071938e-01
-1.99439430e+00 -9.33620512e-01 -8.56461003e-02 2.18265986e+00
1.39374590e+00 5.58746696e-01 1.46369800e-01 4.79006648e-01
4.62422043e-01 -3.09025124e-03 -7.36914456e-01 -6.69919670e-01
-1.26754031e-01 -7.35030472e-02 6.43490911e-01 4.52158242e-01
-1.28777432e+00 7.30146646e-01 4.92458391e+00 1.01862156e+00
-9.85794067e-01 3.52950573e-01 9.82494354e-01 6.90711960e-02
-2.19636425e-01 -1.16597973e-01 -9.08508539e-01 5.55396855e-01
1.36987770e+00 -9.56515148e-02 2.36737758e-01 1.07053947e+00
5.91538623e-02 -5.33169918e-02 -1.20597267e+00 6.49992764e-01
1.67820118e-02 -1.00619507e+00 -3.66857767e-01 -1.59048647e-01
7.31359422e-01 1.94480091e-01 1.09892003e-01 6.38837397e-01
4.88209158e-01 -9.01339054e-01 8.71133029e-01 6.36494040e-01
4.76397842e-01 -9.53935325e-01 1.01415884e+00 7.22850025e-01
-5.86778760e-01 -2.22321495e-01 -4.06079948e-01 3.13350469e-01
-2.47274846e-01 9.50369179e-01 -7.91216493e-01 4.10429925e-01
5.72938979e-01 6.57216251e-01 -8.30640197e-01 8.88889313e-01
-5.34292936e-01 9.64394927e-01 -2.91145682e-01 -3.38883721e-03
1.62528306e-01 -2.67053619e-02 5.15582800e-01 1.41897011e+00
4.00701910e-02 -1.66989982e-01 -2.31349498e-01 7.76394427e-01
-2.34681815e-01 4.24034484e-02 -3.16730201e-01 1.63582310e-01
4.45151746e-01 1.38159454e+00 -5.12696922e-01 -2.62978435e-01
1.12518266e-01 8.67801905e-01 5.46521187e-01 6.86044693e-02
-7.07841456e-01 -3.41639549e-01 1.93602145e-01 -1.18824914e-01
-8.35531801e-02 4.77275640e-01 -5.06559014e-01 -8.32170904e-01
-8.71580690e-02 -8.30734015e-01 4.87978071e-01 -4.55491900e-01
-1.32936001e+00 6.90914214e-01 -1.95956960e-01 -1.07922685e+00
-2.08227932e-01 -4.94710028e-01 -3.54203582e-01 5.67001998e-01
-1.65823698e+00 -6.34381831e-01 1.27698167e-03 3.11112404e-01
2.51691163e-01 -1.20778553e-01 7.57963121e-01 2.84144104e-01
-6.75697327e-01 1.00451791e+00 3.23930770e-01 -1.53878583e-02
7.37035513e-01 -1.62724662e+00 1.13459723e-03 7.06773162e-01
1.06044225e-01 3.80647391e-01 8.87846291e-01 -4.82995600e-01
-6.96957767e-01 -1.37131393e+00 1.07218707e+00 -5.87543249e-01
4.77079928e-01 -1.40680134e-01 -1.25135636e+00 4.45951909e-01
4.56188656e-02 7.43003339e-02 8.11271310e-01 1.33422920e-02
-2.62216210e-01 -1.08336069e-01 -1.31458902e+00 4.32933092e-01
7.89394975e-01 -2.87701815e-01 -4.95281935e-01 4.87716049e-01
7.03165591e-01 -3.94278556e-01 -1.05944443e+00 6.64851785e-01
3.99815500e-01 -1.26942086e+00 4.97890204e-01 -3.76559377e-01
3.21379870e-01 8.34167451e-02 9.11246687e-02 -1.55293465e+00
-3.27724099e-01 -6.85227036e-01 -4.39703435e-01 1.38548636e+00
7.29694307e-01 -6.64711893e-01 7.72523761e-01 8.73332202e-01
-1.01399899e-01 -1.16141224e+00 -9.34395015e-01 -8.93378615e-01
8.73398334e-02 -5.08098662e-01 2.72904605e-01 1.23388684e+00
1.54611990e-01 3.25679481e-01 -4.74357516e-01 -4.97560389e-02
5.08221030e-01 -5.44264674e-01 4.20238942e-01 -1.45022416e+00
-2.26387426e-01 -6.18583798e-01 -2.18321577e-01 -7.21255481e-01
3.74303073e-01 -1.15917647e+00 3.51716071e-01 -1.16630638e+00
2.26718280e-02 -4.04554099e-01 -7.17559993e-01 1.09232843e+00
-3.04781944e-01 1.69399932e-01 1.19425997e-01 1.26633480e-01
-8.60391319e-01 4.68257010e-01 8.38253438e-01 2.12872267e-01
-2.37244621e-01 3.20439011e-01 -9.25671697e-01 7.12086678e-01
1.11057329e+00 -7.39710748e-01 -3.67763638e-01 1.14094801e-01
6.75650895e-01 -5.18064260e-01 3.88363510e-01 -9.61565435e-01
1.47093713e-01 -1.04896143e-01 2.07604915e-01 -3.06511819e-02
-5.98693825e-02 -9.45994556e-01 5.71251586e-02 4.53212827e-01
-9.50569093e-01 -2.92291671e-01 4.60752621e-02 5.38339436e-01
1.58680692e-01 -8.65985930e-01 1.18344104e+00 1.58279925e-03
-4.53139365e-01 8.27365220e-02 -1.76214352e-01 2.33169973e-01
9.00654554e-01 -1.23129889e-01 -3.47526848e-01 -1.22622132e-01
-7.57251799e-01 1.35632113e-01 1.98168889e-01 3.55995834e-01
4.44210351e-01 -1.28160310e+00 -6.47917569e-01 1.76148564e-01
-5.30159920e-02 -1.97325442e-02 5.13997208e-03 1.00408161e+00
-6.12926371e-02 2.30854079e-01 1.16581410e-01 -4.77692097e-01
-1.16478145e+00 4.16277856e-01 5.41261256e-01 -3.35380942e-01
-4.35412884e-01 1.12898684e+00 -2.86468416e-01 -6.58094466e-01
5.75478852e-01 -2.87034005e-01 -3.09504509e-01 1.69418201e-01
2.00080857e-01 3.33733678e-01 3.46276283e-01 -6.69360936e-01
-8.84676278e-02 1.02061346e-01 -2.84861773e-01 1.95125893e-01
1.38616550e+00 -6.63967878e-02 8.92876536e-02 5.53757191e-01
1.27305794e+00 -4.09819007e-01 -1.52529263e+00 -4.62866664e-01
4.28587347e-01 -8.42697322e-02 2.69230306e-01 -1.18435037e+00
-9.98130441e-01 6.82453692e-01 8.96629870e-01 5.99448740e-01
1.13239288e+00 -1.84666649e-01 4.74003196e-01 5.98814309e-01
-2.60293484e-01 -1.60359156e+00 -6.41039163e-02 7.12225437e-01
6.69298649e-01 -1.44567156e+00 -1.32358065e-02 2.93834031e-01
-9.87889051e-01 9.44872141e-01 7.38852322e-01 3.23631853e-01
7.89000869e-01 1.49094695e-02 6.21323436e-02 -6.78131878e-02
-9.28360343e-01 -1.05289370e-01 3.30693096e-01 4.25931185e-01
3.09280396e-01 -5.35665788e-02 -1.83318809e-01 7.52592564e-01
-4.24639225e-01 -2.13006753e-02 3.16860825e-01 6.82767570e-01
-6.23649895e-01 -7.88906276e-01 -1.19257219e-01 6.26398742e-01
-5.39378166e-01 -1.27332183e-02 -5.01844168e-01 4.30840522e-01
5.28422356e-01 8.34611237e-01 -3.45713317e-01 -6.57999814e-01
2.74287194e-01 3.67121309e-01 -4.71926443e-02 -5.00472605e-01
-8.70173991e-01 -1.14265412e-01 -8.57219752e-03 -1.92780256e-01
-6.08389914e-01 -5.66422343e-01 -1.21438420e+00 -1.50333136e-01
-6.85488403e-01 3.64600062e-01 6.69595480e-01 1.07367051e+00
2.45634347e-01 5.25034964e-01 6.31181061e-01 -5.32067418e-01
-8.65114629e-01 -1.03200531e+00 -3.41382682e-01 4.19727355e-01
4.02000129e-01 -5.38448334e-01 -8.43032897e-01 1.84738293e-01] | [9.254318237304688, 3.84490966796875] |
76b73b08-81a6-43a7-a978-fcb96acd31a4 | real-time-incremental-speech-to-speech | null | null | https://aclanthology.org/N12-1048 | https://aclanthology.org/N12-1048.pdf | Real-time Incremental Speech-to-Speech Translation of Dialogs | null | ['Srinivas Bangalore', 'Vivek Kumar Rangarajan Sridhar', 'Ladan Golipour', 'Prakash Kolan', 'Aura Jimenez'] | 2012-06-01 | null | null | null | naacl-2012-6 | ['speech-to-speech-translation'] | ['speech'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.299546241760254, 3.723621129989624] |
bd9cb26e-68aa-41d5-a145-e6bd8e11dfbb | rethinking-audio-visual-synchronization-for | 2206.10421 | null | https://arxiv.org/abs/2206.10421v2 | https://arxiv.org/pdf/2206.10421v2.pdf | Rethinking Audio-visual Synchronization for Active Speaker Detection | Active speaker detection (ASD) systems are important modules for analyzing multi-talker conversations. They aim to detect which speakers or none are talking in a visual scene at any given time. Existing research on ASD does not agree on the definition of active speakers. We clarify the definition in this work and require synchronization between the audio and visual speaking activities. This clarification of definition is motivated by our extensive experiments, through which we discover that existing ASD methods fail in modeling the audio-visual synchronization and often classify unsynchronized videos as active speaking. To address this problem, we propose a cross-modal contrastive learning strategy and apply positional encoding in attention modules for supervised ASD models to leverage the synchronization cue. Experimental results suggest that our model can successfully detect unsynchronized speaking as not speaking, addressing the limitation of current models. | ['ChangShui Zhang', 'Zhiyao Duan', 'You Zhang', 'Abudukelimu Wuerkaixi'] | 2022-06-21 | null | null | null | null | ['audio-visual-synchronization', 'audio-visual-synchronization'] | ['audio', 'computer-vision'] | [ 1.20944746e-01 1.65881217e-01 -2.45929137e-01 -3.22670162e-01
-7.21567869e-01 -6.00615621e-01 5.94032824e-01 -9.43167210e-02
3.13570462e-02 9.48930308e-02 6.47759378e-01 2.75635272e-01
5.94182201e-02 -5.90172447e-02 -3.04377943e-01 -8.73081326e-01
-1.16552375e-01 1.40546709e-01 1.66778788e-01 2.63256170e-02
2.63931006e-01 3.14146608e-01 -1.83011973e+00 5.94585121e-01
4.74327445e-01 3.35275441e-01 1.81540444e-01 9.01804745e-01
-2.36413151e-01 1.14879072e+00 -9.59292591e-01 6.09464087e-02
-2.69483179e-01 -7.74758697e-01 -9.22557592e-01 3.58655393e-01
5.91755331e-01 -4.54128414e-01 -3.59084725e-01 8.29663277e-01
6.36128902e-01 -9.09684633e-04 2.01449886e-01 -1.67473280e+00
-3.90408069e-01 1.01977801e+00 -4.51161683e-01 9.81358528e-01
8.18657458e-01 1.28801465e-01 1.14646542e+00 -1.08305097e+00
3.54256034e-01 1.39589190e+00 3.61578047e-01 6.35421038e-01
-9.27955508e-01 -7.09578574e-01 7.79572070e-01 6.15919650e-01
-1.42201376e+00 -1.19077802e+00 1.04357123e+00 -4.74735320e-01
9.21573937e-01 3.95539016e-01 9.73776579e-01 1.46639204e+00
-1.65483356e-01 1.12870967e+00 5.68986237e-01 -4.81033564e-01
-1.68027077e-02 -1.07770644e-01 1.95603698e-01 3.72689813e-01
-9.66141447e-02 -1.64010569e-01 -1.34102046e+00 5.71766719e-02
3.55194539e-01 -2.18390837e-01 -4.32149082e-01 -2.13848829e-01
-1.48651934e+00 5.96118808e-01 3.72480415e-02 6.83495879e-01
8.02541524e-03 -7.49453902e-02 2.83245534e-01 3.06141019e-01
7.34457910e-01 2.17949897e-01 7.41330087e-02 -3.26568604e-01
-1.05694413e+00 -1.24398105e-01 5.63820779e-01 7.15527654e-01
4.17209268e-01 3.39517929e-02 -2.93671161e-01 6.22432828e-01
4.11591589e-01 2.94915944e-01 4.01179582e-01 -9.92735147e-01
3.52485418e-01 5.19078434e-01 -4.41136062e-02 -9.80483830e-01
-4.64381486e-01 -2.32714757e-01 -2.94808596e-01 -1.57076478e-01
2.99140632e-01 1.93148121e-01 -2.38744557e-01 1.73542213e+00
4.28931177e-01 5.90744019e-01 -1.24815173e-01 9.91151631e-01
1.14892638e+00 3.42960089e-01 -1.61116958e-01 -7.06353903e-01
1.04484141e+00 -1.10339212e+00 -1.20520055e+00 -3.44064951e-01
2.71855146e-01 -8.63554776e-01 1.05671859e+00 1.78583324e-01
-1.24320483e+00 -6.37092292e-01 -9.95630026e-01 1.95527688e-01
-3.46307345e-02 5.40221632e-02 4.02646899e-01 5.54871142e-01
-1.18196511e+00 -1.81305051e-01 -1.14051139e+00 -5.21856904e-01
1.59318849e-01 2.15321645e-01 -2.44620129e-01 3.06476951e-01
-8.97545516e-01 4.59568888e-01 -2.05082446e-01 1.89656377e-01
-1.33001399e+00 -3.80525559e-01 -7.14653134e-01 2.69673970e-02
2.62816519e-01 -5.25896251e-01 1.61588240e+00 -1.58040202e+00
-1.38319016e+00 1.14246941e+00 -7.16402829e-01 -3.03427935e-01
2.89234072e-01 -1.29500940e-01 -6.33460999e-01 6.09364629e-01
4.98485230e-02 4.24595475e-01 1.14300776e+00 -1.14224005e+00
-5.48160136e-01 -3.95256698e-01 3.43044788e-01 5.37619591e-01
-6.05562389e-01 5.14839292e-01 -5.14102876e-01 -6.39539599e-01
4.52266186e-01 -8.64414454e-01 5.36581993e-01 -5.89514002e-02
-3.37724388e-01 -3.50712329e-01 1.28631020e+00 -3.76491964e-01
1.34844148e+00 -2.34799266e+00 1.28056213e-01 -3.25867772e-01
6.90778971e-01 -9.49594155e-02 -3.38874422e-02 4.65332031e-01
-3.14307719e-01 -2.56951720e-01 2.47164413e-01 -7.16642201e-01
-2.21055731e-01 4.05289046e-02 -4.48780119e-01 7.93585896e-01
5.97207472e-02 5.19940078e-01 -1.00770020e+00 -7.30125904e-01
3.31496634e-02 5.80156207e-01 -4.78223681e-01 5.66003442e-01
3.13411057e-01 9.44046676e-01 -8.75888169e-02 6.40175700e-01
4.91218656e-01 -2.54079640e-01 2.11712345e-01 -2.01153174e-01
-3.07139337e-01 6.12853825e-01 -1.06797254e+00 1.55753112e+00
-1.66432157e-01 1.29852879e+00 4.38141912e-01 -1.26210892e+00
5.22796988e-01 8.39825928e-01 7.05049992e-01 -5.47787786e-01
-1.86831743e-01 -1.45171359e-01 2.28559569e-01 -9.54110384e-01
2.31048539e-01 4.20752823e-01 2.10989326e-01 4.06657010e-01
1.03861727e-01 2.41160035e-01 9.97157842e-02 3.74828190e-01
1.10124004e+00 -3.19062263e-01 1.69697583e-01 -7.10014403e-02
5.53234041e-01 -4.09336686e-01 4.94123846e-01 8.11718106e-01
-7.68448889e-01 7.20314205e-01 5.23439288e-01 -1.00704826e-01
-4.98570085e-01 -1.15073693e+00 1.92982048e-01 1.58338952e+00
2.80265063e-01 -5.87010086e-01 -8.88587594e-01 -5.95875740e-01
-5.20263553e-01 3.21847051e-01 -6.75305486e-01 -2.02335212e-02
-7.25820720e-01 -4.70853448e-01 4.58968073e-01 5.48458159e-01
2.26745278e-01 -8.26213777e-01 -5.93426466e-01 -1.31476209e-01
-6.82770193e-01 -1.23040092e+00 -7.07903862e-01 -2.18132153e-01
-2.82042414e-01 -1.27803719e+00 -7.18064785e-01 -9.10640419e-01
6.93620443e-01 8.67915392e-01 1.18767834e+00 1.45769995e-02
-2.14492127e-01 1.17283463e+00 -3.06344062e-01 -3.39914083e-01
-4.57752645e-01 -2.80191541e-01 2.00153068e-01 3.37356925e-01
2.55184889e-01 -5.47403216e-01 -8.10272932e-01 6.09269202e-01
-4.80617195e-01 7.89367110e-02 2.50645727e-03 5.25067091e-01
9.13020223e-02 -3.56063426e-01 5.98208308e-01 -2.12513134e-01
4.18737948e-01 -3.51969719e-01 -2.86165804e-01 1.54900312e-01
-4.58690301e-02 -5.33673465e-01 3.64756919e-02 -6.48185551e-01
-9.90594506e-01 8.62552375e-02 7.46442750e-02 -6.11740589e-01
-1.39167890e-01 1.24104649e-01 -2.23264769e-01 1.54165342e-01
4.06659126e-01 2.35083386e-01 -1.04442485e-01 -1.40542790e-01
-5.64066507e-03 7.14518070e-01 4.30617839e-01 -8.43927860e-02
2.28287026e-01 9.48609054e-01 -6.35584652e-01 -1.21869147e+00
-1.04400504e+00 -8.26779187e-01 -6.96197212e-01 -8.85323226e-01
9.62195694e-01 -1.24119937e+00 -9.73535538e-01 5.14710188e-01
-1.20509887e+00 -1.28345788e-01 6.67609647e-03 7.40526438e-01
-5.72251678e-01 4.52033162e-01 -4.97171879e-01 -1.15577996e+00
1.18246418e-03 -1.28224409e+00 1.24186516e+00 3.98325622e-02
-6.71302736e-01 -7.94272363e-01 3.96286279e-01 6.92820728e-01
3.43784928e-01 -1.39072970e-01 2.07556844e-01 -5.94451427e-01
-5.97226679e-01 1.58839852e-01 3.07447553e-01 -3.96096408e-02
3.54499310e-01 3.09735596e-01 -1.58141923e+00 -4.23704624e-01
2.40459263e-01 1.00677907e-01 7.27544606e-01 5.54620147e-01
9.43861783e-01 -3.57720375e-01 -4.05032694e-01 2.86654681e-01
5.40913880e-01 3.05937469e-01 1.79941207e-01 -2.02920660e-01
6.08475506e-01 8.21532190e-01 4.40728188e-01 4.86050516e-01
4.82963830e-01 1.03120482e+00 6.26048267e-01 -1.39530510e-01
-4.36813414e-01 -1.18132522e-02 8.70946884e-01 1.11909485e+00
1.12960123e-01 -3.54739994e-01 -1.21022701e+00 8.18253934e-01
-1.81709731e+00 -1.28780341e+00 -1.30889118e-02 2.09575415e+00
6.10565066e-01 4.66553383e-02 3.62458646e-01 3.57927769e-01
1.00862825e+00 4.09350842e-01 -4.01611328e-01 8.02413747e-02
-2.62740940e-01 -3.33633423e-01 -3.76694381e-01 5.03213108e-01
-1.14026058e+00 7.04688132e-01 6.96721601e+00 2.48017237e-01
-1.42842245e+00 3.71752888e-01 3.12096387e-01 -5.32402515e-01
6.45578727e-02 -1.46421313e-01 -7.21905112e-01 4.50359017e-01
7.92574406e-01 -1.41881809e-01 2.57357270e-01 5.63517869e-01
6.79914296e-01 -5.93115687e-02 -1.48657691e+00 1.31853712e+00
5.28397083e-01 -1.05855262e+00 -2.71499574e-01 -2.40416303e-01
3.41140151e-01 2.86632758e-02 1.22245729e-01 2.77303206e-03
1.62407309e-02 -8.87429178e-01 9.19997036e-01 5.34196079e-01
2.81016141e-01 -4.95109230e-01 1.95719779e-01 3.09724241e-01
-1.49735892e+00 -2.23783240e-01 3.11947018e-01 -7.11082295e-02
1.47335783e-01 8.37688297e-02 -1.04473817e+00 8.93048570e-02
1.07725966e+00 9.24036980e-01 -5.47204733e-01 8.21576118e-01
-1.58576548e-01 8.88512671e-01 -1.20940402e-01 2.54135370e-01
-2.83111870e-01 2.85423875e-01 1.02958405e+00 1.08747602e+00
1.47373229e-01 -3.74897748e-01 1.03055380e-01 8.41912687e-01
3.60096127e-01 -9.81194600e-02 -6.90873325e-01 -2.10187018e-01
6.80987418e-01 9.12094474e-01 -8.22109640e-01 -2.67233074e-01
-6.74467266e-01 1.03323674e+00 3.58820260e-02 3.88964266e-01
-9.52817619e-01 8.69895443e-02 6.81917667e-01 1.08305871e-01
1.96719393e-01 -4.00662661e-01 9.01524797e-02 -1.30531192e+00
7.82581344e-02 -9.14518416e-01 4.65880692e-01 -9.55902576e-01
-9.45837796e-01 3.67336035e-01 3.16081159e-02 -1.49757695e+00
-3.77968580e-01 4.41066511e-02 -9.45503533e-01 2.98292637e-01
-1.12846899e+00 -1.12698841e+00 -6.98941767e-01 9.04148877e-01
9.79348779e-01 -1.97735772e-01 7.34387100e-01 3.25885594e-01
-8.97932410e-01 6.79778397e-01 -3.34346622e-01 2.37276465e-01
9.88341928e-01 -9.26603198e-01 1.40784889e-01 9.14265454e-01
5.06321788e-01 6.13443553e-01 9.12036598e-01 -1.46257907e-01
-1.37067461e+00 -5.20900548e-01 9.34898376e-01 -4.93620396e-01
7.38479078e-01 -7.01448023e-01 -7.33303130e-01 8.24894011e-01
6.67288601e-01 5.90236066e-03 8.11555266e-01 9.46943462e-02
-2.29547933e-01 -1.62064299e-01 -7.41828561e-01 4.06624526e-01
1.10536671e+00 -8.45943213e-01 -7.22188413e-01 3.36782634e-01
6.52235806e-01 -2.79003739e-01 -3.62323403e-01 1.53730527e-01
4.68093693e-01 -1.40106821e+00 1.09190130e+00 -2.63790816e-01
2.74341851e-02 -2.62830496e-01 -1.20675173e-02 -9.98182714e-01
2.81741079e-02 -7.37588465e-01 -4.41110998e-01 1.53451872e+00
-7.62691274e-02 -4.83855516e-01 4.50764805e-01 -1.41261682e-01
-2.73194224e-01 -1.83315307e-01 -1.23101020e+00 -5.15598416e-01
-5.21749079e-01 -3.88623655e-01 2.77359515e-01 1.34222233e+00
4.80313212e-01 3.73120964e-01 -2.15831935e-01 5.12232900e-01
3.88807833e-01 -6.12300122e-03 7.38427341e-01 -1.14915705e+00
-2.28277832e-01 -4.40654814e-01 -4.31305259e-01 -1.05996108e+00
4.38151717e-01 -3.51100117e-01 -6.99495105e-03 -1.33765793e+00
3.35287660e-01 2.48210683e-01 -2.10748851e-01 3.68979156e-01
4.28357087e-02 1.66143671e-01 -5.16674593e-02 5.17026782e-01
-8.90610635e-01 4.66919094e-01 9.65083361e-01 -3.50376338e-01
-1.70506626e-01 2.46661827e-02 -4.78910744e-01 8.49045932e-01
6.18530452e-01 -2.22343922e-01 -5.90293765e-01 -3.98995191e-01
2.15738133e-01 9.92904231e-02 4.87546176e-01 -9.50294375e-01
4.87365484e-01 1.52946925e-02 4.25255969e-02 -6.69828951e-01
6.04955077e-01 -6.96210563e-01 -2.25800321e-01 2.44415000e-01
-5.87835908e-01 1.75553620e-01 4.03053947e-02 5.03057301e-01
-5.65268159e-01 2.38398239e-01 7.66322672e-01 7.88919926e-02
-6.06887817e-01 -4.48341258e-02 -6.85711682e-01 9.65125337e-02
1.02575862e+00 -2.13395819e-01 -4.08910811e-01 -6.58851922e-01
-1.19591403e+00 6.92912042e-02 8.26236680e-02 7.86425829e-01
4.91150647e-01 -1.14371037e+00 -5.50409913e-01 2.99863368e-01
2.90883720e-01 -3.14101309e-01 5.63900292e-01 1.35692358e+00
-1.92387830e-02 3.73425335e-01 1.84801564e-01 -1.18878746e+00
-1.99518049e+00 4.60965365e-01 5.43997347e-01 2.90814459e-01
-6.71471477e-01 9.43237185e-01 5.69715202e-01 2.02219784e-01
6.73433244e-01 -3.04162621e-01 -5.00003099e-01 5.25653243e-01
6.82647049e-01 3.28894556e-01 1.11377575e-01 -9.90117610e-01
-6.31692290e-01 3.72165531e-01 4.28418890e-02 -1.17849924e-01
9.71441984e-01 -7.52142549e-01 1.38338283e-02 1.08328187e+00
1.03127789e+00 4.54749137e-01 -1.19576192e+00 -1.77185655e-01
-2.09850028e-01 -2.85070866e-01 8.01387504e-02 -2.14756012e-01
-1.09681249e+00 9.80689645e-01 9.67692733e-01 7.14325070e-01
1.09550893e+00 5.91493607e-01 1.91789895e-01 2.84013301e-01
-1.16980039e-01 -1.03072810e+00 6.49818003e-01 3.48698825e-01
1.05342472e+00 -1.27075398e+00 -2.89324671e-01 -5.27116537e-01
-5.79853356e-01 1.04996204e+00 6.99809134e-01 2.71626383e-01
5.68298101e-01 4.65456486e-01 3.60004425e-01 -3.33923310e-01
-9.35692489e-01 -3.38861555e-01 1.47285074e-01 5.83791196e-01
6.78969204e-01 -3.25284868e-01 1.83792949e-01 3.23122710e-01
-2.62751162e-01 -5.46625555e-01 3.73348832e-01 9.40055192e-01
-3.33657950e-01 -5.01558244e-01 -6.38059199e-01 -4.16047841e-01
-3.09570163e-01 1.50139734e-01 -7.65171528e-01 6.01128757e-01
-4.27951152e-03 1.44268906e+00 7.26653755e-01 -2.69244641e-01
2.36217864e-02 1.40722662e-01 5.67821085e-01 -6.11410737e-01
-5.43140709e-01 4.04495686e-01 -3.96415181e-02 -5.05177557e-01
-1.14147007e+00 -8.49219084e-01 -9.00072038e-01 9.74384136e-03
-2.48976827e-01 1.06887795e-01 3.00745666e-01 9.56638277e-01
3.26572627e-01 8.04478824e-01 9.46289778e-01 -8.87192667e-01
3.43920226e-04 -9.64836657e-01 -3.85716349e-01 2.24165589e-01
9.78770614e-01 -7.43212938e-01 -9.54331160e-01 3.94486666e-01] | [14.45088005065918, 5.124743461608887] |
3e3b1397-6904-4099-966a-3ac78f414a54 | direct-posenet-absolute-pose-regression-with | 2104.04073 | null | https://arxiv.org/abs/2104.04073v2 | https://arxiv.org/pdf/2104.04073v2.pdf | Direct-PoseNet: Absolute Pose Regression with Photometric Consistency | We present a relocalization pipeline, which combines an absolute pose regression (APR) network with a novel view synthesis based direct matching module, offering superior accuracy while maintaining low inference time. Our contribution is twofold: i) we design a direct matching module that supplies a photometric supervision signal to refine the pose regression network via differentiable rendering; ii) we modify the rotation representation from the classical quaternion to SO(3) in pose regression, removing the need for balancing rotation and translation loss terms. As a result, our network Direct-PoseNet achieves state-of-the-art performance among all other single-image APR methods on the 7-Scenes benchmark and the LLFF dataset. | ['Victor Prisacariu', 'ZiRui Wang', 'Shuai Chen'] | 2021-04-08 | null | null | null | null | ['camera-relocalization'] | ['computer-vision'] | [ 2.43764564e-01 2.14075178e-01 -1.44805968e-01 -5.32794654e-01
-7.67332673e-01 -5.03475904e-01 8.16920578e-01 -6.08002663e-01
-3.10465872e-01 5.63218176e-01 1.33973556e-02 -1.21892229e-01
1.71400130e-01 -6.62122190e-01 -1.17725778e+00 -3.82173896e-01
3.37322593e-01 5.52999020e-01 3.11647385e-01 -5.29940367e-01
1.37745172e-01 7.73054302e-01 -1.34642982e+00 -2.06034049e-01
6.23175204e-01 1.12437248e+00 -2.79698104e-01 5.20759940e-01
5.33665657e-01 6.63609684e-01 -4.33064163e-01 -6.54721797e-01
7.20888734e-01 -7.34752268e-02 -6.84111118e-01 -6.87228143e-02
1.34615624e+00 -8.04336190e-01 -4.41956371e-01 7.14759707e-01
5.48874438e-01 6.75232857e-02 2.00722367e-01 -1.16732025e+00
-6.04897499e-01 1.07619323e-01 -8.25186551e-01 -3.48956376e-01
5.32460988e-01 2.11677387e-01 1.18101537e+00 -1.05545092e+00
1.07361221e+00 1.51675439e+00 9.04489398e-01 3.15514326e-01
-1.59429884e+00 -6.06038928e-01 1.04216702e-01 1.67433769e-01
-1.46840703e+00 -3.77659023e-01 8.53816569e-01 -1.85368225e-01
1.13998103e+00 1.76289707e-01 8.69080603e-01 1.17732036e+00
2.99315482e-01 4.99100327e-01 1.10378873e+00 -8.56289789e-02
-1.12456478e-01 -4.30335999e-01 -4.26169038e-01 9.83033299e-01
-2.20638476e-02 3.27724904e-01 -8.28943253e-01 2.65372753e-01
1.11212885e+00 -2.06436738e-01 -2.89730400e-01 -8.75070810e-01
-1.33477938e+00 5.98791420e-01 8.37502897e-01 -5.95868170e-01
-1.87250003e-01 7.25038707e-01 1.66980714e-01 3.49163234e-01
6.91978395e-01 4.98758763e-01 -6.02275431e-01 -1.74705870e-02
-9.17359769e-01 3.09297204e-01 7.62600422e-01 1.11990249e+00
1.14696610e+00 1.33612975e-01 -6.16044085e-03 5.83822489e-01
4.15008157e-01 8.19841862e-01 -3.21525961e-01 -1.44238544e+00
5.11129797e-01 2.91538328e-01 -1.52558118e-01 -1.18123138e+00
-4.31530327e-01 -6.27661645e-01 -6.04222596e-01 4.30895537e-01
2.33444851e-02 1.81153730e-01 -7.29007006e-01 1.73825002e+00
3.95018667e-01 3.86913031e-01 -3.08640689e-01 1.03069413e+00
9.38942015e-01 2.44808495e-01 -3.11221719e-01 2.32243374e-01
1.26857233e+00 -1.18204439e+00 -3.76201838e-01 -1.48136318e-01
8.60785991e-02 -8.62468600e-01 1.06222332e+00 4.78257328e-01
-1.17393899e+00 -6.74730480e-01 -1.46624053e+00 -8.06613863e-01
3.69576253e-02 -1.54291876e-02 7.23457694e-01 2.97623634e-01
-1.31935012e+00 7.06620276e-01 -9.18974519e-01 -1.26841769e-01
3.01822811e-01 5.04324913e-01 -5.12743056e-01 7.45334551e-02
-8.84007394e-01 8.71692836e-01 -1.96622368e-02 2.24233031e-01
-6.40736341e-01 -1.14751410e+00 -1.05713117e+00 -2.42028296e-01
4.74756926e-01 -1.34856522e+00 1.15889800e+00 -8.48491907e-01
-1.99336886e+00 1.01893806e+00 -9.03149992e-02 -2.44625002e-01
9.08312380e-01 -5.28215945e-01 7.16340244e-02 3.45856577e-01
-9.91621464e-02 1.28898168e+00 8.78872335e-01 -1.23100400e+00
-1.91226587e-01 -2.75041372e-01 1.70745552e-01 4.76066917e-01
2.09433094e-01 -2.96534419e-01 -8.66695523e-01 -7.53020108e-01
4.06723708e-01 -1.11082721e+00 6.30808473e-02 6.38722241e-01
-4.70561951e-01 3.55111323e-02 9.06246603e-01 -6.87337637e-01
6.11536801e-01 -1.84304786e+00 4.26434070e-01 1.80640548e-01
3.15517515e-01 -2.59749535e-02 -1.54892400e-01 1.54098958e-01
-1.47153094e-01 -2.15262949e-01 3.89763117e-02 -8.41637373e-01
1.90622769e-02 3.45581293e-01 -5.32724917e-01 7.03980565e-01
5.37631392e-01 1.14202178e+00 -7.34325826e-01 -2.45838270e-01
5.05375564e-01 7.86304593e-01 -9.98331308e-01 2.00464681e-01
-2.52907574e-01 6.09711766e-01 4.75599431e-03 6.64286435e-01
8.25416088e-01 -1.57413736e-01 -3.53835188e-02 -7.39514232e-01
-2.31280565e-01 5.54582000e-01 -1.29766726e+00 2.28486872e+00
-4.14347976e-01 5.56975424e-01 3.86827849e-02 -2.68825829e-01
9.74619448e-01 -9.25973579e-02 5.63805878e-01 -6.48565948e-01
7.13165328e-02 -2.16790847e-03 -4.57241148e-01 1.12909451e-02
6.22581244e-01 1.12997681e-01 1.91377103e-01 -1.09509796e-01
1.73713550e-01 -6.45483911e-01 -7.32192546e-02 3.20637645e-03
7.88719535e-01 1.24222386e+00 8.90795663e-02 -1.89369544e-01
5.07351995e-01 -1.29135400e-01 7.15513587e-01 9.85966474e-02
1.73063070e-01 1.22963762e+00 6.80908084e-01 -5.18135905e-01
-9.93269861e-01 -1.36570382e+00 -2.02194050e-01 8.09758604e-01
2.71797895e-01 -5.35489917e-01 -4.30238545e-01 -3.53908449e-01
1.43397957e-01 3.64989698e-01 -4.72789198e-01 5.90313971e-02
-8.71489406e-01 -2.83969790e-01 6.55000746e-01 7.74925530e-01
7.18542397e-01 -5.22440195e-01 -3.96036774e-01 -4.08509448e-02
-1.30031809e-01 -1.27380025e+00 -5.68235636e-01 5.07580563e-02
-9.00600731e-01 -1.12442315e+00 -2.70967275e-01 -5.00969410e-01
6.20903790e-01 2.16087952e-01 1.40829217e+00 -9.83423367e-02
-1.42888308e-01 2.72424489e-01 7.67952725e-02 -1.75704688e-01
3.98269147e-02 7.78897181e-02 -1.06036797e-01 -3.68710101e-01
-8.02136287e-02 -8.00234199e-01 -8.09164286e-01 5.03936708e-01
-5.84119976e-01 4.21927571e-01 4.38702047e-01 6.60947084e-01
1.07103503e+00 -8.80923510e-01 -4.81093787e-02 -5.32681763e-01
-6.98034987e-02 1.45935804e-01 -1.02978337e+00 -8.93766955e-02
-7.56761849e-01 1.70779005e-01 4.19761956e-01 -7.32003599e-02
-9.44642544e-01 3.18026960e-01 -3.58792961e-01 -6.89430177e-01
2.71396726e-01 2.83991918e-02 -1.38115123e-01 -4.49443668e-01
5.51011741e-01 -2.31263310e-01 3.24344933e-01 -5.20177305e-01
7.67253757e-01 -3.09885964e-02 9.64781165e-01 -6.56557739e-01
1.21265829e+00 8.33930492e-01 6.34504855e-01 -5.28919816e-01
-8.85658562e-01 -3.26382905e-01 -9.88035560e-01 -1.20890707e-01
9.76131976e-01 -1.24994421e+00 -1.09988356e+00 4.53901649e-01
-1.18406475e+00 -1.99129879e-01 -3.00035119e-01 4.33652818e-01
-7.65995979e-01 4.57482547e-01 -6.58420146e-01 -2.71004029e-02
-4.80182022e-01 -1.35830152e+00 1.73290884e+00 3.85778403e-04
-1.61950141e-01 -7.18863785e-01 2.15064228e-01 3.27482045e-01
2.30671033e-01 6.21697068e-01 3.61506522e-01 2.61762850e-02
-1.13376093e+00 1.89570159e-01 -5.29014170e-01 3.89269978e-01
-2.01165810e-01 4.17187214e-01 -1.05008662e+00 -4.14661139e-01
-2.70351768e-01 -4.74407822e-01 8.37592900e-01 1.05155095e-01
1.05970371e+00 -2.46612192e-03 1.38683215e-01 1.80958748e+00
1.43926239e+00 -4.46964532e-01 8.53186607e-01 5.03746986e-01
1.26979685e+00 3.29304874e-01 5.13131976e-01 2.28949830e-01
7.77793407e-01 9.60433602e-01 8.69473457e-01 -5.35852253e-01
-4.46535766e-01 -4.64436084e-01 3.39631468e-01 8.60824168e-01
-3.92078698e-01 2.47774407e-01 -5.38248658e-01 -9.43192616e-02
-1.78836668e+00 -6.60828292e-01 -2.49216884e-01 2.07173610e+00
8.24431539e-01 -2.72067171e-02 -7.48442858e-02 -2.40986004e-01
1.98847070e-01 5.07555485e-01 -5.14864266e-01 -2.31905356e-01
-1.89928323e-01 5.83851397e-01 7.76861370e-01 6.60747945e-01
-1.02389419e+00 9.73760486e-01 6.42478895e+00 3.75851572e-01
-1.35053718e+00 -1.20311409e-01 4.71140683e-01 -1.02448821e-01
-3.69117260e-01 1.96623832e-01 -7.94994891e-01 -2.45543092e-01
4.38274860e-01 4.40829486e-01 7.17254162e-01 8.31786752e-01
2.62292195e-02 1.17153578e-01 -1.34237492e+00 1.07364237e+00
3.37037086e-01 -1.28447402e+00 1.39127985e-01 2.15459485e-02
6.55933082e-01 4.02709454e-01 9.54435617e-02 1.10329024e-01
1.25477776e-01 -9.81402159e-01 1.02843904e+00 6.86418295e-01
1.17680287e+00 -7.62984097e-01 3.65593493e-01 -8.60574841e-02
-1.19610143e+00 4.99042988e-01 -2.52315521e-01 6.12371974e-02
2.34647676e-01 4.34818655e-01 -6.69821203e-01 9.37642515e-01
6.72741652e-01 1.06114256e+00 -7.61093378e-01 6.19470477e-01
-6.87765896e-01 1.76558182e-01 -4.45612341e-01 5.77081263e-01
6.99314624e-02 -5.88287115e-01 5.93043327e-01 8.30136418e-01
4.85050045e-02 -3.54355097e-01 1.03133373e-01 1.04746842e+00
-2.78290093e-01 -2.14016572e-01 -2.95401543e-01 6.33215070e-01
3.14075321e-01 1.37412906e+00 -4.14925218e-01 7.99167086e-04
-3.75250340e-01 1.08393228e+00 4.11434144e-01 1.79032609e-01
-8.95614624e-01 -1.14208914e-01 9.33731616e-01 1.43961757e-01
3.67150247e-01 -4.24500883e-01 -4.03742969e-01 -1.28921175e+00
2.64610916e-01 -8.26131165e-01 6.40232768e-03 -1.05763614e+00
-1.01812565e+00 2.86652505e-01 1.95530668e-01 -1.31329751e+00
-4.14345026e-01 -8.33444476e-01 -2.33085662e-01 9.62186754e-01
-1.97087681e+00 -1.60869730e+00 -6.48410380e-01 4.65417594e-01
2.16341376e-01 1.20034516e-01 6.38065338e-01 3.03675056e-01
-4.39215243e-01 5.73653102e-01 -3.72176349e-01 4.51884642e-02
1.13852501e+00 -1.43519592e+00 8.04856896e-01 6.93207383e-01
-3.53001729e-02 8.88239086e-01 3.25914234e-01 -3.88113439e-01
-1.86893034e+00 -1.05721974e+00 4.92054999e-01 -6.99906826e-01
5.10889590e-01 -4.62843150e-01 -4.55641270e-01 9.20311153e-01
2.34113619e-01 3.51936698e-01 9.09569412e-02 -1.88019071e-02
-6.98502004e-01 -4.68030751e-01 -9.72431421e-01 7.62443066e-01
1.32159376e+00 -5.80692530e-01 -4.53035980e-01 5.96575104e-02
9.59514499e-01 -1.15884101e+00 -1.09082913e+00 5.90308726e-01
7.05629170e-01 -1.09671235e+00 1.47609246e+00 -1.01805814e-01
6.31612897e-01 -6.27135873e-01 -1.52382806e-01 -1.13802838e+00
-3.49530727e-01 -9.69101071e-01 -2.13781670e-01 1.02644360e+00
4.86868583e-02 -7.25014329e-01 5.88832080e-01 8.03458914e-02
-2.77237475e-01 -9.92136121e-01 -1.03937352e+00 -5.71184576e-01
-2.49309212e-01 -3.84311050e-01 6.36562884e-01 6.83902383e-01
-8.19579303e-01 4.94875789e-01 -4.54855859e-01 -1.16428472e-02
6.79017365e-01 -1.12424329e-01 1.32486033e+00 -1.11328292e+00
-5.76237142e-01 -2.50247627e-01 -5.49289525e-01 -1.51409030e+00
2.02251047e-01 -8.31828952e-01 -3.75210308e-02 -1.29992473e+00
-1.15022957e-01 -1.48602962e-01 1.07191853e-01 4.04465407e-01
3.01753357e-02 8.36626470e-01 2.63638675e-01 2.47250617e-01
-4.10991102e-01 7.60913074e-01 1.55215633e+00 1.72039688e-01
1.47889778e-02 -2.08304077e-01 -4.77429569e-01 8.55033100e-01
3.01302195e-01 -1.81111842e-01 -3.97174835e-01 -6.01884127e-01
6.71277940e-01 -1.33694172e-01 6.46663189e-01 -8.59721839e-01
-3.24314125e-02 -9.51582938e-02 5.90257943e-01 -9.51268792e-01
6.60647511e-01 -7.31508136e-01 1.97127774e-01 2.96129078e-01
2.14707460e-02 4.26886648e-01 1.92980822e-02 3.80511224e-01
1.28231272e-01 3.40012819e-01 8.25813651e-01 2.32765034e-01
-6.80255473e-01 4.42955911e-01 4.37286913e-01 1.25917524e-01
6.28646255e-01 -2.78120637e-01 -5.87069750e-01 -2.82529444e-01
-2.31851146e-01 2.47094303e-01 9.11397398e-01 5.47068417e-01
4.40787315e-01 -1.45503533e+00 -6.31995142e-01 3.43729019e-01
9.94625464e-02 5.80851734e-01 -3.63884479e-01 9.70525920e-01
-1.20098531e+00 1.07756287e-01 -1.88453645e-01 -9.82084095e-01
-1.15115464e+00 1.56687364e-01 4.45348054e-01 -1.19478978e-01
-7.32228994e-01 7.62615263e-01 6.19527549e-02 -9.39239919e-01
2.22286001e-01 -4.79459792e-01 2.32117444e-01 -1.10401884e-01
8.48552957e-02 6.58428371e-01 2.60309458e-01 -7.02938139e-01
-3.95957798e-01 7.98845112e-01 3.59832436e-01 -2.73994237e-01
1.50378251e+00 2.46504066e-03 -3.98706436e-01 2.13592425e-01
1.47933125e+00 7.05366060e-02 -1.72276640e+00 -1.22252926e-01
-4.95796472e-01 -4.12034124e-01 2.88706925e-02 -6.26833618e-01
-1.11968219e+00 5.82920313e-01 3.52292240e-01 -5.98467708e-01
1.00909722e+00 -2.66670793e-01 7.21596003e-01 6.07276559e-01
2.68732101e-01 -1.10006726e+00 2.32410401e-01 7.76467860e-01
1.24574757e+00 -1.03874660e+00 6.30766809e-01 -6.71166897e-01
-3.96210492e-01 1.24448705e+00 7.39238739e-01 -5.99283218e-01
5.31875193e-01 3.45468283e-01 2.23731741e-01 -2.68303961e-01
-5.98286867e-01 8.03084020e-03 8.51840198e-01 4.56330955e-01
5.15088975e-01 -2.94706821e-01 -1.51116192e-01 -1.43040895e-01
-6.52857184e-01 -2.39161655e-01 4.90547083e-02 8.11004460e-01
1.14963681e-01 -1.13596880e+00 -3.69852394e-01 -1.32013068e-01
-1.91874743e-01 -7.81232193e-02 -5.04218102e-01 1.01935840e+00
1.42723741e-02 4.94539171e-01 2.31251895e-01 -3.55662405e-01
6.63576066e-01 -3.12442660e-01 7.61360586e-01 -3.70070666e-01
-7.14697063e-01 2.63420999e-01 1.31973684e-01 -1.32110751e+00
-6.28857434e-01 -5.08144617e-01 -1.12969136e+00 -4.90089595e-01
-2.46391296e-01 -7.10312426e-01 7.75061131e-01 7.16131389e-01
4.72761512e-01 7.08020091e-01 6.95556879e-01 -1.38900030e+00
-7.42825747e-01 -6.62180305e-01 -8.21187496e-02 4.24634963e-01
3.08736414e-01 -6.70325220e-01 -8.83687139e-02 -4.67162346e-04] | [8.128260612487793, -2.3690221309661865] |
d4b7647d-3da3-4a9f-aca9-c476ced6f96e | pese-event-structure-extraction-using-pointer | 2211.12157 | null | https://arxiv.org/abs/2211.12157v1 | https://arxiv.org/pdf/2211.12157v1.pdf | PESE: Event Structure Extraction using Pointer Network based Encoder-Decoder Architecture | The task of event extraction (EE) aims to find the events and event-related argument information from the text and represent them in a structured format. Most previous works try to solve the problem by separately identifying multiple substructures and aggregating them to get the complete event structure. The problem with the methods is that it fails to identify all the interdependencies among the event participants (event-triggers, arguments, and roles). In this paper, we represent each event record in a unique tuple format that contains trigger phrase, trigger type, argument phrase, and corresponding role information. Our proposed pointer network-based encoder-decoder model generates an event tuple in each time step by exploiting the interactions among event participants and presenting a truly end-to-end solution to the EE task. We evaluate our model on the ACE2005 dataset, and experimental results demonstrate the effectiveness of our model by achieving competitive performance compared to the state-of-the-art methods. | ['Sudeshan Sarkar', 'Alapan Kuila'] | 2022-11-22 | null | null | null | null | ['event-extraction'] | ['natural-language-processing'] | [ 3.95989716e-01 2.12867916e-01 -3.03453386e-01 -5.39085329e-01
-1.00665843e+00 -6.10136271e-01 5.47406495e-01 8.37713480e-01
-4.84405607e-01 8.87522399e-01 8.39434564e-01 -8.34117234e-02
-2.24402070e-01 -8.69653881e-01 -7.76452184e-01 -7.50253052e-02
-4.65338558e-01 5.60944855e-01 4.62724805e-01 1.46856338e-01
-6.64825812e-02 7.55194277e-02 -1.51619065e+00 1.00541270e+00
4.49445993e-01 1.07499063e+00 7.68555552e-02 4.57070410e-01
-4.00862306e-01 1.31063652e+00 -7.59504974e-01 -4.61039603e-01
-1.79979250e-01 -6.69164717e-01 -9.06918108e-01 -4.01647091e-01
-2.66751438e-01 -3.43396574e-01 -1.85550556e-01 6.46377563e-01
5.00847459e-01 -1.96416248e-02 4.57143217e-01 -1.12988842e+00
-1.01872636e-02 1.37276983e+00 -3.75157446e-01 4.35368866e-01
5.88167906e-01 -5.66333890e-01 1.21874607e+00 -9.27246094e-01
8.94697130e-01 1.02554846e+00 4.87578332e-01 2.80294091e-01
-9.76952672e-01 -5.00823259e-01 1.32726341e-01 3.56504381e-01
-1.28873026e+00 -5.32228649e-01 5.58455884e-01 -5.79256052e-03
1.33482635e+00 3.44284713e-01 4.26312178e-01 1.13688874e+00
7.64458105e-02 9.30761576e-01 3.46219033e-01 -3.85892808e-01
3.25744987e-01 -3.84048939e-01 4.36413169e-01 5.05932033e-01
7.60975555e-02 4.99028377e-02 -9.18766737e-01 -3.71649027e-01
4.18814480e-01 -3.37421820e-02 -7.35873133e-02 1.90610334e-01
-1.54550827e+00 4.36264127e-01 8.91800523e-02 1.46988273e-01
-6.79700792e-01 2.81092316e-01 8.41321766e-01 3.07230093e-02
2.70230830e-01 8.84043798e-02 -7.45478988e-01 -3.83267134e-01
-7.65224040e-01 6.07289553e-01 1.05442703e+00 9.77265060e-01
1.84548289e-01 -3.69433433e-01 -7.78814077e-01 5.83646357e-01
-4.97976728e-02 -2.22495139e-01 3.21296364e-01 -6.14454091e-01
9.30763781e-01 7.63908029e-01 3.12820107e-01 -7.02834368e-01
-6.32650495e-01 -3.56696367e-01 -5.49415648e-01 -3.80614281e-01
2.96963453e-01 -4.64275569e-01 -4.64397967e-01 2.06551671e+00
4.33828741e-01 5.73531389e-01 2.09215239e-01 4.77974921e-01
1.17069876e+00 1.05986917e+00 3.29609960e-01 -4.89143729e-01
1.68192041e+00 -6.78543389e-01 -1.09341037e+00 -2.59609789e-01
5.59608519e-01 -5.24214745e-01 3.79098594e-01 2.04886392e-01
-1.37564003e+00 -2.68046260e-01 -8.80294085e-01 -8.08514729e-02
-2.71936860e-02 2.17634067e-01 5.86169541e-01 9.60388873e-03
-3.49096835e-01 4.71671730e-01 -9.65760112e-01 -6.84069991e-02
9.06495377e-02 2.66089231e-01 -3.17186743e-01 4.85216647e-01
-1.40110695e+00 5.79230130e-01 1.06923127e+00 -3.03549375e-02
-7.13164091e-01 -8.17331851e-01 -9.48239505e-01 4.72613186e-01
8.05312157e-01 -5.24388611e-01 1.74087322e+00 -3.47590297e-01
-9.79605317e-01 4.11333591e-01 -6.45656466e-01 -7.71128476e-01
6.92459941e-02 -1.94932699e-01 -7.40010083e-01 2.16805027e-03
9.86122265e-02 4.20202821e-01 1.11185297e-01 -8.69274139e-01
-1.37538767e+00 -1.84846848e-01 -2.08896007e-02 1.86563104e-01
-2.17438210e-02 7.26730406e-01 -7.02248454e-01 -7.40324199e-01
7.35790655e-02 -4.97473449e-01 -6.90648183e-02 -4.47458953e-01
-6.02764249e-01 -6.09953284e-01 4.46979165e-01 -6.34357274e-01
1.93194771e+00 -2.09210873e+00 9.34070796e-02 6.61773011e-02
1.53472573e-01 -1.30811721e-01 2.23098233e-01 8.06717217e-01
-4.21459258e-01 -1.87074132e-02 -2.08836213e-01 -3.68206501e-01
1.72674209e-01 2.29561582e-01 -6.09854937e-01 -5.56384958e-02
2.72961825e-01 8.28425884e-01 -1.05514777e+00 -7.07867444e-01
-1.51054874e-01 3.44097584e-01 -3.82813215e-01 3.63728225e-01
-4.37968254e-01 -1.50791509e-02 -4.71133113e-01 4.03053284e-01
2.13954762e-01 -2.48240978e-01 4.81528491e-01 -3.46066743e-01
-2.20259920e-01 1.01481259e+00 -1.61791861e+00 1.65647268e+00
-1.89086407e-01 4.13195729e-01 -4.07538325e-01 -6.39423788e-01
5.90195119e-01 8.83949220e-01 6.74752593e-01 -6.65857732e-01
-2.93927714e-02 1.45583615e-01 -1.34231910e-01 -4.80768591e-01
5.62006891e-01 1.36276886e-01 -6.00208104e-01 4.35538650e-01
7.61103779e-02 8.84173572e-01 6.99903131e-01 2.58727223e-01
1.25806153e+00 9.97006372e-02 6.82207644e-01 3.42532665e-01
3.25732678e-01 -5.13063595e-02 8.82829964e-01 6.66771531e-01
3.57946694e-01 5.40497541e-01 9.50821221e-01 -5.83356023e-01
-7.16371596e-01 -1.09586978e+00 -1.78664606e-02 8.31827641e-01
-1.48215503e-01 -1.15111113e+00 -5.19402504e-01 -8.46461296e-01
-4.36691999e-01 9.13157761e-01 -5.56798875e-01 4.39200737e-02
-1.04213977e+00 -8.47005129e-01 8.04629207e-01 8.57023060e-01
3.39674324e-01 -1.23853707e+00 -9.62271810e-01 9.39164042e-01
-8.06437850e-01 -1.29097772e+00 -4.89639550e-01 6.10578358e-01
-4.81332392e-01 -1.04242289e+00 -4.82399166e-02 -8.17778230e-01
2.80307859e-01 -6.05505228e-01 1.39095736e+00 -4.53022480e-01
1.80120375e-02 -4.72922057e-01 -3.65198851e-01 -6.34216368e-01
-2.85649180e-01 -1.97041854e-02 -5.22469163e-01 6.62770793e-02
2.48275399e-01 -4.46234435e-01 -4.06586856e-01 1.12115771e-01
-1.03682709e+00 1.49559379e-01 3.34331572e-01 5.06021976e-01
8.08404207e-01 4.13418472e-01 7.57920563e-01 -1.02510488e+00
6.18339360e-01 -6.53609157e-01 -4.43265557e-01 3.88275057e-01
-2.71505952e-01 3.66934448e-01 7.10384429e-01 -2.52214819e-01
-1.34080482e+00 3.10047984e-01 -1.04283825e-01 2.49274254e-01
-2.38546401e-01 9.87752259e-01 -2.79850513e-01 1.03447318e+00
4.67047542e-01 1.72624439e-01 -6.65569603e-01 -5.78386188e-01
2.56245583e-01 3.31733853e-01 9.80777204e-01 -7.75455177e-01
1.92271665e-01 3.45381737e-01 9.35305841e-03 -1.34524509e-01
-1.05088365e+00 -2.58910537e-01 -1.48946464e-01 6.77606370e-03
7.27739990e-01 -9.28813994e-01 -7.60530591e-01 9.16883349e-02
-1.74012387e+00 2.67420143e-01 -5.25397599e-01 5.44160485e-01
-3.13879967e-01 1.06807046e-01 -7.52002180e-01 -6.50422215e-01
-4.43358123e-01 -7.78063655e-01 9.57590640e-01 5.71070239e-02
-5.91941655e-01 -5.28839648e-01 1.96561411e-01 -4.26648110e-01
-2.06615806e-01 5.10443687e-01 1.16945255e+00 -1.00339758e+00
-4.36842203e-01 -2.26818845e-01 -4.50632647e-02 -4.02100772e-01
-1.53699711e-01 -2.28592053e-01 -4.28419232e-01 3.58265728e-01
-2.18800426e-01 7.38886744e-02 7.99700797e-01 2.42095575e-01
1.31463635e+00 -4.43607539e-01 -6.19687259e-01 4.29078251e-01
1.24731100e+00 6.43712997e-01 5.77860534e-01 9.08607766e-02
2.63953507e-01 5.14960945e-01 4.03975815e-01 7.20114052e-01
6.27626300e-01 8.19055438e-01 4.27572787e-01 2.63547804e-02
-7.15881512e-02 -5.69539309e-01 4.06999081e-01 2.91086674e-01
1.42540604e-01 -9.04841185e-01 -6.67682052e-01 7.72640109e-01
-2.19394255e+00 -1.31486619e+00 -4.93073285e-01 2.16639709e+00
1.06379223e+00 3.39483798e-01 2.21783802e-01 3.11734349e-01
6.96955025e-01 1.27659976e-01 -1.66793004e-01 -3.39638740e-01
-7.57506713e-02 3.51898909e-01 3.86813700e-01 5.09211272e-02
-1.20381403e+00 5.20965159e-01 6.39175129e+00 5.67508876e-01
-6.11559093e-01 -3.12969508e-03 4.08017188e-01 -3.69602174e-01
-1.49522766e-01 4.84325550e-02 -1.16119432e+00 5.81504762e-01
1.38346040e+00 -3.15852821e-01 7.31529593e-02 2.90254146e-01
2.47321874e-01 -2.97700055e-02 -1.72380555e+00 8.36697161e-01
-9.04174149e-02 -1.65320730e+00 1.11308871e-02 -3.32613617e-01
2.63397485e-01 -2.62485594e-01 -4.79012758e-01 2.51914501e-01
2.35384971e-01 -7.54148722e-01 1.24564552e+00 3.84570718e-01
4.70002621e-01 -8.97614598e-01 5.88992655e-01 2.84126312e-01
-1.64152551e+00 -2.08842039e-01 1.06617294e-01 -6.55685216e-02
7.94829488e-01 7.30777144e-01 -9.36243773e-01 9.50555444e-01
5.96378207e-01 7.88003623e-01 -1.70438796e-01 1.19821930e+00
-3.20826024e-01 8.50471854e-01 -5.94392717e-01 -5.62573262e-02
6.34407401e-02 3.50502282e-01 6.34220541e-01 1.40874112e+00
4.58695412e-01 2.84218758e-01 1.12466849e-01 8.73264253e-01
-2.99597293e-01 -3.08262650e-02 -2.37304062e-01 -9.10097733e-02
7.11503506e-01 8.50621760e-01 -8.61047387e-01 -4.75971520e-01
-4.86409962e-01 7.20248699e-01 2.53169447e-01 1.45460024e-01
-1.14385676e+00 -5.97400546e-01 2.65517533e-01 -5.30834980e-02
5.46468258e-01 2.06030190e-01 -1.26192585e-01 -1.00719225e+00
4.81312990e-01 -7.74258494e-01 1.06702769e+00 -6.57334566e-01
-9.05864537e-01 6.02701187e-01 3.78923029e-01 -1.04013205e+00
-7.14806795e-01 -1.06396778e-02 -7.02512205e-01 7.25658059e-01
-9.87712979e-01 -8.36308658e-01 1.73268825e-01 5.49033761e-01
5.81672251e-01 1.98018536e-01 8.88164341e-01 7.58661151e-01
-7.10566580e-01 4.54447418e-01 -2.85117537e-01 4.74193752e-01
5.85927486e-01 -1.17305458e+00 5.94665468e-01 9.88211453e-01
3.57911289e-01 4.15565372e-01 5.57783782e-01 -7.78735220e-01
-1.24084365e+00 -1.15363955e+00 1.80331719e+00 -4.77096319e-01
3.47413749e-01 -4.65167284e-01 -6.59995914e-01 9.47979569e-01
3.15854669e-01 -1.76372021e-01 5.63055038e-01 2.22458094e-01
-2.69614756e-01 -1.18809000e-01 -7.03031123e-01 4.77476895e-01
1.17935371e+00 -2.49309748e-01 -1.07127094e+00 1.46912053e-01
9.27172780e-01 -7.61327863e-01 -5.92534423e-01 2.94038653e-01
3.99287730e-01 -6.21979296e-01 1.06403887e+00 -8.47347617e-01
7.70574391e-01 -5.20267189e-01 2.42099818e-02 -9.24567759e-01
-1.22472003e-01 -8.03924084e-01 -6.73269272e-01 1.57738817e+00
7.91494310e-01 -7.14004189e-02 4.57875907e-01 5.13253510e-01
-3.12081784e-01 -6.40641093e-01 -8.83838356e-01 -4.94751692e-01
-7.33825505e-01 -7.93055773e-01 1.02087998e+00 5.89421451e-01
2.09586740e-01 7.65258133e-01 -5.13676047e-01 3.93636703e-01
4.18993860e-01 4.29903835e-01 4.84011918e-02 -1.26878822e+00
-3.66237164e-01 -9.12219733e-02 5.04150204e-02 -7.84828842e-01
9.00157839e-02 -1.14433432e+00 2.93146409e-02 -1.86284244e+00
2.36758187e-01 -1.44999474e-01 -4.78796154e-01 7.41445839e-01
-1.50913194e-01 -3.54704410e-01 -9.70139913e-03 -1.42900959e-01
-8.74444723e-01 7.01412410e-02 5.95552862e-01 4.47612302e-03
-4.17168021e-01 -1.40098095e-01 -6.35179937e-01 6.53841376e-01
4.96440321e-01 -1.05090523e+00 -4.15113240e-01 -5.96958458e-01
7.11472273e-01 7.59079397e-01 3.94193679e-01 -9.00390446e-01
5.73064923e-01 4.00524493e-03 3.94176394e-01 -1.26549935e+00
1.93339616e-01 -9.18914795e-01 5.92077851e-01 2.33440667e-01
-7.26882875e-01 3.94363821e-01 1.43073827e-01 5.80294430e-01
-4.24129575e-01 -3.50617319e-01 -1.50008351e-02 4.48584594e-02
-6.78349495e-01 1.20301425e-01 -2.88973093e-01 3.58914733e-01
9.47552323e-01 2.76897907e-01 -2.11590499e-01 -9.81542915e-02
-7.79144943e-01 2.06520140e-01 -3.12717795e-01 3.28159153e-01
5.27592838e-01 -1.36433482e+00 -9.35493767e-01 6.09607212e-02
1.44967794e-01 4.18814570e-02 -2.36408301e-02 5.07485986e-01
-1.08008713e-01 3.47948313e-01 6.14704862e-02 -6.71475902e-02
-1.25099123e+00 4.89202529e-01 1.55484170e-01 -7.37546623e-01
-7.74834752e-01 6.91747725e-01 8.65977840e-04 9.74643603e-02
6.25551045e-01 -7.78451562e-01 -3.96439463e-01 2.53633171e-01
8.55753720e-01 2.06551790e-01 2.27926493e-01 -2.23439261e-01
-4.44316775e-01 -4.96476563e-03 -7.23570511e-02 -3.83356422e-01
1.39933705e+00 2.45565012e-01 -2.10705414e-01 2.18279019e-01
1.09295559e+00 -3.68807316e-02 -8.19768369e-01 -3.31272483e-01
4.37641621e-01 -2.79459953e-02 -1.70013115e-01 -9.31136370e-01
-8.43476295e-01 3.78398001e-01 9.78475586e-02 1.64265066e-01
1.22450495e+00 2.54549414e-01 1.11775601e+00 3.80222678e-01
8.45764577e-02 -8.22011650e-01 -3.47711742e-01 6.35944605e-01
6.51749790e-01 -6.41449153e-01 -2.66565382e-01 -6.26482427e-01
-4.79049921e-01 1.06294894e+00 3.10945451e-01 2.85150051e-01
5.16233742e-01 7.41512537e-01 -4.95405763e-01 -2.74052233e-01
-1.24621296e+00 -2.48688877e-01 2.75062412e-01 7.45203644e-02
6.28222048e-01 -7.21994936e-02 -8.57222259e-01 1.27780867e+00
-1.10610668e-02 3.10028762e-01 2.49790087e-01 1.15848660e+00
-1.18003696e-01 -1.48432553e+00 -1.85859486e-01 5.38043559e-01
-1.01818156e+00 -1.98482886e-01 -2.80320495e-01 5.05411804e-01
4.04970765e-01 9.71256375e-01 3.69163930e-01 -1.22623704e-01
8.46954048e-01 4.46812153e-01 2.03318045e-01 -6.07630789e-01
-1.03351891e+00 2.33231828e-01 8.38601291e-01 -5.87142527e-01
-2.65717715e-01 -8.10486794e-01 -1.94700289e+00 1.65444404e-01
8.84488449e-02 3.67513895e-01 4.15694684e-01 8.82357240e-01
5.93048036e-01 9.96553957e-01 3.66802484e-01 -1.48472890e-01
-2.25610316e-01 -4.93337333e-01 -1.57858357e-01 4.90601450e-01
2.06916332e-01 -3.61349016e-01 1.60483360e-01 3.08715850e-01] | [9.069357872009277, 9.198006629943848] |
ea70b1c4-97a3-498f-8522-c62e2bcbf0f3 | evaluating-explainable-methods-for-predictive | 2012.04218 | null | https://arxiv.org/abs/2012.04218v1 | https://arxiv.org/pdf/2012.04218v1.pdf | Evaluating Explainable Methods for Predictive Process Analytics: A Functionally-Grounded Approach | Predictive process analytics focuses on predicting the future states of running instances of a business process. While advanced machine learning techniques have been used to increase accuracy of predictions, the resulting predictive models lack transparency. Current explainable machine learning methods, such as LIME and SHAP, can be used to interpret black box models. However, it is unclear how fit for purpose these methods are in explaining process predictive models. In this paper, we draw on evaluation measures used in the field of explainable AI and propose functionally-grounded evaluation metrics for assessing explainable methods in predictive process analytics. We apply the proposed metrics to evaluate the performance of LIME and SHAP in interpreting process predictive models built on XGBoost, which has been shown to be relatively accurate in process predictions. We conduct the evaluation using three open source, real-world event logs and analyse the evaluation results to derive insights. The research contributes to understanding the trustworthiness of explainable methods for predictive process analytics as a fundamental and key step towards human user-oriented evaluation. | ['Renuka Sindhgatta', 'Catarina Moreira', 'Chun Ouyang', 'Mythreyi Velmurugan'] | 2020-12-08 | null | null | null | null | ['predictive-process-monitoring'] | ['time-series'] | [ 2.18034983e-01 8.55083823e-01 -3.19206089e-01 -3.45252961e-01
-1.23401247e-01 -3.88729066e-01 9.52721596e-01 5.96717894e-01
4.03918386e-01 3.40850651e-01 5.25063157e-01 -8.46241236e-01
-6.29039764e-01 -8.27181041e-01 -3.38050812e-01 -5.35907894e-02
1.58149004e-02 7.54555404e-01 -2.38352343e-01 2.49519840e-01
6.34044409e-01 1.93699971e-01 -1.37435579e+00 7.13115752e-01
8.41541290e-01 9.39200282e-01 -2.85339922e-01 6.17398441e-01
-3.88683021e-01 1.67658818e+00 -3.10687959e-01 -6.74393356e-01
2.27979138e-01 -5.84110439e-01 -9.26730156e-01 -3.13104719e-01
-2.12214679e-01 6.30787089e-02 1.11984946e-01 4.37593579e-01
-2.44835094e-01 -2.19752252e-01 7.23122418e-01 -1.74636447e+00
-7.31019735e-01 9.13154781e-01 -1.04459152e-01 1.61842152e-01
5.74713230e-01 5.82861423e-01 1.24672985e+00 -3.63219768e-01
4.32308733e-01 1.38782549e+00 7.87560642e-01 4.95652586e-01
-1.28528357e+00 -5.43659985e-01 1.18778639e-01 9.71106738e-02
-5.59252739e-01 -2.09685385e-01 5.12078881e-01 -6.93225086e-01
9.60663140e-01 6.20229483e-01 1.08713865e+00 9.88621116e-01
6.32875383e-01 9.03285086e-01 1.54638493e+00 -5.99141598e-01
6.84735417e-01 1.95509523e-01 5.35918415e-01 4.08564240e-01
6.65978253e-01 2.03014165e-01 -8.99939001e-01 -2.45862529e-01
4.89227146e-01 4.84305650e-01 1.02488928e-01 -5.33564240e-02
-1.20957696e+00 6.99145794e-01 2.20928982e-01 1.59443274e-01
-8.16885412e-01 4.31808382e-01 1.27148792e-01 3.35420638e-01
6.18301332e-01 1.15735865e+00 -5.40084958e-01 -7.85536766e-01
-8.24933469e-01 2.84664631e-01 1.50098491e+00 8.36511672e-01
3.98298025e-01 -1.47722527e-01 -5.88784933e-01 -1.42605916e-01
7.44972527e-01 1.48372158e-01 2.26598665e-01 -1.15804100e+00
2.72138625e-01 1.12438488e+00 2.77725250e-01 -8.75356257e-01
-2.15314209e-01 -1.66358218e-01 -2.06226796e-01 2.45411426e-01
1.87678352e-01 3.42693143e-02 -7.19753325e-01 8.74246657e-01
-3.30682486e-01 -2.81301290e-02 -4.28394526e-02 6.13098800e-01
4.93511379e-01 7.14901149e-01 8.07339251e-01 -1.99502572e-01
1.14867389e+00 -1.22590721e+00 -8.78771305e-01 -3.40887845e-01
4.17563677e-01 -2.62487948e-01 1.18077517e+00 5.77460766e-01
-1.14280915e+00 -4.96829987e-01 -8.12761843e-01 2.66766459e-01
3.34661789e-02 -4.20973450e-01 1.18751252e+00 7.67477214e-01
-7.76913643e-01 1.01579404e+00 -1.26712930e+00 -4.76954669e-01
6.37134552e-01 4.50267782e-03 -4.99065444e-02 1.07144862e-01
-7.33568251e-01 1.15411150e+00 4.69012827e-01 -1.87153816e-01
-8.67468357e-01 -9.21435833e-01 -2.78916240e-01 6.40133858e-01
3.39747488e-01 -8.75150502e-01 1.42975080e+00 -1.12376416e+00
-1.23102081e+00 2.77091265e-01 -2.34635711e-01 -7.20164418e-01
8.52192581e-01 -4.79278028e-01 -2.96827346e-01 -6.20348267e-02
-1.34377778e-01 1.23113446e-01 2.22802669e-01 -1.33316457e+00
-8.38651299e-01 -3.70834380e-01 -1.50753915e-01 -7.00292215e-02
-7.21545294e-02 7.36265332e-02 1.71082303e-01 2.70575304e-02
3.44116896e-01 -6.80071652e-01 -4.45235968e-01 -3.42736721e-01
-3.99884015e-01 -1.49674639e-01 4.98605162e-01 -7.29830563e-01
1.41459322e+00 -1.72739697e+00 -4.27421868e-01 5.27495444e-01
6.64333701e-01 -2.06301108e-01 5.31275272e-01 8.35736871e-01
1.75113559e-01 9.37880576e-01 2.32242316e-01 -2.28535339e-01
3.42445970e-01 9.54584405e-02 -6.67149067e-01 -1.50656104e-01
2.77357727e-01 1.04089606e+00 -8.40617061e-01 -1.93649173e-01
3.32818568e-01 1.27156615e-01 -1.11096613e-01 3.50422204e-01
-4.03106600e-01 4.96261060e-01 -6.46227598e-01 6.42701268e-01
-4.19161730e-02 -3.73399615e-01 2.12631151e-01 3.13134730e-01
-1.45215049e-01 4.56202716e-01 -7.31572628e-01 8.29376817e-01
-2.35314652e-01 7.51798570e-01 -6.69477403e-01 -4.34577525e-01
1.21169913e+00 5.19388974e-01 4.38112944e-01 -4.84113008e-01
-1.99159965e-01 -8.37788135e-02 1.83815062e-01 -4.27884191e-01
3.91928017e-01 -3.05094063e-01 3.38441432e-01 8.35212648e-01
-3.88334841e-01 -1.20398350e-01 7.81902522e-02 -4.53517027e-02
1.43639660e+00 4.99427050e-01 8.77084911e-01 -2.33351707e-01
1.66049317e-01 3.69406164e-01 5.92608511e-01 7.51786113e-01
-2.95055509e-01 3.31674844e-01 1.07910740e+00 -1.04943752e+00
-1.12588024e+00 -8.28940153e-01 2.13319257e-01 9.37199295e-01
-9.14495962e-04 -9.36458290e-01 -4.82064456e-01 -5.83289564e-01
1.65742680e-01 1.50153232e+00 -6.99720919e-01 2.71132234e-02
1.36173263e-01 -5.40439129e-01 3.43954355e-01 7.14077950e-01
2.12909579e-01 -1.09611177e+00 -8.89468074e-01 4.81122285e-01
3.11622590e-01 -7.55376756e-01 5.92774034e-01 2.76120245e-01
-1.30761254e+00 -1.28407705e+00 4.94837075e-01 2.51579285e-01
3.84157866e-01 -1.09724775e-01 1.42340052e+00 7.79339075e-02
1.71994567e-01 4.51350033e-01 -4.27582771e-01 -1.36084449e+00
-8.78793657e-01 -6.72230124e-02 -3.61710876e-01 -3.54125410e-01
1.05016434e+00 -4.72650439e-01 -4.87969041e-01 3.81863326e-01
-4.72888410e-01 3.95767301e-01 6.31200254e-01 2.97652215e-01
4.25359249e-01 1.09581642e-01 4.13860321e-01 -1.15803123e+00
1.18725038e+00 -5.91423094e-01 -1.24766856e-01 3.41587305e-01
-1.76133966e+00 2.12573171e-01 3.11852932e-01 3.16318125e-02
-1.18928957e+00 -3.57366085e-01 5.33755422e-01 -1.30586743e-01
-2.54423320e-01 7.68573403e-01 3.59307490e-02 4.92537767e-01
8.61553788e-01 -1.73419163e-01 1.07931435e-01 -5.04472479e-02
3.14357758e-01 3.35410327e-01 -8.93736854e-02 -4.92792308e-01
6.54854894e-01 5.49713790e-01 1.80904090e-01 6.81933016e-02
-6.98258936e-01 -3.21551204e-01 -3.03346008e-01 -4.41616744e-01
8.02911222e-01 -6.34061933e-01 -1.04511237e+00 -5.13831437e-01
-7.02055693e-01 -3.73851478e-01 -5.86107910e-01 4.25499052e-01
-8.41278851e-01 -2.95945346e-01 -4.72154588e-01 -1.25727630e+00
-7.17638075e-01 -7.08852112e-01 7.50150323e-01 2.19551817e-01
-1.17247283e+00 -1.27413356e+00 8.47909749e-02 7.37805247e-01
5.58706105e-01 3.61755759e-01 1.00185931e+00 -1.46103013e+00
-7.97679603e-01 -4.98049527e-01 1.16905652e-01 4.90766317e-02
-4.15029339e-02 5.23620248e-01 -9.27657783e-01 3.84245098e-01
3.25027816e-02 3.08750689e-01 3.81607205e-01 1.99512362e-01
1.05534947e+00 -4.61897224e-01 -5.64557850e-01 1.52324289e-01
1.31804383e+00 4.22583252e-01 8.42477977e-01 8.48466039e-01
5.38333118e-01 1.05194664e+00 7.97630787e-01 5.71872056e-01
3.68236631e-01 -8.79589096e-02 3.28734010e-01 3.55652571e-01
3.19770575e-01 -7.19720423e-01 2.70828485e-01 3.86930138e-01
-8.21308136e-01 -8.75551626e-03 -1.61278415e+00 1.03984237e-01
-2.26461291e+00 -1.42313671e+00 -6.23659074e-01 1.75748599e+00
2.13443831e-01 4.83254075e-01 -9.12009701e-02 2.53013492e-01
3.26308310e-01 -3.06074530e-01 -2.18819886e-01 -7.60389209e-01
3.32263082e-01 1.15810245e-01 2.90664673e-01 2.43351653e-01
-8.38500798e-01 7.04941034e-01 6.72402096e+00 2.05092523e-02
-5.71170926e-01 1.41850561e-01 8.47843707e-01 -1.95513636e-01
-6.37347102e-01 6.42274201e-01 -4.70286280e-01 4.08312291e-01
1.36685729e+00 -6.41963005e-01 2.50900745e-01 1.35687196e+00
8.86958003e-01 6.57824129e-02 -1.88908505e+00 5.76909781e-01
-3.39869708e-01 -1.46100867e+00 3.08309216e-02 3.08405668e-01
9.28892195e-01 -5.10516703e-01 -2.43191779e-01 3.62508148e-01
6.66548312e-01 -1.52026665e+00 8.80377829e-01 1.06565928e+00
-2.89733112e-01 -5.51138401e-01 8.28560174e-01 2.69217491e-01
-6.20371997e-01 -5.87889910e-01 -1.92497551e-01 -8.02239358e-01
-3.11385151e-02 5.83270133e-01 -1.52009618e+00 3.49650562e-01
6.10638142e-01 7.30852067e-01 -5.36645353e-01 9.19253707e-01
-5.60904503e-01 1.31057465e+00 1.96160182e-01 -1.02149375e-01
-2.28315573e-02 -3.43210787e-01 3.42230260e-01 9.58369017e-01
2.98655570e-01 -2.50381708e-01 -5.93718700e-02 1.35618651e+00
4.48223144e-01 1.49629444e-01 -6.88185275e-01 -2.41795823e-01
4.92397428e-01 1.14551437e+00 -6.41648531e-01 -3.07655007e-01
-2.99396068e-01 3.76945615e-01 -8.49224180e-02 2.59158492e-01
-4.88953173e-01 5.02721131e-01 5.46357334e-01 6.07612491e-01
-3.32013905e-01 5.38571775e-02 -1.29080534e+00 -5.53203583e-01
-7.39651844e-02 -9.05385196e-01 3.95201743e-01 -1.19146097e+00
-1.35940552e+00 3.73818308e-01 -4.06634472e-02 -9.80004549e-01
-4.60242063e-01 -4.00566280e-01 -1.07243574e+00 1.02147257e+00
-9.79289412e-01 -1.55672944e+00 -7.75142312e-01 -1.31005377e-01
3.14471036e-01 -2.43764713e-01 8.19867134e-01 -6.91698551e-01
-3.59718889e-01 -5.11993885e-01 -1.35311514e-01 -4.06529069e-01
5.06039441e-01 -1.66161716e+00 6.67609394e-01 6.65257275e-01
-6.51682261e-03 9.26880479e-01 9.95637953e-01 -1.07414091e+00
-1.45111859e+00 -9.62821007e-01 8.05806875e-01 -1.08447075e+00
8.19428265e-01 1.71450898e-01 -9.42397833e-01 1.23443782e+00
2.41254017e-01 -5.28981566e-01 1.15849853e+00 5.26598513e-01
-4.12586331e-02 -1.95683286e-01 -1.02920198e+00 5.19038379e-01
8.99067998e-01 -2.72223055e-01 -9.92052555e-01 6.21472858e-02
5.29390991e-01 2.21351624e-01 -1.27437437e+00 7.49995038e-02
5.07408082e-01 -1.14073670e+00 4.41489458e-01 -1.19028819e+00
1.07481694e+00 -2.03912690e-01 1.27767429e-01 -1.23126006e+00
-5.19248784e-01 -8.88616502e-01 -4.82617915e-01 1.21175897e+00
6.22073650e-01 -7.14803636e-01 9.46093082e-01 1.69287944e+00
3.94611917e-02 -7.53167987e-01 -3.92213076e-01 -3.96906823e-01
-2.37584248e-01 -7.05997288e-01 9.22107100e-01 9.34520483e-01
4.46534932e-01 2.05612794e-01 5.37639856e-03 -5.45270257e-02
6.73346400e-01 8.60656127e-02 1.05455363e+00 -1.87037468e+00
-8.41655284e-02 -5.79283714e-01 -6.57640815e-01 -1.92211017e-01
-1.20105840e-01 -6.79077089e-01 -4.66895074e-01 -2.16429806e+00
4.34513271e-01 -4.21806931e-01 -3.53668183e-01 5.76334178e-01
-5.20667374e-01 -5.80601096e-01 6.34721816e-01 7.95166075e-01
-3.90676230e-01 2.91028172e-01 1.16543376e+00 1.46585658e-01
-3.97103190e-01 1.35596231e-01 -1.27975047e+00 8.44006479e-01
1.01036727e+00 -4.84385371e-01 -5.95049024e-01 1.52167186e-01
7.13876009e-01 -1.08279854e-01 4.34848487e-01 -1.06076539e+00
2.26547390e-01 -5.28211951e-01 4.55281585e-01 -1.19225886e-02
-1.24427833e-01 -1.06529546e+00 7.94023871e-01 6.14816725e-01
-7.70687044e-01 2.01246783e-01 -7.77207240e-02 7.14805663e-01
-3.02781492e-01 -2.73245871e-01 -4.60344143e-02 -2.44316995e-01
-6.82836056e-01 1.10871874e-01 -3.16386759e-01 -3.77167314e-01
1.31892300e+00 -5.76957345e-01 -4.82231706e-01 -7.23961115e-01
-6.51475191e-01 1.87914491e-01 3.74065638e-01 2.65441626e-01
2.93890119e-01 -9.24740314e-01 -6.25435412e-01 -3.27398866e-01
2.06460059e-01 -1.77084088e-01 -3.50174189e-01 8.13810170e-01
-8.69234860e-01 4.19082195e-01 -3.77823472e-01 -3.14835578e-01
-8.51908863e-01 5.53248823e-01 1.61298707e-01 -4.70791399e-01
-5.70883930e-01 7.69724101e-02 -1.60453156e-01 -1.92860037e-01
-2.44587690e-01 -6.07010603e-01 -2.03288004e-01 -2.04412982e-01
4.08515781e-01 7.45633125e-01 -2.81731039e-01 2.82705873e-01
-6.79704323e-02 -2.27283239e-01 2.37590447e-01 -3.59824538e-01
1.48927188e+00 4.54125889e-02 -3.19017708e-01 8.89332473e-01
-1.02075739e-02 -9.45735648e-02 -1.21008480e+00 2.99148202e-01
8.40841234e-01 -5.79640448e-01 -1.70777645e-02 -1.17966449e+00
-3.31651986e-01 8.65125895e-01 1.95322946e-01 8.71513963e-01
6.27183437e-01 -1.96052454e-02 -8.84572789e-02 3.42217356e-01
2.18934283e-01 -1.06593883e+00 -1.77799746e-01 -1.80450790e-02
8.25996995e-01 -1.21711195e+00 2.65706152e-01 -5.58150172e-01
-1.16524875e+00 1.35950696e+00 7.26992607e-01 2.63975531e-01
4.31978494e-01 1.34272864e-02 -4.86142077e-02 -4.41196680e-01
-1.20051146e+00 3.10742110e-01 1.38829723e-01 4.91384238e-01
7.89469242e-01 5.98636925e-01 -1.32051602e-01 9.41614509e-01
-5.58903277e-01 3.93932432e-01 5.69815755e-01 9.63346958e-01
-3.94037873e-01 -8.09005916e-01 -5.06869018e-01 9.52505648e-01
-5.87122500e-01 -8.06278437e-02 -6.89573824e-01 9.93066072e-01
-3.28438401e-01 1.26813054e+00 2.08298624e-01 -4.14829701e-01
3.88603538e-01 5.35550058e-01 6.35326356e-02 -6.25784695e-01
-9.62058842e-01 -4.71394330e-01 4.41024601e-01 -7.07350850e-01
-2.71067530e-01 -8.20350945e-01 -1.10294378e+00 -6.82827950e-01
1.19180605e-01 1.88122585e-01 8.21150362e-01 1.10923624e+00
4.89323884e-01 8.71666312e-01 1.17949709e-01 -9.29982066e-02
-6.26861036e-01 -1.25190747e+00 -2.04001069e-01 7.94299781e-01
-4.52582240e-01 -2.80732870e-01 -1.65449008e-01 2.87483335e-01] | [8.605091094970703, 5.896371364593506] |
77e89609-048d-42ed-8981-c4d79acf2c3d | crop-mapping-from-image-time-series-deep | 2102.08820 | null | https://arxiv.org/abs/2102.08820v2 | https://arxiv.org/pdf/2102.08820v2.pdf | Crop mapping from image time series: deep learning with multi-scale label hierarchies | The aim of this paper is to map agricultural crops by classifying satellite image time series. Domain experts in agriculture work with crop type labels that are organised in a hierarchical tree structure, where coarse classes (like orchards) are subdivided into finer ones (like apples, pears, vines, etc.). We develop a crop classification method that exploits this expert knowledge and significantly improves the mapping of rare crop types. The three-level label hierarchy is encoded in a convolutional, recurrent neural network (convRNN), such that for each pixel the model predicts three labels at different level of granularity. This end-to-end trainable, hierarchical network architecture allows the model to learn joint feature representations of rare classes (e.g., apples, pears) at a coarser level (e.g., orchard), thereby boosting classification performance at the fine-grained level. Additionally, labelling at different granularity also makes it possible to adjust the output according to the classification scores; as coarser labels with high confidence are sometimes more useful for agricultural practice than fine-grained but very uncertain labels. We validate the proposed method on a new, large dataset that we make public. ZueriCrop covers an area of 50 km x 48 km in the Swiss cantons of Zurich and Thurgau with a total of 116'000 individual fields spanning 48 crop classes, and 28,000 (multi-temporal) image patches from Sentinel-2. We compare our proposed hierarchical convRNN model with several baselines, including methods designed for imbalanced class distributions. The hierarchical approach performs superior by at least 9.9 percentage points in F1-score. | ['Jan Dirk Wegner', 'Konrad Schindler', 'Constantin Streit', 'Frank Liebisch', 'Gregor Perich', "Stefano D'Aronco", 'Mehmet Ozgur Turkoglu'] | 2021-02-17 | null | null | null | null | ['crop-classification'] | ['miscellaneous'] | [ 2.34360173e-01 2.86117136e-01 -4.16907638e-01 -5.59811831e-01
-3.92287582e-01 -9.16191757e-01 2.18293309e-01 6.68888032e-01
-5.37259839e-02 6.14842355e-01 -1.31243333e-01 -5.30582726e-01
-1.86750576e-01 -1.43259263e+00 -8.67165327e-01 -9.21721399e-01
-4.86903012e-01 2.72300899e-01 7.12961331e-02 -3.06944042e-01
-1.92669302e-01 5.36544144e-01 -2.06131577e+00 6.68632507e-01
7.77768850e-01 1.23880529e+00 4.15804207e-01 7.06457973e-01
4.74108681e-02 7.08315074e-01 -5.09560168e-01 -3.21447551e-02
2.31713504e-01 1.05546666e-02 -6.68597162e-01 8.25486258e-02
3.14378053e-01 -1.84269413e-01 4.16967869e-01 1.13284945e+00
-9.06521734e-03 -2.95714289e-01 9.47354853e-01 -1.11803210e+00
-5.48993528e-01 9.28927898e-01 -9.01402354e-01 -2.54204392e-01
-3.56208354e-01 -1.95662215e-01 1.19390941e+00 -3.91919315e-01
4.41757202e-01 1.16187596e+00 1.01606512e+00 -1.49587065e-01
-1.38010418e+00 -7.56645083e-01 5.67859948e-01 -8.27944726e-02
-1.30492425e+00 1.01688832e-01 1.43345997e-01 -6.39939904e-01
7.09511876e-01 2.77867138e-01 6.95685446e-01 6.13026321e-01
5.21073639e-01 6.26772642e-01 1.27023351e+00 -2.55489618e-01
2.91323543e-01 5.86254001e-02 2.75281876e-01 3.36169243e-01
2.15841323e-01 1.71100736e-01 -7.07130656e-02 -5.35199456e-02
7.14839518e-01 3.59914839e-01 1.22955188e-01 -4.42587107e-01
-1.14758790e+00 1.11571741e+00 1.18816245e+00 3.86153281e-01
-8.87146354e-01 -1.22029446e-02 3.67675066e-01 4.56854790e-01
8.92317176e-01 3.69612157e-01 -1.04669571e+00 7.44650781e-01
-1.21360588e+00 1.54905438e-01 5.48483491e-01 9.34705913e-01
1.00900185e+00 -1.13005854e-01 -1.93403542e-01 7.91646779e-01
2.70333827e-01 6.85891390e-01 1.77430317e-01 -8.96272480e-01
1.77519366e-01 6.26775026e-01 2.94676691e-01 -1.04985011e+00
-7.85659790e-01 -6.63070560e-01 -1.31960821e+00 4.89987284e-01
1.00593165e-01 -2.98859775e-01 -1.46449208e+00 1.74147916e+00
1.01806678e-01 -1.92513138e-01 2.25959823e-01 5.98312676e-01
6.75444901e-01 1.01509190e+00 5.05986154e-01 -1.01581991e-01
1.74396110e+00 -6.25322402e-01 -4.22962695e-01 -2.86320206e-02
7.49650419e-01 -4.75116670e-01 6.22443676e-01 3.24193358e-01
-5.56266129e-01 -7.34336972e-01 -8.21598887e-01 5.61501205e-01
-1.08464897e+00 5.92129767e-01 7.54197001e-01 4.84007627e-01
-1.26088476e+00 6.62041128e-01 -8.72810662e-01 -6.15501761e-01
5.26918411e-01 2.27403566e-01 -3.16489220e-01 2.32215121e-01
-1.31972921e+00 7.87249744e-01 7.52276063e-01 4.72845137e-01
-9.01761234e-01 -6.63905680e-01 -8.92535865e-01 3.48648399e-01
-1.85212754e-02 -1.31331190e-01 1.00974643e+00 -1.19324291e+00
-1.10349107e+00 1.03699148e+00 1.15171455e-01 -7.89282739e-01
5.44142537e-02 -5.43269329e-02 -3.47807407e-01 -1.63620666e-01
4.93208319e-01 1.15177345e+00 6.82318032e-01 -1.43274856e+00
-1.33358741e+00 -4.23056066e-01 1.57585770e-01 7.25267828e-02
-1.56318769e-01 -1.39222518e-01 5.92212617e-01 -7.63944447e-01
4.24555779e-01 -1.03221333e+00 -4.45675164e-01 -3.14905763e-01
-1.29447117e-01 -9.22024474e-02 5.10167897e-01 -7.57782876e-01
9.84039307e-01 -2.07572269e+00 -7.83341378e-02 3.46299440e-01
8.44049826e-02 -1.86426733e-02 -2.25887462e-01 3.03408146e-01
-2.98323244e-01 2.05408305e-01 -5.81796050e-01 4.54622656e-01
3.36646824e-03 2.29137123e-01 -4.98091251e-01 2.45369792e-01
6.56498015e-01 7.67527521e-01 -8.55743289e-01 -1.07262611e-01
2.69855410e-01 3.65756541e-01 -3.36507335e-02 1.65860027e-01
-3.89618635e-01 1.64752886e-01 -3.71031493e-01 1.03081918e+00
1.06829488e+00 -3.44186366e-01 3.16576362e-01 -2.83741683e-01
-5.39185107e-01 -2.55434364e-01 -1.01364732e+00 1.05109501e+00
-4.03797090e-01 4.88452137e-01 -4.64263745e-02 -1.24113894e+00
1.19093597e+00 3.04841101e-01 3.25432897e-01 -2.86581367e-01
-1.61205068e-01 3.18573058e-01 -1.56271994e-01 -6.17954582e-02
6.05753779e-01 2.27510944e-01 -3.58417451e-01 -8.80238935e-02
-2.28484906e-02 -1.40412420e-01 1.41344860e-01 -3.78806353e-01
8.55546296e-01 4.57864106e-01 4.94211167e-01 -6.26361191e-01
1.83241833e-02 4.82152075e-01 5.53722501e-01 7.52334893e-01
-1.04431175e-01 5.12782991e-01 8.53319407e-01 -7.54244089e-01
-9.31157887e-01 -8.89034212e-01 -5.39661825e-01 1.65638411e+00
-1.45836100e-01 1.83421344e-01 -5.58043540e-01 -6.56362712e-01
3.81543070e-01 5.42288601e-01 -1.13222444e+00 1.57636032e-01
-2.66841888e-01 -1.34723771e+00 4.69430774e-01 5.83567023e-01
7.50250041e-01 -1.51730096e+00 -1.00886321e+00 3.38774264e-01
-5.72213791e-02 -7.84592748e-01 5.14822245e-01 1.19696426e+00
-7.89158165e-01 -7.29576409e-01 -9.66662407e-01 -9.60581899e-01
5.37604213e-01 1.33484542e-01 1.63155460e+00 -4.09122974e-01
-9.81390662e-03 -4.29342926e-01 -8.19965065e-01 -5.93801975e-01
-3.30240726e-01 6.52065396e-01 -4.14131433e-01 -1.01214103e-01
5.89781821e-01 -2.38951489e-01 -3.67809802e-01 2.73929745e-01
-9.14083481e-01 -9.80401412e-02 8.13975871e-01 9.47704017e-01
7.32536614e-01 4.00206655e-01 6.67857468e-01 -1.07467079e+00
-1.85813874e-01 -8.32385063e-01 -9.65130508e-01 4.51506346e-01
-3.82889777e-01 -8.48071799e-02 3.98220688e-01 -2.00789109e-01
-8.74545813e-01 5.65118849e-01 4.52582985e-02 5.76082356e-02
-7.34255016e-01 9.35060859e-01 2.25285664e-02 1.13571815e-01
7.75613427e-01 -2.88994789e-01 -3.57023388e-01 -2.92325646e-01
3.06054950e-01 8.31059158e-01 4.80468124e-01 -9.93938074e-02
4.97280389e-01 1.20734170e-01 -1.25654921e-01 -6.62259877e-01
-9.34070468e-01 -3.74699235e-01 -9.21830118e-01 -4.40709181e-02
9.48032081e-01 -1.42445242e+00 -3.63717467e-01 7.19894767e-01
-9.26607549e-01 -6.89471126e-01 -3.27029675e-01 4.31005150e-01
-3.36288869e-01 -3.95893008e-01 -5.63744128e-01 -7.17320502e-01
-2.43593574e-01 -8.32923412e-01 1.51679325e+00 3.26176673e-01
-1.70993477e-01 -7.63566673e-01 -2.31575951e-01 -3.16442162e-01
3.74715805e-01 8.27550173e-01 1.03206670e+00 -3.45433950e-01
-2.13990584e-01 -1.13792457e-01 -6.43444002e-01 2.90062547e-01
2.65480548e-01 1.88007444e-01 -1.14802384e+00 -3.96160990e-01
-3.87764692e-01 -3.93538982e-01 1.30190682e+00 1.02506566e+00
1.03655612e+00 2.65231520e-01 -4.28018689e-01 4.90470618e-01
1.59851789e+00 1.28849059e-01 5.42456150e-01 4.63223666e-01
5.57368457e-01 8.87913167e-01 1.23564637e+00 3.34484905e-01
2.98313439e-01 2.48333201e-01 9.83667433e-01 -6.53951883e-01
3.02641213e-01 2.36707434e-01 2.28173494e-01 2.09217027e-01
-1.64487451e-01 -4.01603520e-01 -9.51206326e-01 8.19017231e-01
-1.91333330e+00 -6.93015873e-01 -1.79487437e-01 2.08096957e+00
7.99092650e-01 -6.90217316e-02 -7.07973018e-02 2.46848479e-01
8.94034445e-01 3.03589463e-01 -5.45423150e-01 -4.09060389e-01
-4.22067255e-01 3.00593883e-01 1.33269620e+00 2.67991900e-01
-1.82011664e+00 1.27983713e+00 5.52468204e+00 6.52335227e-01
-1.05222261e+00 -2.77136773e-01 9.91657317e-01 5.05133390e-01
1.24867812e-01 -1.96525127e-01 -9.22670066e-01 1.63783342e-01
9.37616050e-01 6.25222683e-01 3.15017696e-03 7.79827178e-01
-1.01716816e-01 -3.25528085e-01 -7.35400140e-01 3.66453767e-01
-4.31262791e-01 -1.09107733e+00 -1.69820085e-01 -2.75404733e-02
9.17857468e-01 3.38481724e-01 1.06335744e-01 4.50329095e-01
9.74312723e-01 -1.13608778e+00 7.04519272e-01 2.78911144e-01
9.33169186e-01 -8.45000565e-01 1.30061197e+00 2.19218448e-01
-1.44390416e+00 -2.74572283e-01 -7.15035141e-01 7.07135126e-02
-4.21204120e-01 7.69686282e-01 -3.69809330e-01 7.73350716e-01
1.36244249e+00 9.75613236e-01 -5.70602775e-01 6.01027191e-01
-1.76220369e-02 6.21439338e-01 -4.30815965e-01 2.35039517e-01
6.45575643e-01 -6.15675077e-02 -2.02903092e-01 1.24527228e+00
6.32865608e-01 4.89306226e-02 2.73958355e-01 5.03354192e-01
1.04538381e-01 -8.58589709e-02 -6.84943199e-01 5.09769656e-02
4.59811360e-01 1.42802632e+00 -1.26063466e+00 -2.00523868e-01
-2.09648348e-02 6.69573903e-01 3.30752507e-02 4.08314437e-01
-5.54242849e-01 -7.04587221e-01 5.57267964e-01 -1.64981842e-01
6.32661879e-01 2.66594112e-01 -2.39151835e-01 -7.27693915e-01
-1.52202860e-01 -1.00480449e+00 4.49654937e-01 -8.00525546e-01
-1.21110189e+00 8.02927554e-01 -6.55810162e-02 -1.32321608e+00
-3.87159020e-01 -8.91818345e-01 6.92185983e-02 1.12808836e+00
-1.77942097e+00 -1.62270546e+00 -6.00753129e-01 2.68058274e-02
3.68920743e-01 -1.30187914e-01 1.39797139e+00 -1.70147866e-01
-1.12921298e-01 1.76316723e-01 3.28242928e-01 1.49138480e-01
5.13180792e-01 -1.44874072e+00 5.73334813e-01 4.65821028e-01
-9.47776437e-02 -1.10379130e-01 4.84868586e-01 -6.47798777e-01
-2.97550708e-01 -1.81154895e+00 1.05392730e+00 -3.41134369e-01
5.88185608e-01 -2.19733790e-01 -9.20507908e-01 6.99857116e-01
8.79124328e-02 9.39781964e-02 4.07305777e-01 2.74052680e-01
-5.06051302e-01 -3.85965377e-01 -1.32976615e+00 -7.94844553e-02
5.21808505e-01 -2.71939844e-01 -7.37402886e-02 5.94423532e-01
7.14877427e-01 -2.83613324e-01 -1.19381213e+00 8.16779017e-01
7.53006697e-01 -8.49010587e-01 7.13408709e-01 -4.35663432e-01
6.09647989e-01 -2.82496035e-01 -4.65073884e-01 -1.76591873e+00
-7.69534647e-01 2.80448627e-02 4.45438892e-01 9.97213662e-01
5.38136244e-01 -6.55797243e-01 7.07293153e-01 -4.00092721e-01
3.78711671e-02 -4.56593364e-01 -3.31216365e-01 -6.54814780e-01
4.50341791e-01 6.01839181e-03 1.09721446e+00 1.03077269e+00
-5.59188902e-01 -4.40873578e-02 -1.42782390e-01 8.13063502e-01
3.81620914e-01 5.20875752e-01 2.62934178e-01 -1.86317134e+00
2.01173857e-01 -3.86296034e-01 -5.57204366e-01 -4.42670256e-01
1.66921303e-01 -7.51694381e-01 2.56791025e-01 -1.41341031e+00
-8.93024579e-02 -8.46177757e-01 -5.37325203e-01 9.97565746e-01
-9.96399149e-02 4.35322613e-01 -9.69141796e-02 1.14774577e-01
-7.90587068e-02 -8.49416927e-02 6.25753582e-01 -2.84283787e-01
-3.06081831e-01 2.54591227e-01 -5.34310639e-01 6.18254244e-01
9.76748049e-01 -6.20034277e-01 5.03445454e-02 -5.58540285e-01
3.96175534e-01 -1.23837642e-01 4.98607218e-01 -9.41719711e-01
-4.30573195e-01 -7.28461295e-02 7.30667591e-01 -9.34170187e-01
-2.94494033e-01 -1.04002547e+00 3.14850628e-01 5.75711489e-01
-5.92540085e-01 -1.15357138e-01 3.54991883e-01 2.01638073e-01
-2.90120512e-01 -8.58766362e-02 8.27363670e-01 -1.31251067e-01
-7.20524132e-01 2.69383818e-01 -4.79905248e-01 -4.93196875e-01
9.24406171e-01 5.12255393e-02 -6.14428110e-02 -5.81043698e-02
-9.11777020e-01 4.55290228e-01 2.43839234e-01 6.01207256e-01
-1.27197787e-01 -1.18366718e+00 -1.22530460e+00 5.31577826e-01
4.60413486e-01 1.68476194e-01 -2.33367588e-02 2.90658534e-01
-7.31574714e-01 3.95039469e-01 -6.44049346e-01 -1.06848252e+00
-9.93552268e-01 2.69927442e-01 2.45764673e-01 -5.93326390e-01
-2.51970857e-01 8.42178524e-01 4.10113961e-01 -9.17370617e-01
1.26672179e-01 -8.93278837e-01 -9.32565868e-01 7.84245849e-01
1.83223099e-01 -6.48838207e-02 2.83772826e-01 -7.63645291e-01
-1.77672461e-01 6.94454312e-01 8.58862624e-02 4.29823697e-01
1.54784691e+00 9.50144455e-02 -3.51402462e-01 7.18904853e-01
8.16794336e-01 -6.98329151e-01 -1.27157474e+00 -2.94910651e-02
2.93008208e-01 3.07133924e-02 2.75748521e-01 -9.82151151e-01
-1.19475758e+00 7.75789320e-01 1.22876513e+00 1.02294683e+00
1.40595198e+00 -1.98110655e-01 1.92255631e-01 4.01622593e-01
4.48685676e-01 -9.63214815e-01 -8.02779675e-01 6.09202743e-01
6.45327806e-01 -1.52680838e+00 -2.35754639e-01 -4.10611182e-01
-5.40771067e-01 1.05046296e+00 1.71569392e-01 -2.36199051e-01
8.26987743e-01 4.07773137e-01 3.59698921e-01 -1.68564871e-01
-5.10378122e-01 -5.63884556e-01 -1.12639576e-01 9.17737007e-01
5.80983758e-01 7.95100749e-01 3.04546118e-01 4.13238674e-01
-1.67957976e-01 6.00621067e-02 1.76321208e-01 8.49109232e-01
-6.89145565e-01 -6.32817864e-01 -5.32131135e-01 8.98393869e-01
-3.89877379e-01 -1.96714327e-01 -3.89334410e-02 7.35867739e-01
5.62304080e-01 9.61912811e-01 4.63433206e-01 -1.81239069e-01
3.28637034e-01 -1.36329353e-01 -2.88361274e-02 -7.05360770e-01
-8.78482223e-01 1.17168032e-01 -1.00589551e-01 -3.81201357e-01
-7.64414191e-01 -7.81806648e-01 -9.10770297e-01 -1.21929415e-01
-4.26002145e-01 1.70806751e-01 6.54163957e-01 5.95073283e-01
2.05133751e-01 6.88235223e-01 9.66533601e-01 -1.09426653e+00
-4.14859146e-01 -1.17962039e+00 -1.12447941e+00 -1.23791866e-01
6.13383889e-01 -6.13707244e-01 -2.59591013e-01 2.72089869e-01] | [9.400534629821777, -1.5630769729614258] |
53b49537-1b2e-41d3-a658-e19fced09358 | learning-visible-connectivity-dynamics-for | 2105.10389 | null | https://arxiv.org/abs/2105.10389v4 | https://arxiv.org/pdf/2105.10389v4.pdf | Learning Visible Connectivity Dynamics for Cloth Smoothing | Robotic manipulation of cloth remains challenging for robotics due to the complex dynamics of the cloth, lack of a low-dimensional state representation, and self-occlusions. In contrast to previous model-based approaches that learn a pixel-based dynamics model or a compressed latent vector dynamics, we propose to learn a particle-based dynamics model from a partial point cloud observation. To overcome the challenges of partial observability, we infer which visible points are connected on the underlying cloth mesh. We then learn a dynamics model over this visible connectivity graph. Compared to previous learning-based approaches, our model poses strong inductive bias with its particle based representation for learning the underlying cloth physics; it is invariant to visual features; and the predictions can be more easily visualized. We show that our method greatly outperforms previous state-of-the-art model-based and model-free reinforcement learning methods in simulation. Furthermore, we demonstrate zero-shot sim-to-real transfer where we deploy the model trained in simulation on a Franka arm and show that the model can successfully smooth different types of cloth from crumpled configurations. Videos can be found on our project website. | ['David Held', 'Zixuan Huang', 'YuFei Wang', 'Xingyu Lin'] | 2021-05-21 | null | null | null | null | ['deformable-object-manipulation'] | ['robots'] | [ 6.63269758e-02 2.10662648e-01 -4.01955880e-02 2.03782186e-01
-3.45641583e-01 -5.60695529e-01 5.16623914e-01 -9.02064983e-03
8.59379768e-02 6.18467271e-01 -1.48607016e-01 4.89337407e-02
-1.35952652e-01 -8.28088284e-01 -1.26636839e+00 -7.73316026e-01
-3.88406128e-01 8.59590590e-01 4.00108159e-01 -4.35754150e-01
-7.87680000e-02 6.36793435e-01 -1.37233925e+00 -1.57628004e-02
6.49334371e-01 5.57291508e-01 5.39060950e-01 9.56523418e-01
3.67603540e-01 7.54952013e-01 5.38231991e-02 2.47661412e-01
2.70632684e-01 -2.37708703e-01 -4.95093405e-01 4.69441205e-01
5.83180070e-01 -4.32148010e-01 -8.23669791e-01 8.64435256e-01
4.68088090e-02 8.75186548e-03 6.95129275e-01 -1.21678507e+00
-4.96479064e-01 1.16977096e-01 -4.00949180e-01 -6.45759642e-01
5.57972014e-01 6.94429517e-01 6.88881874e-01 -7.04297245e-01
1.24585032e+00 1.52723622e+00 9.10877466e-01 6.45408213e-01
-1.66245747e+00 -1.97502151e-01 4.10485804e-01 -3.60643230e-02
-9.46040452e-01 -6.30449355e-02 9.45665240e-01 -5.68781495e-01
6.34063482e-01 6.17428776e-03 1.36410320e+00 1.39152265e+00
5.93926132e-01 7.10058153e-01 1.25924218e+00 -3.04953009e-01
3.53298753e-01 -2.95434922e-01 -2.00621903e-01 1.46241343e+00
4.53124754e-02 4.37904775e-01 -4.45545137e-01 -3.32745254e-01
1.39445662e+00 2.43070662e-01 -2.65774429e-01 -1.17796707e+00
-1.49900103e+00 6.70722127e-01 6.79793775e-01 -5.89490592e-01
-5.01059234e-01 8.37383628e-01 -5.66805862e-02 9.86791253e-02
1.37259677e-01 1.24563284e-01 -5.17424107e-01 2.42779218e-02
-5.34831285e-01 7.05632329e-01 1.36750638e+00 1.24824607e+00
7.03348994e-01 5.87960891e-02 2.41763502e-01 1.44586831e-01
6.01600528e-01 8.27402771e-01 -3.11509699e-01 -1.62881064e+00
-1.18955247e-01 3.55456591e-01 4.43637282e-01 -9.51808929e-01
-2.05995336e-01 1.84472412e-01 -7.85441339e-01 9.72069681e-01
3.98103088e-01 -2.40081444e-01 -1.11139333e+00 1.54814887e+00
5.07210732e-01 3.52964908e-01 -1.32029220e-01 1.06747234e+00
3.59282225e-01 6.29802287e-01 -2.05247805e-01 -4.51616384e-02
8.61857891e-01 -9.87632334e-01 -5.50482094e-01 -7.34102130e-02
9.34691876e-02 -4.71612662e-01 1.03132975e+00 3.91599447e-01
-1.10746980e+00 -4.86905038e-01 -9.31504548e-01 2.15081394e-01
-1.51096746e-01 -1.03501342e-01 6.99562728e-01 -1.83913425e-01
-8.80956352e-01 1.20802462e+00 -1.63925612e+00 -6.95550799e-01
3.11279625e-01 3.05990964e-01 -3.46720994e-01 -2.70350307e-01
-4.98445004e-01 9.41411376e-01 -1.67444259e-01 6.28252849e-02
-1.54078901e+00 -8.48113775e-01 -9.23970342e-01 -3.15250218e-01
4.66399193e-01 -1.04310811e+00 1.11777258e+00 -5.46697021e-01
-1.94600356e+00 4.73231465e-01 -8.73618126e-02 -2.95190930e-01
8.15805912e-01 -1.55255228e-01 4.29498374e-01 2.60990322e-01
-3.74770403e-01 8.13458681e-01 1.06641614e+00 -1.89352870e+00
-2.55759150e-01 -9.19456184e-02 3.17590743e-01 1.08959280e-01
1.49921134e-01 -7.19259560e-01 -2.26995632e-01 -4.72662389e-01
7.49241188e-02 -1.59387362e+00 -6.60234928e-01 9.07371938e-01
-2.21684024e-01 5.41629791e-02 1.32532024e+00 -5.94830096e-01
2.59875745e-01 -1.70109785e+00 9.60622609e-01 -4.67625745e-02
1.65899172e-01 -1.78341165e-01 -2.70627588e-02 7.01412976e-01
4.85407501e-01 -3.58988851e-01 -3.76623958e-01 -3.23829979e-01
2.21231356e-01 8.11527371e-01 -2.86597311e-01 7.40737736e-01
2.96081364e-01 8.51356804e-01 -1.22174835e+00 -4.64606583e-01
5.85661054e-01 6.98824942e-01 -4.42170680e-01 2.59013087e-01
-7.92255282e-01 8.96507680e-01 -5.29148340e-01 6.90249264e-01
5.55341959e-01 -3.96179676e-01 4.26740110e-01 -3.40389431e-01
-1.67583331e-01 -2.04474837e-01 -1.18170404e+00 2.13928103e+00
-4.25575167e-01 1.54007286e-01 6.08446419e-01 -7.35577881e-01
6.37828887e-01 2.32990921e-01 6.03652179e-01 1.05213039e-01
-1.61256954e-01 -1.01396352e-01 -2.43233547e-01 -4.63876754e-01
9.36528966e-02 1.11672968e-01 -1.10279685e-02 1.04608543e-01
9.48737040e-02 -7.34394968e-01 -7.23545700e-02 2.70436376e-01
1.42278433e+00 1.16724265e+00 -2.72038877e-01 -3.83425027e-01
-7.60185122e-02 5.94452977e-01 4.50649381e-01 5.43281019e-01
1.04808725e-01 5.99164188e-01 2.34068394e-01 -5.08964598e-01
-1.20430708e+00 -1.34928358e+00 8.02375004e-02 5.08473873e-01
5.29459715e-01 -4.77927417e-01 -4.80036139e-01 -5.66931009e-01
5.54616809e-01 3.72055382e-01 -6.94043696e-01 7.52592906e-02
-8.36831629e-01 -2.86578864e-01 -1.16336718e-01 4.76281911e-01
1.21832713e-01 -1.02048719e+00 -8.65930676e-01 4.05290425e-01
2.48067081e-01 -1.01787031e+00 -1.16454169e-01 3.14760618e-02
-9.03382182e-01 -1.23085058e+00 -5.64771652e-01 -7.44678855e-01
8.75783563e-01 1.74492136e-01 1.06736636e+00 1.03490598e-01
-7.76396751e-01 1.10604322e+00 -1.65019646e-01 -2.92952031e-01
-6.16547942e-01 -3.66756469e-01 1.05167091e-01 -5.05985320e-01
-4.27344143e-01 -8.23136747e-01 -6.82302058e-01 2.25677192e-01
-4.12810564e-01 3.69220525e-01 2.95151740e-01 8.38211119e-01
9.63734567e-01 -1.33730486e-01 -1.19757220e-01 -5.23345947e-01
1.02970041e-01 -2.48027623e-01 -8.67407024e-01 7.43559748e-02
-2.71170855e-01 2.19621155e-02 4.13693011e-01 -8.71595025e-01
-7.77940214e-01 9.19214010e-01 1.97994977e-01 -1.00550973e+00
-2.49815151e-01 1.69225886e-01 2.96740055e-01 -4.21036750e-01
3.45588177e-01 -1.54101014e-01 6.32959783e-01 -3.68466258e-01
5.36423624e-01 -2.13723883e-01 3.43263626e-01 -9.08945203e-01
1.01857281e+00 9.33414996e-01 5.58071077e-01 -8.76063466e-01
-4.89896029e-01 -1.91520061e-02 -9.88379300e-01 -5.33681512e-01
9.20997500e-01 -8.72706294e-01 -1.20631206e+00 4.36220080e-01
-1.17365515e+00 -1.12650585e+00 -5.67627609e-01 3.89844149e-01
-1.03813851e+00 4.13894534e-01 -1.02048373e+00 -7.83739150e-01
1.25262409e-01 -1.18629622e+00 1.31332755e+00 -1.73245549e-01
-3.53534639e-01 -1.23159564e+00 7.74710253e-02 -4.04511690e-01
2.20358133e-01 9.38007236e-01 7.55094647e-01 5.26811063e-01
-9.92455363e-01 -9.60570499e-02 1.88537195e-01 -5.75838201e-02
1.27244398e-01 2.82716364e-01 -5.01761377e-01 -6.54650986e-01
-1.87561303e-01 -4.88540113e-01 8.31227481e-01 4.77651149e-01
8.69859695e-01 -3.41738045e-01 -7.89029598e-01 5.30939460e-01
1.45270288e+00 -5.59390008e-01 2.41493091e-01 -1.50421888e-01
1.06513393e+00 5.82074583e-01 5.78826725e-01 3.86511862e-01
4.81723309e-01 6.76798761e-01 8.71088624e-01 -2.29740724e-01
-4.45232898e-01 -4.43535537e-01 5.64382851e-01 8.73141646e-01
-3.57157499e-01 1.54105514e-01 -8.14396441e-01 3.47013861e-01
-2.23752975e+00 -7.42332935e-01 -2.69978553e-01 1.77466810e+00
4.91839916e-01 3.39158401e-02 1.19023278e-01 -3.44741493e-01
3.64884406e-01 -4.75838967e-03 -9.22092319e-01 2.89946310e-02
2.02190489e-01 -1.11752972e-01 6.80396020e-01 7.99779177e-01
-9.25468326e-01 1.12536240e+00 6.65200472e+00 2.82955021e-01
-9.30486262e-01 1.07011765e-01 -1.93910524e-01 -2.54751388e-02
-1.43456519e-01 3.54714662e-01 -6.52473986e-01 1.95035666e-01
4.69923407e-01 2.42292851e-01 7.62063444e-01 8.83635819e-01
2.67167449e-01 1.07665900e-02 -1.27082694e+00 5.25710583e-01
-8.54346529e-02 -1.50849497e+00 5.33507857e-03 4.58642602e-01
7.51717985e-01 1.51265159e-01 -2.04412192e-01 8.88487399e-02
1.09630227e+00 -7.66264141e-01 9.70435858e-01 1.01483071e+00
5.94755054e-01 -2.54327834e-01 3.29126418e-02 6.46472275e-01
-1.14328718e+00 -5.79391513e-03 -4.86442387e-01 -2.80339688e-01
4.00734425e-01 2.83579379e-01 -3.84814769e-01 3.34533751e-01
6.91936016e-01 1.02859747e+00 2.03405209e-02 6.15791798e-01
-2.45315030e-01 5.76168060e-01 -5.69805562e-01 -3.40429433e-02
-7.63436407e-03 -3.25312525e-01 8.82445276e-01 8.71709526e-01
2.04513758e-01 1.49002112e-02 1.00727439e+00 1.09490573e+00
2.15389222e-01 -5.54094195e-01 -8.53706419e-01 1.72353536e-01
2.50292830e-02 1.33965170e+00 -7.44768798e-01 -2.49318510e-01
-2.62128741e-01 1.19899654e+00 3.87692302e-01 4.34298486e-01
-8.19145441e-01 1.83577880e-01 5.59342384e-01 4.04713690e-01
6.89954638e-01 -9.03446496e-01 2.51081258e-01 -1.28245342e+00
-1.19530760e-01 -6.05691373e-01 -2.58639961e-01 -8.91625226e-01
-1.47851646e+00 7.32342079e-02 1.44564569e-01 -1.15862346e+00
-1.51230767e-01 -8.49772394e-01 -5.04802287e-01 5.82810462e-01
-1.35867643e+00 -1.78460217e+00 -6.45596743e-01 4.33306277e-01
6.20557070e-01 2.24626780e-01 1.25506616e+00 -5.10477483e-01
1.14001648e-03 -2.83656865e-01 8.51037502e-02 -4.67378646e-03
5.15593350e-01 -1.41670537e+00 4.42969173e-01 4.61787850e-01
-1.01431049e-01 4.08493876e-01 9.68505800e-01 -1.08273411e+00
-2.55107999e+00 -1.03498185e+00 -1.05815634e-01 -7.67542124e-01
8.28010380e-01 -5.61911583e-01 -8.55429053e-01 8.80404651e-01
1.66333124e-01 5.47398508e-01 -8.79892632e-02 -2.91702840e-02
-9.96563211e-02 1.11894861e-01 -1.04159975e+00 6.50389433e-01
1.36507368e+00 -1.70643926e-01 -3.87145668e-01 7.02513933e-01
6.65385604e-01 -7.80684352e-01 -1.10080051e+00 4.26304549e-01
8.36791515e-01 -3.06259096e-01 1.06459260e+00 -5.47200263e-01
4.39024955e-01 -3.02791387e-01 -1.86139703e-01 -1.45800459e+00
-4.86774027e-01 -9.33785260e-01 -7.99622893e-01 7.15997875e-01
2.34136377e-02 -2.34175682e-01 9.13479865e-01 3.56434882e-01
-1.02864169e-01 -8.16313505e-01 -6.58748209e-01 -8.52449119e-01
1.15261935e-01 1.50862902e-01 1.90371841e-01 8.14840853e-01
-1.39076650e-01 8.65613818e-02 -4.33472157e-01 5.00104308e-01
1.19005096e+00 5.42630434e-01 1.06570733e+00 -1.32379103e+00
-6.65722728e-01 1.55513912e-01 -2.54600614e-01 -1.14822030e+00
3.63329798e-01 -7.38007247e-01 3.95125926e-01 -1.82553828e+00
8.10705870e-02 -4.94036138e-01 1.31154373e-01 5.07330060e-01
2.58873373e-01 1.51020223e-02 4.25943166e-01 2.74855047e-01
-5.07362723e-01 6.95555508e-01 1.70751822e+00 -2.39438057e-01
-1.09917738e-01 -2.20086381e-01 3.68257374e-01 1.09190309e+00
4.94595468e-01 -4.42083716e-01 -1.43472224e-01 -4.16067630e-01
-5.40712811e-02 4.43716675e-01 1.17098904e+00 -1.05315435e+00
2.81239510e-01 -4.91359293e-01 4.06949431e-01 -4.83313709e-01
8.59672129e-01 -1.01415157e+00 3.11798155e-01 9.26592529e-01
-2.39189014e-01 4.37180065e-02 2.09205523e-01 1.16149330e+00
4.65558857e-01 1.06155843e-01 6.49855256e-01 -6.50335789e-01
-5.28108776e-01 5.55582762e-01 -4.53630030e-01 -4.35704321e-01
9.24677610e-01 8.87375399e-02 -1.88649729e-01 -2.66801327e-01
-1.05300331e+00 1.37897760e-01 1.16410577e+00 2.72611469e-01
6.79587603e-01 -1.21323705e+00 -5.94881237e-01 2.13201944e-04
-1.28557608e-01 1.94662005e-01 1.09087609e-01 6.03559732e-01
-7.02425182e-01 -3.56002539e-01 -4.64219838e-01 -9.89733517e-01
-1.15996385e+00 5.86744249e-01 9.46883336e-02 -6.45923540e-02
-1.48574007e+00 3.46763849e-01 1.67879299e-03 -6.68746412e-01
2.21829340e-01 -2.84492433e-01 5.80618083e-01 -7.14778781e-01
-7.22881854e-02 4.24636841e-01 -4.66408372e-01 -5.24744213e-01
-5.28474897e-02 9.09141183e-01 3.96083295e-01 -1.71786547e-01
1.51769006e+00 5.99321239e-02 -6.00325502e-02 7.05676496e-01
7.24802613e-01 -1.47931829e-01 -2.23230767e+00 8.32051858e-02
-4.63407367e-01 -3.04274201e-01 -5.81447743e-02 -5.43795586e-01
-7.77446270e-01 9.50141668e-01 5.02465963e-01 4.80986722e-02
4.45194006e-01 1.93998613e-03 8.23269546e-01 6.64893985e-01
8.57984245e-01 -8.63219142e-01 2.19723761e-01 4.77363199e-01
1.17107022e+00 -1.14683115e+00 3.37103844e-01 -7.39259660e-01
-3.56649756e-01 1.20799637e+00 5.38034022e-01 -9.27565157e-01
9.71499622e-01 7.17958570e-01 -1.79374650e-01 -4.17188227e-01
-8.60719740e-01 -6.66485950e-02 -4.43802699e-02 8.07728946e-01
-2.58669198e-01 5.27718663e-02 3.69387001e-01 -2.82148957e-01
1.29531100e-01 1.03641607e-01 4.94728416e-01 1.43590593e+00
-4.02799338e-01 -1.15943837e+00 -1.93968862e-01 1.05100855e-01
2.02031806e-01 5.80089748e-01 -1.07634574e-01 9.71465647e-01
-1.51001140e-01 4.77414578e-01 -2.53819022e-02 -1.82021990e-01
4.09271926e-01 -2.03238919e-01 1.25245631e+00 -5.76912403e-01
-1.25590906e-01 8.40800554e-02 -1.27654850e-01 -9.35123205e-01
-4.82808143e-01 -8.09873104e-01 -1.34666538e+00 -3.55929673e-01
-8.20687711e-02 -2.38046154e-01 8.58130455e-01 6.77126110e-01
4.53284711e-01 5.76966584e-01 1.85817510e-01 -1.87678337e+00
-7.24627495e-01 -6.64507031e-01 -4.14610237e-01 6.60955131e-01
5.31328976e-01 -1.03332651e+00 -1.60404071e-01 5.23288608e-01] | [5.178103923797607, 0.05848729610443115] |
ccfd0582-a11a-4912-9166-2023bb353ad9 | biformer-learning-bilateral-motion-estimation | 2304.02225 | null | https://arxiv.org/abs/2304.02225v1 | https://arxiv.org/pdf/2304.02225v1.pdf | BiFormer: Learning Bilateral Motion Estimation via Bilateral Transformer for 4K Video Frame Interpolation | A novel 4K video frame interpolator based on bilateral transformer (BiFormer) is proposed in this paper, which performs three steps: global motion estimation, local motion refinement, and frame synthesis. First, in global motion estimation, we predict symmetric bilateral motion fields at a coarse scale. To this end, we propose BiFormer, the first transformer-based bilateral motion estimator. Second, we refine the global motion fields efficiently using blockwise bilateral cost volumes (BBCVs). Third, we warp the input frames using the refined motion fields and blend them to synthesize an intermediate frame. Extensive experiments demonstrate that the proposed BiFormer algorithm achieves excellent interpolation performance on 4K datasets. The source codes are available at https://github.com/JunHeum/BiFormer. | ['Chang-Su Kim', 'Jintae Kim', 'Junheum Park'] | 2023-04-05 | biformer-learning-bilateral-motion-estimation-1 | https://openaccess.thecvf.com/content/CVPR2023/html/Park_BiFormer_Learning_Bilateral_Motion_Estimation_via_Bilateral_Transformer_for_4K_CVPR_2023_paper.html | https://arxiv.org/pdf/2304.02225.pdf | cvpr-2023-6 | ['video-frame-interpolation', 'motion-estimation'] | ['computer-vision', 'computer-vision'] | [-2.27202117e-01 -4.53034282e-01 -1.71010479e-01 -8.19247141e-02
-6.36562467e-01 -2.43946567e-01 5.24889529e-01 -3.48113656e-01
-1.47935823e-01 8.55133891e-01 4.95725691e-01 -4.92125489e-02
3.41196358e-01 -5.52163661e-01 -7.55657852e-01 -8.07265401e-01
3.33544075e-01 -6.47018999e-02 5.72203636e-01 6.24740086e-02
3.29722077e-01 3.33548069e-01 -9.07383025e-01 4.10947174e-01
8.02775800e-01 1.05225217e+00 1.82951242e-01 9.43572700e-01
3.67428392e-01 9.81223106e-01 -5.31372894e-03 -1.71961561e-01
2.86685556e-01 -6.31089985e-01 -9.22487378e-01 -9.09863934e-02
4.90134329e-01 -8.49738479e-01 -4.86797988e-01 7.21930683e-01
2.63169825e-01 2.20379263e-01 2.98319370e-01 -1.12503731e+00
-4.16724086e-01 1.16030037e-01 -7.76026487e-01 9.52799544e-02
4.03116584e-01 2.88037956e-01 7.14820564e-01 -1.47321260e+00
8.10659349e-01 1.21728289e+00 7.38141537e-01 4.86681849e-01
-1.37349761e+00 -9.26048994e-01 -1.55003086e-01 3.36458713e-01
-1.61880028e+00 -5.79873919e-01 8.45243871e-01 -6.60473824e-01
5.05400836e-01 3.15849781e-01 9.02776897e-01 6.18423343e-01
4.14514631e-01 8.17200720e-01 9.05101240e-01 -9.85153466e-02
-6.17355481e-02 -6.01428807e-01 -3.26124489e-01 6.58485711e-01
-2.53681928e-01 7.94662610e-02 -8.19040954e-01 -1.60187975e-01
1.46080124e+00 -1.79705054e-01 -6.13503158e-01 -1.23064317e-01
-1.67023587e+00 6.01580322e-01 3.72275442e-01 1.63200498e-01
-4.58375216e-01 5.72660565e-01 1.59664229e-01 5.33218160e-02
7.13612378e-01 -3.20541739e-01 -1.19760744e-01 -5.42149730e-02
-1.30406249e+00 5.78869820e-01 4.14042383e-01 9.95449722e-01
1.07462716e+00 -6.85646534e-02 -3.73152673e-01 6.81806326e-01
5.72912335e-01 2.60141164e-01 2.46217906e-01 -1.62997162e+00
4.86034334e-01 2.41730232e-02 3.73613447e-01 -1.18771923e+00
5.32563776e-02 2.15184256e-01 -1.05548322e+00 1.77187875e-01
3.83007586e-01 -1.24788890e-02 -5.93253553e-01 1.26690400e+00
6.39968872e-01 8.60596240e-01 -3.51656944e-01 9.77622569e-01
5.85333407e-01 1.02141798e+00 6.01662099e-02 -1.39536962e-01
9.90193009e-01 -1.16866887e+00 -7.94887364e-01 2.24749386e-01
4.01194990e-01 -1.08916152e+00 5.84889650e-01 2.62532383e-01
-1.45774162e+00 -6.20479167e-01 -6.79200351e-01 -5.16024709e-01
3.90181392e-01 1.23242788e-01 1.59568965e-01 -1.90749660e-01
-1.28778958e+00 7.84734011e-01 -1.10849726e+00 1.11813195e-01
3.76895756e-01 2.31029749e-01 -2.23163307e-01 5.16310669e-02
-1.09152830e+00 3.98108453e-01 2.01438487e-01 3.48937809e-01
-8.57969344e-01 -9.02705252e-01 -9.39479470e-01 -2.47419789e-01
-1.27044633e-01 -1.05950248e+00 1.35117626e+00 -8.56248260e-01
-1.55629075e+00 5.24616897e-01 -8.08160961e-01 -3.34090918e-01
7.42112517e-01 -2.25218713e-01 3.89404483e-02 -3.32234125e-03
4.02408317e-02 8.18883359e-01 1.01353109e+00 -1.18254745e+00
-9.73724186e-01 1.42916828e-01 -4.28563237e-01 1.02497943e-01
1.94365650e-01 -1.19572707e-01 -7.91614234e-01 -1.19997239e+00
1.71160087e-01 -8.11807871e-01 -2.85494775e-01 3.74878347e-01
-2.35894918e-01 -6.22601956e-02 9.58637774e-01 -1.20746207e+00
1.48266900e+00 -2.31313252e+00 4.59092945e-01 1.69518471e-01
3.75389457e-01 2.02532876e-02 9.37690735e-02 7.59202242e-02
6.68391436e-02 -1.79121450e-01 -4.18276876e-01 -5.65223217e-01
-3.46569091e-01 7.49332309e-02 -1.79309502e-01 5.77860534e-01
4.44378983e-03 9.59687233e-01 -1.00043356e+00 -5.70630074e-01
5.12746572e-01 7.38141716e-01 -8.04916739e-01 1.45551488e-01
1.48005977e-01 1.13069439e+00 -3.35671335e-01 5.26353598e-01
9.25507069e-01 -1.11748807e-01 -1.62119880e-01 -5.19631743e-01
-5.01756251e-01 9.42780599e-02 -1.30065918e+00 1.88532341e+00
-2.23779067e-01 8.46614957e-01 2.05439061e-01 -5.68820775e-01
6.68645263e-01 4.44393933e-01 6.62032068e-01 -3.85997772e-01
1.40516032e-02 2.85190105e-01 -6.48490965e-01 -1.04219452e-01
6.68971360e-01 -7.64452145e-02 2.94848830e-01 2.49682143e-02
-1.37113959e-01 -1.33470923e-01 4.27231640e-02 9.59659368e-02
7.15280056e-01 5.03190935e-01 2.99275011e-01 -2.25376502e-01
8.89553010e-01 -3.97036523e-02 1.07374012e+00 1.18780531e-01
-1.75578520e-01 1.03275454e+00 2.56831974e-01 -5.25445104e-01
-1.33264458e+00 -1.04011881e+00 -2.55388543e-02 5.79565287e-01
4.06657100e-01 -5.62567949e-01 -6.54475987e-01 -1.11732900e-01
-2.50093713e-02 1.28498375e-01 -3.78836185e-01 4.19604033e-01
-1.08102107e+00 -1.13375276e-01 2.34796882e-01 6.35235429e-01
7.28123546e-01 -7.99725056e-01 -3.58302176e-01 4.65924680e-01
-6.14327788e-01 -1.01340294e+00 -1.16490090e+00 -7.22900033e-01
-1.13356221e+00 -6.97525561e-01 -1.18784952e+00 -9.46440697e-01
5.84571362e-01 3.01627338e-01 8.68852854e-01 3.76412958e-01
1.69782229e-02 -5.09067401e-02 -1.20819993e-01 4.58397061e-01
-3.32313985e-01 -4.23106074e-01 -2.10358724e-01 2.18191385e-01
-2.88445145e-01 -6.13706768e-01 -1.24353492e+00 5.42195201e-01
-6.94353402e-01 6.90085828e-01 1.37753546e-01 8.19580913e-01
1.02991903e+00 -3.26869935e-01 2.65568998e-02 -4.53147531e-01
3.14663887e-01 -3.25578183e-01 -8.62211108e-01 3.85336429e-02
-1.28753155e-01 -1.86442286e-01 3.41518283e-01 -2.04057977e-01
-1.16714978e+00 2.67953455e-01 -3.99470329e-01 -7.05105722e-01
1.52543351e-01 1.36587679e-01 1.48986354e-01 -2.55014479e-01
1.71694010e-01 3.62237781e-01 -2.25694522e-01 -4.05013114e-01
3.25114965e-01 3.71114075e-01 8.65368307e-01 -5.12747169e-01
8.32675636e-01 7.83787489e-01 -1.18838064e-01 -7.06050754e-01
-1.93007767e-01 -5.06724417e-01 -7.62575090e-01 -4.60126370e-01
1.20464206e+00 -1.12983799e+00 -7.26168334e-01 8.55830312e-01
-1.31859195e+00 -8.21343601e-01 -1.49214342e-01 5.75621128e-01
-6.47239387e-01 5.64511240e-01 -1.03420901e+00 -3.68569493e-01
-4.42317247e-01 -1.34116852e+00 1.09147048e+00 1.83007762e-01
-4.02931362e-01 -1.05716515e+00 2.02282295e-01 2.90904284e-01
2.72032022e-01 3.33846301e-01 2.43050367e-01 5.26103437e-01
-1.12557673e+00 3.22478056e-01 -2.46469349e-01 1.45926982e-01
2.94317126e-01 2.60721207e-01 -6.89076543e-01 -2.28633076e-01
-2.22949177e-01 2.30951160e-01 7.05442727e-01 7.69597352e-01
1.09782302e+00 -3.58843207e-01 -2.45425090e-01 1.16912913e+00
1.27033675e+00 1.84684396e-01 8.29521418e-01 2.61445999e-01
1.13262999e+00 2.38661259e-01 6.65284276e-01 7.04300404e-01
5.45266449e-01 8.39703083e-01 1.92834418e-02 -7.36151114e-02
-3.96269649e-01 -3.29235822e-01 4.08329815e-01 1.19198179e+00
-6.29446149e-01 1.41500473e-01 -7.60184467e-01 6.61234558e-01
-2.10742307e+00 -1.02786231e+00 -3.21678191e-01 2.09969878e+00
9.58089471e-01 -2.69273520e-01 -1.18212467e-02 -1.95703972e-02
7.08924055e-01 2.77847797e-01 -3.05490911e-01 -7.66066983e-02
3.26141939e-02 1.70607090e-01 6.36480689e-01 1.12755120e+00
-9.84589100e-01 1.01910257e+00 5.24430323e+00 9.60050941e-01
-1.08992922e+00 2.97301620e-01 7.51275659e-01 1.38705418e-01
-4.68631804e-01 2.85683215e-01 -6.78417087e-01 6.82504117e-01
5.13184488e-01 -2.95306295e-01 5.24341166e-01 4.15337294e-01
8.04053009e-01 -1.03801139e-01 -7.07345009e-01 9.39523041e-01
-3.71622860e-01 -1.83393514e+00 -2.38371268e-02 -9.07052755e-02
9.12349701e-01 -2.03341216e-01 -1.45774707e-01 -2.13888645e-01
2.37960026e-01 -7.63650835e-01 9.76718605e-01 6.55242503e-01
6.74744308e-01 -7.33428478e-01 2.97018528e-01 1.35813802e-01
-1.79957128e+00 4.04688925e-01 -2.22610593e-01 9.81799737e-02
7.14519143e-01 5.90752006e-01 -3.36476982e-01 5.73684573e-01
8.01748276e-01 1.10040045e+00 -4.89911698e-02 1.09425294e+00
-3.29971373e-01 5.06108701e-01 -9.83722135e-02 7.27367878e-01
9.15408656e-02 -4.69773620e-01 5.36788642e-01 1.20436764e+00
6.09388947e-01 3.85028124e-01 -3.19304317e-02 7.34113216e-01
-3.21951993e-02 9.32862889e-03 -1.49859473e-01 5.08887172e-01
4.72919941e-01 1.21088290e+00 -4.75610256e-01 -4.96952683e-01
-5.64326882e-01 1.33875918e+00 -1.92154106e-03 5.93441606e-01
-9.79803085e-01 -5.67218885e-02 8.58983994e-01 1.96499825e-01
2.99220473e-01 -5.02616763e-01 -2.60748863e-01 -1.43807733e+00
3.34880017e-02 -5.59600830e-01 2.41323456e-01 -8.74016106e-01
-8.43135417e-01 3.02317917e-01 -9.04321373e-02 -1.49123418e+00
-2.43353218e-01 -1.15753546e-01 -6.53391123e-01 1.13637769e+00
-1.43279123e+00 -1.18983006e+00 -7.29823351e-01 9.87470388e-01
7.55664885e-01 5.83873153e-01 1.14191920e-01 6.31599963e-01
-2.19218463e-01 3.11875910e-01 1.27359912e-01 2.28134617e-01
8.20841074e-01 -7.94923902e-01 7.57848442e-01 8.98114383e-01
-2.85106033e-01 3.94830287e-01 4.19961721e-01 -8.70016754e-01
-1.15917504e+00 -1.30384064e+00 9.86325324e-01 -1.91915184e-01
5.57597578e-01 -8.35771933e-02 -1.08530629e+00 9.20259893e-01
3.57266337e-01 5.12434721e-01 1.04663089e-01 -8.44207466e-01
1.67846367e-01 -6.17870465e-02 -9.61110592e-01 6.87510073e-01
8.66672754e-01 -4.21851784e-01 -4.42534778e-03 7.69989714e-02
6.38161898e-01 -8.80260110e-01 -1.21393502e+00 4.05431330e-01
8.21502030e-01 -8.87254477e-01 1.15114677e+00 1.66968763e-01
6.80231452e-01 -8.19783926e-01 -1.13933012e-01 -1.08355594e+00
-6.00887120e-01 -1.01233494e+00 -3.07583153e-01 1.18491888e+00
1.18778631e-01 -5.31091094e-01 7.40441978e-01 5.04862309e-01
-7.28382841e-02 -8.65251184e-01 -9.38588262e-01 -4.25042361e-01
3.44642103e-02 -2.19235092e-01 5.35345137e-01 1.11968684e+00
-2.48640925e-01 -2.12579370e-01 -8.65423381e-01 7.78580755e-02
7.86920726e-01 -4.49660085e-02 7.63342202e-01 -5.92682838e-01
-1.89599857e-01 -2.50156432e-01 -3.69936198e-01 -1.49133861e+00
-1.06405869e-01 -6.42456532e-01 1.40366897e-01 -1.53961039e+00
-1.95374917e-02 -2.76471645e-01 -4.18097787e-02 1.76598206e-01
-3.28681737e-01 5.90366781e-01 1.87636673e-01 5.14664710e-01
-4.26053479e-02 6.15320265e-01 1.71789420e+00 8.43195394e-02
-3.25692862e-01 -2.88409919e-01 -9.45999473e-03 7.92893112e-01
8.05331349e-01 -1.06061809e-01 -1.19861953e-01 -6.08274579e-01
-3.32415968e-01 5.58598638e-01 6.67174101e-01 -9.74740684e-01
3.41555506e-01 -4.88844573e-01 6.46688342e-01 -7.40219295e-01
3.90651733e-01 -5.81987023e-01 6.93412423e-01 5.79684377e-01
-1.23571746e-01 2.87867635e-01 -8.45526718e-03 2.92021781e-01
-4.15757716e-01 2.98410952e-01 1.05588150e+00 1.45289555e-01
-9.87291276e-01 7.42840052e-01 -2.68304914e-01 -2.28557736e-02
9.83042121e-01 -2.43746489e-01 -4.34548818e-02 -3.31509948e-01
-8.40619266e-01 1.73922554e-01 6.85910881e-01 9.27618742e-02
9.13583636e-01 -1.82252717e+00 -7.90797830e-01 2.41603181e-01
-2.84726590e-01 4.11568701e-01 3.52059364e-01 1.30244112e+00
-1.15529180e+00 1.21681966e-01 -5.79354502e-02 -6.92631483e-01
-1.35194445e+00 1.47867173e-01 2.08286747e-01 -5.22208139e-02
-8.92987967e-01 9.05335963e-01 5.21393716e-01 1.41607672e-01
-1.63214922e-01 -5.81859410e-01 1.67645037e-01 -2.10728228e-01
5.52905440e-01 5.70271909e-01 -2.42669865e-01 -9.74982440e-01
-3.41422468e-01 8.94657493e-01 3.13395083e-01 -3.04686338e-01
9.89876747e-01 -3.72774035e-01 -2.62402534e-01 2.09970344e-02
1.23390388e+00 8.40294659e-02 -1.81818843e+00 -4.48049866e-02
-1.93936184e-01 -9.96487498e-01 -5.47967106e-03 2.98610091e-01
-1.30963254e+00 6.87167287e-01 2.59659976e-01 -4.94765222e-01
1.34253931e+00 -2.78238505e-01 1.30227625e+00 -3.98088753e-01
3.01781535e-01 -6.97328627e-01 -2.27450892e-01 5.25230944e-01
8.31594765e-01 -7.82056212e-01 8.46410915e-02 -5.34703255e-01
-4.04262841e-01 1.01117039e+00 4.90005136e-01 -3.79729033e-01
7.11378157e-01 1.79923728e-01 -5.34464419e-02 2.94219196e-01
-7.03487277e-01 2.83353120e-01 4.10095811e-01 2.21084088e-01
6.90535784e-01 -9.16285515e-02 -5.21872044e-01 -3.01169623e-02
4.95588817e-02 5.01553595e-01 2.51327574e-01 8.20409298e-01
-3.02681118e-01 -1.07903290e+00 -5.32783389e-01 -1.52500179e-02
-2.58795857e-01 -1.87327832e-01 5.74846156e-02 5.29511631e-01
1.11258790e-01 5.85489869e-01 3.89628555e-03 -4.02996331e-01
1.52460426e-01 -3.22675794e-01 4.16274607e-01 -7.04497397e-02
-3.41117769e-01 5.14906883e-01 -1.47236302e-01 -1.02882290e+00
-5.15143931e-01 -6.10541344e-01 -1.34630191e+00 -7.06652105e-01
6.62936345e-02 5.14274947e-02 2.03752965e-01 6.36475623e-01
3.52481544e-01 3.82134318e-01 5.56062162e-01 -1.44011652e+00
3.04170787e-01 -4.96066660e-01 -2.71964163e-01 4.51385766e-01
6.66389525e-01 -3.90939802e-01 -2.72091806e-01 6.45857453e-01] | [10.75351333618164, -1.4666666984558105] |
d84315d1-7f00-41ea-b87b-82dee3ae6126 | a-feature-based-approach-for-video | 1605.08470 | null | http://arxiv.org/abs/1605.08470v1 | http://arxiv.org/pdf/1605.08470v1.pdf | A Feature based Approach for Video Compression | It is a high cost problem for panoramic image stitching via image matching
algorithm and not practical for real-time performance. In this paper, we take
full advantage ofHarris corner invariant characterization method light
intensity parallel meaning, translation and rotation, and made a realtime
panoramic image stitching algorithm. According to the basic characteristics and
performance FPGA classical algorithm, several modules such as the feature point
extraction, and matching description is to optimize the feature-based logic.
Real-time optimization system to achieve high precision match. The new
algorithm process the image from pixel domain and obtained from CCD camera
Xilinx Spartan-6 hardware platform. After the image stitching algorithm, will
eventually form a portable interface to output high-definition content on the
display. The results showed that, the proposed algorithm has higher precision
with good real-time performance and robustness. | ['Rajer Sindhu'] | 2016-05-26 | null | null | null | null | ['image-stitching'] | ['computer-vision'] | [ 5.95963836e-01 -8.99294734e-01 -3.57173324e-01 -5.05496468e-03
1.63885832e-01 -5.19944131e-01 4.01287943e-01 -5.30974269e-01
-3.19720715e-01 2.98940897e-01 -6.45612553e-02 -2.75828809e-01
-3.04491609e-01 -9.48677659e-01 -3.48198175e-01 -5.90036213e-01
3.92321825e-01 1.99501485e-01 2.95074642e-01 -1.34820238e-01
8.53839636e-01 5.60803533e-01 -1.63299823e+00 -2.47558892e-01
8.54580164e-01 9.67775941e-01 2.57450610e-01 8.38886201e-01
2.18365505e-01 4.07408029e-01 -3.61226320e-01 -1.15146181e-02
6.39384925e-01 -4.58754539e-01 -2.10481241e-01 2.19735011e-01
4.42642689e-01 -7.23103464e-01 -5.15847802e-01 1.38099957e+00
3.49296302e-01 -3.16550672e-01 2.91552693e-01 -1.50855577e+00
-1.49547875e-01 2.97879398e-01 -9.78807151e-01 3.99201512e-02
3.51675004e-01 2.56970823e-01 3.83112490e-01 -2.62740135e-01
6.94506288e-01 1.09389663e+00 5.91587007e-01 5.17023131e-02
-8.04885447e-01 -1.22267509e+00 -8.43148232e-01 3.18573803e-01
-1.12261677e+00 -1.14448592e-01 8.04110944e-01 -3.96400392e-02
5.64519405e-01 5.22456825e-01 9.76589084e-01 1.81488514e-01
1.17544174e+00 2.19303384e-01 1.47049379e+00 -5.47994614e-01
-3.24836075e-01 -2.78215408e-02 5.15910685e-01 6.13214731e-01
7.07065523e-01 4.70060289e-01 -3.21008824e-02 -7.66111538e-02
1.29686689e+00 5.94918430e-01 -5.17772734e-01 -2.50365794e-01
-1.36550915e+00 4.10261333e-01 3.49782914e-01 4.65784222e-01
3.69954556e-02 3.55391830e-01 1.86767444e-01 3.99748385e-01
-6.65213943e-01 3.22784901e-01 -2.99349576e-02 -2.22485319e-01
-7.34809637e-01 7.89346024e-02 4.64523226e-01 1.11829495e+00
1.02185619e+00 -3.65663394e-02 4.42060053e-01 2.92996258e-01
1.46812871e-01 1.11373019e+00 8.49640846e-01 -5.74190736e-01
2.62283206e-01 6.35900199e-01 -2.62368739e-01 -1.21763265e+00
-3.73066068e-01 1.92843840e-01 -8.35499942e-01 7.80815721e-01
6.28036931e-02 -1.62340373e-01 -8.88359666e-01 8.38878870e-01
1.49967402e-01 1.69191107e-01 2.32465163e-01 1.11680555e+00
6.72633111e-01 1.03949845e+00 -4.05832261e-01 -2.40358055e-01
1.85636735e+00 -7.15337574e-01 -1.02630997e+00 1.81434792e-03
1.54473782e-01 -1.60410881e+00 8.59784842e-01 6.11306369e-01
-6.09549463e-01 -8.32175970e-01 -1.78807569e+00 -7.97241461e-03
-7.20380247e-02 2.41432667e-01 8.45192492e-01 8.77163470e-01
-4.72868145e-01 4.72037554e-01 -4.39952344e-01 -4.26307410e-01
-2.72412598e-01 6.29840195e-01 -4.81091589e-01 6.87490031e-03
-9.77002442e-01 7.96693265e-01 6.62690639e-01 -1.62656307e-01
9.97024700e-02 -5.78000426e-01 -5.56271374e-01 -2.87815053e-02
2.34619491e-02 -7.83150256e-01 1.06289351e+00 -1.24915290e+00
-1.80446768e+00 7.62136221e-01 2.89844722e-01 -3.26972008e-01
2.23628983e-01 -5.59569485e-02 -8.65998328e-01 3.31069499e-01
-2.86499977e-01 3.79963338e-01 9.61289167e-01 -7.65173256e-01
-8.95919263e-01 -3.60864073e-01 -6.22712791e-01 5.08418322e-01
-1.83142070e-02 8.63598101e-03 -3.01961124e-01 -4.16300446e-01
3.81193101e-01 -8.95778358e-01 -1.21919453e-01 -4.50137816e-02
-3.88091914e-02 5.07480681e-01 1.52509058e+00 -3.98219585e-01
1.08286345e+00 -2.07542324e+00 -5.06507695e-01 5.26700258e-01
-2.46188417e-01 4.87483323e-01 4.12448764e-01 3.38756531e-01
-1.44522443e-01 -4.81073707e-01 6.82507530e-02 8.52843285e-01
-4.85869586e-01 -1.29853666e-01 -3.39545697e-01 6.97115123e-01
-4.40868348e-01 6.13472998e-01 -4.09010082e-01 -6.91636145e-01
8.07164073e-01 1.52354553e-01 -1.39405683e-01 -1.07173726e-01
2.57434130e-01 -1.05720535e-02 -4.24144715e-01 6.68379009e-01
1.40082455e+00 4.30761904e-01 -1.86215580e-01 -7.66766131e-01
-5.50206661e-01 -5.64149261e-01 -1.53527176e+00 1.45552135e+00
-3.67772937e-01 1.00724018e+00 1.10100798e-01 -2.36989960e-01
1.41084218e+00 1.28934786e-01 4.48446840e-01 -9.70917463e-01
8.54953527e-01 4.08274353e-01 -9.99421328e-02 -6.76219404e-01
8.86090815e-01 -1.09800018e-01 4.11562175e-02 6.73765898e-01
-3.40745002e-01 -9.45017636e-01 -1.76252976e-01 -8.29727873e-02
4.26466107e-01 -2.92903502e-02 5.19081831e-01 -5.14752448e-01
8.27661932e-01 4.74348187e-01 3.51253510e-01 9.07813385e-03
8.45624357e-02 5.68855524e-01 -6.72973767e-02 -3.63768846e-01
-1.41431701e+00 -6.64771020e-01 -3.43304425e-01 -1.22568257e-01
1.07682478e+00 -1.82311609e-01 -5.66878438e-01 8.28657448e-02
-1.83831051e-01 2.00767964e-01 2.80334111e-02 -7.04602674e-02
-9.33503568e-01 -4.76982474e-01 2.78849691e-01 -1.19052865e-01
1.25394225e+00 -7.17441797e-01 -1.32378578e+00 2.26259932e-01
5.16291678e-01 -9.04745638e-01 -5.64590394e-01 -3.66488010e-01
-1.14127338e+00 -1.32274449e+00 -4.79256868e-01 -1.20658100e+00
4.18964744e-01 7.49001682e-01 4.31888014e-01 2.91204840e-01
-8.95586610e-01 1.00338861e-01 -2.40024477e-02 -1.91128358e-01
-1.98642001e-01 -3.13633204e-01 -2.43291169e-01 -3.07615906e-01
3.31026882e-01 -5.78035116e-01 -8.69600713e-01 4.52078044e-01
-1.07742035e+00 3.79963636e-01 8.56750846e-01 5.46831250e-01
1.61042631e-01 4.20112878e-01 -1.41785741e-01 -4.47866887e-01
3.00488919e-01 5.46778381e-01 -1.40909517e+00 1.25619873e-01
-1.03461945e+00 4.95334668e-03 6.21655524e-01 -1.97864428e-01
-9.92634833e-01 2.17618242e-01 2.44886115e-01 -1.58825696e-01
1.84261993e-01 3.39959972e-02 -1.23904206e-01 -5.57541370e-01
5.91259420e-01 2.93213487e-01 5.49804747e-01 -3.71997319e-02
2.27770180e-01 1.06220937e+00 1.18934047e+00 -9.20478925e-02
1.23467815e+00 6.29731417e-01 3.25635672e-01 -8.76989722e-01
4.70731765e-01 -4.27169889e-01 -3.57688636e-01 -2.33250648e-01
8.46787691e-01 -8.54095042e-01 -9.73612905e-01 9.07889187e-01
-1.04092729e+00 3.02851588e-01 2.64515400e-01 9.08450007e-01
-5.39757431e-01 7.22735047e-01 -4.99684781e-01 -3.21859211e-01
-9.54911411e-01 -1.23810387e+00 4.99018729e-01 9.68262732e-01
1.93700552e-01 -6.11738801e-01 1.65118918e-01 -7.74445087e-02
1.78048879e-01 5.22451460e-01 5.16297758e-01 3.91459763e-01
-9.96092200e-01 -5.98987818e-01 -4.38473910e-01 -1.97191447e-01
2.25198925e-01 5.72156131e-01 -5.76849878e-01 -6.38167486e-02
4.68822837e-01 1.75338015e-01 6.52814865e-01 3.13177198e-01
7.08181858e-01 -1.39925256e-01 -6.53160274e-01 1.06221628e+00
2.29258704e+00 6.86591089e-01 1.24716723e+00 7.23671675e-01
4.42438364e-01 1.05522379e-01 1.22662091e+00 -6.72944728e-03
-2.38003552e-01 6.16008699e-01 1.15702383e-01 -2.76824772e-01
-1.67237848e-01 -2.77818382e-01 3.42959464e-01 4.80376512e-01
1.12382375e-01 7.10015595e-02 -5.05161285e-01 -9.56604332e-02
-1.66973770e+00 -1.03838420e+00 -6.20394409e-01 2.21712780e+00
5.17897248e-01 2.35007182e-02 -2.62181163e-01 4.69099909e-01
1.11332476e+00 -8.60403031e-02 -8.97250772e-02 -5.79685748e-01
-1.54358923e-01 3.15360129e-01 1.35356951e+00 6.22417212e-01
-8.60173881e-01 9.33952510e-01 5.50670433e+00 9.77455199e-01
-1.74617386e+00 -3.49088937e-01 1.23013519e-01 5.65620959e-01
-3.43944907e-01 4.95141447e-01 -7.41350055e-01 4.16136920e-01
2.56190538e-01 -4.34570730e-01 3.86462390e-01 5.75137973e-01
-1.04024373e-01 -6.04569614e-01 -5.06119967e-01 1.67440534e+00
1.53429940e-01 -1.26874197e+00 -1.01488389e-01 -1.47126660e-01
6.17066085e-01 -5.31400263e-01 1.77574173e-01 -5.44646502e-01
-2.74577469e-01 -6.43319607e-01 3.06201816e-01 1.28518462e-01
9.64300632e-01 -8.70364249e-01 6.11336827e-01 4.58044372e-02
-1.17562652e+00 -5.60386963e-02 -6.24359787e-01 -1.53619274e-01
2.78246641e-01 2.91552544e-01 -8.26313853e-01 7.16041207e-01
3.02789122e-01 6.45919740e-01 -3.43522668e-01 1.31571221e+00
6.88617211e-03 1.25605598e-01 -4.69966590e-01 -1.32369980e-01
2.49127641e-01 -6.92789853e-01 4.77831453e-01 7.85855472e-01
6.12681031e-01 3.98066819e-01 -3.26358438e-01 3.83876771e-01
5.97095549e-01 2.72201449e-01 -7.39001393e-01 3.29875916e-01
3.00678611e-01 1.47592044e+00 -1.05820692e+00 -4.56336617e-01
-2.23802075e-01 9.01075542e-01 -7.84969687e-01 -2.45111257e-01
-9.94034171e-01 -1.31281066e+00 1.97822765e-01 7.63175711e-02
-9.07104611e-02 -4.02995378e-01 -6.35897338e-01 -1.07849658e+00
-4.51174527e-01 -9.27432716e-01 3.44563395e-01 -1.09992826e+00
-3.92970353e-01 1.82600692e-01 -6.69300854e-02 -1.75060248e+00
9.57968980e-02 -8.33644986e-01 -8.25926781e-01 6.80277169e-01
-1.24466360e+00 -1.20404005e+00 -7.72984564e-01 9.04237211e-01
3.29288870e-01 -2.37291276e-01 5.08939147e-01 2.72692859e-01
-3.83037269e-01 3.10665816e-01 3.45941037e-01 -4.21771109e-02
1.08166611e+00 -4.74773884e-01 9.04091150e-02 1.19255662e+00
-1.56333148e-01 3.48285586e-01 7.86984205e-01 -6.67231262e-01
-1.78542817e+00 -2.92946726e-01 4.27596539e-01 2.25759253e-01
3.47260475e-01 5.04098944e-02 -2.57721186e-01 4.90777254e-01
7.48499513e-01 -4.28699195e-01 -9.53478664e-02 -8.58356774e-01
-3.10912356e-02 -5.90768874e-01 -1.33815050e+00 7.73536444e-01
5.25812328e-01 -1.65188015e-01 -7.21525550e-01 5.77653721e-02
4.64940012e-01 -6.45224214e-01 -4.91697431e-01 2.92025357e-01
8.87589514e-01 -1.06075335e+00 8.38122070e-01 4.70168978e-01
2.79778570e-01 -7.26732731e-01 1.76230520e-01 -5.83625257e-01
-2.87965149e-01 -1.04201245e+00 9.44962263e-01 9.29840267e-01
5.19535877e-02 -1.02188897e+00 7.66698420e-01 2.80427754e-01
-1.03692204e-01 -4.02316563e-02 -5.79744339e-01 -5.32820642e-01
-5.09037614e-01 2.11106434e-01 8.66400063e-01 7.79172719e-01
3.86810035e-01 4.27767992e-01 -5.01293719e-01 2.56555527e-01
6.41983747e-01 7.47485757e-01 1.18964756e+00 -9.07548010e-01
-1.21309191e-01 -3.82472485e-01 -9.75742996e-01 -9.36762631e-01
-3.49000841e-01 -4.23749119e-01 -4.23748910e-01 -1.00917375e+00
1.05580837e-01 -5.15161335e-01 3.16225529e-01 -1.05434440e-01
6.75329268e-02 5.54294825e-01 1.33241102e-01 4.20027047e-01
2.86603510e-01 6.70937449e-02 1.48200023e+00 -6.65956959e-02
-2.94356823e-01 7.96301365e-02 9.07980278e-02 3.94015789e-01
6.75814569e-01 -1.43064633e-01 -4.23234493e-01 -2.08484828e-01
-1.00143366e-01 4.21555072e-01 2.10035339e-01 -1.40451896e+00
5.34222364e-01 -3.15405697e-01 5.99224031e-01 -8.53923500e-01
2.83979684e-01 -1.58078444e+00 6.60040617e-01 1.54278290e+00
4.99252319e-01 3.07144076e-01 -1.41611660e-03 -8.28733221e-02
-2.51217842e-01 -4.75043893e-01 9.90233839e-01 2.13735271e-02
-1.03107285e+00 1.86632141e-01 -1.87629595e-01 -5.83536327e-01
1.46893585e+00 -9.83319581e-01 -6.49589837e-01 -2.28871763e-01
2.67149240e-01 -1.65945396e-01 1.03494585e+00 1.90746501e-01
7.25868821e-01 -1.34454060e+00 -2.71846950e-01 6.88343763e-01
1.91363264e-02 -3.44387859e-01 4.20360446e-01 6.06546998e-01
-1.83287370e+00 3.01875472e-01 -1.09531784e+00 -6.97917104e-01
-1.75423229e+00 7.11408615e-01 3.40312243e-01 2.94596910e-01
-7.27135539e-01 2.15982378e-01 -3.06846589e-01 2.20756322e-01
-4.58083123e-01 -3.43590051e-01 -2.24540755e-01 -3.43664944e-01
4.94584173e-01 5.64480305e-01 -1.50006428e-01 -6.84543848e-01
-3.20263386e-01 1.74312496e+00 2.55640268e-01 -6.06555045e-01
7.31885731e-01 -1.20992318e-01 -2.39851400e-01 -3.49935770e-01
1.32092011e+00 3.39802653e-01 -8.17179739e-01 1.83384195e-01
-4.71041560e-01 -1.15423322e+00 4.89918254e-02 -4.55714583e-01
-1.05816996e+00 6.08160794e-01 1.15578902e+00 -3.21630329e-01
1.26816165e+00 -9.42475736e-01 1.07800806e+00 3.12149465e-01
4.78056490e-01 -1.13453114e+00 -2.24431932e-01 7.83584919e-03
3.22790772e-01 -1.00133407e+00 7.83856511e-01 -5.67516804e-01
-5.46818316e-01 1.81513989e+00 7.51345873e-01 -6.61945701e-01
6.38219237e-01 6.10074520e-01 4.34018254e-01 -1.10630924e-02
-1.07538417e-01 2.32584864e-01 -9.81949121e-02 4.92811769e-01
-3.13426554e-02 -2.93073822e-02 -8.03120792e-01 -1.17549874e-01
-5.52447975e-01 1.01343907e-01 7.30579317e-01 8.49893451e-01
-8.40415716e-01 -1.10144413e+00 -9.75644767e-01 5.35065085e-02
-4.79883939e-01 2.03240082e-01 9.61395726e-02 1.00640833e+00
3.08139324e-02 7.30171800e-01 2.57538140e-01 -6.01519763e-01
3.21044445e-01 -5.53848565e-01 6.08114541e-01 3.68687332e-01
-8.15391123e-01 1.49820372e-01 -2.96170503e-01 -3.41829032e-01
-1.46585181e-01 -3.33603799e-01 -1.31407928e+00 -6.45303428e-01
-6.17881298e-01 1.57652751e-01 1.01736403e+00 3.33964437e-01
-4.78781536e-02 -9.78422090e-02 9.49313641e-01 -4.61549491e-01
-1.70521632e-01 -5.62129140e-01 -4.69376653e-01 2.05076799e-01
2.55637944e-01 -4.99582291e-02 -1.00169837e-01 -3.99373583e-02] | [9.054574012756348, -2.222778081893921] |
ed190b23-ff0c-43ae-a757-34e16a992042 | from-patches-to-pictures-paq-2-piq-mapping | 1912.10088 | null | https://arxiv.org/abs/1912.10088v1 | https://arxiv.org/pdf/1912.10088v1.pdf | From Patches to Pictures (PaQ-2-PiQ): Mapping the Perceptual Space of Picture Quality | Blind or no-reference (NR) perceptual picture quality prediction is a difficult, unsolved problem of great consequence to the social and streaming media industries that impacts billions of viewers daily. Unfortunately, popular NR prediction models perform poorly on real-world distorted pictures. To advance progress on this problem, we introduce the largest (by far) subjective picture quality database, containing about 40000 real-world distorted pictures and 120000 patches, on which we collected about 4M human judgments of picture quality. Using these picture and patch quality labels, we built deep region-based architectures that learn to produce state-of-the-art global picture quality predictions as well as useful local picture quality maps. Our innovations include picture quality prediction architectures that produce global-to-local inferences as well as local-to-global inferences (via feedback). | ['Praful Gupta', 'Dhruv Mahajan', 'Deepti Ghadiyaram', 'Zhenqiang Ying', 'Haoran Niu', 'Alan Bovik'] | 2019-12-20 | from-patches-to-pictures-paq-2-piq-mapping-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Ying_From_Patches_to_Pictures_PaQ-2-PiQ_Mapping_the_Perceptual_Space_of_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Ying_From_Patches_to_Pictures_PaQ-2-PiQ_Mapping_the_Perceptual_Space_of_CVPR_2020_paper.pdf | cvpr-2020-6 | ['blind-image-quality-assessment'] | ['computer-vision'] | [ 1.12948574e-01 -2.62886137e-01 2.03218028e-01 -7.63448536e-01
-1.23633349e+00 -2.84336627e-01 2.39165470e-01 1.05713025e-01
-1.41312629e-01 4.89350885e-01 7.69734800e-01 -1.03329860e-01
7.32419044e-02 -5.89251280e-01 -6.60770357e-01 -3.74800056e-01
-2.48882905e-01 -4.72039282e-02 4.59584147e-01 -9.61332545e-02
7.78774142e-01 2.31274724e-01 -1.65171647e+00 9.06581759e-01
7.58840501e-01 1.56715822e+00 2.40904957e-01 1.40099788e+00
5.72026193e-01 1.07642663e+00 -6.28940463e-01 -8.46235752e-01
4.39218819e-01 -3.90614331e-01 -5.68434656e-01 1.73357010e-01
1.00007987e+00 -9.84670222e-01 -5.46891630e-01 1.19927740e+00
7.78786957e-01 -1.77001618e-02 4.24361140e-01 -1.12278807e+00
-1.16063452e+00 1.70073479e-01 -5.37922561e-01 8.13589454e-01
6.33931458e-01 9.80354011e-01 1.33794606e+00 -9.33547258e-01
2.55833030e-01 1.52826917e+00 7.15860009e-01 1.43548310e-01
-1.15562606e+00 -4.95479822e-01 -1.58210829e-01 1.02080321e+00
-9.78194356e-01 -9.89106834e-01 3.16372782e-01 -2.82343388e-01
1.17945766e+00 2.46665105e-01 6.11976743e-01 8.85443807e-01
5.88626623e-01 5.12962401e-01 1.00863206e+00 -4.33531851e-02
2.96432942e-01 -3.12892050e-01 -5.31278431e-01 4.03795302e-01
-3.44584674e-01 2.58511543e-01 -6.92772388e-01 2.54858304e-02
1.05589652e+00 -5.19738257e-01 -4.06761974e-01 -9.32695270e-02
-1.19325936e+00 6.02109909e-01 6.21760130e-01 -2.61990726e-01
-4.02063608e-01 2.68527418e-01 1.92936078e-01 7.62488842e-01
7.26170838e-01 3.88245672e-01 -2.77480602e-01 -3.86970252e-01
-1.01011348e+00 3.72717053e-01 3.79619002e-01 6.29730523e-01
6.77987039e-01 1.08626343e-01 -2.52010763e-01 1.08537078e+00
3.00847858e-01 8.91473413e-01 5.68615906e-02 -1.85367215e+00
4.41766500e-01 -3.24169174e-02 6.58251524e-01 -1.56651902e+00
-2.32284814e-01 -1.25721365e-01 -9.27987695e-01 6.50097311e-01
4.57871467e-01 2.52069652e-01 -4.95309591e-01 1.43535233e+00
-1.59009367e-01 2.01072291e-01 -5.00814438e-01 1.25460100e+00
7.48755038e-01 9.04492021e-01 5.87327890e-02 -1.64569139e-01
1.19425905e+00 -8.72141600e-01 -4.17728186e-01 -7.31927855e-03
-1.69316143e-01 -8.84330392e-01 1.21507454e+00 9.97090340e-01
-1.73364532e+00 -1.09667790e+00 -9.14401948e-01 -2.83196747e-01
1.49838582e-01 -3.72005463e-01 -2.14526262e-02 5.90847135e-01
-2.03965306e+00 8.95258844e-01 -2.57773727e-01 -2.26927772e-01
6.63349211e-01 7.71417841e-02 -2.23561406e-01 -5.93002260e-01
-9.81407344e-01 1.00383019e+00 -4.06979561e-01 -1.26591533e-01
-1.17378688e+00 -7.46918559e-01 -6.37873709e-01 1.46918386e-01
7.70147964e-02 -9.23384726e-01 1.40859926e+00 -1.25596023e+00
-1.53370523e+00 9.30557787e-01 -3.39962512e-01 -4.32610810e-01
2.60952950e-01 -2.23297969e-01 -8.43601286e-01 6.37895763e-01
2.58986987e-02 1.13167918e+00 1.21173489e+00 -1.30494022e+00
-1.05732226e+00 1.05100516e-02 1.30375698e-01 4.01877671e-01
6.32536933e-02 3.45195264e-01 -5.35118580e-01 -6.66203141e-01
-2.00283229e-01 -1.19531050e-01 -2.41222885e-02 7.83680975e-01
8.39034840e-02 -1.15268327e-01 2.11356446e-01 -1.17154050e+00
9.41204309e-01 -1.99982405e+00 -7.15583414e-02 -1.81519121e-01
6.19212806e-01 -6.00566564e-04 -7.30976522e-01 8.55855569e-02
9.23891887e-02 -2.93041667e-04 1.85136482e-01 -4.23537314e-01
2.00435638e-01 -1.71253458e-01 -3.03243250e-01 5.42724490e-01
2.53510445e-01 7.84685910e-01 -1.10805058e+00 -3.83927912e-01
5.85678637e-01 2.79279143e-01 -8.99468362e-01 5.84298372e-01
-1.43871188e-01 2.83650041e-01 3.28307509e-01 7.44495451e-01
7.84220457e-01 -3.57317209e-01 -3.74716252e-01 -4.49055582e-01
-1.19070709e-02 2.91328490e-01 -7.69490480e-01 1.47747898e+00
-4.32930440e-01 1.08063769e+00 7.93085471e-02 -3.58745068e-01
6.87080562e-01 3.36026222e-01 4.17138845e-01 -1.32049668e+00
-2.21436575e-01 -6.59244657e-02 -1.18504480e-01 -5.75932741e-01
7.91112840e-01 -1.08322725e-01 3.02161485e-01 3.75938714e-01
1.06000483e-01 -1.15472905e-01 -1.26626208e-01 3.47104715e-03
1.23838866e+00 -3.22221637e-01 2.71358252e-01 -2.33558007e-02
2.81455100e-01 -4.13174540e-01 5.26962817e-01 7.09531784e-01
-6.71526432e-01 1.37201560e+00 4.63051349e-01 -5.66276193e-01
-1.73742640e+00 -1.47846377e+00 1.10039108e-01 1.16825247e+00
3.47716421e-01 -1.63541257e-01 -6.81230009e-01 -7.88935795e-02
-3.06980222e-01 5.72945178e-01 -3.74165475e-01 4.06851247e-02
-2.58060396e-01 -3.19907784e-01 2.29673833e-01 3.72119278e-01
5.48944652e-01 -1.49449027e+00 -3.25096726e-01 1.31463617e-01
-4.72430497e-01 -9.28605855e-01 -6.46040678e-01 -7.04790652e-01
-6.17415905e-01 -9.06854510e-01 -9.70100224e-01 -3.60103697e-01
3.78974751e-02 5.67459047e-01 1.93337727e+00 -4.80665416e-02
-1.05014183e-01 3.99364322e-01 -2.59192467e-01 1.72289059e-01
-5.37880540e-01 -7.72797823e-01 3.60751082e-03 6.38345107e-02
1.60910815e-01 -6.81214929e-01 -1.27442431e+00 6.78686857e-01
-6.52575314e-01 -1.86618175e-02 5.91949701e-01 4.92643684e-01
4.78786051e-01 3.55774343e-01 6.86699629e-01 -8.15130696e-02
6.84861243e-01 -4.89782721e-01 -5.13199985e-01 5.97304888e-02
-5.67258298e-01 -4.59349692e-01 4.73456800e-01 -4.59935740e-02
-1.14688575e+00 -7.82200873e-01 -4.54520464e-01 -4.51012373e-01
-3.59368831e-01 2.11061388e-02 -3.23652387e-01 -2.02744291e-03
7.85593569e-01 -2.30381582e-02 -2.94832915e-01 -4.78224307e-01
5.35425723e-01 6.08363271e-01 1.08143783e+00 -8.54509771e-02
3.54043096e-01 3.39336246e-01 -3.86489809e-01 -5.45511186e-01
-6.10434473e-01 -5.12301683e-01 -3.61288041e-01 -5.60738862e-01
1.04194677e+00 -1.25363100e+00 -9.68808174e-01 8.73229265e-01
-1.09153879e+00 -6.68917596e-01 -3.44083458e-01 -1.17475823e-01
-8.58917713e-01 6.80779934e-01 -9.49178338e-01 -6.29043400e-01
-2.97717631e-01 -1.06979334e+00 1.03873134e+00 4.29680161e-02
-1.81783035e-01 -7.56860018e-01 7.76656121e-02 3.45210940e-01
7.87285566e-01 -3.57917070e-01 8.57492447e-01 3.74920398e-01
-8.34052026e-01 2.03590125e-01 -9.76299703e-01 6.97391450e-01
-6.03466034e-02 -1.50804535e-01 -1.12975943e+00 -2.92548507e-01
2.03359529e-01 -5.31734943e-01 6.27201140e-01 8.10992420e-01
1.31117630e+00 -4.00135130e-01 4.20756072e-01 5.88609219e-01
1.52802682e+00 -4.27722521e-02 1.23946035e+00 5.94632663e-02
4.77284878e-01 3.52086157e-01 4.34737176e-01 7.30390191e-01
6.93321049e-01 7.77882397e-01 5.09226024e-01 -1.01205342e-01
-5.71582377e-01 -2.19671652e-01 4.39026684e-01 7.69655466e-01
1.17872597e-03 -7.07411885e-01 -6.43390894e-01 6.44061446e-01
-1.38099980e+00 -1.33394325e+00 -1.38719156e-02 2.14640260e+00
8.53343189e-01 2.62266546e-01 1.89726844e-01 1.13992736e-01
7.01179504e-01 3.66051584e-01 -7.86086500e-01 -4.73780364e-01
-2.83852607e-01 -2.83644050e-02 2.12873861e-01 5.52122772e-01
-9.58360016e-01 4.77414966e-01 7.70414114e+00 9.41907644e-01
-8.71832728e-01 1.22669421e-01 1.28109574e+00 -1.61596820e-01
-3.56213480e-01 -3.23655903e-01 -2.00757593e-01 6.45175099e-01
1.31823051e+00 2.06687659e-01 7.86452889e-01 5.71176827e-01
7.13726878e-01 -4.79043633e-01 -1.17880964e+00 1.53057122e+00
1.53406009e-01 -1.20603693e+00 5.98770157e-02 -2.85010394e-02
8.22744191e-01 2.83237308e-01 4.52943921e-01 -1.10818148e-01
4.52418685e-01 -1.29028928e+00 9.25457835e-01 1.08816636e+00
1.24851441e+00 -5.87406397e-01 7.43177950e-01 -6.81664646e-02
-9.51002955e-01 -2.24853471e-01 -1.00605679e+00 -2.22788021e-01
4.55503166e-01 7.77455389e-01 -2.24452659e-01 5.50285308e-03
1.20716023e+00 1.06420958e+00 -9.65702713e-01 1.51547182e+00
-1.21351676e-02 5.64854741e-01 1.06113903e-01 6.02704108e-01
-2.21551552e-01 4.90909070e-02 3.80451918e-01 1.06902730e+00
5.38417041e-01 3.43508035e-01 -4.60208833e-01 8.52963507e-01
-1.49347961e-01 -2.69448310e-01 -2.79833704e-01 5.92071950e-01
3.53124380e-01 9.66174304e-01 -2.51403123e-01 -5.34109354e-01
-5.62410176e-01 1.21979463e+00 1.49263173e-01 4.34187710e-01
-6.66371882e-01 7.49193951e-02 9.74599302e-01 3.52053046e-02
2.86500901e-01 5.34518249e-02 -5.56420743e-01 -1.10879719e+00
-1.19226389e-01 -1.34433806e+00 7.80800506e-02 -1.67808855e+00
-2.03578234e+00 5.42703450e-01 -3.43238652e-01 -1.50628805e+00
-1.10598147e-01 -5.29233634e-01 -6.38382435e-01 1.00782132e+00
-1.54544067e+00 -5.49695671e-01 -4.92301852e-01 6.83820546e-01
7.09317625e-01 -1.27573624e-01 6.30710721e-01 3.67139220e-01
4.51732911e-02 6.34882629e-01 1.21315941e-01 -1.85856074e-01
1.22707415e+00 -1.25171793e+00 8.27064037e-01 7.91746855e-01
1.78333484e-02 -1.45973191e-01 8.68332803e-01 -2.17713386e-01
-8.87823284e-01 -8.03364277e-01 8.68694007e-01 -6.67603612e-01
5.29642045e-01 3.38641293e-02 -9.18437481e-01 1.36456102e-01
6.00242496e-01 1.03925303e-01 3.24492782e-01 1.98557004e-02
-5.10254383e-01 -4.44890857e-01 -1.32114351e+00 5.96909523e-01
9.94301379e-01 -8.06942999e-01 -3.16227198e-01 2.27225982e-02
7.56924629e-01 -6.11065142e-02 -1.01801848e+00 4.48480062e-02
6.67707920e-01 -1.94677413e+00 1.23948014e+00 2.50305459e-02
9.98588860e-01 -1.29089564e-01 -4.67856824e-01 -1.47251642e+00
-7.26130247e-01 -4.28625137e-01 -1.21119983e-01 9.36290622e-01
2.31778026e-01 -2.32245773e-01 5.91943085e-01 7.08168328e-01
-3.53277534e-01 -4.34402317e-01 -8.81660104e-01 -3.79040629e-01
-6.17479440e-03 -6.68375731e-01 6.08778715e-01 3.85344923e-01
-1.24366567e-01 1.53194085e-01 -7.88424075e-01 1.16731629e-01
8.35594177e-01 -9.65491906e-02 6.10199332e-01 -8.49240899e-01
-5.85406840e-01 -7.83485115e-01 -7.10475802e-01 -1.59809458e+00
-5.41348398e-01 -3.55214328e-02 3.20970237e-01 -1.54542482e+00
3.29639018e-01 -3.82159412e-01 -4.26005006e-01 -1.09001651e-01
-3.14982027e-01 8.35527480e-01 2.23331720e-01 3.13890666e-01
-1.23551118e+00 5.87534130e-01 1.41670799e+00 -8.96217823e-02
2.95289844e-01 -1.01597495e-01 -5.76904535e-01 5.20012975e-01
5.51085889e-01 -9.56393108e-02 -3.34237546e-01 -8.90987873e-01
6.46400154e-01 4.86104548e-01 6.59551680e-01 -1.36540651e+00
9.98249948e-02 -1.31611526e-01 7.48197794e-01 -6.49896622e-01
5.73620796e-01 -4.34026301e-01 -2.23977119e-02 -1.49213493e-01
-5.06335378e-01 2.69534558e-01 -8.46758336e-02 6.57557666e-01
-4.16211575e-01 1.75920978e-01 1.10506868e+00 -2.21403912e-01
-1.09843349e+00 3.51408750e-01 -2.83266783e-01 1.48064613e-01
3.74399871e-01 -1.39664844e-01 -6.67881489e-01 -1.29678345e+00
-6.11794174e-01 1.27770096e-01 8.19040298e-01 3.02400708e-01
1.10422790e+00 -1.43406880e+00 -1.15566981e+00 -4.52009775e-02
1.15508273e-01 -3.48527789e-01 8.61876726e-01 5.38822412e-01
-7.03468442e-01 2.07231089e-01 -7.07363844e-01 -6.97796881e-01
-9.47037399e-01 8.30478609e-01 2.90778726e-01 5.15224561e-02
-6.15447581e-01 1.06225824e+00 3.84615153e-01 7.74524808e-02
2.09012374e-01 -3.95429105e-01 -1.68614373e-01 -3.47734928e-01
9.99298096e-01 6.42030001e-01 -1.62294909e-01 -7.58295119e-01
1.80979401e-01 5.16600192e-01 4.61748719e-01 -3.28395545e-01
1.15789640e+00 -9.66378987e-01 -4.20257449e-02 4.87754285e-01
1.24454403e+00 -2.83396602e-01 -1.97317541e+00 -3.28514576e-01
-3.48176420e-01 -1.08313560e+00 6.31314367e-02 -1.07604420e+00
-1.06877911e+00 1.14662111e+00 1.03948557e+00 2.95764416e-01
1.54328501e+00 -1.01646706e-01 1.09536576e+00 2.42359750e-02
3.87375683e-01 -8.96346748e-01 6.09579384e-01 1.45803764e-01
1.23579836e+00 -1.62849987e+00 5.72158583e-03 -2.20242199e-02
-6.80787683e-01 8.26989770e-01 4.03254420e-01 -2.85297513e-01
6.25061929e-01 -2.59748966e-01 3.01696271e-01 2.65559983e-02
-1.27709520e+00 9.30403993e-02 5.58838069e-01 9.79498029e-01
2.40265474e-01 -4.29030182e-03 5.73980451e-01 2.49515176e-01
-2.48177245e-01 -3.43464911e-02 4.36251700e-01 2.31935941e-02
-8.41701388e-01 -6.54187620e-01 -4.48753685e-01 5.47693908e-01
-3.37852985e-01 -5.20488441e-01 1.51049644e-01 -2.35578250e-02
1.94283962e-01 1.53425503e+00 4.29464042e-01 -5.04770935e-01
2.28264242e-01 -4.64442790e-01 3.61147344e-01 -1.95476800e-01
-2.98857957e-01 5.88079244e-02 4.44459282e-02 -1.17064321e+00
-3.81730616e-01 -4.81087863e-01 -4.45922762e-01 -9.76141691e-01
3.31088960e-01 -6.62931085e-01 2.43440717e-01 6.62520707e-01
2.28994295e-01 3.36124092e-01 8.80160630e-01 -1.34398329e+00
-3.27446848e-01 -1.20732975e+00 -5.89621484e-01 6.85102046e-01
5.69096267e-01 -1.56022355e-01 -3.40733290e-01 4.30618972e-01] | [11.844594955444336, -1.7765096426010132] |
b66bcdab-15d2-4dff-b878-efe72fab06af | snap-efficient-extraction-of-private | 2208.12348 | null | https://arxiv.org/abs/2208.12348v2 | https://arxiv.org/pdf/2208.12348v2.pdf | SNAP: Efficient Extraction of Private Properties with Poisoning | Property inference attacks allow an adversary to extract global properties of the training dataset from a machine learning model. Such attacks have privacy implications for data owners sharing their datasets to train machine learning models. Several existing approaches for property inference attacks against deep neural networks have been proposed, but they all rely on the attacker training a large number of shadow models, which induces a large computational overhead. In this paper, we consider the setting of property inference attacks in which the attacker can poison a subset of the training dataset and query the trained target model. Motivated by our theoretical analysis of model confidences under poisoning, we design an efficient property inference attack, SNAP, which obtains higher attack success and requires lower amounts of poisoning than the state-of-the-art poisoning-based property inference attack by Mahloujifar et al. For example, on the Census dataset, SNAP achieves 34% higher success rate than Mahloujifar et al. while being 56.5x faster. We also extend our attack to infer whether a certain property was present at all during training and estimate the exact proportion of a property of interest efficiently. We evaluate our attack on several properties of varying proportions from four datasets and demonstrate SNAP's generality and effectiveness. An open-source implementation of SNAP can be found at https://github.com/johnmath/snap-sp23. | ['Jonathan Ullman', 'Florian Tramèr', 'Matthew Jagielski', 'Alina Oprea', 'John Abascal', 'Harsh Chaudhari'] | 2022-08-25 | null | null | null | null | ['inference-attack'] | ['adversarial'] | [ 9.95284542e-02 5.86875565e-02 -5.31161785e-01 -1.50423154e-01
-8.36236298e-01 -1.13135946e+00 2.07182422e-01 3.74439567e-01
-6.76890492e-01 9.16577220e-01 -2.86748201e-01 -8.10107112e-01
-2.35293925e-01 -1.28015316e+00 -1.28945029e+00 -6.44520700e-01
-3.59256357e-01 3.26589555e-01 3.24188173e-01 3.53241414e-01
1.29725621e-03 7.83062637e-01 -9.28372800e-01 8.59627351e-02
2.28559762e-01 8.91976774e-01 -7.35572100e-01 6.28469408e-01
7.10486829e-01 7.92408466e-01 -6.67644322e-01 -7.13832021e-01
3.85947764e-01 1.15628183e-01 -1.04019451e+00 -7.05823839e-01
4.99073148e-01 -8.85763764e-01 -7.01071620e-01 1.40283573e+00
3.12989235e-01 -4.77833837e-01 3.99966478e-01 -2.08028722e+00
-2.63448626e-01 1.21444273e+00 -5.68470418e-01 2.23533213e-01
2.17278272e-01 1.50034860e-01 1.12173462e+00 6.66881651e-02
9.60524231e-02 9.15088594e-01 6.25100732e-01 5.76428354e-01
-1.43273437e+00 -1.61931145e+00 -3.19563746e-01 2.42390782e-01
-1.74093091e+00 -4.16455090e-01 5.48636615e-01 -3.61481830e-02
7.30078697e-01 6.70956969e-01 1.39160588e-01 1.29696155e+00
-6.92342892e-02 9.15329337e-01 9.03283536e-01 8.72282907e-02
2.96259493e-01 1.48810104e-01 2.74007827e-01 5.12026012e-01
8.91541839e-01 2.25610644e-01 -2.97809213e-01 -1.30505455e+00
5.51029503e-01 -5.50547102e-03 -3.95037174e-01 -3.37139040e-01
-7.85837829e-01 8.27257156e-01 1.88931614e-01 -2.68192500e-01
-1.24760434e-01 7.93522775e-01 4.97014910e-01 2.94196367e-01
2.81909287e-01 3.99668694e-01 -9.84111488e-01 4.89845127e-02
-6.20355487e-01 5.35119772e-01 1.16724944e+00 8.53839934e-01
6.10484362e-01 -4.44785506e-01 4.74006608e-02 -1.66432947e-01
2.22548962e-01 8.18388104e-01 -3.59155983e-01 -7.24756002e-01
4.40661460e-01 2.59321779e-01 3.01367313e-01 -8.74831378e-01
-2.46841535e-01 -1.46584868e-01 -6.57888889e-01 -5.81292622e-02
6.65384293e-01 -5.36300421e-01 -3.45928252e-01 2.05871725e+00
4.57879305e-01 6.93135083e-01 1.20572902e-01 2.31637359e-01
1.84638426e-01 4.03312832e-01 3.40187520e-01 2.60908663e-01
1.39064467e+00 -2.59826988e-01 -2.71396190e-01 3.43958408e-01
7.79620767e-01 1.35243833e-02 6.87649906e-01 4.28351045e-01
-8.68520558e-01 2.81342506e-01 -1.07593763e+00 3.74710649e-01
-7.17666388e-01 -4.53418851e-01 9.25028682e-01 1.21620846e+00
-5.09059548e-01 5.08082151e-01 -1.09104431e+00 1.25364810e-01
1.25046647e+00 6.28884912e-01 -3.78073275e-01 1.85433045e-01
-1.56994414e+00 3.98676127e-01 6.81337893e-01 -5.17254710e-01
-1.43340588e+00 -1.23933208e+00 -7.95541048e-01 2.68727571e-01
6.42447114e-01 -4.66497779e-01 1.23695457e+00 -3.79166424e-01
-7.16269016e-01 4.98207539e-01 1.98729560e-01 -9.68808353e-01
6.62648797e-01 -3.06789726e-01 -3.82663548e-01 2.91980535e-01
-1.65135548e-01 1.05269596e-01 5.66971242e-01 -1.05344164e+00
-6.86302960e-01 -2.11656153e-01 5.55030346e-01 -5.95999956e-01
-6.01219714e-01 3.03351194e-01 1.62962675e-01 -3.02875280e-01
-8.23612928e-01 -6.57219172e-01 -1.88074782e-01 2.63899773e-01
-7.89737344e-01 -6.08993955e-02 1.05271769e+00 -2.96343118e-01
1.15935922e+00 -2.00921631e+00 -7.32796431e-01 6.00417078e-01
4.21372265e-01 5.78647792e-01 -2.78212950e-02 5.42312443e-01
1.04617231e-01 6.37945294e-01 -3.80633384e-01 -3.22280824e-02
1.70962617e-01 4.16702330e-02 -7.98142254e-01 9.53577697e-01
1.11612409e-01 9.58550692e-01 -8.25695872e-01 -3.97199661e-01
-1.35855943e-01 5.30272305e-01 -5.56017756e-01 6.57807216e-02
-5.80863357e-01 9.67820212e-02 -7.76386201e-01 5.96435010e-01
1.03795922e+00 -5.31473696e-01 1.95030510e-01 3.31932828e-02
4.61113364e-01 4.03982967e-01 -9.99332786e-01 9.40048993e-01
-1.62436441e-01 2.21561685e-01 -6.81282431e-02 -5.89738250e-01
4.47332352e-01 4.77986664e-01 3.81775618e-01 -2.25427747e-01
2.89118767e-01 -8.06903839e-02 -2.10342497e-01 -1.74352601e-01
5.91863645e-03 7.84457400e-02 -3.10116500e-01 9.51687157e-01
-4.55886662e-01 5.14930367e-01 -3.43417704e-01 2.29926944e-01
1.57743859e+00 -1.55554041e-01 2.28377685e-01 -3.32088739e-01
4.25298274e-01 -1.45537972e-01 4.22589451e-01 1.26134288e+00
-3.51274729e-01 -2.80100256e-02 8.63603711e-01 -5.01522779e-01
-1.09776402e+00 -1.29001689e+00 -2.92378575e-01 7.33069599e-01
1.81409761e-01 -4.91220951e-01 -8.09607744e-01 -1.23893428e+00
4.37308699e-01 7.01299250e-01 -7.50254035e-01 -1.87345013e-01
-2.49135107e-01 -6.31445050e-01 1.60231459e+00 4.26167816e-01
9.45882142e-01 -8.51451278e-01 -7.03378856e-01 -1.85045540e-01
-1.00086719e-01 -1.14678037e+00 -1.94430962e-01 1.76752470e-02
-3.95391643e-01 -1.42078412e+00 8.71473625e-02 -2.83064306e-01
6.76820755e-01 -5.86817376e-02 8.53127658e-01 4.88184392e-01
-8.25196281e-02 1.23714708e-01 -6.96143648e-03 -9.29398000e-01
-4.93359774e-01 4.00787592e-01 1.71136692e-01 -4.14060280e-02
8.46294701e-01 -6.98307395e-01 -3.31187725e-01 2.67834425e-01
-1.28296447e+00 -3.47371697e-01 3.08735460e-01 2.44594157e-01
2.79992878e-01 6.31640911e-01 6.32222712e-01 -1.18634450e+00
4.59094405e-01 -6.56985164e-01 -1.15352702e+00 3.52970392e-01
-2.39840746e-01 9.49409604e-02 9.17288601e-01 -6.81929231e-01
-4.02550012e-01 8.72730911e-02 -5.97280003e-02 -5.00655711e-01
-3.47799182e-01 2.01477349e-01 -4.96029407e-01 -8.55018944e-02
5.84840953e-01 2.25598231e-01 -1.19484581e-01 -3.20793390e-01
1.03851259e-01 5.82025528e-01 4.78457689e-01 -8.63799870e-01
1.33883858e+00 6.93283200e-01 2.81355023e-01 -5.17485321e-01
-9.40235972e-01 -1.64931908e-01 -2.99466640e-01 3.75155866e-01
4.73927498e-01 -6.37522161e-01 -1.56902122e+00 9.49014723e-01
-9.90401030e-01 -7.19225824e-01 -4.34940457e-02 5.71796373e-02
-1.99116617e-01 5.64870894e-01 -5.97482622e-01 -8.49608481e-01
-5.85002005e-01 -7.55676031e-01 6.23499095e-01 -5.46736233e-02
-2.15660647e-01 -1.04618549e+00 -1.32762194e-01 1.59967169e-01
4.20202196e-01 5.93833268e-01 9.02578354e-01 -1.27416015e+00
-5.47792375e-01 -5.21039307e-01 -3.85121822e-01 2.36513004e-01
1.51805297e-01 -8.40656161e-02 -1.19605350e+00 -4.72089142e-01
-1.93288162e-01 -5.76227307e-01 5.42540252e-01 1.24022681e-02
1.76800406e+00 -9.94924366e-01 -3.55280906e-01 7.32133627e-01
1.41002440e+00 -1.35527492e-01 7.34110296e-01 4.83821422e-01
4.64793652e-01 7.64488429e-02 3.67410690e-01 7.03502059e-01
1.43954754e-01 1.66565582e-01 6.29536152e-01 -5.18642366e-02
7.00308621e-01 -6.55579925e-01 1.42058268e-01 -7.29544163e-01
3.81909758e-01 -2.10229978e-01 -7.96963871e-01 4.19562310e-01
-1.62819982e+00 -1.19163811e+00 1.99938819e-01 2.38716340e+00
1.13453031e+00 2.81064063e-01 4.62441832e-01 1.46564022e-01
3.95019919e-01 1.37340963e-01 -6.88702345e-01 -2.65134573e-01
9.25208107e-02 4.13161635e-01 1.46948576e+00 4.93556917e-01
-1.47456896e+00 8.47365499e-01 5.98398542e+00 1.03135264e+00
-8.43446553e-01 6.74433857e-02 7.60364175e-01 -2.38454461e-01
-2.93617368e-01 -7.32970750e-03 -9.32477891e-01 4.16717976e-01
1.22830069e+00 -4.61745948e-01 4.78318989e-01 9.33212519e-01
-2.25292683e-01 -2.70199776e-03 -1.27268147e+00 4.47643250e-01
-3.09567660e-01 -1.37568426e+00 -6.13671578e-02 5.37224293e-01
2.41080791e-01 1.47056893e-01 1.96581557e-01 2.40370631e-02
8.95496309e-01 -1.27082956e+00 5.15336394e-01 1.95153609e-01
8.65755975e-01 -1.37074053e+00 6.17817402e-01 6.34929359e-01
-9.74035740e-01 -1.07164562e-01 -4.54184145e-01 3.02488394e-02
-2.73578554e-01 2.18446016e-01 -8.28191578e-01 3.39039326e-01
7.51482844e-01 1.41352639e-01 -5.68845451e-01 8.28601360e-01
-4.27394450e-01 1.20070839e+00 -8.97714496e-01 -1.92543000e-01
1.91552907e-01 3.75176936e-01 3.59879911e-01 1.04419708e+00
-3.17848146e-01 4.84890044e-02 2.80022640e-02 1.13687086e+00
-5.48567057e-01 -3.88815790e-01 -6.97856903e-01 -4.93212184e-03
8.94722760e-01 9.87614036e-01 -1.58947378e-01 -4.12313402e-01
-1.84920356e-01 5.78920245e-01 2.40357056e-01 1.98553056e-01
-1.39934444e+00 -5.37265182e-01 1.05598831e+00 1.36444241e-01
3.46848100e-01 -6.95912167e-02 -1.21216536e-01 -9.37627375e-01
-8.60777423e-02 -1.00570321e+00 7.63864219e-01 -1.39171571e-01
-1.29759884e+00 1.53164744e-01 4.87182617e-01 -6.72199070e-01
2.69744098e-01 -4.34771597e-01 -7.01506495e-01 8.04594278e-01
-1.49799967e+00 -1.38259995e+00 2.14601040e-01 9.24109399e-01
-4.88666028e-01 9.40273851e-02 1.03125858e+00 1.75458685e-01
-6.98141456e-01 1.35091639e+00 -1.86172351e-01 7.57420719e-01
1.67774722e-01 -9.26058710e-01 9.15136456e-01 1.05600667e+00
2.16821924e-01 6.64167464e-01 6.62862778e-01 -7.63434827e-01
-1.42368126e+00 -1.28697383e+00 6.82718277e-01 -7.48871982e-01
1.06225657e+00 -6.35104656e-01 -9.22023952e-01 1.09278619e+00
-9.53482687e-02 2.10715085e-01 9.56247509e-01 -2.51007248e-02
-1.14531100e+00 -9.57874879e-02 -1.80525136e+00 6.61300242e-01
7.67680824e-01 -7.13973105e-01 -1.54659942e-01 1.84801430e-01
8.18000913e-01 -1.24621494e-02 -1.02355433e+00 3.42721969e-01
8.69592190e-01 -7.01210558e-01 8.88912916e-01 -1.12950051e+00
-2.42836937e-01 -2.24267930e-01 -1.30626678e-01 -4.29107755e-01
-1.12840973e-01 -8.65258455e-01 -5.09338021e-01 1.18739212e+00
7.93124855e-01 -1.12345493e+00 1.14178681e+00 1.05051708e+00
1.12927270e+00 -5.97512960e-01 -1.02564371e+00 -8.95799160e-01
5.29707074e-01 -6.24592781e-01 1.24266684e+00 9.29580450e-01
-1.25490269e-02 -1.18898779e-01 -3.82314980e-01 1.05099225e+00
1.12285197e+00 -1.41448170e-01 1.03572929e+00 -9.92327034e-01
-4.59013820e-01 -8.05824995e-02 -3.46797884e-01 -5.26168883e-01
3.72882724e-01 -6.84797883e-01 -6.26165748e-01 -8.36800873e-01
4.48037148e-01 -5.68752229e-01 -5.04005432e-01 1.31941652e+00
2.16264084e-01 3.30328792e-01 -4.21415456e-02 -6.11286871e-02
-5.21533608e-01 7.43052810e-02 4.80841041e-01 -3.14650238e-01
9.47758108e-02 3.89185339e-01 -9.02230501e-01 5.74130297e-01
1.40830541e+00 -9.49918211e-01 -3.85036588e-01 7.93374926e-02
4.48943824e-01 -2.35879600e-01 7.72881866e-01 -9.68596399e-01
2.17511863e-01 -2.50022292e-01 3.35450828e-01 -5.70470810e-01
7.49726966e-02 -1.15998673e+00 4.36459213e-01 8.42078626e-01
-4.43834066e-01 -1.10733554e-01 4.20193166e-01 5.20518422e-01
5.00717700e-01 -7.77317807e-02 8.64085734e-01 1.33780509e-01
1.12073701e-02 8.82303953e-01 -2.35239029e-01 1.08976305e-01
1.32444036e+00 1.23256035e-01 -8.52911234e-01 -3.54616612e-01
-2.98062176e-01 2.49592662e-01 7.23415554e-01 1.27125625e-03
5.66423416e-01 -9.04331505e-01 -7.07895279e-01 3.23324621e-01
7.43681714e-02 -4.16236185e-02 -1.62028790e-01 6.59451663e-01
-4.25148487e-01 2.36371860e-01 -1.12091087e-01 6.53455257e-02
-1.48956549e+00 1.02082360e+00 4.40773726e-01 -3.82727146e-01
-6.34540558e-01 5.95103502e-01 1.04038514e-01 -1.86637014e-01
4.80680287e-01 -2.39129961e-01 3.57448250e-01 -5.50140917e-01
1.02957726e+00 4.48973238e-01 -4.36433703e-01 -3.01115364e-01
-4.93236810e-01 1.45625668e-02 -2.52468169e-01 -2.69527715e-02
1.17968392e+00 5.01056135e-01 -1.15995362e-01 -2.86001176e-01
1.48573077e+00 2.03873590e-01 -1.08572316e+00 -3.96272033e-01
-1.93475094e-02 -6.18919909e-01 -2.37981111e-01 -7.66281486e-01
-1.17935932e+00 5.89812994e-01 1.73671126e-01 3.91401827e-01
1.07957995e+00 1.24649078e-01 9.58402574e-01 7.27002740e-01
7.22384095e-01 -4.82070029e-01 -4.54856992e-01 -1.53827586e-03
2.80687243e-01 -8.13303947e-01 2.73296922e-01 -2.94418007e-01
-2.78164059e-01 5.80457568e-01 5.00345647e-01 -1.83020905e-01
9.16318238e-01 7.61930406e-01 -2.34417245e-01 -1.68272898e-01
-8.18898797e-01 2.93975711e-01 -4.30945903e-01 5.81894457e-01
-4.73927319e-01 2.56964296e-01 2.87082493e-01 5.85164249e-01
-1.59276709e-01 -1.64335463e-02 5.31298637e-01 9.92432535e-01
-1.20931350e-01 -1.17564571e+00 -1.77348763e-01 4.78167892e-01
-1.32831144e+00 -1.52970687e-01 -3.60114425e-01 8.43558490e-01
-2.23585188e-01 8.03739488e-01 -1.24423236e-01 -4.13320899e-01
-1.11590996e-01 -1.52789280e-01 2.05381230e-01 -3.43979269e-01
-6.49243832e-01 -4.69780952e-01 3.44475687e-01 -6.70256436e-01
-1.53079525e-01 -5.56200087e-01 -1.10867536e+00 -8.68071318e-01
-4.59202379e-01 3.10507804e-01 3.32734495e-01 5.03687739e-01
2.84820944e-01 -3.77627164e-01 9.56300020e-01 3.60153541e-02
-9.93173897e-01 -6.50284946e-01 -8.40640783e-01 1.58248901e-01
5.53041220e-01 -3.48740518e-01 -5.61184645e-01 -3.54686141e-01] | [5.916811943054199, 7.255186080932617] |
a14cf7df-6c23-4ba9-b7c1-a52338c55002 | service-composition-in-the-chatgpt-era | 2305.15788 | null | https://arxiv.org/abs/2305.15788v1 | https://arxiv.org/pdf/2305.15788v1.pdf | Service Composition in the ChatGPT Era | The paper speculates about how ChatGPT-like systems can support the field of automated service composition and identifies new research areas to explore in order to take advantage of such tools in the field of service-oriented composition. | ['Ilche Georgievski', 'Marco Aiello'] | 2023-05-25 | null | null | null | null | ['service-composition'] | ['miscellaneous'] | [ 7.95280375e-03 5.49709499e-01 3.54652628e-02 -8.78624320e-01
-2.56785393e-01 -6.13273859e-01 8.20349336e-01 -4.18589920e-01
3.08837324e-01 1.11212522e-01 6.11559570e-01 -9.20636952e-01
-1.75937235e-01 -7.11118758e-01 2.09018916e-01 -5.54844677e-01
-4.09336574e-02 9.10891712e-01 8.64778101e-01 -7.98639059e-01
3.63784790e-01 2.91209191e-01 -1.57458317e+00 4.26075429e-01
2.56727487e-01 5.16307473e-01 -1.90202966e-01 3.13494861e-01
-1.11055791e+00 8.53759348e-01 -5.36784768e-01 -3.43221754e-01
3.60163301e-01 -7.05936074e-01 -1.13740170e+00 4.19389486e-01
-3.53217304e-01 1.35081202e-01 2.33240604e-01 9.37101603e-01
2.19126016e-01 8.79140720e-02 -4.79427874e-02 -1.58027613e+00
-2.35411927e-01 1.11721134e+00 3.12568903e-01 3.59728336e-02
7.96685815e-01 4.27626729e-01 1.05178404e+00 -1.89611956e-01
8.96269679e-01 1.18232286e+00 4.52880532e-01 7.98341751e-01
-1.03689671e+00 -3.39331061e-01 1.34089455e-01 3.17211390e-01
-8.73678684e-01 -5.34928381e-01 3.31818730e-01 4.47710976e-03
1.20150661e+00 9.16830897e-01 7.07409263e-01 8.08970511e-01
5.54668903e-02 4.43863660e-01 1.21732056e+00 -5.54595351e-01
3.75343293e-01 8.36493894e-02 1.76153183e-01 4.86828268e-01
9.13542062e-02 -4.12595242e-01 -5.23689747e-01 -7.26708591e-01
2.22848937e-01 -3.15149397e-01 1.20877698e-01 -2.65907645e-01
-1.16712451e+00 6.01630032e-01 -2.90949106e-01 1.35887802e+00
-2.00241551e-01 1.32075503e-01 6.75689280e-01 1.06161022e+00
4.04126048e-01 7.19073117e-01 -5.18467188e-01 -9.79583561e-01
-2.97468990e-01 5.42754054e-01 1.93312585e+00 1.09719789e+00
3.29721272e-01 3.76788914e-01 4.17030215e-01 2.65189409e-01
6.90286875e-01 -6.97139502e-02 1.93081975e-01 -1.38529861e+00
-6.53229654e-02 8.76932800e-01 7.96457976e-02 7.19698593e-02
-4.90590371e-02 2.88132906e-01 6.51606619e-01 1.33735403e-01
2.97203511e-01 -5.61334528e-02 -7.54070207e-02 6.15181744e-01
4.55736488e-01 -7.01990649e-02 -1.13111734e-01 9.80998993e-01
7.03841984e-01 2.61756808e-01 -8.89987126e-02 -2.34110832e-01
1.33429515e+00 -1.00771451e+00 -4.59964514e-01 4.05218363e-01
7.45845437e-01 -6.51787937e-01 1.01617539e+00 2.12384015e-01
-1.20397055e+00 7.52291143e-01 -5.59684098e-01 3.85574728e-01
-1.33747429e-01 -1.11516523e+00 1.25482476e+00 9.91727233e-01
-1.35124934e+00 4.53878373e-01 -9.93385077e-01 -9.91704047e-01
-3.25657874e-01 2.55701810e-01 -1.26261571e-02 2.30293900e-01
-5.87883353e-01 7.30114341e-01 -1.43601760e-01 -2.93097436e-01
-8.81948948e-01 -5.93858436e-02 -4.62360293e-01 4.21072841e-01
9.09725189e-01 -3.68028820e-01 2.01902962e+00 -9.58688080e-01
-1.95718050e+00 4.10898119e-01 -3.83430943e-02 -2.51125306e-01
2.93064833e-01 5.06719351e-01 -6.97384715e-01 -1.28589392e-01
7.01714531e-02 -4.46731806e-01 2.31648296e-01 -8.15659404e-01
-1.00535893e+00 -2.61938989e-01 2.18274489e-01 1.16713233e-01
-1.42449468e-01 1.09835708e+00 7.59881660e-02 2.98202634e-01
7.91808665e-02 -8.86545837e-01 -4.57551867e-01 -4.98674929e-01
1.27122626e-01 -7.43024647e-01 8.19959223e-01 -2.35174850e-01
5.61711788e-01 -1.77240193e+00 -1.75773337e-01 5.26941776e-01
-9.62610245e-02 8.02740678e-02 -1.24176584e-01 1.41148758e+00
5.29824972e-01 2.83277422e-01 1.74250603e-01 2.85145253e-01
7.34877706e-01 6.57208979e-01 1.02251530e-01 4.84235398e-02
-3.17175418e-01 3.57245475e-01 -8.37929130e-01 -1.75808266e-01
2.60253362e-02 1.08539931e-01 -3.31383944e-01 2.44152084e-01
-9.63089645e-01 6.18647397e-01 -9.09780443e-01 8.32217395e-01
9.11545679e-02 -1.26826959e-02 1.03586209e+00 7.73057938e-01
-8.07371318e-01 1.00386119e+00 -1.01999795e+00 1.34160078e+00
-2.53972709e-01 1.83249697e-01 5.00931323e-01 -7.22798884e-01
1.01630938e+00 1.23342013e+00 8.34593236e-01 -4.65515316e-01
4.38051850e-01 4.96510535e-01 5.73915422e-01 -8.93185496e-01
1.20107040e-01 -2.07741365e-01 1.70553401e-01 1.37432086e+00
1.95064664e-01 3.37088071e-02 2.81813145e-01 1.60773676e-02
1.45557821e+00 4.05716181e-01 8.30507129e-02 -2.66090810e-01
6.38463974e-01 1.07608184e-01 3.58456582e-01 5.69311738e-01
-5.21417335e-02 -4.02897924e-01 6.14962757e-01 -1.00049579e+00
-9.61810410e-01 -2.62376130e-01 2.96110183e-01 1.43442440e+00
-4.71521243e-02 -8.46879721e-01 -7.33989298e-01 -9.64281142e-01
-1.88119143e-01 8.02316904e-01 1.54332936e-01 6.05312586e-01
-7.71785140e-01 -3.56802106e-01 4.63176548e-01 3.23045224e-01
3.40291746e-02 -1.48908520e+00 -5.77775598e-01 6.57907128e-01
-6.59474209e-02 -1.08247542e+00 9.19547155e-02 -2.03304440e-01
-6.83486879e-01 -1.26390004e+00 -3.27869086e-03 -5.99683821e-01
4.99432720e-02 3.52013499e-01 1.03557253e+00 7.56636262e-01
4.24954027e-01 8.14551532e-01 -9.18358862e-01 -1.52802616e-01
-9.30991769e-01 5.23919702e-01 -3.03257108e-01 -6.82715848e-02
9.92151201e-01 -1.20632136e+00 -1.57786280e-01 9.27800715e-01
-4.68845755e-01 -1.66312814e-01 -8.25249702e-02 1.24945022e-01
-3.68970633e-01 -2.45224252e-01 4.86586541e-01 -9.63691235e-01
7.42323577e-01 -4.84894127e-01 -6.37322366e-01 1.92670628e-01
-9.14185703e-01 -1.89658910e-01 4.96867210e-01 -7.47968107e-02
-1.11895633e+00 -5.08650005e-01 -6.71184480e-01 6.23747051e-01
-1.80049524e-01 3.61671627e-01 -3.32372576e-01 -7.47176945e-01
2.10678667e-01 2.88153619e-01 3.07550460e-01 -8.76708388e-01
2.74966210e-01 6.60381556e-01 -1.96818952e-02 -8.36131394e-01
7.21173763e-01 5.07599711e-01 -3.37382704e-01 -5.28840959e-01
3.43912542e-01 -6.72848940e-01 2.68566042e-01 -1.88684300e-01
1.74085468e-01 1.66951790e-02 -1.26543033e+00 -1.43136755e-01
-7.41653800e-01 -3.43444198e-01 -2.79464602e-01 -1.16432242e-01
-8.09909344e-01 2.79104531e-01 -8.28821361e-01 -1.12001264e+00
-4.81952339e-01 -9.98331547e-01 6.11471653e-01 1.08651817e-01
-6.61162019e-01 -7.56582439e-01 1.02194063e-01 4.82310325e-01
9.80852783e-01 -5.64859986e-01 6.90827310e-01 -1.35590947e+00
-8.20465028e-01 -2.20527917e-01 2.23267779e-01 1.30936980e-01
-4.53321449e-03 1.97520822e-01 -2.83329517e-01 4.52892743e-02
2.81757951e-01 5.55712044e-01 -5.46486735e-01 -5.31769931e-01
3.79271179e-01 -6.46054089e-01 -2.22768560e-01 4.56669368e-02
1.12421155e+00 5.47071040e-01 5.44494390e-01 6.86296105e-01
1.99686140e-01 8.27219844e-01 3.46281290e-01 2.08497480e-01
6.55669510e-01 9.10443604e-01 1.14719786e-01 5.33776879e-01
1.65579766e-02 2.05433015e-02 2.48642325e-01 1.04967427e+00
-3.73360604e-01 2.89187253e-01 -1.27706873e+00 4.64003801e-01
-2.07976174e+00 -1.04919446e+00 -4.94214922e-01 1.18942928e+00
3.07267576e-01 -1.95668906e-01 6.36983395e-01 4.14193898e-01
4.59375829e-01 -9.13698077e-02 6.59912527e-02 -1.14775968e+00
4.16585296e-01 2.31440291e-01 1.52855232e-01 3.46493244e-01
4.64885496e-02 9.60850239e-01 7.18658113e+00 -1.01545200e-01
-9.85169888e-01 4.92889732e-01 -4.81213868e-01 5.40117502e-01
-6.73070431e-01 9.06174362e-01 -4.92867857e-01 3.19683611e-01
1.53028691e+00 -3.95674497e-01 8.33016932e-01 9.58736598e-01
4.34542239e-01 1.83703750e-01 -8.35086763e-01 -2.49294817e-01
-1.71926409e-01 -1.51619208e+00 -2.80854374e-01 1.95147887e-01
5.23342073e-01 2.80188501e-01 -6.97874844e-01 -3.50018591e-02
5.95812678e-01 -2.72770166e-01 6.51101589e-01 2.16293558e-01
-1.09823883e-01 -1.30269527e-01 7.88140476e-01 2.88325518e-01
-8.12485218e-01 -2.09374800e-01 2.14138120e-01 -5.67995191e-01
6.95429504e-01 -1.11159623e-01 -1.33668530e+00 6.22287810e-01
7.69683003e-01 1.92791462e-01 1.92589089e-01 1.21930075e+00
-7.24111870e-02 8.24816823e-01 -1.09078519e-01 -5.19209802e-01
3.64902318e-01 -6.09154522e-01 7.48053014e-01 1.18074238e+00
4.84706879e-01 8.90668780e-02 4.02001679e-01 5.22473872e-01
4.83193606e-01 2.35549256e-01 -4.11043078e-01 -3.62189263e-01
3.48197728e-01 1.11392546e+00 -8.43063295e-01 -5.24516881e-01
-8.47275257e-01 4.30969864e-01 -6.49940789e-01 2.84400254e-01
-1.03480548e-01 1.00434162e-01 7.37196863e-01 6.63725793e-01
3.07917804e-01 -5.79434335e-01 -1.90842316e-01 -1.01394629e+00
-1.36969924e-01 -1.47564566e+00 3.34730119e-01 -6.21371746e-01
-9.87891138e-01 5.71086466e-01 -2.63812840e-01 -7.12478578e-01
-5.74544966e-01 6.37581274e-02 -9.47801650e-01 5.06155849e-01
-9.11637425e-01 -1.26049531e+00 1.47697460e-02 3.25448602e-01
5.76424062e-01 -2.94594705e-01 1.27557731e+00 1.74212992e-01
9.34373811e-02 -1.22024931e-01 -1.80030674e-01 -1.23481110e-01
1.84520677e-01 -9.38513696e-01 1.13973534e+00 5.63344181e-01
3.85553122e-01 4.67186153e-01 1.01820946e+00 -2.94620693e-01
-2.01828432e+00 -2.94769198e-01 1.38096166e+00 -3.66082489e-01
1.11498463e+00 -3.01810294e-01 -7.36586094e-01 9.35946584e-01
1.83116868e-01 -4.19991285e-01 6.20077193e-01 3.40235502e-01
-4.77335274e-01 -2.55976707e-01 -1.32808542e+00 6.00235581e-01
1.46134937e+00 -4.82090861e-01 -4.62120354e-01 4.20505524e-01
8.39184403e-01 -6.31454960e-02 -9.40334678e-01 -1.49002537e-01
6.06643617e-01 -1.02866757e+00 2.19314635e-01 -6.81840658e-01
-5.01927555e-01 -4.78014112e-01 -7.32368156e-02 -8.44830692e-01
1.57221090e-02 -1.34951317e+00 9.73738581e-02 1.13850212e+00
3.46021026e-01 -1.26677132e+00 8.47760558e-01 9.15777981e-01
-4.60452199e-01 -1.67706385e-01 -1.00611866e+00 -9.05841768e-01
-5.18812001e-01 -3.73009682e-01 1.39796627e+00 8.21946025e-01
7.96577275e-01 4.21246849e-02 1.72510192e-01 -4.03808832e-01
1.06991746e-01 8.89438987e-02 7.19821274e-01 -1.22736347e+00
-2.51510203e-01 -3.46812457e-01 -4.68640834e-01 -4.03581083e-01
1.27005681e-01 -8.43205631e-01 7.66233653e-02 -1.62393224e+00
-4.87656474e-01 -5.75852275e-01 7.44747892e-02 5.78975797e-01
8.97289932e-01 -2.44548898e-02 3.41914833e-01 7.58166462e-02
-9.08983469e-01 -1.64746404e-01 9.95372176e-01 1.88083380e-01
6.47712573e-02 4.95057493e-01 -1.08946657e+00 3.29340518e-01
7.20396638e-01 -5.77480555e-01 -2.07381830e-01 -2.38376141e-01
2.82942086e-01 4.15213764e-01 -2.47794777e-01 -6.03773713e-01
2.96915263e-01 -6.94332182e-01 -9.63352621e-01 5.46190202e-01
6.24079537e-03 -1.09103060e+00 7.48880863e-01 2.11213395e-01
-3.06778491e-01 4.17829961e-01 -5.43435216e-01 -1.26205562e-02
-9.01124720e-03 -3.60426664e-01 -1.84774175e-01 -7.41998732e-01
-7.68782973e-01 -2.40662079e-02 -1.31701946e+00 -5.23528039e-01
8.31615627e-01 -2.00209454e-01 -3.69758904e-01 -5.48118651e-01
-6.26442850e-01 2.64708325e-02 8.65532339e-01 5.99543989e-01
1.34322360e-01 -7.02229857e-01 -4.55051988e-01 1.16190098e-01
2.04899788e-01 -1.00738776e+00 -4.60511506e-01 1.09849858e+00
-4.90556479e-01 3.42178673e-01 -4.04095441e-01 -1.85408220e-01
-1.39879656e+00 2.86907971e-01 3.48143041e-01 3.31713647e-01
-1.12842309e+00 1.34947628e-01 -9.10035908e-01 -4.04340565e-01
1.69616103e-01 1.45736486e-01 1.92419589e-01 -6.28289402e-01
8.58681321e-01 2.14768082e-01 4.42985982e-01 -3.61448199e-01
-4.96973693e-01 -2.47240439e-01 3.82673651e-01 -9.04814124e-01
1.47083735e+00 -1.68160424e-01 -8.48856866e-01 4.90835607e-01
3.96040052e-01 -3.21434021e-01 -5.81135035e-01 -1.36103421e-01
8.46319616e-01 -6.86214983e-01 -6.04253411e-01 -8.60607147e-01
-4.49428022e-01 2.24220231e-01 5.12981191e-02 1.03955257e+00
5.39841592e-01 1.48477808e-01 7.16640413e-01 3.24610204e-01
9.65361238e-01 -9.03049648e-01 -5.18285632e-01 5.60833216e-01
6.03935480e-01 -4.14874196e-01 -6.38720155e-01 -5.39183259e-01
-6.57704294e-01 1.22907686e+00 9.20539722e-02 -2.08719343e-01
2.83536732e-01 6.10201776e-01 4.25678968e-01 -6.48162484e-01
-1.30848372e+00 -3.62759143e-01 -5.25133669e-01 1.00634789e+00
4.83720124e-01 3.67162734e-01 -1.03313518e+00 2.52375901e-01
-1.45827666e-01 1.70790002e-01 7.18437135e-01 1.27427900e+00
-4.50431436e-01 -2.09859061e+00 -1.46448240e-01 3.68406057e-01
-4.45866376e-01 1.41196638e-01 -5.82428455e-01 5.95801175e-01
-2.09275454e-01 1.23799777e+00 -1.13228351e-01 -2.88513005e-01
2.84306616e-01 7.57746637e-01 4.11337733e-01 -9.17375565e-01
-1.42437112e+00 2.94176228e-02 1.21595776e+00 -7.70036876e-01
-3.92823935e-01 -9.82287109e-01 -1.17701375e+00 -7.73996294e-01
-2.11328998e-01 9.73325551e-01 1.12179649e+00 1.10103106e+00
2.20260292e-01 1.28180683e-01 4.06693757e-01 -3.58044386e-01
-4.57627177e-01 -7.02699840e-01 -4.05447185e-01 1.50210597e-02
-4.22015369e-01 -1.32166475e-01 -2.89215475e-01 -3.08517516e-01] | [8.634676933288574, 6.951788425445557] |
a68ae607-8c04-41ef-ad2f-207d5cdf768f | tips-text-induced-pose-synthesis | 2207.11718 | null | https://arxiv.org/abs/2207.11718v1 | https://arxiv.org/pdf/2207.11718v1.pdf | TIPS: Text-Induced Pose Synthesis | In computer vision, human pose synthesis and transfer deal with probabilistic image generation of a person in a previously unseen pose from an already available observation of that person. Though researchers have recently proposed several methods to achieve this task, most of these techniques derive the target pose directly from the desired target image on a specific dataset, making the underlying process challenging to apply in real-world scenarios as the generation of the target image is the actual aim. In this paper, we first present the shortcomings of current pose transfer algorithms and then propose a novel text-based pose transfer technique to address those issues. We divide the problem into three independent stages: (a) text to pose representation, (b) pose refinement, and (c) pose rendering. To the best of our knowledge, this is one of the first attempts to develop a text-based pose transfer framework where we also introduce a new dataset DF-PASS, by adding descriptive pose annotations for the images of the DeepFashion dataset. The proposed method generates promising results with significant qualitative and quantitative scores in our experiments. | ['Michael Blumenstein', 'Umapada Pal', 'Saumik Bhattacharya', 'Subhankar Ghosh', 'Prasun Roy'] | 2022-07-24 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 5.05217075e-01 2.27360994e-01 3.18000048e-01 -5.63225269e-01
-8.69889677e-01 -4.65137720e-01 9.34221268e-01 -1.87675301e-02
-5.48340142e-01 6.71462536e-01 2.81994849e-01 3.89210731e-01
-1.88761726e-02 -5.87093711e-01 -7.11911201e-01 -4.47851002e-01
3.70380342e-01 1.26330364e+00 4.44517612e-01 -1.58764631e-01
2.05726370e-01 5.22995412e-01 -1.66483235e+00 1.73555762e-01
5.03035069e-01 9.04594541e-01 1.55769169e-01 5.81790090e-01
2.53215849e-01 2.82003373e-01 -6.98707283e-01 -5.42270124e-01
3.91466677e-01 -5.94051778e-01 -7.69465625e-01 4.70347553e-01
6.36103988e-01 -3.75709027e-01 1.86892580e-02 7.71738470e-01
7.08404183e-01 2.20520407e-01 7.51613855e-01 -1.18655407e+00
-2.10490599e-02 1.79382175e-01 -5.15897095e-01 -3.19842279e-01
6.69355810e-01 1.36153530e-02 6.15412951e-01 -8.34029138e-01
7.97063410e-01 1.25427330e+00 4.69320208e-01 6.85982049e-01
-1.16451573e+00 -4.75965381e-01 1.43722221e-02 -2.21071821e-02
-1.39496529e+00 -3.61463964e-01 9.28332090e-01 -6.31628275e-01
4.40799505e-01 2.22559422e-01 7.91278481e-01 1.26261091e+00
-7.27669746e-02 7.21310794e-01 1.34660757e+00 -5.94674170e-01
1.48934707e-01 2.24757612e-01 -3.27000469e-01 5.37424505e-01
1.08910657e-01 1.09677993e-01 -5.85244358e-01 1.04412295e-01
7.75644958e-01 -1.81700110e-01 -1.82086214e-01 -8.17488611e-01
-1.33079183e+00 6.16480887e-01 5.56903899e-01 1.18085898e-01
-4.41342413e-01 8.89579803e-02 9.32236761e-02 -2.10987926e-01
4.00167346e-01 4.45232809e-01 -2.14740396e-01 1.32462988e-02
-1.09893978e+00 6.56296432e-01 7.58868337e-01 8.98409963e-01
6.61324203e-01 -3.46498251e-01 -2.54047632e-01 4.42824453e-01
2.75722653e-01 3.89416277e-01 2.60433108e-01 -6.82282209e-01
4.77674961e-01 5.07272363e-01 2.78983295e-01 -9.20716822e-01
-3.67365211e-01 -4.71166909e-01 -3.66995245e-01 2.22256258e-01
6.33809388e-01 -2.30702773e-01 -1.05635703e+00 1.56318676e+00
6.74711049e-01 -2.96902414e-02 -2.82055855e-01 1.15943813e+00
7.27890134e-01 4.84914273e-01 -3.08117010e-02 -7.87608325e-03
1.34345305e+00 -9.31981683e-01 -5.99690020e-01 -4.82397825e-01
8.97617191e-02 -9.26220477e-01 7.16332316e-01 4.64433610e-01
-9.79346395e-01 -5.61862230e-01 -9.54465508e-01 -4.72956486e-02
-1.17351763e-01 4.00784701e-01 4.65070844e-01 5.99968612e-01
-8.39982510e-01 2.12932035e-01 -7.12496459e-01 -7.71738529e-01
1.61649123e-01 4.02599424e-01 -4.64919329e-01 8.38124901e-02
-9.49594617e-01 1.04797268e+00 4.59990025e-01 2.14450523e-01
-8.31477225e-01 -2.62576848e-01 -7.08957613e-01 -3.90909106e-01
7.52597094e-01 -9.28073645e-01 1.26251936e+00 -9.96106088e-01
-1.37563884e+00 1.05938554e+00 -1.49899796e-01 -3.25504333e-01
9.93152559e-01 -6.08166814e-01 7.13460743e-02 5.36621213e-02
1.18029043e-01 9.91980433e-01 8.85210991e-01 -1.59759951e+00
-5.75232208e-01 -5.75330079e-01 9.86660123e-02 5.15079260e-01
-9.41251293e-02 -8.61156732e-03 -8.40435922e-01 -6.46156788e-01
3.89891826e-02 -1.15357614e+00 -1.58730000e-01 2.30012104e-01
-4.27164346e-01 -2.07944974e-01 6.88055396e-01 -6.45103276e-01
6.95729375e-01 -1.78149712e+00 5.59585452e-01 1.60577614e-02
4.44312692e-02 3.31606358e-01 8.42188820e-02 6.21635854e-01
6.12311773e-02 -3.13006371e-01 -1.62758678e-01 -7.53891110e-01
-1.16120309e-01 -4.77419198e-02 -2.53120303e-01 5.55062115e-01
1.15701221e-01 8.99937212e-01 -7.28671908e-01 -7.13347673e-01
4.68234509e-01 7.74431705e-01 -4.95461047e-01 4.44261760e-01
-3.52336764e-01 8.34974468e-01 -4.82619852e-01 3.38166475e-01
4.37552363e-01 2.55533963e-01 -1.68564767e-02 -4.39071327e-01
-2.52538896e-03 8.13694764e-03 -1.31226659e+00 1.86948907e+00
-7.24553019e-02 2.23079085e-01 -1.23712599e-01 -6.36783481e-01
8.09704423e-01 3.71959746e-01 6.15470290e-01 -1.94138914e-01
3.37121248e-01 1.97633818e-01 -1.16859861e-01 -2.41530642e-01
6.66835189e-01 -4.12339598e-01 -1.16778016e-01 3.03431064e-01
1.93194807e-01 -2.57598817e-01 1.34848550e-01 -1.60216808e-01
6.28487825e-01 7.83601284e-01 3.20134431e-01 1.14019953e-01
4.44018662e-01 3.98931280e-02 2.16345280e-01 4.13150102e-01
-2.30351433e-01 1.27994883e+00 1.66044503e-01 -4.20382410e-01
-1.06984568e+00 -1.15494156e+00 1.51866823e-01 8.12626123e-01
2.68346947e-02 -3.78473073e-01 -1.01022315e+00 -7.50933886e-01
-2.54858434e-01 4.63290721e-01 -6.83039188e-01 1.04151510e-01
-6.13342702e-01 -3.86377454e-01 4.08418983e-01 4.29761559e-01
6.03114486e-01 -1.20823252e+00 -8.74501646e-01 1.13121331e-01
-5.30861914e-01 -1.18011105e+00 -5.78606129e-01 -2.68457621e-01
-6.17396295e-01 -1.00231135e+00 -9.44132030e-01 -7.08427846e-01
8.73805106e-01 -1.03938788e-01 9.01024818e-01 -1.27004817e-01
-3.19378853e-01 3.54324520e-01 -4.70058322e-01 -5.72406292e-01
-1.23420984e-01 1.04809046e-01 5.33748269e-02 2.84583658e-01
8.99014696e-02 -2.83395588e-01 -6.73188269e-01 4.45102423e-01
-7.99319506e-01 2.88467258e-01 8.46160293e-01 5.22061467e-01
5.99933088e-01 5.03899232e-02 -1.17061948e-02 -7.15148866e-01
3.65896195e-01 1.85599074e-01 -4.31216002e-01 1.55961931e-01
-1.57329902e-01 -4.67640907e-02 2.89219111e-01 -3.98817778e-01
-1.07443285e+00 7.92824626e-01 -1.89354911e-01 -3.90407324e-01
-4.59371626e-01 2.58967161e-01 -3.23633671e-01 9.36288312e-02
7.91805685e-01 3.36614281e-01 -1.13426773e-02 -4.61469442e-01
3.21180820e-01 3.88694376e-01 6.59833729e-01 -6.86543167e-01
1.23063648e+00 5.94224691e-01 1.58749282e-01 -7.48704076e-01
-9.57738817e-01 -4.80666518e-01 -1.13207936e+00 -5.00151455e-01
1.04463112e+00 -7.83973813e-01 -5.65931022e-01 4.31938499e-01
-1.23145843e+00 -4.27879430e-02 -2.40294173e-01 4.78896111e-01
-7.85293102e-01 2.83584714e-01 4.21045348e-02 -8.11378717e-01
-2.00612873e-01 -1.25189435e+00 1.58004296e+00 -4.99554351e-02
-4.73713905e-01 -6.62037253e-01 1.59728244e-01 7.85500467e-01
1.12463549e-01 5.74916780e-01 1.21935822e-01 -3.71092618e-01
-5.07941127e-01 -4.97920692e-01 7.82022700e-02 1.59116253e-01
5.77525236e-02 -3.26652408e-01 -9.58745062e-01 -4.96473670e-01
-1.08937800e-01 -3.89983237e-01 5.87275743e-01 2.58852899e-01
8.43952119e-01 -2.72407159e-02 -4.04055566e-01 3.97645533e-01
1.19807935e+00 -4.54871692e-02 6.22950792e-01 2.28042081e-01
7.42497921e-01 9.50387716e-01 7.91605175e-01 2.86838859e-01
3.97524029e-01 1.22594678e+00 4.14867938e-01 -1.47311956e-01
-3.64545912e-01 -5.82828879e-01 3.20463359e-01 1.87790722e-01
-4.80984807e-01 -3.56221259e-01 -9.65460479e-01 4.24508750e-01
-1.78785026e+00 -8.32873523e-01 5.33772148e-02 2.32039595e+00
5.89404941e-01 1.60602733e-01 3.58752429e-01 1.78359613e-01
7.33121037e-01 -6.31872639e-02 -2.44316563e-01 1.95708126e-01
3.53353739e-01 1.09719969e-01 2.54197150e-01 4.49040681e-01
-1.21268439e+00 9.72077131e-01 5.53975630e+00 4.87153143e-01
-1.01857758e+00 -7.30689839e-02 3.10728699e-01 -1.73199903e-02
1.28872871e-01 1.49789482e-01 -8.22937310e-01 8.48507583e-02
3.98524344e-01 8.45598504e-02 7.08555058e-02 7.11703300e-01
1.33724734e-01 -1.94487259e-01 -1.33777499e+00 1.03900790e+00
4.34664845e-01 -6.80804133e-01 2.06311747e-01 2.06665292e-01
5.43019295e-01 -5.70750773e-01 9.96575430e-02 1.16479769e-01
-2.00766176e-01 -1.08270800e+00 1.08235598e+00 6.49118245e-01
6.35419548e-01 -8.38752508e-01 6.73667729e-01 4.61898744e-01
-1.09633279e+00 2.37439319e-01 2.63455859e-03 -1.36089921e-01
3.63668472e-01 4.27058458e-01 -1.14327180e+00 6.98537827e-01
5.06647289e-01 3.52623612e-01 -7.94884801e-01 1.17833626e+00
-3.68387550e-01 2.49311760e-01 -2.15833306e-01 4.75245975e-02
-5.17030135e-02 4.01528031e-02 7.60169804e-01 9.46811736e-01
3.06287259e-01 -3.37907970e-01 4.64328319e-01 7.86224663e-01
1.88506246e-01 1.36845514e-01 -4.05393600e-01 1.31056368e-01
9.91783068e-02 1.32121444e+00 -9.22169328e-01 -6.49717078e-02
-1.98849924e-02 1.23459172e+00 2.17144921e-01 4.78915237e-02
-8.62330258e-01 -9.64151993e-02 1.34680554e-01 6.46336198e-01
1.77216887e-01 -3.62948149e-01 -8.59872326e-02 -1.00855231e+00
2.30717048e-01 -7.36028552e-01 2.07940236e-01 -9.55167830e-01
-9.63031530e-01 7.03666151e-01 4.85378891e-01 -1.37080657e+00
-5.12200177e-01 -4.16602701e-01 -1.89008281e-01 7.92073071e-01
-9.21418011e-01 -1.68380749e+00 -5.30007780e-01 4.22495544e-01
7.00020313e-01 -4.30524349e-02 6.78861737e-01 1.90561414e-01
-2.79031489e-02 4.17424828e-01 -5.63335776e-01 1.55748308e-01
9.96103406e-01 -1.21251547e+00 3.02769423e-01 9.16074812e-01
2.10100025e-01 5.62777638e-01 1.10310447e+00 -7.66887426e-01
-1.09221578e+00 -1.01977003e+00 8.32387686e-01 -7.07035005e-01
1.90549359e-01 -6.10944748e-01 -3.94006431e-01 6.94255173e-01
-2.67210435e-02 -2.37506479e-01 1.59876838e-01 -2.60653459e-02
1.19869620e-01 -1.31886616e-01 -1.12168944e+00 6.43769324e-01
1.06601989e+00 -1.52035281e-01 -7.48327553e-01 3.30484420e-01
2.12424204e-01 -7.44756818e-01 -4.89814609e-01 4.69255805e-01
6.28071964e-01 -1.07706606e+00 9.87253368e-01 -1.78085268e-01
4.27370846e-01 -6.30637646e-01 5.06075025e-02 -1.28281903e+00
-2.31990442e-02 -4.32784319e-01 2.00178191e-01 1.28958774e+00
1.30840212e-01 -3.73408318e-01 9.22414958e-01 3.73723060e-01
-6.41507562e-03 -4.92950469e-01 -8.47391903e-01 -5.27377248e-01
-1.52285308e-01 -3.27745795e-01 4.67914462e-01 3.77063543e-01
-4.23637241e-01 3.21426421e-01 -6.58615172e-01 1.18969724e-01
8.46272767e-01 7.29187131e-02 1.24412596e+00 -1.26716423e+00
-3.93591583e-01 -9.59808603e-02 -6.85319364e-01 -9.80816782e-01
-1.08888753e-01 -6.37401044e-01 3.33670974e-01 -1.73463607e+00
3.73818994e-01 -9.55156013e-02 3.65721375e-01 3.39224249e-01
-1.06963642e-01 4.31545377e-01 3.60515445e-01 2.89449215e-01
-4.24122810e-01 5.55785477e-01 1.48492670e+00 1.17893077e-01
9.66264214e-03 2.55465478e-01 -5.88540673e-01 7.28898585e-01
6.51868105e-01 -3.59901428e-01 -6.61084473e-01 -2.31777370e-01
1.29308412e-02 -9.15877968e-02 6.75180316e-01 -1.36567605e+00
1.92366809e-01 -9.50246751e-02 6.35305524e-01 -7.76664972e-01
8.25987935e-01 -9.28516805e-01 4.90859985e-01 5.23814142e-01
-1.82841197e-01 -5.20344041e-02 -3.60099995e-03 5.22464871e-01
-8.75927359e-02 3.45692560e-02 8.60491216e-01 -1.96575686e-01
-6.14006519e-01 2.17717156e-01 -1.87948290e-02 -1.11144774e-01
1.36099112e+00 -4.39805746e-01 -8.11247155e-03 -6.73042178e-01
-8.18201125e-01 2.93073338e-02 5.59820056e-01 5.57858169e-01
7.07147360e-01 -1.24858928e+00 -8.79834294e-01 2.21047960e-02
2.15941042e-01 2.49498025e-01 3.39547656e-02 8.23324025e-01
-4.80122745e-01 4.53191638e-01 -2.93729305e-01 -7.16334999e-01
-1.51003706e+00 3.68382841e-01 4.92544979e-01 -2.67780483e-01
-5.35858929e-01 6.83498204e-01 4.59340662e-01 -5.17582536e-01
2.10106328e-01 7.49728009e-02 -2.44804814e-01 5.44624627e-02
2.29131088e-01 2.30401456e-01 1.10825844e-01 -1.26867008e+00
-3.01708549e-01 8.21755767e-01 1.81399770e-02 -5.45616388e-01
1.21370661e+00 7.06144422e-02 9.11898613e-02 2.96256423e-01
9.15387511e-01 -5.12450794e-03 -1.32559085e+00 -7.89935067e-02
-2.89922744e-01 -5.76417744e-01 -2.14119881e-01 -9.81330931e-01
-8.37882519e-01 7.12950468e-01 5.92693508e-01 -3.79506320e-01
1.10375130e+00 1.12117007e-01 6.94358528e-01 3.41809653e-02
7.40026116e-01 -1.06987774e+00 3.91967952e-01 2.51487315e-01
1.27643502e+00 -9.34395313e-01 1.47840381e-01 -7.16071188e-01
-6.35296702e-01 8.28481436e-01 8.54924798e-01 -7.15866983e-02
3.24184150e-01 7.14431051e-03 4.33608554e-02 -2.61851668e-01
-2.68563479e-01 -4.49260980e-01 7.81638801e-01 7.20344126e-01
4.60627854e-01 -6.39405921e-02 -4.94086683e-01 2.15034798e-01
-4.98476684e-01 6.83692694e-02 3.03771377e-01 1.05420887e+00
-4.32438046e-01 -1.33008409e+00 -8.13407481e-01 1.58141643e-01
-2.76903033e-01 4.04060692e-01 -8.40351164e-01 8.35779011e-01
3.19828063e-01 7.60121107e-01 -2.21965864e-01 -4.20407593e-01
7.45703340e-01 5.10231890e-02 9.09950674e-01 -9.27480102e-01
-4.43371028e-01 2.96853632e-01 1.03187049e-02 -3.82935852e-01
-6.13827288e-01 -8.39670002e-01 -1.05669844e+00 6.09576479e-02
-1.63171083e-01 -4.19473462e-02 6.67840421e-01 9.68340933e-01
-4.46404926e-02 3.65599304e-01 2.59718865e-01 -1.46448600e+00
-3.49170804e-01 -1.03283000e+00 -3.65727484e-01 7.58792818e-01
-1.14417309e-02 -9.90852535e-01 -2.80577242e-02 2.39218697e-01] | [7.120259761810303, -0.9948102235794067] |
03172ddd-771c-403c-aa05-cc041b0df89d | machine-translation-aided-bilingual-data-to | null | null | https://aclanthology.org/2020.webnlg-1.13 | https://aclanthology.org/2020.webnlg-1.13.pdf | Machine Translation Aided Bilingual Data-to-Text Generation and Semantic Parsing | We present a system for bilingual Data-ToText Generation and Semantic Parsing. We use a text-to-text generator to learn a single model that works for both languages on each of the tasks. The model is aided by machine translation during both pre-training and fine-tuning. We evaluate the system on WebNLG 2020 data 1 , which consists of RDF triples in English and natural language sentences in English and Russian for both the tasks. We achieve considerable gains over monolingual models, especially on unseen relations and Russian. | ['Rami Al-Rfou', 'Siamak Shakeri', 'Heming Ge', 'Mihir Kale', 'Oshin Agarwal'] | null | null | null | null | acl-webnlg-inlg-2020-12 | ['data-to-text-generation'] | ['natural-language-processing'] | [-2.78902292e-01 1.00513804e+00 -5.86892068e-01 -8.37144613e-01
-1.24656892e+00 -7.82371759e-01 9.31277215e-01 -7.01890439e-02
-4.82490540e-01 1.65444231e+00 6.75556719e-01 -6.57072604e-01
4.47323859e-01 -9.77056265e-01 -1.04455948e+00 2.95891196e-01
1.93152159e-01 1.59918582e+00 1.52176321e-01 -9.82653439e-01
-5.38062632e-01 -1.28736839e-01 -7.93211758e-01 8.81543756e-01
1.14982235e+00 3.61604750e-01 8.08307379e-02 5.38801908e-01
-6.07757270e-01 1.15226769e+00 -6.22108400e-01 -9.38111484e-01
3.05726081e-01 -5.40719569e-01 -1.47633874e+00 -2.69884944e-01
3.52670997e-01 6.93756565e-02 -4.59164739e-01 9.10716057e-01
4.30891782e-01 -2.19948933e-01 4.07402396e-01 -9.44306254e-01
-8.68851542e-01 1.68415844e+00 -1.87388211e-01 2.00253606e-01
6.16401374e-01 -2.08391353e-01 1.39181972e+00 -1.14400697e+00
1.33065712e+00 1.35404086e+00 4.71935660e-01 1.00516450e+00
-1.16864002e+00 -5.98115683e-01 -2.07697526e-01 -6.67627156e-02
-1.22500181e+00 -8.04168582e-01 1.74138844e-01 -1.04932472e-01
1.89279652e+00 -4.08576988e-02 2.60122567e-01 1.29676223e+00
5.20020574e-02 6.02943182e-01 1.01962817e+00 -7.15943933e-01
-3.38612437e-01 2.20160127e-01 -1.33372948e-01 5.91058969e-01
3.15287739e-01 2.33271614e-01 -8.73901129e-01 1.21241555e-01
3.41193885e-01 -1.10974920e+00 -5.00800386e-02 3.62427235e-01
-1.20184386e+00 6.09465897e-01 3.41387898e-01 5.79707995e-02
1.14473291e-02 8.22519660e-02 4.36334312e-01 7.22649813e-01
7.52351820e-01 5.29216886e-01 -1.00848281e+00 7.78141152e-03
-6.23264253e-01 4.83633488e-01 1.23770869e+00 1.68636692e+00
7.88619280e-01 -1.23173177e-01 -1.17408261e-01 1.13235426e+00
3.26292813e-01 6.75578475e-01 5.08993089e-01 -4.78326380e-01
1.62733054e+00 4.81228113e-01 1.17969394e-01 -3.03645171e-02
-4.51325297e-01 -9.61920917e-02 -4.35382396e-01 -5.30329108e-01
2.40666941e-01 -5.93082488e-01 -1.15882337e+00 1.74627781e+00
2.84404665e-01 -4.19715047e-01 1.00361705e+00 3.81499112e-01
1.47193599e+00 5.60506701e-01 4.11036909e-01 3.10484245e-02
1.28600371e+00 -1.14952910e+00 -7.57737398e-01 -7.30804741e-01
1.10111284e+00 -8.21273804e-01 9.18902755e-01 -2.94228226e-01
-1.24168301e+00 -6.70476258e-01 -8.12916815e-01 -6.10391259e-01
-6.68811083e-01 7.10833967e-02 6.28146052e-01 4.08392936e-01
-1.24039423e+00 5.27547479e-01 -4.88762885e-01 -7.77451038e-01
1.31171383e-02 3.03948313e-01 -7.45788276e-01 -1.46460697e-01
-1.86661863e+00 1.07560468e+00 1.09751117e+00 -9.93656516e-02
-7.32111812e-01 -4.63363409e-01 -1.13272524e+00 -4.26051289e-01
4.27391008e-02 -1.02501822e+00 1.49308562e+00 -3.91310722e-01
-1.38037121e+00 1.29573381e+00 -2.26945058e-01 -7.85290062e-01
5.18351495e-01 -3.82062495e-01 -8.16980064e-01 -1.77275240e-01
7.04653740e-01 1.01556695e+00 1.04827419e-01 -1.03225863e+00
-7.40769267e-01 -3.11908662e-01 9.79145616e-02 7.05664098e-01
1.76416740e-01 5.01696050e-01 -4.49472547e-01 -6.39698982e-01
5.01695229e-03 -7.16915190e-01 -3.42413515e-01 -9.85220551e-01
-9.24789429e-01 -4.99154598e-01 3.56592864e-01 -9.51194823e-01
8.49284053e-01 -1.41840672e+00 -1.89181920e-02 -1.34160325e-01
-2.54508644e-01 -8.78624097e-02 -3.21267217e-01 7.00975001e-01
-3.21957320e-01 4.36609209e-01 2.87463255e-02 -3.57251137e-01
-1.91824988e-01 7.17828214e-01 -5.40384471e-01 -1.69181332e-01
4.88278002e-01 1.31257629e+00 -1.01333535e+00 -6.17415547e-01
-2.59638578e-01 -1.54462889e-01 -4.86258984e-01 5.23016155e-01
-7.17841148e-01 7.63897240e-01 -5.03316522e-01 5.73196292e-01
3.24948847e-01 -5.67638427e-02 8.14881086e-01 -4.36961986e-02
8.71328115e-02 1.17581630e+00 -6.39544904e-01 1.96976280e+00
-7.09478140e-01 1.47294611e-01 -2.69173086e-01 -5.61196566e-01
9.41749096e-01 6.84605479e-01 -6.70246333e-02 -9.50802684e-01
-2.43665323e-01 8.89129341e-01 -3.28468800e-01 -7.05783188e-01
9.10351813e-01 -5.28331101e-01 -5.13438523e-01 3.04153115e-01
6.44247353e-01 -4.43799138e-01 7.01803505e-01 4.45002139e-01
8.89628649e-01 7.81343997e-01 3.33893239e-01 -4.46810246e-01
2.96796948e-01 4.50936973e-01 5.81964076e-01 6.60930753e-01
6.84848249e-01 5.00301361e-01 2.88801908e-01 -5.17309368e-01
-1.00729644e+00 -1.01473796e+00 -4.87354882e-02 1.18759716e+00
2.41925307e-02 -7.07444906e-01 -4.61166233e-01 -1.52235436e+00
-4.39657867e-02 1.09770596e+00 -3.87851447e-01 2.75140882e-01
-1.04498041e+00 -1.01254189e+00 9.29853857e-01 6.53519511e-01
3.19383293e-01 -1.28223467e+00 4.97446597e-01 5.51914573e-01
-7.17823207e-01 -1.91541958e+00 -2.01402143e-01 4.33989644e-01
-8.01996887e-01 -1.08140516e+00 -4.60499227e-02 -1.02934027e+00
7.05024362e-01 -4.06703562e-01 2.11571193e+00 -1.64135545e-01
1.87939912e-01 -2.47958541e-01 -4.91088718e-01 -2.01830074e-01
-1.17512727e+00 6.12245619e-01 -2.08096489e-01 -8.12800527e-01
6.39262438e-01 -3.48605871e-01 1.02812499e-01 1.85908005e-01
-5.07074594e-01 4.60405916e-01 2.20519915e-01 8.74140143e-01
4.56112295e-01 -4.35621619e-01 6.24433577e-01 -1.66854787e+00
5.39485693e-01 -5.99076688e-01 -4.87603486e-01 6.65646493e-01
-3.54230851e-01 4.19114769e-01 8.31083953e-01 4.42352027e-01
-1.47151697e+00 -8.98326561e-02 -5.34181058e-01 5.04162014e-01
2.24629976e-02 6.33616090e-01 -4.28730637e-01 4.63506788e-01
1.10919487e+00 -1.67036176e-01 -6.59290850e-01 -8.24568450e-01
1.00643575e+00 8.73059750e-01 7.15373695e-01 -1.35373175e+00
7.44867027e-01 2.87352372e-02 -4.38522190e-01 -4.73265320e-01
-1.25114214e+00 -1.09277233e-01 -5.93045592e-01 3.44638199e-01
8.84123087e-01 -1.60511911e+00 5.73397651e-02 5.83523251e-02
-1.45074594e+00 -6.59824312e-01 -7.18120039e-01 2.62883276e-01
-5.37068367e-01 -6.96082190e-02 -1.01485193e+00 -2.70923916e-02
-6.46580875e-01 -7.84262657e-01 1.26559877e+00 8.19583684e-02
-1.37495548e-01 -1.35587561e+00 4.14699137e-01 4.66838449e-01
-1.05965003e-01 -1.62869588e-01 1.13342416e+00 -9.99250352e-01
-3.15525800e-01 1.39465541e-01 -3.44530791e-01 2.29940191e-01
2.73650475e-02 -3.52767736e-01 -6.21293366e-01 -1.91746280e-01
-5.44665754e-01 -1.10333645e+00 6.80639029e-01 -3.01050097e-01
5.29208899e-01 -3.88016105e-01 -3.46325964e-01 6.86009347e-01
1.42089975e+00 -1.85496807e-01 6.93984568e-01 2.87123680e-01
8.43066454e-01 9.13492322e-01 5.10017633e-01 -1.99545234e-01
1.02042449e+00 6.74641788e-01 -8.13095570e-02 -3.24449062e-01
-6.46948159e-01 -9.83585656e-01 3.28785062e-01 1.12897134e+00
-1.65517572e-02 -5.27095854e-01 -1.21935451e+00 7.86082149e-01
-1.88575506e+00 -6.49313748e-01 -4.07615483e-01 1.80314648e+00
1.63050711e+00 1.36284739e-01 -2.39010125e-01 -8.75211775e-01
5.20830154e-01 8.59811977e-02 -1.04544137e-03 -2.85456240e-01
-7.27666318e-01 7.05382347e-01 6.84096634e-01 1.01251519e+00
-8.77794027e-01 1.90213633e+00 7.08802509e+00 5.35125494e-01
-6.26958191e-01 3.52016121e-01 6.09691858e-01 4.12593395e-01
-7.51513541e-01 5.48142552e-01 -1.49239898e+00 9.31458734e-03
1.19147158e+00 -4.62424815e-01 4.16620374e-01 5.13886750e-01
-2.74979696e-02 3.40783983e-01 -1.15493321e+00 3.74169171e-01
-4.50429171e-02 -1.42501080e+00 4.74066645e-01 -3.42829168e-01
1.00671995e+00 8.23169827e-01 -6.10194862e-01 7.75766075e-01
1.36031771e+00 -1.13980401e+00 7.40600288e-01 2.97225025e-02
1.35069823e+00 -4.00468320e-01 7.94787526e-01 1.86309710e-01
-1.02479815e+00 3.87780696e-01 -3.94621849e-01 3.97728533e-01
4.73448366e-01 4.43507195e-01 -1.06505477e+00 1.19497907e+00
4.81010854e-01 8.84738207e-01 -5.05821645e-01 1.62036881e-01
-1.11323154e+00 3.58654350e-01 -3.62611473e-01 3.76865983e-01
1.74778923e-02 -5.03454134e-02 3.89515311e-01 1.40976787e+00
3.08707654e-01 -1.98268786e-01 4.58270818e-01 5.28157651e-01
-8.59577358e-01 3.90362859e-01 -8.58465493e-01 -1.45372510e-01
4.24512535e-01 1.04807854e+00 -1.60792291e-01 -6.96907520e-01
-6.50238395e-01 8.83507967e-01 8.56673360e-01 6.29805624e-01
-4.15107131e-01 -3.85841787e-01 1.25295937e-01 1.35531038e-01
-1.44383267e-01 -8.24573115e-02 -1.23151444e-01 -1.71614623e+00
7.04609081e-02 -1.18084323e+00 7.73793519e-01 -6.30895495e-01
-1.42509139e+00 1.18730032e+00 4.63992916e-02 -7.40772903e-01
-1.10222745e+00 -7.09959269e-01 -1.62562415e-01 1.28394067e+00
-1.67940938e+00 -1.63271594e+00 1.88343540e-01 5.24438918e-01
6.38428569e-01 -4.52322543e-01 1.06903696e+00 4.53890890e-01
-2.00643554e-01 4.61688250e-01 -1.91857189e-01 5.88035822e-01
9.08605397e-01 -1.66147184e+00 1.35974574e+00 8.76215994e-01
4.65937227e-01 5.63147366e-01 5.95834017e-01 -1.21862411e+00
-7.89775014e-01 -1.36162782e+00 1.97370481e+00 -7.38067091e-01
9.69846189e-01 -7.53463030e-01 -4.19853419e-01 1.23941708e+00
6.48513794e-01 1.39542803e-01 5.66600978e-01 6.50910497e-01
-4.33709949e-01 1.77512482e-01 -8.47902775e-01 4.52312201e-01
1.44159520e+00 -7.06803203e-01 -9.34839368e-01 8.10858130e-01
7.74162173e-01 -1.07761931e+00 -1.00762844e+00 5.74975014e-01
2.67021991e-02 -4.76066589e-01 6.57470405e-01 -1.38526618e+00
7.11467564e-01 -2.49226227e-01 -4.10410076e-01 -1.53421485e+00
3.65393519e-01 -9.01904225e-01 6.50602218e-04 1.19574344e+00
1.38876641e+00 -5.62729239e-01 7.52186179e-01 3.38755965e-01
-2.67883539e-01 -1.88589469e-01 -9.96219754e-01 -8.82258117e-01
4.88535881e-01 -3.48679364e-01 8.51122379e-01 9.13937151e-01
1.83247641e-01 1.31727946e+00 -2.83058047e-01 -1.30223885e-01
5.28989077e-01 1.25521243e-01 9.08094466e-01 -8.03831041e-01
-4.64630067e-01 3.43343556e-01 2.32121244e-01 -1.26252031e+00
4.38156247e-01 -1.51284862e+00 1.60861999e-01 -1.81840134e+00
5.27925044e-02 -6.41863048e-01 3.16524953e-01 7.51250923e-01
-3.55919063e-01 2.39374563e-01 -1.98570326e-01 9.81781557e-02
-5.39546967e-01 7.32653141e-01 1.04213786e+00 1.04536042e-01
-5.57842515e-02 -2.55190164e-01 -8.76924574e-01 4.04262722e-01
5.22168458e-01 -8.37830842e-01 -3.79528880e-01 -1.05554736e+00
6.60404027e-01 3.15941870e-01 -1.90255016e-01 -5.12685120e-01
-1.08162731e-01 -3.72562818e-02 1.74797595e-01 -4.03343588e-01
-5.15811220e-02 -2.60042727e-01 9.81511623e-02 3.64876054e-02
-4.19731081e-01 4.20362502e-01 1.25360101e-01 1.58983663e-01
-4.92658287e-01 -1.73168421e-01 4.10490394e-01 -4.33363736e-01
-5.66030860e-01 1.62033707e-01 2.79995203e-01 1.04751396e+00
3.87447357e-01 4.38789725e-01 -1.04668307e+00 -3.80479574e-01
-9.72978413e-01 5.30532777e-01 2.78379321e-01 6.99038863e-01
1.99426740e-01 -1.33680272e+00 -1.33434474e+00 2.98980862e-01
4.42572981e-01 1.01547204e-01 -2.79371679e-01 2.05043599e-01
-7.90853500e-01 6.24562562e-01 1.08196653e-01 -6.22678623e-02
-7.19427168e-01 4.59913909e-02 4.82906222e-01 -8.56820643e-01
-4.44010079e-01 7.74267495e-01 -2.18603775e-01 -1.24902678e+00
-3.39652956e-01 -4.19235975e-02 -2.79182851e-01 -2.40136787e-01
3.04161876e-01 -1.77294865e-01 2.61050552e-01 -7.55503118e-01
-2.09154770e-01 1.92755535e-01 -3.87511224e-01 -4.54960883e-01
1.28434896e+00 -1.13756090e-01 -4.02207196e-01 1.55799747e-01
8.89769137e-01 3.21401834e-01 -6.12990379e-01 -5.33636689e-01
4.91231978e-01 -7.29987323e-02 -5.58272481e-01 -1.17130423e+00
-7.07542956e-01 4.95139420e-01 -2.39881739e-01 6.28505796e-02
4.93287742e-01 5.65507293e-01 9.38525975e-01 6.53140008e-01
7.20595777e-01 -1.37082946e+00 -3.75738591e-01 1.13952339e+00
8.62071037e-01 -1.27984071e+00 -1.71836674e-01 -7.37595737e-01
-7.20882952e-01 9.63108301e-01 8.36700439e-01 2.57813960e-01
3.91357094e-01 4.50083643e-01 5.43517768e-01 -3.06983888e-01
-9.97846782e-01 -3.85651469e-01 2.17334569e-01 7.73553610e-01
9.90986526e-01 1.77580416e-01 -3.44339073e-01 4.23967212e-01
-7.30475485e-01 -4.10884842e-02 2.21986532e-01 7.74449110e-01
-3.09175346e-02 -1.99726999e+00 3.15141797e-01 2.89138108e-01
-6.76686406e-01 -8.43728483e-01 -4.42711622e-01 1.01773119e+00
-3.18803601e-02 1.09147286e+00 -1.99674100e-01 -2.23889098e-01
5.38490534e-01 1.76261723e-01 6.07389987e-01 -1.38907874e+00
-7.78374255e-01 -2.99147516e-01 1.46916938e+00 -4.55572188e-01
-3.42937857e-01 -3.26040834e-01 -1.50312877e+00 -1.40667604e-02
-5.90620898e-02 7.00304091e-01 5.94408512e-01 1.05796742e+00
3.00985873e-01 2.61659920e-01 4.57660854e-01 4.27124314e-02
-4.49784726e-01 -1.19980764e+00 -2.54295409e-01 4.78990376e-01
-3.16241056e-01 2.36144923e-02 1.02947123e-01 1.32788315e-01] | [10.870941162109375, 9.569014549255371] |
5774911b-650c-49e0-96d8-65ff19890e5e | latentslam-unsupervised-multi-sensor | 2105.03265 | null | https://arxiv.org/abs/2105.03265v1 | https://arxiv.org/pdf/2105.03265v1.pdf | LatentSLAM: unsupervised multi-sensor representation learning for localization and mapping | Biologically inspired algorithms for simultaneous localization and mapping (SLAM) such as RatSLAM have been shown to yield effective and robust robot navigation in both indoor and outdoor environments. One drawback however is the sensitivity to perceptual aliasing due to the template matching of low-dimensional sensory templates. In this paper, we propose an unsupervised representation learning method that yields low-dimensional latent state descriptors that can be used for RatSLAM. Our method is sensor agnostic and can be applied to any sensor modality, as we illustrate for camera images, radar range-doppler maps and lidar scans. We also show how combining multiple sensors can increase the robustness, by reducing the number of false matches. We evaluate on a dataset captured with a mobile robot navigating in a warehouse-like environment, moving through different aisles with similar appearance, making it hard for the SLAM algorithms to disambiguate locations. | ['Jan Steckel', 'Bart Dhoedt', 'Tim Verbelen', 'Wouter Jansen', 'Ozan Çatal'] | 2021-05-07 | null | null | null | null | ['template-matching'] | ['computer-vision'] | [ 3.79457742e-01 -2.34874383e-01 2.87596196e-01 -3.71447861e-01
-4.31129187e-01 -7.51924336e-01 6.19830310e-01 2.81353027e-01
-7.59194970e-01 8.49797845e-01 -1.68042198e-01 8.61289427e-02
-5.60475409e-01 -7.94469059e-01 -6.82830036e-01 -7.01397538e-01
-3.14818859e-01 6.75114572e-01 4.57541734e-01 -2.75997758e-01
4.70684618e-01 9.54752505e-01 -1.94046891e+00 -3.16712946e-01
6.38388991e-01 5.28317273e-01 7.94424593e-01 5.07232249e-01
2.62374934e-02 3.43331248e-01 -4.05409902e-01 3.69552404e-01
2.38959357e-01 -2.38490880e-01 -3.17329854e-01 -1.12169646e-01
4.10190970e-01 7.41382837e-02 -3.04357052e-01 9.25370812e-01
4.79812413e-01 3.43906790e-01 9.04534757e-01 -1.21057618e+00
-1.36432752e-01 1.97774217e-01 -1.08557455e-01 -1.29063517e-01
5.62403679e-01 -1.28972039e-01 4.08726603e-01 -6.74880207e-01
6.58594847e-01 1.32472038e+00 8.23403180e-01 2.20851362e-01
-1.48170626e+00 -4.93251503e-01 -6.74642548e-02 1.50440708e-01
-1.54574764e+00 -5.82668662e-01 5.87364137e-01 -4.01156932e-01
9.24202979e-01 -1.15420058e-01 3.47708225e-01 1.08321559e+00
7.08571732e-01 3.01353663e-01 1.24489987e+00 -3.24128449e-01
7.15871632e-01 -8.95401184e-03 -2.12392330e-01 8.28261971e-01
7.10270882e-01 3.97639722e-01 -6.82417870e-01 7.97105730e-02
8.96693289e-01 3.52568738e-02 -4.78903055e-02 -1.31485510e+00
-1.35816514e+00 8.22574258e-01 7.97451198e-01 2.81849921e-01
-4.84483033e-01 3.07987392e-01 -4.76180948e-02 1.27776012e-01
-3.30090970e-01 7.44170964e-01 -1.94967091e-01 -7.02856705e-02
-7.03409016e-01 1.27969533e-01 7.43292272e-01 1.07789910e+00
1.07268524e+00 1.16894968e-01 7.01461852e-01 5.75990319e-01
3.22369844e-01 9.02986169e-01 6.11860693e-01 -1.13144219e+00
6.73528761e-02 3.81799698e-01 4.47076708e-01 -1.36208057e+00
-8.49066496e-01 -1.26887724e-01 -5.57909012e-01 5.02825201e-01
3.12331676e-01 3.34045924e-02 -1.01369333e+00 1.72638786e+00
-1.16857797e-01 -1.44764278e-02 4.77846265e-01 9.24237609e-01
3.90427232e-01 3.94894987e-01 -1.39188677e-01 7.01738074e-02
1.12865400e+00 -5.20493567e-01 -5.30822754e-01 -9.27157044e-01
5.55844724e-01 -4.78931069e-01 6.67609811e-01 2.95422554e-01
-3.64079207e-01 -4.22126412e-01 -1.31748939e+00 1.34130597e-01
-8.68313849e-01 -1.43250808e-01 7.14792967e-01 4.12053883e-01
-9.69104052e-01 3.34442973e-01 -1.04957736e+00 -1.18194914e+00
-8.40505138e-02 4.58954901e-01 -8.26654017e-01 -1.48550138e-01
-8.60410869e-01 1.37441528e+00 5.14333963e-01 1.40721421e-03
-7.72164762e-01 1.18664652e-01 -1.18836296e+00 -2.51726896e-01
1.00448392e-01 -2.92203188e-01 7.24960685e-01 -2.47651815e-01
-1.40055549e+00 6.55574143e-01 -3.25886071e-01 -7.36671448e-01
2.19052970e-01 -2.10478291e-01 -2.35261634e-01 -9.79842022e-02
3.67014974e-01 8.05979490e-01 5.54585397e-01 -1.46815789e+00
-4.64903742e-01 -4.83343989e-01 -2.12090716e-01 4.14203554e-01
3.00019294e-01 -6.19289756e-01 7.75347501e-02 -1.15003034e-01
9.83253658e-01 -1.10394871e+00 -5.75626552e-01 -6.12073913e-02
8.55753645e-02 4.97056365e-01 6.99682772e-01 -1.11215681e-01
2.50199437e-01 -2.07783532e+00 2.64791965e-01 2.94081151e-01
-4.93123055e-01 -3.49289387e-01 -2.95951396e-01 7.20666885e-01
4.58797067e-01 -4.04761702e-01 -3.47826540e-01 -7.04887658e-02
5.97290136e-02 8.61162186e-01 -3.82243931e-01 7.72229731e-01
-1.03635795e-01 6.68879867e-01 -1.10349047e+00 -2.84693033e-01
5.37867248e-01 3.01788718e-01 -1.12702101e-01 -1.60580531e-01
2.18525976e-01 5.22491336e-01 -1.83478057e-01 5.90921998e-01
5.76907516e-01 4.12952602e-01 1.25555515e-01 9.86254960e-02
-4.72440332e-01 1.96632802e-01 -1.33702421e+00 2.27151656e+00
-6.66080296e-01 8.51569891e-01 1.78457156e-01 -7.95002162e-01
1.37761676e+00 -2.12821096e-01 2.56205201e-01 -9.26726103e-01
-2.08598748e-02 3.47849160e-01 -1.79566011e-01 -1.58218279e-01
7.63399303e-01 -1.34893671e-01 -6.78287625e-01 2.70625919e-01
1.88841641e-01 -5.93657672e-01 -1.92392990e-02 -7.70834535e-02
1.19424868e+00 2.38914832e-01 4.75714207e-01 -4.41157073e-01
3.26374263e-01 2.81711668e-01 4.14702058e-01 1.02679622e+00
-7.77790621e-02 3.36095095e-01 -2.11315736e-01 -3.23841095e-01
-7.13505030e-01 -1.61430848e+00 -2.20344588e-01 7.16630518e-01
9.12720084e-01 -1.65986240e-01 -1.22890688e-01 -4.09904979e-02
2.77915835e-01 7.91830778e-01 -3.24079633e-01 -3.17003042e-01
-4.66637731e-01 -4.43988234e-01 4.44284320e-01 3.96411449e-01
3.56833220e-01 -9.08199489e-01 -1.56657970e+00 4.87214744e-01
-5.90622285e-03 -1.12818372e+00 4.83774662e-01 8.85034740e-01
-9.33442175e-01 -8.93355429e-01 -2.24702135e-01 -8.06777239e-01
7.46179581e-01 4.34774488e-01 6.50525153e-01 -5.68400741e-01
-5.06527603e-01 6.26046956e-01 -2.70982713e-01 -3.92763525e-01
-1.20402813e-01 -1.78488746e-01 7.21608222e-01 -3.46070677e-01
3.24518710e-01 -7.28059947e-01 -1.91929206e-01 6.70512199e-01
-3.61531943e-01 -2.62975603e-01 7.49863446e-01 6.83590293e-01
6.69472456e-01 1.07789867e-01 2.72420555e-01 -3.97162139e-01
4.01083499e-01 -2.22776026e-01 -8.57248127e-01 -6.48084432e-02
-4.61352050e-01 3.53673488e-01 3.01271915e-01 -3.64001095e-01
-5.95791996e-01 7.06683159e-01 4.17371660e-01 1.24408267e-01
-6.18179560e-01 4.81303990e-01 -6.12994544e-02 -5.17833769e-01
8.30853879e-01 5.12307405e-01 -1.01618007e-01 -3.14546287e-01
4.97331589e-01 4.50309247e-01 9.29091871e-01 -3.69074225e-01
1.01066017e+00 5.67541182e-01 4.43599433e-01 -1.18022299e+00
-2.34267171e-02 -5.23903787e-01 -1.06686199e+00 -1.14853054e-01
6.54960096e-01 -9.11260605e-01 -3.98567468e-01 1.39061972e-01
-8.91068041e-01 -1.62102938e-01 -1.16630964e-01 8.07723284e-01
-1.09154880e+00 1.86457500e-01 -1.59436211e-01 -1.02029490e+00
4.39801127e-01 -9.86896515e-01 9.26098228e-01 3.07318926e-01
-3.77560616e-01 -6.95295513e-01 3.27510148e-01 -2.27503255e-01
4.30111527e-01 3.56996387e-01 5.71294665e-01 -4.77303684e-01
-6.98669910e-01 -2.34260276e-01 2.44520530e-02 -4.43213105e-01
3.42511803e-01 -4.82703686e-01 -9.28693712e-01 -5.00733435e-01
-7.57676037e-03 -2.48796836e-01 8.18637311e-01 1.53709948e-01
1.54345140e-01 1.50215194e-01 -6.48435533e-01 4.95171636e-01
1.54102707e+00 5.03346264e-01 5.69040656e-01 6.53055847e-01
2.89474875e-01 8.48116457e-01 9.73789155e-01 1.04466982e-01
1.69754446e-01 6.05770886e-01 6.50224030e-01 1.54359385e-01
2.65591025e-01 -3.84615630e-01 4.03231263e-01 4.72341269e-01
2.81580925e-01 -1.05474561e-01 -1.05384135e+00 6.65155351e-01
-2.07473063e+00 -7.43713200e-01 5.87359890e-02 2.33473277e+00
-1.23967074e-01 4.39609922e-02 -4.23485935e-01 1.01849169e-03
5.27773023e-01 2.39867382e-02 -5.59319735e-01 -3.26263905e-01
-2.37277359e-01 5.08863926e-02 1.01759386e+00 6.92807615e-01
-1.10197747e+00 1.08144844e+00 6.67239237e+00 4.03493755e-02
-9.18386638e-01 -1.05514735e-01 -5.24613619e-01 1.92208067e-01
1.78660661e-01 1.36602402e-01 -5.03813207e-01 1.72951683e-01
9.77537215e-01 1.20898776e-01 3.73747230e-01 8.46444070e-01
-3.43137048e-02 -8.14215064e-01 -1.02627337e+00 1.31082904e+00
1.79662332e-01 -9.24385309e-01 -2.78037757e-01 1.24503583e-01
3.63393426e-01 2.58936226e-01 -1.55912161e-01 1.66034594e-01
3.17436010e-01 -9.97734427e-01 7.03359008e-01 6.11053109e-01
3.55484694e-01 -7.33404100e-01 7.44901657e-01 5.34304440e-01
-1.15534508e+00 -2.92543322e-01 -7.63836443e-01 -3.11473727e-01
2.37005845e-01 1.14094421e-01 -1.12476909e+00 3.66249770e-01
6.25910401e-01 5.02834558e-01 -5.29765368e-01 1.27431881e+00
-2.81310618e-01 -2.31884614e-01 -7.85197794e-01 -2.73017079e-01
2.15339646e-01 -3.58917803e-01 6.02535427e-01 1.05793750e+00
6.53216004e-01 -2.85864651e-01 1.98173821e-01 7.06067085e-01
6.02644920e-01 -2.84001142e-01 -1.26789510e+00 2.02205002e-01
6.92982197e-01 9.42449212e-01 -1.04856026e+00 1.85697511e-01
1.18583374e-01 1.15272784e+00 2.37266794e-02 3.43312740e-01
-4.87735748e-01 -5.97423732e-01 5.97635806e-01 -1.62062973e-01
3.51031184e-01 -9.59550858e-01 -5.41361922e-04 -8.35757315e-01
-3.44134301e-01 -2.08161741e-01 -1.02658160e-01 -1.01480007e+00
-9.15281475e-01 3.90349567e-01 1.19058184e-01 -1.29803193e+00
-5.38912773e-01 -7.40052462e-01 -7.03322813e-02 7.65225470e-01
-1.35535634e+00 -9.71009254e-01 -4.74895447e-01 5.06787062e-01
3.55435669e-01 -1.59064472e-01 1.16709363e+00 -2.10606709e-01
3.28572914e-02 -7.85822943e-02 3.28442693e-01 -1.40559778e-01
7.64658213e-01 -1.19025362e+00 3.18498343e-01 9.51876462e-01
4.71996576e-01 7.42998481e-01 1.06177986e+00 -6.62777543e-01
-1.70969152e+00 -7.98629045e-01 6.91003084e-01 -4.59566623e-01
5.48174202e-01 -6.86648667e-01 -5.82428932e-01 6.63994074e-01
-3.19353372e-01 -2.53956765e-01 3.77279758e-01 1.76085010e-02
-1.77139565e-01 -8.26715305e-02 -1.27872002e+00 5.93209088e-01
1.11106181e+00 -6.70632005e-01 -8.13132048e-01 9.98801962e-02
1.34162098e-01 -2.92629987e-01 -4.67412829e-01 4.09553409e-01
6.59210563e-01 -9.34220254e-01 9.48190272e-01 -9.91739109e-02
-5.44314146e-01 -7.63676703e-01 -5.58689058e-01 -1.33747435e+00
-4.84009176e-01 -2.22527504e-01 3.75583261e-01 9.95676935e-01
1.00339651e-01 -9.48595643e-01 7.06111073e-01 1.28250465e-01
1.65902376e-01 7.00057223e-02 -1.35409153e+00 -1.18326581e+00
-4.66844767e-01 -3.08798015e-01 2.65128970e-01 5.02935469e-01
-4.87863384e-02 3.92678946e-01 -2.45120481e-01 6.83411777e-01
7.83094406e-01 2.54264235e-01 1.01712728e+00 -1.70543981e+00
8.95362720e-02 -6.25492781e-02 -1.08212209e+00 -1.00705993e+00
7.65022784e-02 -6.47311926e-01 8.04865599e-01 -1.63298869e+00
-3.28113079e-01 -6.04891896e-01 -2.33947024e-01 3.67613018e-01
8.18684399e-01 2.75630862e-01 1.02143914e-01 2.95072585e-01
-4.88802224e-01 4.00311977e-01 4.90450174e-01 3.51282209e-03
-1.88232377e-01 -2.61293143e-01 -1.86972871e-01 8.15968275e-01
7.44104743e-01 -6.81738913e-01 -4.18279737e-01 -2.79977351e-01
1.20324180e-01 -1.41063267e-02 4.18398172e-01 -1.65853119e+00
5.37937522e-01 -8.90327245e-02 5.20426929e-01 -8.19983065e-01
8.96302998e-01 -1.13917470e+00 2.51488209e-01 8.40618670e-01
-7.38315657e-02 2.40660444e-01 3.16437066e-01 9.07404006e-01
-5.46959415e-02 -3.85054231e-01 7.03942835e-01 -2.33688921e-01
-1.46242177e+00 -2.87221253e-01 -1.00158656e+00 -6.21126533e-01
9.47099686e-01 -4.94421780e-01 -1.93803921e-01 -5.01154959e-01
-4.65882808e-01 1.62850574e-01 1.08072376e+00 5.17346799e-01
9.14027810e-01 -1.13705337e+00 -1.95132717e-01 5.28057337e-01
4.50448632e-01 -2.07888931e-01 -1.36852875e-01 6.15112543e-01
-8.79062712e-01 6.07847750e-01 -7.77260602e-01 -9.15845156e-01
-9.74526763e-01 6.74008846e-01 2.19018832e-01 3.71795833e-01
-4.01436120e-01 5.61221182e-01 -7.04232976e-02 -7.85961330e-01
1.49569571e-01 -4.62161839e-01 -6.74569905e-02 -3.48253809e-02
1.52930051e-01 1.84422493e-01 -2.27936223e-01 -1.00911796e+00
-8.71405542e-01 9.97681677e-01 4.68162090e-01 -4.08323020e-01
1.17375040e+00 -4.90820706e-01 2.45479122e-02 8.79775047e-01
6.91831410e-01 5.04029496e-03 -1.11720669e+00 4.31837663e-02
4.14575666e-01 -2.87354231e-01 -1.23674758e-01 -6.29413843e-01
-2.52373368e-01 7.98914969e-01 1.11924791e+00 -8.79507959e-02
7.56342292e-01 -1.35738596e-01 3.04098785e-01 1.11083841e+00
1.34130144e+00 -8.41849923e-01 -2.03570262e-01 8.24011087e-01
6.21980786e-01 -1.14078891e+00 1.73809025e-02 2.42257994e-02
-3.52767557e-01 1.10830164e+00 3.06709260e-01 -3.73111457e-01
1.49221122e-01 3.40819091e-01 2.31935844e-01 2.07533240e-02
-3.01618695e-01 -4.06340212e-01 -1.21431425e-01 1.19552529e+00
-1.12486623e-01 4.28271033e-02 -1.30370641e-02 -6.61324114e-02
-3.52986544e-01 -3.91550839e-01 6.09858572e-01 1.37584269e+00
-1.00051260e+00 -9.58772302e-01 -6.32333398e-01 -1.33891821e-01
5.03169477e-01 2.48338610e-01 -5.16995728e-01 1.03467166e+00
1.42600968e-01 9.46037471e-01 2.99516559e-01 -4.77344692e-01
3.79567236e-01 6.48084879e-02 7.96680450e-01 -5.05856156e-01
1.38071924e-01 -1.30138859e-01 -5.78808226e-02 -7.48147786e-01
-4.99172211e-01 -7.29966223e-01 -1.55179334e+00 2.70474821e-01
-8.73869210e-02 2.85033464e-01 1.12277687e+00 7.63958991e-01
3.19880992e-01 1.02264732e-01 1.96680367e-01 -1.25785029e+00
-1.18160628e-01 -6.82799578e-01 -7.94485092e-01 7.99026936e-02
4.29997563e-01 -1.25325406e+00 -3.41163516e-01 -2.36227140e-01] | [7.317629814147949, -2.003607988357544] |
7161e5f6-5d35-47a2-b238-db2fa3c377ca | automl-systems-for-medical-imaging | 2306.04750 | null | https://arxiv.org/abs/2306.04750v2 | https://arxiv.org/pdf/2306.04750v2.pdf | AutoML Systems For Medical Imaging | The integration of machine learning in medical image analysis can greatly enhance the quality of healthcare provided by physicians. The combination of human expertise and computerized systems can result in improved diagnostic accuracy. An automated machine learning approach simplifies the creation of custom image recognition models by utilizing neural architecture search and transfer learning techniques. Medical imaging techniques are used to non-invasively create images of internal organs and body parts for diagnostic and procedural purposes. This article aims to highlight the potential applications, strategies, and techniques of AutoML in medical imaging through theoretical and empirical evidence. | ['Dr. Md Azim Ullah', 'Mofazzal Hossain', 'Sajedul Talukder', 'Md Jahangir Alam', 'Ismail Hossain', 'MD Abdullah Al Nasim', 'Angona Biswas', 'Tasmia Tahmida Jidney'] | 2023-06-07 | null | null | null | null | ['automl', 'architecture-search'] | ['methodology', 'methodology'] | [ 4.62068588e-01 1.95805013e-01 -3.03952008e-01 -4.45016116e-01
-5.29949903e-01 -3.62981170e-01 1.75255522e-01 1.03665449e-01
-4.07779008e-01 3.88219982e-01 -2.91265965e-01 -7.56559670e-01
-3.72817189e-01 -5.19924998e-01 -1.32081389e-01 -5.24944663e-01
-1.10427290e-01 8.10665071e-01 -3.14023316e-01 3.20650429e-01
1.25478029e-01 9.75459218e-01 -9.67896461e-01 5.37972033e-01
6.96956694e-01 1.05108786e+00 3.53955388e-01 8.43918502e-01
-7.17741773e-02 1.13916552e+00 -3.61925453e-01 -1.87056065e-02
2.82540947e-01 -5.85602701e-01 -6.62482858e-01 3.41580033e-01
6.07562773e-02 -4.10558164e-01 -8.70537013e-02 7.14285851e-01
7.23331988e-01 -1.25682056e-01 9.62402403e-01 -4.08596247e-01
-5.36244333e-01 1.42022178e-01 -1.12015992e-01 4.14638519e-01
2.34426036e-01 2.50484020e-01 4.30410616e-02 -6.64433599e-01
4.51387376e-01 7.26810157e-01 8.76382172e-01 6.09301686e-01
-9.75293934e-01 -3.64122748e-01 -7.86472619e-01 1.29817963e-01
-1.14084864e+00 -2.85586327e-01 7.06692755e-01 -7.18611956e-01
1.00702727e+00 3.97714674e-01 8.01588058e-01 3.25581223e-01
6.88566148e-01 4.58277583e-01 1.00329864e+00 -8.70964587e-01
-1.34800992e-03 5.70513546e-01 -1.03325062e-01 1.12350059e+00
2.80520022e-01 4.10868436e-01 -9.12327692e-02 -2.28269279e-01
1.23766506e+00 1.46026120e-01 9.68238562e-02 -3.60375077e-01
-9.01170135e-01 9.42764759e-01 3.71151805e-01 8.34147453e-01
-7.40795195e-01 -4.54339907e-02 3.32915217e-01 7.29291737e-02
-5.32448664e-02 1.02969122e+00 -4.40830916e-01 1.26549751e-01
-7.82636404e-01 -4.92409408e-01 5.68866014e-01 1.58430919e-01
2.27224752e-01 2.20106661e-01 1.27783433e-01 7.09211409e-01
4.24464583e-01 4.41039801e-01 1.11964631e+00 -1.25635457e+00
-3.24358732e-01 6.13810956e-01 -7.59089589e-02 -9.40777421e-01
-6.68334901e-01 -3.49052787e-01 -7.43526340e-01 4.52146053e-01
-7.95019120e-02 -1.74147576e-01 -1.15008545e+00 6.73016965e-01
2.65509427e-01 -3.99345785e-01 8.29062313e-02 6.82818472e-01
9.38450217e-01 2.91869640e-01 2.54579514e-01 -1.96531028e-01
1.19619060e+00 -8.67337525e-01 -7.23166227e-01 -5.96106574e-02
7.66510308e-01 -5.40924311e-01 5.61337113e-01 5.00311852e-01
-1.30975270e+00 -4.95794356e-01 -8.69800091e-01 2.09199667e-01
-1.93402335e-01 3.84744346e-01 9.01052475e-01 9.01755929e-01
-7.04488218e-01 6.58701897e-01 -1.21309316e+00 -9.37096626e-02
5.92695594e-01 6.77757919e-01 -3.34235758e-01 2.06754491e-01
-6.13049150e-01 1.46308076e+00 4.34725732e-01 1.36490371e-02
-5.54334641e-01 -7.39756584e-01 -9.04743075e-01 -2.32083932e-01
-6.41734377e-02 -1.12752247e+00 1.36432827e+00 -1.12335157e+00
-1.42499149e+00 1.29719841e+00 1.30636439e-01 -5.55277824e-01
3.39535028e-01 6.67961463e-02 -3.56642842e-01 6.21365607e-01
-1.61387697e-01 6.52369559e-01 6.42597318e-01 -9.50809419e-01
-4.79566067e-01 -3.44913125e-01 -5.93928814e-01 1.12086959e-01
-1.62094042e-01 5.31127900e-02 -1.47504181e-01 -4.99119848e-01
1.55791759e-01 -8.09832811e-01 -6.43360734e-01 2.91528612e-01
1.94182858e-01 2.81692207e-01 5.20118296e-01 -8.43662083e-01
7.78379023e-01 -2.00258756e+00 -1.59170344e-01 4.74070281e-01
2.71157533e-01 6.21712267e-01 2.77755767e-01 4.33004787e-03
-1.81221038e-01 7.86685273e-02 1.07837416e-01 3.84026825e-01
-6.35593235e-01 2.95524627e-01 5.38386047e-01 3.19205433e-01
-1.83870047e-01 1.32452345e+00 -5.59265077e-01 -8.65025520e-01
9.67695892e-01 4.61818933e-01 -2.77464360e-01 3.14411610e-01
2.43383214e-01 8.59522283e-01 -5.46303391e-01 8.60810995e-01
-9.11982879e-02 -8.95679414e-01 3.69553208e-01 -3.46016765e-01
1.74441949e-01 -1.33448886e-02 -5.20362854e-01 1.58744597e+00
-4.68663067e-01 4.15305138e-01 3.54471779e-03 -1.02478981e+00
7.49133050e-01 6.67379141e-01 9.15112913e-01 -7.05520630e-01
5.42472482e-01 1.02547854e-01 3.60205054e-01 -1.17037427e+00
-4.61443067e-01 -4.93480086e-01 5.25346994e-01 3.64871532e-01
4.94615212e-02 -2.06743255e-01 -3.50322753e-01 -4.67687190e-01
8.28726828e-01 -1.85226053e-01 8.16901803e-01 -1.47532642e-01
4.66729611e-01 3.43289793e-01 3.82541344e-02 7.69305944e-01
-3.48087668e-01 3.21501225e-01 -3.58114600e-01 -1.06613362e+00
-1.02448642e+00 -9.47045028e-01 -2.23437250e-01 5.83321571e-01
-3.15713525e-01 5.19572556e-01 -5.45189261e-01 -4.70619351e-01
-6.35806546e-02 4.83704388e-01 -6.48303092e-01 -1.64707422e-01
-4.14835274e-01 -5.63257337e-01 3.12094152e-01 5.70186198e-01
5.66810481e-02 -1.28202605e+00 -1.36794138e+00 2.79848516e-01
8.04107860e-02 -8.40749145e-01 2.22345106e-02 1.64834112e-02
-1.39213395e+00 -1.09145570e+00 -7.86016107e-01 -1.00912857e+00
8.22795749e-01 -5.93314990e-02 8.82380188e-01 3.51950824e-01
-1.30571091e+00 6.70800328e-01 -1.20213278e-01 -7.09559381e-01
-7.87056744e-01 -1.43068686e-01 -3.16717118e-01 -3.80392641e-01
1.18115135e-01 -4.21054631e-01 -7.25714147e-01 9.86210182e-02
-6.45030081e-01 2.11912483e-01 1.24986243e+00 1.04990649e+00
6.68440819e-01 -1.81849226e-01 3.95247072e-01 -1.00649011e+00
6.12689078e-01 -2.03325853e-01 -3.56216848e-01 4.17043537e-01
-1.09960926e+00 -6.50362447e-02 5.08539528e-02 -3.31352234e-01
-1.05600441e+00 4.29761857e-01 9.39493552e-02 -5.56577027e-01
-3.24957877e-01 8.08690429e-01 7.35034466e-01 -6.92532957e-01
1.09434021e+00 7.10412040e-02 6.74304366e-01 -2.45540157e-01
-7.25543201e-02 4.90071505e-01 6.75914645e-01 -2.98498967e-03
1.95548147e-01 4.55554038e-01 1.61459953e-01 -6.41354501e-01
-3.86462629e-01 -3.33909959e-01 -5.63732445e-01 -3.13516140e-01
8.84030402e-01 -3.96064132e-01 -3.38038057e-01 -9.91586596e-03
-7.56854951e-01 -1.11503109e-01 -4.68961656e-01 9.46393609e-01
-5.13941407e-01 2.76432812e-01 -8.28156888e-01 -6.65015697e-01
-9.76338506e-01 -1.17814779e+00 5.53486526e-01 2.89272130e-01
-5.20293117e-01 -1.25357068e+00 4.46595699e-02 4.71917808e-01
4.07184362e-01 2.67076522e-01 1.05350995e+00 -6.83887303e-01
-3.00732136e-01 -6.50989234e-01 7.37205446e-02 3.05250198e-01
4.94901031e-01 -1.88523561e-01 -7.43292034e-01 6.38861805e-02
1.75273925e-01 -9.91833359e-02 3.01815778e-01 1.00432014e+00
1.26256144e+00 -3.33159268e-01 -5.91567755e-01 3.85480136e-01
1.27632868e+00 9.21222389e-01 5.67984343e-01 1.81039423e-01
2.49767706e-01 6.39701843e-01 3.68206978e-01 2.93231606e-01
-1.51277199e-01 5.48259132e-02 8.81109759e-02 -5.61707258e-01
-1.20626561e-01 2.25646242e-01 -5.02603889e-01 7.62459219e-01
-3.26056093e-01 4.74858195e-01 -1.32474124e+00 4.66401964e-01
-1.40117621e+00 -7.98855424e-01 2.56709814e-01 1.90548646e+00
7.15586245e-01 -9.44013670e-02 -2.52073437e-01 -1.91446289e-01
5.77651858e-01 -9.28766370e-01 -4.88268018e-01 -3.34496140e-01
3.92577678e-01 5.39430976e-01 4.75215703e-01 3.15160275e-01
-1.00132084e+00 3.85137230e-01 8.65141106e+00 3.57064456e-01
-1.32807326e+00 1.12626314e-01 7.92928159e-01 -2.80835642e-03
3.80788334e-02 -5.85929215e-01 -7.27686733e-02 1.87663287e-01
9.74165738e-01 -7.29392916e-02 2.82630563e-01 8.66061687e-01
2.66078830e-01 -1.38960496e-01 -1.08966374e+00 1.10959554e+00
2.56849229e-01 -1.75798368e+00 -8.02399665e-02 -9.31446701e-02
5.51749408e-01 -1.63305461e-01 3.18067282e-01 -1.41013384e-01
-1.69185132e-01 -1.14728522e+00 -6.13995120e-02 7.18734384e-01
8.66459012e-01 -3.42943102e-01 6.69725895e-01 1.68892473e-01
-4.65466410e-01 -1.29183307e-01 1.94019929e-01 -1.19221844e-02
3.17220613e-02 2.41998836e-01 -1.55908847e+00 3.02965492e-01
6.72216296e-01 1.25536501e-01 -3.09773028e-01 1.18186235e+00
1.60620272e-01 3.78695369e-01 -1.39212847e-01 4.70876880e-02
7.37350956e-02 -1.14970729e-01 2.29553565e-01 1.04302704e+00
3.46439667e-02 2.03247696e-01 5.86302616e-02 7.66770363e-01
4.22111154e-01 3.22825700e-01 -7.53921926e-01 -1.26245126e-01
3.63363400e-02 1.25766122e+00 -9.09842670e-01 -6.35777056e-01
-2.94472426e-01 6.76223814e-01 -2.09236473e-01 1.17644193e-02
-6.22786343e-01 -1.24282382e-01 -1.59817368e-01 2.02055380e-01
1.01770267e-01 1.53195232e-01 -7.57658839e-01 -6.14295661e-01
-3.58825415e-01 -9.77537811e-01 4.96408165e-01 -8.04593682e-01
-9.99774337e-01 4.90209550e-01 -2.20384821e-02 -1.01652992e+00
-8.64302754e-01 -7.83769667e-01 -2.59775013e-01 7.13559330e-01
-1.04777074e+00 -1.21325600e+00 -1.27861992e-01 3.02897990e-01
3.52084339e-01 -5.61157525e-01 1.39615881e+00 3.38046789e-01
-1.05751179e-01 2.75462121e-01 7.40596354e-02 1.38255998e-01
4.29308653e-01 -1.06892204e+00 -3.42948288e-01 2.07420871e-01
-9.35438722e-02 3.72381389e-01 4.33877081e-01 -5.53512812e-01
-1.15513992e+00 -7.59686351e-01 5.58087230e-01 -1.86529532e-01
1.27017453e-01 8.35862458e-01 -6.23974681e-01 6.28486216e-01
6.27082959e-02 1.46637401e-02 1.22315598e+00 -2.49178842e-01
1.83662012e-01 -2.48343125e-03 -1.51510680e+00 3.90765965e-01
1.25982597e-01 -5.24074495e-01 -8.20146203e-01 4.55670983e-01
1.87846906e-02 -5.56452870e-01 -1.10771167e+00 7.79597580e-01
1.02457571e+00 -5.27475595e-01 1.11026692e+00 -8.79953802e-01
4.25898761e-01 1.33067563e-01 4.60657299e-01 -9.00724173e-01
-3.55607808e-01 -3.28885138e-01 1.80800930e-02 1.63320661e-01
6.83166385e-01 -6.94253981e-01 9.97425437e-01 1.13291824e+00
-5.33633903e-02 -7.45913744e-01 -7.75726438e-01 -3.11860949e-01
-2.26591271e-03 -1.78651690e-01 -1.21050514e-01 9.01082575e-01
2.69253343e-01 -4.64034490e-02 1.14348633e-02 -3.07772355e-03
4.24891979e-01 1.24950066e-01 1.79832563e-01 -1.24065661e+00
-5.82276881e-01 -4.80237782e-01 -7.35467255e-01 -1.06735542e-01
-3.31526369e-01 -8.03931475e-01 -3.00004542e-01 -1.58733714e+00
2.85972625e-01 -3.98709834e-01 -3.15911621e-01 5.41746318e-01
8.81278813e-02 2.37447008e-01 -2.27227435e-02 4.69178408e-01
-2.24260628e-01 -3.37795794e-01 1.34769988e+00 -1.77159980e-01
-2.13723466e-01 -4.22930494e-02 -4.75287765e-01 9.39835072e-01
1.09199870e+00 -4.76269931e-01 -2.02316985e-01 -2.58951724e-01
-1.02006570e-01 4.05373037e-01 2.52678096e-01 -1.04634607e+00
3.46624613e-01 1.82694346e-02 9.15011346e-01 -2.78553069e-01
2.06539661e-01 -1.14900720e+00 5.54627717e-01 1.26925707e+00
-3.76334608e-01 -3.87018584e-02 3.02891374e-01 7.66928419e-02
-2.56275326e-01 -5.58318377e-01 9.38631654e-01 -8.95332575e-01
-5.63548088e-01 -1.40822500e-01 -9.43373859e-01 -7.07587421e-01
1.38303483e+00 -3.48077059e-01 3.57516646e-01 -2.84076631e-01
-1.32056534e+00 -1.61595583e-01 -1.43420279e-01 -7.02755600e-02
7.80744731e-01 -9.58769560e-01 -4.28054482e-01 2.61472583e-01
-2.07633063e-01 -2.78944224e-01 4.59401786e-01 1.12941647e+00
-1.14233577e+00 6.44157052e-01 -5.49078047e-01 -5.79151809e-01
-1.45800567e+00 6.45722151e-01 9.31596935e-01 -3.98869157e-01
-4.35075223e-01 5.31154990e-01 -7.15380013e-02 -2.22784340e-01
7.18977600e-02 5.39706573e-02 -2.50695050e-01 -5.97195148e-01
5.04899204e-01 3.63625169e-01 2.48647287e-01 -3.03722601e-02
-9.22132358e-02 3.82245451e-01 -2.29336381e-01 -9.35645700e-02
9.75328088e-01 -3.36714014e-02 6.03991374e-02 2.70524800e-01
9.21636760e-01 -5.64347446e-01 -3.97940964e-01 5.80996610e-02
-7.84536973e-02 -2.15497002e-01 3.69903445e-01 -1.46472478e+00
-9.27187204e-01 8.46289694e-01 1.28013361e+00 -1.07675500e-01
1.29795790e+00 5.42804822e-02 3.02326649e-01 5.34832537e-01
1.58610135e-01 -9.61340129e-01 -1.06564527e-02 -3.71353686e-01
6.10853851e-01 -1.47072172e+00 2.29606211e-01 -4.35899079e-01
-6.53853238e-01 1.26628625e+00 3.20706636e-01 -2.63333917e-02
7.18595862e-01 4.49901342e-01 5.41935623e-01 -3.74470174e-01
-2.48202994e-01 1.46585912e-01 5.00448287e-01 6.43740594e-01
5.15355468e-01 9.66115817e-02 -4.14370298e-01 3.63283455e-01
3.45291078e-01 6.08016491e-01 -9.21484381e-02 1.36445189e+00
-4.89112437e-01 -1.00986898e+00 -5.35187602e-01 8.05906534e-01
-9.27592874e-01 -1.19735341e-04 -2.14495376e-01 6.02281392e-01
1.11904703e-01 6.66136324e-01 -2.30215609e-01 -4.16986756e-02
-4.79402719e-03 1.95584744e-01 8.79986703e-01 -4.41237479e-01
-7.62830615e-01 3.16993922e-01 -3.28825153e-02 -3.40622991e-01
-5.62946439e-01 -5.18576622e-01 -1.24602091e+00 4.26316977e-01
-2.58155435e-01 -1.38292182e-03 1.02005553e+00 8.36196780e-01
5.57351291e-01 7.17766225e-01 3.01643640e-01 -6.32660151e-01
-5.09685397e-01 -9.54515159e-01 -3.24549884e-01 1.85700446e-01
1.20206513e-01 -3.50915551e-01 3.17728668e-02 6.27176940e-01] | [14.903812408447266, -2.4608302116394043] |
956ea3a7-4b5b-4787-8f04-24725c4a893b | lmbao-a-landmark-map-for-bundle-adjustment | 2209.08810 | null | https://arxiv.org/abs/2209.08810v1 | https://arxiv.org/pdf/2209.08810v1.pdf | LMBAO: A Landmark Map for Bundle Adjustment Odometry in LiDAR SLAM | LiDAR odometry is one of the essential parts of LiDAR simultaneous localization and mapping (SLAM). However, existing LiDAR odometry tends to match a new scan simply iteratively with previous fixed-pose scans, gradually accumulating errors. Furthermore, as an effective joint optimization mechanism, bundle adjustment (BA) cannot be directly introduced into real-time odometry due to the intensive computation of large-scale global landmarks. Therefore, this letter designs a new strategy named a landmark map for bundle adjustment odometry (LMBAO) in LiDAR SLAM to solve these problems. First, BA-based odometry is further developed with an active landmark maintenance strategy for a more accurate local registration and avoiding cumulative errors. Specifically, this paper keeps entire stable landmarks on the map instead of just their feature points in the sliding window and deletes the landmarks according to their active grade. Next, the sliding window length is reduced, and marginalization is performed to retain the scans outside the window but corresponding to active landmarks on the map, greatly simplifying the computation and improving the real-time properties. In addition, experiments on three challenging datasets show that our algorithm achieves real-time performance in outdoor driving and outperforms state-of-the-art LiDAR SLAM algorithms, including Lego-LOAM and VLOM. | ['Zhifei Duan', 'Xiaojun Tan', 'Nanjie Chen', 'Lu Jie', 'Jinping Wang', 'Letian Zhang'] | 2022-09-19 | null | null | null | null | ['simultaneous-localization-and-mapping'] | ['computer-vision'] | [-2.09575579e-01 -3.61416727e-01 -2.82756597e-01 -5.72411835e-01
-5.56749701e-01 -1.36918649e-01 2.59677052e-01 3.92233998e-01
-7.72627771e-01 8.80663991e-01 -3.75906616e-01 -2.77406424e-01
-2.96762049e-01 -1.10927594e+00 -6.48223698e-01 -5.79800546e-01
8.56342837e-02 8.71080577e-01 5.96037924e-01 -2.80656278e-01
3.58890831e-01 7.13869989e-01 -1.38652813e+00 -1.01399720e+00
1.36544108e+00 6.59671545e-01 4.20752406e-01 -1.37001798e-01
-4.48418915e-01 -1.09938152e-01 -2.07801253e-01 -9.00293887e-02
2.38171220e-01 -1.02679603e-01 -1.13145307e-01 1.80908926e-02
5.86987257e-01 -1.92895368e-01 -2.63220698e-01 1.26913488e+00
5.58308363e-01 2.40913644e-01 1.04234563e-02 -1.38272691e+00
3.70202363e-02 2.85090059e-02 -7.49579012e-01 -5.92380650e-02
1.26728520e-01 -1.16029702e-01 7.94400632e-01 -1.20309293e+00
5.00376821e-01 1.02676570e+00 7.35529780e-01 -1.88597962e-01
-1.33387196e+00 -1.08099020e+00 1.25971949e-02 2.99569786e-01
-2.03276706e+00 -3.47547859e-01 7.49121428e-01 -2.48488739e-01
6.15686357e-01 1.31300062e-01 8.16047311e-01 2.17542082e-01
5.19219995e-01 1.11717649e-01 8.63858938e-01 -3.54066938e-02
1.48167117e-02 2.98845079e-02 2.56570820e-02 9.75849152e-01
8.27401936e-01 7.49851093e-02 -7.02355087e-01 -2.03014109e-02
6.43172681e-01 2.78289616e-01 -1.37926549e-01 -1.05915630e+00
-1.31268907e+00 8.77735317e-01 7.03840077e-01 -1.98930442e-01
-2.58801877e-01 4.48450595e-02 7.32119083e-02 -2.12553088e-02
2.65145838e-01 6.59981370e-02 -3.14727649e-02 -1.07429676e-01
-1.04148948e+00 1.28459439e-01 2.88043946e-01 1.18433201e+00
1.71160960e+00 -5.48233204e-02 4.93499160e-01 6.13967836e-01
4.45981205e-01 7.36354887e-01 2.83571005e-01 -6.95940435e-01
7.13953495e-01 8.02689672e-01 1.72462225e-01 -1.10923779e+00
-5.00639558e-01 -4.53615665e-01 -8.11356008e-01 1.53673440e-01
3.48949060e-02 1.19895622e-01 -7.53305793e-01 1.32023394e+00
5.79994142e-01 4.44520712e-01 -2.03242853e-01 8.50498974e-01
3.73375744e-01 3.97286564e-01 -1.82227850e-01 -3.60820830e-01
1.19263208e+00 -6.50949955e-01 -8.70978534e-01 -5.88143110e-01
5.38242519e-01 -9.23120677e-01 7.41348326e-01 9.45741683e-03
-6.42108560e-01 -7.71558762e-01 -1.52942431e+00 -2.79951125e-01
-1.77640960e-01 -2.31371392e-02 5.07857382e-01 4.26522166e-01
-6.63949847e-01 3.81108105e-01 -9.56044793e-01 -1.79640800e-01
1.43373489e-01 3.37644845e-01 -3.69282722e-01 -2.70445049e-01
-1.09282053e+00 1.11192226e+00 3.45433027e-01 4.85465735e-01
-3.42613131e-01 -4.69519079e-01 -1.13018537e+00 -4.43289101e-01
4.51059759e-01 -4.57089633e-01 6.92569435e-01 2.08538547e-01
-1.24200964e+00 7.16485798e-01 -8.66850972e-01 -3.39178264e-01
6.62655294e-01 -2.65317529e-01 -4.27643150e-01 -3.29574108e-01
5.83896577e-01 5.65639198e-01 4.94192600e-01 -1.24273849e+00
-8.11125755e-01 -5.47427535e-01 -2.32655004e-01 3.43810916e-01
2.55547196e-01 -6.35462880e-01 -6.18215919e-01 -2.64276285e-02
1.39871931e+00 -1.09934855e+00 -4.19286549e-01 -8.97903740e-02
-1.40494004e-01 -1.03158087e-01 7.99070835e-01 -2.14472905e-01
1.17655230e+00 -2.32755661e+00 -7.04234838e-02 3.39519978e-01
2.53890902e-01 -1.58726037e-01 2.18972251e-01 2.50337094e-01
4.09406453e-01 -2.12601364e-01 -3.61143500e-01 -6.87250078e-01
-1.30631462e-01 7.01750875e-01 -2.44467199e-01 8.43158841e-01
-3.63048352e-02 7.90081441e-01 -1.02511179e+00 -8.82326007e-01
4.37358648e-01 2.41972029e-01 -2.56233543e-01 -2.20175669e-01
3.54124308e-01 6.23060167e-01 -4.42081004e-01 6.34807765e-01
1.29369557e+00 4.53703433e-01 -1.67473078e-01 -1.81114167e-01
-8.12538505e-01 6.09330595e-01 -1.65276790e+00 1.96481991e+00
-5.07910430e-01 3.63155693e-01 8.52378085e-02 -5.39627910e-01
1.49276769e+00 -1.94954038e-01 5.49630880e-01 -7.21490562e-01
-1.67707935e-01 6.60092056e-01 -1.79485932e-01 -6.35626772e-03
9.34713602e-01 -1.06203437e-01 -9.24440026e-02 8.97454843e-02
-4.11068082e-01 -7.48107016e-01 9.71379802e-02 -3.62582020e-02
5.46873331e-01 1.50708571e-01 4.57626462e-01 -2.30840340e-01
7.78487325e-01 2.18345463e-01 1.10511744e+00 2.99940497e-01
-3.25268358e-01 4.07289892e-01 -1.78073570e-01 -3.07327509e-01
-7.27044344e-01 -1.41047478e+00 -6.16284370e-01 2.90877342e-01
1.03987885e+00 -5.86956263e-01 -1.25700668e-01 -2.54959941e-01
5.08035243e-01 4.20488715e-01 -6.46008253e-02 -1.56616628e-01
-8.66976023e-01 -5.53121269e-01 2.59730667e-01 2.34693348e-01
8.23499858e-01 -6.71149492e-01 -6.54785991e-01 5.51926434e-01
-2.85896569e-01 -8.68432343e-01 -4.25155878e-01 1.34993151e-01
-1.19814241e+00 -7.82980859e-01 5.71096838e-02 -6.65080130e-01
6.87602878e-01 9.22419906e-01 6.17156148e-01 1.39095128e-01
1.22862823e-01 -4.94024098e-01 4.47158003e-03 -3.44797313e-01
1.33052886e-01 -3.24079162e-03 3.28818232e-01 -1.60028562e-01
5.78214109e-01 -7.06246734e-01 -3.96956921e-01 7.35143185e-01
-3.89477879e-01 -8.24625567e-02 6.56683147e-01 7.46177971e-01
1.18124473e+00 1.33718416e-01 2.03585047e-02 -4.53904748e-01
1.49954036e-01 -2.40374655e-01 -1.05308712e+00 -3.20571601e-01
-9.31974292e-01 -1.01022042e-01 -2.10987702e-02 -6.04296215e-02
-5.25702178e-01 3.79627645e-01 -8.93351510e-02 -2.38911018e-01
2.13425264e-01 5.62126219e-01 -4.64136243e-01 -4.19616282e-01
2.51326919e-01 4.66717988e-01 1.53068528e-01 -5.33741832e-01
2.42072582e-01 5.09866655e-01 6.06314421e-01 -2.73557633e-01
1.38645720e+00 7.50646174e-01 5.32070398e-01 -6.51004732e-01
-5.68552971e-01 -7.00546205e-01 -1.27202678e+00 1.66159850e-02
5.10577977e-01 -1.11734235e+00 -5.80192566e-01 2.61095673e-01
-1.05816150e+00 2.17657968e-01 -1.64768323e-01 9.05643344e-01
-3.18293720e-01 6.00016057e-01 -1.82684347e-01 -6.43723786e-01
1.24331236e-01 -1.44334841e+00 1.04445219e+00 2.13868544e-01
-1.64780300e-02 -5.20003974e-01 2.03418747e-01 1.01638027e-01
4.15415019e-02 1.22117199e-01 3.62947285e-01 1.05606817e-01
-1.18848193e+00 -3.49927992e-01 -2.85396159e-01 -1.69998094e-01
2.17123136e-01 -1.69872224e-01 -6.45333648e-01 -4.38578188e-01
3.69629525e-02 2.70518303e-01 7.06617951e-01 4.42086942e-02
4.10400242e-01 3.24835867e-01 -6.50539696e-01 8.53249550e-01
1.51027918e+00 1.44520521e-01 6.31913483e-01 8.00766587e-01
7.51377106e-01 3.53545427e-01 1.39691281e+00 2.27061644e-01
8.25343430e-01 8.96255791e-01 6.13089919e-01 -4.63494752e-03
2.92819411e-01 -6.09450698e-01 2.40589321e-01 1.06191397e+00
1.55055478e-01 3.86308700e-01 -7.91681707e-01 4.87793386e-01
-1.86904979e+00 -4.98398423e-01 -6.43394470e-01 2.63449168e+00
5.54133236e-01 4.05317098e-01 -2.62543142e-01 1.25678062e-01
7.77773380e-01 4.37422365e-01 -4.45983112e-01 -5.50104491e-02
3.11847590e-02 1.79066479e-01 1.03459728e+00 9.30298269e-01
-9.46494997e-01 1.08644795e+00 4.86280107e+00 6.16173923e-01
-1.02615678e+00 3.55999500e-01 -5.49090266e-01 1.56873867e-01
-1.09473489e-01 5.06297648e-01 -1.42578733e+00 4.84602064e-01
5.40855944e-01 -2.50124007e-01 1.95579186e-01 9.37139690e-01
4.26438153e-01 -5.01098275e-01 -6.24865174e-01 9.80523586e-01
-1.26259848e-01 -1.00182092e+00 -2.04239294e-01 5.10879457e-01
4.59686100e-01 3.10266554e-01 -3.04618508e-01 1.71407893e-01
4.25609481e-03 -4.97141063e-01 8.21192443e-01 2.26560339e-01
7.09186733e-01 -9.37605321e-01 8.09799016e-01 5.78338802e-01
-1.82253075e+00 3.67273957e-01 -8.61939549e-01 -1.80770934e-01
6.72456264e-01 1.00966966e+00 -9.09042895e-01 9.51822281e-01
4.43935275e-01 8.46034646e-01 -6.72262311e-01 1.28537750e+00
-4.63844657e-01 -2.70637535e-02 -6.50714457e-01 2.55848646e-01
2.59603947e-01 -8.50233853e-01 7.48406589e-01 7.52005339e-01
6.50147021e-01 -2.74609834e-01 7.02531695e-01 6.45108581e-01
2.76369572e-01 1.33367732e-01 -6.87865376e-01 6.44111454e-01
1.00637472e+00 1.10859632e+00 -4.59959120e-01 -2.53484786e-01
-2.72368729e-01 7.04430878e-01 1.51670501e-01 3.82285267e-02
-7.95527518e-01 -6.22764766e-01 1.03004146e+00 2.58530021e-01
-1.84060149e-02 -1.19347835e+00 -4.17383671e-01 -9.60480630e-01
2.50423521e-01 -1.23120785e-01 4.32165042e-02 -4.64395285e-01
-6.32718325e-01 3.56998622e-01 -1.17750622e-01 -1.81183851e+00
3.28907855e-02 3.40400985e-03 -5.11008978e-01 1.16598701e+00
-1.93509066e+00 -8.53898346e-01 -8.20837319e-01 3.64559203e-01
3.45053077e-01 3.38638306e-01 3.66788894e-01 5.53252101e-01
-4.99398828e-01 2.15225607e-01 1.58446040e-02 -1.28326401e-01
9.13574219e-01 -1.03546870e+00 6.53221548e-01 1.08092177e+00
2.08054364e-01 8.84159386e-01 6.69535637e-01 -9.78612483e-01
-1.24834526e+00 -1.11462736e+00 1.29788446e+00 -3.04241359e-01
6.04422450e-01 -3.77133638e-01 -1.06740916e+00 8.08242738e-01
-4.81150150e-01 -1.22957043e-02 -9.79928672e-03 -3.76346037e-02
8.53565037e-02 -5.52334487e-01 -9.03091311e-01 4.60700363e-01
1.35792458e+00 -4.90104496e-01 -5.64513147e-01 1.84885174e-01
6.79365098e-01 -8.29438210e-01 -6.42026007e-01 6.52019382e-01
4.33855444e-01 -9.10575271e-01 1.00688589e+00 2.39486262e-01
-5.10720611e-01 -1.09403777e+00 -7.34656379e-02 -1.15252781e+00
-1.88372642e-01 -4.80881006e-01 2.18849167e-01 1.30068982e+00
-6.89445436e-02 -1.19908249e+00 8.84135485e-01 -7.98706040e-02
-3.77306134e-01 -4.74525690e-01 -1.44012189e+00 -9.39226091e-01
-5.31845748e-01 -4.00366813e-01 8.97341430e-01 8.04599822e-01
-4.40678418e-01 3.00519913e-01 -2.48610720e-01 6.42310083e-01
9.06953752e-01 2.86641806e-01 1.37005627e+00 -1.67343533e+00
4.43509549e-01 -1.28004014e-01 -8.20398271e-01 -1.52041566e+00
1.34882450e-01 -7.64218748e-01 4.16088909e-01 -1.42856967e+00
-3.73400092e-01 -1.05200553e+00 -9.29466784e-02 6.71276748e-02
-1.40052170e-01 3.88598293e-01 -4.29369807e-02 5.99999547e-01
-2.06201464e-01 6.93278372e-01 1.21085012e+00 3.29975411e-02
-3.87825102e-01 1.85455278e-01 -1.05312299e-02 7.08372116e-01
5.56298673e-01 -8.00181031e-01 -2.02357635e-01 -3.62540752e-01
-8.07150267e-03 -1.58786759e-01 2.21149668e-01 -1.24767268e+00
4.38361287e-01 -4.27509606e-01 3.16100866e-02 -1.34620738e+00
6.14940226e-01 -9.14627373e-01 3.71110946e-01 7.71074712e-01
6.03693545e-01 3.63577217e-01 -8.40277690e-03 6.33119226e-01
-4.40288156e-01 -3.07310551e-01 7.19698429e-01 2.42164254e-01
-1.02459109e+00 5.39766967e-01 7.88763463e-02 -4.07100499e-01
9.23983872e-01 -5.48099577e-01 -1.24369547e-01 -1.14800166e-02
-3.60264003e-01 6.22040987e-01 9.52912629e-01 2.96548218e-01
7.07676589e-01 -1.53591204e+00 -5.02133727e-01 5.25637329e-01
3.82232875e-01 6.96457863e-01 -1.87330060e-02 1.23582768e+00
-6.93074346e-01 5.04225552e-01 -1.67738542e-01 -1.08400786e+00
-1.10227609e+00 2.77224451e-01 1.39633179e-01 -3.85899805e-02
-7.45894611e-01 5.33129632e-01 -1.22380883e-01 -6.02282763e-01
-1.45004839e-01 -3.54340494e-01 9.95107964e-02 1.51492283e-01
5.44111840e-02 3.99409235e-01 2.00971454e-01 -9.87297952e-01
-6.60629570e-01 1.25100493e+00 2.29035243e-01 -1.61492303e-01
8.90518546e-01 -7.31409132e-01 -3.28549564e-01 5.98125696e-01
1.02598000e+00 6.01110458e-01 -1.03987956e+00 -3.23606074e-01
2.74988592e-01 -9.48647976e-01 4.16545421e-02 2.67497808e-01
-6.87290847e-01 9.81623948e-01 5.74794829e-01 -3.28935117e-01
5.75757205e-01 -3.92262042e-01 1.05998242e+00 5.58257759e-01
1.10926867e+00 -8.55622232e-01 -2.76783705e-01 5.97994328e-01
5.16739011e-01 -1.19099164e+00 5.11670947e-01 -7.14504600e-01
-3.15002501e-01 8.39122772e-01 6.50268495e-01 -1.40096307e-01
5.00237465e-01 -3.35866609e-03 1.92538574e-01 -5.18685952e-03
9.04211849e-02 -3.51190627e-01 -7.53628137e-03 3.58002305e-01
-2.93799993e-02 8.69110897e-02 -5.65322757e-01 -6.48111925e-02
-4.24523443e-01 -1.69523969e-01 3.38191837e-01 1.13618851e+00
-9.01289165e-01 -1.26108134e+00 -4.98957098e-01 7.97333717e-02
3.53423148e-01 5.85104376e-02 1.70124888e-01 1.04784119e+00
4.58332360e-01 6.65957510e-01 3.32937181e-01 -4.05386776e-01
6.21953964e-01 3.14330421e-02 2.32415587e-01 -6.39024436e-01
1.96776316e-01 1.42639741e-01 -2.12866887e-01 -7.20736623e-01
-1.46909177e-01 -8.50985229e-01 -1.67082918e+00 -3.05820256e-01
-6.41139448e-01 4.37770605e-01 9.75368619e-01 8.17237020e-01
3.56650352e-01 1.45585343e-01 7.13623047e-01 -8.52676332e-01
-4.59091246e-01 -7.73847640e-01 -7.40783632e-01 4.47259992e-02
4.00359809e-01 -1.31244671e+00 -2.57364094e-01 -5.02002239e-01] | [7.393918514251709, -2.2320444583892822] |
21576a6b-36e0-4b58-a7c1-466d00afcdc1 | constructing-code-mixed-universal-dependency | 2305.12258 | null | https://arxiv.org/abs/2305.12258v3 | https://arxiv.org/pdf/2305.12258v3.pdf | Constructing Code-mixed Universal Dependency Forest for Unbiased Cross-lingual Relation Extraction | Latest efforts on cross-lingual relation extraction (XRE) aggressively leverage the language-consistent structural features from the universal dependency (UD) resource, while they may largely suffer from biased transfer (e.g., either target-biased or source-biased) due to the inevitable linguistic disparity between languages. In this work, we investigate an unbiased UD-based XRE transfer by constructing a type of code-mixed UD forest. We first translate the sentence of the source language to the parallel target-side language, for both of which we parse the UD tree respectively. Then, we merge the source-/target-side UD structures as a unified code-mixed UD forest. With such forest features, the gaps of UD-based XRE between the training and predicting phases can be effectively closed. We conduct experiments on the ACE XRE benchmark datasets, where the results demonstrate that the proposed code-mixed UD forests help unbiased UD-based XRE transfer, with which we achieve significant XRE performance gains. | ['Tat-Seng Chua', 'Min Zhang', 'Meishan Zhang', 'Hao Fei'] | 2023-05-20 | null | null | null | null | ['relation-extraction'] | ['natural-language-processing'] | [-1.03441626e-01 -7.79392645e-02 -9.22560155e-01 -5.38339794e-01
-1.31210327e+00 -5.71345150e-01 4.90021348e-01 -2.90485501e-01
-5.84688373e-02 9.22242820e-01 3.85267228e-01 -9.15300548e-01
3.47719014e-01 -7.44250536e-01 -8.81167233e-01 -4.08100098e-01
1.64426118e-01 9.07262322e-03 9.20134410e-02 -3.59806418e-01
5.66786751e-02 -5.50759174e-02 -9.09980834e-01 3.78301561e-01
1.28810012e+00 7.05438197e-01 2.01473787e-01 3.29283506e-01
-5.42350888e-01 1.08914340e+00 -2.20209435e-01 -6.31922424e-01
6.41255528e-02 -4.19610769e-01 -9.95594382e-01 -4.44978476e-01
7.83356577e-02 -7.80311674e-02 -1.96168587e-01 1.03114748e+00
1.90314636e-01 -4.82231885e-01 6.77086532e-01 -1.27481377e+00
-1.08207500e+00 1.24767888e+00 -1.04951322e+00 1.01369023e-01
3.75886649e-01 -1.77374601e-01 1.33348215e+00 -9.96364772e-01
8.84892523e-01 1.06131887e+00 6.90709651e-01 7.20592558e-01
-1.10868704e+00 -1.00795829e+00 2.42061168e-01 -8.56671557e-02
-1.34077203e+00 -6.16517723e-01 7.68321216e-01 -5.71667969e-01
1.07164657e+00 1.09073624e-01 -6.60005584e-02 1.17858696e+00
5.82089126e-01 8.09191644e-01 1.31912577e+00 -4.65249538e-01
-2.40668103e-01 2.51353323e-01 4.93755370e-01 8.52105081e-01
1.86681882e-01 1.42036542e-01 -6.04706824e-01 1.02400798e-02
4.07170594e-01 -4.25579399e-01 -4.46448177e-01 -2.53796875e-01
-1.12566209e+00 7.49557793e-01 6.64604604e-01 1.73121199e-01
7.47902915e-02 -2.54631102e-01 7.66623616e-01 4.00277436e-01
5.22154033e-01 1.60484519e-02 -8.75898242e-01 -9.05504525e-02
-4.86863136e-01 -1.32192463e-01 7.26938128e-01 1.64972460e+00
8.50196898e-01 -2.09095463e-01 -1.22105345e-01 1.07641304e+00
4.57173914e-01 5.06430030e-01 4.69856918e-01 -1.51164919e-01
1.10699522e+00 6.23373508e-01 -3.94115269e-01 -5.19878268e-01
-6.56785220e-02 -5.23149729e-01 -9.33596730e-01 -3.44411135e-01
4.84413579e-02 -2.90359229e-01 -5.85895061e-01 1.91646445e+00
2.74018407e-01 -3.71179342e-01 4.57705885e-01 5.84034383e-01
1.03409338e+00 5.99994898e-01 1.65055647e-01 -2.08777234e-01
1.36825442e+00 -1.36245990e+00 -6.23700976e-01 -5.23744762e-01
1.11606860e+00 -7.91050613e-01 1.24587929e+00 -1.17274299e-01
-5.55449605e-01 -4.99794155e-01 -1.30786550e+00 -2.30811328e-01
-1.00503378e-01 3.95546883e-01 6.32542312e-01 5.59840441e-01
-5.50380170e-01 2.08047703e-01 -7.34292686e-01 -6.05587810e-02
3.35604459e-01 -1.85637304e-03 -4.67215002e-01 -3.79900157e-01
-1.30350387e+00 9.20710146e-01 2.25869998e-01 1.08065214e-02
-6.54771447e-01 -6.74550533e-01 -1.07023454e+00 -3.02691787e-01
3.48975092e-01 -3.50909293e-01 1.26691782e+00 -5.88359296e-01
-1.30911338e+00 9.23374891e-01 -4.66279268e-01 -7.35681802e-02
3.66750181e-01 -3.96183580e-01 -5.37486613e-01 -6.05237663e-01
5.27043879e-01 1.75560907e-01 3.36748093e-01 -1.37750244e+00
-6.86424375e-01 -2.39180252e-01 6.76477626e-02 4.30602431e-02
-2.59804577e-01 2.50557363e-01 -5.08206367e-01 -5.10075331e-01
-4.71133590e-02 -1.03348196e+00 1.30334750e-01 -5.37726521e-01
-5.59132993e-01 -4.05177951e-01 6.70118630e-01 -8.49715352e-01
1.70201087e+00 -2.17530370e+00 2.48471156e-01 -1.84317932e-01
1.97892636e-01 -3.04012801e-02 -1.02139123e-01 3.99367630e-01
-1.86044872e-01 3.39928567e-01 -4.59221840e-01 -2.85157084e-01
-1.74569309e-01 4.17712152e-01 -4.31192636e-01 2.67908931e-01
4.88801986e-01 9.06940818e-01 -9.81179237e-01 -8.23520839e-01
-5.05855024e-01 7.67896548e-02 -6.13023579e-01 5.01734078e-01
-8.40343833e-02 5.02019048e-01 -6.28544927e-01 7.60351837e-01
7.28715003e-01 -6.41669556e-02 3.67186606e-01 -1.22364298e-01
-4.09037977e-01 9.18287575e-01 -5.06843567e-01 1.83395267e+00
-1.16450202e+00 4.70864415e-01 -7.68509954e-02 -5.54402232e-01
1.08419240e+00 1.29339725e-01 -1.30635351e-01 -6.60484016e-01
-1.30918056e-01 6.00054085e-01 1.74306586e-01 -2.75851965e-01
4.58610892e-01 -2.20113575e-01 -4.40256059e-01 4.95745391e-01
1.19247325e-02 7.11775199e-02 -1.04444876e-01 3.23934466e-01
1.05004442e+00 7.17137992e-01 5.91938376e-01 -4.90669608e-01
7.36698091e-01 5.39131556e-03 8.01244617e-01 1.67541802e-01
-2.66346764e-02 3.73720407e-01 7.12097526e-01 -2.00531408e-02
-7.12139070e-01 -9.44755912e-01 -4.38126266e-01 1.20319045e+00
2.59196281e-01 -5.74215710e-01 -6.53143644e-01 -1.35565972e+00
-3.82029623e-01 7.94820368e-01 -4.80558902e-01 -3.61762196e-01
-7.96220660e-01 -8.68720889e-01 6.84912384e-01 6.65135086e-01
6.41311347e-01 -7.57288575e-01 1.18345357e-02 2.43476689e-01
-3.73676270e-01 -1.07153070e+00 -7.93832183e-01 4.88559753e-01
-5.09231210e-01 -7.82162547e-01 -4.35390443e-01 -1.03001869e+00
4.60690916e-01 2.29493231e-01 1.38462055e+00 3.10830865e-03
3.15284729e-01 -4.27002460e-01 -6.44553185e-01 1.64785201e-03
-7.83632159e-01 6.25870049e-01 -7.95477703e-02 -3.37891877e-01
4.73450512e-01 -4.89641935e-01 -1.66424647e-01 3.85679901e-01
-4.32675123e-01 3.11231434e-01 7.75658190e-01 9.85837638e-01
3.63130063e-01 -3.45876157e-01 8.89777005e-01 -1.35159039e+00
5.82815349e-01 -8.55682075e-01 -3.92088413e-01 5.32597005e-01
-8.96808803e-01 4.22959656e-01 8.02488148e-01 -2.70566374e-01
-1.35648823e+00 -1.64452121e-01 -4.73351300e-01 4.70989458e-02
2.66551495e-01 8.11552167e-01 -5.79392135e-01 1.50591373e-01
5.86732984e-01 2.41673663e-01 -5.03474176e-01 -4.64729667e-01
5.14976621e-01 1.27863348e+00 4.95988190e-01 -1.26422298e+00
7.87529349e-01 -2.12538883e-01 -4.93936926e-01 -2.32153103e-01
-1.06305337e+00 -2.68483102e-01 -6.39047742e-01 1.66403666e-01
9.38453436e-01 -1.09090400e+00 8.74939039e-02 3.17367494e-01
-1.48450029e+00 -1.78178012e-01 2.28218690e-01 2.94129133e-01
-1.52525693e-01 1.04270056e-01 -1.05163586e+00 -4.24313188e-01
-4.64126885e-01 -1.38385427e+00 1.23628831e+00 9.18329880e-02
-4.90332767e-02 -7.01614976e-01 2.67317832e-01 2.03782246e-01
2.31262416e-01 -1.25463873e-01 1.47614825e+00 -4.50336665e-01
-4.40506876e-01 4.45252657e-02 -6.27326369e-01 4.48190570e-01
3.90715659e-01 4.40832153e-02 -7.70970881e-01 -8.12463313e-02
-1.56915769e-01 -3.85922998e-01 4.53496307e-01 -3.87965828e-01
6.22316897e-01 -1.25640854e-01 -4.29993272e-01 6.24974012e-01
1.52632451e+00 1.29671380e-01 4.39338744e-01 3.66517961e-01
9.62740242e-01 3.93198282e-01 8.23642850e-01 -5.38080335e-02
1.02386439e+00 6.64887607e-01 6.77771121e-02 -4.81549054e-02
-3.49577188e-01 -5.74172676e-01 8.44531059e-01 1.83851159e+00
3.61692323e-03 1.62413731e-01 -1.18503892e+00 6.25721335e-01
-1.46540356e+00 -3.25573593e-01 -4.66559112e-01 1.92877173e+00
1.52361619e+00 2.53004968e-01 -2.72645712e-01 -2.90370077e-01
7.14345992e-01 1.57330707e-01 -3.37380648e-01 -5.65376937e-01
-2.52685733e-02 8.12736601e-02 4.34700012e-01 4.27487582e-01
-8.72963786e-01 1.08303642e+00 4.98346472e+00 9.75757062e-01
-1.17736077e+00 4.51563895e-01 3.48215491e-01 5.68195999e-01
-6.44403398e-01 2.95098841e-01 -1.16255224e+00 3.92119706e-01
9.35730398e-01 -4.47599322e-01 2.31312349e-01 1.02578878e+00
-4.37086463e-01 2.16755748e-01 -1.43447888e+00 6.75587118e-01
-2.81011403e-01 -1.08364606e+00 -1.98119268e-01 1.19282387e-01
7.37047970e-01 3.11411291e-01 -2.05651179e-01 8.55143011e-01
6.36625469e-01 -7.47176826e-01 1.02574992e+00 -7.33669102e-02
1.33279979e+00 -7.42944837e-01 8.04482341e-01 4.82720852e-01
-1.75778949e+00 2.50293165e-01 -3.16982329e-01 9.00277719e-02
-3.03643029e-02 6.03888869e-01 -4.51482385e-01 1.33555615e+00
6.00483418e-01 9.16178048e-01 -4.88561243e-01 1.97889999e-01
-6.14049137e-01 7.93219566e-01 1.05134912e-01 -2.44417191e-02
7.87318125e-02 -4.76683557e-01 3.15173835e-01 1.48598230e+00
3.19939166e-01 -1.61400318e-01 8.44013989e-02 9.26284373e-01
-4.55014557e-01 3.21758568e-01 -7.69246221e-01 8.37514624e-02
5.81561506e-01 1.21557355e+00 -2.00545728e-01 -2.11275741e-01
-1.11445129e+00 8.69422317e-01 8.02760124e-01 6.91301599e-02
-1.06644511e+00 -5.01899898e-01 6.55626178e-01 -3.40361536e-01
1.89729914e-01 -2.50896811e-01 -3.96528155e-01 -1.67462265e+00
2.27054521e-01 -1.07961202e+00 1.37786880e-01 -4.87738520e-01
-1.47117960e+00 1.22195852e+00 -9.27089378e-02 -1.46208763e+00
-1.10598980e-02 -3.49787712e-01 -4.55830216e-01 1.12917137e+00
-1.82411444e+00 -1.40992868e+00 1.41590595e-01 2.92996317e-01
5.91180980e-01 -2.00298056e-01 8.48954856e-01 7.72843957e-01
-9.45309520e-01 9.19710755e-01 4.64522615e-02 5.09381354e-01
7.84757614e-01 -1.17770255e+00 5.15520513e-01 1.05282700e+00
8.86136219e-02 1.06849110e+00 2.29697600e-01 -7.22681165e-01
-1.57287121e+00 -1.40471327e+00 1.13743711e+00 -6.54206574e-01
9.91836131e-01 -7.22008884e-01 -9.30483282e-01 9.87113893e-01
2.86458939e-01 1.78559020e-01 6.16056442e-01 5.32653809e-01
-9.80301738e-01 -6.12758063e-02 -6.75650716e-01 6.22669458e-01
1.38200867e+00 -9.26749885e-01 -8.45323145e-01 7.73227438e-02
1.11832821e+00 -5.00357628e-01 -1.26894426e+00 4.29858893e-01
4.31684464e-01 -7.04533696e-01 4.61339414e-01 -5.38768232e-01
1.18969166e+00 -3.38034809e-01 -4.44852591e-01 -1.38569748e+00
1.84760541e-02 -3.13477755e-01 1.69554770e-01 1.97740364e+00
9.33683395e-01 -6.89142764e-01 1.64869472e-01 2.48206317e-01
-4.11022484e-01 -9.82189417e-01 -8.71445894e-01 -9.30676341e-01
5.76169074e-01 -4.05006677e-01 7.95553803e-01 1.04293716e+00
5.22212051e-02 1.03766608e+00 -3.61221433e-01 4.99907546e-02
3.71985167e-01 6.30542994e-01 8.13035250e-01 -7.51300395e-01
-4.45009321e-01 -1.37598559e-01 2.04225481e-01 -1.38892174e+00
4.85274404e-01 -1.39379704e+00 3.60554338e-01 -1.23145366e+00
4.43994403e-01 -1.05674219e+00 -4.12616953e-02 5.41316509e-01
-4.33129758e-01 -1.78073794e-01 -5.87170795e-02 3.99670571e-01
-2.70777643e-01 6.72546387e-01 1.24734259e+00 -3.34489085e-02
5.99287078e-02 -1.72474697e-01 -8.51725161e-01 4.18470740e-01
5.06518245e-01 -7.85821259e-01 -3.56260121e-01 -6.21412039e-01
2.29925126e-01 2.78572083e-01 -3.91650051e-01 -5.29868245e-01
5.88012673e-03 -1.50990020e-02 -2.87806839e-01 -2.22636119e-01
-3.97672564e-01 -5.31970263e-01 -9.40604880e-02 2.52684385e-01
-2.25082144e-01 5.68285398e-02 1.13331519e-01 4.42336977e-01
-5.11931777e-01 -1.79048553e-01 7.20831096e-01 1.12733610e-01
-6.09600306e-01 2.90191978e-01 1.24861121e-01 4.19340044e-01
8.24406803e-01 2.52509594e-01 -5.97842336e-01 1.06197327e-01
6.21441044e-02 1.48269638e-01 4.14259613e-01 6.21368110e-01
2.51065195e-01 -1.52518976e+00 -8.64505053e-01 2.85080492e-01
5.47863424e-01 1.67473510e-01 -1.25930414e-01 8.89501452e-01
-2.51114458e-01 3.15626770e-01 7.26409629e-03 -4.15949702e-01
-1.09438324e+00 6.29106164e-01 1.60782143e-01 -6.18955910e-01
-5.63873827e-01 9.78476584e-01 4.11727935e-01 -9.18035269e-01
-3.49262983e-01 -1.83741122e-01 -1.90779883e-02 -2.12920740e-01
1.93637177e-01 -1.34221092e-01 2.46054456e-01 -7.21207857e-01
-6.50015295e-01 6.99399650e-01 -3.01186621e-01 1.43809631e-01
1.28824234e+00 -1.87807232e-01 -6.14055455e-01 4.53004569e-01
1.46536183e+00 7.10524797e-01 -8.06785166e-01 -4.70940143e-01
4.03902173e-01 -3.57195526e-01 -1.84787005e-01 -6.74064100e-01
-1.13911974e+00 9.67075467e-01 2.22564712e-02 -4.09607254e-02
1.09286702e+00 3.07963967e-01 1.03890252e+00 -2.14025490e-02
8.91083598e-01 -5.56178808e-01 -6.66656435e-01 6.82947814e-01
7.97073364e-01 -1.16289604e+00 -1.83599383e-01 -8.38122487e-01
-8.51541162e-01 8.78701448e-01 8.26032639e-01 1.20104775e-01
6.51387036e-01 6.65301561e-01 2.16146901e-01 1.24204114e-01
-9.65693653e-01 -5.39379939e-02 2.30857834e-01 6.15251303e-01
1.06679618e+00 2.74418801e-01 -6.52380347e-01 1.12339973e+00
-3.50258499e-01 -9.23721045e-02 4.77603734e-01 9.14000511e-01
7.02663958e-02 -1.61118388e+00 1.10344060e-01 3.88441950e-01
-3.89054775e-01 -6.51944339e-01 -2.00039506e-01 9.01753545e-01
1.42407091e-02 7.73846805e-01 -4.60025370e-01 -7.24855304e-01
3.15515548e-01 6.89401999e-02 4.12850618e-01 -8.33240926e-01
-5.40436089e-01 -1.42086074e-01 4.84978199e-01 -3.28422844e-01
-2.78842658e-01 -5.29074967e-01 -1.47907186e+00 -2.82420188e-01
-5.77431083e-01 2.65238792e-01 4.52044010e-01 1.00811601e+00
2.36448035e-01 6.42056048e-01 7.52395391e-01 -1.72734380e-01
-5.99069595e-01 -1.01717663e+00 -2.88791537e-01 4.23969962e-02
2.53853261e-01 -6.02670550e-01 -2.48539969e-01 9.66371894e-02] | [10.763157844543457, 9.67033576965332] |
50e1f93c-1d8f-46e4-bd4a-7f27e5df7401 | investigating-chain-of-thought-with-chatgpt | 2304.03087 | null | https://arxiv.org/abs/2304.03087v1 | https://arxiv.org/pdf/2304.03087v1.pdf | Investigating Chain-of-thought with ChatGPT for Stance Detection on Social Media | Stance detection predicts attitudes towards targets in texts and has gained attention with the rise of social media. Traditional approaches include conventional machine learning, early deep neural networks, and pre-trained fine-tuning models. However, with the evolution of very large pre-trained language models (VLPLMs) like ChatGPT (GPT-3.5), traditional methods face deployment challenges. The parameter-free Chain-of-Thought (CoT) approach, not requiring backpropagation training, has emerged as a promising alternative. This paper examines CoT's effectiveness in stance detection tasks, demonstrating its superior accuracy and discussing associated challenges. | ['Liwen Jing', 'Yangyang Li', 'Hu Huang', 'Daijun Ding', 'Xianghua Fu', 'BoWen Zhang'] | 2023-04-06 | null | null | null | null | ['stance-detection'] | ['natural-language-processing'] | [-7.99869001e-02 2.59311944e-01 -6.05896592e-01 -5.95995784e-01
-5.80199718e-01 -4.63372231e-01 9.02100205e-01 5.18747747e-01
-7.56429970e-01 7.17944860e-01 5.26026785e-01 -7.08042026e-01
1.88102901e-01 -7.99251139e-01 -3.62259626e-01 -3.24349433e-01
-2.29650587e-02 7.15422392e-01 3.81958991e-01 -6.89149261e-01
5.27158141e-01 -1.14869423e-01 -8.24119508e-01 5.20194888e-01
8.07821453e-01 7.71963179e-01 -4.93590683e-01 5.24389207e-01
-4.49662983e-01 1.21775544e+00 -6.74160123e-01 -9.69984353e-01
-7.19924048e-02 -1.33481756e-01 -8.09374332e-01 -4.73307937e-01
3.57936770e-01 -3.24617922e-01 -3.00426632e-01 7.74976552e-01
6.41256392e-01 6.21568821e-02 3.35381180e-01 -8.78577292e-01
-1.00037229e+00 1.15943277e+00 -5.28103709e-01 5.12777984e-01
1.75338194e-01 1.24261908e-01 1.23129916e+00 -6.79953873e-01
5.50829947e-01 1.53053939e+00 1.12428784e+00 4.68124598e-01
-1.13860428e+00 -4.37046885e-01 2.97509730e-01 2.27815807e-01
-5.27827859e-01 -1.07263133e-01 7.55870283e-01 -6.14589274e-01
1.29115844e+00 -1.37842014e-01 7.49976039e-01 1.63164902e+00
2.79725820e-01 9.14140344e-01 1.42096257e+00 -6.11752748e-01
8.18821490e-02 2.92959273e-01 4.85992014e-01 4.62149799e-01
1.58617839e-01 7.04676583e-02 -5.54938793e-01 -3.99371654e-01
3.13961148e-01 -3.30554634e-01 2.75207132e-01 2.15911150e-01
-9.33806479e-01 1.49116814e+00 5.16203642e-01 3.07677984e-01
-3.41522038e-01 4.15088385e-02 8.68066847e-01 4.59772199e-01
9.97298419e-01 5.46384811e-01 -3.16196233e-01 -2.84678191e-01
-1.01917875e+00 5.08627355e-01 1.02964532e+00 3.58200073e-01
1.17382206e-01 9.74389091e-02 -3.37110609e-01 7.21910119e-01
3.88446540e-01 1.86580092e-01 6.16555393e-01 -6.48616076e-01
6.93616509e-01 6.57524526e-01 -2.82629747e-02 -9.56875682e-01
-5.19577980e-01 -7.25454032e-01 -3.75278205e-01 6.73809871e-02
5.42040408e-01 -5.67000270e-01 -5.92748582e-01 1.61816502e+00
3.01726282e-01 -4.69585747e-01 9.77210999e-02 6.12098038e-01
6.73483312e-01 6.71977460e-01 4.32390392e-01 -1.32635295e-01
1.03278160e+00 -9.56932902e-01 -5.57532609e-01 -7.68608868e-01
5.76494694e-01 -8.27084422e-01 9.22231436e-01 3.76369417e-01
-1.02195954e+00 -2.90718555e-01 -8.19800675e-01 8.40175822e-02
-5.53668261e-01 -5.23070276e-01 6.93485022e-01 8.46780419e-01
-1.06697381e+00 7.03414977e-01 -4.21540231e-01 -2.49838680e-01
6.04359150e-01 4.84198511e-01 2.70958811e-01 3.77421379e-01
-1.65727639e+00 1.32994628e+00 4.54159617e-01 4.00666520e-02
-5.06256461e-01 -4.98143643e-01 -5.07022619e-01 -2.03523681e-01
4.22941715e-01 -5.84290445e-01 1.64532197e+00 -1.37693667e+00
-1.80212760e+00 9.57655072e-01 3.04261684e-01 -9.50286865e-01
7.69549727e-01 -5.92291057e-01 -2.66516328e-01 -4.55283746e-03
-1.07617460e-01 5.47465861e-01 9.88604248e-01 -5.81771255e-01
-4.42595035e-01 9.24574807e-02 4.66740131e-01 1.42032832e-01
-4.81168568e-01 7.80623436e-01 4.64112341e-01 -4.52143461e-01
-4.17018980e-01 -8.28104675e-01 -2.56382376e-01 -4.17523175e-01
-1.95118755e-01 -7.91595340e-01 5.98991573e-01 -8.07226837e-01
1.38660991e+00 -1.72517073e+00 -1.18193425e-01 1.45677596e-01
3.98923278e-01 6.81440949e-01 1.42712146e-01 7.17731416e-01
-7.17997029e-02 1.68615673e-02 3.12594265e-01 -2.20824957e-01
2.35170707e-01 1.00069582e-01 -1.35316834e-01 3.23624194e-01
2.51416177e-01 1.15199697e+00 -9.55935955e-01 -5.54375291e-01
-2.63544335e-03 3.20867419e-01 -8.09190392e-01 7.48802572e-02
-6.44151449e-01 9.24818888e-02 -3.00614834e-01 2.92287052e-01
1.47001937e-01 -4.80520189e-01 3.35238069e-01 7.24656135e-02
-3.96766931e-01 8.15364182e-01 -4.09729749e-01 7.82068670e-01
-1.70771912e-01 8.95105660e-01 -1.57801613e-01 -1.04165554e+00
7.60702670e-01 3.54114443e-01 2.90678084e-01 -9.25762534e-01
6.30407035e-01 3.17746580e-01 3.79844248e-01 -5.66740155e-01
4.32402760e-01 -2.17486396e-01 -1.23442098e-01 5.74691176e-01
-2.09655076e-01 4.13022876e-01 1.63658738e-01 1.23527616e-01
7.47369826e-01 9.95627418e-02 3.54729116e-01 -2.37037838e-01
3.81345630e-01 2.28100106e-01 3.20035100e-01 5.81687570e-01
-4.03755337e-01 5.70439915e-05 4.70652550e-01 -4.53381896e-01
-1.06659341e+00 -3.58961582e-01 1.45449311e-01 1.57652128e+00
-5.54173887e-01 -5.51203609e-01 -9.03222680e-01 -6.87813580e-01
5.12568429e-02 8.55261147e-01 -7.36753404e-01 -1.07901014e-01
-1.11550188e+00 -8.73173416e-01 6.08440518e-01 6.88112020e-01
6.62976623e-01 -1.23751867e+00 -7.62577474e-01 6.08349204e-01
-2.06964076e-01 -9.08903420e-01 -1.21711805e-01 4.07156289e-01
-8.45244288e-01 -7.60109246e-01 -6.12327814e-01 -5.78704476e-01
1.00609109e-01 -1.10374264e-01 1.16026103e+00 -5.19645866e-03
4.48160768e-01 -3.67247537e-02 -2.81535655e-01 -5.44358134e-01
-7.01584220e-01 2.33841151e-01 1.24145024e-01 -3.24181020e-01
7.37696588e-01 -4.64163840e-01 -5.19119322e-01 -1.69426322e-01
-3.95749241e-01 -3.26552205e-02 4.61428791e-01 8.42764378e-01
-1.88025534e-01 -4.89375085e-01 8.00843954e-01 -1.21128404e+00
1.21027648e+00 -5.36494792e-01 -3.81322712e-01 3.40705551e-02
-7.64183044e-01 -2.00027555e-01 5.07002115e-01 -7.39756763e-01
-9.91415083e-01 -8.90319109e-01 -5.59445202e-01 7.41401734e-03
-1.22129306e-01 1.09485006e+00 4.61369276e-01 -1.24982779e-03
1.11994040e+00 -6.07718267e-02 -3.95909026e-02 -3.82232606e-01
3.08139652e-01 7.78812110e-01 1.51425377e-01 -3.30273718e-01
5.68548620e-01 -1.23544002e-03 -7.11930990e-01 -7.54985988e-01
-1.47419846e+00 -2.40684539e-01 -3.79004538e-01 -3.21118444e-01
9.07401741e-01 -8.91983747e-01 -6.30575001e-01 4.51902002e-01
-9.58255827e-01 -7.68516302e-01 9.74856317e-02 3.79720241e-01
-3.52304012e-01 2.28403226e-01 -1.10578477e+00 -7.22514093e-01
-8.53128850e-01 -5.68605602e-01 3.23591948e-01 1.81763202e-01
-8.09480727e-01 -1.43194425e+00 3.10161233e-01 6.05373263e-01
7.92704403e-01 2.63263464e-01 1.08628559e+00 -1.00887203e+00
9.64170545e-02 -3.68231654e-01 -9.85873565e-02 4.66523826e-01
-2.28546575e-01 6.57444494e-03 -8.78176868e-01 -2.31703237e-01
-4.01721708e-02 -6.48722231e-01 5.64490139e-01 5.21039128e-01
5.03188431e-01 -6.75418615e-01 -8.98593515e-02 1.01958737e-01
1.11660063e+00 4.30626608e-02 1.39320672e-01 9.39577579e-01
3.64856511e-01 4.72168207e-01 4.60869431e-01 5.47132231e-02
4.65155423e-01 3.54348004e-01 3.07920903e-01 4.00475832e-03
-4.53361794e-02 -2.52742499e-01 6.94717169e-01 9.16708589e-01
-1.92168266e-01 -1.71175867e-01 -1.02189004e+00 1.72511220e-01
-1.84040534e+00 -1.02021396e+00 -1.57612428e-01 1.42334342e+00
1.11462569e+00 1.08489347e+00 5.78247726e-01 5.24091572e-02
3.91399056e-01 3.12178403e-01 -3.40021938e-01 -1.08697331e+00
-1.55321434e-01 2.52332002e-01 2.43445992e-01 5.47432005e-01
-1.19771111e+00 1.02780843e+00 6.93324423e+00 4.50863332e-01
-1.35125852e+00 3.63416135e-01 4.02119011e-01 1.95310161e-01
-5.44354320e-02 -1.50794819e-01 -1.06049955e+00 3.67886126e-01
1.56668103e+00 -2.07473934e-01 -1.08037487e-01 9.74329829e-01
3.40651363e-01 -4.17057462e-02 -8.74873281e-01 3.91062617e-01
5.06817624e-02 -1.32884657e+00 -5.35446890e-02 -6.01547733e-02
6.10373080e-01 3.47342074e-01 2.31386915e-01 9.09073114e-01
6.20890617e-01 -6.49228692e-01 9.87512767e-01 1.36128888e-01
3.33352208e-01 -4.68777627e-01 8.20667446e-01 6.63810670e-01
-4.34862316e-01 -4.53077048e-01 -3.55942458e-01 -5.31438887e-01
9.56733972e-02 4.12269771e-01 -9.74940479e-01 8.87570754e-02
5.71871221e-01 6.21878445e-01 -5.54385662e-01 7.95817137e-01
-3.94925147e-01 1.38511932e+00 -3.16443950e-01 -5.87698996e-01
9.00274038e-01 1.53931171e-01 4.41420615e-01 1.50947750e+00
-3.33986104e-01 -1.75150554e-03 3.34598303e-01 5.46184957e-01
8.47346783e-02 2.13330582e-01 -2.09008738e-01 -3.70195419e-01
1.89683482e-01 9.61072505e-01 -5.29398143e-01 -5.88211775e-01
-6.01476789e-01 2.68706948e-01 5.87630808e-01 1.17297415e-02
-1.04211092e+00 2.44667679e-01 4.23433512e-01 2.22433761e-01
1.38532117e-01 -2.58655161e-01 -3.16930622e-01 -7.56981432e-01
-5.23923278e-01 -1.33971691e+00 5.44198096e-01 -3.42412919e-01
-1.39361060e+00 5.42309940e-01 1.52183965e-01 -8.00077200e-01
-3.56845528e-01 -6.98029995e-01 -8.61958206e-01 7.21026182e-01
-1.36620545e+00 -1.24301648e+00 2.24844486e-01 4.09138560e-01
8.45729649e-01 -8.83499831e-02 7.97738671e-01 3.41494590e-01
-6.63750648e-01 5.93265593e-01 -3.11985072e-02 3.06175888e-01
7.93337047e-01 -1.30748403e+00 4.99085397e-01 4.16554838e-01
-4.19012249e-01 7.39639282e-01 1.11242461e+00 -7.06298113e-01
-7.30507433e-01 -8.66919219e-01 1.38063693e+00 -3.82527262e-01
1.17032170e+00 -3.39400440e-01 -9.12427664e-01 7.72009254e-01
8.12612593e-01 -7.77161598e-01 8.37124407e-01 4.77979571e-01
-6.08396411e-01 2.45202810e-01 -6.31221533e-01 8.10795486e-01
6.88901544e-01 -4.52420443e-01 -1.18614888e+00 5.21589220e-01
6.02901578e-01 -3.82215858e-01 -6.24289691e-01 1.10487975e-01
4.61734176e-01 -9.17650402e-01 9.11410451e-01 -9.15675819e-01
6.77618861e-01 5.12461722e-01 9.09055620e-02 -1.31765342e+00
-6.50305867e-01 -6.38191104e-01 -3.63414317e-01 9.20180380e-01
6.18665576e-01 -6.19245768e-01 8.83447707e-01 4.01167095e-01
-8.67257044e-02 -9.17860210e-01 -4.03684080e-01 -4.13222909e-01
6.50174916e-01 -1.05697058e-01 4.48312312e-02 1.15225768e+00
2.98611492e-01 9.57400680e-01 -4.38152462e-01 -3.61538380e-01
4.49131221e-01 2.21000895e-01 3.97916108e-01 -1.38162255e+00
-4.33238357e-01 -6.94151402e-01 8.71890634e-02 -8.69002819e-01
9.43646654e-02 -8.42529655e-01 -2.06179693e-01 -1.53814030e+00
3.96214575e-02 -1.15166210e-01 -1.31499231e-01 5.77087641e-01
-2.03374892e-01 4.46353368e-02 6.75429329e-02 1.04064085e-01
-7.05049396e-01 3.34000856e-01 8.60361457e-01 -1.59136370e-01
-1.65835679e-01 2.93223292e-01 -7.64181316e-01 1.05390871e+00
1.41647148e+00 -5.19034088e-01 -1.21869139e-01 -3.63355339e-01
7.83370137e-01 -2.42001146e-01 9.61673707e-02 -8.55854928e-01
1.61640108e-01 -1.37339860e-01 3.89695317e-01 -6.79401457e-01
2.70135343e-01 -2.10285589e-01 -3.45209628e-01 7.49767721e-01
-7.08913565e-01 2.51532227e-01 2.48048499e-01 3.65794271e-01
-1.76189825e-01 -1.67688593e-01 6.90675378e-01 -4.58702564e-01
-5.59980035e-01 -1.04173116e-01 -8.06461513e-01 1.60008058e-01
6.08155012e-01 -2.56427228e-01 -6.29365742e-01 -3.22146177e-01
-8.11198652e-01 3.28383058e-01 -1.88142806e-01 5.42926073e-01
4.00246024e-01 -9.21246290e-01 -8.47392559e-01 -3.37035984e-01
-2.48556510e-01 -5.04803419e-01 -1.62793487e-01 1.20377576e+00
-4.71554667e-01 5.86084485e-01 -2.01055318e-01 -4.35635507e-01
-1.29948485e+00 3.24930698e-01 3.95945251e-01 -5.99750817e-01
-8.04660082e-01 1.05713058e+00 -2.13768750e-01 -5.40356994e-01
2.11993203e-01 -8.18001404e-02 -6.08895957e-01 3.52786392e-01
4.37726110e-01 1.17402688e-01 1.46969478e-03 -3.50961208e-01
-6.51831105e-02 6.47215964e-03 -5.35471380e-01 -8.83178040e-02
1.41857862e+00 -3.63097079e-02 2.65790801e-02 5.96270382e-01
8.99448276e-01 -1.31755546e-01 -9.95660484e-01 -5.69392681e-01
6.15914285e-01 1.18392497e-01 7.70346895e-02 -9.43637609e-01
-4.13618267e-01 9.80230153e-01 2.48923630e-01 4.73071784e-01
4.04345453e-01 -2.32964784e-01 7.87013233e-01 5.49600899e-01
5.57494350e-02 -1.35160840e+00 4.92661744e-01 1.10483348e+00
6.57462656e-01 -1.22888589e+00 -1.16931356e-01 -8.84358585e-02
-4.26548034e-01 1.20319057e+00 6.88644886e-01 -2.96128035e-01
6.92975521e-01 2.56657809e-01 4.20784593e-01 -2.82921493e-01
-9.78206098e-01 1.04978293e-01 3.14242244e-02 2.08847314e-01
8.89627397e-01 -3.73962075e-02 -5.71535349e-01 3.39731485e-01
-5.36957085e-01 1.43754967e-02 2.23957285e-01 1.14283526e+00
-9.39673364e-01 -1.10094440e+00 -1.86399862e-01 6.13169909e-01
-7.99303532e-01 -4.36818510e-01 -7.59196699e-01 8.23501825e-01
-3.07604088e-04 8.97322178e-01 -1.88009068e-01 -2.93972671e-01
3.36053997e-01 4.51781005e-01 3.39047372e-01 -6.59735978e-01
-1.36921751e+00 1.62024066e-01 6.18106961e-01 -3.82543921e-01
-6.06164157e-01 -6.86520696e-01 -7.90594280e-01 -8.31745684e-01
-4.08443868e-01 2.92212129e-01 2.53679723e-01 1.07151115e+00
-1.69770643e-01 5.16739607e-01 2.38552630e-01 -5.93994081e-01
-1.24243295e+00 -1.28547597e+00 1.22772813e-01 2.05257684e-02
2.77199924e-01 -4.75965351e-01 -3.44105698e-02 -1.81096777e-01] | [8.942431449890137, 10.055582046508789] |
3dab00ae-5746-4c2f-897a-ac37298a038a | c2f-tcn-a-framework-for-semi-and-fully | 2212.11078 | null | https://arxiv.org/abs/2212.11078v1 | https://arxiv.org/pdf/2212.11078v1.pdf | C2F-TCN: A Framework for Semi and Fully Supervised Temporal Action Segmentation | Temporal action segmentation tags action labels for every frame in an input untrimmed video containing multiple actions in a sequence. For the task of temporal action segmentation, we propose an encoder-decoder-style architecture named C2F-TCN featuring a "coarse-to-fine" ensemble of decoder outputs. The C2F-TCN framework is enhanced with a novel model agnostic temporal feature augmentation strategy formed by the computationally inexpensive strategy of the stochastic max-pooling of segments. It produces more accurate and well-calibrated supervised results on three benchmark action segmentation datasets. We show that the architecture is flexible for both supervised and representation learning. In line with this, we present a novel unsupervised way to learn frame-wise representation from C2F-TCN. Our unsupervised learning approach hinges on the clustering capabilities of the input features and the formation of multi-resolution features from the decoder's implicit structure. Further, we provide the first semi-supervised temporal action segmentation results by merging representation learning with conventional supervised learning. Our semi-supervised learning scheme, called ``Iterative-Contrastive-Classify (ICC)'', progressively improves in performance with more labeled data. The ICC semi-supervised learning in C2F-TCN, with 40% labeled videos, performs similar to fully supervised counterparts. | ['Angela Yao', 'Rahul Rahaman', 'Dipika Singhania'] | 2022-12-20 | null | null | null | null | ['action-segmentation'] | ['computer-vision'] | [ 9.28095281e-01 3.32103103e-01 -6.49810374e-01 -5.12669444e-01
-1.10449743e+00 -5.67750692e-01 7.56370783e-01 -3.50196630e-01
-3.51234227e-01 5.30102074e-01 6.55646503e-01 1.49455234e-01
-1.35170831e-03 -3.01769018e-01 -8.32930326e-01 -7.55296588e-01
-7.18969554e-02 4.89538521e-01 5.60253799e-01 3.60311344e-02
-5.37078409e-03 2.59786155e-02 -1.50290298e+00 1.03424537e+00
4.45178449e-01 1.27130866e+00 1.07419468e-01 7.24670708e-01
6.84075132e-02 1.49427831e+00 -2.60258436e-01 -1.02688529e-01
3.62723827e-01 -8.41885448e-01 -1.26527953e+00 6.72852993e-01
3.10722560e-01 -2.73875326e-01 -3.24850351e-01 5.19058526e-01
3.95425297e-02 2.91725844e-01 8.04341197e-01 -1.11425078e+00
-2.51274705e-01 7.46451676e-01 -5.48094094e-01 1.47701234e-01
3.64136487e-01 2.08544597e-01 1.01289022e+00 -6.37614250e-01
9.59865034e-01 1.20020783e+00 5.79854727e-01 6.50373518e-01
-1.46609831e+00 -9.23267603e-02 3.88569772e-01 2.67537892e-01
-1.05293882e+00 -4.88537192e-01 4.78968889e-01 -7.43847191e-01
1.08734667e+00 -2.24270821e-02 6.41097188e-01 1.24797606e+00
-9.63849351e-02 1.33308554e+00 1.10663164e+00 -3.01065236e-01
2.42373258e-01 -4.37402338e-01 6.16070069e-02 5.97094417e-01
-5.46816349e-01 2.96484649e-01 -6.79574847e-01 2.93150067e-01
8.30563545e-01 -7.52016604e-02 -1.44168690e-01 -3.87585729e-01
-1.44760239e+00 5.72303593e-01 1.56232581e-01 3.49389166e-01
-1.58445776e-01 6.12657726e-01 8.51543784e-01 2.38116235e-01
4.54836696e-01 2.11014628e-01 -7.65222549e-01 -7.13152111e-01
-1.22557712e+00 -2.38709554e-01 3.81773353e-01 1.08248663e+00
7.17837393e-01 -5.98587235e-03 -4.83847231e-01 6.26846492e-01
3.17460038e-02 1.34910002e-01 6.89287066e-01 -1.73572552e+00
3.23073924e-01 5.14162302e-01 -8.02957192e-02 -2.73210645e-01
-3.18846494e-01 -9.73744094e-02 -6.76063120e-01 1.41674489e-01
6.14247322e-01 -1.02624446e-01 -1.17784333e+00 1.80860090e+00
-1.48163959e-02 4.33660299e-01 1.31150261e-01 6.74336851e-01
4.14278626e-01 5.35323262e-01 2.15328485e-01 -4.33581948e-01
9.80174422e-01 -1.39249587e+00 -7.57545650e-01 -4.02901433e-02
9.36423838e-01 -4.02916610e-01 7.14894891e-01 3.79673421e-01
-9.95517254e-01 -9.01637912e-01 -6.81612313e-01 -1.36720538e-01
-1.86060295e-01 5.26410460e-01 5.02850473e-01 2.71633565e-01
-1.09321082e+00 9.21793878e-01 -1.16224253e+00 -4.44635540e-01
6.96560383e-01 3.87352794e-01 -4.99316990e-01 1.41192630e-01
-8.70487332e-01 6.20674014e-01 7.04997301e-01 -2.25100596e-03
-1.32170749e+00 -3.53907198e-01 -1.04943907e+00 -2.90731281e-01
7.03041494e-01 -3.63457620e-01 1.53198886e+00 -1.57343662e+00
-1.62277758e+00 9.56353128e-01 -3.11202019e-01 -8.48850369e-01
6.02098346e-01 -2.38189608e-01 -2.07771286e-01 6.03362858e-01
3.37258309e-01 1.20464754e+00 1.08914542e+00 -1.11237967e+00
-7.73175120e-01 -9.39430948e-03 2.92102750e-02 4.30545164e-03
2.24312946e-01 -3.50951427e-03 -5.11892557e-01 -8.94705057e-01
-2.51443079e-03 -9.92998481e-01 -4.74390328e-01 4.79790233e-02
-3.31513703e-01 -1.35979816e-01 8.80575955e-01 -5.77759683e-01
1.13065541e+00 -2.35966778e+00 2.96150923e-01 -7.11887330e-02
5.61937243e-02 6.57982603e-02 -2.33938813e-01 2.69866586e-01
-3.42024148e-01 -1.84894264e-01 -5.82557678e-01 -5.10228634e-01
-8.45650882e-02 5.78604460e-01 -1.82420611e-01 4.54441100e-01
3.62582207e-01 1.03720999e+00 -1.04635382e+00 -7.86462486e-01
2.65221536e-01 1.03210822e-01 -5.38509548e-01 2.12473795e-01
-5.57244658e-01 8.32344472e-01 -3.94198060e-01 6.56721890e-01
1.11217007e-01 -3.85124981e-01 2.57712334e-01 -1.71369150e-01
-1.34794667e-01 1.54888436e-01 -8.81783903e-01 2.35022783e+00
9.14168078e-03 6.86397791e-01 -2.48772502e-01 -1.21020651e+00
5.81084609e-01 4.49518561e-01 1.13715160e+00 -6.22644484e-01
7.52471462e-02 1.17598206e-01 -3.54082048e-01 -6.41470373e-01
2.93329477e-01 1.29964715e-02 -3.04660261e-01 5.75019240e-01
5.36632597e-01 2.83565879e-01 6.08397901e-01 2.43793845e-01
1.15356636e+00 1.04318011e+00 3.11869621e-01 -1.07393131e-01
5.64233661e-01 1.66249216e-01 8.10322225e-01 4.88300592e-01
-4.65917528e-01 9.14891303e-01 7.26215303e-01 -2.91942894e-01
-7.91481316e-01 -9.71425533e-01 -5.39414771e-02 1.39683402e+00
-1.65397033e-01 -6.74401164e-01 -1.03703904e+00 -1.08941984e+00
-2.94874728e-01 3.13236505e-01 -9.45896864e-01 -1.15083657e-01
-7.20793307e-01 -2.41914913e-01 6.01499975e-01 1.10181940e+00
6.43965602e-01 -1.14571905e+00 -6.61192119e-01 2.99543560e-01
-4.57758904e-01 -1.40377593e+00 -6.74682915e-01 5.39723873e-01
-9.77258742e-01 -1.28159523e+00 -7.25408018e-01 -8.02012324e-01
5.98806798e-01 -5.58959320e-02 9.61413860e-01 -3.56177151e-01
-8.24913755e-03 6.12280726e-01 -7.29949355e-01 2.63912171e-01
-4.83270884e-01 -1.95605513e-02 -1.00267999e-01 3.99155229e-01
3.16133231e-01 -5.45515656e-01 -4.69858944e-01 5.90481818e-01
-7.97267497e-01 2.69199580e-01 5.91525018e-01 6.62807941e-01
9.39474761e-01 -2.02442631e-01 3.86431307e-01 -8.89886379e-01
-1.26177683e-01 -2.90457964e-01 -2.60911047e-01 1.04958475e-01
-2.28567570e-01 2.39917547e-01 6.52685463e-01 -4.95103776e-01
-1.06896019e+00 8.06185186e-01 -3.62295769e-02 -6.90411866e-01
-3.79015952e-01 2.33288020e-01 -6.55393675e-02 1.91424474e-01
6.83974922e-01 3.68709743e-01 1.09160960e-01 -5.98326623e-01
6.74427450e-01 4.36206222e-01 8.71021390e-01 -5.19566119e-01
4.79796350e-01 6.63057983e-01 -1.78699747e-01 -5.00637829e-01
-9.95414972e-01 -6.12304091e-01 -1.38756108e+00 -5.64921141e-01
1.45158231e+00 -9.33521271e-01 -3.63019317e-01 6.07237518e-01
-9.36447978e-01 -9.39676404e-01 -9.73150313e-01 5.68358541e-01
-1.40666568e+00 4.92273480e-01 -6.90751493e-01 -5.00507057e-01
2.71644205e-01 -1.11071253e+00 1.29756618e+00 -1.67668566e-01
-3.04551333e-01 -8.77457917e-01 2.67060935e-01 6.30403817e-01
-6.86280131e-02 4.12318170e-01 5.58439791e-01 -7.24775612e-01
-5.29562533e-01 1.85999140e-01 -9.66334641e-02 7.55492687e-01
1.93200469e-01 -7.37771913e-02 -9.45049405e-01 -1.45438537e-01
-2.56715238e-01 -6.57132804e-01 1.19567847e+00 5.76326966e-01
1.30796778e+00 -1.05702542e-01 -2.04023317e-01 5.01513362e-01
1.22068930e+00 1.97581798e-01 7.56416798e-01 2.06057861e-01
8.25583756e-01 6.08749926e-01 8.57796371e-01 4.23737437e-01
1.88401163e-01 7.35655546e-01 2.55066365e-01 -1.15465326e-02
-3.97625506e-01 -3.42987090e-01 8.69615674e-01 6.54532552e-01
-4.90688562e-01 5.35536297e-02 -5.78047335e-01 5.36883175e-01
-2.19851732e+00 -1.17285800e+00 -4.56788465e-02 1.85079551e+00
9.23846722e-01 1.81142673e-01 5.28183937e-01 2.89475530e-01
6.69408500e-01 3.72892261e-01 -5.40167987e-01 -2.94836491e-01
-8.06006342e-02 1.82048008e-01 4.16706115e-01 1.93470076e-01
-1.53571939e+00 1.18548214e+00 6.30923843e+00 8.59554112e-01
-6.88844025e-01 2.19480574e-01 6.06939495e-01 -4.19642180e-02
1.61623597e-01 1.17104001e-01 -5.90321422e-01 4.83776063e-01
1.19329143e+00 2.54396290e-01 1.69522762e-01 7.41397440e-01
6.18403435e-01 -2.19101340e-01 -1.42434287e+00 8.64183903e-01
4.13296595e-02 -1.44214225e+00 -6.02976698e-03 -6.11963943e-02
8.80028844e-01 -8.99172425e-02 -1.65532529e-01 3.72962058e-01
3.83201689e-01 -9.20760572e-01 9.41995502e-01 5.32617331e-01
1.04196596e+00 -3.91919881e-01 4.42402631e-01 1.92767963e-01
-1.55100238e+00 -3.17311823e-01 6.47019297e-02 5.28702512e-02
5.39994776e-01 1.48478091e-01 -2.47289866e-01 4.76636410e-01
5.75574636e-01 1.58248425e+00 -5.97065926e-01 7.52411783e-01
-2.43289158e-01 7.27358818e-01 -6.13122769e-02 6.85587585e-01
6.21363521e-01 -2.09440067e-01 2.08086699e-01 1.42809248e+00
-2.92302761e-02 9.97884795e-02 4.80832398e-01 4.41956997e-01
8.69462639e-02 -1.69689789e-01 -5.10243535e-01 -1.91431046e-01
-1.17128879e-01 8.34848225e-01 -9.57381129e-01 -5.58043182e-01
-4.45656359e-01 1.15690851e+00 1.48963839e-01 3.60414863e-01
-9.88158584e-01 2.87103862e-01 3.55298072e-01 1.37130216e-01
6.83284760e-01 -2.06893861e-01 -4.47143912e-02 -1.07961214e+00
-1.61620930e-01 -8.96118462e-01 6.24755859e-01 -7.48447299e-01
-8.52977276e-01 5.31561732e-01 1.79992810e-01 -1.67280006e+00
-6.62142634e-01 -5.54236889e-01 -2.42257893e-01 4.79441732e-02
-1.19731545e+00 -1.32953393e+00 -5.22317700e-02 1.01942551e+00
1.02122414e+00 -1.60963908e-01 7.05407858e-01 2.27798983e-01
-4.63711560e-01 4.01401401e-01 6.23048022e-02 4.63906229e-01
5.88067472e-01 -1.35944927e+00 8.70907158e-02 9.48056459e-01
2.59582281e-01 -7.41739571e-02 2.87532687e-01 -5.39374590e-01
-8.98016870e-01 -1.57388365e+00 6.38989866e-01 -4.59836125e-01
6.68388069e-01 -2.63805747e-01 -5.63743889e-01 1.10309684e+00
1.38415068e-01 1.49835259e-01 7.71453619e-01 -3.88373315e-01
-2.72402436e-01 1.66610293e-02 -8.11246753e-01 3.38282138e-01
1.33481884e+00 -7.62426317e-01 -6.97716832e-01 4.82452899e-01
7.50478029e-01 -2.24190637e-01 -9.99814808e-01 3.97217035e-01
4.46845531e-01 -1.02396822e+00 8.18750441e-01 -7.06792533e-01
7.53734410e-01 -3.60251456e-01 -3.38265061e-01 -7.23743856e-01
-2.97718555e-01 -8.71503592e-01 -2.85052240e-01 9.90729809e-01
3.81898165e-01 -6.88295364e-02 9.43579793e-01 1.81029975e-01
-4.47197288e-01 -7.20337391e-01 -1.04873955e+00 -8.40069115e-01
-1.00221679e-01 -6.85342431e-01 -1.13958530e-01 6.63095355e-01
-4.72736619e-02 4.80056629e-02 -4.67237592e-01 -3.43336314e-01
5.05870700e-01 1.42732129e-01 5.45203209e-01 -8.51130366e-01
-4.89380687e-01 -1.98279560e-01 -6.26648307e-01 -1.61153126e+00
3.96307141e-01 -9.01716352e-01 2.45387986e-01 -1.31646156e+00
1.92165986e-01 -3.20083648e-03 -2.30504274e-01 8.13855529e-01
2.93306977e-01 4.84981537e-01 1.69654056e-01 4.25049841e-01
-1.20715213e+00 5.26416361e-01 1.40636277e+00 -1.53168336e-01
-1.69238627e-01 1.37956619e-01 -1.38670221e-01 8.79952967e-01
5.19109309e-01 -4.07053411e-01 -5.47843516e-01 -2.34616593e-01
-4.27667260e-01 1.72644466e-01 2.02627689e-01 -1.16579866e+00
5.56780351e-03 -1.41861856e-01 2.22622633e-01 -6.49991930e-01
4.15146947e-01 -7.58054018e-01 6.60726726e-02 3.40960443e-01
-6.97675347e-01 -3.27956170e-01 -1.91704229e-01 7.28795588e-01
-4.94532973e-01 -5.69757149e-02 8.79718244e-01 -3.53467911e-01
-1.31265295e+00 2.22451776e-01 -9.04426515e-01 2.57772207e-01
1.19776034e+00 -6.59886062e-01 -1.39922634e-01 -3.56208295e-01
-1.39902759e+00 2.15225458e-01 3.66465062e-01 2.09974542e-01
4.35184509e-01 -1.29770076e+00 -4.65286165e-01 2.11433973e-02
1.14413209e-01 -9.37858317e-03 1.83180436e-01 1.26054525e+00
-2.57600993e-01 4.15279359e-01 -3.28470260e-01 -9.18277025e-01
-1.10243213e+00 5.72104156e-01 4.07111615e-01 -4.23529863e-01
-7.52311945e-01 6.69414461e-01 1.12919502e-01 -1.03968255e-01
3.13604474e-01 -5.36530435e-01 -3.63208950e-01 3.56909364e-01
2.59015113e-01 4.04872626e-01 -2.88316309e-01 -1.09679067e+00
-1.39543295e-01 6.38528943e-01 -2.14443430e-02 -2.00196326e-01
1.21653366e+00 -1.38833821e-01 9.71616805e-02 5.37577212e-01
1.40833187e+00 -5.92057467e-01 -1.99786556e+00 -2.35111341e-01
2.25191072e-01 -2.22161993e-01 -2.67050743e-01 -6.27732873e-01
-1.26833212e+00 6.43695891e-01 4.77151722e-01 -1.28825739e-01
1.37867177e+00 2.25597337e-01 7.08570480e-01 2.54506141e-01
4.22266215e-01 -1.32179463e+00 4.70223010e-01 6.17372453e-01
5.40260971e-01 -1.13473797e+00 -1.62630081e-01 -4.27152514e-01
-1.01869476e+00 1.18324852e+00 5.92058778e-01 -5.24738804e-02
4.63484645e-01 2.72318959e-01 1.60103753e-01 -1.35987595e-01
-7.92878151e-01 -6.40191317e-01 2.20243245e-01 7.21308410e-01
2.25310773e-01 -1.70144424e-01 -1.67364448e-01 5.88154495e-01
1.75987154e-01 3.49643141e-01 3.55813980e-01 1.15333247e+00
-2.93403238e-01 -1.21755791e+00 9.51581821e-02 4.34168339e-01
-2.18862802e-01 1.90657049e-01 -3.21030974e-01 7.32198358e-01
3.57142895e-01 7.87824750e-01 1.75186127e-01 -5.42588174e-01
6.87643811e-02 1.84416607e-01 5.03048837e-01 -7.13879824e-01
-5.53503871e-01 4.87111688e-01 2.73137987e-01 -1.14414763e+00
-1.17858982e+00 -9.82466996e-01 -1.54512084e+00 2.57490963e-01
1.44907475e-01 -2.23104190e-02 1.38285486e-02 1.28549027e+00
1.75670624e-01 5.39232254e-01 6.58003807e-01 -1.02283967e+00
-2.56871819e-01 -8.40304136e-01 -5.67159534e-01 6.27949238e-01
2.75805801e-01 -7.05677509e-01 -2.05106318e-01 8.73970509e-01] | [8.467888832092285, 0.5659775733947754] |
e9392c0a-3939-43f0-8242-c157e8565c60 | cheater-s-bowl-human-vs-computer-search | 2212.03296 | null | https://arxiv.org/abs/2212.03296v1 | https://arxiv.org/pdf/2212.03296v1.pdf | Cheater's Bowl: Human vs. Computer Search Strategies for Open-Domain Question Answering | For humans and computers, the first step in answering an open-domain question is retrieving a set of relevant documents from a large corpus. However, the strategies that computers use fundamentally differ from those of humans. To better understand these differences, we design a gamified interface for data collection -- Cheater's Bowl -- where a human answers complex questions with access to both traditional and modern search tools. We collect a dataset of human search sessions, analyze human search strategies, and compare them to state-of-the-art multi-hop QA models. Humans query logically, apply dynamic search chains, and use world knowledge to boost searching. We demonstrate how human queries can improve the accuracy of existing systems and propose improving the future design of QA models. | ['Jordan Boyd-Graber', 'Andrew Mao', 'Wanrong He'] | 2022-11-15 | null | null | null | null | ['open-domain-question-answering'] | ['natural-language-processing'] | [-1.17992200e-01 1.76528454e-01 -8.80479738e-02 -3.14811707e-01
-1.30104685e+00 -9.33402538e-01 4.96801406e-01 1.38145894e-01
-5.81442058e-01 5.83834589e-01 3.30728650e-01 -8.44835579e-01
-5.27963459e-01 -8.03597391e-01 -1.23916134e-01 2.89511889e-01
3.82091880e-01 1.28373647e+00 7.33919919e-01 -9.63310540e-01
8.31599236e-01 -8.43273401e-02 -1.26807141e+00 5.13607323e-01
1.12973869e+00 5.57343423e-01 4.23957437e-01 8.07119608e-01
-4.12074268e-01 1.01128304e+00 -6.90898180e-01 -6.57111645e-01
1.10632710e-01 -3.47067952e-01 -1.82851660e+00 -8.84881377e-01
4.34730738e-01 -3.38429749e-01 -2.09758461e-01 7.71594465e-01
5.47758996e-01 5.02374709e-01 6.22859597e-01 -9.88159955e-01
-1.11554158e+00 2.91362405e-01 1.41977936e-01 5.58049798e-01
1.24035645e+00 3.61060232e-01 1.10897171e+00 -3.75625402e-01
8.28491926e-01 1.43057799e+00 4.13249761e-01 6.74029231e-01
-1.06895614e+00 -3.10552031e-01 -4.12271529e-01 6.22933447e-01
-1.01153743e+00 -4.50677067e-01 4.50150043e-01 -2.76989341e-01
1.21502233e+00 4.76425469e-01 3.58671129e-01 7.02856660e-01
-6.59727603e-02 6.94057226e-01 8.77271593e-01 -7.86336422e-01
1.97523266e-01 1.57482252e-01 7.31607080e-01 7.93426991e-01
-5.61700985e-02 -7.82833546e-02 -6.62855506e-01 -6.53730810e-01
3.39316815e-01 -3.89465541e-01 -2.61421591e-01 -1.73636302e-01
-1.01533747e+00 9.03704464e-01 3.42737257e-01 3.09439808e-01
-4.21508044e-01 -1.52231038e-01 -2.64832564e-02 5.98223925e-01
-3.68033536e-02 1.30081260e+00 -4.91046667e-01 -5.80200016e-01
-6.28464699e-01 7.82881677e-01 1.55620015e+00 9.26566482e-01
7.21622586e-01 -1.03552485e+00 -4.68973637e-01 9.75729585e-01
4.12848175e-01 4.54520822e-01 5.19502342e-01 -1.58035254e+00
5.00727594e-01 9.35488343e-01 5.48743010e-01 -9.23896670e-01
-9.87660512e-02 6.47569671e-02 1.69509709e-01 -2.02770427e-01
8.07463706e-01 7.91628659e-02 -7.03191161e-01 1.21055698e+00
1.92407161e-01 -8.13983798e-01 -1.27850056e-01 9.90501583e-01
6.63338423e-01 5.06283164e-01 3.06449234e-01 1.94368467e-01
1.67602026e+00 -1.05806911e+00 -7.13341355e-01 -5.73695540e-01
6.75783813e-01 -7.10334182e-01 1.71600556e+00 4.05356854e-01
-1.41617692e+00 -2.90546447e-01 -7.04331815e-01 -8.92628074e-01
-5.80630541e-01 -5.21165669e-01 5.05777657e-01 6.55029476e-01
-1.09836042e+00 3.18961740e-01 -3.83580834e-01 -6.25754237e-01
5.99023513e-02 9.93597433e-02 2.55653471e-01 -4.16044205e-01
-1.55959129e+00 1.40238535e+00 -6.39002956e-03 -4.96243358e-01
-4.99601096e-01 -6.90610766e-01 -4.62760329e-01 4.56523061e-01
6.72445118e-01 -1.19154084e+00 2.17756248e+00 1.33289676e-02
-1.15481007e+00 8.04471910e-01 -3.49017054e-01 -8.60876516e-02
1.31215036e-01 -3.63818914e-01 -9.78216305e-02 3.67551982e-01
3.85988712e-01 4.96531844e-01 2.33726501e-01 -1.00827110e+00
-6.67279959e-01 -6.08046472e-01 5.75765550e-01 3.36958557e-01
-2.19288662e-01 3.59717518e-01 -5.18115044e-01 2.74835099e-02
-1.90959871e-01 -5.49476206e-01 -2.59487957e-01 -1.09315276e-01
1.41890422e-01 -6.81493819e-01 5.35637736e-01 -9.99826491e-01
1.72101700e+00 -1.56225479e+00 -7.61069953e-02 2.33553723e-01
2.30762854e-01 2.02597842e-01 -1.74160346e-01 8.63345861e-01
5.83502352e-01 3.55239332e-01 3.24000716e-01 1.95521235e-01
5.31526089e-01 4.89717629e-03 -1.95261180e-01 -5.27169287e-01
-4.38366592e-01 1.36911404e+00 -1.15368617e+00 -9.15906847e-01
-2.50803262e-01 -2.17148960e-01 -5.93186200e-01 3.36172163e-01
-9.23419535e-01 -1.58620059e-01 -7.57921994e-01 8.34960520e-01
6.13968782e-02 -6.85320497e-01 -1.78570133e-02 2.05885768e-01
2.43859321e-01 9.10683572e-01 -6.77939057e-01 1.94548929e+00
-5.42541683e-01 4.46241021e-01 1.84655208e-02 -2.69428849e-01
5.12404323e-01 1.32675275e-01 -7.08540753e-02 -1.31184781e+00
-2.34840021e-01 1.84202105e-01 -2.35111210e-02 -9.54242945e-01
7.14680970e-01 2.19546497e-01 -8.24668352e-03 1.02879715e+00
-1.13432698e-01 -5.43498039e-01 5.52848041e-01 6.05606198e-01
1.42160428e+00 -2.11839259e-01 3.01821709e-01 -1.33945167e-01
1.78507328e-01 8.80841911e-01 -2.32964307e-01 1.36449981e+00
-1.40159637e-01 -1.78504027e-02 2.68684894e-01 -3.85602444e-01
-7.52287447e-01 -9.04299378e-01 3.44310969e-01 1.81216967e+00
2.35218540e-01 -4.81874138e-01 -1.04978251e+00 -6.38096511e-01
-1.61553845e-01 1.11152148e+00 -4.30313386e-02 1.67551283e-02
-5.14290452e-01 2.07460709e-02 7.43073821e-01 3.16922188e-01
5.77977240e-01 -1.27784514e+00 -1.13295662e+00 1.52440101e-01
-9.87732649e-01 -5.33457994e-01 -7.07883537e-01 -1.81211159e-01
-7.64641762e-01 -1.31105161e+00 -4.51996654e-01 -8.23234379e-01
4.51076142e-02 4.21007484e-01 1.89612448e+00 7.04235792e-01
-2.85828888e-01 8.34085941e-01 -6.06312156e-01 -4.36651707e-01
-4.23810601e-01 2.58522987e-01 -5.27611792e-01 -1.18156230e+00
1.02259493e+00 -2.78217524e-01 -8.51168334e-01 6.44406080e-01
-5.91127098e-01 -1.71137527e-01 3.70301396e-01 7.09912181e-01
-6.11158982e-02 -2.36684099e-01 3.86413097e-01 -6.12715542e-01
1.79428887e+00 -5.56690156e-01 -4.70703512e-01 9.69201326e-01
-1.09706295e+00 1.78336501e-01 1.04442649e-01 -1.49195522e-01
-1.28161407e+00 -5.81655145e-01 1.27080128e-01 2.51937211e-01
-1.88021779e-01 7.13209391e-01 1.70224845e-01 -1.91250160e-01
1.32255781e+00 2.53533460e-02 -9.17442590e-02 -6.49233520e-01
6.83270931e-01 9.08482432e-01 6.16464376e-01 -9.93984878e-01
4.51389551e-01 -8.20215121e-02 -5.54321706e-01 -4.33221161e-01
-6.58769429e-01 -9.39197302e-01 -1.07661024e-01 -8.57568681e-02
5.68036735e-01 -2.59978831e-01 -1.29089284e+00 -7.46310204e-02
-1.36603439e+00 -6.86969876e-01 -2.04990044e-01 -1.74602985e-01
-6.69746816e-01 4.33390588e-01 -6.94429338e-01 -9.10216093e-01
-7.11911798e-01 -7.42683530e-01 1.01428008e+00 6.07691288e-01
-8.26263845e-01 -9.13849890e-01 4.55547094e-01 1.29883289e+00
6.70020282e-01 -8.01134527e-01 1.37808859e+00 -1.06747115e+00
-7.71219671e-01 -1.29080907e-01 -1.67496502e-01 -2.84288853e-01
-2.22063035e-01 -6.21138513e-01 -6.06374741e-01 1.27934247e-01
-6.63316548e-02 -7.20308900e-01 2.48530790e-01 -8.79022256e-02
8.83581758e-01 -4.08759713e-01 -5.39772034e-01 -4.88060899e-02
1.00883818e+00 4.79157239e-01 6.79131925e-01 6.42937541e-01
-2.59379029e-01 8.07164192e-01 7.52365828e-01 8.95905718e-02
7.71582127e-01 6.56220078e-01 -7.42718205e-02 5.54627478e-01
2.76584681e-02 -6.16669655e-01 -1.80703849e-01 1.90266743e-01
-6.55726492e-02 -4.10105407e-01 -1.41152275e+00 8.44158828e-01
-1.83382404e+00 -1.04695153e+00 2.08531275e-01 1.82129264e+00
1.10299671e+00 -5.87304980e-02 -3.85499699e-03 -3.05339754e-01
1.47102654e-01 -2.48564333e-01 -5.24131358e-01 -2.96481937e-01
4.65623677e-01 6.07160747e-01 -5.38249947e-02 7.28366435e-01
-3.05300683e-01 9.83230770e-01 7.64025021e+00 7.42863417e-01
-2.25294858e-01 1.13532611e-03 2.92477012e-01 9.84927490e-02
-4.27028567e-01 1.61534145e-01 -4.83335018e-01 3.69217873e-01
1.09173501e+00 -5.24039447e-01 8.80155265e-01 9.68913972e-01
-1.81494877e-02 -5.70352316e-01 -1.42255139e+00 7.61929572e-01
8.67165253e-02 -1.35062766e+00 1.27291784e-01 5.11219762e-02
3.13306332e-01 -2.81821102e-01 -2.04187021e-01 7.77684271e-01
8.81588042e-01 -1.19620085e+00 1.66248947e-01 8.10104549e-01
2.52721369e-01 -1.35980904e-01 4.59540725e-01 1.02843618e+00
-4.85587090e-01 -3.01996648e-01 -1.53861359e-01 -3.12345535e-01
4.52679425e-01 -1.99456245e-01 -1.05746877e+00 3.39534700e-01
8.04458499e-01 -3.37713569e-01 -8.15460980e-01 1.03801835e+00
-1.05973579e-01 3.10920566e-01 -4.10897434e-01 -7.20725834e-01
-5.40532432e-02 -1.29268348e-01 2.26853460e-01 8.65093172e-01
1.99669287e-01 7.15222836e-01 -5.06799333e-02 1.07705927e+00
6.04785457e-02 -1.40960924e-02 -4.84546214e-01 -1.54133067e-01
8.28270435e-01 8.43894064e-01 -1.03128731e-01 -4.05361146e-01
-3.07726115e-01 8.25155258e-01 3.83343220e-01 4.00068074e-01
-2.28063777e-01 -6.83016658e-01 3.44264299e-01 3.06331187e-01
-3.45118433e-01 -5.07592112e-02 -3.57797980e-01 -9.22769129e-01
2.20082670e-01 -1.64794838e+00 8.71060252e-01 -1.47121119e+00
-1.39318037e+00 2.35703230e-01 2.66275853e-01 -3.91210526e-01
-9.52027380e-01 -4.26466763e-01 -5.56151628e-01 1.12588704e+00
-1.25028241e+00 -8.69239330e-01 -5.72090030e-01 6.24034405e-01
7.57728755e-01 7.04889521e-02 9.12862480e-01 1.20763138e-01
3.50351810e-01 2.72535354e-01 -3.27353984e-01 5.96842952e-02
8.58853221e-01 -1.18843412e+00 7.61131406e-01 2.96576411e-01
2.18548447e-01 1.31500161e+00 6.61873937e-01 -8.02364588e-01
-1.48167074e+00 1.55383021e-01 1.24937189e+00 -1.15838838e+00
6.90306246e-01 5.89303896e-02 -9.98221338e-01 5.38471401e-01
5.62947929e-01 -7.33294725e-01 8.37273121e-01 4.71445441e-01
-2.28064299e-01 3.39853793e-01 -1.16034865e+00 8.41168880e-01
1.14514947e+00 -8.56713414e-01 -1.60316956e+00 5.03415644e-01
7.96750367e-01 -3.94547343e-01 -5.56219935e-01 -1.80290230e-02
7.67172337e-01 -7.67841816e-01 1.29307055e+00 -1.07008088e+00
3.56184304e-01 -5.90457721e-03 -1.14302322e-01 -1.40304053e+00
-2.49755681e-01 -8.29040110e-01 -1.75270423e-01 4.96403188e-01
5.14238417e-01 -4.37896937e-01 7.92061269e-01 1.60565042e+00
1.92903414e-01 -4.80870098e-01 -6.89603448e-01 -5.00315070e-01
2.08623171e-01 -2.55723447e-01 6.23155475e-01 6.73680663e-01
9.03412580e-01 5.44999480e-01 2.47058645e-01 -1.11250311e-01
4.47903335e-01 2.22405400e-02 8.76867115e-01 -1.33034229e+00
-2.86649138e-01 -4.29780632e-01 4.07542974e-01 -1.75601816e+00
-2.84398645e-01 -5.25041282e-01 8.16155411e-03 -1.86394632e+00
2.06083015e-01 -3.65430087e-01 3.53836641e-02 4.48200673e-01
-3.05726528e-01 -2.89563715e-01 3.01099479e-01 2.96959221e-01
-1.08400810e+00 8.94019753e-02 1.25540423e+00 -1.95932060e-01
-2.13106826e-01 -1.41812459e-01 -1.08939981e+00 5.05550921e-01
4.99704003e-01 -2.42840156e-01 -6.20812833e-01 -7.11706102e-01
7.37658501e-01 5.66320240e-01 3.90640795e-01 -8.45489800e-01
1.07726908e+00 -4.62232560e-01 2.51467898e-03 -5.05803287e-01
2.37896860e-01 -6.34334922e-01 -3.02046776e-01 3.05534363e-01
-7.24744916e-01 2.39344761e-01 -3.24524045e-02 3.01079750e-01
-1.97787866e-01 -7.96132565e-01 1.19662598e-01 -6.40022457e-01
-6.92082763e-01 -2.50064105e-01 -3.15756947e-01 6.03033423e-01
5.30588448e-01 -8.03313702e-02 -9.78181541e-01 -8.78458023e-01
-4.10132319e-01 9.61876929e-01 3.90570402e-01 2.81183273e-01
4.28092003e-01 -8.14474702e-01 -1.81745261e-01 -2.14414269e-01
1.49739668e-01 -1.16075218e-01 -6.03344403e-02 1.23784967e-01
-6.86262071e-01 1.00466824e+00 -8.68400857e-02 -5.14045581e-02
-9.86773193e-01 5.32468915e-01 3.87591422e-01 -5.99285781e-01
4.07081209e-02 6.10228121e-01 -3.05328310e-01 -5.81256807e-01
2.62007117e-01 5.03401794e-02 -3.71387869e-01 -2.26786330e-01
6.82497799e-01 7.12585926e-01 1.13954112e-01 3.68540466e-01
-1.59037009e-01 2.61560827e-01 -1.63242459e-01 -6.61236942e-01
8.43110740e-01 -3.33777577e-01 -2.34956816e-01 -1.76774971e-02
6.22744918e-01 -2.06049412e-01 -2.79760510e-01 -3.81619602e-01
4.34856802e-01 -6.42000914e-01 -2.99169332e-01 -1.63062215e+00
4.46582921e-02 7.76839137e-01 2.94520259e-01 5.22167146e-01
8.60665858e-01 3.71690869e-01 9.65817034e-01 1.43305624e+00
7.37070262e-01 -1.29727530e+00 4.19196695e-01 5.68236589e-01
8.76085758e-01 -1.27659166e+00 -1.80701613e-01 -2.26601467e-01
-4.70564395e-01 9.59501207e-01 8.82623553e-01 5.43861687e-01
2.66639590e-01 -1.83284611e-01 4.01193082e-01 -8.25174391e-01
-1.09931171e+00 -1.43733859e-01 2.18750119e-01 6.86539769e-01
3.38040888e-01 -4.33772147e-01 -4.66594338e-01 7.08811522e-01
-4.72092956e-01 6.24698222e-01 1.36903897e-01 1.10199440e+00
-6.56767309e-01 -1.11665678e+00 -4.98979628e-01 5.34008265e-01
-1.82528928e-01 -3.34561557e-01 -7.54414678e-01 6.73640370e-01
-5.23630202e-01 1.59453785e+00 -1.68117851e-01 -1.48396239e-01
6.80503190e-01 7.39913106e-01 3.73934239e-01 -6.49990618e-01
-7.88934827e-01 -6.19042933e-01 4.64600295e-01 -7.11296737e-01
-2.53785644e-02 -3.64178210e-01 -9.06506896e-01 -3.37863922e-01
-3.44162673e-01 7.92607903e-01 5.37108660e-01 8.38702202e-01
5.40630579e-01 -1.63934529e-01 -1.46426484e-01 -1.15057088e-01
-1.24925029e+00 -1.21883321e+00 1.39018092e-02 6.53036892e-01
4.65555266e-02 -4.65987772e-01 -2.31662944e-01 1.46190912e-01] | [11.572220802307129, 7.9421539306640625] |
0a5d8b6e-f7f8-43e6-b7f7-2c346c97d5f9 | ensembling-transformers-for-cross-domain | 2212.05696 | null | https://arxiv.org/abs/2212.05696v1 | https://arxiv.org/pdf/2212.05696v1.pdf | Ensembling Transformers for Cross-domain Automatic Term Extraction | Automatic term extraction plays an essential role in domain language understanding and several natural language processing downstream tasks. In this paper, we propose a comparative study on the predictive power of Transformers-based pretrained language models toward term extraction in a multi-language cross-domain setting. Besides evaluating the ability of monolingual models to extract single- and multi-word terms, we also experiment with ensembles of mono- and multilingual models by conducting the intersection or union on the term output sets of different language models. Our experiments have been conducted on the ACTER corpus covering four specialized domains (Corruption, Wind energy, Equitation, and Heart failure) and three languages (English, French, and Dutch), and on the RSDO5 Slovenian corpus covering four additional domains (Biomechanics, Chemistry, Veterinary, and Linguistics). The results show that the strategy of employing monolingual models outperforms the state-of-the-art approaches from the related work leveraging multilingual models, regarding all the languages except Dutch and French if the term extraction task excludes the extraction of named entity terms. Furthermore, by combining the outputs of the two best performing models, we achieve significant improvements. | ['Senja Pollak', 'Antoine Doucet', 'Andraz Pelicon', 'Matej Martinc', 'Hanh Thi Hong Tran'] | 2022-12-12 | null | null | null | null | ['term-extraction'] | ['natural-language-processing'] | [-9.48735625e-02 -7.92524889e-02 -3.49951774e-01 -4.91257161e-02
-9.67225194e-01 -8.32568407e-01 1.02737427e+00 5.84536612e-01
-8.55936110e-01 8.66440296e-01 2.64398634e-01 -6.87022328e-01
-2.25761190e-01 -4.76586133e-01 -3.76991004e-01 -3.56275916e-01
-2.58653492e-01 5.28260410e-01 -1.85178414e-01 -4.00275320e-01
1.11738324e-01 3.35606158e-01 -9.81242597e-01 3.55266333e-01
9.81741190e-01 4.53857809e-01 8.16764161e-02 3.96250159e-01
-3.85587841e-01 6.86101198e-01 -5.22656024e-01 -3.96502882e-01
1.44770727e-01 -1.23780981e-01 -9.41605866e-01 -3.37237030e-01
1.58803284e-01 8.69493261e-02 5.82400225e-02 7.28073001e-01
4.75920737e-01 -6.80212453e-02 6.68218255e-01 -6.41977966e-01
-7.01353133e-01 8.50264013e-01 -3.22988123e-01 1.07083246e-01
6.29033208e-01 -4.27045599e-02 1.26154900e+00 -8.01114261e-01
1.11715817e+00 1.27632058e+00 4.47290778e-01 2.55622178e-01
-1.22499156e+00 -5.34805894e-01 4.97157544e-01 1.34246290e-01
-1.46524990e+00 -4.38086480e-01 4.00801837e-01 -6.25970662e-01
1.69397020e+00 -7.82916471e-02 2.84028590e-01 1.07419407e+00
3.77531022e-01 5.82776010e-01 1.30826354e+00 -8.60947251e-01
-1.45906219e-02 5.11273324e-01 3.25290740e-01 4.58264619e-01
3.52884233e-01 1.33548304e-01 -5.87516844e-01 -3.80379468e-01
1.19164318e-01 -5.29228151e-01 -2.55199403e-01 8.56959745e-02
-1.25796831e+00 7.15834975e-01 -2.32581869e-01 9.64890480e-01
-5.30881047e-01 -2.57746220e-01 7.95421720e-01 4.69339758e-01
9.93099511e-01 6.58158720e-01 -1.28174543e+00 1.06179476e-01
-9.46447372e-01 2.09868804e-01 1.08909857e+00 8.28010857e-01
5.19517004e-01 -1.43977463e-01 -5.12056723e-02 1.00514817e+00
2.74191618e-01 5.91999710e-01 6.07074261e-01 -3.74650985e-01
6.80871069e-01 6.87818587e-01 -2.88706105e-02 -6.50522053e-01
-5.94102859e-01 -5.02682269e-01 -4.74905074e-01 -3.36375237e-01
2.73099661e-01 -3.33998680e-01 -1.13707221e+00 1.72718990e+00
2.75407374e-01 -2.30390936e-01 4.15566653e-01 4.49805737e-01
1.06188369e+00 6.13488197e-01 6.37380183e-01 -3.35997045e-01
1.77631366e+00 -6.99322522e-01 -8.17430556e-01 -5.63082099e-01
8.15915763e-01 -9.16952550e-01 5.74273229e-01 4.97456104e-01
-7.71303236e-01 -3.56345385e-01 -7.69858837e-01 -3.89936477e-01
-1.13868153e+00 3.73479337e-01 5.26678681e-01 3.24625522e-01
-9.65194106e-01 2.26470202e-01 -4.90455478e-01 -5.11001348e-01
-1.74911305e-01 1.13375232e-01 -6.09153509e-01 -1.01175688e-01
-1.86962950e+00 1.51103246e+00 6.18029356e-01 -8.34597200e-02
-6.69697464e-01 -7.55869150e-01 -1.18560684e+00 -1.92722648e-01
2.84259319e-01 -5.52002132e-01 9.24328685e-01 -4.25520331e-01
-1.06366467e+00 1.24419725e+00 -2.68675089e-01 -5.44263780e-01
1.62545502e-01 -4.58381593e-01 -6.87058985e-01 -1.05554946e-01
3.62118304e-01 3.29627007e-01 2.16539860e-01 -8.67849171e-01
-6.52816057e-01 -3.03425193e-01 2.12145492e-01 2.18251161e-02
-3.38911533e-01 4.51880336e-01 -1.81078032e-01 -7.38532364e-01
-4.41509694e-01 -8.40301335e-01 -2.36439675e-01 -6.02808595e-01
-2.86981195e-01 -5.95957577e-01 1.18290961e-01 -1.11919200e+00
1.59861636e+00 -1.97342825e+00 4.30085361e-01 7.75133893e-02
-1.42771333e-01 2.23854706e-01 -1.97431698e-01 8.68847430e-01
-2.56458551e-01 3.68838012e-01 -2.40510315e-01 -1.76612169e-01
-7.51652420e-02 2.46412575e-01 -1.59368575e-01 2.94279873e-01
3.11073661e-01 6.87410057e-01 -9.05802667e-01 -6.42628491e-01
7.59651065e-02 4.60581332e-01 -2.08868951e-01 -2.25509852e-01
-2.73295581e-01 2.44578704e-01 -5.41481376e-01 5.41482508e-01
2.74868041e-01 1.46771073e-01 5.07027566e-01 -1.23174703e-02
-4.23782647e-01 9.60758328e-01 -8.70646179e-01 2.03787613e+00
-9.62587893e-01 4.19088662e-01 1.03624903e-01 -7.86742508e-01
6.25227988e-01 7.85419703e-01 6.18233740e-01 -7.64513373e-01
2.15404090e-02 5.24919510e-01 1.50294956e-02 -6.15177870e-01
3.70152891e-01 -3.25699180e-01 -3.99190396e-01 4.08959806e-01
4.99066472e-01 -1.19256370e-01 7.06424952e-01 5.66138402e-02
8.64481032e-01 4.06584650e-01 7.66468525e-01 -5.57398558e-01
9.50671077e-01 2.27958500e-01 3.23782563e-01 2.49009699e-01
-3.49317421e-03 8.67078081e-03 3.96753043e-01 -1.83634222e-01
-7.14811683e-01 -7.40454197e-01 -4.29436356e-01 1.51091361e+00
-4.79956716e-01 -7.93467462e-01 -5.21173954e-01 -6.33762002e-01
1.89516954e-02 1.04210067e+00 -4.26819235e-01 -1.37165934e-02
-8.26516211e-01 -8.73336852e-01 8.87522519e-01 7.84412995e-02
1.19125336e-01 -1.11830592e+00 -2.20903382e-01 3.44730258e-01
-4.93673176e-01 -1.50572610e+00 -1.80626541e-01 4.73610222e-01
-5.99642158e-01 -8.53428423e-01 -5.96893311e-01 -9.08868968e-01
3.11282165e-02 -4.79959607e-01 1.43260002e+00 -2.83253998e-01
-1.84362501e-01 1.70888454e-01 -2.86608666e-01 -6.70926392e-01
-4.68642473e-01 5.35773337e-01 3.01081110e-02 -4.52723116e-01
7.05821872e-01 -3.38297784e-01 -9.06584114e-02 -3.04446399e-01
-9.62081969e-01 -2.81971723e-01 5.80490291e-01 6.72752678e-01
3.34217042e-01 -3.27727944e-01 7.76741385e-01 -1.00181830e+00
8.45600367e-01 -7.06101060e-01 -3.54591995e-01 5.61916411e-01
-7.47941673e-01 4.02998775e-01 5.11106074e-01 -3.91087085e-01
-1.05013013e+00 -1.36584893e-01 -2.84502298e-01 3.75931174e-01
-4.39635187e-01 1.05996573e+00 -4.64699566e-02 3.60636264e-01
7.13768899e-01 2.47428313e-01 -5.94774425e-01 -8.08503747e-01
7.13453472e-01 6.83483064e-01 1.56104058e-01 -7.47566462e-01
5.79179704e-01 1.46576896e-01 -1.54354051e-01 -1.01654792e+00
-6.56381071e-01 -8.19016755e-01 -7.91902542e-01 1.68303967e-01
1.05515933e+00 -1.07001233e+00 -1.77674904e-01 4.02455211e-01
-1.52682078e+00 1.07053511e-01 3.15989833e-03 6.05362117e-01
-4.01312001e-02 3.53509963e-01 -7.64609754e-01 -7.34113276e-01
-7.02448547e-01 -8.26236188e-01 1.33930266e+00 -2.48194247e-01
-3.85539442e-01 -1.39748204e+00 4.91213113e-01 7.52129704e-02
2.26147711e-01 2.47832224e-01 1.43222594e+00 -1.11203587e+00
1.40220419e-01 3.02557647e-02 9.12273973e-02 2.75694370e-01
1.92708805e-01 -2.29784846e-01 -8.24873149e-01 -1.36331484e-01
-7.47594908e-02 -2.82339543e-01 8.97935808e-01 5.69388084e-02
1.75992399e-01 -1.23131961e-01 -3.15889955e-01 1.71711385e-01
1.44571412e+00 1.71472624e-01 2.25939065e-01 6.47795856e-01
5.06795108e-01 9.45528030e-01 5.73233664e-01 5.09238578e-02
4.20119405e-01 3.61881018e-01 -1.02183603e-01 -2.43395120e-01
1.04917549e-01 1.02352843e-01 6.55414224e-01 8.45254362e-01
-1.40878251e-02 -2.39375800e-01 -1.14873362e+00 1.14382768e+00
-1.47882283e+00 -4.34408396e-01 -1.96940109e-01 1.94449210e+00
1.31089103e+00 9.71463621e-02 -9.92270634e-02 -2.03724265e-01
1.92858607e-01 6.80253729e-02 -6.45734221e-02 -7.59512722e-01
-1.67209029e-01 7.87514627e-01 3.85283381e-01 5.13648152e-01
-1.30171323e+00 1.18454897e+00 6.28330326e+00 9.05865848e-01
-1.14547634e+00 1.27967611e-01 2.31878698e-01 9.02895853e-02
-2.53277600e-01 9.90491547e-03 -8.78648281e-01 -1.21612074e-02
1.50550008e+00 -2.64641672e-01 2.48922393e-01 6.20786905e-01
2.09895134e-01 -1.41641408e-01 -1.22955620e+00 4.47331160e-01
1.40908554e-01 -7.69599617e-01 -2.10608467e-02 -9.37165096e-02
5.50473630e-01 4.57731992e-01 -3.92898023e-01 5.37108362e-01
1.34837016e-01 -1.04402971e+00 6.11465156e-01 3.13870460e-01
7.80668855e-01 -5.17269969e-01 7.92121768e-01 5.75462282e-01
-1.29639006e+00 2.46592909e-01 3.50625552e-02 2.06423014e-01
2.60555238e-01 7.30788112e-01 -7.73333430e-01 1.21091843e+00
4.32329565e-01 5.73381126e-01 -5.77667296e-01 5.18562913e-01
-5.39841950e-01 6.49145007e-01 -4.15700018e-01 7.58890063e-02
4.33193624e-01 -1.47033513e-01 6.98879898e-01 1.83472371e+00
-1.27447406e-02 -1.85848445e-01 2.31483847e-01 6.37681663e-01
1.17325589e-01 7.23275959e-01 -7.59073377e-01 -3.97529066e-01
-7.27915671e-03 1.13168037e+00 -3.20936680e-01 -4.84390646e-01
-7.05727577e-01 6.28080606e-01 3.09204370e-01 4.52316344e-01
-5.06845236e-01 -5.41209280e-01 6.18469894e-01 -7.21067861e-02
2.38894626e-01 -4.15116251e-01 -3.06920826e-01 -1.25545990e+00
4.49698567e-02 -1.07638359e+00 5.89428842e-01 -3.48394603e-01
-1.29706836e+00 9.55093443e-01 4.28929001e-01 -9.41027403e-01
-7.60935247e-01 -9.02042270e-01 -1.91850308e-02 1.31789970e+00
-1.84291101e+00 -1.50806618e+00 6.80604637e-01 3.73939633e-01
5.91273069e-01 -6.62399977e-02 1.13165641e+00 7.54079163e-01
-2.82331109e-01 3.12028736e-01 -1.94886252e-02 2.56038249e-01
9.66453195e-01 -1.12306464e+00 5.06005228e-01 6.80286646e-01
3.26668173e-01 1.18049920e+00 5.63484132e-01 -8.02801609e-01
-1.00546825e+00 -1.04641867e+00 1.92985177e+00 -4.21105415e-01
1.06223536e+00 -4.17962581e-01 -8.90480578e-01 6.32512867e-01
6.42209709e-01 -5.91987550e-01 8.01468730e-01 4.58738565e-01
-3.39059770e-01 2.91011274e-01 -9.35337961e-01 3.99286896e-01
7.87900150e-01 -7.55459964e-01 -8.88714373e-01 5.46251237e-01
7.60467649e-01 -1.77206069e-01 -1.19226265e+00 4.28733140e-01
6.22257948e-01 -2.29916334e-01 1.00747764e+00 -1.02843726e+00
3.94421577e-01 -1.39439121e-01 9.35441330e-02 -1.34326279e+00
-3.60752717e-02 -3.10130060e-01 1.16494939e-01 1.41386914e+00
1.01646173e+00 -7.05687523e-01 -1.52767047e-01 9.46286991e-02
8.33573192e-02 -6.38348937e-01 -8.80951047e-01 -8.17868173e-01
7.13941693e-01 -6.20178819e-01 2.33193919e-01 9.92137611e-01
1.13195807e-01 8.51277411e-01 -2.06603602e-01 -2.41716951e-01
2.12143570e-01 1.33592889e-01 1.72997743e-01 -1.18802130e+00
-2.31387317e-01 -3.03055853e-01 -2.48719677e-02 -7.02967644e-01
5.70064306e-01 -1.11074018e+00 7.38627762e-02 -1.68730116e+00
-2.30180006e-02 -8.74648616e-02 -3.85904342e-01 6.59853518e-01
-1.93612620e-01 -1.53793871e-01 -2.66639739e-02 -1.23000100e-01
-2.64302671e-01 2.02048197e-01 8.10580254e-01 -8.98256451e-02
-1.30911544e-01 -2.55786151e-01 -6.76297486e-01 7.75140405e-01
5.26217461e-01 -6.11884534e-01 -5.72817996e-02 -6.47677779e-01
3.12793970e-01 -2.23302290e-01 -4.72213626e-02 -3.87099594e-01
7.30939116e-03 -1.05268694e-01 -4.64719422e-02 -2.91538417e-01
7.12631876e-03 -7.42611766e-01 -1.52875334e-01 5.61042845e-01
-3.46189409e-01 2.18081802e-01 4.86565113e-01 2.34209210e-01
-2.85020977e-01 -2.51044065e-01 4.86627907e-01 -3.76378685e-01
-7.00536013e-01 -1.69730470e-01 -6.94167852e-01 2.87943512e-01
7.10704327e-01 3.83696973e-01 -3.34778011e-01 -1.07256537e-02
-8.58178556e-01 4.23558414e-01 4.16195532e-03 7.69092679e-01
2.72073448e-01 -9.21505690e-01 -1.12698305e+00 4.29933518e-02
2.52328098e-01 -4.04840142e-01 -2.91690797e-01 9.07164454e-01
-3.42118621e-01 1.11856174e+00 1.64194793e-01 -2.60197371e-01
-1.19170189e+00 5.15404284e-01 2.27844298e-01 -1.18633962e+00
-7.30630904e-02 5.67438424e-01 9.87672880e-02 -8.31236184e-01
7.29402676e-02 -6.40702188e-01 -6.25321686e-01 3.08456808e-01
9.50964913e-02 3.43396634e-01 5.08871496e-01 -8.45429361e-01
-8.19203973e-01 5.31916380e-01 -8.14698637e-02 -4.32882071e-01
1.34782374e+00 -5.41472621e-03 -5.28956950e-01 6.31220460e-01
1.19130337e+00 1.92759976e-01 1.90112054e-01 -3.10528666e-01
7.73606360e-01 3.87600958e-01 3.77061814e-02 -1.10103762e+00
-5.98880529e-01 6.95096970e-01 2.93590754e-01 8.11383203e-02
1.11887276e+00 8.06160420e-02 6.96565509e-01 3.92173082e-01
5.02843857e-01 -1.12601054e+00 -7.69872844e-01 9.90344107e-01
9.47384596e-01 -1.06181598e+00 2.43669093e-01 -3.95103514e-01
-5.33589065e-01 9.34635520e-01 2.29375690e-01 1.26480153e-02
9.50627267e-01 3.44737768e-01 2.93808788e-01 -3.92512828e-01
-9.21479404e-01 -5.35771251e-01 7.95529127e-01 2.48010501e-01
1.12172854e+00 8.00768286e-02 -1.20761681e+00 6.28122628e-01
-8.08035806e-02 -4.27113287e-02 -5.86837009e-02 1.07005274e+00
-2.39374414e-02 -1.68165481e+00 -2.11813092e-01 2.38543466e-01
-1.07659900e+00 -7.73595870e-01 -8.87391031e-01 1.11268556e+00
4.26861793e-01 1.10826242e+00 -4.80841845e-01 -1.07381478e-01
4.14156795e-01 6.03348017e-01 2.06838652e-01 -9.82777297e-01
-1.19294477e+00 3.19958329e-01 6.82998538e-01 -2.48267502e-01
-7.09927440e-01 -6.38300121e-01 -1.23513639e+00 2.15006158e-01
-4.77136463e-01 2.80537993e-01 7.34019101e-01 1.39651370e+00
1.74673527e-01 5.82218707e-01 1.16240576e-01 -3.47298801e-01
-2.65211314e-01 -1.47911417e+00 -2.40517884e-01 3.07056397e-01
2.26170331e-01 -4.62372631e-01 -3.17728400e-01 1.82031497e-01] | [10.224411964416504, 9.551294326782227] |
3aea6d4a-d8ad-4dd9-ac26-4ab0e34182d7 | face-body-voice-video-person-clustering-with | 2105.09939 | null | https://arxiv.org/abs/2105.09939v1 | https://arxiv.org/pdf/2105.09939v1.pdf | Face, Body, Voice: Video Person-Clustering with Multiple Modalities | The objective of this work is person-clustering in videos -- grouping characters according to their identity. Previous methods focus on the narrower task of face-clustering, and for the most part ignore other cues such as the person's voice, their overall appearance (hair, clothes, posture), and the editing structure of the videos. Similarly, most current datasets evaluate only the task of face-clustering, rather than person-clustering. This limits their applicability to downstream applications such as story understanding which require person-level, rather than only face-level, reasoning. In this paper we make contributions to address both these deficiencies: first, we introduce a Multi-Modal High-Precision Clustering algorithm for person-clustering in videos using cues from several modalities (face, body, and voice). Second, we introduce a Video Person-Clustering dataset, for evaluating multi-modal person-clustering. It contains body-tracks for each annotated character, face-tracks when visible, and voice-tracks when speaking, with their associated features. The dataset is by far the largest of its kind, and covers films and TV-shows representing a wide range of demographics. Finally, we show the effectiveness of using multiple modalities for person-clustering, explore the use of this new broad task for story understanding through character co-occurrences, and achieve a new state of the art on all available datasets for face and person-clustering. | ['Andrew Zisserman', 'Vicky Kalogeiton', 'Andrew Brown'] | 2021-05-20 | null | null | null | null | ['face-clustering'] | ['computer-vision'] | [ 7.20531270e-02 -1.38734967e-01 -1.51004612e-01 -4.66352403e-01
-7.18267918e-01 -7.33631253e-01 7.17828333e-01 -8.64081383e-02
9.63292941e-02 7.95933753e-02 8.23181272e-01 5.84011972e-01
-1.00531720e-01 -2.91254014e-01 -3.04873019e-01 -6.37271583e-01
3.76054794e-02 6.30849123e-01 -1.49849221e-01 -1.93777997e-02
2.43269820e-02 1.12273186e-01 -2.11913443e+00 8.61187100e-01
1.97664529e-01 8.10808241e-01 -2.80625790e-01 7.43834257e-01
2.65742213e-01 6.74560487e-01 -4.13537651e-01 -6.50861979e-01
-1.61712080e-01 -7.24581182e-01 -7.73853004e-01 7.35333145e-01
1.08150685e+00 -4.34855521e-01 -2.60272563e-01 6.16257310e-01
6.30294025e-01 2.33031154e-01 8.98352385e-01 -1.39301801e+00
-3.51391435e-01 5.68479002e-01 -7.38275409e-01 -2.18462974e-01
1.20767987e+00 -1.04719110e-01 8.97869587e-01 -7.31657624e-01
8.49251211e-01 1.69448733e+00 9.82994974e-01 1.05105364e+00
-1.04158688e+00 -4.11165297e-01 1.92235440e-01 2.40020081e-01
-1.45191824e+00 -9.96389925e-01 5.40550888e-01 -9.09730494e-01
5.71908116e-01 3.90125632e-01 7.28193760e-01 1.42246592e+00
-6.37198031e-01 9.47133482e-01 1.01475620e+00 -4.46929812e-01
-1.95140749e-01 2.29086757e-01 1.31454930e-01 8.53032351e-01
-2.82825202e-01 -3.31253469e-01 -8.52759004e-01 -2.08580002e-01
4.43343878e-01 -4.17754278e-02 -3.11451226e-01 -2.03065932e-01
-1.31904173e+00 5.44700563e-01 -2.96906054e-01 2.86476761e-01
1.62859801e-02 1.08384572e-01 5.11694670e-01 -3.41825858e-02
4.67755228e-01 -7.69884735e-02 -5.08900210e-02 -4.61571395e-01
-1.23471272e+00 2.87360519e-01 9.16340828e-01 9.93536592e-01
4.13265407e-01 -2.34786615e-01 -2.88352937e-01 1.10370541e+00
3.16507876e-01 3.63011748e-01 2.23928556e-01 -1.33886635e+00
2.28835255e-01 4.12331998e-01 -3.11048061e-01 -1.03461039e+00
-5.63648582e-01 1.79611042e-01 -5.17282784e-01 -1.47789538e-01
6.10357642e-01 -1.16536275e-01 -6.02629781e-01 1.93314528e+00
5.93863785e-01 2.83447087e-01 -2.94764608e-01 8.62891138e-01
1.48400533e+00 2.84661055e-01 -4.47462834e-02 -3.78466517e-01
1.76141405e+00 -8.97614300e-01 -7.44153082e-01 7.92389289e-02
3.08180839e-01 -7.50193894e-01 8.46850455e-01 4.82217580e-01
-1.24312401e+00 -6.81297779e-01 -4.79388744e-01 1.50625825e-01
-3.36723715e-01 3.74133021e-01 4.76957411e-01 1.15464473e+00
-1.16088188e+00 5.82184970e-01 -2.79883683e-01 -1.09247160e+00
4.02914494e-01 3.13510478e-01 -7.65825033e-01 -2.24869415e-01
-6.28017843e-01 4.98713672e-01 5.45913577e-02 -3.42249215e-01
-7.97718704e-01 -5.18583119e-01 -9.97926414e-01 -3.77223313e-01
3.52279544e-01 -7.98328817e-01 9.81737018e-01 -1.23578286e+00
-1.41483843e+00 1.44875240e+00 -4.59605753e-01 1.69845074e-01
5.85057855e-01 -1.00595914e-01 -7.34736919e-01 5.25320709e-01
1.04637429e-01 9.52079237e-01 1.11503780e+00 -1.64157474e+00
-5.66198111e-01 -6.13003492e-01 -1.68653682e-01 5.04627764e-01
-6.41704977e-01 5.32135963e-01 -1.22513819e+00 -5.85448623e-01
-6.51804432e-02 -1.06822944e+00 4.81826037e-01 -1.14664569e-01
-4.85874474e-01 -2.84105450e-01 1.11776471e+00 -9.92683291e-01
1.33100343e+00 -2.46128750e+00 3.33283216e-01 3.65191624e-02
3.97342980e-01 -2.92492330e-01 6.26544794e-03 4.08025861e-01
-1.30130440e-01 1.44975394e-01 2.98772175e-02 -1.07287860e+00
1.27107292e-01 1.24273011e-02 2.33348325e-01 8.00660253e-01
-2.01273337e-01 6.35194242e-01 -6.19303286e-01 -1.07283545e+00
2.87821293e-01 6.04609489e-01 -5.35556436e-01 -1.82785213e-01
7.60634020e-02 3.96369398e-01 1.23072751e-01 1.05843437e+00
3.10373098e-01 1.18686624e-01 2.66943187e-01 -2.94014096e-01
1.70754686e-01 -4.88872916e-01 -1.34659541e+00 1.58775508e+00
2.26512894e-01 8.39758754e-01 4.56684738e-01 -5.60640275e-01
4.19495106e-01 6.01771832e-01 9.50895846e-01 9.75884125e-02
2.58911848e-01 -2.72594213e-01 -1.88869789e-01 -7.74079919e-01
4.10284191e-01 1.94922425e-02 -1.66952267e-01 4.48015720e-01
2.60401577e-01 1.03004485e-01 4.77640361e-01 3.57192874e-01
7.25158989e-01 2.88531393e-01 1.88163090e-02 -7.00904056e-02
3.78421962e-01 -1.28444314e-01 2.75372624e-01 4.52249259e-01
-3.64934444e-01 1.19704032e+00 4.15727109e-01 2.18209755e-02
-8.23254168e-01 -8.55564237e-01 -1.66728556e-01 1.58136809e+00
-1.45121649e-01 -7.41579413e-01 -1.11279750e+00 -4.50705826e-01
1.27053887e-01 2.46994287e-01 -9.15490746e-01 1.48443326e-01
-2.35004738e-01 -6.45629048e-01 8.70056093e-01 4.63276476e-01
1.88821360e-01 -8.55342209e-01 -1.94942746e-02 -2.47788817e-01
-6.66859329e-01 -1.39012146e+00 -6.54153705e-01 -5.56429505e-01
-4.83327806e-01 -1.26707029e+00 -7.10199535e-01 -7.78764307e-01
6.25041664e-01 1.34971082e-01 1.24187636e+00 1.72463030e-01
-4.16610807e-01 1.43831861e+00 -4.85377967e-01 -1.00195691e-01
-1.72315165e-01 -3.97300214e-01 4.48998153e-01 5.29672503e-01
4.12281305e-01 -3.05255741e-01 -3.41111988e-01 5.50409257e-01
-3.72020841e-01 -9.15137678e-02 4.52339984e-02 3.44138354e-01
2.12649703e-01 3.59580010e-01 2.16474220e-01 -7.90852308e-01
4.21953112e-01 -5.07349730e-01 5.42915821e-01 2.87250608e-01
2.36167610e-01 -9.81471002e-01 2.02084780e-01 -5.52709937e-01
-1.04853606e+00 3.59778106e-01 1.54428989e-01 -6.16588056e-01
-7.78875828e-01 -1.11877052e-02 -4.71746713e-01 9.49714258e-02
4.53243226e-01 1.33851498e-01 2.34733578e-02 -3.52377266e-01
3.91770393e-01 5.89790344e-01 1.02211189e+00 -9.01060402e-01
5.44203401e-01 8.44548225e-01 -2.10973859e-01 -1.28807068e+00
-6.26967549e-01 -9.32084560e-01 -1.13973641e+00 -1.12744546e+00
1.19571817e+00 -1.27792943e+00 -1.14285648e+00 5.97842097e-01
-8.13618124e-01 -1.08352177e-01 -1.19675463e-02 4.13820028e-01
-7.23428309e-01 7.22764909e-01 -7.67783821e-01 -1.16858959e+00
8.79018232e-02 -5.40452957e-01 1.33048546e+00 1.58836190e-02
-7.25572705e-01 -9.20288086e-01 -2.59075433e-01 1.07510626e+00
-3.61268014e-01 7.33716965e-01 6.00540102e-01 -3.69631886e-01
1.16909534e-01 -1.50792181e-01 -6.26651198e-02 -6.93516359e-02
-9.61583778e-02 5.43957293e-01 -1.30332947e+00 -3.12549978e-01
-4.04372960e-01 -5.93709230e-01 7.89055228e-01 5.80213845e-01
1.09370995e+00 1.21399490e-02 -5.77894032e-01 3.69477332e-01
1.01325881e+00 -1.29003584e-01 4.89769816e-01 -1.80377081e-01
1.03188622e+00 1.35015380e+00 4.71028388e-01 8.29289138e-01
7.12535024e-01 7.59321272e-01 9.95552465e-02 6.51724637e-02
-4.76304233e-01 -1.59427255e-01 6.93648040e-01 6.25801682e-01
-6.86666191e-01 -3.02855909e-01 -8.08198512e-01 5.35540640e-01
-1.97589302e+00 -1.52704656e+00 -6.05485857e-01 1.91981912e+00
5.69337249e-01 -3.46477777e-01 1.03278899e+00 3.12819213e-01
1.16912496e+00 -1.36622161e-01 -2.41638482e-01 2.71818973e-02
-3.29876810e-01 -4.62483495e-01 -2.08280176e-01 2.07700282e-01
-1.47567594e+00 7.66911745e-01 6.78577709e+00 8.49840581e-01
-4.93251860e-01 1.27957746e-01 6.68200672e-01 -5.31395853e-01
2.94815805e-02 -3.88352036e-01 -9.53043103e-01 4.83710676e-01
6.40355051e-01 4.91327107e-01 5.19354701e-01 6.47991180e-01
2.02516720e-01 -2.27902681e-01 -1.67164087e+00 1.48956275e+00
9.32705104e-01 -8.28121603e-01 -1.90301955e-01 9.56115872e-02
6.42003119e-01 -6.64204478e-01 1.24381192e-01 2.64495462e-01
-8.54419395e-02 -1.07271433e+00 1.13053882e+00 6.29741490e-01
1.09264350e+00 -7.93892741e-01 3.51181686e-01 3.89571525e-02
-1.49307799e+00 -2.24407509e-01 1.76741511e-01 6.34595286e-03
1.68313473e-01 2.55144179e-01 -4.16026771e-01 3.13567311e-01
1.12300944e+00 9.69352007e-01 -6.70280516e-01 6.77973688e-01
2.81887293e-01 4.95371759e-01 -3.64059657e-01 2.04390332e-01
-2.50372738e-01 -7.15863928e-02 4.09621865e-01 1.67782879e+00
2.55065173e-01 2.93688267e-01 5.93007207e-01 3.83469135e-01
9.69227403e-02 2.04772606e-01 -4.38443273e-01 -1.05668306e-01
4.89350855e-01 1.41737318e+00 -9.27655339e-01 -3.38292271e-01
-5.41045189e-01 1.07312953e+00 -7.75723311e-04 1.15703300e-01
-7.66837955e-01 2.57969350e-01 6.78476632e-01 4.65908438e-01
5.38714975e-02 -1.36044577e-01 -1.87455475e-01 -9.89034235e-01
-1.75276324e-01 -1.14025831e+00 8.20622087e-01 -8.56323004e-01
-1.48253798e+00 1.24258742e-01 1.68538049e-01 -1.08672643e+00
-2.72994131e-01 -4.41576391e-01 -5.51024914e-01 2.70919502e-01
-6.50237799e-01 -1.62232447e+00 -5.20688236e-01 1.24671471e+00
6.50347888e-01 -2.88872063e-01 8.78688574e-01 5.48565447e-01
-8.47238183e-01 7.40543723e-01 -6.55825138e-02 3.97895992e-01
1.07729638e+00 -1.10699689e+00 -2.18266517e-01 4.23514366e-01
1.78519860e-01 5.20495057e-01 6.25039101e-01 -8.23587060e-01
-1.43112862e+00 -7.75631249e-01 5.84532320e-01 -1.00497341e+00
4.21633750e-01 -6.59097791e-01 -3.12442541e-01 5.33628225e-01
2.24494010e-01 -5.17874360e-01 1.21724343e+00 6.21844411e-01
-4.49823469e-01 2.04900637e-01 -1.35394394e+00 5.36244273e-01
1.31406152e+00 -5.86112857e-01 -3.54441285e-01 4.15942460e-01
1.75324604e-01 -3.11697453e-01 -1.10677373e+00 2.14001507e-01
9.01431978e-01 -1.16734481e+00 1.07930839e+00 -3.57224524e-01
7.18280911e-01 -2.52058923e-01 -3.94331306e-01 -8.15317273e-01
-5.66338360e-01 -6.10161185e-01 -3.66710663e-01 1.96604002e+00
9.03792563e-04 2.01943889e-01 9.23774362e-01 5.85367560e-01
-5.75448573e-02 -4.13171321e-01 -5.62493443e-01 -4.43346053e-01
-3.53905588e-01 -5.52479267e-01 2.55890667e-01 1.39304578e+00
1.93443477e-01 2.52541691e-01 -7.32401252e-01 3.77004594e-02
7.70680189e-01 -8.19086954e-02 1.06085789e+00 -1.29657626e+00
-3.75505805e-01 -6.03744149e-01 -4.89421278e-01 -5.73828876e-01
2.59273589e-01 -5.72070003e-01 -3.26638743e-02 -1.42747164e+00
8.54355752e-01 1.03783615e-01 3.99946928e-01 4.56426889e-01
-1.12154819e-01 6.39077485e-01 3.57680649e-01 3.39960307e-01
-9.38751876e-01 1.27552226e-01 9.06147182e-01 -5.74047416e-02
1.05134517e-01 -1.23254701e-01 -6.86939418e-01 1.03749514e+00
3.03928912e-01 -3.01967748e-02 -2.12166503e-01 -1.58368647e-01
-1.95292890e-01 7.06111342e-02 5.15592098e-01 -1.10110247e+00
2.75439888e-01 1.30878270e-01 8.69333804e-01 -7.63612509e-01
9.02101517e-01 -6.68291509e-01 2.03570187e-01 1.27010733e-01
-4.76437323e-02 -2.05700114e-01 4.35954593e-02 7.26700425e-01
-1.40895650e-01 -6.92908242e-02 6.93139672e-01 -3.51632148e-01
-8.59005630e-01 2.54910588e-01 -5.71142256e-01 1.16257787e-01
1.24266291e+00 -7.92696238e-01 -2.30175748e-01 -8.71774554e-01
-1.25753045e+00 1.90849930e-01 7.03235269e-01 5.89618802e-01
4.24991846e-01 -1.43608832e+00 -9.17758226e-01 -1.87352926e-01
2.27774337e-01 -6.08202636e-01 7.18788385e-01 8.49888086e-01
-2.65033897e-02 -1.29186973e-01 -1.14251010e-01 -9.76176798e-01
-2.16429543e+00 5.93750179e-01 1.20755672e-01 4.77066636e-01
-3.24492216e-01 9.83738303e-01 2.61597157e-01 -5.34088314e-01
5.49985409e-01 4.28648889e-01 -6.02066219e-01 7.55809128e-01
6.06550872e-01 6.45220637e-01 -3.55303138e-01 -1.39967656e+00
-5.04971266e-01 1.01873481e+00 3.64326030e-01 -2.17333660e-01
1.13120806e+00 -4.17539120e-01 -2.40372941e-01 7.42664278e-01
1.08262944e+00 1.07516631e-01 -9.45569158e-01 -3.79416421e-02
-2.10214823e-01 -5.59084415e-01 -4.56079811e-01 -6.08107269e-01
-8.61324251e-01 6.54460251e-01 3.77991647e-01 2.26554409e-01
1.07864475e+00 4.00028139e-01 5.01358569e-01 5.34844249e-02
1.25436753e-01 -1.55412257e+00 4.68501836e-01 3.07038844e-01
6.85370624e-01 -1.30529702e+00 2.97222465e-01 -7.12210238e-01
-9.64757204e-01 1.10549462e+00 7.43489802e-01 5.00503600e-01
5.67872941e-01 6.03138767e-02 -6.28452078e-02 -3.33213747e-01
-4.58451182e-01 -4.60415274e-01 5.61678529e-01 8.80618334e-01
6.93033397e-01 1.79895148e-01 3.22247267e-01 7.21506417e-01
-4.87047911e-01 -4.68652338e-01 3.22765023e-01 6.31123900e-01
-3.12778145e-01 -8.18930328e-01 -9.68020499e-01 5.00303566e-01
-5.04525721e-01 1.93450615e-01 -1.05704629e+00 7.27134645e-01
6.61237478e-01 1.45395553e+00 1.50479972e-01 -6.15359962e-01
1.89444527e-01 2.18994305e-01 8.72208178e-01 -8.06414902e-01
-6.65953696e-01 3.30743909e-01 5.19268692e-01 -4.59372550e-01
-9.68880832e-01 -1.48666167e+00 -8.65979671e-01 -9.43388283e-01
-1.60811260e-01 -3.94520164e-01 4.13870841e-01 9.19600010e-01
-2.21115705e-02 2.66405791e-01 3.94508660e-01 -1.12631023e+00
3.03946614e-01 -7.85836816e-01 -7.99686372e-01 9.09010053e-01
-7.98904449e-02 -6.61201715e-01 -1.90475002e-01 6.53188825e-01] | [13.681641578674316, 1.0723369121551514] |
e0732033-b728-4821-af08-241a779a3b18 | case4sr-using-category-sequence-graph-to | null | null | https://www.sciencedirect.com/science/article/pii/S0950705120306870 | https://reader.elsevier.com/reader/sd/pii/S0950705120306870?token=FE79CE03D3E741C761C7E625D69F5A5A88DD0757634725107E20AA8E6CB7ADEA544CD7F8FC3476CE81EE9351078F0277&originRegion=us-east-1&originCreation=20230320035059 | CaSe4SR: Using category sequence graph to augment session-based recommendation | Session-based recommendation aims to predict next item based on users’ anonymous behavior sequence within a short time. Recent studies focus on modeling sequential dependencies or complex relations among items in a session via recurrent/convolutional/graph neural networks. However, the following problems still remain: for short sessions, limited interactions cannot manifest user’s intent clearly; for long sessions, user’s interest may drift but be blurred by complex transitions. Motivated by the observation that different items are often belong to only a few categories or that closely related, in this article, we tackle these challenges by leveraging item category information, which is a concise form of knowledge and readily available in many platforms. We propose a novel method CaSe4SR that utilizes category sequence graph to augment session-based recommendation. In CaSe4SR, we build an item graph and a category graph, from user behavior sequence and item category sequence. The latter summarizes the former at concept level, which reduces item-level user behavior noises and makes user’s interest clearer. Afterwards, graph neural networks are applied on item graph and category graph respectively to learn representations of items and categories. Then two alternative fusion strategies and attention mechanism are designed to integrate them, yielding global embedding of the session, which is further combined with representation of last item to get ultimate session representation. Extensive experiments on real-world datasets show that CaSe4SR outperforms other state-of-the-art methods consistently. Detailed analysis reveals that category sequence graph is beneficial for next-item recommendation in sessions with different lengths. | ['Tao Lian', 'Li Wang', 'Lin Liu'] | 2020-11-06 | null | null | null | knowledge-based-systems-2020-11 | ['session-based-recommendations'] | ['miscellaneous'] | [ 8.13304633e-03 -3.32990110e-01 -6.44252956e-01 -5.14629543e-01
3.00648585e-02 -6.22628868e-01 2.28883594e-01 2.25940749e-01
7.97868799e-03 4.87892777e-01 6.10209882e-01 -1.98783889e-01
-4.11253631e-01 -7.94597328e-01 -5.41153073e-01 -3.53318483e-01
-2.80004114e-01 -2.85694227e-02 -1.77133903e-02 -4.54770803e-01
2.23691493e-01 6.88644722e-02 -1.42112947e+00 4.33708817e-01
9.68274236e-01 9.85660493e-01 1.91065729e-01 3.70742798e-01
-4.52741355e-01 6.34629428e-01 -4.23465282e-01 -3.77328545e-01
-1.20936586e-02 -6.82180107e-01 -4.61891502e-01 2.89309710e-01
1.68179795e-01 -3.43850106e-01 -4.54347998e-01 8.03474665e-01
2.88363785e-01 7.76965916e-01 6.15885258e-01 -1.04624557e+00
-1.51134861e+00 1.08264434e+00 -4.52513099e-01 1.95910648e-01
6.72591090e-01 -1.40203550e-01 1.42571509e+00 -8.30469072e-01
4.10496116e-01 1.04560816e+00 5.65720379e-01 5.80097318e-01
-1.03416598e+00 -5.91339290e-01 1.08892441e+00 3.25832665e-01
-1.24175012e+00 2.82747280e-02 9.35014844e-01 -3.44216526e-01
9.02962446e-01 3.61886144e-01 7.12291002e-01 1.33226728e+00
-6.62220865e-02 8.77967477e-01 3.86142433e-01 8.16004574e-02
5.96160395e-03 3.84147078e-01 9.24489379e-01 4.06213790e-01
3.57546777e-01 -9.02156979e-02 -4.86890376e-01 -9.37910974e-02
7.05232501e-01 8.40122461e-01 -6.12715840e-01 -3.36853385e-01
-8.85614872e-01 9.03916359e-01 9.50697005e-01 4.34938103e-01
-4.67016369e-01 -2.79491395e-01 2.97197402e-01 5.47571599e-01
4.48588461e-01 1.01326808e-01 -3.39037359e-01 3.39886427e-01
-3.75520825e-01 6.76299483e-02 9.86844242e-01 1.03814065e+00
7.24881053e-01 8.58636275e-02 -3.61081600e-01 9.56747353e-01
5.79632580e-01 5.82035631e-02 6.20338917e-01 -1.82833499e-03
3.24644566e-01 8.47398400e-01 4.76419777e-02 -1.30249083e+00
-3.80439132e-01 -1.05333400e+00 -9.66261685e-01 -6.48711264e-01
9.82280597e-02 -7.73999840e-02 -8.27537537e-01 1.63227856e+00
2.64092833e-02 6.36688173e-01 -1.31452337e-01 1.04262996e+00
1.42685747e+00 6.62005126e-01 2.42368698e-01 -4.23503876e-01
1.16803157e+00 -1.12356699e+00 -7.84850001e-01 -7.08353370e-02
5.61886609e-01 -3.17352802e-01 1.01733756e+00 2.58795977e-01
-7.49746978e-01 -9.55288470e-01 -9.85000432e-01 1.74655050e-01
-5.00808954e-01 -3.33611965e-02 1.00414467e+00 5.29766440e-01
-9.12159562e-01 7.05484390e-01 -2.39073694e-01 -5.82302749e-01
2.72630662e-01 5.79016149e-01 -1.34205341e-01 -9.62067097e-02
-1.42487013e+00 2.27792129e-01 2.32926115e-01 3.21495801e-01
-3.24781507e-01 -7.40388930e-01 -9.36377704e-01 4.31517452e-01
4.30199593e-01 -7.14706302e-01 8.91842127e-01 -1.04073954e+00
-1.47439063e+00 2.88423181e-01 -9.61081684e-02 -3.08103889e-01
-1.89286619e-02 -7.71997795e-02 -1.08731413e+00 -5.38840532e-01
-3.04959685e-01 3.51619418e-03 7.66725659e-01 -1.11400235e+00
-7.26056695e-01 -4.76721793e-01 4.00764972e-01 5.48765481e-01
-7.49015570e-01 -3.67072523e-01 -7.58937895e-01 -6.15744114e-01
-6.61152303e-02 -8.01765442e-01 -2.25852400e-01 -6.43692374e-01
-2.50896811e-01 -5.67633331e-01 7.32974052e-01 -5.86383522e-01
1.89431036e+00 -2.10513544e+00 2.45805845e-01 2.89210171e-01
2.76931375e-01 2.63692439e-01 -2.93904692e-01 6.56261802e-01
-1.14907376e-01 1.47989526e-01 2.67581999e-01 -2.35506341e-01
4.69228253e-02 1.29862651e-01 -4.16326225e-01 2.76459575e-01
-4.12560046e-01 1.27422106e+00 -9.15158927e-01 1.04358062e-01
2.76656210e-01 4.98527706e-01 -6.98109388e-01 3.49438250e-01
-6.97877258e-02 3.62050116e-01 -5.39376915e-01 4.40150946e-01
7.97464669e-01 -6.59786463e-01 3.78924906e-01 -3.77643436e-01
2.82836080e-01 4.54430163e-01 -1.16875064e+00 1.79817724e+00
-4.40041661e-01 -5.02430834e-02 -3.35458398e-01 -1.03387904e+00
9.87394929e-01 8.49497691e-02 2.92748123e-01 -6.28272116e-01
4.65934649e-02 -2.30895177e-01 3.15584131e-02 -3.83149058e-01
6.53252363e-01 2.05717757e-01 -1.53606683e-01 5.47381580e-01
3.16413522e-01 7.43522704e-01 7.14262798e-02 4.13589925e-01
8.15621138e-01 1.27703825e-03 3.98817837e-01 1.01896979e-01
7.31114149e-01 -5.80196798e-01 4.22832131e-01 9.66648936e-01
-6.54157922e-02 4.04140770e-01 1.42592683e-01 -2.38690510e-01
-4.24497575e-01 -9.70516741e-01 3.57209831e-01 1.55470347e+00
5.41544735e-01 -5.16872823e-01 -6.50567561e-02 -1.24811709e+00
1.65364325e-01 6.04918063e-01 -9.02763069e-01 -4.41373348e-01
-3.88804048e-01 -5.25096416e-01 -3.14519823e-01 6.06163979e-01
2.07876578e-01 -1.32754791e+00 7.18516171e-01 4.78361845e-01
2.91803218e-02 -6.09327555e-01 -1.05022919e+00 -2.26070762e-01
-7.03065336e-01 -9.10347104e-01 -6.45870745e-01 -9.08846200e-01
6.00400567e-01 9.75509822e-01 1.19856942e+00 3.43184918e-01
1.29394427e-01 4.73435789e-01 -7.83875585e-01 4.93638664e-02
2.19245270e-01 7.29158446e-02 1.76745787e-01 5.61023057e-01
6.99257612e-01 -8.56552958e-01 -1.07434285e+00 4.21020597e-01
-7.51716375e-01 -2.79620945e-01 5.16900182e-01 7.92663455e-01
2.58797109e-01 -1.62070975e-01 9.64713037e-01 -1.43353534e+00
9.55710530e-01 -1.17686915e+00 -3.51927467e-02 2.73275137e-01
-6.64579809e-01 -5.13413429e-01 1.01784945e+00 -6.32063031e-01
-1.06101000e+00 -4.45394665e-01 -5.19697890e-02 -3.43051165e-01
-1.49411976e-01 9.40165043e-01 -1.66312456e-01 1.39623582e-01
5.26012242e-01 3.96183372e-01 -2.94837952e-01 -6.66673779e-01
5.95161736e-01 8.69531512e-01 1.97746754e-01 1.01690382e-01
6.03990197e-01 1.34388462e-01 -6.25597715e-01 -5.25717080e-01
-1.19367290e+00 -1.02557886e+00 -3.12044919e-01 -2.12013349e-01
3.63630414e-01 -9.99459147e-01 -9.16954815e-01 1.61421657e-01
-6.60207748e-01 -4.24058624e-02 -1.98215172e-01 6.24567389e-01
-2.21400205e-02 4.39524800e-01 -7.40102887e-01 -5.81418693e-01
-4.94567037e-01 -5.09411812e-01 5.73664367e-01 7.40390539e-01
2.73315981e-02 -1.37888741e+00 -3.05664390e-02 2.37866834e-01
3.79316956e-01 -3.06515723e-01 5.57800591e-01 -1.07167172e+00
-4.64727640e-01 -4.02738720e-01 -3.34364295e-01 3.76143187e-01
6.57579660e-01 -4.61826771e-01 -4.58663553e-01 -6.21906936e-01
-2.34989330e-01 1.91246629e-01 8.90355408e-01 3.07658732e-01
1.33940542e+00 -4.69364136e-01 -3.65912229e-01 5.07458210e-01
1.44112206e+00 2.53882676e-01 5.08667767e-01 -2.28841186e-01
1.19794929e+00 4.40630674e-01 5.52428126e-01 4.64443505e-01
4.47404891e-01 6.64257348e-01 2.83840001e-01 1.95888579e-01
-1.42662644e-01 -5.52058220e-01 3.86680573e-01 1.23685873e+00
-1.48348257e-01 -8.70883822e-01 4.76015769e-02 3.87093365e-01
-2.28834677e+00 -1.07799840e+00 -2.08089277e-01 2.34809852e+00
1.82330295e-01 -3.69783454e-02 4.49595749e-01 -3.43374282e-01
8.86721849e-01 2.22788259e-01 -7.47212887e-01 -3.01353306e-01
9.04732868e-02 -2.30405733e-01 3.23950648e-01 2.18474716e-01
-1.01641822e+00 8.31439376e-01 5.60396767e+00 7.98019290e-01
-9.39676404e-01 -5.52129708e-02 2.50178248e-01 4.14147899e-02
-6.02960229e-01 -2.81407416e-01 -8.72036934e-01 7.36774743e-01
8.13010931e-01 -3.97611350e-01 6.69517696e-01 7.72671640e-01
2.25267589e-01 6.25222981e-01 -1.02387166e+00 1.03830457e+00
4.16201323e-01 -1.30572474e+00 3.20103943e-01 4.95069735e-02
7.24410236e-01 -2.55396783e-01 2.55130887e-01 9.76158500e-01
4.33275551e-01 -8.50504100e-01 2.84759700e-02 6.63553298e-01
3.57533514e-01 -7.90632725e-01 7.59674549e-01 2.38801941e-01
-1.67123079e+00 -3.48219633e-01 -5.31435609e-01 -3.41309980e-02
1.96890548e-01 2.57698029e-01 -5.85464954e-01 1.09544849e+00
6.35586441e-01 1.67378378e+00 -6.22150242e-01 1.15423667e+00
-1.00172915e-01 9.16583538e-01 4.42636609e-02 -3.08104187e-01
2.07342759e-01 -6.45135224e-01 4.24285084e-01 1.25528073e+00
3.78436744e-01 3.48113269e-01 4.03828084e-01 5.35953045e-01
-3.00263017e-01 4.69874173e-01 -6.08188391e-01 -1.29891053e-01
3.58395547e-01 1.50551760e+00 -4.00822550e-01 -2.73043811e-01
-8.96533608e-01 1.21894479e+00 5.66985488e-01 6.72287345e-01
-8.06746364e-01 -3.34274113e-01 6.76194608e-01 1.22229736e-02
7.25824594e-01 1.15327626e-01 1.01088084e-01 -1.52489901e+00
-1.20979607e-01 -5.97527981e-01 7.93771029e-01 -5.25549531e-01
-1.76034200e+00 4.48338211e-01 -4.55037117e-01 -1.49604416e+00
-1.70656309e-01 -2.41143093e-01 -9.35435653e-01 7.55447745e-01
-1.29521096e+00 -1.33535194e+00 -3.83641452e-01 8.89727175e-01
7.45148957e-01 -1.95559457e-01 7.17575014e-01 5.23769498e-01
-6.74352765e-01 1.15333259e+00 2.27733850e-01 1.77115407e-02
5.08526266e-01 -1.20147133e+00 3.02203745e-01 5.39397836e-01
3.33990753e-01 1.15933001e+00 4.63404566e-01 -8.02995801e-01
-1.50528538e+00 -1.35958791e+00 6.67847455e-01 -3.21853399e-01
6.91534221e-01 -3.93779695e-01 -1.30325711e+00 9.93401468e-01
3.14286388e-02 -1.98467337e-02 1.06473958e+00 7.47996032e-01
-4.56896961e-01 -6.17909580e-02 -7.66497195e-01 6.96736276e-01
1.64045715e+00 -5.82596302e-01 -5.61558127e-01 3.64937693e-01
1.01467168e+00 -7.82341324e-03 -8.76202524e-01 2.05247000e-01
4.44678187e-01 -8.27184618e-01 1.03918505e+00 -1.00932348e+00
-5.43615110e-02 -3.16446662e-01 4.07733545e-02 -1.52869046e+00
-1.04725516e+00 -6.61342800e-01 -6.86189830e-01 1.39577913e+00
2.86623597e-01 -5.29854715e-01 8.94880056e-01 2.88238496e-01
-3.55102897e-01 -6.21255338e-01 -1.74222767e-01 -6.72471225e-01
-4.29784507e-01 -8.47047046e-02 7.20506191e-01 1.04068565e+00
1.89791083e-01 7.83620000e-01 -9.05465186e-01 7.84533694e-02
4.50257152e-01 5.70340633e-01 6.26068294e-01 -1.30986452e+00
-4.46865857e-01 -3.93863440e-01 -3.45126361e-01 -1.46687794e+00
5.36766164e-02 -1.01578832e+00 -2.69556910e-01 -1.70383394e+00
2.75185913e-01 -2.77577728e-01 -1.08815098e+00 2.69697428e-01
-4.57233131e-01 2.69568264e-01 -2.02208720e-02 1.39053077e-01
-1.05281067e+00 4.82209235e-01 1.27410984e+00 -1.10949330e-01
-7.58321524e-01 4.53348428e-01 -1.23585379e+00 4.20214176e-01
6.15117550e-01 4.71883900e-02 -7.75894344e-01 -2.54776422e-03
2.53426135e-01 9.03465843e-04 1.18573317e-02 -6.61908448e-01
2.99425900e-01 -7.96801075e-02 3.10966253e-01 -7.19190001e-01
1.89056993e-01 -7.45742142e-01 2.73317367e-01 8.85679871e-02
-5.02949953e-01 -4.57783043e-01 -1.97215125e-01 1.32267165e+00
-4.57168967e-02 6.33456744e-03 2.13877618e-01 1.37455631e-02
-1.10876667e+00 9.69678819e-01 -9.97965410e-02 -4.66042697e-01
8.27987313e-01 -2.82960683e-01 -1.21969216e-01 -7.12181985e-01
-1.23692513e+00 3.23665321e-01 6.04705811e-02 1.12049091e+00
6.40072405e-01 -1.67930400e+00 -5.87812722e-01 1.55000538e-01
3.97214472e-01 -6.85601950e-01 1.05545032e+00 5.43940067e-01
3.17035794e-01 4.08690840e-01 -1.26997689e-02 -3.73894215e-01
-1.23705971e+00 1.07375550e+00 2.22484451e-02 -3.50770235e-01
-4.57771331e-01 1.04937208e+00 4.98307526e-01 -5.49314916e-01
3.49836558e-01 1.68000069e-02 -1.12405133e+00 2.71678627e-01
7.36247778e-01 2.68288255e-02 -9.01380777e-02 -6.85671031e-01
-2.72606432e-01 4.85566586e-01 -6.44793034e-01 6.53830349e-01
1.30141985e+00 -7.27974653e-01 1.56405330e-01 5.53167045e-01
1.38082886e+00 -3.95369865e-02 -8.87222946e-01 -7.85157025e-01
-4.55993265e-01 -5.71064532e-01 -8.46600085e-02 -7.05080271e-01
-1.24802911e+00 4.46382493e-01 6.15171134e-01 7.87653148e-01
1.02337492e+00 -1.13646917e-01 9.65617537e-01 1.98069438e-01
3.41924727e-01 -8.23829532e-01 5.32579124e-02 3.89160693e-01
8.54653418e-01 -1.37403536e+00 -2.14277599e-02 -6.04500234e-01
-7.66721547e-01 8.45255792e-01 8.42227697e-01 -3.64497185e-01
1.10867751e+00 -8.26248884e-01 -2.50820845e-01 -8.79294649e-02
-6.70848906e-01 -4.51560408e-01 8.70815337e-01 5.40996909e-01
5.21149814e-01 2.31216028e-01 -5.24158061e-01 1.24586487e+00
1.19037092e-01 -4.82569775e-03 3.28555167e-01 6.11277401e-01
-4.35320437e-01 -1.09325123e+00 2.84392118e-01 9.69943166e-01
-2.55956054e-01 -1.87128156e-01 -1.21398516e-01 3.74615222e-01
4.32912484e-02 1.16251659e+00 2.32108980e-02 -9.50551808e-01
4.17819649e-01 -1.85177654e-01 1.04554154e-01 -9.59419787e-01
-8.93920720e-01 4.94832732e-02 -2.47373115e-02 -6.47192895e-01
-1.99100152e-01 -4.76579845e-01 -8.76414418e-01 -4.70000118e-01
-5.98277807e-01 4.66823608e-01 2.91479081e-01 7.91590869e-01
6.57735586e-01 7.72258401e-01 1.05560768e+00 -7.84223914e-01
-2.36626446e-01 -1.00677812e+00 -1.00541866e+00 8.57896864e-01
2.12646306e-01 -5.68119347e-01 -4.57286030e-01 -1.92811042e-01] | [10.198661804199219, 5.613345623016357] |
22326299-1553-45c2-abef-2b2d3f2935c7 | gril-a-2-parameter-persistence-based | 2304.04970 | null | https://arxiv.org/abs/2304.04970v2 | https://arxiv.org/pdf/2304.04970v2.pdf | GRIL: A $2$-parameter Persistence Based Vectorization for Machine Learning | $1$-parameter persistent homology, a cornerstone in Topological Data Analysis (TDA), studies the evolution of topological features such as connected components and cycles hidden in data. It has been applied to enhance the representation power of deep learning models, such as Graph Neural Networks (GNNs). To enrich the representations of topological features, here we propose to study $2$-parameter persistence modules induced by bi-filtration functions. In order to incorporate these representations into machine learning models, we introduce a novel vector representation called Generalized Rank Invariant Landscape (GRIL) for $2$-parameter persistence modules. We show that this vector representation is $1$-Lipschitz stable and differentiable with respect to underlying filtration functions and can be easily integrated into machine learning models to augment encoding topological features. We present an algorithm to compute the vector representation efficiently. We also test our methods on synthetic and benchmark graph datasets, and compare the results with previous vector representations of $1$-parameter and $2$-parameter persistence modules. Further, we augment GNNs with GRIL features and observe an increase in performance indicating that GRIL can capture additional features enriching GNNs. We make the complete code for the proposed method available at https://github.com/soham0209/mpml-graph. | ['Tamal K. Dey', 'Shreyas N. Samaga', 'Soham Mukherjee', 'Cheng Xin'] | 2023-04-11 | null | null | null | null | ['topological-data-analysis'] | ['graphs'] | [-5.46107590e-02 7.30876550e-02 -2.11061910e-01 -2.04591639e-02
-1.69037551e-01 -7.54914343e-01 6.77022338e-01 4.24317569e-01
-7.95690119e-02 6.07909739e-01 1.39433980e-01 -5.07927239e-01
-5.79497576e-01 -1.22294593e+00 -1.03696489e+00 -6.28329337e-01
-9.88915265e-01 1.28581926e-01 2.92976022e-01 -5.98006725e-01
3.18572313e-01 7.32697904e-01 -1.40662539e+00 2.43733358e-02
7.01406419e-01 7.71093130e-01 -4.49033380e-02 7.33197689e-01
-4.27078120e-02 2.51448244e-01 7.66371489e-02 -1.54632464e-01
2.98000664e-01 1.83158088e-02 -9.02780950e-01 -4.77527350e-01
6.08744919e-01 2.27604732e-01 -9.30904448e-01 1.04461241e+00
1.93608895e-01 1.42740652e-01 9.39449370e-01 -1.10648811e+00
-6.21231377e-01 6.21339023e-01 -2.96750307e-01 3.87534559e-01
-5.79830892e-02 -3.99937434e-03 1.34212089e+00 -8.33487749e-01
9.52358603e-01 1.28050935e+00 8.51530075e-01 1.12899952e-01
-1.53214884e+00 -1.83575317e-01 -5.43298051e-02 2.54774243e-01
-1.16964340e+00 5.62371872e-02 1.16631389e+00 -8.93305719e-01
7.63783753e-01 2.61164665e-01 7.18151927e-01 7.21261501e-01
5.47195375e-01 4.59784061e-01 9.48477507e-01 -1.32709667e-01
-1.83629438e-01 -5.21667182e-01 5.94396472e-01 1.40476286e+00
5.01471698e-01 2.97054052e-02 -1.25441208e-01 -2.31253177e-01
1.25378251e+00 1.53097808e-01 -2.77703911e-01 -6.10220790e-01
-1.15955174e+00 1.01584101e+00 9.75724459e-01 3.74156147e-01
-6.91972673e-02 6.30178809e-01 4.19740438e-01 4.51664954e-01
3.70384574e-01 4.10824656e-01 -3.17183465e-01 1.06255993e-01
-2.35137060e-01 1.87329903e-01 4.63060766e-01 6.64103508e-01
1.27984524e+00 -1.78928822e-01 -1.48587480e-01 8.97471011e-01
1.33333549e-01 2.30179384e-01 1.69995338e-01 -8.05823624e-01
1.70352191e-01 1.10157073e+00 -3.85723829e-01 -1.53920794e+00
-6.61820531e-01 -4.71592456e-01 -1.22576237e+00 -4.47206944e-02
2.20485032e-01 5.46163261e-01 -9.30873692e-01 2.03848696e+00
4.85883504e-02 1.08276293e-01 -3.38282645e-01 5.67258775e-01
7.31962502e-01 6.40199006e-01 -2.14028239e-01 1.33058071e-01
1.26629889e+00 -6.69105053e-01 -2.28022948e-01 3.11932176e-01
1.00884545e+00 -3.53826463e-01 1.10365295e+00 -1.51263222e-01
-9.03808117e-01 -3.92118543e-01 -1.11026585e+00 -4.63499278e-02
-7.64256597e-01 -6.45874366e-02 9.08709824e-01 1.92567989e-01
-1.25298846e+00 1.19500411e+00 -8.20368648e-01 -3.76843393e-01
6.10479236e-01 5.21473467e-01 -5.21057010e-01 -3.37557457e-02
-1.28731835e+00 4.90966678e-01 2.47871593e-01 -5.58891594e-02
-5.43807387e-01 -6.40196860e-01 -9.24615502e-01 6.86793029e-02
1.14628583e-01 -7.02578902e-01 3.90830845e-01 -3.33763480e-01
-9.29761410e-01 7.09522247e-01 4.40615527e-02 -3.63073617e-01
1.09548032e-01 3.54301304e-01 -1.07744932e-01 1.60528675e-01
-1.46163493e-01 5.28119981e-01 3.94019037e-01 -8.54653776e-01
9.00238454e-02 -2.66828895e-01 2.32932031e-01 -2.98196346e-01
-2.01197341e-01 -5.85896552e-01 -1.24240205e-01 -8.18641365e-01
2.46545911e-01 -9.27334309e-01 -3.09774548e-01 1.09001420e-01
-3.65608543e-01 -3.27444762e-01 7.52850115e-01 -5.24704337e-01
1.21862519e+00 -1.84102941e+00 5.39765000e-01 5.85475504e-01
7.02530146e-01 2.27919668e-01 -4.86456871e-01 6.45710945e-01
-1.58351764e-01 5.46863437e-01 -6.55191123e-01 2.11085171e-01
1.21042676e-01 5.39301932e-02 -1.68126032e-01 6.47205174e-01
4.15710658e-01 1.11577487e+00 -7.37928569e-01 -1.75010771e-01
1.11076206e-01 5.91425002e-01 -7.53098190e-01 -2.32139647e-01
-3.18448007e-01 2.47708216e-01 -3.25178891e-01 5.77093124e-01
8.37985575e-01 -5.55864811e-01 1.98828220e-01 -3.15912366e-01
-1.64608508e-02 1.09487087e-01 -9.84406769e-01 1.42564440e+00
-2.02064082e-01 8.33428323e-01 -1.48195729e-01 -1.38343191e+00
1.05130994e+00 -2.52930462e-01 4.58474457e-01 -6.31046951e-01
6.04910776e-02 1.32383317e-01 6.47268966e-02 -2.42554054e-01
2.50686467e-01 1.91360489e-01 -1.67379513e-01 2.65336215e-01
2.10670635e-01 4.14795697e-01 3.07450056e-01 5.32371163e-01
1.51415372e+00 -2.76834726e-01 4.01642062e-02 -8.91904473e-01
6.77418530e-01 -1.93416640e-01 4.07002121e-01 4.73621488e-01
2.74643991e-02 1.52333513e-01 9.84315574e-01 -7.30351806e-01
-1.25728333e+00 -1.25398242e+00 -3.18192601e-01 1.07123625e+00
2.36848086e-01 -6.55781865e-01 -3.69485408e-01 -4.82810140e-01
2.01379701e-01 -2.95533359e-01 -8.10136437e-01 -1.45099416e-01
-8.99063289e-01 -7.19367206e-01 5.80300093e-01 5.39050400e-01
4.57087815e-01 -7.97990799e-01 4.32703644e-03 1.52428165e-01
1.16009661e-03 -6.14872873e-01 -4.75316733e-01 1.92108657e-02
-1.07489014e+00 -1.39743733e+00 -5.09589434e-01 -9.80922639e-01
8.02502632e-01 1.06166959e-01 9.31822419e-01 5.69299340e-01
-6.64259136e-01 4.29830968e-01 -1.94906861e-01 4.72429305e-01
-1.85428008e-01 1.65231019e-01 1.35613710e-01 -1.38768509e-01
-3.22015136e-01 -8.63014042e-01 -9.35057223e-01 3.54979903e-01
-1.11039460e+00 1.21736921e-01 4.05130178e-01 9.93564546e-01
6.96239173e-01 -1.10814556e-01 5.64190388e-01 -7.19485521e-01
5.45325041e-01 -4.59788769e-01 -6.84221387e-01 6.09547123e-02
-3.86779696e-01 6.26477540e-01 6.65875793e-01 -3.68607461e-01
-3.36819626e-02 -1.94961444e-01 -7.37705901e-02 -3.36591572e-01
3.53050619e-01 9.40029025e-01 -3.49470600e-02 -5.97885668e-01
5.18418670e-01 3.96356434e-02 1.72599107e-01 -4.76186275e-01
4.55519468e-01 1.04558900e-01 2.94266820e-01 -7.18708336e-01
8.70677054e-01 4.00516242e-01 6.15419865e-01 -1.14427841e+00
-5.77289835e-02 -3.07960868e-01 -6.97982371e-01 -1.05080284e-01
3.54365140e-01 -5.59838295e-01 -7.54369140e-01 5.22432387e-01
-9.48776305e-01 -4.98604178e-01 3.00819986e-02 1.21026941e-01
-8.61431658e-01 6.30679011e-01 -6.89996004e-01 -4.06480402e-01
-2.75762677e-01 -1.08680439e+00 6.89374387e-01 -5.33090383e-02
1.93983957e-01 -1.36926019e+00 5.46061993e-01 -1.79622024e-01
6.56023800e-01 7.68269360e-01 1.69523466e+00 -3.80244851e-01
-9.00172889e-01 -1.73121482e-01 -3.41832131e-01 3.27069610e-02
-7.80636966e-02 5.17464131e-02 -4.28271294e-01 -5.93426287e-01
-6.61662936e-01 5.88158593e-02 1.42659783e+00 4.11142737e-01
1.29475522e+00 -5.82993567e-01 -4.37179208e-01 8.19073677e-01
1.67120504e+00 -4.51528639e-01 7.25545287e-01 2.26801649e-01
7.37945199e-01 5.54236352e-01 2.55042259e-02 5.76519370e-02
3.56881887e-01 5.92297077e-01 4.79285508e-01 1.15654822e-02
-1.69002637e-01 -4.29216117e-01 2.60673463e-01 1.16943812e+00
-3.06646794e-01 9.90627706e-02 -1.08288717e+00 5.37654340e-01
-1.86667824e+00 -8.85124207e-01 -2.41086066e-01 2.00112438e+00
5.54846764e-01 -9.09067467e-02 3.28717917e-01 1.52115688e-01
9.52182472e-01 3.75463396e-01 -4.07305300e-01 -3.46209168e-01
-3.70113611e-01 4.35191989e-01 5.60197890e-01 8.50757718e-01
-1.06215847e+00 1.03014958e+00 5.59574175e+00 7.87480295e-01
-1.23058200e+00 -1.78084359e-01 3.84856790e-01 3.45714658e-01
-6.76904082e-01 9.32088271e-02 -2.54023820e-01 3.26332122e-01
7.11878777e-01 -2.16325372e-01 4.13563520e-01 4.48983073e-01
-1.22916877e-01 5.20574212e-01 -9.67885792e-01 7.55957127e-01
-3.03059101e-01 -2.02108812e+00 3.22757334e-01 4.63551760e-01
5.31294525e-01 1.32261068e-01 1.79648980e-01 2.51112282e-01
7.39652216e-02 -1.23493683e+00 5.47460876e-02 6.56513035e-01
1.08582091e+00 -7.16858983e-01 4.34828907e-01 -4.42229211e-02
-1.75794053e+00 -5.34074642e-02 -6.37358069e-01 -5.43954037e-02
-1.30386069e-01 4.22517687e-01 -5.40497422e-01 5.45534551e-01
4.59487051e-01 9.56262946e-01 -9.57134485e-01 1.20429504e+00
2.52251744e-01 3.86198997e-01 -2.58698970e-01 -1.42617643e-01
1.78451076e-01 -3.69266570e-01 7.78620183e-01 1.14325786e+00
3.47944379e-01 3.87584083e-02 -5.67532964e-02 1.00573361e+00
-3.80748808e-01 2.55616277e-01 -1.12739050e+00 -3.98664504e-01
3.02089334e-01 1.13822234e+00 -8.47409964e-01 -3.72810811e-02
-1.67167615e-02 5.81245363e-01 6.34376168e-01 2.86987662e-01
-6.46796346e-01 -7.37579286e-01 8.98766875e-01 5.00490129e-01
3.93948972e-01 -6.94194317e-01 8.02278146e-02 -1.22254491e+00
-5.78188673e-02 -3.17516327e-01 2.87364036e-01 -3.82309258e-01
-1.41037202e+00 5.93130410e-01 -2.94633117e-02 -9.42736447e-01
2.65266091e-01 -1.10827732e+00 -8.58369827e-01 5.45001447e-01
-1.34160149e+00 -1.16048431e+00 -2.42889926e-01 5.66610932e-01
4.22301143e-02 3.41753364e-02 6.27784193e-01 1.58873528e-01
-4.95081991e-01 5.53750575e-01 3.96884471e-01 3.10821891e-01
1.27309516e-01 -1.25899732e+00 6.26923084e-01 4.75513250e-01
8.47005919e-02 8.19891036e-01 4.28160369e-01 -6.80885136e-01
-1.83817732e+00 -1.28725111e+00 4.95352238e-01 -3.70494932e-01
9.98781681e-01 -5.58900535e-01 -1.09654760e+00 6.87976182e-01
-2.38698289e-01 1.96542755e-01 5.30271113e-01 4.80183214e-02
-5.77197492e-01 6.06502945e-05 -1.01511931e+00 5.16307414e-01
1.42473233e+00 -5.85071921e-01 -1.20306864e-01 2.29220346e-01
1.05762315e+00 -6.90704910e-03 -1.11434782e+00 8.55915666e-01
5.11930823e-01 -6.44218206e-01 1.14317298e+00 -7.51690149e-01
3.00252169e-01 -2.12995321e-01 -4.36524570e-01 -1.16612625e+00
-7.63872981e-01 -5.11577964e-01 -2.04083487e-01 7.07289934e-01
3.78728002e-01 -9.57819402e-01 5.42841375e-01 -1.37525007e-01
-1.27512723e-01 -1.02601039e+00 -1.11951578e+00 -7.55687594e-01
5.45561969e-01 -1.16225399e-01 5.13581693e-01 9.03716683e-01
6.08455911e-02 1.42126098e-01 -3.58420052e-02 6.10992834e-02
7.33384490e-01 9.92089435e-02 6.46637261e-01 -1.61493409e+00
-5.45402654e-02 -8.96014690e-01 -1.08641779e+00 -8.83204043e-01
1.52240366e-01 -1.63298261e+00 -4.92451072e-01 -1.41024363e+00
2.85190940e-01 -6.85423911e-01 -4.99259174e-01 2.54773468e-01
1.43429995e-01 3.33618373e-01 -4.70407354e-03 2.02569455e-01
-4.75357503e-01 7.36715257e-01 1.23077822e+00 -2.81607211e-01
-1.44962206e-01 -3.05879354e-01 -5.20249963e-01 6.16282344e-01
9.15041149e-01 -2.77387984e-02 -5.47509082e-02 -1.37713835e-01
5.26249945e-01 -4.07275371e-02 7.28494704e-01 -9.05268610e-01
2.32786220e-02 -2.81465109e-02 1.86914787e-01 -5.23500860e-01
1.81956664e-01 -3.04070801e-01 1.18975297e-01 6.03999138e-01
-2.12871909e-01 2.79409558e-01 2.88461983e-01 8.43160033e-01
1.76311769e-02 2.20704988e-01 8.27787220e-01 6.15873858e-02
-3.91640753e-01 8.01873028e-01 -2.63997406e-01 -5.12706041e-02
7.54095316e-01 -6.22071102e-02 -9.78463948e-01 -7.72572085e-02
-7.73734748e-01 1.48256481e-01 7.44670153e-01 2.80293256e-01
7.82877147e-01 -1.73423004e+00 -4.76815462e-01 3.07577610e-01
1.13722041e-01 -3.10871482e-01 3.75873119e-01 7.80328870e-01
-9.21496332e-01 4.90526974e-01 -4.75254625e-01 -7.19217420e-01
-1.05412424e+00 5.10282993e-01 3.75924259e-01 -3.75197232e-01
-7.28789926e-01 5.11798322e-01 2.34488562e-01 -5.05284905e-01
9.76110622e-02 -4.09532666e-01 -1.57132760e-01 -1.15625128e-01
7.98525475e-03 5.16656399e-01 -1.45888135e-01 -7.32086241e-01
-3.95938814e-01 1.06512702e+00 1.51762918e-01 2.12576255e-01
1.57801414e+00 1.58177853e-01 -6.13875270e-01 3.37939769e-01
1.54953444e+00 -1.27749920e-01 -9.04587865e-01 -2.37878934e-01
-5.92226768e-03 -2.45918646e-01 -2.81827837e-01 -2.00409114e-01
-1.05061567e+00 9.61764336e-01 6.14618123e-01 3.59062821e-01
7.22041667e-01 3.25284243e-01 5.35706580e-01 5.41245043e-01
4.04876798e-01 -5.50399303e-01 4.75202858e-01 6.82630777e-01
1.09392440e+00 -8.89647305e-01 -2.58340776e-01 -3.93845409e-01
1.00165762e-01 1.16291654e+00 4.15042996e-01 -7.13011563e-01
1.02471006e+00 -3.03699732e-01 -5.69357991e-01 -5.65465927e-01
-7.60343075e-01 -2.46380359e-01 4.51094657e-01 3.20767820e-01
2.69077778e-01 3.31662387e-01 -4.08384234e-01 4.08049107e-01
-1.95133939e-01 -4.48940545e-01 6.76614344e-01 7.74617076e-01
-7.01818168e-01 -1.21786809e+00 9.19435471e-02 7.54756033e-01
-9.83374566e-02 -1.15671113e-01 -4.18154210e-01 6.94500089e-01
-1.22914918e-01 2.29397342e-01 1.24024577e-01 -6.79299235e-01
3.55264395e-02 1.92285294e-03 5.18035471e-01 -3.67216021e-01
-4.94843833e-02 -2.63066322e-01 -1.81819513e-01 -5.36518037e-01
-1.03222392e-01 -2.30200052e-01 -1.08735895e+00 -6.77832067e-01
-9.80483666e-02 9.94347706e-02 5.48040152e-01 4.18785244e-01
7.70094156e-01 3.99044186e-01 6.83683455e-01 -9.17908490e-01
-2.44843923e-02 -8.20172131e-01 -5.24987698e-01 4.47779268e-01
5.44331372e-01 -1.00247669e+00 -6.26117349e-01 -3.04515243e-01] | [6.932113170623779, 6.085634231567383] |
f791710b-9f7d-456d-955d-60a99cffcace | tsetlin-machine-embedding-representing-words | 2301.00709 | null | https://arxiv.org/abs/2301.00709v1 | https://arxiv.org/pdf/2301.00709v1.pdf | Tsetlin Machine Embedding: Representing Words Using Logical Expressions | Embedding words in vector space is a fundamental first step in state-of-the-art natural language processing (NLP). Typical NLP solutions employ pre-defined vector representations to improve generalization by co-locating similar words in vector space. For instance, Word2Vec is a self-supervised predictive model that captures the context of words using a neural network. Similarly, GLoVe is a popular unsupervised model incorporating corpus-wide word co-occurrence statistics. Such word embedding has significantly boosted important NLP tasks, including sentiment analysis, document classification, and machine translation. However, the embeddings are dense floating-point vectors, making them expensive to compute and difficult to interpret. In this paper, we instead propose to represent the semantics of words with a few defining words that are related using propositional logic. To produce such logical embeddings, we introduce a Tsetlin Machine-based autoencoder that learns logical clauses self-supervised. The clauses consist of contextual words like "black," "cup," and "hot" to define other words like "coffee," thus being human-understandable. We evaluate our embedding approach on several intrinsic and extrinsic benchmarks, outperforming GLoVe on six classification tasks. Furthermore, we investigate the interpretability of our embedding using the logical representations acquired during training. We also visualize word clusters in vector space, demonstrating how our logical embedding co-locate similar words. | ['Jivitesh Sharma', 'Rohan Yadav', 'Lei Jiao', 'Ole-Christoffer Granmo', 'Bimal Bhattarai'] | 2023-01-02 | null | null | null | null | ['document-classification'] | ['natural-language-processing'] | [-8.58069137e-02 8.64792019e-02 -4.92515385e-01 -7.80211985e-01
-1.48277804e-01 -5.73455989e-01 7.49322772e-01 8.21146309e-01
-5.08976936e-01 4.48663861e-01 5.95979989e-01 -5.72362125e-01
5.01766130e-02 -1.05444038e+00 -7.86727548e-01 -4.17548299e-01
5.94681054e-02 4.91264641e-01 -4.78939354e-01 -4.51099992e-01
1.64040923e-01 9.66017693e-02 -1.35522854e+00 5.89458048e-01
6.03073955e-01 8.94902527e-01 -9.28132087e-02 3.34399611e-01
-7.72520661e-01 6.90823257e-01 -6.55005455e-01 -5.95042586e-01
-2.06979945e-01 -1.42258584e-01 -7.67078221e-01 -3.21147680e-01
2.67950594e-02 7.49124587e-02 -1.85566574e-01 1.08064115e+00
-2.02153370e-01 2.69934386e-01 8.94772828e-01 -1.15642083e+00
-1.55021811e+00 9.61316645e-01 -1.17890425e-01 -6.36696294e-02
3.78787041e-01 6.56916499e-02 1.84894443e+00 -1.01277244e+00
5.49121559e-01 1.39085090e+00 3.98895442e-01 4.20762211e-01
-1.31750917e+00 -3.63002717e-01 1.33452341e-01 5.45571685e-01
-1.30594921e+00 1.66493341e-01 9.34050322e-01 -4.46953446e-01
1.33040762e+00 2.50833631e-01 7.17567444e-01 1.32237566e+00
3.93193364e-01 7.78324842e-01 6.90445065e-01 -6.98714256e-01
3.98005515e-01 4.35519010e-01 7.03147650e-01 6.30519807e-01
4.32076395e-01 -1.78562716e-01 -3.80336523e-01 -6.36046007e-02
2.25051329e-01 5.08015752e-01 -1.84037760e-01 -2.81813443e-01
-1.17770350e+00 1.44470930e+00 7.70741999e-01 6.52458787e-01
-2.93723732e-01 4.86609787e-01 6.98137939e-01 1.45234913e-01
5.22657216e-01 8.92381072e-01 -5.39250195e-01 2.87111790e-04
-4.16569412e-01 1.36797398e-01 6.76070869e-01 7.23400474e-01
9.87583101e-01 -2.16997811e-03 -1.56372994e-01 7.27784038e-01
5.99906206e-01 3.75547320e-01 9.93799627e-01 -3.56046468e-01
1.93694666e-01 1.01291275e+00 -3.92570347e-01 -1.32316339e+00
-1.92620397e-01 -1.33462772e-01 -8.33106399e-01 -2.37412840e-01
-1.76899999e-01 1.43972933e-01 -7.56810784e-01 1.50176692e+00
-1.16210401e-01 7.37815797e-02 4.41868007e-01 7.83417225e-01
8.76441777e-01 1.07907581e+00 1.39268294e-01 2.06233233e-01
1.71631539e+00 -9.94568527e-01 -9.18951333e-01 -5.00216961e-01
9.36040998e-01 -2.44349450e-01 1.46579051e+00 2.39748850e-01
-4.97640342e-01 -5.01394212e-01 -1.19886649e+00 -3.89927894e-01
-1.13732171e+00 -1.21360861e-01 1.01483262e+00 3.81545395e-01
-5.90412796e-01 6.92907333e-01 -6.15049779e-01 -2.81588405e-01
3.81512910e-01 1.13339372e-01 -6.42711461e-01 -4.89946119e-02
-1.53840077e+00 1.03355014e+00 7.96239316e-01 -1.34325117e-01
-3.07110250e-01 -7.84104764e-01 -1.60851300e+00 3.30778897e-01
-4.24822532e-02 -5.32974064e-01 6.81250572e-01 -9.25865173e-01
-1.24173403e+00 7.70431578e-01 -3.75196785e-01 -7.55136073e-01
-4.37836885e-01 -3.71209741e-01 -6.49836957e-01 -7.56543726e-02
-5.86237498e-02 4.97815669e-01 7.95657218e-01 -1.07268572e+00
-1.57470912e-01 -2.05353707e-01 2.37415388e-01 -9.36552361e-02
-9.55118358e-01 -1.07192144e-01 -1.19252801e-01 -6.93493009e-01
-1.62988991e-01 -5.98578274e-01 -1.18708447e-01 -1.40509918e-01
-3.40015143e-01 -7.50555813e-01 6.58423781e-01 -2.39139467e-01
1.28212047e+00 -2.38433957e+00 1.82616979e-01 2.68798351e-01
4.46986288e-01 8.29009712e-02 -1.54485345e-01 4.35110956e-01
-4.01030153e-01 5.33969879e-01 -2.11524978e-01 -2.69481093e-01
3.86385530e-01 8.43398452e-01 -7.60753512e-01 2.28703320e-01
6.99900329e-01 1.17518556e+00 -1.06857789e+00 -2.43036062e-01
3.94587457e-01 5.38893938e-01 -8.41327548e-01 1.38347745e-01
-4.43877786e-01 -3.99799049e-01 -1.66654333e-01 3.46465051e-01
1.85314670e-01 -3.97915572e-01 3.54548633e-01 -4.06959683e-01
3.27397227e-01 5.43291748e-01 -8.60646188e-01 1.52628970e+00
-7.18371987e-01 1.01444292e+00 -8.15976441e-01 -1.39516556e+00
1.13972759e+00 5.69553524e-02 4.89621349e-02 -3.84160340e-01
4.27322328e-01 -4.56856266e-02 -8.38475749e-02 -4.92924064e-01
8.35773885e-01 -3.28652978e-01 -3.86941642e-01 4.21749622e-01
4.31160122e-01 -2.88215995e-01 2.47680992e-01 2.80381471e-01
9.69718575e-01 -3.19916308e-01 6.46585107e-01 -1.67265907e-01
5.64291596e-01 -4.57418114e-02 3.15777570e-01 3.02245736e-01
7.70109743e-02 3.62581313e-01 7.55475581e-01 -7.54205704e-01
-7.67376900e-01 -1.14778852e+00 -1.63779020e-01 1.09361541e+00
1.39757439e-01 -1.08525896e+00 -3.33106011e-01 -6.31864846e-01
3.05066466e-01 1.19418502e+00 -1.06800628e+00 -4.03072774e-01
-3.32509458e-01 -3.95770371e-01 5.17695487e-01 7.62891889e-01
-2.04661101e-01 -1.14729714e+00 -2.95633316e-01 1.10471748e-01
1.07419945e-01 -9.94717538e-01 -1.26948953e-01 6.83034003e-01
-5.18723071e-01 -8.92408490e-01 -4.83291671e-02 -8.28831673e-01
8.22727203e-01 -1.31991252e-01 1.32759511e+00 3.69831137e-02
-1.45117804e-01 1.93731964e-01 -7.83205867e-01 -4.78130102e-01
-2.61595309e-01 -1.94657147e-01 4.43254918e-01 7.06201419e-02
9.99334991e-01 -3.33194017e-01 -1.40028849e-01 -3.27435493e-01
-1.21054554e+00 -2.61341661e-01 3.11368644e-01 1.17576969e+00
6.97201788e-01 -7.47990906e-02 -1.80057231e-02 -8.98276091e-01
1.07052314e+00 -5.75453699e-01 -1.54818237e-01 2.53827572e-01
-5.75156033e-01 6.97193384e-01 1.07087255e+00 -5.29249012e-01
-3.20980817e-01 -3.93116504e-01 3.18891928e-02 -5.65357089e-01
-8.42911154e-02 8.56160104e-01 -4.21848632e-02 4.16982740e-01
7.61760235e-01 2.38250673e-01 -5.24911247e-02 -1.28050506e-01
9.48859453e-01 6.27938509e-01 3.48612428e-01 -6.28785968e-01
8.26822758e-01 3.45986694e-01 -3.48425508e-01 -8.99386227e-01
-8.28248084e-01 -3.90968621e-01 -3.70259643e-01 4.56527442e-01
1.00505507e+00 -6.38475239e-01 -6.80534303e-01 -3.49664479e-01
-1.58858573e+00 1.05218410e-01 -5.87314427e-01 5.39392054e-01
-2.36946031e-01 3.49584639e-01 -3.82691085e-01 -4.06897604e-01
-4.04973388e-01 -1.00725543e+00 8.37920845e-01 8.99803489e-02
-8.58633697e-01 -1.38963008e+00 2.56522089e-01 -8.20619427e-03
2.97801971e-01 6.06315583e-02 1.46779251e+00 -1.07833207e+00
1.32687822e-01 -2.84102678e-01 -1.24971129e-01 7.28668094e-01
5.74879311e-02 6.76141083e-02 -7.56360948e-01 1.05460972e-01
-2.55928874e-01 -2.70803064e-01 8.73901129e-01 4.90217023e-02
1.46173966e+00 -4.88656521e-01 -2.12825492e-01 6.59580708e-01
1.50667751e+00 -1.11669295e-01 4.69840229e-01 3.65967453e-01
7.47150421e-01 5.89337409e-01 3.06589574e-01 3.50873291e-01
3.24837714e-01 2.11395472e-01 3.69702995e-01 9.44691971e-02
4.51296240e-01 -4.87394631e-01 4.48672175e-01 1.10808837e+00
2.18678892e-01 -2.57731259e-01 -1.03394699e+00 6.79048538e-01
-1.63158786e+00 -6.99583352e-01 3.56317200e-02 1.42346513e+00
9.57383096e-01 2.51819938e-01 -6.73059821e-01 2.44523391e-01
2.65569091e-01 4.09281760e-01 -2.05609515e-01 -9.79248822e-01
-8.91444907e-02 9.13407207e-01 2.85285562e-01 5.48000395e-01
-1.01623380e+00 1.19877756e+00 5.05739737e+00 6.04775906e-01
-1.30578995e+00 7.35773444e-02 3.52199525e-01 1.48431286e-01
-9.28309560e-01 1.33469247e-03 -5.05669355e-01 3.75198632e-01
8.85863066e-01 -3.08709741e-01 2.80929714e-01 1.14028156e+00
-1.77338541e-01 2.73371696e-01 -1.60460043e+00 1.15723860e+00
4.35708582e-01 -1.67961156e+00 4.41856772e-01 -1.79956049e-01
4.44644243e-01 -1.90695256e-01 -1.00127617e-02 7.65693069e-01
2.92699993e-01 -1.47721183e+00 4.75453943e-01 2.33676270e-01
6.76559627e-01 -7.29047596e-01 9.33888018e-01 -2.33678650e-02
-1.09018362e+00 -9.78338346e-03 -7.11985469e-01 -3.01840991e-01
4.23902832e-02 6.90864563e-01 -7.84448326e-01 2.85004467e-01
5.59706509e-01 9.57315922e-01 -5.21441996e-01 -2.63196751e-02
-8.65757585e-01 6.07827485e-01 -6.29610792e-02 -6.74850941e-01
5.00324130e-01 -1.13871664e-01 1.94667891e-01 1.40431499e+00
1.58308018e-02 3.42631675e-02 -1.42601937e-01 1.27777791e+00
-3.65857512e-01 2.40469724e-01 -8.53808105e-01 -5.50302327e-01
2.83871680e-01 1.19741571e+00 -2.84998447e-01 -5.84506273e-01
-3.34946752e-01 1.04424763e+00 4.08055365e-01 3.07202131e-01
-8.94297183e-01 -7.92538822e-01 1.30496037e+00 -2.87917286e-01
2.34642655e-01 -4.48343724e-01 -4.46670979e-01 -1.29439390e+00
2.20971126e-02 -5.98131478e-01 6.18992597e-02 -7.41072178e-01
-1.47717547e+00 6.32691562e-01 -1.40553176e-01 -1.06266439e+00
-3.75202179e-01 -1.31867862e+00 -7.41276741e-01 6.50956452e-01
-1.43528163e+00 -8.54744613e-01 -9.07256380e-02 4.30369824e-01
4.97468323e-01 -3.31674784e-01 1.32037377e+00 -1.07303366e-01
-4.55384284e-01 7.81597257e-01 1.06075853e-01 6.14966035e-01
6.29892290e-01 -1.42185950e+00 3.50892782e-01 5.30655503e-01
7.48308539e-01 1.35762227e+00 6.23343945e-01 -2.02513337e-01
-1.72017860e+00 -1.12613440e+00 1.26864052e+00 -5.00935018e-01
1.08687842e+00 -6.09003246e-01 -1.09667921e+00 8.18316042e-01
4.01270360e-01 3.98632526e-01 1.21812379e+00 4.94672745e-01
-8.31210375e-01 -1.31290615e-01 -6.88073099e-01 7.66004682e-01
5.27287006e-01 -9.31579471e-01 -1.30181098e+00 4.46792752e-01
1.22110581e+00 7.32555687e-02 -7.07406223e-01 5.32130115e-02
4.42092299e-01 -4.50848639e-01 9.84970033e-01 -1.13858759e+00
1.16182554e+00 -2.44206727e-01 -3.61533552e-01 -1.69761992e+00
-3.38667750e-01 -4.71160300e-02 -3.57925951e-01 9.84603167e-01
5.74973762e-01 -6.40200138e-01 6.04816973e-01 3.86372030e-01
7.32998550e-02 -9.41060126e-01 -5.18100798e-01 -7.70239711e-01
2.09463969e-01 -9.05046999e-01 6.67861879e-01 1.24795127e+00
5.92599630e-01 6.48796499e-01 5.76586910e-02 4.57869023e-02
2.17760414e-01 7.96726868e-02 5.36547661e-01 -1.04234540e+00
-3.03752899e-01 -5.02281785e-01 -9.35463846e-01 -9.98811305e-01
8.49133730e-01 -1.44925737e+00 -1.48021067e-02 -1.35102177e+00
-1.22237906e-01 -8.70079622e-02 -5.32889426e-01 6.77674770e-01
-2.09501609e-01 5.47524961e-03 7.64508173e-02 -2.15308443e-01
-3.72701555e-01 8.94777954e-01 6.23255134e-01 -7.74045169e-01
4.57149632e-02 -9.10371840e-01 -8.45969915e-01 7.82103002e-01
6.88562095e-01 -4.05872941e-01 -2.12167069e-01 -7.98892260e-01
5.68830669e-01 -6.60376787e-01 2.70561993e-01 -6.21651232e-01
9.90511999e-02 -2.90725887e-01 2.82555610e-01 -3.21544021e-01
2.79461056e-01 -9.70093846e-01 -5.28979361e-01 3.65257859e-01
-5.22193611e-01 1.93145886e-01 1.22522958e-01 5.13115346e-01
-6.20589077e-01 -3.88914406e-01 3.76151323e-01 6.20199144e-02
-1.07136822e+00 -3.71589586e-02 -2.59145796e-01 6.52111694e-02
1.03123772e+00 -1.28988056e-02 -1.04969457e-01 -1.70416817e-01
-2.67109662e-01 1.79904029e-01 1.92090064e-01 7.22507834e-01
9.47527230e-01 -1.46157372e+00 -3.55072021e-01 4.12532896e-01
6.47927403e-01 -9.62417666e-03 -1.99804544e-01 2.28879198e-01
-7.14829743e-01 5.18652737e-01 -3.52683999e-02 -5.24449885e-01
-9.34934556e-01 6.85838521e-01 -8.71753544e-02 -2.65892148e-01
-5.54227233e-01 1.01570845e+00 1.25088796e-01 -5.53755403e-01
1.31874904e-01 -9.29160595e-01 -4.12065387e-01 9.72644463e-02
6.29328966e-01 -2.91568071e-01 -1.92504704e-01 -3.46090674e-01
-5.86302280e-01 4.58377570e-01 -4.68488336e-02 6.49558902e-02
1.42149007e+00 4.17713434e-01 -5.01198351e-01 7.35037625e-01
1.71233082e+00 -1.06616192e-01 -3.17759812e-01 -3.48347276e-01
2.38866255e-01 -1.33200437e-01 2.14455416e-03 -1.66243166e-01
-7.73750067e-01 1.16211677e+00 1.49508268e-01 1.34423643e-01
5.55183232e-01 1.16650775e-01 7.88281977e-01 8.73369038e-01
3.27197425e-02 -1.05906045e+00 2.15624720e-01 1.09054625e+00
8.74268234e-01 -1.30740070e+00 4.84650545e-02 -1.80325702e-01
-7.31094003e-01 1.49067032e+00 3.38483930e-01 -4.31494862e-01
7.34836996e-01 5.74954748e-02 -1.06445916e-01 -3.09650093e-01
-7.57674932e-01 -7.90683255e-02 4.23863560e-01 4.14192110e-01
6.05232298e-01 4.87971336e-01 -4.53962028e-01 1.07240462e+00
-5.49784839e-01 -3.85152489e-01 3.31717223e-01 7.46707797e-01
-1.81043699e-01 -1.17929363e+00 -1.62169188e-01 5.16289592e-01
-5.01876213e-02 -5.18709958e-01 -3.63867551e-01 5.55430830e-01
1.76382437e-01 7.24681258e-01 4.77800131e-01 -6.06263340e-01
1.56814083e-01 2.86980718e-01 3.11573278e-02 -9.69669223e-01
-5.25550723e-01 -7.15176880e-01 -1.89609170e-01 -4.54343319e-01
-8.26157480e-02 -9.46425796e-02 -1.59557962e+00 -3.02869916e-01
-1.03679076e-01 2.49990076e-01 8.39141846e-01 1.16290092e+00
2.45019495e-01 6.99468315e-01 4.65063095e-01 -3.73119563e-01
-5.17187238e-01 -9.85392332e-01 -4.27829832e-01 8.03971827e-01
1.77796438e-01 -6.04758024e-01 -5.20096302e-01 1.22581430e-01] | [10.266483306884766, 8.630134582519531] |
70c5038f-1ede-436b-96f5-e3a4157315c4 | word-segmentation-on-discovered-phone-units | 2202.11929 | null | https://arxiv.org/abs/2202.11929v2 | https://arxiv.org/pdf/2202.11929v2.pdf | Word Segmentation on Discovered Phone Units with Dynamic Programming and Self-Supervised Scoring | Recent work on unsupervised speech segmentation has used self-supervised models with phone and word segmentation modules that are trained jointly. This paper instead revisits an older approach to word segmentation: bottom-up phone-like unit discovery is performed first, and symbolic word segmentation is then performed on top of the discovered units (without influencing the lower level). To do this, I propose a new unit discovery model, a new symbolic word segmentation model, and then chain the two models to segment speech. Both models use dynamic programming to minimize segment costs from a self-supervised network with an additional duration penalty that encourages longer units. Concretely, for acoustic unit discovery, duration-penalized dynamic programming (DPDP) is used with a contrastive predictive coding model as the scoring network. For word segmentation, DPDP is applied with an autoencoding recurrent neural as the scoring network. The two models are chained in order to segment speech. This approach gives comparable word segmentation results to state-of-the-art joint self-supervised segmentation models on an English benchmark. On French, Mandarin, German and Wolof data, it outperforms previous systems on the ZeroSpeech benchmarks. Analysis shows that the chained DPDP system segments shorter filler words well, but longer words might require some external top-down signal. | ['Herman Kamper'] | 2022-02-24 | null | null | null | null | ['acoustic-unit-discovery'] | ['speech'] | [ 6.39882207e-01 6.36642754e-01 -5.25319278e-01 -4.74837691e-01
-7.66547024e-01 -5.70107281e-01 2.74968445e-01 1.28141657e-01
-8.11091006e-01 4.93472844e-01 1.54948696e-01 -6.54528022e-01
4.19061393e-01 -4.81496841e-01 -6.01480901e-01 -6.37998879e-01
2.13265225e-01 7.07665503e-01 7.11575687e-01 8.46965704e-04
1.88889012e-01 9.11546573e-02 -1.23778331e+00 2.38720134e-01
9.71078694e-01 5.51963329e-01 6.23566091e-01 7.34711170e-01
-4.96149451e-01 3.22466642e-01 -5.45879662e-01 -1.15965400e-02
1.90861106e-01 -7.76922107e-01 -1.04649115e+00 3.32057267e-01
-1.83268666e-01 -7.71314800e-02 2.30774462e-01 8.74032021e-01
4.43377942e-01 3.36756945e-01 5.51935673e-01 -5.46279490e-01
-2.82735050e-01 1.65223074e+00 -2.67222762e-01 1.51568830e-01
1.52568519e-01 -1.43157393e-01 1.39196014e+00 -7.36853063e-01
4.92789268e-01 1.31539369e+00 4.98181254e-01 7.36639977e-01
-1.48782945e+00 -4.02081430e-01 5.77337265e-01 -2.30694890e-01
-1.17845702e+00 -4.11121786e-01 7.00938880e-01 -2.51209259e-01
1.62392616e+00 1.12988196e-01 7.58378565e-01 8.63992870e-01
-1.73083156e-01 1.14050388e+00 9.01739180e-01 -9.43335235e-01
3.73862356e-01 2.13357791e-01 6.67040169e-01 4.88000214e-01
-2.95847178e-01 1.59610018e-01 -2.18916044e-01 1.83850989e-01
4.78374630e-01 -4.98272866e-01 1.19994543e-01 -4.57300693e-02
-1.01135159e+00 9.88624275e-01 -5.28828017e-02 6.99440777e-01
-3.52549314e-01 -6.89127762e-03 5.25267065e-01 1.48193017e-01
5.88755667e-01 4.11911190e-01 -8.01643312e-01 -1.46613657e-01
-1.55209708e+00 -8.50568265e-02 9.54309762e-01 8.56335282e-01
8.68452311e-01 3.47916633e-01 -2.21591339e-01 1.19812346e+00
5.95228076e-01 2.09961578e-01 9.31398213e-01 -7.58696616e-01
2.45116368e-01 2.74740309e-01 -3.78650755e-01 -5.59145845e-02
-3.73919815e-01 -4.27149773e-01 -1.47726655e-01 -1.29393354e-01
3.03299278e-01 -2.76329696e-01 -1.62397695e+00 1.79299891e+00
7.53036439e-02 1.30763248e-01 7.49128833e-02 5.25883734e-01
4.93944287e-01 1.22105730e+00 3.55827063e-01 -5.62037885e-01
1.33956575e+00 -1.31781733e+00 -8.93233180e-01 -5.84102631e-01
8.12053859e-01 -6.29837692e-01 8.75279307e-01 5.44722855e-01
-1.30541360e+00 -7.08685160e-01 -9.75907564e-01 -3.97490105e-03
-6.04105651e-01 1.01591043e-01 2.01721683e-01 9.92557645e-01
-1.06252408e+00 5.82403898e-01 -1.06127715e+00 -3.41138154e-01
-4.58132289e-02 5.90795279e-01 2.40709946e-01 4.59240377e-01
-1.44857287e+00 7.70063758e-01 8.93307388e-01 4.34203260e-02
-7.34216809e-01 -3.16540331e-01 -9.89676952e-01 -1.85783654e-01
3.76446456e-01 -2.83149570e-01 1.50325096e+00 -1.28125298e+00
-1.94601548e+00 8.97691369e-01 -2.92399615e-01 -9.76318598e-01
4.92605120e-02 -2.30979308e-01 -2.82076925e-01 8.15581381e-02
1.19771123e-01 1.14692700e+00 1.07185495e+00 -1.29622710e+00
-6.64502800e-01 4.27930150e-03 -4.21808392e-01 3.10580045e-01
3.95449251e-03 1.35286227e-01 -6.41361952e-01 -8.00396085e-01
1.89989746e-01 -1.02694905e+00 -3.54020268e-01 -1.12393117e+00
-7.13118613e-01 -5.35558403e-01 6.23750329e-01 -8.24704826e-01
1.77110028e+00 -2.06306958e+00 5.23132265e-01 3.72631729e-01
-3.34491223e-01 4.91770685e-01 -9.03492048e-02 2.08085299e-01
-3.62014443e-01 2.22334772e-01 -8.36233020e-01 -8.93157244e-01
8.10366031e-03 7.54556537e-01 -3.04746181e-01 -5.71072148e-03
2.16552585e-01 9.73783672e-01 -7.01237440e-01 -4.93512303e-01
1.65068015e-01 1.67088330e-01 -5.90162694e-01 5.74041866e-02
-5.25480449e-01 1.60185322e-01 6.73910826e-02 5.52939713e-01
2.64707953e-01 3.97242725e-01 1.84107453e-01 1.88735232e-01
-4.69137490e-01 8.41229677e-01 -1.09580803e+00 1.72016668e+00
-4.14004833e-01 2.89840847e-01 4.72234264e-02 -1.28724766e+00
8.37804735e-01 3.96957695e-01 1.93597540e-01 -3.04487824e-01
2.21298888e-01 5.19707978e-01 2.94520885e-01 -1.37516826e-01
4.51967776e-01 -2.84114838e-01 -2.45908901e-01 3.28815430e-01
4.51368481e-01 -5.44906437e-01 3.98631722e-01 2.88689137e-02
7.30347276e-01 3.17300588e-01 1.28794089e-01 -2.22007602e-01
5.53099394e-01 1.24403745e-01 6.48868561e-01 6.76583171e-01
-9.73152444e-02 7.25569546e-01 5.11873007e-01 2.14409307e-01
-8.02167296e-01 -9.79451120e-01 -8.49209204e-02 1.65137243e+00
-2.34162927e-01 -3.72640908e-01 -1.28415203e+00 -6.78663135e-01
-2.75676370e-01 1.09558880e+00 -3.74540210e-01 1.08648092e-01
-1.04947495e+00 -6.36691511e-01 6.71016097e-01 4.69854176e-01
5.90206087e-02 -1.51266062e+00 -2.99583346e-01 6.54005826e-01
-1.33752584e-01 -1.05241156e+00 -5.06558359e-01 1.05132830e+00
-9.18832660e-01 -3.26289117e-01 -9.73701835e-01 -1.36705196e+00
4.42257971e-01 -2.84325421e-01 8.79631817e-01 -1.56725153e-01
1.03983991e-01 1.66409448e-01 -5.91868222e-01 -2.65611112e-01
-6.86033368e-01 6.29101217e-01 -6.28411397e-02 -3.62972096e-02
4.01703984e-01 -2.99471319e-01 -1.47595167e-01 1.59782171e-01
-9.15878654e-01 -1.26165822e-01 5.64174533e-01 8.74391377e-01
9.03082013e-01 -2.59985328e-01 6.36199594e-01 -1.08480692e+00
5.32018363e-01 -2.52015561e-01 -4.82795626e-01 -9.21144485e-02
-6.14427745e-01 6.06209524e-02 5.12404323e-01 -5.78041196e-01
-9.80551600e-01 5.36218703e-01 -6.38228238e-01 -1.98752299e-01
-2.88907707e-01 4.88178223e-01 -3.17559004e-01 4.93846238e-01
4.45938349e-01 2.99113095e-01 -9.91270691e-02 -6.92281067e-01
6.07437253e-01 8.13434541e-01 4.42065328e-01 -2.38243446e-01
5.02379417e-01 -1.25728875e-01 -8.13364744e-01 -1.13132370e+00
-7.11847067e-01 -6.79174483e-01 -9.52616215e-01 1.67340100e-01
1.28833270e+00 -6.72146976e-01 1.11311242e-01 5.12740433e-01
-1.35763931e+00 -8.53224099e-01 -7.05709457e-01 5.47818959e-01
-5.87197721e-01 4.91158187e-01 -8.52097332e-01 -1.00326133e+00
-2.77245283e-01 -1.21317101e+00 1.03865933e+00 1.72403783e-01
-5.16412556e-01 -1.00421882e+00 1.53294191e-01 1.92804039e-01
1.27465740e-01 -4.31355029e-01 8.19943964e-01 -1.23866820e+00
-7.60844871e-02 7.17680156e-02 2.39576325e-01 8.98198247e-01
-3.70763242e-02 5.89738879e-03 -9.48886991e-01 3.15460451e-02
-5.02850711e-02 -2.31996588e-02 1.31932199e+00 6.49547756e-01
7.44304359e-01 -1.04727224e-01 -3.11802000e-01 4.19035792e-01
9.26417112e-01 6.37794137e-01 3.46079558e-01 7.13750049e-02
7.24890232e-01 8.06360543e-01 4.71115947e-01 -6.98185712e-02
2.94790626e-01 3.38924140e-01 2.13315979e-01 -2.49255255e-01
-1.53741777e-01 -1.61649436e-01 6.92789316e-01 1.43691468e+00
2.58608729e-01 -4.37239796e-01 -1.07136500e+00 8.99688005e-01
-1.72317564e+00 -4.76525456e-01 -9.71248895e-02 1.96073067e+00
1.03742242e+00 7.01669037e-01 4.45401281e-01 4.67979401e-01
8.37385416e-01 3.37131858e-01 -4.32575434e-01 -1.04048598e+00
-1.85356557e-01 5.05656183e-01 5.32413125e-01 9.68327105e-01
-1.04467309e+00 1.79012704e+00 6.63088322e+00 1.04654753e+00
-9.71168637e-01 3.48449022e-01 5.37717164e-01 2.19215080e-01
-3.34242135e-01 2.63917804e-01 -1.21905589e+00 3.41930658e-01
1.41377199e+00 5.05998909e-01 3.80017310e-01 8.31361949e-01
2.31071472e-01 -3.55865657e-01 -9.31684613e-01 5.16834199e-01
4.09917124e-02 -9.09780383e-01 -1.20259911e-01 -9.18417722e-02
5.01119971e-01 2.32588381e-01 -2.29286253e-01 4.73518997e-01
2.24759385e-01 -8.17889214e-01 8.28933716e-01 1.64205641e-01
5.24422646e-01 -8.82918119e-01 6.69818521e-01 4.22855824e-01
-1.20636773e+00 1.16939835e-01 -1.11778639e-01 1.63492888e-01
6.82220936e-01 4.90503937e-01 -8.70739818e-01 1.78533405e-01
1.51944637e-01 4.61017370e-01 -3.36651772e-01 6.77960396e-01
-5.91021717e-01 1.35065079e+00 -4.45206374e-01 3.79920378e-03
6.34720445e-01 -2.74476886e-01 7.74977148e-01 1.84037220e+00
-1.33203134e-01 -4.66289036e-02 3.98265183e-01 7.09153116e-01
3.73995483e-01 2.74460375e-01 -5.76406866e-02 -3.15109134e-01
3.37713659e-01 9.55175042e-01 -1.30048275e+00 -5.38673818e-01
-1.38608813e-01 1.36887252e+00 5.78628145e-02 4.35382992e-01
-7.77060390e-01 -5.36809206e-01 2.51095921e-01 7.15736486e-03
5.84203362e-01 -4.09382284e-01 -4.05866623e-01 -6.70925319e-01
-4.88343567e-01 -6.07939363e-01 1.98924690e-01 -3.14964801e-01
-8.37436378e-01 7.30342031e-01 6.27219230e-02 -7.00227857e-01
-7.22776473e-01 -6.13240242e-01 -6.48528814e-01 8.53135049e-01
-1.53647244e+00 -8.02787006e-01 6.27671421e-01 2.49242976e-01
1.03766143e+00 -6.42119721e-02 8.43166828e-01 1.22228444e-01
-7.97883272e-01 4.74914312e-01 -2.58420587e-01 3.20084244e-01
4.38172609e-01 -1.60025311e+00 7.25863934e-01 9.98782098e-01
4.30798084e-01 4.86450821e-01 5.99618554e-01 -9.18597639e-01
-7.25416303e-01 -8.43226492e-01 1.19318914e+00 -1.20827384e-01
6.43805563e-01 -7.14375436e-01 -1.15603399e+00 6.19622588e-01
5.53241670e-01 -3.24253947e-01 8.58933806e-01 -4.77940366e-02
4.23056819e-02 2.54980177e-01 -6.78756475e-01 3.33700538e-01
8.61756265e-01 -4.25742358e-01 -1.08001041e+00 2.17670187e-01
1.38497066e+00 -2.71229655e-01 -4.47479129e-01 1.35420501e-01
4.12105739e-01 -7.63955891e-01 5.97593367e-01 -1.98204666e-01
-1.57492071e-01 -1.94699511e-01 -7.11806193e-02 -1.35702467e+00
-7.63851181e-02 -9.53761101e-01 -5.24662025e-02 1.35199833e+00
1.06956935e+00 -5.93943238e-01 7.23454952e-01 -1.21675255e-02
-5.97983420e-01 -7.52712786e-01 -9.84076917e-01 -9.94877696e-01
1.58794001e-01 -8.00278246e-01 1.65305749e-01 5.31573355e-01
1.11831866e-01 6.16073787e-01 -2.66486350e-02 -1.06405661e-01
2.21235871e-01 -2.08243728e-01 9.96254906e-02 -9.67174947e-01
-5.23640335e-01 -6.40418530e-01 2.70700827e-02 -1.53849888e+00
4.84187990e-01 -9.87736583e-01 7.14969456e-01 -1.53386271e+00
-5.82940757e-01 -2.72069544e-01 -3.21307868e-01 6.58575237e-01
-1.45626524e-02 1.31088376e-01 2.51472861e-01 -8.61238595e-03
-2.10005075e-01 3.54612142e-01 6.76469564e-01 8.29640627e-02
-9.85617399e-01 2.33196348e-01 -3.08810920e-01 8.69204998e-01
9.16838884e-01 -5.35244524e-01 -3.12239051e-01 -2.98271239e-01
-1.14499390e-01 -7.41441920e-02 -3.59993756e-01 -8.89132321e-01
3.10331583e-01 2.65553683e-01 -9.31414068e-02 -1.01306427e+00
2.98002869e-01 -4.70091194e-01 -4.14188921e-01 5.46648264e-01
-2.99844891e-01 -2.18635038e-01 2.26443172e-01 2.43636101e-01
-3.23450089e-01 -7.47773349e-01 9.11367297e-01 -1.76496103e-01
-6.48343980e-01 -8.21487382e-02 -1.02803993e+00 3.40531543e-02
8.20793629e-01 -6.26983523e-01 4.86043960e-01 7.40210637e-02
-1.32434154e+00 2.20158875e-01 -9.00430158e-02 4.18714046e-01
3.92203540e-01 -8.12918305e-01 -5.47592103e-01 4.97416615e-01
-3.11286628e-01 1.62927076e-01 -2.47568578e-01 7.67097652e-01
-2.62994170e-01 5.58782101e-01 3.95772040e-01 -7.31233001e-01
-9.92921531e-01 4.60479617e-01 2.32274547e-01 -4.18823063e-01
-2.48112664e-01 1.31440163e+00 -1.05854936e-01 -6.35389149e-01
5.54179728e-01 -8.16076577e-01 -3.62799764e-01 8.06536674e-01
1.50176985e-02 2.30260342e-01 -2.67909989e-02 -8.28576505e-01
-3.54474515e-01 6.14981890e-01 -2.05520317e-01 -7.63915896e-01
1.01379836e+00 -3.94228697e-01 -7.13221207e-02 1.03818965e+00
9.64511395e-01 1.35801598e-01 -1.07042181e+00 -1.94604218e-01
5.04828155e-01 5.17831743e-01 1.15952305e-01 -7.80150890e-01
-7.28286088e-01 8.88018370e-01 2.89968640e-01 5.87248683e-01
1.08984053e+00 4.33364771e-02 1.11388278e+00 1.98713541e-02
-1.01823799e-01 -1.68657446e+00 -1.81201547e-01 1.09204984e+00
4.75596905e-01 -1.03129172e+00 -4.68474030e-01 -5.66730320e-01
-7.80534565e-01 9.95349050e-01 3.58918160e-01 -2.74593234e-01
7.66902030e-01 3.29086870e-01 -7.08251670e-02 2.98770398e-01
-3.89669687e-01 -7.28826225e-01 3.00732702e-01 3.05531234e-01
4.80449885e-01 2.07528591e-01 -7.61320233e-01 8.37138236e-01
-3.59581590e-01 -5.09904444e-01 2.20208436e-01 8.63776207e-01
-9.59021091e-01 -1.52235007e+00 -4.06105429e-01 2.16958657e-01
-5.35507321e-01 -5.62131226e-01 -4.80547428e-01 5.73261142e-01
4.20821220e-01 1.03090930e+00 2.78026521e-01 -3.27142984e-01
1.76994875e-01 7.02476382e-01 5.98366670e-02 -1.29194427e+00
-9.69927907e-01 9.70289469e-01 1.74411356e-01 -3.80503476e-01
-4.31299955e-01 -7.89431810e-01 -1.84830964e+00 6.62031353e-01
-6.62667155e-01 2.67060041e-01 8.34538221e-01 1.19615030e+00
-1.15569070e-01 7.26588607e-01 5.36392629e-01 -9.56294179e-01
-1.14765920e-01 -1.12961376e+00 -5.40530801e-01 -3.32089186e-01
2.39425451e-01 -1.92983806e-01 -5.38677633e-01 1.11911230e-01] | [14.500436782836914, 6.673360824584961] |
4b2188c1-afcc-4e4a-9e8e-824324064e10 | refinedetlite-a-lightweight-one-stage-object | 1911.08855 | null | https://arxiv.org/abs/1911.08855v2 | https://arxiv.org/pdf/1911.08855v2.pdf | RefineDetLite: A Lightweight One-stage Object Detection Framework for CPU-only Devices | Previous state-of-the-art real-time object detectors have been reported on GPUs which are extremely expensive for processing massive data and in resource-restricted scenarios. Therefore, high efficiency object detectors on CPU-only devices are urgently-needed in industry. The floating-point operations (FLOPs) of networks are not strictly proportional to the running speed on CPU devices, which inspires the design of an exactly "fast" and "accurate" object detector. After investigating the concern gaps between classification networks and detection backbones, and following the design principles of efficient networks, we propose a lightweight residual-like backbone with large receptive fields and wide dimensions for low-level features, which are crucial for detection tasks. Correspondingly, we also design a light-head detection part to match the backbone capability. Furthermore, by analyzing the drawbacks of current one-stage detector training strategies, we also propose three orthogonal training strategies---IOU-guided loss, classes-aware weighting method and balanced multi-task training approach. Without bells and whistles, our proposed RefineDetLite achieves 26.8 mAP on the MSCOCO benchmark at a speed of 130 ms/pic on a single-thread CPU. The detection accuracy can be further increased to 29.6 mAP by integrating all the proposed training strategies, without apparent speed drop. | ['Mengyuan Liu', 'Chen Chen', 'Qi Ju', 'Xiandong Meng', 'Wanpeng Xiao'] | 2019-11-20 | null | null | null | null | ['head-detection'] | ['computer-vision'] | [ 1.19679131e-01 -3.61433983e-01 -1.79163426e-01 -1.92262009e-01
-3.23852569e-01 -2.50705071e-02 3.25281352e-01 3.19104362e-03
-9.27819848e-01 3.36831182e-01 -3.84647012e-01 -3.21005911e-01
1.10313492e-02 -9.12956715e-01 -5.94558716e-01 -8.74585688e-01
1.89560235e-01 2.03610107e-01 1.02561450e+00 -9.53545347e-02
2.60258973e-01 5.74249327e-01 -1.98032033e+00 2.61196822e-01
5.75496376e-01 1.38775933e+00 4.77922827e-01 7.43955016e-01
4.51345332e-02 5.20945907e-01 -3.37638766e-01 -4.44359183e-01
4.42138612e-01 1.69002041e-01 -1.65161192e-01 -3.94404918e-01
6.20933533e-01 -5.07848084e-01 -5.23205996e-01 9.87533271e-01
8.57890606e-01 -2.01738194e-01 5.85018158e-01 -9.66248691e-01
-1.52094647e-01 2.87312418e-01 -9.45468366e-01 6.21088982e-01
-2.35473469e-01 1.15327738e-01 7.56268144e-01 -9.33904648e-01
2.09897920e-01 1.05536926e+00 8.60726893e-01 5.11414886e-01
-7.67078042e-01 -1.06515074e+00 1.16998248e-01 2.40110949e-01
-1.39127183e+00 -3.59701961e-01 3.28334451e-01 -2.61243790e-01
1.11426616e+00 2.62766480e-01 4.29228455e-01 8.11610579e-01
2.11840928e-01 7.02724338e-01 8.02966118e-01 -3.67742807e-01
-5.76524660e-02 4.02951896e-01 2.45057836e-01 8.10443223e-01
8.49094808e-01 2.71135867e-01 -6.14587069e-01 7.99521059e-02
9.74181592e-01 1.74199551e-01 9.23346952e-02 -1.43900931e-01
-1.08362794e+00 7.88333356e-01 5.46586037e-01 2.05853656e-01
-2.67334253e-01 3.13875973e-01 8.27243924e-01 -2.87241396e-02
3.37276548e-01 6.43436331e-04 -3.06049645e-01 6.49913847e-02
-9.18338656e-01 2.12811425e-01 4.89661396e-01 1.06642425e+00
5.36132693e-01 1.57286897e-01 -3.10107589e-01 6.58846796e-01
4.00917113e-01 6.59808397e-01 2.54375309e-01 -5.12729347e-01
6.32462800e-01 5.98219633e-01 3.46619636e-02 -1.02638686e+00
-6.69260085e-01 -1.03434420e+00 -1.09558105e+00 2.79315621e-01
4.14455950e-01 -2.10010763e-02 -6.17753625e-01 1.34418535e+00
5.21398306e-01 4.97262515e-02 -1.67285323e-01 9.75160956e-01
8.81866276e-01 5.00404716e-01 4.48070392e-02 -4.38437499e-02
1.94059002e+00 -1.00695145e+00 -2.76711434e-01 -2.00970545e-01
7.30177224e-01 -7.50933111e-01 8.73040736e-01 4.88175064e-01
-8.70046794e-01 -8.39453638e-01 -1.44341803e+00 1.80728734e-02
-2.62692392e-01 5.43401003e-01 1.01084936e+00 9.84795690e-01
-7.70687640e-01 5.54728270e-01 -8.99811447e-01 -1.88998476e-01
5.80382049e-01 6.34221613e-01 1.89951628e-01 1.87296063e-01
-7.78725624e-01 6.97045147e-01 6.31342232e-01 5.59057593e-02
-6.41433001e-01 -8.35968614e-01 -2.22810984e-01 1.69913575e-01
2.82238424e-01 -6.21820152e-01 1.06023657e+00 -4.32625592e-01
-1.29875922e+00 7.56863058e-01 7.25475773e-02 -5.74908376e-01
6.27999663e-01 -3.17358643e-01 -5.50629020e-01 1.96453044e-03
7.91741908e-02 6.42529607e-01 8.58708203e-01 -7.62686849e-01
-1.31967247e+00 -4.64270681e-01 -1.23911701e-01 2.90978402e-01
-7.94846475e-01 3.34493965e-01 -5.65570235e-01 -3.14230204e-01
1.68488935e-01 -6.62803531e-01 -2.84050196e-01 3.15814048e-01
-2.76651412e-01 -3.32922250e-01 9.48743105e-01 -1.68470085e-01
1.24883533e+00 -2.31730032e+00 -6.20515287e-01 3.04921879e-03
4.06425565e-01 5.83873093e-01 1.20441169e-01 -1.78813756e-01
2.24592894e-01 -4.70443070e-01 4.41993237e-01 -3.51226747e-01
-2.85740290e-02 -1.36709243e-01 -2.03562886e-01 6.61385059e-01
-3.70861627e-02 5.31578779e-01 -7.30094731e-01 -5.27682543e-01
3.07514250e-01 4.64774698e-01 -5.62219620e-01 -1.20568685e-01
2.11325675e-01 9.50843021e-02 -6.27815783e-01 8.02638531e-01
9.96217489e-01 -3.17071497e-01 -2.67631888e-01 -5.68040550e-01
-5.96626639e-01 9.47957933e-02 -1.50433886e+00 1.47991085e+00
-4.37381983e-01 4.72428948e-01 2.78983712e-01 -8.45198870e-01
9.49204922e-01 -1.60068311e-02 2.38736510e-01 -1.07978618e+00
4.22686934e-01 5.58758378e-01 2.61095930e-02 -8.67549852e-02
5.72419822e-01 9.54426378e-02 1.57423824e-01 9.56355259e-02
-9.10063758e-02 5.07945061e-01 3.33548933e-01 1.74706485e-02
1.05690229e+00 -1.33643731e-01 2.76347585e-02 -4.19517249e-01
5.29637456e-01 -9.61302593e-02 5.47459364e-01 1.07630002e+00
-2.80444324e-01 3.69694561e-01 1.84262082e-01 -7.29038000e-01
-1.03798282e+00 -7.55911946e-01 -6.66804254e-01 1.45297277e+00
3.33894849e-01 -3.50086540e-01 -5.13247192e-01 -3.76582980e-01
5.60495108e-02 1.41159624e-01 -1.19020931e-01 1.34596422e-01
-7.22624660e-01 -1.42443442e+00 8.12053323e-01 7.74180174e-01
7.92933166e-01 -8.02817047e-01 -1.26059794e+00 4.43882555e-01
5.03636301e-01 -1.27398241e+00 -2.53472459e-02 4.72458392e-01
-8.76683056e-01 -9.32767093e-01 -6.30876541e-01 -8.56649280e-01
4.76453096e-01 6.56560779e-01 9.60527360e-01 1.54701203e-01
-5.64243436e-01 -2.47394636e-01 -7.00314492e-02 -5.77751040e-01
7.47002959e-02 2.35112488e-01 4.32578743e-01 -1.36785731e-01
3.60465646e-01 -5.48164487e-01 -1.09889877e+00 4.27525252e-01
-5.18751740e-01 2.52314985e-01 9.61676240e-01 7.18348980e-01
5.73272169e-01 -1.07710391e-01 5.00788867e-01 -6.43973649e-01
2.15935543e-01 -7.25070536e-02 -1.04869187e+00 1.61733299e-01
-5.32720268e-01 -6.98200986e-02 7.43201196e-01 -5.72184086e-01
-9.70586479e-01 3.26943219e-01 -4.22429413e-01 -2.05201462e-01
1.05639584e-01 -2.65015125e-01 1.20987758e-01 -5.53034246e-01
9.11266923e-01 1.55300304e-01 -3.40179175e-01 -3.75780165e-01
2.02344730e-01 7.98424065e-01 5.42170763e-01 -4.46496755e-01
5.57200432e-01 7.42370605e-01 2.37536669e-01 -8.18541467e-01
-7.57348657e-01 -8.57720494e-01 -3.48185748e-01 -1.15641579e-01
6.86425924e-01 -1.25885355e+00 -1.21148551e+00 6.22469544e-01
-1.32128632e+00 8.82648304e-02 -2.12922823e-02 5.56068122e-01
-1.02762751e-01 3.08698475e-01 -7.00515687e-01 -8.76515210e-01
-8.08744729e-01 -1.11796749e+00 1.09570801e+00 4.32101160e-01
3.73900115e-01 -4.68667269e-01 -2.85911530e-01 1.25145942e-01
6.36404216e-01 -2.00859353e-01 4.18288231e-01 -4.22788978e-01
-8.02522242e-01 -3.26094389e-01 -1.07408619e+00 2.14972466e-01
-3.74831527e-01 -2.46343270e-01 -1.30417967e+00 -4.41373140e-01
9.22826380e-02 -3.23454916e-01 1.03992712e+00 3.59791249e-01
1.22104919e+00 1.10795006e-01 -7.47245193e-01 9.17154253e-01
1.72807395e+00 1.15766279e-01 4.43365008e-01 3.48640412e-01
7.93939590e-01 1.66134968e-01 5.29228747e-01 7.59302258e-01
8.55610743e-02 7.39079952e-01 5.21996439e-01 -2.17166424e-01
-2.26082519e-01 -3.85386683e-02 6.39392510e-02 8.68114531e-01
-2.10089281e-01 -2.41855189e-01 -7.42577374e-01 2.54391283e-01
-1.78811002e+00 -7.48359859e-01 -5.80054522e-01 2.21148658e+00
3.56077135e-01 6.77183509e-01 1.56615317e-01 1.72854252e-02
7.45755076e-01 9.51945186e-02 -6.43781245e-01 -2.32874960e-01
-5.23890518e-02 1.82691008e-01 1.12556756e+00 -1.22886142e-02
-1.34193623e+00 7.36530364e-01 5.88802576e+00 1.38189721e+00
-1.09483242e+00 5.30658841e-01 4.52179760e-01 -2.55562842e-01
3.64270359e-01 -1.83680579e-01 -1.75345552e+00 3.83458734e-01
8.64213705e-01 2.94518322e-01 -1.63644850e-01 1.25788093e+00
2.83161998e-02 -2.56930798e-01 -9.24939454e-01 1.33487833e+00
-1.33391097e-01 -1.48768866e+00 -1.35563090e-01 1.12003759e-01
4.00934756e-01 3.71738464e-01 -3.71386809e-03 4.00680572e-01
-1.19612016e-01 -5.15708685e-01 7.69043624e-01 -7.22393990e-02
9.63828802e-01 -6.89605355e-01 7.87787020e-01 5.68715930e-01
-1.50746226e+00 -1.92191049e-01 -9.17154789e-01 -2.44783804e-01
7.96768889e-02 8.32389832e-01 -4.85371619e-01 3.33716393e-01
8.62265527e-01 2.58592486e-01 -5.76269865e-01 1.21961391e+00
1.36486977e-01 3.36685628e-01 -7.46678889e-01 -5.86606622e-01
2.57416785e-01 1.75514538e-02 2.57149160e-01 1.62582755e+00
3.38420928e-01 -9.10532549e-02 1.04654223e-01 6.19250596e-01
-6.22558109e-02 6.73413649e-02 -3.20523739e-01 5.42956710e-01
3.97501171e-01 1.61449742e+00 -1.08174002e+00 -3.85270864e-01
-6.88694894e-01 7.97119319e-01 2.48458266e-01 -2.03923926e-01
-1.10117936e+00 -4.84626114e-01 5.82307398e-01 1.94787472e-01
3.00974041e-01 -1.25557721e-01 -5.13148487e-01 -8.57746542e-01
2.95310766e-01 -4.67426538e-01 1.41735718e-01 -1.38186187e-01
-9.96812880e-01 5.54760695e-01 -3.39473486e-01 -1.32403612e+00
3.96533877e-01 -1.05828500e+00 -6.10387683e-01 5.85417151e-01
-1.65270519e+00 -1.01989639e+00 -4.41654503e-01 3.70956957e-01
4.79805082e-01 -1.97297782e-01 5.67854822e-01 9.02810872e-01
-8.07603836e-01 9.17234063e-01 2.95778543e-01 3.61529700e-02
3.67748976e-01 -7.79991984e-01 6.28965735e-01 8.11764181e-01
3.93802039e-02 3.91290307e-01 5.24634302e-01 -3.37190121e-01
-1.56141794e+00 -1.08196211e+00 4.81684953e-01 -4.59990986e-02
5.98138988e-01 -6.74773932e-01 -8.71577561e-01 3.10408091e-03
-2.15194896e-01 4.36415851e-01 2.13293299e-01 1.05831571e-01
-2.80837357e-01 -4.80213523e-01 -9.35401261e-01 3.68171483e-01
1.33345187e+00 -1.82585955e-01 -3.10896169e-02 7.15601742e-01
5.42545080e-01 -5.17245352e-01 -4.33815777e-01 5.08068144e-01
7.63207793e-01 -1.18041790e+00 1.12625134e+00 -6.81633204e-02
-1.20344125e-01 -2.94918656e-01 -2.46424619e-02 -4.38837320e-01
-5.55334210e-01 -2.60833293e-01 -2.04797953e-01 9.53656316e-01
2.04004824e-01 -7.15002298e-01 1.28354394e+00 2.00540815e-02
-3.19445789e-01 -7.62935817e-01 -1.16786528e+00 -1.03052497e+00
-4.34882253e-01 -6.19405508e-01 1.20187260e-01 3.71220410e-01
-3.40984166e-01 3.63271117e-01 -3.43571961e-01 2.63347715e-01
7.94568777e-01 -1.32161185e-01 5.82175195e-01 -1.25971854e+00
-4.41251755e-01 -5.94541550e-01 -6.30385399e-01 -1.39137173e+00
-4.08423066e-01 -6.00448012e-01 -3.85832451e-02 -1.08195603e+00
3.58622134e-01 -8.41372430e-01 -3.52639079e-01 2.11434618e-01
-1.39774168e-02 4.44884121e-01 1.63566500e-01 3.53125602e-01
-9.31317985e-01 2.81103075e-01 9.59419489e-01 1.50641546e-01
-1.45295495e-02 2.09505320e-01 -4.07627225e-01 8.62740040e-01
5.52244604e-01 -6.63874984e-01 -2.03309342e-01 -4.79167074e-01
3.40624809e-01 -1.95175573e-01 3.40820491e-01 -1.69557726e+00
7.25804746e-01 3.43610048e-01 3.89633924e-01 -9.14994717e-01
4.18880790e-01 -6.19644463e-01 -1.90779686e-01 8.56966674e-01
2.34693363e-01 9.30687860e-02 2.95275748e-01 4.85517383e-01
-2.52936594e-02 -3.05552244e-01 9.68837619e-01 2.87985653e-02
-9.47269440e-01 3.36985290e-01 -6.32689968e-02 -1.64541930e-01
1.10280776e+00 -5.09429872e-01 -4.59209591e-01 5.21039128e-01
-6.63880408e-02 1.29701287e-01 3.13127376e-02 2.69549876e-01
4.67646927e-01 -1.13761246e+00 -6.90370798e-01 3.90617222e-01
3.03087719e-02 1.43639743e-01 4.18231159e-01 7.25041628e-01
-8.22956622e-01 6.97136998e-01 -2.94866741e-01 -6.99878275e-01
-1.18383408e+00 4.71355468e-01 1.31427750e-01 -3.12426805e-01
-8.00958633e-01 1.14177585e+00 3.85799170e-01 -5.85616305e-02
5.20961523e-01 -2.89137721e-01 8.31461325e-03 1.28042161e-01
8.54035020e-01 6.26838028e-01 4.10800844e-01 -2.29422882e-01
-5.18443882e-01 6.19949758e-01 -3.37115228e-01 3.01567852e-01
1.02211392e+00 2.81214803e-01 2.57304043e-01 -2.69908663e-02
8.77020240e-01 -2.54933596e-01 -1.29073453e+00 -1.40483662e-01
-5.41754775e-02 -3.71626019e-01 3.25886667e-01 -3.08354437e-01
-1.01372504e+00 1.02458131e+00 1.16491759e+00 4.42460962e-02
1.04269552e+00 -1.00220419e-01 8.14442158e-01 3.96703571e-01
5.45830846e-01 -1.26512003e+00 1.16722696e-01 4.64785248e-01
3.82512391e-01 -1.30582023e+00 2.48425275e-01 -7.22225785e-01
-9.67694595e-02 1.01977754e+00 9.78495479e-01 -1.95721567e-01
4.28490281e-01 6.11404240e-01 -3.82857233e-01 -1.83032945e-01
-5.04023194e-01 -2.52058327e-01 1.92906201e-01 4.72787052e-01
2.49634743e-01 1.58428028e-02 -3.73961717e-01 3.96402866e-01
1.79764554e-01 -1.17315024e-01 8.07126984e-02 7.68006861e-01
-8.17100704e-01 -8.22343647e-01 -4.20725077e-01 7.77193844e-01
-4.61207390e-01 -1.91945300e-01 4.54375982e-01 7.92329013e-01
5.68232536e-01 5.65873742e-01 2.71941960e-01 -4.26520020e-01
4.69163686e-01 -4.51953799e-01 2.93781340e-01 -5.67040682e-01
-7.19794095e-01 -3.97694521e-02 -1.09382108e-01 -5.19893646e-01
-2.14754134e-01 -3.65060061e-01 -1.09065294e+00 -4.74553972e-01
-8.00402343e-01 -4.18143600e-01 9.04037535e-01 7.14498520e-01
3.51323545e-01 5.17824173e-01 3.61920118e-01 -1.03662109e+00
-7.90806532e-01 -8.28071475e-01 -4.81788665e-01 -1.12694196e-01
-5.94856441e-02 -7.95299172e-01 -2.39146441e-01 -6.66716874e-01] | [8.706923484802246, -0.35972973704338074] |
3cf1defe-e845-4aff-9d49-de65cd9a7353 | gm-mlic-graph-matching-based-multi-label | 2104.14762 | null | https://arxiv.org/abs/2104.14762v2 | https://arxiv.org/pdf/2104.14762v2.pdf | GM-MLIC: Graph Matching based Multi-Label Image Classification | Multi-Label Image Classification (MLIC) aims to predict a set of labels that present in an image. The key to deal with such problem is to mine the associations between image contents and labels, and further obtain the correct assignments between images and their labels. In this paper, we treat each image as a bag of instances, and reformulate the task of MLIC as an instance-label matching selection problem. To model such problem, we propose a novel deep learning framework named Graph Matching based Multi-Label Image Classification (GM-MLIC), where Graph Matching (GM) scheme is introduced owing to its excellent capability of excavating the instance and label relationship. Specifically, we first construct an instance spatial graph and a label semantic graph respectively, and then incorporate them into a constructed assignment graph by connecting each instance to all labels. Subsequently, the graph network block is adopted to aggregate and update all nodes and edges state on the assignment graph to form structured representations for each instance and label. Our network finally derives a prediction score for each instance-label correspondence and optimizes such correspondence with a weighted cross-entropy loss. Extensive experiments conducted on various image datasets demonstrate the superiority of our proposed method. | ['Zizhang Wu', 'Gengyu Lyu', 'Yi Jin', 'Songhe Feng', 'He Liu', 'Yanan Wu'] | 2021-04-30 | null | null | null | null | ['multi-label-image-classification'] | ['computer-vision'] | [ 6.60013020e-01 1.26779839e-01 -2.94607580e-01 -6.61201298e-01
-6.38728142e-01 -2.34995633e-01 2.49682799e-01 5.51463664e-01
-8.41716677e-03 3.47274274e-01 -1.98704198e-01 1.17231488e-01
-3.62971455e-01 -9.82230186e-01 -5.45546651e-01 -6.98711276e-01
1.42123654e-01 2.42224634e-01 9.06924680e-02 5.02940953e-01
1.78971723e-01 2.14667037e-01 -1.39655399e+00 2.27244496e-01
6.43234313e-01 1.27677464e+00 3.75862509e-01 3.61479148e-02
-2.57678241e-01 1.11806893e+00 -1.24082230e-01 -4.67159420e-01
6.32646754e-02 -5.07817268e-01 -1.04260194e+00 6.78431034e-01
5.34988105e-01 1.98208503e-02 -2.86859572e-01 1.31680870e+00
3.09859216e-01 1.84247702e-01 8.91146958e-01 -1.56473398e+00
-7.44254410e-01 4.33387011e-01 -9.94091511e-01 -3.33414972e-02
2.35083681e-02 -2.63574511e-01 1.21336246e+00 -7.00993001e-01
4.74930465e-01 1.20787835e+00 5.78977823e-01 3.32426965e-01
-9.85294700e-01 -7.32778788e-01 3.81462634e-01 2.73103267e-01
-1.51908886e+00 -8.28811377e-02 1.02357244e+00 -4.71367806e-01
3.13241810e-01 5.94533011e-02 4.23169136e-01 4.54712749e-01
4.77639511e-02 8.84399652e-01 1.02792847e+00 -3.55233401e-01
-1.25151247e-01 8.85483921e-02 3.10999900e-01 1.18221354e+00
-6.99233487e-02 -3.46281201e-01 -2.42244229e-01 -3.27083394e-02
5.42060673e-01 3.45766664e-01 -2.14060366e-01 -3.94040495e-01
-1.14289463e+00 8.16958606e-01 8.51652920e-01 2.53376871e-01
-4.13738787e-01 3.81069750e-01 5.06304085e-01 1.93159461e-01
5.67935586e-01 -4.73391637e-02 -2.88814336e-01 7.72940457e-01
-6.56611562e-01 -1.03404053e-01 3.14160109e-01 1.11547565e+00
1.28774762e+00 -5.02576530e-01 -2.40982637e-01 1.20374215e+00
7.07246482e-01 5.34610301e-02 5.46392918e-01 -7.84753442e-01
4.51177359e-01 1.12131667e+00 -4.57237631e-01 -1.64195275e+00
-5.10155499e-01 -5.88423371e-01 -9.44684803e-01 -1.42409310e-01
-1.16936803e-01 1.49123982e-01 -6.76306009e-01 1.68896890e+00
5.39588392e-01 6.97984099e-01 -8.09472501e-02 7.49008358e-01
1.22363198e+00 6.86788976e-01 5.90991378e-01 -4.32676882e-01
1.47993243e+00 -1.18221748e+00 -5.78963459e-01 -3.92852336e-01
7.70374775e-01 -4.56676841e-01 6.17188811e-01 -1.03764214e-01
-7.76252210e-01 -7.06111491e-01 -9.41571891e-01 5.34950644e-02
-2.18177199e-01 1.65282875e-01 4.58878666e-01 1.24412045e-01
-9.18563247e-01 5.74957252e-01 -1.71553612e-01 -1.67573825e-01
5.41679859e-01 5.36889732e-01 -5.17706752e-01 -2.42166594e-01
-1.14241648e+00 5.17224848e-01 9.32321608e-01 9.46868658e-02
-6.34476721e-01 -4.47521925e-01 -1.23612070e+00 5.78080479e-04
4.31259036e-01 -5.54413855e-01 7.88027644e-01 -1.27129567e+00
-6.81535959e-01 1.35918176e+00 -7.11543858e-02 -6.99793641e-03
-1.59167461e-02 6.12803340e-01 -6.62191391e-01 2.58718580e-01
4.59988087e-01 8.09533834e-01 7.47441828e-01 -1.70403457e+00
-1.00665581e+00 -5.64102352e-01 1.30504265e-01 4.21153337e-01
-3.12242568e-01 -7.85521194e-02 -5.33910334e-01 -4.16342348e-01
6.15054607e-01 -7.17393041e-01 -9.68254060e-02 -1.49956003e-01
-5.30733526e-01 -4.59841251e-01 7.32529163e-01 -7.73274720e-01
1.27746904e+00 -2.04837513e+00 1.38339728e-01 3.51983249e-01
5.55220246e-01 -1.54305801e-01 -2.90243179e-01 1.76936552e-01
-2.62259632e-01 7.96738192e-02 -3.32355231e-01 -3.84491503e-01
-2.78708696e-01 1.08431861e-01 1.99126527e-01 6.64127231e-01
5.84035888e-02 9.74116445e-01 -9.78799522e-01 -1.07005799e+00
1.66772291e-01 2.49604523e-01 -2.08905131e-01 2.53970176e-01
-2.55713344e-01 4.56911594e-01 -7.33684123e-01 8.34334254e-01
7.94277191e-01 -9.03392613e-01 2.93441325e-01 -6.09983921e-01
3.17983598e-01 -3.39910656e-01 -1.25091600e+00 1.58676088e+00
-5.96426010e-01 1.41455561e-01 -2.02822298e-01 -1.30509675e+00
1.11009681e+00 1.30699769e-01 7.89203465e-01 -6.59609735e-01
2.02493444e-01 1.01994544e-01 -5.79203010e-01 -7.01767147e-01
1.77793086e-01 -1.03195757e-01 -9.11590308e-02 4.68940198e-01
1.94270477e-01 2.71179140e-01 -2.97083519e-02 2.35722974e-01
6.78957045e-01 1.74307395e-02 4.24889326e-01 -5.46993613e-02
1.01427245e+00 -2.84847528e-01 7.60844588e-01 2.24101916e-01
-2.26807162e-01 3.64521801e-01 2.36333534e-01 -5.26057482e-01
-6.30212307e-01 -6.62866533e-01 2.78508961e-02 1.13319182e+00
7.97288060e-01 -2.45373130e-01 -5.95114946e-01 -1.13994992e+00
9.85107720e-02 3.15437943e-01 -6.10970736e-01 -1.55502096e-01
-4.98872817e-01 -7.27078974e-01 6.79647997e-02 1.51210248e-01
8.83541763e-01 -1.24457550e+00 4.40475047e-02 2.32234776e-01
-3.45865071e-01 -8.55288327e-01 -6.36088908e-01 -4.77176607e-02
-5.64385891e-01 -1.27579403e+00 -2.97194570e-01 -1.57899106e+00
9.18944478e-01 3.93909007e-01 1.00182652e+00 5.60854316e-01
-2.46736050e-01 2.93883830e-01 -3.43395948e-01 2.78759837e-01
-3.15340281e-01 -2.48093233e-02 -4.19563651e-01 7.77427554e-01
1.54902190e-01 -4.69918996e-01 -7.70949602e-01 3.64580959e-01
-9.80832696e-01 2.57581383e-01 5.47328055e-01 8.02728832e-01
1.08460486e+00 2.99277931e-01 7.99716830e-01 -1.26251984e+00
4.32669193e-01 -7.95393586e-01 -4.33082968e-01 7.60317206e-01
-8.02516460e-01 -1.00904964e-01 4.24342036e-01 -6.44646734e-02
-7.95135021e-01 4.56068009e-01 1.55128896e-01 -4.24201250e-01
-3.23232636e-02 8.01250279e-01 -3.91638726e-01 -3.22133034e-01
-1.74140092e-02 3.30473095e-01 -7.72033483e-02 -2.50345021e-01
5.82209527e-01 7.48258591e-01 4.15037215e-01 -2.97562331e-01
3.52586091e-01 2.29830906e-01 2.80435026e-01 -1.49722815e-01
-1.29983664e+00 -6.64310932e-01 -7.42510438e-01 -6.02601469e-01
1.24296594e+00 -8.96973968e-01 -9.65057969e-01 4.14387822e-01
-1.01230299e+00 1.06474400e-01 1.72091216e-01 2.37849712e-01
-6.14843607e-01 5.15794814e-01 -5.71733594e-01 -5.37212729e-01
-2.79859632e-01 -1.27249205e+00 1.26331663e+00 5.93534470e-01
3.14873397e-01 -1.45931172e+00 -1.21736512e-01 5.53633571e-01
-1.50126040e-01 5.03687143e-01 1.24270272e+00 -5.97678423e-01
-5.76814055e-01 -1.94151893e-01 -8.74024212e-01 2.52660036e-01
2.36616939e-01 -3.32542062e-01 -8.71555686e-01 -5.10877192e-01
-1.33469040e-02 -3.67480904e-01 8.60092759e-01 2.92333782e-01
1.54180443e+00 -3.65961611e-01 -7.33338416e-01 5.81351697e-01
1.98737848e+00 1.90315053e-01 1.57994092e-01 2.93155462e-01
1.14719260e+00 8.08435857e-01 7.46758819e-01 1.96780249e-01
7.63965309e-01 6.45569026e-01 7.91511893e-01 -2.00587526e-01
-2.03509718e-01 -3.19462031e-01 -2.61685848e-01 9.72726285e-01
4.65558201e-01 -4.39870745e-01 -7.04189658e-01 3.61787647e-01
-2.11099982e+00 -7.06630766e-01 -2.37137601e-01 1.89774811e+00
6.21333122e-01 -3.24766308e-01 -8.69097263e-02 -1.34795710e-01
1.45666051e+00 3.73922020e-01 -7.08411634e-01 1.22430570e-01
3.47064249e-03 -2.14855254e-01 5.12424052e-01 3.29725504e-01
-1.47332895e+00 8.42922568e-01 5.19593811e+00 1.09557331e+00
-7.91133642e-01 2.77503222e-01 1.16179550e+00 4.55976695e-01
-2.39753380e-01 7.30524436e-02 -6.93386078e-01 5.40759206e-01
4.77900565e-01 1.94159150e-02 3.85640770e-01 6.92709267e-01
-1.13661394e-01 1.83322251e-01 -8.81428659e-01 1.26458085e+00
2.95115381e-01 -1.20809853e+00 3.12415004e-01 -2.02843938e-02
6.63707852e-01 -4.13934350e-01 -1.24036022e-01 1.29587576e-01
2.74772346e-01 -9.68105912e-01 5.35554469e-01 6.31225467e-01
9.13767219e-01 -8.50941002e-01 5.75522959e-01 2.43442014e-01
-1.78310025e+00 -2.37584457e-01 -4.91710544e-01 3.65289271e-01
-4.58691381e-02 4.40632612e-01 -4.50913548e-01 9.04844403e-01
1.92514241e-01 1.26883078e+00 -7.51469254e-01 9.75072563e-01
-1.07288301e-01 1.99333698e-01 3.29373002e-01 1.53257757e-01
3.69419783e-01 -4.74779963e-01 -3.90884243e-02 8.92912745e-01
1.32266924e-01 6.98237792e-02 7.97120214e-01 7.21731603e-01
-3.76668185e-01 5.38404047e-01 -4.48545396e-01 2.30363905e-01
6.30578041e-01 1.61677396e+00 -9.28355753e-01 -2.96030760e-01
-6.49718821e-01 1.18036544e+00 7.98356056e-01 1.91593751e-01
-8.92084301e-01 -3.14575702e-01 1.15473099e-01 -2.24467352e-01
-4.13645208e-02 3.79293293e-01 -1.14266679e-01 -8.85199189e-01
9.62222368e-03 -4.69642490e-01 8.91662240e-01 -8.77447724e-01
-1.44663703e+00 5.39929807e-01 -3.27444673e-01 -1.35255015e+00
2.52903283e-01 -1.24568053e-01 -4.99130785e-01 8.67619932e-01
-1.68442416e+00 -1.77139568e+00 -6.38515472e-01 6.55312181e-01
4.89107788e-01 -8.91610384e-02 7.44461000e-01 5.27991354e-01
-6.75328434e-01 5.73406756e-01 -9.49312560e-03 1.62757277e-01
3.95040363e-01 -1.05253184e+00 -1.26642272e-01 4.94963557e-01
1.34236932e-01 8.96402448e-02 8.84572715e-02 -7.05169439e-01
-8.58510911e-01 -1.74261093e+00 9.12422299e-01 7.80980587e-02
4.45750237e-01 3.27455252e-02 -7.93782473e-01 7.35059142e-01
4.01754342e-02 3.69305104e-01 6.56499684e-01 -2.74841487e-01
-3.12123537e-01 -2.29800820e-01 -1.24578071e+00 2.41928354e-01
1.16363275e+00 -5.73365629e-01 -2.39611492e-01 8.55472028e-01
1.01262391e+00 -2.20222890e-01 -1.15865242e+00 4.97862399e-01
1.59114659e-01 -7.26419926e-01 9.60180044e-01 -3.36627632e-01
5.39500058e-01 -4.86326009e-01 -1.21611424e-01 -1.34539533e+00
-6.54955208e-01 1.83525339e-01 3.58459294e-01 1.54055524e+00
4.11321431e-01 -5.80344379e-01 6.65898860e-01 4.08730984e-01
-8.06595758e-02 -9.64016080e-01 -6.51788235e-01 -3.55764806e-01
-3.24353576e-01 -3.91896907e-03 8.26614320e-01 1.38558161e+00
-2.37962499e-01 3.92517328e-01 -4.14952040e-01 3.92809659e-01
8.45975876e-01 4.81735349e-01 1.45361274e-01 -1.32973015e+00
-1.54072732e-01 -3.48148376e-01 -7.10648537e-01 -7.46182323e-01
6.84018850e-01 -1.59265995e+00 -6.06248453e-02 -1.81407118e+00
6.12485409e-01 -1.00225127e+00 -6.71482086e-01 6.24241710e-01
-4.68667418e-01 4.11213040e-01 5.74424230e-02 4.96737659e-01
-1.09080029e+00 4.63588059e-01 1.26364267e+00 -5.62961936e-01
2.04153121e-01 -1.22421056e-01 -7.46117532e-01 4.92512673e-01
6.12439036e-01 -5.82097828e-01 -6.51581943e-01 -4.04208302e-01
1.07885815e-01 4.06019926e-01 3.62064213e-01 -9.12041247e-01
3.06724131e-01 -1.74556300e-01 2.56072879e-01 -4.07317191e-01
4.89791147e-02 -9.67720866e-01 3.45497370e-01 3.33184987e-01
-6.99836552e-01 -1.31515861e-01 -2.47886732e-01 8.28911364e-01
-3.95169705e-01 -5.14571309e-01 9.10669744e-01 -2.47799233e-01
-1.17446744e+00 8.74593139e-01 3.35006773e-01 4.99121621e-02
1.49904549e+00 -1.42333418e-01 -2.94998229e-01 5.37954457e-02
-9.28741932e-01 5.43722689e-01 3.52828026e-01 3.03332239e-01
7.72973001e-01 -1.80923378e+00 -5.45346677e-01 1.83698595e-01
5.53120971e-01 7.42489472e-03 5.77147305e-01 4.64875579e-01
-4.35780495e-01 -3.54463495e-02 -5.40682226e-02 -5.52850127e-01
-1.43769288e+00 7.06035674e-01 5.26504278e-01 -3.74657840e-01
-5.59818208e-01 9.93060470e-01 4.16616350e-01 -5.58928013e-01
9.13542882e-02 2.96752542e-01 -7.14945078e-01 3.11265439e-01
3.29273969e-01 1.13140583e-01 -2.32395716e-02 -1.18542099e+00
-3.70270163e-01 9.57361102e-01 -1.20388240e-01 4.71964538e-01
1.13333952e+00 -4.60360587e-01 -6.09685898e-01 3.30935508e-01
1.82263565e+00 -6.56876266e-01 -9.70907390e-01 -5.88080406e-01
1.44133782e-02 -4.46088940e-01 1.42937526e-01 -4.77069527e-01
-1.68370593e+00 5.06699681e-01 6.63107336e-01 1.11610852e-01
1.25048757e+00 1.36229545e-01 8.45633030e-01 -1.14029840e-01
2.80418426e-01 -7.75844574e-01 2.34031707e-01 -4.49267663e-02
4.62814808e-01 -1.38421071e+00 -9.20699015e-02 -6.38233781e-01
-5.99398017e-01 9.93625939e-01 7.74296701e-01 -8.13246891e-02
9.04321551e-01 -3.88923168e-01 -9.83230174e-02 -6.59526408e-01
-3.12041998e-01 -2.62094855e-01 3.07966471e-01 4.03286189e-01
2.64667869e-01 2.21600786e-01 -3.24859172e-01 3.36486012e-01
4.78676349e-01 -2.08261013e-01 7.88389072e-02 5.97528517e-01
-6.53642178e-01 -1.00440145e+00 2.81233378e-02 7.59226441e-01
-1.60543412e-01 -5.52847683e-02 -4.66564968e-02 3.27579021e-01
4.79028374e-01 1.08368599e+00 8.58053565e-02 -7.31960714e-01
9.48814303e-02 -6.71932772e-02 1.56976730e-01 -6.87520325e-01
-5.70939362e-01 -2.25530252e-01 -2.02884138e-01 -3.63849461e-01
-7.23250568e-01 -4.57110763e-01 -1.41465127e+00 -3.08170659e-03
-5.47510207e-01 5.21914810e-02 5.63000679e-01 9.25340056e-01
2.56508976e-01 7.66294003e-01 1.04831576e+00 -4.72112328e-01
-1.64480314e-01 -6.30795419e-01 -1.00754976e+00 9.10891831e-01
2.14299224e-02 -7.22358286e-01 -2.59301275e-01 1.02206565e-01] | [9.791433334350586, 4.047863960266113] |
e6427afa-89a8-4dde-b784-15a1fee9c1e4 | graph-neural-networks-for-contextual-asr-with | 2305.18824 | null | https://arxiv.org/abs/2305.18824v1 | https://arxiv.org/pdf/2305.18824v1.pdf | Graph Neural Networks for Contextual ASR with the Tree-Constrained Pointer Generator | The incorporation of biasing words obtained through contextual knowledge is of paramount importance in automatic speech recognition (ASR) applications. This paper proposes an innovative method for achieving end-to-end contextual ASR using graph neural network (GNN) encodings based on the tree-constrained pointer generator method. GNN node encodings facilitate lookahead for future word pieces in the process of ASR decoding at each tree node by incorporating information about all word pieces on the tree branches rooted from it. This results in a more precise prediction of the generation probability of the biasing words. The study explores three GNN encoding techniques, namely tree recursive neural networks, graph convolutional network (GCN), and GraphSAGE, along with different combinations of the complementary GCN and GraphSAGE structures. The performance of the systems was evaluated using the Librispeech and AMI corpus, following the visual-grounded contextual ASR pipeline. The findings indicate that using GNN encodings achieved consistent and significant reductions in word error rate (WER), particularly for words that are rare or have not been seen during the training process. Notably, the most effective combination of GNN encodings obtained more than 60% WER reduction for rare and unseen words compared to standard end-to-end systems. | ['Phil Woodland', 'Chao Zhang', 'Guangzhi Sun'] | 2023-05-30 | null | null | null | null | ['automatic-speech-recognition'] | ['speech'] | [ 6.64726853e-01 4.26961571e-01 6.28448948e-02 -1.37218639e-01
-7.04450548e-01 -4.99550313e-01 7.44253337e-01 2.84371108e-01
-4.48547989e-01 4.75347966e-01 6.01693928e-01 -1.10964656e+00
1.21161424e-01 -7.46131897e-01 -4.64379996e-01 -4.72930402e-01
-1.15941875e-01 3.48244220e-01 -3.78896073e-02 -5.88567793e-01
3.69718730e-01 6.55886471e-01 -1.37548780e+00 5.29684365e-01
4.21294659e-01 5.78103960e-01 5.64863920e-01 1.23117292e+00
-4.60492283e-01 7.08890557e-01 -1.10488510e+00 -3.63675296e-01
-4.58346792e-02 -3.18770558e-01 -6.33088827e-01 -2.19111785e-01
4.94239658e-01 -1.20015517e-02 -7.16361463e-01 7.24897206e-01
9.18616891e-01 5.35865545e-01 3.32320750e-01 -6.23580277e-01
-8.46673965e-01 1.22442353e+00 -4.56043407e-02 6.33060336e-01
4.90651876e-01 2.91821420e-01 1.31164670e+00 -9.24806774e-01
5.21106601e-01 1.54166889e+00 4.98486817e-01 8.51657629e-01
-1.09832859e+00 -4.61555958e-01 4.28992778e-01 3.05826992e-01
-1.45252442e+00 -8.44873011e-01 5.92755258e-01 -5.53224282e-03
2.00186706e+00 3.95625859e-01 5.81004202e-01 1.25222933e+00
6.15242682e-02 6.75348282e-01 5.97936749e-01 -9.97247934e-01
3.01659405e-01 -4.73210901e-01 1.07149392e-01 5.01285195e-01
1.21364787e-01 4.09441829e-01 -9.02371883e-01 1.62604958e-01
3.58952790e-01 -5.91459274e-01 -3.04008275e-01 3.57651651e-01
-9.68300641e-01 7.66539216e-01 4.39245403e-01 2.77816117e-01
-5.05910337e-01 3.76365900e-01 4.21555907e-01 3.46419103e-02
5.37638843e-01 2.23917678e-01 -3.44676465e-01 -3.40414494e-01
-9.69629765e-01 -1.89045386e-03 5.46400905e-01 1.25712168e+00
3.49405140e-01 7.58939147e-01 -7.58324802e-01 8.84807289e-01
5.47477126e-01 5.63833654e-01 7.14420021e-01 -2.00494424e-01
8.36360753e-01 3.67744416e-01 -4.77285326e-01 -6.15164578e-01
-1.28359839e-01 -6.89192832e-01 -5.11560440e-01 -4.47476888e-03
1.58107147e-01 -4.38407138e-02 -1.83898950e+00 1.67182374e+00
-1.05876504e-02 9.56460740e-03 2.52432823e-01 4.96274054e-01
1.13488519e+00 1.09075356e+00 4.31203693e-01 4.74523790e-02
1.36524796e+00 -8.92905772e-01 -9.46271360e-01 -7.66685426e-01
9.26825285e-01 -5.83289146e-01 1.14000535e+00 1.22547582e-01
-1.02020073e+00 -5.06341815e-01 -1.15070653e+00 -2.10014790e-01
-6.53708339e-01 -1.38192952e-01 3.11460912e-01 8.99717510e-01
-1.66153347e+00 3.00643265e-01 -4.83352482e-01 -2.78535783e-01
3.05848420e-01 2.04482853e-01 -9.69875082e-02 -1.89829975e-01
-1.38441479e+00 9.97499228e-01 6.71254396e-01 3.85814279e-01
-8.86082172e-01 -5.13210893e-01 -1.06276262e+00 2.15561122e-01
1.51758522e-01 -3.05955708e-01 1.42232192e+00 -8.03227127e-01
-1.40006852e+00 6.19035542e-01 -4.67046767e-01 -7.36024499e-01
4.48621511e-02 -8.35590437e-02 -6.09731436e-01 1.02952856e-03
-3.89240831e-01 8.33329201e-01 8.43479574e-01 -9.73366201e-01
-4.07023162e-01 -2.29906663e-01 -2.16286823e-01 4.74343330e-01
3.44288200e-02 1.42399194e-02 -2.84975260e-01 -8.67298067e-01
6.43860176e-02 -6.74589157e-01 -1.64982155e-01 -6.56242609e-01
-4.33781058e-01 -4.19512212e-01 7.17158079e-01 -1.17529941e+00
1.79761124e+00 -2.09743547e+00 -1.82337984e-02 2.92583942e-01
9.31789204e-02 7.71124661e-01 -3.09056580e-01 6.14159226e-01
-2.40692452e-01 2.95383334e-01 -8.19775835e-02 -1.64115965e-01
-8.66201222e-02 4.08250570e-01 -4.30412203e-01 -4.54312637e-02
2.47971579e-01 1.19689012e+00 -1.05665207e+00 -2.05080763e-01
4.09637153e-01 6.87960684e-01 -2.82328546e-01 1.71351328e-01
-4.76704866e-01 -1.33631125e-01 1.76642329e-01 4.25449699e-01
1.63408518e-01 1.56261683e-01 2.30463594e-01 2.04719722e-01
1.88477576e-01 1.17296910e+00 -7.92970479e-01 1.45299530e+00
-7.05218971e-01 9.72892225e-01 -2.24818572e-01 -5.32295644e-01
1.18045199e+00 5.66821814e-01 -5.22917688e-01 -1.09501576e+00
-1.10643648e-01 8.39177668e-02 2.31609598e-01 7.87412189e-03
1.03417051e+00 -1.44979402e-01 2.08652750e-01 3.74626629e-02
1.34437889e-01 -8.83472934e-02 1.81246281e-01 4.26896989e-01
1.31162417e+00 -1.45959795e-01 4.41031963e-01 4.70778309e-02
5.48572958e-01 -2.54260957e-01 -1.12762954e-02 8.57260227e-01
-6.94735050e-02 5.95621169e-01 2.96368212e-01 -2.29534626e-01
-1.10170257e+00 -1.01334643e+00 3.18947405e-01 1.31090868e+00
-4.24280524e-01 -7.11789012e-01 -7.99136162e-01 -5.82383692e-01
-3.65592718e-01 1.71156049e+00 -3.81797761e-01 -2.85167009e-01
-9.37791407e-01 -1.36272490e-01 8.41672182e-01 7.11296618e-01
-6.24729227e-03 -1.64954007e+00 -3.88877839e-01 3.94527614e-01
-1.61124736e-01 -1.10138273e+00 -4.33361113e-01 5.08418918e-01
-8.49653304e-01 -4.24155682e-01 -5.78525841e-01 -6.88858688e-01
5.17665148e-01 2.27000788e-01 1.25970578e+00 2.98923492e-01
-1.57671705e-01 3.33000541e-01 -6.52657449e-01 -3.71818662e-01
-8.59123170e-01 2.36742660e-01 -4.90679383e-01 -3.74892116e-01
5.39312184e-01 -2.04056010e-01 -4.08566654e-01 -1.68090872e-02
-6.93251789e-01 1.26116663e-01 4.55966711e-01 7.04132020e-01
5.18558741e-01 -3.65859985e-01 5.77255428e-01 -6.47389531e-01
9.16502357e-01 -2.51228303e-01 -3.72059911e-01 2.11175203e-01
-7.10350037e-01 2.77955294e-01 5.46200514e-01 -3.49260509e-01
-1.06915390e+00 -2.68730909e-01 -6.44067645e-01 -2.14262694e-01
-1.65778339e-01 5.52759469e-01 -1.94791377e-01 5.03896713e-01
8.82078111e-01 3.70752484e-01 -1.77182660e-01 -1.41743869e-01
8.84630620e-01 9.67124403e-01 3.56937319e-01 -3.34529132e-01
3.52568150e-01 -1.80406675e-01 -2.67157316e-01 -1.21865547e+00
-4.02102053e-01 -3.75772506e-01 -2.76971817e-01 -4.10389930e-01
7.83507884e-01 -8.03565264e-01 -6.45823628e-02 2.98082650e-01
-1.40883410e+00 -6.81959391e-01 -5.94103932e-01 1.88058436e-01
-2.76990712e-01 2.58262157e-01 -3.50678802e-01 -1.21258879e+00
-7.06006408e-01 -1.04796815e+00 1.05237412e+00 8.39913916e-03
-5.13746798e-01 -1.05516005e+00 -2.48950973e-01 1.68588221e-01
5.99916875e-01 -3.33730429e-01 1.18962908e+00 -1.07049930e+00
-4.82904077e-01 -1.58645496e-01 -6.37381822e-02 3.94138694e-01
-1.32199988e-01 -1.40060693e-01 -1.05645263e+00 -3.57847720e-01
-5.29682040e-01 2.14793995e-01 7.43705750e-01 3.60789895e-01
8.90554607e-01 -4.82169807e-01 -2.39845440e-01 3.48084569e-01
1.11461818e+00 6.79637074e-01 1.04019356e+00 2.32358262e-01
9.35449481e-01 4.81981456e-01 2.40013599e-01 4.65481021e-02
1.57591611e-01 5.98718643e-01 6.09323919e-01 2.82049745e-01
-9.22332048e-01 -6.39299452e-01 6.76806927e-01 1.03556085e+00
1.82014480e-01 -1.10615468e+00 -1.09720731e+00 8.76292527e-01
-1.31717527e+00 -7.25536287e-01 -2.19350114e-01 2.19030213e+00
5.90144038e-01 3.36450636e-01 -3.25919718e-01 3.64505678e-01
1.08558428e+00 5.85883021e-01 -2.26167023e-01 -1.16685867e+00
-2.28493005e-01 7.21882105e-01 5.49417317e-01 8.07775557e-01
-3.69674325e-01 1.49073410e+00 6.52556276e+00 1.11943471e+00
-9.45670307e-01 2.26832088e-03 5.89483142e-01 -8.92145485e-02
-6.54274821e-01 -2.48018224e-02 -1.11070335e+00 1.95051670e-01
1.72607374e+00 5.91592118e-03 7.26171613e-01 7.59833455e-01
1.78405732e-01 2.29066581e-01 -9.02696013e-01 9.00779903e-01
1.33109137e-01 -1.28873289e+00 4.94726658e-01 -1.11028560e-01
2.36207232e-01 1.84945494e-01 7.59354830e-02 4.12113190e-01
4.36773866e-01 -1.27397299e+00 1.04883885e+00 1.08858995e-01
1.08731270e+00 -8.42237294e-01 4.34711486e-01 -1.09226756e-01
-1.23942852e+00 -2.02228688e-02 -2.47063726e-01 -4.49924394e-02
3.37186217e-01 3.70833009e-01 -1.47617853e+00 3.33109170e-01
1.96108967e-01 2.52429128e-01 -5.99529088e-01 7.29014337e-01
-8.18358421e-01 1.20637417e+00 -1.26635581e-01 -4.63395774e-01
4.51007485e-01 2.55896956e-01 8.74822557e-01 1.82291925e+00
4.67980802e-01 2.98143793e-02 -5.63450277e-01 5.13805091e-01
-1.26676708e-01 2.36247212e-01 -7.53557742e-01 -3.30810457e-01
8.14479947e-01 8.76702785e-01 -7.14995503e-01 -3.32425267e-01
-1.46935761e-01 8.87505770e-01 4.76975977e-01 4.73629236e-01
-5.43897271e-01 -4.96426702e-01 5.20280302e-01 1.77649468e-01
6.21863246e-01 -4.56447333e-01 -4.17920649e-01 -5.04908562e-01
-1.00089654e-01 -1.08407462e+00 3.76668662e-01 -1.18061590e+00
-7.94165969e-01 8.97059679e-01 -7.19350204e-02 -6.33136392e-01
-5.42921364e-01 -7.62457967e-01 -5.17326772e-01 1.27834499e+00
-1.34456360e+00 -1.01599097e+00 6.41964823e-02 3.63191009e-01
8.90156984e-01 -3.58153880e-01 8.31542075e-01 4.25637588e-02
-3.18872511e-01 7.95686066e-01 -3.98219436e-01 1.22372404e-01
2.10466441e-02 -1.26719987e+00 1.21439636e+00 1.27779806e+00
5.46255767e-01 6.35848939e-01 5.84671736e-01 -9.15241718e-01
-1.21377289e+00 -1.19311595e+00 1.34368348e+00 -2.86557853e-01
4.86607760e-01 -5.90841234e-01 -9.57329035e-01 7.01066971e-01
4.35203463e-01 -2.08055869e-01 3.97002727e-01 8.83586034e-02
-5.02356291e-01 1.66152894e-01 -7.66845644e-01 9.18144584e-01
1.38360763e+00 -8.08639586e-01 -8.05680037e-01 1.17591016e-01
1.23690951e+00 -5.95455110e-01 -1.86262026e-01 1.56706870e-01
2.63786465e-01 -5.98886430e-01 8.58933985e-01 -6.67195559e-01
2.71671772e-01 -1.34648487e-01 -4.73738641e-01 -1.55459654e+00
-2.68982202e-01 -8.08508456e-01 -5.26605725e-01 1.26090086e+00
8.26303482e-01 -5.29707193e-01 6.12170875e-01 2.02807024e-01
-7.13359892e-01 -5.76257586e-01 -1.22651637e+00 -6.91536069e-01
-1.06619328e-01 -1.03478956e+00 6.72694147e-01 3.25491965e-01
-3.39757651e-01 6.96730614e-01 5.27262222e-04 2.07313985e-01
1.06937647e-01 -7.88123190e-01 3.73008579e-01 -6.03715360e-01
-9.93364211e-03 -5.17311931e-01 -6.76638246e-01 -1.20679355e+00
1.37277365e-01 -1.25276971e+00 3.63038421e-01 -1.94545507e+00
-4.94760364e-01 -2.72182703e-01 -3.69610608e-01 5.18087924e-01
-2.78217793e-01 -2.56952018e-01 2.40604237e-01 -3.59930843e-01
-1.14093408e-01 3.83130580e-01 9.69058633e-01 -1.21552743e-01
-2.55085856e-01 -2.20256373e-01 -5.91988385e-01 2.56089211e-01
9.07037437e-01 -3.51416409e-01 -6.40028536e-01 -6.57547295e-01
2.76144415e-01 6.50736392e-02 2.71412224e-01 -9.13255990e-01
2.87727803e-01 1.94692880e-01 1.69131786e-01 -9.99578118e-01
2.16845050e-01 -5.49668372e-01 -3.35271917e-02 2.11676970e-01
-6.34793401e-01 4.60938245e-01 6.16842926e-01 4.84702647e-01
-3.63764167e-02 -3.51623982e-01 5.68416715e-01 -4.26225662e-02
-8.46030653e-01 -1.47294059e-01 -7.81931221e-01 2.49389380e-01
5.76663673e-01 -5.13612926e-01 -3.52341950e-01 -6.57057464e-01
-6.93118215e-01 -2.87975430e-01 -3.11393291e-01 6.15297854e-01
1.06169558e+00 -1.09196734e+00 -6.76824212e-01 3.73245865e-01
2.03418747e-01 6.81794137e-02 9.64829996e-02 1.39836326e-01
-6.49672210e-01 6.56219602e-01 2.97173828e-01 -4.27706540e-01
-1.41247404e+00 4.90058213e-01 2.84433097e-01 -3.77348214e-01
-6.94932401e-01 1.14808655e+00 -7.66160935e-02 -2.96157002e-01
4.91317958e-01 -4.31104779e-01 -1.54887244e-01 5.66731133e-02
5.69341660e-01 4.94016826e-01 5.86735606e-01 -8.90244305e-01
-3.02103996e-01 -5.02396375e-02 -3.17053169e-01 -5.88194609e-01
1.03552389e+00 -1.87940404e-01 2.26280287e-01 3.52333486e-01
8.19479167e-01 7.49645708e-03 -8.63359213e-01 -1.67263463e-01
1.93810508e-01 -1.55024067e-01 4.29157197e-01 -1.24582481e+00
-8.28320205e-01 1.19285941e+00 5.87093771e-01 2.19693944e-01
9.07730997e-01 -7.85241947e-02 7.72578716e-01 2.22594455e-01
1.04143262e-01 -1.09792984e+00 -2.36066863e-01 9.49268281e-01
9.93371129e-01 -5.43442428e-01 -3.13459933e-01 -1.62668809e-01
-5.45058131e-01 1.19179356e+00 3.81710231e-01 2.76417047e-01
3.04194510e-01 1.03881612e-01 1.92694291e-01 -1.44695953e-01
-9.25919175e-01 -3.33551288e-01 3.69824797e-01 8.45212460e-01
5.73628068e-01 3.01010013e-01 -2.72037148e-01 1.49206311e-01
-4.50475335e-01 -6.09816134e-01 3.38766068e-01 8.39475989e-01
-4.78648245e-01 -9.26336706e-01 -2.51656502e-01 3.21594298e-01
-3.68134826e-01 -9.52043355e-01 -3.82989407e-01 6.42379820e-01
-2.77362913e-01 1.19174302e+00 1.34120017e-01 -3.76229852e-01
5.17371416e-01 5.47513664e-01 2.92086899e-01 -9.29823577e-01
-8.30887079e-01 -1.26702875e-01 5.80670178e-01 -4.03436631e-01
2.26971537e-01 -4.19595748e-01 -1.37730694e+00 -8.21985975e-02
-4.91798848e-01 -1.05686456e-01 1.09646380e+00 8.17760706e-01
4.29343313e-01 1.07175708e+00 1.78413585e-01 -6.90941155e-01
-3.17240387e-01 -1.17742169e+00 -1.30973503e-01 1.54517710e-01
2.58256108e-01 -1.89622134e-01 -3.59697700e-01 -1.62156582e-01] | [14.312844276428223, 6.8173418045043945] |
6023095f-a191-4511-b0af-d0e5073bc6c8 | an-event-driven-compressive-neuromorphic | 2205.13292 | null | https://arxiv.org/abs/2205.13292v1 | https://arxiv.org/pdf/2205.13292v1.pdf | An Event-Driven Compressive Neuromorphic System for Cardiac Arrhythmia Detection | Wearable electrocardiograph (ECG) recording and processing systems have been developed to detect cardiac arrhythmia to help prevent heart attacks. Conventional wearable systems, however, suffer from high energy consumption at both circuit and system levels. To overcome the design challenges, this paper proposes an event-driven compressive ECG recording and neuromorphic processing system for cardiac arrhythmia detection. The proposed system achieves low power consumption and high arrhythmia detection accuracy via system level co-design with spike-based information representation. Event-driven level-crossing ADC (LC-ADC) is exploited in the recording system, which utilizes the sparsity of ECG signal to enable compressive recording and save ADC energy during the silent signal period. Meanwhile, the proposed spiking convolutional neural network (SCNN) based neuromorphic arrhythmia detection method is inherently compatible with the spike-based output of LC-ADC, hence realizing accurate detection and low energy consumption at system level. Simulation results show that the proposed system with 5-bit LC-ADC achieves 88.6\% reduction of sampled data points compared with Nyquist sampling in the MIT-BIH dataset, and 93.59\% arrhythmia detection accuracy with SCNN, demonstrating the compression ability of LC-ADC and the effectiveness of system level co-design with SCNN. | ['Mohamad Sawan', 'Jie Yang', 'Fengshi Tian', 'Jinbo Chen'] | 2022-05-26 | null | null | null | null | ['arrhythmia-detection'] | ['medical'] | [ 7.77673900e-01 -6.01020217e-01 1.79516464e-01 -6.15412109e-02
-4.80456263e-01 -5.53353131e-01 -2.12201640e-01 4.12290215e-01
-5.18191338e-01 6.04404569e-01 -7.25225583e-02 4.91837747e-02
-7.69107789e-02 -4.16002929e-01 -3.29192072e-01 -6.33958101e-01
-1.56698212e-01 -6.03162467e-01 -2.09957585e-01 4.71270263e-01
1.98609263e-01 3.65216345e-01 -1.19552052e+00 2.84511238e-01
8.14477742e-01 1.60527158e+00 -4.14682850e-02 8.09933841e-01
5.17832160e-01 4.37915832e-01 -8.31593513e-01 5.46953194e-02
7.83868730e-02 -9.90619779e-01 3.81249577e-01 -4.90656853e-01
-3.37404311e-02 -5.32321155e-01 -7.56787241e-01 9.35086071e-01
1.10175729e+00 -7.49448597e-01 6.13683939e-01 -9.92539227e-01
-3.01092863e-01 5.41375995e-01 -4.15607005e-01 5.73339760e-01
1.41215369e-01 3.01362664e-01 2.60704421e-02 -7.75346696e-01
1.70934767e-01 1.86279759e-01 1.38486516e+00 1.67905658e-01
-1.08718228e+00 -1.01213765e+00 -9.43000674e-01 1.10016413e-01
-1.83049309e+00 -3.10803831e-01 9.97601926e-01 -3.81980576e-02
1.25081778e+00 2.90078312e-01 1.36868870e+00 8.80387306e-01
8.79591227e-01 1.89424261e-01 1.05509806e+00 -1.84393764e-01
6.51076972e-01 -4.45323437e-01 1.55091628e-01 3.39018285e-01
9.22575831e-01 -1.01530641e-01 -8.95809054e-01 -5.36082566e-01
1.03153288e+00 4.57741320e-01 -4.88801062e-01 5.47972143e-01
-1.11140430e+00 4.74467836e-02 3.34604442e-01 3.60029846e-01
-7.05991089e-01 7.84888506e-01 6.20527864e-01 1.19560517e-01
-3.85632634e-01 2.88831949e-01 7.84538239e-02 -3.82312387e-01
-1.30570424e+00 -4.07140963e-02 5.55499732e-01 8.40548813e-01
-1.06049843e-01 8.54608417e-01 -3.29371065e-01 1.88562110e-01
1.73731655e-01 7.20160484e-01 9.58402395e-01 -8.64208281e-01
8.02607611e-02 6.21043801e-01 -1.60681590e-01 -7.96686471e-01
-5.56700230e-01 -7.25809693e-01 -1.61218464e+00 -3.29585522e-01
2.98535936e-02 -3.06123465e-01 -5.39874375e-01 1.53233409e+00
-1.17915966e-01 4.89108354e-01 4.31155205e-01 6.44132018e-01
7.61926532e-01 5.04847109e-01 -5.69170937e-02 -7.06160307e-01
1.47930574e+00 1.20057508e-01 -1.15279639e+00 -3.93071324e-02
2.09060520e-01 -2.80512750e-01 6.66703641e-01 3.67858946e-01
-1.16913402e+00 -5.31299055e-01 -1.55410063e+00 1.30970806e-01
5.34539819e-01 4.93516773e-01 2.05308840e-01 8.06837142e-01
-8.31731856e-01 3.17076236e-01 -9.97830868e-01 -1.11075431e-01
8.05490136e-01 4.11326855e-01 2.82858700e-01 5.09196937e-01
-1.06565535e+00 2.95309812e-01 8.56157541e-02 6.00510538e-02
-6.74635231e-01 -8.77573431e-01 -3.28124523e-01 3.94451886e-01
-6.96406245e-01 -6.58352792e-01 9.61281717e-01 -6.26791894e-01
-1.25783157e+00 3.43378544e-01 -1.57320909e-02 -1.08421981e+00
1.91790685e-01 6.17099330e-02 -7.14495003e-01 5.79449713e-01
-3.35757911e-01 7.70919723e-03 7.61060178e-01 -3.17848384e-01
6.84026033e-02 -5.91957092e-01 -9.34498191e-01 3.70273180e-02
-7.10270345e-01 -3.28092456e-01 1.95632264e-01 -8.83118689e-01
5.11921167e-01 -5.76419175e-01 -8.30011219e-02 1.43196017e-01
1.67864542e-02 5.37069440e-01 8.74782979e-01 -4.70771223e-01
1.78300512e+00 -2.50742602e+00 -5.19470811e-01 2.72636592e-01
1.97121888e-01 4.71300960e-01 4.30653006e-01 5.59115708e-01
3.30661982e-01 -2.97768176e-01 -7.45079100e-01 2.40028441e-01
-5.30412257e-01 -2.15250865e-01 -3.54175061e-01 7.88450003e-01
2.23925352e-01 1.10909379e+00 -3.21003407e-01 -9.33606774e-02
-1.89672917e-01 8.23800504e-01 -2.10386992e-01 -1.08556941e-01
4.82821524e-01 4.38036263e-01 -3.55554432e-01 8.03838253e-01
5.34135282e-01 -2.09113359e-01 4.62553054e-01 -5.91575921e-01
6.56441506e-03 2.14694992e-01 -1.18618166e+00 1.91312909e+00
4.47281376e-02 5.93147695e-01 2.44445018e-02 -5.80948353e-01
1.29321015e+00 7.86424935e-01 3.68917495e-01 -1.00291729e+00
3.04288268e-01 6.32070184e-01 2.10632905e-01 -5.02554536e-01
-4.09468599e-02 -1.11921005e-01 -2.38661900e-01 3.80205214e-01
-9.00934413e-02 8.60283524e-02 -6.11350477e-01 5.10731805e-03
1.52372193e+00 -3.21551450e-02 4.80932832e-01 -6.06893480e-01
2.19529137e-01 -1.94961652e-01 8.21157515e-01 5.42496324e-01
-3.86898726e-01 3.61436903e-01 -3.13211535e-03 -6.11494184e-01
-9.48606074e-01 -1.05429196e+00 -3.55909109e-01 -1.62498891e-01
2.74114877e-01 -4.42665368e-01 -9.72298682e-01 4.24941748e-01
-1.02242291e-01 2.27549955e-01 3.34432833e-02 -4.69225019e-01
-7.22671390e-01 -7.11774826e-01 1.74815011e+00 7.65511692e-01
1.01126075e+00 -8.70990396e-01 -1.94929445e+00 7.71250963e-01
-1.05820969e-01 -1.01188552e+00 -2.10598126e-01 4.16713446e-01
-1.27004886e+00 -5.02900183e-01 -5.14006615e-01 -7.84822166e-01
4.63026971e-01 -3.96300405e-01 5.20596564e-01 -6.69139847e-02
-8.47577095e-01 1.21379822e-01 2.22732112e-01 -6.32766008e-01
1.68259859e-01 -4.63200420e-01 4.41322029e-01 -3.62818688e-02
5.54761112e-01 -1.09959221e+00 -1.34596288e+00 -2.24667326e-01
-7.79189050e-01 -3.55163753e-01 6.68446720e-01 5.84038079e-01
7.72911668e-01 -1.46673277e-01 1.30067801e+00 -2.68304259e-01
7.05642939e-01 -8.92016590e-02 -4.15290564e-01 -1.19870618e-01
-6.38028264e-01 -4.75720733e-01 9.63411808e-01 -5.91077626e-01
-4.07007098e-01 4.89939928e-01 -1.05176575e-01 -3.14293504e-01
2.37198457e-01 2.31925875e-01 2.29247306e-02 -6.25796691e-02
8.66534650e-01 8.23825419e-01 1.41088650e-01 4.38996367e-02
-4.75179166e-01 9.52247322e-01 1.00421214e+00 -9.70712490e-03
7.93320537e-02 4.21967238e-01 3.03579062e-01 -8.42823505e-01
2.45698914e-01 -1.60830662e-01 -4.74402122e-02 -1.62641332e-01
7.02029824e-01 -1.54124546e+00 -9.65780735e-01 7.23017216e-01
-1.11001194e+00 8.98369327e-02 -3.61773163e-01 7.71650374e-01
-5.10873497e-01 2.80651838e-01 -1.05622709e+00 -1.25496733e+00
-1.46280980e+00 -6.70679808e-01 5.51842391e-01 2.97128648e-01
-5.62958360e-01 -9.05393660e-02 -1.56686023e-01 -2.17735901e-01
7.31098950e-01 8.35941613e-01 7.23252416e-01 -2.28683621e-01
-4.13712829e-01 -4.86154079e-01 2.14127898e-01 1.32260509e-02
-6.52462766e-02 -5.53872705e-01 -1.21229839e+00 -4.30162400e-01
8.98067951e-01 -1.11676075e-01 4.80103642e-01 6.57836974e-01
1.18642664e+00 -3.53081346e-01 -3.72898012e-01 8.07543874e-01
1.86206484e+00 4.80645597e-01 1.06054628e+00 -5.22537351e-01
4.03604865e-01 -2.56844342e-01 -6.63721263e-02 7.36859024e-01
1.23826684e-02 4.98936363e-02 5.09569719e-02 7.51371607e-02
-4.46160026e-02 -1.42476469e-01 3.20744455e-01 1.45014596e+00
1.13674201e-01 -1.19575463e-01 -8.30798626e-01 5.29338360e-01
-1.51230586e+00 -8.46663415e-01 -5.84942877e-01 2.30376029e+00
8.76424670e-01 -1.20775355e-02 -8.31566006e-03 8.03542614e-01
6.62259519e-01 -5.85854173e-01 -1.15378773e+00 -4.85920340e-01
-3.28435928e-01 6.52106464e-01 6.54557705e-01 -3.51806968e-01
-3.88607740e-01 -1.97443947e-01 5.48624277e+00 4.19279933e-01
-1.35233128e+00 2.40490481e-01 4.30074304e-01 -4.23956305e-01
5.61790727e-02 -4.87143785e-01 -4.98776257e-01 7.83840299e-01
1.46520615e+00 -3.22929621e-01 9.67481807e-02 4.69019353e-01
2.20911443e-01 2.01577041e-02 -1.05429816e+00 1.55938625e+00
-2.08909184e-01 -1.44293940e+00 -7.09404647e-02 -7.13449866e-02
4.40488487e-01 -2.97928870e-01 8.94454401e-03 -3.48468810e-01
-1.04201198e+00 -9.89053607e-01 6.14039719e-01 7.20607817e-01
1.67738581e+00 -1.00498390e+00 6.23605847e-01 4.76192087e-01
-1.54732776e+00 -3.68046761e-01 -2.62675703e-01 -4.43377554e-01
5.43845296e-02 9.74028409e-01 -5.61155140e-01 1.89962775e-01
7.77866304e-01 4.79979426e-01 -1.19622886e-01 9.72440660e-01
5.12135983e-01 1.07018697e+00 -6.39678955e-01 -2.21753538e-01
-2.83374697e-01 1.95703283e-02 5.11314213e-01 1.28543258e+00
8.56058478e-01 6.09628558e-01 -2.40020454e-01 8.83213162e-01
-2.31212094e-01 -3.72492135e-01 -6.83402359e-01 1.62532534e-02
1.09193850e+00 1.08527172e+00 -7.16059208e-01 -3.79997134e-01
-2.93231010e-02 8.62409592e-01 -3.08488041e-01 -1.89108141e-02
-9.14814949e-01 -1.21577764e+00 1.88062459e-01 4.28909600e-01
-9.66722742e-02 -3.32791567e-01 -1.09098148e+00 -6.37814462e-01
3.94330829e-01 -6.82332635e-01 -2.47044191e-02 -3.78819227e-01
-6.40774310e-01 5.57438374e-01 -8.31075132e-01 -1.64130819e+00
3.97394523e-02 8.56444240e-02 -6.84928238e-01 8.55608582e-01
-9.48760509e-01 -5.51505864e-01 -5.34685314e-01 7.05743790e-01
2.01638609e-01 1.38598874e-01 1.17088115e+00 4.70176011e-01
-1.00848600e-01 7.54176497e-01 -1.36973953e-03 8.22978541e-02
3.11297029e-01 -6.66125953e-01 3.12987328e-01 1.03552318e+00
-2.33568057e-01 7.68502414e-01 1.23149119e-01 -9.11630630e-01
-2.13474989e+00 -1.30224526e+00 6.35365486e-01 1.79633602e-01
-1.93688516e-02 -4.95827138e-01 -9.56672490e-01 -6.74326494e-02
1.35733753e-01 1.95610803e-03 9.94038463e-01 -1.19365656e+00
-3.88025269e-02 -5.11944413e-01 -1.54470015e+00 5.05776048e-01
8.12377095e-01 -7.16774940e-01 -3.48999411e-01 -1.62273467e-01
3.87637645e-01 -1.35631159e-01 -1.08388186e+00 7.31860936e-01
9.31828260e-01 -6.40770257e-01 7.02679336e-01 3.90264392e-01
1.88386440e-01 -6.80402994e-01 -1.37318522e-01 -5.53125560e-01
-2.97854125e-01 -9.28860962e-01 -5.45898676e-01 1.07913780e+00
9.18488875e-02 -5.92949688e-01 6.36351049e-01 1.75675899e-01
-8.64219964e-02 -9.38605189e-01 -1.19360662e+00 -7.76975691e-01
-4.80215430e-01 -1.83254167e-01 3.19868058e-01 4.08587337e-01
4.27458137e-01 1.32040799e-01 -2.01828584e-01 1.70143500e-01
8.97946537e-01 -1.92514539e-01 -1.65901944e-01 -1.01376629e+00
-2.93271482e-01 -3.20403278e-02 -7.40209937e-01 -5.62131822e-01
-8.79400194e-01 -8.70860636e-01 -5.49535602e-02 -1.18101788e+00
4.78266738e-02 -9.53345820e-02 -6.78691745e-01 6.91402480e-02
1.66331488e-03 9.94555175e-01 -1.87403142e-01 3.25045645e-01
-1.58983409e-01 2.16561124e-01 5.91783881e-01 1.44842252e-01
-4.29436028e-01 -4.26682562e-01 -2.95063674e-01 6.79428995e-01
8.71022105e-01 -5.17634511e-01 -3.81779432e-01 -3.25003684e-01
1.05701491e-01 5.65454721e-01 5.30825436e-01 -1.96709800e+00
7.90912867e-01 7.85820186e-01 1.14339864e+00 -7.18410015e-01
3.74842465e-01 -7.09529817e-01 8.65798652e-01 1.38145876e+00
-1.61539704e-01 5.10233283e-01 3.38007629e-01 6.31503642e-01
-3.12293284e-02 2.68057138e-01 7.86755860e-01 1.53189912e-01
-7.09986873e-03 6.18408024e-02 -8.89184952e-01 -7.34402388e-02
1.06575465e+00 -8.92160296e-01 -1.06384262e-01 1.86442241e-01
-3.63489270e-01 -2.65192449e-01 1.75460964e-01 -4.07663435e-01
1.17678714e+00 -1.40921402e+00 -5.25589943e-01 6.43219531e-01
-1.44021921e-02 -3.32238972e-01 5.75421989e-01 1.11824632e+00
-8.26440573e-01 2.73729563e-01 -7.92115211e-01 -8.94141376e-01
-1.00008357e+00 -6.63629919e-03 1.17558338e-01 1.88908175e-01
-1.02807093e+00 5.29120088e-01 -6.40277207e-01 8.72647703e-01
3.01161647e-01 -4.24226373e-01 2.36490965e-01 -2.11551502e-01
7.49545157e-01 4.84769523e-01 1.28062144e-01 8.95636454e-02
-7.01947749e-01 7.16140866e-01 5.89775205e-01 -8.99321884e-02
1.13449144e+00 -4.17619338e-03 3.81473973e-02 6.36273563e-01
9.71956789e-01 -3.37367475e-01 -1.07191086e+00 7.85724893e-02
-2.24037871e-01 4.31878604e-02 -2.25832388e-02 -8.68914962e-01
-7.81784654e-01 8.89923513e-01 1.18684268e+00 -6.14556819e-02
1.78017783e+00 -9.03452754e-01 1.32606685e+00 2.71706700e-01
6.21810913e-01 -8.66939962e-01 1.44809019e-02 -2.01286167e-01
5.24816215e-01 -2.40363494e-01 7.89300650e-02 1.29679412e-01
-4.06573921e-01 1.28238571e+00 2.44417265e-01 -6.67605221e-01
5.69800138e-01 9.35260296e-01 -1.44831598e-01 -1.46851450e-01
-5.47095537e-01 5.90985477e-01 -2.11242005e-01 5.88420451e-01
3.72143894e-01 -2.42688283e-02 -8.45504165e-01 1.22093809e+00
8.73977169e-02 7.06092656e-01 7.05268383e-01 1.24498641e+00
-4.19189513e-01 -3.64719898e-01 -1.69461459e-01 1.00570714e+00
-9.28340614e-01 -1.95584595e-01 -1.71914026e-01 -1.78629741e-01
1.83569849e-01 6.82429194e-01 1.93573967e-01 -4.00030077e-01
7.43859932e-02 3.42983335e-01 5.29257834e-01 -1.44755393e-01
-1.21542525e+00 3.63184363e-01 -3.55055839e-01 -5.25756657e-01
-4.07829955e-02 -3.51054132e-01 -1.60579801e+00 1.73190445e-01
-2.03123227e-01 -1.92469612e-01 8.80209744e-01 4.51795965e-01
1.00275493e+00 9.51923788e-01 2.61520922e-01 -1.39113009e-01
-5.72064519e-01 -8.58300149e-01 -8.54395628e-01 -9.06687453e-02
5.51283419e-01 2.46406525e-01 -5.76156229e-02 3.48660588e-01] | [13.947436332702637, 3.149134874343872] |
cc327bf9-f2bb-4f78-94cb-18edfc0565a9 | cg-bert-conditional-text-generation-with-bert | 2004.01881 | null | https://arxiv.org/abs/2004.01881v1 | https://arxiv.org/pdf/2004.01881v1.pdf | CG-BERT: Conditional Text Generation with BERT for Generalized Few-shot Intent Detection | In this paper, we formulate a more realistic and difficult problem setup for the intent detection task in natural language understanding, namely Generalized Few-Shot Intent Detection (GFSID). GFSID aims to discriminate a joint label space consisting of both existing intents which have enough labeled data and novel intents which only have a few examples for each class. To approach this problem, we propose a novel model, Conditional Text Generation with BERT (CG-BERT). CG-BERT effectively leverages a large pre-trained language model to generate text conditioned on the intent label. By modeling the utterance distribution with variational inference, CG-BERT can generate diverse utterances for the novel intents even with only a few utterances available. Experimental results show that CG-BERT achieves state-of-the-art performance on the GFSID task with 1-shot and 5-shot settings on two real-world datasets. | ['Chenwei Zhang', 'Philip Yu', 'Congying Xia', 'Hoang Nguyen', 'Jiawei Zhang'] | 2020-04-04 | null | null | null | null | ['conditional-text-generation'] | ['natural-language-processing'] | [ 3.65559965e-01 2.76516944e-01 -9.29467380e-02 -6.28841162e-01
-1.11867869e+00 -6.19832836e-02 9.02459502e-01 -2.62315512e-01
-1.74732789e-01 6.62989855e-01 5.22231996e-01 -3.45912874e-02
5.05296767e-01 -5.54612100e-01 -2.40646571e-01 -4.79324460e-01
3.23912591e-01 9.58515406e-01 -1.01380050e-01 -2.19673827e-01
1.46441773e-01 -3.74750227e-01 -1.57141542e+00 2.54119575e-01
8.81452203e-01 6.17079616e-01 5.65498710e-01 9.60750580e-01
-2.55253941e-01 9.07662094e-01 -7.65602171e-01 -1.13489822e-01
-3.20830308e-02 -6.13184810e-01 -7.84857035e-01 2.13137880e-01
-3.36469449e-02 -7.75804281e-01 -3.09334606e-01 8.10647309e-01
5.10723889e-01 6.16462171e-01 1.27051699e+00 -1.44186068e+00
-6.23595357e-01 5.89348793e-01 -3.40941310e-01 1.51616782e-01
7.06122220e-01 3.89454067e-01 1.39352787e+00 -1.14493632e+00
6.64221585e-01 1.64641476e+00 1.50841072e-01 1.00985980e+00
-1.19502997e+00 -5.22535801e-01 2.80099332e-01 4.01069410e-02
-1.26766360e+00 -4.67979103e-01 5.78293979e-01 -5.11533141e-01
1.05292165e+00 1.28970027e-01 -2.87279231e-03 1.60431635e+00
1.48286792e-02 1.40191019e+00 6.65366709e-01 -4.56593305e-01
7.05907047e-01 -1.04415054e-02 7.28271008e-01 4.37280893e-01
-1.71859041e-01 -1.91900671e-01 -6.73786402e-01 -2.78732747e-01
2.64077336e-01 1.98934466e-01 -2.64442831e-01 1.28093258e-01
-1.13127899e+00 1.36690271e+00 -1.60921931e-01 8.62396881e-02
-3.38377029e-01 -5.00135832e-02 4.66417313e-01 2.43599564e-01
9.00213599e-01 3.39747488e-01 -1.85927555e-01 -2.45921612e-01
-9.05251861e-01 3.90079468e-01 9.24015105e-01 1.26174152e+00
8.41714263e-01 -1.94148161e-02 -9.61758316e-01 9.44848835e-01
5.42567849e-01 5.24128556e-01 7.13709116e-01 -5.60602009e-01
5.43643773e-01 3.22317690e-01 3.76055479e-01 -3.34718078e-01
-2.27516010e-01 7.59218484e-02 -6.62880003e-01 -2.12674499e-01
4.97076055e-03 -3.85152638e-01 -1.42775905e+00 1.83870649e+00
3.83799523e-01 3.81684214e-01 2.88010627e-01 6.73222601e-01
9.65249538e-01 1.22655332e+00 1.47363722e-01 -3.24009508e-01
1.37990010e+00 -1.27415967e+00 -7.80336201e-01 -5.61277628e-01
7.30858088e-01 -4.44065601e-01 1.40467906e+00 -5.74117117e-02
-6.98372245e-01 -5.08598387e-01 -9.01590407e-01 -1.07987992e-01
-1.45790726e-01 5.40669337e-02 3.60127836e-01 5.61886311e-01
-9.34976339e-01 -8.20101798e-02 -6.70476437e-01 -4.48592216e-01
2.05720156e-01 -2.93088436e-01 1.38699055e-01 -4.68117744e-01
-1.44541252e+00 4.61598426e-01 5.60983956e-01 -4.89654630e-01
-1.56450355e+00 -7.12619781e-01 -1.16110158e+00 1.37752309e-01
5.99914670e-01 -8.71313512e-01 1.82908213e+00 -3.35106641e-01
-1.54215872e+00 7.45598614e-01 -6.29927754e-01 -5.25638223e-01
4.24813151e-01 -1.91775799e-01 -2.67984360e-01 9.72177237e-02
6.75270796e-01 8.63420069e-01 1.19009769e+00 -1.19951522e+00
-6.57906711e-01 -2.21630007e-01 -1.07099917e-02 3.93366605e-01
-1.07527860e-01 -1.20927952e-01 -3.57725888e-01 -6.26309752e-01
-2.86094964e-01 -8.17068160e-01 -2.44828716e-01 -4.64531690e-01
-7.51993239e-01 -8.14386427e-01 1.04961252e+00 -3.57763380e-01
1.01873219e+00 -2.06970549e+00 -6.03025891e-02 -4.92127031e-01
1.26125887e-01 2.35267103e-01 -3.00772548e-01 5.32652497e-01
5.12367189e-01 9.65011418e-02 -1.76339403e-01 -1.07482827e+00
4.78594154e-01 3.34242642e-01 -8.85707200e-01 -1.42268464e-01
1.99094415e-01 9.85621750e-01 -1.12953222e+00 -4.30754125e-01
2.42044508e-01 2.38107726e-01 -4.53897595e-01 8.02931249e-01
-9.08337653e-01 2.58598745e-01 -6.96067333e-01 4.87522572e-01
4.93965894e-01 -5.73425233e-01 -1.53136089e-01 3.32497180e-01
2.59271353e-01 5.62393010e-01 -8.38601708e-01 1.74499464e+00
-5.73274851e-01 3.82258713e-01 -1.26575381e-01 -7.00668335e-01
7.13969350e-01 5.41112840e-01 -3.08562312e-02 -2.30944276e-01
1.20690450e-01 -1.31489262e-01 -4.71860886e-01 -4.44871336e-01
6.97996914e-01 -5.81334591e-01 -5.35547495e-01 1.06290603e+00
5.50815165e-01 -3.46874952e-01 3.66398424e-01 7.21677005e-01
1.13777399e+00 -2.05372110e-01 4.19763654e-01 3.75512429e-02
1.16555072e-01 -2.64196247e-01 3.41899246e-01 1.41912687e+00
-3.56564641e-01 8.51956427e-01 2.87438184e-01 -6.16686866e-02
-7.03522503e-01 -1.19347036e+00 1.82320848e-01 1.47930145e+00
1.36973426e-01 -3.29604626e-01 -5.42504013e-01 -9.85149860e-01
-4.43205953e-01 1.64396882e+00 -6.29492939e-01 -2.18217641e-01
-2.90157795e-02 -5.58968902e-01 2.86540657e-01 2.39016056e-01
3.85653704e-01 -1.32075894e+00 -6.54798865e-01 2.23651290e-01
-6.38405561e-01 -1.12187874e+00 -8.14584553e-01 2.21567571e-01
-2.53410101e-01 -5.07659554e-01 -8.54661047e-01 -7.19753623e-01
4.99733001e-01 4.73145902e-01 1.20457339e+00 -2.00859591e-01
-4.08715814e-01 5.12062609e-01 -8.26062799e-01 -5.95954537e-01
-6.69031560e-01 -1.20595582e-01 -5.43164723e-02 1.42773971e-01
5.61811924e-01 -1.44980386e-01 -2.01696396e-01 9.72626433e-02
-9.66342688e-01 4.68467176e-01 3.58638674e-01 1.03032935e+00
1.46684393e-01 -1.02486610e-01 9.25143719e-01 -7.78754056e-01
7.63394654e-01 -8.22663248e-01 2.44814926e-03 3.16940218e-01
-3.30028296e-01 5.67088947e-02 6.78453982e-01 -3.23679954e-01
-1.61973643e+00 -4.12329137e-01 -3.49533737e-01 -5.30278265e-01
-6.19441152e-01 5.47410786e-01 -6.58270344e-02 8.78424942e-01
3.90298963e-01 5.51622808e-01 -3.16565096e-01 -3.68423671e-01
5.53160787e-01 1.04828417e+00 2.41000980e-01 -3.68935913e-01
4.27653611e-01 3.83908510e-01 -6.53910816e-01 -1.23808002e+00
-1.44945359e+00 -1.05229199e+00 -3.47667992e-01 -1.63836524e-01
1.10483158e+00 -1.09983790e+00 -6.45890012e-02 5.62627614e-01
-1.33576059e+00 -5.66033959e-01 -4.69612449e-01 3.16243708e-01
-6.61937237e-01 2.15237558e-01 -6.13287568e-01 -1.34033680e+00
-5.76024830e-01 -1.12506509e+00 1.73189318e+00 2.28903934e-01
-4.94779289e-01 -1.01290739e+00 2.60560691e-01 3.80048752e-01
3.70117440e-03 -2.52217948e-01 8.61584485e-01 -1.16865706e+00
-4.93401080e-01 -1.80946887e-01 6.70196582e-03 1.40707225e-01
1.53857246e-01 -4.12852168e-01 -1.17617261e+00 -2.22226396e-01
4.23993498e-01 -1.07218480e+00 1.15527809e+00 2.83083528e-01
5.89883626e-01 -2.97408015e-01 -4.45319742e-01 7.37681240e-02
1.09620595e+00 3.78096879e-01 5.04982173e-01 -4.01834309e-01
5.12257636e-01 4.53399748e-01 9.92603719e-01 8.42092752e-01
4.56932396e-01 5.17875731e-01 5.06811813e-02 2.35394210e-01
1.31086642e-02 -5.07622540e-01 4.45843935e-01 5.87305665e-01
6.41166747e-01 -9.42669392e-01 -7.35641837e-01 7.11621165e-01
-1.95953321e+00 -1.04943836e+00 2.12962359e-01 1.80968547e+00
8.91402304e-01 1.72487065e-01 -5.46051487e-02 -2.33799577e-01
8.08512509e-01 5.21412015e-01 -6.52203918e-01 -1.71283722e-01
2.90821373e-01 -3.01371999e-02 -4.29653913e-01 5.55629730e-01
-1.19468987e+00 1.02560902e+00 6.14716721e+00 1.12408245e+00
-6.85564280e-01 2.72267103e-01 8.80816758e-01 -3.90392132e-02
-4.66849148e-01 -1.84217259e-01 -1.14702725e+00 6.20150566e-01
9.87725973e-01 -3.83351624e-01 -3.32566276e-02 1.06502867e+00
2.36473501e-01 -1.81234747e-01 -1.16266668e+00 8.19110870e-01
4.32416141e-01 -1.11593807e+00 2.54460573e-01 -2.63816714e-01
9.29392278e-01 5.13407923e-02 -5.55976527e-03 1.01492369e+00
7.60701299e-01 -8.42739522e-01 4.87333626e-01 2.60233760e-01
8.35844278e-01 -3.76969963e-01 5.86182177e-01 1.00187862e+00
-1.00062466e+00 1.04061462e-01 -3.52246732e-01 -9.91926715e-02
7.99165130e-01 6.20814145e-01 -1.31152797e+00 3.73581856e-01
2.23300338e-01 6.35617614e-01 1.68933196e-03 5.72087467e-01
-3.14724892e-01 8.85886014e-01 -1.51862144e-01 -3.66284072e-01
4.96974826e-01 -8.71768594e-02 7.50128031e-01 1.15392685e+00
3.38148892e-01 3.02101851e-01 5.75371563e-01 1.12223065e+00
-2.20173895e-01 -1.48010746e-01 -6.15524471e-01 -2.85976321e-01
4.24680948e-01 1.05723774e+00 -6.62921607e-01 -8.71059895e-01
-3.50660861e-01 1.32993579e+00 2.79979616e-01 5.03436565e-01
-8.22088182e-01 -2.42802978e-01 6.32438242e-01 -4.41606522e-01
3.51602882e-01 1.32280350e-01 -7.55344480e-02 -1.52318704e+00
-4.35187459e-01 -5.48383832e-01 5.39553285e-01 -9.57062900e-01
-1.56674111e+00 4.70476121e-01 2.22047314e-01 -9.16616380e-01
-8.10095966e-01 -2.57989436e-01 -1.06244218e+00 7.73075461e-01
-1.25429535e+00 -1.15276814e+00 -1.19810104e-01 3.58191878e-01
1.63137221e+00 -1.42050520e-01 9.28684175e-01 -2.94398248e-01
-5.13164341e-01 4.21922296e-01 9.52621251e-02 -1.62064448e-01
6.08793080e-01 -1.21972346e+00 8.79608989e-01 8.14861536e-01
1.81786686e-01 4.21080679e-01 9.15690303e-01 -8.26758742e-01
-9.45224583e-01 -1.39547420e+00 1.04869854e+00 -2.61821121e-01
4.13032860e-01 -7.18773246e-01 -7.86684394e-01 9.15212035e-01
2.63194203e-01 -4.99820054e-01 8.28245580e-01 -3.59036848e-02
-3.03863704e-01 8.30410004e-01 -1.15342677e+00 8.07800233e-01
8.66155386e-01 -4.62496221e-01 -9.83664870e-01 5.91725230e-01
1.28976572e+00 -1.07261844e-01 -1.28659189e-01 1.74673751e-01
-2.40943190e-02 -8.66969287e-01 8.23299050e-01 -6.10631168e-01
5.81908703e-01 3.25103886e-02 -2.09868014e-01 -1.43191814e+00
-1.38636976e-01 -6.97109222e-01 -3.21079493e-01 1.21932507e+00
2.86846697e-01 -4.53636706e-01 5.43447256e-01 7.13455200e-01
-3.69134992e-01 -5.93302727e-01 -8.14207137e-01 -7.65594125e-01
-2.68436551e-01 -5.44158518e-01 4.02867198e-01 6.29657269e-01
1.42392486e-01 1.15210629e+00 -7.15688109e-01 -1.17802002e-01
6.38611317e-01 3.40325654e-01 9.66302454e-01 -9.79541361e-01
-4.57445383e-01 1.06053188e-01 -4.03827280e-02 -1.66951632e+00
5.06533682e-01 -6.69997811e-01 8.69315684e-01 -1.77520978e+00
5.43659270e-01 -1.52185902e-01 9.85333323e-02 4.06965077e-01
-8.10435236e-01 2.75112279e-02 9.19122919e-02 2.13400096e-01
-1.13782513e+00 1.10515583e+00 9.84094858e-01 -3.36258382e-01
-2.21092448e-01 1.23580635e-01 -7.04851568e-01 6.88858926e-01
5.15255213e-01 -4.15802240e-01 -8.68942440e-01 3.55828628e-02
-4.50937271e-01 2.23610416e-01 -2.44553667e-02 -7.51163840e-01
-3.44787799e-02 -3.75665009e-01 -1.15187541e-01 -9.62013662e-01
6.30845070e-01 -1.24157384e-01 -2.90743887e-01 2.85195827e-01
-8.40581298e-01 -8.78154397e-01 -3.15278441e-01 9.88749325e-01
-4.17532884e-02 -6.75848484e-01 5.89083791e-01 -2.17416063e-01
-9.60220516e-01 3.56411189e-01 -7.26037621e-01 6.36067748e-01
1.04693639e+00 1.46628261e-01 -3.32188547e-01 -8.62103343e-01
-5.91279328e-01 4.55136478e-01 1.30919784e-01 5.27734816e-01
8.40688705e-01 -1.07049215e+00 -8.22597027e-01 2.91209459e-01
5.46420455e-01 1.44391268e-01 6.40268505e-01 2.38190085e-01
1.81098029e-01 5.86781859e-01 5.26046932e-01 -6.33514881e-01
-9.94172633e-01 6.59592688e-01 -2.37145543e-01 -7.37473428e-01
-5.61815083e-01 1.18101084e+00 9.05058920e-01 -5.67383468e-01
1.64771453e-01 6.86767336e-04 -2.08068430e-01 6.14663959e-03
9.33301508e-01 1.92874119e-01 -3.65192980e-01 -6.66796923e-01
-1.93000492e-02 -2.09136263e-01 -5.10200143e-01 -4.47660625e-01
8.73478591e-01 -3.70975941e-01 3.30209255e-01 1.01974308e+00
1.25385451e+00 -5.76184869e-01 -1.38327014e+00 -2.64779657e-01
-3.01001519e-01 -4.75563228e-01 -1.14694931e-01 -7.03388870e-01
-1.76163793e-01 8.41313541e-01 8.96784514e-02 3.50380570e-01
6.69818401e-01 3.62234771e-01 1.08655858e+00 4.45010304e-01
5.52943826e-01 -1.00087667e+00 7.57369995e-01 1.08524919e+00
9.16475236e-01 -1.48420215e+00 -5.01219928e-01 -2.11424708e-01
-1.13629079e+00 5.87380171e-01 6.91019654e-01 2.56194532e-01
5.72175622e-01 -4.46831202e-03 1.32875517e-01 -2.31900007e-01
-1.16915417e+00 -3.87803525e-01 1.88997000e-01 5.41235387e-01
1.71345428e-01 2.16333196e-01 -1.70805771e-02 6.95939243e-01
-1.65732093e-02 -1.24341831e-01 5.87503970e-01 9.39876735e-01
-7.24850178e-01 -6.96629047e-01 -1.73247963e-01 8.85341048e-01
-1.59706548e-01 -3.59302521e-01 -2.50939250e-01 2.34712854e-01
-4.14615929e-01 1.44340217e+00 1.08006619e-01 -3.41960937e-01
-2.83732623e-01 6.77984893e-01 -7.58833289e-02 -1.33262002e+00
2.45818831e-02 2.22536042e-01 9.70692188e-02 -1.61388919e-01
-6.56676888e-02 -5.18534184e-01 -1.26604450e+00 2.53362864e-01
-5.42492688e-01 2.40685552e-01 3.71545494e-01 1.34898257e+00
3.34663689e-01 5.16908646e-01 7.27065623e-01 -9.77162242e-01
-9.08182979e-01 -1.34489083e+00 -7.72554338e-01 5.42205989e-01
4.44594562e-01 -6.62155509e-01 -6.38260365e-01 1.98150426e-01] | [12.090154647827148, 7.667134761810303] |
d4618911-e603-47df-8393-31af3a2f3b55 | learning-calibrated-guidance-for-object | 2103.11399 | null | https://arxiv.org/abs/2103.11399v4 | https://arxiv.org/pdf/2103.11399v4.pdf | Learning Calibrated-Guidance for Object Detection in Aerial Images | Object detection is one of the most fundamental yet challenging research topics in the domain of computer vision. Recently, the study on this topic in aerial images has made tremendous progress. However, complex background and worse imaging quality are obvious problems in aerial object detection. Most state-of-the-art approaches tend to develop elaborate attention mechanisms for the space-time feature calibrations with arduous computational complexity, while surprisingly ignoring the importance of feature calibrations in channel-wise. In this work, we propose a simple yet effective Calibrated-Guidance (CG) scheme to enhance channel communications in a feature transformer fashion, which can adaptively determine the calibration weights for each channel based on the global feature affinity correlations. Specifically, for a given set of feature maps, CG first computes the feature similarity between each channel and the remaining channels as the intermediary calibration guidance. Then, re-representing each channel by aggregating all the channels weighted together via the guidance operation. Our CG is a general module that can be plugged into any deep neural networks, which is named as CG-Net. To demonstrate its effectiveness and efficiency, extensive experiments are carried out on both oriented object detection task and horizontal object detection task in aerial images. Experimental results on two challenging benchmarks (DOTA and HRSC2016) demonstrate that our CG-Net can achieve the new state-of-the-art performance in accuracy with a fair computational overhead. The source code has been open sourced at https://github.com/WeiZongqi/CG-Net | ['Mingqiang Wei', 'Qixiang Geng', 'Dong Liang', 'Zongqi Wei', 'Huiyu Zhou', 'Liyan Zhang', 'Dong Zhang'] | 2021-03-21 | null | null | null | null | ['object-detection-in-aerial-images'] | ['computer-vision'] | [ 2.03670755e-01 -6.25581801e-01 9.71560925e-02 -3.50622028e-01
-4.36228395e-01 -3.54396820e-01 1.99709982e-01 -1.64215505e-01
-3.79890680e-01 2.88262844e-01 -2.07697764e-01 -3.00015837e-01
-2.13660017e-01 -8.63149047e-01 -5.96195936e-01 -1.03822637e+00
-1.63437188e-01 -1.94427222e-01 6.40784562e-01 -1.71238169e-01
7.65819550e-02 2.82280326e-01 -1.28257382e+00 1.67064533e-01
7.88584054e-01 1.38635075e+00 3.06439757e-01 5.39063871e-01
1.42040685e-01 2.22853035e-01 -3.24423671e-01 -2.78150946e-01
7.08471656e-01 -3.58598173e-01 -4.78775471e-01 3.15974534e-01
4.37078565e-01 -6.69750094e-01 -4.54944551e-01 1.45190728e+00
6.47677720e-01 -1.47222087e-01 3.08211952e-01 -1.18173575e+00
-5.64322412e-01 5.66713989e-01 -1.00818861e+00 4.54496026e-01
-1.96474105e-01 4.35393929e-01 9.19337094e-01 -1.03726280e+00
5.82114384e-02 1.09934151e+00 5.58345377e-01 2.37629354e-01
-9.46486413e-01 -1.06528497e+00 6.26508832e-01 1.97199613e-01
-1.65075386e+00 -1.15666829e-01 5.42054057e-01 -3.66672248e-01
2.97730088e-01 2.06898570e-01 7.75611460e-01 4.78656352e-01
-4.88204658e-02 8.67016673e-01 8.11513662e-01 -2.35232636e-01
-1.26228169e-01 -6.90644383e-02 1.42338529e-01 8.79024208e-01
4.26158965e-01 1.01964705e-01 -2.20501587e-01 2.19301600e-02
9.01187599e-01 2.75113821e-01 -6.63061202e-01 -3.44128251e-01
-1.35311580e+00 7.24812448e-01 9.97217119e-01 4.91937622e-02
-2.20950767e-01 2.14545146e-01 2.69110560e-01 1.94704264e-01
1.75528005e-01 1.49549097e-01 -5.07176876e-01 4.97825533e-01
-5.21972656e-01 2.30939656e-01 3.19107413e-01 1.19687879e+00
7.66733944e-01 -2.05289096e-01 -4.29472148e-01 6.48813963e-01
3.53647232e-01 4.96324003e-01 2.27968231e-01 -4.45956111e-01
5.45106411e-01 7.99136043e-01 1.31149694e-01 -1.29529953e+00
-3.59460831e-01 -7.40215957e-01 -9.98862684e-01 -7.88506400e-03
3.50342870e-01 -4.96252745e-01 -7.98885643e-01 1.40277696e+00
6.22958183e-01 1.98824137e-01 -1.81497633e-01 1.32454681e+00
9.35656369e-01 7.42769420e-01 -8.47498402e-02 3.24110799e-02
1.40894103e+00 -1.03214025e+00 -2.63177991e-01 -1.94303602e-01
3.83138061e-01 -8.31490576e-01 8.87237132e-01 2.33209848e-01
-6.32650912e-01 -7.66417623e-01 -1.18250239e+00 1.80412531e-01
-1.88620538e-01 7.41249144e-01 8.90578747e-01 4.97005045e-01
-5.90679467e-01 3.26199532e-01 -7.10920095e-01 -3.62500429e-01
7.01626003e-01 4.51519281e-01 -1.41392842e-01 -1.79111332e-01
-9.99133110e-01 1.38865545e-01 5.66198170e-01 4.54699129e-01
-9.57442939e-01 -4.82776582e-01 -4.90761161e-01 3.81698161e-02
7.01956153e-01 -5.28225899e-01 1.07118988e+00 -9.91610527e-01
-1.25267351e+00 3.50325882e-01 2.85412848e-01 -2.21483976e-01
4.67705905e-01 -2.64542907e-01 -3.64532322e-01 -5.95090501e-02
6.47109672e-02 9.04354095e-01 8.88171852e-01 -9.65406477e-01
-1.19682682e+00 -4.65417087e-01 2.76003003e-01 4.05244350e-01
-7.00854003e-01 -4.18400206e-03 -1.00566864e+00 -6.56339943e-01
3.72986823e-01 -9.18961048e-01 -2.74185389e-01 6.09220743e-01
-5.11955619e-01 -1.62756220e-02 7.33978570e-01 -3.88322741e-01
1.34075034e+00 -2.30985260e+00 1.46224111e-01 1.87146381e-01
6.82328343e-02 4.07608479e-01 -1.61455825e-01 2.97102220e-02
2.63542421e-02 -2.06871644e-01 -4.74026650e-01 2.10596517e-01
-2.75180399e-01 -3.48667175e-01 -2.75496155e-01 5.91045737e-01
2.85796911e-01 6.91929162e-01 -8.97987247e-01 -3.96769702e-01
1.98851317e-01 3.40645343e-01 -5.06867111e-01 1.44311339e-01
2.09722836e-02 2.79885650e-01 -7.52763808e-01 1.10128474e+00
1.01120532e+00 -3.32583040e-01 4.81030606e-02 -5.39589763e-01
-3.08108151e-01 -3.40460420e-01 -1.42165482e+00 1.54922318e+00
-7.29565928e-03 5.51646650e-01 4.85950299e-02 -8.04310799e-01
8.40444922e-01 -1.44763350e-01 2.05362499e-01 -3.65698963e-01
4.72384095e-01 1.70851976e-01 3.08128178e-01 -3.81670564e-01
3.34791183e-01 3.61387521e-01 -2.49909572e-02 -1.31201176e-02
-2.40886789e-02 1.28125489e-01 7.44122416e-02 1.81527510e-01
7.21605182e-01 -6.03671409e-02 3.33618104e-01 -2.82943934e-01
6.53111696e-01 1.77759789e-02 4.33414400e-01 6.22814178e-01
-2.04830617e-01 4.16746289e-01 1.02393717e-01 -4.72774029e-01
-5.64468622e-01 -7.02449560e-01 -2.40621254e-01 9.53050077e-01
7.22396731e-01 -3.67406040e-01 -7.49480486e-01 -6.01731062e-01
8.67016986e-02 -1.20273963e-01 -6.78414464e-01 -9.76860821e-02
-1.53599113e-01 -1.16107380e+00 3.68492514e-01 6.66291773e-01
1.15466297e+00 -7.94166684e-01 -6.73499346e-01 2.08351970e-01
-4.50885110e-02 -9.13161218e-01 -5.64926565e-01 7.22304657e-02
-6.75466716e-01 -1.09723866e+00 -9.18386221e-01 -7.20661819e-01
8.65309656e-01 1.00057578e+00 4.96364564e-01 5.05786359e-01
-6.11233592e-01 1.02981694e-01 -4.83049154e-01 -5.85339069e-01
5.80908000e-01 1.68091297e-01 -2.08893165e-01 3.57322544e-01
4.07360315e-01 -1.68276861e-01 -9.76513743e-01 6.71427369e-01
-9.02526677e-01 -4.05290797e-02 8.42153072e-01 1.06297779e+00
7.16869235e-01 2.86814451e-01 7.12317303e-02 -7.26959765e-01
4.11656410e-01 -3.38347495e-01 -1.09679079e+00 1.43671513e-01
-2.66929090e-01 -3.89792264e-01 4.21010852e-01 -3.52202594e-01
-7.79500782e-01 4.43644017e-01 8.77755061e-02 -2.55507648e-01
-1.09828949e-01 5.06391764e-01 -3.97653967e-01 -5.60475588e-01
4.28567797e-01 3.15462261e-01 -3.61528069e-01 -1.69111490e-01
1.36608034e-01 7.16209948e-01 4.59098071e-01 -3.69247973e-01
1.05809784e+00 5.51357269e-01 -9.36887339e-02 -6.76237583e-01
-9.86181915e-01 -6.24856889e-01 -6.04960203e-01 -3.46921563e-01
6.74324095e-01 -1.22969878e+00 -6.77930951e-01 8.07856679e-01
-8.58626068e-01 -2.00235248e-01 2.36835212e-01 4.81746852e-01
1.14599138e-01 3.86599988e-01 -2.25373939e-01 -5.91984093e-01
-3.34109336e-01 -1.26707828e+00 1.00373650e+00 7.09516943e-01
5.44678330e-01 -4.38482225e-01 -4.14579093e-01 -5.42484820e-02
3.48907918e-01 1.01492427e-01 5.40841162e-01 -1.10805243e-01
-1.11157262e+00 -3.43524069e-01 -8.69064808e-01 2.99154848e-01
3.27744260e-02 4.52814959e-02 -9.80039716e-01 -3.72500718e-01
-4.50826645e-01 -1.46955922e-01 1.14287448e+00 4.53478158e-01
1.42794955e+00 7.83651695e-02 -4.89926010e-01 1.01123524e+00
1.50995541e+00 3.03583562e-01 4.96155977e-01 4.34656441e-01
7.82187462e-01 4.11624640e-01 1.02630925e+00 6.33257985e-01
2.15070903e-01 6.67919576e-01 7.35010087e-01 -3.21324229e-01
2.85033192e-02 -1.27340227e-01 8.72104913e-02 4.70367223e-01
-1.49378106e-01 -2.24563450e-01 -5.98976076e-01 4.11971211e-01
-1.94314277e+00 -6.40396416e-01 -3.32904041e-01 2.14232421e+00
4.99989122e-01 7.56408721e-02 -1.28331304e-01 6.92949519e-02
8.02597165e-01 1.18646361e-01 -6.42369151e-01 5.98430276e-01
-7.18107149e-02 -2.36159787e-01 8.63730252e-01 4.09828871e-02
-1.67616951e+00 9.46704209e-01 5.01298761e+00 9.09196734e-01
-1.17513609e+00 -1.23810567e-01 6.39876604e-01 1.49595574e-01
4.23899591e-01 -2.07933802e-02 -1.14003730e+00 6.01247013e-01
3.28621790e-02 1.17554963e-01 4.85629559e-01 1.01676345e+00
-1.97156817e-01 -7.38479346e-02 -6.34007633e-01 1.12389958e+00
-5.10588139e-02 -1.16015494e+00 1.07341275e-01 -1.48413941e-01
6.05573714e-01 4.32349415e-03 1.36867270e-01 7.90144950e-02
5.96970171e-02 -5.50113797e-01 7.07361996e-01 1.68551862e-01
8.26960385e-01 -7.35987961e-01 1.05660355e+00 1.40130594e-01
-1.73972487e+00 -4.79200602e-01 -8.22486043e-01 3.44140902e-02
-9.85033959e-02 4.50407952e-01 -3.11774373e-01 6.69231415e-01
1.13922143e+00 8.83711874e-01 -7.79178858e-01 1.54179573e+00
-2.39177480e-01 4.69999582e-01 -2.07645729e-01 -4.39155176e-02
4.23991740e-01 -1.77633211e-01 2.90378243e-01 1.14367819e+00
4.72491294e-01 5.62153935e-01 4.79380935e-01 4.95910734e-01
-1.15730546e-01 2.58700788e-01 -2.23869577e-01 8.90991315e-02
5.24215519e-01 1.68375874e+00 -9.99344051e-01 -1.97673336e-01
-5.82996666e-01 9.20713067e-01 6.05979972e-02 1.34386107e-01
-9.68916953e-01 -8.10570002e-01 4.51079786e-01 -1.26670867e-01
5.79450369e-01 -1.74008310e-02 1.01617031e-01 -1.01680827e+00
9.29877385e-02 -7.19725311e-01 4.50036734e-01 -6.55538321e-01
-1.26772273e+00 7.49887943e-01 1.60018187e-02 -1.66699767e+00
7.44201839e-01 -8.51063907e-01 -7.07780421e-01 7.39929974e-01
-1.73638546e+00 -1.13478982e+00 -1.12892067e+00 7.95053959e-01
5.33742130e-01 -1.97705194e-01 4.06945735e-01 4.75008845e-01
-8.24011087e-01 5.71487606e-01 -5.86475991e-02 3.77010912e-01
6.69851005e-01 -8.21377575e-01 3.70354325e-01 1.05067754e+00
-1.04350798e-01 4.69483018e-01 3.27641308e-01 -4.09452170e-01
-1.35080898e+00 -1.40790856e+00 2.58706331e-01 4.50295433e-02
5.49495757e-01 -3.17921996e-01 -8.38363469e-01 4.31976706e-01
2.85506640e-02 4.15979803e-01 4.68590409e-01 -2.81155407e-01
-1.14356868e-01 -3.87940079e-01 -5.85789204e-01 3.13866794e-01
1.22591424e+00 -1.15037858e-01 4.81999554e-02 5.01616180e-01
6.58247828e-01 -4.90796745e-01 -4.83346462e-01 4.52218145e-01
5.62570691e-01 -9.93536294e-01 9.64675725e-01 -1.69644237e-01
6.77661002e-02 -1.00825608e+00 -2.89791793e-01 -1.11786866e+00
-6.24174893e-01 -3.89187574e-01 3.61742586e-01 1.16970396e+00
2.63828695e-01 -6.07960343e-01 4.51081991e-01 1.45378038e-01
-1.92953750e-01 -8.83022964e-01 -4.51829970e-01 -5.87614119e-01
-5.26733875e-01 -1.36860803e-01 7.84958422e-01 7.30015218e-01
-5.29240847e-01 2.41398230e-01 -4.33688164e-01 6.83212638e-01
5.83000064e-01 2.97186762e-01 1.02110553e+00 -1.22997177e+00
-3.59181643e-01 -3.22427243e-01 -4.60456789e-01 -1.46461976e+00
-3.99110168e-01 -6.15707815e-01 2.38451034e-01 -1.33891034e+00
3.06476951e-01 -4.99714524e-01 -3.46691549e-01 5.02632022e-01
-4.56944674e-01 4.43122238e-01 3.08668703e-01 1.90153822e-01
-6.51244044e-01 6.17204905e-01 1.24116373e+00 -3.98938358e-01
-1.35972396e-01 3.56530696e-02 -8.98326337e-01 7.19557643e-01
6.35112166e-01 -4.72224683e-01 -3.57724011e-01 -7.44954586e-01
-9.59761962e-02 -3.47935885e-01 5.76707125e-01 -1.24796891e+00
4.68450874e-01 -4.13840637e-02 7.71115422e-01 -6.13205135e-01
1.73567489e-01 -1.00114155e+00 -1.84638728e-03 5.57038844e-01
-9.28507224e-02 -3.07583250e-02 3.44648838e-01 7.77784109e-01
-3.22348952e-01 -3.77460290e-03 9.10228789e-01 -8.23308527e-02
-1.05695283e+00 7.85708129e-01 1.07153445e-01 -2.91441530e-01
1.25860631e+00 -1.24739662e-01 -3.26483041e-01 -3.03414408e-02
-1.98847488e-01 4.62212145e-01 2.28952229e-01 2.57471770e-01
7.16833174e-01 -1.21824920e+00 -6.80653453e-01 3.85637850e-01
3.83078903e-01 1.62523046e-01 4.68042970e-01 9.61503923e-01
-4.38652039e-01 3.76892447e-01 -3.56897563e-01 -7.59939373e-01
-1.35662699e+00 3.69566739e-01 4.14069980e-01 1.20572440e-01
-4.67664361e-01 1.32232332e+00 6.63531125e-01 -1.26023561e-01
3.84067118e-01 -4.35442418e-01 -2.75902987e-01 -2.83925310e-02
7.75404036e-01 1.07461363e-01 -1.45275369e-01 -4.68526840e-01
-4.06011850e-01 8.51790011e-01 -1.63168415e-01 4.22022849e-01
1.09865630e+00 -1.04098782e-01 -6.60404190e-02 -1.17101818e-01
8.53815317e-01 -2.87299544e-01 -1.55096161e+00 -3.57897609e-01
-5.20719051e-01 -9.46707428e-01 3.54041010e-01 -6.43637300e-01
-1.56279361e+00 1.06398571e+00 9.33565497e-01 2.70345937e-02
1.45333576e+00 -3.04599732e-01 5.66199183e-01 5.26554048e-01
4.25700635e-01 -6.90777183e-01 8.14082325e-02 2.21720204e-01
7.41201580e-01 -1.31586325e+00 2.59326279e-01 -7.32633650e-01
-6.13420427e-01 1.14425683e+00 9.41760659e-01 -2.09759519e-01
7.75599658e-01 -1.15075114e-03 4.34692018e-02 -1.65413350e-01
-2.99944222e-01 -3.56573254e-01 2.23753899e-01 4.83242244e-01
1.55624136e-01 1.28677666e-01 -8.92446935e-02 7.92352736e-01
3.47586960e-01 -5.29299416e-02 1.51166156e-01 8.09486508e-01
-6.02471650e-01 -8.06776822e-01 -3.27671707e-01 5.10117948e-01
-2.24040762e-01 -3.00152302e-01 -3.35673809e-01 8.66239488e-01
3.74620825e-01 6.25736117e-01 -7.85967782e-02 -6.37125850e-01
4.14677173e-01 -6.61957800e-01 1.70903608e-01 -3.93109620e-01
-4.83319789e-01 2.76654094e-01 -3.41240555e-01 -4.93231505e-01
-5.45364976e-01 -5.19041717e-01 -1.09683096e+00 4.74019498e-02
-9.48793054e-01 -3.86537053e-02 4.04654086e-01 5.38124442e-01
3.33537340e-01 6.87037408e-01 5.79353392e-01 -9.28197742e-01
-5.31153381e-01 -8.35528076e-01 -6.16837800e-01 3.66098583e-02
1.93252772e-01 -8.95863593e-01 -1.77743360e-01 3.45697403e-02] | [8.831212043762207, -0.8222801089286804] |
29e3dfbb-cc4f-4b92-99ef-3095eb33673a | a-holistic-cascade-system-benchmark-and-human | 2301.10606 | null | https://arxiv.org/abs/2301.10606v1 | https://arxiv.org/pdf/2301.10606v1.pdf | A Holistic Cascade System, benchmark, and Human Evaluation Protocol for Expressive Speech-to-Speech Translation | Expressive speech-to-speech translation (S2ST) aims to transfer prosodic attributes of source speech to target speech while maintaining translation accuracy. Existing research in expressive S2ST is limited, typically focusing on a single expressivity aspect at a time. Likewise, this research area lacks standard evaluation protocols and well-curated benchmark datasets. In this work, we propose a holistic cascade system for expressive S2ST, combining multiple prosody transfer techniques previously considered only in isolation. We curate a benchmark expressivity test set in the TV series domain and explored a second dataset in the audiobook domain. Finally, we present a human evaluation protocol to assess multiple expressive dimensions across speech pairs. Experimental results indicate that bi-lingual annotators can assess the quality of expressive preservation in S2ST systems, and the holistic modeling approach outperforms single-aspect systems. Audio samples can be accessed through our demo webpage: https://facebookresearch.github.io/speech_translation/cascade_expressive_s2st. | ['Peng-Jen Chen', 'Ann Lee', 'Yossi Adi', 'Elizabeth Salesky', 'Hongyu Gong', 'Changhan Wang', 'Justine Kao', 'Benjamin Peloquin', 'Wen-Chin Huang'] | 2023-01-25 | null | null | null | null | ['speech-to-speech-translation'] | ['speech'] | [-3.59433666e-02 2.90116072e-01 -3.56221676e-01 -5.26041865e-01
-1.64331377e+00 -6.35922253e-01 4.17700052e-01 -5.28411448e-01
-2.24718116e-02 4.80379820e-01 6.07689798e-01 -1.11311367e-02
2.13972732e-01 -5.95637262e-02 -3.74856055e-01 -4.09569681e-01
4.36650485e-01 5.96673608e-01 1.42133445e-01 -5.27094364e-01
-4.38531756e-01 -8.05034786e-02 -1.09889233e+00 6.69987857e-01
5.83329439e-01 1.07284832e+00 -5.70796430e-02 8.13247561e-01
1.02915972e-01 8.37420046e-01 -7.36944079e-01 -6.59681022e-01
6.08920567e-02 -6.37311101e-01 -8.37800741e-01 -1.39957771e-01
3.51501644e-01 -3.40170175e-01 4.32666652e-02 9.49786663e-01
1.05974257e+00 4.53111827e-02 2.65816122e-01 -1.47943389e+00
-5.03886521e-01 1.20747852e+00 5.88937337e-03 5.16284928e-02
5.60301840e-01 1.73473790e-01 1.21877468e+00 -9.10930693e-01
5.77130497e-01 1.16710687e+00 6.30266070e-01 7.56387353e-01
-1.38930082e+00 -1.06190968e+00 -1.77625522e-01 -1.29799038e-01
-1.26247931e+00 -1.15232992e+00 1.02467263e+00 -3.05322200e-01
9.29912508e-01 6.48415625e-01 5.37168205e-01 1.81209970e+00
-6.15063548e-01 9.69871461e-01 1.29004908e+00 -2.67833591e-01
-4.23100591e-02 5.02945065e-01 -4.22875360e-02 3.31110030e-01
-7.52396047e-01 2.23686203e-01 -1.20921087e+00 1.44189924e-01
2.49268755e-01 -8.81771982e-01 -2.88844496e-01 3.08760583e-01
-1.26362062e+00 6.22234464e-01 -6.22213446e-02 6.03960335e-01
4.56162915e-02 -1.34656467e-02 9.73375797e-01 7.54319966e-01
7.95592666e-01 2.35236675e-01 -5.98869443e-01 -7.90437102e-01
-1.15534306e+00 2.41987616e-01 1.09813178e+00 1.20243931e+00
1.97830349e-01 2.29171336e-01 -2.88139403e-01 1.22236657e+00
1.19314678e-01 5.95725596e-01 5.61405599e-01 -1.25147676e+00
7.37665057e-01 2.33281568e-01 -2.71510988e-01 -3.42386723e-01
-2.85833716e-01 -3.13802153e-01 -4.77406919e-01 -1.80517033e-01
2.22315103e-01 -2.39179626e-01 -8.90012011e-02 1.94256210e+00
2.17412919e-01 -1.69114411e-01 1.05932146e-01 1.02007139e+00
1.14843678e+00 7.98400879e-01 1.51118383e-01 -4.90942061e-01
1.35434628e+00 -1.34934008e+00 -9.31752384e-01 1.73370287e-01
6.65248334e-01 -1.10597885e+00 1.71490943e+00 4.23207194e-01
-1.43190587e+00 -5.25304496e-01 -8.96545768e-01 -3.86321723e-01
-6.35400936e-02 3.47261518e-01 2.24688962e-01 6.00152612e-01
-1.11639333e+00 2.28084266e-01 -8.15304279e-01 -2.82005280e-01
-1.55983085e-03 2.84534782e-01 -3.70746553e-01 5.81698596e-01
-1.35737884e+00 6.71655774e-01 1.43735528e-01 -3.97971988e-01
-7.70907104e-01 -9.43101287e-01 -5.78197539e-01 -2.74863951e-02
3.82351100e-01 -5.33060133e-01 2.07988524e+00 -1.45320261e+00
-2.10324287e+00 9.49475229e-01 -1.49687752e-02 -3.86797369e-01
7.10739017e-01 -3.07410836e-01 -7.31920779e-01 2.20359862e-01
1.61858082e-01 6.26665950e-01 6.96249366e-01 -1.04400086e+00
-3.57096344e-01 -1.38439864e-01 -4.08844724e-02 3.09290141e-01
-6.36168778e-01 5.61172426e-01 -4.87882316e-01 -8.29306960e-01
-2.94143856e-01 -9.60559964e-01 5.62264085e-01 -1.66634768e-01
-6.07991219e-01 -2.44013667e-01 7.52359211e-01 -7.92377293e-01
1.59874284e+00 -2.23436356e+00 3.47835153e-01 -2.02151909e-01
1.06242515e-01 9.72815901e-02 2.31188256e-02 4.57691312e-01
-7.50342980e-02 1.17192857e-01 -1.03949413e-01 -9.05563474e-01
5.47824144e-01 -1.53445765e-01 -3.22722435e-01 1.89762358e-02
-2.10882276e-02 8.44613016e-01 -6.91595614e-01 -7.14826286e-01
-2.19227560e-02 5.80433846e-01 -5.51458418e-01 2.65129149e-01
-3.17444950e-01 6.85471356e-01 -2.77354658e-01 7.53771365e-01
2.30940893e-01 3.44972797e-02 1.83212444e-01 -2.27357551e-01
-1.39207736e-01 7.90314913e-01 -6.26635909e-01 1.82646310e+00
-9.14869010e-01 7.19804883e-01 1.86207384e-01 -5.16423047e-01
8.71201158e-01 9.04460788e-01 5.98764181e-01 -7.16377378e-01
2.75056541e-01 4.89476413e-01 -7.24048913e-02 -5.31063199e-01
4.93614316e-01 -4.53156978e-01 -3.63275439e-01 4.57736135e-01
3.10779184e-01 -4.47200894e-01 7.72984996e-02 -4.26291451e-02
8.98678303e-01 1.24780230e-01 3.12714785e-01 -1.27086550e-01
3.91890705e-01 -1.78591266e-01 5.68165481e-01 6.89340159e-02
-4.79669005e-01 6.67151749e-01 5.28414071e-01 8.62865448e-02
-1.04737568e+00 -1.16467941e+00 -8.42046738e-02 1.76301527e+00
-3.52273375e-01 -8.02004874e-01 -1.20341206e+00 -4.45598334e-01
-4.36093479e-01 8.46006036e-01 -4.03696418e-01 1.72969326e-01
-5.59926867e-01 -1.93186417e-01 1.24237347e+00 4.73573893e-01
4.89674479e-01 -1.08094931e+00 -1.49000555e-01 1.81117579e-01
-8.98843110e-01 -1.41921842e+00 -9.03880835e-01 2.68365536e-02
-4.06020969e-01 -5.62098384e-01 -6.36070907e-01 -6.83979630e-01
-2.14435175e-01 -4.15874213e-01 1.31612444e+00 -3.28899086e-01
3.46103489e-01 2.78259426e-01 -5.09690225e-01 -3.45440418e-01
-1.06882298e+00 4.78472441e-01 2.50193447e-01 -1.55932412e-01
5.07537723e-01 -8.35201442e-01 -2.72745162e-01 5.59638083e-01
-4.15611625e-01 2.59008557e-01 2.79559970e-01 6.88575447e-01
4.67119157e-01 -5.17165303e-01 8.49075913e-01 -5.41172743e-01
8.69941771e-01 -3.90379399e-01 -3.47509116e-01 -6.12808540e-02
-4.62022692e-01 -5.08836329e-01 5.21623671e-01 -7.06143260e-01
-1.10221326e+00 -2.77747959e-01 -6.24986708e-01 -5.27467906e-01
-1.92565039e-01 2.38609999e-01 -5.96755028e-01 4.23963785e-01
6.82206631e-01 1.02706745e-01 -1.26830280e-01 -5.35730183e-01
2.78528810e-01 1.15728164e+00 5.21903574e-01 -8.41798842e-01
4.38295126e-01 6.18854165e-02 -6.58024371e-01 -9.45035219e-01
-8.49692702e-01 -2.74684936e-01 -3.91705513e-01 -4.07510787e-01
7.22528100e-01 -1.12876439e+00 -7.00311959e-01 2.66187161e-01
-1.00487232e+00 -6.53784931e-01 -4.45909023e-01 5.40008485e-01
-1.10307562e+00 -7.74778277e-02 -8.19726050e-01 -6.38456821e-01
-6.40515268e-01 -1.27185178e+00 1.41486812e+00 -2.60128140e-01
-8.14519465e-01 -8.06463599e-01 3.68269831e-01 9.63801205e-01
3.87130737e-01 2.04236675e-02 4.92415518e-01 -8.74752998e-01
-4.43414263e-02 1.53444847e-02 2.48333529e-01 5.98072767e-01
-1.10816903e-01 7.21519738e-02 -1.48092735e+00 3.76194827e-02
-1.22192334e-02 -6.98795497e-01 3.49338353e-01 1.53450012e-01
7.20070243e-01 -6.05296612e-01 2.12015137e-01 4.48096037e-01
5.66331923e-01 2.54576132e-02 3.39937985e-01 9.06636715e-02
4.13275808e-01 5.75263619e-01 6.18305147e-01 3.84692013e-01
5.07566273e-01 1.29648209e+00 -2.47584715e-01 -3.57272029e-02
-5.47081769e-01 -4.23509806e-01 9.90052879e-01 1.63974178e+00
-1.09642275e-01 -3.21886837e-01 -8.30345452e-01 4.30266678e-01
-1.59878719e+00 -9.16253805e-01 2.34833017e-01 1.75007832e+00
1.30844915e+00 6.25071600e-02 5.89521170e-01 3.15108448e-01
4.87995088e-01 8.13764334e-02 -2.95296818e-01 -5.82450926e-01
-2.30622187e-01 -3.51556912e-02 -1.61196426e-01 6.03466630e-01
-8.74357939e-01 1.18031859e+00 5.33024693e+00 1.27851546e+00
-1.34544182e+00 6.98900700e-01 6.41871572e-01 -6.93276823e-01
-3.67599428e-01 -2.04643279e-01 -7.64851451e-01 4.45147336e-01
1.40846455e+00 -2.94953763e-01 5.03175378e-01 8.04713249e-01
6.46251202e-01 5.68950117e-01 -1.22197735e+00 1.06286204e+00
-3.09458207e-02 -9.43143189e-01 -2.38579646e-01 -4.63593528e-02
5.70948124e-01 2.33721703e-01 3.51168096e-01 5.61628699e-01
-6.48831427e-02 -7.40174651e-01 1.14551640e+00 2.94461101e-01
1.15711117e+00 -5.41974127e-01 3.65131438e-01 2.12526903e-01
-1.19238067e+00 9.47861001e-02 3.90907288e-01 3.26051444e-01
3.94425780e-01 1.41120881e-01 -8.23222697e-01 3.14518005e-01
8.92705262e-01 5.51692843e-01 -8.35760310e-02 1.81473434e-01
-2.96110660e-01 1.08846891e+00 -4.01978135e-01 2.00920110e-03
6.54137582e-02 -7.86479041e-02 1.07351232e+00 1.53921854e+00
3.10792476e-01 4.78594229e-02 6.41471744e-02 9.02962387e-01
-4.05878991e-01 5.57077229e-01 -5.50591052e-01 -3.46568763e-01
5.18789828e-01 1.35059237e+00 -1.82983622e-01 -3.23743105e-01
-3.43297839e-01 7.99527764e-01 1.70830771e-01 1.80838972e-01
-9.79300797e-01 1.36988446e-01 6.99914813e-01 2.82179832e-01
-1.42140865e-01 -4.40777652e-02 -2.35661194e-01 -1.12700152e+00
1.71268389e-01 -1.34033287e+00 1.61919802e-01 -8.98333013e-01
-1.12812650e+00 1.01668525e+00 8.54080915e-02 -1.45108902e+00
-7.38855422e-01 -1.58536717e-01 -3.84007037e-01 5.93471229e-01
-1.04019034e+00 -1.60061705e+00 -5.19013628e-02 6.52186155e-01
9.97902811e-01 -3.97586942e-01 9.96734738e-01 6.55391753e-01
-4.67603475e-01 1.03382766e+00 -3.24150860e-01 1.44221336e-01
1.03246832e+00 -1.04899895e+00 1.37441978e-01 4.04886633e-01
2.25321412e-01 3.95558387e-01 8.69545102e-01 -2.67288327e-01
-9.61055279e-01 -8.56585205e-01 1.03434408e+00 -4.95179117e-01
1.09558523e+00 -6.17140472e-01 -7.87882805e-01 7.61669934e-01
5.22535026e-01 -1.89302683e-01 1.03659606e+00 3.07034165e-01
-4.34521347e-01 -1.58708557e-01 -9.57609892e-01 6.38141692e-01
1.04122686e+00 -8.69415283e-01 -3.04988265e-01 2.11444706e-01
1.18891644e+00 -4.61223781e-01 -1.45496619e+00 2.60456234e-01
7.55572200e-01 -8.92492950e-01 6.83956623e-01 -4.51179385e-01
5.36597788e-01 -5.92909306e-02 -6.47769153e-01 -1.28490865e+00
3.15283209e-01 -1.07115841e+00 -1.40988424e-01 1.76868081e+00
7.97254920e-01 -3.89240324e-01 2.77592540e-01 2.70972103e-01
-5.24715543e-01 -5.79225421e-01 -1.17501545e+00 -9.43116963e-01
1.39609575e-01 -7.26822257e-01 5.14234841e-01 9.49500978e-01
3.98137689e-01 8.64823759e-01 -5.27433038e-01 -1.32723913e-01
2.36704797e-01 6.64682165e-02 9.28584218e-01 -7.45831013e-01
-5.70165753e-01 -6.75392926e-01 -1.40491232e-01 -6.63565516e-01
4.28370386e-01 -1.11410213e+00 -6.21865056e-02 -7.38633454e-01
4.54254150e-02 -1.69647083e-01 9.09645855e-03 6.00130796e-01
2.50199169e-01 5.28837204e-01 3.40908438e-01 1.52705997e-01
-5.33276200e-01 8.74838889e-01 1.12248409e+00 -1.55824170e-01
-5.08028209e-01 1.34114146e-01 -6.73210919e-01 7.38624930e-01
1.18211329e+00 -3.77195805e-01 -5.66454589e-01 -2.94798255e-01
4.55047376e-02 3.46708924e-01 1.22767664e-01 -9.44050908e-01
5.84705062e-02 1.93389922e-01 -3.29164386e-01 -2.92312413e-01
8.37354600e-01 -6.35535419e-01 2.28153899e-01 -1.74577266e-01
-5.73887944e-01 -5.86268194e-02 3.30480993e-01 -9.84070152e-02
-5.56687117e-01 1.05442412e-01 7.68941104e-01 2.41091073e-01
-2.37484246e-01 1.63877457e-01 -1.76658958e-01 4.35281068e-01
6.14694774e-01 1.26736820e-01 -2.76824743e-01 -7.10884333e-01
-1.06391251e+00 -5.82688190e-02 1.88164011e-01 6.27948642e-01
3.75555396e-01 -1.51456821e+00 -9.70296800e-01 -1.44104391e-01
4.60463881e-01 -5.65562248e-01 1.48783982e-01 1.25043190e+00
-1.13340341e-01 2.75429785e-01 -4.28322367e-02 -5.62417150e-01
-1.61185861e+00 1.93686709e-01 4.75894094e-01 -1.26184449e-01
-6.12003028e-01 6.98245704e-01 1.41409427e-01 -8.66713464e-01
2.14824706e-01 -3.31426382e-01 2.08709359e-01 1.52984366e-01
3.73915404e-01 4.61314321e-01 9.26922560e-02 -1.01597750e+00
-2.76635855e-01 3.67580086e-01 1.94999218e-01 -7.81916440e-01
1.25337696e+00 -3.78002226e-01 1.89864367e-01 9.48589206e-01
1.38030767e+00 2.36895531e-01 -9.27698195e-01 -2.63764918e-01
-1.10095948e-01 5.29776476e-02 3.09483688e-02 -1.00241697e+00
-9.30473089e-01 9.28152323e-01 3.32929671e-01 1.93566471e-01
1.36205876e+00 2.69684464e-01 9.54110146e-01 2.68239856e-01
-7.29573444e-02 -1.50281155e+00 1.15064003e-01 4.79438126e-01
1.40166020e+00 -1.21708989e+00 -5.01823008e-01 -4.13906425e-01
-1.18003738e+00 9.28877473e-01 5.05898237e-01 4.37010914e-01
6.50144517e-01 7.30089188e-01 4.96189117e-01 -1.47657981e-02
-1.10073316e+00 9.24237445e-03 3.29030901e-01 3.45182359e-01
8.74253035e-01 4.15846467e-01 -9.86376256e-02 1.08722794e+00
-1.10294771e+00 1.11150555e-02 4.93394881e-02 3.34409088e-01
-9.90513787e-02 -1.08071280e+00 -1.62875831e-01 -5.60050495e-02
-6.35187209e-01 -2.46013314e-01 -6.15024507e-01 8.69827271e-01
-1.48710564e-01 1.01933384e+00 -1.76450223e-01 -6.19930685e-01
5.72547793e-01 2.85576403e-01 2.73747802e-01 -2.42693663e-01
-9.06708956e-01 6.43848121e-01 6.35413349e-01 -7.23613501e-01
-4.92440730e-01 -7.50269413e-01 -1.12528229e+00 -3.90366733e-01
2.93265022e-02 2.01602638e-01 8.40190828e-01 6.89732432e-01
1.41879559e-01 6.45056307e-01 6.50193691e-01 -6.83808744e-01
-5.51759124e-01 -1.32827115e+00 -4.91950214e-01 2.25480691e-01
3.09979677e-01 -3.62665325e-01 -4.32505876e-01 3.28407645e-01] | [14.72860336303711, 6.814136981964111] |
7738328f-7f4d-41d2-af3c-07fd8328728b | a-collaborative-ranking-model-with-multiple | 1807.04210 | null | http://arxiv.org/abs/1807.04210v2 | http://arxiv.org/pdf/1807.04210v2.pdf | A Collaborative Ranking Model with Multiple Location-based Similarities for Venue Suggestion | Recommending venues plays a critical rule in satisfying users' needs on
location-based social networks. Recent studies have explored the idea of
adopting collaborative ranking (CR) for recommendation, combining the idea of
learning to rank and collaborative filtering. However, CR suffers from the
sparsity problem, mainly because it associates similar users based on exact
matching of the venues in their check-in history. Even though research in
collaborative filtering has shown that considering auxiliary information such
as geographical influence, helps the model to alleviate the sparsity problem,
the same direction still needs to be explored in CR. In this work, we present a
CR framework that focuses on the top of the ranked list while integrating an
arbitrary number of similarity functions between venues as it learns the
model's parameters. We further introduce three example similarity measures
based on venues' contents and locations. Incorporating cross-venue similarity
measures into the model enhances the latent associations between users as
similar venues are also taken into account while associating users with each
other. Our experiments on the TREC Contextual Suggestion dataset show that our
proposed CR model beats other state-of-the-art venue suggestion methods. | ['Crestani Fabio', 'Rafailidis Dimitrios', 'Aliannejadi Mohammad'] | 2018-07-13 | null | null | null | null | ['collaborative-ranking'] | ['graphs'] | [-0.34413484 -0.13673797 -0.4768377 -0.62631994 -0.323619 -0.41484648
0.8635267 0.32908255 -0.45075557 0.42930728 0.87175804 -0.20689593
-0.9305001 -0.9102382 -0.3096905 -0.5277272 -0.46690547 0.4567289
0.5819189 -0.4346539 0.8446847 0.05357393 -1.5477307 0.42048007
1.2166778 0.5951441 0.5097179 0.1373525 -0.36420342 0.4801996
-0.23374133 -0.33127558 0.46919525 -0.14905262 -0.49377248 -0.29246646
0.48969755 -0.24390115 -0.56315386 0.7438883 0.404698 0.8639648
0.7744463 -0.90189666 -0.91118413 1.0624541 -0.43354252 0.2215368
0.3594403 -0.7573527 1.404975 -0.8073385 0.29637113 1.1458105
0.5134426 0.11421196 -0.8438576 -0.5151966 0.70330733 0.21175171
-1.4750974 0.08638114 0.6741759 -0.1518158 0.59471846 0.5554879
0.5405253 0.39324495 -0.17205492 0.6321999 0.8484811 -0.17332174
0.10870871 0.24911965 0.20849887 0.2711464 0.1315252 -0.10642341
-0.651725 -0.6591005 0.68838817 0.816103 -0.12707788 -0.30902037
-0.8820002 0.83386075 0.78022027 0.70409197 -0.29210004 -0.15270974
0.03819716 0.2694962 0.73017305 0.6117314 -0.30201912 0.35200796
-0.9704512 0.07035177 0.70159215 0.8362375 0.862341 -0.5352103
-0.38479108 1.146875 0.7093986 0.09309127 0.25098476 -0.87385637
0.3375194 0.51760983 0.2948846 -1.3762482 -0.18732022 -1.0012002
-0.7634547 -0.5851738 0.33761713 -0.01917243 -0.49643615 1.470949
0.2500871 0.8217843 -0.45937985 1.0602113 0.87323296 0.55171627
0.11012053 -0.11905207 1.0772742 -1.0725119 -0.53033406 0.00596385
0.47964913 -0.86701626 0.6426715 0.27605733 -0.80001473 -0.6415938
-0.5255544 0.19321255 -0.49502024 0.02223817 1.1898806 0.58675283
-1.3986094 0.77028483 -0.2841418 -0.5928865 -0.02523185 0.30189186
0.23963013 -0.21591045 -1.498514 0.62084323 0.03554273 -0.10798945
-0.39384073 -0.5976314 -0.29698846 0.46537554 0.58554685 -0.7667455
0.6630795 -0.63391566 -1.0161613 0.0597263 -0.4019311 -0.1371595
0.18269953 -0.2871241 -0.8335419 -0.17132826 0.14643644 0.38458622
0.42906132 -1.3419956 -1.116258 -0.15735337 0.47370657 0.60928
-0.72512937 -0.03624706 -1.087072 -0.5309092 0.57331777 -0.8242548
-0.7241112 -0.5107955 0.01648438 -0.60237104 0.3859692 -0.26325983
1.7994733 -1.7816083 -0.1694404 1.0442996 0.26562792 -0.09017987
-0.14176224 0.8653608 0.5988411 0.15412807 0.4912763 -0.25322643
-0.04789401 0.100587 -0.20621826 0.21026696 -0.7203894 0.7602549
-1.2960205 -0.41392976 0.04594362 0.51227623 -0.8423116 -0.10138073
0.13086805 0.57536066 -0.8414043 0.37673536 0.64833343 -0.5012284
0.5625365 -0.00952196 -0.18772353 0.5636217 -1.2250696 1.7247255
-0.35635287 0.08920386 -0.10381547 -0.83894664 1.0210129 0.11448871
0.7532388 -0.6275385 -0.38837618 0.262477 -0.04123751 -0.2251177
1.0386806 0.42829356 0.02632261 0.77652043 -0.26915225 0.52662396
0.21278292 0.9513789 0.9240814 -0.06801449 0.15447241 -0.40390533
0.6063723 -0.38502818 0.37055108 1.426268 0.23381183 0.33869988
-0.11357824 -0.15413535 -0.74660134 -0.86925834 -0.02106336 1.5077624
0.43027624 -0.85406363 -0.09945364 -0.89816874 0.3013092 0.47440556
-0.6439781 0.15386206 -0.34248826 -0.6257732 -0.12962769 0.4133726
0.23794393 -0.48320195 0.54986346 0.16970131 -0.4020393 -0.48713943
-0.8434117 -0.11255675 -1.0160239 -0.7506157 -0.8609157 -0.63891226
0.9026 1.247244 1.0822369 0.6283062 0.6167835 0.5233959
-0.79242086 0.20118779 0.49931204 0.1708265 0.16580003 0.13794221
0.6036328 -0.9075667 -1.0934049 1.1557868 -0.6569502 -0.29501012
0.44075242 0.60074985 0.2584996 0.34785992 0.9836311 -1.3666587
0.6858906 -1.0550174 -0.15143049 0.3619083 -1.0174664 -0.20247874
0.4846478 -0.38437402 -1.2288802 -0.3152035 -0.05605844 0.20526932
0.21251503 0.90170723 0.0774613 -0.0632276 0.49446678 0.02081933
-0.6883462 -0.7449661 0.63474995 0.72918594 0.04225679 -0.49744752
0.7500578 0.5239407 -0.22876751 -0.4669001 -1.0836966 -1.5131147
-0.48189947 -0.32541963 0.09015705 -1.0322393 -0.6888794 -0.20888962
-0.59363896 0.22651252 0.1904014 0.81649655 0.12952666 0.70096666
-0.40263945 -0.9550304 0.07040478 -0.4859047 0.58436817 0.29371068
0.03878172 -1.2467599 0.1642353 0.56298786 0.7642478 -0.6273541
0.5454761 -1.0928563 -0.8751715 -0.23132975 -0.32739028 -0.26806995
0.29524004 -0.5248518 -0.675338 -0.31918547 -0.3181161 0.3962463
1.2209309 0.5011391 1.1346617 -0.30450326 -0.65662646 0.1871018
1.2773156 0.08995106 0.7157821 0.15707506 0.6783079 0.8094165
1.1107088 0.6398005 0.6809887 0.7695294 0.35096174 -0.05616764
0.31147468 -0.6375889 0.01283163 0.81555295 -0.49593002 -0.56938463
-0.34578386 0.4624944 -2.3046203 -1.1292201 -0.3211871 2.5796988
0.31542084 -0.03178564 0.2882772 -0.10745995 0.88515645 0.14646323
-0.18594894 0.09294666 -0.00776452 -0.21172181 0.65311867 0.65420324
-0.9332267 0.86530226 5.366245 1.0572215 -0.6126869 0.14091353
0.5454163 0.02639153 -0.8050249 0.30250832 -1.0453767 0.53224105
0.6908624 -0.04604916 0.44411403 0.6608754 0.5306685 -0.3464168
-0.73984766 0.556388 0.28538305 -1.2953175 0.01745562 0.1822007
1.1962684 -0.00696041 0.05925606 0.4330162 0.5383993 -0.5791444
0.15652996 0.84270215 0.06917348 -0.6802659 0.80706674 0.5004625
-1.4706392 -0.38566017 -0.7260122 -0.15010518 0.30807993 0.8054257
-0.8429888 0.8810268 0.7829583 1.164882 -0.5493015 1.5644529
-0.10966366 0.6967037 -0.22826535 -0.1979503 0.39408737 -0.737884
0.24590673 0.95411634 0.7515718 0.30449495 0.42008594 0.22795884
0.09135566 0.6186926 -0.38619936 0.2905845 0.7664636 1.3154578
-0.8410121 -0.34559196 -0.6707285 0.63045996 0.10834008 0.45680887
-0.53051126 -0.0930342 0.22568798 0.5745422 0.29733875 -0.29610646
-0.27190426 -0.9653174 -0.3552996 -0.31215286 0.61740154 -0.6226697
-1.398552 0.1827161 -0.10939968 -1.5367103 0.1219808 0.04771462
-0.79133946 0.8682161 -1.609629 -1.0434239 -0.17234044 0.6302999
0.31498432 -0.12010597 0.5046248 0.82496107 -0.03112311 0.54554695
0.5737929 -0.3543953 0.9585165 -1.1057786 0.08102744 0.60657114
0.4108683 1.4524534 0.4508663 -1.0631747 -1.0619899 -0.9674129
1.5604124 -0.38938674 0.6776623 -0.04973081 -0.79144585 0.3712283
0.06862747 -0.36435026 1.0909826 0.9724386 -0.27519962 -0.18854748
-0.92218107 0.5822842 1.4847484 -0.38785127 -0.40993398 0.4142876
0.37371707 0.3410972 -0.8443415 0.09656425 0.68600005 -0.8815347
1.3922246 -0.29362738 0.1518965 -0.5580576 -0.08998604 -1.3625731
-0.79157007 -0.41638887 -0.070815 1.3438288 0.5690159 -0.58442914
0.97145665 0.6490932 -0.18698046 -0.5665279 -0.39690158 -0.5471022
-0.41424626 -0.07505237 0.5700316 1.1433915 0.36728722 0.38161132
-0.9195618 0.19030559 0.33561411 0.1635319 0.6724552 -1.6537281
-0.48182744 -0.20381273 0.19143197 -1.822021 -0.33193028 -1.0762551
-0.3094285 -1.8015337 0.49110767 -1.2718192 -0.85517275 0.23224337
-0.3627616 0.25598162 -0.02864911 0.6797054 -1.2774451 0.28772986
1.2514316 0.11098452 -0.52646184 0.6043467 -1.2069852 0.47312388
0.55668277 -0.5027244 -0.71684414 -0.28407198 0.8435698 0.15161674
-0.16765796 -0.51070887 0.87578386 -0.2378337 0.1896834 -0.95916903
0.41292295 -0.98328924 0.23986748 0.02394494 -0.82975847 -0.24131352
-0.5880007 1.1876556 -0.15009995 -0.29233545 0.09331229 -0.16837452
-0.69357127 0.33439288 -0.38890386 -0.47940964 0.38790873 -0.12050579
-0.21645325 -0.81522226 -0.81730103 0.6276523 0.34724754 0.43503156
0.58892286 -1.301919 -0.61814934 -0.20376551 0.24824712 -0.5586432
0.5729541 0.94527394 0.3532161 0.69939786 0.24763213 -0.24946636
-1.2891501 0.4090965 -0.2553774 -0.52171445 -0.18158671 0.8430819
0.1121292 -0.4414006 0.4445101 0.27408552 -1.0301107 0.24635364
0.46811077 0.51076895 -0.077975 -0.39037648 -0.12031133 0.4138122
-0.4264773 0.09813 1.30564 -0.72937196 -0.04773799 0.02178787
0.6478036 0.5803427 -0.718109 -0.87335676 0.23987354 -0.9556001
0.17695072 -0.8658363 -1.1534894 0.22059819 0.28423283 0.34606415
0.8135355 -0.03330157 0.7235092 0.5018351 0.76425874 -1.1566948
0.07785142 0.5967599 0.34402677 -1.1594216 0.30108148 -0.70225054
-0.7024556 0.6523298 0.40966073 -0.1448726 1.126063 -0.5789791
-0.25454465 -0.10204343 -0.6375725 -0.47028396 0.76573026 0.21495631
0.83554196 0.04807147 -1.184252 0.6427692 0.1647128 -0.2154309
0.24511057 0.7383452 -0.79812264 -1.5165101 -0.16099903 0.9675685
-0.44449502 -0.4910898 -0.15332772 0.18316986 0.16120337 1.3151777
-0.08513168 -0.64779 -0.17164645 -0.31317014 -0.04384691 -0.8556269
-1.0535785 0.50209934 0.33679536 -0.37955853 -0.6728673 -0.68971735
-0.8753696 -0.40042347 -0.8166713 0.993525 0.64611644 0.7725175
0.48746145 0.16584824 1.0086535 -0.67123085 -0.23355827 -0.91004956
-0.98427004 0.36033618 -0.22784562 -0.80757606 -0.5313516 -0.6072888 ] | [10.04401683807373, 5.6491923332214355] |
1452004c-78e1-4711-9959-b3d0bb6f475b | segmenting-unseen-industrial-components-in-a | 2002.03501 | null | https://arxiv.org/abs/2002.03501v3 | https://arxiv.org/pdf/2002.03501v3.pdf | Segmenting Unseen Industrial Components in a Heavy Clutter Using RGB-D Fusion and Synthetic Data | Segmentation of unseen industrial parts is essential for autonomous industrial systems. However, industrial components are texture-less, reflective, and often found in cluttered and unstructured environments with heavy occlusion, which makes it more challenging to deal with unseen objects. To tackle this problem, we present a synthetic data generation pipeline that randomizes textures via domain randomization to focus on the shape information. In addition, we propose an RGB-D Fusion Mask R-CNN with a confidence map estimator, which exploits reliable depth information in multiple feature levels. We transferred the trained model to real-world scenarios and evaluated its performance by making comparisons with baselines and ablation studies. We demonstrate that our methods, which use only synthetic data, could be effective solutions for unseen industrial components segmentation. | ['Kyoobin Lee', 'Seungjun Choi', 'Seunghyeok Back', 'Jongwon Kim', 'Raeyoung Kang'] | 2020-02-10 | null | null | null | null | ['unseen-object-instance-segmentation'] | ['computer-vision'] | [ 5.51946282e-01 3.08011800e-01 3.04344654e-01 -2.58523673e-01
-7.05031574e-01 -9.21015263e-01 2.72926033e-01 -2.13460803e-01
7.55384937e-02 5.56383491e-01 -3.68744105e-01 -4.70131561e-02
1.85427666e-01 -7.81667054e-01 -8.61343086e-01 -7.28411853e-01
3.89165938e-01 5.75912595e-01 5.00468075e-01 2.16907766e-02
5.52091971e-02 6.67933881e-01 -1.83365679e+00 3.12676966e-01
7.81122804e-01 1.10348749e+00 4.49547946e-01 2.96085477e-01
1.74376532e-01 3.14186037e-01 -9.18590784e-01 -2.39692986e-01
6.76183999e-01 -1.46948904e-01 -3.69482368e-01 5.92878580e-01
5.64688325e-01 -4.62116003e-01 -2.38675669e-01 1.04877555e+00
2.74491340e-01 7.96451792e-02 4.87633258e-01 -1.17295098e+00
-3.58537704e-01 2.25062042e-01 -6.17050648e-01 -5.11525691e-01
2.50252217e-01 5.52748203e-01 4.20279264e-01 -5.37499785e-01
7.77477324e-01 1.26517797e+00 4.95934725e-01 6.68314815e-01
-1.25567949e+00 -6.26510799e-01 4.77269202e-01 -2.89806008e-01
-1.08678579e+00 -2.41692573e-01 8.99472117e-01 -2.88236260e-01
4.54425454e-01 1.28421605e-01 6.78986907e-01 1.56609583e+00
1.32598996e-01 9.36557114e-01 1.04532588e+00 -1.87929198e-01
4.21605498e-01 1.18762806e-01 -2.70055801e-01 5.66778004e-01
6.04076147e-01 3.73690754e-01 -2.50380218e-01 2.94940889e-01
1.07013190e+00 2.46323925e-02 -1.44216955e-01 -8.57300282e-01
-1.44763839e+00 5.94120383e-01 4.25850600e-01 -2.17828631e-01
-2.43133351e-01 7.59293735e-02 5.87577932e-02 3.43406200e-02
4.29624349e-01 6.91681981e-01 -6.81117833e-01 -4.30274718e-02
-5.91739118e-01 2.42179617e-01 8.68432701e-01 1.35535049e+00
6.99914932e-01 -3.03239631e-03 -8.08324218e-02 7.29848266e-01
3.95130634e-01 6.54158831e-01 7.55196065e-02 -1.27947342e+00
3.26322764e-01 5.59346676e-01 1.43528908e-01 -5.29587030e-01
-2.17250556e-01 -2.48109639e-01 -5.65723777e-01 8.20045769e-01
4.98807251e-01 -3.31758469e-01 -1.42144561e+00 1.26208770e+00
5.30852556e-01 -1.53075578e-02 9.98255014e-02 1.14552474e+00
6.89521074e-01 1.14753127e-01 -2.26468667e-01 1.28743380e-01
1.16582930e+00 -9.78192389e-01 -5.91965497e-01 -3.46037000e-01
1.25921294e-01 -1.03110194e+00 1.16743731e+00 7.77548671e-01
-7.72629440e-01 -5.94824493e-01 -1.17422938e+00 1.28715396e-01
-3.05082083e-01 1.63797066e-01 6.99044883e-01 7.33234346e-01
-4.15457934e-01 7.48132110e-01 -1.07530904e+00 -2.80122429e-01
6.07388318e-01 2.17142671e-01 -1.46220192e-01 -2.13366047e-01
-6.81633413e-01 4.63669509e-01 4.88335490e-01 3.78776580e-01
-1.33483505e+00 -4.21467513e-01 -9.72664177e-01 -5.17216444e-01
8.39956880e-01 -5.01350880e-01 1.22157228e+00 -4.94424999e-01
-1.87714696e+00 6.30172849e-01 9.19744745e-02 -1.48386389e-01
7.44784296e-01 -5.36899686e-01 -1.09835118e-01 2.26550233e-02
-1.41404979e-02 7.44441032e-01 1.03613818e+00 -1.69213414e+00
-5.33558130e-01 -3.41799200e-01 1.26990527e-01 -7.37810805e-02
3.07486445e-01 -6.57450974e-01 -4.06674773e-01 -7.23736227e-01
4.86855984e-01 -9.99424100e-01 -5.62936366e-01 1.95330121e-02
-8.19313824e-01 2.54170358e-01 1.22841847e+00 -3.99377167e-01
2.11226627e-01 -2.44540644e+00 -2.57810205e-03 1.60777763e-01
1.29673615e-01 -2.73749884e-03 -1.24161996e-01 -8.53978842e-02
1.79397672e-01 -9.46884677e-02 -3.15038800e-01 -1.95886180e-01
1.73502281e-01 2.59834349e-01 -2.71191537e-01 3.06086749e-01
5.75549126e-01 8.03067505e-01 -8.87851179e-01 -3.71001959e-01
8.48983169e-01 4.08592612e-01 -3.75153959e-01 2.09803686e-01
-7.51887918e-01 7.73216605e-01 -6.96450889e-01 1.30225921e+00
8.57281089e-01 3.34763117e-02 9.04449224e-02 -5.29760480e-01
1.28204122e-01 -2.00622737e-01 -1.32894504e+00 1.92577493e+00
-4.71274585e-01 4.72394168e-01 3.37040842e-01 -4.64313895e-01
8.97995710e-01 9.04802904e-02 3.95014912e-01 -5.51851928e-01
2.63277531e-01 1.04285449e-01 -3.01463544e-01 -3.78402025e-01
4.26917821e-01 9.25342292e-02 -1.60424128e-01 -1.23050489e-01
-3.35743725e-01 -1.02909410e+00 1.19500145e-01 -1.11684375e-01
1.28555703e+00 7.69590378e-01 -3.04980040e-01 1.18738420e-01
-7.14752749e-02 1.24150112e-01 7.55292594e-01 6.46920741e-01
-4.47694473e-02 1.22280288e+00 3.56333315e-01 -4.96055990e-01
-8.87428582e-01 -1.52174675e+00 -1.40148208e-01 3.02194029e-01
7.72386014e-01 -1.60968706e-01 -8.34133506e-01 -9.90981936e-01
2.43427232e-01 8.18369210e-01 -4.27350134e-01 -2.05122218e-01
-4.42615420e-01 -5.49711645e-01 -5.64344265e-02 4.89713788e-01
5.96707404e-01 -1.13619792e+00 -8.60555470e-01 3.74610037e-01
6.24873228e-02 -1.58289957e+00 1.19072478e-02 5.84472537e-01
-8.66752028e-01 -1.44172049e+00 -4.96851921e-01 -5.63288987e-01
6.76928639e-01 2.37725884e-01 1.39843035e+00 -3.59219998e-01
-7.53777564e-01 3.09520185e-01 -6.16006136e-01 -7.28704393e-01
-4.06104505e-01 1.25913456e-01 -2.63515025e-01 -2.57685214e-01
1.19457148e-01 -2.76446313e-01 -7.79439211e-01 7.16083109e-01
-9.93707240e-01 1.36627313e-02 8.84855807e-01 6.23567581e-01
8.61020982e-01 2.04076022e-01 2.24294186e-01 -8.78950119e-01
3.00966829e-01 -5.75908944e-02 -8.52630198e-01 1.21117704e-01
-2.86987871e-01 5.79847908e-03 3.01367521e-01 -6.30874038e-01
-1.08446491e+00 6.26996517e-01 2.20677480e-02 -6.97747886e-01
-6.07007623e-01 -3.27384293e-01 -6.29797399e-01 7.79365599e-02
5.25019109e-01 -3.23313862e-01 -5.53213954e-02 -5.59135020e-01
5.00000417e-01 5.83505511e-01 5.23298681e-01 -8.78599584e-01
9.63900805e-01 6.67698383e-01 -7.44512752e-02 -9.39596772e-01
-7.65659094e-01 -5.97050525e-02 -6.66419446e-01 -2.56760180e-01
9.84434068e-01 -7.29742765e-01 -5.71867704e-01 4.92067814e-01
-1.09112060e+00 -6.43755317e-01 -8.44939291e-01 4.80836153e-01
-6.04495645e-01 3.04935426e-01 -5.54285049e-01 -7.65481353e-01
1.82189614e-01 -1.56304014e+00 1.75408971e+00 1.77357793e-01
-9.88280177e-02 -4.52576041e-01 -4.67026055e-01 5.22724986e-01
1.10693842e-01 8.55090737e-01 6.49991274e-01 -2.96786517e-01
-1.11150730e+00 -2.07167834e-01 -2.44626775e-01 5.92775702e-01
4.81617808e-01 4.41035666e-02 -1.20411766e+00 -1.88151345e-01
-7.18899369e-02 -5.06226659e-01 8.99363339e-01 3.08531225e-01
1.43996811e+00 3.18091303e-01 -5.02770483e-01 5.69286883e-01
9.73405302e-01 1.97656795e-01 7.44917154e-01 2.90458143e-01
8.90399873e-01 6.30596578e-01 1.07549489e+00 4.18891460e-01
-1.88793361e-01 5.41627407e-01 9.59275961e-01 -4.47528601e-01
-3.84019047e-01 1.15632033e-02 1.73633844e-01 2.71610051e-01
1.99353788e-02 -4.51452672e-01 -4.41166222e-01 3.56741309e-01
-1.66856945e+00 -2.85955697e-01 -6.99903443e-02 2.12141562e+00
6.19181752e-01 4.69711810e-01 -2.85404712e-01 -9.40707251e-02
5.98141253e-01 -9.01877433e-02 -9.08563018e-01 8.61634761e-02
-2.53230691e-01 3.37273300e-01 6.69790387e-01 7.96713904e-02
-1.23794329e+00 1.03881919e+00 6.40463352e+00 5.09615839e-01
-8.66054773e-01 -2.25562930e-01 4.77833837e-01 -1.22649588e-01
-2.00761557e-01 -2.01804236e-01 -6.24465227e-01 4.21092600e-01
4.01230961e-01 5.34403384e-01 3.90935749e-01 1.06284869e+00
-6.61321729e-02 -2.09225938e-01 -1.21918535e+00 8.75840247e-01
-1.56964123e-01 -8.11470628e-01 -1.05408072e-01 2.07587659e-01
9.17412102e-01 1.75022092e-02 1.41793638e-01 4.97098714e-02
9.29508626e-01 -8.55181634e-01 8.30691755e-01 2.55211562e-01
6.38910472e-01 -5.29698849e-01 7.72245526e-01 -2.58279052e-02
-1.02210820e+00 9.98278409e-02 -2.75972337e-01 3.97076517e-01
2.59024471e-01 1.02859521e+00 -9.52393889e-01 6.09265089e-01
7.47028053e-01 5.87257624e-01 -4.97803777e-01 8.53570819e-01
-4.46098715e-01 3.05729866e-01 -5.12258530e-01 3.01799923e-01
-9.85205099e-02 -1.51758000e-01 4.78866071e-01 7.69678891e-01
4.26673591e-01 -3.75046343e-01 3.52607280e-01 1.12485564e+00
-8.25326890e-02 -5.12261569e-01 -7.58159995e-01 -4.53537703e-02
2.30429813e-01 1.31850386e+00 -1.20827997e+00 -5.74430563e-02
-2.48433620e-01 1.19729590e+00 -1.53001890e-01 4.76858795e-01
-8.96265805e-01 -4.04762864e-01 9.09737110e-01 1.27900064e-01
5.41773319e-01 -3.52621198e-01 -1.05947852e-01 -1.18077624e+00
1.39593989e-01 -8.74310374e-01 3.51534598e-02 -8.09398830e-01
-1.51410592e+00 5.48987448e-01 3.08153033e-02 -1.39040208e+00
-1.09713271e-01 -1.05932617e+00 -3.82905826e-02 4.43682790e-01
-1.39140952e+00 -1.21870697e+00 -8.67760420e-01 2.23599568e-01
1.00213766e+00 -8.62114970e-03 6.65284991e-01 -4.96939495e-02
-6.50528371e-01 7.38860071e-02 4.78339642e-02 -9.76198316e-02
7.11817324e-01 -1.37692535e+00 5.10427535e-01 6.07628882e-01
4.78458293e-02 3.20595920e-01 7.21152961e-01 -8.20220113e-01
-1.42044544e+00 -1.29624534e+00 -5.69518626e-01 -5.78665614e-01
3.16119730e-01 -6.80660367e-01 -7.29775131e-01 5.73650897e-01
9.61996764e-02 2.00380281e-01 2.54258305e-01 -1.20007314e-01
-7.89579749e-02 -6.27382025e-02 -1.34523273e+00 5.31335831e-01
1.15723836e+00 -9.46252942e-02 -3.71631801e-01 4.03916955e-01
8.54056239e-01 -7.70055771e-01 -8.62260222e-01 8.98108721e-01
6.91861212e-01 -8.09194386e-01 1.09111989e+00 -8.08171704e-02
2.47431859e-01 -6.22780859e-01 -3.24376494e-01 -1.38152099e+00
-3.04968492e-03 -5.43841422e-01 -7.26565048e-02 1.16936147e+00
4.32892501e-01 -4.58153129e-01 1.15476549e+00 6.64260745e-01
-3.41380328e-01 -2.99996287e-01 -7.03399479e-01 -8.59714627e-01
-1.89912766e-01 -3.87560874e-01 7.51065493e-01 5.94811618e-01
-6.98055089e-01 3.64454091e-02 9.35135130e-03 2.95795739e-01
7.79160142e-01 2.68954754e-01 9.29719210e-01 -1.42479122e+00
-2.10856453e-01 -6.51827380e-02 -4.84130263e-01 -9.65443790e-01
1.58873647e-01 -3.37934226e-01 5.52727461e-01 -1.58756864e+00
-2.96128690e-01 -5.27413487e-01 -1.39164418e-01 2.09345907e-01
1.37237146e-01 4.38119859e-01 -1.51900472e-02 -1.11037716e-01
-5.64830899e-01 6.32456422e-01 1.59007239e+00 -5.82950056e-01
-2.47992426e-01 1.34218618e-01 -3.88714582e-01 8.42893720e-01
9.59863067e-01 -2.44375795e-01 -4.97948229e-01 -5.01014292e-01
-1.40830770e-01 -3.81561339e-01 5.19911408e-01 -1.24165583e+00
-2.17951387e-01 -1.74986780e-01 8.51962030e-01 -6.94114506e-01
5.83219647e-01 -1.16503882e+00 5.71258105e-02 2.65500754e-01
9.03434232e-02 -3.15020621e-01 3.42661440e-01 6.43895924e-01
-2.17553191e-02 -1.34236723e-01 5.80971241e-01 -3.72012287e-01
-7.76468575e-01 4.55092154e-02 -2.36701369e-01 -1.16934940e-01
1.29119790e+00 -4.59381551e-01 -1.28078640e-01 -2.42556259e-02
-7.69840956e-01 3.43204290e-02 9.15814757e-01 6.18276954e-01
8.00042152e-01 -1.10083663e+00 -4.65386242e-01 6.39488876e-01
3.16975147e-01 8.37549210e-01 -6.07661568e-02 2.77411193e-01
-7.82921433e-01 -4.58211936e-02 -1.76898837e-02 -8.65495563e-01
-8.34367812e-01 6.35147154e-01 2.78155774e-01 9.58818123e-02
-7.59365737e-01 5.66559017e-01 4.84967768e-01 -5.82580686e-01
2.89838672e-01 -9.34435368e-01 3.63419861e-01 -2.63660878e-01
4.06889133e-02 2.69697905e-01 2.70724654e-01 -7.71229789e-02
-1.49789333e-01 6.49076581e-01 -2.75292806e-02 5.25895394e-02
1.16622150e+00 -2.93918978e-02 2.75524497e-01 4.13665533e-01
7.88578153e-01 4.94169556e-02 -1.88424659e+00 1.37458459e-01
-9.91520956e-02 -6.45695031e-01 3.12574431e-02 -8.74630213e-01
-1.45270908e+00 6.74942136e-01 7.30777323e-01 1.69602796e-01
1.02083945e+00 2.06688017e-01 8.06921005e-01 4.35199082e-01
6.48013711e-01 -1.16752577e+00 2.58105218e-01 1.38667867e-01
7.38775730e-01 -1.44782603e+00 9.65293199e-02 -9.89763558e-01
-4.50305104e-01 1.17650437e+00 8.90046537e-01 2.44175605e-02
4.56683576e-01 7.03666568e-01 3.63161594e-01 -2.81026483e-01
-3.64367306e-01 -2.90151209e-01 -6.85572177e-02 9.60537374e-01
6.95377067e-02 -8.76823366e-02 4.64435041e-01 2.54846275e-01
-1.73820302e-01 -1.49687722e-01 2.40302995e-01 1.22061026e+00
-1.05619371e-01 -1.25583673e+00 -4.29436117e-01 3.54200393e-01
-1.45491853e-01 4.14249361e-01 -4.58540082e-01 1.02004051e+00
3.65967631e-01 1.07370353e+00 7.53543526e-02 -5.60610235e-01
7.80774772e-01 -3.07101697e-01 6.79461658e-01 -7.02176571e-01
-2.79193580e-01 2.25014150e-01 -2.45781168e-02 -8.43991756e-01
-1.96791261e-01 -7.10969746e-01 -1.03720176e+00 2.48005554e-01
-5.37066936e-01 -2.30810165e-01 8.50907445e-01 6.48451090e-01
3.22531730e-01 1.13213205e+00 4.85904753e-01 -1.05412591e+00
-3.54415685e-01 -8.70443881e-01 -7.23197818e-01 5.64970851e-01
1.89630032e-01 -9.22871470e-01 -4.77725238e-01 2.57142842e-01] | [6.382167816162109, -0.9982771277427673] |
70fe408e-551b-46ac-9c1b-9603af03b414 | compression-of-deep-neural-networks-for-image | 1701.04923 | null | http://arxiv.org/abs/1701.04923v1 | http://arxiv.org/pdf/1701.04923v1.pdf | Compression of Deep Neural Networks for Image Instance Retrieval | Image instance retrieval is the problem of retrieving images from a database
which contain the same object. Convolutional Neural Network (CNN) based
descriptors are becoming the dominant approach for generating {\it global image
descriptors} for the instance retrieval problem. One major drawback of
CNN-based {\it global descriptors} is that uncompressed deep neural network
models require hundreds of megabytes of storage making them inconvenient to
deploy in mobile applications or in custom hardware. In this work, we study the
problem of neural network model compression focusing on the image instance
retrieval task. We study quantization, coding, pruning and weight sharing
techniques for reducing model size for the instance retrieval problem. We
provide extensive experimental results on the trade-off between retrieval
performance and model size for different types of networks on several data sets
providing the most comprehensive study on this topic. We compress models to the
order of a few MBs: two orders of magnitude smaller than the uncompressed
models while achieving negligible loss in retrieval performance. | ['Ling-Yu Duan', 'Antoine Veillard', 'Olivier Morère', 'Jie Lin', 'Vijay Chandrasekhar', 'Tomaso Poggio', 'Qianli Liao'] | 2017-01-18 | null | null | null | null | ['image-instance-retrieval'] | ['computer-vision'] | [ 4.29684639e-01 -2.86861867e-01 -5.12669146e-01 -3.53142083e-01
-8.49777460e-01 -2.39091828e-01 3.94120157e-01 6.11309469e-01
-9.11881208e-01 3.82814050e-01 -4.01020259e-01 -1.15325131e-01
-5.24746716e-01 -1.02865326e+00 -9.08129930e-01 -5.14350176e-01
-1.36921585e-01 4.39575881e-01 3.40772837e-01 -9.00977175e-04
4.99089420e-01 7.93742120e-01 -2.07875657e+00 3.04426253e-01
5.55245131e-02 1.76811445e+00 5.53312540e-01 7.34284937e-01
-1.52994886e-01 9.47627366e-01 -9.53707457e-01 -2.37989709e-01
5.64442337e-01 5.09234294e-02 -7.44280815e-01 -2.91012168e-01
9.58832145e-01 -8.49452674e-01 -8.30765307e-01 1.04963696e+00
7.12096095e-01 2.27425080e-02 4.37396169e-01 -1.19824588e+00
-4.56309021e-01 3.99973512e-01 -5.23038325e-04 6.07908070e-01
-2.99493939e-01 -4.50090557e-01 7.71109462e-01 -5.40579855e-01
6.93904221e-01 1.15078676e+00 3.85731369e-01 4.51628298e-01
-8.00324559e-01 -8.03707838e-01 -3.10870498e-01 6.56453371e-01
-1.81449711e+00 -6.84983253e-01 1.87521845e-01 1.73407108e-01
1.51269257e+00 4.34799492e-01 6.21097684e-01 3.58494490e-01
1.90212294e-01 6.63684368e-01 3.35486233e-01 -4.04625684e-01
3.57068449e-01 -1.83207959e-01 2.35522792e-01 6.61461949e-01
4.80316550e-01 -2.40239292e-01 -7.99722731e-01 -8.56723711e-02
1.02010190e+00 2.61069715e-01 -7.26569593e-02 1.02456473e-01
-8.39538574e-01 7.88596690e-01 6.63480401e-01 2.29546517e-01
-4.90373254e-01 1.03972304e+00 7.57040679e-01 6.65522873e-01
2.99958289e-01 2.06255630e-01 -4.13444787e-01 -2.33582392e-01
-1.34359586e+00 4.78053600e-01 6.01651430e-01 1.42048049e+00
7.69021213e-01 -2.63492228e-03 1.02510557e-01 9.74129796e-01
-6.70863241e-02 4.98471022e-01 7.89123058e-01 -1.20234299e+00
4.34950113e-01 3.51148576e-01 -2.77845055e-01 -1.16673064e+00
8.79531875e-02 -4.32184756e-01 -1.06428158e+00 -3.02442044e-01
-1.69067323e-01 3.31000000e-01 -1.03855312e+00 1.33242106e+00
-1.94048993e-02 9.87276584e-02 -2.06203669e-01 6.00035667e-01
9.53819633e-01 8.32630932e-01 -4.61329781e-02 8.22993815e-02
1.17759264e+00 -1.06821156e+00 -5.24578750e-01 4.52029333e-02
9.18542087e-01 -6.60364985e-01 4.61776882e-01 3.71807665e-01
-1.53072810e+00 -5.92141867e-01 -1.24721241e+00 -4.95035946e-01
-5.77179730e-01 6.18437566e-02 4.08040792e-01 4.05610800e-01
-1.49342430e+00 8.70299816e-01 -9.55755949e-01 -1.32437900e-01
4.76964831e-01 8.34534824e-01 -3.84973317e-01 -3.32971990e-01
-9.79560375e-01 7.79482901e-01 6.11633599e-01 -8.78069997e-02
-8.84791791e-01 -8.09888065e-01 -6.43787444e-01 5.52988172e-01
-1.39453262e-01 -1.97014004e-01 1.08948898e+00 -8.65799189e-01
-8.89277279e-01 7.96663463e-01 -1.44098893e-01 -1.08193338e+00
-3.70221548e-02 -2.32189652e-02 -1.30017638e-01 6.35565519e-01
-3.37741643e-01 1.21484029e+00 8.58682215e-01 -6.69053853e-01
-5.87616086e-01 -3.98889154e-01 1.78565204e-01 2.88607739e-02
-9.71986592e-01 1.07696205e-01 -8.82366657e-01 -6.21369421e-01
5.12836054e-02 -8.62516284e-01 5.60012534e-02 4.76861864e-01
-3.03941239e-02 -3.55578154e-01 8.96807909e-01 -4.81919795e-01
1.34398592e+00 -2.06526160e+00 -4.16337326e-02 2.72002757e-01
2.87843764e-01 6.92093432e-01 -4.45568562e-01 4.01085049e-01
1.73523471e-01 3.27195942e-01 1.90200672e-01 -3.71837914e-01
-6.87222481e-02 4.70272392e-01 -2.94719309e-01 2.59471506e-01
6.92059286e-03 9.34366763e-01 -3.26979578e-01 -6.88827574e-01
-1.10975884e-01 8.48189056e-01 -4.25836772e-01 3.57615612e-02
9.06908140e-02 -7.21237719e-01 -3.71133924e-01 4.78210837e-01
6.10780418e-01 -4.01971430e-01 -2.38692127e-02 -2.97858953e-01
2.62060791e-01 3.66487026e-01 -8.71268809e-01 1.76247394e+00
-5.55429399e-01 1.13534498e+00 -1.36888593e-01 -1.15405703e+00
8.68910611e-01 2.82599568e-01 3.65682065e-01 -1.19341671e+00
1.55333996e-01 4.04880434e-01 -2.73187190e-01 -1.25628665e-01
8.27477753e-01 2.66803145e-01 1.49103835e-01 3.94175798e-01
2.05036610e-01 5.47644868e-02 4.40499932e-01 1.32954031e-01
1.29553163e+00 -6.65272176e-01 -2.09863007e-01 -2.37797961e-01
7.91272521e-02 2.38585249e-02 9.07414258e-02 9.51527655e-01
8.58225580e-03 6.34860635e-01 3.04115564e-01 -7.74254799e-01
-1.39533925e+00 -5.09745359e-01 -2.01399535e-01 9.44988132e-01
1.31191328e-01 -7.30990350e-01 -8.53158712e-01 1.36681572e-01
-5.77133931e-02 -1.40112266e-02 -4.03205007e-01 -5.20884335e-01
-7.35971808e-01 -3.78348231e-01 9.76618588e-01 5.98832488e-01
8.09401929e-01 -9.56421614e-01 -1.30890858e+00 1.23057649e-01
1.41794950e-01 -1.13506782e+00 -1.74119785e-01 2.67468184e-01
-1.28527367e+00 -7.48796880e-01 -1.02308357e+00 -9.45613563e-01
5.79524577e-01 4.76337761e-01 1.21384728e+00 9.28781033e-01
-6.47056997e-01 3.63724828e-01 -2.31435359e-01 -5.21977246e-01
-5.18202335e-02 4.25839871e-01 -2.48851761e-01 -6.34872735e-01
4.94181395e-01 -3.90553474e-01 -8.57005239e-01 -4.89097536e-02
-1.63506424e+00 -3.78458858e-01 5.06794214e-01 6.12873733e-01
8.91365469e-01 3.50513101e-01 3.58361632e-01 -5.19838035e-01
7.45669961e-01 -2.55584717e-01 -7.63118505e-01 3.06833476e-01
-6.97241008e-01 1.19823188e-01 3.51380557e-01 -5.62386453e-01
-1.79355428e-01 -1.31689832e-01 1.76740870e-01 -9.39137399e-01
1.74498618e-01 6.34241343e-01 4.87836838e-01 -3.37410748e-01
3.79740953e-01 4.23695594e-01 3.81811410e-02 -4.17877942e-01
-4.05616648e-02 5.43028951e-01 1.86450943e-01 -2.76042014e-01
3.00350338e-01 3.02584678e-01 2.66295314e-01 -8.99799347e-01
-2.77515352e-01 -4.35311049e-01 -4.32664216e-01 6.28520101e-02
5.91928065e-01 -1.05833673e+00 -6.14748955e-01 3.92407805e-01
-1.22914183e+00 -4.32595998e-01 -1.81465477e-01 8.97431374e-02
-4.33831841e-01 1.16015442e-01 -9.03535187e-01 -3.45523596e-01
-7.37297177e-01 -1.20543206e+00 1.32835948e+00 -2.53433269e-02
1.92472503e-01 -6.18165672e-01 -3.71592879e-01 7.48748109e-02
9.49587286e-01 -1.44546688e-01 1.12866557e+00 -5.23546457e-01
-1.02161205e+00 -6.38382018e-01 -6.00458324e-01 5.02309799e-01
-3.55601043e-01 -2.38129914e-01 -6.65137410e-01 -4.80897933e-01
-2.38233224e-01 -6.22168422e-01 1.02366245e+00 3.12183887e-01
1.83885777e+00 -3.50641966e-01 -2.13630289e-01 5.29540300e-01
1.94543517e+00 3.92917931e-01 9.89228666e-01 4.21676755e-01
2.97369570e-01 2.59814113e-01 4.40614223e-01 5.34169734e-01
-5.50830830e-03 7.02541471e-01 5.46379328e-01 1.24421000e-01
-2.97614336e-01 8.54083300e-02 -1.15359321e-01 9.99636889e-01
1.63264908e-02 -5.32868028e-01 -9.33981776e-01 8.38499725e-01
-1.53357458e+00 -7.17655420e-01 4.21442956e-01 2.31327033e+00
7.24705756e-01 -4.39927541e-02 -4.19120878e-01 3.61156404e-01
5.25583088e-01 3.12305782e-02 -3.27093452e-01 -4.70306307e-01
8.21468160e-02 8.85015547e-01 9.43074167e-01 2.21980706e-01
-8.20942700e-01 8.54581654e-01 6.30743313e+00 1.33362556e+00
-1.40353489e+00 1.43912032e-01 8.73042166e-01 -6.97424591e-01
1.77556962e-01 -3.06270003e-01 -1.05379784e+00 2.64201075e-01
1.46659887e+00 -1.52451798e-01 3.44954818e-01 8.39327574e-01
-3.19719374e-01 -6.94616884e-02 -1.13792396e+00 1.42968917e+00
9.67727602e-02 -1.77443159e+00 5.07906914e-01 2.72878379e-01
6.35449469e-01 4.05635804e-01 2.27839828e-01 2.36245152e-02
-4.32967097e-01 -1.24643922e+00 8.10985684e-01 4.37048703e-01
1.12053311e+00 -9.74078000e-01 7.82721519e-01 1.14883073e-01
-1.10953629e+00 4.71986225e-03 -1.04035294e+00 1.94276899e-01
-1.95453823e-01 3.82219046e-01 -3.96166384e-01 -1.20308660e-01
1.06108630e+00 4.87516284e-01 -6.30924344e-01 1.10898209e+00
7.88368881e-01 2.25853816e-01 -7.55045950e-01 -1.74220338e-01
4.75289196e-01 3.96901637e-01 -8.07849169e-02 1.13224423e+00
5.63261390e-01 3.27461772e-02 -3.30245495e-01 3.61013681e-01
-6.38150692e-01 -8.42157155e-02 -6.35430455e-01 -6.58483058e-02
7.41029441e-01 8.61431479e-01 -7.90628016e-01 -6.80945992e-01
-9.91184786e-02 9.38991785e-01 2.77337760e-01 1.42550185e-01
-5.68343401e-01 -9.45255041e-01 5.20471632e-01 1.09747708e-01
6.47529364e-01 -3.61009210e-01 2.49004945e-01 -7.02793181e-01
2.72260547e-01 -8.34011018e-01 9.31511051e-04 -7.10485458e-01
-6.11986876e-01 7.40351558e-01 3.35825890e-01 -9.56998885e-01
-3.13376009e-01 -6.34615242e-01 8.70791078e-02 6.48153603e-01
-1.82257068e+00 -6.00032568e-01 -6.56477273e-01 6.39711380e-01
6.60064220e-01 -2.84787625e-01 9.66073155e-01 8.45276475e-01
-2.59586394e-01 8.47012877e-01 3.39480281e-01 3.77664529e-02
2.52407521e-01 -4.91483182e-01 2.57739335e-01 2.24671945e-01
2.07817256e-01 8.35023820e-01 2.68695235e-01 -2.41133481e-01
-1.69385445e+00 -1.18114471e+00 1.04194331e+00 1.53275207e-01
3.49752039e-01 -1.99636474e-01 -7.39871323e-01 4.31194842e-01
1.24578923e-01 3.09382379e-01 4.57349330e-01 -6.16229892e-01
-4.14764702e-01 -4.08589423e-01 -1.12502277e+00 3.52595210e-01
9.09568965e-01 -7.62725353e-01 1.82429835e-01 5.95398128e-01
7.31867135e-01 -3.39558393e-01 -1.25782096e+00 1.48222655e-01
6.37885869e-01 -7.01548159e-01 1.25155783e+00 -4.87040609e-01
6.35588050e-01 2.94965297e-01 -5.71556389e-01 -4.79204118e-01
1.55369714e-01 -1.08181894e-01 -3.35914999e-01 7.70334005e-01
2.59251088e-01 -2.30347723e-01 1.05575573e+00 7.12197781e-01
2.34677970e-01 -1.07556248e+00 -1.16807210e+00 -8.80522728e-01
2.34711114e-02 -3.54606062e-01 6.78836405e-01 4.94654924e-01
-6.43397629e-01 -2.36171022e-01 -5.18222637e-02 -2.23200917e-01
5.62083483e-01 -1.65325075e-01 5.46338439e-01 -1.01663148e+00
-1.03364654e-01 -3.03940237e-01 -9.48354363e-01 -1.30929565e+00
1.40843332e-01 -8.41046154e-01 -1.63787112e-01 -1.38605392e+00
2.58334279e-01 -7.09156930e-01 -4.89335358e-01 4.99284953e-01
7.79833496e-01 7.88255274e-01 5.10299087e-01 5.16597807e-01
-7.89975226e-01 2.52699286e-01 8.55867565e-01 -4.68096823e-01
3.07606548e-01 -5.05293012e-01 -2.92130619e-01 1.35355592e-01
7.25571215e-01 -7.67421663e-01 -6.86044633e-01 -1.33686018e+00
3.75820726e-01 2.61196867e-02 5.14792204e-01 -1.37049985e+00
5.98819613e-01 3.58106583e-01 3.07877272e-01 -6.89374745e-01
8.13758016e-01 -9.70634401e-01 4.08516750e-02 5.35045087e-01
-7.64137208e-01 7.25821614e-01 3.82178992e-01 4.95101929e-01
-6.98918223e-01 -6.87947989e-01 6.03442132e-01 -3.45695227e-01
-6.11304522e-01 6.37400806e-01 -3.05664420e-01 -2.66120106e-01
5.40684581e-01 -2.91936100e-01 -2.71688372e-01 -6.02320492e-01
-4.33258146e-01 -1.40196025e-01 1.60074696e-01 3.39396864e-01
1.08099413e+00 -1.27214336e+00 -3.32278907e-01 4.40762602e-02
-4.62290347e-02 1.04397304e-01 5.27650528e-02 3.18142354e-01
-1.33624113e+00 1.03272378e+00 -2.68709630e-01 -6.40180588e-01
-1.55338883e+00 3.69377226e-01 3.37575525e-01 -2.97265232e-01
-3.97732198e-01 9.76210773e-01 -2.05735147e-01 1.72191441e-01
6.27264977e-01 -3.89283627e-01 1.83013573e-01 -1.72013357e-01
8.81626487e-01 5.13888419e-01 5.25754511e-01 -4.93171513e-01
-1.74626503e-02 4.59973186e-01 -4.05906171e-01 6.23752251e-02
1.47616041e+00 1.02224588e-01 -5.21343768e-01 -1.53040690e-02
1.98652709e+00 -1.09527493e+00 -7.36182690e-01 -1.64150432e-01
-5.02988845e-02 -4.12269711e-01 7.16605723e-01 -1.46113768e-01
-1.51172948e+00 1.04488492e+00 1.10750067e+00 1.90796793e-01
1.38198435e+00 -1.96784735e-01 1.29124069e+00 1.20810187e+00
6.36731446e-01 -1.26912260e+00 2.12361533e-02 4.59726453e-01
8.30900371e-01 -7.30279684e-01 2.87671298e-01 -6.76760152e-02
2.55570918e-01 1.22032440e+00 2.53293216e-01 -6.69504285e-01
9.34675634e-01 1.60352230e-01 -1.70906901e-01 -4.05533850e-01
-1.00150692e+00 2.53827661e-01 1.16284072e-01 2.52297789e-01
2.08302602e-01 -3.09852272e-01 -2.00716898e-01 -1.81342512e-01
-1.33782119e-01 1.05617210e-01 1.37109846e-01 1.36117935e+00
-4.87527013e-01 -1.20172763e+00 6.46746382e-02 8.14276040e-01
-7.78253138e-01 -4.27297682e-01 -1.06700599e-01 7.34959185e-01
-1.18156746e-01 5.91196477e-01 5.98257601e-01 -2.87192166e-01
4.19948921e-02 -1.12942591e-01 6.10779405e-01 -3.57820690e-01
-7.14255929e-01 -1.29012495e-01 -1.64113924e-01 -7.73163617e-01
-5.53527057e-01 -7.33176544e-02 -1.08855355e+00 -5.38761497e-01
-4.82084244e-01 -1.32067487e-01 1.16212344e+00 6.39048994e-01
7.21391618e-01 3.06829989e-01 6.61254749e-02 -9.36219811e-01
-4.48843658e-01 -7.93532491e-01 -5.24186492e-01 3.15081775e-01
3.38038176e-01 -2.81759083e-01 -7.62815550e-02 1.08074188e-01] | [8.560998916625977, 3.0139849185943604] |
5898ed09-a054-4ea5-b61f-3bf6a5a1550a | a-large-scale-study-on-unsupervised | 2104.14558 | null | https://arxiv.org/abs/2104.14558v1 | https://arxiv.org/pdf/2104.14558v1.pdf | A Large-Scale Study on Unsupervised Spatiotemporal Representation Learning | We present a large-scale study on unsupervised spatiotemporal representation learning from videos. With a unified perspective on four recent image-based frameworks, we study a simple objective that can easily generalize all these methods to space-time. Our objective encourages temporally-persistent features in the same video, and in spite of its simplicity, it works surprisingly well across: (i) different unsupervised frameworks, (ii) pre-training datasets, (iii) downstream datasets, and (iv) backbone architectures. We draw a series of intriguing observations from this study, e.g., we discover that encouraging long-spanned persistency can be effective even if the timespan is 60 seconds. In addition to state-of-the-art results in multiple benchmarks, we report a few promising cases in which unsupervised pre-training can outperform its supervised counterpart. Code is made available at https://github.com/facebookresearch/SlowFast | ['Kaiming He', 'Ross Girshick', 'Bo Xiong', 'Haoqi Fan', 'Christoph Feichtenhofer'] | 2021-04-29 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Feichtenhofer_A_Large-Scale_Study_on_Unsupervised_Spatiotemporal_Representation_Learning_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Feichtenhofer_A_Large-Scale_Study_on_Unsupervised_Spatiotemporal_Representation_Learning_CVPR_2021_paper.pdf | cvpr-2021-1 | ['self-supervised-action-recognition', 'unsupervised-pre-training'] | ['computer-vision', 'methodology'] | [ 2.66884774e-01 -3.75062436e-01 -6.67938948e-01 -2.91032284e-01
-7.24523902e-01 -6.53331935e-01 7.67281592e-01 -1.87400818e-01
-2.55243361e-01 5.64285636e-01 6.81798697e-01 -2.43802443e-01
-3.75249237e-01 -3.29745501e-01 -6.26503527e-01 -7.17208624e-01
-6.53321743e-01 -1.62682593e-01 3.33625525e-01 -2.59808481e-01
2.93593138e-01 2.95627534e-01 -1.73724544e+00 3.37812304e-01
4.80133951e-01 1.07197821e+00 1.42400250e-01 5.49620152e-01
4.43551391e-01 1.11155415e+00 -5.72281443e-02 1.11511908e-01
3.67546618e-01 -2.43863225e-01 -8.71430993e-01 3.32406610e-01
4.46676105e-01 -2.71128446e-01 -1.06582236e+00 7.05194533e-01
1.94365099e-01 5.06114125e-01 4.57605600e-01 -1.27757621e+00
-7.98640966e-01 2.78244793e-01 -4.12038863e-01 7.16209590e-01
3.80262703e-01 4.40334409e-01 1.11410499e+00 -8.44343424e-01
8.40067625e-01 8.07075560e-01 5.11641145e-01 3.00356030e-01
-1.18070054e+00 -2.76756555e-01 5.39066672e-01 3.81649047e-01
-1.22389209e+00 -7.84268498e-01 6.38824999e-01 -5.33666372e-01
1.08641970e+00 2.68645227e-01 5.24094284e-01 1.45254755e+00
7.52938911e-02 9.38696265e-01 1.09347665e+00 -3.16608459e-01
7.24668428e-02 -3.16175401e-01 1.63916454e-01 5.77541769e-01
-1.47291884e-01 2.40381673e-01 -8.87766540e-01 -1.27451378e-03
6.86963618e-01 3.13962638e-01 -3.31402212e-01 -5.19631624e-01
-1.49158978e+00 7.67195106e-01 3.22609842e-01 5.97681582e-01
-7.51863793e-02 3.68584156e-01 6.35634184e-01 5.66527784e-01
4.24217582e-01 3.83100152e-01 -5.59054017e-01 -5.65651655e-01
-1.02516568e+00 9.05327871e-02 5.87001622e-01 1.10517871e+00
7.93499351e-01 6.20653406e-02 -1.29533648e-01 4.82119113e-01
-1.96800962e-01 8.40951279e-02 6.43439531e-01 -1.34770989e+00
4.11742032e-01 4.09671478e-02 -3.63147557e-02 -9.02546823e-01
-2.97132313e-01 -3.05485964e-01 -6.33624911e-01 -1.58973441e-01
4.33423281e-01 -6.33035973e-02 -8.57396185e-01 1.93124926e+00
-1.57398701e-01 6.67040765e-01 -1.98770776e-01 9.51300740e-01
6.40535116e-01 5.91121793e-01 -9.22698975e-02 -3.75781506e-01
9.48123336e-01 -1.39795601e+00 -4.74927068e-01 -1.33167595e-01
7.78814554e-01 -4.15970534e-01 1.12688792e+00 3.46505433e-01
-9.13597226e-01 -4.65877146e-01 -7.54031360e-01 -1.29204005e-01
-4.51802254e-01 3.57755413e-03 1.12812746e+00 1.13073125e-01
-1.34429133e+00 8.40868354e-01 -1.05990398e+00 -1.02928495e+00
2.84391195e-01 1.19020492e-01 -5.45455992e-01 -4.51864749e-02
-8.95557940e-01 4.69151169e-01 7.75534809e-02 -1.81264609e-01
-1.12794137e+00 -6.54200554e-01 -7.86829770e-01 -1.39847949e-01
7.63383746e-01 -5.95107377e-01 1.22422993e+00 -1.23883331e+00
-1.34660876e+00 8.09000671e-01 -3.77345502e-01 -5.64222634e-01
4.48395431e-01 -5.91961920e-01 -2.99022526e-01 4.50408787e-01
2.48382837e-01 5.87688804e-01 7.89800882e-01 -8.51381361e-01
-5.16226172e-01 -7.91866332e-02 2.73508698e-01 1.35324612e-01
-6.11052871e-01 1.67251050e-01 -7.02049434e-01 -1.02077961e+00
-5.34000695e-02 -1.08247900e+00 -2.52782106e-01 2.62056775e-02
-3.05256307e-01 -2.46587649e-01 8.37546229e-01 -3.16213161e-01
1.32798600e+00 -2.33836174e+00 3.50220978e-01 -1.12370282e-01
1.74558893e-01 -9.74255279e-02 -1.71217144e-01 7.59929836e-01
-3.45405608e-01 2.16657057e-01 -1.25303596e-01 -3.25651050e-01
-2.08696928e-02 1.54578730e-01 -5.65672219e-01 6.15345597e-01
2.69204289e-01 1.00216079e+00 -1.09215450e+00 -3.14815551e-01
2.50232339e-01 1.25387162e-01 -6.01073384e-01 8.43848661e-02
-8.71703327e-02 8.52739930e-01 -5.11260152e-01 8.31694901e-01
1.10309951e-01 -7.10578084e-01 7.72957802e-02 -1.16027832e-01
-2.32557163e-01 4.48600888e-01 -7.15552449e-01 2.23165488e+00
-4.46371995e-02 9.85524178e-01 -2.31493041e-01 -1.47669876e+00
1.83019117e-01 3.65658551e-01 1.03572690e+00 -6.02842152e-01
-1.47542164e-01 -8.38047490e-02 -1.60352230e-01 -7.92271435e-01
5.43374538e-01 1.89916089e-01 6.18085964e-03 5.54256737e-01
4.77518141e-01 3.51368099e-01 5.09365857e-01 3.94061387e-01
1.33931541e+00 5.41574359e-01 1.85832277e-01 -3.60391200e-01
7.39520267e-02 -1.20194934e-01 5.23838937e-01 8.98914814e-01
-3.96702856e-01 7.03448415e-01 4.69236016e-01 -6.06224120e-01
-9.21739876e-01 -1.14346004e+00 -4.43803631e-02 1.51462197e+00
7.73713961e-02 -8.69403362e-01 -1.82745308e-01 -6.05732083e-01
-8.63818750e-02 1.63682833e-01 -8.32503378e-01 4.25714962e-02
-5.38403273e-01 -4.20397401e-01 4.47223604e-01 8.09862733e-01
2.46040627e-01 -8.53182375e-01 -5.17276525e-01 -4.07425873e-02
-2.45865330e-01 -1.42667222e+00 -4.58031267e-01 2.48663932e-01
-1.06949842e+00 -8.24390411e-01 -7.18401015e-01 -7.15657711e-01
3.86117131e-01 7.59077609e-01 1.19270301e+00 2.47842342e-01
-1.46725386e-01 8.00999880e-01 -5.43003380e-01 1.30193457e-01
3.23817611e-01 1.96297422e-01 3.04245412e-01 -5.53902574e-02
2.45318130e-01 -1.02459037e+00 -8.54532480e-01 4.38203067e-01
-8.66137326e-01 -2.18488872e-02 2.66965598e-01 5.68628013e-01
5.13839066e-01 -2.19026297e-01 6.31725967e-01 -5.96162260e-01
2.07859039e-01 -9.41171885e-01 -2.06555933e-01 -4.23293281e-03
-4.11679178e-01 -2.28639871e-01 5.21303475e-01 -5.40038526e-01
-6.31573379e-01 -7.72135630e-02 2.96360970e-01 -7.99419165e-01
-2.65129060e-01 7.09465563e-01 4.55251396e-01 1.11973815e-01
7.14651823e-01 5.02264142e-01 -1.53311826e-02 -4.64581162e-01
5.37106335e-01 2.19215915e-01 6.18614197e-01 -8.36490512e-01
7.94161201e-01 8.58153641e-01 -1.57024845e-01 -9.17333543e-01
-7.36866474e-01 -6.86630189e-01 -7.47251391e-01 -1.17029905e-01
7.61520803e-01 -1.21596026e+00 -2.88013428e-01 2.07038999e-01
-6.63725853e-01 -8.84671569e-01 -3.21580082e-01 5.65008283e-01
-9.36830699e-01 4.54635769e-01 -7.19947278e-01 -4.74661678e-01
1.98769823e-01 -9.38868761e-01 1.00836825e+00 1.47180334e-01
-1.90465167e-01 -1.06754100e+00 5.74853234e-02 1.36305287e-01
4.09538627e-01 3.13568383e-01 4.61098164e-01 -4.08844650e-01
-8.20461035e-01 1.70599371e-01 -1.32747829e-01 2.60885149e-01
3.04154098e-01 1.84324682e-01 -9.17705238e-01 -5.74135959e-01
-1.98625386e-01 -6.17983878e-01 1.33528078e+00 3.96426558e-01
1.43292463e+00 -2.04799592e-01 -3.77692640e-01 8.86892259e-01
1.31957579e+00 -1.21588178e-01 6.09654605e-01 4.51404870e-01
4.78151351e-01 4.32081610e-01 7.07757533e-01 4.10679013e-01
3.98510337e-01 8.20967436e-01 3.21197718e-01 8.35275576e-02
-9.31576416e-02 -7.99858421e-02 5.86724341e-01 8.50980461e-01
-6.75867975e-01 -2.28747681e-01 -6.61838591e-01 8.13361287e-01
-2.27288842e+00 -1.27964294e+00 2.53025085e-01 2.04421377e+00
5.31348646e-01 4.08887900e-02 5.71997881e-01 -1.85951293e-01
4.06635195e-01 7.65637934e-01 -6.44645393e-01 7.47848898e-02
-2.74362534e-01 9.10340995e-02 2.96457529e-01 -5.13953455e-02
-1.48953617e+00 1.01551294e+00 6.84044266e+00 6.48581207e-01
-1.39813650e+00 3.08796316e-01 5.82105875e-01 -6.06049359e-01
-1.02147631e-01 1.58009544e-01 -2.12922797e-01 6.56199396e-01
9.65037704e-01 -3.26287031e-01 6.10119343e-01 6.87710643e-01
3.21279228e-01 -3.43997478e-02 -1.40416479e+00 1.03441393e+00
7.33414665e-02 -1.63789058e+00 -2.86826968e-01 5.35349250e-02
9.19305801e-01 6.35163903e-01 2.80342013e-01 2.68620700e-01
5.81755154e-02 -1.10856104e+00 7.98893332e-01 4.88694012e-01
9.31078851e-01 -2.59637862e-01 2.65565813e-01 1.87389985e-01
-1.39183080e+00 -3.20848167e-01 -2.18841597e-01 -3.48096341e-01
1.88578472e-01 2.86668450e-01 1.78952500e-01 7.21397936e-01
1.08529484e+00 1.58650875e+00 -7.26742446e-01 1.09302711e+00
-1.47184953e-01 7.90811181e-01 -3.64041001e-01 4.55525845e-01
5.11435747e-01 -1.58355355e-01 5.65343678e-01 1.45874667e+00
4.25073415e-01 1.52641788e-01 2.65980631e-01 4.14849102e-01
-7.54908845e-02 -1.86524734e-01 -1.06988072e+00 -2.72388488e-01
2.73898661e-01 1.18978202e+00 -6.96679354e-01 -3.25462997e-01
-8.71096671e-01 9.79377866e-01 5.54737389e-01 7.69708633e-01
-1.03515637e+00 -5.20780310e-02 6.85020328e-01 1.26384601e-01
5.78944802e-01 -6.39184415e-01 1.75923914e-01 -1.65564454e+00
2.26174384e-01 -7.34409928e-01 5.81217587e-01 -8.52745354e-01
-1.22341037e+00 4.13910002e-01 1.28709689e-01 -1.57151830e+00
-4.61538732e-01 -4.63196784e-01 -7.32331634e-01 1.67245224e-01
-1.60681927e+00 -9.51627851e-01 -3.10869783e-01 8.98123205e-01
7.98284709e-01 -2.16561094e-01 7.24024475e-01 4.09676880e-01
-6.50771558e-01 4.39609826e-01 2.67760634e-01 1.73855051e-01
8.82871747e-01 -1.06122363e+00 3.05410266e-01 9.89087164e-01
6.19823158e-01 6.75155818e-01 5.57874501e-01 -2.29772806e-01
-1.69571614e+00 -1.08986139e+00 4.47630942e-01 -6.66051388e-01
1.15505838e+00 -3.32344621e-01 -7.26509273e-01 1.29276443e+00
4.33229208e-01 5.11389256e-01 6.42348647e-01 3.29935104e-01
-5.30126512e-01 6.19950481e-02 -4.41158712e-01 6.23788655e-01
1.62055862e+00 -7.92665184e-01 -4.58889097e-01 6.51348889e-01
7.13227987e-01 -2.99523383e-01 -9.80148733e-01 3.63171339e-01
6.84738457e-01 -1.16542017e+00 9.68740940e-01 -1.01631331e+00
7.37012088e-01 -2.03667611e-01 -3.14773470e-01 -1.02169788e+00
-5.79096377e-01 -1.15118897e+00 -6.29410744e-01 9.14689481e-01
2.36230120e-01 -6.04370832e-01 7.47961640e-01 3.55618417e-01
-2.73246288e-01 -1.18852484e+00 -8.24170947e-01 -1.26856923e+00
5.31573966e-02 -5.00746250e-01 7.63113722e-02 1.07534111e+00
2.79432386e-01 1.36523873e-01 -5.84961891e-01 -2.16414910e-02
4.17405903e-01 2.54389435e-01 7.57234216e-01 -7.82418370e-01
-3.84524554e-01 -4.10136700e-01 -5.78790545e-01 -1.64513326e+00
1.35676414e-01 -7.28928804e-01 -2.22216412e-01 -1.28058946e+00
4.09576327e-01 -3.34601700e-01 -7.30190039e-01 5.93358219e-01
7.00082555e-02 3.28588992e-01 3.34457964e-01 6.55229628e-01
-1.21135437e+00 4.86282021e-01 9.47926760e-01 -6.48397356e-02
-8.03332925e-02 -2.53711730e-01 -8.41656983e-01 7.03231394e-01
7.46653557e-01 -3.12418938e-01 -6.30369186e-01 -6.92540705e-01
2.31515970e-02 -9.19720624e-03 4.85516876e-01 -9.78556156e-01
2.75709540e-01 -4.40365821e-01 4.92344014e-02 -2.98735052e-01
4.13632303e-01 -4.73508716e-01 -1.07111268e-01 2.90029347e-02
-4.18966919e-01 2.10741073e-01 1.49860248e-01 8.37825477e-01
-2.99722672e-01 2.54869074e-01 4.88220066e-01 -2.11519673e-01
-1.26449192e+00 5.95297337e-01 -3.75599295e-01 3.11589450e-01
9.57937360e-01 -3.85110795e-01 -6.32616818e-01 -8.06334138e-01
-7.66712487e-01 1.98873416e-01 6.19592428e-01 7.01987505e-01
5.17371058e-01 -1.40460563e+00 -6.15944028e-01 -1.46454617e-01
3.37445647e-01 -3.55176061e-01 2.41522014e-01 1.44465160e+00
-1.76989302e-01 5.56177497e-01 4.92694788e-03 -9.49042320e-01
-8.62221837e-01 7.80805171e-01 -5.34587838e-02 -1.14088312e-01
-8.70123804e-01 6.02921069e-01 4.92870688e-01 1.00004688e-01
3.59058321e-01 -3.28486145e-01 5.60097955e-02 -9.72217768e-02
5.07556498e-01 2.05474094e-01 -1.93042621e-01 -7.66864300e-01
-4.85539198e-01 6.90623283e-01 -4.86835763e-02 -6.87950989e-03
1.64668989e+00 -3.03642720e-01 1.95493311e-01 7.41464138e-01
1.32762837e+00 -8.36774707e-02 -1.67360723e+00 -3.60505372e-01
6.45506606e-02 -5.76433778e-01 -1.77539304e-01 -3.36725622e-01
-1.18711662e+00 6.42533422e-01 3.41270089e-01 3.78611058e-01
1.32032883e+00 2.45673463e-01 6.01976871e-01 5.50045907e-01
5.18485487e-01 -1.00180292e+00 4.17096585e-01 6.30727470e-01
8.75884771e-01 -1.42027318e+00 2.43121922e-01 -1.73291206e-01
-6.89522743e-01 1.08297718e+00 3.94814521e-01 -4.44458336e-01
8.25491548e-01 -1.05140656e-01 -3.09708297e-01 -2.64521390e-01
-1.20339632e+00 -3.66914093e-01 1.61012128e-01 5.58251143e-01
5.54808676e-01 -2.83345670e-01 -2.53689103e-03 3.83385539e-01
1.18177868e-01 3.22526455e-01 4.00892079e-01 1.15604162e+00
-1.29873306e-01 -7.62812376e-01 1.89562798e-01 4.20873404e-01
-5.89968085e-01 -9.82227549e-02 -4.78540063e-02 9.39098299e-01
-2.32290581e-01 9.87494707e-01 1.69711798e-01 -4.13352609e-01
6.08644150e-02 -1.49990529e-01 5.38782656e-01 -6.85902119e-01
-1.92198098e-01 4.07858081e-02 2.25736722e-01 -1.26392460e+00
-9.82418418e-01 -7.56663382e-01 -9.57766712e-01 -2.89962322e-01
7.87098482e-02 -4.54130322e-02 9.89837125e-02 9.43872869e-01
6.68982029e-01 2.16296196e-01 6.89686954e-01 -1.12264347e+00
-1.48776501e-01 -9.03097808e-01 -5.92641771e-01 6.18228316e-01
5.66398561e-01 -7.74665534e-01 -4.85291183e-01 4.17291701e-01] | [8.754661560058594, 0.7314515113830566] |
aa3649a3-a147-446a-98f0-6b0de61dd4fe | distributed-coordination-of-multi-microgrids | 2305.04456 | null | https://arxiv.org/abs/2305.04456v1 | https://arxiv.org/pdf/2305.04456v1.pdf | Distributed Coordination of Multi-Microgrids in Active Distribution Networks for Provisioning Ancillary Services | We propose a distributed optimization framework that coordinates multiple microgrids in an Active Distribution Network (ADN) for provisioning passive voltage support based ancillary services while satisfying operational constraints. Specifically, we exploit the reactive power support capability of the inverters and the flexibility offered by storage systems available with microgrids for provisioning ancillary service support to the transmission grid. We develop novel mixed integer inequalities to represent the set of feasible active and reactive power exchange with the transmission grid that ensure passive voltage support. The proposed alternating direction method of multipliers (ADMM) based algorithm is fully distributed, and does not require the presence of a centralized entity to achieve coordination among the microgrids. We present detailed numerical results on the IEEE 33-bus distribution test system to demonstrate the effectiveness of the proposed approach and examine convergence behavior of the distributed algorithm for different choice of hyper-parameters. | ['Prabodh Bajpai', 'Ashish R. Hota', 'Abhishek Mishra', 'Arghya Mallick'] | 2023-05-08 | null | null | null | null | ['distributed-optimization'] | ['methodology'] | [-5.72781384e-01 -6.77792430e-02 -3.56324196e-01 -6.44899383e-02
-2.69840330e-01 -1.19708514e+00 1.58000246e-01 2.27516606e-01
2.22178310e-01 1.43623650e+00 -1.84596270e-01 -2.44303301e-01
-8.15984607e-01 -8.00294459e-01 8.56410563e-02 -1.28211164e+00
-6.15939200e-01 4.87185478e-01 -4.92607027e-01 -4.68889832e-01
1.29695505e-01 7.75904834e-01 -6.50914669e-01 -5.38428843e-01
1.24252439e+00 1.21668780e+00 -3.81444171e-02 1.53056579e-03
4.61336851e-01 2.26506338e-01 -1.08442080e+00 2.65029877e-01
6.20897114e-01 -3.54859591e-01 -1.27411649e-01 5.24040818e-01
-5.16947627e-01 -4.76903200e-01 -1.09292030e-01 1.40463209e+00
6.96365952e-01 3.08371544e-01 3.99375468e-01 -2.26783395e+00
-4.39578086e-01 6.96154773e-01 -1.03959739e+00 3.99446785e-01
2.47234434e-01 2.28537619e-01 9.98293459e-01 -5.22688448e-01
3.60883743e-01 4.80501592e-01 -5.47990017e-03 -1.25046507e-01
-1.33297408e+00 -4.17686462e-01 3.98125559e-01 5.03903627e-01
-1.56866670e+00 -4.05456722e-02 8.82197261e-01 -5.36654964e-02
1.07642329e+00 5.43455720e-01 1.13135695e+00 -2.83260763e-01
2.94252336e-01 2.55083621e-01 8.44538450e-01 -1.83213249e-01
6.84727848e-01 -4.63783108e-02 6.25623688e-02 -1.03434222e-02
8.40142250e-01 -6.98496878e-01 -1.82610348e-01 -4.22837526e-01
3.22986782e-01 -3.05135936e-01 -8.61685038e-01 -5.47977388e-01
-8.47321451e-01 1.00796580e+00 3.13775063e-01 3.57730478e-01
-5.95704019e-01 -2.54739374e-01 1.01813875e-01 5.12484670e-01
4.21216547e-01 2.42234290e-01 -4.74453926e-01 3.90971094e-01
-1.03142345e+00 8.76722187e-02 8.89673173e-01 1.13832474e+00
3.50873679e-01 1.01043177e+00 1.76883087e-01 1.89487591e-01
4.15972143e-01 4.40570861e-01 1.80424735e-01 -9.20115054e-01
3.19001377e-01 5.89424074e-01 3.99911642e-01 -6.16032958e-01
-4.00247157e-01 -9.52871740e-01 -9.94511247e-01 5.70494235e-01
1.65926829e-01 -7.28344440e-01 1.28427014e-01 1.33506656e+00
3.32092941e-01 -2.73489207e-01 -3.88164259e-02 9.08964157e-01
-1.27099067e-01 1.02267015e+00 -5.20599782e-01 -1.11424839e+00
9.63375568e-01 -7.46476650e-01 -1.13752568e+00 3.03171962e-01
5.10683417e-01 -5.39647996e-01 -1.35219827e-01 4.60385233e-01
-1.55115521e+00 2.34402806e-01 -1.46964765e+00 6.35873556e-01
-2.27383956e-01 -5.64137697e-02 2.91393250e-01 4.42541093e-01
-1.23929751e+00 1.46841347e-01 -5.49179673e-01 2.48309195e-01
4.77969879e-03 6.47265971e-01 -2.99163997e-01 3.59495759e-01
-9.14084256e-01 1.07197464e+00 5.32301784e-01 5.74206650e-01
-4.42784727e-01 -8.48608077e-01 -7.80506551e-01 4.29714471e-01
2.65089095e-01 -6.83097363e-01 8.13645661e-01 -6.70347333e-01
-1.22674823e+00 -7.64011070e-02 1.52379721e-01 -6.02334499e-01
4.23274875e-01 6.34526312e-01 -4.47671741e-01 3.55837256e-01
-5.23543321e-02 -1.34984225e-01 1.58540741e-01 -9.60648179e-01
-5.78550637e-01 -2.00719848e-01 9.96772051e-02 3.57347697e-01
-4.17897552e-01 -1.71247244e-01 4.63500351e-01 -6.95315778e-01
4.06221449e-02 -3.73107046e-01 -6.60451710e-01 -4.79822867e-02
-4.19501126e-01 -2.76488513e-01 1.02305901e+00 -7.61226058e-01
1.19859827e+00 -1.77073860e+00 5.10076880e-01 8.36903036e-01
-8.62798840e-02 -2.81520814e-01 1.24753006e-01 1.01568949e+00
-1.71946064e-01 -2.73542076e-01 -1.28774360e-01 -2.14063928e-01
3.98241490e-01 5.50508022e-01 9.20200124e-02 9.37105358e-01
-2.04080254e-01 3.68974388e-01 -7.14512110e-01 1.14168540e-01
2.81427830e-01 1.07918710e-01 -3.37835312e-01 -4.33052219e-02
-7.85348862e-02 3.21581364e-01 -3.18656743e-01 5.62078655e-01
1.07759511e+00 -1.74961001e-01 6.09821796e-01 -2.37943932e-01
-4.67331648e-01 -5.41535392e-02 -1.87087131e+00 1.49456477e+00
-2.52312332e-01 4.46550488e-01 1.03527772e+00 -1.38865089e+00
4.08454716e-01 5.07332265e-01 9.43845034e-01 -6.17935777e-01
9.40853264e-03 2.34288752e-01 2.55458862e-01 1.30171608e-02
-3.96186998e-03 2.36476853e-01 1.97844014e-01 6.66045010e-01
2.01446533e-01 -8.12728256e-02 9.18733239e-01 2.92590559e-01
5.13934314e-01 -5.05701900e-01 6.06974542e-01 -1.09064305e+00
1.05786407e+00 -1.66839465e-01 1.05296612e+00 -7.94597119e-02
-3.49567793e-02 -3.45633954e-01 6.21583879e-01 1.20996192e-01
-8.51761639e-01 -9.90470648e-01 -2.50708044e-01 2.72220641e-01
5.57581149e-02 2.99132685e-03 -2.55307883e-01 -1.89856231e-01
1.86331391e-01 9.69532073e-01 -1.12622872e-01 4.29433137e-01
-4.12519127e-01 -1.25457513e+00 -5.15954614e-01 1.72582120e-01
1.60579026e-01 3.14806364e-02 -3.36785108e-01 3.86032015e-01
2.91146994e-01 -8.40691149e-01 -4.29710776e-01 3.98667157e-01
-5.26462495e-01 -8.87636185e-01 -6.06216788e-01 -9.05752599e-01
1.21799719e+00 -1.19336337e-01 8.44172359e-01 -3.51148546e-02
-9.34143290e-02 2.40413532e-01 2.30565608e-01 4.64536063e-02
1.24520816e-01 -2.01941088e-01 4.05684888e-01 -2.58245587e-01
-5.66267848e-01 -1.08672702e+00 -6.43384099e-01 2.41088122e-01
-7.58781374e-01 -1.67117193e-02 -7.93711096e-02 5.57867765e-01
5.10347962e-01 9.11437452e-01 1.43041706e+00 5.76565750e-02
7.16647685e-01 -9.47925270e-01 -1.59927189e+00 2.05340818e-01
-7.16656506e-01 -4.00126547e-01 9.28907394e-01 -8.19844473e-03
-6.68193698e-01 -2.05389008e-01 1.91365823e-01 9.43544433e-02
5.13478756e-01 5.86286545e-01 -7.55563736e-01 -7.65735507e-01
-5.50279796e-01 2.93239295e-01 -3.24772038e-02 -4.02489156e-01
3.57162714e-01 2.95450091e-01 3.19845617e-01 -4.37167376e-01
1.31845057e+00 3.91617179e-01 5.32002091e-01 -3.97338241e-01
-1.73518080e-02 -8.53281319e-02 -2.59387016e-01 -3.11184704e-01
2.58078575e-01 -1.04946351e+00 -1.38481236e+00 2.60097206e-01
-7.20445454e-01 1.48619875e-01 -2.75759369e-01 4.06051189e-01
-2.37000167e-01 4.91344422e-01 -4.19272333e-01 -7.91149080e-01
-6.61952615e-01 -1.13890684e+00 -1.09253945e-02 5.05015135e-01
2.76144326e-01 -1.42625344e+00 -2.19702929e-01 -2.76447576e-03
7.45657384e-01 8.35816324e-01 9.13098633e-01 -4.06972557e-01
-8.37622523e-01 9.27350968e-02 1.85205460e-01 4.92046386e-01
4.68706936e-01 6.61990866e-02 -1.03049770e-01 -1.18822646e+00
1.97370678e-01 2.84693569e-01 -4.69064474e-01 3.69072318e-01
5.58631182e-01 -1.00634289e+00 -4.96419556e-02 5.48808753e-01
2.36853838e+00 3.69194418e-01 -1.04456186e-01 3.92205745e-01
5.28473593e-02 2.42928103e-01 2.65292019e-01 1.09184349e+00
7.29489267e-01 6.03659868e-01 7.89478004e-01 -1.13242239e-01
7.39390135e-01 7.88644910e-01 2.25539133e-01 8.70676517e-01
1.50884986e-01 -5.80118001e-01 -2.19807088e-01 7.47297108e-01
-1.86947453e+00 -6.43628418e-01 -6.81832060e-02 1.85348105e+00
6.07659698e-01 1.13926129e-02 1.08736254e-01 5.87842882e-01
5.22002816e-01 -2.33250141e-01 -5.59343517e-01 -5.65808952e-01
-5.69600523e-01 1.17055923e-01 7.24353850e-01 5.35089314e-01
-3.90061826e-01 -5.18529832e-01 5.22711611e+00 5.58970928e-01
-8.77469301e-01 9.57194716e-02 4.92913812e-01 -4.49124962e-01
-3.54824841e-01 -8.68217051e-02 -3.23495328e-01 7.75642991e-01
8.58347595e-01 -1.40236533e+00 5.95001757e-01 6.05053246e-01
1.10087442e+00 -3.79308164e-01 -8.91438305e-01 4.57322031e-01
-2.78518111e-01 -1.35561013e+00 -2.59350628e-01 3.46119225e-01
1.40710413e+00 -5.67213595e-02 -2.26882234e-01 -5.23976922e-01
1.89437166e-01 -2.57971108e-01 8.75744522e-01 1.42687127e-01
-4.54065353e-02 -1.46290374e+00 6.78394198e-01 2.41158172e-01
-1.23466873e+00 -4.69248533e-01 9.21967849e-02 -2.36122265e-01
9.75868464e-01 8.89461279e-01 -3.38884652e-01 1.21122301e+00
3.46773893e-01 3.41146260e-01 -7.85135180e-02 1.39963424e+00
-3.79452974e-01 2.38057286e-01 -5.19753754e-01 3.29925239e-01
2.58612454e-01 -7.84248292e-01 6.08384013e-01 4.38281268e-01
2.16828704e-01 2.21058503e-01 5.02491415e-01 7.64661789e-01
-7.59174749e-02 9.69546214e-02 -3.89575586e-02 3.24951380e-01
7.63502598e-01 1.60077429e+00 -5.25335014e-01 -2.63577521e-01
-5.07262766e-01 3.62410933e-01 -3.42734069e-01 8.11972618e-01
-6.22328818e-01 -4.10197347e-01 1.22160959e+00 -3.48088205e-01
1.33643404e-01 -6.64918244e-01 -6.00485981e-01 -1.11257553e+00
2.91399151e-01 -6.12136841e-01 5.66452384e-01 -6.32963955e-01
-1.47442520e+00 1.34648174e-01 1.06970504e-01 -1.18708909e+00
-7.83955514e-01 -1.14865020e-01 -1.09391618e+00 1.47661269e+00
-2.04430556e+00 -7.59641290e-01 1.62293583e-01 7.01295435e-01
3.30856711e-01 -1.11797936e-01 5.88750064e-01 6.18298352e-01
-1.09459293e+00 3.47477764e-01 6.52683914e-01 -4.89707142e-02
-3.10334265e-01 -1.25560355e+00 -4.09675330e-01 1.53593254e+00
-6.51086748e-01 3.11810732e-01 9.38570499e-01 -8.83459747e-02
-1.87270415e+00 -5.52067876e-01 5.04080355e-01 8.38499904e-01
1.15485096e+00 -1.96393907e-01 -4.22461718e-01 5.64697087e-01
1.36049592e+00 -7.07521886e-02 8.93938899e-01 -9.25295711e-01
5.05936742e-01 -4.19479758e-01 -1.62505627e+00 3.37941915e-01
6.81488961e-02 7.35880584e-02 -1.01765044e-01 5.74639618e-01
2.51501590e-01 -1.79398403e-01 -1.23429286e+00 3.20809156e-01
-1.16017550e-01 -1.77092209e-01 8.88321280e-01 -1.45229444e-01
-4.53827858e-01 -1.04440308e+00 -3.73796761e-01 -1.84972548e+00
-3.33741635e-01 -1.02883792e+00 -4.67295945e-01 1.29031718e+00
2.85626560e-01 -1.30188799e+00 3.79770160e-01 4.73237008e-01
-1.55778244e-01 -6.72345281e-01 -1.45318913e+00 -8.53733003e-01
6.14947826e-02 3.83336484e-01 9.05526519e-01 1.22368312e+00
7.22785711e-01 -9.69126001e-02 8.51041600e-02 9.28370655e-01
1.24339044e+00 4.24152166e-01 -1.54206842e-01 -7.61362433e-01
-3.29221010e-01 -3.43936026e-01 -3.81966412e-01 -1.87004358e-01
1.57653853e-01 -6.78058743e-01 -8.88219297e-01 -2.04711819e+00
-5.02086759e-01 -1.77397549e-01 -3.72818649e-01 4.47072595e-01
3.63287330e-01 2.06974640e-01 3.57653558e-01 -5.97354509e-02
-8.79985169e-02 5.90544403e-01 8.44230235e-01 -2.83426046e-01
4.53999527e-02 4.73567434e-02 -4.74658966e-01 4.21302885e-01
8.87813151e-01 1.88561127e-01 -7.81394780e-01 -4.64399368e-01
2.30106369e-01 6.48942471e-01 -3.29810023e-01 -7.17562616e-01
4.49877948e-01 -2.77447224e-01 3.14711988e-01 -8.66043985e-01
9.54488590e-02 -1.48542571e+00 5.66119373e-01 8.40158403e-01
2.67056733e-01 8.20585430e-01 1.92671269e-02 -5.28497472e-02
-1.99784204e-01 -6.81752563e-02 7.73403287e-01 3.78662199e-01
-4.44223255e-01 4.15184237e-02 -7.38771975e-01 -4.27566797e-01
1.57130873e+00 7.08404928e-02 -5.41737020e-01 -5.15469074e-01
-8.78206074e-01 1.37442660e+00 3.96492779e-01 -2.08361223e-02
1.14511952e-01 -1.18335140e+00 -7.52896249e-01 1.15057148e-01
-6.26604974e-01 -3.77620488e-01 7.69760832e-02 8.54469478e-01
-5.37831903e-01 5.74888945e-01 -4.17150348e-01 -2.90040106e-01
-7.53488064e-01 2.65532166e-01 6.09294653e-01 -5.97622469e-02
-1.49757400e-01 1.67307213e-01 -6.54203773e-01 3.43802840e-01
-5.39993532e-02 -2.35490337e-01 -1.53851807e-01 5.87014914e-01
3.64890665e-01 8.85417879e-01 4.19174492e-01 -3.06201726e-01
-4.42684174e-01 1.24451891e-02 4.59105164e-01 -1.39038647e-02
1.69389391e+00 -6.83449507e-01 -8.60801756e-01 -1.95924744e-01
1.04810119e+00 -8.59950632e-02 -8.91123354e-01 2.35490561e-01
-3.32279652e-01 -2.20145300e-01 2.38046378e-01 -8.28650117e-01
-1.76193619e+00 -6.47228435e-02 2.02595040e-01 6.42499804e-01
1.59605002e+00 -7.41289139e-01 2.78257996e-01 4.60186340e-02
7.68277586e-01 -9.50500846e-01 -7.97132850e-01 2.44612247e-03
5.38707197e-01 -4.56698835e-01 2.18628526e-01 -4.46824640e-01
8.25998485e-02 1.18966067e+00 3.80035043e-01 -2.74894983e-01
7.20680535e-01 5.68742812e-01 -4.94989842e-01 2.00042158e-01
-9.70219493e-01 2.65166879e-01 -1.54405743e-01 3.08344573e-01
5.08605242e-02 1.72972754e-01 -1.20042419e+00 2.44277999e-01
9.57179815e-02 -2.69033670e-01 1.28256214e+00 1.21082139e+00
-7.49521777e-02 -1.04866731e+00 -3.71070683e-01 3.40491652e-01
-5.01099706e-01 1.33347034e-01 4.62250471e-01 4.22849596e-01
1.88137367e-01 1.26226962e+00 1.43505603e-01 8.16418886e-01
2.64456093e-01 -2.40356833e-01 4.14221853e-01 -1.43062234e-01
-5.00061691e-01 2.68127799e-01 -3.18348370e-02 -4.26830985e-02
-2.26100489e-01 -6.39112830e-01 -1.42835176e+00 -4.62815732e-01
-4.84228730e-01 9.88273025e-01 8.32846105e-01 9.21293914e-01
-2.06046700e-02 5.72280943e-01 1.35595715e+00 -7.05796897e-01
-6.27605796e-01 -5.45092523e-01 -1.18299568e+00 -5.92146754e-01
3.03965598e-01 -2.23464414e-01 -6.66256368e-01 -3.30652118e-01] | [5.701694011688232, 2.587869167327881] |
0c1d2a33-e2e9-4a60-9e91-ef0c8943649f | a-robust-pose-transformational-gan-for-pose | 2001.01259 | null | https://arxiv.org/abs/2001.01259v1 | https://arxiv.org/pdf/2001.01259v1.pdf | A Robust Pose Transformational GAN for Pose Guided Person Image Synthesis | Generating photorealistic images of human subjects in any unseen pose have crucial applications in generating a complete appearance model of the subject. However, from a computer vision perspective, this task becomes significantly challenging due to the inability of modelling the data distribution conditioned on pose. Existing works use a complicated pose transformation model with various additional features such as foreground segmentation, human body parsing etc. to achieve robustness that leads to computational overhead. In this work, we propose a simple yet effective pose transformation GAN by utilizing the Residual Learning method without any additional feature learning to generate a given human image in any arbitrary pose. Using effective data augmentation techniques and cleverly tuning the model, we achieve robustness in terms of illumination, occlusion, distortion and scale. We present a detailed study, both qualitative and quantitative, to demonstrate the superiority of our model over the existing methods on two large datasets. | ['Deepak Mishra', 'Arnab Karmakar'] | 2020-01-05 | null | null | null | null | ['foreground-segmentation'] | ['computer-vision'] | [ 5.38833022e-01 1.88561931e-01 3.95667821e-01 -2.82583505e-01
-5.27732909e-01 -6.14706635e-01 5.71890056e-01 -4.94018644e-01
-1.26092374e-01 7.47901142e-01 -8.61596093e-02 1.36251673e-01
1.55550748e-01 -5.34126341e-01 -7.08769500e-01 -6.92349851e-01
5.57737589e-01 5.30051827e-01 2.57657349e-01 -1.92512408e-01
-6.38159737e-02 6.13742232e-01 -1.52109098e+00 -3.07554126e-01
7.03762650e-01 6.75888479e-01 -7.85536841e-02 5.82472801e-01
3.52088451e-01 2.38918260e-01 -7.11722076e-01 -7.42307305e-01
7.88758755e-01 -6.02227151e-01 -5.27975440e-01 5.94972849e-01
7.56879449e-01 -4.47728395e-01 -2.62569904e-01 8.43549967e-01
7.09586024e-01 1.61109164e-01 6.22845769e-01 -1.33102822e+00
-3.36786628e-01 -2.57544100e-01 -9.47726011e-01 -3.03954720e-01
5.16116381e-01 2.17897654e-01 4.95697051e-01 -4.79423344e-01
7.87348926e-01 1.21126425e+00 5.76064825e-01 8.12756419e-01
-1.22961211e+00 -5.75373530e-01 1.23625964e-01 -4.76176590e-02
-1.19176149e+00 -3.15603763e-01 1.12847495e+00 -3.79624635e-01
1.64985344e-01 4.33763117e-01 8.12942684e-01 1.17096949e+00
2.73295213e-02 5.67858875e-01 1.40628290e+00 -6.38416946e-01
9.75540280e-02 8.02372321e-02 -2.72671402e-01 7.58637726e-01
2.43516549e-01 4.63762246e-02 -2.18876317e-01 -1.33495852e-01
1.21525288e+00 1.21404463e-02 -1.84856176e-01 -8.54894161e-01
-9.84298527e-01 6.49033666e-01 3.33010852e-01 -1.73979014e-01
-1.71869695e-01 1.48219779e-01 7.02519715e-02 -1.05790710e-02
3.45449179e-01 3.25443089e-01 -4.39378977e-01 8.81203935e-02
-7.39145637e-01 3.00983071e-01 5.25384307e-01 1.15172434e+00
6.38507426e-01 4.39692736e-02 -1.37778476e-01 7.57814586e-01
2.63361514e-01 5.08402109e-01 1.97490096e-01 -1.07549036e+00
2.40412518e-01 6.38656259e-01 1.08707100e-01 -1.02953267e+00
-2.76737273e-01 -4.25216496e-01 -9.00252581e-01 3.51894617e-01
5.29811561e-01 -3.16984624e-01 -1.33519387e+00 1.90532994e+00
9.12370682e-01 1.06708221e-01 -1.75379321e-01 9.60870922e-01
7.77781367e-01 3.36316764e-01 -8.45138803e-02 -1.53581858e-01
1.43528616e+00 -9.03523922e-01 -6.88107014e-01 -3.27507585e-01
1.21341817e-01 -9.59993839e-01 1.09091675e+00 4.69369113e-01
-1.08944070e+00 -5.78464866e-01 -9.99681056e-01 -3.01461965e-01
-2.50421581e-03 4.11690682e-01 7.46207774e-01 7.70686269e-01
-8.33475769e-01 2.07359001e-01 -7.54402101e-01 -5.28651416e-01
2.63981313e-01 5.66864371e-01 -5.77162981e-01 1.11253457e-02
-7.68796444e-01 7.52315760e-01 -6.92326855e-03 1.80969626e-01
-6.94395900e-01 -4.26996738e-01 -1.02423954e+00 -3.88411731e-01
4.07896012e-01 -1.04626119e+00 1.08538401e+00 -8.83168340e-01
-1.83185625e+00 9.63463247e-01 -8.08317512e-02 -4.96680252e-02
1.03928030e+00 -1.79056197e-01 1.66099846e-01 4.87331450e-02
-2.80731082e-01 8.82696927e-01 1.00647235e+00 -1.37508166e+00
-1.93377838e-01 -5.63661695e-01 2.29222447e-01 4.24099267e-01
-1.58293754e-01 -6.12885691e-02 -8.54779363e-01 -7.26085901e-01
2.48756409e-01 -1.29883158e+00 -4.16617334e-01 1.90375537e-01
-4.01681572e-01 5.64645454e-02 8.24911773e-01 -8.95019710e-01
7.34753191e-01 -1.96030188e+00 3.03846985e-01 1.15493514e-01
1.45129457e-01 1.75006956e-01 3.83436196e-02 1.37983710e-01
8.72214511e-02 -8.29833969e-02 -2.13519499e-01 -5.63534379e-01
-7.81977773e-02 1.53919771e-01 -1.62167847e-02 5.71331978e-01
1.22574314e-01 8.04764688e-01 -5.10775208e-01 -7.33043909e-01
4.79922622e-01 8.15501273e-01 -6.20184004e-01 4.40573603e-01
-7.74825662e-02 9.07276273e-01 -3.84483069e-01 4.56620216e-01
7.59818494e-01 6.45034341e-03 8.44040811e-02 -3.77387136e-01
4.00621533e-01 -2.72246361e-01 -1.27927566e+00 1.71360600e+00
-4.41418469e-01 2.69597888e-01 9.91842970e-02 -7.09672570e-01
8.31043303e-01 3.10001194e-01 4.18597460e-01 -3.28471750e-01
3.69288951e-01 -1.62987895e-02 3.82960625e-02 -2.97643870e-01
1.58633932e-01 -2.35134229e-01 -4.89046378e-03 2.45861426e-01
-2.62021087e-02 -2.60201305e-01 1.22797273e-01 4.02885899e-02
7.74719238e-01 6.33835137e-01 2.02774882e-01 8.29528719e-02
4.50560004e-01 -2.27183312e-01 6.91281617e-01 2.64243901e-01
-8.56670961e-02 1.16231740e+00 2.28422344e-01 -4.28565860e-01
-1.27419245e+00 -9.17383313e-01 -4.51182993e-03 6.01240158e-01
1.66041806e-01 -1.79345250e-01 -1.00666869e+00 -6.94041610e-01
-2.42970854e-01 2.64677078e-01 -6.57227814e-01 -1.56594545e-01
-7.54604995e-01 -8.71058822e-01 2.77659327e-01 5.35369992e-01
6.20514810e-01 -9.59228098e-01 -6.51527345e-01 -2.03058198e-01
-2.13824719e-01 -1.32385242e+00 -6.33906543e-01 -3.81855220e-01
-8.70571911e-01 -9.48226750e-01 -8.79344523e-01 -7.62763441e-01
1.18838727e+00 2.52538938e-02 8.83556306e-01 1.14301234e-01
-7.78661788e-01 3.10979128e-01 -6.91462085e-02 -4.58997607e-01
-2.14797184e-01 -2.08023906e-01 -5.91948591e-02 1.24197491e-01
-8.12079683e-02 -6.40873790e-01 -8.83552194e-01 4.05872822e-01
-8.53084326e-01 3.06788653e-01 6.29178286e-01 7.20666230e-01
4.87655520e-01 3.15132900e-03 2.76386172e-01 -9.69649851e-01
3.03266913e-01 1.96976230e-01 -6.20153427e-01 2.65558153e-01
-1.75927848e-01 -1.12825982e-01 4.31344002e-01 -5.28049231e-01
-1.26830935e+00 6.09131932e-01 -9.19116661e-02 -4.12139356e-01
-3.03206086e-01 -1.79832414e-01 -5.33140421e-01 -2.73536325e-01
5.24735212e-01 1.96338385e-01 1.36808857e-01 -5.35659254e-01
4.72584605e-01 3.08200777e-01 5.62195837e-01 -6.75124526e-01
1.22508740e+00 5.65532804e-01 3.48503590e-01 -7.34218776e-01
-7.01912463e-01 -2.64085740e-01 -1.05215371e+00 -1.59494296e-01
7.72625089e-01 -8.42508078e-01 -6.96200550e-01 5.20110488e-01
-1.08728671e+00 -9.29411792e-04 -2.03908592e-01 3.73543710e-01
-6.13346159e-01 5.56359589e-01 -4.35021400e-01 -7.79593110e-01
-3.94446671e-01 -1.02597868e+00 1.40587902e+00 1.97463483e-01
-2.11978123e-01 -8.21673334e-01 -7.23005831e-02 7.53905177e-01
1.90899253e-01 7.36390412e-01 6.31954134e-01 -8.41875374e-02
-6.39729798e-01 -3.47079039e-01 2.93603744e-02 4.05100495e-01
2.91108549e-01 8.43365714e-02 -8.96127224e-01 -5.29594600e-01
9.78321284e-02 -3.51847112e-01 2.84528434e-01 2.68300235e-01
1.06954670e+00 -3.05376440e-01 -2.16114268e-01 7.17417419e-01
1.34734607e+00 1.05465010e-01 7.32811034e-01 2.82285698e-02
1.00660038e+00 7.03325629e-01 6.20737970e-01 3.37087631e-01
1.61274105e-01 8.00835967e-01 2.00971931e-01 -5.20778954e-01
-4.44506764e-01 -3.56396645e-01 5.00062965e-02 4.44652557e-01
-3.75134885e-01 -8.84159654e-02 -5.94779730e-01 2.78335780e-01
-1.58998585e+00 -6.50623739e-01 4.14645746e-02 2.49484563e+00
9.33042526e-01 -1.18224002e-01 2.18494639e-01 1.31987348e-01
7.13559031e-01 -2.32025713e-01 -4.02202815e-01 -5.08560762e-02
1.07240669e-01 2.88842529e-01 3.45889062e-01 4.15162176e-01
-1.03337240e+00 9.15199339e-01 6.51498604e+00 6.07295513e-01
-1.09447956e+00 -9.00109410e-02 7.07157731e-01 -1.83403101e-02
-1.06128991e-01 3.96206451e-04 -7.08458245e-01 2.64738470e-01
3.11872572e-01 1.03004180e-01 2.92689741e-01 7.24653602e-01
6.77149445e-02 -3.36745158e-02 -1.05348253e+00 1.20637107e+00
3.99198949e-01 -6.80689216e-01 1.21916980e-01 1.94382027e-01
7.21699417e-01 -7.54504502e-01 6.51998222e-02 6.62227273e-02
1.98354959e-01 -9.41996336e-01 5.94592392e-01 2.99886853e-01
7.29808033e-01 -6.57777429e-01 5.37951529e-01 3.99667948e-01
-8.08842957e-01 2.95122951e-01 -3.25226158e-01 -5.49211837e-02
2.77002454e-01 3.56588155e-01 -8.08160543e-01 6.28785074e-01
3.67782980e-01 2.35441953e-01 -8.77895117e-01 9.49935913e-01
-3.75785381e-01 2.88360566e-01 -5.38809240e-01 2.93552667e-01
-2.88660169e-01 -3.52960795e-01 4.13154632e-01 6.55745983e-01
3.13299149e-01 1.83751062e-01 2.44220510e-01 7.18093157e-01
1.03571843e-02 1.86154172e-01 -6.57576680e-01 3.52981538e-01
3.09370849e-02 1.58260429e+00 -7.56351709e-01 -1.72941685e-01
-1.37760252e-01 1.28854895e+00 6.26865029e-02 3.30222666e-01
-9.56685901e-01 -1.30557492e-01 3.11801195e-01 4.04736429e-01
4.36127707e-02 -3.14431459e-01 -2.02816620e-01 -1.19369578e+00
3.61238420e-01 -9.05720413e-01 7.39960149e-02 -7.53235221e-01
-1.03425944e+00 6.49062514e-01 1.62520349e-01 -1.29482555e+00
-3.42036426e-01 -5.11640608e-01 -4.46687907e-01 7.46817112e-01
-1.04700053e+00 -1.73358405e+00 -6.17931962e-01 6.43620431e-01
4.23186719e-01 6.22182451e-02 6.33050919e-01 3.28652948e-01
-5.91244638e-01 7.86745906e-01 -4.85062867e-01 3.19489576e-02
8.93928111e-01 -1.23524988e+00 3.40866923e-01 8.34842026e-01
7.66456267e-03 6.47361994e-01 9.21251297e-01 -7.02231228e-01
-1.31424069e+00 -8.97862554e-01 5.58696985e-01 -7.03812838e-01
6.34323284e-02 -6.75093710e-01 -4.26413387e-01 7.38681376e-01
3.08665186e-01 8.89755413e-02 4.71944511e-01 -5.63405454e-02
5.54494502e-04 -1.67795599e-01 -1.41082609e+00 8.99955988e-01
1.28187644e+00 -3.01161446e-02 -4.42977041e-01 3.10530663e-01
4.78325278e-01 -6.74624205e-01 -7.32081950e-01 7.63057947e-01
6.43280804e-01 -7.25465119e-01 9.83955383e-01 -5.25191724e-01
2.39098281e-01 -5.29071212e-01 1.81740791e-01 -1.10493958e+00
-2.09716842e-01 -7.69365668e-01 9.44542065e-02 1.37347698e+00
1.15832880e-01 -5.26178360e-01 1.06685138e+00 9.49533582e-01
2.73639679e-01 -6.29100144e-01 -7.24731028e-01 -6.33643627e-01
-1.30073830e-01 2.59864470e-03 4.32579339e-01 6.53268158e-01
-4.40594196e-01 4.71850395e-01 -8.76168013e-01 2.37605780e-01
8.04472983e-01 9.64640602e-02 1.32233989e+00 -1.26557505e+00
-3.50842178e-01 1.85955703e-01 -6.55010343e-01 -8.61783803e-01
-1.21643305e-01 -3.62332642e-01 6.23281524e-02 -1.42962646e+00
4.23873156e-01 -3.33723068e-01 1.27556145e-01 4.48426872e-01
-3.00675541e-01 7.04161465e-01 2.26570070e-01 9.47232097e-02
-3.07592630e-01 5.66029072e-01 1.74109340e+00 1.13769881e-01
-6.68442920e-02 1.37373030e-01 -7.02377260e-01 9.79226649e-01
7.70636380e-01 -2.46494070e-01 -8.24080706e-01 -4.52939093e-01
5.33877723e-02 -7.79544711e-02 5.66444397e-01 -9.30073678e-01
-8.09851810e-02 -1.03345215e-01 8.59457016e-01 -3.86204511e-01
6.82492733e-01 -9.01166379e-01 4.65654701e-01 3.26295674e-01
1.45592317e-02 1.02367818e-01 1.51813373e-01 5.12393236e-01
1.01233134e-02 -1.18943892e-01 9.31929111e-01 -1.74967244e-01
-3.03143173e-01 4.81948644e-01 2.10496485e-01 -1.01257786e-01
1.31101263e+00 -3.71789426e-01 5.17486706e-02 -4.99813557e-01
-8.53878140e-01 -8.15427378e-02 8.61961126e-01 4.42107797e-01
4.50440645e-01 -1.41958928e+00 -6.54041886e-01 4.51977342e-01
-2.49987300e-02 3.09493721e-01 2.94425756e-01 7.17293382e-01
-6.70288026e-01 1.25090748e-01 -3.82579148e-01 -6.19513154e-01
-1.61363745e+00 5.43433905e-01 2.36297563e-01 -2.05007538e-01
-5.36776423e-01 5.39730012e-01 6.24307275e-01 -6.06904209e-01
1.89694896e-01 -5.93934879e-02 1.15383742e-02 -3.16824585e-01
1.41500011e-01 2.54938155e-01 -3.08680143e-02 -7.88069785e-01
-2.63794750e-01 1.04414773e+00 -2.94226613e-02 -2.56496847e-01
1.03580427e+00 -1.05404951e-01 9.87878963e-02 -5.32634780e-02
1.00683045e+00 1.66089073e-01 -1.50054407e+00 7.74128712e-04
-5.68271041e-01 -7.82711148e-01 -3.38025302e-01 -7.41089046e-01
-1.29812932e+00 9.15814281e-01 6.84057057e-01 -4.26391780e-01
1.16710711e+00 -6.23586141e-02 7.27529168e-01 5.97976288e-03
5.55346310e-01 -1.03486884e+00 2.25793600e-01 -1.62048321e-02
1.08651507e+00 -1.25977194e+00 1.96462408e-01 -9.66772318e-01
-5.66792130e-01 8.27108800e-01 9.39603865e-01 -1.95277736e-01
4.54482108e-01 1.09853238e-01 2.26269826e-01 6.80523925e-03
-3.55207503e-01 -3.10235079e-02 5.26414216e-01 6.85404837e-01
4.53171045e-01 -1.57011345e-01 -4.62787062e-01 1.69585079e-01
-4.22065467e-01 -1.09973863e-01 3.70869309e-01 9.52525616e-01
3.40492427e-02 -1.47115457e+00 -4.87284333e-01 7.83671811e-02
-4.86469924e-01 3.16323042e-01 -5.60078263e-01 8.49943221e-01
2.00702563e-01 7.28818238e-01 -2.11902812e-01 -1.45295128e-01
4.44165856e-01 -5.75305670e-02 9.98120725e-01 -6.90330625e-01
-2.53098696e-01 4.33195889e-01 -8.68161768e-02 -2.86159694e-01
-5.16215682e-01 -6.96530640e-01 -9.44805264e-01 -1.41166717e-01
-3.46795321e-01 -2.81541765e-01 5.83667278e-01 7.88919747e-01
3.00968289e-01 4.26298648e-01 4.33448672e-01 -9.45747793e-01
-5.39402902e-01 -1.02093720e+00 -3.87882710e-01 9.47662294e-01
8.17140751e-03 -7.76190937e-01 -2.56230712e-01 3.68630081e-01] | [11.6555814743042, -0.9100406765937805] |
c367f1e8-bf54-4ca6-bded-9ed8dda9660f | learning-relationships-for-multi-view-3d | null | null | http://openaccess.thecvf.com/content_ICCV_2019/html/Yang_Learning_Relationships_for_Multi-View_3D_Object_Recognition_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Yang_Learning_Relationships_for_Multi-View_3D_Object_Recognition_ICCV_2019_paper.pdf | Learning Relationships for Multi-View 3D Object Recognition | Recognizing 3D object has attracted plenty of attention recently, and view-based methods have achieved best results until now. However, previous view-based methods ignore the region-to-region and view-to-view relationships between different view images, which are crucial for multi-view 3D object representation. To tackle this problem, we propose a Relation Network to effectively connect corresponding regions from different viewpoints, and therefore reinforce the information of individual view image. In addition, the Relation Network exploits the inter-relationships over a group of views, and integrates those views to obtain a discriminative 3D object representation. Systematic experiments conducted on ModelNet dataset demonstrate the effectiveness of our proposed methods for both 3D object recognition and retrieval tasks.
| [' Liwei Wang', 'Ze Yang'] | 2019-10-01 | null | null | null | iccv-2019-10 | ['3d-object-recognition'] | ['computer-vision'] | [-4.41290379e-01 -5.38828552e-01 -3.56745392e-01 -6.54956162e-01
-2.95074373e-01 -6.18826389e-01 5.89305222e-01 -1.83798194e-01
1.65481746e-01 9.03291926e-02 2.99221516e-01 2.97169298e-01
-3.28274071e-01 -6.47596240e-01 -3.57879132e-01 -4.86806571e-01
3.14452648e-01 1.78439662e-01 6.09548271e-01 2.90231705e-02
3.38930994e-01 1.02267051e+00 -1.30908823e+00 3.93792421e-01
1.76957756e-01 1.09519482e+00 3.42219442e-01 1.76724233e-02
-2.13022336e-01 3.12140763e-01 -3.81518155e-01 1.41869141e-02
3.64918023e-01 -1.53006345e-01 -4.69978631e-01 5.62021852e-01
5.70878625e-01 -5.38712502e-01 -6.12358510e-01 1.06665313e+00
6.00316346e-01 -3.45389731e-02 5.19323409e-01 -1.07500207e+00
-7.88738430e-01 3.84718068e-02 -9.75685656e-01 3.80095363e-01
4.63011771e-01 -1.68436736e-01 9.11178410e-01 -1.23464978e+00
7.35973597e-01 1.45542443e+00 4.56786677e-02 3.28732729e-01
-8.33224118e-01 -5.41097581e-01 6.48527324e-01 2.65470982e-01
-1.32059216e+00 -2.87127614e-01 1.22519779e+00 -1.69116661e-01
9.63360369e-01 -2.58022733e-02 1.05634105e+00 7.14097381e-01
9.38625708e-02 1.13065827e+00 1.29409361e+00 -5.91139346e-02
-2.94445693e-01 1.22768894e-01 -1.19557027e-02 6.34506404e-01
3.28171611e-01 -3.07465177e-02 -5.86844563e-01 -2.04734337e-02
1.10177290e+00 7.57746875e-01 -4.71383542e-01 -1.02441406e+00
-1.03967047e+00 5.33016443e-01 5.67923844e-01 3.39751840e-01
-3.33202302e-01 -2.71894932e-01 3.75374913e-01 3.29640619e-02
7.25831747e-01 -1.19474076e-01 -2.98268974e-01 2.18322471e-01
-5.73389530e-01 -1.31920651e-01 4.71227378e-01 1.18565881e+00
6.43816233e-01 -1.13975897e-01 2.31280357e-01 1.08399689e+00
5.90156436e-01 5.15385628e-01 7.91257769e-02 -5.68631649e-01
6.70514107e-01 1.09461880e+00 -2.22236887e-01 -1.43058801e+00
-1.40110001e-01 -4.38986599e-01 -9.85207200e-01 2.13960595e-02
-1.75662003e-02 6.35139346e-01 -8.64694118e-01 1.30078554e+00
5.65248966e-01 -2.41364405e-01 -1.89514518e-01 1.18694651e+00
1.05307090e+00 5.18308222e-01 -4.17017490e-01 -2.22537681e-01
1.07735872e+00 -8.94925714e-01 -4.14848298e-01 -1.85460299e-01
2.03147173e-01 -8.41407180e-01 5.69039106e-01 1.66706771e-01
-1.28515887e+00 -6.50952339e-01 -9.89678025e-01 -7.16245621e-02
-2.40220457e-01 -1.24926127e-01 4.98263121e-01 2.66033918e-01
-7.83205628e-01 -1.13180615e-02 -6.44320369e-01 -3.12233508e-01
6.10992551e-01 2.46407047e-01 -7.85332263e-01 -7.21368432e-01
-7.99157321e-01 8.28784108e-01 4.44093913e-01 3.07348341e-01
-7.86733687e-01 -3.74788016e-01 -6.65814757e-01 -1.18766492e-02
4.38779354e-01 -6.55134976e-01 7.02415645e-01 -4.40788180e-01
-1.00996149e+00 1.12389028e+00 -3.18369955e-01 2.72943795e-01
2.99717337e-01 -1.34150177e-01 -5.78750968e-01 4.93875861e-01
-1.11718953e-01 4.43401605e-01 6.48786843e-01 -1.76543939e+00
-5.10647953e-01 -9.36996043e-01 3.02142918e-01 5.00520587e-01
-2.69244283e-01 -6.52949587e-02 -9.75385070e-01 -2.38308728e-01
1.01811671e+00 -6.72982991e-01 1.12066343e-01 1.87791437e-01
-2.57239491e-01 -3.06366950e-01 1.33373654e+00 -3.85706663e-01
8.41694772e-01 -2.24066615e+00 2.57954001e-01 2.56656557e-01
3.31369311e-01 3.50876361e-01 -9.23037827e-02 4.46902871e-01
-1.77794561e-01 -1.66199505e-01 2.75286645e-01 -6.00376241e-02
-4.11517233e-01 3.41848880e-01 -1.26247957e-01 5.29905438e-01
-1.14737190e-01 8.06343496e-01 -7.50827253e-01 -7.72207081e-01
4.01993662e-01 4.80213523e-01 -4.51317042e-01 3.39055926e-01
7.12833926e-02 3.65519196e-01 -8.55981231e-01 8.73465896e-01
1.02411592e+00 -4.30349469e-01 -5.49790859e-02 -5.74596226e-01
3.73410136e-02 7.04373270e-02 -1.02061975e+00 1.97273731e+00
-3.63346964e-01 1.56892315e-01 3.56771126e-02 -1.01517248e+00
1.14357805e+00 1.52854800e-01 6.75912201e-01 -5.64281642e-01
-8.24001338e-03 7.85042644e-02 -2.44071007e-01 -6.05337679e-01
2.25540981e-01 -1.15433754e-02 1.92968652e-01 4.94302809e-01
-6.79419935e-02 -1.32541209e-01 -2.29123250e-01 3.19006681e-01
4.32224661e-01 1.55575842e-01 5.61782837e-01 2.24314868e-01
6.87433302e-01 -3.87376964e-01 4.62681383e-01 3.26172531e-01
-2.07698196e-01 8.51994514e-01 3.49424392e-01 -6.13950968e-01
-8.46846521e-01 -1.02997017e+00 -8.46190602e-02 3.86394620e-01
7.80240059e-01 -2.16516063e-01 -9.90573689e-02 -1.09852183e+00
1.34955257e-01 1.58707619e-01 -3.96760941e-01 -1.82114035e-01
-5.63245595e-01 -7.21929297e-02 7.39104822e-02 5.93325198e-01
7.91645110e-01 -6.09747827e-01 -4.73297030e-01 3.94972712e-02
-1.16565071e-01 -1.18149650e+00 -6.01836681e-01 -1.86314538e-01
-1.28134346e+00 -1.12793767e+00 -8.56592655e-01 -8.36382031e-01
1.03507996e+00 1.32247281e+00 1.03764272e+00 1.40263453e-01
-1.41062692e-01 7.11419880e-01 -2.99896181e-01 -1.74613193e-01
-6.17769593e-03 -2.29217991e-01 7.72948489e-02 3.26328874e-02
4.04980004e-01 -8.51127088e-01 -8.30517769e-01 7.97790885e-01
-9.65053678e-01 7.68984929e-02 7.38318384e-01 8.72039378e-01
8.75496745e-01 -1.63743928e-01 2.76408166e-01 -5.84547222e-01
1.86792742e-02 -1.43429086e-01 -5.87945580e-01 6.36513829e-01
-3.12641442e-01 -3.18493873e-01 2.27509320e-01 -2.58774281e-01
-1.00164557e+00 2.76717041e-02 9.69265327e-02 -1.11050677e+00
-3.53436947e-01 4.14433688e-01 -7.94344127e-01 -2.03116670e-01
-1.17710590e-01 5.40774703e-01 1.98695228e-01 -5.41690886e-01
3.65367740e-01 5.35433590e-01 -7.17617273e-02 -3.03183109e-01
6.56491280e-01 8.60500216e-01 4.42413464e-02 -5.54615676e-01
-1.21243429e+00 -7.89123237e-01 -1.01142156e+00 -5.00063241e-01
7.17059672e-01 -1.09471631e+00 -5.24641633e-01 3.70840162e-01
-1.36155701e+00 6.26412094e-01 -1.85776949e-02 6.39804244e-01
-2.65917420e-01 5.39194167e-01 -2.94589341e-01 -5.25321007e-01
-1.14812285e-01 -1.12433541e+00 1.10690403e+00 3.67085874e-01
3.40963632e-01 -8.62029850e-01 -2.58497030e-01 3.12929243e-01
3.76911759e-02 -9.12987292e-02 1.02839768e+00 -5.91133237e-01
-9.98042524e-01 -4.32346046e-01 -6.97974980e-01 3.63300323e-01
3.91953290e-01 -2.71307915e-01 -8.32845092e-01 -2.82834917e-01
2.70128697e-01 -1.54601529e-01 8.47386420e-01 3.32364470e-01
1.15766454e+00 1.16436183e-01 -6.12475336e-01 3.61382246e-01
1.36798608e+00 2.38193512e-01 3.76169056e-01 -8.89137536e-02
9.00399566e-01 5.93422115e-01 5.95770538e-01 3.27085227e-01
2.10247695e-01 7.83107698e-01 6.35890543e-01 -1.20671568e-02
-7.70533308e-02 -4.26385552e-01 -1.19761765e-01 1.33827293e+00
-2.08508477e-01 -2.18473583e-01 -5.09551823e-01 3.68074328e-01
-1.58981824e+00 -1.12204123e+00 8.69714990e-02 2.09266710e+00
1.31786704e-01 2.73794740e-01 5.43049276e-02 -3.09676260e-01
7.90748417e-01 6.27130985e-01 -6.82100892e-01 1.62349015e-01
-1.43894449e-01 -5.01982152e-01 1.74413212e-02 -1.64638132e-01
-8.60942483e-01 6.77574813e-01 6.11282110e+00 8.66859019e-01
-1.09813702e+00 2.63839420e-02 4.64182049e-01 -1.38511613e-01
-4.35345978e-01 1.35591358e-01 -8.88394535e-01 -1.15487225e-01
-3.33295345e-01 4.46253382e-02 6.87611429e-03 1.02710819e+00
-2.01180447e-02 -1.87008515e-01 -1.13076496e+00 1.25843215e+00
6.61116838e-01 -1.07308006e+00 5.58324397e-01 3.00076336e-01
7.88362682e-01 -1.45579189e-01 6.19283058e-02 -2.04965174e-02
-2.16952711e-01 -5.60380280e-01 4.06810373e-01 5.90738118e-01
4.03809726e-01 -7.32013464e-01 5.90205669e-01 4.89739031e-01
-1.50301611e+00 1.88941881e-01 -6.44423008e-01 2.50982940e-01
4.01938081e-01 8.48293483e-01 -3.91943753e-01 1.14468551e+00
7.28061199e-01 1.32535946e+00 -5.23335457e-01 1.13761342e+00
-1.15748532e-01 -2.31233835e-02 -1.84797809e-01 6.06388971e-03
2.83763111e-01 -3.57480347e-01 7.11612284e-01 5.28900146e-01
2.99649924e-01 2.98188329e-01 2.72921681e-01 7.76502430e-01
5.98374195e-02 -2.91163977e-02 -1.09763813e+00 1.81244954e-01
3.21953863e-01 1.17218900e+00 -8.73431563e-01 -3.66050214e-01
-8.43184769e-01 8.80147159e-01 4.01478291e-01 3.30051214e-01
-5.18292189e-01 -4.67650034e-02 3.30190778e-01 3.64591219e-02
6.46169543e-01 -5.46087086e-01 8.98734108e-03 -1.31014442e+00
2.34201461e-01 -6.57252133e-01 4.25870597e-01 -9.81789231e-01
-1.54886484e+00 3.99770379e-01 3.68481964e-01 -1.84981883e+00
3.16487074e-01 -3.89387429e-01 -2.70128369e-01 8.64092052e-01
-1.48648214e+00 -1.22885787e+00 -3.10007572e-01 5.88251412e-01
6.32103622e-01 -2.01278001e-01 4.47418988e-01 2.88460821e-01
-1.75735369e-01 1.34026602e-01 -8.46799612e-02 2.48746425e-02
7.22228765e-01 -7.32261777e-01 1.29936397e-01 4.26810175e-01
3.74332517e-01 9.04504538e-01 -7.27174953e-02 -5.33665895e-01
-1.61038518e+00 -8.46658230e-01 6.46601081e-01 -4.24182653e-01
5.57638258e-02 -1.24130748e-01 -1.05393624e+00 3.68996054e-01
1.98892385e-01 3.84916484e-01 5.52235603e-01 1.45769805e-01
-8.11838090e-01 -2.96570510e-01 -8.07918012e-01 3.77397805e-01
1.38170600e+00 -6.40924633e-01 -6.52498484e-01 1.46979898e-01
4.45408642e-01 -6.15282714e-01 -7.64048219e-01 6.40471697e-01
8.45712543e-01 -1.35957825e+00 1.35300648e+00 -6.32671773e-01
4.79701728e-01 -4.37269568e-01 -3.95493209e-01 -1.31610072e+00
-2.90558934e-01 2.27133512e-01 -2.22339109e-02 1.30919409e+00
-6.46704761e-03 -6.84891701e-01 6.73003495e-01 1.81563362e-01
-1.38297498e-01 -1.02739847e+00 -9.09831047e-01 -8.41520905e-01
-3.02443594e-01 -1.61380962e-01 5.29777348e-01 8.53258789e-01
-3.35426539e-01 4.16542828e-01 -3.51114392e-01 3.86971176e-01
5.22202849e-01 9.87347066e-01 9.29780900e-01 -1.16862822e+00
-2.44503826e-01 -4.78443086e-01 -8.01999152e-01 -1.65993154e+00
-5.09624323e-03 -1.01975620e+00 -2.47380823e-01 -1.81313288e+00
6.80543959e-01 -4.03685272e-01 -3.37033182e-01 1.29089534e-01
-8.94171372e-02 2.14312404e-01 2.91616023e-01 4.85671133e-01
-7.95670927e-01 6.99286819e-01 1.88619888e+00 -1.06628016e-01
-1.25379954e-02 4.13871631e-02 -4.73575950e-01 7.93939769e-01
4.03616428e-01 -3.81230623e-01 -5.47583222e-01 -5.54693758e-01
-1.42397597e-01 2.53296763e-01 5.34697652e-01 -6.65910244e-01
2.61153907e-01 -2.14700192e-01 8.24045122e-01 -1.36020887e+00
6.93084180e-01 -1.24269187e+00 1.07174747e-01 1.40989453e-01
-1.52950464e-02 1.46837413e-01 -8.97668153e-02 1.00553823e+00
-5.96519351e-01 4.02075090e-02 6.06427908e-01 -4.86929417e-01
-7.46022999e-01 8.07167113e-01 2.04194978e-01 3.77975926e-02
9.92911279e-01 -4.74750102e-01 -1.35600179e-01 -3.63030344e-01
-5.46337247e-01 2.38028347e-01 5.16619802e-01 6.26296580e-01
1.04769135e+00 -1.58618176e+00 -3.61314416e-01 4.59173977e-01
4.74198431e-01 2.63528317e-01 6.33376956e-01 8.14329386e-01
-1.63781971e-01 4.57493126e-01 -3.64003658e-01 -1.03953803e+00
-1.54820442e+00 6.69544339e-01 1.75446123e-01 -1.83053657e-01
-8.12517464e-01 7.00044036e-01 6.69469059e-01 -3.02551121e-01
2.24923983e-01 -4.20367010e-02 -3.05378258e-01 3.04049999e-02
2.70464808e-01 6.15648702e-02 -5.01894206e-02 -8.26270461e-01
-4.86030668e-01 1.14555573e+00 -5.16412437e-01 2.17646100e-02
1.38277292e+00 -3.07847768e-01 -1.30979165e-01 6.49031579e-01
1.49910402e+00 -1.19635880e-01 -1.03280759e+00 -5.63023210e-01
-4.89536792e-01 -1.03331947e+00 -5.77212945e-02 -4.01403427e-01
-1.36318970e+00 1.22635627e+00 3.92871559e-01 2.21152306e-01
1.11969674e+00 2.34668940e-01 7.09173799e-01 3.55369747e-01
7.17226386e-01 -6.67432308e-01 5.02776504e-01 4.41794723e-01
9.61669385e-01 -1.17751610e+00 4.83863413e-01 -7.19927967e-01
-4.53043073e-01 1.02553248e+00 8.37802172e-01 1.49465539e-02
8.93534660e-01 -4.23888087e-01 -4.24674712e-02 -5.00043511e-01
-5.30808926e-01 -7.40392879e-02 5.21495998e-01 5.71708858e-01
1.44973218e-01 -2.62848377e-01 -3.48465070e-02 1.98274195e-01
6.69638634e-01 -2.46044010e-01 -4.73227724e-02 1.07030988e+00
-3.22845161e-01 -1.11093855e+00 -2.75804996e-01 3.50019604e-01
-8.80974978e-02 2.26468712e-01 -4.71032262e-01 9.72966969e-01
-1.84937909e-01 7.17626750e-01 -6.51123971e-02 -3.17315996e-01
6.03602588e-01 -6.40110821e-02 7.31032610e-01 -5.23355722e-01
-2.03634009e-01 5.08722961e-01 -2.99009562e-01 -4.54775155e-01
-1.08512366e+00 -6.08544588e-01 -8.56317759e-01 2.32450701e-02
-6.62400246e-01 -5.77079877e-02 3.87397438e-01 8.60846758e-01
5.46839535e-01 2.14771509e-01 1.06323051e+00 -9.33553934e-01
-3.89414251e-01 -6.16439104e-01 -8.28757405e-01 3.65721941e-01
3.91988866e-02 -9.51857924e-01 -3.10410947e-01 -2.30504468e-01] | [8.141435623168945, -3.819246530532837] |
149c28af-3ca4-4593-a1a6-a2f06a7a40c1 | prediction-of-sea-surface-temperature-using | 1705.06861 | null | http://arxiv.org/abs/1705.06861v1 | http://arxiv.org/pdf/1705.06861v1.pdf | Prediction of Sea Surface Temperature using Long Short-Term Memory | This letter adopts long short-term memory(LSTM) to predict sea surface
temperature(SST), which is the first attempt, to our knowledge, to use
recurrent neural network to solve the problem of SST prediction, and to make
one week and one month daily prediction. We formulate the SST prediction
problem as a time series regression problem. LSTM is a special kind of
recurrent neural network, which introduces gate mechanism into vanilla RNN to
prevent the vanished or exploding gradient problem. It has strong ability to
model the temporal relationship of time series data and can handle the
long-term dependency problem well. The proposed network architecture is
composed of two kinds of layers: LSTM layer and full-connected dense layer.
LSTM layer is utilized to model the time series relationship. Full-connected
layer is utilized to map the output of LSTM layer to a final prediction. We
explore the optimal setting of this architecture by experiments and report the
accuracy of coastal seas of China to confirm the effectiveness of the proposed
method. In addition, we also show its online updated characteristics. | ['Junyu Dong', 'Qin Zhang', 'Hui Wang', 'Guoqiang Zhong', 'Xin Sun'] | 2017-05-19 | null | null | null | null | ['time-series-regression'] | ['time-series'] | [-1.89286843e-01 -5.32304525e-01 7.83736911e-03 -4.03707504e-01
-2.02206075e-01 -2.83437580e-01 4.88762259e-01 -6.23720348e-01
-4.14503366e-01 6.24229074e-01 1.46070898e-01 -9.11340892e-01
2.92746305e-01 -7.01956570e-01 -6.38224125e-01 -1.03652096e+00
-2.97770768e-01 -3.56677145e-01 6.64765909e-02 -4.83963132e-01
1.38472795e-01 3.60655695e-01 -1.20527327e+00 2.02587731e-02
9.54394460e-01 1.25780261e+00 4.14350688e-01 5.16754925e-01
-3.02824974e-01 9.89349782e-01 -3.04986566e-01 2.13128924e-01
1.08275369e-01 -2.21485019e-01 -5.66549540e-01 -5.02265215e-01
-3.64059448e-01 -2.69453615e-01 -3.94289136e-01 7.22953439e-01
5.54601908e-01 3.08987945e-01 2.80140042e-01 -6.20405257e-01
-4.34783608e-01 6.47937000e-01 -5.73546112e-01 7.56053865e-01
-3.18925291e-01 -3.33380774e-02 5.61232448e-01 -7.84369707e-01
4.64739688e-02 8.83786023e-01 9.62004900e-01 3.91575426e-01
-4.67734963e-01 -7.99117506e-01 4.64153409e-01 9.27191079e-02
-1.05332172e+00 -3.32120270e-01 7.22395122e-01 -1.55050635e-01
1.19276083e+00 2.48539552e-01 7.80629754e-01 9.67209756e-01
7.23440468e-01 8.03691924e-01 1.10762846e+00 -1.65617585e-01
-9.00699049e-02 -2.27801800e-01 2.89772272e-01 2.69462079e-01
-3.56308252e-01 3.20358425e-01 -2.61548966e-01 1.07508242e-01
6.91613674e-01 3.97101015e-01 -2.24102840e-01 6.78043842e-01
-8.74287486e-01 5.80514073e-01 7.81612396e-01 5.60488820e-01
-4.84383583e-01 1.16147369e-01 4.24793333e-01 6.20292842e-01
9.23324823e-01 2.82570392e-01 -1.00495839e+00 -1.74937934e-01
-8.72035742e-01 -2.74847150e-01 6.12153649e-01 6.06878459e-01
3.77743840e-01 9.39278901e-01 1.65969387e-01 9.45688009e-01
3.34421813e-01 7.77577460e-01 1.16679835e+00 -8.05117115e-02
4.25534964e-01 2.30781376e-01 -4.56453413e-02 -8.36814940e-01
-7.22388208e-01 -7.43340135e-01 -1.21826792e+00 -3.65527660e-01
-1.29082620e-01 -8.62839222e-01 -1.12301254e+00 1.47843587e+00
1.15644865e-01 8.76964509e-01 4.09502029e-01 9.99901116e-01
8.40706170e-01 1.51565623e+00 1.50013797e-03 -5.04166126e-01
8.95654082e-01 -1.09298778e+00 -8.60058784e-01 -3.95310849e-01
7.89365828e-01 -6.19532228e-01 5.55215120e-01 -6.86143115e-02
-7.02714741e-01 -5.01667798e-01 -8.84208858e-01 7.10298959e-03
-7.83498287e-01 1.85533375e-01 5.86734116e-01 3.63562964e-02
-1.12575579e+00 8.25846314e-01 -1.18799591e+00 -3.97244990e-01
-3.82947117e-01 5.01320541e-01 -2.25212518e-03 4.84153926e-01
-1.82860887e+00 1.09371197e+00 2.35080257e-01 1.28784716e+00
-5.26324034e-01 -5.40247321e-01 -8.50098848e-01 -2.01850999e-02
-3.27119678e-01 -2.57380486e-01 1.35948360e+00 -1.20141661e+00
-1.85137916e+00 1.77080095e-01 -5.39628327e-01 -5.96533239e-01
4.94756475e-02 -2.26694837e-01 -9.13615167e-01 -5.91830850e-01
-4.47914720e-01 -7.21880049e-03 6.75510347e-01 -5.20181239e-01
-6.25221789e-01 -3.13085526e-01 -6.90598011e-01 1.09084710e-01
-4.99197394e-01 3.46157178e-02 -1.04922988e-01 -7.72621274e-01
4.45166141e-01 -9.33116496e-01 -6.73974335e-01 -7.39956260e-01
-4.25832979e-02 -2.21777633e-01 1.04545891e+00 -9.30452287e-01
1.47211504e+00 -2.26159072e+00 -2.28040278e-01 1.26124516e-01
-4.33835924e-01 3.99451941e-01 -2.27114692e-01 4.50225025e-01
-2.17225775e-01 1.27421319e-01 -1.22267164e-01 -3.18160683e-01
-3.45302284e-01 4.76090193e-01 -1.03430891e+00 5.74206054e-01
1.47559941e-01 9.11681116e-01 -8.13608825e-01 -1.00664765e-01
1.62694812e-01 6.80558264e-01 4.23188180e-01 3.76330644e-01
-8.20715204e-02 4.69388962e-01 -6.54449999e-01 2.94377506e-01
8.42109919e-01 -1.20688200e-01 9.04961750e-02 1.29758403e-01
-8.94996464e-01 6.29215002e-01 -5.53173900e-01 1.25940871e+00
-7.47562110e-01 7.60034800e-01 -2.97586739e-01 -6.04014754e-01
1.42086935e+00 5.91291666e-01 1.85837075e-01 -1.20147133e+00
2.15146303e-01 3.93169701e-01 -2.74434417e-01 -6.38640523e-01
6.75563276e-01 -3.36092234e-01 1.32562621e-02 4.03909683e-01
-3.75212044e-01 1.45792261e-01 -5.40670931e-01 -4.76226538e-01
3.79702181e-01 2.95142680e-01 -1.83538228e-01 -2.69227773e-01
4.27206457e-01 -1.60034761e-01 9.19102430e-01 5.05277395e-01
1.06948867e-01 4.26161021e-01 3.07933837e-01 -1.07173610e+00
-9.17351842e-01 -3.06880772e-01 -2.00616434e-01 1.13468182e+00
-6.48944899e-02 9.45333168e-02 -4.10806760e-02 -1.61672279e-01
-2.66403437e-01 6.71939492e-01 -7.54328310e-01 -1.02954723e-01
-7.44775236e-01 -1.04646277e+00 5.03945112e-01 6.16621315e-01
7.62988865e-01 -1.22663033e+00 -2.88257897e-01 4.56685513e-01
-5.28854132e-02 -7.50040412e-01 -3.06497276e-01 5.26018620e-01
-1.39563608e+00 -5.43576181e-01 -8.66402745e-01 -1.01474929e+00
4.04145807e-01 2.69589126e-01 7.96921670e-01 -6.48644045e-02
6.11395657e-01 -5.64858615e-01 -3.40581030e-01 -4.14425910e-01
5.31315245e-02 3.72307211e-01 -1.31039564e-02 3.26868184e-02
2.71144807e-01 -6.62755370e-01 -7.26473212e-01 3.96595687e-01
-7.79646218e-01 6.01127818e-02 3.76693368e-01 8.13420057e-01
4.38218921e-01 1.96666066e-02 4.54285830e-01 -5.05325019e-01
4.43588406e-01 -8.55762661e-01 -9.03064489e-01 2.07401156e-01
-6.05901361e-01 8.80244747e-02 1.09648740e+00 -2.68904865e-01
-1.08913565e+00 -1.89192723e-02 -3.41862917e-01 -4.68289286e-01
5.16182005e-01 1.15922821e+00 5.78192711e-01 -8.32460150e-02
-8.10662135e-02 9.04394567e-01 -7.90834576e-02 -8.69173944e-01
-8.17134529e-02 8.49265516e-01 2.85467416e-01 3.67145007e-03
3.54614466e-01 1.57521293e-01 -2.69140035e-01 -6.68432236e-01
-8.74946415e-01 -2.47402132e-01 -5.64865649e-01 -1.95977628e-01
5.87292314e-01 -1.16097498e+00 -6.77340865e-01 1.01622427e+00
-1.01830113e+00 -6.64144218e-01 2.45420322e-01 6.01986229e-01
7.45813027e-02 -9.60295349e-02 -9.91207838e-01 -8.25231612e-01
-8.66305888e-01 -6.64909184e-01 6.53667510e-01 5.08680046e-01
5.08655787e-01 -1.45780754e+00 3.23550880e-01 -4.56334293e-01
9.09340382e-01 2.18069896e-01 3.88980567e-01 -3.83572131e-01
-4.28018570e-02 -2.80303448e-01 3.47652175e-02 4.29135799e-01
1.77032545e-01 2.58530706e-01 -1.04365206e+00 -4.55745012e-01
3.43555331e-01 1.34470955e-01 1.23479068e+00 6.06026232e-01
1.16195452e+00 -3.50014299e-01 -1.64580107e-01 1.04838383e+00
1.40918136e+00 4.21735764e-01 7.59603977e-01 6.10689342e-01
7.95138776e-01 2.19555900e-01 4.32785481e-01 3.19187582e-01
6.45071149e-01 1.76164463e-01 3.06289822e-01 -4.58763659e-01
4.73682910e-01 -2.33601585e-01 6.31768286e-01 1.65168154e+00
-1.51710346e-01 -1.58758968e-01 -9.75963831e-01 5.55791557e-01
-1.87598681e+00 -8.75232756e-01 -4.23220664e-01 1.99788558e+00
5.93196630e-01 1.19049057e-01 -3.33151937e-01 -1.46419495e-01
5.05929828e-01 7.16277063e-01 -5.31983495e-01 -9.16316926e-01
-2.29761109e-01 5.16144000e-02 8.03113163e-01 2.40840927e-01
-1.24949205e+00 1.23087263e+00 6.87090921e+00 4.51947540e-01
-2.14122295e+00 -4.25934270e-02 7.52190769e-01 1.51775256e-01
-1.02878615e-01 -1.83896884e-01 -7.94725955e-01 7.26659238e-01
1.54543209e+00 1.75352413e-02 1.40006199e-01 6.18290424e-01
8.41278613e-01 2.95771837e-01 -4.11954820e-01 5.20033062e-01
-3.76314223e-01 -1.09128213e+00 -9.34291780e-02 -3.01190138e-01
8.37446451e-01 6.88962817e-01 2.84457803e-01 5.16111672e-01
3.01054657e-01 -1.03522050e+00 3.40719372e-01 8.53911638e-01
6.19167328e-01 -7.60423362e-01 1.15197766e+00 3.57569695e-01
-1.60712230e+00 -2.58252472e-01 -5.23600936e-01 -5.66658616e-01
-1.06759377e-01 4.50154960e-01 -5.44739246e-01 6.39032900e-01
8.46541166e-01 1.42352045e+00 -2.36083284e-01 1.11665940e+00
-1.11893177e-01 1.00564134e+00 -4.99267161e-01 -2.30477542e-01
8.45620573e-01 -2.40909874e-01 9.32872817e-02 1.20018351e+00
7.51644135e-01 3.08962494e-01 -1.21029370e-01 1.25534639e-01
1.43397480e-01 1.84134841e-02 -5.87472975e-01 -5.48964813e-02
3.21815670e-01 9.34065580e-01 -2.24089444e-01 -1.96881533e-01
-4.74725574e-01 6.33910894e-01 1.76774159e-01 7.08179653e-01
-6.73653185e-01 -4.40340817e-01 5.08597314e-01 -3.49846214e-01
6.76384985e-01 -4.23535258e-01 -4.01601374e-01 -1.24264991e+00
-1.97501332e-02 -4.46616173e-01 4.53103721e-01 -1.09716594e+00
-1.10583711e+00 9.55166519e-01 -7.44834840e-01 -1.54594946e+00
-2.70190060e-01 -4.74112451e-01 -1.17588854e+00 1.38720882e+00
-2.19218421e+00 -1.06133902e+00 -2.11874302e-02 5.80664337e-01
5.40393472e-01 1.47808596e-01 8.10888886e-01 1.40110388e-01
-1.10712612e+00 3.53020161e-01 8.77907336e-01 2.15111539e-01
2.33911142e-01 -8.84382844e-01 8.18948925e-01 1.02404547e+00
-6.55422390e-01 6.34954751e-01 5.96489191e-01 -6.27061486e-01
-1.56369317e+00 -1.38994038e+00 1.17103517e+00 3.01854700e-01
9.54491794e-01 -1.54489949e-01 -1.20457685e+00 9.24895525e-01
1.88656956e-01 2.41405115e-01 3.47687036e-01 1.37972385e-01
-8.06848407e-02 -5.10830164e-01 -5.64640820e-01 3.51073235e-01
3.82067502e-01 -5.39563656e-01 -5.35044789e-01 2.39895761e-01
9.35231209e-01 -6.77852988e-01 -8.70406449e-01 5.08019984e-01
5.79900503e-01 -4.71782506e-01 3.93424749e-01 -5.63694596e-01
4.03410167e-01 -2.66389757e-01 9.38089266e-02 -1.51641929e+00
-2.82931417e-01 -6.66814923e-01 -7.46533275e-02 8.93657148e-01
7.75195777e-01 -1.07485282e+00 4.08400685e-01 4.63635296e-01
-5.13953805e-01 -9.03744578e-01 -9.48043585e-01 -6.01297975e-01
3.04471284e-01 -3.47115576e-01 9.29999173e-01 1.16558254e+00
-6.93583265e-02 1.10729346e-02 -9.68868136e-01 4.26343739e-01
-4.16378267e-02 5.14200151e-01 2.50142604e-01 -9.05731738e-01
2.95272619e-01 -2.89753169e-01 5.41480631e-02 -1.39782655e+00
1.50148734e-01 -4.05165255e-01 2.22874403e-01 -1.47679627e+00
-3.63726497e-01 -3.94294977e-01 -1.03123176e+00 5.58087111e-01
-1.88848034e-01 -1.12625800e-01 -1.87311411e-01 2.91719913e-01
-2.28055045e-01 8.84900689e-01 1.38756633e+00 7.04985186e-02
-5.86675465e-01 4.03598815e-01 -2.21288621e-01 3.41662854e-01
9.54483271e-01 -2.94752657e-01 -3.55556458e-02 -9.25348818e-01
3.67308438e-01 5.02498686e-01 3.64424884e-02 -7.13249445e-01
5.42448282e-01 -2.65367895e-01 5.29006600e-01 -8.34147394e-01
7.08816797e-02 -7.60931611e-01 1.21337608e-01 4.90484238e-01
-2.13137999e-01 4.67769146e-01 5.37372708e-01 2.44894072e-01
-5.63819885e-01 2.96969235e-01 4.44024712e-01 -7.95412734e-02
-1.09196699e+00 5.88737845e-01 -6.65680468e-01 -5.56318939e-01
4.38221008e-01 -3.68917041e-04 -2.80882627e-01 -2.27493986e-01
-6.56586111e-01 7.98925102e-01 -1.30505055e-01 6.48694575e-01
5.84559560e-01 -1.14177382e+00 -5.54891050e-01 5.21049201e-01
-3.39533865e-01 -1.19065315e-01 5.28696477e-01 8.83556902e-01
-5.68720400e-01 5.26707768e-01 4.38759848e-02 -4.12515283e-01
-7.87413538e-01 2.72918582e-01 9.63747621e-01 -1.54279590e-01
-8.54247630e-01 8.13145816e-01 -1.58633500e-01 -5.45754433e-01
1.19433582e-01 -5.96662581e-01 -6.69649601e-01 -7.06494302e-02
6.01355195e-01 2.13897638e-02 7.02996254e-02 -4.64072615e-01
-3.20519149e-01 7.87593067e-01 -2.26357523e-02 1.63423419e-01
1.74133480e+00 -3.20240468e-01 -4.48910475e-01 9.56518173e-01
1.26618004e+00 -6.22370005e-01 -1.24123573e+00 -3.79840195e-01
-1.58976819e-02 2.22555920e-02 5.57616413e-01 -6.51898980e-01
-1.44725800e+00 9.68355119e-01 6.21267736e-01 2.92438060e-01
1.28814697e+00 -7.19596028e-01 1.31569469e+00 4.44861352e-01
6.22156858e-02 -9.33830976e-01 -6.68938577e-01 1.29358160e+00
6.13083303e-01 -1.07588923e+00 -3.62453192e-01 4.57252383e-01
-4.56662446e-01 1.46057761e+00 5.22762775e-01 -1.42245471e-01
1.04558837e+00 2.55193204e-01 4.78855699e-01 1.02530271e-01
-1.14786375e+00 -1.15173839e-01 1.14854515e-01 -1.89721167e-01
5.37164986e-01 1.50692657e-01 -1.96455613e-01 4.06743556e-01
-3.41452956e-01 1.37084201e-01 4.03278410e-01 9.45097625e-01
-4.33257848e-01 -6.02592111e-01 -2.25146532e-01 4.06874955e-01
-5.42652369e-01 -4.03411806e-01 1.48621663e-01 4.07514066e-01
-4.48657513e-01 1.02183759e+00 4.99626756e-01 -7.51216710e-01
1.81567073e-01 1.76360495e-02 -4.10792828e-01 -7.66130537e-02
-9.87218618e-01 1.14249237e-01 -2.17601597e-01 -3.06568980e-01
-2.51062274e-01 -4.63519186e-01 -1.30693507e+00 -6.11030757e-01
-3.28508437e-01 5.45694172e-01 9.08767760e-01 1.11459005e+00
3.13897341e-01 5.77307224e-01 1.40232813e+00 -7.29191661e-01
-2.79859364e-01 -1.17205918e+00 -8.20869386e-01 -2.46595025e-01
1.03320098e+00 8.82432214e-04 -5.39086521e-01 -2.74023771e-01] | [6.634230136871338, 2.8899483680725098] |
a9288eaf-f012-4b1f-b11f-4b145d8672e7 | pre-trained-language-meaning-models-for | 2306.00124 | null | https://arxiv.org/abs/2306.00124v1 | https://arxiv.org/pdf/2306.00124v1.pdf | Pre-Trained Language-Meaning Models for Multilingual Parsing and Generation | Pre-trained language models (PLMs) have achieved great success in NLP and have recently been used for tasks in computational semantics. However, these tasks do not fully benefit from PLMs since meaning representations are not explicitly included in the pre-training stage. We introduce multilingual pre-trained language-meaning models based on Discourse Representation Structures (DRSs), including meaning representations besides natural language texts in the same model, and design a new strategy to reduce the gap between the pre-training and fine-tuning objectives. Since DRSs are language neutral, cross-lingual transfer learning is adopted to further improve the performance of non-English tasks. Automatic evaluation results show that our approach achieves the best performance on both the multilingual DRS parsing and DRS-to-text generation tasks. Correlation analysis between automatic metrics and human judgements on the generation task further validates the effectiveness of our model. Human inspection reveals that out-of-vocabulary tokens are the main cause of erroneous results. | ['Johan Bos', 'Malvina Nissim', 'Huiyuan Lai', 'Chunliu Wang'] | 2023-05-31 | null | null | null | null | ['cross-lingual-transfer', 'drs-parsing'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.30395377e-01 8.66256535e-01 -2.13583454e-01 -5.44642687e-01
-1.13277614e+00 -5.70052266e-01 9.88465130e-01 3.40157807e-01
-5.71924567e-01 8.99554908e-01 7.72865772e-01 -5.27367234e-01
2.03045696e-01 -7.83843219e-01 -6.76375985e-01 -3.66127521e-01
4.73454595e-01 4.78662640e-01 1.52578160e-01 -6.22751057e-01
2.46920735e-01 -1.26787588e-01 -1.16762877e+00 7.88108051e-01
1.16239834e+00 3.81160200e-01 5.18429518e-01 3.09815913e-01
-7.37118959e-01 1.12769485e+00 -8.55243087e-01 -4.99555618e-01
-2.86529034e-01 -5.30301690e-01 -1.12537324e+00 -2.18379557e-01
2.29354147e-02 6.72770441e-02 3.42827946e-01 9.09540892e-01
4.96214628e-01 1.25425253e-02 7.39956081e-01 -6.52724922e-01
-8.43103588e-01 1.32864320e+00 -2.85534710e-01 -8.78846422e-02
5.61476767e-01 -1.92497820e-01 1.15197039e+00 -8.32274556e-01
9.17942047e-01 1.73447645e+00 4.57874447e-01 6.90524459e-01
-1.30354798e+00 -3.43487203e-01 2.33591080e-01 1.16787799e-01
-1.13174713e+00 -4.92634863e-01 7.05796957e-01 -3.79386812e-01
1.40756023e+00 -4.82271761e-02 2.19378263e-01 1.23408246e+00
-4.92204651e-02 6.85520351e-01 1.06330431e+00 -9.94399309e-01
1.10780716e-01 4.44332898e-01 -8.31359923e-02 2.83875883e-01
3.22667927e-01 -2.80700505e-01 -4.20568764e-01 6.20134622e-02
6.03888333e-01 -8.31975698e-01 -1.19434796e-01 -2.22718969e-01
-1.34483552e+00 1.25763381e+00 1.66802332e-01 8.77152205e-01
-3.42655420e-01 4.05961291e-05 6.97119474e-01 4.64657634e-01
6.42478168e-01 6.91621840e-01 -6.21205986e-01 9.22821164e-02
-5.70907712e-01 2.85576731e-01 6.34409428e-01 7.81907022e-01
5.40765524e-01 1.12738684e-01 -3.91313553e-01 1.19949663e+00
3.44308496e-01 5.30519664e-01 9.08153713e-01 -8.48131716e-01
6.95926964e-01 9.19254243e-01 3.95123921e-02 -6.89679384e-01
-1.75132334e-01 -2.22987369e-01 -4.02156830e-01 -2.96785355e-01
4.11697656e-01 -1.39996141e-01 -5.62556744e-01 1.93294203e+00
3.01613569e-01 -4.84979242e-01 6.76124752e-01 6.50433064e-01
9.37705219e-01 7.54923940e-01 6.13228142e-01 -2.53071189e-01
1.46281326e+00 -7.35410690e-01 -8.38048577e-01 -4.31029499e-01
1.22783422e+00 -7.40567148e-01 1.49877501e+00 2.46207379e-02
-1.13423753e+00 -5.66214323e-01 -8.94338429e-01 -3.15311849e-01
-4.37178910e-01 4.58510429e-01 5.03957391e-01 4.95053679e-01
-8.97782147e-01 3.34329635e-01 -4.44752991e-01 -4.99301821e-01
-8.54998976e-02 6.89192787e-02 -3.34542155e-01 -1.03811480e-01
-1.59614289e+00 1.26779270e+00 7.09667206e-01 -7.62270987e-02
-5.15770555e-01 -2.82745957e-01 -1.14103246e+00 -2.14498773e-01
4.00031120e-01 -5.14607310e-01 1.36443663e+00 -1.19879568e+00
-1.50603199e+00 1.24968815e+00 -1.19648397e-01 -4.58891004e-01
2.68364519e-01 -3.59769166e-01 -3.56001318e-01 -6.94128349e-02
4.98740464e-01 7.66298652e-01 6.55225933e-01 -1.20059574e+00
-5.39994359e-01 -5.88613562e-02 2.12341398e-01 3.55047911e-01
-2.00804487e-01 1.15663424e-01 2.47188024e-02 -8.25432003e-01
-1.07174374e-01 -7.58468330e-01 -8.68136585e-02 -7.98473537e-01
-3.08755875e-01 -7.51409292e-01 3.21572870e-01 -7.73148775e-01
1.41428161e+00 -1.85126042e+00 9.86853689e-02 -1.27692088e-01
-4.62361842e-01 3.12964797e-01 -2.42559552e-01 5.28526247e-01
-2.58418899e-02 2.58036733e-01 -3.20287257e-01 -2.94971377e-01
8.80298391e-02 5.82649946e-01 -4.82477725e-01 -1.40211895e-01
5.32239735e-01 9.79747176e-01 -9.58252311e-01 -5.52312493e-01
1.85136914e-01 3.43600780e-01 -4.04006392e-01 3.00344497e-01
-5.64591229e-01 5.66842437e-01 -5.20406544e-01 9.40254480e-02
2.05098808e-01 -5.44140190e-02 5.84408283e-01 8.14547762e-02
-1.55163750e-01 1.17937756e+00 -9.02092695e-01 1.88973355e+00
-8.29932630e-01 2.60524929e-01 -2.77120173e-01 -1.13500345e+00
7.90430546e-01 5.79145193e-01 -9.52612683e-02 -8.11518073e-01
7.54794180e-02 4.09635842e-01 2.85865426e-01 -5.61255693e-01
8.21691871e-01 -4.77587730e-01 -4.62808430e-01 4.59388226e-01
1.62782043e-01 -1.91842914e-01 2.63304144e-01 1.25162303e-01
6.55554831e-01 5.27538478e-01 7.56373048e-01 -4.96523082e-01
9.54050779e-01 3.41177255e-01 3.74416053e-01 4.61558163e-01
2.93981522e-01 5.15413880e-01 5.47062695e-01 -1.36279464e-01
-8.18437636e-01 -7.39318132e-01 -9.23732594e-02 1.36059809e+00
-5.69947101e-02 -4.23893064e-01 -1.08581495e+00 -8.21970999e-01
-2.86430001e-01 1.51555192e+00 -5.50572753e-01 8.18348899e-02
-8.11930537e-01 -5.47087789e-01 5.57126939e-01 3.78224194e-01
3.57277244e-02 -1.56010044e+00 -4.99504179e-01 5.97440004e-01
-5.77553511e-01 -1.42600250e+00 8.78823251e-02 -6.77249506e-02
-7.63789535e-01 -9.24865186e-01 -4.61110771e-01 -8.34311664e-01
5.79228938e-01 -3.82063538e-02 1.26194036e+00 6.27860948e-02
2.53970534e-01 1.62185714e-01 -5.35470784e-01 -4.00161654e-01
-1.21342480e+00 3.03105354e-01 -3.06473672e-01 -3.02504390e-01
4.88501996e-01 -2.17045486e-01 -1.02300778e-01 -1.50929272e-01
-9.25277114e-01 2.00897232e-01 5.18068254e-01 7.81228483e-01
4.31476831e-01 -5.30989468e-01 1.03161848e+00 -1.14773464e+00
9.71803725e-01 -3.29359829e-01 -2.87830383e-01 4.42091376e-01
-4.67158824e-01 6.25604451e-01 5.80422103e-01 -3.12471628e-01
-1.59612596e+00 -3.72224271e-01 -4.09650654e-01 3.40275198e-01
-1.90941542e-01 6.64075613e-01 -3.81201982e-01 6.70562446e-01
7.68502593e-01 -2.59608507e-01 -1.42597184e-01 -5.34558415e-01
6.53609455e-01 6.05600893e-01 2.70605803e-01 -8.54515851e-01
3.85142922e-01 2.65362789e-03 -4.72583860e-01 -7.54674673e-01
-1.27339864e+00 -2.31934845e-01 -4.39270556e-01 8.11785236e-02
9.98803556e-01 -1.12228441e+00 -2.84234099e-02 2.05388237e-02
-1.54080760e+00 -3.84236336e-01 -4.58911985e-01 4.13313061e-01
-4.15554285e-01 2.52183884e-01 -5.68717897e-01 -7.68440545e-01
-5.50146043e-01 -8.85404229e-01 1.16969681e+00 -2.45468002e-02
-6.69980466e-01 -1.36520064e+00 7.22709820e-02 3.83229911e-01
4.26188976e-01 2.21420020e-01 1.50689578e+00 -9.18824077e-01
-1.23062722e-01 1.55636892e-01 -1.53187528e-01 3.73030007e-01
1.06246561e-01 -4.04172063e-01 -1.05752754e+00 5.30083291e-02
-6.51450902e-02 -6.64838850e-01 5.81165075e-01 3.29539329e-01
5.05723953e-01 -3.28547239e-01 -5.67086264e-02 -7.89004788e-02
1.34808671e+00 -1.66795671e-01 6.43759906e-01 5.17692089e-01
6.08863473e-01 1.08609211e+00 7.95041323e-01 5.41518629e-02
8.17088485e-01 6.09041572e-01 9.63073894e-02 -2.79627778e-02
-2.72049606e-01 -4.89851922e-01 9.11323845e-01 1.06847489e+00
6.83272555e-02 -2.03841493e-01 -7.57622659e-01 7.90367901e-01
-1.97502339e+00 -6.43250883e-01 -4.44746733e-01 2.23306942e+00
1.19433630e+00 2.65899271e-01 -3.18365455e-01 -1.79811403e-01
5.90344310e-01 1.67291641e-01 -4.45846654e-02 -7.61161089e-01
-3.09445977e-01 3.78163844e-01 1.16250608e-02 7.33138025e-01
-8.34398985e-01 1.68545747e+00 5.67188454e+00 9.02532816e-01
-9.87526119e-01 3.19608569e-01 3.29106331e-01 4.97394413e-01
-7.35691547e-01 2.01471612e-01 -8.22563350e-01 4.08305153e-02
1.08667803e+00 -7.02738836e-02 8.44909251e-02 8.44661415e-01
2.51664102e-01 -1.44429848e-01 -1.10642672e+00 5.32459259e-01
1.51636777e-02 -1.13425434e+00 4.68112946e-01 -2.14832008e-01
7.00159132e-01 1.18520604e-02 -4.87069041e-01 6.85444057e-01
5.07361650e-01 -9.10461128e-01 8.36166322e-01 2.11282507e-01
8.41383576e-01 -6.31312072e-01 1.04824305e+00 4.07225043e-01
-9.87135291e-01 3.44110668e-01 -5.35135329e-01 -1.23848960e-01
3.69078398e-01 4.53367680e-01 -1.05538619e+00 6.73705459e-01
8.69026408e-02 4.51148063e-01 -6.33783996e-01 1.59446690e-02
-8.80493164e-01 6.53165102e-01 -8.95859022e-03 -1.11246988e-01
2.21633434e-01 3.76958437e-02 4.67483908e-01 1.53603220e+00
3.00562292e-01 -2.10915610e-01 2.70303190e-01 1.05422139e+00
9.44023393e-03 6.82786107e-01 -6.49202406e-01 -2.83305287e-01
3.01325023e-01 1.12550259e+00 -5.21638930e-01 -5.46742201e-01
-3.97057652e-01 6.81316078e-01 5.18478632e-01 1.58804059e-01
-4.86580998e-01 -1.54777661e-01 5.03792405e-01 2.33120143e-01
1.64204329e-01 -1.74533576e-01 -3.13500583e-01 -1.10345113e+00
-9.17571634e-02 -9.09139872e-01 2.47097775e-01 -6.86327517e-01
-1.25154257e+00 7.54655898e-01 1.77097276e-01 -9.32897568e-01
-9.61130202e-01 -5.67317367e-01 -3.79578471e-01 9.40358579e-01
-1.91506815e+00 -1.47576046e+00 2.15048224e-01 3.65630150e-01
7.15493143e-01 -1.32269263e-01 1.27981544e+00 -3.17960866e-02
-1.43634126e-01 4.14779782e-01 -3.50882888e-01 1.34311527e-01
7.95452118e-01 -1.24883223e+00 5.04474938e-01 8.31319869e-01
8.90213996e-02 7.66718626e-01 7.09558249e-01 -6.26603127e-01
-8.51094365e-01 -9.31591332e-01 1.62356246e+00 -3.88979107e-01
7.80164182e-01 -2.97892332e-01 -1.12153709e+00 5.09169102e-01
5.89088202e-01 -5.70095003e-01 8.22588980e-01 3.21025401e-01
-1.82470784e-01 3.64891291e-01 -9.29666996e-01 5.82699537e-01
8.43432903e-01 -6.50294185e-01 -1.32028437e+00 2.28734344e-01
9.67012405e-01 -1.77165717e-01 -7.16146708e-01 2.19915614e-01
1.13562256e-01 -6.48902416e-01 6.58133566e-01 -5.35509169e-01
6.68525696e-01 -1.90749168e-01 -1.49944291e-01 -1.35655093e+00
1.08014710e-01 -4.30355728e-01 3.37746054e-01 1.47610843e+00
7.33283997e-01 -5.47737420e-01 1.50757521e-01 3.94161999e-01
-1.24966241e-01 -2.13824272e-01 -7.48590231e-01 -5.75110376e-01
3.73941571e-01 -5.04289389e-01 4.63962793e-01 1.04575980e+00
3.07488531e-01 1.12991190e+00 8.13490450e-02 -2.34800518e-01
4.07627344e-01 5.40203750e-02 6.80525661e-01 -1.26372695e+00
-3.65175396e-01 -2.21186519e-01 2.45174259e-01 -8.61110449e-01
6.32704020e-01 -1.14531255e+00 2.03007519e-01 -1.83079576e+00
-5.99012189e-02 -5.84107161e-01 -1.07171230e-01 6.46485746e-01
-3.72053921e-01 -2.00810850e-01 9.44238827e-02 4.30704132e-02
-3.90121430e-01 6.90177023e-01 1.15397251e+00 1.85932249e-01
-3.69914919e-01 -3.84526610e-01 -9.83422399e-01 8.56901169e-01
7.87012041e-01 -6.74675584e-01 -4.79661018e-01 -6.09816611e-01
3.28845322e-01 -3.84210385e-02 -2.97144987e-02 -4.95151550e-01
-3.78063112e-01 -1.61704630e-01 -8.80870745e-02 -2.41911054e-01
7.19844550e-02 -4.16021347e-01 -2.56454140e-01 2.31135592e-01
-6.38093054e-01 7.67178908e-02 2.54073054e-01 9.28723346e-03
-3.57496649e-01 -6.80929482e-01 5.19341111e-01 -2.87271947e-01
-7.03281462e-01 -5.92913389e-01 -5.27851284e-01 3.91974628e-01
5.47346115e-01 -9.58361849e-02 -2.80388683e-01 -2.82416701e-01
-6.73190534e-01 3.51324566e-02 3.58751863e-01 7.47525096e-01
2.19205320e-01 -1.16972911e+00 -7.91001260e-01 4.41859849e-02
3.64498913e-01 1.51904255e-01 -4.73690778e-02 6.04794085e-01
-3.15085441e-01 6.18501186e-01 3.56441401e-02 -3.73918444e-01
-1.03422976e+00 2.57929236e-01 4.70686294e-02 -7.91625381e-01
-4.75596607e-01 5.86862803e-01 3.35117072e-01 -8.29010546e-01
1.05683925e-03 -6.03321373e-01 -6.53136611e-01 9.82804000e-02
5.04080355e-01 -4.30400595e-02 4.99891713e-02 -8.50334883e-01
-1.40380815e-01 2.30376944e-01 -1.50624722e-01 -4.24262702e-01
1.30627406e+00 -3.17415297e-01 -2.14598954e-01 7.04086900e-01
8.63700867e-01 1.78967491e-01 -8.06992054e-01 -3.20223242e-01
6.84098840e-01 1.54941589e-01 -8.63503814e-02 -8.94491911e-01
-3.45547140e-01 1.02072644e+00 3.21246684e-02 5.85288666e-02
6.81449592e-01 1.35852590e-01 8.07676673e-01 4.07618642e-01
3.33197892e-01 -1.46955836e+00 3.11052147e-02 9.13520217e-01
1.14125729e+00 -1.21026242e+00 -2.86798805e-01 -4.98865366e-01
-1.02721500e+00 1.01959407e+00 5.28600216e-01 -1.08943351e-01
1.23412870e-01 3.25926505e-02 2.26958096e-01 -4.42851288e-03
-8.32432330e-01 -3.09963048e-01 1.81261182e-01 5.94426394e-01
1.07882178e+00 3.17254633e-01 -8.98335874e-01 8.28280628e-01
-4.55432594e-01 -1.46555707e-01 5.09886980e-01 8.43501210e-01
-4.88990337e-01 -1.62745190e+00 -2.25444317e-01 -1.78673089e-01
-5.87423503e-01 -4.41776812e-01 -5.34739554e-01 9.18224812e-01
1.58879533e-02 1.03804314e+00 -1.09847620e-01 -5.56146633e-03
3.52490246e-01 5.08227587e-01 4.68276680e-01 -9.84943569e-01
-8.87027800e-01 2.15306893e-01 6.74314737e-01 -3.76175225e-01
-7.28928804e-01 -5.76113462e-01 -1.64466035e+00 1.13515548e-01
-2.82617837e-01 3.41743380e-01 6.68081343e-01 1.22404730e+00
2.19236299e-01 5.39213181e-01 1.90950736e-01 -5.33027828e-01
-6.25019908e-01 -1.40084636e+00 -1.20206453e-01 5.42126179e-01
-1.45987868e-01 -3.87205690e-01 -1.81708559e-01 2.75947928e-01] | [10.83299732208252, 9.283388137817383] |
0531b940-9c47-4a10-982d-e9a44807d4f5 | toward-personalized-medicine-in-connectomic | 2109.12327 | null | https://arxiv.org/abs/2109.12327v1 | https://arxiv.org/pdf/2109.12327v1.pdf | Toward Personalized Medicine in Connectomic Deep Brain Stimulation | At the group-level, deep brain stimulation leads to significant therapeutic benefit in a multitude of neurological and neuropsychiatric disorders. At the single-patient level, however, symptoms may sometimes persist despite "optimal" electrode placement at established treatment coordinates. This may be partly explained by limitations of disease-centric strategies that are unable to account for heterogeneous phenotypes and comorbidities observed in clinical practice. Instead, tailoring electrode placement and programming to individual patients' symptom profiles may increase the fraction of top responding patients. Here, we propose a three-step, circuit-based framework that aims to develop patient-specific treatment targets that address the unique symptom constellation prevalent in each patient. First, we describe how a symptom network target library could be established by mapping beneficial or undesirable DBS effects to distinct circuits based on (retrospective) group-level data. Second, we suggest ways of matching the resulting symptom networks to circuits defined in the individual patient (template matching). Third, we introduce network blending as a strategy to calculate optimal stimulation targets and parameters by selecting and weighting a set of symptom-specific networks based on the symptom profile and subjective priorities of the individual patient. We integrate the approach with published literature and conclude by discussing limitations and future challenges. | ['Andreas Horn', 'Clemens Neudorfer', 'Michael D. Fox', 'Helen S. Mayberg', 'Andrea A. Kühn', 'Carsten Finke', 'Shan H. Siddiqi', 'Nanditha Rajamani', 'Barbara Hollunder'] | 2021-09-25 | null | null | null | null | ['template-matching'] | ['computer-vision'] | [ 4.02220339e-01 6.69821575e-02 -5.01520932e-01 -2.19549045e-01
-6.30247235e-01 -9.09446418e-01 2.89308667e-01 4.90281701e-01
-1.84234649e-01 8.36711407e-01 5.41496634e-01 -7.58541152e-02
-1.06059802e+00 -4.07842398e-01 -1.13595128e-01 -4.27918583e-01
-1.54242173e-01 8.43911648e-01 -6.31544366e-02 -1.21204279e-01
2.94373989e-01 9.24998403e-01 -1.08066595e+00 4.86743063e-01
9.88849163e-01 7.83402264e-01 4.20573294e-01 -2.97777206e-02
1.36870509e-02 -3.62456329e-02 -5.44605315e-01 9.48007703e-02
-4.07893136e-02 -7.86020279e-01 -5.36991775e-01 -2.52575129e-01
-9.26137790e-02 1.61916809e-03 -1.29927590e-01 9.29913759e-01
1.04554284e+00 -3.42993677e-01 7.76977003e-01 -9.06885624e-01
-1.13174744e-01 8.93563807e-01 -2.29513630e-01 2.21783683e-01
9.71581265e-02 2.06698373e-01 5.67746937e-01 -5.29350460e-01
9.15677488e-01 7.98801303e-01 4.96424615e-01 7.39439428e-01
-1.81625211e+00 -5.86336315e-01 2.21901521e-01 3.02728433e-02
-1.31837273e+00 -4.74726200e-01 5.57592750e-01 -5.20105660e-01
8.83793056e-01 4.26136643e-01 1.34682524e+00 1.01051819e+00
2.42873743e-01 5.66179976e-02 9.30012047e-01 1.31072491e-01
2.95380265e-01 -1.27349392e-01 -1.24341957e-01 -1.78023010e-01
3.56456071e-01 -2.48952612e-01 -5.28906047e-01 -5.44842780e-01
7.37582922e-01 -1.44083366e-01 -5.83997548e-01 -6.31296992e-01
-1.24044299e+00 6.12970948e-01 3.07297647e-01 8.87318969e-01
-7.64924109e-01 3.54694463e-02 5.36141336e-01 -1.58951342e-01
1.54739708e-01 8.38989019e-01 -5.41627645e-01 7.89351296e-03
-1.40224600e+00 4.01779920e-01 3.60067874e-01 4.46753055e-01
1.88904554e-01 -1.26876548e-01 -6.58716619e-01 1.17382741e+00
4.52676192e-02 1.20434836e-02 5.08484006e-01 -9.46244001e-01
1.92075372e-01 5.92624307e-01 -1.05404951e-01 -7.13687837e-01
-1.03709304e+00 -9.15983200e-01 -6.89451396e-01 1.41219869e-01
1.25890359e-01 -3.97102326e-01 -6.87939167e-01 2.10761857e+00
-1.44549176e-01 -3.02911490e-01 -3.80702764e-01 7.07920730e-01
3.69167179e-01 -1.44242123e-01 2.28609309e-01 -5.09085298e-01
1.18926501e+00 -2.49455392e-01 -6.31417096e-01 -5.24743974e-01
7.84733772e-01 -9.06106457e-02 8.06372643e-01 5.39206803e-01
-1.32889330e+00 3.26154262e-01 -7.96018660e-01 7.36640692e-01
-1.64593861e-01 9.89370570e-02 4.14180309e-01 6.49934947e-01
-1.48283398e+00 6.80168629e-01 -7.32992291e-01 -7.71531761e-01
7.61315167e-01 8.55547965e-01 -1.62161767e-01 2.22688839e-01
-1.06782985e+00 1.33320165e+00 2.87362665e-01 1.34952471e-01
-7.29488134e-01 -1.01666665e+00 -1.18966877e-01 1.67414024e-01
8.47454295e-02 -1.11416149e+00 6.82833970e-01 -8.70642543e-01
-1.15991187e+00 7.90208042e-01 4.63382248e-03 -1.19007505e-01
3.66173059e-01 5.71714580e-01 -3.16989958e-01 2.42930040e-01
8.81894305e-02 6.32892191e-01 9.88392085e-02 -1.12167442e+00
-1.66418567e-01 -7.13119864e-01 -2.77806401e-01 4.49847162e-01
-2.07195580e-01 2.88805932e-01 -1.85801312e-01 -6.47810936e-01
2.41936147e-01 -6.65445447e-01 -6.62149310e-01 3.79047133e-02
-2.93913424e-01 3.84516209e-01 1.79333434e-01 -3.69133532e-01
1.19358122e+00 -1.96501875e+00 6.57507479e-01 6.50958240e-01
3.95401984e-01 2.59808134e-02 -1.46620616e-01 7.74479806e-01
-4.35178250e-01 2.10438713e-01 -3.54020774e-01 1.73165336e-01
-2.03954294e-01 -1.96733579e-01 1.65462390e-01 4.65647161e-01
2.58645564e-02 8.74493957e-01 -7.05960155e-01 -1.27620742e-01
1.94357961e-01 4.47891980e-01 -7.76922345e-01 -2.11373687e-01
8.19333941e-02 7.50669777e-01 -4.69705135e-01 6.19770288e-01
3.75807762e-01 -7.49644265e-03 8.72431993e-01 -2.04303294e-01
3.63894626e-02 3.70528936e-01 -8.88244927e-01 1.84318006e+00
-1.55551463e-01 1.96996123e-01 4.23971385e-01 -1.19737005e+00
7.26638556e-01 4.81442779e-01 9.15706694e-01 -3.73824507e-01
2.81486183e-01 5.95945716e-01 6.94955945e-01 -1.81303084e-01
-2.65086591e-01 -4.57042545e-01 2.92548705e-02 3.38694960e-01
1.36216044e-01 -1.54590895e-02 8.56160596e-02 -5.78670353e-02
1.46483707e+00 -3.75181884e-01 2.19733909e-01 -5.93568742e-01
1.04593918e-01 1.26489267e-01 4.07263517e-01 7.10494578e-01
-2.65831828e-01 7.69867480e-01 9.56781924e-01 3.34870696e-01
-6.70873761e-01 -1.05812347e+00 -5.49172461e-01 5.31838596e-01
-1.58929616e-01 -1.50383234e-01 -8.00122678e-01 -1.37961835e-01
2.39314996e-02 6.00752592e-01 -6.67096615e-01 -5.29276431e-01
-1.76174000e-01 -1.10402560e+00 6.83643043e-01 3.58909607e-01
-1.69493377e-01 -1.09064674e+00 -4.32825029e-01 4.51279789e-01
1.05402209e-01 -5.11199713e-01 -1.04274563e-01 5.38881242e-01
-1.12277484e+00 -8.55537772e-01 -9.39378083e-01 -6.47955954e-01
7.11875796e-01 -2.94639528e-01 6.57707572e-01 -6.15905188e-02
-2.73542643e-01 2.43826151e-01 -7.02643320e-02 2.38232445e-02
-5.57458065e-02 9.33146477e-02 2.43046060e-01 -2.30327040e-01
-2.02483982e-01 -1.00166500e+00 -1.02125752e+00 3.36001590e-02
-7.74088621e-01 -1.80144444e-01 7.28054702e-01 5.90309441e-01
6.71896040e-01 -2.71435976e-01 9.94632304e-01 -7.00041294e-01
1.13845706e+00 -6.95988894e-01 -1.13804430e-01 2.40887254e-01
-7.04937696e-01 -3.09846073e-01 2.88213372e-01 -5.59693515e-01
-5.55521369e-01 -3.21893580e-02 1.19709205e-02 -2.27936164e-01
-1.39165908e-01 7.85976052e-01 -2.93186247e-01 -3.87928188e-02
6.75877154e-01 9.32547543e-03 1.30855426e-01 -3.22796017e-01
2.16513351e-01 4.57166284e-01 3.34324449e-01 -4.38068032e-01
-9.40911565e-03 2.90805906e-01 -1.02627233e-01 -1.93062320e-01
-1.41141161e-01 1.12757729e-02 -4.66633409e-01 -3.03399742e-01
6.36969984e-01 -5.63772261e-01 -6.56531632e-01 1.33084238e-01
-9.84487832e-01 -5.91244161e-01 -2.60074645e-01 5.58475494e-01
-6.79943740e-01 -3.45060945e-01 -1.78719059e-01 -4.68616128e-01
-5.16676247e-01 -1.47010517e+00 7.94604599e-01 -4.14784066e-02
-8.72865498e-01 -1.02359998e+00 3.08143586e-01 -2.21954018e-01
5.61232865e-01 3.62862468e-01 1.35751009e+00 -7.55861223e-01
-9.42873582e-02 -3.30452740e-01 -9.32590738e-02 -1.17684416e-01
3.49757999e-01 -3.33556950e-01 -5.77589869e-01 -1.20572016e-01
-1.09444760e-01 2.47213263e-02 2.36647978e-01 1.09441793e+00
1.04124439e+00 1.57110736e-01 -7.15357482e-01 5.16919196e-01
1.45125914e+00 4.24858928e-01 6.21844828e-01 2.07487240e-01
3.23164433e-01 9.34697449e-01 -1.13236234e-01 2.06441283e-01
-3.01510375e-02 8.93754065e-01 2.55213946e-01 3.24956626e-02
-2.09586143e-01 6.20050691e-02 1.00914590e-01 1.88634157e-01
7.32353181e-02 -1.91681400e-01 -1.02118111e+00 7.02314854e-01
-1.65660322e+00 -9.24442410e-01 -1.47563711e-01 2.21451569e+00
7.31167316e-01 1.58989027e-01 4.87312615e-01 -1.60895452e-01
8.88807774e-01 -2.79126257e-01 -7.29120314e-01 -4.52488512e-01
-2.76178509e-01 2.95901388e-01 5.97078204e-01 2.50923812e-01
-6.62223175e-02 5.92672348e-01 7.93795252e+00 7.59163916e-01
-1.38947093e+00 2.72588015e-01 5.48141778e-01 -6.97489142e-01
-5.42077541e-01 8.86441469e-02 -3.98021311e-01 4.58315283e-01
9.91176903e-01 -4.60250735e-01 7.34878123e-01 4.82418090e-02
8.21513593e-01 -7.55538419e-02 -1.13706815e+00 8.16481113e-01
-2.15185568e-01 -1.64128256e+00 -1.41339049e-01 1.94980606e-01
4.01210040e-01 3.67840260e-01 -4.80672978e-02 -2.83834726e-01
-1.48502290e-01 -1.14631605e+00 6.92423224e-01 6.57671571e-01
1.17865038e+00 -5.63646674e-01 3.45049769e-01 1.04259290e-01
-6.63187504e-01 -2.16041222e-01 1.96493819e-01 3.18068594e-01
2.86225230e-01 5.79939127e-01 -5.78286707e-01 3.43119383e-01
5.10929585e-01 4.42852139e-01 -2.77458131e-01 1.42425835e+00
1.59771532e-01 3.44816536e-01 -3.24317843e-01 9.35394540e-02
-6.78464770e-02 -2.64220148e-01 6.18185818e-01 8.71337891e-01
5.76012611e-01 1.46318763e-01 -5.03683150e-01 1.35850441e+00
2.49700636e-01 2.34213948e-01 -6.63613737e-01 -1.51610449e-01
4.95884657e-01 1.40365064e+00 -1.02416551e+00 3.12789418e-02
-4.64651585e-02 6.66682899e-01 1.98837683e-01 4.39719141e-01
-3.65944594e-01 -2.50160068e-01 7.13548779e-01 3.37900668e-01
-5.91671914e-02 2.61812150e-01 -8.67829442e-01 -7.34263062e-01
-9.92656574e-02 -9.14760768e-01 2.07916871e-01 -8.48581851e-01
-9.70421612e-01 5.68711281e-01 1.63712099e-01 -9.29357409e-01
-1.37373999e-01 -4.78681445e-01 -8.37274194e-01 1.11863923e+00
-5.22825420e-01 -7.22113013e-01 2.09773883e-01 3.19465101e-01
-8.66185129e-02 3.92978229e-02 8.81908059e-01 4.52684104e-01
-5.87545216e-01 4.13646191e-01 1.55038342e-01 -5.35599947e-01
6.31311834e-01 -8.19050193e-01 -4.22879130e-01 2.28625029e-01
-5.19401729e-01 6.29690766e-01 8.45956504e-01 -8.59540939e-01
-1.05976415e+00 -6.70211732e-01 6.89281523e-01 -2.24483505e-01
7.98551142e-01 -4.22899604e-01 -7.45892465e-01 4.32762474e-01
8.40092525e-02 -4.09166098e-01 7.13063955e-01 2.43327558e-01
3.14970702e-01 -1.28004119e-01 -1.53979588e+00 9.74367678e-01
1.31017447e+00 -4.09627080e-01 -1.40100405e-01 4.33547974e-01
1.78403914e-01 -1.53154075e-01 -1.11011326e+00 4.56296861e-01
5.21476388e-01 -7.53856063e-01 7.15729475e-01 -4.64037448e-01
1.74178571e-01 9.82277021e-02 1.54686779e-01 -1.67489040e+00
-4.74780321e-01 -4.97584343e-01 5.30895889e-01 1.12074399e+00
7.26633132e-01 -7.18322039e-01 7.23868608e-01 9.11766529e-01
-2.55186558e-01 -1.10505831e+00 -8.29114914e-01 -5.19627512e-01
4.45240825e-01 -3.75416964e-01 6.08001649e-01 7.61420369e-01
7.79123187e-01 6.25040308e-02 7.24282488e-02 -1.87187687e-01
1.09144621e-01 -2.72587359e-01 -1.54021040e-01 -1.10761094e+00
-2.21049637e-01 -1.07066202e+00 -3.94041866e-01 2.57126032e-03
-1.05979025e-01 -1.33667433e+00 -2.19289586e-01 -1.97687817e+00
3.34426761e-01 -6.28430545e-01 -4.92212266e-01 5.60864508e-01
3.35410804e-01 1.43328160e-01 -2.02274486e-01 -1.07820621e-02
1.32149458e-01 2.57147163e-01 1.00439203e+00 1.54906750e-01
-5.99554420e-01 -2.94246048e-01 -1.32791662e+00 4.90408897e-01
9.92088914e-01 -6.03647172e-01 -6.34735227e-01 -3.21513444e-01
3.03740025e-01 3.31160635e-01 2.12455004e-01 -8.71104360e-01
1.37596264e-01 -3.42605412e-01 3.27729911e-01 -3.06121826e-01
2.70635337e-01 -6.70730770e-01 6.08792543e-01 6.76041901e-01
-3.49104673e-01 -8.12685117e-02 4.70666319e-01 1.82032764e-01
2.05436721e-01 -1.82358682e-01 7.37935781e-01 1.81472674e-01
3.65721993e-02 2.35390231e-01 -8.22987914e-01 -6.05442934e-02
9.38029051e-01 -3.51704210e-01 -3.23182523e-01 -4.06208813e-01
-1.25315034e+00 1.69527620e-01 4.63790387e-01 1.54847935e-01
4.28971142e-01 -1.34637177e+00 -5.84391057e-01 -1.41723365e-01
5.98493032e-02 -5.27780712e-01 5.64329922e-01 1.47494447e+00
-2.67689437e-01 5.84554076e-01 -6.14112616e-01 -4.00080323e-01
-9.21396971e-01 2.12301001e-01 8.34135473e-01 -4.98179607e-02
-4.84117031e-01 6.96153522e-01 3.14429849e-01 -3.59376609e-01
7.88719207e-02 -4.19484898e-02 -3.12735051e-01 3.13749552e-01
1.66570500e-01 5.95332459e-02 3.84517789e-01 -4.92658257e-01
-7.75506437e-01 4.13212031e-01 1.51799127e-01 -5.04072368e-01
1.56058490e+00 1.43367741e-02 -3.12050939e-01 1.67951182e-01
8.63587916e-01 -7.26680979e-02 -7.14871645e-01 4.83119279e-01
-1.50266752e-01 -1.80254746e-02 1.39264613e-01 -1.25268364e+00
-1.38141131e+00 4.57455724e-01 7.51940250e-01 -8.54992047e-02
1.32821858e+00 2.62155563e-01 1.87615827e-01 -1.56540751e-01
3.49595845e-01 -9.21103597e-01 -6.61782801e-01 -1.39399514e-01
1.12851489e+00 -1.75521463e-01 -1.12969160e-01 -2.27201283e-01
-5.19592166e-01 6.75871670e-01 3.13392043e-01 -1.56606942e-01
6.94407284e-01 2.21205056e-01 -6.51635751e-02 -6.35213017e-01
-7.02634275e-01 -5.33565506e-02 1.53862357e-01 8.51199448e-01
4.93488699e-01 1.63092121e-01 -1.07098663e+00 1.06462419e+00
-6.40196353e-02 1.29290223e-01 3.82938594e-01 7.37978995e-01
-2.29941756e-01 -1.67058837e+00 -3.60030025e-01 1.21349287e+00
-2.94920504e-01 -2.41433576e-01 -7.47405469e-01 4.52370644e-01
2.94210821e-01 6.50442660e-01 6.04123883e-02 -3.76176178e-01
5.13262570e-01 1.11116335e-01 7.97626853e-01 -8.18475008e-01
-9.78840828e-01 4.85314101e-01 2.31480032e-01 -4.16252732e-01
-1.39425740e-01 -1.06806576e+00 -1.39622772e+00 8.50430429e-02
-4.86087799e-02 -1.46984503e-01 6.01533711e-01 1.03471398e+00
7.66585767e-01 7.83998847e-01 2.84742862e-02 -9.30869758e-01
-1.39149889e-01 -6.98894560e-01 -1.01099062e+00 -1.46448895e-01
-1.25486240e-01 -8.73398483e-01 -1.71399161e-01 -6.54589772e-01] | [13.034975051879883, 3.4074392318725586] |
1a871528-f263-4712-8e33-271f8b63fc69 | amodal-instance-segmentation-with-kins | null | null | http://openaccess.thecvf.com/content_CVPR_2019/html/Qi_Amodal_Instance_Segmentation_With_KINS_Dataset_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Qi_Amodal_Instance_Segmentation_With_KINS_Dataset_CVPR_2019_paper.pdf | Amodal Instance Segmentation With KINS Dataset | Amodal instance segmentation, a new direction of instance segmentation, aims to segment each object instance involving its invisible, occluded parts to imitate human ability. This task requires to reason objects' complex structure. Despite important and futuristic, this task lacks data with large-scale and detailed annotations, due to the difficulty of correctly and consistently labeling invisible parts, which creates the huge barrier to explore the frontier of visual recognition. In this paper, we augment KITTI with more instance pixel-level annotation for 8 categories, which we call KITTI INStance dataset (KINS). We propose the network structure to reason invisible parts via a new multi-task framework with Multi-View Coding (MVC), which combines information in various recognition levels. Extensive experiments show that our MVC effectively improves both amodal and inmodal segmentation. The KINS dataset and our proposed method will be made publicly available.
| [' Jiaya Jia', ' Xiaoyong Shen', ' Shu Liu', ' Li Jiang', 'Lu Qi'] | 2019-06-01 | null | null | null | cvpr-2019-6 | ['amodal-instance-segmentation'] | ['computer-vision'] | [ 5.38205385e-01 3.04067254e-01 -3.71772289e-01 -2.61643231e-01
-5.71595252e-01 -6.92281485e-01 3.26105475e-01 -5.54113925e-01
4.07481976e-02 6.47368312e-01 7.65552595e-02 -1.27874255e-01
-2.93840077e-02 -3.18556339e-01 -7.29422569e-01 -6.82182550e-01
5.00226080e-01 4.70725536e-01 3.74762803e-01 6.65326416e-02
3.71073276e-01 3.43632936e-01 -1.48082101e+00 7.04959214e-01
1.09511793e+00 1.18341768e+00 3.49627465e-01 4.24751759e-01
-3.77296150e-01 7.54530787e-01 -4.10627156e-01 -5.50265849e-01
1.44777894e-01 -3.16400647e-01 -1.19579959e+00 4.34392422e-01
1.02002263e+00 -8.70785564e-02 1.75129816e-01 1.11522222e+00
2.16593206e-01 -7.61430711e-02 7.29419529e-01 -1.44774985e+00
-7.05196381e-01 6.77886903e-01 -9.14801300e-01 2.01304764e-01
1.82300046e-01 1.91764534e-02 9.04077172e-01 -8.35938692e-01
6.78555131e-01 1.19851899e+00 5.95505238e-01 6.49601281e-01
-1.24515069e+00 -4.84801769e-01 5.10083437e-01 1.82040185e-01
-1.07695687e+00 -2.35896215e-01 8.92164528e-01 -4.81131017e-01
5.98419845e-01 4.13894475e-01 7.22724438e-01 1.27217209e+00
-9.01580304e-02 1.31838834e+00 1.52743316e+00 -7.82303661e-02
-9.32780001e-03 -9.43680629e-02 3.99167717e-01 6.90702319e-01
2.10680693e-01 -3.18650275e-01 -3.98938179e-01 3.86582226e-01
9.93638098e-01 2.52167523e-01 -2.65404046e-01 -4.71665114e-01
-1.40008450e+00 2.46396348e-01 6.71815813e-01 2.75127053e-01
-1.88456133e-01 1.27380714e-01 1.58932790e-01 -3.75160947e-02
3.97944003e-01 5.32118566e-02 -5.72407365e-01 5.89486659e-02
-8.97068083e-01 -2.69393742e-01 4.62498337e-01 8.68990898e-01
7.89546311e-01 4.65483628e-02 -3.00974399e-01 9.97838318e-01
1.28363818e-01 6.79980695e-01 2.81247526e-01 -1.29983866e+00
6.46954596e-01 1.04766858e+00 -2.24963382e-01 -8.30209613e-01
-4.67480570e-01 -4.38720465e-01 -1.04097831e+00 2.70694494e-01
6.06381834e-01 1.06790975e-01 -1.38641083e+00 1.41405141e+00
2.91779965e-01 2.30369391e-03 -2.57809516e-02 1.19395375e+00
1.08022797e+00 4.42845851e-01 -8.55608061e-02 2.29103491e-01
1.52700925e+00 -1.44921279e+00 -6.06474459e-01 -5.66210926e-01
2.36283839e-01 -6.43651009e-01 1.16981721e+00 5.39385796e-01
-1.09792614e+00 -7.37578332e-01 -6.47718549e-01 -2.48347580e-01
-5.89348435e-01 3.26750189e-01 9.99160528e-01 5.30743420e-01
-1.03003323e+00 8.43711644e-02 -5.08892655e-01 -1.72464848e-01
9.58096862e-01 1.44086063e-01 -5.69519758e-01 -2.15506122e-01
-7.54467845e-01 5.33483624e-01 4.49478656e-01 4.60638255e-01
-8.77688766e-01 -4.15782601e-01 -8.88245404e-01 7.86685124e-02
6.95045948e-01 -6.91988289e-01 7.90313125e-01 -1.29955637e+00
-1.14247763e+00 1.00515676e+00 -2.43630514e-01 -8.62490907e-02
4.70762402e-01 -5.70287220e-02 -4.43589129e-02 2.10156515e-01
2.45841935e-01 1.18588984e+00 1.04605114e+00 -1.72793055e+00
-4.09847438e-01 -6.56631291e-01 1.46704599e-01 1.82351723e-01
-1.82410493e-01 -3.17064494e-01 -8.44473839e-01 -7.18369544e-01
5.43839335e-01 -7.60998607e-01 -1.36607334e-01 -3.62965800e-02
-9.29528236e-01 -2.98475564e-01 7.54993677e-01 -6.41594887e-01
7.96580791e-01 -2.22990227e+00 5.01706839e-01 8.75440761e-02
5.98708034e-01 1.35308787e-01 -1.93727061e-01 -5.75552657e-02
-2.34488696e-02 1.67540103e-01 -4.85228062e-01 -3.78854811e-01
-1.40138134e-01 3.77049387e-01 -2.31619298e-01 9.91684273e-02
8.14714879e-02 1.26444256e+00 -6.70212269e-01 -9.11274791e-01
2.01057732e-01 3.36027563e-01 -4.14841950e-01 1.81265380e-02
-1.77278638e-01 6.31547868e-01 -4.08600479e-01 1.21240628e+00
8.59511852e-01 -6.22419715e-01 -1.42615795e-01 -4.47240114e-01
9.80124250e-03 -4.47125167e-01 -9.81749833e-01 1.83747852e+00
-2.73095369e-01 6.26454473e-01 1.48957714e-01 -1.08842897e+00
6.54033840e-01 -3.84747535e-02 3.45241249e-01 -7.61142313e-01
6.34874254e-02 5.49342856e-02 -1.10807113e-01 -6.07052386e-01
2.59170145e-01 3.30354750e-01 -1.00602083e-01 2.20901042e-01
-5.97245879e-02 6.34065419e-02 1.80526361e-01 3.97476941e-01
6.53934479e-01 3.61500531e-01 -2.60832142e-02 -7.99358934e-02
5.84863663e-01 1.37860447e-01 5.44964373e-01 6.84259355e-01
-3.32676768e-01 8.78865361e-01 6.02554083e-01 -5.06959677e-01
-5.95237553e-01 -1.16820335e+00 -2.38756359e-01 9.73392725e-01
5.69103360e-01 1.26659900e-01 -9.22202408e-01 -9.00038898e-01
8.27451348e-02 2.11331993e-01 -8.16125631e-01 2.60307670e-01
-3.07306081e-01 -5.02186418e-01 4.55613583e-01 8.49093258e-01
8.07714999e-01 -1.20351195e+00 -5.97335994e-01 -2.81907052e-01
-5.77733338e-01 -1.31176853e+00 -5.74369490e-01 8.63827690e-02
-9.61752594e-01 -1.21076977e+00 -1.17872715e+00 -9.27145123e-01
9.61024821e-01 4.81447160e-01 9.42460120e-01 -7.87153691e-02
-5.23267210e-01 4.58488345e-01 -1.27778232e-01 -2.86961365e-02
9.25079137e-02 5.86851649e-02 -3.92543346e-01 2.22022206e-01
-1.41261250e-01 -3.93790722e-01 -6.96078062e-01 3.87741774e-01
-8.96588206e-01 6.13268852e-01 9.50491071e-01 6.68688774e-01
8.66769850e-01 -3.60752434e-01 4.26445037e-01 -1.06509137e+00
1.71864107e-01 -8.28087181e-02 -3.22089106e-01 7.21307814e-01
-3.87890041e-01 -5.59871644e-02 2.95809716e-01 -3.25910777e-01
-1.17140615e+00 2.09831983e-01 1.94778144e-01 -5.80852032e-01
-4.91545439e-01 1.83352306e-01 -3.19215745e-01 -2.10504100e-01
1.80885464e-01 2.45140493e-01 -9.24770311e-02 -4.24891770e-01
7.12334931e-01 4.89122331e-01 7.92737424e-01 -5.61327875e-01
3.27710956e-01 5.95621049e-01 -1.01050235e-01 -5.68862438e-01
-1.19296551e+00 -5.29157579e-01 -9.53531325e-01 -5.42416632e-01
1.25222671e+00 -8.09929132e-01 -9.02079880e-01 6.18510962e-01
-1.25375104e+00 -5.11121273e-01 -1.06341600e-01 -4.61714715e-02
-5.41385233e-01 5.45930326e-01 -5.30275643e-01 -8.18689823e-01
-2.02091962e-01 -1.15263033e+00 1.38812971e+00 3.67680758e-01
1.93060502e-01 -9.42472875e-01 -3.18116367e-01 1.15613484e+00
2.28872135e-01 2.45726585e-01 6.92564070e-01 -2.74869591e-01
-9.83611643e-01 5.24203256e-02 -9.47085202e-01 3.19153041e-01
-4.72590625e-02 -1.15566298e-01 -1.14638507e+00 9.82599892e-03
-3.10336798e-01 -5.11703551e-01 1.36331713e+00 5.74432254e-01
1.70909023e+00 2.12213360e-02 -4.78742063e-01 6.37468219e-01
1.21827066e+00 -1.03385141e-03 6.84552252e-01 -2.08798293e-02
1.28406751e+00 7.50893116e-01 3.46414506e-01 1.65240820e-02
5.42199075e-01 4.67150867e-01 5.55830300e-01 -5.20117044e-01
-4.96364951e-01 -8.21818970e-03 6.68950677e-02 7.15908825e-01
-4.78715688e-01 -2.38200977e-01 -9.64859962e-01 4.00956959e-01
-1.91293776e+00 -1.05050182e+00 -2.73567975e-01 1.70673549e+00
4.58486736e-01 1.05507128e-01 1.92114040e-02 6.91214800e-02
7.48536170e-01 8.45547765e-02 -7.27872550e-01 -1.99184224e-01
-2.82761306e-01 -2.59900749e-01 4.00043815e-01 3.64347100e-01
-1.10659552e+00 1.04721951e+00 6.35074520e+00 9.16789114e-01
-8.24585974e-01 1.96481228e-01 1.04781306e+00 4.43092555e-01
-2.96451420e-01 -1.30111739e-01 -7.73088038e-01 4.81573045e-01
1.77350804e-01 5.04106581e-01 4.33111578e-01 7.30170846e-01
-2.50627726e-01 -3.53893220e-01 -8.66639853e-01 1.20755482e+00
5.63949645e-01 -1.17839813e+00 1.14389725e-01 1.16517104e-01
8.87274623e-01 -2.02340439e-01 3.86354864e-01 1.35197848e-01
2.59963144e-02 -8.99464786e-01 7.78924704e-01 7.03878939e-01
7.49247670e-01 -3.98676932e-01 5.20801604e-01 2.57659793e-01
-1.38869631e+00 -2.52783000e-01 -1.74771622e-02 1.18925735e-01
1.00035697e-01 5.21679342e-01 -2.66871333e-01 5.42890310e-01
8.39359164e-01 9.40662444e-01 -8.69850278e-01 8.62096310e-01
-1.17193446e-01 2.58742809e-01 2.82307360e-02 3.18695754e-01
2.92146981e-01 -4.13102299e-01 1.58500060e-01 9.34981883e-01
-5.44959791e-02 2.96741053e-02 3.80563945e-01 1.17158604e+00
-1.26159359e-02 -3.05074036e-01 -6.19950354e-01 -5.28190797e-03
4.59434977e-03 1.49251461e+00 -1.49698818e+00 -5.54493845e-01
-3.69203687e-01 1.57522774e+00 4.19263124e-01 6.40077591e-01
-8.88708472e-01 -1.32373810e-01 3.19401026e-01 -3.29748869e-01
2.93524981e-01 -4.52649184e-02 -6.95562422e-01 -1.38569510e+00
3.18790972e-02 -6.48943245e-01 4.61095244e-01 -1.05568099e+00
-1.44066072e+00 5.36542356e-01 1.49130514e-02 -9.80534434e-01
2.72145420e-01 -9.29837584e-01 -4.43973929e-01 5.25876701e-01
-1.46399140e+00 -1.56662071e+00 -6.83405876e-01 5.87250054e-01
9.21806037e-01 1.29416004e-01 2.65462011e-01 3.08066994e-01
-9.12113070e-01 5.26670098e-01 -2.97055334e-01 1.97138026e-01
4.94245321e-01 -1.24677217e+00 1.75625220e-01 6.58085644e-01
3.75321954e-01 3.22071731e-01 2.04976663e-01 -6.49027467e-01
-1.14375091e+00 -7.84713149e-01 6.78933680e-01 -8.37614715e-01
3.75417799e-01 -4.09214467e-01 -7.63483047e-01 6.49163008e-01
1.73203409e-01 1.42896667e-01 3.41253161e-01 7.52445534e-02
-5.30324399e-01 -3.04321740e-02 -9.54181969e-01 4.95693237e-01
1.48324573e+00 -4.13730949e-01 -5.46210587e-01 3.47839981e-01
5.82895696e-01 -4.86365646e-01 -6.24108613e-01 5.29299378e-01
5.93101025e-01 -1.09048009e+00 1.30517423e+00 -3.18720520e-01
4.82224196e-01 -2.05532819e-01 -1.56793401e-01 -8.99323165e-01
-5.22966124e-02 -1.07000358e-02 -1.14437893e-01 1.19338143e+00
4.52457905e-01 -6.08281314e-01 8.01843286e-01 5.48815370e-01
-3.78798693e-01 -8.29345882e-01 -8.93580914e-01 -6.30647063e-01
-1.80161893e-01 -3.72894198e-01 3.20225090e-01 8.60529602e-01
-2.71517575e-01 3.75625402e-01 -3.95857066e-01 1.28645137e-01
7.59572923e-01 6.93072438e-01 4.96098042e-01 -1.51912427e+00
-9.84073430e-02 -6.61792219e-01 -3.37588698e-01 -1.09797335e+00
1.54865205e-01 -9.93062973e-01 1.50799686e-02 -1.84550571e+00
5.98825395e-01 -2.27913454e-01 -2.62802601e-01 8.24421644e-01
-2.39022478e-01 7.77064562e-01 2.27382869e-01 2.95945048e-01
-1.17743695e+00 4.53930080e-01 1.76517785e+00 -3.79979759e-01
-3.99762094e-02 -2.20050246e-01 -6.13172233e-01 7.68859923e-01
5.45048892e-01 -8.81413072e-02 -4.77142632e-01 -5.49496233e-01
1.35202795e-01 6.34477437e-02 7.44522870e-01 -1.03241372e+00
1.39042214e-01 -1.54533699e-01 8.43205929e-01 -9.23812270e-01
4.98637915e-01 -9.17124093e-01 -3.33322696e-02 4.24103469e-01
-2.18452409e-01 -4.14998904e-02 9.07366127e-02 7.66247630e-01
-3.00971597e-01 1.31526932e-01 5.25669336e-01 -1.23146303e-01
-1.08864963e+00 3.55891436e-01 -1.33521959e-01 1.12769701e-01
1.11542284e+00 -5.65128922e-01 -5.40281117e-01 -2.05884799e-02
-9.25336659e-01 4.67070818e-01 3.98063183e-01 3.93739820e-01
7.94685841e-01 -1.15457368e+00 -2.39204705e-01 3.10543299e-01
2.79537916e-01 1.37102664e-01 6.57513261e-01 1.14862394e+00
-3.01181018e-01 3.94202560e-01 -3.66511345e-01 -9.30047989e-01
-1.14952636e+00 6.18929267e-01 2.29961410e-01 1.24250419e-01
-7.46437132e-01 8.74030948e-01 6.15753114e-01 -4.50265199e-01
3.97039503e-01 -5.98279655e-01 -4.19943362e-01 2.06369430e-01
2.14414433e-01 3.52422059e-01 -4.03056353e-01 -6.60173893e-01
-3.16942126e-01 8.78530443e-01 -7.18053384e-03 1.32847965e-01
8.05065274e-01 -4.08229440e-01 -2.46854201e-01 5.51774025e-01
8.37549031e-01 -4.42143559e-01 -1.48475230e+00 9.19451863e-02
-9.82501209e-02 -4.18906689e-01 -1.59353629e-01 -1.01529098e+00
-1.37443662e+00 1.11557794e+00 6.49717629e-01 1.12775922e-01
1.07834661e+00 3.30256790e-01 8.67549479e-01 3.34827602e-01
3.49038303e-01 -1.06441212e+00 4.85050589e-01 3.80934447e-01
6.99614704e-01 -1.54481554e+00 -2.75506675e-01 -7.08986044e-01
-9.99176681e-01 9.57234144e-01 1.05555499e+00 3.22534204e-01
3.72145087e-01 3.03229336e-02 2.77096361e-01 -3.95364344e-01
-2.93800592e-01 -6.05417550e-01 5.30904710e-01 6.67670548e-01
1.80256397e-01 1.56363044e-02 5.68879209e-02 6.12422466e-01
4.16297376e-01 -3.06645602e-01 1.42016381e-01 5.12982190e-01
-4.17340726e-01 -7.93885767e-01 -4.31949079e-01 4.12843555e-01
-2.00024247e-01 1.65579394e-02 -7.21502662e-01 6.69808865e-01
4.73167747e-01 7.56503224e-01 1.57176957e-01 -1.83292016e-01
7.10588843e-02 7.62015730e-02 3.75954807e-01 -4.09299195e-01
-1.94067538e-01 -5.22322431e-02 -3.14301401e-01 -6.94112062e-01
-7.21104562e-01 -3.88051659e-01 -1.30825067e+00 1.08766861e-01
-1.54390574e-01 -2.34694302e-01 5.77750444e-01 1.02932274e+00
3.42426986e-01 6.97128356e-01 3.80313337e-01 -1.13853681e+00
8.11170489e-02 -6.51918709e-01 -4.29103792e-01 5.84447801e-01
2.69532442e-01 -8.13402236e-01 -2.25153998e-01 1.90503135e-01] | [9.694302558898926, 0.3177073299884796] |
83a7a280-1439-423d-88e3-5df89d2da98c | timestamp-supervised-action-segmentation-with | 2206.15031 | null | https://arxiv.org/abs/2206.15031v4 | https://arxiv.org/pdf/2206.15031v4.pdf | Timestamp-Supervised Action Segmentation with Graph Convolutional Networks | We introduce a novel approach for temporal activity segmentation with timestamp supervision. Our main contribution is a graph convolutional network, which is learned in an end-to-end manner to exploit both frame features and connections between neighboring frames to generate dense framewise labels from sparse timestamp labels. The generated dense framewise labels can then be used to train the segmentation model. In addition, we propose a framework for alternating learning of both the segmentation model and the graph convolutional model, which first initializes and then iteratively refines the learned models. Detailed experiments on four public datasets, including 50 Salads, GTEA, Breakfast, and Desktop Assembly, show that our method is superior to the multi-layer perceptron baseline, while performing on par with or better than the state of the art in temporal activity segmentation with timestamp supervision. | ['Quoc-Huy Tran', 'M. Zeeshan Zia', 'Andrey Konin', 'Shakeeb Siddiqui', 'Awais Ahmed', 'Sanjay Haresh', 'Hamza Khan'] | 2022-06-30 | null | null | null | null | ['action-segmentation'] | ['computer-vision'] | [ 5.47997236e-01 2.85353214e-01 -3.66209626e-01 -5.85057020e-01
-7.01648533e-01 -6.78988695e-01 7.07442164e-01 5.94255999e-02
-5.78797579e-01 6.31612897e-01 5.16478777e-01 -1.72615111e-01
3.69035453e-01 -4.76595193e-01 -1.02072489e+00 -6.49622262e-01
-3.09649408e-01 4.88354951e-01 4.18769658e-01 2.67901748e-01
-1.84301976e-02 2.55288500e-02 -1.24530935e+00 5.60788810e-01
4.88174081e-01 1.07426584e+00 -5.49376123e-02 1.00031543e+00
1.31120920e-01 1.52256835e+00 -3.94413829e-01 -3.86802107e-02
2.81866729e-01 -5.01085162e-01 -1.16868401e+00 2.82872468e-01
4.20900851e-01 -5.28949678e-01 -5.69611430e-01 2.61755824e-01
1.86015248e-01 4.11137640e-01 2.21722677e-01 -1.12886322e+00
-2.82494009e-01 8.88156772e-01 -7.02405035e-01 5.13090909e-01
5.05979836e-01 2.79903293e-01 1.13842714e+00 -4.02539521e-01
9.01513815e-01 1.06546843e+00 7.01020002e-01 4.34003413e-01
-1.38404179e+00 -4.34935927e-01 6.81434333e-01 -2.83626094e-03
-1.06070018e+00 -2.89298326e-01 5.42078316e-01 -4.70275134e-01
1.24437523e+00 -2.27809817e-01 1.07713640e+00 1.42268348e+00
-9.75761265e-02 1.22332919e+00 7.69341528e-01 -1.22606300e-01
7.52977133e-02 -7.31642842e-01 1.45108834e-01 9.52124834e-01
-3.97501647e-01 -7.75363445e-02 -7.66001999e-01 1.02662649e-02
1.01345229e+00 -1.93100691e-01 -1.74514696e-01 -2.39333019e-01
-1.65926456e+00 4.81970578e-01 3.05713594e-01 1.99623153e-01
-3.60476553e-01 7.40735173e-01 4.01279539e-01 -8.76091644e-02
6.63282692e-01 1.75490722e-01 -4.96583700e-01 -6.92023039e-01
-1.26154947e+00 2.22108752e-01 9.46194053e-01 9.52565074e-01
7.54551172e-01 -1.26888901e-01 -5.86492240e-01 2.89023995e-01
2.20871344e-01 1.03032261e-01 1.64374173e-01 -1.39849424e+00
5.19385397e-01 4.30755526e-01 8.42147022e-02 -3.52789074e-01
-5.63078046e-01 -1.20329067e-01 -4.35075611e-01 -1.98445722e-01
6.60624206e-01 -3.79030436e-01 -1.26657796e+00 1.76333356e+00
3.91073823e-01 1.13742137e+00 -2.71817476e-01 6.43185556e-01
7.02194750e-01 7.94107556e-01 2.16424853e-01 -5.24648279e-02
9.81959581e-01 -1.58063269e+00 -5.78170002e-01 -2.61926383e-01
7.55811810e-01 -3.51771384e-01 9.35932636e-01 2.81924218e-01
-1.36364806e+00 -7.30468750e-01 -8.38546753e-01 -4.03521389e-01
5.60990311e-02 -1.79882310e-02 8.49607885e-01 2.37538099e-01
-1.35599113e+00 8.29104900e-01 -1.70021462e+00 -2.85239965e-01
7.48257279e-01 3.90826017e-01 -8.29448327e-02 1.46274641e-01
-6.14130735e-01 3.64455700e-01 6.23472631e-01 -5.30982465e-02
-1.40819681e+00 -8.61264110e-01 -1.19010961e+00 3.00606452e-02
4.41199780e-01 -8.48024309e-01 1.67792869e+00 -1.02649689e+00
-1.63270903e+00 6.78389013e-01 -3.61929208e-01 -8.45826328e-01
6.76855803e-01 -2.54394591e-01 2.11958677e-01 2.98628956e-01
3.90177995e-01 1.14430892e+00 6.12188220e-01 -6.82329237e-01
-9.15525019e-01 1.31599158e-01 3.48342419e-01 3.13213408e-01
2.00439587e-01 4.48667631e-03 -1.09471166e+00 -5.60039401e-01
-2.68840104e-01 -1.00951385e+00 -3.98061246e-01 -2.61935353e-01
-4.61426795e-01 -3.82807195e-01 8.01449955e-01 -7.79607534e-01
1.04245138e+00 -1.95666242e+00 3.86261910e-01 1.85586065e-02
2.87886947e-01 -3.42872411e-01 -2.65913665e-01 2.19797313e-01
-9.00777057e-02 -1.26033485e-01 -4.10089910e-01 -9.88535225e-01
-1.58055022e-01 4.49924737e-01 1.33061307e-02 4.69168931e-01
3.31724077e-01 1.03074741e+00 -1.23791230e+00 -4.52114761e-01
2.29672343e-01 5.70998967e-01 -6.26556754e-01 3.84906292e-01
-7.10115314e-01 1.06564307e+00 -2.28153974e-01 4.86513078e-01
7.68653676e-02 -7.81891584e-01 3.56080592e-01 5.66895753e-02
-9.59641859e-02 6.85747683e-01 -1.01203537e+00 2.43594170e+00
-2.25502759e-01 6.70403898e-01 -1.63039282e-01 -8.90357792e-01
3.10492039e-01 3.58241439e-01 8.91118467e-01 -4.72662866e-01
1.33744150e-01 -3.96872371e-01 -4.11808282e-01 -3.98829758e-01
3.78756374e-01 3.52901459e-01 -2.24590331e-01 7.48745739e-01
4.27193850e-01 2.19978705e-01 7.37234890e-01 4.96800721e-01
1.45935249e+00 1.10630631e+00 -1.77452639e-01 9.30647478e-02
2.51791418e-01 -2.94908625e-03 7.96249747e-01 6.49666369e-01
-5.19665964e-02 8.40888798e-01 9.88656580e-01 -6.76040471e-01
-9.55801785e-01 -9.10999775e-01 4.77547675e-01 1.55880809e+00
5.80001287e-02 -7.82587171e-01 -9.19142008e-01 -1.06090307e+00
-2.26986110e-01 4.75654334e-01 -8.01450610e-01 1.87476516e-01
-1.05682969e+00 -4.29848224e-01 7.13465214e-01 1.05048835e+00
7.10498691e-01 -1.18570852e+00 -7.15124846e-01 4.11918461e-01
-5.64151049e-01 -1.53921759e+00 -8.10043693e-01 1.33295670e-01
-8.96496356e-01 -1.24657631e+00 -5.46834171e-01 -6.93030953e-01
4.48202282e-01 1.30929232e-01 1.50883949e+00 1.44536838e-01
1.95675530e-02 4.60795522e-01 -2.61103600e-01 -4.39268574e-02
-2.42485702e-01 5.57085097e-01 -4.81033981e-01 3.31180021e-02
7.40520060e-02 -6.17713332e-01 -6.75050318e-01 -6.97310269e-02
-6.51261568e-01 4.93903548e-01 1.12988353e-01 5.90380311e-01
7.23705828e-01 -3.78696799e-01 1.40793279e-01 -1.19841349e+00
1.22072682e-01 -4.98941362e-01 -7.52888381e-01 6.83239996e-02
-2.75478721e-01 9.02844891e-02 3.71064335e-01 -1.66316122e-01
-1.16598010e+00 5.32772839e-01 -5.35500795e-02 -5.22680819e-01
-3.00716847e-01 4.90661621e-01 1.31103158e-01 2.31407240e-01
4.38955635e-01 -1.14567755e-02 -2.88208514e-01 -5.18692791e-01
5.93775570e-01 -3.23822275e-02 9.00759280e-01 -7.27726817e-01
5.89059174e-01 7.80078948e-01 -1.40683219e-01 -5.25126696e-01
-9.87178087e-01 -6.16968274e-01 -1.03654432e+00 -1.33642361e-01
1.31376243e+00 -1.21447504e+00 -5.76499581e-01 7.18195736e-01
-1.19253099e+00 -1.30180681e+00 -4.31500047e-01 2.26592466e-01
-7.61280358e-01 4.49871540e-01 -1.11353111e+00 -4.42160636e-01
-2.49237433e-01 -9.86021161e-01 1.53092992e+00 3.18089306e-01
-3.04434150e-01 -1.21241903e+00 1.58337682e-01 6.77972496e-01
-3.79389450e-02 6.56876087e-01 4.32181478e-01 -4.42387372e-01
-1.00992560e+00 1.79173172e-01 -1.69443533e-01 2.99654063e-02
3.19719017e-02 1.95712015e-01 -9.69612598e-01 -2.84563750e-01
-4.81892198e-01 -4.58703816e-01 1.10135651e+00 6.64724708e-01
1.28734350e+00 -7.78586343e-02 -4.69094247e-01 1.00603282e+00
9.01108801e-01 -1.91588197e-02 6.46027565e-01 1.85113668e-01
1.21465003e+00 2.16971189e-01 6.18463039e-01 4.85787243e-01
7.89478600e-01 4.65592325e-01 3.36110741e-01 -3.55210543e-01
-1.29209444e-01 -2.00886905e-01 3.04415733e-01 3.93486679e-01
-3.71144086e-01 -3.58186424e-01 -8.56338501e-01 8.16382229e-01
-2.36090636e+00 -1.05559099e+00 -2.98881419e-02 1.78845143e+00
7.67372966e-01 1.92743063e-01 6.48019314e-01 -5.15803061e-02
4.91020560e-01 6.79048896e-01 -5.07297099e-01 -1.61772236e-01
6.26140311e-02 3.84297997e-01 5.53463280e-01 4.96432006e-01
-1.55846548e+00 1.32902753e+00 6.76096630e+00 3.04513484e-01
-8.87304187e-01 2.33655006e-01 7.53917158e-01 -4.81252521e-01
-3.78351659e-02 2.43804649e-01 -6.43609762e-01 4.02411520e-01
1.17095220e+00 1.90475732e-01 6.79535866e-01 5.59894562e-01
4.35846746e-01 -1.67724371e-01 -1.44172609e+00 7.28880465e-01
-2.95136124e-03 -1.68324649e+00 -4.18293506e-01 -6.81138262e-02
1.03297353e+00 5.42609870e-01 -3.88236284e-01 2.15845406e-01
1.07294476e+00 -9.40135121e-01 9.39722776e-01 3.40380937e-01
5.08132756e-01 -5.65413594e-01 4.30899829e-01 2.78647393e-01
-1.52566552e+00 2.02632770e-01 2.74732411e-01 -4.01074618e-01
5.52600980e-01 3.08626652e-01 -9.92750227e-01 4.59433913e-01
8.67269933e-01 1.44615042e+00 -5.30645669e-01 9.53658342e-01
-6.54558957e-01 8.91795456e-01 -4.05323863e-01 5.30119658e-01
6.17214859e-01 -6.70751780e-02 1.64978400e-01 1.46078360e+00
-6.10949360e-02 -4.91615906e-02 6.72736406e-01 7.36747801e-01
-2.93386698e-01 -4.26234603e-01 -4.42983299e-01 -3.15046549e-01
2.33972013e-01 1.40396762e+00 -1.20970583e+00 -7.60350466e-01
-6.85399711e-01 1.15685546e+00 3.92066866e-01 4.74046767e-01
-1.36209989e+00 1.09362520e-01 4.88263488e-01 3.84222642e-02
4.84994203e-01 -4.54166979e-01 -1.66913956e-01 -1.21748757e+00
-8.50307494e-02 -7.11963832e-01 8.73909950e-01 -6.81305647e-01
-8.13212633e-01 2.79417366e-01 1.33861050e-01 -7.29925036e-01
-4.70381349e-01 -6.80756941e-02 -7.33051419e-01 5.84784925e-01
-1.39945412e+00 -1.43659377e+00 -6.36353254e-01 6.49458051e-01
7.66386151e-01 5.51278770e-01 3.73647422e-01 3.57886553e-01
-7.57808030e-01 2.42586970e-01 -5.16998768e-01 4.23385888e-01
4.83287990e-01 -1.61674607e+00 1.10677242e+00 1.19857585e+00
4.03042734e-01 1.71595141e-01 2.09719002e-01 -7.98114240e-01
-1.05218291e+00 -1.40041614e+00 8.13602746e-01 -6.50732934e-01
5.96020937e-01 -6.53999209e-01 -6.77054822e-01 1.45883524e+00
5.57728648e-01 2.86658764e-01 5.64758658e-01 1.19169772e-01
-2.52223313e-01 2.34269634e-01 -6.38098180e-01 4.65056628e-01
1.56400800e+00 -4.37881470e-01 -3.43501627e-01 5.19807518e-01
9.31590199e-01 -1.06631839e+00 -9.66251194e-01 1.15906589e-01
4.67151910e-01 -9.01619971e-01 8.79627109e-01 -5.62119842e-01
4.11595047e-01 -3.21584433e-01 1.50301397e-01 -9.19687033e-01
-4.45961267e-01 -1.08314621e+00 -3.54037642e-01 1.23978364e+00
4.84033495e-01 -1.68996483e-01 1.22515774e+00 5.64697444e-01
-4.07574266e-01 -5.33560276e-01 -8.76446128e-01 -4.61423546e-01
-3.83392990e-01 -5.56031048e-01 4.62919086e-01 9.08108473e-01
-2.23463878e-01 5.18892467e-01 -4.01809514e-01 1.04394145e-01
4.77739573e-01 1.16113499e-01 9.14379954e-01 -9.26604211e-01
-6.08535230e-01 -1.23090722e-01 -5.72731793e-02 -1.57367909e+00
3.68163258e-01 -7.78287351e-01 2.13877603e-01 -1.80196404e+00
1.55494407e-01 -1.81757703e-01 -4.01548147e-01 8.92820418e-01
-1.55025065e-01 3.96200746e-01 1.38392478e-01 2.41849516e-02
-1.24856627e+00 2.25974634e-01 1.16557050e+00 -2.17187002e-01
-5.36531746e-01 -4.24688309e-02 -5.11791766e-01 8.61059427e-01
5.62656522e-01 -4.61360872e-01 -5.37711382e-01 -7.79525638e-01
1.02182403e-01 1.93426311e-01 4.09593642e-01 -1.12692821e+00
1.37977943e-01 -6.78537041e-02 3.44694942e-01 -7.92588949e-01
3.84091735e-01 -4.09947932e-01 3.64141278e-02 2.86540329e-01
-3.82024676e-01 1.26062244e-01 2.31751613e-02 7.90735185e-01
7.55646825e-03 3.26210588e-01 4.23963547e-01 -3.15038353e-01
-7.97760546e-01 5.61909199e-01 -4.45460796e-01 2.51249999e-01
8.47028613e-01 -2.52458572e-01 -2.51389951e-01 -4.64782536e-01
-8.73479724e-01 6.19154930e-01 4.43159103e-01 4.49419618e-01
2.74267551e-02 -1.23324740e+00 -5.42214692e-01 -1.14796218e-02
-1.91529945e-01 5.98593056e-01 -3.57222855e-02 1.14978135e+00
-5.94487429e-01 1.60549656e-01 8.48025084e-02 -1.11949790e+00
-1.12056792e+00 3.86407703e-01 2.40880758e-01 -6.08932793e-01
-8.57249737e-01 1.03602576e+00 1.49652779e-01 -1.24604575e-01
4.63884711e-01 -1.04748476e+00 2.87258457e-02 9.72663332e-03
2.99014479e-01 4.93889660e-01 -1.40841618e-01 -5.49685299e-01
-3.07300985e-01 3.77534389e-01 2.43850313e-02 -3.27106506e-01
1.40165722e+00 -4.36246581e-02 -1.69886053e-02 5.13492465e-01
1.02641368e+00 -3.77041101e-01 -2.13153100e+00 -1.40512139e-01
1.56873673e-01 -2.27265865e-01 -1.95774361e-01 -5.00001967e-01
-1.25192583e+00 6.81676328e-01 1.09604955e-01 1.20551154e-01
1.12158716e+00 7.68821314e-02 1.14116240e+00 2.49242917e-01
2.29345739e-01 -1.09169745e+00 4.21251237e-01 7.01201379e-01
2.13984981e-01 -1.07328236e+00 -2.38250121e-01 -4.13976520e-01
-5.35906434e-01 9.02415693e-01 5.71421266e-01 -3.16570431e-01
2.20657304e-01 4.62943673e-01 -5.92760816e-02 -2.45035887e-01
-9.08861220e-01 -2.71787375e-01 1.12516314e-01 5.80976665e-01
6.11769438e-01 -2.40618631e-01 1.01698093e-01 3.85840029e-01
6.38152286e-02 4.73363370e-01 2.73008645e-01 1.21667469e+00
-1.15961507e-01 -1.07802725e+00 6.15673773e-02 3.36026192e-01
-6.50025785e-01 -8.32994506e-02 -3.05921495e-01 7.27712572e-01
2.86546946e-01 9.96037543e-01 4.34647113e-01 -2.51456112e-01
1.80028677e-01 2.25443438e-01 6.47268713e-01 -8.67232561e-01
-8.87158811e-01 4.24648464e-01 6.29035413e-01 -1.20529425e+00
-7.47878075e-01 -8.06396127e-01 -1.56392813e+00 -1.44282624e-01
2.12670714e-02 -3.29294950e-02 1.14871912e-01 1.13291073e+00
4.20372933e-01 8.51552248e-01 2.54357874e-01 -1.32787430e+00
3.60507071e-01 -8.46664250e-01 -9.88931432e-02 5.84130406e-01
2.59684831e-01 -3.49042416e-01 -3.62251513e-02 6.58561051e-01] | [8.579451560974121, 0.494861364364624] |
64dc04eb-1476-4ddf-8e03-56120c7bd142 | monte-carlo-q-learning-for-general-game | 1802.05944 | null | http://arxiv.org/abs/1802.05944v2 | http://arxiv.org/pdf/1802.05944v2.pdf | Monte Carlo Q-learning for General Game Playing | After the recent groundbreaking results of AlphaGo, we have seen a strong
interest in reinforcement learning in game playing. General Game Playing (GGP)
provides a good testbed for reinforcement learning. In GGP, a specification of
games rules is given. GGP problems can be solved by reinforcement learning.
Q-learning is one of the canonical reinforcement learning methods, and has been
used by (Banerjee & Stone, IJCAI 2007) in GGP. In this paper we implement
Q-learning in GGP for three small-board games (Tic-Tac-Toe, Connect Four, Hex),
to allow comparison to Banerjee et al. As expected, Q-learning converges,
although much slower than MCTS. Borrowing an idea from MCTS, we enhance
Q-learning with Monte Carlo Search, to give QM-learning. This enhancement
improves the performance of pure Q-learning. We believe that QM-learning can
also be used to improve performance of reinforcement learning further for
larger games, something which we will test in future work. | ['Aske Plaat', 'Hui Wang', 'Michael Emmerich'] | 2018-02-16 | null | null | null | null | ['board-games'] | ['playing-games'] | [-4.17616248e-01 1.38633296e-01 -1.06185839e-01 2.47366577e-01
-8.27653408e-01 -7.30028808e-01 4.42562133e-01 -1.02755344e-02
-6.25382364e-01 1.23536301e+00 7.20195249e-02 -7.32184708e-01
-3.27159941e-01 -1.17384899e+00 -5.68216205e-01 -6.06768787e-01
-5.38134217e-01 3.85325372e-01 6.66070282e-01 -8.32044601e-01
4.58408654e-01 6.98810965e-02 -1.42121017e+00 1.80155560e-01
5.69879413e-01 4.08363670e-01 1.96896538e-01 1.11749732e+00
9.86544937e-02 1.36722982e+00 -5.33874691e-01 -4.34243441e-01
5.09705603e-01 -8.10229182e-01 -1.19641995e+00 -1.02825254e-01
-2.29150504e-01 -3.32980335e-01 -3.42395872e-01 7.30183601e-01
6.35953486e-01 4.55175608e-01 4.00970876e-01 -1.28907657e+00
-2.16611959e-02 5.59424698e-01 -6.38716459e-01 2.91958928e-01
5.56242049e-01 5.45813978e-01 1.18945765e+00 -2.23349601e-01
4.79142517e-01 1.11982381e+00 5.29092073e-01 7.09838867e-01
-8.77219677e-01 -7.17642188e-01 -9.70734358e-02 4.69530672e-01
-8.36072505e-01 2.05782577e-01 4.36459363e-01 2.27233730e-02
1.22078454e+00 1.93124250e-01 1.37000024e+00 7.17719197e-01
3.51093799e-01 1.24479616e+00 1.64226389e+00 -6.77308679e-01
8.17912877e-01 -5.09144485e-01 -5.57526886e-01 6.54067099e-01
-2.99814224e-01 9.34044123e-01 -4.10940766e-01 -1.16786197e-01
1.09172475e+00 -3.67416948e-01 2.89991438e-01 -6.70830548e-01
-7.18987584e-01 1.36336958e+00 2.91949719e-01 1.30233780e-01
-3.25572580e-01 7.44502127e-01 5.10135055e-01 8.67974699e-01
1.12215154e-01 7.54953623e-01 -5.87472260e-01 -1.01982081e+00
-8.88614237e-01 1.00535822e+00 1.04806471e+00 5.34404695e-01
7.30137765e-01 2.26785257e-01 8.88401046e-02 6.68969989e-01
2.27056935e-01 1.75842360e-01 4.68145818e-01 -1.58988297e+00
2.41243199e-01 -3.96193974e-02 2.08148763e-01 -2.08303928e-01
-4.86858487e-01 -5.72009347e-02 -3.55749249e-01 9.00266826e-01
5.41299999e-01 -5.69297850e-01 -5.03880262e-01 1.49536014e+00
1.86210528e-01 4.51026380e-01 1.44391671e-01 6.08307481e-01
1.97211221e-01 7.81850040e-01 -1.76122710e-02 -3.42006266e-01
9.63420153e-01 -1.02525461e+00 -2.86573768e-01 -3.16153318e-02
8.60722184e-01 -5.31583071e-01 9.68828380e-01 8.87095511e-01
-1.37008965e+00 -2.65092611e-01 -1.00056207e+00 6.20063126e-01
-1.40206963e-01 -8.57029736e-01 1.14404202e+00 1.10079777e+00
-1.39608037e+00 9.08654332e-01 -8.73474717e-01 -4.45174873e-01
3.21815431e-01 5.72570980e-01 1.20235130e-01 5.32489037e-03
-1.48728371e+00 1.05832696e+00 7.13865757e-01 -5.34892857e-01
-1.07655132e+00 -2.51020342e-01 -7.31715262e-01 -1.56218350e-01
9.19003665e-01 -5.29926717e-01 2.04900646e+00 -7.39715874e-01
-2.25014544e+00 5.23435771e-01 5.22399664e-01 -7.37012684e-01
5.85310817e-01 4.05307412e-02 -4.87260055e-03 8.56631771e-02
3.85401919e-02 8.21161628e-01 4.63422865e-01 -1.07533967e+00
-1.03528261e+00 -3.00583486e-02 8.24666977e-01 5.39716899e-01
3.72190326e-01 1.31906435e-01 -4.14737239e-02 -4.07496005e-01
-4.37744290e-01 -8.68136942e-01 -7.42074370e-01 -5.25048137e-01
1.52746141e-01 -4.56704736e-01 5.79177141e-02 -2.38685613e-03
1.10775042e+00 -1.77427590e+00 -8.55218545e-02 1.97284982e-01
1.29727125e-01 2.26577491e-01 -3.77519250e-01 9.10900056e-01
-7.55440667e-02 4.29445393e-02 -2.15933956e-02 2.99610436e-01
2.89252639e-01 7.93862820e-01 1.22506693e-01 1.91109419e-01
-8.04407150e-02 1.17291367e+00 -1.26816249e+00 -4.31113958e-01
3.89881074e-01 -3.45727324e-01 -1.10453141e+00 1.42405868e-01
-3.81545812e-01 1.50963858e-01 -4.06632334e-01 3.22785854e-01
2.87709028e-01 1.11970946e-01 2.75565207e-01 7.92211771e-01
-2.24289387e-01 3.20619941e-01 -1.54962730e+00 1.61309230e+00
-4.47707087e-01 2.94370174e-01 -7.44668618e-02 -1.32801425e+00
5.84273398e-01 3.99279267e-01 6.78849816e-01 -1.14882410e+00
-4.07263339e-02 8.34453478e-02 3.79544348e-01 -4.20570314e-01
5.70364416e-01 -6.74264908e-01 -4.11266088e-01 6.17142498e-01
2.09637836e-01 -7.84051657e-01 8.44636858e-01 1.46493182e-01
1.27547383e+00 4.97022063e-01 6.63703263e-01 -3.66539806e-01
1.95614502e-01 1.75819322e-01 2.82681018e-01 1.30073369e+00
-4.09549952e-01 1.65763825e-01 7.87471712e-01 -4.60525900e-01
-8.85640264e-01 -1.16431034e+00 2.16464832e-01 1.42369580e+00
6.23559300e-03 -7.60705769e-01 -6.94012165e-01 -6.54546976e-01
-1.20650187e-01 6.21532202e-01 -6.24060929e-01 6.58494746e-03
-6.45774782e-01 -7.31278896e-01 4.29762453e-01 5.30503333e-01
4.55193609e-01 -1.75884569e+00 -8.22270036e-01 7.85985768e-01
2.47098476e-01 -4.14011836e-01 -1.10430811e-02 6.26777411e-01
-9.21192706e-01 -1.25333166e+00 -6.50481939e-01 -6.44312680e-01
-2.52263337e-01 5.26015721e-02 1.33650017e+00 1.85963482e-01
5.75468503e-03 5.73889136e-01 -7.18440592e-01 -5.36845803e-01
-6.35379434e-01 1.35515079e-01 -7.41384104e-02 -9.46908951e-01
5.91897033e-02 -6.88238800e-01 -5.60687423e-01 1.86272159e-01
-8.51089835e-01 3.75905656e-03 3.25564295e-01 1.11510265e+00
2.01520026e-01 2.89480209e-01 5.41377842e-01 -9.52029645e-01
9.95259881e-01 -3.01260203e-01 -7.58447111e-01 -2.74624258e-01
-4.64854568e-01 1.47210687e-01 7.00991035e-01 -1.52250350e-01
-5.70031941e-01 -3.05407643e-01 -6.42969608e-01 3.04501116e-01
-1.10466117e-02 6.22662663e-01 3.23295921e-01 -2.26946488e-01
6.93226159e-01 1.96433291e-01 -7.94635788e-02 -1.49692828e-03
5.58051050e-01 2.43907347e-01 5.42156324e-02 -9.04673636e-01
5.72778344e-01 -6.54488429e-02 -8.63429084e-02 -3.72265339e-01
-4.95410204e-01 -2.39904255e-01 -1.18038505e-01 -2.87564635e-01
5.99430084e-01 -7.55937874e-01 -1.31926668e+00 2.50834256e-01
-5.67229390e-01 -1.10142779e+00 -8.08538020e-01 3.91061485e-01
-1.40536058e+00 2.11747497e-01 -8.91229987e-01 -8.93636167e-01
2.26273879e-01 -1.01910543e+00 5.43261588e-01 5.20343959e-01
-4.04185057e-02 -1.08765781e+00 7.24071562e-01 2.52573431e-01
2.39702106e-01 -2.44372897e-02 5.43366134e-01 -3.70083809e-01
-3.08922619e-01 2.44572982e-01 2.34335005e-01 1.14197537e-01
-1.96404785e-01 -1.77249089e-01 -4.50688392e-01 -4.97444540e-01
-2.74285138e-01 -8.15226436e-01 5.90707183e-01 5.90712607e-01
9.52609420e-01 -8.31069052e-02 1.90760419e-01 1.69570357e-01
1.78111446e+00 6.16454065e-01 9.19466317e-01 1.10952508e+00
7.06121251e-02 2.00314313e-01 8.87583196e-01 7.72344410e-01
4.88400161e-01 5.36100805e-01 6.39178753e-01 -9.30896550e-02
3.37936610e-01 -3.33717465e-01 4.99558657e-01 7.01261699e-01
-4.29306597e-01 -6.94968775e-02 -9.21181321e-01 4.53660607e-01
-2.09354925e+00 -1.28556025e+00 1.28852859e-01 1.81522977e+00
9.46045041e-01 5.01019716e-01 1.00921774e+00 3.65236610e-01
2.68919051e-01 1.17892690e-01 -3.66836488e-01 -1.01155674e+00
2.09413856e-01 1.10401404e+00 5.20714760e-01 7.38371432e-01
-7.96249866e-01 1.31866097e+00 7.22123098e+00 1.13421977e+00
-6.33405507e-01 1.28634661e-01 4.00448829e-01 -1.04490876e-01
-6.40481710e-02 3.53818446e-01 -2.02874511e-01 4.56167966e-01
9.91273105e-01 -3.73466104e-01 8.28254223e-01 7.42210507e-01
4.14895028e-01 -5.88987291e-01 -7.42300630e-01 7.50712335e-01
-5.47322571e-01 -1.26982367e+00 -5.18965125e-01 1.87915310e-01
9.18263495e-01 2.87679676e-02 -2.09502831e-01 1.20829260e+00
1.39986944e+00 -1.08873940e+00 3.82224292e-01 -2.78393090e-01
3.03776562e-01 -1.43349254e+00 7.57970691e-01 5.09223521e-01
-1.11242878e+00 -2.50475109e-01 -5.40940344e-01 -8.04990411e-01
4.24604304e-03 -1.37236357e-01 -6.18849933e-01 7.48665631e-01
4.79780555e-01 4.51761633e-01 -2.61927873e-01 1.35915613e+00
-6.14515245e-01 9.89004314e-01 -1.22221895e-01 -5.03027141e-01
8.58474076e-01 -3.18666428e-01 1.84543822e-02 9.16524351e-01
-5.75223714e-02 2.81540424e-01 5.11007845e-01 4.10186827e-01
3.06886435e-01 1.52719617e-01 -4.23614800e-01 1.76302150e-01
1.29698098e-01 9.91826832e-01 -8.38861644e-01 -2.35014945e-01
-4.83466566e-01 7.37358928e-01 3.03656936e-01 1.26145914e-01
-7.90601015e-01 -2.27427781e-01 5.50822139e-01 -4.42002714e-02
5.15195310e-01 -1.98449105e-01 1.22252941e-01 -8.75244558e-01
-5.95585585e-01 -1.36688912e+00 4.23535287e-01 -7.62055576e-01
-1.00974381e+00 5.21512851e-02 1.26619503e-01 -1.13795948e+00
-8.65354955e-01 -6.41763210e-01 -7.42209375e-01 5.56094766e-01
-1.47084796e+00 -6.86850488e-01 2.82214433e-01 7.16725349e-01
6.59191728e-01 -1.28124952e-01 8.36131096e-01 -1.97736725e-01
-1.73051387e-01 2.63713092e-01 3.82468432e-01 2.99202744e-02
3.68814796e-01 -1.89767754e+00 6.08817101e-01 3.75487417e-01
4.34925020e-01 -5.36415912e-02 8.11659634e-01 -3.01754892e-01
-1.32010925e+00 -4.64797199e-01 1.84460133e-01 -4.62843291e-02
1.01831353e+00 -7.34563842e-02 -4.24517006e-01 4.34770316e-01
6.30353987e-01 -2.92561412e-01 6.51681185e-01 1.97663665e-01
1.97548866e-01 1.13645278e-01 -7.70514071e-01 8.51290464e-01
8.45202565e-01 -3.31847548e-01 -7.55981326e-01 1.69469893e-01
2.49554619e-01 -6.07384145e-01 -7.63251781e-01 -3.47937010e-02
3.91584665e-01 -1.42887390e+00 7.77116477e-01 -7.20680714e-01
6.27917886e-01 -1.12144269e-01 1.73633561e-01 -1.87423348e+00
-4.90707517e-01 -1.02678120e+00 -3.21463197e-02 5.87869525e-01
1.11500032e-01 -4.92802739e-01 1.27648270e+00 1.75480656e-02
9.12659615e-02 -9.77059305e-01 -9.43847537e-01 -1.04283643e+00
9.20672834e-01 -7.10480869e-01 4.30271864e-01 6.41178548e-01
6.99472129e-01 1.54733330e-01 -4.93349731e-01 -4.83135194e-01
4.16996658e-01 -3.25167067e-02 8.22493613e-01 -8.52652073e-01
-1.09143019e+00 -6.21658385e-01 -3.59681517e-01 -1.13038301e+00
-8.30718800e-02 -6.97324753e-01 3.53689156e-02 -1.47472644e+00
6.36838982e-03 -2.04568416e-01 -3.65781039e-01 5.18589973e-01
-1.64653420e-01 3.32706511e-01 6.65361822e-01 -2.70673871e-01
-9.84741807e-01 4.42920178e-01 1.79733145e+00 2.27177888e-01
-1.70967415e-01 2.60520548e-01 -6.82818770e-01 5.39840639e-01
1.21930838e+00 -4.73800659e-01 -4.23896343e-01 2.44307697e-01
5.02199531e-01 6.58709884e-01 5.91159537e-02 -1.08932853e+00
6.64433241e-02 -6.03398502e-01 3.62388343e-02 -9.51507464e-02
1.03312405e-02 -4.16374445e-01 -3.00525188e-01 9.50851560e-01
-1.84860960e-01 3.33101273e-01 3.26707721e-01 1.95097849e-01
-2.84529120e-01 -6.62594378e-01 6.39693141e-01 -6.61153197e-01
-8.60070825e-01 1.41785815e-01 -1.23273540e+00 4.59482491e-01
1.06867766e+00 -5.08728385e-01 1.40640184e-01 -8.23527634e-01
-8.73285055e-01 4.94628280e-01 4.20926899e-01 -1.07452430e-01
3.84818494e-01 -1.21734798e+00 -6.63614750e-01 -9.42269415e-02
-1.85055077e-01 -3.03992629e-01 5.78573532e-02 5.33706367e-01
-7.78683662e-01 8.85560662e-02 -6.48539484e-01 -5.09866700e-02
-9.95218933e-01 4.46675867e-01 5.27596176e-01 -9.42450404e-01
-4.39663142e-01 5.79205394e-01 -1.71636328e-01 -4.98069435e-01
-3.97120863e-02 -1.69765636e-01 -4.80813235e-02 -2.58817524e-01
4.13646281e-01 2.41440549e-01 -1.24035552e-01 3.60210747e-01
-8.38287026e-02 1.85455292e-01 -9.79222879e-02 -6.73278034e-01
1.55836129e+00 2.11601004e-01 3.30154359e-01 3.30111444e-01
5.22046089e-01 -2.00599179e-01 -1.42276168e+00 2.01423809e-01
1.67388707e-01 -2.01326057e-01 -1.83959529e-01 -8.48352969e-01
-8.52073491e-01 7.50445724e-01 5.71327090e-01 6.73012435e-01
1.11197376e+00 -2.79937714e-01 4.79746252e-01 3.27875614e-01
1.00262034e+00 -1.36554730e+00 4.27718937e-01 9.05903637e-01
4.42810684e-01 -9.87974107e-01 -4.18927744e-02 1.53334990e-01
-7.63688743e-01 1.16687965e+00 6.57645702e-01 -7.35726714e-01
3.09316933e-01 5.70990205e-01 -8.69049802e-02 -8.94491151e-02
-1.08274388e+00 -7.46903241e-01 -5.88410020e-01 8.58292997e-01
3.27008277e-01 1.18087806e-01 -6.01081550e-01 1.75182536e-01
-5.33599555e-01 3.54601920e-01 8.80656600e-01 1.31846046e+00
-6.80320859e-01 -1.82470500e+00 -3.26364219e-01 3.96704942e-01
-3.96613687e-01 -2.23404154e-01 -1.69900402e-01 1.19252276e+00
7.44927600e-02 9.96036530e-01 -6.12604804e-02 -2.16117397e-01
2.95128554e-01 -1.57085687e-01 1.20253909e+00 -5.99605024e-01
-1.09786379e+00 5.79865761e-02 1.45448193e-01 -7.45402038e-01
-4.82557267e-01 -5.54104090e-01 -1.20749879e+00 -7.38081992e-01
-1.38205335e-01 5.96335053e-01 2.43776217e-01 1.02519512e+00
-3.82847071e-01 4.84295517e-01 6.47200286e-01 -6.59923732e-01
-7.16450751e-01 -7.17035651e-01 -1.13328362e+00 1.63897708e-01
-9.46199447e-02 -6.72651470e-01 -7.62028173e-02 -4.92237896e-01] | [3.573046922683716, 1.5335273742675781] |
f4756435-3ad4-4f93-abb3-35f925944885 | broad-linguistic-complexity-analysis-for | null | null | https://aclanthology.org/2021.bea-1.5 | https://aclanthology.org/2021.bea-1.5.pdf | Broad Linguistic Complexity Analysis for Greek Readability Classification | This paper explores the linguistic complexity of Greek textbooks as a readability classification task. We analyze textbook corpora for different school subjects and textbooks for Greek as a Second Language, covering a very wide spectrum of school age groups and proficiency levels. A broad range of quantifiable linguistic complexity features (lexical, morphological and syntactic) are extracted and calculated. Conducting experiments with different feature subsets, we show that the different linguistic dimensions contribute orthogonal information, each contributing towards the highest result achieved using all linguistic feature subsets. A readability classifier trained on this basis reaches a classification accuracy of 88.16% for the Greek as a Second Language corpus. To investigate the generalizability of the classification models, we also perform cross-corpus evaluations. We show that the model trained on the most varied text collection (for Greek as a school subject) generalizes best. In addition to advancing the state of the art for Greek readability analysis, the paper also contributes insights on the role of different feature sets and training setups for generalizable readability classification. | ['Detmar Meurers', 'Maria Giagkou', 'Savvas Chatzipanagiotidis'] | null | null | null | null | eacl-bea-2021-4 | ['cross-corpus'] | ['computer-vision'] | [-1.34251997e-01 6.71037957e-02 -9.41447765e-02 -1.75166056e-01
-1.10118604e+00 -8.79326880e-01 7.63338089e-01 7.59428740e-01
-5.68153799e-01 5.28560996e-01 6.10550880e-01 -2.94464886e-01
-7.28831887e-01 -9.13004518e-01 -1.76349327e-01 -2.26995066e-01
4.47148353e-01 2.91380882e-01 2.95751840e-01 -3.77662003e-01
5.45539498e-01 2.66222686e-01 -1.93214488e+00 7.53134415e-02
1.56652260e+00 5.75021386e-01 2.40251333e-01 6.77910149e-01
-1.29457250e-01 5.22123933e-01 -9.14091706e-01 -7.45789587e-01
7.18615577e-02 -5.24164796e-01 -1.08930147e+00 -8.66984129e-02
9.61488366e-01 2.01657802e-01 -3.35804641e-01 8.61581266e-01
4.16433394e-01 2.19495520e-01 1.03754342e+00 -6.21967733e-01
-9.68803287e-01 8.00306678e-01 1.37930050e-01 4.13398683e-01
9.71623778e-01 9.02831033e-02 1.38662159e+00 -4.38480675e-01
5.74943781e-01 1.00089705e+00 4.74980831e-01 2.22514644e-01
-9.79838669e-01 -2.12559044e-01 -2.25995913e-01 3.47679168e-01
-1.11881030e+00 -1.79180607e-01 4.49679971e-01 -7.47524559e-01
9.60204959e-01 4.72432494e-01 7.64932752e-01 8.73445511e-01
1.81657344e-01 6.04308784e-01 1.72981942e+00 -8.71766150e-01
-6.87100226e-03 3.12899739e-01 8.37565184e-01 6.19739413e-01
6.10380888e-01 -1.63860261e-01 -6.35927081e-01 3.20324093e-01
9.12797824e-02 -5.75051248e-01 -4.87121254e-01 1.30216122e-01
-1.17775655e+00 7.78333068e-01 -1.23201624e-01 6.56903327e-01
2.27943048e-01 -6.03769422e-01 3.72924864e-01 9.33773637e-01
2.17469111e-01 9.18687403e-01 -6.02402687e-01 -6.30864024e-01
-1.01215792e+00 2.08922535e-01 1.28961504e+00 9.39229727e-01
3.35807472e-01 -2.94341087e-01 -3.62552911e-01 1.10823464e+00
1.16808541e-01 6.08256757e-01 8.43630254e-01 -5.37921548e-01
7.32998490e-01 1.02265060e+00 -4.82393235e-01 -8.15181673e-01
-6.86463594e-01 -3.82477909e-01 -1.97906464e-01 4.85786758e-02
1.01069522e+00 8.34619999e-02 -3.60148132e-01 1.53535199e+00
-8.42952877e-02 -8.15220296e-01 -1.23977419e-02 4.13231105e-01
1.27467525e+00 4.42535996e-01 1.07782803e-01 4.59364466e-02
1.70599091e+00 -8.16070020e-01 -3.64865094e-01 1.18042856e-01
9.26077485e-01 -9.88499045e-01 1.58204079e+00 7.20358193e-01
-1.39510727e+00 -7.84128726e-01 -1.16337812e+00 -4.82097894e-01
-6.90928996e-01 3.25530738e-01 2.64229596e-01 1.20975077e+00
-9.85280991e-01 6.48198843e-01 -5.98195791e-01 -5.49650908e-01
-6.62356913e-02 1.87051386e-01 -3.68351668e-01 1.07924260e-01
-9.03729200e-01 1.06256986e+00 2.39747852e-01 -7.54970431e-01
1.66259900e-01 -7.86722600e-01 -8.47660303e-01 3.13896626e-01
-9.84977037e-02 -3.75354588e-01 9.47525978e-01 -6.78621411e-01
-1.72234857e+00 1.29547179e+00 3.28513086e-01 1.98527724e-01
4.95822698e-01 -3.88159603e-02 -4.60096300e-01 -2.37219948e-02
2.11407185e-01 1.44623928e-02 1.00206062e-01 -4.73898888e-01
-7.92852402e-01 -2.68585801e-01 -2.78870147e-02 1.26457319e-01
-7.93280363e-01 1.76585808e-01 -1.53227031e-01 -7.30312347e-01
1.06127262e-02 -7.88615346e-01 5.71529925e-01 -5.80917060e-01
-1.08442813e-01 -7.06409574e-01 -2.45336205e-01 -9.59490776e-01
1.88476872e+00 -1.81325483e+00 4.50394183e-01 4.29596245e-01
2.36484379e-01 3.18186097e-02 -1.31942034e-01 5.66503882e-01
1.29466772e-01 2.66227812e-01 2.96557456e-01 -9.49938446e-02
6.09583557e-01 -2.60874093e-01 2.69088387e-01 2.35842317e-01
1.25412479e-01 8.92738342e-01 -8.16930771e-01 -4.87161756e-01
1.90414369e-01 6.59072399e-02 -5.81877708e-01 1.74247194e-02
1.25903279e-01 1.53078020e-01 -3.61731082e-01 4.04549956e-01
2.65009612e-01 -1.20632336e-01 1.75911747e-02 3.28273118e-01
-3.43839288e-01 7.88507462e-01 -9.90045607e-01 1.51554108e+00
-9.54244494e-01 9.08347070e-01 -7.03655958e-01 -5.52138567e-01
1.09837914e+00 -6.57809824e-02 -1.79384410e-01 -1.06635249e+00
2.62246162e-01 5.06863832e-01 5.11238158e-01 -7.21798897e-01
6.28144979e-01 3.81388336e-01 -4.49677020e-01 5.13646007e-01
3.64679545e-01 -2.67610103e-01 8.58177900e-01 -4.87285154e-03
1.24819243e+00 3.66040543e-02 5.22445917e-01 -9.86417472e-01
9.86882746e-01 -2.87946582e-01 -7.21280649e-02 6.92957044e-01
-4.34828661e-02 4.73367542e-01 5.39346159e-01 -5.28589115e-02
-9.48659003e-01 -1.25479341e+00 -5.28602064e-01 1.43102527e+00
-2.78221339e-01 -6.58837378e-01 -9.08376276e-01 -4.37056184e-01
-7.31539279e-02 6.76797569e-01 -4.93699819e-01 -2.02876866e-01
-4.71928537e-01 -5.02808511e-01 4.90942657e-01 2.81993926e-01
4.11129028e-01 -9.98168528e-01 -7.58087456e-01 -4.01409790e-02
-2.27755100e-01 -9.88414824e-01 -1.85722649e-01 1.10984959e-01
-4.70002651e-01 -1.19190717e+00 -4.76579666e-01 -1.07559586e+00
1.95691586e-01 -2.27560624e-01 1.66486537e+00 4.84243840e-01
-1.87233657e-01 7.31261313e-01 -6.14350557e-01 -3.02336425e-01
-8.24737191e-01 7.10774064e-01 -8.62073153e-02 -8.17405939e-01
8.42683911e-01 -2.94693589e-01 -1.95882708e-01 -3.51707712e-02
-5.19774973e-01 -3.91050547e-01 3.73828441e-01 5.94212711e-01
1.56141622e-02 -3.74558009e-02 4.58524644e-01 -6.14972174e-01
1.00476229e+00 -3.17380309e-01 -5.41613936e-01 5.45192719e-01
-4.60503131e-01 1.61993116e-01 8.31053436e-01 -4.90612328e-01
-8.79040003e-01 -5.85086763e-01 -2.55449742e-01 7.11066306e-01
-4.11289960e-01 4.57975239e-01 -2.49327913e-01 -1.85669050e-01
9.03620660e-01 1.63960487e-01 -2.41136551e-01 -6.60029113e-01
1.22845098e-01 8.29512537e-01 3.06392401e-01 -1.00864911e+00
6.84388220e-01 -2.89435565e-01 5.21616079e-02 -1.26164520e+00
-9.77553129e-01 -2.72819936e-01 -1.08244145e+00 -1.27278358e-01
6.81731582e-01 -7.08835959e-01 -8.97467375e-01 4.81664479e-01
-5.72090507e-01 -3.42083931e-01 -3.80418718e-01 6.93467081e-01
-6.04548693e-01 5.46271265e-01 -6.23615921e-01 -3.45221072e-01
9.18712765e-02 -1.17615771e+00 9.19675231e-01 4.35827911e-01
-7.41095781e-01 -1.39045036e+00 3.58768106e-01 6.23330534e-01
8.99432693e-03 1.40465200e-01 1.33357239e+00 -1.03969049e+00
-2.08865076e-01 -3.24471965e-02 2.23340113e-02 1.75346360e-01
8.74798149e-02 2.39876211e-01 -8.20710719e-01 -1.78602725e-01
-1.07262775e-01 -6.29310369e-01 8.08266401e-01 1.17689833e-01
7.37763464e-01 -3.24411020e-02 2.33859390e-01 3.36562097e-01
1.35093606e+00 -2.21956953e-01 5.29206038e-01 8.26641798e-01
4.66868192e-01 8.48839164e-01 4.03692871e-01 1.72682196e-01
6.35954201e-01 5.56533813e-01 -5.06101549e-01 4.43709165e-01
-4.67318237e-01 -4.56887744e-02 4.84164536e-01 1.55608046e+00
-7.00670183e-02 -2.38842025e-01 -1.18299687e+00 5.96045852e-01
-1.04314184e+00 -8.09061944e-01 -2.21015394e-01 2.26107526e+00
9.81922388e-01 1.52783096e-01 5.84313750e-01 6.78163648e-01
1.88119158e-01 -1.66449517e-01 2.76154131e-01 -8.37004423e-01
-2.56830215e-01 7.86234319e-01 1.27756298e-01 6.63009763e-01
-8.20320666e-01 8.82109761e-01 6.18119097e+00 9.67408597e-01
-7.51215994e-01 -2.03632727e-01 2.01994687e-01 4.87395003e-02
-3.69447738e-01 -4.03860927e-01 -9.75066245e-01 4.31255132e-01
1.31446481e+00 -3.31928581e-01 1.89257681e-01 2.77538896e-01
-2.82624960e-02 -3.82891655e-01 -1.34323120e+00 6.86215937e-01
3.14311475e-01 -6.58738196e-01 -2.71358993e-02 9.65030938e-02
9.29110944e-01 -1.49101466e-01 2.29985207e-01 4.26883191e-01
1.94124326e-01 -1.08467412e+00 9.30924654e-01 5.44268847e-01
6.86212599e-01 -7.76796460e-01 4.48291957e-01 2.27439344e-01
-8.74635398e-01 -3.32285672e-01 -4.09884393e-01 -6.30644500e-01
-4.99954641e-01 2.28068754e-01 -2.27926835e-01 4.19479519e-01
6.52443349e-01 4.95176166e-01 -1.42762077e+00 1.00916052e+00
-2.38063723e-01 8.41179430e-01 -3.11988562e-01 -6.37018919e-01
7.31757209e-02 -3.27765197e-01 3.40906203e-01 1.43382585e+00
4.28140938e-01 1.09038815e-01 2.32515167e-02 6.11504316e-01
1.13691300e-01 6.86010063e-01 -3.44137490e-01 -4.94062863e-02
5.52904665e-01 1.05759346e+00 -8.70552957e-01 -7.48885348e-02
-8.66472602e-01 6.38378024e-01 7.17510223e-01 -5.58784455e-02
-4.65195656e-01 -7.93949127e-01 2.49207795e-01 8.62308070e-02
-5.38026094e-02 -3.08666438e-01 -6.51723027e-01 -1.24924397e+00
1.65839270e-01 -9.27679777e-01 4.05097008e-01 -1.96219847e-01
-1.45577836e+00 2.49781743e-01 3.41843837e-03 -1.20626891e+00
-2.28540525e-02 -1.08903062e+00 -6.17313623e-01 1.01777327e+00
-1.35090458e+00 -8.24103355e-01 -5.74151039e-01 2.87302434e-01
5.32639742e-01 -5.53285182e-01 7.02426910e-01 1.14890866e-01
-4.51765001e-01 1.07599938e+00 3.83000731e-01 1.37506634e-01
6.69904292e-01 -1.70632601e+00 2.51431406e-01 6.83263123e-01
2.72485048e-01 6.75012231e-01 4.03265297e-01 -3.88173074e-01
-9.57125902e-01 -6.31119072e-01 1.46878731e+00 -8.93909097e-01
1.02820444e+00 -2.07850158e-01 -8.53847206e-01 2.87998617e-01
4.12239730e-01 -9.86136138e-01 1.10390794e+00 4.72082794e-01
-3.40736896e-01 2.90522635e-01 -8.44050467e-01 7.00742602e-01
1.04076779e+00 -4.89060998e-01 -1.08331299e+00 2.71057129e-01
3.50348026e-01 -2.22759560e-01 -1.73406422e+00 -1.09581620e-01
5.29266417e-01 -1.15405011e+00 7.10644782e-01 -3.10099781e-01
8.68589759e-01 2.23114416e-01 -7.86335468e-02 -1.40509558e+00
-4.67661977e-01 -3.26187789e-01 2.65970439e-01 1.42299485e+00
4.90486741e-01 -5.92674136e-01 5.59466422e-01 5.44457495e-01
-1.08588785e-01 -7.78993666e-01 -9.00200963e-01 -9.53401685e-01
1.08085620e+00 -1.26060799e-01 5.95513105e-01 8.94964755e-01
5.57517707e-01 4.52944875e-01 5.51500559e-01 -3.68975639e-01
2.17600659e-01 6.41857181e-03 5.76590002e-01 -1.71128380e+00
-9.11458433e-02 -1.14661348e+00 -7.58782029e-01 -7.53328979e-01
3.92051041e-01 -1.29316831e+00 -3.82858425e-01 -1.23442101e+00
2.86891669e-01 -1.18326768e-01 1.78955883e-01 8.06860775e-02
-4.61157858e-01 1.08258411e-01 3.95821303e-01 -1.81591213e-01
-5.40934026e-01 1.13205083e-01 1.34999549e+00 9.93559510e-02
-3.06549579e-01 9.90863051e-03 -7.03268170e-01 6.15326822e-01
9.39363062e-01 3.62529531e-02 -4.15129185e-01 -3.41362804e-01
5.35247505e-01 -3.24955612e-01 -6.24039993e-02 -1.32015026e+00
3.89155969e-02 -5.22623248e-02 5.04024386e-01 -3.85766216e-02
-1.00050770e-01 -5.11888385e-01 -5.19658685e-01 3.40363652e-01
-4.81734276e-01 2.73754537e-01 1.52515039e-01 -1.18014283e-01
-6.72698170e-02 -6.02853775e-01 8.22483957e-01 -3.18950228e-02
-3.97851437e-01 -2.72612333e-01 -3.90261441e-01 8.48817766e-01
6.59849167e-01 -5.20682395e-01 -4.17412102e-01 -6.47778809e-02
-7.25811422e-01 1.02948539e-01 6.92819297e-01 6.40417337e-01
-9.00852829e-02 -1.07894373e+00 -1.01078725e+00 2.88814068e-01
3.26554418e-01 -5.64838529e-01 -7.08102435e-02 5.33110976e-01
-6.22031450e-01 7.83119202e-01 -3.62963498e-01 -3.23392719e-01
-1.37809503e+00 2.17637360e-01 1.33529365e-01 -5.48061788e-01
-3.14800799e-01 5.81153750e-01 -2.75944442e-01 -6.42514586e-01
9.12214443e-02 -6.32085741e-01 -6.53671205e-01 4.37802255e-01
3.61287862e-01 1.07388949e+00 4.26995784e-01 -8.35556924e-01
9.94398147e-02 1.13516796e+00 8.87015387e-02 -1.08994350e-01
1.07837844e+00 -2.73604184e-01 3.75275835e-02 7.65680254e-01
1.10160840e+00 6.54780924e-01 -4.97335643e-01 -4.96199317e-02
4.67504650e-01 -2.76672572e-01 -2.81045765e-01 -1.06503558e+00
-5.94295800e-01 8.21637213e-01 2.96192676e-01 3.98062438e-01
9.68864560e-01 -7.56827518e-02 2.44819194e-01 4.10145760e-01
2.35026568e-01 -1.46466577e+00 3.14352824e-03 8.63515973e-01
7.60233939e-01 -1.10967982e+00 1.04747146e-01 -4.74094540e-01
-3.55718076e-01 1.39454913e+00 5.39840102e-01 -7.97046572e-02
5.78633547e-01 -1.52956501e-01 -4.30539018e-03 -6.26718923e-02
-3.46928149e-01 -3.14025283e-01 8.37568998e-01 5.28776050e-01
9.93822753e-01 1.74694866e-01 -1.10065830e+00 8.77503335e-01
-1.40080142e+00 -4.01010722e-01 9.12778258e-01 4.89481717e-01
-5.06944120e-01 -1.22754490e+00 -2.94667363e-01 7.73112237e-01
-5.66261172e-01 -2.38098249e-01 -5.65470755e-01 1.31131363e+00
-1.13289636e-02 1.11635387e+00 1.68358266e-01 -2.02685803e-01
5.40694177e-01 2.27310091e-01 1.00932014e+00 -7.67889261e-01
-1.16268754e+00 -5.21423399e-01 7.43718371e-02 -1.08206064e-01
-8.49589780e-02 -1.03813183e+00 -9.24332857e-01 -5.70382297e-01
-4.05026466e-01 1.62571877e-01 2.91933954e-01 9.49330688e-01
-4.20413375e-01 6.26708448e-01 4.31467980e-01 -2.63033628e-01
-8.54390025e-01 -1.13803685e+00 -6.77824080e-01 6.53874815e-01
1.82466328e-01 -4.64235306e-01 -5.54393589e-01 -1.73205703e-01] | [11.006858825683594, 10.121245384216309] |
42a817d4-1d97-448a-8731-ea64cdc56138 | formality-style-transfer-for-noisy-user | null | null | https://aclanthology.org/D19-5502 | https://aclanthology.org/D19-5502.pdf | Formality Style Transfer for Noisy, User-generated Conversations: Extracting Labeled, Parallel Data from Unlabeled Corpora | Typical datasets used for style transfer in NLP contain aligned pairs of two opposite extremes of a style. As each existing dataset is sourced from a specific domain and context, most use cases will have a sizable mismatch from the vocabulary and sentence structures of any dataset available. This reduces the performance of the style transfer, and is particularly significant for noisy, user-generated text. To solve this problem, we show a technique to derive a dataset of aligned pairs (style-agnostic vs stylistic sentences) from an unlabeled corpus by using an auxiliary dataset, allowing for in-domain training. We test the technique with the Yahoo Formality Dataset and 6 novel datasets we produced, which consist of scripts from 5 popular TV-shows (Friends, Futurama, Seinfeld, Southpark, Stargate SG-1) and the Slate Star Codex online forum. We gather 1080 human evaluations, which show that our method produces a sizable change in formality while maintaining fluency and context; and that it considerably outperforms OpenNMT{'}s Seq2Seq model directly trained on the Yahoo Formality Dataset. Additionally, we publish the full pipeline code and our novel datasets. | ['Alan W. black', 'Isak Czeresnia Etinger'] | 2019-11-01 | null | null | null | ws-2019-11 | ['formality-style-transfer'] | ['natural-language-processing'] | [ 3.14245045e-01 1.27638131e-01 1.26466781e-01 -8.50598872e-01
-9.58326399e-01 -1.18529499e+00 5.52255690e-01 -3.86093736e-01
-5.49228728e-01 1.14759946e+00 5.65899432e-01 -1.49641484e-01
2.49336988e-01 -4.39950436e-01 -6.87903225e-01 -2.37814099e-01
5.15150905e-01 9.04906988e-01 -4.61827256e-02 -6.27973258e-01
2.31801897e-01 4.50808741e-02 -9.07865763e-01 6.00964367e-01
9.56837177e-01 5.67335010e-01 3.95226061e-01 8.55230033e-01
-3.76589596e-01 6.96780860e-01 -1.07727325e+00 -6.57027245e-01
1.76131159e-01 -8.01380634e-01 -1.17272878e+00 -6.63080812e-02
9.01097119e-01 -1.36302516e-01 -1.02388158e-01 8.22000504e-01
7.34277606e-01 2.66067777e-02 6.10474288e-01 -8.58102977e-01
-9.28280890e-01 1.00386226e+00 -3.56809273e-02 3.62083524e-01
4.58486289e-01 3.48333180e-01 1.03355241e+00 -5.26880443e-01
1.21581495e+00 1.40355134e+00 4.68231380e-01 1.00559664e+00
-1.48153281e+00 -5.85124969e-01 -2.97464877e-02 -1.89206854e-01
-9.75847125e-01 -6.25037611e-01 7.11247742e-01 -2.77410150e-01
9.32785034e-01 2.05202505e-01 3.65438938e-01 1.90420341e+00
-1.15824595e-01 9.10519063e-01 1.52764630e+00 -4.27150667e-01
8.69502202e-02 3.05752963e-01 -6.28368109e-02 2.45109931e-01
-3.24470937e-01 -5.39604127e-02 -8.42810571e-01 1.36685267e-01
5.44568062e-01 -7.29174018e-01 -3.76862377e-01 1.17001899e-01
-1.41416442e+00 6.39649510e-01 -2.80814692e-02 4.48099762e-01
2.88950890e-01 -2.16294229e-02 7.80044854e-01 9.05882895e-01
5.38280606e-01 9.32456017e-01 -8.88648510e-01 -6.87931716e-01
-8.92525911e-01 5.21070600e-01 1.08605933e+00 1.46184039e+00
5.27636349e-01 1.64262224e-02 -4.25172478e-01 1.05816746e+00
-1.83618903e-01 7.12364495e-01 7.55583107e-01 -9.36865807e-01
8.31258059e-01 1.46435499e-01 -1.03597781e-02 -5.95089257e-01
-2.80355334e-01 -1.92131311e-01 -5.47829866e-01 -3.13180357e-01
7.51380742e-01 -4.35230434e-01 -6.99239969e-01 1.98663855e+00
1.89680792e-02 -5.22008955e-01 1.91925675e-01 5.10797083e-01
1.12225473e+00 6.44808531e-01 -4.54935655e-02 1.21827470e-02
1.11693513e+00 -1.00098205e+00 -8.92837346e-01 -4.66643363e-01
8.66583347e-01 -1.08220220e+00 1.90896845e+00 5.10327578e-01
-1.13362908e+00 -6.91529214e-01 -8.43093574e-01 -4.12283868e-01
-3.47508132e-01 3.50762494e-02 2.79903054e-01 6.48162246e-01
-1.05366540e+00 8.13454866e-01 -1.96909532e-01 -6.92172945e-01
2.23920211e-01 -9.02198721e-03 -3.09093058e-01 4.47897352e-02
-1.22938836e+00 1.07426083e+00 6.06033504e-01 -3.14806521e-01
-6.17386043e-01 -9.72682953e-01 -8.67627859e-01 -2.64370918e-01
2.44025275e-01 -6.04906976e-01 1.59536755e+00 -1.52166688e+00
-1.88140666e+00 1.21054411e+00 8.06275234e-02 -2.80869871e-01
7.95906782e-01 -3.51255327e-01 -6.23044193e-01 -1.21502690e-01
1.83496654e-01 9.72835243e-01 7.26432979e-01 -9.28444266e-01
-4.41562384e-01 -1.00177899e-01 8.24141353e-02 2.78002053e-01
-2.71222442e-01 1.86252117e-01 -3.82032543e-02 -1.05345893e+00
-5.57087600e-01 -9.71419036e-01 1.83732882e-01 -3.31456840e-01
-4.97712076e-01 -1.76729500e-01 6.56100035e-01 -9.96217430e-01
1.17849243e+00 -2.20052338e+00 3.20817232e-01 -2.12923735e-01
-1.58438653e-01 2.24064291e-01 -4.19713318e-01 4.75375116e-01
1.08185612e-01 2.78768748e-01 -4.26237166e-01 -2.44950563e-01
8.62960052e-03 3.76199484e-01 -2.89755613e-01 -2.43590046e-02
3.31040710e-01 1.02481663e+00 -1.13584554e+00 -5.27010262e-01
-1.94409519e-01 -2.95310561e-02 -6.79883599e-01 3.64731669e-01
-5.49240053e-01 7.78685689e-01 -5.62198162e-02 2.81349242e-01
4.39262062e-01 4.32672836e-02 1.47562459e-01 7.66086485e-03
3.99967469e-03 7.30015337e-01 -8.46084416e-01 2.43376946e+00
-8.51578891e-01 8.09225023e-01 -2.13752210e-01 -2.86788940e-01
1.07656395e+00 3.48876953e-01 -1.63814813e-01 -7.35676944e-01
1.25559464e-01 2.79355854e-01 7.14438874e-03 -5.68003058e-01
7.82928526e-01 -3.32582921e-01 -5.43371797e-01 4.90012348e-01
4.62242037e-01 -8.12981904e-01 5.10561407e-01 2.11269543e-01
1.22255981e+00 3.36168587e-01 6.20833226e-02 -5.84607780e-01
2.86440939e-01 2.00798154e-01 5.29696047e-01 6.54954851e-01
-3.43067378e-01 9.31547880e-01 5.88238895e-01 -2.80144572e-01
-1.34629679e+00 -1.19918358e+00 2.44345870e-02 1.19817388e+00
-3.86207491e-01 -4.87144947e-01 -1.05990124e+00 -1.01333117e+00
-2.91151017e-01 1.22142458e+00 -5.98408580e-01 -1.21033583e-02
-7.70175934e-01 -1.24280594e-01 9.29999888e-01 2.86880791e-01
5.62360644e-01 -1.42017329e+00 -1.20273419e-01 1.74547225e-01
-4.88007784e-01 -1.10497057e+00 -9.12162066e-01 5.78105077e-02
-7.49464273e-01 -5.64380884e-01 -6.49476230e-01 -7.65068352e-01
2.53304303e-01 -3.06381226e-01 1.80999935e+00 -3.44150841e-01
2.14870766e-01 1.15969971e-01 -5.06169558e-01 -4.96664554e-01
-9.84266341e-01 5.19003391e-01 -1.63201913e-01 -3.99436355e-01
4.04460430e-01 -5.67949891e-01 -7.52559751e-02 3.02457418e-02
-9.92033422e-01 2.59588391e-01 4.15349245e-01 1.01534474e+00
1.55964345e-01 -4.86930192e-01 8.28738213e-01 -1.42506802e+00
7.21952081e-01 -3.08560610e-01 -3.16083640e-01 1.47150874e-01
-1.81487352e-01 2.09675863e-01 9.25821185e-01 -3.07927817e-01
-1.39055347e+00 5.84667251e-02 -2.38706097e-01 1.39202201e-03
-4.92990524e-01 1.60539091e-01 -5.51973820e-01 4.79548305e-01
1.00913274e+00 8.28855038e-02 -1.48142681e-01 -7.59488285e-01
7.33746886e-01 9.59153175e-01 7.49147892e-01 -9.13975179e-01
8.04789424e-01 6.12224266e-02 -5.81291318e-01 -8.80069196e-01
-1.04248416e+00 -1.91512909e-02 -8.56924117e-01 5.63613847e-02
7.41414964e-01 -6.90777361e-01 -1.05121411e-01 3.43057990e-01
-1.58336794e+00 -6.72035575e-01 -5.94761789e-01 6.46566823e-02
-7.73128629e-01 1.78601637e-01 -6.54007554e-01 -8.16706121e-02
-4.58043247e-01 -9.73418295e-01 9.10961747e-01 1.57160595e-01
-9.70906436e-01 -1.17610586e+00 4.14297044e-01 3.83617222e-01
3.94152224e-01 1.50329337e-01 1.16432559e+00 -8.38566601e-01
5.44042587e-02 3.64861608e-01 -2.75653624e-03 8.01335931e-01
2.32325181e-01 -2.85989642e-02 -1.01060462e+00 -1.81662947e-01
-1.39372855e-01 -7.24979103e-01 3.45460862e-01 -1.31373510e-01
9.80069458e-01 -3.90053391e-01 3.03558826e-01 5.87694228e-01
1.18225193e+00 2.15013465e-03 5.37160814e-01 3.07807386e-01
7.11391270e-01 7.14025199e-01 4.40421790e-01 1.86081603e-01
2.60728985e-01 5.46391726e-01 -2.83787251e-01 1.89731552e-04
-4.44518775e-01 -5.21372855e-01 7.46963084e-01 1.20058715e+00
4.22308564e-01 -3.61827344e-01 -7.48998165e-01 7.67170668e-01
-1.36981547e+00 -8.71958852e-01 -1.12697616e-01 1.96854496e+00
1.60534596e+00 2.56298214e-01 2.30814666e-01 -1.60097480e-01
5.69826484e-01 1.22851327e-01 -3.54554534e-01 -1.00973463e+00
-4.91629034e-01 4.76607084e-01 2.48309597e-01 4.55516011e-01
-8.21728528e-01 1.32394135e+00 6.66794586e+00 9.71968651e-01
-1.02960110e+00 1.53836489e-01 5.91363549e-01 -4.29919839e-01
-7.24419892e-01 -1.10468931e-01 -7.61061966e-01 7.91520119e-01
1.17295206e+00 -2.15262994e-01 6.62202895e-01 6.88486636e-01
2.88857460e-01 5.70121966e-02 -1.55313694e+00 8.45928311e-01
1.46813139e-01 -1.12043846e+00 -2.41788514e-02 -3.65229815e-01
9.12037075e-01 2.15430245e-01 -3.44701037e-02 5.29130340e-01
5.14034092e-01 -1.05720484e+00 1.00357306e+00 2.62497902e-01
1.11345696e+00 -7.39483416e-01 4.06767130e-01 2.93112487e-01
-2.68672079e-01 3.92102659e-01 -1.57136381e-01 -1.29922226e-01
2.25329518e-01 6.19531512e-01 -8.97484362e-01 3.04051906e-01
6.83062434e-01 6.20721042e-01 -8.51247132e-01 3.88772279e-01
-5.75694084e-01 8.11146021e-01 -2.03112647e-01 -1.71466798e-01
1.29712388e-01 -9.18718278e-02 6.92433059e-01 1.50339854e+00
-9.06052161e-03 -2.08783805e-01 -9.21367779e-02 1.03867114e+00
-4.15826529e-01 1.90644234e-01 -8.47776532e-01 -3.12306017e-01
4.83653694e-01 1.28893173e+00 -4.94359225e-01 -5.48090458e-01
-5.18215060e-01 1.39714432e+00 5.50849676e-01 2.14995369e-01
-5.51037252e-01 -5.92764080e-01 6.04448617e-01 -7.85400569e-02
2.01279968e-01 -2.23131388e-01 -6.15417480e-01 -1.25395179e+00
1.38691172e-01 -1.49313259e+00 2.90668845e-01 -8.11503232e-01
-1.63870025e+00 8.11899245e-01 -8.93548802e-02 -1.07589614e+00
-3.28886807e-01 -5.47447324e-01 -4.35281813e-01 1.06462753e+00
-1.15876865e+00 -1.12814033e+00 -1.14555322e-01 3.35944504e-01
9.90840971e-01 -2.48320445e-01 7.98485696e-01 2.96217740e-01
-3.86916965e-01 7.79988706e-01 1.88779727e-01 2.82848716e-01
1.36292684e+00 -1.65552020e+00 9.01654124e-01 6.70258164e-01
2.05812708e-01 4.03321505e-01 9.52269137e-01 -7.28723645e-01
-9.77084279e-01 -1.09813082e+00 1.31339359e+00 -9.02870834e-01
7.72893667e-01 -7.21656024e-01 -7.78534949e-01 8.74880672e-01
7.27950215e-01 -4.04466212e-01 6.93640649e-01 1.85240626e-01
-3.82089168e-01 1.11031123e-01 -1.26998031e+00 8.89352441e-01
1.37524998e+00 -5.75624287e-01 -1.00825489e+00 4.45165217e-01
1.03015745e+00 -6.39638960e-01 -8.69852841e-01 7.30946511e-02
3.08318228e-01 -8.42660129e-01 3.12076807e-01 -9.21121120e-01
9.65253711e-01 -9.01796371e-02 -8.60408098e-02 -1.92507136e+00
-1.83926105e-01 -8.66605222e-01 3.62164855e-01 1.92963767e+00
7.67796755e-01 -2.18154103e-01 5.21642923e-01 4.21311319e-01
-2.67938554e-01 -2.33125329e-01 -6.93420649e-01 -1.02205610e+00
6.37575269e-01 -3.14335346e-01 5.73277235e-01 1.04627883e+00
-5.39207160e-02 1.03356445e+00 -2.16088027e-01 -5.47158182e-01
9.85946357e-02 -1.18750989e-01 8.91604424e-01 -7.76045620e-01
-6.49236083e-01 -4.08585459e-01 3.62212397e-02 -1.01523721e+00
4.26670045e-01 -1.23277938e+00 9.10283998e-02 -1.12443650e+00
-8.41095597e-02 -4.77069736e-01 3.25385958e-01 3.52819115e-01
-1.03730522e-01 2.26501748e-01 3.19437981e-01 -1.65172100e-01
-4.72448111e-01 5.51106691e-01 1.48499274e+00 -1.69389680e-01
-2.95433879e-01 -3.01445544e-01 -7.46774852e-01 5.67181170e-01
8.73948216e-01 -4.56632584e-01 -4.76224959e-01 -9.34903860e-01
4.60428715e-01 -2.30806440e-01 -7.33563304e-02 -8.36740077e-01
-3.66791904e-01 -1.40151352e-01 3.30697507e-01 -2.82764614e-01
-7.23968223e-02 -4.56592113e-01 -2.03676298e-01 2.75737289e-02
-7.99125731e-01 1.28260404e-01 3.65514189e-01 1.02352850e-01
-5.39433956e-02 -4.34988022e-01 8.26398969e-01 -3.79210621e-01
-6.17966890e-01 -1.15293421e-01 -4.06730503e-01 9.28415716e-01
5.35850942e-01 5.91604151e-02 -3.98827016e-01 -3.87019217e-01
-4.61622268e-01 7.53659904e-02 6.67145669e-01 6.98344409e-01
1.14576947e-02 -1.43204558e+00 -9.94652092e-01 9.80069190e-02
3.12042862e-01 8.19760337e-02 9.24146250e-02 4.16272908e-01
-5.45535207e-01 1.01681679e-01 -5.54964066e-01 -3.69893909e-01
-8.40076923e-01 4.16143060e-01 1.13708287e-01 -2.21327633e-01
-4.59758550e-01 9.24870074e-01 1.50086880e-01 -9.99728978e-01
-1.57807276e-01 -4.61525738e-01 6.03210218e-02 -4.43590526e-03
4.55775887e-01 3.59503925e-01 2.48222098e-01 -4.80513871e-01
1.52039491e-02 8.39008540e-02 -2.37842664e-01 -6.14793420e-01
1.11271119e+00 -2.70802081e-01 -5.65366969e-02 8.40828300e-01
1.24687004e+00 2.41445780e-01 -1.19495356e+00 -3.42612237e-01
2.90697157e-01 -3.22981417e-01 -4.11145985e-01 -1.19979393e+00
-6.51178598e-01 7.60603786e-01 2.27060288e-01 3.23789939e-02
9.61640656e-01 6.65318668e-02 1.02503598e+00 3.83594334e-01
1.92220852e-01 -1.53878951e+00 4.85614389e-02 1.09217978e+00
1.18812752e+00 -1.09924412e+00 -5.04341006e-01 -2.15711728e-01
-8.70474815e-01 1.05362594e+00 6.88704610e-01 -5.02378829e-02
2.50201106e-01 2.42372006e-01 4.45127040e-01 1.47823349e-01
-7.46395707e-01 8.28997120e-02 7.96316713e-02 8.57369781e-01
9.05843735e-01 1.23087503e-01 -4.83681202e-01 7.91647136e-01
-1.05737662e+00 -1.62284568e-01 8.02341461e-01 9.09484267e-01
-5.75493202e-02 -1.40629125e+00 -2.33277366e-01 4.60689753e-01
-4.58305687e-01 -4.25633401e-01 -8.60478759e-01 7.05238461e-01
-1.33231450e-02 7.87245929e-01 1.23154491e-01 -1.70736417e-01
5.50377131e-01 2.93613285e-01 7.47969627e-01 -9.31788087e-01
-8.80459845e-01 -2.09383026e-01 6.35460317e-01 -3.84355009e-01
-1.94990244e-02 -7.70593941e-01 -9.52022135e-01 -4.15071279e-01
1.80679321e-01 -2.80956663e-02 5.03803313e-01 1.18081701e+00
1.97599262e-01 4.86527652e-01 4.84466434e-01 -6.29986584e-01
-5.04293501e-01 -1.38208342e+00 -5.86770117e-01 7.77500331e-01
2.82385163e-02 -2.27471232e-01 -1.83098584e-01 4.01846200e-01] | [11.528836250305176, 9.634202003479004] |
730d9369-b1f1-454e-89d0-3b2c68cbdab3 | unsupervised-learning-of-probabilistic | 1903.03545 | null | https://arxiv.org/abs/1903.03545v2 | https://arxiv.org/pdf/1903.03545v2.pdf | Unsupervised Learning of Probabilistic Diffeomorphic Registration for Images and Surfaces | Classical deformable registration techniques achieve impressive results and offer a rigorous theoretical treatment, but are computationally intensive since they solve an optimization problem for each image pair. Recently, learning-based methods have facilitated fast registration by learning spatial deformation functions. However, these approaches use restricted deformation models, require supervised labels, or do not guarantee a diffeomorphic (topology-preserving) registration. Furthermore, learning-based registration tools have not been derived from a probabilistic framework that can offer uncertainty estimates. In this paper, we build a connection between classical and learning-based methods. We present a probabilistic generative model and derive an unsupervised learning-based inference algorithm that uses insights from classical registration methods and makes use of recent developments in convolutional neural networks (CNNs). We demonstrate our method on a 3D brain registration task for both images and anatomical surfaces, and provide extensive empirical analyses. Our principled approach results in state of the art accuracy and very fast runtimes, while providing diffeomorphic guarantees. Our implementation is available at http://voxelmorph.csail.mit.edu. | ['Mert R. Sabuncu', 'Guha Balakrishnan', 'Adrian V. Dalca', 'John Guttag'] | 2019-03-08 | null | null | null | null | ['constrained-diffeomorphic-image-registration', 'deformable-medical-image-registration', 'diffeomorphic-medical-image-registration'] | ['computer-vision', 'medical', 'medical'] | [ 1.24554597e-01 1.39409110e-01 -1.08662926e-01 -6.42399549e-01
-1.15433872e+00 -3.64007950e-01 6.21895671e-01 1.82645738e-01
-5.12156904e-01 5.86986363e-01 1.19033143e-01 -4.65784513e-04
-4.32510227e-01 -7.04204381e-01 -6.05457664e-01 -7.29713976e-01
-4.35073435e-01 7.61507392e-01 2.98557371e-01 3.09966914e-02
1.19718812e-01 6.17306471e-01 -8.76397491e-01 -2.09500298e-01
8.33445549e-01 7.34834552e-01 -1.52943367e-02 3.04223776e-01
1.71255991e-01 1.40218243e-01 9.60520878e-02 -2.30143666e-01
2.97238082e-01 -4.20198321e-01 -1.13602269e+00 -2.71595597e-01
6.46197677e-01 -2.61489749e-01 -4.13660377e-01 1.04126227e+00
6.19652212e-01 2.38433406e-01 8.05426598e-01 -1.05207348e+00
-7.94801950e-01 5.25426030e-01 -3.28937203e-01 3.15452516e-01
6.06358722e-02 -2.78895289e-01 6.72552288e-01 -8.52041841e-01
7.22246349e-01 7.82718837e-01 1.09923983e+00 7.02864110e-01
-1.53634143e+00 -4.00529385e-01 -3.80109638e-01 -1.05223507e-01
-1.35677183e+00 -5.54716051e-01 8.19978178e-01 -6.40580833e-01
7.98982799e-01 2.06589475e-01 5.90688527e-01 6.76137567e-01
5.65126836e-01 3.56171191e-01 1.46173561e+00 -2.04448804e-01
1.52438596e-01 -4.18954760e-01 1.64265241e-02 8.32423568e-01
1.21412762e-01 2.42105871e-01 -3.17682892e-01 -3.89157325e-01
1.14792681e+00 -8.58499669e-03 -3.54714572e-01 -7.51519620e-01
-1.28067112e+00 8.72284055e-01 6.47800207e-01 5.42010188e-01
-2.52039343e-01 3.66045356e-01 1.03833221e-01 3.51565704e-02
8.56345832e-01 8.09089839e-02 -1.78228259e-01 1.45050272e-01
-1.27857554e+00 2.45664239e-01 7.60786414e-01 6.42276049e-01
7.02875614e-01 -1.89405426e-01 1.97234526e-01 6.41650259e-01
6.14918470e-01 4.07680273e-01 5.01261532e-01 -1.15997791e+00
-1.39681771e-01 5.28893396e-02 -2.50543207e-01 -1.06967902e+00
-7.67328084e-01 -1.52019382e-01 -9.78924274e-01 5.03007948e-01
5.44988632e-01 1.88906118e-01 -6.79467320e-01 1.81956649e+00
3.64519060e-01 3.66143316e-01 -4.36348140e-01 8.50304723e-01
5.28479874e-01 -1.06928982e-01 -4.15410697e-02 -2.74792701e-01
1.12335742e+00 -4.69921231e-01 -6.38768375e-01 2.52776563e-01
7.20810235e-01 -7.54441679e-01 6.39994681e-01 8.12770575e-02
-1.40107334e+00 -1.31923720e-01 -7.73913980e-01 -1.33129552e-01
-6.96216151e-02 -3.34344059e-01 6.72387004e-01 8.16264808e-01
-1.54068351e+00 1.20496070e+00 -1.53977764e+00 -2.46805727e-01
8.22661340e-01 6.02188408e-01 -5.68601310e-01 8.61288458e-02
-7.94215560e-01 1.14474499e+00 -4.69923578e-02 1.57959908e-01
-7.87238538e-01 -9.10251677e-01 -9.66890395e-01 -4.85104382e-01
-1.54881254e-01 -8.49234700e-01 1.01229596e+00 -6.14349604e-01
-1.43162882e+00 1.22172534e+00 -6.01460114e-02 -4.86117929e-01
6.43006206e-01 -7.54477270e-03 1.36818334e-01 2.59947032e-01
5.01396554e-03 7.32201457e-01 6.07232690e-01 -1.02880514e+00
2.80451566e-01 -5.27455151e-01 -2.61001468e-01 7.33157527e-03
-5.07548973e-02 2.57652521e-01 1.06242962e-01 -7.04421878e-01
5.36367416e-01 -1.05509400e+00 -3.40622455e-01 4.88437712e-01
-1.90784618e-01 -5.16060106e-02 3.54539156e-01 -8.34416747e-01
4.43713844e-01 -1.82793581e+00 1.83396488e-01 2.67113000e-01
5.28281391e-01 -9.47828442e-02 2.28269413e-01 -2.17477679e-02
-1.73781410e-01 1.17305763e-01 -9.42356825e-01 -6.33058369e-01
1.02878340e-01 2.16224954e-01 1.59463231e-02 1.23908782e+00
4.80069816e-02 1.01533568e+00 -9.12331402e-01 -6.72063410e-01
2.57279396e-01 9.70451117e-01 -6.13066494e-01 9.93651897e-03
2.08870888e-01 8.44269097e-01 -3.32619101e-01 3.61912578e-01
8.36940348e-01 -8.56876671e-02 1.90763429e-01 -3.40367526e-01
-8.41218978e-02 2.87444383e-01 -8.71625721e-01 2.28779602e+00
-3.74464571e-01 4.36700553e-01 1.58723503e-01 -1.22398067e+00
7.76523769e-01 3.41450810e-01 1.02833390e+00 -1.13755509e-01
3.09762776e-01 4.92112517e-01 -4.41896692e-02 -1.72903731e-01
5.68352640e-02 -5.29875994e-01 2.58414716e-01 7.13921845e-01
2.86150634e-01 -4.92411345e-01 -2.62680948e-01 -2.94116661e-02
1.19094706e+00 6.32469058e-01 1.67014480e-01 -8.34754288e-01
2.33398810e-01 -1.38341501e-01 5.07689774e-01 4.37936187e-01
-3.89430672e-01 9.04693186e-01 4.62535396e-02 -5.68283141e-01
-1.07603693e+00 -1.39720213e+00 -7.92039156e-01 4.42193538e-01
5.96963509e-04 -9.03633311e-02 -1.03386569e+00 -5.33839941e-01
-1.81579739e-01 9.58589613e-02 -6.96594536e-01 -5.44477813e-02
-8.37060571e-01 -1.03012669e+00 6.54080153e-01 4.99381155e-01
2.41169199e-01 -8.58421206e-01 -4.21810418e-01 2.29151219e-01
-4.18519452e-02 -9.24295843e-01 -6.12020314e-01 -1.53464243e-01
-1.42512834e+00 -1.17019069e+00 -8.06740999e-01 -8.71969700e-01
9.28943276e-01 -1.91287592e-01 1.03262925e+00 1.93599477e-01
-5.74707866e-01 4.97527927e-01 1.94198340e-02 7.68128857e-02
-3.73138100e-01 4.58261184e-02 3.65414441e-01 -2.80041873e-01
1.31548285e-01 -1.15391564e+00 -6.00412965e-01 4.08663124e-01
-7.82756627e-01 -2.56922990e-01 2.23131165e-01 8.51134121e-01
8.92676234e-01 -3.02775681e-01 3.89786482e-01 -7.60267377e-01
4.04660106e-01 -2.22240269e-01 -5.80636621e-01 1.49838272e-02
-8.11473846e-01 2.98694015e-01 1.23734824e-01 -2.46112019e-01
-8.46438587e-01 2.25263581e-01 -3.79739285e-01 -4.25083399e-01
-3.12996507e-01 3.70851040e-01 1.55255690e-01 -6.94952726e-01
7.80408442e-01 2.41343513e-01 4.12430555e-01 -5.87828875e-01
3.64140183e-01 1.91657826e-01 6.26895487e-01 -9.86775875e-01
9.42447245e-01 9.40038264e-01 1.93993539e-01 -5.89217722e-01
-4.76057291e-01 -1.29759967e-01 -1.20452702e+00 -1.48101926e-01
9.97417808e-01 -4.87671316e-01 -5.02960205e-01 4.73369181e-01
-1.06683183e+00 -6.03717566e-01 -4.10803586e-01 8.18404317e-01
-1.02461433e+00 5.36582291e-01 -7.45378077e-01 -4.21619564e-01
-4.08936232e-01 -1.21813893e+00 9.95507658e-01 -2.58998983e-02
-2.65955359e-01 -1.37266672e+00 4.36250478e-01 1.04603931e-01
7.51426816e-01 5.59437454e-01 6.20574236e-01 -6.14748836e-01
-5.27313113e-01 -5.97657226e-02 -1.01144664e-01 1.18978329e-01
3.05488765e-01 -3.16924274e-01 -8.97291005e-01 -3.48674655e-01
3.05120647e-01 -5.35003319e-02 6.82375908e-01 7.10658848e-01
1.18895268e+00 -1.13549948e-01 -3.27078819e-01 9.72979188e-01
1.37593961e+00 -3.92505735e-01 6.45851314e-01 4.91978675e-02
7.43551016e-01 6.75037444e-01 1.53307691e-02 -5.27482070e-02
3.73396337e-01 7.67326593e-01 2.78484315e-01 3.45839076e-02
-4.27296549e-01 2.39074677e-01 1.61683515e-01 1.11035740e+00
-6.09073162e-01 6.24230683e-01 -1.08073330e+00 5.35834670e-01
-1.74426210e+00 -8.66542637e-01 -3.25649321e-01 2.32344747e+00
1.26729536e+00 -1.11760296e-01 7.61929527e-02 -1.92613393e-01
6.12245023e-01 -1.20763645e-01 -2.39813328e-01 7.19049498e-02
9.93435532e-02 7.41085589e-01 5.40482998e-01 7.67973244e-01
-1.24907625e+00 6.75383270e-01 6.76476383e+00 5.16002297e-01
-8.24588776e-01 7.38843441e-01 4.54806596e-01 9.52107310e-02
-3.62037033e-01 2.74202190e-02 -3.46269906e-01 1.67288646e-01
8.77737880e-01 -2.08203375e-01 3.77797931e-01 5.42854548e-01
3.54992040e-02 1.32067174e-01 -1.26570690e+00 8.33649814e-01
1.00639485e-01 -1.48509979e+00 -4.76536989e-01 3.10759306e-01
7.57856369e-01 4.96793449e-01 -3.09876669e-02 -4.39049006e-01
3.49633634e-01 -1.18839645e+00 3.92334402e-01 8.63862932e-01
7.45315969e-01 -4.88989800e-01 5.46333313e-01 3.72916460e-02
-1.03409779e+00 7.54010499e-01 -4.31704670e-01 2.44373977e-01
2.93096781e-01 7.80097425e-01 -3.31110448e-01 6.73959553e-01
6.95351124e-01 8.02192211e-01 -1.69284508e-01 1.23194492e+00
-1.25358656e-01 4.42444831e-01 -5.53342223e-01 4.79305148e-01
-1.77375942e-01 -4.08561528e-01 6.58645272e-01 1.04566169e+00
2.47940168e-01 1.99689254e-01 1.90215811e-01 1.31006813e+00
-4.45647575e-02 9.96404216e-02 -5.76025665e-01 4.68805492e-01
1.25227839e-01 1.42504168e+00 -1.00423145e+00 1.69621453e-01
-1.64962590e-01 7.83717036e-01 2.96701223e-01 -9.25763175e-02
-6.14703357e-01 -6.01289161e-02 5.97585797e-01 2.86745936e-01
-2.40347400e-01 -7.43363917e-01 -3.10078800e-01 -1.17755532e+00
2.29909699e-02 -2.69644916e-01 1.00336559e-01 -4.50983793e-01
-1.51726580e+00 4.85886693e-01 2.14858830e-01 -9.53632295e-01
-1.80445582e-01 -5.88145673e-01 -5.58772445e-01 9.83957887e-01
-1.49228573e+00 -1.09446537e+00 -2.38196865e-01 5.51275074e-01
-1.10024167e-02 2.36549377e-01 1.10005808e+00 4.49953228e-01
-9.89195853e-02 4.40742105e-01 2.16136184e-02 3.80859882e-01
8.06180537e-01 -1.45314372e+00 2.09669277e-01 7.14375496e-01
1.83533385e-01 7.81364143e-01 4.25420314e-01 -6.37595713e-01
-1.45498204e+00 -9.32567537e-01 9.24797714e-01 -6.33125365e-01
8.71172369e-01 -2.17706442e-01 -1.07990873e+00 8.92120421e-01
6.73053637e-02 6.68213904e-01 7.97751129e-01 4.84034345e-02
-2.43380234e-01 2.12466925e-01 -1.50475109e+00 2.24486858e-01
1.25141096e+00 -5.96865356e-01 -7.81475663e-01 6.38863504e-01
3.68166417e-01 -7.27567434e-01 -1.47991431e+00 3.27514350e-01
5.51674068e-01 -7.68232048e-01 1.19011950e+00 -3.68282825e-01
1.45811185e-01 -1.19837470e-01 -4.11970615e-02 -1.22412729e+00
-1.88742906e-01 -7.32729614e-01 1.07547387e-01 9.29111898e-01
9.36163440e-02 -8.91069829e-01 5.45176327e-01 8.99173021e-01
-4.77127045e-01 -6.32615030e-01 -1.41049814e+00 -1.02350998e+00
7.16306329e-01 -4.80464399e-01 2.75558114e-01 1.22749782e+00
7.49396011e-02 -4.65024501e-01 -1.95441432e-02 1.42999142e-01
1.29055846e+00 -1.22081470e-02 1.64906129e-01 -1.37164843e+00
-7.92614892e-02 -5.68949819e-01 -7.01018393e-01 -5.12052894e-01
5.34207821e-01 -1.47717702e+00 2.86671013e-01 -1.37010849e+00
2.19697937e-01 -8.45499456e-01 -2.28626505e-01 7.71726012e-01
2.60930330e-01 7.06835389e-01 -3.30936462e-01 6.35982513e-01
-1.20130993e-01 3.95796210e-01 1.01210105e+00 3.27736326e-02
2.93955561e-02 -1.84983701e-01 -2.33391389e-01 9.08639967e-01
1.04424572e+00 -7.61398971e-01 -1.05014160e-01 -3.68876189e-01
1.08362518e-01 -1.96302250e-01 8.46877635e-01 -9.26543713e-01
3.16354513e-01 1.47373989e-01 2.61294782e-01 -4.63327132e-02
1.74593762e-01 -6.05635345e-01 2.13797659e-01 5.81702769e-01
-2.92160630e-01 -1.02837250e-01 1.64884161e-02 2.99774230e-01
1.18802100e-01 -1.40565306e-01 1.07747245e+00 7.38375122e-03
-9.67383608e-02 7.20796824e-01 -5.19411676e-02 2.06315264e-01
5.88059843e-01 1.24488413e-01 -1.23714348e-02 -2.44340058e-02
-1.15399718e+00 -2.77995914e-01 6.13374293e-01 -5.05836681e-02
6.48538113e-01 -1.59899163e+00 -9.57568944e-01 2.15097163e-02
-2.63834029e-01 8.28658864e-02 1.40538961e-01 1.56284750e+00
-6.70094192e-01 1.55404255e-01 -4.01596487e-01 -8.27406049e-01
-9.53754961e-01 2.24657491e-01 6.33046567e-01 -9.68453512e-02
-8.77151608e-01 7.09974349e-01 -4.74893935e-02 -7.69509554e-01
-3.01393479e-01 -2.59555161e-01 9.00766924e-02 -3.34971100e-01
3.64685863e-01 2.30889753e-01 2.40529537e-01 -9.16575730e-01
-6.37405753e-01 8.61184180e-01 2.51393288e-01 -2.79076129e-01
1.56827581e+00 4.25699428e-02 -5.45959771e-01 1.54022709e-01
1.30841458e+00 -1.83757037e-01 -1.17995107e+00 -5.00416875e-01
8.72111693e-03 -3.26469272e-01 3.30675900e-01 -3.75148326e-01
-1.21316576e+00 8.14142585e-01 7.68129826e-01 1.34616360e-01
8.37522805e-01 2.81442672e-01 8.14520478e-01 2.15885267e-02
6.53839409e-01 -6.72271967e-01 -4.48641598e-01 2.41169423e-01
7.66066015e-01 -1.26796424e+00 1.53715655e-01 -5.61619461e-01
1.31542444e-01 1.19496691e+00 6.53263703e-02 -7.09102988e-01
1.09561884e+00 5.98807096e-01 -9.21545699e-02 -4.07352746e-01
-4.21642736e-02 4.96503571e-03 6.38518691e-01 8.96492422e-01
6.66732848e-01 8.84994343e-02 -5.77740431e-01 3.56291384e-01
-2.27247953e-01 8.82060751e-02 2.45136350e-01 9.75448191e-01
-2.93625921e-01 -1.44789386e+00 -2.39086881e-01 3.57325137e-01
-5.53178608e-01 -2.50248343e-01 2.03819051e-02 6.60555840e-01
-1.49349466e-01 3.09164792e-01 1.69185445e-01 1.35233514e-02
-1.50877908e-01 1.98395073e-01 1.22106397e+00 -6.38750315e-01
-3.38871807e-01 1.55078143e-01 -2.87438542e-01 -8.06589544e-01
-9.67177570e-01 -1.13510454e+00 -1.56407893e+00 -3.10103804e-01
-2.62315333e-01 8.71357024e-02 8.23602557e-01 1.20844901e+00
2.77387142e-01 2.26851299e-01 4.31808263e-01 -1.29814565e+00
-5.25543749e-01 -7.77917206e-01 -5.52542627e-01 2.04118460e-01
2.48402301e-02 -7.77656674e-01 -2.53656596e-01 2.11138785e-01] | [13.968268394470215, -2.5367794036865234] |
95d04527-51a5-4dc3-9a2b-afd6aa7d320f | mimamo-net-integrating-micro-and-macro-motion | 1911.09784 | null | https://arxiv.org/abs/1911.09784v1 | https://arxiv.org/pdf/1911.09784v1.pdf | MIMAMO Net: Integrating Micro- and Macro-motion for Video Emotion Recognition | Spatial-temporal feature learning is of vital importance for video emotion recognition. Previous deep network structures often focused on macro-motion which extends over long time scales, e.g., on the order of seconds. We believe integrating structures capturing information about both micro- and macro-motion will benefit emotion prediction, because human perceive both micro- and macro-expressions. In this paper, we propose to combine micro- and macro-motion features to improve video emotion recognition with a two-stream recurrent network, named MIMAMO (Micro-Macro-Motion) Net. Specifically, smaller and shorter micro-motions are analyzed by a two-stream network, while larger and more sustained macro-motions can be well captured by a subsequent recurrent network. Assigning specific interpretations to the roles of different parts of the network enables us to make choice of parameters based on prior knowledge: choices that turn out to be optimal. One of the important innovations in our model is the use of interframe phase differences rather than optical flow as input to the temporal stream. Compared with the optical flow, phase differences require less computation and are more robust to illumination changes. Our proposed network achieves state of the art performance on two video emotion datasets, the OMG emotion dataset and the Aff-Wild dataset. The most significant gains are for arousal prediction, for which motion information is intuitively more informative. Source code is available at https://github.com/wtomin/MIMAMO-Net. | ['Zhaokang Chen', 'Didan Deng', 'Bertram Shi', 'Yuqian Zhou'] | 2019-11-21 | null | null | null | null | ['video-emotion-recognition'] | ['computer-vision'] | [-1.26111448e-01 -4.06318009e-01 -2.23381400e-01 -3.54777575e-01
-1.19773842e-01 -4.98818338e-01 4.22448963e-01 -1.93445422e-02
-3.76291394e-01 4.59598839e-01 3.83439451e-01 1.14404134e-01
8.48906264e-02 -5.00800550e-01 -3.93716067e-01 -7.34471142e-01
-3.18630219e-01 -3.54379714e-01 3.69913988e-02 -3.29714984e-01
1.69697940e-01 3.52537155e-01 -1.57289076e+00 4.18702185e-01
4.83904392e-01 1.42052615e+00 -1.22912172e-02 8.79689217e-01
-2.52215583e-02 9.93162036e-01 -3.29515636e-01 -7.63860345e-02
8.14337656e-02 -5.97791970e-01 -6.56400561e-01 -6.94746375e-02
2.45496288e-01 -2.53266394e-01 -4.61366028e-01 7.17145026e-01
5.44885576e-01 4.15549994e-01 4.14179891e-01 -1.34765875e+00
-2.41684571e-01 2.53621519e-01 -4.44538772e-01 5.81977189e-01
3.82943660e-01 2.90657222e-01 1.09439898e+00 -8.14702392e-01
6.72767460e-01 1.21812284e+00 5.95400929e-01 5.11987090e-01
-8.80060136e-01 -5.25054395e-01 4.40148443e-01 7.06490517e-01
-1.05387533e+00 -5.53444624e-01 1.07368815e+00 -3.67408901e-01
9.16001678e-01 1.89843148e-01 1.06416178e+00 1.33702230e+00
5.37594557e-01 9.57844257e-01 8.89021337e-01 -4.82009072e-03
1.14706688e-01 -2.47423783e-01 -2.63967440e-02 6.47217810e-01
-2.89078534e-01 1.05124608e-01 -9.18529987e-01 2.39695668e-01
7.28545725e-01 1.10212795e-01 -6.25677466e-01 -1.63014501e-01
-1.15350008e+00 7.61587858e-01 3.50373030e-01 3.72869253e-01
-4.95927900e-01 4.09064591e-01 6.07241631e-01 4.09450889e-01
5.78996062e-01 2.76773095e-01 -4.86705214e-01 -9.58160400e-01
-8.54790092e-01 -1.18772693e-01 5.36559224e-01 4.98383284e-01
7.43435442e-01 1.90439627e-01 1.05994204e-02 6.38596237e-01
2.16876507e-01 2.45354876e-01 7.00796425e-01 -1.35605240e+00
1.66860580e-01 2.24667639e-01 -7.30035501e-03 -1.35891855e+00
-5.28887689e-01 -2.33365551e-01 -9.58832741e-01 7.13139251e-02
4.62921351e-01 -3.99722844e-01 -5.43478191e-01 2.01041293e+00
1.92999169e-01 6.04670167e-01 -3.58979702e-02 1.24810135e+00
7.29941964e-01 7.17860460e-01 -1.77956343e-01 -5.05376279e-01
1.24414444e+00 -1.07330406e+00 -9.92707014e-01 -2.55870521e-02
6.70768559e-01 -6.55065596e-01 8.70892584e-01 3.80450666e-01
-9.67777550e-01 -6.35757506e-01 -8.41827571e-01 -2.01210324e-02
-3.62432241e-01 -1.73352122e-01 7.28299558e-01 2.56866753e-01
-1.05155087e+00 7.36331403e-01 -1.11092436e+00 -3.42069119e-01
2.65523195e-01 2.26739466e-01 -2.59561419e-01 3.36232096e-01
-1.28573596e+00 4.42910075e-01 -8.40818882e-02 4.61577743e-01
-4.91336703e-01 -5.22487283e-01 -9.73717690e-01 5.00229746e-02
2.11072773e-01 -7.30445862e-01 1.17904401e+00 -1.40105700e+00
-1.78082168e+00 3.98182601e-01 -7.39444375e-01 -3.51430565e-01
3.24844986e-01 -4.30848926e-01 -4.60953742e-01 6.75142229e-01
-2.68138200e-01 8.31767499e-01 1.04306757e+00 -8.26027393e-01
-4.56529915e-01 -2.40391612e-01 6.29710928e-02 3.45552802e-01
-5.86995006e-01 2.97742970e-02 -3.45593750e-01 -7.80543387e-01
2.20911484e-02 -9.07634258e-01 -1.23508377e-02 1.30126774e-01
-2.02537086e-02 -6.54023960e-02 1.07309568e+00 -3.61679733e-01
1.49206686e+00 -2.13448143e+00 3.31056982e-01 -1.02876462e-01
3.32169056e-01 1.92422450e-01 -2.01406762e-01 1.85268670e-01
-1.85492948e-01 3.40497168e-03 1.63356170e-01 -3.49257737e-01
-1.70847863e-01 5.00679761e-02 -3.12666088e-01 4.20401007e-01
2.55827278e-01 8.81339252e-01 -9.33356166e-01 -2.98624575e-01
1.79626703e-01 7.17534423e-01 -6.40636086e-01 1.66753978e-01
7.57609308e-02 6.17456377e-01 -3.37536901e-01 4.97260004e-01
3.37531924e-01 -3.79242450e-01 -1.40724257e-01 -4.72882807e-01
-2.24491060e-01 3.16217095e-01 -1.00067365e+00 1.73849893e+00
-4.34202462e-01 1.12461758e+00 2.56459620e-02 -8.64035845e-01
7.80911028e-01 5.57311952e-01 9.00675535e-01 -8.13178480e-01
3.49754214e-01 -6.89376071e-02 5.90250231e-02 -6.96517467e-01
5.06265104e-01 -1.06161624e-01 1.51634067e-01 3.46852303e-01
7.81872571e-02 3.29526484e-01 3.49188924e-01 1.62653789e-01
9.60853040e-01 2.11350024e-01 2.29347482e-01 -1.92324687e-02
4.58563536e-01 -5.13246655e-01 1.05654824e+00 4.14146125e-01
-6.63154960e-01 6.49993658e-01 6.01313531e-01 -5.52830040e-01
-5.57645321e-01 -7.05769241e-01 1.74236104e-01 1.04293585e+00
4.30952996e-01 -7.18595862e-01 -4.31352913e-01 -3.83472890e-01
-2.84966677e-01 1.01225100e-01 -6.36107087e-01 -3.38784486e-01
-6.70513511e-01 -5.33552408e-01 4.10213023e-01 7.20004320e-01
3.76499921e-01 -1.22493100e+00 -1.21495914e+00 3.28480482e-01
-5.31513035e-01 -1.25736892e+00 -6.68439925e-01 7.19237179e-02
-9.05248582e-01 -7.99272299e-01 -4.90642816e-01 -3.84883881e-01
1.28393084e-01 2.65733182e-01 9.76981103e-01 -3.98788154e-02
-2.57598519e-01 4.85457540e-01 -5.65536678e-01 -1.78584039e-01
2.98691809e-01 -1.69512913e-01 1.07078321e-01 4.36782479e-01
4.16142732e-01 -8.75470936e-01 -8.69894624e-01 2.86946267e-01
-7.96791136e-01 7.95963854e-02 1.28012806e-01 8.71708333e-01
4.24270034e-01 -1.60740942e-01 2.71150172e-01 -3.33084494e-01
4.31043416e-01 -3.41917187e-01 -1.60863265e-01 -2.19593182e-01
-2.37826318e-01 -1.03243098e-01 7.82479823e-01 -7.29528844e-01
-9.46555793e-01 -1.62180841e-01 -1.13638461e-01 -8.79930079e-01
-1.98535904e-01 4.58511412e-01 5.17552644e-02 8.42985287e-02
1.41694143e-01 5.78832300e-03 -2.47464478e-02 -6.43892288e-02
1.41962722e-01 2.42871478e-01 2.94533253e-01 -3.15754533e-01
2.39715368e-01 6.55557454e-01 9.59759858e-03 -1.09139180e+00
-6.73564970e-01 -6.28924310e-01 -5.36378682e-01 -5.96471667e-01
9.51681614e-01 -8.61961484e-01 -1.12154150e+00 7.03390360e-01
-1.09089720e+00 -5.61536670e-01 -1.89863622e-01 6.78208113e-01
-6.14680707e-01 2.31713548e-01 -1.05367398e+00 -7.44792879e-01
-3.15889210e-01 -1.11872947e+00 1.03488338e+00 3.84656131e-01
-4.79134619e-01 -1.23603356e+00 6.15704209e-02 -6.72976598e-02
4.70975101e-01 3.16350877e-01 3.72895151e-01 -5.37597761e-02
-3.00545961e-01 2.60499924e-01 -4.54887785e-02 6.79712072e-02
1.62261978e-01 3.13074887e-01 -1.00337863e+00 -8.58840197e-02
8.09176564e-02 -4.47327793e-01 1.07595325e+00 6.87774897e-01
1.29141605e+00 -2.46837065e-01 -3.90924327e-02 1.05726004e+00
1.10280681e+00 3.49263549e-01 7.55482554e-01 2.75828004e-01
8.64474833e-01 6.48097336e-01 7.47078896e-01 7.59271264e-01
4.52134728e-01 8.77926528e-01 3.47882420e-01 4.63139974e-02
7.96045288e-02 9.41492245e-02 7.39793479e-01 1.11765611e+00
-2.54479825e-01 -2.77961344e-01 -7.41896629e-01 3.95702183e-01
-2.03216672e+00 -1.41214907e+00 -2.40590386e-02 1.74847722e+00
5.52711308e-01 3.88854295e-02 2.13296503e-01 6.40825257e-02
3.80695999e-01 6.64339900e-01 -5.73250175e-01 -4.95896578e-01
-3.01191092e-01 4.83201891e-02 -4.36080061e-02 6.33465707e-01
-1.09918189e+00 8.92027974e-01 5.13609648e+00 5.31906664e-01
-1.78233266e+00 -2.13185903e-02 7.35787630e-01 -5.31499982e-01
-2.08648011e-01 -9.01308954e-02 -5.58713853e-01 6.30217791e-01
9.87038970e-01 -1.55253010e-02 2.72799134e-01 3.41833413e-01
7.00839996e-01 -1.67030692e-01 -9.29425061e-01 1.22055328e+00
4.52261278e-03 -1.31905425e+00 -1.88406542e-01 -1.07194178e-01
3.80441666e-01 -6.22952767e-02 1.00592338e-01 1.54137269e-01
-3.65038723e-01 -8.25725734e-01 7.85223544e-01 6.53024614e-01
6.15800083e-01 -7.06955254e-01 7.15970993e-01 6.26323968e-02
-1.69268262e+00 -5.29024079e-02 -1.64376259e-01 -2.58187592e-01
3.90356958e-01 7.02293754e-01 -3.49556990e-02 3.32278311e-01
9.33818877e-01 1.39547849e+00 -3.01313609e-01 7.08886147e-01
-3.09324618e-02 6.18584394e-01 -3.26886177e-01 -5.68080097e-02
2.58387923e-01 -3.71765167e-01 6.08281016e-01 1.45220804e+00
3.90674472e-01 2.91204423e-01 4.49935980e-02 5.73981762e-01
8.04806352e-02 -1.76685289e-01 -5.55162966e-01 -1.62878454e-01
2.22677082e-01 1.52729499e+00 -6.26186013e-01 -2.94642895e-01
-3.55325937e-01 1.10392785e+00 1.81912065e-01 6.52233005e-01
-9.68792140e-01 -4.25787300e-01 1.15890968e+00 -2.59472519e-01
3.78235996e-01 -3.66467923e-01 -9.33184028e-02 -1.59380770e+00
4.76867780e-02 -7.41081655e-01 3.81381363e-01 -7.50959814e-01
-9.68778789e-01 7.76328921e-01 -3.64600539e-01 -1.38884497e+00
-5.09497344e-01 -6.55553758e-01 -7.68173814e-01 4.59836721e-01
-1.52867198e+00 -6.50592208e-01 -6.20355308e-01 7.70972073e-01
5.46673715e-01 2.46408641e-01 6.55862510e-01 2.99284905e-01
-6.63356662e-01 5.66729307e-01 -3.23783994e-01 1.68470874e-01
8.36998105e-01 -1.02512002e+00 -6.15832135e-02 8.74912262e-01
1.87962085e-01 3.61815423e-01 7.38803685e-01 -2.06228629e-01
-1.42524397e+00 -9.24548209e-01 7.63092458e-01 -2.28385031e-01
7.68390238e-01 -3.68627399e-01 -8.66394877e-01 4.42959458e-01
2.57475525e-01 2.87436962e-01 6.60844624e-01 7.18885362e-02
-2.83559203e-01 -2.34854490e-01 -6.56202316e-01 5.70828140e-01
1.00133121e+00 -7.20283091e-01 -1.87528893e-01 -2.92902231e-01
5.71432531e-01 -3.51314723e-01 -8.96429896e-01 5.34410775e-01
8.91535819e-01 -1.33539486e+00 7.11245298e-01 -4.36692297e-01
7.44330406e-01 -1.95151493e-01 -1.07878134e-01 -1.32987010e+00
-2.57896900e-01 -9.08715010e-01 -5.47984600e-01 7.90058553e-01
1.52004376e-01 -5.31215847e-01 7.92167783e-01 2.47705370e-01
-5.92906512e-02 -1.14267552e+00 -7.32389688e-01 -5.09668648e-01
-3.65297168e-01 -6.30509913e-01 9.02282149e-02 1.11059225e+00
3.11639249e-01 4.12972063e-01 -5.64292908e-01 -4.64619659e-02
1.90898597e-01 3.57718736e-01 5.78765631e-01 -9.47832227e-01
-4.01799887e-01 -7.30718315e-01 -5.64430237e-01 -1.37971663e+00
2.15327978e-01 -4.12933469e-01 -6.77001616e-03 -1.14381135e+00
-5.25129884e-02 3.51602882e-02 -4.34401602e-01 4.49782968e-01
-1.68190345e-01 4.59831893e-01 3.95202219e-01 8.51431787e-02
-7.38479435e-01 1.05222344e+00 1.35367155e+00 9.47330594e-02
-4.60723370e-01 -2.88009584e-01 -3.62107903e-01 7.94744015e-01
8.39041471e-01 -1.20390847e-01 -3.73963296e-01 -3.73261720e-01
2.12415352e-01 3.39136243e-01 2.88389415e-01 -9.22915995e-01
2.64710397e-01 -1.45802483e-01 4.12684292e-01 -3.27323943e-01
7.84141421e-01 -6.03889108e-01 -9.81241316e-02 2.42813155e-01
-3.70984256e-01 3.71461838e-01 3.09110403e-01 5.00835717e-01
-6.44798517e-01 2.59522915e-01 6.74596071e-01 1.06564596e-01
-8.85872662e-01 5.90108633e-01 -5.52627742e-01 1.05966099e-01
8.30184042e-01 -4.40858364e-01 -3.24390054e-01 -8.24920058e-01
-9.19300199e-01 1.48287460e-01 2.06285164e-01 5.73846161e-01
7.60084689e-01 -1.37365735e+00 -3.21282953e-01 1.17810056e-01
-2.07706559e-02 -3.62537980e-01 3.54646444e-01 1.30109715e+00
-3.23057652e-01 4.21514809e-01 -3.38707060e-01 -8.34194183e-01
-1.23052728e+00 9.42839682e-02 4.54403758e-01 -9.86476839e-02
-5.16993463e-01 9.14362550e-01 3.15580130e-01 1.05582863e-01
1.70415327e-01 -4.45676237e-01 -3.69392991e-01 4.52287465e-01
6.12221420e-01 4.30312097e-01 -2.45397553e-01 -9.10771847e-01
-3.72318566e-01 7.97650814e-01 6.52289763e-02 -1.46575183e-01
1.38194549e+00 -4.21427339e-01 -8.87758434e-02 9.19593871e-01
1.56098342e+00 -1.69280514e-01 -1.66987479e+00 -1.85884293e-02
-2.50486314e-01 -4.14710909e-01 8.40072483e-02 -3.45936358e-01
-1.36465859e+00 1.26157713e+00 6.53354347e-01 1.31190971e-01
1.62036383e+00 -2.28153989e-01 1.01760173e+00 2.42507324e-01
1.42765880e-01 -1.05050325e+00 6.00678682e-01 8.48609686e-01
6.08149827e-01 -1.20011270e+00 -4.63721842e-01 -2.15781853e-01
-6.33682728e-01 1.40196133e+00 7.14897573e-01 -7.39068538e-03
7.30300665e-01 2.85469443e-01 2.73157209e-01 -2.02250108e-01
-1.27737534e+00 -1.53661117e-01 3.75211984e-01 7.25168139e-02
7.31842697e-01 -1.04937352e-01 -5.16064726e-02 3.36673200e-01
-1.63899720e-01 1.78986005e-02 4.77128774e-01 7.81912029e-01
-2.12475434e-01 -8.70604634e-01 -7.80304894e-02 3.19874525e-01
-4.79612917e-01 8.71488303e-02 -5.28534353e-02 5.07865906e-01
6.84866086e-02 9.65219617e-01 3.72833312e-01 -5.90244174e-01
5.98459914e-02 -1.97281912e-02 4.80859369e-01 -1.50911525e-01
-4.09786284e-01 4.00109917e-01 4.68800850e-02 -1.22694576e+00
-9.08231676e-01 -6.45942330e-01 -1.44390929e+00 -3.30201358e-01
-2.92212870e-02 6.03800230e-02 2.52231359e-01 9.38098431e-01
5.72566569e-01 6.59159839e-01 6.80154920e-01 -1.14539051e+00
8.95943120e-02 -8.46080303e-01 -4.45856363e-01 6.47112489e-01
6.50818944e-01 -5.65782428e-01 -5.14418483e-01 2.58449852e-01] | [13.529342651367188, 1.8160613775253296] |
d4ce851b-e9d2-465f-8ee5-b2347fd68b0c | a-novel-two-level-causal-inference-framework | 2304.04755 | null | https://arxiv.org/abs/2304.04755v1 | https://arxiv.org/pdf/2304.04755v1.pdf | A Novel Two-level Causal Inference Framework for On-road Vehicle Quality Issues Diagnosis | In the automotive industry, the full cycle of managing in-use vehicle quality issues can take weeks to investigate. The process involves isolating root causes, defining and implementing appropriate treatments, and refining treatments if needed. The main pain-point is the lack of a systematic method to identify causal relationships, evaluate treatment effectiveness, and direct the next actionable treatment if the current treatment was deemed ineffective. This paper will show how we leverage causal Machine Learning (ML) to speed up such processes. A real-word data set collected from on-road vehicles will be used to demonstrate the proposed framework. Open challenges for vehicle quality applications will also be discussed. | ['Anqi He', 'Kamran Paynabar', 'Devesh Upadhyay', 'Milad Zafar Nezhad', 'Thi Tu Trinh Tran', 'Huanyi Shui', 'Qian Wang'] | 2023-03-31 | null | null | null | null | ['causal-inference', 'causal-inference'] | ['knowledge-base', 'miscellaneous'] | [ 3.28313828e-01 3.05317640e-02 -1.02178228e+00 -3.55981141e-01
-8.10253561e-01 -2.33331352e-01 5.21649957e-01 1.70471475e-01
-2.72794254e-02 7.66065240e-01 4.88401115e-01 -7.98564553e-01
-6.28740847e-01 -8.70794058e-01 -4.88791198e-01 -4.96701777e-01
9.26717743e-02 1.97847441e-01 -1.68003604e-01 -5.61947487e-02
6.73584700e-01 5.49625397e-01 -1.64656067e+00 2.56343365e-01
7.30749905e-01 5.05563915e-01 -2.50513311e-02 3.19692254e-01
-7.90749565e-02 6.94171369e-01 -3.26272905e-01 -3.18641186e-01
7.72659481e-02 -3.33465517e-01 -8.22219312e-01 1.96317196e-01
3.07500362e-01 -4.00750309e-01 -2.60567963e-01 7.29290068e-01
2.98528880e-01 -1.47196591e-01 8.66478205e-01 -1.79266715e+00
-3.37677181e-01 5.40351093e-01 -3.11087608e-01 2.93600172e-01
1.54823065e-01 4.70262736e-01 1.08858895e+00 -7.51378596e-01
3.85033786e-01 1.63119066e+00 2.97764212e-01 5.17019153e-01
-9.70566988e-01 -1.17803884e+00 5.34975119e-02 6.40848160e-01
-1.04160523e+00 -5.56872189e-01 6.09783053e-01 -8.00418675e-01
8.81249487e-01 5.80382682e-02 6.35149658e-01 1.02206075e+00
6.50967658e-01 6.09246433e-01 1.05515826e+00 -2.56295681e-01
3.15756410e-01 -1.74647167e-01 4.53548849e-01 3.64107400e-01
3.60059649e-01 3.82162094e-01 -5.25858521e-01 -2.13683203e-01
1.84484005e-01 1.01748534e-01 1.05014168e-01 8.40876773e-02
-4.47160006e-01 1.03486478e+00 -1.07206786e-02 -8.99574310e-02
-5.48247576e-01 4.48131174e-01 4.68795240e-01 2.85051852e-01
3.30564946e-01 3.64572257e-01 -6.03102624e-01 -5.17018199e-01
-7.33511806e-01 4.89323318e-01 4.50020015e-01 7.15118289e-01
6.87943757e-01 -1.10852249e-01 -2.39066511e-01 7.46746361e-01
4.36000735e-01 3.98023158e-01 -1.39588803e-01 -1.23605335e+00
2.28959501e-01 5.29678106e-01 7.46513829e-02 -7.56382763e-01
-1.97992399e-01 5.56051917e-02 -3.86024863e-01 1.99053675e-01
2.35857666e-02 3.22488789e-03 -7.39749908e-01 1.32288635e+00
1.55391321e-01 4.77140456e-01 -3.05943578e-01 4.87393916e-01
5.12983680e-01 4.82518464e-01 5.56310415e-01 -5.97119093e-01
1.20987821e+00 -4.30721521e-01 -1.20300710e+00 -4.85609859e-01
6.81858599e-01 -7.21733212e-01 9.76516843e-01 5.14028132e-01
-8.59010518e-01 -2.10782871e-01 -1.02748537e+00 1.68142289e-01
-2.78583974e-01 -1.70395330e-01 6.43088043e-01 7.24105597e-01
-5.04828155e-01 5.17089009e-01 -4.50698227e-01 -2.10127383e-01
6.24333441e-01 3.70337814e-01 -1.59143373e-01 -4.41959947e-01
-1.33993995e+00 1.06291223e+00 -7.47034028e-02 1.19515650e-01
-1.22887182e+00 -1.23756742e+00 -6.43137515e-01 -1.73523858e-01
5.82818747e-01 -6.82395399e-01 1.31917691e+00 -3.16684812e-01
-1.13299215e+00 2.99794406e-01 -4.75287855e-01 -2.46538356e-01
1.25961145e-02 -4.28512514e-01 -4.69999373e-01 -4.42950755e-01
3.23005974e-01 3.41051817e-01 7.01551616e-01 -1.17202103e+00
-1.08729756e+00 -4.27423209e-01 -5.66653721e-02 -3.58289033e-01
-4.26870793e-01 2.73783088e-01 -3.07990909e-01 -2.81669945e-01
-2.57216126e-01 -1.12549162e+00 -2.37461939e-01 -6.11500084e-01
-2.36195475e-01 -6.44816160e-01 9.72916961e-01 -6.22900903e-01
1.81008494e+00 -1.94317389e+00 -4.94262949e-02 1.23193480e-01
3.37154418e-01 -9.65056121e-02 -1.59151837e-01 5.81208527e-01
-4.67626035e-01 5.93795419e-01 1.08696789e-01 1.16864227e-01
-9.06158313e-02 4.30318773e-01 -3.19296092e-01 5.43027461e-01
6.50359809e-01 6.68237090e-01 -1.02187991e+00 -3.88835877e-01
4.78671640e-01 1.82882249e-01 -5.83752632e-01 1.18530743e-01
1.65568013e-02 5.51518938e-03 -4.62383568e-01 7.96062291e-01
5.03942132e-01 1.31700769e-01 1.51066229e-01 -3.03003103e-01
-3.26926708e-01 8.18974853e-01 -1.10524321e+00 7.30592072e-01
-6.94909573e-01 7.21090555e-01 -2.34021693e-01 -9.36890841e-01
6.40094757e-01 1.18967026e-01 7.55726278e-01 -9.60152626e-01
2.36875698e-01 4.50652689e-02 1.60620242e-01 -9.53631699e-01
1.81389153e-01 -2.67471135e-01 -1.42131418e-01 2.50247270e-01
-2.55494326e-01 -5.23489080e-02 2.15833828e-01 -1.90287512e-02
1.52691722e+00 -3.38782042e-01 3.19579065e-01 -2.25004572e-02
2.17671141e-01 7.75280744e-02 9.45030272e-01 4.43240434e-01
-4.36715901e-01 7.70231560e-02 7.31062651e-01 -1.28069907e-01
-7.71601379e-01 -7.73494780e-01 -2.27920771e-01 8.68914247e-01
-1.05059743e-01 -6.55492067e-01 -3.41141611e-01 -8.32988501e-01
2.57767171e-01 1.26238549e+00 -5.73024809e-01 -4.80920106e-01
-3.98522407e-01 -6.11799717e-01 1.68210149e-01 3.98080826e-01
-3.53248939e-02 -7.87732601e-01 -2.34024301e-01 7.62967542e-02
-1.28442675e-01 -8.67711663e-01 -9.34176818e-02 -4.56586154e-03
-8.55552971e-01 -1.34093750e+00 2.35827923e-01 -5.95134683e-02
2.97856480e-01 5.11763275e-01 1.02621388e+00 1.70738414e-01
-2.57860214e-01 4.84037623e-02 -7.58347139e-02 -5.29582322e-01
-6.11049592e-01 -3.30727369e-01 1.73756227e-01 -4.29064631e-01
7.07787275e-01 -2.64793515e-01 -4.87537414e-01 4.17938262e-01
-5.67648053e-01 -2.94824839e-01 6.88069344e-01 7.21193254e-01
4.34000134e-01 6.11435175e-01 8.72505128e-01 -8.10733914e-01
9.46983457e-01 -9.33395386e-01 -5.00168979e-01 1.63654208e-01
-1.22689939e+00 -1.62384793e-01 -1.27162904e-01 -3.10011834e-01
-8.67407203e-01 -1.41374648e-01 -1.35349140e-01 -2.52735317e-01
-1.87605783e-01 6.16682589e-01 -3.56369555e-01 5.19151747e-01
4.39183384e-01 -7.15904713e-01 -8.87739286e-02 -3.54043871e-01
5.38394868e-01 9.24609125e-01 1.65432870e-01 -4.25009459e-01
6.44636214e-01 1.91408589e-01 6.21733181e-02 -5.35211563e-01
-6.36333108e-01 -5.75622618e-01 -3.91267121e-01 -6.41482890e-01
7.79909670e-01 -7.98673749e-01 -9.26555514e-01 -8.05824921e-02
-1.05874312e+00 -2.95862556e-01 9.51164588e-03 4.68391895e-01
-5.47427416e-01 -1.59382060e-01 -2.75166363e-01 -9.46466923e-01
5.49637415e-02 -1.33717728e+00 8.89554262e-01 5.49804866e-02
-6.49148464e-01 -6.57253563e-01 -5.45882955e-02 7.44455457e-01
7.69072399e-02 -2.05444977e-01 1.38299882e+00 -2.49285579e-01
-5.12214363e-01 -2.88470954e-01 -1.76711112e-01 2.34312639e-01
6.26145422e-01 4.33537096e-01 -6.78395987e-01 1.11679018e-01
-1.81288078e-01 1.86861716e-02 5.79062521e-01 7.74579704e-01
1.06660259e+00 -3.07591140e-01 -5.51180303e-01 -1.01470329e-01
1.12380099e+00 5.10403395e-01 8.52227390e-01 2.76144296e-01
5.57239175e-01 1.03257763e+00 1.28172338e+00 4.04835075e-01
4.61572856e-01 6.59783065e-01 4.96290088e-01 -5.32997819e-03
-3.96103442e-01 -1.30884007e-01 4.81485099e-01 5.32426119e-01
2.05217957e-01 -2.67685682e-01 -1.15767682e+00 7.16174364e-01
-1.88316035e+00 -1.05100644e+00 -7.03960419e-01 1.91852784e+00
6.65546358e-01 4.62851793e-01 1.23665407e-01 3.74589652e-01
4.28173631e-01 -4.94402468e-01 -2.50002950e-01 -6.93167329e-01
4.26072091e-01 -1.03166789e-01 7.53815413e-01 5.96163392e-01
-9.75438714e-01 9.53123748e-01 7.78225994e+00 8.81265283e-01
-7.83840179e-01 1.56400993e-01 7.10924566e-01 -1.07781895e-01
-4.36195999e-01 3.74928087e-01 -8.10220063e-01 3.05528015e-01
1.34287238e+00 -3.49075586e-01 1.85573101e-01 5.98361135e-01
1.28905559e+00 -2.10364550e-01 -1.20430219e+00 5.24791479e-01
-3.82009536e-01 -1.29390216e+00 -1.52681870e-02 3.10188353e-01
5.89837670e-01 -1.05025649e-01 -5.00795357e-02 3.00555974e-01
4.84864086e-01 -1.12646031e+00 3.71644884e-01 6.24342084e-01
4.25270319e-01 -9.89028454e-01 5.57801545e-01 1.33708060e-01
-7.87899673e-01 -6.51300192e-01 -1.15984961e-01 -3.56237292e-01
3.13430786e-01 8.11854422e-01 -1.06899750e+00 3.81506413e-01
7.59264290e-01 8.17343354e-01 -3.89924079e-01 1.01676762e+00
-3.40573400e-01 1.11823070e+00 2.04316348e-01 1.23407483e-01
-6.66522309e-02 -1.32812530e-01 2.85286427e-01 1.15096414e+00
3.73390317e-01 3.29322666e-02 1.44288009e-02 7.25377798e-01
3.40572387e-01 -8.69521946e-02 -9.31840420e-01 -2.46197686e-01
5.75031996e-01 1.08262551e+00 -3.37353706e-01 -8.20810571e-02
-7.23575592e-01 1.74713686e-01 -2.14336917e-01 3.69477868e-01
-8.58041286e-01 1.22837998e-01 1.19030392e+00 5.25347650e-01
-8.54569525e-02 -8.36035684e-02 -8.60170901e-01 -1.28912747e-01
-4.16605890e-01 -1.10304523e+00 3.40550900e-01 -6.58007920e-01
-1.29814398e+00 -2.24208564e-01 4.32670891e-01 -1.18348110e+00
-2.38521233e-01 -3.86997879e-01 -3.72222543e-01 7.40794897e-01
-1.37333798e+00 -8.55698884e-01 -7.68199116e-02 9.66418236e-02
7.87220120e-01 -1.20838642e-01 3.97852242e-01 7.40127563e-01
-8.64885151e-01 2.63132811e-01 -4.24754351e-01 -4.12478030e-01
7.67677963e-01 -8.38619649e-01 -4.12539765e-02 5.83342433e-01
-5.90713620e-01 6.83349371e-01 1.14308274e+00 -1.15172482e+00
-1.71640897e+00 -1.09423947e+00 1.18252659e+00 -4.25576240e-01
1.00750434e+00 -1.58228189e-01 -8.26105893e-01 2.94865698e-01
2.52172053e-01 -4.67816263e-01 7.99836934e-01 5.99040627e-01
-1.98435783e-01 -3.21828365e-01 -8.68850827e-01 7.64279783e-01
8.52836549e-01 -4.14753586e-01 -3.89924407e-01 3.12446654e-01
7.23732948e-01 3.56643319e-01 -7.86815286e-01 7.92751729e-01
4.10621345e-01 -4.10231203e-01 7.19549894e-01 -8.60373199e-01
6.69381678e-01 -1.86923265e-01 -5.20501621e-02 -1.15052366e+00
-4.23218250e-01 -5.39157987e-01 5.01643941e-02 1.41763389e+00
4.35943305e-01 -1.31680489e-01 4.80053723e-01 9.96788323e-01
-9.16161165e-02 -8.19041491e-01 -9.19157684e-01 -5.39424658e-01
1.58694968e-01 -1.16876507e+00 5.20402849e-01 8.59430552e-01
-1.20102972e-01 6.35982156e-01 -4.11871523e-01 3.39869022e-01
5.42108893e-01 -3.42971802e-01 8.44716311e-01 -1.21450436e+00
3.11209187e-02 -6.51546776e-01 -3.46802741e-01 -4.93637562e-01
4.25762028e-01 -5.53586423e-01 1.43270910e-01 -1.65023506e+00
3.44186038e-01 -4.56293195e-01 -3.23123723e-01 7.01589108e-01
-3.96710694e-01 -2.29104921e-01 -1.59051046e-01 -1.71359405e-01
-9.82903168e-02 4.47008252e-01 9.25253808e-01 -3.97547901e-01
-1.23326860e-01 7.78368190e-02 -9.04897213e-01 6.26188934e-01
8.71982515e-01 -8.49516213e-01 -6.14157915e-01 -7.76944822e-03
3.72367293e-01 1.81774229e-01 4.29992378e-01 -3.01010072e-01
6.18388355e-02 -1.09079123e+00 -2.97144324e-01 -6.95138037e-01
-8.98879543e-02 -8.75279665e-01 3.19933176e-01 5.59329748e-01
-3.53534967e-01 2.34409511e-01 4.19240177e-01 6.10955238e-01
-3.47957909e-02 -3.74564022e-01 4.24020499e-01 5.13063729e-01
-7.02532470e-01 1.27610162e-01 -7.14868426e-01 -3.88092548e-01
1.12165570e+00 7.78897852e-02 -3.94068211e-01 -3.59175384e-01
-3.27787817e-01 4.79249030e-01 -5.80767579e-02 1.00131500e+00
6.85780764e-01 -1.46312535e+00 -7.59301186e-01 -2.51823604e-01
1.75584987e-01 -7.20707178e-01 2.23325312e-01 6.67234480e-01
2.20423251e-01 4.36978370e-01 1.07851170e-01 -2.89678037e-01
-1.52955711e+00 6.99012220e-01 1.73952296e-01 -7.09440485e-02
-2.94276565e-01 3.76064390e-01 1.06243022e-01 9.85682830e-02
1.64786682e-01 -8.09725225e-02 -4.31294203e-01 2.96544433e-01
4.64263380e-01 8.61351669e-01 3.66271973e-01 -4.00755495e-01
-4.72929001e-01 2.62222916e-01 1.90372437e-01 -7.94003159e-02
1.38146412e+00 -1.94045007e-01 -2.91185945e-01 5.31158566e-01
1.15147018e+00 -1.81302026e-01 -8.40562403e-01 2.81382203e-01
3.03421766e-01 -6.83803976e-01 6.11564755e-01 -6.89558327e-01
-1.09439063e+00 5.97166181e-01 8.73563826e-01 1.30216360e-01
1.19466674e+00 -4.88349125e-02 5.25179207e-01 3.20466422e-02
1.69343323e-01 -1.25369966e+00 -1.22164689e-01 2.49687284e-01
1.08007395e+00 -1.14104152e+00 2.37431690e-01 -6.82487547e-01
-2.64718324e-01 9.25218225e-01 5.39999902e-01 2.00878963e-01
1.07558441e+00 4.24728394e-01 -2.63809830e-01 -5.37678123e-01
-1.31206775e+00 -2.37875938e-01 2.83825457e-01 2.63360947e-01
5.64968824e-01 2.53017724e-01 -8.27545404e-01 5.00439227e-01
1.03440970e-01 1.82954535e-01 5.31913459e-01 8.94005716e-01
-2.91437000e-01 -1.56542754e+00 -3.97321403e-01 7.43672669e-01
-3.78112882e-01 -4.77892580e-03 -3.90967578e-01 7.85231411e-01
4.39708441e-01 1.69254100e+00 -1.26292795e-01 -7.17248499e-01
6.66560292e-01 1.13332152e-01 1.93464458e-01 -5.28232336e-01
-2.83639342e-01 1.47345334e-01 3.63626033e-01 -7.88640618e-01
-3.58989030e-01 -9.55676019e-01 -1.28084898e+00 -5.39429665e-01
-3.06415290e-01 -2.99753159e-01 7.23860085e-01 1.16181684e+00
2.99148798e-01 9.66280580e-01 8.51291120e-01 -2.86572307e-01
-2.32667238e-01 -7.61804819e-01 5.60647584e-02 2.73634762e-01
3.42485696e-01 -1.24093187e+00 -4.29198623e-01 3.25852334e-02] | [7.849998950958252, 5.3227152824401855] |
64fb67db-e3b1-4aff-95de-df64615ea9ef | a-unified-implicit-dialog-framework-for | 1802.04358 | null | http://arxiv.org/abs/1802.04358v1 | http://arxiv.org/pdf/1802.04358v1.pdf | A Unified Implicit Dialog Framework for Conversational Search | We propose a unified Implicit Dialog framework for goal-oriented, information
seeking tasks of Conversational Search applications. It aims to enable dialog
interactions with domain data without replying on explicitly encoded the rules
but utilizing the underlying data representation to build the components
required for dialog interaction, which we refer as Implicit Dialog in this
work. The proposed framework consists of a pipeline of End-to-End trainable
modules. A centralized knowledge representation is used to semantically ground
multiple dialog modules. An associated set of tools are integrated with the
framework to gather end users' input for continuous improvement of the system.
The goal is to facilitate development of conversational systems by identifying
the components and the data that can be adapted and reused across many end-user
applications. We demonstrate our approach by creating conversational agents for
several independent domains. | ['Song Feng', 'Kshitij P. Fadnis', 'Sunil Shashidhara', 'R. Chulaka Gunasekara', 'Lazaros C. Polymenakos'] | 2018-02-12 | null | null | null | null | ['conversational-search'] | ['natural-language-processing'] | [-5.51252849e-02 8.56092691e-01 2.34508708e-01 -6.16091311e-01
-5.05143881e-01 -8.70085716e-01 9.61588919e-01 6.14124276e-02
-2.58898795e-01 6.01575851e-01 4.41220939e-01 -3.87427002e-01
-3.70545954e-01 -5.62759757e-01 1.40778124e-01 -1.08562939e-01
2.54773587e-01 1.13982880e+00 4.48451996e-01 -8.03046227e-01
3.55541140e-01 2.48056695e-01 -1.43639863e+00 7.95386255e-01
1.01477408e+00 5.33981144e-01 6.28464997e-01 8.76964033e-01
-7.48615801e-01 1.23715866e+00 -5.43466389e-01 -4.48268175e-01
-6.00377321e-02 -5.26139140e-01 -1.65928817e+00 8.75511020e-02
-2.11716831e-01 -5.38424909e-01 3.31273109e-01 5.99780262e-01
2.55101621e-01 5.85603952e-01 5.23231983e-01 -1.30138958e+00
-1.97654441e-01 8.81529629e-01 6.82832062e-01 -4.01483804e-01
7.78097093e-01 3.98799181e-01 1.05925691e+00 -7.60447800e-01
8.56176019e-01 1.48149669e+00 2.07041189e-01 1.09312212e+00
-1.03381217e+00 -1.41373565e-02 8.84611011e-02 7.39851967e-02
-7.34760523e-01 -8.02414060e-01 8.51035953e-01 -6.34800434e-01
1.24318933e+00 3.82415146e-01 5.93221411e-02 8.91517818e-01
-4.90227222e-01 4.83875513e-01 7.60079503e-01 -7.28053272e-01
3.00088823e-01 9.33279455e-01 9.42443967e-01 6.97163165e-01
-3.70483637e-01 -1.91933274e-01 -6.24235928e-01 -4.44112718e-01
4.58478391e-01 -2.64442116e-01 -7.69049721e-03 -5.37589014e-01
-6.05246902e-01 9.06891882e-01 2.12943867e-01 7.02984869e-01
-4.29698944e-01 -3.63711447e-01 2.55728930e-01 5.03927350e-01
1.75819211e-02 6.77448809e-01 -5.91944933e-01 -2.21541092e-01
-1.57504708e-01 5.16278923e-01 1.77490389e+00 1.25872493e+00
9.89077151e-01 -6.59103036e-01 -3.36096823e-01 1.01251400e+00
5.96465945e-01 -5.11596464e-02 5.08740664e-01 -1.04948056e+00
3.75594616e-01 1.57314456e+00 6.64595783e-01 -3.71830732e-01
-4.76889759e-01 2.83329546e-01 -9.39829927e-03 3.24801117e-01
5.17574787e-01 -2.84748971e-01 -4.83657181e-01 1.60845745e+00
4.43813562e-01 -3.78428847e-01 5.56654394e-01 5.49232602e-01
1.27317989e+00 1.64826006e-01 3.93695325e-01 4.82335053e-02
1.68702400e+00 -1.20991015e+00 -7.85292149e-01 -1.91535980e-01
8.52269292e-01 -6.06947958e-01 1.45030713e+00 -5.86041156e-03
-1.25447190e+00 -5.96857429e-01 -6.83866739e-01 -3.87078643e-01
-8.94254982e-01 -1.30240500e-01 2.93814003e-01 5.48999548e-01
-1.23876560e+00 2.30117932e-01 -4.27820742e-01 -7.45424330e-01
-3.28989923e-01 5.31578839e-01 -6.03809133e-02 4.47879404e-01
-1.14396644e+00 1.19688308e+00 5.57921886e-01 -2.44397700e-01
-7.19520092e-01 -4.27228421e-01 -8.17640603e-01 3.15409720e-01
5.25675893e-01 -1.00929332e+00 2.02805257e+00 -7.97782421e-01
-2.18567753e+00 8.70185733e-01 -1.32181451e-01 -3.33253980e-01
3.69036794e-01 -4.24655288e-01 8.76416564e-02 -1.53218031e-01
-8.85074660e-02 5.75616360e-01 3.50838929e-01 -1.36097908e+00
-8.08703959e-01 -3.10208559e-01 6.50188804e-01 6.16341650e-01
-2.80114383e-01 1.15297221e-01 -6.04910791e-01 1.20104097e-01
-4.68582481e-01 -7.87752986e-01 -3.47956717e-01 -6.69848800e-01
-1.78960666e-01 -6.13215506e-01 9.50660586e-01 -6.24858379e-01
1.03390098e+00 -1.53656137e+00 5.60927808e-01 -4.26969342e-02
2.38394186e-01 2.04906225e-01 -7.44936690e-02 8.53389144e-01
3.96432936e-01 -1.55997604e-01 -1.54131129e-01 -5.58670163e-01
5.27642250e-01 1.84720427e-01 -1.83906689e-01 -6.98844671e-01
2.38432717e-02 8.53517532e-01 -9.42375660e-01 -4.46973771e-01
5.83587110e-01 -1.66627765e-02 -6.78277612e-01 1.15518248e+00
-9.66590345e-01 6.99891210e-01 -6.81286514e-01 2.15576410e-01
3.67982328e-01 -2.74961978e-01 6.23041987e-01 5.11690304e-02
-2.34647065e-01 7.96791375e-01 -1.05132675e+00 2.09659195e+00
-8.55344892e-01 4.69463207e-02 4.71950680e-01 -6.43060863e-01
9.83019590e-01 5.73533952e-01 1.22084655e-01 -3.51549983e-01
2.57197708e-01 -9.06829089e-02 -1.22669503e-01 -8.92337084e-01
6.84856653e-01 2.83997238e-01 -2.85808086e-01 8.52579474e-01
4.77703035e-01 -2.01011539e-01 1.46984920e-01 4.23258364e-01
9.21920419e-01 7.59122297e-02 4.49586838e-01 -2.54626781e-01
1.10369670e+00 3.90391409e-01 -1.52608305e-01 7.47081101e-01
1.48430556e-01 -3.58197391e-01 3.18854183e-01 -3.97114426e-01
-6.34798706e-01 -8.96488726e-01 2.54576713e-01 1.67031527e+00
2.13992074e-02 -5.20408511e-01 -8.76799226e-01 -9.46377218e-01
-3.91754538e-01 1.01938379e+00 -3.79323363e-01 1.56924710e-01
-2.97114789e-01 7.68384803e-03 3.71165067e-01 1.50092468e-01
6.55279756e-01 -1.62823999e+00 -6.51230216e-01 3.54911596e-01
-3.02821279e-01 -9.57298100e-01 -6.77106902e-02 1.22411095e-01
-7.11512506e-01 -1.24508071e+00 -2.28698730e-01 -8.59849155e-01
2.45309860e-01 -6.37127971e-03 1.57307172e+00 3.48303527e-01
4.92473692e-02 9.39682543e-01 -6.14108443e-01 -3.34633261e-01
-9.95425582e-01 4.17702854e-01 -5.05742908e-01 -2.28596434e-01
7.56675541e-01 -5.18480301e-01 -3.85244042e-01 5.91252148e-01
-7.58752167e-01 3.34759831e-01 2.16122195e-01 8.23500156e-01
-3.51200372e-01 -5.47824383e-01 7.24745572e-01 -1.13449991e+00
1.56622791e+00 -4.78974611e-01 -6.67502701e-01 6.13832653e-01
-4.37970668e-01 4.16936040e-01 2.37212375e-01 -1.65039957e-01
-1.73023784e+00 2.56877631e-01 -3.07767093e-01 3.58983636e-01
-7.05525637e-01 5.13158441e-01 -4.07594383e-01 1.69341758e-01
8.79410148e-01 -1.46221248e-02 -1.06495351e-01 -8.29943597e-01
9.38091338e-01 1.16979265e+00 5.10842323e-01 -9.64118123e-01
3.45141977e-01 -2.16275960e-01 -6.92629099e-01 -7.32988536e-01
-4.55700427e-01 -8.04935575e-01 -9.61444616e-01 -2.09397048e-01
1.10873413e+00 -4.69652623e-01 -1.00128329e+00 1.29648775e-01
-1.42645943e+00 -7.97776222e-01 -2.46463731e-01 -1.35732114e-01
-7.66224504e-01 2.07359478e-01 -4.58383620e-01 -1.21845722e+00
-7.54371047e-01 -1.30294371e+00 8.25507641e-01 5.55223703e-01
-6.84340179e-01 -1.36358237e+00 6.42468929e-01 5.64678550e-01
5.51221371e-01 -4.38055575e-01 1.02448153e+00 -1.42391932e+00
-5.33630252e-01 1.46099418e-01 1.00928918e-01 9.25479904e-02
4.02220637e-01 -4.92390633e-01 -1.25163794e+00 2.37122267e-01
6.28682971e-02 -6.05456650e-01 -2.92224437e-02 -1.70467749e-01
4.44593787e-01 -5.09643197e-01 -4.32352245e-01 -1.49911925e-01
8.34183276e-01 5.58525324e-01 4.54111487e-01 2.73775965e-01
7.25945085e-02 1.24631357e+00 5.59721768e-01 2.20172748e-01
5.67058980e-01 1.13256836e+00 1.12891931e-03 5.35936356e-02
7.25767240e-02 -1.07833907e-01 2.62881100e-01 4.22924191e-01
1.35644883e-01 -1.58817664e-01 -9.36880887e-01 5.16679287e-01
-2.20807552e+00 -7.79193461e-01 1.97928578e-01 1.85065746e+00
9.79613900e-01 -8.28363374e-02 2.85784960e-01 -5.17090976e-01
4.00864780e-01 -3.02285641e-01 -4.84839559e-01 -6.42413259e-01
5.44143319e-01 3.13742191e-01 -5.28459728e-01 1.28166986e+00
-7.46841967e-01 1.39435494e+00 5.94637632e+00 2.82356776e-02
-4.81898367e-01 7.77308717e-02 1.41817704e-01 6.44162297e-01
-1.60137877e-01 3.36446732e-01 -8.36913049e-01 -6.42247349e-02
1.02452850e+00 -4.71478164e-01 5.84093809e-01 1.26934588e+00
1.83870226e-01 1.26298726e-01 -1.50133824e+00 3.53851587e-01
-3.56066674e-01 -1.41398907e+00 7.41774961e-02 -2.32520819e-01
1.04455948e-01 -2.59018302e-01 -3.46295595e-01 6.62357211e-01
1.07395649e+00 -8.33530188e-01 4.66587814e-03 5.41055262e-01
2.32021078e-01 -1.77729428e-01 6.66717172e-01 7.57182717e-01
-9.76781607e-01 -6.91545606e-02 -3.22155468e-02 -9.30598155e-02
1.92524448e-01 -3.66374552e-01 -1.82831299e+00 7.30988503e-01
4.01250124e-01 1.95242271e-01 -4.73251164e-01 5.47599733e-01
-1.99446172e-01 1.01304255e-01 -2.11091295e-01 -2.05704272e-01
1.58736825e-01 -3.37422997e-01 4.90019441e-01 1.39166367e+00
-5.73197342e-02 2.84601420e-01 4.63788062e-01 1.08254337e+00
6.93606138e-02 3.24669540e-01 -8.29640329e-01 4.18384820e-02
5.51637650e-01 1.53313208e+00 -1.82730198e-01 -5.57375968e-01
-4.65892524e-01 8.83088768e-01 3.97313446e-01 2.51128852e-01
-1.55220315e-01 -3.70950252e-01 5.95578015e-01 -1.06167220e-01
-1.58915102e-01 -8.92717242e-02 1.27983727e-02 -1.00857103e+00
-3.43687445e-01 -1.16739476e+00 7.60421515e-01 -9.41892266e-01
-1.31259704e+00 1.01920497e+00 2.89348096e-01 -7.17597663e-01
-1.08061171e+00 -6.74495757e-01 -1.05963624e+00 1.37254846e+00
-9.75129664e-01 -1.32918763e+00 -8.20550919e-01 9.78335142e-01
1.05199862e+00 -4.80715096e-01 1.41272807e+00 -1.60527036e-01
-1.85626164e-01 2.86199719e-01 -6.18620753e-01 1.21598512e-01
6.68273568e-01 -1.36009848e+00 2.43896842e-01 3.17726105e-01
-4.37187217e-02 1.16610551e+00 8.36470604e-01 -6.71139657e-01
-1.12932181e+00 -5.08384168e-01 8.12080085e-01 -9.36852753e-01
6.36143267e-01 -4.20475245e-01 -9.45735693e-01 8.93740594e-01
1.00703275e+00 -1.10060704e+00 8.56778562e-01 5.01489639e-01
-6.29521534e-02 4.14592028e-01 -1.18315160e+00 6.18978977e-01
8.10550451e-01 -6.17352545e-01 -1.45420098e+00 3.44859183e-01
8.03014517e-01 -3.08792055e-01 -7.40517437e-01 6.45795539e-02
2.84690499e-01 -9.45063531e-01 9.73436534e-01 -9.96923268e-01
3.60581540e-02 -8.34583789e-02 5.95951304e-02 -1.20776212e+00
-7.41993356e-03 -9.26734388e-01 -1.20530270e-01 1.47454965e+00
6.74306870e-01 -5.71185172e-01 7.26011038e-01 1.37300265e+00
-1.96104988e-01 4.75586839e-02 -4.94976103e-01 -1.67617798e-01
-6.84118718e-02 -2.47277692e-01 5.49254298e-01 8.16142023e-01
9.39470351e-01 1.20961523e+00 -3.46427783e-02 2.00924769e-01
2.03292876e-01 2.31300488e-01 1.40725458e+00 -1.52075219e+00
-4.20182228e-01 -4.41628903e-01 2.90977865e-01 -1.51306832e+00
1.57169253e-01 -8.59420121e-01 1.67766258e-01 -1.55761421e+00
-3.61859351e-02 -6.48718476e-01 -4.08832729e-02 4.36574310e-01
-2.13989183e-01 -7.52858222e-01 4.32045832e-02 2.49743238e-01
-7.02016354e-01 4.91110712e-01 7.33348727e-01 -1.47856370e-01
-1.02710104e+00 3.45687300e-01 -7.98212230e-01 7.49617398e-01
8.28539908e-01 -1.74869910e-01 -9.04517293e-01 -1.73241928e-01
-2.48058692e-01 2.25827858e-01 2.74597973e-01 -9.27652836e-01
6.26195490e-01 -3.19777638e-01 -2.11875305e-01 -5.61946452e-01
5.69597483e-01 -9.52037692e-01 8.29544887e-02 3.16207767e-01
-8.84304106e-01 5.56487450e-03 2.40329146e-01 1.03634328e-01
-1.24278575e-01 -7.24167407e-01 3.71131033e-01 -3.79992932e-01
-7.44236350e-01 -2.94034779e-01 -5.73287487e-01 -2.81844825e-01
9.76133108e-01 1.52854428e-01 -4.57807034e-01 -6.82839096e-01
-1.10459876e+00 5.59134007e-01 3.81961405e-01 5.56847334e-01
3.32692742e-01 -7.53615379e-01 -3.27379733e-01 -1.05912410e-01
3.60437512e-01 -1.60826147e-01 5.68862744e-02 7.85449967e-02
-3.06753606e-01 6.91784501e-01 -4.29758042e-01 -4.83418405e-01
-1.53103161e+00 4.73733634e-01 5.55280089e-01 -8.00921500e-01
-3.37184280e-01 6.29753947e-01 2.61536300e-01 -9.45484877e-01
4.72040623e-01 -1.63119793e-01 -8.42623413e-01 -5.02597913e-02
6.79202318e-01 -1.48026561e-02 3.19785066e-02 -1.43965364e-01
-1.17063805e-01 2.07996257e-02 -1.57434717e-01 -6.67880237e-01
1.21303737e+00 -3.52473408e-01 -3.07081565e-02 2.73214072e-01
3.96933466e-01 -4.24275324e-02 -8.44227731e-01 -4.54905838e-01
6.02845728e-01 5.38837053e-02 -4.55972165e-01 -1.32045221e+00
-1.84044406e-01 7.71729112e-01 6.15438282e-01 6.77681327e-01
8.25141251e-01 3.11347902e-01 1.97277769e-01 1.24123144e+00
3.89444321e-01 -1.03005815e+00 1.56264305e-01 7.71875739e-01
1.10086894e+00 -1.16698349e+00 -3.39010298e-01 -4.43017364e-01
-1.00560248e+00 1.19154906e+00 9.05527949e-01 4.05576974e-01
4.99468565e-01 2.43415594e-01 3.63377094e-01 -5.68757355e-01
-1.13100612e+00 -5.11593103e-01 2.99037367e-01 8.15867841e-01
5.54262578e-01 -3.76607388e-01 -4.08676714e-01 1.02457798e+00
-1.15877531e-01 1.86995983e-01 6.54129013e-02 1.14457095e+00
-6.75095439e-01 -1.74487102e+00 -1.35898357e-02 -1.61323339e-01
2.24640906e-01 -9.33646485e-02 -1.21148491e+00 7.21120298e-01
-1.60959795e-01 1.39926255e+00 -2.02949524e-01 -4.58947927e-01
7.19719231e-01 9.41358387e-01 6.75334930e-02 -9.81804192e-01
-1.24536359e+00 -2.87222207e-01 7.99155176e-01 -4.32432950e-01
-5.10767698e-01 -1.74174666e-01 -1.02996349e+00 1.31530166e-01
-1.29721612e-01 7.68499196e-01 5.88263929e-01 9.94053006e-01
5.65021574e-01 2.80637562e-01 3.07755500e-01 -6.20045900e-01
-3.96851957e-01 -1.50648880e+00 3.21704805e-01 6.93241775e-01
-1.07578583e-01 -6.35592222e-01 1.43726030e-02 5.48873961e-01] | [12.7999267578125, 7.93665885925293] |
7d7ba734-72e7-447e-bc88-5c5d078c74ab | probing-neural-representations-of-scene | 2303.06367 | null | https://arxiv.org/abs/2303.06367v1 | https://arxiv.org/pdf/2303.06367v1.pdf | Probing neural representations of scene perception in a hippocampally dependent task using artificial neural networks | Deep artificial neural networks (DNNs) trained through backpropagation provide effective models of the mammalian visual system, accurately capturing the hierarchy of neural responses through primary visual cortex to inferior temporal cortex (IT). However, the ability of these networks to explain representations in higher cortical areas is relatively lacking and considerably less well researched. For example, DNNs have been less successful as a model of the egocentric to allocentric transformation embodied by circuits in retrosplenial and posterior parietal cortex. We describe a novel scene perception benchmark inspired by a hippocampal dependent task, designed to probe the ability of DNNs to transform scenes viewed from different egocentric perspectives. Using a network architecture inspired by the connectivity between temporal lobe structures and the hippocampus, we demonstrate that DNNs trained using a triplet loss can learn this task. Moreover, by enforcing a factorized latent space, we can split information propagation into "what" and "where" pathways, which we use to reconstruct the input. This allows us to beat the state-of-the-art for unsupervised object segmentation on the CATER and MOVi-A,B,C benchmarks. | ['Caswell Barry', 'Christian F. Doeller', 'Markus Frey'] | 2023-03-11 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Frey_Probing_Neural_Representations_of_Scene_Perception_in_a_Hippocampally_Dependent_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Frey_Probing_Neural_Representations_of_Scene_Perception_in_a_Hippocampally_Dependent_CVPR_2023_paper.pdf | cvpr-2023-1 | ['unsupervised-object-segmentation'] | ['computer-vision'] | [ 3.57005805e-01 2.80555338e-01 2.12421566e-01 -4.61026490e-01
1.43512636e-01 -5.15412688e-01 9.85013306e-01 -2.27152511e-01
-6.10630035e-01 3.18457544e-01 3.09346080e-01 -5.40537946e-02
-1.22667082e-01 -7.88007200e-01 -9.55097377e-01 -7.42630363e-01
2.03344017e-01 6.11866295e-01 1.02709040e-01 -1.51505426e-01
2.91024953e-01 6.38706446e-01 -1.39200878e+00 7.15023816e-01
5.16641378e-01 8.66221189e-01 3.68959159e-01 3.32467914e-01
1.91720307e-01 8.70680273e-01 -5.04446886e-02 -1.36473790e-01
2.59793192e-01 -7.76991129e-01 -1.14514959e+00 -3.04373205e-02
4.05845702e-01 -2.17158824e-01 -5.12355626e-01 8.96397233e-01
1.35652721e-01 5.00716902e-02 9.02069390e-01 -9.87146556e-01
-7.90148675e-01 5.47573924e-01 -2.90817499e-01 4.42264885e-01
1.05855167e-01 1.64514944e-01 1.16821992e+00 -7.69761026e-01
1.01636839e+00 1.37408280e+00 4.14973259e-01 5.82068980e-01
-1.80028617e+00 -3.88274193e-01 2.59152770e-01 3.34507257e-01
-8.82135153e-01 -3.72575432e-01 6.06948316e-01 -5.95595539e-01
1.22158182e+00 -8.11003149e-02 1.11300743e+00 1.21538413e+00
5.57412326e-01 8.80435884e-01 1.36809456e+00 -1.61361620e-01
9.50456411e-02 1.79215502e-02 4.91601452e-02 5.10383070e-01
-1.17799006e-02 3.76980394e-01 -5.69828212e-01 3.85633856e-01
9.49328303e-01 1.59646079e-01 -3.20559084e-01 -5.32538652e-01
-1.36716568e+00 9.09723997e-01 1.20900965e+00 3.74981076e-01
-4.65581805e-01 2.42772534e-01 1.54758647e-01 1.27399042e-01
1.34849578e-01 7.06083179e-01 -3.74153167e-01 5.68924904e-01
-1.02005434e+00 2.53857940e-01 4.41176385e-01 5.56240439e-01
6.75266743e-01 1.09017052e-01 -2.36091122e-01 5.57800412e-01
5.73465049e-01 -1.38994427e-02 4.30330068e-01 -1.23082602e+00
1.42113417e-01 5.35768032e-01 -2.77831733e-01 -6.58819973e-01
-7.94499874e-01 -6.37422621e-01 -8.06399226e-01 3.85225087e-01
5.72573662e-01 3.93021911e-01 -1.04042912e+00 1.93907785e+00
-2.71140158e-01 -2.99610317e-01 -4.23891693e-02 1.07899523e+00
4.52446699e-01 3.00248623e-01 1.93577141e-01 5.51107377e-02
1.22481930e+00 -9.10276294e-01 -2.30408400e-01 -8.28761399e-01
4.85060960e-01 -4.11603332e-01 9.41314220e-01 2.79155821e-01
-1.50333989e+00 -5.58277071e-01 -9.57222700e-01 -4.95433956e-01
-5.56985497e-01 -4.79980484e-02 7.85170436e-01 1.78357005e-01
-1.47357392e+00 6.39060318e-01 -9.85836744e-01 -6.64269209e-01
1.11757994e+00 3.40589464e-01 -4.86681283e-01 2.80133933e-02
-8.37194026e-01 1.14652646e+00 5.78469157e-01 -1.10221421e-02
-1.38054478e+00 -8.80282283e-01 -7.34929681e-01 3.79444987e-01
-3.23518544e-01 -1.30637848e+00 9.07753766e-01 -1.32205403e+00
-1.20889640e+00 1.36874592e+00 -1.10487297e-01 -8.24904978e-01
1.84203476e-01 9.50590521e-02 2.22580522e-01 4.54113364e-01
1.29808679e-01 1.53167713e+00 7.07325995e-01 -1.30447078e+00
-2.48035178e-01 -8.18007410e-01 -1.69164233e-03 1.25765860e-01
2.27590397e-01 3.46949771e-02 1.80382356e-02 -7.49741197e-01
5.75137258e-01 -1.06842542e+00 -1.26381278e-01 1.84464246e-01
-3.57273012e-01 -9.10418704e-02 4.53137070e-01 -8.35412920e-01
4.83624220e-01 -2.19590235e+00 6.43882513e-01 -7.34914169e-02
3.70267153e-01 -1.26148954e-01 -2.36717865e-01 1.03269726e-01
-5.89031637e-01 1.29255027e-01 -7.11512029e-01 -3.37137848e-01
2.23418772e-01 3.45731825e-01 -6.78783238e-01 5.44045210e-01
3.21719170e-01 1.23764205e+00 -5.56226850e-01 5.93107715e-02
2.55887471e-02 5.29990256e-01 -1.02389371e+00 -3.26378755e-02
-4.00546163e-01 8.23646963e-01 1.76410347e-01 1.90311521e-01
5.15722275e-01 -2.92730659e-01 -1.65899824e-02 -1.49941102e-01
-1.67464703e-01 4.24424917e-01 -3.57535869e-01 2.09773278e+00
-1.91553459e-01 1.03675675e+00 -8.02914277e-02 -1.43068290e+00
4.27895635e-01 1.51289269e-01 2.57458299e-01 -1.28322053e+00
3.52794230e-01 1.20602809e-02 4.55888122e-01 -1.35501906e-01
-1.98073089e-01 -3.72272074e-01 1.39317915e-01 5.36333144e-01
6.80699468e-01 -1.77795753e-01 1.60565302e-01 3.23645115e-01
8.19720447e-01 3.53014112e-01 1.88523307e-02 -3.68933916e-01
1.43999606e-01 -8.92538503e-02 3.60413820e-01 6.95670903e-01
-1.63688183e-01 8.61959219e-01 7.51207709e-01 -4.03441101e-01
-1.16012442e+00 -1.37740731e+00 -2.52158016e-01 1.06147432e+00
-2.05489010e-01 2.34939590e-01 -9.26701725e-01 -2.36967504e-01
-2.68268675e-01 1.01665652e+00 -1.04265988e+00 -6.05178356e-01
-4.18340176e-01 -7.55701780e-01 6.46617889e-01 5.52655935e-01
7.48916030e-01 -1.60026002e+00 -9.67511237e-01 1.26093611e-01
3.38365100e-02 -1.13151336e+00 1.01945922e-01 4.29090291e-01
-8.90813172e-01 -1.06135535e+00 -8.19047689e-01 -9.41021681e-01
7.31466651e-01 5.33364117e-02 1.25156271e+00 -3.21294695e-01
-5.96786797e-01 5.10480940e-01 1.51023179e-01 -2.96619594e-01
-9.48166773e-02 9.08637643e-02 -1.87220588e-01 -3.52243371e-02
2.49227077e-01 -8.10235977e-01 -8.16752791e-01 2.51294017e-01
-1.19951975e+00 4.94285345e-01 5.14145434e-01 7.59980798e-01
5.05600810e-01 -4.18495357e-01 2.40214124e-01 -7.91271567e-01
1.07624665e-01 -6.71723247e-01 -4.31603014e-01 -7.22956955e-02
-4.54905063e-01 3.40085328e-01 6.08278334e-01 2.46482342e-02
-1.28538930e+00 -2.22874321e-02 -1.20794706e-01 -5.85137345e-02
-5.11313617e-01 3.24905992e-01 -7.28255287e-02 -1.88254654e-01
6.43636048e-01 4.17915046e-01 -2.57759362e-01 -3.41984570e-01
4.20172513e-01 -2.65471697e-01 6.73863709e-01 -2.04354942e-01
5.43708205e-01 1.09100568e+00 7.13476837e-02 -5.38209677e-01
-1.09145427e+00 -2.36225888e-01 -9.75303829e-01 6.58798069e-02
1.64081168e+00 -6.63720608e-01 -6.88892961e-01 3.77421260e-01
-1.49355221e+00 -5.88664711e-01 -4.76115167e-01 7.01311111e-01
-1.13058066e+00 -1.91350147e-01 -6.31512523e-01 -8.54472890e-02
1.52335539e-01 -9.91582990e-01 7.58491874e-01 2.38802120e-01
-1.13739423e-01 -1.10968447e+00 1.92494988e-01 3.58516216e-01
3.53749961e-01 5.45705110e-02 1.42240572e+00 -4.21317488e-01
-7.75909722e-01 3.55934173e-01 -6.11145198e-01 1.23974979e-01
-3.89333695e-01 -3.13653916e-01 -1.31194365e+00 -1.46538600e-01
2.18704090e-01 -3.82869691e-01 1.85803139e+00 5.94123423e-01
1.15544081e+00 -3.81915629e-01 -2.90645123e-01 1.15513623e+00
1.36951017e+00 3.15877125e-02 8.94636929e-01 3.54467571e-01
5.07585347e-01 1.03243911e+00 -2.02564225e-01 -6.22479953e-02
4.11838710e-01 3.58101785e-01 6.70868039e-01 -1.93138942e-01
-3.90815079e-01 -1.10783793e-01 2.34359443e-01 3.55744541e-01
-1.73941746e-01 -3.37592652e-03 -9.75459099e-01 7.28595257e-01
-1.62611175e+00 -1.18738914e+00 1.83737114e-01 1.86761284e+00
7.18544245e-01 1.86562419e-01 -2.80071557e-01 -1.58168543e-02
5.73071130e-02 1.97430119e-01 -7.42581129e-01 -4.77150679e-01
-3.87169659e-01 3.97133172e-01 2.26416558e-01 2.95211613e-01
-8.96180034e-01 1.18420875e+00 6.49664927e+00 1.87117413e-01
-1.11108279e+00 1.68314248e-01 5.89705825e-01 -3.32129910e-03
-3.81283432e-01 1.98876083e-01 -5.07699907e-01 -4.08135690e-02
9.32829916e-01 1.88557386e-01 9.05455649e-01 4.13720757e-01
2.57547013e-02 -1.31774917e-01 -1.42472041e+00 7.74699092e-01
2.32294261e-01 -1.46320033e+00 3.04809839e-01 8.75797197e-02
8.25496078e-01 4.03219461e-01 5.55190146e-01 2.83161730e-01
2.57807791e-01 -1.41911495e+00 8.21576178e-01 7.82366753e-01
3.05273384e-01 -2.91706026e-01 1.21063046e-01 4.19203550e-01
-7.03916907e-01 -1.10357091e-01 -6.38323784e-01 -4.30312268e-02
-3.28198150e-02 5.80289364e-02 -5.34151435e-01 -1.78465098e-01
8.02413166e-01 8.88604879e-01 -7.22237289e-01 1.06479979e+00
-3.02839279e-01 5.28931141e-01 -4.03521806e-02 5.26517689e-01
4.00227934e-01 -1.71252474e-01 6.25624657e-01 9.64346111e-01
7.09236860e-02 7.03104287e-02 -4.00377035e-01 1.66287804e+00
-3.05717230e-01 -8.72906819e-02 -8.01206946e-01 6.09906279e-02
-2.95788020e-01 1.15840280e+00 -1.02987885e+00 3.17268670e-02
-5.21708839e-02 1.20126307e+00 5.77710986e-01 9.41801429e-01
-7.05171227e-01 8.39902088e-02 5.02963722e-01 1.31509840e-01
5.04471660e-01 -2.75374979e-01 -4.85146075e-01 -1.05364430e+00
-4.07070875e-01 -4.91073579e-01 1.36364400e-01 -1.23012424e+00
-9.87461388e-01 5.46604633e-01 -1.65417477e-01 -6.21840239e-01
-2.85150200e-01 -9.57003176e-01 -5.28319538e-01 7.94935167e-01
-1.62054348e+00 -1.25324392e+00 -2.14147910e-01 7.51867533e-01
4.41829115e-01 -3.38336937e-02 9.44073260e-01 2.91151516e-02
-3.38193566e-01 1.75582260e-01 -1.31531700e-01 1.91904530e-01
4.24931765e-01 -1.26245940e+00 2.90190160e-01 7.02974558e-01
6.13956869e-01 6.62507057e-01 5.36969781e-01 -1.14167430e-01
-8.92483830e-01 -1.03049934e+00 7.41360962e-01 -4.88821149e-01
3.85646969e-01 -5.83275795e-01 -9.00350094e-01 1.12512243e+00
4.32266563e-01 1.96151361e-01 6.06397271e-01 -1.92620426e-01
-7.49128878e-01 7.80948699e-02 -1.00048542e+00 6.37561142e-01
1.14613688e+00 -7.59940982e-01 -8.33383977e-01 3.37324023e-01
5.80987751e-01 1.59615844e-01 -5.72495282e-01 8.52655545e-02
4.96828616e-01 -1.34343040e+00 1.17210579e+00 -9.57011580e-01
7.35777974e-01 -2.27208342e-02 -2.13475913e-01 -1.52402937e+00
-7.17865646e-01 -9.42775235e-02 2.74654955e-01 6.92650378e-01
4.16989982e-01 -7.47584820e-01 5.84613919e-01 1.62742883e-01
-3.08792442e-01 -4.38062876e-01 -9.41619277e-01 -2.40898907e-01
3.93425167e-01 -2.84491956e-01 -8.99347514e-02 8.50041330e-01
-3.19215775e-01 4.90354151e-01 2.47026220e-01 -1.61310285e-01
6.41018867e-01 8.27901531e-03 4.05107886e-02 -1.30787778e+00
-2.85922050e-01 -8.35988343e-01 -3.29809695e-01 -8.62127423e-01
5.01845598e-01 -1.50063038e+00 -1.31787121e-01 -1.96090496e+00
4.32019562e-01 6.88787624e-02 -5.01432836e-01 6.51741326e-01
4.49342310e-01 6.24669731e-01 4.23217386e-01 2.91878879e-01
-4.65963870e-01 5.08995891e-01 1.31060982e+00 -2.00377062e-01
1.68908358e-01 -2.54604697e-01 -6.62200093e-01 1.15403938e+00
8.19781601e-01 -6.01376653e-01 -2.87509978e-01 -8.16064060e-01
5.73136687e-01 -1.64518073e-01 9.64551985e-01 -1.13862669e+00
3.13184470e-01 1.38269573e-01 8.77902210e-01 -2.35696822e-01
2.45614126e-01 -7.54905343e-01 -2.40045398e-01 5.39653480e-01
-5.90266526e-01 -2.64665693e-01 3.15021157e-01 4.12401438e-01
-3.23200613e-01 1.92069896e-02 1.02332222e+00 -4.65720534e-01
-6.49498403e-01 3.86793554e-01 -6.22192085e-01 2.70993084e-01
7.12992728e-01 -3.14683676e-01 -4.61706340e-01 -6.98981062e-02
-1.03975606e+00 -1.06904969e-01 3.28856915e-01 2.14902043e-01
6.63161457e-01 -1.15107250e+00 -7.62723804e-01 5.06209373e-01
-4.31854688e-02 -2.73885518e-01 2.69794941e-01 1.16001952e+00
-5.84168851e-01 8.83150339e-01 -9.92894650e-01 -6.51859462e-01
-5.28583109e-01 5.83858788e-01 7.31318295e-01 -1.53968081e-01
-3.15968692e-01 1.16054201e+00 9.75767672e-01 -5.24186254e-01
-2.80316733e-02 -4.26609367e-01 -4.89473879e-01 1.95247531e-01
1.02328509e-01 8.34133029e-02 -1.91774502e-01 -6.76464796e-01
-2.65305966e-01 4.42447454e-01 1.16746470e-01 -3.03795427e-01
1.47945988e+00 -8.59450269e-03 -4.60356116e-01 4.85172659e-01
1.26338506e+00 -5.06353855e-01 -1.47729611e+00 -7.22583905e-02
-2.45194603e-02 1.21821864e-02 1.35125712e-01 -9.04535294e-01
-1.16824722e+00 1.60542917e+00 5.45019865e-01 1.49361985e-02
1.22505009e+00 1.20585047e-01 4.96625841e-01 2.62664557e-01
2.31796920e-01 -9.17013228e-01 2.18742251e-01 7.51624346e-01
1.24792016e+00 -1.20329750e+00 -3.37934583e-01 1.54324338e-01
-5.78468680e-01 9.43080902e-01 4.26040471e-01 -5.22544682e-01
6.50295258e-01 -1.39361560e-01 -2.61209130e-01 -2.11813375e-01
-8.23218048e-01 -2.56013960e-01 6.03031099e-01 4.55385447e-01
4.30416465e-01 -6.76831082e-02 1.46271111e-02 6.15924537e-01
-2.50323772e-01 -3.06685448e-01 3.17687422e-01 4.23867524e-01
-4.56826091e-01 -6.14623368e-01 -7.82945827e-02 4.37963754e-01
-4.18431401e-01 -2.03431815e-01 -4.35186535e-01 7.45094657e-01
2.91250736e-01 2.83030421e-01 4.50979948e-01 1.56695127e-01
1.07593080e-02 2.21710995e-01 9.33845401e-01 -7.84820974e-01
-5.29828846e-01 -3.47285643e-02 -4.75163698e-01 -7.57511735e-01
-5.16057849e-01 -6.69301748e-01 -1.50697052e+00 5.84553517e-02
3.99456084e-01 -3.26853126e-01 5.03758192e-01 1.10145533e+00
1.47466257e-01 8.00128043e-01 2.37136558e-01 -1.07877684e+00
-1.55066565e-01 -8.61170292e-01 -5.59690416e-01 5.97468555e-01
2.85651863e-01 -5.82003474e-01 -2.95048118e-01 3.21620524e-01] | [10.10061264038086, 2.538435220718384] |
35c69435-0ad2-4b3f-92ea-f8e1cbc0fc18 | ldsa-learning-dynamic-subtask-assignment-in | 2205.02561 | null | https://arxiv.org/abs/2205.02561v3 | https://arxiv.org/pdf/2205.02561v3.pdf | LDSA: Learning Dynamic Subtask Assignment in Cooperative Multi-Agent Reinforcement Learning | Cooperative multi-agent reinforcement learning (MARL) has made prominent progress in recent years. For training efficiency and scalability, most of the MARL algorithms make all agents share the same policy or value network. However, in many complex multi-agent tasks, different agents are expected to possess specific abilities to handle different subtasks. In those scenarios, sharing parameters indiscriminately may lead to similar behavior across all agents, which will limit the exploration efficiency and degrade the final performance. To balance the training complexity and the diversity of agent behavior, we propose a novel framework to learn dynamic subtask assignment (LDSA) in cooperative MARL. Specifically, we first introduce a subtask encoder to construct a vector representation for each subtask according to its identity. To reasonably assign agents to different subtasks, we propose an ability-based subtask selection strategy, which can dynamically group agents with similar abilities into the same subtask. In this way, agents dealing with the same subtask share their learning of specific abilities and different subtasks correspond to different specific abilities. We further introduce two regularizers to increase the representation difference between subtasks and stabilize the training by discouraging agents from frequently changing subtasks, respectively. Empirical results show that LDSA learns reasonable and effective subtask assignment for better collaboration and significantly improves the learning performance on the challenging StarCraft II micromanagement benchmark and Google Research Football. | ['Houqiang Li', 'Jiangcheng Zhu', 'Wengang Zhou', 'Xunhan Hu', 'Jian Zhao', 'Mingyu Yang'] | 2022-05-05 | null | null | null | null | ['starcraft-ii'] | ['playing-games'] | [-2.21533895e-01 -1.44003153e-01 -3.64218354e-01 -9.26294029e-02
-2.36295715e-01 -6.37497008e-01 4.17841256e-01 -7.10124895e-02
-5.19127846e-01 9.73005652e-01 2.18079150e-01 1.00891767e-02
-3.68413985e-01 -5.31918824e-01 -4.94164884e-01 -1.04991663e+00
-9.52638090e-02 5.70748091e-01 4.59112704e-01 -5.05417883e-01
1.29121775e-03 1.91145971e-01 -1.35226381e+00 6.14084974e-02
1.27565134e+00 7.89937377e-01 6.68581784e-01 2.26338923e-01
4.96680774e-02 1.30469120e+00 -9.06061709e-01 -1.77506626e-01
4.57034260e-01 -3.01686794e-01 -8.31264079e-01 1.72604397e-01
2.28481833e-02 -1.72241196e-01 -4.69198912e-01 1.01177585e+00
4.69824702e-01 6.26834929e-01 4.36429203e-01 -1.61224246e+00
-5.12970805e-01 1.00018847e+00 -6.45742893e-01 2.31839746e-01
1.87278360e-01 5.03781796e-01 9.43857789e-01 -2.12995216e-01
1.56491280e-01 1.14057517e+00 2.04328224e-01 7.36275136e-01
-9.84449387e-01 -8.20751607e-01 8.68657351e-01 4.63483393e-01
-1.14283454e+00 -1.18996762e-01 7.50664055e-01 -2.80895501e-01
6.54588461e-01 6.43842518e-02 8.58266115e-01 9.68093812e-01
2.18274564e-01 1.00931180e+00 9.48545098e-01 5.60075492e-02
3.75776052e-01 -3.77797708e-02 -2.14418918e-02 8.01096857e-01
3.02212238e-01 8.61180946e-02 -3.97897780e-01 -1.30644128e-01
8.85146260e-01 5.64698130e-02 -9.83425304e-02 -9.05188978e-01
-1.39689898e+00 8.31106126e-01 4.45701241e-01 1.04904279e-01
-5.90667486e-01 8.44841003e-02 6.15119159e-01 6.91850841e-01
-2.08635896e-01 7.63460636e-01 -4.96878088e-01 -4.17712051e-03
-8.65951926e-02 4.55978662e-01 6.08139753e-01 6.93812966e-01
9.48402345e-01 2.35992461e-01 -4.45362121e-01 9.69704449e-01
1.80961505e-01 2.89999068e-01 9.20439899e-01 -1.04990518e+00
7.22944081e-01 9.44966912e-01 1.89774692e-01 -8.05773914e-01
-5.09056270e-01 -6.39997482e-01 -7.95506120e-01 4.90771413e-01
2.70701557e-01 -6.32236540e-01 -6.70498013e-01 1.86056340e+00
4.07425404e-01 4.00359660e-01 4.94434416e-01 1.12119818e+00
5.77367127e-01 5.15333891e-01 2.03847557e-01 -2.85910517e-01
1.11612630e+00 -1.27112782e+00 -3.69144261e-01 -3.67449343e-01
6.40526772e-01 -3.54507297e-01 8.95300686e-01 3.95840377e-01
-9.30126309e-01 -6.14373982e-01 -1.08574629e+00 7.07405627e-01
5.44014089e-02 -2.11050939e-02 8.10153008e-01 3.58818322e-01
-5.37667632e-01 1.77624211e-01 -7.99830019e-01 2.95954943e-01
3.72737885e-01 6.04052246e-01 -1.11450016e-01 2.69711941e-01
-1.26963305e+00 7.79804766e-01 7.39865780e-01 -2.41397932e-01
-1.45664418e+00 -4.00531590e-01 -7.15187192e-01 1.20132498e-01
8.54145885e-01 -6.20837748e-01 1.38844824e+00 -1.23841190e+00
-1.82559991e+00 1.65847301e-01 4.24299330e-01 -4.65117753e-01
3.17140788e-01 -3.48642245e-02 -2.95027405e-01 -2.42977023e-01
2.25330070e-01 4.70753163e-01 1.00212097e+00 -1.27942944e+00
-1.14920914e+00 -1.40548989e-01 6.53312683e-01 1.04987073e+00
-5.79494298e-01 -2.42393464e-01 -2.33995020e-01 -7.43169367e-01
-2.19310507e-01 -1.02416408e+00 -5.71536303e-01 -7.11554229e-01
-1.82263479e-01 -5.83765686e-01 6.81008637e-01 1.77539941e-02
1.02701080e+00 -1.91129911e+00 8.69755924e-01 1.30924180e-01
3.40049595e-01 2.59960234e-01 -4.10537720e-01 1.35745078e-01
2.51117080e-01 -4.72418159e-01 1.36428058e-01 3.90608824e-04
4.23209481e-02 5.02095640e-01 -6.47777542e-02 2.18591720e-01
-2.21528217e-01 7.81244934e-01 -1.13176584e+00 -4.40237015e-01
1.49507552e-01 -2.08185911e-01 -6.30767584e-01 4.96559292e-01
-3.55802447e-01 6.55924678e-01 -9.74778116e-01 4.60652798e-01
4.95557159e-01 -2.10899785e-01 3.65574569e-01 5.84685057e-02
9.31803063e-02 -7.72715285e-02 -1.50417709e+00 1.66673541e+00
-6.77200615e-01 1.66209951e-01 1.31203920e-01 -1.20058155e+00
9.46258843e-01 2.06744134e-01 6.07960761e-01 -8.49292517e-01
4.42570187e-02 7.78179690e-02 4.94764715e-01 -4.00045902e-01
4.41211820e-01 1.67998284e-01 -2.20628217e-01 5.77692151e-01
-4.41485308e-02 -2.45450772e-02 3.06710631e-01 -2.16407366e-02
8.84417534e-01 -8.51057097e-02 3.44540954e-01 -7.04437345e-02
7.72168159e-01 2.81676725e-02 9.06780422e-01 9.87151444e-01
-4.20715213e-01 -5.00063226e-02 2.82238245e-01 -7.23450780e-01
-7.74311781e-01 -9.14712846e-01 4.99211997e-01 1.72395909e+00
6.46065414e-01 -2.46365070e-02 -4.03888226e-01 -1.10352576e+00
1.31355286e-01 2.88720965e-01 -5.14744341e-01 -4.55358922e-01
-8.88906181e-01 -8.95146608e-01 3.11658263e-01 4.17888582e-01
6.91417038e-01 -1.55012739e+00 -8.77409101e-01 2.95759052e-01
-1.49529648e-03 -7.61622965e-01 -6.33295476e-01 4.31205660e-01
-5.03008544e-01 -1.15164423e+00 -8.99978936e-01 -1.18251824e+00
6.06042683e-01 3.77317220e-01 9.08705831e-01 6.81807920e-02
2.23480240e-01 9.17811394e-02 -5.94233632e-01 -1.84784442e-01
-3.33298802e-01 4.42285657e-01 4.39526647e-01 3.51522602e-02
1.60625666e-01 -5.72702944e-01 -5.44539630e-01 6.36196256e-01
-7.36368179e-01 6.42864257e-02 7.86382079e-01 1.01311970e+00
2.18970478e-01 3.61224979e-01 7.29207933e-01 -3.95409524e-01
1.01705205e+00 -5.80432475e-01 -6.51420712e-01 2.97032118e-01
-5.26707172e-01 1.65255979e-01 1.23123813e+00 -9.90429580e-01
-9.23281908e-01 1.37947444e-02 4.15120870e-01 -6.57084346e-01
1.15832664e-01 5.13475299e-01 -2.01138750e-01 -1.21140346e-01
5.67667603e-01 6.49976909e-01 1.37470901e-01 -1.16107523e-01
1.52697772e-01 4.98085797e-01 3.21682036e-01 -8.59575689e-01
7.55527258e-01 5.15449494e-02 -1.87261760e-01 -2.00374812e-01
-7.17641473e-01 -1.32298976e-01 -6.08557053e-02 -3.15946519e-01
4.71281558e-01 -1.04986227e+00 -1.20480967e+00 6.60205007e-01
-7.55059302e-01 -7.85169244e-01 -3.27142954e-01 5.17066538e-01
-6.76297247e-01 4.03155863e-01 -3.45803231e-01 -4.49292779e-01
-8.84370357e-02 -1.62694705e+00 4.69559640e-01 7.01083541e-01
1.13323197e-01 -7.84587502e-01 3.19799125e-01 2.63416559e-01
3.03816378e-01 -2.61507571e-01 7.79762626e-01 -8.01852703e-01
-4.39310163e-01 3.66051942e-01 6.19090460e-02 1.56613544e-01
2.88426220e-01 -6.23352706e-01 -1.90495774e-01 -7.80575871e-01
-2.66427577e-01 -8.16793203e-01 5.33023179e-01 2.58343488e-01
9.69241202e-01 -4.26308781e-01 -3.29783022e-01 4.54247802e-01
9.72108006e-01 5.88838816e-01 2.27030888e-01 7.90407002e-01
7.39688158e-01 5.25576949e-01 8.24452519e-01 6.69424832e-01
4.99037683e-01 8.86912346e-01 7.26285815e-01 3.85545075e-01
1.12466171e-01 -2.22882450e-01 5.86166203e-01 6.62873149e-01
-2.10795641e-01 -2.47753665e-01 -5.89872718e-01 4.29710299e-01
-2.35124278e+00 -9.42662418e-01 4.70695466e-01 2.11887527e+00
8.69676769e-01 1.16194926e-01 4.13595527e-01 -3.18565369e-01
6.18675530e-01 3.53221416e-01 -1.01737142e+00 -7.98431560e-02
-1.11577742e-01 -3.89242411e-01 4.09625679e-01 3.01149964e-01
-1.05841911e+00 1.05191469e+00 5.30562592e+00 1.09015000e+00
-9.70784128e-01 1.05167083e-01 3.88299257e-01 -1.98093355e-01
-1.34305647e-02 -2.12639272e-01 -8.45302999e-01 7.94234097e-01
3.01652402e-01 -3.45152199e-01 7.60332167e-01 9.28909600e-01
9.16712359e-02 7.61462525e-02 -7.72666395e-01 9.58024442e-01
-1.42091691e-01 -1.18989897e+00 1.98448569e-01 -1.57395184e-01
1.06136429e+00 9.62864782e-04 1.22020170e-01 9.42761958e-01
1.08339942e+00 -8.89709294e-01 6.45584702e-01 2.92961419e-01
1.34784281e-01 -1.05783963e+00 6.35717809e-01 6.27398193e-01
-1.35291207e+00 -7.11822391e-01 -5.79752445e-01 -7.26533011e-02
-1.01742305e-01 -1.60245076e-01 -4.38141882e-01 6.35645628e-01
5.74628532e-01 5.26924729e-01 -4.30532664e-01 1.01970410e+00
-2.15161189e-01 2.13110551e-01 3.11718937e-02 -3.52861345e-01
5.42719781e-01 -2.86142558e-01 7.25853622e-01 4.42055643e-01
2.03707186e-03 -4.93725687e-02 1.10499597e+00 2.08716944e-01
1.29469074e-02 1.99800223e-01 -2.39566445e-01 4.82946187e-02
8.41733634e-01 1.09791458e+00 -2.73748249e-01 -3.34466606e-01
-4.87970352e-01 8.68815064e-01 8.11819196e-01 4.12588596e-01
-9.07897413e-01 -3.23754340e-01 8.90039325e-01 -4.27579015e-01
1.67545959e-01 -1.23822488e-01 3.45435664e-02 -1.11057937e+00
-2.60081817e-03 -1.32567859e+00 6.13741338e-01 -3.44442546e-01
-1.25173628e+00 8.01770866e-01 -1.22588530e-01 -1.51986980e+00
-3.94112051e-01 -2.71694452e-01 -7.79308736e-01 6.76979005e-01
-1.40455234e+00 -8.77053738e-01 -4.19511735e-01 7.80872345e-01
8.52559745e-01 -1.09090185e+00 4.73768741e-01 -7.01788589e-02
-8.18790853e-01 7.32370377e-01 2.06122845e-01 2.70995386e-02
4.86974597e-01 -1.14770639e+00 -4.67086174e-02 3.52281809e-01
-4.74506952e-02 4.52802241e-01 6.50331616e-01 -6.20500684e-01
-1.32865894e+00 -1.03172302e+00 -1.41930431e-02 -3.53210680e-02
7.38559604e-01 4.12167870e-02 -8.11123133e-01 6.40151083e-01
2.01665029e-01 -1.97557837e-01 3.73824328e-01 1.91358268e-01
-1.46010267e-02 -2.32289314e-01 -7.47819126e-01 9.70097542e-01
8.75155807e-01 2.06548125e-02 -6.75589383e-01 3.59333545e-01
7.05163717e-01 -3.88883144e-01 -6.83669209e-01 4.05997187e-01
4.27709192e-01 -8.12684357e-01 8.39276671e-01 -8.74274015e-01
1.90197706e-01 -5.54002404e-01 1.43021852e-01 -1.97714090e+00
-7.51769364e-01 -7.26475239e-01 -3.69484156e-01 7.39280164e-01
1.45112917e-01 -6.50819957e-01 1.18801498e+00 2.72915930e-01
-2.85597950e-01 -8.49820137e-01 -9.71185565e-01 -1.05636561e+00
1.20968394e-01 2.04473272e-01 9.24865961e-01 9.17515397e-01
2.35950112e-01 4.89088833e-01 -6.58068895e-01 1.95306838e-01
4.43313092e-01 3.47971469e-01 9.39358234e-01 -1.20552754e+00
-6.37704790e-01 -6.54155731e-01 -2.17257813e-01 -1.36223185e+00
5.22405624e-01 -8.85869622e-01 -7.87524581e-02 -1.21102166e+00
4.16279227e-01 -9.13800895e-01 -5.91461062e-01 7.22444773e-01
-4.58814055e-01 -1.62655056e-01 2.41589040e-01 4.07519162e-01
-1.14641666e+00 7.63029933e-01 1.60698020e+00 -3.13471109e-01
-7.52668440e-01 1.93967581e-01 -7.99208045e-01 5.73392093e-01
9.44219232e-01 -2.39773974e-01 -7.92437971e-01 -4.37705964e-01
1.61513671e-01 1.73742771e-01 1.47805750e-01 -1.13676012e+00
4.11184520e-01 -7.48012066e-01 2.50469148e-01 3.50903943e-02
1.87013984e-01 -8.56707454e-01 6.74627349e-02 8.44284356e-01
-4.94335324e-01 1.15919054e-01 -2.07924331e-03 5.91438234e-01
-7.77203217e-02 -4.10873801e-01 7.40281045e-01 -2.29635075e-01
-1.02294302e+00 5.02486289e-01 -6.11110687e-01 1.00744642e-01
1.54377210e+00 -7.47608691e-02 -3.21194261e-01 -3.54366571e-01
-5.81204236e-01 1.06002069e+00 2.73511320e-01 6.21390283e-01
4.33155924e-01 -1.38627946e+00 -8.95636976e-01 1.18992850e-01
1.25376284e-01 5.70910163e-02 5.48129559e-01 5.46252370e-01
-8.58254880e-02 2.39684984e-01 -5.72158456e-01 -2.60731310e-01
-1.19202876e+00 6.37621284e-01 5.71947038e-01 -6.80860162e-01
-3.22884768e-01 7.93370843e-01 4.83336478e-01 -5.81354022e-01
3.52233142e-01 1.52162358e-01 -7.74068117e-01 -5.23376802e-04
4.02150840e-01 4.77950037e-01 -2.40808174e-01 -4.83253837e-01
-2.15362310e-01 2.91932702e-01 -7.59172320e-01 2.59088695e-01
1.00362742e+00 -5.32688461e-02 3.06814730e-01 8.73025879e-03
5.10129571e-01 -1.88173339e-01 -1.64821911e+00 -4.39184129e-01
-3.60581845e-01 -2.89055645e-01 -3.24871577e-02 -8.29438508e-01
-1.32208049e+00 1.55601010e-01 3.96082252e-01 3.92249614e-01
1.00864542e+00 5.70051782e-02 5.82294643e-01 7.42679656e-01
6.64183915e-01 -1.33192945e+00 6.13355994e-01 7.49268234e-01
7.85455704e-01 -1.16866326e+00 -2.89673209e-02 -1.11012205e-01
-1.26989353e+00 8.87137413e-01 1.28846729e+00 -1.80609092e-01
6.42999634e-02 5.65335117e-02 9.85284373e-02 -1.71508864e-01
-9.34228122e-01 -1.48680106e-01 5.65434247e-02 8.04562867e-01
5.70670422e-03 2.64495581e-01 -2.26226315e-01 6.50549471e-01
-3.26981507e-02 -2.72458792e-01 4.70887214e-01 8.46561849e-01
-6.80733740e-01 -1.18019474e+00 -2.55540937e-01 5.73293567e-01
-8.37895870e-02 2.61561424e-01 -3.24764326e-02 5.27900517e-01
1.32489175e-01 9.21427131e-01 -1.16985561e-02 -4.67870057e-01
3.78083676e-01 -3.32287341e-01 4.33720529e-01 -5.89799702e-01
-8.47830772e-01 -1.83768004e-01 -1.43378675e-01 -4.00507092e-01
-3.13837975e-01 -5.61305821e-01 -1.31900716e+00 -1.73259571e-01
-1.37827009e-01 5.12143612e-01 -1.33929849e-01 8.55118692e-01
3.09837699e-01 8.81209731e-01 1.03683484e+00 -6.78916156e-01
-9.57080722e-01 -8.37763608e-01 -6.19456351e-01 2.40999684e-01
1.98936164e-01 -1.13353407e+00 -1.13912478e-01 -4.72435504e-01] | [3.7860519886016846, 1.9293774366378784] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.