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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
a7c93832-1460-4f43-a4d2-e0f50538ebcc | dip-differentiable-interreflection-aware | 2212.04705 | null | https://arxiv.org/abs/2212.04705v1 | https://arxiv.org/pdf/2212.04705v1.pdf | DIP: Differentiable Interreflection-aware Physics-based Inverse Rendering | We present a physics-based inverse rendering method that learns the illumination, geometry, and materials of a scene from posed multi-view RGB images. To model the illumination of a scene, existing inverse rendering works either completely ignore the indirect illumination or model it by coarse approximations, leading to sub-optimal illumination, geometry, and material prediction of the scene. In this work, we propose a physics-based illumination model that explicitly traces the incoming indirect lights at each surface point based on interreflection, followed by estimating each identified indirect light through an efficient neural network. Furthermore, we utilize the Leibniz's integral rule to resolve non-differentiability in the proposed illumination model caused by one type of environment light -- the tangent lights. As a result, the proposed interreflection-aware illumination model can be learned end-to-end together with geometry and materials estimation. As a side product, our physics-based inverse rendering model also facilitates flexible and realistic material editing as well as relighting. Extensive experiments on both synthetic and real-world datasets demonstrate that the proposed method performs favorably against existing inverse rendering methods on novel view synthesis and inverse rendering. | ['Ming-Hsuan Yang', 'Sifei Liu', 'Xueting Li', 'Youming Deng'] | 2022-12-09 | null | null | null | null | ['inverse-rendering'] | ['computer-vision'] | [ 5.98746240e-01 -1.02386773e-01 5.67684233e-01 -5.92500329e-01
-3.02897215e-01 -6.11956358e-01 5.15541911e-01 -3.84817243e-01
1.19119458e-01 5.84927440e-01 1.48621649e-01 6.69410378e-02
3.89706306e-02 -9.28955674e-01 -1.04841292e+00 -7.26410329e-01
6.71498537e-01 3.71611476e-01 -1.17678247e-01 -2.66854554e-01
2.16014579e-01 6.32597327e-01 -1.61696863e+00 2.42542922e-01
1.01506507e+00 1.10970616e+00 2.46905729e-01 6.07713819e-01
-1.03732675e-01 6.07537270e-01 -2.14825213e-01 -2.78771400e-01
5.01060605e-01 -1.38437897e-01 -3.99339855e-01 2.51328081e-01
6.27893090e-01 -7.93146372e-01 -2.66373038e-01 9.30380166e-01
3.33613306e-01 2.46138155e-01 4.50774491e-01 -8.59162807e-01
-7.85459220e-01 -3.63692999e-01 -8.05879533e-01 -5.72108567e-01
5.33213139e-01 1.18610911e-01 7.29498506e-01 -1.01545894e+00
5.20227849e-01 1.19841456e+00 6.09134376e-01 2.16355264e-01
-1.12387693e+00 -6.85610831e-01 4.08923984e-01 4.71232235e-02
-1.21433544e+00 -2.96341777e-01 1.34224510e+00 -2.51903176e-01
4.91308600e-01 4.21298206e-01 1.05087960e+00 7.54673004e-01
2.76484847e-01 3.76070380e-01 1.30699253e+00 -3.62810969e-01
5.42613342e-02 1.99363425e-01 -1.87390953e-01 9.29509878e-01
-1.40837133e-01 3.93875599e-01 -5.20779133e-01 -1.68996155e-01
1.02646255e+00 2.81129420e-01 -6.55440450e-01 -5.73859274e-01
-1.10161006e+00 2.75905073e-01 5.46952426e-01 -5.15640795e-01
-3.37253034e-01 9.59501415e-02 -2.69481470e-03 -2.42081106e-01
7.85621762e-01 1.54228434e-01 -4.60958362e-01 2.97222435e-01
-3.47847223e-01 5.78155406e-02 6.74956322e-01 9.97069359e-01
1.06467998e+00 -1.26800304e-02 1.90646891e-02 8.57351065e-01
6.70270979e-01 7.51763403e-01 -2.89718509e-01 -1.42551351e+00
3.66241723e-01 5.80067754e-01 4.43469733e-01 -9.43875074e-01
-2.75960982e-01 -6.43230617e-01 -8.80358338e-01 4.43476111e-01
1.44795865e-01 1.31901160e-01 -9.59165156e-01 1.59347212e+00
6.65995181e-01 5.05519569e-01 -2.83038467e-01 1.12353277e+00
7.93354809e-01 7.41083026e-01 -4.64565337e-01 -1.05544344e-01
1.12480390e+00 -9.48886454e-01 -6.72046661e-01 3.82120651e-03
-3.75825353e-02 -9.62897837e-01 1.30951834e+00 6.11578166e-01
-1.20569527e+00 -3.78944874e-01 -8.96037638e-01 -5.36873996e-01
2.29626596e-02 1.79648250e-01 5.45086145e-01 2.99717635e-01
-8.19610596e-01 5.99273086e-01 -6.39243245e-01 2.71603405e-01
9.98525470e-02 1.68063134e-01 6.02592826e-02 -3.80357534e-01
-8.50158393e-01 4.32212263e-01 -2.55296677e-01 4.37546045e-01
-5.53289533e-01 -1.17447424e+00 -6.40979111e-01 -8.20040796e-03
3.46171856e-01 -1.17484784e+00 8.61988485e-01 -1.02848136e+00
-2.04365706e+00 6.62526608e-01 -3.25380266e-01 5.17208815e-01
4.86034095e-01 -3.67995501e-01 -1.48548633e-01 8.61415789e-02
-2.34668866e-01 1.95142880e-01 8.67926657e-01 -2.10098886e+00
-3.04172993e-01 -3.80907089e-01 2.77203947e-01 5.89311123e-01
-5.15676588e-02 -5.99379838e-01 -6.03635728e-01 -3.96186233e-01
3.42555344e-01 -7.41339147e-01 -1.99486557e-02 6.29426181e-01
-5.33953428e-01 4.81594682e-01 8.22250307e-01 -7.67806232e-01
5.86739421e-01 -1.89802396e+00 8.50610435e-02 2.43027031e-01
3.05969894e-01 -7.83528686e-02 -2.23634139e-01 2.30826169e-01
1.22978901e-02 -3.33548278e-01 -1.63697630e-01 -6.36118710e-01
-1.30823836e-01 2.19257042e-01 -4.89427179e-01 5.14578462e-01
-2.55132079e-01 6.07299924e-01 -9.17759776e-01 -2.17651576e-01
6.62005723e-01 1.33723462e+00 -7.59715617e-01 4.81818140e-01
-2.68464893e-01 9.77925003e-01 -4.21340704e-01 4.52277184e-01
1.06721485e+00 -1.65847808e-01 -1.83075145e-01 -8.09472322e-01
-3.70450705e-01 1.94805115e-01 -1.08713305e+00 1.98700249e+00
-1.11893153e+00 3.42717856e-01 3.28531206e-01 -4.17972684e-01
8.12874496e-01 5.27645908e-02 4.80637014e-01 -8.40828538e-01
9.05632228e-02 -3.71304303e-02 -6.05670631e-01 -2.11889192e-01
5.20552337e-01 -1.27996191e-01 6.61521077e-01 4.20554310e-01
-5.32968462e-01 -4.33352977e-01 -4.47669953e-01 -9.38845053e-02
5.64064205e-01 7.93445706e-01 -2.08057195e-01 3.74310501e-02
7.01057196e-01 -4.12958413e-01 6.58578992e-01 2.98195481e-01
4.92958665e-01 9.60296929e-01 -7.95654356e-02 -5.87760508e-01
-1.00025606e+00 -1.43373704e+00 -7.77021423e-03 8.87893438e-01
6.00995243e-01 8.34968910e-02 -7.42735386e-01 -4.45576534e-02
-2.18631491e-01 8.36563826e-01 -4.42601949e-01 -6.26678690e-02
-6.97170675e-01 -5.14051080e-01 -2.95200974e-01 1.54074907e-01
7.31633604e-01 -6.92715764e-01 -4.70058143e-01 -7.11832866e-02
-3.54898751e-01 -1.21951854e+00 -6.61724031e-01 -3.14346522e-01
-7.75031388e-01 -1.02504361e+00 -5.58002353e-01 -3.46537858e-01
9.03579772e-01 6.06112301e-01 1.23162234e+00 1.26866058e-01
-3.31830412e-01 6.04132533e-01 4.66308333e-02 -2.36751080e-01
-1.77053779e-01 -4.86523420e-01 -3.59327674e-01 5.40183902e-01
-3.89321774e-01 -8.14204574e-01 -1.16737759e+00 4.14791882e-01
-7.69277215e-01 8.44828427e-01 1.86362535e-01 6.15686476e-01
1.01991880e+00 -1.66571274e-01 -2.31780902e-01 -8.43329132e-01
1.97693780e-01 -1.47397950e-01 -9.26432014e-01 4.62379605e-01
-5.24579823e-01 -1.45809650e-01 8.56839895e-01 -3.06388229e-01
-1.88630319e+00 2.39560127e-01 -4.24905457e-02 -6.81907773e-01
-1.25461340e-01 -1.20655097e-01 -4.12350506e-01 -3.71355474e-01
2.30392486e-01 2.96799481e-01 -2.97465801e-01 -6.09396756e-01
3.63549352e-01 1.91791669e-01 5.31651974e-01 -7.40877211e-01
1.01090991e+00 1.12750947e+00 3.67092729e-01 -9.25952613e-01
-1.14598048e+00 -2.11282060e-01 -5.48287630e-01 -3.32530111e-01
7.65170157e-01 -1.01293862e+00 -1.08730674e+00 4.97736037e-01
-1.39440584e+00 -4.64510918e-01 -3.54446799e-01 2.98513502e-01
-6.08083546e-01 3.59178364e-01 -5.20577550e-01 -7.26161718e-01
-3.52971435e-01 -1.23449814e+00 1.50500441e+00 2.25454912e-01
2.66487867e-01 -1.10157371e+00 7.01247528e-02 5.49672961e-01
3.19815338e-01 3.24286163e-01 8.76369834e-01 7.29331136e-01
-1.19722462e+00 9.98457745e-02 -5.93878686e-01 3.03882688e-01
2.85432994e-01 1.77246392e-01 -1.26191521e+00 3.24185602e-02
2.40768850e-01 2.77030542e-02 5.36318123e-01 4.02403951e-01
1.45640457e+00 -1.86869144e-01 -4.63665510e-03 1.52213371e+00
1.68689871e+00 3.22258882e-02 5.16543567e-01 8.59471112e-02
1.29702294e+00 6.39846742e-01 5.02751410e-01 7.36930788e-01
6.95547819e-01 6.90654457e-01 8.55013847e-01 -5.37098229e-01
-3.15397292e-01 -3.97175968e-01 4.96384427e-02 8.65260065e-01
-2.55837858e-01 -3.64664227e-01 -4.34799224e-01 -1.47954464e-01
-1.40323734e+00 -5.12545168e-01 -3.89285475e-01 2.30736566e+00
5.44722140e-01 -4.04257566e-01 -4.53567535e-01 -2.50888199e-01
4.35235083e-01 9.62477475e-02 -7.63998985e-01 -3.34417582e-01
-5.01792356e-02 1.25246316e-01 5.19144177e-01 8.53320420e-01
-5.17488241e-01 6.57675385e-01 5.53762293e+00 4.31408048e-01
-1.29327416e+00 7.09110796e-02 5.77360213e-01 -1.19491369e-01
-1.12609863e+00 5.46016730e-02 -4.54776794e-01 2.43174449e-01
1.63151339e-01 1.19450361e-01 1.08089900e+00 3.48919690e-01
4.29992080e-01 -1.17074430e-01 -9.01420414e-01 9.96468663e-01
5.35249822e-02 -1.14600325e+00 2.59418309e-01 7.65256211e-02
7.96147525e-01 -6.49730265e-02 2.25643858e-01 -3.35169882e-01
9.52585936e-02 -5.80465436e-01 8.87407899e-01 9.12748039e-01
9.59828377e-01 -6.83709264e-01 1.15350813e-01 3.00641716e-01
-1.18024254e+00 1.39749303e-01 -2.40761191e-01 6.15192503e-02
4.11461532e-01 7.46602416e-01 -2.41299838e-01 5.80160499e-01
4.89147305e-01 7.26918578e-01 -1.39836326e-01 6.84120357e-01
-4.07165408e-01 1.78214520e-01 -3.26135039e-01 4.46302384e-01
-2.11834341e-01 -8.85796607e-01 5.18781841e-01 7.08421469e-01
3.06707948e-01 4.36949581e-01 1.16053477e-01 1.41796875e+00
-1.34108961e-01 -7.46337175e-02 -2.79884964e-01 5.20437539e-01
1.36600032e-01 1.44147766e+00 -4.86115873e-01 -4.85358387e-02
-4.23045844e-01 1.21297216e+00 2.72972345e-01 1.02713192e+00
-8.97971034e-01 -7.50655010e-02 6.18571639e-01 3.09236318e-01
-1.88734382e-01 -9.08731669e-02 -5.44842422e-01 -1.30878472e+00
2.28428215e-01 -3.78385305e-01 -3.85729194e-01 -1.39722168e+00
-1.05216801e+00 5.38090646e-01 -8.79191607e-02 -1.29067910e+00
2.44346127e-01 -6.33308589e-01 -5.46939552e-01 1.09202766e+00
-1.90773427e+00 -1.39095640e+00 -7.43345618e-01 5.88105261e-01
3.93061787e-01 4.81342465e-01 7.52094805e-01 2.33409807e-01
-3.13943177e-01 4.23315577e-02 2.19087675e-01 -3.32917720e-01
7.10418701e-01 -9.91894901e-01 2.58615732e-01 4.85570252e-01
-3.43341589e-01 5.61393678e-01 5.23565710e-01 -4.64625478e-01
-1.80082726e+00 -1.06400967e+00 4.93836030e-02 -1.60579070e-01
1.66157618e-01 -3.82719785e-01 -9.06032860e-01 5.39851487e-01
1.18345432e-01 3.49174052e-01 3.20653975e-01 -1.43501669e-01
-3.65568221e-01 -4.52702880e-01 -1.20042205e+00 5.61904132e-01
1.30827892e+00 -5.90886652e-01 7.94983730e-02 4.84529465e-01
7.45280385e-01 -8.96304369e-01 -7.42444754e-01 3.80434543e-01
8.48814726e-01 -1.36452246e+00 1.38171327e+00 7.70751908e-02
7.32626140e-01 -3.50182325e-01 -1.06200799e-01 -1.39104164e+00
-2.54779965e-01 -5.87243021e-01 -1.79937094e-01 1.08151329e+00
-4.75383103e-02 -8.58647466e-01 6.21825576e-01 8.53290856e-01
-4.70826656e-01 -8.59608769e-01 -5.53696394e-01 -3.76278162e-01
-4.04539496e-01 -3.41727108e-01 7.86178291e-01 8.54631424e-01
-9.36671436e-01 1.10052235e-01 -4.07475352e-01 5.72207987e-01
1.25532508e+00 6.13914132e-01 8.30140829e-01 -1.10178769e+00
-4.97786641e-01 -2.32482497e-02 3.86889398e-01 -1.41684628e+00
2.39388674e-01 -5.41939437e-01 1.90347001e-01 -1.63280666e+00
1.20980844e-01 -6.71770513e-01 -4.48873304e-02 -2.07776017e-02
-1.03238635e-01 2.82404006e-01 -7.12199416e-03 -5.86662721e-03
-1.53468847e-01 7.77069509e-01 1.84414196e+00 9.78676155e-02
-3.01027447e-01 -1.34904701e-02 -4.76109207e-01 1.11562836e+00
5.03429353e-01 -1.96043551e-01 -6.07138038e-01 -9.56059873e-01
5.42557120e-01 1.33920625e-01 6.20825648e-01 -5.47347009e-01
-1.08010083e-01 -4.47649688e-01 5.41453838e-01 -6.34732842e-01
8.70039165e-01 -1.06583166e+00 4.48087871e-01 3.27162780e-02
-1.67517185e-01 -3.15621644e-01 -3.84478793e-02 6.50308967e-01
2.03049287e-01 1.58117712e-01 7.95399547e-01 -1.78973779e-01
-2.70853162e-01 6.10940695e-01 4.03151751e-01 -6.90951720e-02
5.47367156e-01 -4.14576799e-01 -2.47861102e-01 -4.70514625e-01
-2.80140340e-01 4.46561575e-02 9.27725554e-01 1.20749697e-01
8.83642733e-01 -1.23299670e+00 -4.63357598e-01 6.03180528e-01
-5.31288795e-02 3.52143109e-01 5.02288043e-01 7.06242740e-01
-7.79395938e-01 -1.87823743e-01 8.92359763e-02 -6.71354949e-01
-1.17214465e+00 3.06786686e-01 5.29649734e-01 -1.21339858e-02
-9.18514132e-01 6.93125308e-01 1.00015080e+00 -6.97412968e-01
8.87824520e-02 -3.04124117e-01 2.16894433e-01 -5.91199517e-01
3.43232930e-01 5.22819936e-01 -1.55669034e-01 -7.62636781e-01
-5.98574057e-02 1.20040798e+00 3.22338164e-01 -1.34246081e-01
1.46380854e+00 -5.43122709e-01 -2.26438537e-01 4.88809198e-01
1.40010607e+00 2.07797229e-01 -1.65465868e+00 -1.21563233e-01
-1.05723786e+00 -8.31820726e-01 3.79833788e-01 -7.27232218e-01
-1.34788668e+00 1.08414567e+00 4.70955312e-01 -3.05865169e-01
1.44189560e+00 -5.36257625e-01 9.79619503e-01 2.70391345e-01
4.17418391e-01 -1.06731319e+00 -9.49835852e-02 4.29926842e-01
1.07900703e+00 -1.00904691e+00 3.69771332e-01 -9.39583004e-01
-2.33434796e-01 1.16252005e+00 5.73878407e-01 -1.05903380e-01
7.16561735e-01 1.68915763e-01 1.29625246e-01 -1.71597600e-01
-4.35748070e-01 4.04236317e-01 4.87091064e-01 4.15029526e-01
3.76137495e-01 -2.11535109e-04 3.03839892e-01 -1.03098534e-01
-1.16905216e-02 7.13770315e-02 4.94115472e-01 5.93891382e-01
-1.05075434e-01 -8.54236841e-01 -5.03140211e-01 1.85368538e-01
5.90048619e-02 -2.45633405e-02 -1.41153604e-01 4.79831874e-01
2.36018777e-01 5.39593935e-01 6.00383542e-02 -1.66793406e-01
5.50674379e-01 -4.14333075e-01 7.28437185e-01 -5.37558377e-01
-3.29110056e-01 3.47654641e-01 -2.67606437e-01 -9.45874214e-01
-4.61020648e-01 -2.71325976e-01 -1.34525073e+00 -2.41069481e-01
-4.03742820e-01 -4.05049294e-01 9.21306729e-01 7.96218574e-01
4.18310016e-01 5.79139233e-01 9.29107904e-01 -1.31280208e+00
-1.49346277e-01 -3.12691212e-01 -6.21807575e-01 4.80888784e-01
6.01239681e-01 -6.67567849e-01 -5.58699012e-01 3.21787260e-02] | [9.69948959350586, -3.0710480213165283] |
bfaefe29-f4fe-4527-97e7-6a8771201be0 | recallm-an-architecture-for-temporal-context | 2307.02738 | null | https://arxiv.org/abs/2307.02738v2 | https://arxiv.org/pdf/2307.02738v2.pdf | RecallM: An Architecture for Temporal Context Understanding and Question Answering | The ideal long-term memory mechanism for Large Language Model (LLM) based chatbots, would lay the foundation for continual learning, complex reasoning and allow sequential and temporal dependencies to be learnt. Creating this type of memory mechanism is an extremely challenging problem. In this paper we explore different methods of achieving the effect of long-term memory. We propose a new architecture focused on creating adaptable and updatable long-term memory for AGI systems. We demonstrate through various experiments the benefits of the RecallM architecture, particularly the improved temporal understanding of knowledge it provides. | ['Hugo Latapie', 'Brandon Kynoch'] | 2023-07-06 | null | null | null | null | ['continual-learning', 'question-answering'] | ['methodology', 'natural-language-processing'] | [-8.34920228e-01 2.64378071e-01 -3.40205103e-01 -3.34961236e-01
-2.98993975e-01 -5.96146941e-01 9.32867467e-01 -6.56722635e-02
-4.46986794e-01 1.05240417e+00 1.72904283e-01 -4.99100119e-01
-3.20136815e-01 -1.04802835e+00 -4.65356022e-01 -2.28073686e-01
-6.41832650e-01 8.98688078e-01 8.76387119e-01 -5.49971223e-01
5.33855915e-01 5.12080431e-01 -1.04081345e+00 6.21461034e-01
2.13868707e-01 3.82220000e-01 5.69821954e-01 8.23468506e-01
-6.99438572e-01 1.84612536e+00 -7.76192665e-01 -1.32123351e-01
-3.04012895e-01 -2.83492029e-01 -1.77721906e+00 -4.82050240e-01
-4.41080302e-01 -2.02948004e-01 -3.93228620e-01 2.68621027e-01
1.71573013e-01 6.19024694e-01 3.00649852e-01 -1.07444978e+00
-6.52221918e-01 1.12500644e+00 8.93256962e-02 3.18051457e-01
5.50009012e-01 3.99264634e-01 6.29074395e-01 -5.76160312e-01
9.88908410e-01 1.40307570e+00 6.47484124e-01 9.80706930e-01
-7.73503661e-01 -3.25542241e-01 2.55024761e-01 6.40017211e-01
-1.06849682e+00 -3.14178824e-01 3.71181339e-01 -7.17423186e-02
1.77943790e+00 1.05420016e-01 5.88599503e-01 1.08713651e+00
7.90059090e-01 7.31957555e-01 1.09682560e+00 -7.11700439e-01
5.18467790e-03 1.77654237e-01 7.40928233e-01 8.55792880e-01
-2.36580789e-01 2.79834688e-01 -1.09957993e+00 4.11264114e-02
8.45640361e-01 1.10504150e-01 3.75985950e-01 3.19636285e-01
-6.48747325e-01 6.12141907e-01 4.53982025e-01 8.97533238e-01
-3.42138052e-01 6.43992543e-01 3.44911247e-01 1.00478244e+00
1.56997547e-01 5.35606503e-01 -4.94424045e-01 -6.02311850e-01
-5.20926476e-01 -1.55082285e-01 1.31916702e+00 8.48201871e-01
6.79121912e-01 -2.38445655e-01 -2.84134179e-01 5.50983012e-01
4.78927493e-01 2.34094821e-02 6.72782779e-01 -9.18656588e-01
2.13201940e-01 5.72200477e-01 1.94331229e-01 -3.84701669e-01
-7.49275684e-01 -1.89476654e-01 -3.78825366e-01 3.78141850e-01
3.98990184e-01 -1.05050458e-02 -7.29555309e-01 1.74233615e+00
1.79183520e-02 7.77056590e-02 2.08659977e-01 3.15190136e-01
5.75286210e-01 8.46259534e-01 3.38821739e-01 -4.17691141e-01
8.66261601e-01 -1.17887306e+00 -8.53858829e-01 -3.83821517e-01
9.37919915e-01 -4.58875179e-01 1.06379879e+00 2.15719640e-02
-1.19000340e+00 -2.62671888e-01 -7.08589971e-01 -1.06112584e-02
-6.28729105e-01 -7.81704664e-01 9.92640674e-01 3.28145504e-01
-1.34725845e+00 7.07667589e-01 -1.01655805e+00 -5.11276066e-01
-1.46731123e-01 4.38624114e-01 -2.16665491e-01 2.39594892e-01
-1.55528498e+00 1.62992227e+00 5.19124508e-01 2.09321722e-01
-1.15566003e+00 -2.40777791e-01 -3.88578027e-01 5.50655015e-02
3.95088971e-01 -4.14612085e-01 1.51922047e+00 -6.74758792e-01
-1.82770348e+00 8.44425321e-01 2.87634823e-02 -7.52153039e-01
2.86916047e-01 -1.20409854e-01 -1.23801343e-01 -1.64895833e-01
-3.97994876e-01 3.58609676e-01 4.75970060e-01 -8.54495466e-01
-2.09663257e-01 -2.56358296e-01 4.33919817e-01 -1.59692794e-01
-3.07515979e-01 2.94349343e-01 -9.01123434e-02 -1.72492623e-01
-3.07969183e-01 -9.40088034e-01 -4.27072614e-01 -2.75469214e-01
4.34901327e-01 -4.79065388e-01 7.68791199e-01 -4.83898222e-01
1.49431169e+00 -1.73920226e+00 -1.42624080e-01 1.53270736e-01
1.12721376e-01 5.72169185e-01 -3.03015292e-01 1.05563962e+00
3.73889416e-01 1.15140542e-01 5.06625056e-01 -1.02754310e-02
2.47815549e-02 7.00359464e-01 -1.97873265e-01 -3.20716053e-01
-3.87294665e-02 1.30853176e+00 -8.01719129e-01 -5.88145375e-01
4.96995747e-02 2.02130787e-02 -3.05473715e-01 7.31037617e-01
-7.68889844e-01 4.13720995e-01 -4.37508762e-01 3.76336016e-02
6.64300397e-02 -5.20171046e-01 5.84341764e-01 7.41840661e-01
-4.02580261e-01 5.97577870e-01 -8.21881890e-01 1.87472641e+00
-8.87244582e-01 5.94724596e-01 1.39179066e-01 -5.99672198e-01
1.14733779e+00 7.91133106e-01 -1.59727014e-03 -1.03873527e+00
-2.40562931e-02 2.17531137e-02 4.63572443e-02 -6.17947876e-01
4.75795031e-01 -3.43402267e-01 -1.12162903e-01 1.13695431e+00
3.21750253e-01 1.32182643e-01 1.92259774e-01 5.08686543e-01
1.18632352e+00 -1.14038937e-01 2.27339178e-01 -3.79927814e-01
4.19827551e-01 1.69380102e-02 1.20312177e-01 1.16098809e+00
-2.25055397e-01 -1.75982296e-01 3.66809130e-01 -8.31544638e-01
-5.50758302e-01 -7.72242248e-01 3.81507516e-01 1.66709602e+00
5.23487106e-02 -5.60473621e-01 -4.22875166e-01 -5.53756893e-01
-5.06582916e-01 8.11879039e-01 -4.62122381e-01 -5.61916411e-01
-7.84303486e-01 -2.84938812e-01 6.18964255e-01 7.64432430e-01
7.25003481e-01 -1.68997955e+00 -6.74939215e-01 6.28543377e-01
-1.58102542e-01 -3.62741709e-01 -3.42322230e-01 6.19019747e-01
-1.13412762e+00 -6.52784526e-01 -6.14847541e-02 -8.67116809e-01
-1.25573948e-01 3.19510140e-02 1.43730986e+00 6.69337034e-01
4.44547981e-02 4.03462946e-01 -4.51967180e-01 -2.76013017e-01
-7.34018981e-01 3.91048104e-01 -2.39104033e-01 -6.84274793e-01
1.62410572e-01 -1.00125575e+00 -1.07990913e-01 2.89947242e-01
-6.38152480e-01 3.33597064e-02 5.66029489e-01 9.82346416e-01
-9.64723229e-02 -2.45219871e-01 9.75915849e-01 -1.08857548e+00
8.51494491e-01 -5.25194585e-01 -4.22437191e-01 7.48650908e-01
-9.16275263e-01 6.83946133e-01 3.66329670e-01 -5.22493124e-01
-1.59899306e+00 -3.85576546e-01 -4.56526041e-01 2.76013881e-01
1.58199325e-01 9.44697499e-01 9.94755104e-02 -1.58951223e-01
8.14005554e-01 1.97126731e-01 7.52331167e-02 -5.64505994e-01
6.47856116e-01 5.83174229e-01 1.83881655e-01 -1.06008768e+00
-5.84429177e-03 -1.68100756e-03 -2.32452944e-01 -4.31894928e-01
-7.17366934e-01 -3.30605924e-01 -8.09352815e-01 -4.62868989e-01
6.67722881e-01 -3.16702753e-01 -9.99920368e-01 4.56955880e-01
-1.56260788e+00 -1.31329679e+00 -3.65671903e-01 1.11028217e-01
-6.20408416e-01 -7.27435946e-02 -1.25521803e+00 -9.34585810e-01
-4.53459710e-01 -4.37727809e-01 1.84270173e-01 1.84044257e-01
-5.75733662e-01 -1.50420356e+00 5.40155351e-01 1.98073506e-01
1.07112145e+00 -5.15430748e-01 9.42350328e-01 -6.59887016e-01
-8.66233587e-01 8.46185684e-02 1.09802797e-01 -1.02285646e-01
-3.39523911e-01 -3.07718873e-01 -6.95836842e-01 -1.46314070e-01
2.73890570e-02 -8.19758415e-01 7.82228649e-01 -8.06188285e-02
5.34119666e-01 -5.24185181e-01 -2.84562051e-01 -1.71893328e-01
1.19217765e+00 4.65261102e-01 8.44165504e-01 4.19988930e-01
1.27723485e-01 5.15492678e-01 7.02127695e-01 3.80842626e-01
4.83015388e-01 5.81479371e-01 1.61929876e-01 5.12111723e-01
-3.86244625e-01 -1.77267730e-01 4.38389212e-01 1.17451060e+00
-3.27920437e-01 -1.93477884e-01 -1.27904367e+00 5.53482592e-01
-2.28279614e+00 -1.26622987e+00 1.01070395e-02 1.74421048e+00
1.25594187e+00 2.98502982e-01 -1.04488723e-01 -3.41558099e-01
4.20979977e-01 1.07208736e-01 -2.01649338e-01 -9.68847394e-01
2.88198501e-01 4.07582492e-01 1.54838702e-02 1.03277445e+00
-4.04974729e-01 1.34628415e+00 7.91785431e+00 7.32623398e-01
-1.09843802e+00 6.33089602e-01 1.23156726e-01 4.62293476e-02
-3.88166942e-02 2.20802248e-01 -7.32718110e-01 -3.67842764e-02
1.80511916e+00 -2.57850200e-01 5.51493585e-01 6.19393170e-01
2.17340309e-02 -3.17202806e-01 -9.88135219e-01 5.31606019e-01
-4.59469587e-01 -1.70213878e+00 -3.56896222e-02 -3.92237633e-01
5.30922711e-01 7.07613006e-02 -2.30767757e-01 7.67717004e-01
7.36033559e-01 -1.04461050e+00 7.04685092e-01 1.05166352e+00
2.95605779e-01 -6.22859120e-01 6.04560316e-01 8.38128924e-01
-1.16944754e+00 -2.72519410e-01 -1.17515467e-01 -7.52295196e-01
2.59403288e-02 -5.63050546e-02 -6.60792053e-01 2.35720068e-01
4.28628296e-01 3.71665627e-01 -5.35289466e-01 6.28422737e-01
-3.14727426e-01 6.64427698e-01 -1.23096317e-01 -6.05156064e-01
2.89562881e-01 1.14244513e-01 9.21724439e-02 1.34543240e+00
-2.20157743e-01 2.20538795e-01 2.76052475e-01 7.03258395e-01
1.64251612e-03 -2.38557801e-01 -7.50803351e-01 -4.76042069e-02
5.90703130e-01 7.03327775e-01 -5.50174534e-01 -4.67665017e-01
-2.81962842e-01 6.67230308e-01 9.34916735e-01 7.53231421e-02
-6.90002203e-01 -7.00823441e-02 1.18453719e-01 -5.86001500e-02
-1.00640893e-01 -6.50171876e-01 -3.62255692e-01 -9.85054672e-01
-9.26151201e-02 -6.06745958e-01 7.66830981e-01 -7.77672470e-01
-9.83233988e-01 5.98759472e-01 -1.94924116e-01 -3.76015753e-01
-6.58606470e-01 -2.84033149e-01 -1.00114107e+00 7.41385698e-01
-1.25357413e+00 -1.52519679e+00 -2.00405605e-02 9.14633512e-01
5.94947278e-01 3.16019878e-02 1.40876877e+00 2.17049852e-01
-5.87456152e-02 3.73349994e-01 -4.80689444e-02 -7.09274560e-02
5.73459864e-01 -1.00977397e+00 -9.02556255e-02 4.77941245e-01
2.20253155e-01 1.06618774e+00 6.30346417e-01 -5.49232006e-01
-1.53035152e+00 -1.06594980e+00 1.24685109e+00 -8.08257043e-01
8.97081256e-01 -2.60256857e-01 -1.04715049e+00 1.21945524e+00
4.81395990e-01 -5.61196148e-01 3.42041790e-01 6.08774066e-01
-5.36282420e-01 -2.24410489e-01 -8.30987215e-01 3.54913026e-01
1.07235599e+00 -8.78621042e-01 -9.83216763e-01 3.70990217e-01
7.89191961e-01 -2.78774090e-02 -8.00123930e-01 -2.26416409e-01
6.92520797e-01 -8.08042586e-01 6.72542393e-01 -8.25046062e-01
-1.29725365e-02 1.73794568e-01 2.92323619e-01 -1.03831387e+00
-3.77637357e-01 -9.69144523e-01 -3.48242283e-01 1.26122570e+00
7.45372534e-01 -8.79535973e-01 6.58371627e-01 1.07202435e+00
-1.46302581e-01 -4.19739366e-01 -8.56512785e-01 -1.24812579e+00
3.15921396e-01 -5.54773688e-01 3.40680808e-01 7.31724501e-01
7.78042316e-01 6.68785989e-01 -7.77765810e-01 -2.44342476e-01
-9.06773731e-02 4.40413922e-01 5.84960818e-01 -8.66119802e-01
-4.91681784e-01 -2.39773870e-01 -1.68140251e-02 -1.06061983e+00
2.53063977e-01 -9.16550994e-01 1.81639165e-01 -1.67115760e+00
3.03454995e-01 -4.71569330e-01 -4.12811697e-01 8.76824737e-01
2.52838016e-01 -2.77320504e-01 1.41615242e-01 2.91947633e-01
-1.36553216e+00 3.04866195e-01 1.10398054e+00 -5.85907400e-02
-3.72803569e-01 -5.91087937e-02 -2.42157310e-01 4.45222437e-01
1.00056767e+00 -7.58051455e-01 -5.25270224e-01 -5.35784960e-01
4.28449124e-01 6.34352148e-01 1.60784334e-01 -7.99408853e-01
8.92450869e-01 -5.45282543e-01 -4.08111155e-01 -2.73824900e-01
4.02002245e-01 -6.34967804e-01 2.37864122e-01 9.19769347e-01
-7.03150630e-01 1.79761276e-01 2.16935784e-01 3.87427360e-01
-3.50667626e-01 -3.84840935e-01 7.10115910e-01 -6.42293513e-01
-1.11488402e+00 8.74489918e-02 -9.85037327e-01 -3.75483707e-02
8.55222106e-01 3.19793969e-01 -5.09727001e-01 -5.30619323e-01
-1.15422010e+00 3.93314213e-01 4.35988642e-02 5.04411280e-01
5.99756718e-01 -1.02276051e+00 -2.93321520e-01 -7.16679618e-02
-1.68699533e-01 -6.30391598e-01 1.30554140e-01 6.28809392e-01
-4.29380953e-01 6.18701398e-01 -5.73760748e-01 -5.15096746e-02
-1.24489689e+00 7.22761989e-01 3.92408073e-01 -9.29849446e-01
-5.25969088e-01 8.71039331e-01 -4.45073932e-01 -5.82679689e-01
2.60753661e-01 -1.49255106e-02 -3.49633425e-01 -8.30246210e-02
6.86040044e-01 2.95230538e-01 4.52775694e-02 1.16990313e-01
-3.82425815e-01 -2.33979803e-02 -7.17964098e-02 -4.17671889e-01
1.53464985e+00 -3.63591939e-01 -7.19259143e-01 9.73102570e-01
6.82596862e-01 -4.27733988e-01 -7.74933517e-01 -2.02710003e-01
7.99851775e-01 -9.99837518e-02 -3.92066509e-01 -1.08422554e+00
-5.71121097e-01 1.02533662e+00 3.68506193e-01 3.48114341e-01
5.86381674e-01 1.99192092e-01 7.90096462e-01 1.10418463e+00
9.61750329e-01 -1.22563839e+00 5.63729525e-01 1.29976368e+00
7.84159124e-01 -8.03871393e-01 -6.19069397e-01 2.85690695e-01
-2.96597302e-01 1.40434837e+00 8.02404106e-01 -7.81390592e-02
6.84250832e-01 5.71526527e-01 3.06277543e-01 -5.70629597e-01
-1.78004932e+00 -9.99618769e-02 -3.96733761e-01 6.03389025e-01
5.84107339e-01 -3.10244132e-02 -6.18118584e-01 5.68807065e-01
2.26456344e-01 1.79231033e-01 5.66984653e-01 1.33700800e+00
-7.34756410e-01 -1.49954641e+00 -1.92730024e-01 5.34862541e-02
-1.70302853e-01 5.23769762e-03 -4.76771563e-01 6.56218648e-01
-3.52538675e-01 1.26708186e+00 -3.53801936e-01 -2.71334827e-01
-6.50800765e-02 6.02141321e-01 6.00077987e-01 -8.03984821e-01
-9.43449140e-01 -6.50590956e-01 6.06365383e-01 -6.54718816e-01
-2.56731898e-01 -4.73665707e-02 -1.66070867e+00 -6.04161024e-01
-5.14237285e-01 6.23713970e-01 4.39000398e-01 1.12787426e+00
2.97506779e-01 4.27928150e-01 3.67768914e-01 -3.10064763e-01
-5.56386292e-01 -1.15229797e+00 -4.96378362e-01 3.43714446e-01
-1.36227533e-01 -3.05381060e-01 5.05898595e-02 -6.30193055e-02] | [12.5944242477417, 7.869400501251221] |
fbd1c6e7-67ef-475c-9351-661760284694 | beyond-human-parts-dual-part-aligned | 1910.10111 | null | https://arxiv.org/abs/1910.10111v1 | https://arxiv.org/pdf/1910.10111v1.pdf | Beyond Human Parts: Dual Part-Aligned Representations for Person Re-Identification | Person re-identification is a challenging task due to various complex factors. Recent studies have attempted to integrate human parsing results or externally defined attributes to help capture human parts or important object regions. On the other hand, there still exist many useful contextual cues that do not fall into the scope of predefined human parts or attributes. In this paper, we address the missed contextual cues by exploiting both the accurate human parts and the coarse non-human parts. In our implementation, we apply a human parsing model to extract the binary human part masks \emph{and} a self-attention mechanism to capture the soft latent (non-human) part masks. We verify the effectiveness of our approach with new state-of-the-art performances on three challenging benchmarks: Market-1501, DukeMTMC-reID and CUHK03. Our implementation is available at https://github.com/ggjy/P2Net.pytorch. | ['Jinge Yao', 'Yuhui Yuan', 'Kai Han', 'Lang Huang', 'Jianyuan Guo', 'Chao Zhang'] | 2019-10-22 | beyond-human-parts-dual-part-aligned-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Guo_Beyond_Human_Parts_Dual_Part-Aligned_Representations_for_Person_Re-Identification_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Guo_Beyond_Human_Parts_Dual_Part-Aligned_Representations_for_Person_Re-Identification_ICCV_2019_paper.pdf | iccv-2019-10 | ['human-parsing'] | ['computer-vision'] | [-5.24107553e-02 8.41873735e-02 2.20893591e-04 -5.89071691e-01
-7.85783410e-01 -3.92766923e-01 5.59346318e-01 -1.26339048e-01
-6.04722738e-01 8.54985833e-01 3.04111511e-01 2.26337433e-01
3.91689807e-01 -4.03302997e-01 -5.01052976e-01 -4.25224394e-01
3.68671209e-01 4.74553287e-01 4.50696409e-01 -2.04754621e-02
-7.06031024e-02 2.02075794e-01 -1.56565034e+00 2.60044307e-01
8.51527095e-01 7.72142708e-01 -9.01795402e-02 5.46158254e-01
7.07486942e-02 3.55212420e-01 -6.38905406e-01 -9.25643027e-01
4.85051572e-01 -2.69777864e-01 -7.66303003e-01 1.08294465e-01
5.51831186e-01 -3.91610146e-01 -1.84172481e-01 1.35644686e+00
6.66190147e-01 -4.33856547e-02 4.36737508e-01 -1.18400216e+00
-5.56839168e-01 4.95353043e-01 -7.91356325e-01 2.54003853e-01
5.06838858e-01 2.62868911e-01 7.25728214e-01 -8.65343392e-01
5.08713484e-01 1.40506589e+00 7.07644463e-01 8.56950641e-01
-9.36995685e-01 -9.68735993e-01 3.51179928e-01 2.91664779e-01
-1.52419043e+00 -6.13383591e-01 5.65502524e-01 -3.83583307e-01
6.66191995e-01 3.30322742e-01 3.49302143e-01 1.32828355e+00
-1.24863066e-01 9.67185259e-01 1.27906251e+00 -3.65654200e-01
-1.37737393e-01 3.20563048e-01 6.08731091e-01 6.51328862e-01
3.79551321e-01 -5.75613044e-02 -1.96794569e-01 -1.91265926e-01
6.53355420e-01 5.43130934e-02 -3.07865679e-01 8.68886560e-02
-9.81931865e-01 4.69729483e-01 4.36843246e-01 2.24699676e-01
-2.63593107e-01 -2.03771263e-01 3.34356487e-01 -2.71847337e-01
3.90484124e-01 7.98589289e-02 -5.19321859e-01 6.91506127e-03
-8.46370757e-01 2.73737162e-01 6.96035862e-01 1.22723770e+00
7.74995565e-01 -3.35885584e-01 -4.08710271e-01 7.55627632e-01
1.53287813e-01 3.39492321e-01 4.29422647e-01 -6.43091023e-01
5.17843604e-01 7.69263983e-01 2.41636723e-01 -5.99893808e-01
-4.72527027e-01 -5.42765617e-01 -8.48592699e-01 -8.88446569e-02
5.69392979e-01 -1.14956476e-01 -1.20351660e+00 1.62279832e+00
6.16306603e-01 1.08963013e-01 -3.48127633e-01 1.06775260e+00
1.01927590e+00 2.50851750e-01 5.34463644e-01 2.18992054e-01
1.93987012e+00 -1.24601209e+00 -6.25459552e-01 -5.39011896e-01
1.88978016e-01 -7.61502922e-01 9.87615705e-01 6.89329728e-02
-8.58567238e-01 -9.33570802e-01 -7.22725451e-01 -3.06944847e-01
-5.13278544e-01 5.03759146e-01 3.69105786e-01 6.65210545e-01
-7.16122091e-01 2.85047024e-01 -7.11511254e-01 -5.65322042e-01
4.64934647e-01 3.20197552e-01 -5.72178423e-01 -1.50790453e-01
-1.11360061e+00 6.38240218e-01 4.07809675e-01 2.37473994e-01
-3.91952276e-01 -3.27206194e-01 -8.21542621e-01 -8.07392448e-02
5.33286929e-01 -7.89929569e-01 1.18275392e+00 -1.03870976e+00
-1.13775456e+00 1.13433850e+00 -2.97289014e-01 -3.01098347e-01
8.78487349e-01 -5.25058746e-01 -4.55661029e-01 1.50153726e-01
2.91925728e-01 8.80931556e-01 6.35379493e-01 -1.21546638e+00
-6.90439284e-01 -5.57518184e-01 -1.08781934e-01 9.28789303e-02
-7.75077045e-02 4.12195146e-01 -1.05006933e+00 -8.97307038e-01
-2.14522526e-01 -1.08930755e+00 -1.37319252e-01 -2.31184110e-01
-8.75141561e-01 -3.30393046e-01 4.71434593e-01 -1.12136650e+00
1.07896638e+00 -2.03351259e+00 -7.19383806e-02 -3.88853960e-02
2.24150732e-01 4.28322792e-01 6.80693146e-03 -5.51480381e-03
5.47372177e-02 2.06620008e-01 -3.81874621e-01 -8.60711575e-01
1.14532307e-01 -4.90441732e-02 9.70930532e-02 3.70206416e-01
1.39077097e-01 9.93358314e-01 -5.12605727e-01 -7.58027434e-01
5.75598441e-02 5.43033004e-01 -2.70894617e-01 2.35709280e-01
1.76742658e-01 5.08732796e-01 -5.37192285e-01 9.95602608e-01
9.03316200e-01 -1.07496165e-01 -7.30557814e-02 -2.07437649e-01
5.05673327e-02 3.37915421e-02 -1.20257676e+00 1.34767926e+00
2.24103510e-01 1.16522051e-01 3.09231013e-01 -3.50438893e-01
5.35966396e-01 7.44092325e-03 1.40600994e-01 -6.37768388e-01
3.32822323e-01 7.60141760e-02 -3.44672985e-02 -2.70843714e-01
5.53566396e-01 8.74518976e-02 -1.68997809e-01 9.50484443e-03
4.31618840e-02 7.27479696e-01 6.71255141e-02 2.49077973e-04
8.76862884e-01 2.61461735e-01 6.50576591e-01 -3.50489914e-01
5.99893451e-01 -1.32902116e-01 1.04895949e+00 8.08157802e-01
-8.14586103e-01 1.03447688e+00 3.43885899e-01 -2.56688923e-01
-9.28280234e-01 -7.92828918e-01 -1.44499511e-01 1.16131210e+00
2.31272981e-01 -5.22040665e-01 -1.14143384e+00 -9.15915132e-01
-7.41048232e-02 5.69425285e-01 -8.36540282e-01 2.10442677e-01
-7.94945955e-01 -7.74403393e-01 7.39205718e-01 7.13818431e-01
7.95292318e-01 -1.25135660e+00 -4.58570212e-01 5.15744016e-02
-3.78683984e-01 -1.43069983e+00 -8.23229611e-01 -2.39734769e-01
-2.96399057e-01 -1.19857419e+00 -1.14104521e+00 -6.75488055e-01
7.49172986e-01 1.87169552e-01 1.08686900e+00 2.80588984e-01
-4.69526261e-01 1.24419890e-01 -4.14418399e-01 -4.06932831e-01
-1.22428527e-02 2.19585747e-01 3.85382809e-02 8.71587917e-02
6.48775041e-01 -5.99202327e-02 -7.56364584e-01 6.81322217e-01
-4.09823656e-01 2.69535512e-01 6.43109918e-01 5.68050742e-01
7.87468493e-01 -6.76185936e-02 3.56169045e-01 -1.08159554e+00
2.02976987e-01 -3.86062741e-01 -3.55967015e-01 4.43541288e-01
-1.76898137e-01 -2.95691133e-01 3.78467828e-01 -5.47958672e-01
-1.15616846e+00 3.20377082e-01 -2.67955542e-01 -2.83922046e-01
-7.26500034e-01 -2.12478012e-01 -8.56296599e-01 1.90050006e-01
3.11263651e-01 1.37470767e-01 -3.43990982e-01 -7.83252835e-01
1.46738067e-01 7.47659564e-01 8.02622855e-01 -7.32820988e-01
6.92993641e-01 3.65116566e-01 -3.97178680e-01 -4.62579548e-01
-7.96265721e-01 -7.71101356e-01 -9.05868828e-01 9.45583582e-02
1.16731203e+00 -1.06669283e+00 -5.35417140e-01 5.95095158e-01
-1.05525029e+00 -2.32726738e-01 4.09946851e-02 1.33513466e-01
-2.07146462e-02 7.17956722e-01 -9.28565562e-01 -8.68215263e-01
-7.01440632e-01 -1.19201791e+00 1.31991637e+00 6.01803005e-01
-1.38303131e-01 -3.15558076e-01 -2.23580554e-01 5.76406837e-01
7.68774971e-02 3.01879108e-01 3.06486845e-01 -8.49503040e-01
-3.91122490e-01 -4.78913724e-01 -4.62532669e-01 1.04285724e-01
-6.97429478e-02 -2.46745437e-01 -1.09813511e+00 -2.18202367e-01
-3.11765760e-01 -5.40122502e-02 1.01006007e+00 1.77702144e-01
1.09247863e+00 -3.57798517e-01 -5.64024866e-01 5.31635702e-01
1.01388884e+00 -1.23728454e-01 7.63570070e-01 1.96774632e-01
9.42680418e-01 7.29238927e-01 5.29259086e-01 3.32348138e-01
6.66648448e-01 8.79564762e-01 -8.36030990e-02 -1.55685067e-01
-2.62017429e-01 -3.00186545e-01 2.13617265e-01 2.37047613e-01
-2.36708671e-01 -3.28609526e-01 -1.00402582e+00 3.98397714e-01
-1.95837057e+00 -9.33753252e-01 -2.77215362e-01 2.13440561e+00
7.08529472e-01 3.41223180e-01 4.53476816e-01 -2.29430974e-01
1.20869541e+00 -1.85677439e-01 -4.57009435e-01 2.82014281e-01
-2.08411679e-01 2.54049767e-02 5.32951415e-01 4.21458244e-01
-1.48419225e+00 1.09398317e+00 5.16297150e+00 8.13748062e-01
-4.93278116e-01 2.94854552e-01 8.04989696e-01 1.39652967e-01
2.22632632e-01 -1.99222639e-01 -1.43657517e+00 6.88420713e-01
5.63735425e-01 3.07044029e-01 1.12832591e-01 8.19763005e-01
-2.10624874e-01 -1.46331191e-01 -9.54274654e-01 1.01661599e+00
1.35820165e-01 -6.31842911e-01 -2.23025039e-01 2.84243286e-01
5.07576048e-01 -3.26978028e-01 -2.29622826e-01 6.61586821e-01
3.43378559e-02 -9.48314726e-01 8.96155298e-01 6.34373128e-01
7.09190845e-01 -5.97053587e-01 9.81968701e-01 3.59690040e-01
-1.38114691e+00 3.87729928e-02 -3.99244964e-01 1.12189755e-01
2.20856592e-01 3.99353117e-01 -6.14568710e-01 5.65595150e-01
9.59315300e-01 3.14129621e-01 -1.07958400e+00 1.15672410e+00
-3.24996948e-01 5.60140133e-01 -3.09656054e-01 2.56271899e-01
-2.11792961e-01 1.77562669e-01 4.63352978e-01 1.48327482e+00
1.39752209e-01 5.01896918e-01 3.76824945e-01 8.87296081e-01
-1.24857739e-01 1.28299832e-01 -7.53107890e-02 2.96084851e-01
3.94626260e-01 1.48529935e+00 -9.08970654e-01 -5.52938819e-01
-5.89441895e-01 1.21625412e+00 3.18592519e-01 2.91056871e-01
-9.82692420e-01 -1.77197382e-01 7.71218538e-01 3.26039433e-01
1.86549902e-01 -4.57484163e-02 -2.85112441e-01 -1.46644688e+00
1.67074516e-01 -9.55476820e-01 7.66273975e-01 -5.52348137e-01
-1.50347471e+00 8.24789643e-01 6.47918582e-02 -1.02080214e+00
-3.45781483e-02 -5.40004313e-01 -4.94792104e-01 9.67129886e-01
-1.17217028e+00 -1.55450094e+00 -6.39631867e-01 5.33291698e-01
6.13758445e-01 -4.98436056e-02 5.63958883e-01 5.10669291e-01
-1.01720989e+00 9.94634271e-01 -5.96746027e-01 5.70490599e-01
9.31680083e-01 -1.19712424e+00 7.09057152e-01 1.15533447e+00
-1.94820955e-01 8.50121260e-01 6.92756236e-01 -9.76009250e-01
-7.19001532e-01 -9.58406925e-01 9.06699717e-01 -7.96509087e-01
1.83072224e-01 -6.82003975e-01 -1.06289876e+00 7.91647375e-01
2.38492429e-01 1.42741427e-01 5.68920732e-01 1.98050961e-02
-3.83339822e-01 2.23392099e-01 -1.38501799e+00 4.69713390e-01
1.34953356e+00 -2.64773250e-01 -5.23647249e-01 -3.06658130e-02
6.34868801e-01 -4.59968090e-01 -4.74892259e-01 5.75052559e-01
5.35355389e-01 -9.46343422e-01 1.14632773e+00 -3.91111463e-01
2.18176991e-02 -4.89590466e-01 -1.11049719e-01 -7.90382624e-01
-5.54041266e-01 -3.63233626e-01 -7.74738267e-02 1.76249933e+00
1.57652736e-01 -6.02906168e-01 7.14348257e-01 1.11672819e+00
3.81292179e-02 -4.88319635e-01 -8.40690434e-01 -5.18401265e-01
-1.38068646e-01 -1.45104125e-01 7.26603329e-01 6.88731849e-01
-3.15798074e-01 1.70067072e-01 -4.64975744e-01 2.80120373e-01
8.30112576e-01 -7.03865886e-02 8.32990289e-01 -1.09763908e+00
-3.55347008e-01 -5.28638959e-01 -3.41882676e-01 -8.20586860e-01
3.48814279e-01 -6.11161947e-01 1.55996621e-01 -1.34003127e+00
7.49241650e-01 -3.70648891e-01 -4.20634210e-01 8.77920032e-01
-7.12141395e-01 3.85057181e-01 4.25446063e-01 2.90176570e-01
-7.46892691e-01 4.12864834e-01 9.86200213e-01 -5.64127229e-03
-9.72584710e-02 1.50569752e-01 -8.88227642e-01 7.59795964e-01
1.01370323e+00 -3.41718942e-01 1.57370642e-01 -2.66234875e-01
-3.32844406e-01 -3.46911401e-01 7.57315278e-01 -1.16793370e+00
2.68219262e-01 1.15139171e-01 8.33774507e-01 -6.01236582e-01
3.72540355e-01 -5.45084596e-01 1.49571896e-01 2.17209578e-01
-1.73382107e-02 1.92904577e-01 1.63554877e-01 4.76100594e-01
3.43406759e-02 -1.41813904e-01 7.73212016e-01 -2.70684600e-01
-8.99098992e-01 3.05367887e-01 1.45813137e-01 -1.46851372e-02
8.27117622e-01 -1.58416763e-01 -5.22614717e-01 -8.52072909e-02
-6.35222614e-01 3.42285872e-01 6.47487104e-01 4.90930378e-01
4.10472035e-01 -1.13572514e+00 -6.68531895e-01 1.38652995e-01
1.41020894e-01 -2.38847718e-01 3.91249299e-01 7.60968685e-01
-2.57836878e-01 3.38473678e-01 -3.55472475e-01 -1.98847532e-01
-1.54910326e+00 6.97735369e-01 4.29166973e-01 -1.30245268e-01
-7.18377352e-01 8.72140765e-01 5.51244736e-01 -4.66418624e-01
3.36630344e-01 -1.07425474e-01 -2.68263787e-01 1.01811597e-02
7.52730370e-01 4.38997209e-01 -3.10860693e-01 -1.15108669e+00
-7.72842705e-01 5.32603204e-01 -1.23105645e-01 5.59366271e-02
9.12349582e-01 -2.36221701e-01 1.38025954e-01 -1.33481756e-01
9.64565337e-01 2.09302157e-01 -1.28904402e+00 -2.70476818e-01
9.58465040e-02 -3.38551581e-01 -4.55902964e-01 -9.41102266e-01
-8.98142815e-01 6.87726498e-01 8.35383952e-01 -2.05077648e-01
9.69388843e-01 2.26851314e-01 9.85506833e-01 -1.36513576e-01
4.94526863e-01 -1.10138249e+00 -2.56563574e-01 2.12934732e-01
8.81035268e-01 -1.49967468e+00 3.68710496e-02 -8.40251803e-01
-8.42752457e-01 7.04801977e-01 1.06839836e+00 1.51087210e-01
5.22139966e-01 8.73062089e-02 9.38680172e-02 1.61657199e-01
-1.83036685e-01 -7.13078737e-01 6.28884792e-01 4.06759113e-01
4.77054805e-01 3.74466777e-01 -3.36294591e-01 1.27615952e+00
-2.59511143e-01 -1.42369553e-01 1.15957625e-01 7.26541758e-01
-2.88193941e-01 -1.05374753e+00 -6.49059594e-01 3.35743845e-01
-8.85982692e-01 -8.65884684e-03 -6.15612328e-01 8.69132102e-01
5.42608917e-01 8.89265954e-01 -1.78106636e-01 -2.85167307e-01
3.32159847e-01 2.20032588e-01 4.25954849e-01 -5.56508303e-01
-7.02295363e-01 2.48319462e-01 1.93801403e-01 -5.24423420e-01
-2.21530557e-01 -8.57716024e-01 -1.22773242e+00 -1.70030102e-01
-3.27841081e-02 -1.34839416e-01 1.26729146e-01 7.42537677e-01
3.94284904e-01 2.39191353e-01 -1.01978332e-01 -9.36377823e-01
-1.69816554e-01 -1.13759267e+00 -2.44867951e-01 6.32787108e-01
7.38320872e-02 -6.90770626e-01 -3.73149812e-02 -1.21015646e-02] | [14.712675094604492, 0.8450649380683899] |
27139a42-4ac7-439f-8c31-ed43115ab08d | phase-only-image-based-kernel-estimation-for | 1811.10185 | null | http://arxiv.org/abs/1811.10185v3 | http://arxiv.org/pdf/1811.10185v3.pdf | Phase-only Image Based Kernel Estimation for Single-image Blind Deblurring | The image blurring process is generally modelled as the convolution of a blur
kernel with a latent image. Therefore, the estimation of the blur kernel is
essentially important for blind image deblurring. Unlike existing approaches
which focus on approaching the problem by enforcing various priors on the blur
kernel and the latent image, we are aiming at obtaining a high quality blur
kernel directly by studying the problem in the frequency domain. We show that
the auto-correlation of the absolute phase-only image can provide faithful
information about the motion (e.g. the motion direction and magnitude, we call
it the motion pattern in this paper.) that caused the blur, leading to a new
and efficient blur kernel estimation approach. The blur kernel is then refined
and the sharp image is estimated by solving an optimization problem by
enforcing a regularization on the blur kernel and the latent image. We further
extend our approach to handle non-uniform blur, which involves spatially
varying blur kernels. Our approach is evaluated extensively on synthetic and
real data and shows good results compared to the state-of-the-art deblurring
approaches. | ['Miaomiao Liu', 'Richard Hartley', 'Liyuan Pan', 'Yuchao Dai'] | 2018-11-26 | null | null | null | null | ['single-image-blind-deblurring', 'blind-image-deblurring'] | ['computer-vision', 'computer-vision'] | [ 3.06235790e-01 -5.38178086e-01 4.08876091e-01 -8.63880962e-02
-3.75726014e-01 -6.02387547e-01 6.13192081e-01 -5.43513238e-01
-3.25963646e-01 8.65919232e-01 5.67762852e-01 6.88217729e-02
-2.91774571e-01 -2.52531260e-01 -5.67170799e-01 -1.03339207e+00
6.89424342e-03 -2.01746330e-01 9.64785367e-02 2.10513547e-01
5.01730621e-01 2.30165794e-01 -1.11178553e+00 -2.84518562e-02
1.21061516e+00 7.11796939e-01 4.72361237e-01 8.84530723e-01
4.02107447e-01 1.00978696e+00 -6.37350976e-01 7.73247425e-03
7.85284415e-02 -7.15408802e-01 -9.13822293e-01 4.51033443e-01
4.20648903e-01 -5.60627937e-01 -6.26991212e-01 1.49232483e+00
2.63363630e-01 1.32961199e-01 6.05683327e-01 -6.26167953e-01
-9.30573463e-01 2.19992280e-01 -7.12720573e-01 4.27843720e-01
3.63860875e-01 -5.83400801e-02 4.48462635e-01 -9.66589451e-01
4.65850145e-01 9.84400630e-01 6.92609370e-01 1.93861753e-01
-1.49087501e+00 -6.28853515e-02 -4.66644555e-01 5.13760567e-01
-1.48284090e+00 -7.54398525e-01 8.91578555e-01 -6.02838397e-01
2.04025343e-01 4.28395778e-01 2.01137275e-01 7.33195484e-01
3.41270417e-01 3.65979940e-01 1.55154169e+00 -5.99318504e-01
2.73610443e-01 -7.22300634e-02 1.67001247e-01 3.58520776e-01
2.51373351e-01 2.75303185e-01 -2.52386034e-01 -2.35429972e-01
1.18426263e+00 -1.25719085e-01 -1.25095832e+00 -4.41641271e-01
-1.42752624e+00 3.71445984e-01 5.77869415e-01 4.37728643e-01
-6.49217308e-01 2.54866153e-01 -1.07355967e-01 9.79360491e-02
6.08514369e-01 5.40983379e-01 -9.37911943e-02 -1.97975427e-01
-1.29653192e+00 2.19986275e-01 8.68296564e-01 4.80455458e-01
7.85315216e-01 -1.61246389e-01 -3.58428240e-01 8.88501644e-01
3.36427838e-01 4.93461281e-01 4.41579103e-01 -1.06143475e+00
2.48155296e-02 -1.78579479e-01 8.95793915e-01 -6.77795291e-01
1.59137860e-01 -2.53380805e-01 -1.03635812e+00 3.06707919e-01
7.64852405e-01 -3.86128426e-01 -1.00660801e+00 1.55932868e+00
-3.16952430e-02 7.44975984e-01 1.14708148e-01 1.36255944e+00
2.25101158e-01 6.48219228e-01 -4.58971113e-01 -5.12821376e-01
1.50570655e+00 -1.00900185e+00 -1.22575021e+00 -3.60862464e-01
-1.34594977e-01 -1.36932027e+00 5.27337849e-01 1.92865163e-01
-1.25116658e+00 -5.55949748e-01 -9.41028178e-01 8.58513862e-02
1.95762485e-01 2.11007580e-01 2.14731380e-01 4.13373679e-01
-1.25119567e+00 6.42784119e-01 -6.80529296e-01 -2.74273187e-01
7.14528710e-02 4.85334471e-02 -2.28962645e-01 -4.84194756e-01
-1.17979813e+00 1.28215444e+00 1.41735524e-01 3.58190715e-01
-6.29737437e-01 -5.86624801e-01 -7.74355769e-01 9.15477127e-02
6.70013055e-02 -8.54288101e-01 1.20856369e+00 -1.08578944e+00
-1.52888703e+00 4.68992978e-01 -5.53786397e-01 -3.83785695e-01
6.10440612e-01 -5.65247357e-01 -1.84671327e-01 2.41311327e-01
-6.25322983e-02 9.93163809e-02 1.53920114e+00 -1.44488633e+00
-1.28740892e-01 2.55122911e-02 -1.92999970e-02 2.18722209e-01
2.52807200e-01 2.70146560e-02 -4.07862395e-01 -9.92013335e-01
1.03060119e-01 -8.76277506e-01 -3.20134126e-02 -2.37633973e-01
-2.29853526e-01 5.61317086e-01 8.73762071e-01 -1.03289068e+00
1.30108082e+00 -2.29472995e+00 3.88950020e-01 -3.02432001e-01
3.34383279e-01 2.74091274e-01 1.40725493e-01 3.66281599e-01
-2.73851246e-01 -4.41344023e-01 -5.94952941e-01 -2.94919550e-01
-3.57940763e-01 6.47269934e-03 -4.83660609e-01 9.73108172e-01
1.29043683e-02 8.22627962e-01 -1.00487411e+00 -4.15307283e-03
4.06771809e-01 6.85554802e-01 -2.37506941e-01 6.16844058e-01
2.76319355e-01 7.08252132e-01 -8.96105245e-02 9.21461061e-02
1.30186236e+00 -3.34464133e-01 -1.14221044e-01 -4.75446433e-01
-4.17445958e-01 -2.22536489e-01 -1.08434713e+00 1.66516554e+00
-4.37190205e-01 9.30495679e-01 5.25099456e-01 -5.81596315e-01
5.22362232e-01 4.67426568e-01 2.22183704e-01 -6.38970882e-02
-9.56818536e-02 1.88151196e-01 -2.67383885e-02 -6.48302615e-01
5.16903222e-01 -2.96385109e-01 3.17695290e-01 3.12029034e-01
-1.77661896e-01 -2.42385909e-01 -8.63719583e-02 -1.58884674e-02
1.06957078e+00 9.60561484e-02 1.75115094e-01 -7.70656288e-01
6.38839066e-01 -2.32289046e-01 1.46446293e-02 7.84822762e-01
-1.81612760e-01 9.51068342e-01 1.96708232e-01 -1.04348674e-01
-1.08187699e+00 -1.02631760e+00 -2.67803490e-01 2.01194078e-01
5.70436895e-01 -7.13898912e-02 -1.27852547e+00 -1.43732414e-01
-1.86554119e-01 6.15584791e-01 -6.83289230e-01 -1.59397677e-01
-4.37358081e-01 -9.31076348e-01 9.41463783e-02 1.00258470e-01
8.92770469e-01 -5.67743123e-01 -4.99031216e-01 2.59148926e-01
-4.66035992e-01 -1.14925683e+00 -1.11689878e+00 -1.30257130e-01
-6.71187401e-01 -1.13969541e+00 -1.48722947e+00 -6.34007573e-01
8.10382843e-01 6.21240079e-01 7.39573538e-01 -2.31302604e-01
-1.79735541e-01 2.93405741e-01 -8.75561982e-02 1.85129747e-01
-4.42927659e-01 -7.11745441e-01 -1.94030292e-02 5.54949403e-01
-1.09521322e-01 -4.86487687e-01 -8.91121864e-01 3.03128511e-01
-1.01679254e+00 2.37675503e-01 6.48188591e-01 9.96618867e-01
-1.90234318e-01 3.51260334e-01 1.04320712e-01 -5.68857014e-01
8.15516353e-01 -2.78127462e-01 -7.59699523e-01 8.62835124e-02
-5.41366220e-01 3.08772862e-01 2.87976056e-01 -6.01484001e-01
-1.65145707e+00 -1.45803139e-01 3.82004112e-01 -6.40669346e-01
-1.86200827e-01 3.12269777e-01 1.99751742e-02 -2.43130133e-01
6.49146080e-01 3.45654547e-01 -4.39503376e-04 -7.31285214e-01
5.17216384e-01 7.42905617e-01 8.98907781e-01 -2.11764902e-01
1.04582870e+00 5.32382727e-01 -9.36080590e-02 -1.00337255e+00
-6.51330292e-01 -6.37953222e-01 -6.59553468e-01 -4.91218977e-02
9.13650632e-01 -8.65996063e-01 -4.01044309e-01 1.16363108e+00
-1.53090084e+00 -2.30882138e-01 -1.71074048e-01 8.14897180e-01
-5.56152046e-01 7.48180270e-01 -1.00882328e+00 -8.19911301e-01
9.85878557e-02 -1.24004555e+00 8.31139266e-01 2.32012376e-01
-7.65237734e-02 -1.18743539e+00 2.90491104e-01 1.30613104e-01
8.76740336e-01 -9.17580500e-02 6.33884966e-01 4.10020977e-01
-8.71347010e-01 6.19949177e-02 -7.31999695e-01 4.37649310e-01
6.34893954e-01 -5.05363643e-01 -1.25790203e+00 -2.39731580e-01
7.27769852e-01 3.96459639e-01 9.99558568e-01 9.14948046e-01
7.26821601e-01 -4.59733158e-01 -1.83860138e-01 7.69960344e-01
1.58462787e+00 -5.38798198e-02 8.82647634e-01 1.89931124e-01
7.94454873e-01 3.58529627e-01 4.26979125e-01 2.03839779e-01
-1.41095385e-01 9.05005395e-01 1.86374262e-01 -1.70151949e-01
-4.31096345e-01 1.94986030e-01 2.24330589e-01 5.85571706e-01
-3.04980814e-01 -1.03753410e-01 -5.46008885e-01 7.08381951e-01
-1.96899664e+00 -1.00000656e+00 -3.34577620e-01 2.36386228e+00
1.10304904e+00 -2.99682677e-01 -4.69911486e-01 -1.15921296e-01
9.13322031e-01 3.58751982e-01 -2.83555269e-01 3.13389972e-02
-7.93078623e-04 1.94907673e-02 6.07044041e-01 1.13328445e+00
-1.10902870e+00 6.95332527e-01 6.86278009e+00 5.97485542e-01
-1.08683991e+00 1.54079860e-02 3.32399696e-01 4.36195105e-01
-6.49978369e-02 1.27064154e-01 -1.10728517e-01 7.32364714e-01
6.88062012e-01 -3.03553641e-01 1.06004751e+00 3.03914338e-01
6.35519326e-01 -5.08651733e-01 -8.67706716e-01 1.18438518e+00
-7.87320659e-02 -9.86737370e-01 -4.78426456e-01 1.38337716e-01
8.16222131e-01 -2.58046001e-01 2.13570409e-02 -4.81497020e-01
1.09825410e-01 -9.14318562e-01 7.52621651e-01 9.91116107e-01
9.00634110e-01 -2.47717187e-01 8.80773187e-01 1.73225641e-01
-7.84449935e-01 3.04641247e-01 -2.52417982e-01 -2.40031451e-01
3.03161353e-01 1.13234890e+00 -4.49922293e-01 5.52257717e-01
6.02727354e-01 7.61879086e-01 -3.49443853e-01 1.40150964e+00
-3.12540978e-01 5.55038452e-01 1.85243800e-01 6.57036126e-01
-9.05480906e-02 -5.64677536e-01 8.36556137e-01 1.25884593e+00
6.00687146e-01 1.52043834e-01 -4.05832469e-01 1.17500556e+00
6.24595545e-02 -6.72899425e-01 -2.36919433e-01 2.46355325e-01
1.13585159e-01 1.11474180e+00 -4.16106910e-01 -3.59850675e-01
-3.22689623e-01 1.73139775e+00 -1.31210774e-01 9.38584447e-01
-7.18345046e-01 -3.43148082e-01 9.01521146e-01 -1.35996968e-01
4.42755491e-01 -1.90459415e-01 -9.64250565e-02 -1.59425533e+00
-1.56649560e-01 -5.32220542e-01 -2.96941608e-01 -1.31725323e+00
-1.32782483e+00 6.30808234e-01 6.76158592e-02 -1.13682783e+00
-2.30336756e-01 -4.40523207e-01 -7.10517228e-01 1.69775307e+00
-1.76991582e+00 -7.60583699e-01 -6.43329918e-01 4.98820543e-01
4.28495884e-01 4.32683885e-01 4.95211929e-01 9.39430371e-02
-1.55045956e-01 -2.59042591e-01 5.28464973e-01 -1.88853294e-01
1.09410441e+00 -1.50784171e+00 3.48768651e-01 1.24860597e+00
-3.28032494e-01 7.79970288e-01 1.32440484e+00 -6.01905406e-01
-1.27812564e+00 -8.52488935e-01 6.77783549e-01 -4.10289913e-01
8.46996427e-01 -9.71623436e-02 -1.16854584e+00 5.05079389e-01
7.62437344e-01 2.29499251e-01 6.27341866e-02 -4.80679512e-01
1.07352426e-02 1.45411268e-01 -9.96518016e-01 3.08506012e-01
6.29646838e-01 -6.16969764e-01 -8.24309170e-01 -3.87643017e-02
5.24199605e-01 -5.33033788e-01 -5.34972548e-01 1.63608029e-01
3.62805665e-01 -1.06150556e+00 1.01380861e+00 -2.51680277e-02
3.33074421e-01 -6.92681849e-01 1.81346402e-01 -1.82297134e+00
-8.21615756e-01 -9.07562673e-01 -3.91256511e-01 1.07437599e+00
-2.41036609e-01 -5.94138145e-01 3.41864109e-01 4.33281124e-01
8.53102356e-02 -1.28151581e-01 -5.72758019e-01 -5.06735504e-01
-5.53897917e-01 1.25844404e-01 2.12164298e-01 1.05343902e+00
-1.71543524e-01 2.17029721e-01 -7.71828353e-01 5.64895093e-01
1.05130708e+00 2.16592446e-01 3.32659245e-01 -8.12776983e-01
-4.12984163e-01 -1.86078697e-01 -9.12723392e-02 -1.56489050e+00
-1.89723343e-01 -1.58238634e-01 4.85815734e-01 -1.32930458e+00
3.00106198e-01 -4.25142907e-02 -1.05310813e-01 -2.54741400e-01
-6.38901830e-01 2.20910519e-01 -9.17754322e-02 6.69396520e-01
-9.89971384e-02 3.61359835e-01 1.45349991e+00 4.94558401e-02
-9.52482373e-02 1.04504190e-01 -4.81080949e-01 7.60808170e-01
3.82819414e-01 -1.53854102e-01 -2.93024272e-01 -5.34862638e-01
-3.21047723e-01 2.05640793e-01 7.80646980e-01 -8.38462353e-01
4.36626196e-01 -9.36403498e-02 3.90089244e-01 -1.45527527e-01
3.50253105e-01 -8.12589347e-01 3.53498876e-01 3.55111361e-01
-8.78117234e-02 -3.90353948e-01 9.82716084e-02 6.41429067e-01
-4.08569008e-01 -5.81098437e-01 1.00155544e+00 -1.48331210e-01
-5.27710438e-01 -2.06006002e-02 -4.30227131e-01 -2.06002951e-01
6.32906020e-01 -4.38578315e-02 -5.05809009e-01 -5.86385667e-01
-7.54247844e-01 -4.29664195e-01 8.69091332e-01 4.03730199e-02
5.13604581e-01 -1.17000461e+00 -7.14534342e-01 2.88069248e-01
-2.44842261e-01 -2.93032110e-01 3.31059784e-01 1.09508872e+00
-6.89855278e-01 3.13742489e-01 -2.13344954e-02 -4.26560938e-01
-1.20407891e+00 7.76631892e-01 6.06510460e-01 -4.99623828e-02
-3.46934825e-01 8.04065824e-01 6.42699778e-01 5.20493746e-01
-1.44720480e-01 -3.38654310e-01 -7.14446753e-02 -3.19982737e-01
7.40633368e-01 5.29561520e-01 -2.03410655e-01 -8.10949922e-01
-2.19249830e-01 7.39226282e-01 1.50474653e-01 -2.47993499e-01
1.18714404e+00 -6.39574647e-01 -6.33894622e-01 1.70825243e-01
1.16820717e+00 4.94626850e-01 -1.74899232e+00 -3.05880010e-01
-3.69622372e-02 -9.03752208e-01 5.20771384e-01 -6.92274153e-01
-7.97975659e-01 6.81100011e-01 7.94900954e-01 4.23177719e-01
1.34445632e+00 2.94854748e-03 4.61840868e-01 -4.38254207e-01
1.87096357e-01 -4.43389624e-01 -5.15672304e-02 3.25865507e-01
1.03441012e+00 -1.05724275e+00 1.13190010e-01 -6.13584816e-01
-2.45174259e-01 1.18099129e+00 -8.01494643e-02 -1.90354034e-01
5.95331192e-01 2.14146644e-01 1.31793544e-01 2.37025111e-03
-2.03338280e-01 -1.55061364e-01 6.44523084e-01 4.07974094e-01
4.15281326e-01 -1.36872068e-01 -1.40891299e-01 1.76418386e-02
2.68604994e-01 4.01524544e-01 7.90647388e-01 6.52592838e-01
-4.39563364e-01 -9.89076912e-01 -9.56260383e-01 -5.49888425e-02
-6.07737005e-01 -3.31378311e-01 -1.19500468e-02 1.33335710e-01
6.27454966e-02 9.75339770e-01 -1.15163207e-01 1.35831743e-01
8.14225450e-02 -1.22241303e-01 7.59387016e-01 -3.46429348e-01
1.19537875e-01 4.88575399e-01 -1.68504462e-01 -4.13096637e-01
-5.92064023e-01 -6.75003588e-01 -5.32750189e-01 -1.92153662e-01
-5.20671546e-01 3.15454751e-01 4.63390499e-01 7.47334063e-01
1.26082838e-01 4.57772583e-01 6.29511535e-01 -1.08842301e+00
-4.76594061e-01 -1.23196137e+00 -8.56984019e-01 5.06981969e-01
1.09098148e+00 -4.01488274e-01 -8.13014984e-01 5.69386899e-01] | [11.616364479064941, -2.7486751079559326] |
eca5fc9b-e6bd-410f-968b-d7a748f9647a | graph-neural-networks-for-molecules | 2209.05582 | null | https://arxiv.org/abs/2209.05582v2 | https://arxiv.org/pdf/2209.05582v2.pdf | Graph Neural Networks for Molecules | Graph neural networks (GNNs), which are capable of learning representations from graphical data, are naturally suitable for modeling molecular systems. This review introduces GNNs and their various applications for small organic molecules. GNNs rely on message-passing operations, a generic yet powerful framework, to update node features iteratively. Many researches design GNN architectures to effectively learn topological information of 2D molecule graphs as well as geometric information of 3D molecular systems. GNNs have been implemented in a wide variety of molecular applications, including molecular property prediction, molecular scoring and docking, molecular optimization and de novo generation, molecular dynamics simulation, etc. Besides, the review also summarizes the recent development of self-supervised learning for molecules with GNNs. | ['Amir Barati Farimani', 'Zijie Li', 'Yuyang Wang'] | 2022-09-12 | null | null | null | null | ['molecular-property-prediction'] | ['miscellaneous'] | [ 1.77901149e-01 -2.01709613e-01 -7.00109720e-01 -1.45310029e-01
3.70798200e-01 -2.50956357e-01 3.55627030e-01 9.75120962e-01
-1.54329538e-01 1.08389318e+00 -3.87197793e-01 -9.64879930e-01
-4.20173258e-01 -1.25729156e+00 -6.85666203e-01 -8.02547634e-01
-9.72549736e-01 4.04109687e-01 9.40722600e-02 -3.96580130e-01
4.94420081e-01 1.33473182e+00 -1.05695868e+00 1.86263487e-01
8.19734871e-01 6.72245383e-01 1.37917250e-01 8.83540213e-01
-3.86302441e-01 8.92207086e-01 -5.50028622e-01 -1.99596286e-01
-2.56414950e-01 -2.44288713e-01 -4.77177292e-01 -4.56002861e-01
3.86936754e-01 3.11887264e-01 -7.42086351e-01 9.63929772e-01
6.94298625e-01 4.47512627e-01 1.12863243e+00 -1.00543427e+00
-8.85825872e-01 4.98077393e-01 -9.56158936e-02 2.60805339e-01
6.18551791e-01 7.01860040e-02 8.01934421e-01 -8.48877609e-01
8.16129386e-01 1.21908128e+00 7.34026849e-01 3.63386363e-01
-1.10126972e+00 -7.23937452e-01 8.42275172e-02 6.12898529e-01
-1.27207971e+00 3.92172113e-02 8.10906947e-01 -3.51681292e-01
1.56041813e+00 1.29125595e-01 7.17451870e-01 7.33260572e-01
1.04135263e+00 2.81630725e-01 6.64246559e-01 -3.71418029e-01
4.12136346e-01 -3.68039906e-01 2.36722946e-01 9.07600105e-01
5.85882187e-01 1.90904960e-01 -5.86520135e-01 -5.08241177e-01
9.65966344e-01 3.09906363e-01 -1.69941429e-02 -7.22660661e-01
-7.71607101e-01 1.16554928e+00 1.14256644e+00 2.24506691e-01
-3.68782580e-01 2.10591584e-01 4.45521951e-01 4.92711097e-01
2.82358173e-02 7.63393879e-01 -4.12716329e-01 4.50435698e-01
-2.32664078e-01 7.99142867e-02 9.20478582e-01 7.80995488e-01
6.72541142e-01 4.19619977e-01 2.16535091e-01 3.37594241e-01
4.58985686e-01 2.60683179e-01 4.77925450e-01 1.74207259e-02
3.38136315e-01 8.34464252e-01 -4.50828344e-01 -1.32783258e+00
-1.04595959e+00 -1.70847356e-01 -1.49969995e+00 1.59512714e-01
-2.12069035e-01 -3.22729349e-02 -1.19733167e+00 1.14764953e+00
3.83976340e-01 -4.14258018e-02 1.98584870e-01 2.93886960e-01
1.76300240e+00 1.03769803e+00 5.23080707e-01 -2.70191461e-01
7.32890844e-01 -8.35993528e-01 -5.47664642e-01 4.24671203e-01
7.49940872e-01 -4.76247340e-01 1.29226506e-01 3.97889584e-01
-8.93007457e-01 -5.74440122e-01 -1.22035801e+00 1.03701450e-01
-1.13859165e+00 -1.90804049e-01 1.60447764e+00 7.30927587e-01
-8.78994644e-01 1.35068727e+00 -8.67080867e-01 -2.16917500e-01
6.37883961e-01 1.17412686e+00 -6.14764273e-01 8.50227922e-02
-1.36585748e+00 9.53172028e-01 9.95651007e-01 1.89011469e-01
-9.54244971e-01 -2.84142524e-01 -1.08439612e+00 -1.92173004e-01
-1.17209814e-01 -7.21328080e-01 6.36528611e-01 -3.42233866e-01
-1.68527782e+00 3.60197991e-01 5.78501448e-02 -5.61637044e-01
-2.86469102e-01 5.77669859e-01 -6.81610823e-01 -6.57956004e-02
-4.51782644e-01 3.97414565e-01 4.86486971e-01 -5.97243488e-01
1.41038015e-01 -3.65146965e-01 -3.75760980e-02 2.64585644e-01
-8.29296932e-02 -3.41456890e-01 2.84166217e-01 -4.12940741e-01
7.72080794e-02 -5.32213211e-01 -8.91317487e-01 -3.22667539e-01
-6.73252761e-01 -5.12748539e-01 7.22988784e-01 -1.41173616e-01
1.30430079e+00 -1.40196502e+00 3.06389987e-01 8.82739544e-01
6.80548072e-01 5.95246851e-01 -3.35176289e-01 9.16682422e-01
-6.81382120e-01 6.65693283e-02 3.44460225e-03 5.41404963e-01
-3.88766140e-01 -4.59406467e-04 1.55214548e-01 3.68303627e-01
1.18211461e-02 1.17279184e+00 -9.75704253e-01 -1.12105072e-01
4.76957768e-01 4.54379141e-01 -1.87131003e-01 7.23658204e-02
-5.96346498e-01 2.75330752e-01 -7.08590984e-01 6.98993862e-01
7.59957433e-01 -6.82142377e-01 5.80424964e-01 -8.24400112e-02
1.05900795e-03 3.51495594e-01 -1.04862368e+00 1.44197428e+00
6.74168169e-02 2.96638489e-01 -5.56437969e-01 -1.16806912e+00
1.12847435e+00 2.24424809e-01 4.58727181e-01 -6.51240110e-01
7.58255497e-02 7.05954805e-02 3.57108325e-01 -3.23740721e-01
1.41997278e-01 5.28067686e-02 4.54856753e-01 1.98202729e-01
3.69962573e-01 -3.79078537e-02 3.95915180e-01 1.47591248e-01
9.89772618e-01 -2.30970204e-01 8.55827630e-01 1.74195558e-01
8.30578148e-01 -3.11402492e-02 -7.64059201e-02 7.57298768e-01
1.47583529e-01 -2.01989308e-01 4.12539750e-01 -1.09365726e+00
-9.44720745e-01 -9.78598833e-01 -1.65777817e-01 1.05533540e+00
1.07132830e-01 -7.66313732e-01 -3.60137463e-01 -5.91969430e-01
2.18940526e-01 -2.71408819e-02 -5.04225194e-01 -4.37523603e-01
-4.35089290e-01 -1.16823459e+00 2.72139043e-01 2.21924379e-01
4.55801971e-02 -1.38081384e+00 3.01700324e-01 6.70925617e-01
8.92792881e-01 -5.47212780e-01 1.05664432e-01 7.87886858e-01
-1.34834385e+00 -1.37737513e+00 -3.51131767e-01 -1.13965368e+00
7.81029820e-01 2.63301492e-01 9.53134596e-01 3.35385144e-01
-4.77172732e-01 -2.95085102e-01 2.75019556e-02 -4.31423336e-01
-5.73543727e-01 3.71291459e-01 3.45060855e-01 -5.98264754e-01
1.48457959e-01 -9.20598805e-01 -3.43767911e-01 2.82013621e-02
-8.35062981e-01 -9.98547077e-02 6.85573518e-01 8.26803625e-01
6.62556767e-01 1.25643745e-01 5.30446053e-01 -1.12662148e+00
1.01709628e+00 -4.35956240e-01 -6.88326895e-01 3.69409114e-01
-5.65579414e-01 3.34895730e-01 1.10706878e+00 -4.52404171e-01
-3.18334460e-01 5.65969013e-02 -4.65360105e-01 2.54991390e-02
-2.47289583e-01 9.83786821e-01 -1.37872905e-01 -1.05236101e+00
9.95882452e-01 1.46801099e-01 -9.97434631e-02 -2.55501717e-01
4.23999459e-01 2.97867030e-01 -2.21724715e-02 -4.84740168e-01
5.59385002e-01 -1.31309658e-01 9.14042234e-01 -1.13824034e+00
-1.23265617e-01 -2.47348949e-01 -9.08332944e-01 1.56343564e-01
5.60363352e-01 -5.60369909e-01 -1.34224939e+00 4.81756866e-01
-1.25343096e+00 -2.55111277e-01 2.34248906e-01 6.21144235e-01
-2.74982154e-01 5.56921780e-01 -8.25916946e-01 -2.68036425e-01
-5.37388325e-01 -1.00247133e+00 3.07266891e-01 5.50961435e-01
1.23841196e-01 -1.63410854e+00 4.05237138e-01 -3.50563228e-01
3.33901674e-01 5.55947185e-01 1.52036798e+00 -9.25211132e-01
-7.41265774e-01 -3.24666321e-01 -7.27678537e-02 -1.65980071e-01
3.42245847e-01 2.06441060e-01 -5.66354573e-01 -4.16379243e-01
-7.52804875e-01 -2.29401052e-01 7.87474275e-01 6.49664819e-01
1.33502138e+00 -4.17892903e-01 -8.35864604e-01 7.14184821e-01
1.36892176e+00 7.93999434e-01 6.24080896e-01 4.60011661e-02
1.04110289e+00 -5.45471236e-02 -5.72222732e-02 2.09883660e-01
2.15252806e-02 3.48201632e-01 6.41406000e-01 -2.98493475e-01
8.33771229e-02 -3.29552859e-01 2.26849541e-01 8.68396103e-01
-5.06047249e-01 -7.19603658e-01 -7.60262072e-01 -4.88811970e-01
-1.69171619e+00 -9.94253099e-01 -2.62145519e-01 1.99499464e+00
7.41851389e-01 1.04460359e-01 1.38969734e-01 -8.40137303e-02
7.98013031e-01 2.96224833e-01 -8.68341208e-01 -8.70929360e-01
-1.75901681e-01 8.37767482e-01 5.47963440e-01 5.41320264e-01
-1.15817928e+00 1.10734129e+00 7.87164736e+00 8.79772544e-01
-1.38099277e+00 -4.97068316e-01 5.68872869e-01 5.88640392e-01
-1.05157167e-01 -2.02528715e-01 -7.16012180e-01 1.99040174e-01
1.18330383e+00 -1.61673710e-01 2.18350813e-01 8.38644564e-01
8.79303273e-03 2.53607571e-01 -1.01342821e+00 1.03858244e+00
-1.70021862e-01 -2.23975158e+00 9.11638141e-01 1.99507680e-02
9.02763903e-01 7.73114676e-04 3.63622531e-02 6.04780717e-03
3.26435417e-01 -1.68783855e+00 -2.99970865e-01 5.58199644e-01
6.46435797e-01 -1.00759268e+00 6.31502151e-01 -6.86011240e-02
-1.48623765e+00 2.38943785e-01 -7.17349529e-01 -3.91822457e-01
-9.37987790e-02 4.15384889e-01 -1.14936066e+00 7.90456653e-01
9.34141353e-02 1.16220367e+00 -7.35343516e-01 1.52908933e+00
-2.83193827e-01 2.04312265e-01 -8.39157552e-02 -9.14908528e-01
4.31813240e-01 -6.50580227e-01 2.40888909e-01 9.98866022e-01
-1.58010542e-01 6.59257025e-02 3.29928607e-01 5.50687671e-01
-3.80607784e-01 3.58882219e-01 -9.65643048e-01 -4.70636606e-01
2.31429577e-01 1.04237366e+00 -8.63619447e-01 -3.04782540e-01
-5.43311052e-02 5.53465486e-01 4.19735998e-01 4.07411724e-01
-6.04336679e-01 -1.02668929e+00 5.10527074e-01 -1.45349011e-01
8.14899728e-02 -6.35228455e-01 2.14732960e-01 -6.98846519e-01
-6.30914569e-01 -7.66710758e-01 5.15923858e-01 -6.60706222e-01
-1.13693523e+00 5.06599307e-01 -2.94732720e-01 -9.22023177e-01
8.92355666e-02 -1.60049987e+00 -8.96621704e-01 6.01768076e-01
-1.48741043e+00 -7.80802608e-01 3.99649404e-02 6.69682026e-01
-1.93728924e-01 -4.86136347e-01 1.16513872e+00 2.31449679e-01
-7.33385503e-01 2.98165679e-01 5.51702857e-01 -9.34660621e-03
3.71614754e-01 -1.19986296e+00 6.51951611e-01 1.17717326e-01
3.50507230e-01 1.00256228e+00 2.41605937e-01 -7.86402047e-01
-1.71491706e+00 -1.03374851e+00 5.63210189e-01 -1.18045926e-01
6.37087762e-01 -2.47619599e-01 -8.48709106e-01 2.85184592e-01
-8.79554227e-02 -3.37848766e-03 9.70429957e-01 2.03606740e-01
-1.83339953e-01 -5.68848429e-03 -9.14727330e-01 7.05350280e-01
1.17391825e+00 -4.28403854e-01 -7.54539967e-02 1.11160862e+00
5.92734873e-01 -8.09457839e-01 -1.28598142e+00 3.91718298e-01
4.18914884e-01 -6.65920436e-01 1.37178171e+00 -1.55516052e+00
1.53991915e-02 -4.00071561e-01 3.26039881e-01 -1.21326852e+00
-6.71246886e-01 -8.40277314e-01 -5.50928712e-01 6.07943088e-02
7.16900945e-01 -6.10436678e-01 1.09393096e+00 1.29747078e-01
-1.68378890e-01 -8.36030543e-01 -8.52007866e-01 -5.96525609e-01
-1.06839016e-01 4.72245319e-03 6.80137575e-01 8.46652746e-01
1.62083715e-01 8.07305634e-01 -2.31350675e-01 6.80627301e-03
3.93380970e-01 4.25311029e-02 6.25818968e-01 -1.50007486e+00
5.97148016e-02 -7.33498454e-01 -1.11386704e+00 -1.06532979e+00
1.00579873e-01 -1.27896726e+00 -6.96706295e-01 -1.79281831e+00
9.02969018e-02 -3.59294355e-01 -4.75405663e-01 4.41156983e-01
2.84319729e-01 -7.57139698e-02 -3.06271762e-01 -8.72937962e-02
-8.48802149e-01 5.06443858e-01 1.50631535e+00 -4.88747597e-01
-5.99263251e-01 2.84247518e-01 -3.65739286e-01 5.76188624e-01
8.28032434e-01 -4.14252907e-01 -2.97327191e-01 1.20372057e-01
5.26984930e-01 -8.41062441e-02 1.11557215e-01 -9.13600743e-01
5.17360806e-01 -5.00146866e-01 9.56125319e-01 -7.88072228e-01
2.05952317e-01 -6.00779414e-01 2.53234118e-01 9.79558051e-01
-1.00474805e-01 3.17597032e-01 2.79824674e-01 7.01313019e-01
-1.84761599e-01 -2.89395034e-01 6.54885650e-01 -2.39223212e-01
-6.84345722e-01 1.16701317e+00 -6.33588195e-01 -6.98215246e-01
9.05579984e-01 -5.17884851e-01 -2.61874795e-01 -8.81917104e-02
-9.31701422e-01 6.11127317e-02 1.34646595e-01 7.66050369e-02
9.82188821e-01 -1.45520246e+00 3.63405198e-02 3.14554095e-01
2.00560000e-02 -1.47804871e-01 1.39449835e-01 3.66327077e-01
-1.10284901e+00 9.18876112e-01 -4.19760168e-01 -3.66945922e-01
-1.26507437e+00 7.88458824e-01 5.35709739e-01 -4.93245125e-01
-9.67028290e-02 7.39496648e-01 -1.05923735e-01 -6.70946240e-01
3.59798551e-01 -4.73185807e-01 -4.57363069e-01 -2.78358310e-01
3.15329403e-01 2.84904242e-01 3.61654997e-01 -2.46334597e-01
-4.78097677e-01 6.28077626e-01 -3.29965502e-01 8.16707373e-01
1.60292923e+00 6.56450331e-01 -4.06947762e-01 6.01218417e-02
1.27294540e+00 -4.09490407e-01 -5.90555310e-01 -3.59545611e-02
6.29691035e-02 2.15161428e-01 -2.79357165e-01 -5.49258113e-01
-5.86810172e-01 9.63632464e-01 5.72179735e-01 2.06209108e-01
3.83490562e-01 -2.76915014e-01 4.73431826e-01 1.30762112e+00
1.91465184e-01 -8.32818151e-01 3.24236780e-01 1.01332581e+00
6.81672931e-01 -1.10770273e+00 5.83310664e-01 -2.67463267e-01
9.40306261e-02 1.90528917e+00 5.08438885e-01 -9.85784903e-02
9.25397456e-01 -2.41245136e-01 -5.84667444e-01 -6.34566128e-01
-5.44356048e-01 1.03348434e-01 5.24336636e-01 1.02512407e+00
6.96285486e-01 6.36397228e-02 -3.29174511e-02 -1.05843218e-02
2.68986020e-02 -2.57494330e-01 3.14913213e-01 9.81003523e-01
-6.66900933e-01 -1.45256007e+00 -1.93871483e-01 5.69035649e-01
1.04510948e-01 -4.04104531e-01 -6.69228017e-01 7.49855340e-01
-2.96258628e-02 4.78483737e-01 -3.53771001e-01 -3.23412448e-01
2.71417886e-01 -2.35471308e-01 9.06905711e-01 -6.85588539e-01
-4.52759832e-01 -6.22582912e-01 -6.26894683e-02 -2.89425820e-01
-4.91277665e-01 6.35012016e-02 -1.33645248e+00 -6.34333551e-01
-5.79915643e-01 5.42369306e-01 7.10447609e-01 5.45274198e-01
5.91496408e-01 5.58398724e-01 7.56906152e-01 -9.46587145e-01
1.59062445e-01 -6.81054831e-01 -5.99589050e-01 -9.86246616e-02
2.45741338e-01 -6.22543752e-01 8.52871463e-02 -3.45331311e-01] | [5.158682823181152, 5.819759368896484] |
39f00452-c522-49f1-8b7c-ec6d14da2faf | short-term-and-long-term-context-aggregation-1 | 2009.05721 | null | https://arxiv.org/abs/2009.05721v1 | https://arxiv.org/pdf/2009.05721v1.pdf | Short-Term and Long-Term Context Aggregation Network for Video Inpainting | Video inpainting aims to restore missing regions of a video and has many applications such as video editing and object removal. However, existing methods either suffer from inaccurate short-term context aggregation or rarely explore long-term frame information. In this work, we present a novel context aggregation network to effectively exploit both short-term and long-term frame information for video inpainting. In the encoding stage, we propose boundary-aware short-term context aggregation, which aligns and aggregates, from neighbor frames, local regions that are closely related to the boundary context of missing regions into the target frame. Furthermore, we propose dynamic long-term context aggregation to globally refine the feature map generated in the encoding stage using long-term frame features, which are dynamically updated throughout the inpainting process. Experiments show that it outperforms state-of-the-art methods with better inpainting results and fast inpainting speed. | ['DaCheng Tao', 'Mingming Gong', 'Ang Li', 'Ramamohanarao Kotagiri', 'Rui Zhang', 'Jianzhong Qi', 'Xingjun Ma', 'Shanshan Zhao'] | 2020-09-12 | short-term-and-long-term-context-aggregation | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/2723_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123490698.pdf | eccv-2020-8 | ['video-inpainting'] | ['computer-vision'] | [ 2.93520927e-01 -4.93885845e-01 -2.38019437e-01 -3.86111587e-01
-4.95334208e-01 -2.68003047e-01 2.29676142e-01 2.18442261e-01
-2.93400228e-01 9.56366718e-01 5.23528397e-01 2.04100087e-01
5.28896265e-02 -7.64370382e-01 -8.21125865e-01 -5.22627711e-01
6.33358434e-02 -1.10729583e-01 5.02879083e-01 -1.17820129e-01
4.81477529e-01 5.72736859e-01 -1.50036919e+00 6.78888917e-01
8.17702830e-01 1.08309388e+00 6.47071898e-01 5.88300586e-01
-3.26985717e-01 9.72124159e-01 -5.91053247e-01 1.78359285e-01
3.59398037e-01 -7.20501184e-01 -5.21435738e-01 3.42339814e-01
6.71176672e-01 -6.21429563e-01 -5.01883268e-01 9.11814153e-01
2.56948680e-01 2.89195240e-01 -1.47928476e-01 -7.48585641e-01
-4.64345068e-01 2.21724406e-01 -8.58169854e-01 4.95532930e-01
6.16734564e-01 1.27979353e-01 6.47457242e-01 -1.07912242e+00
1.12386632e+00 1.13413644e+00 5.37289858e-01 4.84926552e-01
-9.26919222e-01 -6.48184657e-01 4.49990988e-01 5.08076727e-01
-1.37399542e+00 -4.64348465e-01 1.13916647e+00 -1.28619298e-01
6.72451019e-01 3.40252161e-01 8.48439038e-01 5.35496473e-01
4.69783366e-01 7.38450825e-01 8.07509303e-01 -3.76421511e-01
2.08488807e-01 -7.31474996e-01 -3.66409183e-01 6.51575804e-01
-2.88473040e-01 1.37916461e-01 -9.49572206e-01 -1.86847076e-01
1.26100004e+00 2.90192187e-01 -4.96218354e-01 -1.98642582e-01
-1.31915772e+00 5.90845704e-01 2.44331926e-01 4.08930480e-01
-7.21991956e-01 2.08164886e-01 3.09217900e-01 3.91882151e-01
7.92235076e-01 2.26201579e-01 -2.82950699e-01 -2.22585380e-01
-1.50626314e+00 3.10466915e-01 3.45904827e-01 1.01903796e+00
1.05304539e+00 -9.87867787e-02 -5.34038007e-01 9.32458043e-01
-2.50080861e-02 7.47318640e-02 2.67005235e-01 -1.42117655e+00
5.98489761e-01 4.23521757e-01 1.86114475e-01 -1.17963076e+00
1.39769778e-01 -1.08391769e-01 -7.24849284e-01 9.59559754e-02
-5.52582778e-02 -2.38930918e-02 -8.86208534e-01 1.50460076e+00
4.94463176e-01 9.41260040e-01 -1.72686860e-01 9.53571260e-01
6.58571839e-01 9.56849098e-01 -1.27732024e-01 -7.35454202e-01
1.01153731e+00 -1.35217917e+00 -9.43531394e-01 -2.46185094e-01
1.87574342e-01 -1.04727578e+00 6.49482667e-01 2.46587455e-01
-1.43708706e+00 -7.04312563e-01 -8.81885946e-01 -4.28699434e-01
2.13838220e-01 -2.26222321e-01 3.26274276e-01 -1.61871284e-01
-1.01690924e+00 6.52036428e-01 -7.16663718e-01 -2.14962326e-02
4.46775019e-01 1.10613205e-01 -5.02684951e-01 -5.88149130e-01
-9.67984200e-01 4.88251150e-01 3.34598154e-01 1.75665826e-01
-9.25407350e-01 -8.39997649e-01 -9.84187007e-01 3.51809375e-02
6.69718802e-01 -7.42404222e-01 9.54510152e-01 -1.14139795e+00
-1.19042027e+00 3.13924342e-01 -7.74256587e-01 -3.02937418e-01
3.64345700e-01 -4.90568280e-01 -2.34713495e-01 5.68560064e-01
2.59862572e-01 8.24815452e-01 1.18496513e+00 -1.13854599e+00
-9.14202273e-01 -1.30176708e-01 2.32972428e-02 4.79246378e-01
-1.02855429e-01 1.00622229e-01 -9.42675889e-01 -1.42914140e+00
3.75950664e-01 -4.66561705e-01 -2.52681524e-01 5.42415738e-01
1.10673457e-02 1.45713970e-01 1.50657308e+00 -9.69285786e-01
1.64525735e+00 -2.31876636e+00 4.22532976e-01 -5.23819476e-02
1.89757735e-01 3.67624730e-01 -2.78181583e-01 3.25608194e-01
-1.74136343e-03 -1.94302768e-01 -3.74727517e-01 -6.03830278e-01
-6.36478186e-01 5.91531515e-01 -5.23217618e-01 7.54265636e-02
1.15053862e-01 7.73852646e-01 -1.18851948e+00 -6.43844366e-01
5.42973578e-01 6.61808252e-01 -6.04658306e-01 3.41136098e-01
-2.90629208e-01 6.72509074e-01 -2.97855794e-01 6.70923412e-01
7.41928577e-01 1.53357208e-01 5.60553037e-02 -2.84994036e-01
-3.24457109e-01 -2.15072073e-02 -9.24675465e-01 2.25226879e+00
-4.90240157e-01 8.40522408e-01 2.31320098e-01 -6.38772011e-01
7.01680481e-01 7.14201778e-02 6.75969720e-01 -5.92858613e-01
-1.73759863e-01 8.68434533e-02 -5.40031612e-01 -5.03579259e-01
8.36548507e-01 4.86272126e-02 3.68012875e-01 2.21298054e-01
-2.26824015e-01 3.01081967e-02 4.03723001e-01 1.95382103e-01
9.67451274e-01 3.60835016e-01 3.33056152e-01 1.21148331e-02
6.42671883e-01 -4.20937449e-01 9.19662058e-01 4.45434541e-01
-1.29532203e-01 1.21082914e+00 2.90943146e-01 -7.45222926e-01
-8.76109183e-01 -6.53344214e-01 3.63576263e-01 9.21190381e-01
5.44031799e-01 -8.43915462e-01 -7.43427634e-01 -6.23290062e-01
-2.02151880e-01 3.66062015e-01 -6.48977757e-01 -8.22319314e-02
-1.07011938e+00 -1.48595080e-01 -5.52966371e-02 5.27021348e-01
6.81911707e-01 -1.15052712e+00 -6.65538251e-01 6.95120573e-01
-5.44551492e-01 -1.08780098e+00 -1.00640500e+00 -2.88893729e-01
-1.29577374e+00 -9.22731757e-01 -9.72084403e-01 -8.77047122e-01
6.73035264e-01 5.56213558e-01 1.01099730e+00 5.62187791e-01
-1.25398323e-01 -1.02469854e-01 -5.33844829e-01 3.34594578e-01
-3.48543674e-01 -2.66281873e-01 -4.48513299e-01 8.39535519e-02
-9.24038067e-02 -8.01809847e-01 -8.18310916e-01 4.39086825e-01
-1.21670008e+00 3.10137331e-01 4.42811847e-01 9.29089785e-01
1.13452280e+00 4.50087618e-03 3.45871955e-01 -6.44339025e-01
2.52868205e-01 -2.26452053e-01 -4.29145336e-01 3.27305317e-01
-1.72674701e-01 -5.83266616e-02 5.55808663e-01 -3.70006055e-01
-1.07558584e+00 -5.78347258e-02 -2.34102547e-01 -8.51131499e-01
3.23431417e-02 3.19979131e-01 -1.81727976e-01 -1.24746479e-01
1.31110743e-01 3.36907208e-01 -2.24532098e-01 -6.43891573e-01
1.41532794e-01 1.37501612e-01 6.73884869e-01 -5.42381108e-01
5.14315426e-01 6.39448524e-01 -2.99833640e-02 -7.12315023e-01
-8.34768832e-01 -3.31523985e-01 -4.88138795e-01 -2.74218678e-01
6.95065558e-01 -9.91961718e-01 5.31332158e-02 5.11399150e-01
-1.27183473e+00 -3.47471237e-01 -5.57092845e-01 2.46888772e-01
-5.87106526e-01 8.38520169e-01 -7.77747631e-01 -2.39767149e-01
-2.82661080e-01 -1.16595137e+00 1.22964692e+00 2.79205620e-01
-2.07639634e-04 -7.97224224e-01 -1.18805818e-01 7.31995702e-02
3.27322990e-01 3.90372753e-01 5.17110407e-01 3.83188099e-01
-8.94409359e-01 3.03890049e-01 -2.25462973e-01 3.16047490e-01
5.22233129e-01 3.67787806e-03 -4.42507327e-01 -3.26098561e-01
-4.26213630e-02 2.58388460e-01 1.01185751e+00 6.26771986e-01
1.30099678e+00 -3.25107902e-01 -3.92071962e-01 8.66603255e-01
1.29058528e+00 2.51250595e-01 9.91590261e-01 2.37412930e-01
9.35766220e-01 4.00210798e-01 1.20939755e+00 6.25432670e-01
1.37002826e-01 7.44819701e-01 3.29149067e-01 6.38130009e-02
-5.58559716e-01 -2.10161418e-01 3.37874025e-01 7.48843849e-01
-9.79292765e-02 -1.59431830e-01 -4.13134724e-01 6.33371890e-01
-2.07472491e+00 -1.16667736e+00 1.20630890e-01 2.15522242e+00
1.05241489e+00 4.39296924e-02 -1.42414331e-01 -8.71918648e-02
1.08852839e+00 5.27677655e-01 -3.55810642e-01 -2.14761004e-01
-1.52070031e-01 1.47997752e-01 1.89591661e-01 8.53573859e-01
-9.98431861e-01 1.01732874e+00 6.06628370e+00 1.10606349e+00
-1.06084001e+00 2.15854988e-01 7.01341569e-01 -3.38081896e-01
-3.10057521e-01 3.09802860e-01 -4.28038329e-01 6.23363137e-01
2.91944385e-01 5.93952537e-02 4.08806205e-01 5.42360127e-01
4.15510297e-01 -5.39557159e-01 -9.55183208e-01 1.05780637e+00
2.77700871e-01 -1.66497302e+00 2.28924915e-01 1.75019670e-02
1.10773420e+00 -3.41968387e-01 -2.66786247e-01 -1.82127729e-01
-5.69584310e-01 -5.29498100e-01 8.52500558e-01 6.73074067e-01
8.56691241e-01 -1.13159966e+00 4.95669723e-01 1.24095991e-01
-1.54461491e+00 -1.47330627e-01 -2.86760151e-01 -1.12920374e-01
5.44552505e-01 9.60721910e-01 -2.07240164e-01 6.34444714e-01
5.51739514e-01 1.09949648e+00 -4.53320920e-01 1.17508197e+00
-2.76727647e-01 3.81759763e-01 -2.80418873e-01 7.03898847e-01
2.20973417e-01 -2.17206657e-01 7.33815730e-01 1.05364013e+00
5.54500759e-01 3.97338957e-01 2.59068668e-01 6.36094868e-01
-1.10711612e-01 -1.30896538e-01 -3.95164371e-01 3.64337206e-01
4.14340675e-01 9.42621171e-01 -7.78692067e-01 -6.77335620e-01
-4.25303817e-01 1.35280120e+00 -5.04533686e-02 3.80473375e-01
-9.55166638e-01 -4.08676237e-01 6.75399065e-01 1.72719911e-01
6.55655742e-01 -3.26394796e-01 -4.14098836e-02 -1.25367236e+00
3.88897568e-01 -5.59126616e-01 2.16530174e-01 -8.87841940e-01
-8.20435464e-01 5.24241567e-01 4.94029820e-02 -1.42547905e+00
-3.24096113e-01 6.63849190e-02 -5.91279328e-01 6.34986997e-01
-1.73986006e+00 -1.15097392e+00 -5.27002811e-01 6.68953300e-01
1.17428195e+00 1.39309436e-01 2.49632031e-01 4.76475388e-01
-2.26566657e-01 2.63493776e-01 4.58576679e-02 -6.68674633e-02
9.04030919e-01 -6.73581004e-01 3.81242871e-01 1.18944001e+00
-3.58233415e-02 4.81467903e-01 6.26497209e-01 -1.03986406e+00
-1.12791288e+00 -1.32978714e+00 9.57607985e-01 2.26786807e-01
2.79783159e-02 -3.93985622e-02 -9.69434142e-01 5.34785569e-01
1.38358802e-01 5.48127174e-01 1.90176591e-01 -5.59446454e-01
-1.11387163e-01 -2.99826354e-01 -1.23826611e+00 6.14915431e-01
1.23533535e+00 -5.13754129e-01 -4.88881618e-01 7.29466081e-02
8.01349759e-01 -5.90758681e-01 -5.83467782e-01 2.92337149e-01
6.25483394e-01 -1.22665882e+00 1.01681745e+00 -4.55009758e-05
6.32133126e-01 -6.05008841e-01 -1.25309139e-01 -1.01024997e+00
-3.23984265e-01 -9.74546313e-01 -6.76087499e-01 1.20012403e+00
-2.60334641e-01 -1.86319441e-01 8.07196975e-01 2.44893909e-01
-3.82509619e-01 -8.57519150e-01 -1.03074551e+00 -5.34046769e-01
-4.39610273e-01 -2.83831805e-01 5.68197548e-01 8.44145775e-01
-3.84699851e-01 -1.95339799e-01 -6.04615808e-01 -2.46667098e-02
5.04544497e-01 4.68341649e-01 4.74863440e-01 -8.15096498e-01
-1.79154381e-01 -9.11626965e-02 -3.95147592e-01 -1.38730001e+00
7.63834342e-02 -1.67943671e-01 2.91719943e-01 -1.53360021e+00
1.57668769e-01 -3.08847278e-01 -2.61624902e-01 4.43848789e-01
-4.69359875e-01 5.88476300e-01 3.36581320e-01 1.07960112e-01
-7.02399909e-01 5.46333849e-01 1.41009486e+00 -2.91766245e-02
-3.17925096e-01 -4.18658406e-01 -3.99903893e-01 7.05516458e-01
6.47386849e-01 -5.54589152e-01 -4.29224193e-01 -6.80027127e-01
-2.02708822e-02 4.35859323e-01 3.90596449e-01 -1.17380166e+00
1.77905202e-01 -4.65215176e-01 5.09517074e-01 -8.97750854e-01
4.87370789e-01 -8.46189141e-01 3.20047855e-01 7.61750117e-02
-1.61660329e-01 3.55889708e-01 1.11583941e-01 7.03232765e-01
-6.04473233e-01 -3.31029266e-01 7.60206878e-01 -2.03538805e-01
-1.05869830e+00 5.02026558e-01 -2.68691033e-01 -2.67467499e-02
1.16102183e+00 -3.50666314e-01 1.57991335e-01 -4.56968874e-01
-7.39985168e-01 1.10804677e-01 8.74852777e-01 4.66981947e-01
1.11635065e+00 -1.40019512e+00 -5.93338609e-01 3.68539602e-01
-1.65477753e-01 3.19146186e-01 6.12899542e-01 6.81178808e-01
-7.55306244e-01 -5.99761121e-02 -3.10771853e-01 -6.39776528e-01
-1.55626500e+00 6.59722447e-01 -9.80855152e-02 -2.35110849e-01
-8.16110730e-01 8.99644315e-01 4.00919944e-01 5.17448962e-01
5.69964685e-02 -5.71160614e-01 1.30261287e-01 -6.15814626e-02
9.99922633e-01 3.39621454e-01 -1.13297470e-01 -8.74263644e-01
-1.44954145e-01 9.11569715e-01 -2.40686715e-01 -9.59254354e-02
1.22620118e+00 -5.47850728e-01 -3.86174321e-01 9.01790261e-02
1.19669032e+00 1.40779555e-01 -1.82363713e+00 -1.18398532e-01
-2.64073461e-01 -1.21327507e+00 1.33715704e-01 -5.62938869e-01
-1.50046384e+00 5.96245587e-01 4.35230523e-01 -2.88955450e-01
1.66247106e+00 -2.47214451e-01 1.36322916e+00 1.88271590e-02
5.05753040e-01 -1.17587471e+00 1.68552980e-01 4.21867847e-01
1.06920755e+00 -7.94874847e-01 3.91562641e-01 -7.78488398e-01
-2.21410066e-01 1.22096884e+00 6.19433641e-01 -2.62429893e-01
3.55378866e-01 3.33337337e-01 -9.44794640e-02 1.31255105e-01
-7.54491568e-01 9.16270688e-02 1.08620703e-01 5.06761312e-01
2.82044917e-01 -3.97206366e-01 -4.76090282e-01 -9.95495766e-02
4.15948600e-01 4.11231995e-01 4.64343518e-01 1.43801832e+00
-4.30018365e-01 -1.38672101e+00 -5.47791958e-01 2.16894358e-01
-4.23338383e-01 -2.52825379e-01 -5.58937415e-02 4.08832639e-01
4.17689979e-01 8.74264777e-01 1.25687271e-01 -5.73500991e-02
1.05232894e-01 -1.65669695e-01 6.02100134e-01 -5.33548176e-01
-5.67230046e-01 4.99850690e-01 -1.16808608e-01 -1.03367162e+00
-7.48664081e-01 -6.07490182e-01 -1.23202968e+00 -1.65809676e-01
-2.75848895e-01 2.64442917e-02 2.34586090e-01 7.31370389e-01
7.97223091e-01 4.64185089e-01 5.50181091e-01 -1.23475432e+00
3.47391129e-01 -7.65664637e-01 -3.51771802e-01 4.44008321e-01
7.27643371e-01 -5.78427911e-01 -6.88670948e-02 4.23022598e-01] | [10.815571784973145, -1.3959252834320068] |
0918ce95-0fa2-42f3-a0b5-a9b4f8be85bb | broaden-your-views-for-self-supervised-video | 2103.16559 | null | https://arxiv.org/abs/2103.16559v3 | https://arxiv.org/pdf/2103.16559v3.pdf | Broaden Your Views for Self-Supervised Video Learning | Most successful self-supervised learning methods are trained to align the representations of two independent views from the data. State-of-the-art methods in video are inspired by image techniques, where these two views are similarly extracted by cropping and augmenting the resulting crop. However, these methods miss a crucial element in the video domain: time. We introduce BraVe, a self-supervised learning framework for video. In BraVe, one of the views has access to a narrow temporal window of the video while the other view has a broad access to the video content. Our models learn to generalise from the narrow view to the general content of the video. Furthermore, BraVe processes the views with different backbones, enabling the use of alternative augmentations or modalities into the broad view such as optical flow, randomly convolved RGB frames, audio or their combinations. We demonstrate that BraVe achieves state-of-the-art results in self-supervised representation learning on standard video and audio classification benchmarks including UCF101, HMDB51, Kinetics, ESC-50 and AudioSet. | ['Corentin Tallec', 'Florian Strub', 'Ross Hemsley', 'Andrew Zisserman', 'Aäron van den Oord', 'Jean-bastien Grill', 'Michal Valko', 'Florent Altché', 'Viorica Patraucean', 'Mateusz Malinowski', 'Luyu Wang', 'Jean-Baptiste Alayrac', 'Pauline Luc', 'Adrià Recasens'] | 2021-03-30 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Recasens_Broaden_Your_Views_for_Self-Supervised_Video_Learning_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Recasens_Broaden_Your_Views_for_Self-Supervised_Video_Learning_ICCV_2021_paper.pdf | iccv-2021-1 | ['self-supervised-action-recognition'] | ['computer-vision'] | [ 2.44530931e-01 -1.46128535e-01 -3.72880071e-01 -1.23345166e-01
-5.30049086e-01 -6.91153586e-01 6.55076206e-01 -2.53887504e-01
-2.28127599e-01 4.52984124e-01 4.71246958e-01 2.86644340e-01
1.30075961e-01 -5.25150657e-01 -9.31353748e-01 -8.01470935e-01
-2.04579502e-01 1.68864951e-01 2.29468092e-01 -1.16979562e-01
-1.14522249e-01 1.16972081e-01 -1.93209589e+00 7.85528541e-01
2.31041424e-02 1.00896060e+00 -3.83940563e-02 1.03153217e+00
-4.54693623e-02 1.21601093e+00 -2.16576159e-01 -2.78173629e-02
3.81695628e-01 -4.99593645e-01 -9.25340414e-01 4.85489309e-01
9.74746227e-01 -4.30576533e-01 -8.72358263e-01 7.13020921e-01
2.26049960e-01 8.45409557e-02 5.00437796e-01 -1.67275655e+00
-3.88510883e-01 3.78626972e-01 -6.38391376e-01 4.41729844e-01
6.39368713e-01 -1.34353954e-02 9.48922455e-01 -8.76613736e-01
1.00775576e+00 1.20457661e+00 4.03728694e-01 5.97327709e-01
-1.22554338e+00 -5.81810594e-01 3.98533762e-01 4.49786633e-01
-9.23446357e-01 -7.35856235e-01 8.25062573e-01 -6.34222269e-01
8.17397714e-01 8.24818090e-02 8.87595534e-01 1.45503438e+00
-2.01248154e-01 9.95806396e-01 1.25088644e+00 -4.12799776e-01
1.00860968e-01 4.71238270e-02 -1.99899413e-02 6.97568238e-01
-2.90111989e-01 5.15705384e-02 -1.09599030e+00 5.73364086e-02
9.18077826e-01 1.56146690e-01 -5.14765263e-01 -9.18812811e-01
-1.43027854e+00 5.60444593e-01 1.81245804e-01 9.48756114e-02
-2.04475090e-01 1.01103537e-01 6.70113206e-01 5.88005543e-01
4.89127010e-01 3.47775430e-03 -4.92678881e-01 -3.04094970e-01
-1.10375619e+00 -1.79444458e-02 7.15590477e-01 9.16322351e-01
7.60994136e-01 1.05516903e-01 -9.23651736e-03 5.07846534e-01
2.55131871e-01 3.15940291e-01 6.65567696e-01 -1.19042611e+00
6.07470155e-01 2.02470735e-01 -1.58584371e-01 -5.17182529e-01
-8.07993859e-02 -2.32940465e-01 -7.51960695e-01 2.95807987e-01
3.52569968e-01 5.29722311e-02 -1.08120823e+00 1.84386158e+00
2.60264307e-01 6.22133076e-01 4.10473436e-01 6.37256026e-01
1.00603664e+00 7.02452898e-01 -1.89211935e-01 -3.24887842e-01
1.13808715e+00 -1.33866966e+00 -6.17821038e-01 -1.86573312e-01
2.69879818e-01 -5.32487392e-01 7.46387601e-01 6.64759457e-01
-1.26753557e+00 -8.71916890e-01 -1.11675823e+00 -5.49101755e-02
-3.58317494e-01 6.09542318e-02 3.83440942e-01 2.06038520e-01
-1.11157060e+00 6.69965982e-01 -8.62208843e-01 -4.97266799e-01
5.49111843e-01 1.91538081e-01 -1.02069795e+00 -2.33811811e-01
-9.39314127e-01 5.07653713e-01 2.98846573e-01 -3.38676363e-01
-1.11728847e+00 -7.29500294e-01 -1.11984015e+00 -8.28844607e-02
3.49087566e-01 -6.75599456e-01 9.71197605e-01 -1.32154596e+00
-1.36263871e+00 1.13556981e+00 -1.34761572e-01 -5.78260064e-01
4.76545781e-01 -6.85150445e-01 -4.51332539e-01 7.09760487e-01
1.00726582e-01 1.03209329e+00 1.62493789e+00 -1.19325650e+00
-7.07126260e-01 -3.98032844e-01 2.84154207e-01 3.40110749e-01
-4.44941372e-01 -2.05380321e-01 -6.74526155e-01 -6.57457948e-01
-4.99012731e-02 -1.04208851e+00 2.74639040e-01 1.18120037e-01
-6.44701645e-02 1.52668551e-01 1.37067592e+00 -5.34182906e-01
1.07502997e+00 -2.45355201e+00 8.62444639e-01 -2.42745802e-01
2.68096566e-01 1.43896177e-01 -2.89886028e-01 5.11680901e-01
-6.73503101e-01 -2.55356908e-01 7.26808459e-02 -5.09128094e-01
-4.55389082e-01 3.57208043e-01 -4.63972390e-01 6.81555331e-01
7.68699124e-02 6.02274537e-01 -1.13738906e+00 -4.65995312e-01
4.12068814e-01 5.37995219e-01 -6.46319389e-01 5.48148513e-01
9.96862948e-02 6.09490693e-01 1.59025759e-01 4.77427840e-01
3.85153711e-01 -3.01896036e-01 7.20943436e-02 -5.71735024e-01
1.04702704e-01 1.76249191e-01 -1.16268313e+00 2.32692266e+00
-3.89623135e-01 8.16507697e-01 1.45202186e-02 -1.23756087e+00
4.59906697e-01 5.52611351e-01 8.64595890e-01 -3.39684576e-01
-1.63976252e-01 -7.43699074e-02 -2.91528881e-01 -7.82278538e-01
1.67911753e-01 6.33305237e-02 2.38342464e-01 5.33601224e-01
9.31152225e-01 -7.33805299e-02 3.32598716e-01 4.12002832e-01
1.01757538e+00 8.23578596e-01 4.37903374e-01 1.85325280e-01
5.87126493e-01 -3.41465384e-01 1.60246760e-01 3.32264096e-01
-2.12594554e-01 8.38792682e-01 3.67479354e-01 -4.89002913e-01
-1.05712426e+00 -1.13149476e+00 2.55447686e-01 1.35221589e+00
-1.03020407e-01 -7.95333743e-01 -7.30562508e-01 -9.84222114e-01
-3.38211060e-02 1.41171142e-01 -9.58073676e-01 -9.51567739e-02
-5.00070214e-01 -1.94690689e-01 2.65101612e-01 6.23595595e-01
4.16859031e-01 -9.23025608e-01 -8.91496003e-01 -4.70236642e-03
-4.40517634e-01 -1.56437373e+00 -3.41917306e-01 5.00076771e-01
-9.79222178e-01 -1.19295752e+00 -7.30021834e-01 -6.83433115e-01
5.12516022e-01 5.29262960e-01 1.08985209e+00 -1.47248387e-01
-2.05281451e-01 1.15325367e+00 -5.32527447e-01 -1.44591093e-01
-2.68308103e-01 -1.29759535e-01 1.81820408e-01 3.84211928e-01
2.86483392e-02 -8.90353739e-01 -4.19838965e-01 1.52480632e-01
-9.22979474e-01 1.35788247e-01 1.46702796e-01 8.99916053e-01
5.16471684e-01 -3.33162546e-01 1.70769438e-01 -7.58865476e-01
-2.87295789e-01 -5.76001644e-01 -8.80573913e-02 1.06884360e-01
-3.79540958e-02 1.20074622e-01 5.32038331e-01 -6.33983970e-01
-7.55746186e-01 4.32151377e-01 2.47913957e-01 -1.15840411e+00
-3.11871618e-01 2.24861383e-01 -2.93836407e-02 1.67836443e-01
5.85617244e-01 1.84498176e-01 2.84010947e-01 -3.08404058e-01
7.00287998e-01 4.40382630e-01 1.02338612e+00 -3.80622178e-01
8.37162852e-01 1.00595415e+00 -4.81010005e-02 -9.15068984e-01
-8.93728793e-01 -8.79398704e-01 -1.16378975e+00 -4.65522289e-01
8.90239239e-01 -1.22472286e+00 -1.50656059e-01 4.32766199e-01
-8.91180575e-01 -2.10845217e-01 -6.98797822e-01 5.85312605e-01
-1.15912139e+00 4.80707258e-01 -4.15095210e-01 -4.59959060e-01
-1.29063904e-01 -1.09247828e+00 1.31561255e+00 3.24799865e-02
-3.69187444e-01 -9.58754122e-01 2.05693960e-01 4.51611489e-01
3.96850556e-02 2.17095226e-01 4.40202087e-01 -6.76103771e-01
-2.61184126e-01 5.29499836e-02 1.16447665e-01 5.16678274e-01
4.16868389e-01 8.25676173e-02 -1.45192277e+00 -5.34432232e-01
4.11083475e-02 -8.21175933e-01 1.22637880e+00 2.56952196e-01
1.08209062e+00 -1.79725677e-01 -2.20189705e-01 9.37458813e-01
1.33955526e+00 1.97085701e-02 6.71507180e-01 4.46388632e-01
7.78155982e-01 6.22172415e-01 4.82480764e-01 4.33192402e-01
1.95103005e-01 6.09618306e-01 6.80564463e-01 -1.20073631e-01
-3.01624715e-01 -3.07252288e-01 7.12435365e-01 8.57308567e-01
-3.24877441e-01 -1.09753259e-01 -3.94164950e-01 5.51905632e-01
-1.99840140e+00 -1.31471992e+00 4.22865570e-01 2.17754221e+00
6.21387482e-01 6.91324174e-02 4.30605590e-01 5.59163988e-01
5.09208143e-01 5.04301131e-01 -5.39262712e-01 9.19150338e-02
-1.96930006e-01 2.12787002e-01 2.12242469e-01 1.02982938e-01
-1.63637698e+00 6.08233511e-01 6.26796770e+00 3.65148157e-01
-1.25245690e+00 1.27039820e-01 3.47744882e-01 -4.80722845e-01
1.79661781e-01 -2.09139474e-02 -4.60460722e-01 2.24563956e-01
9.92067933e-01 2.54859835e-01 5.10603249e-01 6.66034341e-01
-7.25403205e-02 2.82147583e-02 -1.71748269e+00 1.34244013e+00
4.93976742e-01 -1.37612724e+00 2.39309371e-02 -4.92569245e-02
7.17544854e-01 1.01823255e-01 1.21143863e-01 1.86886460e-01
5.76325171e-02 -9.96905982e-01 9.10903633e-01 6.18390143e-01
1.00524259e+00 -3.44581038e-01 4.36023474e-01 8.97224024e-02
-1.37931132e+00 -2.73470283e-01 7.08985317e-04 9.38870832e-02
1.11375980e-01 -3.96032855e-02 -2.64904380e-01 6.54268801e-01
1.15645635e+00 1.45059812e+00 -5.65346658e-01 7.33653963e-01
2.56333388e-02 4.49225724e-01 -1.17451914e-01 8.59595716e-01
2.32165590e-01 -7.09191561e-02 5.30419409e-01 1.28480113e+00
2.25605264e-01 -1.60932645e-01 2.48493284e-01 2.57162452e-01
-3.98115925e-02 -1.82901248e-01 -8.01391900e-01 -1.71911523e-01
1.51928782e-01 1.28547788e+00 -5.54371595e-01 -6.76637948e-01
-8.78137648e-01 1.17338836e+00 2.95613468e-01 4.82957244e-01
-8.36935818e-01 -2.26404928e-02 6.05270684e-01 1.49913043e-01
7.31444895e-01 -4.86857072e-02 4.93333995e-01 -1.43427515e+00
-1.61886394e-01 -1.14363146e+00 6.10838115e-01 -1.07571781e+00
-1.08168089e+00 4.45272177e-01 1.72017589e-01 -1.83307469e+00
-4.26819444e-01 -6.23593390e-01 -2.91972011e-01 4.16656017e-01
-1.54610729e+00 -1.28998542e+00 -4.55373526e-01 1.05164158e+00
9.46389496e-01 -5.83935082e-01 8.29601765e-01 1.79452866e-01
-2.61392027e-01 1.93227559e-01 -5.42193763e-02 2.60940701e-01
1.09228802e+00 -1.28057432e+00 1.32962480e-01 4.96977150e-01
8.08242500e-01 2.79098183e-01 4.82497483e-01 -2.45948061e-01
-1.60550320e+00 -8.87109697e-01 3.13724905e-01 -4.04192775e-01
7.20406771e-01 -2.66484201e-01 -7.51886845e-01 8.99556100e-01
6.37718201e-01 6.43697500e-01 7.21626163e-01 -2.75560886e-01
-8.83372784e-01 -1.83536217e-01 -7.48191535e-01 2.70204574e-01
1.09206557e+00 -1.02791774e+00 -7.11823344e-01 2.72622347e-01
5.97700238e-01 -5.98935008e-01 -7.80503273e-01 2.92402536e-01
8.65015388e-01 -1.21170211e+00 1.22307384e+00 -1.08427930e+00
7.18337476e-01 -2.29290307e-01 -4.81094986e-01 -1.41424370e+00
-7.60525689e-02 -7.38145351e-01 -8.08685005e-01 1.13183093e+00
-9.69976410e-02 3.19911470e-03 7.46995568e-01 -2.39556015e-01
8.08512792e-02 -6.14066362e-01 -8.88067842e-01 -6.31179810e-01
-2.86857635e-01 -3.13520789e-01 1.48339987e-01 1.03067553e+00
2.47970462e-01 4.41709101e-01 -6.60788357e-01 -5.87446801e-02
6.16199017e-01 2.75732633e-02 9.30044889e-01 -1.22405481e+00
-4.66827780e-01 -1.00491479e-01 -6.41282499e-01 -1.03075993e+00
3.00253093e-01 -7.36644745e-01 -2.58954704e-01 -1.33472073e+00
2.98931837e-01 2.66909629e-01 -3.75734031e-01 4.12090868e-01
1.25343576e-01 3.60126197e-01 4.29425061e-01 3.83215189e-01
-7.22280324e-01 2.90302455e-01 1.09372520e+00 -3.36489737e-01
-9.47516635e-02 -2.57510066e-01 -3.57257396e-01 9.10413623e-01
4.30314511e-01 -1.44355863e-01 -7.00696588e-01 -3.93162668e-01
6.85268641e-02 2.66987532e-01 4.21315312e-01 -1.08988380e+00
2.26547673e-01 1.25880599e-01 5.57918489e-01 -6.11882448e-01
6.42361045e-01 -9.92950559e-01 1.24864750e-01 2.36370116e-01
-4.41177368e-01 2.69052595e-01 1.65758595e-01 8.12045693e-01
-4.83089626e-01 -7.46371970e-02 6.75495744e-01 -3.97717923e-01
-8.13105226e-01 2.93903083e-01 -2.81037003e-01 1.68263316e-01
1.00510252e+00 -2.89394289e-01 -2.73632228e-01 -6.95796192e-01
-1.11481643e+00 -1.78400502e-02 4.40601796e-01 6.31854832e-01
6.36110604e-01 -1.33299088e+00 -4.26759332e-01 3.66736382e-01
1.42822996e-01 -8.93827081e-02 2.84607559e-01 8.71715367e-01
-2.80523270e-01 3.08659464e-01 -6.83008015e-01 -9.33530748e-01
-1.49103451e+00 9.20682311e-01 1.49188280e-01 -1.49044380e-01
-6.46438599e-01 7.28316844e-01 3.96478623e-01 -1.15649989e-02
5.36761522e-01 -2.72471339e-01 -5.16277790e-01 6.23168647e-01
5.76720238e-01 3.55539083e-01 -6.93840310e-02 -1.05185008e+00
-1.89661339e-01 8.17159593e-01 9.55837499e-03 -2.40593657e-01
1.45182157e+00 -2.46996999e-01 1.19093671e-01 8.82978559e-01
1.41992640e+00 -6.36597201e-02 -1.67519891e+00 -5.39140582e-01
-3.92377406e-01 -3.04687947e-01 -2.86606047e-02 -2.64174372e-01
-1.34198725e+00 1.04947758e+00 5.70036411e-01 1.72911361e-01
1.32312465e+00 4.95176911e-02 4.80617255e-01 1.90814421e-01
1.68329939e-01 -9.85700130e-01 5.66163242e-01 4.07815754e-01
8.11269581e-01 -1.25261021e+00 2.64565736e-01 -3.04050505e-01
-7.50475109e-01 1.45076931e+00 5.51336646e-01 -1.18433587e-01
6.76658034e-01 2.94998854e-01 3.10394075e-02 -1.54359907e-01
-1.02601683e+00 -3.01767379e-01 3.01126450e-01 8.78920496e-01
4.48378116e-01 -4.76719916e-01 4.90220994e-01 1.32646918e-01
-1.22491784e-01 1.80658828e-02 6.38097823e-01 1.13263714e+00
1.38049414e-02 -8.67820024e-01 -3.93313527e-01 2.56365925e-01
-4.01237071e-01 3.58515116e-03 -2.70332038e-01 9.32565987e-01
4.78442833e-02 6.72215223e-01 3.76088232e-01 -2.86323249e-01
6.14241697e-02 4.31117564e-01 8.83134782e-01 -6.97035909e-01
-2.28247821e-01 4.59358692e-01 -4.41972353e-02 -7.86554217e-01
-1.31514227e+00 -8.20874274e-01 -9.67720091e-01 2.91719526e-01
5.29854931e-02 -4.41932864e-02 3.14989597e-01 9.94555593e-01
1.50073484e-01 4.53503102e-01 6.33749723e-01 -1.35558736e+00
-1.79042950e-01 -8.27389359e-01 -4.94903833e-01 7.79044330e-01
8.80342960e-01 -8.44913840e-01 -4.04644907e-01 7.64584601e-01] | [9.322710990905762, 0.8865151405334473] |
29f32eb7-f100-4f9f-90d4-01712305292c | towards-linked-hypernyms-dataset-20 | null | null | https://aclanthology.org/L14-1552 | https://aclanthology.org/L14-1552.pdf | Towards Linked Hypernyms Dataset 2.0: complementing DBpedia with hypernym discovery | This paper presents a statistical type inference algorithm for ontology alignment, which assigns DBpedia entities with a new type (class). To infer types for a specific entity, the algorithm first identifies types that co-occur with the type the entity already has, and subsequently prunes the set of candidates for the most confident one. The algorithm has one parameter for balancing specificity/reliability of the resulting type selection. The proposed algorithm is used to complement the types in the LHD dataset, which is RDF knowledge base populated by identifying hypernyms from the free text of Wikipedia articles. The majority of types assigned to entities in LHD 1.0 are DBpedia resources. Through the statistical type inference, the number of entities with a type from DBpedia Ontology is increased significantly: by 750 thousand entities for the English dataset, 200.000 for Dutch and 440.000 for German. The accuracy of the inferred types is at 0.65 for English (as compared to 0.86 for LHD 1.0 types). A byproduct of the mapping process is a set of 11.000 mappings from DBpedia resources to DBpedia Ontology classes with associated confidence values. The number of the resulting mappings is an order of magnitude larger than what can be achieved with standard ontology alignment algorithms (Falcon, LogMapLt and YAM++), which do not utilize the type co-occurrence information. The presented algorithm is not restricted to the LHD dataset, it can be used to address generic type inference problems in presence of class membership information for a large number of instances. | ['Ond{\\v{r}}ej Zamazal', "Tom{\\'a}{\\v{s}} Kliegr"] | 2014-05-01 | null | null | null | lrec-2014-5 | ['hypernym-discovery'] | ['natural-language-processing'] | [-1.40924662e-01 7.46929049e-01 -2.60930151e-01 -2.71258384e-01
-9.72146019e-02 -6.45236254e-01 6.18782282e-01 8.93610537e-01
-8.26602817e-01 1.52303290e+00 -1.49215013e-01 -1.62100583e-01
-4.80555862e-01 -1.61689854e+00 -8.22554290e-01 -3.03422719e-01
-9.38549414e-02 1.23366153e+00 5.09150684e-01 -2.81752855e-01
1.37327621e-02 3.23668361e-01 -2.23792934e+00 3.34733546e-01
1.04869664e+00 1.05258512e+00 3.32293183e-01 -1.81361027e-02
-6.48174942e-01 5.13303339e-01 -5.85395217e-01 -5.91068923e-01
2.66407225e-02 1.25419006e-01 -1.06210077e+00 -6.16799593e-01
3.74505818e-01 1.72679558e-01 2.02155173e-01 1.24159896e+00
3.86352926e-01 -6.90460876e-02 7.83634603e-01 -1.36708355e+00
-2.32636511e-01 6.09322071e-01 -5.69866151e-02 -1.08800735e-02
8.31343055e-01 -7.07418144e-01 1.04200113e+00 -9.77993071e-01
1.16025269e+00 1.13438761e+00 6.57201767e-01 2.18847185e-01
-8.88493955e-01 -4.73710895e-01 -4.18471813e-01 1.87787786e-01
-1.62152946e+00 -1.41815305e-01 -1.03687895e-02 -6.28090024e-01
1.11645508e+00 3.08215857e-01 6.08975530e-01 4.96827513e-01
-1.98101416e-01 -1.01461388e-01 1.11615920e+00 -8.38933051e-01
3.63255858e-01 8.73375416e-01 4.10642296e-01 4.66298938e-01
1.20468271e+00 -1.87398687e-01 -3.25840443e-01 -3.62262011e-01
2.08618402e-01 -6.41206205e-01 1.50615975e-01 -3.48013878e-01
-9.47749376e-01 5.46957493e-01 3.75096165e-02 3.50170076e-01
-4.86000925e-01 -5.01916230e-01 3.98872703e-01 1.61795512e-01
1.24929003e-01 5.30264139e-01 -7.20287979e-01 4.85857055e-02
-5.24216056e-01 7.31224418e-01 1.12754738e+00 1.65649796e+00
1.22222722e+00 -3.96432489e-01 3.58872771e-01 9.63495016e-01
3.83926362e-01 6.09030664e-01 2.68402934e-01 -5.42668223e-01
3.67493540e-01 1.25083673e+00 5.25447726e-01 -9.09534574e-01
-4.93615448e-01 1.27352804e-01 -3.87359172e-01 1.43748326e-02
5.64543903e-01 -1.25262171e-01 -6.68665588e-01 1.60811460e+00
7.22126365e-01 -2.93465286e-01 5.57021320e-01 3.15696537e-01
1.09886813e+00 2.76658058e-01 3.00148159e-01 -2.88555205e-01
1.84963524e+00 4.59800363e-02 -7.51874983e-01 6.10337965e-02
6.76235139e-01 -7.18514621e-01 4.06119823e-01 6.29725158e-02
-7.91602731e-01 -3.86367619e-01 -8.98636639e-01 2.03451619e-01
-1.15741241e+00 -2.25995824e-01 5.13615251e-01 7.87960947e-01
-5.42920649e-01 3.12933415e-01 -1.55707955e-01 -8.88554335e-01
6.45385357e-03 4.67157543e-01 -6.00652933e-01 1.03917234e-01
-1.82084322e+00 1.32197547e+00 1.35061049e+00 -2.88730025e-01
-2.13282350e-02 -7.66731620e-01 -6.16421163e-01 -1.84824511e-01
4.03516918e-01 -7.37816691e-01 5.36556304e-01 -5.77930093e-01
-5.20198464e-01 1.04661250e+00 -1.29821748e-01 -5.77038169e-01
2.94729203e-01 1.16957098e-01 -1.05116248e+00 -2.31622532e-02
5.09506822e-01 3.31028014e-01 1.48423566e-02 -1.31408823e+00
-1.30525279e+00 -3.12482387e-01 3.55752975e-01 9.82498750e-02
-2.58159876e-01 -9.29256976e-02 -2.09978878e-01 -3.39988768e-01
-1.64960530e-02 -7.46508002e-01 1.43629581e-01 -5.25748193e-01
-7.58925676e-02 -3.98908556e-01 1.96057603e-01 -5.43917418e-01
1.51504290e+00 -1.60448110e+00 -3.50596495e-02 7.20570505e-01
1.27275772e-02 1.23462655e-01 4.86816555e-01 5.94480217e-01
-8.08539614e-02 2.17258021e-01 -2.89434880e-01 6.04842782e-01
2.03927442e-01 5.37201881e-01 -9.08435732e-02 -2.81561725e-02
-1.53375342e-01 1.34403765e-01 -9.14125681e-01 -6.10951364e-01
-2.41540074e-02 2.94258073e-02 -5.58857501e-01 1.58257242e-02
-3.92464936e-01 -3.55328083e-01 -5.25345922e-01 5.24061084e-01
8.79915655e-01 1.86583489e-01 6.95467114e-01 -5.74471295e-01
-4.19702083e-01 3.88972402e-01 -1.69919479e+00 1.12576866e+00
-3.42288762e-01 -1.23536594e-01 -4.57047671e-01 -8.66489291e-01
1.31249464e+00 4.93401974e-01 7.53393471e-01 -5.27097464e-01
-1.95122048e-01 7.80351400e-01 -1.33861199e-01 -7.67050564e-01
8.00278068e-01 -1.52019039e-01 -2.67922282e-01 1.23902075e-02
3.84999186e-01 2.28550300e-01 8.71161938e-01 -1.29143417e-01
7.61631370e-01 2.05322027e-01 6.87975287e-01 -6.46059930e-01
9.17710721e-01 4.66873169e-01 7.67298877e-01 5.93075752e-01
5.19651949e-01 -2.29335740e-01 5.30239761e-01 -5.85939527e-01
-1.34495175e+00 -1.01784122e+00 -7.88748205e-01 4.58618492e-01
2.92303741e-01 -5.34631252e-01 -5.60347855e-01 -5.43578804e-01
3.68304193e-01 8.08300674e-01 -3.87789726e-01 2.81545192e-01
-2.51380086e-01 -1.02186608e+00 7.41682887e-01 -6.26884028e-02
4.77261692e-01 -1.12406921e+00 -4.15005416e-01 2.89254814e-01
-3.36320549e-01 -1.34069180e+00 8.04819047e-01 6.22379519e-02
-4.39054549e-01 -1.33552682e+00 -3.65629070e-03 -7.60783494e-01
5.71186781e-01 -6.85979068e-01 1.24075568e+00 2.72713378e-02
3.47001031e-02 7.04099089e-02 -6.09562516e-01 -4.95257616e-01
-7.51798987e-01 2.37713486e-01 4.49942321e-01 -2.62791902e-01
9.86578941e-01 -5.36433756e-01 1.85514018e-01 3.57522279e-01
-7.59513497e-01 -2.92780906e-01 3.33050311e-01 5.94461381e-01
5.96391678e-01 4.79533076e-01 9.18408334e-01 -1.49682748e+00
2.03037843e-01 -9.43152010e-01 -7.72261918e-01 5.29186428e-01
-9.65569794e-01 1.22353010e-01 4.60031569e-01 -3.17213461e-02
-1.17454302e+00 -1.03621311e-01 -8.21214169e-02 3.48242193e-01
-4.36081499e-01 9.94478166e-01 -4.66179967e-01 2.65261054e-01
7.21436501e-01 -2.58195907e-01 -2.83522904e-01 -5.80446839e-01
3.38642746e-02 7.59307504e-01 4.93342519e-01 -1.07662117e+00
7.32095063e-01 -6.51556719e-03 2.09892750e-01 -8.92426670e-01
-4.76552695e-01 -3.66108328e-01 -9.23736870e-01 -1.29953101e-01
8.61014366e-01 -9.47823524e-01 -6.03746235e-01 1.49395689e-01
-9.47556436e-01 1.86416790e-01 -3.01793903e-01 6.03944540e-01
-4.80755180e-01 2.74190575e-01 1.41127864e-02 -7.36641705e-01
3.12522128e-02 -5.59233069e-01 4.65435177e-01 7.21648559e-02
-3.51747513e-01 -9.64601636e-01 9.34107825e-02 1.59208030e-01
-1.25374049e-01 4.51585650e-01 1.47818422e+00 -1.10247517e+00
-3.12454134e-01 -4.22008961e-01 -3.35969999e-02 -4.06508148e-02
1.59063637e-01 -7.16276765e-02 -6.15253568e-01 2.00744346e-02
-6.26321793e-01 1.26169696e-01 1.22445457e-01 -1.59918487e-01
3.89878392e-01 -4.35360789e-01 -4.28410798e-01 1.19011335e-01
1.81198192e+00 3.60005826e-01 9.75091159e-01 9.41337883e-01
4.88368154e-01 8.04313183e-01 1.22972333e+00 4.92919922e-01
4.83609557e-01 9.03808594e-01 2.04617530e-01 2.34929949e-01
1.83969334e-01 -1.44848153e-01 -1.46531940e-01 4.59242463e-01
-5.32133341e-01 -2.52613127e-01 -1.18360162e+00 5.79877257e-01
-1.85773396e+00 -1.40542495e+00 -4.86290514e-01 2.77850699e+00
1.02519822e+00 1.72370091e-01 3.99647683e-01 1.63747206e-01
9.45520163e-01 -5.04276514e-01 -3.32261174e-04 -3.80573452e-01
-2.01617539e-01 3.42544019e-01 8.43680739e-01 7.38593876e-01
-9.26278055e-01 8.45617473e-01 5.32113123e+00 5.14873385e-01
-2.66391635e-01 -1.31034225e-01 -2.90643573e-01 4.70889211e-01
-4.63594556e-01 3.18381935e-01 -1.30051625e+00 6.97399259e-01
8.72952878e-01 -5.90407372e-01 3.02776873e-01 6.62130594e-01
-1.27860844e-01 -3.44688386e-01 -8.23857367e-01 4.85976309e-01
-2.22193897e-01 -1.30996919e+00 1.38705552e-01 4.12877351e-01
5.14419436e-01 -2.28013292e-01 -7.63300061e-01 2.04258159e-01
4.29456890e-01 -7.48671353e-01 8.17430317e-01 7.51608729e-01
1.03639197e+00 -9.50432897e-01 1.14967620e+00 1.38379306e-01
-1.16185594e+00 -1.96177028e-02 -8.06666136e-01 1.01714350e-01
6.45964826e-03 8.25223148e-01 -1.07249689e+00 1.13791203e+00
9.12261009e-01 3.47889602e-01 -4.75074679e-01 1.03155279e+00
-3.27662355e-03 8.71103704e-02 -6.29113674e-01 -8.73961076e-02
-2.66689479e-01 -3.75846297e-01 6.09261632e-01 1.25757170e+00
6.49467707e-01 3.38002928e-02 -9.70820859e-02 6.79650426e-01
6.81997240e-02 3.25236529e-01 -6.98603213e-01 1.18409619e-01
1.24140716e+00 9.96117592e-01 -3.12309980e-01 -6.60030723e-01
-5.12639284e-01 2.71641850e-01 2.53900468e-01 1.45314664e-01
-6.97623610e-01 -7.37636983e-01 5.14698923e-01 4.11293209e-01
1.84682757e-01 2.52752244e-01 4.91198376e-02 -8.49126816e-01
2.10900083e-01 -8.57744992e-01 8.65529299e-01 -4.48878080e-01
-1.23596370e+00 6.00440800e-01 5.54065764e-01 -1.24177980e+00
-4.95317012e-01 -7.24265933e-01 2.16529995e-01 1.08503807e+00
-1.21068609e+00 -9.54025269e-01 -3.30186248e-01 3.19268107e-01
-2.85359889e-01 -4.22093064e-01 1.35710979e+00 8.19644213e-01
-1.83555946e-01 5.14582872e-01 -5.07679358e-02 1.30252570e-01
6.79945886e-01 -1.40347350e+00 -1.60317328e-02 6.15575016e-01
-5.15017807e-01 8.04185212e-01 9.90862131e-01 -9.11837339e-01
-9.57743227e-01 -9.16895926e-01 1.58964002e+00 -4.28797364e-01
6.79198563e-01 -9.91840735e-02 -9.15523469e-01 5.87930501e-01
-1.05114155e-01 -2.44107261e-01 6.65723681e-01 3.28511149e-01
-4.67503041e-01 -2.85682052e-01 -1.61451590e+00 1.32198140e-01
9.61330533e-01 -2.53205895e-01 -1.00116742e+00 3.41119796e-01
3.15274119e-01 -1.44958660e-01 -1.82296789e+00 5.33219039e-01
7.09657490e-01 -8.02058160e-01 1.00301397e+00 -5.02081692e-01
9.09624323e-02 -7.01067388e-01 -4.21278685e-01 -9.63684559e-01
-1.29661083e-01 9.48930383e-02 4.05717902e-02 1.69823873e+00
8.92797351e-01 -8.25285316e-01 4.07784462e-01 7.45369613e-01
-4.46616001e-02 -2.36797169e-01 -1.02806604e+00 -9.42481339e-01
-9.68960673e-02 1.73679128e-01 1.07458282e+00 1.31199932e+00
3.52155000e-01 1.56722873e-01 -1.48837850e-03 5.22072613e-01
7.02294409e-01 4.01101038e-02 7.97357202e-01 -1.81534743e+00
-1.06045648e-01 -1.12145968e-01 -8.26137304e-01 -2.64578853e-02
4.53395918e-02 -1.02075481e+00 -4.03945863e-01 -1.63625872e+00
-2.27857947e-01 -1.27655613e+00 -1.91434562e-01 6.85410500e-01
1.68154880e-01 1.74925327e-02 -1.93726197e-01 1.15256995e-01
9.45967808e-02 -1.34795442e-01 5.63872457e-01 1.26115367e-01
-4.00585271e-02 -1.96576878e-01 -4.26409036e-01 7.77440429e-01
7.22565532e-01 -6.81781948e-01 -5.99679463e-02 1.11012012e-01
9.67209458e-01 -2.26433817e-02 1.37866884e-01 -9.79754090e-01
3.46745960e-02 -2.57403016e-01 1.21906787e-01 -6.20316327e-01
1.23128876e-01 -1.18053305e+00 9.08320725e-01 2.95273870e-01
3.25337164e-02 8.45574960e-02 1.02246307e-01 2.76650637e-01
-2.90396869e-01 -9.74058390e-01 6.07161582e-01 -1.42031655e-01
-1.28182149e+00 -5.26040830e-02 -2.25699797e-01 1.29264280e-01
9.56774950e-01 -4.86026615e-01 -3.71350795e-01 3.69054735e-01
-8.92712772e-01 -9.29838344e-02 7.48971403e-01 2.92949766e-01
-5.16347401e-02 -1.47273207e+00 -5.08201182e-01 -1.12852074e-01
5.14865160e-01 -4.54759859e-02 -1.28998488e-01 4.69987035e-01
-6.15811646e-01 2.06374764e-01 -6.28730118e-01 -1.66777804e-01
-1.07280231e+00 6.50962591e-01 2.08735496e-01 -6.85450658e-02
-3.53684694e-01 2.91386664e-01 -4.10298496e-01 -6.59151495e-01
-1.13252565e-01 1.10850953e-01 -7.73118973e-01 6.92228317e-01
2.59169608e-01 4.88384068e-01 3.92124474e-01 -8.20283055e-01
-7.77533889e-01 5.12859106e-01 1.58761665e-01 -7.35116675e-02
1.50861228e+00 6.43760487e-02 -6.34243429e-01 4.56336498e-01
7.06700802e-01 4.65611428e-01 -1.43095493e-01 -1.92228146e-02
4.34292674e-01 -3.95850152e-01 -6.21003926e-01 -8.78039777e-01
-5.39102674e-01 -2.49192491e-02 4.71214980e-01 4.66029406e-01
8.25041294e-01 -5.51467128e-02 1.61563531e-01 3.99234086e-01
8.57728779e-01 -1.30470884e+00 -9.48336840e-01 5.67931950e-01
5.15098155e-01 -7.72746205e-01 3.55094314e-01 -7.98361063e-01
-3.03912163e-01 1.12366247e+00 5.77867508e-01 5.43596409e-02
4.82636571e-01 3.30355853e-01 -3.49000335e-01 -4.19313252e-01
-4.89191234e-01 -4.42648500e-01 1.13316260e-01 8.32984030e-01
5.08561254e-01 1.95193708e-01 -1.05313909e+00 4.97998446e-01
-4.57399160e-01 1.10024936e-01 7.23971963e-01 7.33991206e-01
-7.08604872e-01 -1.41863918e+00 -3.40343803e-01 6.58899665e-01
-3.08001488e-01 -1.66954264e-01 -9.62390080e-02 1.26554835e+00
8.12292933e-01 7.93969870e-01 2.68042356e-01 -1.53459668e-01
5.79392076e-01 3.10993612e-01 5.64089060e-01 -5.64405203e-01
-4.07507271e-01 -5.16853690e-01 1.02232647e+00 -1.85145121e-02
-7.49328077e-01 -6.27912998e-01 -1.40486968e+00 -3.76442194e-01
-4.58831221e-01 4.54968512e-01 4.38485831e-01 8.56930673e-01
2.25862354e-01 1.33698598e-01 1.75503209e-01 1.51513955e-02
-1.04736304e-03 -6.25774503e-01 -6.20839715e-01 6.30887091e-01
-3.72978389e-01 -1.31338668e+00 -2.92304814e-01 8.44824016e-02] | [9.211627006530762, 8.086844444274902] |
482b895e-0b9a-4407-817d-0231d9056041 | black-box-node-injection-attack-for-graph | 2202.09389 | null | https://arxiv.org/abs/2202.09389v1 | https://arxiv.org/pdf/2202.09389v1.pdf | Black-box Node Injection Attack for Graph Neural Networks | Graph Neural Networks (GNNs) have drawn significant attentions over the years and been broadly applied to vital fields that require high security standard such as product recommendation and traffic forecasting. Under such scenarios, exploiting GNN's vulnerabilities and further downgrade its classification performance become highly incentive for adversaries. Previous attackers mainly focus on structural perturbations of existing graphs. Although they deliver promising results, the actual implementation needs capability of manipulating the graph connectivity, which is impractical in some circumstances. In this work, we study the possibility of injecting nodes to evade the victim GNN model, and unlike previous related works with white-box setting, we significantly restrict the amount of accessible knowledge and explore the black-box setting. Specifically, we model the node injection attack as a Markov decision process and propose GA2C, a graph reinforcement learning framework in the fashion of advantage actor critic, to generate realistic features for injected nodes and seamlessly merge them into the original graph following the same topology characteristics. Through our extensive experiments on multiple acknowledged benchmark datasets, we demonstrate the superior performance of our proposed GA2C over existing state-of-the-art methods. The data and source code are publicly accessible at: https://github.com/jumxglhf/GA2C. | ['Liang Zhao', 'Yanfang Ye', 'Yujie Fan', 'Mingxuan Ju'] | 2022-02-18 | null | null | null | null | ['product-recommendation'] | ['miscellaneous'] | [ 2.50147998e-01 4.37300563e-01 -4.04874980e-01 1.76541746e-01
-8.39438215e-02 -8.08446348e-01 5.33044219e-01 -4.61951643e-03
-8.13774168e-02 6.26098275e-01 -3.88724715e-01 -1.04019833e+00
-2.45794244e-02 -1.08688557e+00 -8.86729956e-01 -6.80086613e-01
-2.11581171e-01 -1.33791063e-02 2.16066644e-01 -3.49467158e-01
2.08524495e-01 4.76159364e-01 -6.48465037e-01 -4.35588837e-01
8.91353250e-01 7.79416502e-01 -4.65424508e-01 4.90845203e-01
2.15159416e-01 5.78450680e-01 -6.60562634e-01 -8.22632432e-01
6.75079465e-01 -3.06794614e-01 -5.86532056e-01 -1.09463930e-01
-9.93718300e-03 -2.30065033e-01 -8.06632638e-01 1.54979634e+00
4.71567512e-01 -3.77693512e-02 1.30151987e-01 -1.80178726e+00
-8.89578938e-01 1.00113440e+00 -9.88352358e-01 8.13730061e-02
1.16363026e-01 5.98541677e-01 6.81938350e-01 -7.31015876e-02
3.61813098e-01 1.23992324e+00 4.87238139e-01 7.59398341e-01
-1.14414549e+00 -1.00911355e+00 4.88583416e-01 7.24470541e-02
-1.14465523e+00 -1.96839437e-01 1.14045274e+00 -4.23287787e-02
4.99959230e-01 2.34616995e-01 4.81674492e-01 1.66075814e+00
1.70810461e-01 6.21155024e-01 9.13823307e-01 -2.20369726e-01
2.95434445e-01 -9.35389698e-02 -3.64625789e-02 6.09093726e-01
5.45587122e-01 4.05358732e-01 1.48906022e-01 -2.90492773e-01
8.34997833e-01 2.77418524e-01 -2.91283101e-01 -3.65719110e-01
-6.71690464e-01 9.38829720e-01 8.80315185e-01 -1.39296278e-01
-4.71760094e-01 5.52646816e-01 3.74300241e-01 2.34232545e-01
1.32658795e-01 4.20467883e-01 -1.50697976e-01 5.18093407e-02
-3.27194005e-01 2.32177041e-02 8.28926742e-01 8.11009943e-01
4.18144912e-01 4.77252454e-01 1.56380236e-01 2.01010898e-01
2.44541079e-01 4.30796713e-01 1.63177028e-01 -6.24891877e-01
5.33866048e-01 7.12869883e-01 -2.29203627e-01 -1.46698797e+00
-1.68107077e-01 -6.42699838e-01 -1.12356818e+00 2.62938738e-01
4.45884198e-01 -6.04645550e-01 -9.95822608e-01 1.73287797e+00
5.82153201e-01 7.40109026e-01 -2.56939773e-02 8.05764437e-01
4.52907026e-01 5.98144054e-01 1.93618849e-01 1.22582063e-01
1.04830360e+00 -1.04884076e+00 -4.19601679e-01 -1.76626727e-01
4.95941818e-01 -2.53869265e-01 8.95398855e-01 3.42894822e-01
-8.05289149e-01 -2.14761004e-01 -1.13878512e+00 6.42757714e-01
-6.53360963e-01 -4.13765222e-01 7.97346056e-01 1.05187380e+00
-9.59237576e-01 7.19578981e-01 -9.58477259e-01 -2.02788934e-01
7.57587433e-01 5.04351854e-01 -1.59084633e-01 5.18152565e-02
-1.56170368e+00 3.40423167e-01 3.76445204e-01 5.78700364e-01
-1.36521220e+00 -5.81014514e-01 -8.10889781e-01 4.59259339e-02
1.06123126e+00 -4.78909612e-01 9.25568581e-01 -7.58196354e-01
-1.58975589e+00 2.83349127e-01 7.50563443e-01 -6.75558627e-01
6.62863374e-01 -3.23469006e-02 -5.64644098e-01 -1.36584295e-02
-3.76092970e-01 4.01742399e-01 9.50783968e-01 -1.22207773e+00
-2.59203911e-01 -2.26338446e-01 6.60453975e-01 -1.42258823e-01
-4.79606539e-01 -1.82405144e-01 -3.67500931e-01 -7.58765876e-01
-3.97693694e-01 -1.08783054e+00 -6.97953284e-01 -9.50867236e-02
-1.07277322e+00 8.66893902e-02 1.06868911e+00 -4.30785418e-01
1.31540060e+00 -1.92301440e+00 -5.35056321e-03 5.42767644e-01
5.80069542e-01 7.81378329e-01 -1.86455369e-01 5.85820079e-01
-1.48782954e-01 6.92776024e-01 -7.07463026e-02 3.56890038e-02
8.20366517e-02 5.24272211e-02 -4.01763737e-01 5.87544203e-01
6.26219362e-02 1.12294161e+00 -9.12871361e-01 -9.45326388e-02
1.73806071e-01 3.63853842e-01 -4.88556564e-01 1.82240605e-02
-2.45777383e-01 4.51601446e-01 -9.06192839e-01 6.66783452e-01
9.57261741e-01 -3.89253765e-01 2.73500621e-01 3.76382321e-02
5.40464520e-01 -2.50410974e-01 -1.03539848e+00 1.21248174e+00
-2.80472517e-01 5.08193076e-02 1.84405684e-01 -1.01281583e+00
8.14102113e-01 6.48661191e-03 7.83491731e-02 -5.75478494e-01
4.01230007e-01 -1.74187317e-01 3.78458828e-01 -1.35682002e-01
5.50918914e-02 2.00764969e-01 -1.70760006e-01 4.30864602e-01
-3.58897984e-01 1.99551716e-01 -5.70324250e-02 5.37463605e-01
1.45750415e+00 -1.88051373e-01 2.38198653e-01 4.65337262e-02
6.16066337e-01 -3.38257849e-01 4.65474039e-01 9.36203361e-01
-3.86740535e-01 6.89730272e-02 1.02648616e+00 -2.99882501e-01
-7.74232507e-01 -1.04895508e+00 4.19604152e-01 5.87935269e-01
4.59268689e-01 -3.15688014e-01 -9.17389572e-01 -1.28912866e+00
1.27461001e-01 6.69675946e-01 -7.18241215e-01 -6.89060271e-01
-5.32035589e-01 -7.21598148e-01 9.91858542e-01 3.36818099e-01
6.64392948e-01 -1.27585745e+00 -2.24748194e-01 1.36978135e-01
3.17087710e-01 -1.13819993e+00 -3.83918196e-01 -1.72765955e-01
-5.13791859e-01 -1.20410717e+00 -3.87253076e-01 -5.22938907e-01
7.89888024e-01 1.21728435e-01 7.53056884e-01 5.77035487e-01
-3.11579764e-01 1.90217689e-01 -4.63477582e-01 -2.31265411e-01
-5.36752582e-01 3.98328632e-01 -5.51685644e-03 2.30605617e-01
-1.08993605e-01 -9.30870712e-01 -8.17005992e-01 4.08725411e-01
-1.05701315e+00 -1.22679465e-01 5.45080245e-01 6.06182098e-01
1.31524757e-01 4.70884353e-01 7.96416879e-01 -1.38045526e+00
7.13886321e-01 -6.70165777e-01 -9.32239652e-01 1.18286356e-01
-7.38675594e-01 -1.00040689e-01 1.13984489e+00 -6.88623369e-01
-6.05321348e-01 -3.13023567e-01 -1.59962296e-01 -7.80096948e-01
-9.52491611e-02 4.40256417e-01 -5.84878564e-01 -4.37958449e-01
4.67065692e-01 -3.24410237e-02 4.31710891e-02 -1.58464968e-01
4.73797530e-01 2.35582232e-01 3.06774139e-01 -6.29380465e-01
1.53753722e+00 2.72675961e-01 2.57180929e-01 -3.80314440e-01
-4.43314075e-01 1.32391721e-01 -2.05001216e-02 -2.33601362e-01
4.51124072e-01 -4.16929156e-01 -1.18691123e+00 6.24304056e-01
-8.50283444e-01 -5.18632233e-01 3.46597880e-02 -3.27521674e-02
-2.13725045e-02 5.04142344e-01 -6.02003038e-01 -7.86448002e-01
-3.50582212e-01 -1.26750016e+00 5.78080654e-01 4.58280385e-01
3.06806773e-01 -1.23985398e+00 -1.42658070e-01 2.60159940e-01
6.53293610e-01 8.19189131e-01 9.76608813e-01 -9.30511236e-01
-7.85103917e-01 -3.50682795e-01 -1.46950036e-01 1.90299466e-01
1.23807989e-01 -3.03491000e-02 -6.21703684e-01 -6.06526077e-01
-3.01562697e-01 -2.93159395e-01 5.11938870e-01 -1.74266342e-02
1.55581188e+00 -7.60407448e-01 -5.04412413e-01 8.13211679e-01
1.48127687e+00 2.89416462e-01 8.40627909e-01 2.60220855e-01
1.06294465e+00 3.78142893e-01 1.62706226e-01 2.57411540e-01
1.00650273e-01 4.20922428e-01 1.12402570e+00 -2.88446546e-01
3.00283313e-01 -7.09831893e-01 2.57912844e-01 2.64242023e-01
8.70182067e-02 -8.54077578e-01 -7.95786679e-01 8.20714682e-02
-1.78665745e+00 -7.27535069e-01 2.10044429e-01 1.95314193e+00
4.52693462e-01 6.07538939e-01 1.44551843e-01 3.97430547e-02
1.05586767e+00 4.43328351e-01 -9.62955415e-01 -2.68287867e-01
1.89392403e-01 1.75135076e-01 9.27260816e-01 3.77844036e-01
-1.05380023e+00 1.17262983e+00 4.98515224e+00 1.04637849e+00
-1.12535655e+00 -1.09967545e-01 9.56997573e-01 1.08760126e-01
-2.52267361e-01 2.59102106e-01 -5.47646999e-01 5.53456903e-01
9.81093645e-01 -2.25144863e-01 7.87525415e-01 7.60628819e-01
1.42479479e-01 4.24918562e-01 -6.62054718e-01 6.13486350e-01
-2.32730597e-01 -1.12875128e+00 9.78762582e-02 4.03369367e-01
4.93896842e-01 -2.71653265e-01 4.00261223e-01 4.82479602e-01
8.79277229e-01 -1.14308989e+00 2.92083561e-01 7.93368071e-02
6.78559422e-01 -9.31258261e-01 4.60731745e-01 3.63188654e-01
-1.01385546e+00 -2.63004452e-01 -2.42536962e-01 1.61271855e-01
2.61053771e-01 4.37382638e-01 -6.99589074e-01 7.71552265e-01
4.24352229e-01 5.41085660e-01 -7.88363874e-01 7.71309435e-01
-5.39794683e-01 1.05435848e+00 -2.66489029e-01 -1.56847201e-02
3.84678364e-01 -2.98645258e-01 7.50410616e-01 5.42010844e-01
3.44352648e-02 -9.63592809e-03 3.49132925e-01 9.53532815e-01
-4.21188474e-01 -8.62724110e-02 -7.22305834e-01 -3.53670090e-01
6.20927215e-01 1.55964267e+00 -8.64666581e-01 4.14081737e-02
-2.69424736e-01 7.63690889e-01 3.90601665e-01 5.54339707e-01
-1.39325082e+00 -6.26122832e-01 4.56080824e-01 1.74524218e-01
2.54147589e-01 1.25802323e-01 6.36039814e-03 -9.99366760e-01
9.16664675e-02 -1.11148787e+00 3.62152308e-01 -3.61945391e-01
-1.36891580e+00 7.09722459e-01 -1.55636176e-01 -1.04975927e+00
4.63466346e-02 -5.54370582e-01 -1.13634908e+00 5.73856413e-01
-1.34436786e+00 -1.29795575e+00 -1.55688569e-01 6.75666332e-01
-4.00507310e-03 -1.97259143e-01 3.88723433e-01 2.39370078e-01
-1.11206567e+00 1.19012582e+00 -1.15279578e-01 4.41752613e-01
3.15564424e-01 -1.02088702e+00 9.14489567e-01 1.22419691e+00
-6.11737818e-02 7.04964697e-01 6.31854773e-01 -1.06579638e+00
-1.66308892e+00 -1.30268061e+00 -2.05119833e-01 -1.40633807e-01
1.10526955e+00 -7.50475943e-01 -8.14153731e-01 7.21638560e-01
2.99566239e-01 4.05422449e-01 4.83225845e-02 -2.86001235e-01
-3.87063354e-01 -9.72674862e-02 -1.29395163e+00 9.94562089e-01
1.22834921e+00 -1.89805374e-01 3.16638321e-01 2.46095344e-01
1.08968008e+00 -3.84872079e-01 -6.64869428e-01 4.16375220e-01
2.32988998e-01 -6.67652845e-01 1.00925875e+00 -9.22630131e-01
1.32490203e-01 -3.10852051e-01 2.39319623e-01 -1.38534439e+00
-6.21268600e-02 -1.22982728e+00 -4.10819322e-01 1.30108130e+00
5.30713737e-01 -1.14911270e+00 1.15989769e+00 4.08461869e-01
2.04180330e-01 -9.43240762e-01 -6.99844301e-01 -8.07349205e-01
1.54160872e-01 -5.46908230e-02 8.49536777e-01 9.14371550e-01
-1.59234062e-01 1.49065003e-01 -5.52969873e-01 5.36243439e-01
8.16516817e-01 -2.01280981e-01 8.95030320e-01 -8.46777320e-01
-5.51751375e-01 -5.48971295e-01 -5.44349134e-01 -7.85142183e-01
3.79285544e-01 -8.45020413e-01 -3.44697237e-01 -1.01875353e+00
-1.65886804e-01 -5.41772068e-01 -3.54742646e-01 6.48376226e-01
-2.43921652e-01 1.33782029e-01 3.28484684e-01 -1.60540968e-01
-5.67227721e-01 4.93510634e-01 1.26954687e+00 -3.36090744e-01
-3.27944010e-02 3.63440961e-01 -1.10098064e+00 4.91061300e-01
1.12031603e+00 -5.04342914e-01 -7.50617385e-01 2.47591604e-02
2.72259444e-01 2.60074645e-01 6.18937671e-01 -8.17779064e-01
1.76820040e-01 -2.08749488e-01 1.26112282e-01 -7.41425008e-02
-2.87323911e-02 -1.01027107e+00 2.56538600e-01 7.21642256e-01
-3.21957558e-01 2.23770231e-01 1.36635780e-01 1.02196705e+00
1.41091421e-01 -5.81383668e-02 7.17972934e-01 1.85201894e-02
-3.76264304e-01 7.96634793e-01 -5.94289601e-02 1.20230645e-01
1.38899553e+00 -6.70942366e-02 -7.20892370e-01 -6.33458376e-01
-5.75968266e-01 5.18000484e-01 4.99929190e-01 5.03960550e-01
5.10503232e-01 -1.24268723e+00 -4.37730283e-01 2.02814221e-01
-1.36731431e-01 -1.51218072e-01 3.41184705e-01 6.50588930e-01
-4.64847833e-01 3.56591493e-02 -8.09180811e-02 -1.34962454e-01
-8.06204438e-01 1.16285515e+00 4.84279633e-01 -6.09258235e-01
-5.76334953e-01 3.81560057e-01 4.07077312e-01 -5.75820863e-01
3.11531514e-01 1.48445427e-01 -1.96754873e-01 -6.51325345e-01
1.76507279e-01 3.82703304e-01 -1.76132679e-01 -1.78801507e-01
-2.49890953e-01 6.45793974e-02 -4.15879250e-01 2.13253573e-01
1.08394027e+00 -5.47236344e-03 1.20258346e-01 -3.35595906e-01
9.82726216e-01 -1.31885437e-02 -1.20993817e+00 -5.18464074e-02
-2.37550169e-01 -4.27281022e-01 -1.09085612e-01 -8.12609017e-01
-1.70337307e+00 7.48751462e-01 4.24252898e-01 5.78572869e-01
1.13126397e+00 -2.42517665e-01 9.77139652e-01 2.52455622e-01
4.64576364e-01 -5.34253120e-01 1.46479011e-01 9.28455070e-02
4.54332620e-01 -1.03547657e+00 -9.29803178e-02 -6.41598821e-01
-5.45902491e-01 8.23116481e-01 1.07840550e+00 -4.97637659e-01
8.43433142e-01 2.33651921e-01 -1.20273195e-01 -2.70339370e-01
-5.23507357e-01 2.36391857e-01 -1.89534619e-01 7.10225224e-01
-3.06912154e-01 5.17350361e-02 2.35519595e-02 5.97582936e-01
2.73120999e-02 -3.93123627e-01 8.38585794e-01 8.55558515e-01
8.58348384e-02 -1.26779246e+00 -1.97186247e-01 3.85212392e-01
-6.88403249e-01 -1.00074962e-01 -4.15015459e-01 1.02138460e+00
-2.57082969e-01 8.69181573e-01 -4.45353806e-01 -6.40121758e-01
2.49028057e-01 -4.96601462e-01 1.74400602e-02 -2.29289755e-01
-6.11748874e-01 -1.30467758e-01 -1.37348520e-03 -6.01856351e-01
1.28349781e-01 -3.05014282e-01 -8.95750761e-01 -7.07615793e-01
-4.67101812e-01 2.13178217e-01 3.66804898e-01 5.83994985e-01
4.74054128e-01 6.04339719e-01 1.04595613e+00 -6.82777941e-01
-9.16992188e-01 -7.29153216e-01 -5.21336317e-01 3.57737929e-01
1.80372939e-01 -6.53577983e-01 -5.70326269e-01 -5.02419233e-01] | [6.102791786193848, 7.328357219696045] |
8ade0c9c-bdb8-4cdd-8256-fa659a78a1d4 | a-novel-deep-learning-based-approach-for | 2208.03408 | null | https://arxiv.org/abs/2208.03408v2 | https://arxiv.org/pdf/2208.03408v2.pdf | A novel deep learning-based approach for sleep apnea detection using single-lead ECG signals | Sleep apnea (SA) is a type of sleep disorder characterized by snoring and chronic sleeplessness, which can lead to serious conditions such as high blood pressure, heart failure, and cardiomyopathy (enlargement of the muscle tissue of the heart). The electrocardiogram (ECG) plays a critical role in identifying SA since it might reveal abnormal cardiac activity. Recent research on ECG-based SA detection has focused on feature engineering techniques that extract specific characteristics from multiple-lead ECG signals and use them as classification model inputs. In this study, a novel method of feature extraction based on the detection of S peaks is proposed to enhance the detection of adjacent SA segments using a single-lead ECG. In particular, ECG features collected from a single lead (V2) are used to identify SA episodes. On the extracted features, a CNN model is trained to detect SA. Experimental results demonstrate that the proposed method detects SA from single-lead ECG data is more accurate than existing state-of-the-art methods, with 91.13% classification accuracy, 92.58% sensitivity, and 88.75% specificity. Moreover, the further usage of features associated with the S peaks enhances the classification accuracy by 0.85%. Our findings indicate that the proposed machine learning system has the potential to be an effective method for detecting SA episodes. | ['Cuong Do', 'Huy-Hieu Pham', 'Huy-Khiem Le', 'Thao Nguyen', 'Anh-Tu Nguyen'] | 2022-08-05 | null | null | null | null | ['sleep-apnea-detection'] | ['medical'] | [ 4.31847125e-01 -4.05945748e-01 -5.25552556e-02 -1.42519444e-01
-3.74743074e-01 -2.44782090e-01 -3.64491343e-01 2.86708862e-01
-3.30187917e-01 6.87141776e-01 -2.74177164e-01 -2.82485783e-01
-3.04373261e-02 -6.07451737e-01 1.30992264e-01 -8.96137416e-01
-1.52039811e-01 -1.42989084e-01 1.55322319e-02 3.98140512e-02
2.04469576e-01 5.82074523e-01 -1.13431895e+00 -9.57783163e-02
1.01180959e+00 1.45835841e+00 -1.44294242e-03 6.09435976e-01
3.49922836e-01 1.57114968e-01 -1.07691276e+00 1.92385256e-01
5.96280321e-02 -9.65065777e-01 -1.83964342e-01 -4.91348729e-02
-2.16124117e-01 -1.06613658e-01 4.76935320e-02 7.98822761e-01
8.53591561e-01 -1.80596292e-01 2.70919353e-01 -9.65939641e-01
1.42590716e-01 1.50983423e-01 -5.06686270e-01 9.65176582e-01
2.13072494e-01 1.66357458e-01 5.11472821e-01 -6.68322563e-01
-1.14540935e-01 1.63999677e-01 8.45240235e-01 3.48899156e-01
-1.00669301e+00 -7.88153052e-01 -7.49095082e-01 3.84774476e-01
-1.50651336e+00 -2.56465822e-01 1.18104279e+00 -2.02445537e-01
8.12065721e-01 5.23000062e-01 1.33389139e+00 4.33563530e-01
6.32791042e-01 3.85436743e-01 1.16984367e+00 -4.35753912e-01
1.00074686e-01 -5.44542298e-02 1.50907665e-01 6.99013054e-01
4.45035607e-01 -4.41669151e-02 -2.83208162e-01 -2.33228728e-01
6.38041377e-01 4.00816411e-01 -4.81346995e-01 1.27909020e-01
-1.06337821e+00 6.17243171e-01 3.03842455e-01 5.71556628e-01
-5.71014404e-01 -2.37851903e-01 4.75323290e-01 1.20439515e-01
1.11263648e-01 7.18671739e-01 -3.65152121e-01 -3.93627703e-01
-1.19897246e+00 -2.08127707e-01 4.86443430e-01 3.17213058e-01
5.02896249e-01 2.44521320e-01 -2.50682831e-01 6.78830743e-01
3.24725688e-01 6.15578234e-01 6.70052171e-01 -8.37340534e-01
1.14433788e-01 9.16948497e-01 -2.64591407e-02 -9.93422568e-01
-8.91409218e-01 -9.63080823e-01 -1.00797272e+00 -2.33074382e-01
5.10597080e-02 -1.30308539e-01 -5.06855905e-01 1.21625412e+00
8.74629393e-02 1.74377576e-01 1.23037897e-01 8.71670604e-01
1.00952446e+00 4.29013491e-01 -6.87513277e-02 -7.06901312e-01
1.40857005e+00 -3.99116009e-01 -8.48795176e-01 -1.41355559e-01
2.11115524e-01 -5.18372357e-01 8.12570691e-01 5.36122978e-01
-8.57800007e-01 -5.95293403e-01 -1.23528135e+00 3.97180170e-01
2.24287529e-02 5.48111320e-01 6.65961653e-02 8.16520095e-01
-6.85807467e-01 4.94258791e-01 -1.00903916e+00 -3.12245548e-01
5.14508605e-01 5.02538264e-01 -6.06961828e-03 3.02569062e-01
-1.10139477e+00 7.82048583e-01 5.91946095e-02 3.99158895e-01
-4.72321182e-01 -3.57536256e-01 -8.04147363e-01 1.72777563e-01
4.59409840e-02 -6.03204668e-01 7.67750859e-01 -6.69168711e-01
-1.14142299e+00 7.05802619e-01 -3.88793916e-01 -4.31353450e-01
6.83374479e-02 -1.80639982e-01 -7.76198626e-01 5.75634360e-01
8.33200067e-02 -2.46469229e-01 8.59028041e-01 -5.79495668e-01
-4.66924071e-01 -7.13288248e-01 -3.26354980e-01 -6.60815686e-02
-3.23403090e-01 1.85606539e-01 3.03181987e-02 -4.86127853e-01
3.52499574e-01 -7.80232131e-01 -2.58403748e-01 -2.99336344e-01
-3.53739947e-01 -4.84480672e-02 6.75783217e-01 -7.28281260e-01
1.70688975e+00 -2.38885307e+00 -2.18474343e-01 4.07261223e-01
6.62750363e-01 6.54003263e-01 5.57808399e-01 2.36397356e-01
2.73359299e-01 1.22680739e-01 -4.48911101e-01 5.68631887e-02
-6.00783825e-01 1.51479304e-01 2.86870509e-01 6.25778496e-01
1.69787928e-01 9.05736566e-01 -6.89586580e-01 -4.28894669e-01
3.42854202e-01 5.60779452e-01 1.43646300e-01 3.08245420e-01
6.69613600e-01 7.86812603e-01 -2.78008878e-01 6.09051883e-01
3.93392861e-01 -3.71713340e-01 2.17396513e-01 -3.03796381e-01
-1.64005026e-01 4.35963064e-01 -7.43217170e-01 1.42607737e+00
-1.80168271e-01 6.87501729e-01 -2.97508687e-01 -1.02558541e+00
1.29582453e+00 4.89192188e-01 5.43515980e-01 -6.80985868e-01
3.94848526e-01 2.91075826e-01 4.04286236e-01 -1.01946628e+00
-2.84683883e-01 -2.72086531e-01 1.97627679e-01 2.49508649e-01
-2.92863488e-01 1.19769141e-01 9.96911153e-02 -2.22329587e-01
1.04376268e+00 -4.19993073e-01 8.44426513e-01 -2.84672618e-01
8.35300803e-01 -4.12975639e-01 1.01541054e+00 6.19047701e-01
-4.82117891e-01 4.94592756e-01 1.71878174e-01 -7.51034021e-01
-5.36044359e-01 -7.76265740e-01 -3.05326760e-01 5.95130362e-02
2.08706141e-01 -5.66683590e-01 -5.07329524e-01 -5.95583558e-01
-2.13713601e-01 2.99915582e-01 -4.44604635e-01 -4.85413849e-01
-6.51613951e-01 -9.76852775e-01 6.74170017e-01 8.23417783e-01
7.90994346e-01 -1.22577715e+00 -1.55841839e+00 3.65772009e-01
-3.95411164e-01 -9.10490334e-01 -8.41224045e-02 1.12618998e-01
-1.32136750e+00 -1.29931223e+00 -6.17147923e-01 -6.00699961e-01
5.31460404e-01 1.47056133e-01 7.92517781e-01 5.03895819e-01
-7.20104158e-01 -3.96166705e-02 -2.89894253e-01 -6.30593777e-01
-2.13377908e-01 -6.54887185e-02 7.50784501e-02 3.06358039e-01
5.56001425e-01 -8.54945958e-01 -1.10199380e+00 2.20677167e-01
-4.40321892e-01 -3.10852200e-01 7.83543169e-01 5.40655136e-01
6.41278684e-01 -1.08329561e-02 9.64296401e-01 -5.95894039e-01
6.30065560e-01 -3.08609694e-01 -2.01699659e-01 -1.40698999e-01
-1.09416950e+00 -5.93803883e-01 7.00402141e-01 -1.48840114e-01
-4.57639515e-01 -2.13234220e-02 -2.84536898e-01 -3.14556330e-01
-4.74626601e-01 5.12947679e-01 -8.77957616e-04 5.62506020e-02
6.05098486e-01 5.91258943e-01 2.50967085e-01 -2.92379975e-01
-6.38717651e-01 8.65732372e-01 5.38967431e-01 1.13036335e-01
4.47200269e-01 3.64787638e-01 2.92680651e-01 -1.15189099e+00
-6.41247511e-01 -8.99598539e-01 -6.87811553e-01 -2.89552927e-01
1.06189907e+00 -7.77713656e-01 -6.22966707e-01 4.10785258e-01
-8.18284333e-01 5.86959571e-02 -1.66201308e-01 7.15745747e-01
-1.39599070e-01 5.83569646e-01 -4.19809431e-01 -1.05154049e+00
-1.10431898e+00 -8.31283331e-01 6.59313500e-01 4.86360371e-01
-5.25237024e-01 -7.19053328e-01 2.15990301e-02 3.23896259e-01
6.28270030e-01 6.02935553e-01 8.71485472e-01 -6.27775371e-01
3.81144919e-02 -3.79898190e-01 2.45866448e-01 5.13465464e-01
5.44133782e-01 -4.36622016e-02 -8.29295933e-01 -2.97296703e-01
4.87152457e-01 3.22447836e-01 5.33473551e-01 5.38213968e-01
8.87009799e-01 -4.79626991e-02 -2.80900538e-01 7.09868133e-01
1.39624798e+00 8.86828899e-01 7.78099775e-01 3.07907283e-01
4.12225157e-01 -8.88179988e-02 4.73077238e-01 5.96154690e-01
7.12196082e-02 3.03847820e-01 3.76441896e-01 -5.18256485e-01
9.21033993e-02 4.48292553e-01 1.37290061e-01 8.08136880e-01
-4.47047859e-01 2.38453239e-01 -8.95995319e-01 5.36505818e-01
-1.45934856e+00 -8.13992739e-01 -6.62481308e-01 2.29871273e+00
6.32747293e-01 1.61250442e-01 4.17239428e-01 8.27021062e-01
7.13966548e-01 -1.55281067e-01 -5.03254473e-01 -2.82715023e-01
1.57344341e-02 6.82009280e-01 -3.85844931e-02 -2.46327996e-01
-9.53496754e-01 1.61044057e-02 6.17153740e+00 -1.71323135e-01
-1.58424914e+00 7.37827420e-02 3.22325796e-01 2.38274205e-02
3.19376677e-01 -3.45545053e-01 -4.43266839e-01 7.61340022e-01
9.34003890e-01 -3.95396072e-03 5.15826233e-02 6.86022341e-01
6.12472653e-01 -7.15320110e-02 -6.02577150e-01 1.33342040e+00
2.42718577e-01 -1.04828155e+00 -5.75135708e-01 -1.59443796e-01
1.67369619e-01 -2.47792631e-01 -2.37831026e-01 -5.43420985e-02
-1.02137172e+00 -9.07446027e-01 2.78344572e-01 5.71229875e-01
1.15063488e+00 -8.37181628e-01 1.24085188e+00 2.85007805e-01
-1.28223419e+00 -4.08832967e-01 -8.88596401e-02 -3.39325458e-01
-6.27297023e-03 8.13997090e-01 -9.15248752e-01 3.22564006e-01
9.12393987e-01 1.09158099e+00 -7.81442821e-01 1.50969172e+00
-4.88059819e-01 1.14718759e+00 -2.98560858e-01 -3.57278809e-02
-3.70458663e-01 -1.31747305e-01 6.59953952e-01 1.06383133e+00
3.89452726e-01 2.66330838e-01 6.16795383e-02 9.58345890e-01
2.40162939e-01 1.22359470e-01 -6.16003990e-01 2.95246661e-01
3.14519167e-01 1.37970090e+00 -8.52423549e-01 -3.16833258e-01
-4.93806303e-01 4.94580597e-01 -4.14026469e-01 1.84488401e-01
-6.74450040e-01 -7.90513873e-01 2.04769656e-01 3.10969979e-01
1.28495395e-01 4.97805886e-02 -6.39017344e-01 -9.43712294e-01
1.21726744e-01 -7.28121221e-01 3.25589657e-01 -5.05110562e-01
-9.27283525e-01 6.86922550e-01 -3.62039626e-01 -1.46814525e+00
-8.78363475e-02 7.60799227e-03 -1.18582249e+00 9.24437225e-01
-1.37503171e+00 -6.81956112e-01 -6.88778400e-01 6.56694472e-01
4.19584572e-01 -4.28846702e-02 1.05243838e+00 3.82641256e-01
-7.95974433e-01 4.52390850e-01 -3.44046414e-01 2.68915445e-01
3.37264627e-01 -1.31938112e+00 -1.27586478e-03 1.04248059e+00
-1.32617190e-01 7.64372408e-01 3.95799518e-01 -4.83700752e-01
-1.11360180e+00 -9.86530960e-01 1.01577926e+00 -2.74034701e-02
-8.77365693e-02 7.75867328e-02 -7.60108113e-01 2.99442768e-01
-9.38032568e-02 1.98616415e-01 1.11598456e+00 -3.64037693e-01
3.85157585e-01 -6.06006026e-01 -1.16967678e+00 3.91520411e-02
3.38008910e-01 -3.45004171e-01 -6.84007704e-01 -3.65715176e-01
2.21525207e-02 -2.06854910e-01 -1.11319387e+00 6.91019356e-01
8.01065385e-01 -9.84723330e-01 6.85274661e-01 -1.62126124e-01
1.14019439e-01 -4.21821952e-01 3.63745004e-01 -1.08622849e+00
-3.76737505e-01 -7.57101595e-01 -3.25081497e-01 9.36676621e-01
3.07944447e-01 -7.16359317e-01 4.87515450e-01 1.79497808e-01
-3.26167911e-01 -9.64345157e-01 -7.20911801e-01 -5.28709114e-01
-6.34502947e-01 -1.60372898e-01 4.38904226e-01 6.61058664e-01
6.73138350e-02 5.46549201e-01 -5.06247915e-02 9.59173366e-02
4.21598554e-01 2.52510995e-01 3.50811958e-01 -1.63471079e+00
4.87670526e-02 -9.53129008e-02 -5.96706092e-01 -3.53930295e-01
-4.73125219e-01 -6.70026243e-01 -1.24501050e-01 -1.66666794e+00
5.48626594e-02 -3.46944064e-01 -1.03920448e+00 4.02189612e-01
-4.01391715e-01 7.24622130e-01 4.16675285e-02 1.56798631e-01
-3.50177318e-01 1.26967952e-01 1.05782735e+00 5.98158203e-02
-6.29719436e-01 6.55139089e-01 -5.96044779e-01 7.14296997e-01
1.23652494e+00 -6.37935221e-01 -3.51162851e-01 3.31241488e-01
1.53916925e-01 3.33290726e-01 3.44290197e-01 -1.36296976e+00
-6.16666675e-02 3.17863643e-01 7.45944619e-01 -5.09023309e-01
1.93073764e-01 -7.43244946e-01 1.59724593e-01 9.44420636e-01
-5.74384294e-02 1.57192573e-01 1.64280042e-01 2.89344221e-01
-2.70381778e-01 -6.60985112e-02 8.41486633e-01 -1.77768454e-01
-3.80697876e-01 6.84277713e-02 -7.39004612e-01 -4.83569875e-02
9.81615663e-01 -3.92776728e-01 4.89481464e-02 -1.35050207e-01
-8.84948552e-01 -1.83041081e-01 2.09593344e-02 1.75111257e-02
1.07239938e+00 -1.04285371e+00 -5.57799160e-01 6.17152989e-01
1.68836802e-01 -1.34151921e-01 1.17084563e-01 1.64209127e+00
-5.69954693e-01 2.30831876e-01 -3.67476165e-01 -9.19260144e-01
-1.57246041e+00 2.10904464e-01 4.91266251e-01 -3.07321269e-02
-9.53043520e-01 6.05449677e-01 -4.66499567e-01 5.08003771e-01
6.58653378e-02 -4.18846428e-01 -7.74830639e-01 -4.64669541e-02
6.29061937e-01 6.08017683e-01 2.70297647e-01 -4.21559393e-01
-5.96485198e-01 7.53094077e-01 4.72451359e-01 4.12505865e-01
1.12280536e+00 -1.92479327e-01 -3.65296006e-01 5.70982873e-01
8.79524648e-01 1.53570995e-01 -6.74785733e-01 1.52779892e-01
-2.61200935e-01 -3.44937295e-01 -3.33511233e-02 -9.21117544e-01
-1.28563881e+00 1.23077798e+00 1.07730699e+00 5.80872595e-01
1.52949584e+00 -2.61509418e-01 1.19007957e+00 3.69363949e-02
3.41832519e-01 -6.25042796e-01 -6.01732768e-02 -2.37182871e-01
5.39812684e-01 -9.49640632e-01 -1.14365434e-02 -1.74163625e-01
-5.15782952e-01 1.53122365e+00 4.29059833e-01 -1.53103068e-01
6.89127326e-01 1.83896981e-02 3.18673313e-01 -3.75861734e-01
-1.61091566e-01 -1.32360473e-01 2.57114798e-01 5.70446789e-01
4.03803259e-01 1.31135985e-01 -7.83155739e-01 8.22376072e-01
-8.35255906e-02 2.66557485e-01 5.12728333e-01 9.73968983e-01
-5.70533454e-01 -7.60940909e-01 -2.03923136e-01 8.96781802e-01
-9.78741407e-01 -8.23030174e-02 -2.61347592e-01 4.54297125e-01
4.74653780e-01 1.35134804e+00 -6.36004135e-02 -5.43103278e-01
2.76607037e-01 3.83090556e-01 2.48767540e-01 -6.51308715e-01
-8.16508949e-01 2.80086130e-01 -1.05791308e-01 -4.82937336e-01
-5.51093757e-01 -4.91194069e-01 -1.46998131e+00 4.23669457e-01
-3.40603948e-01 3.68108541e-01 5.97752631e-01 9.98345196e-01
4.94949341e-01 6.87136710e-01 7.35810816e-01 -3.06860358e-01
-1.59046918e-01 -1.10073054e+00 -8.64529490e-01 2.87578017e-01
8.19884896e-01 -5.23992181e-01 -4.77552384e-01 1.07389130e-01] | [14.085192680358887, 3.2255005836486816] |
6fc655d6-48c4-4c79-bed4-e9075e7e1a20 | winning-the-lottery-with-continuous-1 | 1912.04427 | null | https://arxiv.org/abs/1912.04427v4 | https://arxiv.org/pdf/1912.04427v4.pdf | Winning the Lottery with Continuous Sparsification | The search for efficient, sparse deep neural network models is most prominently performed by pruning: training a dense, overparameterized network and removing parameters, usually via following a manually-crafted heuristic. Additionally, the recent Lottery Ticket Hypothesis conjectures that, for a typically-sized neural network, it is possible to find small sub-networks which, when trained from scratch on a comparable budget, match the performance of the original dense counterpart. We revisit fundamental aspects of pruning algorithms, pointing out missing ingredients in previous approaches, and develop a method, Continuous Sparsification, which searches for sparse networks based on a novel approximation of an intractable $\ell_0$ regularization. We compare against dominant heuristic-based methods on pruning as well as ticket search -- finding sparse subnetworks that can be successfully re-trained from an early iterate. Empirical results show that we surpass the state-of-the-art for both objectives, across models and datasets, including VGG trained on CIFAR-10 and ResNet-50 trained on ImageNet. In addition to setting a new standard for pruning, Continuous Sparsification also offers fast parallel ticket search, opening doors to new applications of the Lottery Ticket Hypothesis. | ['Pedro Savarese', 'Michael Maire', 'Hugo Silva'] | 2019-12-10 | null | http://proceedings.neurips.cc/paper/2020/hash/83004190b1793d7aa15f8d0d49a13eba-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/83004190b1793d7aa15f8d0d49a13eba-Paper.pdf | neurips-2020-12 | ['ticket-search'] | ['methodology'] | [ 3.16202074e-01 7.29576290e-01 -3.92228812e-01 -3.51906031e-01
-4.27763402e-01 -1.80832054e-02 3.63234073e-01 -4.54032689e-01
-6.44223928e-01 9.30958390e-01 1.05847158e-01 -2.25041375e-01
-5.51738918e-01 -7.64464319e-01 -9.63648915e-01 -5.58778882e-01
-2.82829612e-01 8.32101345e-01 1.18767560e-01 9.29029584e-02
-1.03681788e-01 5.03441811e-01 -1.59234762e+00 3.06925982e-01
4.57326382e-01 1.15355945e+00 3.71976167e-01 1.14582427e-01
-9.92230922e-02 8.09392393e-01 -4.08462703e-01 -4.13950264e-01
6.55464113e-01 -3.63931835e-01 -8.89733553e-01 3.54400985e-02
7.60219872e-01 -3.82279307e-02 -6.52006269e-01 1.15375149e+00
2.22447403e-02 2.22912490e-01 2.68501580e-01 -9.68602717e-01
-2.14699954e-01 1.25329852e+00 -4.40818429e-01 4.16197151e-01
-5.66533148e-01 3.73243615e-02 1.22874188e+00 -7.49592841e-01
7.99125195e-01 1.14707649e+00 1.16650844e+00 6.88310087e-01
-1.58384478e+00 -9.33574796e-01 5.08510947e-01 -1.36845961e-01
-1.46402514e+00 -9.29661810e-01 5.09026110e-01 -6.86990619e-02
1.44767153e+00 7.51618817e-02 1.01755977e+00 1.10352027e+00
-3.30769658e-01 7.15276718e-01 6.68923974e-01 -3.31287622e-01
3.62888187e-01 -3.95624936e-02 1.86864778e-01 1.04673851e+00
6.84413075e-01 1.22465923e-01 -5.86248100e-01 -2.52413034e-01
9.40691888e-01 1.59216776e-01 -3.72939795e-01 -3.71214360e-01
-8.61451745e-01 1.06116009e+00 3.94357562e-01 3.47791731e-01
-3.60901356e-01 3.63860637e-01 4.82328266e-01 5.34409344e-01
4.48188961e-01 7.56264627e-01 -7.61400521e-01 1.27036959e-01
-1.78192627e+00 2.54939646e-01 1.18192947e+00 9.39063668e-01
8.88110995e-01 5.63310742e-01 1.37627974e-01 9.08599079e-01
1.16432972e-01 -1.85160249e-01 4.88058656e-01 -1.13809884e+00
6.06301904e-01 3.30153584e-01 -4.83382225e-01 -7.52151072e-01
-3.89680117e-01 -9.78528559e-01 -1.25610304e+00 4.77195308e-02
2.20435634e-01 -3.17788839e-01 -9.79062140e-01 1.91158783e+00
-3.13817430e-03 5.93694806e-01 -1.24372907e-01 5.34934521e-01
7.10742295e-01 3.89203876e-01 -2.04168018e-02 -6.97598234e-02
9.14413989e-01 -1.15326154e+00 -9.63475853e-02 -5.57656527e-01
8.45124900e-01 -1.61571175e-01 7.65909314e-01 7.20960915e-01
-1.40563393e+00 -2.76100308e-01 -9.89108562e-01 1.14709243e-01
-6.68169186e-02 8.59503001e-02 1.04112709e+00 5.96832037e-01
-1.38514841e+00 1.19066048e+00 -1.08118355e+00 -1.90281406e-01
1.16261876e+00 7.46480763e-01 -4.33187246e-01 -2.26771116e-01
-8.17237735e-01 8.58165622e-01 6.73635364e-01 2.99298819e-02
-1.25989759e+00 -9.86116827e-01 -7.91230917e-01 4.42011774e-01
4.30413902e-01 -1.04917347e+00 1.05524552e+00 -1.10822928e+00
-1.28646660e+00 9.27646220e-01 -1.21059462e-01 -1.38194835e+00
7.64073581e-02 7.52369463e-02 -5.50207905e-02 2.60852367e-01
-1.97009705e-02 9.68252361e-01 9.74536538e-01 -9.70757484e-01
-5.20608246e-01 -2.09198475e-01 -5.69234714e-02 7.62705728e-02
-4.53417927e-01 -9.35543478e-02 -4.34425771e-01 -6.65623724e-01
3.22230220e-01 -7.49681532e-01 -7.87692130e-01 -1.45333797e-01
-6.21565640e-01 -1.55394450e-01 5.89699268e-01 -2.17730612e-01
1.06498003e+00 -1.87184644e+00 2.18776107e-01 5.30623734e-01
7.88671315e-01 2.43106321e-01 -3.96968812e-01 -1.17002599e-01
-3.22962582e-01 3.02442163e-01 -1.95274681e-01 -7.64109612e-01
-1.09430484e-01 4.79287535e-01 -4.24294323e-01 5.29807270e-01
5.59057891e-02 7.93909609e-01 -4.41096902e-01 -4.17406112e-01
-2.22294927e-01 3.78585935e-01 -9.44583952e-01 -3.96144509e-01
-1.66767731e-01 -2.67332971e-01 -2.97952354e-01 6.01659477e-01
4.74795848e-01 -5.70619345e-01 2.72290677e-01 -1.58314988e-01
1.04646914e-01 5.08223414e-01 -1.14162314e+00 1.60243630e+00
-2.35210821e-01 6.07247233e-01 4.85139698e-01 -1.73028219e+00
7.65295982e-01 2.00424045e-01 4.27573353e-01 -4.57532406e-01
2.64053375e-01 2.62551337e-01 -1.35187358e-01 -2.05995180e-02
1.26120672e-01 -1.40128776e-01 4.31165963e-01 3.80652934e-01
7.12758899e-01 3.51646319e-02 3.80838573e-01 2.90077060e-01
1.57318401e+00 -3.71876061e-01 5.01461476e-02 -4.84820515e-01
-1.25405073e-01 4.65088338e-02 6.56698287e-01 1.22508848e+00
1.46976247e-01 6.34948611e-01 6.60867214e-01 -6.78085327e-01
-1.10695648e+00 -6.59828305e-01 -1.46617278e-01 1.00843549e+00
-4.47352290e-01 -4.36072111e-01 -8.60010564e-01 -6.50632322e-01
-8.20825100e-02 3.72403860e-01 -7.70842314e-01 -4.32009324e-02
-6.85390770e-01 -9.78273749e-01 6.74382031e-01 4.67906713e-01
5.94345868e-01 -9.82314408e-01 -5.49019933e-01 3.38194221e-01
2.95517981e-01 -1.08813643e+00 2.54619513e-02 9.61220741e-01
-1.54000235e+00 -9.71044958e-01 -7.56284237e-01 -1.02578521e+00
8.11142743e-01 3.71119916e-01 1.49171793e+00 5.46016753e-01
-3.26096922e-01 -1.99444350e-02 8.70748144e-03 -2.07669199e-01
1.15071625e-01 6.02578759e-01 1.75244771e-02 -4.17039603e-01
2.63076007e-01 -9.63654399e-01 -5.02921939e-01 -8.04691687e-02
-6.92324877e-01 9.50655118e-02 7.48095989e-01 1.01970303e+00
8.38189244e-01 2.23619178e-01 5.26180923e-01 -1.21297765e+00
5.21995306e-01 -6.69599950e-01 -6.42559052e-01 8.81958455e-02
-5.93421400e-01 2.04418898e-01 7.31480777e-01 -5.69684386e-01
-6.46116555e-01 1.73171237e-01 -2.41984919e-01 -9.33535576e-01
-1.34434298e-01 6.35665476e-01 1.54515952e-01 -5.14068842e-01
8.26109290e-01 1.33763731e-01 4.66539301e-02 -6.66155756e-01
2.52062052e-01 -2.26646423e-01 4.10571784e-01 -5.72984278e-01
7.91204453e-01 7.37256765e-01 -1.23897828e-02 -8.76787961e-01
-1.20825338e+00 -3.09217662e-01 -3.10526788e-01 4.08551902e-01
3.09057832e-01 -1.14842498e+00 -3.90195161e-01 6.60642385e-02
-1.02494872e+00 -6.87273085e-01 -8.77265811e-01 3.53923291e-01
-5.16413033e-01 4.43462171e-02 -6.24670327e-01 -3.61137718e-01
-4.80230242e-01 -8.97769511e-01 6.86655402e-01 -7.03781918e-02
-2.95852214e-01 -8.96960497e-01 1.01794209e-02 6.73781261e-02
4.93706197e-01 -1.67672232e-01 9.19634581e-01 -1.05569065e+00
-6.99171126e-01 -7.38591477e-02 -4.03789788e-01 4.48166907e-01
-6.03513718e-01 -3.56874436e-01 -8.28325331e-01 -3.34169149e-01
1.77375078e-01 -5.76642692e-01 1.59579170e+00 8.21515441e-01
1.48899817e+00 -7.86666214e-01 -6.33038282e-01 1.24876440e+00
1.62250209e+00 -2.88176179e-01 6.45140588e-01 3.48066330e-01
5.22214592e-01 2.82324821e-01 -2.13006601e-01 4.22017455e-01
-1.02256127e-01 1.05606303e-01 6.46565557e-01 -1.95582688e-01
-1.66618943e-01 -2.69308448e-01 -2.41378248e-01 5.96379220e-01
-2.07582280e-01 8.12338591e-02 -6.72366202e-01 7.23684192e-01
-1.70042944e+00 -1.19467556e+00 3.66633654e-01 1.91462374e+00
1.01234961e+00 4.70593363e-01 -8.55638236e-02 1.60421535e-01
3.92031640e-01 2.58581936e-01 -5.41522264e-01 -6.33550361e-02
-2.78929442e-01 8.91187966e-01 9.32311177e-01 3.09749871e-01
-1.02790105e+00 1.22192311e+00 7.23150444e+00 1.19254446e+00
-8.91478777e-01 3.15641254e-01 9.74513054e-01 -7.19481885e-01
-2.74856925e-01 2.08385568e-02 -1.38383102e+00 7.91822821e-02
9.49695826e-01 3.09617799e-02 8.17070663e-01 1.35591674e+00
8.67801234e-02 1.09296761e-01 -1.15288174e+00 1.08201456e+00
2.09817197e-03 -2.19889665e+00 -3.73057015e-02 5.21753207e-02
1.10789502e+00 7.51449108e-01 3.40447649e-02 5.65364480e-01
6.23351693e-01 -1.40968215e+00 7.14325309e-01 1.01560391e-01
5.57404518e-01 -5.84843993e-01 3.77151608e-01 3.32993299e-01
-1.04975426e+00 -1.65137336e-01 -8.15828204e-01 7.37183243e-02
-8.14459324e-02 8.92983556e-01 -6.65505290e-01 -1.91963226e-01
9.17184174e-01 7.55583823e-01 -3.42545748e-01 1.38422477e+00
-6.48788065e-02 9.16530669e-01 -7.91107655e-01 8.86884779e-02
5.42201817e-01 -9.58539322e-02 3.51210654e-01 1.26007342e+00
3.27481598e-01 -1.76046994e-02 -3.94582339e-02 1.13031113e+00
-5.58974206e-01 -2.51504362e-01 -7.62347698e-01 -7.94815198e-02
5.72689533e-01 1.00842249e+00 -9.76161301e-01 -4.42705333e-01
-2.49635085e-01 3.82561833e-01 7.42641628e-01 6.05506122e-01
-6.52746141e-01 -3.16234678e-01 4.80077296e-01 2.67140269e-01
8.41521025e-01 -5.14398813e-02 -6.12805724e-01 -1.22003603e+00
-3.10906827e-01 -1.15784192e+00 3.07464927e-01 -3.07944000e-01
-1.03657842e+00 8.01456094e-01 -1.35856554e-01 -5.71966350e-01
9.07485560e-03 -4.35110271e-01 -7.11489618e-01 5.69770217e-01
-1.50685585e+00 -7.49385595e-01 1.00284591e-01 8.20327461e-01
7.52284288e-01 -4.07993346e-01 7.54358709e-01 3.26914579e-01
-7.18407214e-01 7.20868945e-01 1.35388598e-01 -7.52339587e-02
-4.23745215e-02 -6.61996365e-01 2.89736778e-01 6.19197547e-01
6.46997631e-01 7.88114846e-01 6.14349723e-01 -4.68756586e-01
-1.06193626e+00 -1.15579021e+00 8.23739886e-01 3.50717247e-01
7.69803822e-01 -2.31690243e-01 -9.77690339e-01 1.00339937e+00
2.38903351e-02 1.39316902e-01 3.60008150e-01 4.69693393e-01
-4.21157718e-01 -8.40784013e-02 -1.16099811e+00 6.15202665e-01
1.50363004e+00 -1.82424590e-01 -6.02263510e-01 6.38541102e-01
7.45377421e-01 -1.94379792e-01 -5.16464651e-01 3.75248611e-01
3.04882437e-01 -9.67524648e-01 1.16999054e+00 -8.51968229e-01
3.27748239e-01 4.70937490e-01 -1.08376570e-01 -1.16461468e+00
-3.22297543e-01 -9.34322655e-01 -4.06999648e-01 5.06064892e-01
5.83985448e-01 -6.50709033e-01 1.61161435e+00 5.16447902e-01
-3.74974459e-01 -1.20360768e+00 -1.13568497e+00 -9.58043158e-01
-9.02624354e-02 -4.40719843e-01 5.03004432e-01 9.02024567e-01
-2.33743325e-01 2.63383657e-01 -1.86166540e-01 -4.76069003e-02
8.64982009e-01 -2.35545218e-01 4.26365376e-01 -1.52649629e+00
-5.26062965e-01 -9.01307642e-01 -1.90825164e-01 -1.29740167e+00
4.30403322e-01 -9.89175081e-01 -2.77347624e-01 -1.26070476e+00
1.62657142e-01 -7.83283174e-01 -2.65351295e-01 8.71189654e-01
6.92269146e-01 3.65454167e-01 3.80331650e-02 2.86040127e-01
-7.69478083e-01 4.17769730e-01 9.58402276e-01 -1.70253351e-01
-1.83970690e-01 9.09500122e-02 -1.01221633e+00 1.14238501e+00
8.01903188e-01 -8.55084538e-01 -5.41488707e-01 -6.49163067e-01
4.02784199e-01 -1.89625204e-01 5.01768410e-01 -1.24885964e+00
5.18818438e-01 1.69609655e-02 1.56677634e-01 -4.66532230e-01
4.90388781e-01 -6.53136849e-01 2.38030497e-02 4.68877703e-01
-5.38614392e-01 -2.14164406e-01 2.31435865e-01 5.89893222e-01
-1.43492371e-01 -7.92912662e-01 1.01300550e+00 -6.68686152e-01
-4.27560508e-01 7.40824997e-01 -1.10372968e-01 4.45249826e-01
5.60935020e-01 -5.23876727e-01 -2.81874955e-01 -2.13278811e-02
-9.35898960e-01 -5.38240448e-02 5.04186228e-02 -2.58774281e-01
6.82717800e-01 -8.83105338e-01 -4.47882891e-01 3.39753151e-01
-6.99997723e-01 3.73990685e-01 1.53281018e-01 8.61123323e-01
-3.67672354e-01 5.27497053e-01 4.11740169e-02 -2.11334586e-01
-1.01851666e+00 1.76195890e-01 5.49536884e-01 -6.39937878e-01
-9.38724935e-01 1.60389256e+00 1.82757154e-01 -6.14549480e-02
8.12438250e-01 -4.01165694e-01 1.13091089e-01 -9.92029831e-02
3.29101413e-01 1.25494614e-01 2.69151777e-01 -7.43377302e-03
-5.44495620e-02 1.65255547e-01 -2.59427726e-01 8.19222480e-02
1.92875886e+00 2.63529181e-01 -5.94301783e-02 -2.16933116e-01
1.33814061e+00 -5.05897939e-01 -1.34670031e+00 -4.66746509e-01
-1.08978741e-01 -2.15995222e-01 4.17943001e-01 -4.64427352e-01
-1.79517257e+00 6.13338053e-01 1.53979240e-02 2.94052541e-01
8.92895103e-01 1.05647929e-01 7.82856941e-01 1.00288689e+00
4.24380124e-01 -1.19701362e+00 -2.33958159e-02 8.72451901e-01
7.19350219e-01 -9.52744126e-01 2.16797784e-01 -1.83263972e-01
-3.13588679e-01 8.52717638e-01 4.52702969e-01 -5.88106155e-01
1.06435633e+00 5.96767783e-01 -7.29134560e-01 -4.51451302e-01
-1.12297773e+00 6.16514273e-02 -3.43261138e-02 5.90150476e-01
-3.35291699e-02 -2.90180475e-01 2.18964785e-01 7.60947466e-01
-4.50733751e-01 1.43953070e-01 6.52127201e-03 6.79965615e-01
-8.54208171e-01 -6.37697697e-01 -3.31774056e-02 1.15667701e+00
-4.89725411e-01 -6.33578241e-01 2.84749735e-02 8.69340658e-01
2.01219872e-01 4.05729026e-01 1.64396450e-01 -1.72395214e-01
6.56409040e-02 3.13937292e-02 5.23467839e-01 -9.07059968e-01
-6.77061141e-01 3.95574886e-03 3.27448905e-01 -7.39367843e-01
-2.23849908e-01 -4.51267898e-01 -9.49992895e-01 -3.98368448e-01
-4.89200383e-01 1.61610201e-01 4.96468306e-01 1.08154702e+00
3.96440297e-01 3.14692497e-01 1.77445516e-01 -1.20584404e+00
-8.42347741e-01 -7.68432736e-01 -8.29765201e-01 -9.50098187e-02
8.01775753e-02 -5.34733474e-01 -7.60990798e-01 -1.96808740e-01] | [8.547319412231445, 3.3171210289001465] |
6cf09fe1-29ac-47ec-b83e-60931c17ab31 | 190408338 | 1904.08338 | null | http://arxiv.org/abs/1904.08338v1 | http://arxiv.org/pdf/1904.08338v1.pdf | OCKELM+: Kernel Extreme Learning Machine based One-class Classification using Privileged Information (or KOC+: Kernel Ridge Regression or Least Square SVM with zero bias based One-class Classification using Privileged Information) | Kernel method-based one-class classifier is mainly used for outlier or
novelty detection. In this letter, kernel ridge regression (KRR) based
one-class classifier (KOC) has been extended for learning using privileged
information (LUPI). LUPI-based KOC method is referred to as KOC+. This
privileged information is available as a feature with the dataset but only for
training (not for testing). KOC+ utilizes the privileged information
differently compared to normal feature information by using a so-called
correction function. Privileged information helps KOC+ in achieving better
generalization performance which is exhibited in this letter by testing the
classifiers with and without privileged information. Existing and proposed
classifiers are evaluated on the datasets from UCI machine learning repository
and also on MNIST dataset. Moreover, experimental results evince the advantage
of KOC+ over KOC and support vector machine (SVM) based one-class classifiers. | ['Chandan Gautam', 'M. Tanveer', 'Aruna Tiwari'] | 2019-04-13 | null | null | null | null | ['one-class-classifier'] | ['methodology'] | [-6.11417517e-02 -3.79516572e-01 -4.15695906e-01 -3.68492812e-01
-2.91666657e-01 -2.72165745e-01 6.33676112e-01 4.28131968e-01
-5.28523266e-01 1.05075228e+00 -1.91289008e-01 -3.05977434e-01
-4.81592536e-01 -6.35349810e-01 -4.22576189e-01 -8.70393515e-01
-4.66924071e-01 -1.18934833e-01 1.21347144e-01 -1.51440769e-01
5.80583811e-01 8.21455002e-01 -1.92185271e+00 3.15676033e-01
1.21320295e+00 8.25694263e-01 -3.70889962e-01 4.53148484e-01
-2.24348828e-02 4.90835875e-01 -7.30098248e-01 1.22673094e-01
2.80722708e-01 -2.08500385e-01 -3.43089402e-01 -4.31864411e-01
9.63501632e-02 1.89271182e-01 1.50204808e-01 7.86737382e-01
1.45361260e-01 2.58832425e-01 1.18647039e+00 -1.64336789e+00
-6.97721779e-01 1.87571004e-01 -3.52547616e-01 1.51406690e-01
3.13952386e-01 -1.52846202e-01 3.72514963e-01 -9.57556605e-01
3.56261730e-01 7.40728021e-01 4.82815415e-01 4.65880632e-01
-1.02006280e+00 -5.54466546e-01 -1.44685298e-01 4.77891386e-01
-1.49361503e+00 2.10037649e-01 9.89208162e-01 -5.10582387e-01
1.00458443e+00 5.10766089e-01 5.30315518e-01 8.81765425e-01
6.44142687e-01 6.65044904e-01 1.52833331e+00 -6.18041813e-01
5.94160378e-01 6.85369670e-01 8.82215202e-01 4.98657852e-01
5.31903505e-01 4.34283525e-01 -5.55002689e-01 -6.37301981e-01
2.11617544e-01 3.52962345e-01 -3.33861381e-01 -2.54465640e-01
-8.22834551e-01 9.13833678e-01 1.68787077e-01 3.54258627e-01
-3.04638594e-01 -4.36638713e-01 6.12977326e-01 7.57584333e-01
4.38396752e-01 4.54579651e-01 -6.40192032e-01 -2.19255075e-01
-6.37269616e-01 9.09220949e-02 8.20187747e-01 7.88195193e-01
6.39064729e-01 5.59259832e-01 4.35087383e-02 7.93193996e-01
5.15835509e-02 3.94540340e-01 1.21362746e+00 3.66177224e-02
5.74845821e-02 1.00031984e+00 -2.36949418e-02 -7.89975643e-01
-4.76218730e-01 -5.69997609e-01 -6.77129507e-01 6.09966695e-01
1.54113069e-01 1.43144846e-01 -9.93899167e-01 1.21092820e+00
1.54406220e-01 4.67992604e-01 5.56681275e-01 5.03266752e-01
6.55650437e-01 5.65885663e-01 -7.63831735e-02 -1.87211603e-01
9.62606013e-01 -5.94630241e-01 -4.91436660e-01 4.22513098e-01
9.96539414e-01 -5.72188199e-01 1.02203369e+00 9.24086750e-01
-2.22239587e-02 -4.01862592e-01 -1.33102989e+00 6.61539316e-01
-1.18536913e+00 1.60566032e-01 9.06873226e-01 1.09313393e+00
-5.59085608e-01 6.41218781e-01 -4.76963431e-01 -5.30274808e-01
2.79450297e-01 3.38351786e-01 -8.61112475e-01 9.53479931e-02
-1.04435778e+00 9.56945240e-01 7.83138037e-01 -2.14416429e-01
-4.37172383e-01 -4.99430060e-01 -8.51292491e-01 -1.99050620e-01
1.38096288e-01 4.69481081e-01 4.23738658e-01 -7.45041132e-01
-1.37970543e+00 4.95283663e-01 1.57729611e-01 -7.23388672e-01
3.48631591e-01 -2.66104221e-01 -9.44510162e-01 -8.69579688e-02
-1.23930998e-01 -1.46453172e-01 1.04557920e+00 -1.15474200e+00
-4.16979015e-01 -4.88820612e-01 -6.66620255e-01 -2.02189147e-01
-1.37073383e-01 -5.22163585e-02 3.62700731e-01 -9.25810099e-01
2.26057217e-01 -8.43648553e-01 3.93065423e-01 -5.49498141e-01
-3.97997528e-01 -4.11401689e-01 1.56627500e+00 -5.64315379e-01
1.42176759e+00 -2.04637527e+00 -5.74130893e-01 8.11540663e-01
-2.36463591e-01 7.26756930e-01 2.96665162e-01 4.92468715e-01
-6.23356462e-01 -2.46413499e-02 -1.49119303e-01 3.40284437e-01
-5.24245322e-01 2.84712017e-01 -4.38628078e-01 6.84782326e-01
1.83334425e-01 4.77737188e-01 -5.19653380e-01 -2.31515482e-01
3.52944404e-01 9.77624282e-02 -1.32396007e-02 -2.46098153e-02
4.11206573e-01 3.84605020e-01 -3.01400691e-01 1.13615847e+00
9.15433407e-01 3.31266850e-01 -3.75573426e-01 1.73616514e-01
-3.92808877e-02 -6.63618624e-01 -1.28913164e+00 7.75812328e-01
-3.02259862e-01 2.30106279e-01 -6.01675391e-01 -1.23352456e+00
1.52158272e+00 2.65721142e-01 7.19233975e-02 -2.02899680e-01
1.11965783e-01 5.33188283e-01 -4.39582281e-02 -3.30285043e-01
2.35913366e-01 1.59815818e-01 3.94589566e-02 8.02486166e-02
1.36092633e-01 5.47346294e-01 -1.42151967e-01 1.47554092e-02
7.37770617e-01 9.16766152e-02 1.04705954e+00 -5.72923005e-01
1.05500579e+00 4.63477224e-02 6.70754850e-01 8.03839505e-01
-4.81569916e-01 -3.57910097e-02 3.98085952e-01 -3.72928917e-01
-6.04913175e-01 -9.08131480e-01 -7.56246448e-01 6.89776182e-01
-2.00980395e-01 -9.01317075e-02 -3.12357903e-01 -9.00746584e-01
6.11669421e-01 8.02010000e-01 -8.12001228e-01 -4.23515677e-01
-2.56567776e-01 -6.64908707e-01 4.65199292e-01 4.44709092e-01
4.71344620e-01 -9.36910510e-01 -3.74712110e-01 -4.21950184e-02
7.47535646e-01 -5.37940145e-01 7.09957257e-02 5.57343364e-01
-1.02322042e+00 -1.25933778e+00 -5.05490065e-01 -3.07160050e-01
6.58523500e-01 1.12128422e-01 3.69295120e-01 -5.80801256e-02
-6.82809293e-01 3.34465235e-01 -8.37757409e-01 -8.40508044e-01
-3.58085513e-01 -1.58124730e-01 5.16287982e-01 3.93266112e-01
7.53773987e-01 -5.86139202e-01 -5.52484281e-02 3.43269944e-01
-6.57387555e-01 -6.85510099e-01 6.62240326e-01 1.08670700e+00
2.49062255e-01 1.95666656e-01 1.28423154e+00 -1.24289393e+00
6.20436311e-01 -7.20975876e-01 -6.97213113e-01 2.82722294e-01
-1.24911129e+00 7.10295811e-02 9.29331958e-01 -5.94322026e-01
-9.71401572e-01 -7.44663775e-02 2.86387384e-01 -4.48972404e-01
-3.23188454e-01 3.38871300e-01 -1.90370634e-01 -3.48049909e-01
1.04863238e+00 4.60214615e-01 6.62684068e-02 -7.32835710e-01
-9.47265252e-02 1.18723333e+00 5.08943081e-01 -6.89333558e-01
8.78236294e-01 2.49122545e-01 4.09236044e-01 -1.00203109e+00
-4.31234002e-01 -7.36809671e-01 -7.50436008e-01 -1.01901315e-01
5.61783791e-01 -6.17907822e-01 -8.15810919e-01 8.27679574e-01
-4.56175566e-01 4.12999123e-01 -1.30093649e-01 9.90002334e-01
-3.01293552e-01 2.07080066e-01 -8.54087695e-02 -1.42282844e+00
-4.46600735e-01 -8.33101571e-01 3.97718519e-01 4.38768387e-01
4.73833419e-02 -9.50011432e-01 -1.35009706e-01 8.45755041e-02
4.23983216e-01 8.51936877e-01 9.65262949e-01 -1.51769865e+00
-1.53959498e-01 -1.12965357e+00 1.12762533e-01 6.85039520e-01
1.99453875e-01 4.14760783e-02 -1.02411819e+00 -3.69704068e-01
-1.05282679e-01 -1.23682998e-01 8.03044617e-01 -1.19040616e-01
1.06756604e+00 -2.10038975e-01 -1.32504433e-01 6.84699953e-01
1.71503222e+00 4.76698607e-01 7.15243459e-01 7.04527915e-01
4.04331863e-01 3.34055007e-01 1.00956261e+00 4.39824462e-01
-3.56458783e-01 5.14130712e-01 1.74221724e-01 2.06149772e-01
6.39954627e-01 -6.29749224e-02 3.01662773e-01 3.01224440e-01
-1.30749032e-01 3.05450499e-01 -8.92809272e-01 8.44039246e-02
-1.82193661e+00 -8.80826116e-01 -5.19362152e-01 2.64546299e+00
7.15322495e-01 2.35692471e-01 1.93025675e-02 5.92478871e-01
6.18499517e-01 -3.68751645e-01 -5.76092124e-01 -8.54110897e-01
-2.15038121e-01 4.66589421e-01 6.19288146e-01 3.05100709e-01
-1.26597619e+00 7.23893702e-01 4.80349064e+00 1.17037308e+00
-1.18985081e+00 -2.36742005e-01 2.81929493e-01 2.31807217e-01
3.44493955e-01 2.14008257e-01 -8.59469891e-01 7.16079056e-01
9.56629157e-01 -1.24963872e-01 2.09366247e-01 1.64189839e+00
5.77929616e-02 -6.91204011e-01 -8.01910818e-01 1.15399957e+00
1.25997975e-01 -8.92304122e-01 8.70476756e-03 1.37341440e-01
5.31681716e-01 -5.11395752e-01 -9.61626470e-02 7.92384982e-01
-2.93506593e-01 -8.84111345e-01 1.91154718e-01 8.98126006e-01
6.95528328e-01 -1.09904790e+00 1.24577332e+00 2.74462521e-01
-9.02794182e-01 -2.11186498e-01 -4.53527749e-01 -1.36156734e-02
-6.72026098e-01 5.16253829e-01 -7.68270135e-01 9.43188667e-01
5.79602301e-01 6.22590125e-01 -1.14378786e+00 1.01052082e+00
4.96917963e-02 9.17803049e-01 -1.41945392e-01 -8.01593512e-02
-2.22145766e-02 -2.81070322e-01 6.44613922e-01 9.39278781e-01
2.29659170e-01 -1.94991350e-01 1.11508600e-01 4.48595345e-01
5.21507680e-01 6.12525105e-01 -8.10126245e-01 1.94683254e-01
6.17769480e-01 1.13273609e+00 -5.87967813e-01 -3.85286719e-01
-2.76061028e-01 1.09497523e+00 1.13208748e-01 3.81028175e-01
-5.28384507e-01 -1.08987582e+00 3.98371369e-01 -1.29655853e-01
-7.01715052e-02 -7.31206611e-02 -3.11876655e-01 -1.07820976e+00
-4.20092493e-02 -5.25865495e-01 5.67532063e-01 -3.89458716e-01
-1.41806257e+00 5.68246007e-01 3.35617572e-01 -1.65626419e+00
-5.30853719e-02 -9.44740176e-01 -9.87712145e-01 1.10788691e+00
-1.27923870e+00 -1.20843828e+00 -3.87860149e-01 7.92016983e-01
7.19837993e-02 -9.73527372e-01 8.75840247e-01 -2.81819075e-01
-7.30571389e-01 9.32886600e-01 5.69429994e-01 4.78912219e-02
9.65966284e-01 -1.47210586e+00 -4.46882099e-01 7.65740454e-01
-4.23851490e-01 1.00044465e+00 5.06035686e-01 -8.72745156e-01
-1.04955864e+00 -1.15895808e+00 6.17090702e-01 -4.77330625e-01
3.99669915e-01 -1.28781512e-01 -1.26823545e+00 4.36164051e-01
-4.70589757e-01 3.14099729e-01 1.21824396e+00 7.89400414e-02
-6.30617261e-01 -3.05701256e-01 -1.66902757e+00 2.24382356e-01
1.47627041e-01 -3.68824631e-01 -6.02662146e-01 3.71876135e-02
3.79831970e-01 -2.66330852e-03 -1.02125239e+00 4.27781343e-01
6.20897770e-01 -8.80273461e-01 7.78675795e-01 -8.67409170e-01
-3.05384427e-01 -4.40714628e-01 -4.86609966e-01 -1.04932618e+00
8.10783952e-02 -3.52660328e-01 -3.25758874e-01 1.24425280e+00
3.28679740e-01 -1.44551194e+00 6.88736260e-01 5.87664187e-01
-2.07540933e-02 -1.04819572e+00 -8.30964267e-01 -1.55641747e+00
1.21182404e-01 -4.03553069e-01 5.60263276e-01 1.24132204e+00
1.37649134e-01 -1.65308043e-01 -6.24867439e-01 4.48765978e-02
7.35552132e-01 -2.67830402e-01 7.74631262e-01 -1.66264725e+00
-1.40694469e-01 1.33510962e-01 -1.02371836e+00 4.34499234e-01
1.88283458e-01 -1.02028418e+00 -7.17507184e-01 -6.61535800e-01
-2.58050933e-02 -4.75694418e-01 -5.61503530e-01 6.48406804e-01
-2.42718473e-01 2.97972206e-02 2.63431873e-02 5.64867556e-01
4.41745259e-02 3.43493462e-01 4.95224446e-01 2.20907137e-01
-7.71289408e-01 3.26591522e-01 -3.65880370e-01 5.26660919e-01
1.23989344e+00 -3.96228850e-01 -3.84003162e-01 1.01552713e+00
-3.85355026e-01 -2.77362078e-01 5.44371903e-01 -1.31013608e+00
-6.23500794e-02 -4.06887084e-01 8.25169325e-01 -8.12635422e-01
7.97776207e-02 -8.00364017e-01 -2.04337999e-01 6.02157652e-01
-2.68459976e-01 -1.47127777e-01 3.69560882e-03 1.00365806e+00
-1.54149219e-01 -4.64166284e-01 1.09983039e+00 1.64223745e-01
-8.94750476e-01 -9.85978991e-02 -2.97220945e-01 -4.24599826e-01
1.57223189e+00 -7.27626920e-01 -4.63294625e-01 -7.96679929e-02
-9.87781405e-01 1.74589851e-03 4.56258744e-01 1.95862219e-01
8.73043954e-01 -1.41418159e+00 -4.53746408e-01 6.74050033e-01
8.86272848e-01 -7.62171268e-01 2.03904524e-01 7.09139943e-01
-4.57466573e-01 7.52934933e-01 -3.94284725e-01 -3.57171476e-01
-1.58999181e+00 6.96008325e-01 1.30451307e-01 -9.02213454e-02
-5.78346789e-01 4.39496279e-01 -5.57817757e-01 -4.84150797e-01
1.54100865e-01 -1.33073658e-01 -5.50451458e-01 -2.15001628e-01
4.18333352e-01 7.28099704e-01 3.24911654e-01 -5.53780913e-01
-5.67382455e-01 3.99058044e-01 -5.28421581e-01 2.14309812e-01
9.29220378e-01 5.77298164e-01 -1.61598340e-01 6.30268216e-01
1.44759107e+00 7.01032728e-02 -5.21259546e-01 -1.89620346e-01
4.87836927e-01 -7.00382710e-01 -1.82546869e-01 -1.02667654e+00
-3.13348889e-01 5.47207296e-01 1.12586212e+00 -1.75385609e-01
1.00190985e+00 -5.05779445e-01 1.84458360e-01 8.50077868e-01
5.90378106e-01 -1.28472674e+00 -1.40194625e-01 4.18673158e-01
8.92907202e-01 -1.36810315e+00 8.10031518e-02 -2.44722947e-01
-8.95075798e-01 1.32801878e+00 7.41870761e-01 -3.55407953e-01
1.18018222e+00 -2.00496733e-01 -3.60881835e-02 1.05052881e-01
-5.71075261e-01 -5.06617539e-02 2.66670763e-01 1.00477087e+00
2.51167446e-01 3.13677609e-01 -7.69172370e-01 8.44537318e-01
-6.97497055e-02 -1.22254334e-01 6.35133445e-01 1.36220443e+00
-7.48512864e-01 -1.09714580e+00 -6.25622511e-01 9.53794539e-01
-2.06566766e-01 1.12589933e-01 -2.25975215e-01 9.71677423e-01
1.11389674e-01 9.78594184e-01 -4.06705678e-01 -6.21976793e-01
1.05920680e-01 5.96578777e-01 -1.40529871e-01 -3.77796352e-01
-4.65419561e-01 -5.22687852e-01 -1.32557422e-01 -5.60712874e-01
1.03190288e-01 -4.09617096e-01 -1.28581750e+00 3.74352224e-02
-5.91354311e-01 4.87988889e-01 9.57182348e-01 7.46932983e-01
2.53651768e-01 1.61906853e-01 9.88897741e-01 -1.67329744e-01
-7.99089968e-01 -1.01218700e+00 -1.23189378e+00 3.54377151e-01
2.04800859e-01 -1.01201963e+00 -7.56020963e-01 -3.05375636e-01] | [8.189534187316895, 3.883599281311035] |
1310afd8-ced7-4b0f-8cb1-0e151f1fd885 | bidirectional-copy-paste-for-semi-supervised | 2305.00673 | null | https://arxiv.org/abs/2305.00673v1 | https://arxiv.org/pdf/2305.00673v1.pdf | Bidirectional Copy-Paste for Semi-Supervised Medical Image Segmentation | In semi-supervised medical image segmentation, there exist empirical mismatch problems between labeled and unlabeled data distribution. The knowledge learned from the labeled data may be largely discarded if treating labeled and unlabeled data separately or in an inconsistent manner. We propose a straightforward method for alleviating the problem - copy-pasting labeled and unlabeled data bidirectionally, in a simple Mean Teacher architecture. The method encourages unlabeled data to learn comprehensive common semantics from the labeled data in both inward and outward directions. More importantly, the consistent learning procedure for labeled and unlabeled data can largely reduce the empirical distribution gap. In detail, we copy-paste a random crop from a labeled image (foreground) onto an unlabeled image (background) and an unlabeled image (foreground) onto a labeled image (background), respectively. The two mixed images are fed into a Student network and supervised by the mixed supervisory signals of pseudo-labels and ground-truth. We reveal that the simple mechanism of copy-pasting bidirectionally between labeled and unlabeled data is good enough and the experiments show solid gains (e.g., over 21% Dice improvement on ACDC dataset with 5% labeled data) compared with other state-of-the-arts on various semi-supervised medical image segmentation datasets. Code is available at https://github.com/DeepMed-Lab-ECNU/BCP}. | ['Yan Wang', 'Wei Shen', 'Qingli Li', 'Duowen Chen', 'Yunhao Bai'] | 2023-05-01 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Bai_Bidirectional_Copy-Paste_for_Semi-Supervised_Medical_Image_Segmentation_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Bai_Bidirectional_Copy-Paste_for_Semi-Supervised_Medical_Image_Segmentation_CVPR_2023_paper.pdf | cvpr-2023-1 | ['semi-supervised-medical-image-segmentation'] | ['computer-vision'] | [ 3.97674739e-01 6.12154126e-01 -6.61888301e-01 -7.93322325e-01
-1.02791488e+00 -6.81902587e-01 1.19734518e-01 -3.13987911e-01
-3.26131016e-01 8.70911956e-01 -6.64380565e-02 -2.73220003e-01
2.01324269e-01 -3.90682817e-01 -8.94644320e-01 -1.15122759e+00
3.73289675e-01 6.68299973e-01 1.12352163e-01 3.03243279e-01
-2.50435978e-01 1.68203060e-02 -1.00751567e+00 3.64297688e-01
1.00732088e+00 8.21831942e-01 3.39527518e-01 3.68683457e-01
-2.75426298e-01 9.22457576e-01 -2.47742787e-01 -1.80909052e-01
3.11664969e-01 -7.42205739e-01 -1.11022842e+00 6.42565966e-01
4.44237053e-01 -2.96694726e-01 -3.51599962e-01 1.39780676e+00
2.08680049e-01 -3.65534067e-01 5.84568381e-01 -1.45500791e+00
-8.96056354e-01 8.82178605e-01 -1.05662680e+00 -1.31011391e-02
-3.77440929e-01 2.29581058e-01 6.05460584e-01 -7.08691776e-01
8.09377670e-01 8.92308414e-01 2.99433947e-01 8.59618783e-01
-1.32914448e+00 -8.36611271e-01 3.78488630e-01 -9.86735597e-02
-1.17369533e+00 -1.05686836e-01 8.84395480e-01 -4.77296293e-01
-8.82558003e-02 1.17810994e-01 5.65730095e-01 1.08288682e+00
-1.96694255e-01 1.35132074e+00 1.38449907e+00 -3.32634598e-01
1.08747520e-01 3.12736660e-01 5.89874327e-01 8.14548075e-01
1.97679847e-01 2.78792143e-01 -3.53453338e-01 9.31072757e-02
7.32449949e-01 2.99385846e-01 -4.13622797e-01 -5.59406579e-01
-1.43880451e+00 5.04929721e-01 5.09351432e-01 2.29157165e-01
-1.41587347e-01 9.40279067e-02 1.49952278e-01 2.54992932e-01
5.57598412e-01 1.50038041e-02 -6.65655494e-01 3.01881611e-01
-1.13022935e+00 -3.41403633e-01 5.02272725e-01 1.27121735e+00
9.76129591e-01 -4.59884964e-02 -1.15384199e-01 6.88022614e-01
5.70817947e-01 6.56624377e-01 5.08282185e-01 -1.12037706e+00
2.25858003e-01 5.53932607e-01 -3.67394984e-02 -5.42114913e-01
-1.41301692e-01 -4.55489397e-01 -9.25456047e-01 6.98786676e-02
8.45563173e-01 -4.63632792e-01 -1.44018376e+00 1.95048273e+00
6.11118019e-01 4.78493094e-01 -8.77401140e-03 9.62786734e-01
9.20843065e-01 4.67878431e-01 -7.32434765e-02 -2.31677592e-01
1.20011091e+00 -1.39001167e+00 -9.18253601e-01 -3.74174714e-01
6.69074953e-01 -6.89665437e-01 1.04816723e+00 4.66220379e-01
-9.92979407e-01 -4.51017827e-01 -9.58626032e-01 -1.35300159e-01
-1.51112229e-01 4.51490581e-01 3.25859368e-01 4.72961545e-01
-9.24980581e-01 3.81888449e-01 -1.15024209e+00 1.14008794e-02
8.97883534e-01 2.44020939e-01 -3.55988801e-01 -3.65926415e-01
-7.97667265e-01 3.55079949e-01 3.69262129e-01 6.64898306e-02
-1.23376942e+00 -7.14863896e-01 -6.95182681e-01 -2.67498702e-01
5.40213525e-01 -3.70221496e-01 1.25867867e+00 -1.42656422e+00
-1.15819263e+00 1.17559624e+00 -7.78443292e-02 -9.40623134e-02
7.19598353e-01 -8.73950869e-02 -8.88587236e-02 3.10201883e-01
3.42294574e-01 1.18956721e+00 7.95547247e-01 -1.78601551e+00
-6.68820262e-01 -5.78425348e-01 -4.31164980e-01 1.31686702e-01
-1.87729955e-01 -3.98223698e-01 -6.24071538e-01 -8.09477031e-01
6.34880662e-01 -1.20440805e+00 -1.42803654e-01 1.06959455e-01
-7.97212362e-01 8.03498998e-02 9.10776675e-01 -5.51141381e-01
8.37575078e-01 -2.21687436e+00 1.90924257e-02 4.00302112e-02
3.01604509e-01 -7.26916268e-02 -9.82734859e-02 -2.02067584e-01
-4.02970999e-01 -6.46156818e-02 -6.69820487e-01 -3.51446569e-01
-3.45776528e-01 5.57554007e-01 -4.00975466e-01 7.51563728e-01
3.70448269e-02 1.02575338e+00 -1.30500948e+00 -8.65342140e-01
-3.62044126e-02 3.58205646e-01 -1.65522322e-01 2.32325941e-01
-2.02778742e-01 1.00875163e+00 -4.63085383e-01 8.37011874e-01
7.80076981e-01 -6.85032904e-01 3.00816566e-01 -2.03482687e-01
4.59443063e-01 -9.62455794e-02 -1.06585383e+00 1.89602149e+00
-4.08793353e-02 4.92938191e-01 2.15658769e-01 -1.10194135e+00
7.15134740e-01 4.86594468e-01 5.42397082e-01 -3.49731952e-01
2.76361525e-01 3.37458909e-01 -1.02627829e-01 -5.43123960e-01
1.99437188e-03 -3.65464836e-01 2.10440189e-01 7.94407368e-01
3.45546603e-01 -1.12817973e-01 5.30717894e-02 3.04198176e-01
5.46851397e-01 3.96653861e-01 -2.10448921e-01 -3.63217115e-01
2.11567432e-01 2.98141599e-01 8.74383390e-01 4.96556312e-01
-5.85606694e-01 9.81427908e-01 5.16255796e-01 -1.55251622e-01
-7.03683257e-01 -1.23506474e+00 -1.91334501e-01 9.86077070e-01
5.71620524e-01 3.09907973e-01 -9.77295041e-01 -1.19782293e+00
-1.22840337e-01 5.24989426e-01 -7.65048921e-01 -1.07496360e-03
-4.02984470e-01 -7.67034471e-01 5.25968790e-01 6.92533672e-01
4.86857712e-01 -9.40322042e-01 -1.67037785e-01 -2.48485491e-01
-2.86992103e-01 -9.86011803e-01 -8.11056912e-01 5.70788324e-01
-1.13990724e+00 -1.07468402e+00 -9.76978600e-01 -1.14908659e+00
1.42481506e+00 4.50261652e-01 9.81615305e-01 -3.67013030e-02
-9.37724859e-02 1.84542343e-01 -9.86923948e-02 -2.16801807e-01
-4.51870829e-01 -3.93014774e-02 -2.09129304e-01 1.58841223e-01
2.59037703e-01 -4.06633943e-01 -6.71065807e-01 6.11516893e-01
-1.04174721e+00 3.03200155e-01 4.90021735e-01 1.30919206e+00
1.04740536e+00 -2.78693736e-01 5.68728447e-01 -1.59221196e+00
-6.87254146e-02 -6.56996906e-01 -4.71829206e-01 4.40460533e-01
-7.01400459e-01 -7.53141567e-02 2.96114951e-01 -5.61726034e-01
-1.24917519e+00 4.96316344e-01 2.78438509e-01 -7.86353469e-01
-4.69735086e-01 9.93754193e-02 -2.28283957e-01 2.99032986e-01
4.97498274e-01 2.25114748e-01 2.61221170e-01 -3.12425196e-01
7.38552272e-01 8.10912848e-01 8.34948003e-01 -7.20787108e-01
6.26663983e-01 9.16244864e-01 -4.95964557e-01 -3.42682391e-01
-1.22690821e+00 -5.05288720e-01 -8.61890435e-01 -1.09457478e-01
1.03669810e+00 -1.00886083e+00 -1.58302039e-01 5.96708655e-01
-8.10321569e-01 -7.46132672e-01 -6.86123967e-01 5.13279736e-01
-4.23333704e-01 2.60738134e-01 -8.35540891e-01 -2.30260655e-01
-9.05723870e-02 -1.50266016e+00 1.06556046e+00 5.35903752e-01
-1.51596377e-02 -1.12277842e+00 -1.46606222e-01 7.09639788e-01
-1.03828438e-01 1.39958680e-01 7.67613471e-01 -8.97745311e-01
-5.64981341e-01 -1.14293240e-01 -3.65942568e-01 6.55092299e-01
5.87760031e-01 3.61209437e-02 -1.10551667e+00 -2.91032642e-01
1.87572569e-01 -6.72597349e-01 9.08997655e-01 5.61313093e-01
1.21943486e+00 -2.33497526e-02 -5.37452161e-01 6.51703894e-01
1.01496494e+00 3.16124819e-02 1.58034042e-01 -1.23795196e-01
9.58427191e-01 8.85970354e-01 7.57683635e-01 7.43288640e-03
3.85773361e-01 -1.09608015e-02 3.10608298e-01 -5.96843600e-01
-3.96469712e-01 -3.62190425e-01 9.91861746e-02 8.97353411e-01
5.69124401e-01 -1.61113411e-01 -9.98678207e-01 7.78553069e-01
-1.83182120e+00 -5.41882396e-01 -1.90946192e-01 2.02557182e+00
1.36550307e+00 3.69817093e-02 -6.48398772e-02 -1.61196619e-01
9.54962373e-01 -1.46376519e-02 -9.13165808e-01 3.64918739e-01
-7.12448284e-02 -2.33210132e-01 6.62152588e-01 5.38120091e-01
-1.16133380e+00 8.30024540e-01 5.81861210e+00 8.36675167e-01
-1.31330597e+00 3.08322370e-01 1.43086338e+00 -1.75999910e-01
-4.92723405e-01 -8.70440062e-03 -5.94163656e-01 5.41721702e-01
4.72042292e-01 1.49795145e-01 1.02816522e-01 9.24381256e-01
3.98979634e-02 -1.55883864e-01 -1.29208636e+00 8.38695347e-01
9.96542815e-03 -1.28855634e+00 -3.57601866e-02 4.28318530e-02
1.19732785e+00 2.53887534e-01 2.54339099e-01 -4.53320630e-02
6.44173920e-01 -8.56926501e-01 8.37314785e-01 1.66351736e-01
9.34627652e-01 -2.81243145e-01 7.36520171e-01 5.88132143e-01
-6.66502237e-01 2.26634040e-01 -8.13837722e-02 4.39840555e-01
6.95036501e-02 7.54922390e-01 -6.59850717e-01 3.44606072e-01
6.28041804e-01 8.90373766e-01 -3.55339438e-01 5.39948404e-01
-5.35977364e-01 1.00945210e+00 -2.26149619e-01 5.37247777e-01
3.37061733e-01 -5.72986782e-01 8.02018121e-02 9.40030217e-01
-1.36354774e-01 7.24395737e-02 4.30637032e-01 1.00871968e+00
-3.59437078e-01 -2.48377427e-01 -3.82279873e-01 8.50256067e-03
3.99096787e-01 1.33860481e+00 -1.17399287e+00 -6.76624537e-01
-3.83690387e-01 8.88947845e-01 8.48380625e-02 6.53466463e-01
-8.92918825e-01 8.88643861e-02 -1.88767686e-02 -2.06310395e-02
-6.71838177e-03 2.10392803e-01 -6.76575959e-01 -1.11506152e+00
-6.97348565e-02 -7.27640450e-01 5.03662527e-01 -8.05628419e-01
-1.47737610e+00 5.38984954e-01 1.02023289e-01 -1.37025809e+00
1.45371467e-01 -3.61903310e-01 -3.79023165e-01 7.40036130e-01
-1.50071537e+00 -1.16539896e+00 -4.11031812e-01 5.84911227e-01
5.16189933e-01 3.60788256e-02 3.59172046e-01 4.51390237e-01
-5.57106793e-01 4.84442502e-01 1.96353883e-01 5.23397386e-01
9.91469979e-01 -1.56732881e+00 -5.99922687e-02 7.60650277e-01
3.90712082e-01 4.62873459e-01 2.95203745e-01 -6.92567885e-01
-1.05740166e+00 -1.16768599e+00 4.68733042e-01 -3.88789147e-01
5.35524309e-01 -2.06138402e-01 -1.11103690e+00 1.03667843e+00
4.12433743e-01 5.79608083e-01 8.62714827e-01 -4.00386542e-01
-2.93126404e-01 -7.33140483e-02 -1.32762182e+00 4.93608534e-01
7.81524777e-01 -3.37849796e-01 -6.41550243e-01 5.26131332e-01
9.76999164e-01 -6.54937863e-01 -5.53172827e-01 3.30261886e-01
2.64561415e-01 -8.04566324e-01 6.57007337e-01 -5.56253970e-01
5.03168881e-01 -4.85778421e-01 -1.01213790e-01 -1.15979505e+00
2.42179021e-01 -5.59357405e-01 1.97551638e-01 1.23220003e+00
5.55067539e-01 -5.97885430e-01 1.20358396e+00 6.30792618e-01
-1.45838216e-01 -1.07564950e+00 -7.32756138e-01 -5.51365256e-01
4.24986362e-01 -2.32671425e-01 2.44336277e-01 1.35535622e+00
-1.76715568e-01 2.26624280e-01 -1.80419564e-01 2.92930365e-01
7.87239492e-01 3.12342644e-01 3.65130723e-01 -7.66050577e-01
-2.42714882e-01 -1.96822271e-01 1.94188088e-01 -1.29778373e+00
1.35799497e-01 -1.17824590e+00 3.75380635e-01 -1.28613460e+00
4.71538812e-01 -6.48670137e-01 -3.53520155e-01 7.97157109e-01
-3.55222225e-01 4.03072268e-01 -6.60481229e-02 3.41780424e-01
-4.68291253e-01 4.57004666e-01 1.77669048e+00 -4.29731131e-01
-8.92964303e-02 -1.08905025e-01 -7.27001429e-01 9.40451801e-01
7.88905323e-01 -8.79951417e-01 -8.15879583e-01 -7.09432185e-01
-3.76891643e-01 1.44474134e-01 2.16552436e-01 -4.44280833e-01
3.02880853e-01 -1.71768650e-01 2.65235126e-01 -5.06366372e-01
-1.88999251e-01 -9.19177830e-01 -1.59963369e-01 4.89691615e-01
-5.89552522e-01 -3.34770590e-01 -1.36626244e-01 6.95141435e-01
-4.21068549e-01 -1.59853518e-01 1.04854512e+00 -1.24553651e-01
-4.19702321e-01 4.33488756e-01 1.38906986e-01 4.26469266e-01
1.11080682e+00 -1.52563199e-01 -5.37772834e-01 -1.64772764e-01
-1.14031327e+00 6.32186174e-01 4.59563375e-01 1.65277824e-01
5.04222810e-01 -1.12617922e+00 -5.56849897e-01 3.24769229e-01
-3.20888422e-02 6.45535827e-01 3.29463124e-01 1.08022845e+00
-5.35417676e-01 -4.68986556e-02 -4.89029149e-03 -1.13316357e+00
-1.03725708e+00 6.09718561e-01 4.37943369e-01 -9.36267078e-02
-5.06307483e-01 1.11832452e+00 5.69216192e-01 -8.13872099e-01
4.99328494e-01 -4.80213732e-01 2.39810795e-01 3.46034393e-02
2.70268142e-01 2.54353553e-01 -2.81426698e-01 -4.56778705e-01
-1.92715988e-01 4.78111893e-01 -2.38840908e-01 -6.90023601e-02
1.13549197e+00 -3.16867679e-01 -1.42702222e-01 6.38154149e-01
1.22339463e+00 -2.33697489e-01 -1.65838444e+00 -4.17442948e-01
-4.62628342e-02 -3.29613090e-01 -1.58537924e-02 -9.22305286e-01
-1.75498700e+00 1.04745102e+00 7.54989684e-01 -7.75307119e-02
1.02415431e+00 4.00234640e-01 7.26828694e-01 9.14253443e-02
1.77917317e-01 -9.66310561e-01 2.70642698e-01 -1.01184919e-01
3.94405335e-01 -1.67738175e+00 -1.33085445e-01 -5.93490779e-01
-1.12429428e+00 7.59901702e-01 6.63096011e-01 -1.52174219e-01
9.05974150e-01 5.72165847e-01 7.54511535e-01 -2.49101207e-01
-6.19336963e-01 1.84661467e-02 9.66472626e-02 5.70955038e-01
4.88966733e-01 1.31808266e-01 8.76537859e-02 6.51845276e-01
1.70899510e-01 1.76807091e-01 5.07255077e-01 8.69891405e-01
-1.52697012e-01 -9.97921109e-01 -3.48478854e-01 3.49689156e-01
-3.42847258e-01 1.00905681e-02 -2.93224275e-01 7.83439398e-01
2.73840725e-01 7.92873144e-01 3.83089185e-02 -4.17542011e-02
-1.10595211e-01 -1.48107465e-02 3.03184032e-01 -7.60612071e-01
-2.96856582e-01 5.62496662e-01 -4.43358868e-01 -3.25148225e-01
-5.38701236e-01 -5.64990759e-01 -1.76356614e+00 2.36125112e-01
-3.65327388e-01 1.13101043e-01 2.59708315e-01 8.46867621e-01
1.93555743e-01 6.80841625e-01 6.54052138e-01 -5.00165701e-01
-6.07069552e-01 -7.72847950e-01 -8.24045956e-01 5.39041042e-01
3.29131365e-01 -4.12608594e-01 -4.11609322e-01 6.38933599e-01] | [14.62393569946289, -1.9951493740081787] |
14679da2-3931-4755-a064-d92eb988f48b | an-efficient-multilinear-optimization | 1511.02667 | null | http://arxiv.org/abs/1511.02667v2 | http://arxiv.org/pdf/1511.02667v2.pdf | An Efficient Multilinear Optimization Framework for Hypergraph Matching | Hypergraph matching has recently become a popular approach for solving
correspondence problems in computer vision as it allows to integrate
higher-order geometric information. Hypergraph matching can be formulated as a
third-order optimization problem subject to the assignment constraints which
turns out to be NP-hard. In recent work, we have proposed an algorithm for
hypergraph matching which first lifts the third-order problem to a fourth-order
problem and then solves the fourth-order problem via optimization of the
corresponding multilinear form. This leads to a tensor block coordinate ascent
scheme which has the guarantee of providing monotonic ascent in the original
matching score function and leads to state-of-the-art performance both in terms
of achieved matching score and accuracy. In this paper we show that the lifting
step to a fourth-order problem can be avoided yielding a third-order scheme
with the same guarantees and performance but being two times faster. Moreover,
we introduce a homotopy type method which further improves the performance. | ['Francesco Tudisco', 'Antoine Gautier', 'Quynh Nguyen', 'Matthias Hein'] | 2015-11-09 | null | null | null | null | ['hypergraph-matching'] | ['graphs'] | [ 2.76761174e-01 3.94612193e-01 9.47029516e-02 1.28112137e-01
-8.49481404e-01 -5.09711981e-01 4.59804237e-01 4.62471366e-01
-4.26316738e-01 2.98930436e-01 -1.40263304e-01 -4.28460181e-01
-3.44599694e-01 -8.47654223e-01 -8.04108918e-01 -6.21278465e-01
-7.68189281e-02 8.01183581e-01 5.08777142e-01 -2.57099986e-01
3.66918296e-01 8.03030431e-01 -1.63310611e+00 -8.26202258e-02
9.54846740e-01 9.40846324e-01 2.99225580e-02 7.60754585e-01
-5.94426412e-03 1.43793657e-01 7.16999024e-02 -6.22315407e-01
7.30015337e-01 -3.36224735e-01 -1.10783243e+00 2.31688753e-01
1.17135191e+00 3.57490368e-02 -2.70119399e-01 1.15197694e+00
2.71276265e-01 4.96690609e-02 3.41206580e-01 -1.52529430e+00
-2.33469427e-01 5.04828915e-02 -6.74685717e-01 -3.29715788e-01
5.44472814e-01 -4.19352770e-01 1.29970324e+00 -6.89850330e-01
8.15154850e-01 1.05796957e+00 6.61144793e-01 1.33814812e-01
-1.69836080e+00 -1.79770470e-01 -4.20231223e-01 4.11860079e-01
-1.29714155e+00 1.24873981e-01 6.98010206e-01 -6.51433170e-01
6.90434158e-01 5.91968179e-01 8.09859097e-01 -7.17096543e-03
-2.02489849e-02 4.80854481e-01 9.67802882e-01 -7.27163374e-01
5.38789667e-02 2.97489334e-02 4.45468932e-01 1.05693817e+00
4.43980426e-01 2.12247431e-01 -1.23578578e-01 -2.48653233e-01
7.04426408e-01 -2.88211852e-01 -2.85136700e-01 -8.01621616e-01
-1.46565235e+00 9.98462200e-01 8.07796359e-01 3.99552852e-01
-2.48122349e-01 4.21997122e-02 2.39383444e-01 2.06804946e-01
2.48016223e-01 4.64153618e-01 -5.52868471e-02 3.15824062e-01
-7.44923770e-01 4.58187580e-01 1.11830282e+00 8.52048159e-01
1.13706172e+00 -4.05835211e-01 1.12493992e-01 4.96759564e-01
1.65685806e-02 2.61356533e-01 -2.50352204e-01 -1.04743433e+00
4.04794127e-01 7.74090350e-01 -1.30184710e-01 -1.42358673e+00
-6.14094913e-01 -3.88655216e-01 -8.20104539e-01 4.77089107e-01
6.63278341e-01 4.21695530e-01 -5.77702641e-01 1.60644388e+00
6.48437858e-01 8.51845145e-02 -2.18251422e-01 1.05766535e+00
4.09825891e-01 6.03596628e-01 -4.41968381e-01 -1.93387255e-01
1.44747972e+00 -9.05104876e-01 -2.93361187e-01 4.21206176e-01
6.76495850e-01 -8.79283786e-01 7.28960812e-01 4.06165212e-01
-1.06032467e+00 -2.77146429e-01 -1.24295735e+00 -3.21691364e-01
-8.09770972e-02 -1.28300831e-01 5.99569082e-01 5.69306493e-01
-1.30396903e+00 7.92825639e-01 -4.91239041e-01 -5.15531421e-01
-3.03682089e-02 6.23201489e-01 -7.75982380e-01 7.38461837e-02
-9.31059837e-01 8.63413215e-01 4.06624585e-01 -1.16311327e-01
1.29026845e-01 -1.03975105e+00 -7.48413265e-01 7.09764138e-02
6.53857708e-01 -8.38793278e-01 8.07382047e-01 -4.71124262e-01
-1.44207764e+00 1.22797787e+00 -2.14979574e-01 -2.36313581e-01
8.45404208e-01 1.61563113e-01 1.27406448e-01 3.44485164e-01
1.05890915e-01 5.40947020e-01 4.34311002e-01 -1.06848752e+00
-5.22012293e-01 -6.31549001e-01 4.87526119e-01 1.43186837e-01
-2.47202575e-01 -3.39119464e-01 -6.41999543e-01 -5.15984476e-01
6.82478786e-01 -1.49752760e+00 -3.38932961e-01 2.33370826e-01
-4.03993517e-01 -1.99537516e-01 4.71666366e-01 -5.70168912e-01
1.06833255e+00 -1.85455132e+00 8.18332672e-01 5.28878570e-01
4.04724747e-01 8.91219750e-02 5.74383838e-03 5.46062112e-01
-2.01937869e-01 -1.09775484e-01 -3.86079997e-01 -2.64827520e-01
-5.15210675e-03 1.87111527e-01 2.33503938e-01 7.46263504e-01
3.74229886e-02 7.65089452e-01 -7.27095246e-01 -5.72135925e-01
2.70290881e-01 3.58677328e-01 -8.67134213e-01 -5.68499342e-02
1.01939458e-02 3.58815044e-01 -1.06428914e-01 1.18021756e-01
9.53731954e-01 -1.93453491e-01 1.01193674e-01 -4.74964887e-01
-2.70198256e-01 -1.86397702e-01 -1.63056695e+00 1.71663427e+00
-3.88739198e-01 5.16134024e-01 1.45915672e-01 -1.15050387e+00
8.09747517e-01 1.17912419e-01 8.82944345e-01 -5.32853603e-01
4.11692113e-02 3.73345464e-01 -2.36363322e-01 -3.15331280e-01
5.34364641e-01 -2.39176929e-01 3.71576473e-02 3.80155779e-02
-2.62880921e-01 -1.32512659e-01 3.12959582e-01 2.69284695e-01
8.94584835e-01 -2.42839620e-01 2.28638619e-01 -7.91199446e-01
9.21964824e-01 2.01729253e-01 3.00334096e-01 4.32456434e-01
5.12050569e-01 5.34001172e-01 7.24436820e-01 -4.12266076e-01
-1.22751307e+00 -9.29749072e-01 -2.59381860e-01 5.01308203e-01
2.41514802e-01 -4.60614383e-01 -8.58399510e-01 -3.45287114e-01
2.15409368e-01 9.92906168e-02 -6.83770359e-01 1.58243880e-01
-8.28234494e-01 -4.41372544e-01 1.13751374e-01 2.78634042e-01
2.92979002e-01 -4.58197862e-01 -5.59541047e-01 1.03389390e-01
-1.75308570e-01 -1.28714466e+00 -6.21074021e-01 -1.20415725e-01
-9.56612349e-01 -1.38422656e+00 -9.37717795e-01 -7.83468962e-01
9.15066957e-01 1.98763058e-01 8.51273596e-01 4.38840836e-01
-5.23362875e-01 1.46797881e-01 -1.70294940e-01 1.19991884e-01
-3.59970242e-01 1.99462116e-01 -1.80456638e-01 2.37828210e-01
-7.58895874e-02 -4.19319183e-01 -3.61664563e-01 2.51592994e-01
-1.14717484e+00 1.40594631e-01 3.98845732e-01 6.94703162e-01
6.75718844e-01 -8.93459916e-02 3.32475640e-02 -8.66664588e-01
1.44939765e-01 4.18251976e-02 -1.32815194e+00 2.09843934e-01
-7.46815622e-01 5.75082183e-01 6.31581187e-01 -1.24058217e-01
-4.26047415e-01 6.54611051e-01 1.32094081e-02 -2.19808787e-01
3.03087115e-01 4.93786037e-01 -2.20640004e-01 -8.39040995e-01
4.31745201e-01 -5.38146719e-02 1.08112760e-01 -5.41085303e-01
5.01642227e-01 1.12000331e-01 7.81722784e-01 -5.74945569e-01
1.12724650e+00 6.27446771e-01 1.02305257e+00 -9.31788325e-01
-5.89217424e-01 -9.28605378e-01 -8.90466034e-01 -2.34230474e-01
8.56876135e-01 -3.40952873e-01 -1.15322804e+00 3.59256923e-01
-1.23483264e+00 2.26467848e-01 8.58362466e-02 3.08491588e-01
-8.02691102e-01 7.53837943e-01 -4.57173228e-01 -4.74199176e-01
-1.02615967e-01 -1.24112332e+00 1.04258323e+00 -5.06747775e-02
2.08618596e-01 -8.74347687e-01 2.14640975e-01 3.11711222e-01
4.97549810e-02 5.72681010e-01 1.06284153e+00 -1.69351369e-01
-1.04560840e+00 -1.73363075e-01 -5.60990810e-01 -2.53808856e-01
-4.34373111e-01 -2.47979134e-01 -4.59841281e-01 -3.36274385e-01
-2.03458354e-01 3.37600201e-01 8.73301566e-01 1.98844060e-01
6.28308237e-01 -4.24853295e-01 -2.79386848e-01 8.09947193e-01
1.97983444e+00 -3.37941557e-01 5.73310673e-01 4.88213629e-01
8.31188142e-01 8.92546356e-01 4.92488146e-01 9.55433920e-02
4.33931768e-01 1.46137190e+00 3.88466448e-01 -3.12677205e-01
-1.13103241e-01 -4.37697172e-02 -2.68137783e-01 7.25346923e-01
-1.06949493e-01 3.65671694e-01 -7.56547272e-01 4.74270463e-01
-2.22060561e+00 -8.03790689e-01 -8.78246665e-01 2.61061001e+00
5.53950608e-01 -2.28430733e-01 5.00974059e-01 3.68401647e-01
6.68736339e-01 -3.02582651e-01 -1.24052964e-01 -3.46951395e-01
1.56349525e-01 1.19345129e-01 7.74956226e-01 1.01756012e+00
-9.53822255e-01 6.60792530e-01 5.87586975e+00 5.21868348e-01
-9.90981817e-01 1.63783487e-02 -8.00002217e-02 4.62897360e-01
-3.13070774e-01 2.72391975e-01 -6.82544529e-01 1.23921052e-01
3.23633075e-01 -3.14509988e-01 5.08305430e-01 5.15881956e-01
-2.84939945e-01 -2.44701177e-01 -1.24076760e+00 1.08024669e+00
1.66926622e-01 -1.28106546e+00 -3.10957860e-02 5.15703321e-01
7.52817214e-01 -4.04314697e-01 -1.71986848e-01 -2.74317831e-01
-3.39683563e-01 -6.66983664e-01 4.96977180e-01 2.72332966e-01
7.09199727e-01 -7.93705225e-01 4.67916191e-01 2.66224146e-01
-1.49209058e+00 2.15399534e-01 -3.52534443e-01 9.62625295e-02
2.91716397e-01 6.31950557e-01 -4.98774499e-01 1.00051498e+00
3.25717837e-01 1.70699567e-01 -3.83785546e-01 1.55875051e+00
2.18018696e-01 -2.58182555e-01 -5.58510900e-01 3.59542333e-02
3.02507073e-01 -7.66054153e-01 8.13377738e-01 1.07921803e+00
3.29930156e-01 1.84035599e-01 4.00820732e-01 6.04055047e-01
1.20022148e-01 3.86382312e-01 -5.19983232e-01 5.21764159e-01
-3.50635424e-02 1.28822851e+00 -8.70787799e-01 -1.61372021e-01
-3.21091056e-01 1.16768098e+00 4.52725053e-01 1.15676276e-01
-6.00834489e-01 -4.56252903e-01 4.46179539e-01 3.28779429e-01
3.28349620e-01 -2.30966404e-01 -2.60533631e-01 -1.18122625e+00
4.86536235e-01 -6.11935318e-01 4.23753619e-01 -1.95301041e-01
-9.66531992e-01 7.27144361e-01 1.61998019e-01 -1.03552771e+00
-3.18444997e-01 -8.68347645e-01 -1.31673932e-01 7.79723644e-01
-1.43590522e+00 -1.03040981e+00 -3.39080960e-01 5.35699546e-01
7.42463619e-02 3.49218756e-01 7.60254622e-01 6.29324496e-01
-1.27099335e-01 6.15232766e-01 -1.06255136e-01 -1.58266544e-01
3.16291451e-01 -1.53638101e+00 9.47220251e-02 8.46184850e-01
9.66179743e-02 3.07217717e-01 9.26978171e-01 -3.47568870e-01
-1.57363284e+00 -7.65497208e-01 1.37838674e+00 -2.55463004e-01
7.43199229e-01 -2.03854010e-01 -1.04401910e+00 3.78013343e-01
-1.90265462e-01 -9.75567326e-02 1.41416118e-01 7.93762207e-02
-4.10349458e-01 -2.31666937e-01 -1.12994313e+00 4.81877863e-01
1.02520490e+00 -5.08065999e-01 -3.51022720e-01 4.95598733e-01
6.20548785e-01 -5.71411967e-01 -9.85773623e-01 3.32830131e-01
5.83898425e-01 -8.61836255e-01 1.03847086e+00 -3.90617490e-01
4.09047194e-02 -5.38320065e-01 -1.40266135e-01 -9.73964632e-01
-4.36323762e-01 -9.33329523e-01 9.87183303e-02 9.11637604e-01
3.44534129e-01 -7.35292256e-01 8.51679444e-01 6.12113774e-01
5.55092581e-02 -8.61834288e-01 -1.00448990e+00 -1.06296492e+00
-7.51058832e-02 -1.20849177e-01 2.40588367e-01 8.86477411e-01
2.55192161e-01 2.20529437e-01 -3.25439423e-01 3.65774900e-01
1.05137622e+00 3.60688567e-01 7.75753558e-01 -1.58099830e+00
-4.50600088e-01 -6.56570971e-01 -9.83697712e-01 -1.13435280e+00
2.57136554e-01 -1.18680859e+00 -5.09631857e-02 -1.44077849e+00
3.11575949e-01 -6.35285378e-01 1.84201766e-02 1.75727502e-01
-9.29047093e-02 5.27683914e-01 5.08298516e-01 1.68589447e-02
-1.44768268e-01 1.01803122e-02 1.08871603e+00 4.13764380e-02
-1.75632656e-01 -8.02727789e-03 -3.81657451e-01 2.99345344e-01
4.44177955e-01 -4.21578199e-01 -4.07559946e-02 -2.38465324e-01
3.88084322e-01 3.57633591e-01 5.08338571e-01 -1.03055811e+00
5.74490249e-01 1.64991930e-01 -3.53031337e-01 -6.20209813e-01
3.47952843e-01 -1.01161575e+00 5.16837656e-01 6.64886951e-01
-1.81842417e-01 3.40946943e-01 1.96549058e-01 3.81631017e-01
-1.22440211e-01 -4.72677350e-01 8.32529783e-01 1.78277791e-01
-4.81914699e-01 4.08432633e-01 1.96325183e-01 -1.48072556e-01
1.15046442e+00 -3.08341682e-01 -6.49612471e-02 -2.45413572e-01
-7.49587476e-01 2.30564639e-01 8.06962550e-01 1.97822958e-01
3.54776382e-01 -1.44264877e+00 -8.98714185e-01 4.33558449e-02
1.12420589e-01 -3.00228029e-01 1.43569127e-01 1.20670140e+00
-8.86328816e-01 7.28225648e-01 -1.87676370e-01 -8.17748666e-01
-1.65077603e+00 8.19282055e-01 2.29650229e-01 -5.60111701e-01
-6.93676949e-01 4.63551164e-01 1.49617940e-01 -3.50631803e-01
2.03085884e-01 2.14815084e-02 7.14096501e-02 -9.36528668e-03
2.97234237e-01 7.06709325e-01 3.72382551e-01 -9.52163219e-01
-3.82677883e-01 1.24834549e+00 2.84675330e-01 -1.39338329e-01
1.07826555e+00 9.73212346e-02 -5.51517367e-01 -5.87489121e-02
1.73094964e+00 -1.23037295e-02 -6.72128260e-01 -2.04696223e-01
2.02475682e-01 -5.14042079e-01 8.53272453e-02 -2.99995542e-01
-1.06678069e+00 6.39201105e-01 3.24788511e-01 4.53396797e-01
1.14555502e+00 1.42760694e-01 9.18367028e-01 2.94523358e-01
5.32733560e-01 -6.74682021e-01 -1.13212451e-01 3.32679093e-01
9.81239259e-01 -9.86996591e-01 1.16313621e-01 -1.11332929e+00
-2.36630421e-02 1.33638251e+00 2.46278971e-01 -4.20666099e-01
6.47932887e-01 -7.71840662e-02 -1.68172732e-01 -1.08103894e-01
-1.64448112e-01 -4.90413100e-01 8.89858186e-01 3.30736190e-01
2.10916504e-01 1.47663087e-01 -9.01200175e-01 -3.74759525e-01
-3.57261509e-01 -2.37651154e-01 2.30202064e-01 2.70125479e-01
-2.15886697e-01 -1.56992090e+00 -4.50777769e-01 8.57959315e-02
-3.66816521e-01 3.77507508e-02 -3.31983954e-01 8.48831236e-01
-8.49484876e-02 8.10731351e-01 -1.79946035e-01 -2.42875963e-01
6.09642088e-01 -2.62255877e-01 9.11706865e-01 -3.08787704e-01
-4.57542181e-01 4.87929992e-02 -2.55143102e-02 -8.60047638e-01
-3.71187389e-01 -5.18145621e-01 -1.28739130e+00 -4.71566975e-01
-4.24362421e-01 2.76995212e-01 7.16623843e-01 7.76506007e-01
2.56950527e-01 1.86523959e-01 5.92614174e-01 -8.36732745e-01
-4.57265645e-01 -3.64316761e-01 -4.60930437e-01 5.31131744e-01
3.91391456e-01 -6.18545592e-01 -1.77044436e-01 -1.61320984e-01] | [8.099360466003418, -2.190068244934082] |
9190ab08-8b27-43e9-86d6-2d5de9f9ab4a | improving-object-counting-with-heatmap | 1803.05494 | null | http://arxiv.org/abs/1803.05494v2 | http://arxiv.org/pdf/1803.05494v2.pdf | Improving Object Counting with Heatmap Regulation | In this paper, we propose a simple and effective way to improve one-look
regression models for object counting from images. We use class activation map
visualizations to illustrate the drawbacks of learning a pure one-look
regression model for a counting task. Based on these insights, we enhance
one-look regression counting models by regulating activation maps from the
final convolution layer of the network with coarse ground-truth activation maps
generated from simple dot annotations. We call this strategy heatmap regulation
(HR). We show that this simple enhancement effectively suppresses false
detections generated by the corresponding one-look baseline model and also
improves the performance in terms of false negatives. Evaluations are performed
on four different counting datasets --- two for car counting (CARPK, PUCPR+),
one for crowd counting (WorldExpo) and another for biological cell counting
(VGG-Cells). Adding HR to a simple VGG front-end improves performance on all
these benchmarks compared to a simple one-look baseline model and results in
state-of-the-art performance for car counting. | ['Ian Stavness', 'Shubhra Aich'] | 2018-03-14 | null | null | null | null | ['object-counting'] | ['computer-vision'] | [-4.38311845e-02 -6.37714043e-02 4.01804596e-01 -3.61520022e-01
-3.70018661e-01 -3.52517337e-01 8.37844074e-01 4.41282004e-01
-1.06455517e+00 7.88954914e-01 -1.54498473e-01 -3.86075288e-01
5.61236501e-01 -9.18840706e-01 -8.44524443e-01 -5.65842688e-01
1.68059781e-01 4.19686109e-01 6.45081997e-01 -3.29337493e-02
3.58720571e-01 5.81905425e-01 -1.55756867e+00 5.65832257e-01
5.08435309e-01 5.01514852e-01 -1.11060411e-01 1.13907039e+00
-3.13248307e-01 1.34183764e+00 -9.83649909e-01 -4.54325885e-01
-1.13354720e-01 -1.71168864e-01 -5.45382917e-01 -8.18956077e-01
7.03199804e-01 -3.24625492e-01 -3.03290933e-02 7.71545291e-01
4.81354684e-01 3.73370200e-02 8.17414582e-01 -1.10903084e+00
-6.12494707e-01 3.26508075e-01 -9.99337256e-01 7.01513588e-01
-5.97327463e-02 4.43835855e-01 3.92508119e-01 -9.51712906e-01
4.80867803e-01 1.41625190e+00 1.01711845e+00 8.83902848e-01
-1.40627944e+00 -8.39726329e-01 2.50047892e-01 -3.99532586e-01
-1.24987018e+00 -1.84108526e-01 8.48227646e-03 -7.13603795e-01
1.39396846e+00 2.19026253e-01 5.45879245e-01 8.91744435e-01
-1.88381881e-01 6.15878999e-01 1.22758234e+00 -4.70459133e-01
3.32110286e-01 5.90819418e-02 7.66429305e-01 8.82821798e-01
6.24244869e-01 -3.72571051e-02 -4.12736475e-01 -3.78012434e-02
9.06467974e-01 -9.10710990e-02 1.13221996e-01 2.63076182e-02
-1.21114790e+00 8.82548392e-01 8.01165760e-01 5.12236893e-01
-1.00596272e-03 8.71746063e-01 5.04327655e-01 -3.88247728e-01
7.13196933e-01 4.77263272e-01 -2.65936583e-01 5.85502945e-02
-9.71370161e-01 3.16006631e-01 7.35818207e-01 7.46357441e-01
7.05192387e-01 1.29085451e-01 -1.09870934e+00 6.69163644e-01
-1.48206830e-01 3.17419082e-01 6.66719978e-04 -8.73394072e-01
3.09818327e-01 7.00616300e-01 3.41300219e-01 -7.25258410e-01
-7.54607916e-01 -3.51574451e-01 -9.11011934e-01 5.84260523e-01
1.10151100e+00 -5.28128035e-02 -1.18096578e+00 1.67457318e+00
4.96219024e-02 3.57651353e-01 -5.30935228e-01 7.84076750e-01
9.92372036e-01 6.05498970e-01 6.17627740e-01 1.72938526e-01
1.55439711e+00 -1.11968279e+00 -3.93044740e-01 -2.92164683e-01
9.34001744e-01 1.22167945e-01 1.40381277e+00 2.40580902e-01
-8.94196987e-01 -5.64785540e-01 -1.16439378e+00 -4.91740167e-01
-1.13208532e+00 5.77763617e-01 8.08206320e-01 7.78043747e-01
-1.07923853e+00 5.28703094e-01 -9.83713806e-01 -4.66693133e-01
1.05661428e+00 3.65601689e-01 -1.07010536e-01 2.20173776e-01
-6.50790036e-01 1.05075681e+00 3.26336361e-02 -1.71129674e-01
-7.83793092e-01 -9.67382729e-01 -4.78085399e-01 2.18391240e-01
4.69065495e-02 -7.24452436e-01 1.25004756e+00 -2.99787343e-01
-1.07346070e+00 1.30957687e+00 -3.55545431e-01 -5.58628440e-01
9.37604189e-01 -1.76065192e-01 3.23713094e-01 -9.08108726e-02
3.21808428e-01 1.24576378e+00 3.01137686e-01 -1.35739851e+00
-5.40783644e-01 -3.42032075e-01 7.86662474e-03 -5.04082739e-01
-1.54830366e-01 1.46888718e-01 -3.20735067e-01 -5.35666764e-01
-6.27019942e-01 -3.74785751e-01 -1.51919082e-01 3.10340673e-01
-3.89855951e-01 -2.75775403e-01 7.26337790e-01 -4.65304255e-01
1.10109115e+00 -1.74214613e+00 -4.76130366e-01 -1.15886919e-01
5.13076127e-01 5.13320088e-01 -7.69791231e-02 -2.27101907e-01
-2.11252034e-01 4.62617457e-01 -4.01952147e-01 -7.89313436e-01
-3.94573152e-01 1.15002066e-01 -7.72714317e-02 5.29497921e-01
7.98645079e-01 1.18811095e+00 -1.21800411e+00 -6.12259209e-01
4.18818891e-01 9.19785976e-01 -3.11868757e-01 -1.19818963e-01
-2.47004718e-01 6.72324181e-01 8.36519524e-02 6.10279858e-01
8.12868059e-01 -4.85255510e-01 -2.70518899e-01 7.95651749e-02
-3.52120429e-01 -8.52140933e-02 -6.94635630e-01 1.18552744e+00
-5.74306846e-01 7.96060860e-01 -3.63037139e-01 -5.33831358e-01
9.25846040e-01 -2.75371552e-01 -4.14966196e-02 -7.34473050e-01
3.65657121e-01 1.61767881e-02 -1.47362143e-01 1.08990237e-01
3.53988677e-01 5.84207103e-02 6.76267147e-02 2.91107923e-01
3.32733870e-01 4.04260084e-02 6.89317822e-01 2.38058805e-01
1.27350307e+00 9.96225104e-02 4.26210582e-01 -4.78313982e-01
6.01096869e-01 8.77548978e-02 2.52177477e-01 1.14028788e+00
-3.13922226e-01 8.15547228e-01 9.06462193e-01 -8.19224954e-01
-9.31679726e-01 -8.14124882e-01 -1.83451492e-02 1.44827986e+00
-2.43926600e-01 -3.31026524e-01 -1.23291409e+00 -5.80362082e-01
6.87303990e-02 9.93704081e-01 -1.33209634e+00 1.58873349e-01
-1.00324476e+00 -1.02304101e+00 1.12640059e+00 1.00393844e+00
7.75273561e-01 -1.02989137e+00 -1.08152378e+00 -5.03380932e-02
5.45532349e-03 -1.08094680e+00 -1.26426414e-01 6.01199746e-01
-6.28285468e-01 -1.08058023e+00 -9.98977721e-01 -3.71278971e-01
7.04424024e-01 2.81291813e-01 1.64166009e+00 6.30194128e-01
-8.86799455e-01 2.10357774e-02 7.56935626e-02 -1.00632906e+00
-3.59855413e-01 1.79247215e-01 -4.19375926e-01 -4.87544715e-01
7.78469443e-01 -1.70298949e-01 -5.28600097e-01 2.63086766e-01
-5.86570740e-01 -1.34394551e-02 8.22016820e-02 7.42122412e-01
5.86955488e-01 -7.85170853e-01 4.95869726e-01 -1.10052145e+00
4.86930579e-01 -1.58994421e-01 -1.11398387e+00 2.44981155e-01
-2.57506430e-01 1.18097417e-01 6.30727947e-01 -4.69162554e-01
-8.93675148e-01 2.79891223e-01 2.11803660e-01 -3.52636337e-01
4.33388725e-02 -3.24878991e-01 4.05433148e-01 -1.69688445e-02
1.16800356e+00 -1.40188396e-01 -2.47846246e-01 -1.90238699e-01
3.93250227e-01 6.35424554e-02 8.92507255e-01 -5.03257751e-01
3.74917507e-01 8.30905139e-01 1.62499636e-01 -6.07862353e-01
-9.15913939e-01 -6.23066187e-01 -8.07812154e-01 -3.69691372e-01
1.37623227e+00 -7.81400084e-01 -1.27189112e+00 5.65350235e-01
-1.63890171e+00 -7.10554361e-01 -4.87316847e-01 -2.44229659e-02
-3.92728388e-01 -2.72056669e-01 -7.19753981e-01 -1.14210451e+00
-3.12461257e-01 -7.81275868e-01 1.39819932e+00 3.62776667e-01
-1.33648977e-01 -7.44121373e-01 2.26806089e-01 -2.31083184e-01
6.31397009e-01 5.31457484e-01 7.97404051e-01 -4.97236818e-01
-4.82940406e-01 -1.37667254e-01 -8.58187973e-01 7.23703727e-02
-4.26533937e-01 2.80579478e-01 -1.77996874e+00 1.89800382e-01
-5.84947646e-01 -3.26967865e-01 1.64349580e+00 6.27761781e-01
1.58194935e+00 8.47377330e-02 -6.25104368e-01 5.76994777e-01
1.37342966e+00 -3.70832026e-01 9.13911641e-01 1.17895409e-01
7.37209141e-01 3.81722599e-01 3.59624684e-01 2.10360795e-01
2.37821981e-01 4.09769535e-01 5.43564320e-01 -5.64108729e-01
-2.33744919e-01 -3.82900774e-01 -2.26080865e-01 6.88273832e-02
-6.24417007e-01 -3.13391149e-01 -1.07018900e+00 3.99680763e-01
-2.00563192e+00 -1.08779657e+00 -5.92149854e-01 2.27924871e+00
5.05739510e-01 6.41274229e-02 5.46762168e-01 5.08637987e-02
7.09969819e-01 -5.87931871e-02 -2.98064977e-01 -5.01643419e-01
-3.27995121e-01 3.40272039e-01 8.57212305e-01 4.03665334e-01
-1.27233577e+00 1.02187455e+00 6.58317947e+00 6.63137257e-01
-8.67755234e-01 3.39922994e-01 1.35174727e+00 -6.50831461e-02
3.75828117e-01 -2.29454249e-01 -1.04799867e+00 3.47786278e-01
8.01878989e-01 4.41989541e-01 3.90777946e-01 8.20995569e-01
1.12200782e-01 -5.52444577e-01 -1.32582808e+00 1.04295754e+00
-2.46933684e-01 -1.74014664e+00 -8.44111741e-02 -9.43913758e-02
4.42491233e-01 1.69077724e-01 -3.55537534e-01 7.49920130e-01
7.44857788e-01 -1.46797168e+00 7.16894150e-01 7.89262593e-01
9.90335941e-01 -5.85503280e-01 7.64114797e-01 5.30818820e-01
-1.13084674e+00 1.16934516e-01 -5.06709337e-01 -1.78354442e-01
-8.16984102e-02 5.83580673e-01 -7.61405289e-01 -2.43104607e-01
8.92914176e-01 2.57167369e-01 -1.02106845e+00 1.03872824e+00
-2.06578895e-01 5.72062850e-01 -4.78045911e-01 -4.12864834e-01
9.06581357e-02 9.85214114e-02 2.78283171e-02 2.20080614e+00
-6.62516430e-02 3.12814750e-02 -3.41084421e-01 1.56179464e+00
-3.01972210e-01 -3.67183447e-01 -6.82752490e-01 3.44079852e-01
3.72646570e-01 1.51578104e+00 -1.39372659e+00 -7.06550658e-01
-2.43516639e-01 7.58167326e-01 7.64814317e-01 1.12345949e-01
-1.17967820e+00 -2.74306357e-01 4.06097740e-01 3.39156449e-01
1.47493437e-01 9.15217325e-02 -6.44780457e-01 -7.25401282e-01
-2.73852855e-01 -2.02162012e-01 3.24326187e-01 -9.74365294e-01
-1.29289615e+00 1.84198752e-01 2.51224607e-01 -7.19007015e-01
1.28731094e-02 -1.03986847e+00 -8.97129476e-01 9.08079922e-01
-1.41488349e+00 -1.17707455e+00 -9.49857712e-01 2.94588327e-01
2.27710992e-01 3.31792146e-01 9.01913285e-01 2.91839391e-01
-4.69945729e-01 6.60246730e-01 -5.55576861e-01 5.23720145e-01
4.14586544e-01 -1.54486370e+00 8.37099135e-01 5.96045017e-01
4.54940163e-02 6.08802855e-01 3.23860556e-01 -4.61018145e-01
-5.51525176e-01 -1.33886182e+00 7.66858578e-01 -8.56929064e-01
1.86993137e-01 -7.97724724e-01 -9.86827910e-01 4.46853578e-01
-1.37697339e-01 6.99741483e-01 2.29310930e-01 -1.04670227e-01
-7.50961065e-01 2.82091171e-01 -1.49914420e+00 4.39216882e-01
1.10317409e+00 -2.18027443e-01 -4.77275327e-02 4.06338304e-01
6.28229916e-01 -1.53376818e-01 -1.49839163e-01 6.95065632e-02
5.81039071e-01 -1.13218355e+00 1.02121460e+00 -5.91008842e-01
4.38848823e-01 -5.55168688e-01 1.75496504e-01 -1.05514300e+00
-3.63827795e-01 -1.46011561e-01 -3.45402882e-02 1.07139111e+00
4.15730715e-01 -4.81789500e-01 7.93664634e-01 2.64498711e-01
1.72395304e-01 -4.64439064e-01 -9.12262678e-01 -7.61129558e-01
6.09738827e-01 -2.51310199e-01 4.60646093e-01 7.84435034e-01
-1.77890837e-01 5.11338353e-01 4.38665822e-02 -1.57482550e-01
4.45415378e-01 -7.63683200e-01 1.05807972e+00 -1.25031769e+00
-9.67464894e-02 -6.50902927e-01 -3.94380629e-01 -8.26089323e-01
-2.12010905e-01 -6.85604274e-01 7.85836279e-02 -1.56312501e+00
5.54398417e-01 -1.76364467e-01 -3.08568031e-03 6.73723936e-01
-4.35723871e-01 6.26453638e-01 3.94687772e-01 -3.67039554e-02
-6.56634748e-01 -1.15242593e-01 9.77292717e-01 -6.04255795e-02
4.22123596e-02 -2.06289545e-01 -5.91115296e-01 9.42138672e-01
8.07500541e-01 -6.49336040e-01 9.10712034e-02 -6.04375303e-01
3.55667531e-01 -3.57395351e-01 9.80177224e-01 -1.19753981e+00
2.70749897e-01 3.15506727e-01 7.11804926e-01 -7.13329315e-01
3.87288600e-01 -3.08598042e-01 -6.04574442e-01 6.45898163e-01
-2.64997125e-01 8.36652070e-02 6.15320027e-01 6.50950074e-01
2.32795522e-01 -1.53609723e-01 1.17794824e+00 -4.48575139e-01
-2.83464849e-01 -3.25599819e-01 -2.80261457e-01 2.30415940e-01
8.12214375e-01 -5.03275573e-01 -1.01986754e+00 -1.47147954e-01
-5.08887112e-01 1.60922691e-01 1.92467287e-01 -5.02741709e-03
1.69162571e-01 -8.84224176e-01 -8.43096733e-01 -3.13717267e-03
3.58473957e-01 1.64039016e-01 -3.83078098e-01 8.50636065e-01
-8.76491129e-01 4.45165545e-01 -1.50357932e-01 -8.47483575e-01
-1.17258012e+00 5.19787967e-01 5.88996768e-01 -8.02994609e-01
-2.62498081e-01 1.16751349e+00 4.62749928e-01 -7.46637523e-01
3.01098585e-01 -1.04966414e+00 -3.65119278e-01 -8.24451074e-02
9.26952779e-01 7.91052163e-01 2.45175853e-01 -2.44126961e-01
-5.53695917e-01 4.01659697e-01 1.34818286e-01 5.09834476e-02
1.22217405e+00 4.36139107e-01 1.65298253e-01 7.36856401e-01
8.19221973e-01 -1.69319481e-01 -1.34797549e+00 3.04225206e-01
7.01983571e-02 -4.55581874e-01 -2.82979965e-01 -1.03094137e+00
-1.01386678e+00 1.39732921e+00 7.04350770e-01 1.38616916e-02
5.32255888e-01 -1.11011080e-01 -2.41263166e-01 5.65422654e-01
3.70677441e-01 -9.72313643e-01 1.47390008e-01 6.25127792e-01
7.12346077e-01 -1.37370098e+00 2.67256834e-02 -3.90573919e-01
-3.98946255e-01 1.06969416e+00 1.05990422e+00 -5.06167829e-01
3.80108953e-01 8.48967791e-01 -1.86189741e-01 -4.06552404e-01
-5.59043586e-01 -2.31496334e-01 -2.68121600e-01 8.58778775e-01
7.50574052e-01 6.28061593e-02 -6.50873920e-03 3.62719357e-01
3.57834846e-02 4.46723580e-01 4.36648130e-01 7.86939144e-01
-4.09739047e-01 -2.52209246e-01 -4.86635029e-01 6.52567267e-01
-4.08322006e-01 -1.95646316e-01 -7.06480265e-01 1.35614562e+00
1.84008792e-01 7.41991460e-01 6.91570640e-01 3.39779481e-02
2.88939893e-01 -3.90604958e-02 4.99121666e-01 -6.70657694e-01
-1.01273715e+00 -3.62731874e-01 1.67102087e-02 -5.21482944e-01
-3.99686813e-01 -2.06670478e-01 -1.41168833e+00 -5.28049231e-01
-6.31780624e-01 -5.62531590e-01 6.28567457e-01 5.44515014e-01
5.32059111e-02 9.16298211e-01 -3.11874896e-01 -1.03640389e+00
-2.59549081e-01 -1.23879397e+00 -2.27151364e-01 8.26727450e-02
1.48535550e-01 -8.13166738e-01 -2.23821193e-01 1.71297714e-02] | [8.713135719299316, 0.006962099578231573] |
3f7936fa-2bb4-490d-bb29-1772cec28da0 | incorporating-linguistic-constraints-into | null | null | https://aclanthology.org/P19-1515 | https://aclanthology.org/P19-1515.pdf | Incorporating Linguistic Constraints into Keyphrase Generation | Keyphrases, that concisely describe the high-level topics discussed in a document, are very useful for a wide range of natural language processing tasks. Though existing keyphrase generation methods have achieved remarkable performance on this task, they generate many overlapping phrases (including sub-phrases or super-phrases) of keyphrases. In this paper, we propose the parallel Seq2Seq network with the coverage attention to alleviate the overlapping phrase problem. Specifically, we integrate the linguistic constraints of keyphrase into the basic Seq2Seq network on the source side, and employ the multi-task learning framework on the target side. In addition, in order to prevent from generating overlapping phrases of keyphrases with correct syntax, we introduce the coverage vector to keep track of the attention history and to decide whether the parts of source text have been covered by existing generated keyphrases. Experimental results show that our method can outperform the state-of-the-art CopyRNN on scientific datasets, and is also more effective in news domain. | ['Yuxiang Zhang', 'Jing Zhao'] | 2019-07-01 | null | null | null | acl-2019-7 | ['keyphrase-generation'] | ['natural-language-processing'] | [ 1.40772134e-01 2.73786243e-02 -2.19913065e-01 -2.66411398e-02
-1.04955196e+00 -5.90768874e-01 6.49044275e-01 3.42000365e-01
-5.31580031e-01 1.18471873e+00 9.71389472e-01 -3.33172321e-01
1.35946527e-01 -8.66856098e-01 -9.19651508e-01 -7.81539559e-01
1.40212342e-01 2.56062627e-01 3.34221035e-01 -3.42740089e-01
5.53602755e-01 3.96497548e-02 -1.10041940e+00 5.84682345e-01
8.54485333e-01 4.67810065e-01 5.22129893e-01 4.41339016e-01
-8.81491482e-01 7.02200770e-01 -8.93874407e-01 -2.42132396e-01
-8.40719789e-02 -4.67382014e-01 -7.48909652e-01 -3.20636094e-01
4.64947164e-01 -2.34078303e-01 -3.56240660e-01 1.07782364e+00
6.35613859e-01 1.14459895e-01 2.90741265e-01 -1.12189853e+00
-6.26195014e-01 1.31098962e+00 -8.58082652e-01 6.08380914e-01
2.84928858e-01 -2.11212322e-01 1.35377944e+00 -8.00834298e-01
6.86936438e-01 1.43683243e+00 1.80863008e-01 4.91359383e-01
-8.23697388e-01 -1.01561642e+00 5.59936285e-01 1.21023901e-01
-1.05861402e+00 -5.19813858e-02 6.94075108e-01 9.11501702e-03
7.97991574e-01 1.72486424e-01 5.76270819e-01 1.20167613e+00
7.34465837e-01 1.11061883e+00 7.89133668e-01 -1.37957439e-01
-4.30547856e-02 -6.28942549e-02 3.25766683e-01 4.38644171e-01
1.60250917e-01 -2.24072263e-01 -7.51937151e-01 -3.77275378e-01
6.08172119e-01 -1.24857575e-01 -4.49848831e-01 2.83384681e-01
-1.84397662e+00 8.39137256e-01 7.41259605e-02 3.62000614e-01
-5.14610052e-01 2.08352461e-01 6.70611084e-01 1.74020037e-01
6.37547314e-01 6.86352670e-01 -6.84781671e-01 -5.72287925e-02
-8.28921258e-01 8.29552650e-01 8.71877611e-01 1.15742433e+00
5.75549603e-01 -2.64403731e-01 -1.02709222e+00 6.17792547e-01
-5.44135608e-02 3.11612636e-01 6.17645562e-01 -4.06530350e-01
7.22279251e-01 4.71481621e-01 1.62123084e-01 -9.01881993e-01
-2.77947187e-01 -3.79991621e-01 -1.00344229e+00 -6.71777606e-01
-5.72355054e-02 -6.26182795e-01 -9.79604959e-01 1.58582258e+00
3.43859881e-01 2.16955796e-01 9.74584445e-02 6.66654766e-01
1.06410170e+00 1.57895792e+00 6.19432144e-02 -4.14036185e-01
1.82540250e+00 -1.18277776e+00 -9.95040894e-01 -2.32921168e-01
2.44461939e-01 -9.04661298e-01 9.81189191e-01 2.80761868e-01
-9.37529922e-01 -5.16341865e-01 -7.02717602e-01 -2.91357875e-01
-3.68102551e-01 8.46864209e-02 4.56116021e-01 -3.24791707e-02
-8.19108725e-01 3.88130337e-01 -3.94903392e-01 -6.69300705e-02
2.02831432e-01 -4.77048121e-02 6.50701597e-02 1.17628276e-01
-1.69003272e+00 4.76059169e-01 8.25035751e-01 -7.31682256e-02
-8.08340251e-01 -1.14842200e+00 -7.70471215e-01 3.58961493e-01
7.61311173e-01 -6.21351838e-01 1.33398509e+00 -3.29482824e-01
-1.15773106e+00 4.68686193e-01 -3.23349029e-01 -4.05471087e-01
1.87165320e-01 -4.51953769e-01 -3.37849557e-01 1.10937364e-01
4.48695451e-01 7.85290241e-01 7.16792285e-01 -8.55646372e-01
-1.05223143e+00 2.06946313e-01 8.85569677e-02 4.01937455e-01
-2.95721710e-01 2.29695186e-01 -7.50049174e-01 -1.26940489e+00
-2.56890088e-01 -6.18837714e-01 -3.55096698e-01 -3.23342144e-01
-8.75232875e-01 -6.72410369e-01 6.25703037e-01 -6.14304900e-01
1.39095652e+00 -1.98274374e+00 3.02824438e-01 -1.20297223e-01
1.17767565e-01 1.45372272e-01 -3.08533788e-01 6.78706527e-01
1.17085680e-01 4.59628314e-01 -1.37250438e-01 -6.51301630e-03
-1.72194138e-01 3.56501229e-02 -9.81041908e-01 -8.61347243e-02
2.03860700e-01 8.46291780e-01 -1.16085553e+00 -6.41822755e-01
-4.60691422e-01 1.67476937e-01 -2.32638463e-01 4.10459906e-01
-9.11283553e-01 2.86744863e-01 -1.01119947e+00 1.60113037e-01
4.96649653e-01 -2.48034000e-01 -3.71180773e-01 -2.06024244e-01
-4.16326582e-01 7.25589693e-01 -9.88414705e-01 1.78800309e+00
-3.77199590e-01 3.64378601e-01 -6.98840395e-02 -6.14176333e-01
7.61779308e-01 4.96350646e-01 3.74636918e-01 -4.11138862e-01
-9.21008363e-02 -3.77905183e-02 -1.25371441e-01 -2.50373542e-01
8.36666226e-01 2.64801597e-03 -3.51858616e-01 8.51798654e-01
1.03477299e-01 -1.63249239e-01 6.24105513e-01 6.04491055e-01
8.69875193e-01 1.02659456e-01 2.57826269e-01 -5.44363081e-01
6.94721520e-01 -3.78448609e-03 7.33141363e-01 9.86948431e-01
3.51472735e-01 5.17012715e-01 9.10650253e-01 -3.87204260e-01
-1.07139659e+00 -4.58266646e-01 2.17081800e-01 1.14663184e+00
1.33081093e-01 -6.61191761e-01 -5.28923273e-01 -8.15865934e-01
-1.29492879e-01 7.06591964e-01 -7.07941651e-01 -2.42024474e-02
-8.38797748e-01 -9.25580740e-01 5.23139000e-01 3.77356708e-01
4.08839136e-01 -1.41869175e+00 -2.76324242e-01 6.13863349e-01
-4.97804552e-01 -1.04845774e+00 -1.05821395e+00 1.54179439e-01
-3.53615195e-01 -7.38034546e-01 -1.28872454e+00 -8.45637441e-01
4.45913643e-01 3.64437133e-01 1.15913606e+00 -1.16793841e-01
-1.62990447e-02 -1.41764596e-01 -6.14256322e-01 -6.59870625e-01
-2.09253684e-01 4.99333829e-01 -3.04136515e-01 -2.23969430e-01
2.35505223e-01 -2.54930168e-01 -3.52746367e-01 -1.19024850e-01
-1.06070471e+00 3.51883769e-01 7.68126070e-01 8.66826594e-01
7.13284910e-01 1.97868109e-01 7.11745620e-01 -1.21374011e+00
1.02771056e+00 -5.77179074e-01 -4.98218358e-01 3.74511719e-01
-4.11747366e-01 4.88934129e-01 9.62689757e-01 -4.41234112e-01
-1.15394652e+00 -5.15474916e-01 -2.22229674e-01 -5.11902533e-02
8.79346356e-02 8.38175893e-01 -1.10627733e-01 5.16695023e-01
2.59042531e-01 7.70952046e-01 -5.04239380e-01 -5.40164769e-01
4.08677042e-01 3.89466554e-01 3.06023210e-01 -7.37076938e-01
8.29463661e-01 2.30115265e-01 -1.97068706e-01 -6.72497213e-01
-1.42949247e+00 -5.90193748e-01 -1.52021587e-01 2.23614529e-01
8.21927607e-01 -1.05487752e+00 -3.42896849e-01 4.17069554e-01
-1.67157352e+00 8.99662375e-02 -1.22268973e-02 1.53328553e-01
-4.69912104e-02 4.83427554e-01 -7.43076801e-01 -4.08890903e-01
-1.02451086e+00 -1.02476501e+00 1.29300284e+00 5.45816183e-01
-4.84218597e-02 -5.90515435e-01 8.93477052e-02 -1.90303728e-01
2.37299904e-01 3.36963795e-02 1.20687127e+00 -9.46969748e-01
-5.86906135e-01 2.29079768e-01 -3.00903469e-01 -7.04381168e-02
2.21266165e-01 -1.10700682e-01 -5.28960526e-01 -1.86595216e-01
-2.66327739e-01 -2.98025876e-01 1.30862415e+00 1.50448695e-01
1.42155015e+00 -7.25943148e-01 -5.21642506e-01 2.88689256e-01
1.05464709e+00 1.92143202e-01 3.99393678e-01 2.39368290e-01
7.79411077e-01 6.88084126e-01 5.84169030e-01 3.11292529e-01
4.73323464e-01 2.40324214e-01 1.25077171e-02 -7.42886961e-02
1.64449990e-01 -4.83495563e-01 4.25849050e-01 9.26713705e-01
5.09301543e-01 -6.24317646e-01 -6.82560146e-01 6.39103413e-01
-1.72597480e+00 -9.22881782e-01 -1.92492679e-01 1.66882837e+00
1.38094175e+00 4.48703945e-01 -9.80030373e-02 -2.91124761e-01
7.36643136e-01 6.82341218e-01 -3.66700888e-01 -2.68144906e-01
-1.72182962e-01 1.72607616e-01 3.83464217e-01 3.49336028e-01
-1.10174131e+00 1.25294161e+00 5.67447996e+00 1.05620039e+00
-9.99758303e-01 -2.00294003e-01 5.62920570e-01 7.43913800e-02
-7.24128604e-01 -1.13131903e-01 -1.41077387e+00 5.37569940e-01
6.37259781e-01 -7.26689398e-01 4.92382944e-02 6.88116908e-01
2.38711700e-01 -4.54849526e-02 -8.92054498e-01 4.71484393e-01
1.13015935e-01 -1.49829614e+00 5.48861086e-01 -2.18140364e-01
7.38550425e-01 -1.50634438e-01 -1.86385289e-01 4.66237158e-01
4.12471741e-01 -6.60789549e-01 6.16921663e-01 3.92029285e-01
3.91398519e-01 -8.91575754e-01 8.37783098e-01 5.30521631e-01
-1.15777206e+00 3.03694785e-01 -6.67415440e-01 2.93729454e-01
3.26835126e-01 9.42876160e-01 -6.53726876e-01 6.77212238e-01
5.73603094e-01 6.24489605e-01 -2.06426963e-01 9.76351321e-01
-6.32978022e-01 6.22169435e-01 -1.79060102e-01 -6.44826114e-01
6.60485148e-01 2.03770343e-02 6.83290541e-01 1.57429564e+00
2.59564519e-01 2.89171338e-01 3.71462941e-01 9.51529086e-01
-3.78572941e-01 3.46533388e-01 -2.40540341e-01 -4.88079756e-01
4.51473117e-01 1.49648106e+00 -6.19394541e-01 -6.01067066e-01
-3.11746091e-01 9.16470528e-01 1.61697790e-02 2.62125582e-01
-6.35608912e-01 -8.84449482e-01 4.73477811e-01 -2.60496259e-01
3.56155902e-01 -3.92850041e-02 9.76214930e-02 -1.20946097e+00
5.36778308e-02 -1.08515978e+00 3.82583678e-01 -7.96632051e-01
-1.36599994e+00 6.47898316e-01 1.85778476e-02 -8.12267721e-01
2.47914605e-02 -2.31100202e-01 -9.83050466e-01 1.10364830e+00
-1.71023130e+00 -1.09540474e+00 -2.96817650e-03 1.73321024e-01
9.36749876e-01 2.78076273e-03 6.50573254e-01 1.44121489e-02
-4.79628354e-01 2.76905477e-01 9.13197268e-03 2.55026668e-01
9.37472880e-01 -1.43077040e+00 9.29821134e-01 8.47272336e-01
2.11893637e-02 8.53463054e-01 7.98860133e-01 -9.57017422e-01
-1.29751372e+00 -1.16264176e+00 1.18893766e+00 1.17654866e-02
7.74554372e-01 -5.55322409e-01 -1.02228630e+00 4.55477834e-01
5.88035464e-01 -4.73079741e-01 5.65112293e-01 1.65672544e-02
-2.00680569e-01 -3.16480100e-02 -4.27759737e-01 9.17563200e-01
6.31978214e-01 -1.98540449e-01 -9.31012571e-01 5.42113781e-01
1.52236795e+00 -6.82841659e-01 -4.87442762e-01 2.44719446e-01
3.57189655e-01 -3.11606228e-01 7.26114690e-01 -6.97039187e-01
9.83829677e-01 -2.09562913e-01 4.77507889e-01 -1.58362305e+00
-4.99289483e-01 -8.72098029e-01 -7.08849728e-02 1.41696119e+00
6.65175557e-01 -1.83496669e-01 6.08666182e-01 -7.25943968e-02
-2.93298513e-01 -8.61369193e-01 -6.83102667e-01 -3.24053496e-01
1.97922722e-01 2.63653733e-02 6.60603344e-01 8.04218471e-01
1.48802325e-01 8.73473287e-01 -6.47282898e-01 -7.00227097e-02
3.73829216e-01 3.72979522e-01 5.97885251e-01 -9.14774477e-01
-1.74664959e-01 -4.45653468e-01 5.09986579e-01 -1.51056993e+00
2.91586906e-01 -7.79297650e-01 3.16538870e-01 -1.75418913e+00
3.87084395e-01 -1.98434159e-01 -2.47536436e-01 5.17105222e-01
-8.60125721e-01 -3.70560259e-01 6.57093823e-02 1.10298157e-01
-5.59262812e-01 7.78936088e-01 1.61868870e+00 -3.53425205e-01
-1.43557355e-01 -9.50997621e-02 -1.04518652e+00 3.87022853e-01
6.42827988e-01 -6.27949595e-01 -3.26453090e-01 -4.91896719e-01
5.37730694e-01 2.03741983e-01 -1.80282705e-02 -5.44800758e-01
4.26770419e-01 -4.78581160e-01 1.58321023e-01 -1.05741954e+00
-1.79942057e-01 -3.65760446e-01 -3.97790045e-01 3.53724748e-01
-8.05252373e-01 1.99545562e-01 4.67626691e-01 6.32560372e-01
-1.80986598e-01 -2.85409123e-01 3.35197777e-01 -5.90584576e-01
-5.48268318e-01 6.02925181e-01 -5.74028194e-01 3.28434736e-01
7.62174368e-01 5.69745123e-01 -5.54932296e-01 -2.13076577e-01
-1.10985905e-01 7.93658912e-01 -9.43971649e-02 6.99476600e-01
5.88730574e-01 -1.09880161e+00 -1.07961118e+00 -5.49145378e-02
1.57745108e-01 4.53367203e-01 1.21577762e-01 3.08494449e-01
-3.68663967e-01 8.22280586e-01 -1.12541821e-02 2.14389283e-02
-9.31976259e-01 6.38992310e-01 -7.69113302e-02 -6.29191458e-01
-7.49736547e-01 1.04936969e+00 4.52900976e-01 -2.84078360e-01
3.93799186e-01 -6.24743581e-01 -4.74816620e-01 2.67724961e-01
1.00002027e+00 -8.76956508e-02 -1.25988826e-01 -1.22443981e-01
-3.83936949e-02 2.82899797e-01 -9.02639925e-01 -7.63118938e-02
1.20568419e+00 8.90709832e-02 -4.89168793e-01 5.32200515e-01
1.03784752e+00 2.51663804e-01 -8.19969833e-01 -4.49523032e-01
2.97706854e-02 1.45493492e-01 1.16973169e-01 -7.90654898e-01
-8.04167211e-01 7.77108610e-01 -2.59428024e-01 2.00907201e-01
8.43600690e-01 5.79329245e-02 1.32430804e+00 4.68900800e-01
1.49965540e-01 -9.55964923e-01 7.58287162e-02 8.18828106e-01
9.70549166e-01 -9.29039717e-01 2.14248553e-01 -4.80086416e-01
-6.23013794e-01 1.07799351e+00 7.90120363e-01 1.33167338e-02
4.64658648e-01 2.86009520e-01 -8.16948265e-02 -1.56591043e-01
-1.11610723e+00 1.15362272e-01 3.08982909e-01 9.09065157e-02
5.32049179e-01 -3.30596805e-01 -7.67607093e-01 7.60003269e-01
-2.94128269e-01 -1.71987727e-01 7.28015006e-01 8.07823956e-01
-7.64617801e-01 -1.08291292e+00 -2.27449343e-01 5.60732305e-01
-1.00824678e+00 -7.29481518e-01 -5.05742073e-01 3.61154348e-01
3.29930754e-03 5.88832319e-01 5.75933941e-02 -1.33212522e-01
1.63585663e-01 8.63770843e-02 -2.97249630e-02 -9.70796227e-01
-8.97561967e-01 3.90865564e-01 1.05701827e-01 -1.87162846e-01
-2.19177961e-01 -4.57021773e-01 -1.19867241e+00 -1.62450328e-01
-3.05907667e-01 7.45947242e-01 2.38952279e-01 1.01355052e+00
4.25723851e-01 8.71611774e-01 5.42276919e-01 -4.85340178e-01
-7.24082649e-01 -1.26111853e+00 -4.07447666e-01 9.33124963e-03
5.26061654e-01 -5.31748608e-02 -6.59905970e-02 9.77692977e-02] | [12.323901176452637, 9.02120590209961] |
00256ce1-0d6a-4062-a96d-bd7f7db42bc1 | maniskill2-a-unified-benchmark-for | 2302.04659 | null | https://arxiv.org/abs/2302.04659v1 | https://arxiv.org/pdf/2302.04659v1.pdf | ManiSkill2: A Unified Benchmark for Generalizable Manipulation Skills | Generalizable manipulation skills, which can be composed to tackle long-horizon and complex daily chores, are one of the cornerstones of Embodied AI. However, existing benchmarks, mostly composed of a suite of simulatable environments, are insufficient to push cutting-edge research works because they lack object-level topological and geometric variations, are not based on fully dynamic simulation, or are short of native support for multiple types of manipulation tasks. To this end, we present ManiSkill2, the next generation of the SAPIEN ManiSkill benchmark, to address critical pain points often encountered by researchers when using benchmarks for generalizable manipulation skills. ManiSkill2 includes 20 manipulation task families with 2000+ object models and 4M+ demonstration frames, which cover stationary/mobile-base, single/dual-arm, and rigid/soft-body manipulation tasks with 2D/3D-input data simulated by fully dynamic engines. It defines a unified interface and evaluation protocol to support a wide range of algorithms (e.g., classic sense-plan-act, RL, IL), visual observations (point cloud, RGBD), and controllers (e.g., action type and parameterization). Moreover, it empowers fast visual input learning algorithms so that a CNN-based policy can collect samples at about 2000 FPS with 1 GPU and 16 processes on a regular workstation. It implements a render server infrastructure to allow sharing rendering resources across all environments, thereby significantly reducing memory usage. We open-source all codes of our benchmark (simulator, environments, and baselines) and host an online challenge open to interdisciplinary researchers. | ['Hao Su', 'Rui Chen', 'Zhiao Huang', 'Pengwei Xie', 'Xiaodi Yuan', 'Yunchao Yao', 'Xinyue Wei', 'Stone Tao', 'Yihe Tang', 'Tongzhou Mu', 'Xiqiang Liu', 'Zhan Ling', 'Xuanlin Li', 'Fanbo Xiang', 'Jiayuan Gu'] | 2023-02-09 | null | null | null | null | ['robot-manipulation'] | ['robots'] | [-2.25888208e-01 -3.10562730e-01 -1.54897735e-01 2.86268204e-01
-8.03228915e-02 -8.83077264e-01 5.79994380e-01 -2.29004800e-01
-4.71561223e-01 6.61234736e-01 -2.32413441e-01 -3.55052471e-01
-1.88097715e-01 -6.76551640e-01 -8.95628512e-01 -4.91242319e-01
-4.54569459e-01 8.32352102e-01 4.81640130e-01 -8.04040194e-01
-3.93282212e-02 6.21212780e-01 -1.94992173e+00 2.09267795e-01
4.54606801e-01 9.89311576e-01 4.43838239e-01 8.79268408e-01
3.29202712e-01 4.22036350e-01 -6.47063732e-01 7.48815611e-02
5.98475754e-01 -9.22329128e-02 -4.24816668e-01 -2.23568454e-01
2.45445594e-01 -3.73603284e-01 -2.89907545e-01 5.49380660e-01
8.19764495e-01 4.66893315e-01 3.10791761e-01 -1.70374656e+00
-2.50822216e-01 3.04099977e-01 -3.12143594e-01 4.01841365e-02
7.09584117e-01 1.13124132e+00 3.98420542e-01 -4.98710185e-01
8.32015216e-01 1.48751688e+00 4.31146175e-01 7.08644331e-01
-1.04185855e+00 -6.17697358e-01 3.22045714e-01 1.89087000e-02
-9.27643836e-01 -1.60663366e-01 3.01360041e-01 -2.80921549e-01
1.27050376e+00 5.15711784e-01 1.36871314e+00 1.90104723e+00
4.84070480e-01 7.07408965e-01 1.16210270e+00 1.96772031e-02
4.76618201e-01 -5.13950408e-01 -3.06995392e-01 7.16043353e-01
2.23239496e-01 4.06503648e-01 -5.86619377e-01 1.17688384e-02
1.21132827e+00 -4.87681516e-02 -2.15142190e-01 -7.94130802e-01
-1.82475078e+00 2.15347201e-01 4.58216786e-01 -1.63611352e-01
-3.05726558e-01 7.83797622e-01 5.29184699e-01 4.40916032e-01
-1.86686143e-01 6.15655005e-01 -6.31116450e-01 -6.53284311e-01
-3.02373052e-01 1.14829350e+00 9.14284885e-01 1.43817043e+00
3.50713760e-01 1.14531063e-01 -2.30348378e-01 1.98082402e-01
-5.80202937e-02 7.78904855e-01 9.44397002e-02 -1.44237697e+00
4.31802988e-01 4.84086603e-01 3.82012278e-01 -8.68490458e-01
-8.96731853e-01 -1.72850713e-01 -7.15329707e-01 1.04278529e+00
3.66595626e-01 -2.15009838e-01 -8.85067642e-01 1.38901877e+00
6.87270284e-01 -2.05653794e-02 -2.10552350e-01 1.28788137e+00
8.00821900e-01 5.18659949e-01 -7.32982457e-02 3.41907173e-01
1.30393255e+00 -1.31618094e+00 -5.01996338e-01 -1.85743049e-01
3.97023827e-01 -4.63426203e-01 1.70566559e+00 7.08857000e-01
-1.42386091e+00 -5.95622480e-01 -9.51521516e-01 -2.04056635e-01
-5.27328074e-01 -1.73362896e-01 8.14687490e-01 1.68616623e-01
-1.00210547e+00 7.07508028e-01 -1.36080050e+00 -4.01645362e-01
-3.45550254e-02 5.36063015e-01 -2.73178935e-01 1.26999140e-01
-7.90494502e-01 1.14667714e+00 3.87793034e-01 8.65075216e-02
-1.31956100e+00 -8.79534125e-01 -6.05641663e-01 -1.66967645e-01
8.03660035e-01 -1.10121691e+00 1.32719231e+00 -5.81020772e-01
-1.96255481e+00 5.99959195e-01 5.18540323e-01 -2.23682374e-01
9.92609203e-01 -4.95217264e-01 -5.06600440e-02 -3.55291404e-02
-6.43947870e-02 8.96202087e-01 6.95423484e-01 -1.28835106e+00
-3.54591638e-01 -1.12378158e-01 5.16790211e-01 4.74945754e-01
2.88488746e-01 -9.92460772e-02 -5.77413678e-01 -6.02589130e-01
-4.14536089e-01 -1.23384619e+00 -2.36907229e-01 4.76888061e-01
-2.51400352e-01 8.69570896e-02 1.04331100e+00 -2.18229517e-01
5.68164289e-01 -1.96532118e+00 8.57637644e-01 1.46066519e-02
1.44450054e-01 1.29462704e-01 -5.15976772e-02 6.10734880e-01
2.00944424e-01 -5.82926869e-02 1.65839437e-02 -3.28824043e-01
4.29194808e-01 3.70133221e-01 -2.03620493e-01 3.67382109e-01
-2.02861175e-01 1.02985120e+00 -1.10247982e+00 -3.41589063e-01
6.22892141e-01 3.17992687e-01 -7.58519948e-01 1.90585762e-01
-7.55184650e-01 8.45376551e-01 -4.38956916e-01 8.03139925e-01
3.90164375e-01 -2.34862734e-02 -8.32746178e-02 -7.94917941e-02
-3.54776978e-01 -9.94111691e-03 -1.39523339e+00 2.26297259e+00
-6.00694537e-01 3.22414398e-01 5.85123122e-01 -2.43899569e-01
4.98282611e-01 -1.24850664e-02 5.32012641e-01 -5.53265691e-01
3.51905316e-01 1.68254420e-01 -1.38759930e-02 -4.99570996e-01
6.02670074e-01 3.72093558e-01 -2.39764631e-01 2.21136048e-01
-1.01336457e-01 -7.88961351e-01 3.10251534e-01 -2.39338160e-01
1.43037176e+00 8.28489184e-01 -3.77783291e-02 -2.78097779e-01
-7.58749247e-02 3.91147524e-01 2.41583094e-01 7.50195801e-01
-1.57277912e-01 3.40891480e-01 4.48924601e-01 -6.30758345e-01
-1.15965903e+00 -1.33445442e+00 2.55494833e-01 1.39626491e+00
5.32496870e-01 -5.33784449e-01 -6.02519453e-01 -8.29577148e-02
2.20416114e-01 6.11784518e-01 -4.67885911e-01 -1.20243974e-01
-7.24134624e-01 -1.95634380e-01 4.87660199e-01 6.17256641e-01
5.93989789e-01 -1.49892116e+00 -1.63690794e+00 3.63038220e-02
1.68859020e-01 -1.03750205e+00 -5.84024638e-02 7.45930970e-02
-5.56655586e-01 -1.22428536e+00 -5.23971856e-01 -3.82918060e-01
1.71445534e-01 1.22237571e-01 1.22231078e+00 7.96149820e-02
-3.69409680e-01 6.73855722e-01 -4.92702574e-01 -3.88390213e-01
-1.14717744e-01 1.45533621e-01 2.87934661e-01 -8.31387639e-01
-5.69024146e-01 -6.59438252e-01 -6.46552861e-01 5.82493901e-01
-6.72566652e-01 4.72333878e-01 3.89745951e-01 6.04684830e-01
5.53180456e-01 -4.58548725e-01 -1.93858489e-01 -4.78706807e-02
6.38036788e-01 -2.84774035e-01 -8.38260591e-01 6.19866624e-02
3.25113162e-02 -3.88676763e-01 6.21307135e-01 -7.86204755e-01
-7.54681766e-01 -2.13728681e-01 7.49774277e-02 -7.61565983e-01
-1.83007941e-01 1.25810981e-01 -4.92335437e-03 -1.83803394e-01
7.45004117e-01 3.36491205e-02 8.48331954e-03 -2.87124336e-01
4.81085181e-01 1.75942898e-01 6.09357595e-01 -1.20792758e+00
5.77183306e-01 4.98432457e-01 1.79932684e-01 -5.79250157e-01
-1.36955172e-01 1.26465484e-01 -4.96369660e-01 -4.78890747e-01
9.45460916e-01 -6.65245056e-01 -1.62533963e+00 5.51807761e-01
-1.01468062e+00 -1.30062580e+00 -3.45447093e-01 4.69110757e-01
-1.18056095e+00 -1.10945433e-01 -6.09261394e-01 -5.26591122e-01
-2.31239453e-01 -1.51307845e+00 1.47394347e+00 1.51842952e-01
-4.02982324e-01 -5.01925528e-01 2.07781121e-02 -1.22517860e-02
6.10072315e-01 8.10748816e-01 6.09006882e-01 1.76581237e-02
-6.74956620e-01 8.69168416e-02 1.88863248e-01 -2.83575982e-01
-3.18592787e-01 2.16602504e-01 -4.09167737e-01 -7.33077526e-01
-5.52533031e-01 -7.12529778e-01 2.71559477e-01 1.19045973e-01
1.47081041e+00 -1.86513234e-02 -4.47747141e-01 8.86985123e-01
9.03111935e-01 1.09197401e-01 4.45294648e-01 6.24295712e-01
6.35530233e-01 2.28124723e-01 8.66465569e-01 6.28683329e-01
4.00280029e-01 1.09835184e+00 1.11307144e+00 -1.34123251e-01
4.47900705e-02 1.74459685e-02 5.40846050e-01 4.93220866e-01
-7.10710049e-01 -2.18979731e-01 -1.05206811e+00 5.80002889e-02
-1.83700538e+00 -8.01660061e-01 -5.47679216e-02 2.10836005e+00
5.70271671e-01 2.43548661e-01 4.79061067e-01 -4.64696698e-02
3.51309568e-01 1.75036117e-03 -8.66725385e-01 -4.37129587e-01
2.84312606e-01 3.94828916e-01 4.57405627e-01 9.93274897e-02
-7.65199721e-01 1.20402336e+00 6.22092056e+00 7.33013868e-01
-1.15297914e+00 -3.88191827e-02 -4.92168963e-02 -7.08840907e-01
1.83085397e-01 -2.10399345e-01 -5.14001369e-01 4.92752433e-01
6.90862715e-01 -2.40324121e-02 9.76523578e-01 1.15244257e+00
2.81709939e-01 -2.45366648e-01 -1.31053936e+00 9.46095169e-01
-2.82465369e-01 -1.27315927e+00 -2.96517819e-01 -2.28148662e-02
4.37510878e-01 2.70681828e-01 1.92643151e-01 6.73279703e-01
7.34372437e-01 -1.20366633e+00 1.16398954e+00 4.86027747e-01
9.77063239e-01 -6.14303708e-01 2.62466580e-01 6.31717324e-01
-1.15251851e+00 -1.06788889e-01 -1.64065167e-01 -4.02864486e-01
2.04668760e-01 -1.80078998e-01 -1.84159055e-01 4.60837781e-01
1.21230972e+00 4.96321887e-01 -1.89654738e-01 8.57157171e-01
-6.30655438e-02 -4.39357460e-02 -6.77350760e-01 -3.10628325e-01
3.48921537e-01 -1.20333038e-01 6.69808924e-01 8.48245203e-01
2.17335239e-01 4.40256037e-02 5.78309298e-01 7.89361775e-01
3.20921659e-01 -3.07140052e-01 -6.05579615e-01 2.59978861e-01
3.52180123e-01 1.19307613e+00 -9.07021403e-01 -2.01617867e-01
4.71285954e-02 9.64601934e-01 2.96695292e-01 2.52053052e-01
-1.42072845e+00 -1.65552333e-01 1.15696049e+00 1.98608145e-01
4.47988361e-02 -9.25656796e-01 8.01723916e-03 -1.05348229e+00
1.27159089e-01 -1.35482645e+00 -1.06214173e-01 -1.08914316e+00
-6.82435215e-01 2.79462785e-01 4.61472899e-01 -1.14265978e+00
-5.56330383e-02 -8.94285083e-01 -4.68662769e-01 4.53512549e-01
-9.97006714e-01 -1.10573542e+00 -9.50166821e-01 7.64868736e-01
5.31129837e-01 9.00481492e-02 8.45292091e-01 1.24735758e-01
-5.18949509e-01 1.80270523e-01 -7.79008195e-02 -3.17916095e-01
4.13840353e-01 -1.08051956e+00 6.86908841e-01 2.24217534e-01
-4.11778390e-01 5.64158142e-01 7.74420261e-01 -6.02970421e-01
-2.12462378e+00 -8.01297367e-01 -3.94176006e-01 -7.01689541e-01
6.57567441e-01 -7.31757879e-01 -4.23007458e-01 8.53680134e-01
8.52250531e-02 1.93388507e-01 -6.73038810e-02 -2.11661786e-01
2.44727045e-01 6.58856332e-02 -1.06744003e+00 1.07812822e+00
1.73802793e+00 1.00104228e-01 -3.07792068e-01 4.62714761e-01
8.01729977e-01 -1.47067690e+00 -9.97701287e-01 5.32515228e-01
7.79670656e-01 -1.06647122e+00 1.13771510e+00 -5.64800739e-01
5.60482621e-01 -4.53151137e-01 -6.34377822e-03 -1.43568122e+00
-1.77496418e-01 -8.74580562e-01 -5.08987248e-01 5.09625375e-01
-1.34133652e-01 -4.19496000e-01 5.51419854e-01 1.94591209e-01
-3.76873046e-01 -9.67928529e-01 -9.41559851e-01 -9.50103462e-01
-1.83645517e-01 -5.65379560e-01 8.16972375e-01 5.31766295e-01
-6.04536869e-02 -1.82684094e-01 -3.17816019e-01 -9.14499536e-02
3.15977573e-01 3.39221433e-02 1.51862919e+00 -6.62154734e-01
-6.02161467e-01 -6.82580352e-01 -2.95545161e-01 -1.08352232e+00
2.38681193e-02 -6.93639278e-01 -2.89199352e-02 -1.49289262e+00
-4.87238497e-01 -6.02806449e-01 2.67189980e-01 7.35683262e-01
2.86798447e-01 -1.27941910e-02 6.78255081e-01 1.72843546e-01
-6.31839991e-01 7.20654726e-01 1.71385837e+00 5.14919311e-02
-2.24706367e-01 -3.06313913e-02 3.12839705e-03 7.91037560e-01
7.48706996e-01 -3.88637893e-02 -3.78832370e-01 -6.82706118e-01
4.16391432e-01 1.98257610e-01 9.53265131e-01 -1.65827858e+00
-4.58309241e-03 -6.25563979e-01 3.14575344e-01 -3.67265344e-01
7.50247478e-01 -9.48061109e-01 4.97982204e-01 7.37694800e-01
-2.18509361e-01 6.53645039e-01 3.77328873e-01 3.95986140e-01
3.99948448e-01 4.26927924e-01 5.21972120e-01 -4.64930981e-01
-7.19551921e-01 3.50845307e-01 -2.76437283e-01 1.54286966e-01
1.55148160e+00 -3.04461926e-01 -5.57356477e-01 -1.53135821e-01
-7.55056024e-01 5.86040735e-01 9.15971756e-01 7.39096999e-01
3.93105030e-01 -1.19061720e+00 -3.61808419e-01 -1.09538632e-02
-3.58656887e-03 4.82683033e-01 1.12352863e-01 7.61993766e-01
-1.16737103e+00 9.17788595e-03 -7.17641950e-01 -6.60536587e-01
-1.03747582e+00 7.26744592e-01 3.64795148e-01 -9.38446298e-02
-1.13917065e+00 5.41341722e-01 -6.11974709e-02 -6.55301630e-01
3.23022634e-01 -8.26940835e-01 3.29930693e-01 -5.49770832e-01
2.60513037e-01 6.96331739e-01 -6.83876798e-02 -2.13215619e-01
-3.45032930e-01 5.70966423e-01 5.37907481e-01 -1.66075692e-01
1.23569727e+00 4.81429100e-01 1.01283021e-01 5.16556561e-01
4.97144580e-01 -2.68191457e-01 -1.59573996e+00 5.06773353e-01
-5.69078445e-01 -3.69098991e-01 -3.48405302e-01 -8.72554362e-01
-8.37368429e-01 7.71512270e-01 4.38361406e-01 -1.49146691e-01
8.16780031e-01 -1.75485179e-01 6.79403305e-01 7.66593635e-01
9.88300264e-01 -1.20527816e+00 2.72736847e-01 7.30855763e-01
1.46113884e+00 -8.19649041e-01 4.21586037e-02 -2.98952132e-01
-5.48242509e-01 1.00351524e+00 1.20447791e+00 -3.79993856e-01
2.05115795e-01 7.69744694e-01 -1.85197532e-01 -2.55868822e-01
-8.84918034e-01 -1.57733306e-01 -1.10742897e-01 7.93684483e-01
-6.44811094e-02 1.40994370e-01 -8.17708299e-03 2.56812602e-01
-6.86363816e-01 4.73594368e-02 2.74979621e-01 1.37757754e+00
-1.64881110e-01 -7.24873245e-01 -3.88220578e-01 1.90227494e-01
6.93879724e-02 3.68352503e-01 -1.50164470e-01 1.29479229e+00
3.19043905e-01 3.98192912e-01 7.17370659e-02 -4.31819439e-01
6.92095339e-01 -4.64845330e-01 9.50644314e-01 -4.90931481e-01
-1.12656093e+00 -3.71759295e-01 -4.51561734e-02 -1.26637387e+00
-5.53582683e-02 -4.28549588e-01 -1.32983577e+00 -8.05028260e-01
9.74434018e-02 -4.51368719e-01 8.43702912e-01 6.51986718e-01
5.32315969e-01 8.38487804e-01 -1.14343151e-01 -1.96664071e+00
-5.95121384e-01 -7.74837971e-01 -1.69518799e-01 2.85945356e-01
5.49858548e-02 -1.31024754e+00 -3.52697559e-02 -3.34494084e-01] | [4.623555660247803, 0.7832876443862915] |
2c7cabd7-4831-4384-8e9d-a7333085d71e | exploiting-unsupervised-data-for-emotion | 2010.01908 | null | https://arxiv.org/abs/2010.01908v2 | https://arxiv.org/pdf/2010.01908v2.pdf | Exploiting Unsupervised Data for Emotion Recognition in Conversations | Emotion Recognition in Conversations (ERC) aims to predict the emotional state of speakers in conversations, which is essentially a text classification task. Unlike the sentence-level text classification problem, the available supervised data for the ERC task is limited, which potentially prevents the models from playing their maximum effect. In this paper, we propose a novel approach to leverage unsupervised conversation data, which is more accessible. Specifically, we propose the Conversation Completion (ConvCom) task, which attempts to select the correct answer from candidate answers to fill a masked utterance in a conversation. Then, we Pre-train a basic COntext- Dependent Encoder (Pre-CODE) on the ConvCom task. Finally, we fine-tune the Pre-CODE on the datasets of ERC. Experimental results demonstrate that pre-training on unsupervised data achieves significant improvement of performance on the ERC datasets, particularly on the minority emotion classes. | ['Irwin King', 'Michael R. Lyu', 'Wenxiang Jiao'] | 2020-10-02 | null | https://aclanthology.org/2020.findings-emnlp.435 | https://aclanthology.org/2020.findings-emnlp.435.pdf | findings-of-the-association-for-computational | ['emotion-recognition-in-conversation'] | ['natural-language-processing'] | [ 1.78095922e-01 2.25146934e-01 5.67368232e-02 -1.01659656e+00
-9.97049212e-01 -2.32583106e-01 3.12296540e-01 -5.05594769e-03
-2.81394213e-01 5.72358668e-01 6.92607999e-01 -2.57086337e-01
5.23892283e-01 -2.84325272e-01 -1.31915703e-01 -5.66980660e-01
1.33815527e-01 7.54321739e-02 -6.00452304e-01 -2.13726461e-01
1.09657206e-01 -1.83208287e-02 -1.51653886e+00 8.33180070e-01
1.00874054e+00 1.26518238e+00 -9.02245753e-03 9.08611417e-01
-4.06491190e-01 1.41868210e+00 -6.20939672e-01 -6.46672010e-01
-3.81655455e-01 -7.14468181e-01 -9.91285980e-01 5.03255069e-01
-2.87487000e-01 -1.39384970e-01 -7.58115053e-02 7.41493285e-01
4.46611851e-01 4.35987562e-01 6.53922558e-01 -1.32223821e+00
-1.24508455e-01 5.94905913e-01 -3.86903882e-02 -2.29624927e-01
6.76022708e-01 -2.55037248e-01 1.19766450e+00 -1.35813463e+00
3.36686611e-01 1.32397068e+00 3.66506249e-01 9.15275693e-01
-7.63061047e-01 -6.30238175e-01 3.69937122e-01 1.84894651e-01
-1.04522526e+00 -7.03850925e-01 1.13009834e+00 -2.22874969e-01
1.04218125e+00 3.86929929e-01 3.58803302e-01 1.36606407e+00
-1.16596170e-01 1.08171177e+00 9.85213161e-01 -4.64024931e-01
5.83176970e-01 3.40527028e-01 1.78605437e-01 3.00540894e-01
-8.40237081e-01 -3.00685108e-01 -9.45759952e-01 -1.88857168e-01
-2.54014790e-01 -1.47921488e-01 -4.75135505e-01 2.88927525e-01
-5.71698070e-01 1.08784401e+00 4.16941904e-02 1.25163123e-01
-4.31031674e-01 -3.06183398e-01 6.92904234e-01 6.26969755e-01
8.73246014e-01 2.60041982e-01 -5.97514272e-01 -6.44614160e-01
-5.35794914e-01 -6.94023371e-02 1.28726935e+00 8.05377126e-01
5.73464155e-01 -2.07263991e-01 -9.63524356e-02 1.17350757e+00
1.83635056e-01 1.99738055e-01 5.07331431e-01 -8.50116372e-01
6.24013782e-01 7.53078878e-01 2.83060260e-02 -7.30602980e-01
-4.46123570e-01 -2.39257917e-01 -8.89087796e-01 -1.56883687e-01
-3.66395712e-02 -8.51036310e-01 -4.59782809e-01 1.65272880e+00
3.05298150e-01 6.82334241e-04 6.65035009e-01 8.37415576e-01
9.47621047e-01 9.59302425e-01 1.77126378e-01 -4.58770216e-01
1.15661359e+00 -1.28928745e+00 -1.09283614e+00 -4.87311184e-01
7.44441569e-01 -5.64734161e-01 9.73536730e-01 4.30596292e-01
-7.53252327e-01 -3.24978113e-01 -8.07193458e-01 -2.92968545e-02
-2.84518540e-01 2.26755023e-01 6.73632205e-01 5.66511452e-01
-8.89623582e-01 -4.55824435e-02 -4.10775781e-01 -7.65109584e-02
9.94160399e-03 8.88174400e-02 -2.40567267e-01 7.50178769e-02
-1.34940517e+00 7.61403143e-01 1.07349515e-01 4.30143505e-01
-6.06226444e-01 -2.52602637e-01 -1.11364841e+00 3.72204006e-01
3.12992334e-01 8.31455663e-02 1.47850871e+00 -1.37439740e+00
-2.09888339e+00 7.96252072e-01 -7.21046686e-01 -3.63843679e-01
2.58325785e-01 -4.46362011e-02 -6.30391181e-01 1.49937019e-01
-2.09875777e-01 7.51029015e-01 8.92991722e-01 -1.17554891e+00
-6.45742357e-01 -7.43215233e-02 -4.20032740e-02 4.82807040e-01
-3.60545337e-01 2.82915354e-01 -4.04845446e-01 -3.77355129e-01
-9.72307548e-02 -8.99828494e-01 -1.54036865e-01 -6.31244242e-01
-3.66267592e-01 -6.49353087e-01 7.07331121e-01 -6.51506305e-01
1.18587422e+00 -2.48326564e+00 1.16861947e-01 1.28870845e-01
-6.83606640e-02 -1.64515793e-01 -2.41894513e-01 6.04972720e-01
-2.60351062e-01 -9.69350152e-03 -2.34272525e-01 -9.38103378e-01
2.01705992e-01 2.31726930e-01 -5.40338695e-01 4.86281998e-02
5.87852120e-01 6.92190349e-01 -5.38946509e-01 -4.20202762e-01
-6.20172508e-02 3.98532003e-01 -7.74012804e-01 6.97721303e-01
-2.02253208e-01 5.75048923e-01 -4.02822256e-01 4.15061742e-01
5.58185875e-01 -4.20991071e-02 4.45403159e-01 2.68981934e-01
-8.18246454e-02 6.82139337e-01 -8.17477226e-01 1.39363682e+00
-8.78151059e-01 7.94433951e-01 3.26393604e-01 -1.18071640e+00
1.19047761e+00 7.77993917e-01 3.32816660e-01 -5.47449112e-01
2.97214210e-01 1.23542473e-01 -1.26129657e-01 -6.81253850e-01
6.37416601e-01 -3.29279631e-01 -6.20145082e-01 4.82952446e-01
1.45791829e-01 -4.90461104e-02 -4.85284552e-02 2.18931973e-01
9.08935189e-01 -4.51031566e-01 1.01837672e-01 5.15180752e-02
9.25065756e-01 -3.35251331e-01 6.83378160e-01 4.42071408e-01
-4.04195726e-01 4.31429535e-01 8.44835222e-01 -1.44475609e-01
-2.71596849e-01 -3.31225961e-01 4.44418676e-02 1.43151891e+00
-1.55725345e-01 -5.95032215e-01 -9.57236946e-01 -8.88205588e-01
-4.52931195e-01 7.33816504e-01 -6.77336216e-01 -2.30780497e-01
-2.79943109e-01 -4.93697375e-01 2.32282415e-01 4.73692417e-01
3.49497467e-01 -1.41747749e+00 -3.81381035e-01 1.71277672e-01
-8.86299372e-01 -1.26104748e+00 -5.99692523e-01 6.68560266e-01
-4.49742943e-01 -7.46918082e-01 -2.55931824e-01 -1.06024647e+00
6.78773940e-01 -4.74289618e-03 1.09439075e+00 1.00397058e-01
2.77078837e-01 3.27108800e-01 -8.03037524e-01 -4.25887972e-01
-3.95736516e-01 1.93126872e-01 -2.49429718e-01 5.49372494e-01
8.86112630e-01 -2.92985231e-01 -3.57965380e-01 1.44638270e-01
-4.31850255e-01 9.35795158e-02 2.37855434e-01 9.01511669e-01
9.94815491e-03 3.30638736e-02 1.23451483e+00 -7.62246132e-01
9.59181666e-01 -3.53441775e-01 4.70525585e-02 1.86917052e-01
-1.35089159e-01 -1.24721266e-01 6.31164074e-01 -2.70963371e-01
-1.40804720e+00 1.71755850e-01 -5.86752057e-01 -3.46851498e-02
-2.54203051e-01 7.52389133e-01 -4.26704198e-01 5.09935558e-01
1.20250471e-01 1.54954717e-01 -2.15551898e-01 -3.44768077e-01
7.88024068e-02 1.41502273e+00 4.04567152e-01 -4.99095857e-01
2.46700104e-02 1.45078879e-02 -9.52478588e-01 -8.30885887e-01
-1.17683387e+00 -6.90778196e-01 -2.70812869e-01 -5.52281559e-01
1.08976865e+00 -1.19846559e+00 -8.93331110e-01 2.65482992e-01
-1.27219033e+00 -4.93511230e-01 -3.20947319e-02 4.78753626e-01
-4.18152869e-01 -1.03193887e-01 -7.93133795e-01 -1.42721057e+00
-4.63660479e-01 -1.04598665e+00 1.19271684e+00 2.93447345e-01
-4.92254972e-01 -8.89079392e-01 -2.61706673e-03 6.13507628e-01
2.46314406e-01 -8.29028785e-02 7.25807250e-01 -9.64091420e-01
1.25875697e-01 -3.07508111e-01 1.28709406e-01 6.30699277e-01
-8.22537541e-02 -2.97186315e-01 -1.40396965e+00 -6.26841038e-02
4.60889250e-01 -9.89430487e-01 7.68293679e-01 -1.16788410e-01
1.38840580e+00 -1.10092983e-01 1.38272315e-01 1.15416437e-01
7.69104421e-01 4.39770281e-01 6.23005748e-01 -2.38412082e-01
1.24668762e-01 1.18089688e+00 6.85398102e-01 6.76105738e-01
6.44777834e-01 2.96009690e-01 5.70534095e-02 -5.28695323e-02
4.23177510e-01 -1.71487778e-01 6.60287499e-01 1.39359069e+00
3.58378202e-01 -4.33731824e-01 -5.99222004e-01 3.55835557e-01
-1.90436089e+00 -7.44327605e-01 1.29045010e-01 1.58109021e+00
1.10944390e+00 6.95548058e-02 -2.53958315e-01 2.27045581e-01
7.00815856e-01 2.48904988e-01 -4.87595350e-01 -8.26340497e-01
6.55789897e-02 2.37361446e-01 -3.53283286e-01 7.56793797e-01
-1.10305882e+00 8.54000032e-01 5.88651180e+00 4.16028053e-01
-1.14127028e+00 1.24970391e-01 1.07855725e+00 -6.69734180e-02
-2.15374306e-01 -1.96407840e-01 -6.37992501e-01 4.88891631e-01
1.23840809e+00 4.32835035e-02 3.74657273e-01 1.03396130e+00
4.25824553e-01 -4.10361700e-02 -1.24349964e+00 1.18928683e+00
3.61985654e-01 -7.38700390e-01 -5.44967175e-01 -2.12760329e-01
6.47883654e-01 -2.35279039e-01 -2.31588304e-01 1.15720773e+00
1.63912967e-01 -8.69489312e-01 5.38583934e-01 3.19708526e-01
4.73427057e-01 -1.07467604e+00 1.00058663e+00 5.00580370e-01
-8.36342514e-01 -6.86091930e-02 -1.82349443e-01 -2.66775995e-01
1.76716849e-01 5.57459652e-01 -6.45697773e-01 1.30753517e-01
6.31921053e-01 5.20403445e-01 -1.79346979e-01 1.80372462e-01
-3.75815213e-01 9.62269068e-01 5.54602146e-02 -3.32678318e-01
2.63970226e-01 -1.83685645e-01 8.04976448e-02 1.46636808e+00
4.38509136e-02 4.05746877e-01 3.31530869e-01 3.43129784e-01
-5.91743410e-01 3.23173434e-01 -1.43730685e-01 -3.66699807e-02
3.80081207e-01 1.50030923e+00 -3.04285616e-01 -3.85110855e-01
-5.37840784e-01 1.24788666e+00 6.27566576e-01 5.56943059e-01
-7.85226583e-01 -5.49048007e-01 5.96601605e-01 -6.68970942e-01
3.14237893e-01 2.31298640e-01 -7.80946687e-02 -1.32718658e+00
1.55188531e-01 -1.18538129e+00 2.41181180e-01 -7.36665487e-01
-1.30721712e+00 7.58886516e-01 -7.29975581e-01 -9.59313154e-01
-6.13285065e-01 -5.52180946e-01 -1.10999310e+00 7.12165952e-01
-1.51408684e+00 -6.53506875e-01 -3.50470155e-01 5.54080606e-01
9.62098956e-01 -4.68136221e-02 1.04782057e+00 2.68370867e-01
-7.46806085e-01 6.51345015e-01 -6.94355443e-02 3.72482300e-01
8.07862282e-01 -1.28870714e+00 -6.30409271e-02 4.95948821e-01
-2.61315346e-01 3.56220275e-01 4.93635058e-01 -3.64322007e-01
-1.33337557e+00 -9.97756004e-01 1.42758858e+00 -1.55544102e-01
4.54948038e-01 -8.90601218e-01 -8.46046746e-01 6.03787303e-01
4.96919841e-01 -3.60331744e-01 1.24246097e+00 5.24335504e-01
-1.74227700e-01 2.86280867e-02 -9.50772882e-01 5.41995823e-01
4.78960484e-01 -1.08818030e+00 -5.55444777e-01 1.39805883e-01
8.02364767e-01 -2.03498781e-01 -7.48008430e-01 1.61221638e-01
3.44160587e-01 -7.95315504e-01 3.62188429e-01 -6.79313183e-01
7.27053642e-01 2.85399914e-01 -3.00701112e-01 -1.55213165e+00
2.47040540e-01 -7.93972969e-01 -5.22064045e-02 1.62690485e+00
6.54709220e-01 -4.58850831e-01 6.68579578e-01 8.73741448e-01
-2.44752333e-01 -8.27166140e-01 -7.47168720e-01 3.24551240e-02
-6.60255253e-02 -7.37326086e-01 4.84543771e-01 9.62360680e-01
8.02411318e-01 6.96650386e-01 -5.38397491e-01 -1.09874979e-01
6.39446266e-03 2.70192981e-01 7.49382734e-01 -9.60029244e-01
-2.20199674e-01 -4.95974310e-02 1.89011976e-01 -1.20327950e+00
8.59080613e-01 -7.97520518e-01 7.47987688e-01 -1.17415035e+00
1.07197896e-01 -1.41582981e-01 -1.36241794e-01 3.59597892e-01
-4.76706624e-01 -1.11706644e-01 5.36402166e-02 -1.80632696e-01
-9.29388225e-01 1.00232708e+00 8.51004243e-01 -2.14325324e-01
-3.66657227e-01 8.43910798e-02 -9.42959368e-01 6.69850707e-01
8.94223213e-01 -3.68797511e-01 -3.66710097e-01 7.97010120e-03
1.07493006e-01 4.29577053e-01 -1.70618482e-02 -5.83550453e-01
2.48167917e-01 -1.41715314e-02 2.30918005e-01 -8.19568932e-01
5.33918023e-01 -7.04576254e-01 -5.76555133e-01 -5.16915098e-02
-9.46765184e-01 -2.49749899e-01 -3.36573669e-03 3.71462077e-01
-7.20510244e-01 -4.28976059e-01 4.11807328e-01 1.96001932e-01
-4.39582348e-01 -1.16706237e-01 -9.77381408e-01 1.62436128e-01
5.48982501e-01 3.75290394e-01 -3.62046212e-02 -1.05507040e+00
-7.96543956e-01 7.61332154e-01 -2.27965817e-01 6.31357610e-01
6.49279833e-01 -1.20576346e+00 -7.82959640e-01 1.80485472e-01
3.14278752e-01 -2.76974924e-02 3.46207917e-01 6.64513707e-01
1.22899458e-01 3.94846797e-01 4.56141353e-01 -3.14579338e-01
-1.25206542e+00 2.82720923e-01 3.89108151e-01 -3.69196475e-01
-3.44750464e-01 7.29065776e-01 2.15252727e-01 -8.37749004e-01
6.99072957e-01 -2.58743525e-01 -3.78307074e-01 2.45768130e-01
6.61193371e-01 -1.05828680e-01 1.88734800e-01 -5.51731527e-01
-4.59847182e-01 -7.99211189e-02 -1.27375007e-01 -3.61740023e-01
1.35128486e+00 -3.65379870e-01 -1.49931625e-01 5.09990692e-01
1.57159019e+00 4.54734974e-02 -1.23503602e+00 -2.09258527e-01
2.55482107e-01 1.29403114e-01 7.42578208e-02 -7.97903299e-01
-8.63576055e-01 9.50299501e-01 1.76750779e-01 2.69984007e-01
1.26653147e+00 8.55038036e-03 8.37092102e-01 7.21251428e-01
-1.45833656e-01 -1.54966950e+00 2.44201675e-01 8.68440688e-01
1.05098104e+00 -1.49415743e+00 -6.36606336e-01 -4.00803357e-01
-1.19782650e+00 1.04024363e+00 7.26103067e-01 2.53288358e-01
7.40634501e-01 4.22627568e-01 4.90375847e-01 -1.82354271e-01
-1.40034783e+00 -1.27625108e-01 -1.49345165e-02 1.56518579e-01
6.86353505e-01 1.03171863e-01 -2.04953969e-01 1.13872278e+00
-3.67094934e-01 -2.40335956e-01 4.22493011e-01 9.74932611e-01
-4.07618284e-01 -9.02254045e-01 -1.26153633e-01 1.91878095e-01
-5.43168485e-01 -5.51294200e-02 -9.19025421e-01 1.49299800e-01
-1.22279339e-01 1.80922484e+00 2.43204057e-01 -5.76485336e-01
3.36659968e-01 7.29778230e-01 -3.52515578e-01 -4.59921896e-01
-8.42036963e-01 1.85477793e-01 6.63589418e-01 -3.27826470e-01
-5.29565454e-01 -4.70613629e-01 -1.28625369e+00 1.27522290e-01
-3.95385832e-01 8.47520351e-01 7.26133287e-01 1.10047626e+00
3.22976857e-01 5.79173923e-01 1.34100354e+00 -6.25325918e-01
-4.66714680e-01 -1.26660717e+00 -3.64770532e-01 4.39869255e-01
3.10965687e-01 -2.74250776e-01 -6.82743788e-01 1.72509521e-01] | [13.130273818969727, 6.0322184562683105] |
d8fb8b25-404e-4541-bbf8-e965a4c14a36 | ranking-in-contextual-multi-armed-bandits | 2207.00109 | null | https://arxiv.org/abs/2207.00109v1 | https://arxiv.org/pdf/2207.00109v1.pdf | Ranking in Contextual Multi-Armed Bandits | We study a ranking problem in the contextual multi-armed bandit setting. A learning agent selects an ordered list of items at each time step and observes stochastic outcomes for each position. In online recommendation systems, showing an ordered list of the most attractive items would not be the best choice since both position and item dependencies result in a complicated reward function. A very naive example is the lack of diversity when all the most attractive items are from the same category. We model position and item dependencies in the ordered list and design UCB and Thompson Sampling type algorithms for this problem. We prove that the regret bound over $T$ rounds and $L$ positions is $\Tilde{O}(L\sqrt{d T})$, which has the same order as the previous works with respect to $T$ and only increases linearly with $L$. Our work generalizes existing studies in several directions, including position dependencies where position discount is a particular case, and proposes a more general contextual bandit model. | ['Arnaud Doucet', 'George Deligiannidis', 'Amitis Shidani'] | 2022-06-30 | null | null | null | null | ['thompson-sampling'] | ['methodology'] | [-1.27326682e-01 -1.49320632e-01 -8.30901802e-01 -4.54659313e-01
-9.66153443e-01 -9.80083287e-01 -9.54173207e-02 3.84869993e-01
-7.86379039e-01 1.15703022e+00 1.16190396e-03 -5.06583333e-01
-9.55258846e-01 -8.73131752e-01 -1.11877882e+00 -8.29097092e-01
-3.67256463e-01 9.60323989e-01 1.34035975e-01 -1.13848493e-01
4.38769013e-01 2.42314130e-01 -1.33844650e+00 3.39047104e-01
8.21078837e-01 1.38820672e+00 1.23597667e-01 7.11030960e-01
-2.75572658e-01 5.54933965e-01 -4.73772794e-01 -8.50802898e-01
7.83197045e-01 -6.35416567e-01 -8.28497887e-01 -1.01154625e-01
1.64237052e-01 -5.60497761e-01 -8.03366378e-02 1.19473624e+00
1.97735116e-01 3.25557739e-01 3.70711476e-01 -1.01272666e+00
-5.34188926e-01 1.25704813e+00 -9.16960716e-01 3.70009929e-01
3.77530418e-02 -6.67147160e-01 1.65803480e+00 -2.27952257e-01
3.13250631e-01 1.04417157e+00 3.63220364e-01 2.68309653e-01
-1.12498796e+00 -4.41778004e-01 8.77778828e-01 2.40703374e-01
-7.20285594e-01 1.57953650e-01 6.16732597e-01 -1.52640760e-01
4.45609748e-01 7.68278241e-01 9.50914681e-01 4.54240382e-01
-1.60675824e-01 1.21036696e+00 1.27883065e+00 -6.18224084e-01
5.89326441e-01 3.86818685e-02 7.07334816e-01 3.64130229e-01
8.93057227e-01 4.60572578e-02 -6.01060867e-01 -3.13010126e-01
5.87027967e-01 5.13416231e-01 -6.43128902e-03 -6.72040462e-01
-7.86993325e-01 1.02119195e+00 3.97679746e-01 -1.81049071e-02
-4.77165937e-01 4.94780749e-01 1.12875924e-01 5.54751456e-01
5.63082516e-01 3.33296180e-01 -5.90888381e-01 -1.90346062e-01
-8.71540785e-01 3.07240933e-01 7.42505372e-01 1.15575433e+00
5.48797309e-01 -4.90222067e-01 -3.34789902e-01 8.76244187e-01
-1.06435999e-01 5.41905701e-01 -9.43368301e-02 -1.01451838e+00
8.87587309e-01 1.89438850e-01 9.67165172e-01 -3.46176147e-01
-4.47197556e-01 -7.95812488e-01 -3.30143899e-01 -6.30716458e-02
7.87996769e-01 -3.80538732e-01 -5.39885879e-01 1.72738183e+00
2.16156855e-01 -3.53314012e-01 -3.89248848e-01 9.32596803e-01
2.68928055e-02 4.73599225e-01 -3.28930020e-01 -7.80909419e-01
1.14329135e+00 -1.05304420e+00 -5.53838074e-01 -3.46620679e-01
5.35606802e-01 -4.71450597e-01 8.02607834e-01 7.61000931e-01
-1.49453318e+00 1.59751683e-01 -8.71115029e-01 1.49431899e-01
2.02666502e-02 -1.24929823e-01 1.21371531e+00 9.60198283e-01
-7.28032947e-01 7.18079507e-01 -5.34174562e-01 -1.00668766e-01
3.29812497e-01 3.96326900e-01 4.04137999e-01 -2.67264456e-01
-8.37680280e-01 5.32494247e-01 -1.17655955e-01 1.92142084e-01
-6.47266865e-01 -3.73580724e-01 -1.33782089e-01 1.97167486e-01
8.69439900e-01 -3.49976212e-01 1.48652959e+00 -8.43120694e-01
-1.36233115e+00 3.81904393e-01 -1.18734173e-01 -6.89717591e-01
7.77079761e-01 -3.64318430e-01 8.77209753e-02 -3.10649902e-01
7.66701773e-02 -3.42909321e-02 3.29236358e-01 -9.16986704e-01
-1.22661698e+00 -6.74577534e-01 7.27170765e-01 4.08371836e-01
-2.62985706e-01 -1.28945142e-01 -3.28757316e-01 -3.88563573e-01
2.17854396e-01 -1.16729617e+00 -4.26714033e-01 -5.86761057e-01
-2.74840593e-01 -2.04090938e-01 -3.89848143e-01 -1.06354259e-01
1.33510232e+00 -1.70656121e+00 1.41169205e-02 3.66068572e-01
-2.73061901e-01 -4.47895229e-01 -7.02290051e-03 5.53031623e-01
3.69914591e-01 2.11195678e-01 4.26413208e-01 -1.49562135e-01
3.09856892e-01 1.52494997e-01 -4.66661513e-01 4.13687170e-01
-8.77687752e-01 7.41239965e-01 -8.45125377e-01 -2.74933893e-02
-2.58043587e-01 -4.66577172e-01 -9.19081330e-01 -2.40985125e-01
-2.97994971e-01 -2.01571703e-01 -6.09422863e-01 3.28267872e-01
7.03922093e-01 -3.80204201e-01 4.50197875e-01 2.33550280e-01
-1.11105129e-01 5.02098560e-01 -1.45458567e+00 1.32972240e+00
-4.38557684e-01 1.75375752e-02 5.79746477e-02 -9.33422029e-01
3.94334048e-01 -1.31944707e-02 4.80777949e-01 -6.78414345e-01
2.23186575e-02 3.76496077e-01 -1.94297116e-02 -1.26532450e-01
5.19733131e-01 -3.74761641e-01 -3.34011078e-01 8.43449295e-01
-4.36071813e-01 2.78311640e-01 3.23660254e-01 2.20184729e-01
9.55734134e-01 -1.84963852e-01 1.69305637e-01 -3.61687183e-01
-1.92844018e-01 6.53788075e-02 7.60606647e-01 1.63857555e+00
6.87899888e-02 1.20918795e-01 8.00578594e-01 -6.85877264e-01
-8.92301857e-01 -8.75738800e-01 5.26472852e-02 1.86059761e+00
6.85947657e-01 1.13310009e-01 -4.56961960e-01 -7.35051572e-01
3.03451449e-01 9.31178093e-01 -9.47233617e-01 3.04119766e-01
-3.01142991e-01 -1.03579032e+00 -3.60554218e-01 5.41924119e-01
1.65817499e-01 -7.02724218e-01 -7.49736011e-01 3.50189179e-01
-6.92281947e-02 -5.67723930e-01 -7.39358723e-01 6.99014366e-01
-1.07246220e+00 -8.77859175e-01 -7.38365531e-01 -4.34452504e-01
6.69991374e-01 5.29366910e-01 9.63956237e-01 -2.28006616e-01
2.38910228e-01 7.99013153e-02 -5.29356122e-01 -7.32252240e-01
3.58382523e-01 6.50220513e-02 8.78992379e-02 -8.02082717e-02
3.40191513e-01 -1.18452996e-01 -1.14888632e+00 3.79955471e-01
-6.12159848e-01 -1.34855092e-01 5.21467984e-01 9.72265303e-01
6.72748327e-01 -9.86883268e-02 6.07501447e-01 -1.42212605e+00
4.95994300e-01 -3.41857702e-01 -1.01363015e+00 5.20809829e-01
-5.70119858e-01 1.28759146e-01 7.27124393e-01 -4.18806046e-01
-8.64430726e-01 -1.03446424e-01 3.13795298e-01 1.71987534e-01
4.11700100e-01 6.03426874e-01 2.52951264e-01 3.13851833e-01
3.90608549e-01 1.29602596e-01 -3.51106912e-01 -6.45189881e-01
3.41950208e-01 3.71312559e-01 -1.50894627e-01 -7.40986586e-01
2.52967447e-01 4.98675555e-01 -1.07855443e-03 -1.59280539e-01
-1.38359427e+00 -2.87419081e-01 9.65101272e-03 -3.39672379e-02
7.25530535e-02 -6.69296265e-01 -1.29051220e+00 -1.42494500e-01
-6.91410840e-01 -3.94103348e-01 -5.00994444e-01 7.83088088e-01
-7.81323910e-01 1.37314990e-01 -6.22785270e-01 -1.36069107e+00
-1.45921424e-01 -9.78704929e-01 5.49816966e-01 1.45842791e-01
3.29917401e-01 -3.75356019e-01 -8.34946036e-02 5.26040375e-01
2.68474281e-01 -2.99630791e-01 9.43338335e-01 -7.88352966e-01
-8.78405571e-01 -3.42946589e-01 7.69433677e-02 -5.26463687e-02
-1.40461057e-01 -5.87111294e-01 -2.11700112e-01 -4.73357111e-01
-8.79015848e-02 -3.95043075e-01 8.86517584e-01 1.00567007e+00
9.45795476e-01 -6.20020866e-01 -3.18557233e-01 1.83995426e-01
1.47891974e+00 7.43816018e-01 2.23520517e-01 5.27617931e-01
4.23231833e-02 3.36958289e-01 9.36187387e-01 7.62876630e-01
1.56384185e-01 8.00933957e-01 5.16138911e-01 2.56801963e-01
6.18533015e-01 -6.98658749e-02 1.09537810e-01 3.19828779e-01
-1.38556540e-01 -4.52318847e-01 -3.09695244e-01 6.99486971e-01
-2.10422230e+00 -1.05937266e+00 1.14367947e-01 2.87253261e+00
7.37953305e-01 3.45891744e-01 6.16701126e-01 6.67443126e-02
7.82070696e-01 -1.34416968e-01 -8.30483079e-01 -6.58704579e-01
7.93712959e-02 -4.51050214e-02 1.07192957e+00 6.35421157e-01
-7.98580706e-01 6.22884154e-01 6.45289135e+00 9.94683266e-01
-6.38679087e-01 1.81280300e-01 9.92140770e-01 -1.28891098e+00
-3.57624233e-01 1.48954440e-03 -1.01067150e+00 6.61817133e-01
5.62331617e-01 -2.18486011e-01 8.72169435e-01 9.93503749e-01
1.66741405e-02 -4.70831931e-01 -1.30842447e+00 8.92030895e-01
-2.06235871e-01 -1.28176045e+00 -2.54691631e-01 4.43229079e-01
1.02832830e+00 -1.35483742e-02 2.07937598e-01 1.92197964e-01
1.01149547e+00 -5.57433486e-01 1.03335702e+00 5.74676916e-02
4.26702619e-01 -1.19993603e+00 6.28569543e-01 6.52159929e-01
-7.21774817e-01 -5.94774425e-01 -5.62838733e-01 -3.53642613e-01
2.18016446e-01 5.89410007e-01 -3.51501822e-01 5.17068565e-01
9.05376911e-01 2.16215383e-02 1.18110627e-01 1.35454750e+00
7.84587041e-02 4.41456169e-01 -6.51599169e-01 -7.17682123e-01
4.65940088e-01 -4.71366137e-01 2.89448023e-01 7.02735245e-01
4.47492659e-01 3.75961304e-01 2.04460651e-01 2.34237090e-02
-1.99256033e-01 2.89198637e-01 -7.71168992e-02 4.41096336e-01
6.45897448e-01 8.67516816e-01 -9.03992355e-01 -3.50146472e-01
-3.36204976e-01 6.82448030e-01 4.74612325e-01 1.24398641e-01
-9.91847932e-01 -1.41245604e-01 3.50839972e-01 3.73127684e-02
8.95545661e-01 1.90764248e-01 -5.29560030e-01 -9.16578710e-01
1.66475385e-01 -4.17121679e-01 8.39479029e-01 -2.54650086e-01
-1.35256481e+00 2.90670067e-01 5.67122847e-02 -1.09530127e+00
1.44363604e-02 -4.42562133e-01 -2.80699227e-02 4.29069817e-01
-1.41629004e+00 -5.23360074e-01 3.61547887e-01 3.86642724e-01
5.71061313e-01 1.81003168e-01 4.48058903e-01 9.69501510e-02
-4.63609129e-01 7.75037885e-01 8.92126977e-01 -3.69984597e-01
3.45583022e-01 -1.39353335e+00 -1.98591784e-01 5.66770971e-01
2.78228253e-01 6.90939307e-01 9.64056671e-01 -3.47811252e-01
-1.50290334e+00 -7.01308846e-01 5.43988883e-01 -3.09389710e-01
4.84355599e-01 -2.51021028e-01 -1.03863969e-01 8.12560916e-01
2.45490950e-02 -1.24537073e-01 9.15132225e-01 7.45859981e-01
-3.25513661e-01 -6.70630395e-01 -1.12137806e+00 6.01049364e-01
1.36756849e+00 4.93847281e-02 -2.22282082e-01 4.25211519e-01
3.67460459e-01 -3.78843278e-01 -6.59084857e-01 1.43105447e-01
9.72984910e-01 -1.12745988e+00 6.79338574e-01 -6.96792960e-01
8.40440392e-02 2.33663712e-02 -3.20781112e-01 -1.24922252e+00
-5.10746181e-01 -7.13799238e-01 -5.53083979e-02 6.94822013e-01
9.50947165e-01 -6.14749968e-01 1.16566086e+00 9.58701730e-01
2.68675268e-01 -8.92973602e-01 -1.20444715e+00 -1.05408454e+00
2.53300190e-01 -2.86861092e-01 7.06799984e-01 4.72198904e-01
2.84792364e-01 1.82130516e-01 -8.39161575e-01 -5.00972681e-02
6.95051193e-01 1.01131117e+00 5.85003436e-01 -1.15831268e+00
-8.28870237e-01 -5.79470217e-01 4.41642255e-01 -1.71387076e+00
-5.65941274e-01 -6.84691191e-01 -1.01424307e-02 -1.55349362e+00
7.29169190e-01 -9.06950712e-01 -1.00449359e+00 2.38137513e-01
2.70871650e-02 1.89113896e-02 3.67335409e-01 1.19538583e-01
-1.19299209e+00 1.03693359e-01 1.11278820e+00 -4.91064452e-02
-2.99954236e-01 5.07251978e-01 -1.19182575e+00 4.24670219e-01
5.01691222e-01 -6.19800866e-01 -4.34198171e-01 -6.54004931e-01
8.81108820e-01 6.52889669e-01 -2.77166694e-01 -4.17977780e-01
2.83500671e-01 -6.16003931e-01 1.89347297e-01 -9.11212504e-01
3.38856667e-01 -8.93994987e-01 2.63846695e-01 4.90741521e-01
-8.40187430e-01 3.16545740e-02 -2.11910069e-01 9.43720520e-01
3.53871882e-01 -4.94433314e-01 4.85359848e-01 -3.32075179e-01
1.17918169e-02 3.59397113e-01 -1.41993612e-02 -1.37640238e-01
1.08716512e+00 -1.90562621e-01 -5.22881150e-01 -6.04102612e-01
-6.45803273e-01 3.51448566e-01 3.40041310e-01 5.28980233e-02
3.23257968e-02 -1.17270947e+00 -4.04828489e-01 -3.81271720e-01
4.20504063e-03 -1.80081189e-01 5.07876098e-01 6.93150342e-01
-1.36435524e-01 6.37571633e-01 6.46371916e-02 -2.25574315e-01
-1.05232346e+00 8.53716969e-01 1.08026853e-02 -7.59902060e-01
-6.05695508e-02 1.26224136e+00 2.03912422e-01 1.79403782e-01
3.99074554e-01 -3.17224979e-01 5.69498632e-03 2.75150806e-01
5.47350883e-01 5.82702577e-01 4.86300029e-02 1.98949024e-01
-3.00032228e-01 3.51284444e-01 -6.08834565e-01 -3.86142910e-01
1.58193946e+00 -3.03453654e-01 -1.07128862e-02 5.02892733e-01
6.94843173e-01 2.66759068e-01 -1.28270507e+00 -4.12923545e-01
3.30419801e-02 -7.09741175e-01 -5.50529622e-02 -9.80057478e-01
-1.05346608e+00 3.86036813e-01 3.51206809e-01 9.89053369e-01
9.71287489e-01 5.74388504e-02 4.13195312e-01 4.82452691e-01
1.14513326e+00 -1.44598281e+00 -1.68481529e-01 4.09192502e-01
6.86520278e-01 -8.96258295e-01 2.31697515e-01 -1.17654316e-01
-6.87378168e-01 5.55343509e-01 3.17413032e-01 -3.24343741e-01
6.42124355e-01 -2.05598742e-01 -4.71253455e-01 9.52270478e-02
-8.78231585e-01 -2.33563155e-01 -1.19612277e-01 -9.42639932e-02
3.70077103e-01 4.54077959e-01 -9.90647614e-01 8.64564896e-01
-2.40551621e-01 -3.75833474e-02 5.75019300e-01 9.15320098e-01
-6.50361896e-01 -1.31474972e+00 -3.84286314e-01 1.11462235e+00
-1.05081594e+00 -1.48544237e-01 -8.16497058e-02 2.96705425e-01
-2.15077743e-01 1.11847556e+00 9.55537036e-02 -2.28580534e-02
3.03333551e-01 -5.18836603e-02 7.82575607e-01 -5.28622746e-01
-6.05645359e-01 4.37228262e-01 1.72111422e-01 -2.78784484e-01
-3.59516114e-01 -7.39547014e-01 -6.73123121e-01 -3.84167641e-01
-6.48737609e-01 6.74264431e-01 5.64353883e-01 7.39040911e-01
3.20898861e-01 1.35668650e-01 9.96677876e-01 -5.60411453e-01
-7.61297047e-01 -7.22523570e-01 -1.11045206e+00 3.88538539e-01
3.75304401e-01 -7.51082480e-01 -2.32329339e-01 -6.67158842e-01] | [4.6173577308654785, 3.366774797439575] |
ffdca9ee-6d15-4245-96d9-71f456e0cc20 | dynamic-atomic-column-detection-in | 2302.00816 | null | https://arxiv.org/abs/2302.00816v1 | https://arxiv.org/pdf/2302.00816v1.pdf | Dynamic Atomic Column Detection in Transmission Electron Microscopy Videos via Ridge Estimation | Ridge detection is a classical tool to extract curvilinear features in image processing. As such, it has great promise in applications to material science problems; specifically, for trend filtering relatively stable atom-shaped objects in image sequences, such as Transmission Electron Microscopy (TEM) videos. Standard analysis of TEM videos is limited to frame-by-frame object recognition. We instead harness temporal correlation across frames through simultaneous analysis of long image sequences, specified as a spatio-temporal image tensor. We define new ridge detection algorithms to non-parametrically estimate explicit trajectories of atomic-level object locations as a continuous function of time. Our approach is specially tailored to handle temporal analysis of objects that seemingly stochastically disappear and subsequently reappear throughout a sequence. We demonstrate that the proposed method is highly effective and efficient in simulation scenarios, and delivers notable performance improvements in TEM experiments compared to other material science benchmarks. | ['David S. Matteson', 'Peter A. Crozier', 'Andrew M. Thomas', 'Yuchen Xu'] | 2023-02-02 | null | null | null | null | ['object-recognition'] | ['computer-vision'] | [ 3.19876164e-01 -8.00028443e-01 3.40122581e-01 5.25175594e-02
-7.87534535e-01 -5.71998000e-01 6.26076519e-01 2.52524823e-01
-7.52506137e-01 5.71677327e-01 -5.22539198e-01 -1.97931379e-01
-1.21940069e-01 -3.91954601e-01 -5.84512293e-01 -1.29084933e+00
-2.95743525e-01 3.99205595e-01 3.89525950e-01 1.83999240e-01
5.36574304e-01 9.78202403e-01 -1.25306082e+00 2.22383305e-01
1.56371206e-01 1.02086353e+00 3.35492074e-01 9.81043935e-01
1.11704603e-01 6.75928116e-01 -2.04078823e-01 6.33822083e-02
-3.20822410e-02 -2.71764603e-02 -7.26647735e-01 4.69021559e-01
3.53086799e-01 -4.87881228e-02 -2.73095161e-01 8.17201078e-01
1.97897971e-01 3.41973633e-01 7.67519474e-01 -8.25074792e-01
-2.34851211e-01 -1.32462054e-01 -1.00680339e+00 8.09244215e-01
1.40884295e-01 2.85521805e-01 7.64237463e-01 -1.18480611e+00
9.12800491e-01 1.01537919e+00 6.32781744e-01 1.53690159e-01
-1.62678611e+00 -7.18384632e-04 1.06019024e-02 4.12548184e-01
-1.01694930e+00 -4.88317221e-01 1.03896761e+00 -8.07375610e-01
8.13673139e-01 3.35191250e-01 3.43351394e-01 6.74704969e-01
7.35687435e-01 6.88324690e-01 1.04106843e+00 -2.06331015e-01
2.39273325e-01 -4.51262474e-01 3.71618450e-01 7.60115743e-01
6.84411153e-02 -2.39144042e-01 -5.48374236e-01 -2.99669445e-01
9.32978928e-01 3.54625881e-01 -2.67245043e-02 -4.17215198e-01
-1.80102849e+00 4.86106068e-01 1.10424303e-01 3.29770237e-01
-5.39526880e-01 3.51869315e-01 5.20262957e-01 1.68770105e-01
5.87476492e-01 2.76755422e-01 -1.86750352e-01 -2.89647341e-01
-1.13867223e+00 5.12199819e-01 1.46822870e-01 6.67297363e-01
7.42104709e-01 -4.18999419e-02 -1.39890552e-01 5.58161974e-01
-1.28646672e-01 5.03449619e-01 1.06430933e-01 -1.09247565e+00
2.70659681e-02 3.57536077e-01 3.16416472e-01 -8.82945716e-01
-4.77437407e-01 -7.42430016e-02 -9.56458092e-01 2.75667995e-01
5.83645165e-01 2.34549925e-01 -6.58716261e-01 1.24773085e+00
5.38750887e-01 1.98760465e-01 -4.57635015e-01 7.96399057e-01
5.12557268e-01 8.64867032e-01 -3.34950447e-01 -7.48780668e-01
1.41327536e+00 -6.75653398e-01 -6.49517775e-01 3.38113189e-01
2.83409983e-01 -7.26157427e-01 8.27777803e-01 3.66344810e-01
-1.29627085e+00 -2.75816590e-01 -6.01498306e-01 4.87580188e-02
-1.90362141e-01 1.02483205e-01 5.25965631e-01 7.52014443e-02
-7.55352676e-01 7.53235877e-01 -1.36656225e+00 -1.12713203e-01
6.10209167e-01 3.56214732e-01 -1.72423527e-01 1.42151520e-01
-1.06022954e-01 9.60912630e-02 -2.34844536e-01 2.04651356e-01
-9.72108901e-01 -8.75061214e-01 -4.98991460e-01 -3.16741228e-01
2.75092274e-01 -4.47479844e-01 1.12083936e+00 -5.30577838e-01
-1.16877747e+00 8.95514667e-01 -8.69839549e-01 -3.11530203e-01
4.52179730e-01 4.85207886e-02 -1.76678061e-01 4.69192863e-01
2.25641578e-01 -5.17747439e-02 1.30000651e+00 -1.30253172e+00
-5.79261184e-01 -5.29410601e-01 -4.35011476e-01 -2.92248517e-01
-1.28687531e-01 2.86954492e-01 -3.37449282e-01 -7.03675926e-01
2.06255108e-01 -8.74665320e-01 -4.62122709e-01 8.72846469e-02
-4.50860560e-01 -1.17325924e-01 1.32214248e+00 -2.97396630e-01
1.00448775e+00 -1.95130384e+00 2.33448908e-01 1.79465357e-02
5.95760703e-01 -1.70528755e-01 3.01860303e-01 4.67545658e-01
1.61312178e-01 -2.66978502e-01 -4.50731903e-01 -4.81893659e-01
-5.53613067e-01 -2.62894183e-01 -2.90135056e-01 1.06068647e+00
2.28610605e-01 8.56778800e-01 -1.00503683e+00 -4.75517660e-01
2.95589864e-01 3.61748725e-01 -3.83805782e-01 -4.22320440e-02
-1.78417534e-01 8.51089299e-01 -6.06273711e-01 7.45796204e-01
7.04553008e-01 -7.29032218e-01 -2.25970447e-01 -3.19533378e-01
-6.78003788e-01 -2.39100203e-01 -7.42321253e-01 1.32193005e+00
-1.44484416e-01 8.14114749e-01 3.04212451e-01 -1.26444995e+00
5.23763239e-01 1.32135406e-01 1.19045889e+00 -3.26038182e-01
4.50867005e-02 5.51502481e-02 -2.36905694e-01 -5.48523903e-01
5.76564729e-01 -1.09548122e-01 6.82693198e-02 4.51054335e-01
-3.10928077e-01 -3.56357731e-02 4.32153583e-01 1.27914801e-01
1.52185094e+00 -2.28047848e-01 -4.97882068e-02 -6.15941107e-01
7.85377860e-01 2.05667734e-01 2.64508575e-01 7.57056773e-01
-1.91050500e-01 6.21692717e-01 2.88860798e-01 -7.46284664e-01
-1.51963031e+00 -1.33482432e+00 -3.46980959e-01 8.25113773e-01
3.00587177e-01 -1.98650226e-01 -4.95452166e-01 -2.62507528e-01
6.15438819e-02 1.12478092e-01 -5.17897010e-01 2.92903602e-01
-9.64224339e-01 -9.65496182e-01 -1.44981027e-01 3.23133081e-01
9.23839286e-02 -1.21095479e+00 -9.58883584e-01 5.80051959e-01
2.20067352e-01 -1.23695385e+00 -6.34010315e-01 4.85583544e-02
-1.10438776e+00 -9.63499606e-01 -7.79129803e-01 -6.64666772e-01
8.81883085e-01 6.08710825e-01 9.78012025e-01 -2.40195612e-03
-9.68799353e-01 7.67257154e-01 -3.18803918e-03 2.34154180e-01
-2.58907795e-01 -2.87972271e-01 2.53663659e-01 4.18367565e-01
-5.78589812e-02 -5.44340014e-01 -8.69594812e-01 5.49866855e-01
-9.96747136e-01 1.25707775e-01 7.67331421e-02 1.01846635e+00
1.09521937e+00 2.52071559e-01 2.84409344e-01 -6.79678679e-01
4.06956017e-01 -2.52885997e-01 -9.18055654e-01 5.16913459e-02
-3.36086810e-01 -1.50290474e-01 8.00138772e-01 -4.98226464e-01
-1.04495251e+00 1.23944014e-01 2.64632195e-01 -7.07132280e-01
6.51526451e-02 2.98438042e-01 5.82975566e-01 -3.13654363e-01
1.45940438e-01 4.83940780e-01 9.90399271e-02 -3.67968917e-01
-1.02046728e-01 1.15882061e-01 9.06786144e-01 -1.00468767e+00
8.32468808e-01 1.33952832e+00 3.78177851e-01 -1.41038370e+00
-3.58779997e-01 -9.16564703e-01 -7.28721440e-01 -5.02187073e-01
8.97350550e-01 -4.37233061e-01 -1.31220591e+00 4.96747911e-01
-1.20234776e+00 -3.24856311e-01 -1.41543910e-01 3.09935451e-01
-8.12019587e-01 4.93272662e-01 -1.05150533e+00 -9.32446539e-01
-2.59330690e-01 -1.17952251e+00 1.43449318e+00 8.94980878e-02
5.22043183e-02 -1.09927416e+00 1.25034288e-01 1.82836086e-01
1.80324689e-01 3.29572797e-01 1.03657794e+00 1.79949090e-01
-9.48334754e-01 -1.67216554e-01 -1.22155853e-01 -1.05314866e-01
3.92879516e-01 3.64300251e-01 -6.41175926e-01 -6.68336749e-01
3.36087614e-01 1.19417951e-01 8.42947483e-01 8.56486619e-01
1.48615360e+00 -2.24960178e-01 -4.98068720e-01 5.94306946e-01
1.40286613e+00 1.96832076e-01 3.62775564e-01 4.01426941e-01
8.38890672e-01 6.08322978e-01 5.74538350e-01 6.27068996e-01
-1.40663356e-01 8.59972298e-01 9.26370546e-02 4.43336740e-02
2.40083203e-01 4.34324533e-01 3.56878102e-01 9.41904724e-01
-5.27273357e-01 -2.09642500e-02 -7.86170363e-01 7.30513334e-01
-1.81798875e+00 -1.33783221e+00 -5.71714520e-01 2.26254272e+00
5.26692450e-01 1.36636244e-02 3.42866242e-01 -2.76509505e-02
7.32769668e-01 1.69019178e-01 -8.44625235e-01 1.31115839e-01
-2.27455005e-01 -2.76236702e-02 5.56122959e-01 1.85782373e-01
-9.60746825e-01 5.62665939e-01 6.65337944e+00 7.35507607e-01
-1.25779843e+00 1.69817999e-01 6.53440595e-01 -1.87678963e-01
-2.29848847e-01 -1.74284711e-01 -6.67527437e-01 4.89491999e-01
6.64479971e-01 -2.76374787e-01 3.96404237e-01 4.85902160e-01
6.31191373e-01 -1.41652197e-01 -1.07001233e+00 1.28939092e+00
-2.89391369e-01 -1.96571159e+00 -1.20976761e-01 9.46911797e-02
7.75777042e-01 5.10447323e-02 3.30741018e-01 -4.49626714e-01
-1.35391757e-01 -6.13770247e-01 7.22326219e-01 5.83682716e-01
5.58002412e-01 -5.87225080e-01 2.61084706e-01 1.33911490e-01
-1.40236509e+00 1.69964045e-01 -2.46390432e-01 1.75812349e-01
6.00343525e-01 7.99727201e-01 -6.87387526e-01 3.11471283e-01
8.17196131e-01 9.20979440e-01 -3.16042602e-01 1.00105548e+00
6.44238353e-01 5.91309607e-01 -3.37010473e-01 1.26284316e-01
3.16720098e-01 -6.50837004e-01 9.24107432e-01 1.11841750e+00
3.79129469e-01 1.21622890e-01 8.73021036e-02 8.38390708e-01
6.01486452e-02 -1.82129890e-01 -5.80945194e-01 -1.79109141e-01
9.43942964e-02 1.40565717e+00 -1.38641250e+00 -2.40662545e-01
-6.04756296e-01 9.45574105e-01 1.69862270e-01 4.40259606e-01
-5.49534142e-01 7.47577520e-03 6.29593492e-01 5.40250003e-01
6.25075638e-01 -8.25159490e-01 -1.69963703e-01 -1.15105546e+00
1.80580288e-01 -4.69280332e-01 1.74654901e-01 -5.87433398e-01
-1.40366983e+00 4.34959143e-01 -1.17550559e-01 -1.15870702e+00
4.37237062e-02 -7.69306540e-01 -8.41446459e-01 1.83478028e-01
-1.12336767e+00 -9.00390804e-01 -2.42451072e-01 7.67613590e-01
8.76035988e-01 2.69076955e-02 3.26550782e-01 1.39116496e-01
-6.49657190e-01 -1.25914682e-02 8.10638785e-01 -1.63272232e-01
2.89097607e-01 -1.08717966e+00 3.72349590e-01 8.08772385e-01
1.05361797e-01 7.11246550e-01 1.07160628e+00 -6.47846937e-01
-2.10821605e+00 -9.14340615e-01 1.30080014e-01 -6.04484320e-01
9.84836280e-01 -4.13430244e-01 -1.11406732e+00 4.08277184e-01
-1.50567695e-01 4.53822672e-01 1.70145392e-01 -4.56465811e-01
8.60552341e-02 -6.62077591e-02 -9.96297956e-01 4.88507003e-01
8.60529125e-01 -5.18703759e-01 -9.60734487e-03 4.75942969e-01
4.15689349e-01 -2.02910841e-01 -1.01680338e+00 4.30467665e-01
5.23182690e-01 -9.00323272e-01 1.19915581e+00 -5.60337007e-01
4.97008801e-01 -5.63181162e-01 -2.80319061e-02 -6.66934073e-01
-2.89208323e-01 -1.09338534e+00 -2.29164854e-01 7.74452448e-01
9.91330519e-02 -3.97851288e-01 1.08619094e+00 2.20754579e-01
-1.72531784e-01 -7.85573184e-01 -1.22535825e+00 -1.05840313e+00
-2.06012547e-01 -3.52071702e-01 7.12140426e-02 6.60472333e-01
-2.26520061e-01 1.05886109e-01 -2.62984186e-01 2.60000080e-01
1.20318413e+00 6.23104990e-01 6.45299137e-01 -1.09430218e+00
-1.09148070e-01 -5.75719059e-01 -5.11165857e-01 -9.83418941e-01
6.67360146e-03 -6.19865477e-01 2.73774654e-01 -9.65001822e-01
5.51540315e-01 -4.29510385e-01 -2.95309395e-01 -2.89209396e-01
-4.57434282e-02 2.46001944e-01 -8.91738981e-02 5.17136157e-01
-7.86979079e-01 5.08166134e-01 1.29541576e+00 -2.84749180e-01
-7.96566680e-02 1.53367832e-01 2.25000620e-01 6.26579821e-01
3.81033987e-01 -4.97477770e-01 -7.47904330e-02 -1.97595149e-01
8.25641677e-02 2.83054709e-01 6.24062836e-01 -8.17953706e-01
3.62693757e-01 -3.05314839e-01 1.63865343e-01 -1.07594228e+00
5.01868367e-01 -7.99284339e-01 2.53331363e-01 2.93298155e-01
-2.67656092e-02 4.00371045e-01 2.97804121e-02 7.85226643e-01
-1.60631351e-02 -1.26341149e-01 9.32122707e-01 4.43866178e-02
-5.06747484e-01 8.06025147e-01 -7.04148352e-01 -1.41883373e-01
1.17422926e+00 -2.60765076e-01 -2.26868495e-01 -1.26598537e-01
-6.03020668e-01 -4.01932336e-02 8.58697593e-01 -1.77189276e-01
8.35291922e-01 -1.26804686e+00 -4.53961223e-01 -3.29954922e-02
6.13798201e-02 4.15854976e-02 5.38735747e-01 1.33878028e+00
-5.98857701e-01 3.25572014e-01 -1.42951142e-02 -1.15375197e+00
-1.41663587e+00 6.51962996e-01 6.18943572e-02 -2.70952940e-01
-1.08441079e+00 6.27348304e-01 4.07782316e-01 1.87709764e-01
-2.42124274e-01 -4.05314684e-01 1.30361333e-01 -2.69818664e-01
6.61315858e-01 5.76061606e-01 1.98771536e-01 -8.02412570e-01
-2.51932263e-01 8.25155377e-01 -3.65283310e-01 -6.60849661e-02
1.69605470e+00 -3.53946388e-01 -4.36540633e-01 8.98921132e-01
1.36445951e+00 2.05424875e-02 -1.69536972e+00 -3.62360656e-01
2.67084360e-01 -4.33750629e-01 -4.93573807e-02 2.09908217e-01
-9.62912798e-01 6.53943837e-01 3.91290396e-01 4.48476911e-01
8.22553456e-01 3.40224504e-01 9.09860313e-01 5.72884083e-01
4.88513231e-01 -9.41119909e-01 3.35051626e-01 3.56514543e-01
5.99530578e-01 -1.21349883e+00 2.76548952e-01 -3.34683120e-01
-1.74205348e-01 1.40049052e+00 3.90456505e-02 -2.43750781e-01
5.70539653e-01 3.61215919e-01 -4.52157944e-01 -6.62231207e-01
-8.38167846e-01 2.60754853e-01 8.64143297e-02 2.80079693e-01
3.08840483e-01 -7.56551977e-03 4.45190072e-02 -5.05676270e-02
2.14789912e-01 -3.22680384e-01 6.60636425e-01 1.12519908e+00
-3.96694779e-01 -7.69929588e-01 -5.16081572e-01 7.16304421e-01
-6.44602239e-01 1.15984082e-01 2.90283233e-01 4.97251093e-01
-4.76791501e-01 5.25967121e-01 2.62875378e-01 3.75663489e-02
4.85280268e-02 -2.15662330e-01 6.47439241e-01 -2.74870872e-01
-4.52441461e-02 2.00073242e-01 -3.03592861e-01 -5.89927614e-01
-6.45614862e-01 -1.30422926e+00 -1.30832815e+00 -4.03013051e-01
-8.28655213e-02 -1.06640913e-01 6.65389478e-01 9.90457416e-01
9.32286307e-03 5.74842811e-01 8.54241967e-01 -1.30936456e+00
-7.19270930e-02 -3.88885289e-01 -6.09551847e-01 5.35449862e-01
8.41309607e-01 -7.16121376e-01 -4.10239697e-01 5.58343112e-01] | [12.214000701904297, -2.6167595386505127] |
4264f803-2ca2-4529-9703-5ef4402d876b | a-useful-criterion-on-studying-consistent | 2109.14950 | null | https://arxiv.org/abs/2109.14950v2 | https://arxiv.org/pdf/2109.14950v2.pdf | A useful criterion on studying consistent estimation in community detection | In network analysis, developing a unified theoretical framework that can compare methods under different models is an interesting problem. This paper proposes a partial solution to this problem. We summarize the idea of using separation condition for a standard network and sharp threshold of Erd\"os-R\'enyi random graph to study consistent estimation, compare theoretical error rates and requirements on network sparsity of spectral methods under models that can degenerate to stochastic block model as a four-step criterion SCSTC. Using SCSTC, we find some inconsistent phenomena on separation condition and sharp threshold in community detection. Especially, we find original theoretical results of the SPACL algorithm introduced to estimate network memberships under the mixed membership stochastic blockmodel were sub-optimal. To find the formation mechanism of inconsistencies, we re-establish theoretical convergence rates of this algorithm by applying recent techniques on row-wise eigenvector deviation. The results are further extended to the degree corrected mixed membership model. By comparison, our results enjoy smaller error rates, lesser dependence on the number of communities, weaker requirements on network sparsity, and so forth. Furthermore, separation condition and sharp threshold obtained from our theoretical results match classical results, which shows the usefulness of this criterion on studying consistent estimation. | ['Huan Qing'] | 2021-09-30 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [ 3.38704228e-01 1.53978735e-01 -4.17494923e-01 1.09413955e-02
1.10867210e-02 -5.35794973e-01 1.72056481e-01 -2.99541861e-01
-1.14735030e-01 8.59979033e-01 -1.82642072e-01 -2.30656072e-01
-7.40531802e-01 -7.59741008e-01 -3.19128662e-01 -8.57872784e-01
-2.79635757e-01 5.02160490e-01 3.69053036e-01 -1.22432381e-01
3.96630704e-01 4.34688747e-01 -1.01171756e+00 -2.24069715e-01
1.17383909e+00 4.82196182e-01 1.72842115e-01 5.99241674e-01
-4.57751714e-02 3.95405263e-01 -3.27969283e-01 -2.68875688e-01
5.27825177e-01 -7.06848145e-01 -7.21738696e-01 4.66387391e-01
1.63997024e-01 1.72983944e-01 -4.45162863e-01 1.57736278e+00
3.26224536e-01 -2.29675636e-01 9.57406938e-01 -1.58774722e+00
-4.85666901e-01 1.08787715e+00 -1.11651790e+00 3.96926731e-01
3.34762752e-01 -2.10107952e-01 8.99099946e-01 -5.99933445e-01
7.83889532e-01 1.12002027e+00 1.04108596e+00 2.12822646e-01
-1.48613429e+00 -8.41482341e-01 9.31866840e-02 2.15941697e-01
-1.97389042e+00 -3.87476057e-01 6.73651636e-01 -4.08466578e-01
3.22908759e-01 4.81115162e-01 5.73633015e-01 6.28457963e-01
-1.34626701e-01 2.74825990e-01 1.25435269e+00 -4.88485098e-01
-5.98322637e-02 2.36450464e-01 4.40760761e-01 7.37546980e-01
1.15111315e+00 -2.55196840e-01 -1.89076483e-01 -3.15257639e-01
9.42218363e-01 -1.88423246e-01 -5.42266965e-01 -5.25007367e-01
-1.04697239e+00 7.30337560e-01 1.70681506e-01 8.06204617e-01
-4.43489812e-02 1.44569248e-01 7.72474334e-02 5.72884738e-01
3.03724855e-01 7.05577582e-02 -1.42891243e-01 4.10986036e-01
-1.25389671e+00 -1.30272359e-01 9.89306331e-01 1.19654965e+00
6.80068195e-01 1.30647719e-01 2.13364065e-01 6.26826108e-01
2.78495610e-01 7.04064250e-01 -5.84916510e-02 -1.16806734e+00
4.08739477e-01 3.25530082e-01 -2.25619182e-01 -1.53000903e+00
-2.13495061e-01 -9.81377244e-01 -1.44399917e+00 -1.93763703e-01
5.20313084e-01 -8.06063116e-02 -3.30519587e-01 1.76447737e+00
1.14301011e-01 4.66804326e-01 -8.55102763e-02 7.87841320e-01
4.74523008e-01 4.79105532e-01 -5.59472382e-01 -1.10086882e+00
9.32899714e-01 -5.64717412e-01 -9.12136316e-01 4.41877633e-01
6.10727429e-01 -5.10469019e-01 4.22369778e-01 5.56849003e-01
-1.04263949e+00 -2.67233610e-01 -1.02740943e+00 8.45083892e-01
1.07933767e-02 4.52540107e-02 5.06640494e-01 1.09992266e+00
-1.36485088e+00 7.12632596e-01 -3.30996633e-01 -7.26551354e-01
-1.29511371e-01 4.37915415e-01 -3.14302474e-01 -2.88260486e-02
-1.05600989e+00 5.49499094e-01 5.29455960e-01 2.74355382e-01
-4.16128010e-01 -3.06176543e-01 -4.36289549e-01 1.05549112e-01
5.25853992e-01 -5.25770128e-01 3.39320391e-01 -1.15715075e+00
-9.94813263e-01 6.54310346e-01 -1.56230241e-01 -4.40238833e-01
5.37279487e-01 4.94172215e-01 -5.38154960e-01 3.77353430e-01
3.49314332e-01 -8.29964206e-02 6.74864173e-01 -1.36925590e+00
-5.08966148e-02 -1.97003141e-01 -3.60416591e-01 -3.43943238e-02
-3.15662086e-01 2.23552119e-02 -2.42197245e-01 -5.42845905e-01
8.32446456e-01 -8.47867608e-01 -4.32522625e-01 -3.14842910e-01
-5.21259308e-01 -1.61410004e-01 6.44681275e-01 -3.86473745e-01
1.61652792e+00 -2.08231521e+00 3.97143751e-01 1.13237274e+00
6.74941599e-01 -1.51126459e-01 -3.16520557e-02 6.61851466e-01
-4.10604864e-01 5.53256929e-01 -3.98777217e-01 2.81896472e-01
-2.04068080e-01 5.38413152e-02 2.34495774e-02 9.33862507e-01
-3.03127259e-01 -2.51138322e-02 -7.47955680e-01 -8.23959589e-01
-2.53192127e-01 -3.77142802e-02 -5.71924686e-01 -3.50415945e-01
4.47402984e-01 4.24859710e-02 -3.15588266e-01 5.50482810e-01
1.05041552e+00 -6.50103450e-01 8.40283513e-01 -3.79451245e-01
-8.80217627e-02 -2.91472167e-01 -2.03743052e+00 9.96685743e-01
8.89647603e-02 5.50163388e-01 6.85351908e-01 -1.46105611e+00
8.53637099e-01 4.65650201e-01 5.89624465e-01 2.05345050e-01
2.10383683e-01 3.89294237e-01 2.98119366e-01 -3.18369180e-01
1.89772338e-01 6.94723800e-02 2.12761372e-01 4.75374371e-01
-3.12568471e-02 1.72909290e-01 6.07708395e-01 7.22450376e-01
1.14856327e+00 -4.36622232e-01 3.62357229e-01 -9.73541737e-01
7.12731779e-01 -1.80071697e-01 5.79425395e-01 9.55801487e-01
-1.46135330e-01 3.19720656e-01 8.12979758e-01 2.23051444e-01
-1.01340878e+00 -8.54273558e-01 -2.12291777e-01 5.60613275e-01
4.08110797e-01 -4.89956319e-01 -6.64726257e-01 -3.74466240e-01
-3.97140458e-02 5.52374199e-02 -3.95014971e-01 1.68408722e-01
-3.33118886e-01 -1.12197828e+00 4.91143078e-01 2.82858033e-02
6.51871026e-01 -3.78616363e-01 3.70245457e-01 3.26218158e-02
-3.48537743e-01 -1.02085769e+00 -4.61349845e-01 6.00345731e-02
-9.95489240e-01 -1.43431354e+00 -6.87821567e-01 -9.57464755e-01
8.77369165e-01 6.21064782e-01 9.49242771e-01 5.49281240e-01
6.13895059e-02 3.82801086e-01 -3.38457495e-01 3.91937643e-01
-4.72930521e-01 2.04958051e-01 4.79711950e-01 -4.31229509e-02
-5.73571846e-02 -9.94168878e-01 -4.28478092e-01 5.49634993e-01
-6.68407261e-01 -1.82140529e-01 5.63828528e-01 5.70519686e-01
1.39048591e-01 4.58494306e-01 6.56931043e-01 -8.23907971e-01
7.33148813e-01 -6.69121921e-01 -6.22100830e-01 2.66259074e-01
-1.13365221e+00 1.58894658e-01 4.21713233e-01 -3.13621968e-01
-7.52312839e-01 -1.67090282e-01 3.37449133e-01 -2.06113294e-01
4.29530025e-01 5.57948947e-01 -6.58828840e-02 -2.92029619e-01
5.99838734e-01 2.46137232e-01 1.68216586e-01 -2.39550680e-01
1.96002513e-01 7.33837426e-01 2.58218408e-01 -5.28870225e-01
1.03626454e+00 4.91730511e-01 2.91209936e-01 -9.91430342e-01
-4.20153469e-01 -7.84904480e-01 -4.82212573e-01 -2.96666235e-01
3.75515133e-01 -8.29514921e-01 -8.45983982e-01 2.38330781e-01
-1.03809583e+00 1.87729463e-01 2.23826990e-01 5.13884962e-01
-4.00296897e-01 1.10996735e+00 -7.38390565e-01 -1.09798896e+00
-2.17440709e-01 -6.33690476e-01 2.93256223e-01 -2.38690883e-01
5.08383848e-02 -1.06527591e+00 1.96359172e-01 2.84649469e-02
1.32752284e-01 7.02825189e-02 6.48514569e-01 -5.45963407e-01
-6.24993145e-01 1.21743329e-01 -6.63110435e-01 2.01437965e-01
-1.31254777e-01 2.63162255e-01 -2.32323825e-01 -5.92303693e-01
-2.85407528e-02 3.58196646e-01 9.47396576e-01 5.16530097e-01
9.78124738e-01 -3.20332646e-01 -7.30179071e-01 5.86743712e-01
1.91144073e+00 -1.48215219e-01 6.37945056e-01 1.96021423e-02
3.88250947e-01 4.38711315e-01 3.16232502e-01 4.15996611e-01
-8.81798863e-02 4.44310278e-01 2.94588000e-01 -8.43403414e-02
7.78352050e-03 6.83699697e-02 2.58629888e-01 1.52589393e+00
-5.01709342e-01 -4.04419541e-01 -6.81778848e-01 5.34585774e-01
-1.94737720e+00 -1.40025568e+00 -8.43342841e-01 2.10388136e+00
7.52985537e-01 1.61222160e-01 3.88380319e-01 4.12060171e-01
1.50718510e+00 4.95802127e-02 3.60860787e-02 -1.19711636e-02
-4.67345744e-01 -1.39912903e-01 9.58370626e-01 6.06531382e-01
-5.82534552e-01 5.52790344e-01 7.14075327e+00 1.25810516e+00
-5.86790740e-01 1.37426257e-01 3.22082639e-01 4.33619678e-01
-4.78126973e-01 4.51581210e-01 -6.83926642e-01 4.53167707e-01
6.19517386e-01 -2.54386932e-01 3.61556768e-01 5.57718635e-01
3.25548649e-01 -2.08581701e-01 -5.85416555e-01 1.00883067e+00
1.00933798e-01 -1.10416937e+00 -9.81122032e-02 3.50769341e-01
8.78211617e-01 -3.83315265e-01 -4.17732120e-01 -1.45969883e-01
5.00276089e-01 -9.02522862e-01 1.94612786e-01 4.57989722e-01
8.59505057e-01 -5.24042785e-01 7.62211025e-01 3.69839698e-01
-1.46204042e+00 1.12228885e-01 -5.21566868e-01 -2.77610034e-01
-1.22673862e-01 8.73741210e-01 -7.07668126e-01 8.86006236e-01
3.92762601e-01 8.69440436e-01 -2.84742415e-01 1.20284092e+00
2.64996797e-01 7.69656837e-01 -6.24581635e-01 -1.68804452e-01
-9.38182101e-02 -7.68380702e-01 7.70883739e-01 1.22578955e+00
4.99736339e-01 1.28370017e-01 1.75738886e-01 7.91855097e-01
1.52834073e-01 2.77653247e-01 -7.15734363e-01 1.31030649e-01
6.91533446e-01 1.09981501e+00 -1.26062882e+00 -4.07000571e-01
-1.81672335e-01 6.34694636e-01 6.25315905e-02 5.15298367e-01
-8.55280757e-01 -4.61004972e-01 3.89256291e-02 5.32292724e-01
2.54779279e-01 -4.21718061e-01 -3.60685706e-01 -1.27769077e+00
-1.85442716e-01 -8.34223628e-01 3.36478919e-01 -4.73649621e-01
-1.19439423e+00 3.53460222e-01 3.34828913e-01 -1.28231990e+00
1.32744059e-01 -4.72890854e-01 -5.69387496e-01 4.95470762e-01
-9.66519177e-01 -5.53780615e-01 -8.24757144e-02 7.46143460e-01
1.13445923e-01 -2.67852962e-01 2.38546923e-01 4.62229460e-01
-7.47735918e-01 4.51390833e-01 4.61381018e-01 1.03329219e-01
4.71265733e-01 -9.26160038e-01 -4.66466427e-01 1.20456672e+00
-3.62859992e-03 7.40174115e-01 1.03634310e+00 -8.45262110e-01
-9.82639134e-01 -5.82422912e-01 7.31358647e-01 -4.83893715e-02
9.20168757e-01 -1.76789641e-01 -6.38610840e-01 4.50027585e-01
4.26347733e-01 -3.67681503e-01 4.88388926e-01 2.59965062e-01
-2.28599370e-01 -1.83887765e-01 -1.08719373e+00 5.75329900e-01
1.57256961e+00 -2.02462494e-01 -3.20035368e-01 4.20472741e-01
4.12207305e-01 3.29974741e-01 -8.70867729e-01 5.14618456e-01
3.70015293e-01 -1.09707046e+00 8.26368630e-01 -1.26804531e-01
-7.86336660e-02 -4.32714075e-01 -1.04652792e-01 -8.60041559e-01
-6.25777185e-01 -8.32886338e-01 2.01850265e-01 1.04251134e+00
3.99718761e-01 -8.13122094e-01 8.70795608e-01 -2.56530166e-01
2.01257065e-01 -3.00556421e-01 -9.06903863e-01 -1.07216203e+00
-1.81180835e-01 -5.66396415e-02 1.14212133e-01 1.47818112e+00
3.12019527e-01 2.16808721e-01 -4.66531157e-01 1.89038098e-01
1.03111339e+00 8.12354311e-02 5.80901682e-01 -1.55422819e+00
-4.15177524e-01 -5.26757836e-01 -2.20571145e-01 -1.01978660e+00
1.76306054e-01 -9.65892315e-01 -3.11902910e-01 -1.35201871e+00
6.02934241e-01 -6.89014137e-01 -1.25990093e-01 -7.74602890e-02
1.42803162e-01 2.61870682e-01 7.89288804e-02 5.68573356e-01
-7.28367269e-01 1.16390340e-01 1.09654486e+00 6.43119961e-02
-1.07381791e-01 1.44392475e-01 -4.47710752e-01 7.34048128e-01
7.37296700e-01 -6.31455839e-01 -4.72376972e-01 1.78743348e-01
6.32057667e-01 4.36209083e-01 2.15457454e-01 -1.17383564e+00
3.85172099e-01 -2.29843184e-01 -1.65301785e-02 -5.84272921e-01
5.52984029e-02 -1.00037527e+00 6.19200766e-01 8.67444932e-01
-7.97545016e-02 6.03889301e-02 -4.58500147e-01 9.44825947e-01
-3.03870044e-03 -6.77645564e-01 6.24468923e-01 -1.99071199e-01
-3.43596727e-01 1.63882896e-01 -5.41431427e-01 1.45775378e-01
8.17636728e-01 -8.49050164e-01 -1.78468913e-01 -6.54226184e-01
-1.12492347e+00 2.24944562e-01 5.19364953e-01 -4.09382612e-01
4.12868440e-01 -1.23556936e+00 -8.45473945e-01 9.58044678e-02
-4.26251113e-01 -5.74163496e-01 5.07932715e-02 1.43082392e+00
-6.58045411e-01 1.75864711e-01 -1.42640188e-01 -7.48060524e-01
-1.45094883e+00 5.45599699e-01 4.11429733e-01 -2.75055200e-01
-3.71586881e-03 5.82567811e-01 -5.95499650e-02 -2.36937299e-01
-2.22756900e-02 -1.57249887e-02 -1.00918368e-01 2.29318626e-02
-1.42823324e-01 7.11002767e-01 -3.17731649e-01 -4.60855633e-01
-3.26160997e-01 7.73709357e-01 3.13219488e-01 -1.06819812e-02
1.04415703e+00 -5.08398592e-01 -6.32066429e-01 1.55423209e-01
8.90992701e-01 5.44811189e-01 -5.67337155e-01 -2.36281216e-01
-7.49229640e-02 -3.47511232e-01 -2.80208081e-01 -2.33203825e-02
-1.18791938e+00 1.74621493e-01 5.01546741e-01 8.91509652e-01
9.27216351e-01 -6.20550616e-03 9.75185111e-02 3.67460668e-01
4.85899717e-01 -1.07392561e+00 -1.41817063e-01 3.20517004e-01
6.94471419e-01 -8.26171994e-01 4.23485994e-01 -1.28226817e+00
-1.00488700e-01 1.19471192e+00 4.50313717e-01 -5.13921261e-01
8.89065981e-01 1.39655933e-01 -6.89107835e-01 -1.89349666e-01
-4.75396007e-01 -3.50969791e-01 -7.04859048e-02 6.24114096e-01
3.73358935e-01 -2.11084019e-02 -9.94380713e-01 3.21033627e-01
-1.56597905e-02 -2.39629030e-01 9.52704012e-01 2.87232608e-01
-7.15987146e-01 -9.17167068e-01 -5.88268876e-01 4.45266932e-01
-3.04218411e-01 -2.60153949e-01 -3.76109272e-01 1.04612982e+00
4.35375161e-02 9.36241746e-01 -2.23782957e-01 -3.31862152e-01
-1.29153028e-01 -2.60563850e-01 4.89819437e-01 -3.57827842e-01
-2.27538809e-01 3.19369823e-01 2.94638276e-01 -6.59530684e-02
-9.83431518e-01 -6.76072896e-01 -7.44031191e-01 -8.77629995e-01
-1.00042975e+00 7.55699158e-01 8.73874724e-02 6.53224647e-01
1.63291126e-01 2.11277068e-01 5.78906298e-01 -3.12860727e-01
-5.56465149e-01 -1.02058208e+00 -1.07990456e+00 8.06675702e-02
-7.92044625e-02 -5.98193288e-01 -9.35220778e-01 -6.72437623e-02] | [6.976063251495361, 5.200781345367432] |
08e5f8aa-f4ee-41af-af4c-137ada9108c9 | mvp-unified-motion-and-visual-self-supervised | 2003.00667 | null | https://arxiv.org/abs/2003.00667v1 | https://arxiv.org/pdf/2003.00667v1.pdf | MVP: Unified Motion and Visual Self-Supervised Learning for Large-Scale Robotic Navigation | Autonomous navigation emerges from both motion and local visual perception in real-world environments. However, most successful robotic motion estimation methods (e.g. VO, SLAM, SfM) and vision systems (e.g. CNN, visual place recognition-VPR) are often separately used for mapping and localization tasks. Conversely, recent reinforcement learning (RL) based methods for visual navigation rely on the quality of GPS data reception, which may not be reliable when directly using it as ground truth across multiple, month-spaced traversals in large environments. In this paper, we propose a novel motion and visual perception approach, dubbed MVP, that unifies these two sensor modalities for large-scale, target-driven navigation tasks. Our MVP-based method can learn faster, and is more accurate and robust to both extreme environmental changes and poor GPS data than corresponding vision-only navigation methods. MVP temporally incorporates compact image representations, obtained using VPR, with optimized motion estimation data, including but not limited to those from VO or optimized radar odometry (RO), to efficiently learn self-supervised navigation policies via RL. We evaluate our method on two large real-world datasets, Oxford Robotcar and Nordland Railway, over a range of weather (e.g. overcast, night, snow, sun, rain, clouds) and seasonal (e.g. winter, spring, fall, summer) conditions using the new CityLearn framework; an interactive environment for efficiently training navigation agents. Our experimental results, on traversals of the Oxford RobotCar dataset with no GPS data, show that MVP can achieve 53% and 93% navigation success rate using VO and RO, respectively, compared to 7% for a vision-only method. We additionally report a trade-off between the RL success rate and the motion estimation precision. | ['Marvin Chancán', 'Michael Milford'] | 2020-03-02 | null | null | null | null | ['radar-odometry'] | ['robots'] | [-2.61680514e-01 -3.28235626e-01 -2.28444368e-01 -2.24595293e-01
-7.42642105e-01 -7.50539362e-01 8.02336991e-01 -9.68634859e-02
-1.07957029e+00 1.01440084e+00 -1.71610907e-01 -3.42906237e-01
-7.12843612e-02 -9.88021791e-01 -1.07734430e+00 -7.17541456e-01
-3.27995002e-01 5.57543814e-01 5.25150299e-01 -5.58136046e-01
2.95665979e-01 5.60988545e-01 -1.72911000e+00 -6.02776825e-01
8.19135070e-01 6.98734224e-01 8.02688539e-01 9.71338928e-01
1.20100163e-01 6.24236763e-01 -1.74431339e-01 4.40340012e-01
2.75295287e-01 6.14652112e-02 -5.03868282e-01 -1.34177551e-01
2.01779693e-01 -1.97073892e-01 -5.39914489e-01 7.93595612e-01
4.36052382e-01 5.06733477e-01 5.40981531e-01 -1.27351213e+00
-3.40911716e-01 1.69467721e-02 -3.69967371e-01 9.71348584e-02
4.47253555e-01 5.44447780e-01 6.97631538e-01 -7.04734862e-01
9.07483160e-01 1.01320231e+00 8.26215625e-01 3.93201977e-01
-1.04075122e+00 -3.13112348e-01 2.62931198e-01 3.78401160e-01
-1.31339216e+00 -3.48734081e-01 2.66963989e-01 -4.41958696e-01
1.20918274e+00 -1.07424989e-01 6.54795527e-01 1.36456239e+00
4.79952306e-01 5.37566960e-01 1.03615594e+00 -1.08375251e-01
5.52515149e-01 -1.44254686e-02 -2.45881915e-01 8.56711805e-01
1.99675113e-01 5.87070465e-01 -5.95913649e-01 2.08511636e-01
7.90677428e-01 -2.18310021e-02 -3.98643494e-01 -8.30117226e-01
-1.47291315e+00 8.82077932e-01 9.40066993e-01 -2.79701203e-02
-5.47969043e-01 6.10628724e-01 3.79856676e-02 4.12486762e-01
-9.46722478e-02 2.87399411e-01 -6.42327726e-01 -3.90523821e-01
-7.73288190e-01 3.49582762e-01 8.59524310e-01 1.04578757e+00
1.27228844e+00 4.06199157e-01 4.32908177e-01 6.35474086e-01
6.68321431e-01 1.27559805e+00 6.28143370e-01 -1.12929738e+00
5.11849940e-01 1.74033269e-02 6.62594318e-01 -1.09343731e+00
-9.37917113e-01 -1.60061732e-01 -6.35795236e-01 5.08874595e-01
4.17527318e-01 -2.46990830e-01 -1.29519820e+00 1.85781324e+00
3.50575984e-01 3.20462734e-01 6.66617990e-01 1.08925152e+00
7.79876232e-01 6.82110250e-01 -9.18730646e-02 2.12695509e-01
1.09684968e+00 -1.15544033e+00 -3.34121555e-01 -9.44467962e-01
7.98888922e-01 -2.45596200e-01 8.67103755e-01 1.60721034e-01
-3.41418803e-01 -5.84242404e-01 -1.29845142e+00 2.06501275e-01
-6.87484980e-01 -4.84693944e-02 5.64791739e-01 2.51617372e-01
-1.45394051e+00 4.89649415e-01 -1.14732039e+00 -9.42752421e-01
-3.56512740e-02 3.34996551e-01 -7.13868320e-01 -4.04712528e-01
-1.30757523e+00 1.24208808e+00 3.65090877e-01 2.86577940e-01
-1.39060009e+00 -7.80904964e-02 -1.41210485e+00 -3.90210658e-01
2.51655549e-01 -7.54701436e-01 1.06869364e+00 -7.00916529e-01
-1.50127721e+00 5.40202141e-01 -2.41216242e-01 -8.38532865e-01
3.68083656e-01 -2.49937072e-01 -2.88518846e-01 -2.43705995e-02
3.29739630e-01 1.10950291e+00 5.13354242e-01 -1.53787458e+00
-9.53367770e-01 -2.64701456e-01 1.15436718e-01 6.27369285e-01
5.04196882e-01 -8.35162997e-01 -4.61829841e-01 7.01011643e-02
5.45430899e-01 -1.31681883e+00 -7.28787184e-01 -3.23999487e-02
6.09113425e-02 1.83834016e-01 4.86966014e-01 -5.15826643e-01
2.90770710e-01 -1.92849386e+00 1.06855281e-01 1.97939262e-01
-2.24895209e-01 -1.44733071e-01 -2.63495177e-01 4.83421564e-01
6.89947128e-01 -2.81784475e-01 -3.03653419e-01 -2.97546417e-01
1.09974325e-01 1.06429398e+00 -2.23138615e-01 7.31160760e-01
-1.41512632e-01 1.00008523e+00 -1.15117967e+00 -1.80555940e-01
4.45372194e-01 3.14999789e-01 -3.26233804e-01 -1.34836525e-01
-1.51122376e-01 8.33146989e-01 -2.67344862e-01 7.02333510e-01
4.97378260e-01 7.37103820e-03 3.43131498e-02 4.23158199e-01
-4.84724551e-01 5.62494658e-02 -1.17829335e+00 2.15285063e+00
-7.67095089e-01 9.49346721e-01 9.65688080e-02 -7.21026421e-01
1.06457615e+00 -4.44972999e-02 5.99414408e-02 -1.30036056e+00
-2.81008303e-01 5.05378664e-01 -3.86162341e-01 -5.30601859e-01
8.99606466e-01 1.76050276e-01 -3.14834237e-01 -2.32913092e-01
6.25058338e-02 -2.97416151e-01 -1.26197249e-01 3.16695496e-02
1.24950957e+00 5.41331589e-01 3.02211136e-01 2.97634266e-02
3.67285013e-01 6.20005786e-01 6.88761652e-01 1.33180261e+00
-4.92186874e-01 5.32989681e-01 -8.47300142e-02 -4.71363157e-01
-9.36895847e-01 -1.18826044e+00 5.53916283e-02 9.71958935e-01
8.73431027e-01 9.31194201e-02 -7.73974210e-02 -5.27679324e-01
1.82904571e-01 4.71261591e-01 -4.26487207e-01 -1.04394637e-01
-6.57306015e-01 -7.71128297e-01 4.98416752e-01 4.28993523e-01
7.10153699e-01 -1.21499074e+00 -9.17219341e-01 4.34742808e-01
-2.81724036e-01 -1.25429225e+00 3.16448838e-01 3.63909960e-01
-7.94966936e-01 -7.42050886e-01 -5.24152875e-01 -6.88611865e-01
3.24609160e-01 6.25515223e-01 9.32889223e-01 -1.94710106e-01
2.47572020e-01 8.74678671e-01 -5.16288340e-01 1.30175240e-02
-2.18715277e-02 3.04083619e-02 3.35389644e-01 -4.15036857e-01
-5.66626675e-02 -6.20531619e-01 -6.10403240e-01 4.85214651e-01
-4.52579439e-01 -1.50547728e-01 8.07383895e-01 1.02018225e+00
6.70540154e-01 -4.17235523e-01 3.12695771e-01 -3.69930506e-01
-1.34410992e-01 -8.28554690e-01 -9.01384950e-01 -1.77855745e-01
-6.48199737e-01 1.25376567e-01 2.39828020e-01 -2.60070592e-01
-6.99469924e-01 3.35680276e-01 -2.16221422e-01 -1.16953500e-01
-4.58334684e-01 6.85523212e-01 1.66498736e-01 -5.27893007e-01
9.02031004e-01 7.55084991e-01 -1.45969585e-01 -2.64408812e-02
5.64164162e-01 4.44791079e-01 8.93388093e-01 -2.84410894e-01
1.04559076e+00 8.91976893e-01 5.72376549e-02 -1.01290929e+00
-2.04937041e-01 -6.88189149e-01 -5.33228695e-01 -1.97843343e-01
9.09711361e-01 -1.35777617e+00 -4.36440855e-01 4.46382731e-01
-8.46296966e-01 -7.80733883e-01 2.44470686e-02 8.35462749e-01
-8.77172470e-01 4.03159559e-01 -2.01340720e-01 -9.61297154e-01
-4.08778079e-02 -1.00637293e+00 1.04143453e+00 5.24461091e-01
1.51578963e-01 -1.05945122e+00 4.57385153e-01 4.90088239e-02
4.33916152e-01 4.21659410e-01 1.35126695e-01 -2.09670708e-01
-1.05322671e+00 -7.90572837e-02 -1.73436373e-01 -2.93102622e-01
-1.84992522e-01 -4.76140559e-01 -9.32835400e-01 -5.25738657e-01
-4.32687879e-01 -3.33870947e-01 1.22478306e+00 4.13331658e-01
5.49008697e-02 -1.23544991e-01 -5.90431154e-01 9.24098849e-01
1.85427427e+00 3.41928482e-01 7.36469686e-01 1.17200494e+00
7.67034054e-01 4.00513589e-01 1.13148117e+00 3.85875940e-01
1.06148505e+00 6.92464173e-01 1.13833630e+00 -3.61804850e-03
2.67795235e-01 -2.85896569e-01 5.72804034e-01 4.10519570e-01
-8.52282941e-02 -2.41653368e-01 -1.07679343e+00 8.90000641e-01
-2.29586506e+00 -7.03558803e-01 -6.06474429e-02 2.30842924e+00
1.01698235e-01 2.38228545e-01 -3.44374418e-01 -3.21596265e-01
1.75452098e-01 3.50616127e-01 -5.33204973e-01 -1.03436530e-01
-3.41632575e-01 -2.41640344e-01 1.23037148e+00 7.67036140e-01
-1.18535507e+00 1.34104931e+00 5.25802660e+00 3.06001812e-01
-1.09928727e+00 1.86707258e-01 -3.57873887e-01 2.85770684e-01
-2.60580838e-01 2.21995533e-01 -8.32826853e-01 2.48708248e-01
1.16850638e+00 5.29485047e-01 5.42784154e-01 1.09085262e+00
2.87113786e-01 -6.39758110e-01 -6.28634989e-01 1.14036465e+00
-2.00671822e-01 -1.33428109e+00 -5.60386717e-01 2.18949467e-02
5.68681955e-01 9.61955190e-01 -5.63895218e-02 7.98114121e-01
8.18316877e-01 -8.99361730e-01 9.45301652e-01 5.88461339e-01
3.05141509e-01 -6.71091437e-01 9.00758982e-01 6.22317493e-01
-1.33782697e+00 -1.79274634e-01 -4.15105820e-01 -9.03303400e-02
2.54395008e-01 1.33602992e-01 -9.54370916e-01 8.85169804e-01
9.63916004e-01 9.89268422e-01 -3.81069034e-01 1.22866166e+00
-4.48207825e-01 2.45465979e-01 -5.39595246e-01 -2.12979481e-01
8.26925933e-01 -1.04725003e-01 7.54912198e-01 9.61834669e-01
4.56159055e-01 -3.15978140e-01 3.83480877e-01 4.32598621e-01
4.62983102e-01 -3.04631561e-01 -1.12700105e+00 6.16249800e-01
5.71260333e-01 1.19173694e+00 -5.02102196e-01 -1.39390394e-01
-3.43977422e-01 9.57677662e-01 3.79745603e-01 6.79671526e-01
-8.15378606e-01 -2.30062678e-01 9.52785075e-01 -3.76108110e-01
5.63173413e-01 -1.04222608e+00 1.13463059e-01 -1.07366240e+00
-1.77822053e-01 -4.22980458e-01 3.41383629e-02 -1.09316778e+00
-5.72445452e-01 4.27880108e-01 -3.37770194e-01 -1.45191061e+00
-5.19620657e-01 -6.41110301e-01 -2.58158714e-01 7.75426626e-01
-2.15213370e+00 -1.13043547e+00 -6.59143150e-01 3.92768174e-01
4.11990196e-01 -3.41810249e-02 7.27572978e-01 5.00892708e-03
-1.57655656e-01 -1.69847738e-02 4.83081996e-01 -4.85398360e-02
5.41395605e-01 -1.37286508e+00 6.51486576e-01 9.04132843e-01
4.04365182e-01 1.71002135e-01 8.52777600e-01 -5.45068383e-01
-1.76878321e+00 -1.19471884e+00 6.44002974e-01 -4.57116842e-01
5.34396648e-01 -3.01501065e-01 -7.50135779e-01 8.03510845e-01
-2.08818451e-01 8.83204639e-02 -9.46985334e-02 -1.10878661e-01
-1.99242905e-01 -1.02707028e-01 -1.10071504e+00 8.59149933e-01
1.00504434e+00 -3.26880068e-01 -4.13029522e-01 1.97515354e-01
7.73157060e-01 -7.41935849e-01 -5.77248871e-01 5.07922471e-01
6.27053797e-01 -8.86787772e-01 1.12670374e+00 -2.23285839e-01
-2.20311701e-01 -7.41220057e-01 -5.97750604e-01 -1.50647771e+00
-2.79406756e-01 -1.61720812e-01 8.88652876e-02 6.97647691e-01
5.32035470e-01 -9.42947626e-01 7.40639687e-01 -8.51899683e-02
-2.06843272e-01 -1.74043670e-01 -1.31716204e+00 -8.92080486e-01
-3.15355837e-01 -6.88232422e-01 2.18030110e-01 6.96734071e-01
-5.10844529e-01 -3.67760509e-02 -5.65379739e-01 9.56399202e-01
5.01266778e-01 1.55132497e-03 1.15520799e+00 -1.06395411e+00
-1.12383083e-01 -1.54864974e-03 -8.92205119e-01 -1.28921700e+00
1.25297606e-01 -5.93606591e-01 7.94823706e-01 -1.89384425e+00
-4.96048391e-01 -5.54736674e-01 -1.88035890e-01 4.82975096e-01
3.13538671e-01 1.09909303e-01 -8.44460875e-02 3.94139975e-01
-7.99832284e-01 7.54045725e-01 9.32269573e-01 -1.69520512e-01
-3.96518230e-01 -8.15126300e-02 7.88366646e-02 5.93531609e-01
6.49427235e-01 -4.65532422e-01 -2.48604611e-01 -6.45618558e-01
3.72899413e-01 3.05376112e-01 6.63184762e-01 -1.51521420e+00
5.93109071e-01 -2.99835980e-01 3.84526461e-01 -7.42201030e-01
5.89840591e-01 -7.41763294e-01 2.19736788e-02 7.37427235e-01
2.96314687e-01 1.81102648e-01 3.24342400e-01 1.17704296e+00
-2.16025025e-01 3.34865646e-03 6.04347706e-01 -3.64573598e-01
-1.92889810e+00 1.77376032e-01 -7.99852312e-01 -9.44898874e-02
6.98306322e-01 -3.82458776e-01 -4.79681909e-01 -6.56550586e-01
-7.29411721e-01 7.84377158e-01 6.43262327e-01 5.43278754e-01
9.36317742e-01 -1.22321236e+00 -4.13616985e-01 1.71321034e-01
3.80320430e-01 1.74026027e-01 3.05678993e-01 1.04038835e+00
-8.51869464e-01 4.42090780e-01 -3.49819154e-01 -1.09020936e+00
-7.30126202e-01 2.91630417e-01 4.35004085e-01 -1.19626507e-01
-7.04443038e-01 6.24704003e-01 1.06017785e-02 -1.03645694e+00
-2.70137116e-02 -4.74842340e-01 -3.24156255e-01 -2.23427922e-01
2.29064122e-01 1.89015314e-01 -9.67492685e-02 -7.08341002e-01
-5.51244497e-01 8.41619372e-01 4.71846551e-01 -5.85405529e-01
1.05555749e+00 -7.15544105e-01 4.09490943e-01 6.38355136e-01
7.66481936e-01 -3.34832758e-01 -1.61987662e+00 -2.51265615e-01
2.92621642e-01 -1.70152083e-01 1.06895398e-02 -4.77014542e-01
-7.21192658e-01 5.72285891e-01 9.23834801e-01 -1.38659582e-01
6.67820096e-01 -8.80114809e-02 5.81725359e-01 8.82964909e-01
1.03830636e+00 -9.97460127e-01 -1.65074483e-01 1.07562649e+00
6.17559969e-01 -1.65551591e+00 -3.34686548e-01 2.46515647e-01
-8.22208643e-01 8.64351511e-01 5.64210057e-01 -2.69032240e-01
3.37498397e-01 -1.19063057e-01 4.23540831e-01 8.35102051e-02
-7.15725899e-01 -7.83540964e-01 -6.53930753e-02 1.05386293e+00
-4.97853607e-01 4.78686858e-03 3.61799568e-01 1.13317408e-01
-1.61966264e-01 -2.57095307e-01 6.29277468e-01 1.37253594e+00
-9.69947338e-01 -4.17505682e-01 -5.05646348e-01 -1.36768609e-01
2.79642761e-01 5.85783012e-02 1.24398172e-01 1.21744978e+00
-3.95240933e-02 9.43715453e-01 1.42670438e-01 -5.21708786e-01
3.08242232e-01 -2.79231489e-01 3.75154465e-01 -2.31234491e-01
-1.01403281e-01 -2.77453065e-01 2.99461126e-01 -9.61817145e-01
-3.86545926e-01 -6.48697734e-01 -1.56351244e+00 -1.87129378e-01
-4.70944680e-02 -2.54375413e-02 1.02617049e+00 1.13590944e+00
2.52954036e-01 2.84975022e-01 4.50809330e-01 -1.40861082e+00
-2.63938934e-01 -7.92571366e-01 -5.78627944e-01 -1.97846293e-01
8.53128672e-01 -1.05815887e+00 -4.98222709e-01 -3.95544440e-01] | [7.427687168121338, -1.9276379346847534] |
f4386f99-57f8-4dd7-8b7d-d156cd209be3 | performance-analysis-of-empirical-open-1 | 2306.16547 | null | https://arxiv.org/abs/2306.16547v1 | https://arxiv.org/pdf/2306.16547v1.pdf | Performance Analysis of Empirical Open-Circuit Voltage Modeling in Lithium Ion Batteries, Part-2: Data Collection Procedure | This paper is the second part of a series of papers about empirical approaches to open circuit voltage (OCV) modeling and its performance comparison in lithium-ion batteries. The first part of the series introduced various sources of uncertainties in the OCV models and established a theoretical relationship between uncertainties and the performance of a battery management system. In this paper, clearly defined approaches for low-rate OCV data collection are defined and described in detail. The data collection is designed with consideration to several parameters that affect the experimental time. Firstly, a more suitable method to fully charge the battery at different C-Rates is defined. Secondly, the OCV characterization following the full charge is described for various performance comparisons. Finally, optimal and efficient resistance estimation profiles are discussed. From the voltage, current and time data recorded using the procedure described in this paper, the OCV-SOC relationship is characterized and its uncertainties are modeled in the third part of this series of papers. | ['Balakumar Balasingam', 'James Nguyen', 'Prarthana Pillai'] | 2023-06-28 | null | null | null | null | ['management'] | ['miscellaneous'] | [-3.81186932e-01 -7.09343553e-01 -4.77112412e-01 -2.47663364e-01
-2.39798412e-01 -6.57561600e-01 5.69255590e-01 6.83890402e-01
-4.89207447e-01 1.30324340e+00 -4.77634102e-01 -4.90023375e-01
-5.17372668e-01 -7.59326994e-01 -6.17936671e-01 -6.76735520e-01
3.63738462e-02 6.68764710e-01 2.72465706e-01 -2.76982099e-01
6.94543540e-01 1.10004997e+00 -1.72089112e+00 -4.01575029e-01
1.16099310e+00 1.24558938e+00 3.73792887e-01 5.79477251e-01
-1.09590687e-01 1.70827225e-01 -5.17800510e-01 -7.90411308e-02
-4.09952283e-01 -2.51990944e-01 -4.89561528e-01 -5.48482895e-01
-6.40528440e-01 -7.07593784e-02 -2.01901585e-01 7.70726800e-01
5.80600202e-01 -1.24508943e-02 1.03777587e+00 -1.61887717e+00
1.22508213e-01 6.95862770e-01 2.88006216e-01 4.42837268e-01
3.52152467e-01 -2.09424347e-01 1.69788063e-01 -9.69903290e-01
4.59100294e-04 7.34754205e-01 5.20558238e-01 4.29795533e-01
-1.25589788e+00 -6.01395607e-01 -3.14726681e-01 4.23123598e-01
-1.65294147e+00 -2.54245073e-01 7.35494912e-01 -3.25395435e-01
1.30767870e+00 7.18088388e-01 1.17009735e+00 2.09119692e-01
6.62555575e-01 2.68548131e-01 1.25863326e+00 -4.74324703e-01
6.19438827e-01 8.18571866e-01 6.34973168e-01 -5.78678250e-01
1.15988529e+00 -1.34936171e-02 9.56725478e-02 -1.16600394e-01
1.95182905e-01 -4.53778625e-01 -3.58854145e-01 -4.40764546e-01
4.93019074e-02 3.02687734e-01 -3.28459330e-02 7.35175252e-01
1.34036586e-01 6.74637482e-02 6.22812033e-01 -2.89407223e-01
-1.49987906e-01 2.39462957e-01 -1.80141926e-01 -5.21058142e-01
-7.06773043e-01 3.28777552e-01 1.16549015e+00 1.20271218e+00
3.12695295e-01 2.14993134e-01 7.38867745e-03 5.97438872e-01
5.19502938e-01 7.42447615e-01 3.27653646e-01 -4.30052131e-01
2.37134069e-01 2.63395041e-01 5.59327483e-01 1.37502074e-01
-3.10587287e-01 -9.20196325e-02 -3.18179816e-01 2.81745978e-02
-1.88861112e-03 -8.97675678e-02 -6.80898964e-01 5.86839855e-01
-1.69712961e-01 -4.48962122e-01 3.94066423e-01 3.16112220e-01
9.90325749e-01 7.23648787e-01 6.87415004e-02 -7.71169484e-01
1.09712541e+00 -1.52010292e-01 -1.37600982e+00 1.55661225e-01
7.31103942e-02 -2.58897305e-01 6.11562788e-01 2.27878466e-01
-1.68332684e+00 -2.44935840e-01 -1.53741825e+00 1.35212183e-01
-6.49829447e-01 -1.62192807e-01 4.48633991e-02 1.00197136e+00
-9.84594941e-01 8.50364029e-01 -5.24259567e-01 -2.60382295e-01
7.43623972e-02 4.98385996e-01 2.73042679e-01 3.12715203e-01
-1.42036402e+00 1.51720715e+00 3.76496613e-01 3.95086884e-01
-7.07833707e-01 -7.99224854e-01 -7.30997741e-01 -1.11169189e-01
-1.62461892e-01 6.87465966e-02 1.19883454e+00 1.03444271e-01
-1.40123093e+00 3.64125490e-01 -5.16455948e-01 -5.89728177e-01
6.04868650e-01 1.09454893e-01 -5.93900204e-01 -4.13889624e-02
-4.14736301e-01 -1.17067389e-01 1.33003369e-01 -1.62748682e+00
-3.71307403e-01 -1.17780060e-01 -4.16735977e-01 -1.29397362e-01
9.20192376e-02 -9.52354968e-02 -3.29741925e-01 -1.00901835e-02
-2.02530593e-01 -6.64398134e-01 1.16525114e-01 -5.33615530e-01
7.16744810e-02 -5.66121876e-01 5.88790715e-01 -2.64840156e-01
1.29175007e+00 -1.62898779e+00 8.56836699e-03 6.66421056e-01
-5.12842655e-01 2.71604538e-01 8.25597882e-01 9.08586442e-01
2.84372792e-02 4.16139185e-01 -2.71370888e-01 -1.64965317e-02
1.00134164e-01 4.21284974e-01 -1.74456343e-01 6.89794838e-01
-1.83085464e-02 9.58105505e-01 -6.86042249e-01 -2.80818157e-02
3.23135227e-01 5.48488081e-01 2.51214087e-01 9.92833897e-02
2.81263918e-01 1.67410951e-02 -1.42311975e-01 5.65946758e-01
1.01341867e+00 5.46462953e-01 2.54884243e-01 -5.99068344e-01
-7.63362825e-01 3.38714600e-01 -1.03555477e+00 7.07648873e-01
-3.93688619e-01 5.21414995e-01 -3.65641490e-02 -9.87505317e-01
1.28291094e+00 2.56577492e-01 4.12605613e-01 -1.01783872e+00
5.61747968e-01 9.74217474e-01 -2.02943668e-01 -5.80610812e-01
6.85973346e-01 -4.11505044e-01 2.88183659e-01 -1.38511881e-01
-6.78425506e-02 -8.59405696e-01 5.60333908e-01 1.00451544e-01
2.96522230e-02 -7.69189820e-02 -1.44617409e-01 -1.18549490e+00
9.25345480e-01 -8.02528933e-02 2.00640291e-01 3.71523857e-01
-1.78211674e-01 1.28128985e-02 4.93282020e-01 2.48523965e-01
-1.31715286e+00 -5.74255764e-01 -1.21479392e+00 -1.18147619e-01
9.94402468e-01 -2.35664174e-01 -4.76359695e-01 4.02650386e-01
5.24969399e-01 1.36369240e+00 -3.22448164e-01 -2.79765815e-01
-3.27440321e-01 -7.74881601e-01 3.12563479e-01 1.02289402e+00
1.93609014e-01 -4.39805269e-01 -5.01039624e-01 1.11717716e-01
4.07364890e-02 -9.37526643e-01 2.98878253e-01 7.79621184e-01
-1.22485781e+00 -9.52417314e-01 -4.46571499e-01 -4.23819423e-01
4.54560518e-01 -8.24695230e-02 9.75800276e-01 1.81443900e-01
-2.58399457e-01 5.39211154e-01 -1.04648275e-02 -1.09819484e+00
-5.09438217e-01 -4.13258165e-01 2.81455129e-01 -6.02376580e-01
6.03136659e-01 -4.42492738e-02 -5.37271202e-01 4.73493457e-01
-7.13719487e-01 -8.29518676e-01 9.00417287e-03 2.45546371e-01
8.53099763e-01 1.90858752e-01 9.19330895e-01 -1.75155833e-01
7.79480040e-01 -3.71099770e-01 -9.08795297e-01 2.04781786e-01
-1.40179539e+00 6.60779998e-02 3.27975422e-01 -1.86362430e-01
-9.98187661e-01 -4.64623451e-01 -4.92546827e-01 -1.99777689e-02
1.59299761e-01 1.20837964e-01 -6.19934201e-01 -3.02984774e-01
2.36183330e-01 3.39199156e-01 3.87117505e-01 -5.48902214e-01
-4.20202464e-01 8.11786056e-01 5.54875433e-01 -5.50541759e-01
6.20540917e-01 2.71010548e-01 1.85459822e-01 -8.16374719e-01
1.35639563e-01 -4.36623603e-01 -5.20784318e-01 -4.80850220e-01
5.99006236e-01 -9.41630304e-01 -1.02085507e+00 6.54199719e-01
-5.50158381e-01 -4.17252153e-01 -4.09430027e-01 4.33164120e-01
-2.75722116e-01 2.35282540e-01 -3.11038941e-01 -1.57381749e+00
-6.09247208e-01 -1.49293697e+00 4.24712896e-01 6.30372703e-01
1.18776806e-01 -1.29259765e+00 -2.24685550e-01 -2.70459633e-02
5.00885010e-01 -1.74994141e-01 6.37287259e-01 -4.59804803e-01
-1.74467131e-01 -4.62648898e-01 1.31463185e-01 6.00649595e-01
-2.28944406e-01 2.98457909e-02 -8.55776131e-01 -7.70026505e-01
3.14709514e-01 4.37709503e-03 4.42561418e-01 2.98174262e-01
1.19157290e+00 2.71996886e-01 -7.67696619e-01 2.16318607e-01
2.18600035e+00 9.19546783e-01 1.29012418e+00 5.01418769e-01
6.10099062e-02 1.11493707e-01 8.30281138e-01 2.86901623e-01
2.01346502e-01 4.32621747e-01 5.62683344e-01 3.05046469e-01
4.48308080e-01 -5.12188971e-02 1.17679656e-01 4.92090821e-01
-2.91485399e-01 -4.14963335e-01 -4.55404818e-01 5.74193358e-01
-1.22917378e+00 -6.30696952e-01 -1.01983321e+00 2.70321107e+00
6.75762117e-01 3.08271289e-01 2.54852980e-01 8.17074776e-01
5.79774082e-01 -5.06853461e-01 -5.14668763e-01 -8.93875539e-01
-3.00305784e-01 1.28188804e-01 1.11181164e+00 6.66651011e-01
-3.68755430e-01 -2.39610627e-01 7.45816326e+00 5.82136750e-01
-1.14088190e+00 -5.15267849e-02 2.16489390e-01 1.93410739e-02
-9.18135703e-01 2.40694910e-01 -1.32177103e+00 7.83515990e-01
1.47236133e+00 -6.80212617e-01 7.43587390e-02 2.51266599e-01
2.71218657e-01 -8.84331048e-01 -1.05174148e+00 9.34234798e-01
9.33301672e-02 -1.07047439e+00 -1.41104370e-01 5.87079003e-02
4.35349613e-01 -2.10916534e-01 -5.52817702e-01 1.56364486e-01
-6.79467320e-01 -7.73652613e-01 9.77259636e-01 9.25399542e-01
1.16907883e+00 -1.09061038e+00 1.16575348e+00 3.48907523e-02
-1.24064207e+00 -4.26370770e-01 -5.02241671e-01 7.89015442e-02
4.84482139e-01 5.53148031e-01 -2.40184501e-01 8.82737458e-01
8.92325282e-01 6.71613291e-02 -5.81781387e-01 1.50499439e+00
1.79214716e-01 5.01219511e-01 -5.40728927e-01 -8.76057684e-01
-3.13275129e-01 -6.08753502e-01 3.29784930e-01 9.39374566e-01
4.72643673e-01 1.84801757e-01 -7.24354565e-01 8.82023573e-01
2.82368004e-01 7.96843693e-02 -4.03820962e-01 -8.34674109e-04
8.11829746e-01 9.44338620e-01 -3.80947381e-01 -3.76177341e-01
-1.65408120e-01 1.48967341e-01 -4.80473489e-01 3.68011624e-01
-9.04337823e-01 -7.31371164e-01 2.43670195e-01 7.85233736e-01
-8.20762441e-02 1.44444751e-02 -5.91217041e-01 -5.89410365e-01
8.66117924e-02 1.13632776e-01 3.20095606e-02 -8.03610504e-01
-9.40046549e-01 8.84168074e-02 8.64806414e-01 -9.51298714e-01
2.38846257e-01 -7.34194577e-01 -9.05113518e-01 1.26069963e+00
-1.62599838e+00 -9.32446197e-02 -4.52451646e-01 1.02847256e-01
2.49106094e-01 6.21651746e-02 4.75870669e-01 5.38450778e-01
-5.36524057e-01 6.10825479e-01 1.01025498e+00 -6.13870203e-01
1.85922340e-01 -1.30328405e+00 -1.64314747e-01 3.65186155e-01
-1.14230347e+00 3.11373591e-01 1.07556701e+00 -9.44642723e-01
-2.00185537e+00 -5.19469440e-01 9.01910722e-01 -1.48270324e-01
7.00035036e-01 -4.31892365e-01 -1.22418404e+00 1.56346083e-01
2.09870726e-01 -3.03584278e-01 3.39589894e-01 -7.66192317e-01
7.77413666e-01 -5.89299440e-01 -1.20843613e+00 1.10289656e-01
3.29448372e-01 -5.81076503e-01 -4.37943757e-01 4.61598039e-02
5.70834614e-02 -4.68105406e-01 -1.36059082e+00 7.21039712e-01
7.56905019e-01 -4.65155840e-01 8.76062334e-01 -3.90760526e-02
-4.59097356e-01 -3.30579758e-01 1.14001490e-01 -9.33988214e-01
1.13819994e-01 -3.38276237e-01 9.15676206e-02 1.52070498e+00
5.57533324e-01 -1.00015116e+00 1.57646149e-01 9.10468280e-01
-2.45732993e-01 -8.82263243e-01 -1.05267298e+00 -8.84649098e-01
5.48636079e-01 -4.07861233e-01 5.43534696e-01 -4.83236797e-02
4.47285652e-01 -6.59559071e-02 2.36330748e-01 -5.90754785e-02
5.90311050e-01 -3.96228015e-01 2.72841662e-01 -1.52043450e+00
5.47394872e-01 -7.78032839e-02 -3.81713778e-01 -3.89608502e-01
-1.66554093e-01 -7.15728939e-01 1.31297559e-01 -2.10424590e+00
9.22745541e-02 -7.18913674e-01 -4.13119406e-01 -4.90449339e-01
-4.06321958e-02 6.27043396e-02 1.66352764e-02 3.04517716e-01
1.64961159e-01 5.60182154e-01 7.39111841e-01 -2.38374457e-01
-3.57112557e-01 1.99702069e-01 -6.57305941e-02 1.14321813e-01
9.23959672e-01 -5.07337749e-02 -6.41203403e-01 2.42006332e-01
2.46331543e-01 5.04627451e-03 4.40129131e-01 -1.15422988e+00
1.42533854e-01 1.36445358e-01 5.05619705e-01 -1.31882787e+00
2.81354487e-01 -1.26665616e+00 8.30588162e-01 6.21330321e-01
3.18823218e-01 1.25487119e-01 6.64969027e-01 3.92893106e-01
1.76064327e-01 -1.09937525e+00 9.35376167e-01 3.57198358e-01
-4.85685319e-01 -2.61422545e-01 -6.25899613e-01 -3.74002129e-01
1.44740760e+00 -5.42145193e-01 -3.22088927e-01 -1.45630147e-02
-6.78824484e-01 5.51418602e-01 6.34724319e-01 1.90096825e-01
5.68180203e-01 -1.35370922e+00 -1.08244434e-01 8.08584094e-02
1.55077815e-01 -8.96512941e-02 2.97634810e-01 1.00984609e+00
-5.54716349e-01 6.29988432e-01 -5.19509278e-02 -8.00120234e-01
-1.14953792e+00 6.78835869e-01 1.07062638e+00 2.99018174e-01
-2.24535868e-01 2.65172571e-01 -7.57315099e-01 3.61790389e-01
1.99431241e-01 -3.01982552e-01 -6.69987500e-01 3.02698195e-01
3.11274379e-01 7.81033218e-01 6.20008707e-01 -7.25515425e-01
-5.27117372e-01 7.23318934e-01 5.95347643e-01 -1.54253542e-01
9.53289449e-01 -3.90251905e-01 -4.27617759e-01 1.18157053e+00
1.07734752e+00 -2.56411731e-01 -8.52550149e-01 4.63224411e-01
-1.01858422e-01 -2.76092201e-01 3.95210348e-02 -6.84841871e-01
-6.43719077e-01 6.49259448e-01 8.32588792e-01 3.55580688e-01
1.00284445e+00 -1.16480999e-01 3.19102824e-01 -1.64186247e-02
4.78025198e-01 -1.66221511e+00 -6.05228066e-01 2.00174764e-01
1.11558855e+00 -6.56507254e-01 4.21605110e-01 -3.68774891e-01
-4.06679630e-01 1.15303636e+00 5.96682131e-01 2.24643528e-01
8.97836685e-01 7.61166096e-01 -3.74787003e-01 8.59188437e-02
-4.13613945e-01 2.23328069e-01 6.62276894e-02 4.48576182e-01
3.02258909e-01 6.99625723e-03 -1.36622512e+00 7.15048611e-01
2.89688379e-01 1.04749016e-01 8.15706253e-01 1.27248776e+00
-5.82548201e-01 -1.09385788e+00 -6.52654946e-01 5.17371714e-01
-3.70139807e-01 4.26370859e-01 1.27628064e-02 7.85821915e-01
-1.69621170e-01 1.15232337e+00 3.82016391e-01 -5.58383018e-02
1.13223135e+00 2.07080945e-01 5.18233716e-01 2.74730027e-01
-3.54879737e-01 6.34897128e-02 3.48108351e-01 -1.16662152e-01
-2.46637002e-01 -5.55384934e-01 -1.92735374e+00 -1.99698076e-01
-9.13568974e-01 1.05336654e+00 1.63459551e+00 8.26395750e-01
-1.50647461e-01 6.86955512e-01 6.03982747e-01 -1.05630398e+00
-5.07341862e-01 -8.47453356e-01 -1.13485575e+00 2.16272834e-04
6.08624937e-03 -8.69197845e-01 -7.32310057e-01 -6.85049832e-01] | [6.289844512939453, 2.749812602996826] |
01b68053-4282-4f0a-a62a-1fde70221cfc | a-demand-driven-perspective-on-generative | 2307.04292 | null | https://arxiv.org/abs/2307.04292v1 | https://arxiv.org/pdf/2307.04292v1.pdf | A Demand-Driven Perspective on Generative Audio AI | To achieve successful deployment of AI research, it is crucial to understand the demands of the industry. In this paper, we present the results of a survey conducted with professional audio engineers, in order to determine research priorities and define various research tasks. We also summarize the current challenges in audio quality and controllability based on the survey. Our analysis emphasizes that the availability of datasets is currently the main bottleneck for achieving high-quality audio generation. Finally, we suggest potential solutions for some revealed issues with empirical evidence. | ['Ben Sangbae Chon', 'Keunwoo Choi', 'Hyeongi Moon', 'Minsung Kang', 'Sangshin Oh'] | 2023-07-10 | null | null | null | null | ['audio-generation'] | ['audio'] | [ 4.26745981e-01 -3.10111754e-02 -1.17552161e-01 -1.04381748e-01
-9.41222310e-01 -5.49244046e-01 8.92690942e-02 -2.03176737e-01
-1.43218726e-01 6.53828621e-01 4.93538052e-01 -2.38639116e-01
-5.01949787e-01 -2.20472857e-01 -3.97028565e-01 -1.43494725e-01
2.95641134e-03 9.53580812e-02 -1.11832224e-01 -2.90452570e-01
6.05208874e-01 4.00638700e-01 -1.95572793e+00 4.12011564e-01
7.31117845e-01 9.78094220e-01 2.55333841e-01 6.96563542e-01
3.11439514e-01 1.02997041e+00 -1.10897124e+00 -9.87424850e-02
8.12641233e-02 -4.85823065e-01 -7.24055290e-01 1.25819772e-01
3.55832040e-01 -4.89782810e-01 -5.98441437e-02 7.92048752e-01
7.74994791e-01 1.48137346e-01 4.44861025e-01 -1.46027327e+00
-4.29188460e-01 9.13567364e-01 -1.84223801e-02 5.01822293e-01
6.45849049e-01 2.82077342e-01 1.38620389e+00 -5.80662251e-01
4.39828247e-01 1.01815510e+00 3.18892151e-01 2.68391967e-01
-1.20733452e+00 -9.85232234e-01 1.92166388e-01 3.34379107e-01
-1.24111784e+00 -9.40037549e-01 8.72119427e-01 -5.72566450e-01
8.51377845e-01 4.24769431e-01 1.00958705e+00 1.01444292e+00
1.65140644e-01 5.52691281e-01 9.82933044e-01 -3.97073865e-01
2.65772611e-01 6.39613420e-02 -1.87520534e-01 3.48852463e-02
1.23053834e-01 1.63732171e-01 -1.04346657e+00 -1.79567069e-01
9.25299346e-01 -1.06039608e+00 -3.59489679e-01 9.55141033e-04
-1.05879021e+00 4.39649493e-01 -1.00455321e-01 6.89491868e-01
-4.15201962e-01 2.99369544e-01 1.29449099e-01 5.38702726e-01
7.52033368e-02 1.21972179e+00 -1.48090631e-01 -8.15903962e-01
-9.33536589e-01 5.96953273e-01 8.06518197e-01 6.52559340e-01
4.03938182e-02 6.93316221e-01 1.90863550e-01 8.53473961e-01
2.63100147e-01 2.33976126e-01 1.49888471e-01 -1.72634983e+00
2.24678308e-01 -7.18863457e-02 3.93162459e-01 -1.10686302e+00
-2.92966366e-01 -5.89744747e-01 -1.54407382e-01 -4.23014024e-03
3.91699880e-01 -3.38153392e-01 -3.84324491e-01 1.42293501e+00
-3.37245375e-01 -4.52275686e-02 -2.14108109e-01 9.52163517e-01
4.28822458e-01 6.77036643e-01 -4.22547422e-02 -5.04076183e-01
9.51388180e-01 -6.10405564e-01 -1.14277434e+00 -3.55972052e-01
7.05965012e-02 -1.08251154e+00 1.21207559e+00 1.00419283e+00
-1.40732920e+00 -5.36521018e-01 -1.04377842e+00 3.22345048e-01
3.63194108e-01 -7.30531290e-02 7.27234364e-01 7.14591682e-01
-9.35734332e-01 4.42400128e-01 -6.93616390e-01 -1.09299660e-01
-9.48355347e-02 2.57946998e-01 2.33397409e-01 2.02486157e-01
-9.88266289e-01 7.56295204e-01 5.42273037e-02 5.32936677e-02
-8.08862209e-01 -7.70219862e-01 -4.28367943e-01 -1.85262978e-01
3.22344691e-01 -2.22426683e-01 1.73044944e+00 -7.23598123e-01
-1.64337826e+00 1.57208756e-01 2.65353948e-01 -1.21986501e-01
1.09684400e-01 -2.88237631e-01 -6.57523215e-01 2.07958221e-01
1.41587272e-01 5.15139341e-01 6.44539058e-01 -1.19592881e+00
-1.08144891e+00 -6.65142834e-02 1.17735621e-02 1.72699884e-01
-3.45045537e-01 3.87739509e-01 1.00895287e-02 -7.57225335e-01
5.81101403e-02 -8.57478321e-01 -1.89794466e-01 -4.90404546e-01
8.91971216e-02 -1.37972489e-01 3.24217141e-01 -5.08125842e-01
1.57167149e+00 -2.35132217e+00 2.02595323e-01 1.33997530e-01
1.51573494e-01 -2.79540926e-01 -1.56465724e-01 6.66579902e-01
3.30540389e-02 1.92346796e-01 2.78327376e-01 1.90670684e-01
4.35229503e-02 -2.05554396e-01 -6.39434218e-01 1.58713832e-01
1.16304077e-01 3.91504526e-01 -8.04311097e-01 -3.16440254e-01
1.88009012e-02 2.87693411e-01 -7.20231831e-01 1.90542489e-01
4.51790635e-03 4.65478778e-01 -4.62379277e-01 8.62398326e-01
-5.07477745e-02 1.44953176e-01 1.44756526e-01 -1.66201442e-01
-5.95470071e-01 7.22001791e-01 -1.28915012e+00 1.45243025e+00
-3.60783160e-01 9.30337548e-01 5.04300952e-01 -8.77470553e-01
8.41325819e-01 6.82534039e-01 6.78109407e-01 -6.77048087e-01
2.84046173e-01 5.09482384e-01 6.04006350e-01 -5.14613867e-01
5.58690488e-01 -2.59861022e-01 -6.51292652e-02 6.20250583e-01
-7.59806558e-02 -9.52465653e-01 4.60578799e-01 -2.01599538e-01
8.68943512e-01 -2.57018864e-01 -1.15700200e-01 -4.97343689e-01
-2.64390893e-02 1.37640730e-01 6.31778181e-01 6.19021714e-01
-4.36922550e-01 4.19579387e-01 4.54199612e-01 -1.04626328e-01
-8.66634905e-01 -9.84988928e-01 -2.55475752e-02 1.08102632e+00
-2.33695671e-01 -5.77445149e-01 -6.78856373e-01 1.82762027e-01
-1.99430093e-01 9.03864145e-01 -6.47202507e-02 2.65996028e-02
-4.15161431e-01 -2.72008359e-01 4.97766763e-01 4.71481919e-01
-1.85862519e-02 -1.02545106e+00 -9.89839733e-01 4.86905426e-01
-6.25618458e-01 -1.02321148e+00 -1.69132277e-01 5.55254817e-02
-7.98237801e-01 -7.31735528e-01 -2.50454038e-01 -5.97074389e-01
1.99135151e-02 2.28794307e-01 8.59344244e-01 -2.29381576e-01
-3.17179352e-01 5.81457794e-01 -3.04313570e-01 -6.39297426e-01
-3.58759791e-01 1.12142637e-01 2.59691000e-01 -5.16994715e-01
1.02577344e-01 -9.28833485e-01 -2.60847032e-01 4.83260721e-01
-6.80379331e-01 -1.83338165e-01 6.19491220e-01 3.60017389e-01
3.37763309e-01 5.80527604e-01 1.03770506e+00 -2.26135850e-01
1.41233552e+00 -2.12406978e-01 -4.29702669e-01 -8.38008597e-02
-5.98165691e-01 -3.93453926e-01 2.80334532e-01 -6.75911427e-01
-7.77849257e-01 -1.74515426e-01 -4.37253676e-02 7.81550109e-02
-8.83157477e-02 6.63449287e-01 -2.59005763e-02 4.56061848e-02
8.42249036e-01 -2.08263204e-01 -4.39475738e-02 -3.14803630e-01
9.12607759e-02 9.85162675e-01 3.22417051e-01 -9.40091312e-01
2.81675100e-01 -3.19628827e-02 -3.29452842e-01 -1.26010227e+00
-5.90019166e-01 9.56011042e-02 -2.36882493e-01 -7.22354889e-01
4.09895867e-01 -9.17711914e-01 -4.18320537e-01 2.15342835e-01
-8.10020924e-01 -4.36113477e-01 -5.04675031e-01 8.30261409e-01
-8.15193832e-01 -2.49796331e-01 -7.46713519e-01 -1.05490077e+00
1.46778030e-02 -1.44408488e+00 6.75099492e-01 2.19506875e-01
-9.87856448e-01 -4.52108413e-01 -1.62772119e-01 7.89987743e-01
5.23847342e-01 -1.33666486e-01 7.01959193e-01 -1.36571571e-01
-5.21992207e-01 1.21428750e-01 2.84133255e-01 2.93795556e-01
2.40122110e-01 4.46284711e-01 -9.08243656e-01 -1.96517497e-01
2.36274496e-01 -5.92384040e-01 4.95455787e-02 6.86750352e-01
9.24081564e-01 3.96588957e-03 1.68239236e-01 1.53199181e-01
8.27115834e-01 6.64021611e-01 3.30583453e-01 3.22568893e-01
6.28599450e-02 9.08321142e-01 9.35771585e-01 7.21082628e-01
2.31898367e-01 3.78290206e-01 -8.95272847e-03 4.08328891e-01
-2.04710532e-02 -2.90441215e-01 2.24518180e-01 1.45719504e+00
-2.76732683e-01 -1.59112856e-01 -7.46623158e-01 6.03519976e-01
-1.37010491e+00 -7.99477398e-01 1.02220051e-01 1.84498131e+00
7.26203322e-01 4.60710645e-01 4.57191378e-01 7.88434803e-01
5.60067296e-01 -6.68656603e-02 -2.22014114e-01 -5.01553833e-01
1.52410448e-01 2.51052737e-01 6.42017126e-02 4.01740491e-01
-5.93366146e-01 7.63771355e-01 9.06476784e+00 4.34123248e-01
-9.54054832e-01 -3.70022416e-01 2.12838337e-01 -3.27862352e-01
-4.83040094e-01 -3.23956124e-02 -3.84354353e-01 1.81570783e-01
1.15723598e+00 -7.12982476e-01 7.80929863e-01 5.39770842e-01
6.55215859e-01 -1.78291321e-01 -1.05353475e+00 8.74458909e-01
1.01773459e-02 -8.96045327e-01 -2.39074573e-01 8.88793021e-02
4.85783905e-01 -5.04037976e-01 4.80694115e-01 -3.52428965e-02
7.97150731e-02 -9.82389271e-01 1.09256136e+00 2.41153792e-01
5.45702517e-01 -8.13836396e-01 1.52947292e-01 1.09447323e-01
-1.03019655e+00 -5.11341572e-01 -8.31793249e-02 -8.45513880e-01
3.91091824e-01 4.83565062e-01 -6.48403227e-01 7.36751929e-02
8.00285280e-01 4.50576782e-01 -1.00742564e-01 1.02468967e+00
-2.30121344e-01 1.02029812e+00 -3.82663250e-01 -1.49460942e-01
-9.06773061e-02 2.89701149e-02 5.39570332e-01 8.78607213e-01
3.93711329e-01 3.57229054e-01 1.13519557e-01 8.80047679e-01
3.26291174e-01 1.25939891e-01 -6.24739051e-01 -7.08251774e-01
8.49194825e-01 8.42357874e-01 -5.63699305e-01 2.22372770e-01
-1.27797127e-01 2.68195987e-01 -2.36036241e-01 2.55829692e-01
-5.00087678e-01 -6.28604352e-01 8.63049924e-01 1.97080195e-01
-1.99941754e-01 -5.75056911e-01 -3.61671150e-01 -7.47490525e-01
-1.25812069e-01 -1.47530055e+00 6.27045184e-02 -9.19255733e-01
-9.02370811e-01 2.76400298e-01 1.54809147e-01 -9.92614448e-01
-5.35744011e-01 -2.70642996e-01 -8.32823515e-02 5.26894510e-01
-8.36866200e-01 -5.04481018e-01 4.37336266e-02 1.11478768e-01
7.29476571e-01 -1.36800915e-01 5.89684904e-01 5.71796238e-01
-4.54709560e-01 3.82751614e-01 -4.39542949e-01 -4.38284338e-01
6.77871048e-01 -8.23428452e-01 2.53706843e-01 5.21509349e-01
1.94315255e-01 6.29381418e-01 1.20746136e+00 -5.66314220e-01
-1.62284040e+00 -2.32465908e-01 7.75060296e-01 -2.80607522e-01
1.00503552e+00 -5.55456579e-02 -4.22681332e-01 6.24826670e-01
2.35044703e-01 -9.28230464e-01 9.52843606e-01 2.19687656e-01
1.93464145e-01 -2.90232897e-01 -8.69668007e-01 6.86784327e-01
1.08212936e+00 -6.26923203e-01 -7.41925120e-01 -7.94326216e-02
7.68262506e-01 -3.37457329e-01 -1.00723994e+00 3.11037600e-01
9.42568183e-01 -9.44607675e-01 7.37436354e-01 -3.58553320e-01
4.92033035e-01 -1.63409069e-01 -3.52170706e-01 -1.40635681e+00
-4.65441018e-01 -8.51548731e-01 2.85144180e-01 1.27392685e+00
3.90902638e-01 -2.53147095e-01 6.37754560e-01 8.61805201e-01
-2.90428400e-01 -3.47447336e-01 -7.30069160e-01 -6.20324373e-01
-8.89820084e-02 -1.00852025e+00 4.16298687e-01 6.50434434e-01
4.79288667e-01 4.85822767e-01 -3.44684839e-01 -1.17767051e-01
5.74924290e-01 -5.50186597e-02 6.16574466e-01 -1.20822978e+00
-2.77950376e-01 -5.47211111e-01 -4.11235541e-01 -8.54252756e-01
-3.93373556e-02 -1.77736804e-01 1.00321151e-01 -1.45825064e+00
-2.17603877e-01 -3.12505662e-01 -3.43936652e-01 2.01317971e-03
3.40346396e-01 2.53230423e-01 4.36428219e-01 -1.08270548e-01
-1.25494108e-01 2.12990478e-01 1.22497725e+00 -3.49132456e-02
-4.54857022e-01 1.98348030e-01 -1.24954295e+00 7.06000090e-01
1.01358271e+00 -1.74764350e-01 -7.81524003e-01 -6.40134096e-01
3.15239251e-01 2.23422796e-01 -1.54425412e-01 -1.33071470e+00
2.12505266e-01 -5.86272061e-01 -2.98528690e-02 -5.74691594e-01
6.15340412e-01 -9.36878979e-01 4.52262133e-01 2.88541704e-01
-6.62020743e-01 3.18200737e-01 2.41152316e-01 7.08264112e-02
-3.30173343e-01 -2.02003062e-01 6.01644099e-01 1.60373971e-01
-4.34844553e-01 -5.67441463e-01 -9.76786017e-01 2.47071549e-01
7.84506381e-01 2.45383959e-02 3.49822380e-02 -7.38147259e-01
-5.43535292e-01 2.42050737e-01 1.09443910e-01 5.11541426e-01
6.03264928e-01 -1.25737154e+00 -5.50700784e-01 4.72818390e-02
-1.56144962e-01 -4.93032873e-01 1.31977826e-01 4.36607599e-01
-4.23034042e-01 5.73266685e-01 -5.11269689e-01 -2.90419251e-01
-1.15880227e+00 8.49445909e-03 5.84740490e-02 3.53944063e-01
-2.74446219e-01 8.05043221e-01 -3.20502847e-01 2.13061854e-01
5.99535465e-01 -3.82952601e-01 -1.22000426e-01 9.04735103e-02
6.79079592e-01 6.63441658e-01 4.28985395e-02 -3.46135169e-01
-1.74692795e-01 3.34517747e-01 3.60378712e-01 -8.77523780e-01
1.30285943e+00 -2.89617684e-02 -7.12537766e-03 1.08581376e+00
4.68873709e-01 1.96481243e-01 -8.68189514e-01 3.90718848e-01
1.99432224e-02 -8.00779760e-01 3.22606921e-01 -7.17471242e-01
-9.07702744e-01 7.46091366e-01 3.40189010e-01 3.81293267e-01
1.25531197e+00 -2.16862619e-01 5.34224331e-01 2.32381627e-01
5.25440395e-01 -1.63723826e+00 2.83719122e-01 1.95896521e-01
1.04388249e+00 -4.02220398e-01 3.13259475e-02 -4.85379577e-01
-7.92535245e-01 9.26887155e-01 6.37977421e-01 2.21062303e-01
7.71684527e-01 5.94070792e-01 1.99058264e-01 -2.24998523e-03
-1.05879664e+00 4.26103137e-02 8.30710605e-02 7.28508651e-01
8.85680556e-01 8.31656680e-02 -5.79522610e-01 6.68337822e-01
-7.60821223e-01 3.38400334e-01 7.42941380e-01 1.09914088e+00
-8.15638244e-01 -1.28147030e+00 -4.78903413e-01 5.60009062e-01
-6.75623059e-01 1.73093706e-01 -5.75854003e-01 5.59232950e-01
-6.60179481e-02 1.66565835e+00 3.72117423e-02 -7.36496866e-01
7.51823604e-01 -2.56025176e-02 5.65187514e-01 -5.30114472e-01
-4.99999911e-01 5.24186254e-01 5.35317719e-01 -4.55502987e-01
-3.64708245e-01 -8.70716929e-01 -8.47087383e-01 -2.60285944e-01
-5.24153650e-01 3.82503182e-01 7.46392965e-01 5.60293198e-01
3.05999070e-01 8.02244902e-01 5.51946998e-01 -4.78403389e-01
-6.61085248e-01 -9.22379434e-01 -9.24656749e-01 -1.22679010e-01
9.65634808e-02 -7.36386418e-01 -5.30659199e-01 1.71330005e-01] | [15.47815990447998, 5.804722309112549] |
0790d9e7-d530-4edc-be41-28b061e0e29a | top-down-rst-parsing-utilizing-granularity | null | null | https://doi.org/10.1609/aaai.v34i05.6321 | https://ojs.aaai.org/index.php/AAAI/article/view/6321/6177 | Top-Down RST Parsing Utilizing Granularity Levels in Documents | Some downstream NLP tasks exploit discourse dependency trees converted from RST trees. To obtain better discourse dependency trees, we need to improve the accuracy of RST trees at the upper parts of the structures. Thus, we propose a novel neural top-down RST parsing method. Then, we exploit three levels of granularity in a document, paragraphs, sentences and Elementary Discourse Units (EDUs), to parse a document accurately and efficiently. The parsing is done in a top-down manner for each granularity level, by recursively splitting a larger text span into two smaller ones while predicting nuclearity and relation labels for the divided spans. The results on the RST-DT corpus show that our method achieved the state-of-the-art results, 87.0 unlabeled span score, 74.6 nuclearity labeled span score, and the comparable result with the state-of-the-art, 60.0 relation labeled span score. Furthermore, discourse dependency trees converted from our RST trees also achieved the state-of-the-art results, 64.9 unlabeled attachment score and 48.5 labeled attachment score. | ['Masaaki Nagata', 'Manabu Okumura', 'Hidetaka Kamigaito', 'Tsutomu Hirao', 'Naoki Kobayashi'] | 2020-04-03 | null | null | null | null | ['discourse-parsing'] | ['natural-language-processing'] | [-4.83788177e-02 9.86795306e-01 -5.87586939e-01 -3.37305725e-01
-1.01469922e+00 -6.40596807e-01 4.45131063e-01 3.86174530e-01
-1.16189957e-01 1.22593522e+00 8.25328112e-01 -4.43710268e-01
3.11501473e-01 -9.76216674e-01 -5.03138363e-01 -4.79813099e-01
-1.33379400e-01 6.85059130e-01 4.90795642e-01 -3.01530480e-01
2.61666864e-01 1.10718317e-01 -1.12261605e+00 8.63804579e-01
9.66448724e-01 7.04583645e-01 3.99657972e-02 6.73007190e-01
-6.05587661e-01 1.33238339e+00 -8.18008304e-01 -3.60147536e-01
-3.58953565e-01 -6.08229339e-01 -1.37259173e+00 -1.60927400e-01
-5.91812655e-02 -3.03252012e-01 -1.35808930e-01 9.66179371e-01
3.60936671e-02 2.35659257e-01 8.75502646e-01 -4.64557827e-01
-7.31736004e-01 1.43272042e+00 -7.38197446e-01 3.29798281e-01
3.07752818e-01 -8.28446329e-01 1.27495956e+00 -6.99318469e-01
8.59569609e-01 1.54390419e+00 3.53279203e-01 5.95305085e-01
-9.66627240e-01 -4.58975196e-01 4.98413026e-01 5.32022864e-02
-6.52862966e-01 -3.81364405e-01 6.79274619e-01 -4.24177676e-01
1.29046977e+00 6.78807124e-02 1.79777399e-01 8.38166118e-01
1.26810461e-01 7.82699108e-01 9.36699212e-01 -9.12586033e-01
1.50175020e-01 -2.58165389e-01 8.69984508e-01 7.93072462e-01
1.26491502e-01 -4.37137008e-01 -3.18338603e-01 1.87509999e-01
5.49749851e-01 -6.22197509e-01 -7.14475960e-02 5.49340367e-01
-8.83389294e-01 9.35578525e-01 4.40034121e-01 8.30237210e-01
-2.46271074e-01 -2.91218281e-01 6.49903536e-01 1.70426920e-01
5.65662384e-01 2.59298354e-01 -4.57364321e-01 -1.50840476e-01
-6.07139885e-01 -5.47204874e-02 1.07144403e+00 9.38957334e-01
4.83208150e-01 -1.40037149e-01 -5.32583237e-01 8.87858331e-01
2.85855502e-01 9.85091403e-02 5.79735279e-01 -1.20650649e+00
1.09669423e+00 8.54416490e-01 -9.24307555e-02 -7.09719121e-01
-5.79417229e-01 -3.50078046e-02 -1.03228736e+00 -2.54380047e-01
5.36697924e-01 -4.92291957e-01 -8.95974696e-01 1.91366744e+00
4.21239436e-01 -4.71254319e-01 5.90945423e-01 5.55538177e-01
1.16989803e+00 1.16868806e+00 2.13226005e-01 -6.89927995e-01
1.80794799e+00 -1.32318830e+00 -1.06465554e+00 -3.04767519e-01
1.10083854e+00 -6.22290313e-01 8.40444863e-01 1.83841810e-01
-1.21944761e+00 -5.51541507e-01 -1.10564506e+00 -4.32111919e-01
3.61063741e-02 3.25813264e-01 4.14755464e-01 3.11477929e-01
-7.76480675e-01 6.77452922e-01 -1.00295472e+00 -9.99906883e-02
2.33637810e-01 1.30684674e-01 -2.17843696e-01 2.83011466e-01
-1.39253831e+00 9.63309288e-01 8.61843169e-01 -1.78212658e-01
-3.19233060e-01 -2.18323916e-01 -9.11635876e-01 3.46009642e-01
3.82822990e-01 -3.37912560e-01 1.59501886e+00 -6.04939342e-01
-1.71907008e+00 9.61524904e-01 -2.99374849e-01 -3.53567809e-01
3.84215564e-02 -4.36158180e-01 -1.95229836e-02 1.97605327e-01
3.50461334e-01 4.96676624e-01 -2.81740632e-02 -1.10740852e+00
-8.61863494e-01 -5.19414961e-01 5.25618553e-01 4.39486980e-01
-2.65302122e-01 2.19385564e-01 -2.10077673e-01 -5.55818737e-01
3.88825357e-01 -7.37401187e-01 -1.00119123e-02 -8.28544796e-01
-6.76957071e-01 -9.35523391e-01 4.92490947e-01 -1.08020258e+00
1.70385051e+00 -1.85360837e+00 3.90857548e-01 -4.83552188e-01
1.98520645e-01 2.47961342e-01 1.34672433e-01 5.64364374e-01
-1.32041037e-01 2.52030104e-01 -1.59829602e-01 -3.90012920e-01
-1.71662584e-01 3.32205772e-01 -4.38866317e-02 -9.21324566e-02
2.72690624e-01 6.48604989e-01 -8.69210064e-01 -8.75615716e-01
-2.03117937e-01 -1.21107481e-01 -3.09757084e-01 4.60347921e-01
-4.45454866e-01 3.98465395e-01 -6.44297123e-01 5.09235978e-01
1.97117209e-01 -3.30502093e-01 7.49114156e-01 5.97906411e-02
-2.18065903e-01 8.48014116e-01 -6.93118870e-01 1.65856302e+00
-2.95463592e-01 6.08108342e-01 7.55152628e-02 -1.04208374e+00
1.19409239e+00 6.00269914e-01 -3.37708555e-02 -3.77957135e-01
4.43753213e-01 2.07227096e-01 2.44867012e-01 -6.90466821e-01
6.25304878e-01 -3.46344560e-01 -5.52500248e-01 4.20847297e-01
-8.62638727e-02 3.25987935e-01 6.65937603e-01 8.93172547e-02
1.03866255e+00 8.63893926e-02 6.34912729e-01 -3.13660562e-01
8.37245345e-01 1.29162669e-01 7.39632308e-01 1.68440834e-01
2.25739982e-02 4.04128283e-01 1.14126050e+00 -2.07279503e-01
-9.89713073e-01 -7.16130316e-01 -5.81420287e-02 1.35692418e+00
-2.33567148e-01 -4.34799463e-01 -1.23824310e+00 -8.77759218e-01
-6.15595043e-01 8.46096754e-01 -6.87096238e-01 3.78119797e-01
-1.49279237e+00 -6.25948250e-01 7.06870079e-01 9.64862943e-01
7.75498450e-01 -1.33495593e+00 -2.65306890e-01 4.65379894e-01
-6.52907491e-01 -1.15509760e+00 1.55006163e-02 2.06076443e-01
-1.00979710e+00 -9.36864913e-01 -5.42842507e-01 -1.33928239e+00
4.57814872e-01 -2.78934747e-01 9.85317528e-01 6.29502982e-02
7.02764571e-01 -7.96721697e-01 -7.73997188e-01 -1.26752704e-01
-7.40256786e-01 6.71316385e-01 -2.85816789e-01 -6.17790937e-01
3.19386482e-01 -5.36835134e-01 -6.82423785e-02 -8.98084864e-02
-3.82671982e-01 1.03638165e-01 4.25104260e-01 9.80738997e-01
3.66570383e-01 -2.61922479e-01 8.80541027e-01 -1.15109837e+00
6.39635205e-01 -3.01795512e-01 -2.67641902e-01 4.19407457e-01
-2.52859533e-01 1.43677518e-01 8.50256026e-01 -1.86100215e-01
-1.55455601e+00 -3.65364939e-01 -4.12818044e-01 2.68146008e-01
-2.55942911e-01 7.34145522e-01 -3.81812453e-01 9.87538934e-01
6.88403785e-01 -4.29226905e-01 -4.87055093e-01 -5.52577257e-01
4.70092565e-01 9.38774288e-01 5.92242479e-01 -8.78304124e-01
-5.72203484e-04 -3.45530473e-02 -1.11657336e-01 -6.97277665e-01
-1.49213612e+00 -1.50350332e-01 -9.29211974e-01 2.21802995e-01
1.28349924e+00 -6.92728698e-01 -5.28539479e-01 2.00498700e-01
-1.84643781e+00 -2.75060862e-01 -1.95284620e-01 4.02331144e-01
-3.33555877e-01 5.36167622e-01 -1.45299435e+00 -7.26356149e-01
-6.56198204e-01 -8.28251481e-01 8.00050497e-01 5.22158384e-01
-5.62876344e-01 -8.95771742e-01 3.65415484e-01 4.68670070e-01
-3.41796130e-01 4.97837037e-01 1.40319133e+00 -9.72428083e-01
-7.25935586e-03 1.61117807e-01 -4.65426207e-01 2.57213145e-01
2.13865846e-01 -2.02508673e-01 -8.94852042e-01 1.18664354e-01
1.02059759e-01 -4.03860629e-01 7.17209220e-01 2.16981679e-01
5.05463660e-01 -3.57636899e-01 -5.28413177e-01 1.00369435e-02
9.41599309e-01 4.99502301e-01 6.07963204e-01 4.88646060e-01
6.01027906e-01 9.78715003e-01 8.09966981e-01 -2.24194340e-02
6.18172646e-01 3.58379245e-01 2.72034723e-02 2.01413155e-01
-2.82738566e-01 -1.82008281e-01 3.80746007e-01 1.41883910e+00
-3.82864982e-01 -5.45133770e-01 -1.07359695e+00 6.31133139e-01
-2.12536955e+00 -7.74969339e-01 -4.96950060e-01 1.48867309e+00
1.18110621e+00 5.29291332e-01 1.09112831e-02 1.59604207e-01
1.13805890e+00 3.10181350e-01 -1.96234271e-01 -7.98478782e-01
5.88591322e-02 3.90669964e-02 -1.78293303e-01 6.94910824e-01
-1.10555947e+00 1.28059006e+00 6.05134583e+00 6.61767006e-01
-6.32302999e-01 8.93515646e-02 8.13049734e-01 2.63506442e-01
3.20864469e-02 -6.50227070e-02 -1.18054724e+00 2.17233449e-01
1.25724173e+00 -4.42392789e-02 -1.48136750e-01 8.20341527e-01
-1.13784254e-01 -6.69060200e-02 -1.13903773e+00 3.88576269e-01
-4.79417704e-02 -1.36016655e+00 -1.34866685e-01 -7.86331892e-02
7.70378351e-01 -1.69159368e-01 -7.04789102e-01 6.91448450e-01
5.47830105e-01 -8.51756811e-01 4.59647536e-01 5.88826835e-02
7.44246960e-01 -6.70685232e-01 1.05241585e+00 6.58765614e-01
-1.27037501e+00 9.29862335e-02 -3.27078730e-01 -4.89529222e-01
5.71683168e-01 4.79364455e-01 -8.07594895e-01 7.51723468e-01
4.14598256e-01 3.94436836e-01 3.03597152e-02 8.35175142e-02
-8.92620087e-01 6.99084640e-01 -1.17829911e-01 -3.04642200e-01
4.53553140e-01 -1.99765742e-01 3.56331319e-01 1.47892404e+00
1.90081056e-02 7.40293205e-01 2.67331213e-01 3.41808617e-01
-4.40105259e-01 9.78554115e-02 -2.70934731e-01 6.09799996e-02
7.41318524e-01 1.03852463e+00 -9.37519789e-01 -7.44982839e-01
-2.00187683e-01 6.28556550e-01 9.38207984e-01 2.11467352e-02
-7.60363817e-01 -6.10541701e-01 8.80858824e-02 -2.42502227e-01
4.10885602e-01 -1.78150699e-01 -4.42038089e-01 -9.67153132e-01
1.00689895e-01 -5.62145412e-01 6.13775730e-01 -3.77967745e-01
-1.19875073e+00 1.03109252e+00 1.82291433e-01 -7.52037764e-01
-4.96608824e-01 -5.70306838e-01 -7.58022130e-01 6.60438597e-01
-1.34373617e+00 -1.01844323e+00 -7.01015117e-03 -1.03227511e-01
9.80906188e-01 8.57866034e-02 1.10596240e+00 -3.86221297e-02
-8.05349767e-01 3.25748086e-01 5.15201800e-02 7.13943958e-01
4.90850359e-01 -1.41208696e+00 4.79139864e-01 7.08897650e-01
-3.14290822e-01 4.03616250e-01 3.41329455e-01 -7.86386490e-01
-3.96595776e-01 -7.12446272e-01 1.57177985e+00 -2.63938099e-01
8.52514982e-01 -1.17739178e-01 -1.11272824e+00 9.02490377e-01
6.62732661e-01 -5.84455073e-01 6.28619313e-01 5.41453660e-01
-1.50765166e-01 2.72433072e-01 -9.84826684e-01 4.23625529e-01
1.04170060e+00 -2.77830660e-01 -1.52870750e+00 2.24678487e-01
1.20227206e+00 -7.51655102e-01 -1.18086970e+00 4.49170291e-01
3.82759035e-01 -8.50164652e-01 5.37178397e-01 -6.99567199e-01
1.04464948e+00 3.30499969e-02 -3.21953565e-01 -9.71613407e-01
-6.03600562e-01 -3.42671931e-01 -4.46156442e-01 1.69881618e+00
7.87960410e-01 -3.03934872e-01 9.46024597e-01 5.60797870e-01
-5.92711151e-01 -9.03266788e-01 -1.00326526e+00 -5.51653028e-01
6.21547341e-01 1.90228462e-01 4.76336002e-01 1.05484843e+00
6.68853402e-01 1.30495310e+00 1.54267490e-01 -5.60782254e-02
1.72077879e-01 3.70070577e-01 2.77096152e-01 -1.28294134e+00
-2.58195847e-01 -4.18633103e-01 3.13477516e-01 -1.33517134e+00
5.02113163e-01 -7.36601770e-01 1.34480432e-01 -1.90800869e+00
1.05681285e-01 -2.40901202e-01 -2.79292073e-02 4.81046975e-01
-4.13074344e-01 -2.01322302e-01 2.63818115e-01 4.70407575e-01
-4.83651817e-01 4.66731638e-01 1.32235038e+00 -6.75043166e-02
-4.59237754e-01 -2.96520174e-01 -5.89368522e-01 1.17222500e+00
1.06568313e+00 -6.40535116e-01 -1.48206741e-01 -5.31948447e-01
-9.98799354e-02 5.98148406e-01 -5.86055636e-01 -5.99708498e-01
2.08358213e-01 -1.91102058e-01 1.39521092e-01 -1.04934990e+00
3.65625054e-01 -1.15461722e-01 -5.12030542e-01 5.07113934e-01
-3.93544316e-01 8.69556442e-02 8.99670795e-02 1.21121958e-01
-3.52630049e-01 -7.50842869e-01 5.15163898e-01 -3.26691091e-01
-2.82738149e-01 -4.66591179e-01 -4.40672457e-01 3.28228116e-01
9.73968148e-01 2.69476883e-02 -8.01166534e-01 1.76871628e-01
-1.05412364e+00 1.96292311e-01 -2.83402186e-02 1.77487791e-01
1.24988034e-01 -1.16076243e+00 -8.29041898e-01 -2.83045828e-01
-4.12873328e-01 6.57113492e-01 -8.49976316e-02 4.46633071e-01
-5.41187942e-01 6.76864088e-01 -1.01917095e-01 -3.67069989e-01
-1.40804815e+00 3.37978929e-01 -1.42769635e-01 -1.14062142e+00
-6.08708441e-01 8.07792783e-01 3.51551414e-01 -3.16357344e-01
1.71106189e-01 -5.81607103e-01 -9.59000766e-01 4.61711973e-01
6.12979114e-01 4.81303215e-01 -1.27332229e-02 -4.83241022e-01
-1.69361800e-01 5.46664953e-01 -3.71739328e-01 -2.39332736e-01
1.28268087e+00 -2.01826379e-01 -4.51628804e-01 6.39999449e-01
9.26216424e-01 3.62476148e-02 -8.72521818e-01 -9.43025351e-02
4.15128231e-01 1.45078465e-01 -2.45359734e-01 -6.49952650e-01
-4.12061065e-01 8.25768232e-01 -2.59248734e-01 6.41431451e-01
8.02428424e-01 2.74389237e-01 9.52065289e-01 6.93136156e-01
4.70340513e-02 -1.05943704e+00 6.54969364e-02 1.12604368e+00
8.09013844e-01 -9.15200770e-01 3.57021540e-02 -9.05236900e-01
-5.81578732e-01 1.21145451e+00 8.25499773e-01 -2.50970602e-01
3.38596731e-01 3.61696333e-01 -1.13252383e-02 -1.53627843e-01
-7.81247258e-01 6.74717650e-02 7.46117756e-02 4.48372364e-01
1.08710647e+00 1.58642292e-01 -6.47167087e-01 1.02515376e+00
-4.94757771e-01 -3.71348321e-01 5.28496027e-01 9.47016776e-01
-9.15937543e-01 -1.40139782e+00 -4.17285204e-01 1.18387096e-01
-7.04378009e-01 -4.80456986e-02 -5.50243497e-01 9.34723735e-01
-1.57756478e-01 1.12792218e+00 9.30087045e-02 -9.30696726e-02
4.16127890e-01 2.33225971e-01 4.66848195e-01 -1.03957105e+00
-7.13476539e-01 1.83122545e-01 1.11505020e+00 -2.43243048e-04
-5.80694497e-01 -5.62732816e-01 -2.03386879e+00 -3.80794555e-01
-5.59747219e-01 4.56422150e-01 1.40426099e-01 1.14359212e+00
3.70683707e-02 8.35335612e-01 4.45408881e-01 -8.28014314e-01
-4.58963722e-01 -1.44556797e+00 -2.23105654e-01 7.35289827e-02
1.05255574e-01 -4.61123139e-01 -2.69586071e-02 1.42064705e-01] | [10.735918998718262, 9.417052268981934] |
f8c83014-4fff-4d32-94df-57eeaea39aac | valuation-of-public-bus-electrification-with | 2209.12107 | null | https://arxiv.org/abs/2209.12107v1 | https://arxiv.org/pdf/2209.12107v1.pdf | Valuation of Public Bus Electrification with Open Data | This research provides a novel framework to estimate the economic, environmental, and social values of electrifying public transit buses, for cities across the world, based on open-source data. Electric buses are a compelling candidate to replace diesel buses for the environmental and social benefits. However, the state-of-art models to evaluate the value of bus electrification are limited in applicability because they require granular and bespoke data on bus operation that can be difficult to procure. Our valuation tool uses General Transit Feed Specification, a standard data format used by transit agencies worldwide, to provide high-level guidance on developing a prioritization strategy for electrifying a bus fleet. We develop physics-informed machine learning models to evaluate the energy consumption, the carbon emissions, the health impacts, and the total cost of ownership for each transit route. We demonstrate the scalability of our tool with a case study of the bus lines in the Greater Boston and Milan metropolitan areas. | ['Carlo Papa', 'Erika Mellekas', 'Christian Zulberti', 'Luigi Lanuzza', 'Giuseppe Ferrara', 'Sergio Gambacorta', 'David Rodriguez', 'Akshat Jain', 'Scott J. Moura', 'Soomin Woo', 'Upadhi Vijay'] | 2022-09-25 | null | null | null | null | ['physics-informed-machine-learning'] | ['graphs'] | [-8.56066525e-01 -7.28928521e-02 -4.80618477e-01 -9.14007351e-02
-1.11121500e+00 -3.71268749e-01 4.37753201e-01 3.40206057e-01
-2.18976930e-01 1.13921928e+00 3.74944836e-01 -7.28327751e-01
-4.32540178e-01 -1.44515932e+00 -3.44747692e-01 -6.27807558e-01
-2.23667268e-02 5.70536554e-01 -7.71172047e-02 -3.11507940e-01
-3.16319279e-02 5.38485050e-01 -1.14866364e+00 -2.49854311e-01
9.68380451e-01 6.84397459e-01 9.10664722e-03 1.64900646e-01
1.10096164e-01 3.76710743e-01 -3.53810072e-01 -1.68388069e-01
1.67406902e-01 8.12521353e-02 -6.72013998e-01 -5.13529778e-01
-3.66430432e-01 -6.28857613e-01 -4.63409513e-01 6.15110397e-01
2.50463605e-01 2.88757056e-01 7.13810444e-01 -1.91693819e+00
1.44754693e-01 9.07786846e-01 -9.49496254e-02 2.37637267e-01
1.42959476e-01 4.79864836e-01 1.14526260e+00 3.94150950e-02
1.59630224e-01 8.16738307e-01 4.53035355e-01 -2.57986546e-01
-1.05679548e+00 -6.00380480e-01 3.15598585e-02 4.23190951e-01
-1.55762923e+00 -6.85821697e-02 5.47590256e-01 -5.22517383e-01
1.65559196e+00 8.64012420e-01 1.23357713e+00 5.10884225e-01
5.19127071e-01 4.34735805e-01 1.10398293e+00 -5.33562712e-02
3.43007088e-01 1.17176555e-01 1.35095656e-01 1.68791249e-01
4.67885345e-01 4.25113797e-01 -3.16838920e-02 -5.48270205e-03
-1.95554599e-01 -1.80948436e-01 2.98415929e-01 5.22377901e-02
-8.21359098e-01 6.37844265e-01 7.28925526e-01 1.34246334e-01
-3.67883444e-01 5.12550950e-01 2.40456238e-01 -7.55484728e-03
5.92380285e-01 1.81938946e-01 -4.49097842e-01 -3.02726150e-01
-9.99457359e-01 3.94385219e-01 6.21944845e-01 8.15542221e-01
6.63182795e-01 1.29512563e-01 8.87124166e-02 3.47220272e-01
7.64023304e-01 7.51886368e-01 -2.79192001e-01 -7.70840824e-01
6.62893593e-01 2.91478723e-01 6.00442946e-01 -4.32244927e-01
-8.06579769e-01 -2.42840648e-01 -1.71390235e-01 -2.82115843e-02
3.23547840e-01 -4.30761576e-01 -4.56744105e-01 1.07146847e+00
1.48121640e-01 -1.82951495e-01 -3.15550357e-01 4.48108435e-01
5.67115605e-01 9.89303112e-01 5.73631585e-01 4.20766026e-01
9.81252730e-01 -9.05927241e-01 -3.59783322e-01 2.10452422e-01
8.44253242e-01 2.24689487e-02 7.71843016e-01 1.82869688e-01
-1.29021275e+00 3.31099220e-02 -7.81468391e-01 1.26765803e-01
-1.02394378e+00 -4.61035758e-01 7.40754783e-01 1.10664833e+00
-5.94500124e-01 3.41810942e-01 -1.10929608e+00 -2.19196483e-01
2.96245962e-01 5.01685552e-02 3.18443120e-01 1.25077352e-01
-1.26593256e+00 1.51908767e+00 3.33585173e-01 -3.09835047e-01
-1.10807371e+00 -1.15078497e+00 -7.79247463e-01 6.71023548e-01
-5.26079349e-02 -2.99909323e-01 1.16938043e+00 7.36749992e-02
-8.59972715e-01 4.41255681e-02 5.22318125e-01 -4.18647379e-01
5.47746480e-01 5.24886370e-01 -1.00134122e+00 -2.88979679e-01
1.09730929e-01 5.31296909e-01 -2.31591538e-01 -9.35019493e-01
-1.10179174e+00 1.92980934e-02 3.51984262e-01 -1.10205509e-01
-2.34850958e-01 -1.46393433e-01 1.58777550e-01 -1.03362031e-01
-7.21630275e-01 -9.30628479e-01 -3.50376278e-01 -6.96334839e-01
-3.35175008e-01 -4.08261418e-01 4.16341931e-01 -1.00205815e+00
1.43280423e+00 -1.85716915e+00 -5.19633293e-01 6.92234039e-01
-4.64288682e-01 -4.10249144e-01 8.98112834e-04 7.07723439e-01
1.19204372e-01 4.58603352e-01 8.33138451e-02 2.38848969e-01
5.95036447e-01 1.20372154e-01 7.65901357e-02 3.52817416e-01
1.26038522e-01 7.74915218e-01 -1.11326420e+00 4.03058119e-02
8.11898470e-01 2.57010937e-01 -3.26420397e-01 -4.62963253e-01
-2.45315835e-01 3.50153476e-01 -4.43314403e-01 5.21701276e-01
6.76172078e-01 2.17334554e-01 4.92088348e-02 -1.30900159e-01
-6.67329252e-01 9.40021336e-01 -9.72944319e-01 1.12898350e+00
-9.87375915e-01 7.44287968e-01 -2.32460424e-01 -6.46734297e-01
4.50055122e-01 1.84799004e-02 9.45444763e-01 -9.66545761e-01
2.81233471e-02 2.78314710e-01 8.04850832e-02 -5.47027886e-01
5.79938889e-01 1.12119270e-02 -5.39264917e-01 7.04208136e-01
-4.47895676e-01 -6.45821512e-01 6.85526550e-01 -9.35353637e-02
1.03446639e+00 -1.44690350e-02 -8.50860849e-02 -6.35242760e-01
1.05704568e-01 4.25511926e-01 2.15423003e-01 -3.43513101e-01
6.85254708e-02 -2.96496868e-01 3.76744002e-01 -2.22639799e-01
-1.19248843e+00 -1.15912282e+00 -4.48601216e-01 6.44981146e-01
2.39047557e-02 -1.13475651e-01 -7.88643301e-01 -4.00725842e-01
2.75040060e-01 1.73264921e+00 1.01892255e-01 -2.44387761e-02
-1.93849459e-01 -1.04318857e+00 2.83730626e-01 4.90192264e-01
5.13503015e-01 3.96437617e-03 -7.31769264e-01 2.91937590e-01
-5.44622958e-01 -7.67670631e-01 -3.01223576e-01 -1.57702401e-01
-3.66036385e-01 -1.00029564e+00 -1.76141813e-01 -2.40056049e-02
5.80436170e-01 1.52557239e-01 1.05993950e+00 2.44630966e-02
-5.21286391e-02 4.09018308e-01 2.36337170e-01 -6.76115513e-01
-2.72038102e-01 3.80744666e-01 -4.21412975e-01 -5.86283088e-01
5.62610626e-01 -1.10706888e-01 -5.87818563e-01 7.40168095e-01
-5.15721858e-01 -6.27885312e-02 -4.03492413e-02 1.99108884e-01
4.36998218e-01 1.06245565e+00 9.21790123e-01 -1.18359782e-01
6.31426990e-01 -1.17753303e+00 -1.04897392e+00 3.45737524e-02
-9.00996685e-01 -1.81475952e-01 1.50610050e-02 3.87403846e-01
-1.32872546e+00 -3.24402303e-01 -3.35267305e-01 8.75033379e-01
-5.87602295e-02 1.15665483e+00 -3.05557400e-01 3.40350647e-03
8.17662403e-02 -5.52994013e-01 -6.24618948e-01 -4.23005849e-01
1.85245454e-01 6.49934351e-01 1.44726083e-01 -6.27003014e-01
8.09224606e-01 4.96338695e-01 1.90280214e-01 -7.24533260e-01
6.62881806e-02 -4.29607540e-01 -3.43863398e-01 -7.03980684e-01
6.97706282e-01 -1.17097938e+00 -6.73175275e-01 2.62736589e-01
-4.92388576e-01 -5.97068369e-01 -3.09490174e-01 7.17929184e-01
-4.49029446e-01 -1.75970390e-01 3.83059308e-03 -9.29939747e-01
1.64440140e-01 -1.24947250e+00 1.47121504e-01 1.28896805e-02
-3.38223934e-01 -1.12243557e+00 1.53589621e-01 7.12570667e-01
6.99398339e-01 6.36012316e-01 1.35330737e+00 -1.17695667e-01
-1.07869399e+00 -2.74575561e-01 -3.24922353e-01 1.38836995e-01
-3.32447380e-04 3.47497106e-01 -6.01399481e-01 -3.84492248e-01
-6.99141979e-01 9.25572291e-02 5.34514844e-01 5.73176920e-01
1.11903882e+00 -6.34427741e-02 -6.61624789e-01 1.38644932e-03
1.79233742e+00 2.92939752e-01 8.99283886e-01 5.78154325e-01
7.86055997e-02 6.13480031e-01 6.88169479e-01 5.82737982e-01
1.05851376e+00 3.39162260e-01 6.98954880e-01 -2.12232918e-02
-1.00552537e-01 -3.35776061e-01 2.27971390e-01 6.77896976e-01
-3.76915932e-01 -4.80085254e-01 -1.31088495e+00 1.02472651e+00
-1.20661199e+00 -8.34433258e-01 -5.15084684e-01 2.18456173e+00
2.55427778e-01 4.85255390e-01 5.11786640e-01 7.23260716e-02
2.33881757e-01 -1.29394531e-01 -4.10415262e-01 -6.28560841e-01
2.82160491e-01 1.51612803e-01 1.33171141e+00 2.73722559e-01
-3.25192094e-01 1.39422372e-01 7.20647049e+00 7.94104517e-01
-7.22200930e-01 4.95506674e-01 7.15892017e-01 -5.17455578e-01
-1.12901151e+00 1.76045895e-01 -6.18220985e-01 6.72219098e-01
1.96890736e+00 -4.91908550e-01 5.84778130e-01 3.12083155e-01
1.00607646e+00 -6.11534119e-01 -9.52594817e-01 1.77489042e-01
-7.25109100e-01 -1.44574225e+00 -5.18516064e-01 4.08426613e-01
9.48309183e-01 6.74678564e-01 -8.94461349e-02 4.16605949e-01
6.61469758e-01 -7.49015749e-01 9.95341539e-01 6.05845630e-01
5.43368280e-01 -1.33139801e+00 6.46758854e-01 3.37965153e-02
-1.39401925e+00 -4.97112125e-01 1.54256031e-01 -6.31162971e-02
6.62599325e-01 5.25064409e-01 -4.49057221e-01 6.48180783e-01
9.15624678e-01 3.22855264e-01 -1.60229102e-01 1.31000614e+00
-1.15436196e-01 6.87316835e-01 -6.36792243e-01 -1.72329232e-01
5.69565296e-01 -3.68892312e-01 1.36776507e-01 9.39422071e-01
6.73309326e-01 -2.11326763e-01 -2.12287262e-01 1.04776394e+00
9.13582277e-03 -2.83562273e-01 -3.55547577e-01 -9.56076309e-02
7.32003748e-01 1.15087497e+00 -4.41237539e-01 1.45494223e-01
-7.90335834e-01 -3.34626108e-01 -4.19629544e-01 4.53859687e-01
-1.36846483e+00 -5.48095882e-01 7.70729065e-01 3.68454784e-01
3.09389625e-02 -4.58754063e-01 -4.04534936e-01 -5.65649450e-01
-5.18076420e-01 -2.02954650e-01 6.04609698e-02 -9.88701820e-01
-8.91335666e-01 -4.79007035e-01 8.03115845e-01 -8.86346161e-01
-7.09559098e-02 -4.56263661e-01 -8.13095331e-01 1.01230395e+00
-2.09637523e+00 -1.08732796e+00 -2.14021131e-01 2.31560841e-01
2.22240821e-01 1.18759543e-01 5.11771500e-01 8.65498543e-01
-7.72628248e-01 1.32733449e-01 6.34019852e-01 -3.09418231e-01
-9.66122970e-02 -1.04698503e+00 5.77362120e-01 3.32040191e-01
-5.67187726e-01 -1.45065799e-01 4.86295015e-01 -6.48170650e-01
-1.22211230e+00 -1.02229691e+00 7.99555242e-01 -2.21869156e-01
1.08906996e+00 -1.46768503e-02 -4.48863655e-02 7.69566596e-01
6.14609182e-01 -9.72438633e-01 8.41107845e-01 1.07094862e-01
4.39934850e-01 -4.11490321e-01 -1.69526637e+00 5.24504483e-01
5.43772638e-01 -2.88062245e-01 -1.68474659e-01 4.24117893e-01
1.70583144e-01 2.65464168e-02 -1.41536272e+00 -2.16234848e-01
6.05396330e-01 -4.85410094e-01 8.50492477e-01 -3.68706942e-01
2.34326404e-02 -2.25153610e-01 -1.94220364e-01 -1.80405295e+00
-3.94812584e-01 -2.30827019e-01 3.97438556e-01 1.39682996e+00
8.78832579e-01 -9.45489228e-01 3.45329404e-01 1.41600907e+00
-4.50069755e-01 -4.31625605e-01 -1.37421131e+00 -9.58506763e-01
4.28059787e-01 -7.19917953e-01 2.02954650e+00 6.15926087e-01
1.21842287e-01 -3.83014172e-01 2.19525382e-01 3.45042467e-01
6.06656015e-01 -3.74403715e-01 4.11439091e-01 -1.06940973e+00
4.36278582e-01 -6.53947294e-01 -2.88179725e-01 -2.13663176e-01
1.67343631e-01 -9.67703044e-01 -4.74206686e-01 -2.08408046e+00
1.80021808e-01 -9.17070389e-01 -5.81432521e-01 4.81423050e-01
8.16813648e-01 -1.38919085e-01 5.95953576e-02 -3.07000011e-01
2.20780358e-01 6.01889431e-01 6.58870578e-01 -8.86824429e-01
2.05930233e-01 5.14291264e-02 -4.66811895e-01 7.16236591e-01
9.87544060e-01 -4.95972157e-01 -4.24178928e-01 -6.74720109e-01
7.97296703e-01 -1.49047568e-01 4.71574306e-01 -1.11433375e+00
-1.54140994e-01 -9.17016029e-01 -5.93932718e-02 -1.24130297e+00
3.12830843e-02 -1.39153290e+00 1.13026011e+00 7.33258009e-01
-1.27899051e-01 -6.70216093e-03 3.25084120e-01 2.36083701e-01
1.04197882e-01 -3.16379994e-01 5.35977066e-01 2.98445914e-02
-5.45644581e-01 2.64907628e-01 -1.03554785e+00 -3.43033284e-01
1.42413318e+00 -1.13295175e-01 -5.91454506e-01 -2.05538556e-01
-3.65055919e-01 8.47348988e-01 4.12708282e-01 2.71038204e-01
3.25566418e-02 -1.54417098e+00 -6.67820156e-01 -1.05818249e-01
-7.91736245e-02 -5.29496491e-01 3.18234533e-01 6.73056602e-01
-7.74619401e-01 1.02493906e+00 -4.68903869e-01 -2.85610650e-02
-4.31857079e-01 3.28123897e-01 5.56819856e-01 -3.52638394e-01
-2.28065804e-01 6.84108511e-02 -5.56736350e-01 -2.69113749e-01
-3.03392440e-01 -1.01371384e+00 2.18383074e-01 2.82356143e-01
1.22457288e-01 1.49606180e+00 4.52241808e-01 -4.43351388e-01
-2.97307253e-01 -1.13777583e-02 7.51101553e-01 -1.89985335e-01
1.41963863e+00 -7.20830739e-01 -2.10131202e-02 2.41612777e-01
9.33338344e-01 -1.12884521e-01 -1.09098518e+00 5.35918236e-01
-4.14799191e-02 -4.21114296e-01 8.38806629e-01 -1.16584682e+00
-1.24192250e+00 5.17524302e-01 4.82600093e-01 6.91923678e-01
9.92544591e-01 -2.19510630e-01 1.22258031e+00 1.74015999e-01
6.55558407e-01 -1.76660120e+00 -8.44108343e-01 -1.12162016e-01
5.65691292e-01 -7.88354874e-01 1.59527883e-01 4.83140834e-02
-7.09969103e-02 6.99402392e-01 2.81200141e-01 2.21633151e-01
1.15227520e+00 1.44488424e-01 -6.84694290e-01 -5.82515001e-02
-5.07081628e-01 -1.39698209e-02 -1.12518400e-01 7.02802360e-01
-2.02893894e-02 8.72584581e-01 -3.83398682e-01 3.84836584e-01
-2.06227109e-01 3.81523669e-02 7.19882011e-01 7.11472034e-01
-2.98704028e-01 -1.03937066e+00 -3.92707109e-01 7.89180577e-01
4.48129363e-02 -5.19735133e-03 3.23509693e-01 1.09004760e+00
2.43202448e-01 1.34220684e+00 2.47626767e-01 -1.04639314e-01
5.39705396e-01 -2.06965327e-01 3.61675739e-01 -8.38448480e-02
-4.59630519e-01 -3.10887545e-01 8.11640441e-01 -1.74520507e-01
-4.40810680e-01 -9.41475213e-01 -1.46098316e+00 -1.03069699e+00
-3.45783561e-01 1.42568454e-01 1.29796326e+00 9.18481350e-01
1.27149627e-01 8.24908793e-01 8.79739761e-01 -7.20618308e-01
-2.39359617e-01 -6.70101166e-01 -9.66303051e-01 -1.67957619e-02
1.46716520e-01 -1.01136124e+00 -3.87271404e-01 -5.57301939e-01] | [5.905526638031006, 2.1311838626861572] |
1abb5674-b3d0-428b-bba4-407d5c5489f1 | three-dimensional-generative-adversarial-nets | 1911.08105 | null | https://arxiv.org/abs/1911.08105v3 | https://arxiv.org/pdf/1911.08105v3.pdf | Three-dimensional Generative Adversarial Nets for Unsupervised Metal Artifact Reduction | The reduction of metal artifacts in computed tomography (CT) images, specifically for strong artifacts generated from multiple metal objects, is a challenging issue in medical imaging research. Although there have been some studies on supervised metal artifact reduction through the learning of synthesized artifacts, it is difficult for simulated artifacts to cover the complexity of the real physical phenomena that may be observed in X-ray propagation. In this paper, we introduce metal artifact reduction methods based on an unsupervised volume-to-volume translation learned from clinical CT images. We construct three-dimensional adversarial nets with a regularized loss function designed for metal artifacts from multiple dental fillings. The results of experiments using 915 CT volumes from real patients demonstrate that the proposed framework has an outstanding capacity to reduce strong artifacts and to recover underlying missing voxels, while preserving the anatomical features of soft tissues and tooth structures from the original images. | ['Yuichiro Imai', 'Megumi Nakao', 'Keiho Imanishi', 'Nobuhiro Ueda', 'Tadaaki Kirita', 'Tetsuya Matsuda'] | 2019-11-19 | null | null | null | null | ['metal-artifact-reduction'] | ['medical'] | [ 5.33395946e-01 3.47939044e-01 5.28889835e-01 -3.38038892e-01
-1.17927408e+00 1.20302476e-01 4.70706671e-02 -1.25922978e-01
-1.27289742e-01 8.03544462e-01 1.65248603e-01 1.05978604e-02
-2.70795554e-01 -7.46638000e-01 -8.45949769e-01 -9.74009514e-01
-2.87406147e-01 5.49855649e-01 1.25370607e-01 1.35670587e-01
3.79369929e-02 6.46697700e-01 -8.33121955e-01 2.95955002e-01
1.08831906e+00 8.63616765e-01 1.13302968e-01 1.39759824e-01
1.33659929e-01 1.03624737e+00 -6.97518110e-01 -3.17669511e-01
3.36277127e-01 -3.89165819e-01 -6.72227263e-01 2.47669026e-01
4.92245495e-01 -3.64580154e-01 -4.85024333e-01 1.17422926e+00
5.52497566e-01 -1.87063605e-01 8.48445892e-01 -8.83037090e-01
-8.45845580e-01 6.72811210e-01 -6.96073890e-01 9.24633369e-02
-5.18838428e-02 1.46540418e-01 2.43230090e-01 -8.65829706e-01
6.81914866e-01 9.80171084e-01 7.89025068e-01 7.86160648e-01
-1.27579069e+00 -6.35181427e-01 -5.37508428e-01 -1.88754842e-01
-1.09499979e+00 2.27499694e-01 8.90040457e-01 -5.35925746e-01
2.68233150e-01 4.57470536e-01 5.82888901e-01 9.38346267e-01
8.62222552e-01 4.29555476e-01 1.31284201e+00 -2.24931836e-01
3.24930757e-01 8.55433866e-02 1.62105616e-02 6.79239810e-01
5.33439755e-01 1.74372986e-01 -9.90009010e-02 -5.13051867e-01
1.11777365e+00 2.46370137e-01 -5.11717200e-01 -2.60856986e-01
-6.98650360e-01 7.80699313e-01 6.37117565e-01 4.17713881e-01
-5.20400167e-01 3.03392380e-01 2.34784633e-01 -6.50694733e-03
6.34176970e-01 4.35604990e-01 -1.43476516e-01 5.86843729e-01
-7.12267101e-01 3.36569846e-01 3.45784128e-01 7.84705102e-01
2.53986984e-01 4.03382003e-01 -1.30146921e-01 7.48719394e-01
3.61993253e-01 6.60425484e-01 5.54899454e-01 -6.83837831e-01
2.90394515e-01 3.24963719e-01 -2.14727655e-01 -7.20524490e-01
-2.55595475e-01 -3.81747514e-01 -1.06509066e+00 6.52895570e-01
3.86147201e-01 -3.06286439e-02 -1.36080396e+00 1.41630149e+00
5.37153304e-01 2.23594815e-01 -2.59203613e-01 9.01500762e-01
8.34448397e-01 2.58415729e-01 2.29534864e-01 -3.72082263e-01
1.23976183e+00 -5.49061596e-01 -9.97058272e-01 -1.90969676e-01
2.11804554e-01 -8.22359443e-01 1.01909578e+00 5.22590876e-01
-1.52562082e+00 -2.31802419e-01 -1.09463811e+00 5.36568351e-02
1.71818942e-01 -5.03990412e-01 6.01877868e-01 6.67380691e-01
-5.61774969e-01 8.42222929e-01 -1.05847383e+00 3.11489522e-01
8.34263027e-01 4.08429205e-01 -9.81505439e-02 -3.84214856e-02
-1.12694442e+00 9.47521985e-01 -3.48496169e-01 1.89513683e-01
-9.45402205e-01 -1.30035114e+00 -7.36883700e-01 -4.23370898e-01
1.31745577e-01 -6.11847341e-01 1.10798597e+00 -7.70511746e-01
-1.31695819e+00 7.92698264e-01 3.80105197e-01 -3.40285391e-01
1.01374960e+00 -2.68841445e-01 -3.79534602e-01 3.19423646e-01
2.36270681e-01 2.42990460e-02 7.98953354e-01 -1.68338430e+00
1.76344797e-01 -7.07957566e-01 -6.90415740e-01 -1.85602397e-01
-1.15540616e-01 -1.95756555e-02 -2.07273886e-02 -1.15418243e+00
5.21611273e-01 -1.07405424e+00 -7.60292768e-01 5.44143021e-01
-5.15549600e-01 3.31274033e-01 5.58543265e-01 -1.00417864e+00
6.63739800e-01 -2.05794072e+00 -1.61126807e-01 4.34931040e-01
2.29683682e-01 -1.73990175e-01 1.87355176e-01 6.80374773e-03
-1.20262928e-01 6.71755672e-02 -1.05085933e+00 -3.36484641e-01
-4.17340606e-01 2.08951935e-01 -2.03696825e-02 7.79345512e-01
-2.52659284e-02 7.20390260e-01 -7.80988276e-01 -6.96973026e-01
1.92902878e-01 8.30759346e-01 -4.93934333e-01 1.39576361e-01
-1.69611685e-02 9.70270634e-01 -7.41211653e-01 5.35969257e-01
1.14773226e+00 -1.67276829e-01 -7.73373097e-02 -1.16147436e-01
3.42273504e-01 -1.17670640e-01 -9.07116771e-01 1.68657351e+00
-5.44380784e-01 2.41810456e-01 2.33689487e-01 -3.52617532e-01
4.91833180e-01 6.56637251e-01 1.14423633e+00 -7.40022659e-01
2.95150071e-01 5.44790983e-01 -1.09286597e-02 -7.49630630e-01
-1.14339799e-01 -8.69158924e-01 1.95769608e-01 4.67538804e-01
-3.85474682e-01 -9.26078081e-01 -4.96787578e-01 -7.32511953e-02
1.22561729e+00 -2.13906243e-01 -3.37465167e-01 -3.90372217e-01
2.47201428e-01 -8.95202607e-02 4.98401999e-01 2.91113049e-01
-1.01412334e-01 1.10569227e+00 -2.97732145e-01 -5.71434438e-01
-1.17520154e+00 -1.55432582e+00 -8.14808846e-01 -1.01499133e-01
4.86538820e-02 4.57747877e-01 -9.69018757e-01 -6.33622169e-01
-3.04402877e-02 5.52262843e-01 -7.00531006e-01 -2.25684986e-01
-1.06408930e+00 -1.10584950e+00 4.30146515e-01 4.97786462e-01
3.41519296e-01 -1.38163364e+00 -4.32071000e-01 4.38707262e-01
-2.23118484e-01 -1.02984798e+00 -6.52269900e-01 2.54980206e-01
-1.33460522e+00 -1.09725344e+00 -8.77666891e-01 -9.83066082e-01
1.10028231e+00 -1.62183195e-01 1.12210202e+00 3.55657011e-01
-1.11304116e+00 1.00329652e-01 -1.75402071e-02 -6.84401512e-01
-7.98520863e-01 -7.11235702e-01 -6.46221936e-02 -5.90644516e-02
-9.64849517e-02 -8.29846740e-01 -8.48884881e-01 1.96504340e-01
-1.38665128e+00 -1.90760776e-01 3.90179753e-01 8.34866583e-01
8.69951010e-01 3.52370769e-01 3.09717387e-01 -1.10016274e+00
4.66184944e-01 -2.25266531e-01 -2.11800978e-01 -5.43258004e-02
-5.05469859e-01 1.54838085e-01 4.89609629e-01 -4.82532799e-01
-1.35502267e+00 8.06960538e-02 -2.56164998e-01 -3.13791931e-01
2.52889041e-02 -1.17496967e-01 -4.09438610e-02 -4.49177593e-01
8.03706229e-01 1.98729992e-01 -6.89001335e-03 -3.83308113e-01
-3.51325087e-02 2.73103476e-01 4.48263079e-01 -6.24815762e-01
9.77321446e-01 9.39475000e-01 3.29652041e-01 -8.04656208e-01
-6.02582157e-01 5.26362062e-02 -3.59557211e-01 -6.95078522e-02
9.75768149e-01 -6.06359601e-01 -4.44542378e-01 5.76098621e-01
-1.10097253e+00 -2.69583821e-01 -6.23332620e-01 6.16405249e-01
-6.30898714e-01 6.15812123e-01 -1.20091820e+00 -4.52499658e-01
-7.81544030e-01 -1.45052016e+00 8.26008677e-01 -5.82868941e-02
-4.71892096e-02 -8.80364299e-01 2.65646964e-01 5.11325955e-01
3.86140525e-01 7.06667483e-01 1.25515223e+00 3.81284207e-02
-6.39186561e-01 -5.61123155e-02 2.30815217e-01 6.35277987e-01
4.63221520e-01 -2.44955093e-01 -9.21685457e-01 -2.63096154e-01
9.55587327e-01 -2.93433845e-01 5.33133090e-01 8.36365283e-01
1.47913539e+00 -2.34588534e-01 -2.07524374e-01 8.01072061e-01
1.71759713e+00 4.43640739e-01 1.02228427e+00 2.23428831e-01
6.17905855e-01 5.24403155e-01 2.42879599e-01 3.22476178e-01
-3.14402074e-01 4.34090227e-01 6.59078717e-01 -4.44934875e-01
-4.90565807e-01 3.19636017e-02 -2.09249839e-01 6.47961497e-01
-1.66971147e-01 1.52086556e-01 -7.62573779e-01 6.03025496e-01
-1.16414440e+00 -5.82597554e-01 -6.29847229e-01 2.10985947e+00
9.86988246e-01 7.84659311e-02 -5.91501355e-01 1.94694236e-01
7.39897966e-01 -3.64286005e-01 -6.24508739e-01 -2.91004833e-02
1.17578983e-01 8.44065011e-01 5.18715680e-01 4.67118621e-01
-7.29622424e-01 3.36730301e-01 6.87278748e+00 6.43759549e-01
-9.96526420e-01 4.82868493e-01 7.00265229e-01 -1.27396718e-01
-5.31970024e-01 -5.54070830e-01 1.92680284e-02 5.04989862e-01
4.28345323e-01 3.06577444e-01 -4.62930016e-02 7.57913828e-01
3.02790821e-01 9.47770700e-02 -9.01743710e-01 7.54974961e-01
-5.42292744e-02 -9.80721116e-01 1.69849575e-01 9.37612131e-02
1.02901161e+00 -1.91520110e-01 2.80795366e-01 -3.86610448e-01
3.90000731e-01 -1.15000153e+00 5.02460063e-01 2.23051369e-01
6.46985710e-01 -7.16338217e-01 4.84070003e-01 -1.00440748e-01
-6.38935626e-01 2.19884127e-01 -3.48547339e-01 4.56534743e-01
4.97648239e-01 9.91812229e-01 -8.39894295e-01 3.87318641e-01
7.18421280e-01 5.45357727e-02 -1.55745327e-01 1.20320010e+00
-1.91646770e-01 3.95097315e-01 -2.43754387e-01 4.65391308e-01
2.38963403e-02 -2.61376023e-01 4.34542805e-01 7.65036464e-01
3.79720986e-01 3.10191423e-01 -2.20838577e-01 9.69693422e-01
-1.59069166e-01 2.46738300e-01 -5.94396293e-01 8.15173328e-01
-1.52811268e-02 9.99208629e-01 -8.54735792e-01 -1.86392367e-01
-2.05513135e-01 9.84462798e-01 -1.76035821e-01 1.63999736e-01
-9.96126115e-01 4.43060845e-02 4.67058927e-01 6.45604968e-01
-2.65455753e-01 4.53292504e-02 -8.63560259e-01 -7.90552974e-01
2.78976828e-01 -7.62116015e-01 1.45014688e-01 -7.24754870e-01
-1.54106450e+00 9.38609123e-01 -2.38506868e-01 -1.25748992e+00
3.72483134e-01 -4.34867561e-01 -7.11732328e-01 5.57501197e-01
-1.20967603e+00 -1.04343235e+00 -4.12263125e-01 8.06306660e-01
5.33908248e-01 3.09685618e-01 6.28596187e-01 5.88805914e-01
-1.10759214e-02 5.60342848e-01 1.37217626e-01 1.03183471e-01
5.59178889e-01 -1.18821144e+00 1.01813637e-01 5.20369947e-01
-6.23873957e-02 2.98642009e-01 8.40289295e-01 -1.00979853e+00
-1.07576454e+00 -1.18981409e+00 3.51245522e-01 -3.74584705e-01
4.82874364e-01 -4.01096828e-02 -1.02734685e+00 7.88798988e-01
2.04201311e-01 4.41687435e-01 6.00690424e-01 -8.05349827e-01
-2.25814842e-02 4.38796543e-02 -1.90931261e+00 5.02836227e-01
8.44718874e-01 -9.89925936e-02 -5.69831252e-01 7.20055997e-01
5.99852920e-01 -4.93375689e-01 -1.17588866e+00 6.03655159e-01
3.10093939e-01 -5.77159524e-01 1.37434876e+00 -4.69209522e-01
7.33792722e-01 6.69568330e-02 1.43693060e-01 -1.29179561e+00
-2.69067079e-01 -3.04136872e-01 3.98012727e-01 6.65903330e-01
2.91835994e-01 -5.31640708e-01 1.00396574e+00 8.21543992e-01
-5.41716754e-01 -6.50534868e-01 -1.25812137e+00 -7.10819602e-01
6.31618559e-01 -4.05360639e-01 3.82386833e-01 9.95402396e-01
-3.69687021e-01 -4.22784299e-01 -2.59088695e-01 4.93383229e-01
1.35973227e+00 -1.78526551e-01 2.16009215e-01 -1.09703052e+00
-3.72031212e-01 -7.86066130e-02 -2.25866348e-01 -3.04565549e-01
-1.78603545e-01 -8.46825480e-01 3.98484230e-01 -1.53565097e+00
4.27970588e-01 -7.20262170e-01 -2.00549379e-01 1.30721256e-01
-1.43913403e-01 6.70764923e-01 -2.77546197e-02 2.76812822e-01
4.01595622e-01 5.98425031e-01 2.17609072e+00 -5.95506728e-01
-5.71150705e-02 9.72534250e-03 -2.21510082e-01 1.07420957e+00
5.54221094e-01 -1.04948151e+00 -3.12434673e-01 -6.28451467e-01
-2.22510114e-01 3.63135934e-01 4.36435819e-01 -1.20608890e+00
-1.32883474e-01 -1.47236750e-01 3.78029615e-01 -3.94710183e-01
3.41917187e-01 -1.21164453e+00 5.79898238e-01 9.88880634e-01
-1.41873255e-01 -2.00760454e-01 1.60306349e-01 3.64085138e-01
-1.77855417e-01 -2.73036987e-01 1.28922856e+00 -5.42771578e-01
2.44179711e-01 5.77537417e-01 -2.92217195e-01 4.25528735e-02
9.63775277e-01 -6.33521006e-02 1.60961583e-01 -5.90147562e-02
-1.04862416e+00 -3.80786747e-01 2.96316832e-01 9.19424295e-02
9.93887305e-01 -1.43101597e+00 -8.51790607e-01 1.32768407e-01
-2.06786230e-01 4.13660794e-01 5.45979798e-01 6.25976682e-01
-9.79188740e-01 -9.72042829e-02 -2.06825778e-01 -5.10757864e-01
-1.17716098e+00 5.46316862e-01 6.03056967e-01 -6.62885189e-01
-9.99479651e-01 8.33624601e-01 5.94687045e-01 -1.94270656e-01
-4.21748846e-05 -6.19840622e-01 3.88630629e-01 -6.76624715e-01
4.16174263e-01 3.50707173e-01 4.95290339e-01 -3.78876567e-01
-1.22595750e-01 8.48973274e-01 -1.32991940e-01 4.41673882e-02
1.66956270e+00 1.39473483e-01 -8.81352201e-02 7.72387115e-03
1.06863248e+00 -1.57271454e-03 -1.05113876e+00 -2.21444309e-01
-2.73408115e-01 -5.32294214e-01 8.03023949e-02 -7.81377673e-01
-1.73184538e+00 7.15223134e-01 1.08502650e+00 -1.12461306e-01
9.81972158e-01 -4.89489846e-02 1.24230635e+00 -2.34389171e-01
7.25063920e-01 -9.00569975e-01 5.41398942e-01 -2.96820432e-01
1.19744563e+00 -1.10047281e+00 1.19983017e-01 -9.42899525e-01
-3.67694318e-01 8.08141887e-01 5.40048301e-01 -4.12380993e-01
1.00750697e+00 6.73083067e-01 4.17146951e-01 -5.83133876e-01
9.80538800e-02 5.79235375e-01 -4.03547995e-02 5.61192393e-01
3.27952057e-01 7.58747160e-02 -4.92566258e-01 3.90971720e-01
-6.56190440e-02 1.98489986e-02 6.21139765e-01 1.14685130e+00
-6.55706227e-02 -1.03255987e+00 -8.50824058e-01 3.96607131e-01
-1.01305842e+00 -1.63358390e-01 8.19962844e-02 7.90565908e-01
1.79900080e-01 7.01825738e-01 -3.52080137e-01 1.97794408e-01
4.71157223e-01 -2.60116398e-01 7.77992666e-01 -6.97966933e-01
-8.85811269e-01 1.77641511e-01 -4.68418330e-01 -4.50922430e-01
-2.33329028e-01 -5.23552120e-01 -1.60181606e+00 -3.96254621e-02
-3.57713580e-01 -8.70729387e-02 6.71317160e-01 6.21505916e-01
-2.53088474e-01 9.32881057e-01 6.80660725e-01 -6.18192315e-01
-6.25485301e-01 -1.07460010e+00 -1.06158519e+00 1.02624929e+00
3.16180021e-01 -5.71090519e-01 -5.98161221e-01 1.27417639e-01] | [13.46764087677002, -2.546825408935547] |
7417e00a-e61d-433a-866e-411bd9168172 | residue-based-natural-language-adversarial-1 | 2204.10192 | null | https://arxiv.org/abs/2204.10192v2 | https://arxiv.org/pdf/2204.10192v2.pdf | Residue-Based Natural Language Adversarial Attack Detection | Deep learning based systems are susceptible to adversarial attacks, where a small, imperceptible change at the input alters the model prediction. However, to date the majority of the approaches to detect these attacks have been designed for image processing systems. Many popular image adversarial detection approaches are able to identify adversarial examples from embedding feature spaces, whilst in the NLP domain existing state of the art detection approaches solely focus on input text features, without consideration of model embedding spaces. This work examines what differences result when porting these image designed strategies to Natural Language Processing (NLP) tasks - these detectors are found to not port over well. This is expected as NLP systems have a very different form of input: discrete and sequential in nature, rather than the continuous and fixed size inputs for images. As an equivalent model-focused NLP detection approach, this work proposes a simple sentence-embedding "residue" based detector to identify adversarial examples. On many tasks, it out-performs ported image domain detectors and recent state of the art NLP specific detectors. | ['Mark Gales', 'Vyas Raina'] | 2022-04-17 | null | https://aclanthology.org/2022.naacl-main.281 | https://aclanthology.org/2022.naacl-main.281.pdf | naacl-2022-7 | ['adversarial-attack-detection', 'adversarial-attack-detection'] | ['computer-vision', 'knowledge-base'] | [ 6.75067484e-01 1.26455277e-01 2.28491679e-01 -4.30107936e-02
-6.06089890e-01 -1.09874845e+00 1.17434466e+00 1.02922134e-01
-4.05682445e-01 1.89230144e-01 -3.92568037e-02 -5.14867067e-01
4.04846132e-01 -8.13461304e-01 -8.22892964e-01 -4.86767650e-01
4.95560430e-02 1.06608838e-01 4.49853212e-01 -4.82138008e-01
4.43607450e-01 9.18479860e-01 -1.03218853e+00 7.26188660e-01
1.23164527e-01 6.97400689e-01 -4.51037824e-01 1.25613630e+00
5.60060609e-03 9.27271485e-01 -9.21185553e-01 -8.17460537e-01
9.18022454e-01 -2.78706640e-01 -6.59248710e-01 -3.96618433e-02
9.34326470e-01 -4.31287259e-01 -1.03342474e+00 1.58180225e+00
4.99040633e-01 -4.02914852e-01 8.43380034e-01 -1.45919871e+00
-1.12074184e+00 3.06662858e-01 -3.65617186e-01 4.65072781e-01
5.97573757e-01 7.96795011e-01 5.31403065e-01 -7.08866179e-01
6.50762796e-01 1.66386294e+00 5.36942601e-01 7.82293856e-01
-1.34480667e+00 -7.36609578e-01 -1.31355748e-01 1.44919410e-01
-1.08641815e+00 -3.13797414e-01 7.92556345e-01 -6.14383280e-01
1.20070350e+00 3.95746320e-01 1.38674840e-01 1.60272479e+00
6.89008653e-01 8.17786813e-01 1.33247709e+00 -6.68911517e-01
5.56199700e-02 5.86373866e-01 -1.73394248e-01 5.34816206e-01
-1.83634628e-02 5.34389913e-01 -1.03814252e-01 -3.55403304e-01
5.68833947e-01 -1.73368812e-01 -1.69042394e-01 -1.53990343e-01
-9.28077042e-01 1.02528882e+00 3.86707842e-01 4.11434919e-01
-9.16077122e-02 1.95059031e-01 8.46038818e-01 8.77004445e-01
1.33252084e-01 7.71207273e-01 -2.55677938e-01 2.83880472e-01
-8.16756189e-01 2.30720058e-01 8.43752444e-01 9.96327043e-01
3.66121590e-01 2.17520460e-01 -3.39138776e-01 2.01009929e-01
2.17342570e-01 4.51954991e-01 5.29565215e-01 -2.78572530e-01
5.03891826e-01 4.02871460e-01 -2.80876964e-01 -1.45262992e+00
-3.79462466e-02 9.18807164e-02 -7.29149222e-01 8.87573957e-01
5.21031618e-01 -2.42985748e-02 -1.16372311e+00 1.28818357e+00
-3.40638980e-02 -2.32862309e-01 3.05468529e-01 6.90052927e-01
5.24488568e-01 8.66679370e-01 2.98307747e-01 2.17256740e-01
1.56034029e+00 -7.37272441e-01 -2.88054943e-01 -4.43736464e-01
3.81620973e-01 -1.08797395e+00 9.99007463e-01 3.46930236e-01
-8.14020813e-01 -5.57600498e-01 -1.25729883e+00 2.22189531e-01
-1.03084755e+00 -3.79620194e-01 -1.33992229e-02 1.13900805e+00
-1.00103629e+00 5.62321484e-01 -3.70302141e-01 -4.42341596e-01
5.42805254e-01 2.97862530e-01 -4.92190421e-01 -1.30553275e-01
-1.46889818e+00 1.17747533e+00 5.05650699e-01 -3.58716697e-01
-1.09431410e+00 -5.36817253e-01 -7.98739254e-01 -2.03213751e-01
3.11277155e-02 -4.40536916e-01 1.05057442e+00 -1.60658765e+00
-1.19532073e+00 1.36234999e+00 4.22425300e-01 -9.32383716e-01
9.94229734e-01 1.20513737e-01 -7.47167706e-01 4.65368122e-01
-3.29246372e-01 7.10419416e-01 1.83944988e+00 -1.14496708e+00
-1.55124962e-01 -1.98069915e-01 1.26485974e-01 -2.13493347e-01
-5.38789749e-01 6.66037738e-01 1.33835211e-01 -9.74345565e-01
-4.64886487e-01 -7.96153843e-01 -3.88359845e-01 4.30247217e-01
-6.23915195e-01 1.86883762e-01 1.25785041e+00 -4.69913304e-01
9.27214742e-01 -2.30246568e+00 -3.55026364e-01 1.33369312e-01
2.45316297e-01 8.96530092e-01 -3.30958009e-01 8.29142869e-01
-5.09675205e-01 5.27372301e-01 -2.38980725e-01 8.35494399e-02
3.34591806e-01 9.31500196e-02 -8.14578831e-01 8.01798165e-01
6.88086450e-01 1.01686478e+00 -6.80139840e-01 -5.39371371e-01
4.29772168e-01 3.08053672e-01 -2.53469437e-01 1.02284670e-01
-9.69292298e-02 -3.13133411e-02 -3.21069688e-01 5.83685577e-01
8.57946217e-01 3.23386222e-01 -3.91084403e-01 -1.09668970e-01
1.81990966e-01 -2.04519600e-01 -9.13371921e-01 1.06223488e+00
-9.22192931e-02 1.14427125e+00 -1.31903207e-02 -1.10802329e+00
7.97682583e-01 4.15276349e-01 -8.87216777e-02 -5.90229809e-01
9.71625969e-02 1.17422931e-01 1.63550183e-01 -5.81210077e-01
3.17492157e-01 -7.04043880e-02 -1.45586044e-01 6.57927766e-02
9.73262787e-02 -2.12690696e-01 -8.90239999e-02 3.50824952e-01
1.68958890e+00 -2.68505067e-01 6.37162626e-01 -9.32621881e-02
8.96438718e-01 2.31080115e-01 -9.70458090e-02 1.33759141e+00
-7.00044274e-01 7.41146207e-01 5.61286688e-01 -6.18413210e-01
-1.38048160e+00 -1.23898900e+00 -2.47151300e-01 7.53630400e-01
-6.39436990e-02 -1.77545354e-01 -8.75087202e-01 -9.51125503e-01
4.45335507e-02 5.71224630e-01 -6.07572794e-01 -4.80769426e-01
-5.65644979e-01 -4.02673692e-01 1.21760750e+00 2.52269328e-01
3.59068215e-01 -1.47253156e+00 -5.13709009e-01 1.61756396e-01
6.22142911e-01 -1.24323547e+00 -4.44613665e-01 1.05614193e-01
-5.20046115e-01 -9.72573817e-01 -6.10502422e-01 -1.01983976e+00
7.36974537e-01 -2.22373575e-01 1.21648669e+00 -1.21511683e-01
-8.89372587e-01 6.84515476e-01 -3.47128183e-01 -8.76772165e-01
-1.38681257e+00 -4.05697346e-01 -2.19588690e-02 2.33900221e-03
7.44335413e-01 -2.78215647e-01 -4.63999301e-01 -4.20672335e-02
-1.45426989e+00 -5.61974227e-01 8.14060330e-01 6.67369425e-01
1.00089118e-01 2.03625992e-01 2.58131623e-01 -8.77532899e-01
1.01299226e+00 -2.12259769e-01 -4.22614992e-01 1.77553803e-01
-4.56091344e-01 -6.93699941e-02 1.19406736e+00 -1.03648913e+00
-4.69702244e-01 3.49439889e-01 -2.59720117e-01 -8.43047678e-01
-8.43851268e-01 -2.47996636e-02 -2.24888936e-01 -4.43585962e-01
1.19867671e+00 5.66190898e-01 7.10966066e-02 3.61043885e-02
3.99210125e-01 6.00910366e-01 5.15754044e-01 -2.05869690e-01
1.52233016e+00 3.33558649e-01 -6.49133027e-02 -1.05061257e+00
-8.67902786e-02 -3.34470689e-01 -6.27145231e-01 -1.23667076e-01
8.25261474e-01 -6.26372159e-01 -3.47350955e-01 7.62619734e-01
-1.34998894e+00 8.54377225e-02 -4.82453316e-01 1.00413142e-02
-4.97961730e-01 8.13968480e-01 -8.25170338e-01 -8.22725654e-01
-3.20671231e-01 -1.24896443e+00 9.00280654e-01 -2.54817665e-01
-2.34873876e-01 -1.00029624e+00 4.70803194e-02 2.71305814e-02
3.89721096e-01 6.21411502e-01 1.00987220e+00 -1.04578781e+00
-3.46691906e-01 -9.27278578e-01 -2.10801974e-01 7.76428282e-01
-1.49442062e-01 2.05952123e-01 -1.18838441e+00 -3.80781740e-01
3.22467417e-01 -3.19582790e-01 7.40353048e-01 -2.17246413e-02
7.22637296e-01 -8.62821162e-01 -1.38694122e-01 2.41601318e-01
1.76903427e+00 1.07554972e-01 1.04843116e+00 5.53168952e-01
4.94196028e-01 5.06640077e-01 2.94702232e-01 1.34543970e-01
-6.20780945e-01 4.63105887e-01 7.38273323e-01 -2.97500253e-01
-1.68767769e-03 -2.12732658e-01 9.42365587e-01 1.60998814e-02
7.72302210e-01 -6.13386333e-01 -1.08655512e+00 4.81445163e-01
-1.55589664e+00 -1.23367643e+00 -1.68478042e-01 1.66207969e+00
8.31553102e-01 5.59806943e-01 -1.54874856e-02 1.47376806e-01
8.47886264e-01 3.36045027e-01 -5.04101098e-01 -1.11460066e+00
-1.81325570e-01 3.87896031e-01 9.40435410e-01 9.12272707e-02
-1.65886497e+00 1.00129795e+00 6.76670551e+00 9.73525763e-01
-1.05616856e+00 -2.63493508e-02 4.95594561e-01 2.58537263e-01
1.53709829e-01 -2.05839276e-01 -4.10332680e-01 5.23496270e-01
9.50020790e-01 -8.19965526e-02 1.27847850e-01 1.14038944e+00
1.90165155e-02 4.04335320e-01 -1.37087548e+00 8.62675548e-01
1.92973971e-01 -1.23099494e+00 4.68597054e-01 1.29710898e-01
4.40339118e-01 -2.08957382e-02 5.19288480e-01 1.55733496e-01
3.09163719e-01 -1.37661541e+00 7.38624990e-01 4.04388815e-01
5.18704116e-01 -6.60740376e-01 6.73153460e-01 4.37513411e-01
-7.32867181e-01 -1.62062779e-01 -6.04655564e-01 -1.66864600e-02
-1.84510276e-01 1.26616955e-01 -9.31751370e-01 -3.45749371e-02
5.53992629e-01 2.88616270e-01 -8.13556790e-01 5.93338668e-01
-1.74552172e-01 6.29558384e-01 -2.34439746e-01 -1.15748094e-02
5.90965807e-01 2.58305341e-01 9.35703158e-01 1.71774673e+00
-1.88811228e-01 -2.59481013e-01 2.51881540e-01 9.60857451e-01
-8.09355974e-02 1.04215994e-01 -1.29958284e+00 -3.43786865e-01
1.52558446e-01 1.08096576e+00 -6.04500234e-01 -1.96011215e-01
-4.89020735e-01 1.37459683e+00 -1.81694299e-01 1.05674796e-01
-9.16590869e-01 -4.74473268e-01 8.63918424e-01 2.12987423e-01
2.50096321e-01 -1.33696347e-01 -2.09706780e-02 -8.95992935e-01
1.42492473e-01 -1.31274903e+00 2.31273651e-01 -4.92479622e-01
-1.85575151e+00 5.70687890e-01 -1.85281411e-01 -1.33472681e+00
-2.65647054e-01 -1.24543488e+00 -9.32807267e-01 8.59640837e-01
-1.15112627e+00 -1.34050703e+00 2.58128434e-01 8.63973439e-01
6.75094306e-01 -6.53706193e-01 9.50473249e-01 -6.49410114e-03
-2.21433148e-01 9.49164629e-01 1.48875028e-01 7.87982464e-01
6.92792654e-01 -1.23966742e+00 9.04170096e-01 1.34898329e+00
3.63403559e-01 3.99317414e-01 9.82959449e-01 -3.39957058e-01
-1.47096014e+00 -1.24364090e+00 4.82431561e-01 -8.38860571e-01
1.00360334e+00 -5.39304435e-01 -8.80857408e-01 4.74659801e-01
4.86923993e-01 2.76996285e-01 3.13157350e-01 -8.18851590e-01
-7.65981078e-01 2.77491689e-01 -1.58876276e+00 7.97950625e-01
5.23712516e-01 -1.01163316e+00 -7.35524118e-01 6.90158188e-01
6.21278703e-01 -2.14933574e-01 -6.82479978e-01 1.03242010e-01
3.10535729e-01 -7.78972030e-01 1.43777633e+00 -8.79731536e-01
6.40355170e-01 -3.16259444e-01 -5.03574982e-02 -9.35546279e-01
-3.73542577e-01 -7.92788327e-01 -1.45142376e-01 1.20122731e+00
3.50662827e-01 -7.08473325e-01 7.09859252e-01 5.23878098e-01
2.42396846e-01 -2.89346606e-01 -8.95549834e-01 -9.39499736e-01
3.75461608e-01 -4.06436086e-01 9.25808102e-02 9.13258076e-01
-2.85292268e-01 3.59647423e-02 -3.55348319e-01 4.52939153e-01
5.55476665e-01 -4.46864873e-01 7.42758751e-01 -6.77270651e-01
-3.41575861e-01 -6.00146413e-01 -1.36815560e+00 -4.48818415e-01
1.64975181e-01 -7.27945626e-01 -1.05621666e-01 -8.03541541e-01
-1.62881330e-01 7.03884885e-02 -2.03648686e-01 3.86691570e-01
5.85225038e-02 5.77215016e-01 5.32244503e-01 2.66096413e-01
-1.08840548e-01 -1.58485502e-01 7.92862535e-01 -7.87247419e-01
2.89877146e-01 -4.95470278e-02 -3.80464792e-01 8.30890179e-01
9.23211455e-01 -8.48875403e-01 -2.85639405e-01 -1.57384932e-01
-2.27218438e-02 -5.34465075e-01 7.29936540e-01 -1.21045721e+00
2.23273158e-01 -9.96326581e-02 6.36831760e-01 -1.60650492e-01
3.70280445e-02 -1.06203401e+00 -1.55839950e-01 8.28648925e-01
-4.68257964e-01 1.26094073e-01 3.77435982e-01 7.50171423e-01
-2.53127486e-01 -5.99248290e-01 1.30657041e+00 -6.86661601e-01
-9.30222929e-01 2.49089926e-01 -9.23276126e-01 9.73139107e-02
1.46128142e+00 -5.56476116e-01 -3.30703557e-01 -2.64863878e-01
-6.32905066e-01 -3.38606149e-01 6.30982280e-01 7.22701907e-01
7.93160677e-01 -9.32010233e-01 -8.87247801e-01 2.21131921e-01
2.50562608e-01 -5.54259777e-01 -9.20861959e-02 2.27453768e-01
-1.01096082e+00 3.41601551e-01 -4.82938737e-01 -4.71197248e-01
-1.49785292e+00 1.35549128e+00 4.78854537e-01 -3.24387580e-01
-6.57604635e-01 7.93604493e-01 4.13488477e-01 -3.05915862e-01
1.00352690e-01 8.39611515e-02 6.32891282e-02 -1.46165848e-01
6.71371937e-01 -1.19970679e-01 -4.47543450e-02 -6.73587263e-01
-3.64901781e-01 5.34683131e-02 -4.62361455e-01 1.54353291e-01
8.41410220e-01 3.67840856e-01 9.12053064e-02 5.06823249e-02
1.48733234e+00 -5.26231416e-02 -9.41379964e-01 -1.46149009e-01
-1.58423092e-02 -5.11768520e-01 -2.38670275e-01 -7.08830059e-01
-7.10738659e-01 1.09292912e+00 9.60657001e-01 6.50400519e-01
1.05803311e+00 -7.17485249e-02 6.30936265e-01 4.19178575e-01
2.54103038e-02 -9.53362584e-01 2.35458434e-01 3.86142910e-01
1.09869397e+00 -1.25624669e+00 -3.61817144e-02 -2.96116561e-01
-4.77223188e-01 1.44678080e+00 5.90292573e-01 -5.51489830e-01
5.29488266e-01 5.50735652e-01 1.33684367e-01 -1.53258204e-01
-3.25415432e-01 2.08767548e-01 9.47811827e-02 1.09506679e+00
1.99830458e-02 -1.59438729e-01 3.80749702e-02 -6.49817735e-02
-1.29812524e-01 -5.47512472e-01 6.72060251e-01 1.01663506e+00
-1.64720878e-01 -1.22702610e+00 -7.49147475e-01 2.69722164e-01
-7.66203344e-01 -4.69069779e-01 -9.67469215e-01 9.97752130e-01
-1.95667036e-02 8.11877549e-01 -6.91197962e-02 -4.51082826e-01
3.74831676e-01 9.36711580e-02 4.03144896e-01 -5.73077917e-01
-1.20105839e+00 -6.56835258e-01 -3.67598265e-01 -4.88951921e-01
-1.31744102e-01 -5.10769546e-01 -6.48285985e-01 -3.68464589e-01
-1.64386496e-01 -3.58986259e-01 5.42800665e-01 7.55211949e-01
1.39769405e-01 2.11069196e-01 7.87665546e-01 -7.93755651e-01
-1.04801571e+00 -7.12830067e-01 -3.48254591e-01 7.71732509e-01
5.12655318e-01 5.83605170e-02 -6.09125376e-01 2.75299460e-01] | [5.878303050994873, 8.059139251708984] |
a44102e2-0bba-4374-8b97-979e0c3775ad | qaf-frame-semantics-based-question | null | null | https://aclanthology.org/W16-4412 | https://aclanthology.org/W16-4412.pdf | QAF: Frame Semantics-based Question Interpretation | Natural language questions are interpreted to a sequence of patterns to be matched with instances of patterns in a knowledge base (KB) for answering. A natural language (NL) question answering (QA) system utilizes meaningful patterns matching the syntac-tic/lexical features between the NL questions and KB. In the most of KBs, there are only binary relations in triple form to represent relation between two entities or entity and a value using the domain specific ontology. However, the binary relation representation is not enough to cover complex information in questions, and the ontology vocabulary sometimes does not cover the lexical meaning in questions. Complex meaning needs a knowledge representation to link the binary relation-type triples in KB. In this paper, we propose a frame semantics-based semantic parsing approach as KB-independent question pre-processing. We will propose requirements of question interpretation in the KBQA perspective, and a query form representation based on our proposed format QAF (Ques-tion Answering with the Frame Semantics), which is supposed to cover the requirements. In QAF, frame semantics roles as a model to represent complex information in questions and to disambiguate the lexical meaning in questions to match with the ontology vocabu-lary. Our system takes a question as an input and outputs QAF-query by the process which assigns semantic information in the question to its corresponding frame semantic structure using the semantic parsing rules. | ['Key-Sun Choi', 'Younggyun Hahm', 'Sangha Nam'] | 2016-12-01 | null | null | null | ws-2016-12 | ['knowledge-base-question-answering'] | ['natural-language-processing'] | [ 1.25340730e-01 6.04276836e-01 -1.13342963e-01 -8.33710372e-01
-4.02832270e-01 -5.09720266e-01 4.07592803e-01 4.74252701e-01
-1.61953136e-01 7.29685068e-01 4.80244547e-01 -4.98541355e-01
-5.83040118e-01 -1.57337689e+00 -5.90114951e-01 2.87834555e-01
6.57186866e-01 5.65212905e-01 1.27792859e+00 -1.09739220e+00
-4.01770361e-02 -7.87706152e-02 -1.95482743e+00 1.13407278e+00
7.89317667e-01 1.25995576e+00 5.74663222e-01 2.62213558e-01
-1.49555361e+00 1.39868319e+00 -5.09136200e-01 -4.47284937e-01
-2.65047789e-01 -6.09530628e-01 -1.86089802e+00 -9.58538875e-02
-2.46414803e-02 2.64955223e-01 3.65751795e-02 1.13885009e+00
-1.80904225e-01 1.06161349e-01 1.56643376e-01 -1.26154470e+00
-5.92022598e-01 5.46128094e-01 5.31098425e-01 1.28359675e-01
1.42337263e+00 -5.83279729e-01 1.13616920e+00 -4.14520860e-01
7.98087358e-01 1.80449033e+00 2.82748282e-01 6.19108558e-01
-4.09750730e-01 1.67359244e-02 7.98185021e-02 8.68062496e-01
-1.01560307e+00 -1.27703354e-01 4.06105638e-01 -1.77079514e-01
1.27021170e+00 6.37165785e-01 6.84925556e-01 1.53450638e-01
2.14036673e-01 2.03645051e-01 8.65841925e-01 -9.79620397e-01
2.80992627e-01 2.87257582e-01 1.07874990e+00 5.65094650e-01
3.07685643e-01 -6.07709706e-01 -4.31845248e-01 -2.75538981e-01
4.44988489e-01 -3.26774418e-01 -2.25121647e-01 8.59539136e-02
-6.38443589e-01 8.58404040e-01 1.51752561e-01 7.76923716e-01
-3.00681561e-01 -1.03403643e-01 3.41082722e-01 5.39695323e-01
-4.37385380e-01 3.73007864e-01 -6.60062253e-01 2.24533662e-01
9.52932835e-02 5.14995337e-01 1.29881930e+00 1.24264121e+00
1.06981039e+00 -5.25493979e-01 -1.15854166e-01 4.84082341e-01
8.68692517e-01 4.44251150e-01 7.70139754e-01 -1.00220680e+00
2.68783569e-01 1.70083976e+00 3.29088569e-01 -1.06405914e+00
-4.21948671e-01 3.66980016e-01 7.10271969e-02 -5.97655475e-01
4.00125951e-01 2.72361249e-01 -5.92894495e-01 1.65761149e+00
9.54296827e-01 -1.68569520e-01 7.81945765e-01 8.01870465e-01
1.65642405e+00 7.66360223e-01 2.89646894e-01 -2.15082780e-01
2.52898383e+00 -4.17495400e-01 -1.34562588e+00 -2.77445287e-01
8.78513157e-01 -8.83491576e-01 1.34057331e+00 -2.45212987e-01
-6.63132787e-01 -3.67087334e-01 -7.74632871e-01 -5.70439100e-01
-7.20905423e-01 -2.63016969e-01 3.73881221e-01 6.16699219e-01
-6.16855562e-01 -3.11473042e-01 -9.81850773e-02 -9.11326230e-01
-2.48175532e-01 2.21958116e-01 -3.79731774e-01 -2.85288841e-01
-1.92761016e+00 9.16169941e-01 1.27636933e+00 -2.31474236e-01
-1.90620750e-01 -3.45946670e-01 -1.18463850e+00 3.09664100e-01
9.85802948e-01 -9.47143793e-01 1.31104982e+00 -7.97311664e-01
-1.15525305e+00 8.21408510e-01 -6.10570550e-01 -5.68972826e-01
-4.06307191e-01 1.50041208e-01 -1.00277817e+00 4.30032521e-01
4.52426493e-01 1.98487923e-01 1.96776509e-01 -9.45422471e-01
-9.21238422e-01 -4.74058837e-01 1.00073242e+00 2.59294152e-01
1.10228062e-01 5.36844134e-01 -3.07526588e-01 2.99918912e-02
4.81131464e-01 -2.02698961e-01 2.23602355e-01 -5.40232718e-01
1.72173157e-01 -6.46685958e-01 9.14543092e-01 -5.94085157e-01
1.61387455e+00 -1.59921157e+00 -4.57272947e-01 2.45694548e-01
-1.00856371e-01 1.19414531e-01 1.88762188e-01 6.63868368e-01
-1.54890865e-03 1.93549335e-01 -2.42739860e-02 9.11685526e-01
2.39873454e-01 1.02620625e+00 -6.46272480e-01 -5.80616474e-01
3.13133784e-02 9.07786965e-01 -7.12386549e-01 -8.34045053e-01
-4.92243394e-02 -2.58247793e-01 -5.06636798e-01 3.95768613e-01
-1.09338164e+00 -8.94392952e-02 -9.98030424e-01 4.34994102e-01
4.68126655e-01 -2.76855409e-01 5.39399326e-01 -5.08544564e-01
2.30911747e-01 5.43968260e-01 -1.54086816e+00 1.44398534e+00
-5.47122300e-01 -1.64655253e-01 -7.52683580e-02 -8.95822048e-01
1.18521619e+00 6.66183889e-01 -5.03822193e-05 -6.57091856e-01
6.23876639e-02 4.67693746e-01 -1.41413838e-01 -1.35211110e+00
3.27207923e-01 -5.77075958e-01 -2.98819959e-01 4.12445329e-02
2.07983658e-01 -3.77755344e-01 4.22317594e-01 6.46690726e-02
8.87809753e-01 1.70629933e-01 7.65053749e-01 -4.11376894e-01
1.35732698e+00 6.90586209e-01 5.60875833e-01 4.42691654e-01
4.27902162e-01 -8.21915828e-03 6.88769400e-01 -6.91241264e-01
-5.92356324e-01 -6.08297765e-01 -2.39689782e-01 9.08610821e-01
4.92756188e-01 -7.32843876e-01 -8.05507958e-01 -6.27140045e-01
-3.74382526e-01 9.54902053e-01 -1.56186640e-01 1.55971036e-03
-6.26066804e-01 -2.55784631e-01 3.71011764e-01 1.32988682e-02
7.11068153e-01 -1.31553900e+00 -9.30663645e-01 4.38265800e-01
-5.48426569e-01 -1.41200471e+00 3.24071854e-01 -2.34721184e-01
-5.22092402e-01 -1.71604156e+00 3.59395951e-01 -1.09174883e+00
3.50078344e-01 4.85742539e-02 1.11637187e+00 3.91530693e-01
2.56168187e-01 5.81038654e-01 -1.02882743e+00 -6.27578497e-01
-5.66456318e-01 -2.10849494e-01 -4.77297843e-01 6.97181448e-02
1.02988791e+00 -9.44154561e-02 -1.74775213e-01 5.01596034e-01
-1.56626284e+00 -1.41486079e-01 -1.61666766e-01 3.91887188e-01
5.96869409e-01 4.59033139e-02 8.35926652e-01 -8.48985493e-01
8.84628415e-01 -6.07095838e-01 -4.44924623e-01 1.00556624e+00
-1.68674380e-01 3.55867743e-01 6.57960713e-01 2.11863458e-01
-1.50733078e+00 -3.93264264e-01 -3.42619091e-01 2.64972895e-01
-4.64799851e-01 9.22029018e-01 -9.30916786e-01 2.37258494e-01
8.23822021e-01 1.01427804e-03 -1.67567268e-01 -3.37562621e-01
3.56644273e-01 5.17662883e-01 3.28677773e-01 -1.16247392e+00
4.34612483e-01 1.89810619e-01 2.45072186e-01 -4.51458395e-01
-1.07414019e+00 -6.49436533e-01 -3.24830681e-01 8.87379574e-04
1.31303608e+00 -5.37844419e-01 -8.38263988e-01 -3.99782479e-01
-1.52404857e+00 3.52152884e-01 -7.94364393e-01 4.31741834e-01
-7.41191983e-01 5.69743335e-01 -1.11302085e-01 -8.28325629e-01
-1.27366915e-01 -8.43572915e-01 8.68869245e-01 4.36006546e-01
-1.16838329e-01 -8.40367377e-01 -8.84343237e-02 6.00581646e-01
5.68418764e-02 3.09019703e-02 1.63874102e+00 -1.12236583e+00
-2.93076366e-01 1.01681687e-02 -8.33168849e-02 9.64571014e-02
2.61683971e-01 -5.94642460e-01 -5.47470331e-01 6.77910089e-01
4.17652100e-01 -1.94092438e-01 9.92510282e-03 -1.15822963e-01
5.51191390e-01 -5.28903484e-01 6.01854809e-02 -1.87364832e-01
1.62878096e+00 7.72112966e-01 9.77850914e-01 4.96965975e-01
4.97380234e-02 1.20645928e+00 1.00343359e+00 2.58984463e-03
8.35634828e-01 5.71914315e-01 1.26598794e-02 5.72361290e-01
-5.32258004e-02 -4.70572382e-01 -1.69428706e-01 6.18711531e-01
1.51639253e-01 -1.91843901e-02 -1.17800975e+00 3.13337743e-01
-2.16410542e+00 -1.04710948e+00 -6.42788589e-01 1.82135868e+00
8.83310854e-01 -1.87330499e-01 -2.60957748e-01 1.57007173e-01
7.37348378e-01 -2.74154365e-01 -7.20863715e-02 -4.87947106e-01
-3.17455262e-01 3.65984261e-01 -2.03296170e-01 9.14133787e-01
-5.11142671e-01 1.24102712e+00 4.63352251e+00 9.23301637e-01
-4.09765273e-01 2.89045751e-01 -1.88906282e-01 7.87882090e-01
-8.78108621e-01 7.14885712e-01 -9.78426516e-01 -9.13753510e-02
7.97890902e-01 -6.28983736e-01 2.77899057e-01 5.91015279e-01
2.03431278e-01 -2.16338918e-01 -7.52046943e-01 7.92708993e-01
-7.73250451e-03 -1.58940136e+00 6.92993045e-01 -5.02577960e-01
1.38275037e-02 -9.13276374e-01 -8.92649710e-01 3.33378404e-01
-2.18709577e-02 -6.91628993e-01 6.84126258e-01 8.45361829e-01
4.19054806e-01 -4.08228964e-01 1.15367079e+00 5.18906116e-01
-1.68047714e+00 -2.68655062e-01 -8.89436781e-01 -1.63079441e-01
3.66479963e-01 2.23122939e-01 -8.15221012e-01 1.41386247e+00
8.07000816e-01 -6.63207425e-03 -3.38351071e-01 5.41302919e-01
-3.31147283e-01 2.87322134e-01 -4.94351447e-01 -2.07185790e-01
9.54758674e-02 -2.47083947e-01 2.41389349e-01 7.70421505e-01
5.03422260e-01 9.91003036e-01 1.61066517e-01 5.95705211e-01
1.31204516e-01 8.61066818e-01 -4.08450305e-01 2.53585547e-01
4.91656035e-01 6.42767131e-01 -5.56439340e-01 -7.17427909e-01
-6.41653359e-01 1.83402494e-01 -2.95980662e-01 2.60596633e-01
-5.76169431e-01 -4.86434549e-01 2.43888244e-01 4.06678736e-01
-1.06120773e-01 2.30068862e-01 1.24707364e-01 -1.22810423e+00
3.95492852e-01 -9.75508213e-01 9.18957412e-01 -1.20429206e+00
-9.87393916e-01 7.82986224e-01 6.06295407e-01 -9.09173429e-01
-2.99100250e-01 -5.51429868e-01 -3.91210645e-01 9.73675191e-01
-1.56349528e+00 -1.21418905e+00 -4.08810467e-01 1.08404052e+00
3.69399518e-01 1.51092112e-01 1.11367691e+00 1.43213794e-01
-2.16898117e-02 -2.54796296e-01 -9.55556750e-01 -1.83470279e-01
2.96163768e-01 -9.37292278e-01 -1.93130761e-01 6.49371386e-01
-6.86653331e-02 8.17248583e-01 6.92450643e-01 -8.57112169e-01
-1.41466618e+00 -6.67757571e-01 1.55896831e+00 -3.29601794e-01
6.22070789e-01 1.03893988e-01 -1.38496912e+00 9.28591311e-01
3.25682729e-01 -7.02205077e-02 9.63967621e-01 -3.66562158e-01
-3.34801525e-01 -1.40255913e-01 -1.46115804e+00 2.56560743e-01
9.00543571e-01 -5.92586517e-01 -1.89373624e+00 3.67673844e-01
1.48649395e+00 -2.89335668e-01 -1.09703314e+00 5.26794016e-01
2.22138301e-01 -7.89831996e-01 8.08315992e-01 -1.10486591e+00
-1.56935275e-01 -1.03108454e+00 -8.51026595e-01 -3.08532894e-01
1.65204957e-01 -9.30671990e-02 1.53562322e-01 1.18527293e+00
5.28914034e-01 -7.14774609e-01 4.78447616e-01 9.66332078e-01
-1.10595115e-01 -3.77031982e-01 -1.26906419e+00 -4.50828731e-01
-4.16741341e-01 -8.15555155e-01 1.38342357e+00 8.22816074e-01
3.16952080e-01 7.38429308e-01 3.02876323e-01 4.32656169e-01
-1.33841202e-01 3.99654418e-01 4.86051589e-01 -1.47110951e+00
-8.02445561e-02 1.49184778e-01 -4.56773400e-01 -1.03686988e+00
2.52836794e-01 -8.41731727e-01 -4.42821085e-01 -2.19235301e+00
-5.30390441e-01 -2.88359433e-01 3.75935465e-01 5.43879032e-01
3.86560075e-02 -6.13807082e-01 4.14302915e-01 1.38171494e-01
-5.47164559e-01 3.74721527e-01 1.30559206e+00 -5.87861724e-02
-2.73782045e-01 -2.48385474e-01 -6.88557267e-01 7.65731812e-01
5.42847931e-01 -4.96811777e-01 -8.68637800e-01 -1.96076229e-01
9.34767246e-01 6.03616476e-01 1.96402371e-01 -8.20016742e-01
4.17774439e-01 -7.63223708e-01 -4.15048897e-01 -2.25789681e-01
1.76126733e-01 -1.42239141e+00 3.32108587e-01 3.56059521e-01
-2.32532114e-01 8.63205269e-02 -5.92658706e-02 1.07138641e-01
-7.83521950e-01 -1.11892807e+00 3.67646962e-01 -4.17187423e-01
-1.15319633e+00 -2.35877886e-01 -3.07541072e-01 4.34218973e-01
9.10018921e-01 -3.11988741e-01 -3.77464682e-01 -3.17223638e-01
-9.55558717e-01 4.87543792e-01 6.94617480e-02 3.30301464e-01
7.12780714e-01 -1.19656372e+00 -1.79243430e-01 5.24169579e-02
4.20533180e-01 2.27527395e-01 4.08887208e-01 1.63486764e-01
-8.92943799e-01 6.72617435e-01 -1.56246737e-01 -1.87022805e-01
-9.64117944e-01 6.86070621e-01 6.35846376e-01 -3.62436116e-01
-1.80335209e-01 3.49989623e-01 1.15966931e-01 -6.32717192e-01
-1.92413375e-01 -7.94060230e-01 -1.15468931e+00 -1.12883123e-02
7.10225046e-01 -5.39339334e-02 8.66129249e-03 -9.50070143e-01
-4.31308568e-01 8.43954504e-01 6.81727469e-01 -1.29935861e-01
6.56006277e-01 -3.50955397e-01 -6.79535270e-01 2.03011990e-01
6.93331361e-01 8.52603465e-02 1.68763116e-01 -4.97390211e-01
6.06350422e-01 -2.94883311e-01 -5.57820380e-01 -5.94224691e-01
-2.30902389e-01 4.99598384e-01 3.14692169e-01 8.21918368e-01
1.09231079e+00 6.12987578e-01 9.66855824e-01 8.54339242e-01
7.41210997e-01 -8.51687670e-01 -2.14768127e-01 9.01840448e-01
1.19294703e+00 -5.84261358e-01 -3.96602929e-01 -1.10682380e+00
-4.72778529e-01 1.60793602e+00 7.66899407e-01 3.35232019e-01
6.67203188e-01 -1.42011922e-02 1.68073654e-01 -7.73968101e-01
-6.85044110e-01 -7.16051519e-01 1.05058491e-01 5.48319280e-01
8.45175982e-03 -1.96305722e-01 -1.05905497e+00 1.16316497e+00
-5.39704561e-01 3.00492704e-01 4.55849439e-01 1.19087076e+00
-1.29369211e+00 -1.55268121e+00 -8.97901416e-01 1.81767985e-01
-4.58075047e-01 1.14271320e-01 -3.07146430e-01 7.25180566e-01
6.68293238e-01 1.52679718e+00 3.17872092e-02 -2.17492387e-01
9.83371019e-01 5.65471232e-01 2.89520055e-01 -9.78573561e-01
-7.25863993e-01 -4.39866602e-01 5.71943521e-01 -4.94453490e-01
-8.62562358e-01 2.25182436e-02 -2.07352376e+00 1.20833963e-01
-2.18337491e-01 8.04908156e-01 3.15246046e-01 1.47202587e+00
1.68661729e-01 1.37220502e-01 1.09781481e-01 7.42671371e-01
-8.90460089e-02 -5.47824740e-01 -2.03270882e-01 6.33102536e-01
-2.29887143e-01 -5.53784966e-01 -5.21769151e-02 1.71379060e-01] | [10.415282249450684, 8.042506217956543] |
e21af7d8-49e0-4a3e-aedc-2457494fabfb | the-overview-of-the-nlm-chem-biocreative-vii | null | null | https://biocreative.bioinformatics.udel.edu/resources/publications/bc-vii-workshop-proceedings/ | https://biocreative.bioinformatics.udel.edu/media/store/files/2021/TRACK2_pos_01_BC7_submission_223.pdf | The overview of the NLM-Chem BioCreative VII track: full-text chemical identification and indexing in PubMed articles | The BioCreative NLM-Chem track calls for a community effort to fine-tune automated recognition of chemical names in biomedical literature. Chemical names are one of the most searched biomedical entities in PubMed and – as highlighted during the COVID-19 pandemic – their identification may significantly advance research in multiple biomedical subfields. While previous community challenges focused on identifying chemical names mentioned in titles and abstracts, the full text contains valuable additional detail. We organized the BioCreative NLM-Chem track to call for a community effort to address automated chemical entity recognition in full-text articles. The track consisted of two tasks: 1) Chemical Identification task, and 2) Chemical Indexing prediction task. For the Chemical Identification task, participants were expected to predict with high accuracy all chemicals mentioned in recently published full-text articles, both span (i.e., named entity recognition) and normalization (i.e., entity linking) using MeSH. For the Chemical Indexing task, participants identified which chemicals should be indexed as topics for the article's topic terms in the NLM article and indexing, i.e., appear in the listing of MeSH terms for the document. | ['Zhiyong Lu', 'Rezarta Islamaj', 'Robert Leaman'] | 2021-11-08 | null | null | null | biocreative-vii-challenge-evaluation-workshop | ['chemical-entity-recognition', 'chemical-indexing'] | ['medical', 'natural-language-processing'] | [ 3.86247426e-01 2.66712129e-01 -6.92600310e-01 1.01229055e-02
-9.56561685e-01 -9.49132979e-01 4.56324875e-01 1.06112564e+00
-4.98316914e-01 1.28710687e+00 3.73917788e-01 -4.49493349e-01
-2.38821268e-01 -4.96493518e-01 -8.68795335e-01 -5.40362477e-01
7.33790025e-02 7.16099918e-01 -3.11493039e-01 4.23927367e-01
5.31748414e-01 7.31511652e-01 -9.94590998e-01 1.15596369e-01
1.02654922e+00 4.60590690e-01 9.13668200e-02 1.31706804e-01
-5.26299894e-01 4.49331999e-01 -6.41823769e-01 -3.43637496e-01
-3.98036867e-01 -3.50038320e-01 -9.34035420e-01 -6.84639394e-01
2.05969200e-01 6.67966127e-01 3.14774294e-03 1.16002166e+00
5.41434288e-01 -8.84916484e-02 9.87944365e-01 -1.06021821e+00
-5.24400353e-01 7.77158499e-01 -3.46744388e-01 1.97418213e-01
7.35599101e-01 -1.53993815e-01 1.09302449e+00 -1.06536424e+00
1.26357496e+00 1.00459552e+00 7.00908124e-01 4.88799453e-01
-1.25280643e+00 -1.04680634e+00 -1.06073551e-01 -9.44518447e-02
-1.63958490e+00 -4.81361359e-01 6.04849160e-02 -9.04430389e-01
1.33900583e+00 3.43601823e-01 3.25214922e-01 9.96493697e-01
5.26945829e-01 2.06574649e-01 6.20061517e-01 -1.46431029e-01
3.14506829e-01 2.85171837e-01 8.85435864e-02 4.03800905e-01
8.99812460e-01 -1.80279896e-01 -5.60700774e-01 -7.29583979e-01
3.64727303e-02 -7.44557083e-02 -3.20725530e-01 2.44817242e-01
-1.46278036e+00 5.58150113e-01 2.07329810e-01 5.33055723e-01
-7.33448267e-01 -3.22654024e-02 5.43829739e-01 -3.08480173e-01
4.91062820e-01 1.33813190e+00 -8.28248739e-01 2.85111636e-01
-1.04899204e+00 1.47556484e-01 1.07794714e+00 1.13137662e+00
5.27966321e-01 -6.85077369e-01 -5.36042899e-02 7.05969155e-01
3.93095106e-01 2.56724894e-01 4.99588072e-01 -3.74772310e-01
2.16812670e-01 7.46306598e-01 1.29591227e-01 -7.82437444e-01
-8.53111088e-01 -2.16094822e-01 -5.89879453e-01 -6.09945238e-01
1.85521796e-01 -3.14436436e-01 -9.56899047e-01 1.49453866e+00
3.55564594e-01 1.50883093e-01 2.82082856e-01 3.28435928e-01
1.27291548e+00 5.14961720e-01 1.06056297e+00 -5.23893297e-01
1.91854632e+00 -2.97465771e-01 -1.01927972e+00 1.53561428e-01
8.79055202e-01 -1.05270576e+00 -4.22480106e-02 1.28836185e-01
-8.22014093e-01 -3.67610976e-02 -1.10102642e+00 -7.25237727e-02
-1.32309937e+00 -1.52460173e-01 6.29768193e-01 2.01373160e-01
-7.27595508e-01 6.24856532e-01 -5.61596632e-01 -6.74639523e-01
4.23877269e-01 4.45286572e-01 -6.19857311e-01 -1.48115188e-01
-1.61856794e+00 1.34762537e+00 6.33187115e-01 -2.46761084e-01
-4.83037770e-01 -1.34743547e+00 -7.40860522e-01 -1.55672655e-01
2.98630923e-01 -8.87486517e-01 9.27724004e-01 4.05179448e-02
-4.81587529e-01 1.18742979e+00 -3.50313544e-01 -3.10239166e-01
-1.91543788e-01 -7.31597468e-02 -9.08616304e-01 2.21298039e-01
8.39991748e-01 8.28463674e-01 6.06751144e-02 -7.42525041e-01
-8.30052853e-01 -2.57632881e-01 -4.17237878e-01 3.02063048e-01
-1.92852959e-01 2.58860558e-01 -3.06788653e-01 -7.87264705e-01
-1.37378976e-01 -8.97893727e-01 -4.31200713e-01 -3.65176708e-01
-6.62270665e-01 -5.40247798e-01 3.04150701e-01 -5.20838261e-01
1.23819590e+00 -1.81820846e+00 -5.55077046e-02 2.67858893e-01
4.98093247e-01 -2.14362234e-01 2.24027231e-01 7.64951408e-01
-5.19926965e-01 9.08506989e-01 1.37162387e-01 4.02987242e-01
-1.40626714e-01 -3.25641334e-01 -2.39674717e-01 6.42694294e-01
2.55768448e-01 9.23422813e-01 -1.15482724e+00 -6.77935004e-01
-4.20595080e-01 4.44758475e-01 -1.43602759e-01 -2.06543714e-01
-3.22478801e-01 3.50192189e-01 -8.86705041e-01 1.00280094e+00
3.73995572e-01 -4.75406736e-01 1.72529995e-01 -5.82535028e-01
-4.75297123e-01 5.64674258e-01 -8.33097279e-01 1.41603899e+00
-2.29262337e-02 2.02947468e-01 -8.50759000e-02 -5.14657557e-01
5.97613633e-01 8.53182614e-01 1.01906860e+00 -2.81823635e-01
-1.86586872e-01 6.93331242e-01 -3.82256247e-02 -2.86324918e-01
1.81008682e-01 -1.22034736e-01 -8.61101896e-02 1.13670915e-01
-4.94795665e-03 1.96476519e-01 4.58321840e-01 2.59933591e-01
1.21732771e+00 -6.54906854e-02 6.80389941e-01 -6.84768081e-01
7.43854821e-01 6.19789422e-01 4.94029552e-01 5.64157248e-01
1.96349964e-01 -1.07921600e-01 2.33063236e-01 -1.40488178e-01
-6.32127583e-01 -4.44225639e-01 -6.63113236e-01 7.36175954e-01
-1.45058230e-01 -5.24091542e-01 -5.29847383e-01 -3.81089181e-01
1.67351604e-01 5.22165954e-01 -7.13311017e-01 -7.25927278e-02
-3.05016249e-01 -1.02481019e+00 8.35500658e-01 1.09074481e-01
-3.63098979e-02 -1.08269382e+00 -1.26645803e-01 5.10498166e-01
-1.13484077e-01 -9.80366886e-01 -7.73839831e-01 7.06966579e-01
-5.27540386e-01 -1.40980995e+00 -1.09340382e+00 -9.03499305e-01
6.32270575e-01 -3.32376778e-01 9.90852833e-01 -1.85598582e-01
-6.61044002e-01 2.15054810e-01 -4.33182903e-02 -9.95697796e-01
-4.12084758e-01 2.25484222e-01 8.54480788e-02 -5.81676960e-01
7.18440890e-01 -1.24019653e-01 -7.92554259e-01 5.53573333e-02
-6.69827878e-01 -1.44232750e-01 6.44551873e-01 5.08029521e-01
9.39480543e-01 -1.62521541e-01 7.44016886e-01 -1.19092178e+00
5.19244730e-01 -8.77817094e-01 -6.09219074e-01 4.13303614e-01
-1.03358448e+00 1.66701496e-01 8.42851400e-02 -4.47487742e-01
-4.12027061e-01 3.75763714e-01 -2.94048876e-01 2.81660497e-01
-3.46173346e-01 1.21345913e+00 -3.83725286e-01 7.18040764e-02
6.20751023e-01 -9.97136980e-02 -5.72952211e-01 -6.15672290e-01
2.46642560e-01 5.68939209e-01 3.86374474e-01 -4.14298922e-01
5.30408680e-01 -1.60006151e-01 4.11612391e-01 -6.28915668e-01
-6.85550511e-01 -8.92831624e-01 -3.66798878e-01 3.35229993e-01
1.42732489e+00 -1.12093747e+00 -1.05167425e+00 -5.82394302e-02
-1.27778947e+00 3.61619353e-01 1.88979775e-01 7.55772889e-01
1.24986254e-01 8.91203061e-02 -5.55247247e-01 -3.74477714e-01
-7.13327467e-01 -1.07711804e+00 9.20947433e-01 2.26690903e-01
-9.15498137e-01 -1.06292605e+00 3.02935898e-01 1.12939358e-01
-1.38648646e-02 4.98932928e-01 1.56175590e+00 -1.46493185e+00
-3.36529501e-02 -2.98853070e-01 -1.06791578e-01 -7.68254399e-01
3.98810118e-01 -9.79493186e-03 -7.71991730e-01 2.36812875e-01
-4.44738716e-01 2.74969310e-01 6.45298243e-01 4.28177983e-01
8.12061965e-01 -4.35990691e-01 -1.24867010e+00 2.81090856e-01
1.18887174e+00 8.43416512e-01 3.36169630e-01 4.69693542e-01
7.55704105e-01 6.59685373e-01 3.47460628e-01 1.76485050e-02
-2.30054744e-03 4.88244802e-01 -3.65200453e-02 -3.46672833e-01
1.23867102e-01 -2.85965174e-01 -2.04378814e-01 2.72171170e-01
3.35149884e-01 -2.99392283e-01 -1.25367117e+00 5.52493870e-01
-1.31008148e+00 -6.46102965e-01 -4.89625067e-01 2.06238914e+00
1.49316537e+00 -5.12553453e-02 -5.03039993e-02 -3.66387725e-01
8.47312748e-01 -4.78641540e-01 -7.97244668e-01 -1.01606727e-01
-1.68589041e-01 4.20293599e-01 8.48752856e-01 3.93684357e-01
-1.14910805e+00 8.46653938e-01 6.47265482e+00 7.26854265e-01
-6.90550506e-01 -1.63439870e-01 6.15829527e-01 2.88698822e-01
-1.44683838e-01 3.18329670e-02 -1.37813282e+00 3.91522676e-01
1.18168998e+00 -7.91699350e-01 -4.77457345e-02 5.95340133e-01
2.60050327e-01 -2.62113456e-02 -1.47460830e+00 7.67998040e-01
-2.38380983e-01 -1.80544043e+00 2.30709061e-01 3.43953073e-01
5.49019217e-01 -4.99569513e-02 -2.27998570e-01 2.37617432e-03
2.03438461e-01 -1.23743403e+00 3.04735750e-01 5.70607543e-01
1.11026978e+00 -4.97176290e-01 6.63073719e-01 -4.76791933e-02
-1.21697235e+00 3.94878358e-01 -1.17088519e-01 7.95430779e-01
3.33757587e-02 9.56987977e-01 -9.93417680e-01 6.89964771e-01
5.12316823e-01 9.70743835e-01 -3.98920834e-01 1.47864962e+00
-9.98485163e-02 4.11372751e-01 8.57370421e-02 -2.74411529e-01
3.85755450e-02 1.99347287e-01 4.71767157e-01 1.57450259e+00
2.08070695e-01 2.97467679e-01 2.58906811e-01 8.06810081e-01
-5.61399102e-01 5.93605280e-01 -4.25585270e-01 -9.90446806e-01
8.94956112e-01 1.10927570e+00 -9.85590219e-01 -4.80789572e-01
-1.69790357e-01 5.87144196e-01 -3.88911963e-01 2.38074094e-01
-3.39599192e-01 -7.89435744e-01 6.16665483e-01 5.85501008e-02
-6.38364181e-02 4.64729548e-01 -2.01363951e-01 -6.25497937e-01
-7.07997620e-01 -8.80306423e-01 7.23252177e-01 -6.41724765e-01
-1.44895482e+00 5.40384829e-01 6.66317716e-02 -1.01368463e+00
6.96880296e-02 -6.26181543e-01 -8.06377921e-03 1.24719608e+00
-1.00450766e+00 -8.67214262e-01 3.64558637e-01 2.19975904e-01
1.81002289e-01 1.25552177e-01 1.31847215e+00 7.30981946e-01
-4.87156212e-01 3.97881031e-01 2.17407435e-01 5.76644018e-02
1.38372302e+00 -1.17395639e+00 2.90940642e-01 -1.55822588e-02
-2.01047406e-01 1.22005391e+00 9.17227030e-01 -1.42518651e+00
-1.20238924e+00 -1.19464123e+00 1.62931287e+00 -5.86899757e-01
9.19362187e-01 -2.92653680e-01 -7.90072322e-01 3.29533130e-01
1.00396082e-01 -6.19158268e-01 1.37082982e+00 3.63935642e-02
-3.08648348e-01 6.09176099e-01 -1.08015764e+00 5.26471436e-01
6.62918270e-01 -4.18187827e-01 -6.70769632e-01 1.03214216e+00
9.52063382e-01 -4.72841948e-01 -1.48736274e+00 2.04482645e-01
4.49626267e-01 7.63575375e-01 1.00084662e+00 -8.99276793e-01
2.46664122e-01 -5.19248426e-01 2.36393601e-01 -9.39986467e-01
-4.60089147e-01 -5.02702713e-01 1.05842657e-01 1.15769136e+00
1.15124190e+00 -6.67015851e-01 4.28248197e-01 7.18161702e-01
-6.93312883e-02 -6.07952416e-01 -7.24625587e-01 -3.79500717e-01
1.52781904e-01 2.22043559e-01 4.39393908e-01 1.19350004e+00
4.62934464e-01 7.61347234e-01 8.52069706e-02 1.47884309e-01
4.23194587e-01 -1.84192404e-01 -1.36533096e-01 -1.67960334e+00
2.47949973e-01 -6.46129966e-01 -2.05887914e-01 -4.05886114e-01
2.45673787e-02 -1.18451273e+00 -2.22667083e-02 -1.98907351e+00
5.16445577e-01 2.01938096e-02 -4.76586193e-01 8.67672265e-01
-3.22788537e-01 -1.01465121e-01 -4.92487460e-01 3.50452304e-01
-5.27951896e-01 -2.82508552e-01 6.75667226e-01 -4.50291246e-01
-2.31068954e-01 -2.78669000e-01 -1.19938576e+00 3.57848287e-01
2.76544452e-01 -1.06426442e+00 9.84506309e-02 3.79761159e-01
6.83689356e-01 -5.69928437e-02 -8.79670680e-02 -6.01965845e-01
5.32468915e-01 -5.54888666e-01 7.23169446e-01 -8.91948760e-01
-1.77845024e-02 -6.34717047e-01 7.32020617e-01 8.71878445e-01
-7.42435098e-01 7.48371184e-02 6.31744802e-01 5.08633256e-01
7.04225600e-02 -3.18611443e-01 4.59561616e-01 -4.02164787e-01
-2.75531143e-01 3.29053581e-01 -8.29557836e-01 1.04891300e-01
7.39738405e-01 -4.10692468e-02 -5.52090049e-01 1.69923723e-01
-9.16863501e-01 3.71110111e-01 2.40119800e-01 3.62051934e-01
2.35626474e-01 -1.05525947e+00 -6.64781272e-01 -6.78145528e-01
4.21787411e-01 -3.65010172e-01 -1.05218105e-01 9.56062555e-01
-2.52915621e-01 9.79014993e-01 1.83725372e-01 -8.75024777e-03
-1.23974597e+00 9.37010467e-01 2.14560255e-01 -4.91629094e-01
-2.49616653e-01 8.23374629e-01 4.10005987e-01 -2.45865002e-01
4.71424729e-01 -1.05582796e-01 -6.75191641e-01 4.98624057e-01
7.18245029e-01 7.38515705e-03 3.44881952e-01 -6.73711360e-01
-1.05526185e+00 3.74928087e-01 -3.54363292e-01 1.46826074e-01
1.42130518e+00 3.83221567e-01 -5.36116421e-01 3.96556109e-01
1.45770609e+00 1.05409913e-01 1.69207416e-02 4.75643910e-02
7.77377486e-01 4.86465216e-01 -2.08054006e-01 -1.46046746e+00
-4.24695283e-01 2.90755749e-01 5.58992505e-01 -3.25436056e-01
5.71631074e-01 1.43390864e-01 4.32199150e-01 5.25175989e-01
-6.08017296e-02 -8.62660646e-01 -5.15515685e-01 4.56820041e-01
8.40961158e-01 -8.49006057e-01 6.44614637e-01 -5.58922887e-01
-1.88537553e-01 9.70609248e-01 1.77014232e-01 7.68881500e-01
7.30047286e-01 2.20348790e-01 -1.48591802e-01 -9.11141098e-01
-5.88228524e-01 7.85427392e-02 7.76744127e-01 3.30205142e-01
9.96918976e-01 -5.71175814e-02 -8.62426758e-01 8.32564533e-01
2.79361755e-01 -5.52641600e-02 1.81427458e-03 9.46021795e-01
-3.08250785e-01 -1.25839043e+00 -3.94469291e-01 7.14639366e-01
-1.10450041e+00 -8.27224016e-01 -9.55326200e-01 4.97745275e-01
1.37179956e-01 1.05579948e+00 -5.04590452e-01 -9.76735577e-02
4.36064750e-01 3.89551342e-01 -2.23864373e-02 -9.29741859e-01
-9.27124560e-01 4.68618333e-01 3.32502395e-01 -1.79648638e-01
-4.94368255e-01 -6.63847208e-01 -1.65956795e+00 1.56363666e-01
-4.57923532e-01 8.74681294e-01 8.58015358e-01 8.84660840e-01
5.72793067e-01 5.35846233e-01 -5.46778478e-02 -2.64425635e-01
8.60737935e-02 -8.52648079e-01 -3.75529021e-01 -6.95615858e-02
-1.18186690e-01 -5.53001881e-01 -1.82485253e-01 4.29355502e-01] | [8.486650466918945, 8.729165077209473] |
5b180a27-c231-4b18-aaa2-91c220d66f22 | mirrornet-bio-inspired-adversarial-attack-for-1 | 2007.12881 | null | https://arxiv.org/abs/2007.12881v3 | https://arxiv.org/pdf/2007.12881v3.pdf | MirrorNet: Bio-Inspired Camouflaged Object Segmentation | Camouflaged objects are generally difficult to be detected in their natural environment even for human beings. In this paper, we propose a novel bio-inspired network, named the MirrorNet, that leverages both instance segmentation and mirror stream for the camouflaged object segmentation. Differently from existing networks for segmentation, our proposed network possesses two segmentation streams: the main stream and the mirror stream corresponding with the original image and its flipped image, respectively. The output from the mirror stream is then fused into the main stream's result for the final camouflage map to boost up the segmentation accuracy. Extensive experiments conducted on the public CAMO dataset demonstrate the effectiveness of our proposed network. Our proposed method achieves 89% in accuracy, outperforming the state-of-the-arts. Project Page: https://sites.google.com/view/ltnghia/research/camo | ['Tam V. Nguyen', 'Thanh-Toan Do', 'Minh-Triet Tran', 'Khanh-Duy Nguyen', 'Trung-Nghia Le', 'Jinnan Yan'] | 2020-07-25 | mirrornet-bio-inspired-adversarial-attack-for | null | null | pattern-recognition-journal-2020-7 | ['camouflage-segmentation', 'camouflaged-object-segmentation'] | ['computer-vision', 'computer-vision'] | [ 4.92977083e-01 1.60697788e-01 -2.07672983e-01 7.16653392e-02
-2.65114158e-01 -6.48012459e-01 4.16453272e-01 -3.82486135e-01
-3.37522268e-01 5.55005848e-01 -9.23150256e-02 -2.00929418e-01
5.37003994e-01 -7.90327251e-01 -8.12146723e-01 -5.67479312e-01
4.06866044e-01 6.65431693e-02 6.24202907e-01 4.56784479e-02
3.13732505e-01 2.50469416e-01 -1.19466126e+00 3.05532198e-02
1.08848464e+00 1.19715238e+00 2.20234469e-01 5.79058111e-01
-1.44506559e-01 5.38891494e-01 -5.90802252e-01 -2.96519399e-01
7.57169664e-01 -5.30481458e-01 -4.23302054e-01 7.21280053e-02
6.04079843e-01 -4.95451361e-01 -5.50853491e-01 1.31214833e+00
2.16884986e-01 3.29934731e-02 3.37221354e-01 -1.33192611e+00
-7.96374202e-01 5.04033506e-01 -6.55525327e-01 3.06216419e-01
-7.46748969e-02 5.34042418e-01 6.64747775e-01 -7.41678596e-01
7.59946406e-01 1.02801979e+00 4.40193534e-01 7.63064802e-01
-9.19598579e-01 -1.16920078e+00 2.23089635e-01 4.88441903e-03
-1.05342984e+00 -3.63800049e-01 8.23460042e-01 -3.05141747e-01
3.57541710e-01 1.24359727e-01 9.79578316e-01 7.54601061e-01
3.01490296e-02 1.09443939e+00 1.00604963e+00 3.27941105e-02
2.06339329e-01 -3.67269486e-01 -8.09939429e-02 7.32273221e-01
5.26899040e-01 1.34044200e-01 -3.68690610e-01 1.16342574e-01
9.83609557e-01 2.20204115e-01 -4.81172562e-01 -1.37521401e-01
-1.10307729e+00 5.26757956e-01 7.92754292e-01 8.96387175e-02
-3.57856005e-01 4.23828483e-01 2.89501667e-01 8.83248262e-03
5.09041429e-01 4.73303705e-01 -3.17585766e-01 2.54208595e-02
-1.04440570e+00 2.74712652e-01 6.99221373e-01 9.90169048e-01
6.17110193e-01 4.93679978e-02 -2.31983617e-01 6.56982481e-01
1.45707175e-01 3.83077919e-01 3.11137795e-01 -1.07204306e+00
3.32347244e-01 7.64021575e-01 1.30445793e-01 -9.62508440e-01
-5.70870787e-02 -3.18489075e-01 -8.58038723e-01 3.81580107e-02
5.81079483e-01 -3.10651422e-01 -1.40822840e+00 1.47391319e+00
5.63198745e-01 8.50610435e-01 -2.76015669e-01 1.11838245e+00
8.10091972e-01 6.84414923e-01 -1.43297344e-01 2.10239157e-01
1.17226398e+00 -1.52254343e+00 -5.80159843e-01 -3.76414269e-01
4.72935438e-01 -6.90307200e-01 7.83719182e-01 4.65449184e-01
-1.00108552e+00 -4.04407471e-01 -1.22352695e+00 -7.75142908e-02
-1.95750743e-01 4.96466048e-02 6.55738890e-01 4.43904400e-01
-8.01177323e-01 6.86312139e-01 -8.47087145e-01 -2.67092362e-02
1.11299658e+00 2.51752645e-01 -1.67328957e-02 -1.69035032e-01
-7.93698132e-01 3.71041238e-01 4.25524294e-01 1.51902497e-01
-1.16062593e+00 -7.94448018e-01 -5.83850086e-01 2.71754209e-02
7.65634418e-01 -5.84642231e-01 1.26944387e+00 -1.03303421e+00
-1.36742139e+00 7.11513877e-01 -7.18327165e-02 -6.58539295e-01
7.03919590e-01 -3.96533370e-01 -1.01553693e-01 3.77992868e-01
1.18783735e-01 1.09161139e+00 1.07185447e+00 -1.30494285e+00
-1.01394987e+00 -1.48157522e-01 -1.04453422e-01 -5.03919721e-02
-8.79843310e-02 -3.59011084e-01 -8.01345587e-01 -7.72203743e-01
6.83810115e-02 -9.81095493e-01 -2.91554481e-01 4.03259575e-01
-8.06606472e-01 -2.43337220e-03 1.22506070e+00 -6.52394772e-01
9.60237741e-01 -2.26200557e+00 -2.06444100e-01 8.33033398e-02
4.66052920e-01 6.54054046e-01 -2.72665888e-01 -7.49466754e-03
9.12547112e-02 2.07776234e-01 -3.62880260e-01 -3.14100057e-01
-2.34418288e-01 -4.42094803e-02 -3.80245060e-01 7.10442841e-01
1.55409172e-01 1.24953353e+00 -1.01187897e+00 -4.13149536e-01
2.96609253e-01 3.93886536e-01 -3.95056695e-01 5.72728589e-02
-3.96488011e-01 6.95925176e-01 -4.59526777e-01 8.83867443e-01
9.11482036e-01 -2.68346041e-01 -6.00688159e-02 2.25493815e-02
-7.26186186e-02 1.27754929e-02 -7.80091465e-01 1.64919364e+00
1.08870894e-01 6.37270927e-01 2.68861204e-01 -6.02358997e-01
8.53264868e-01 6.05926923e-02 4.30905074e-01 -8.26609015e-01
5.69239497e-01 2.71179497e-01 1.04014359e-01 -1.56154469e-01
5.18052220e-01 2.41907969e-01 1.07512087e-01 3.77124220e-01
-1.06039420e-01 -1.84104875e-01 3.75091463e-01 2.41083488e-01
9.92962837e-01 2.49312282e-01 -1.67371138e-04 -1.25033468e-01
3.51035297e-01 1.48412421e-01 8.52823973e-01 6.69242978e-01
-6.46938860e-01 6.22607827e-01 5.04514933e-01 -2.76087254e-01
-1.09523082e+00 -9.14866328e-01 2.92835116e-01 6.55649066e-01
8.46575737e-01 -9.29429755e-02 -1.12909353e+00 -7.72184730e-01
2.58599948e-02 4.81216550e-01 -5.30480564e-01 -1.71246678e-01
-6.69344008e-01 -2.42231444e-01 6.73081994e-01 3.91816497e-01
1.02264261e+00 -1.19378066e+00 -6.86098695e-01 -2.88252514e-02
-1.24441549e-01 -1.11294234e+00 -8.97368371e-01 -3.35311711e-01
-7.47680187e-01 -1.08147728e+00 -9.36035216e-01 -7.83675432e-01
7.53653347e-01 6.27630472e-01 6.57683611e-01 3.69215667e-01
-2.14565650e-01 -1.20856106e-01 -3.64945173e-01 -5.05530953e-01
-3.30700517e-01 3.31474990e-02 -4.27451968e-01 1.09667085e-01
-2.22143680e-02 -6.31013751e-01 -1.15339148e+00 3.40027720e-01
-1.15836680e+00 2.32289314e-01 5.76337934e-01 6.73676312e-01
7.37863481e-01 -1.61404580e-01 4.58928734e-01 -9.67668712e-01
3.47616166e-01 -4.19591308e-01 -8.65646839e-01 -1.86679155e-01
-4.57026303e-01 -4.07682925e-01 6.11756921e-01 -6.41802549e-01
-7.18441606e-01 1.72983631e-01 3.05768140e-02 -7.09937572e-01
-2.59195864e-01 -1.69259775e-02 -4.61672917e-02 -2.32594132e-01
3.10530961e-01 1.12146303e-01 6.63902611e-02 -4.70975965e-01
5.31135142e-01 4.35310870e-01 9.39079404e-01 -1.81396320e-01
1.04121268e+00 9.61068451e-01 -1.67799577e-01 -6.28789604e-01
-8.50179613e-01 -4.25030589e-01 -3.67245942e-01 -3.89145881e-01
8.98082018e-01 -7.10487962e-01 -6.86793923e-01 9.22744513e-01
-1.21432424e+00 -5.37470281e-01 -3.89288902e-01 -2.74878982e-02
-2.50145674e-01 2.73310691e-01 -6.84871256e-01 -5.95177829e-01
-6.43092632e-01 -1.01792836e+00 1.01500189e+00 8.69690955e-01
1.05387732e-01 -7.11280763e-01 -1.83082476e-01 6.30098522e-01
3.39185059e-01 4.21220839e-01 5.50538361e-01 -6.83311284e-01
-1.00601637e+00 -1.26952857e-01 -5.53587556e-01 3.87815684e-01
1.30200922e-01 -4.18985588e-03 -9.03638661e-01 -2.51741320e-01
-1.98505506e-01 -1.01585448e-01 1.15640986e+00 4.59220469e-01
1.32316792e+00 -1.66514069e-01 -4.49119240e-01 7.37651944e-01
1.34611654e+00 3.19395185e-01 7.91011333e-01 2.05899090e-01
1.01828623e+00 1.51865929e-01 6.95794523e-01 2.21703336e-01
1.72364160e-01 2.85664380e-01 7.42001176e-01 -3.42400670e-01
-4.52366054e-01 -4.76730883e-01 1.94489166e-01 6.43269002e-01
1.79139256e-01 -5.54279625e-01 -6.35459304e-01 7.82605052e-01
-1.78473425e+00 -6.53562963e-01 -2.28117615e-01 2.02938437e+00
4.87248182e-01 2.89543390e-01 9.35015734e-03 -1.55313671e-01
9.94339585e-01 5.04441738e-01 -1.01116014e+00 -1.34406492e-01
-4.90621924e-02 2.86308825e-01 8.30134213e-01 1.98616654e-01
-1.15563011e+00 1.14485407e+00 5.53279877e+00 1.10053658e+00
-1.22374165e+00 2.67546862e-01 7.11897075e-01 -1.42578751e-01
-1.31121293e-01 8.09473395e-02 -5.43710291e-01 6.76307559e-01
5.95391095e-01 -2.37035891e-03 4.07196373e-01 7.03475952e-01
2.74798781e-01 -3.83404613e-01 -5.71187019e-01 7.30782330e-01
-5.80157787e-02 -1.67867827e+00 -6.17552474e-02 2.17508093e-01
1.05028713e+00 1.99733421e-01 2.59591043e-01 -5.22269569e-02
2.76334882e-01 -9.11448359e-01 7.99040675e-01 2.16094956e-01
8.03502738e-01 -6.95940912e-01 4.65271503e-01 1.64522305e-01
-1.17450929e+00 -1.44874882e-02 -3.41360942e-02 1.24983154e-02
3.53243679e-01 7.89524853e-01 -7.13805676e-01 3.51790637e-01
6.16194904e-01 8.46378803e-01 -4.48842794e-01 1.31152439e+00
-3.15697074e-01 8.62020433e-01 -3.00594985e-01 1.37481272e-01
3.90091360e-01 -3.66046339e-01 8.45360994e-01 8.64070833e-01
7.56571442e-02 -5.79129644e-02 2.97838271e-01 1.10660827e+00
-6.26117349e-01 -1.83096930e-01 -4.28186208e-01 -2.47407123e-01
6.24596000e-01 1.43649805e+00 -1.24706686e+00 -5.75214446e-01
-2.71836907e-01 1.15489280e+00 6.80190846e-02 3.52669060e-01
-9.38314140e-01 -5.43168187e-01 6.87302768e-01 -7.20512420e-02
6.62587523e-01 -1.13195151e-01 -5.59809029e-01 -1.12945986e+00
-2.62081865e-02 -7.41275549e-01 1.35128438e-01 -6.96927965e-01
-1.12855494e+00 3.03067058e-01 -3.74403745e-01 -1.15348470e+00
4.62294221e-01 -5.05506933e-01 -8.59114170e-01 5.63685060e-01
-1.44617295e+00 -1.09545863e+00 -4.54955220e-01 2.89055169e-01
6.62707865e-01 6.97069392e-02 6.92211688e-02 3.13734770e-01
-7.86593556e-01 3.58743787e-01 8.25693607e-02 3.48120600e-01
5.06662786e-01 -9.40944195e-01 8.89329433e-01 1.12399364e+00
-7.00053722e-02 2.65891403e-01 4.51249659e-01 -1.05500698e+00
-1.13109636e+00 -1.32036817e+00 2.11817399e-01 -2.81755831e-02
6.04019046e-01 -4.20298696e-01 -9.61317062e-01 4.42797780e-01
8.24300423e-02 1.45306736e-01 1.32348180e-01 -7.39442885e-01
-3.94217014e-01 8.20245594e-02 -1.34012544e+00 7.78142154e-01
1.26957619e+00 -8.12543854e-02 -2.63672858e-01 2.32120574e-01
1.22484934e+00 -6.77737296e-01 -5.24022520e-01 2.12426841e-01
6.89038515e-01 -8.20457935e-01 7.39169419e-01 -2.36746758e-01
8.34020019e-01 -4.53078061e-01 1.45344600e-01 -1.04635584e+00
-8.93111974e-02 -7.34566987e-01 -1.41137719e-01 1.12262642e+00
2.09273443e-01 -8.51994216e-01 1.13817680e+00 2.76409000e-01
-3.36440533e-01 -9.91389155e-01 -9.04711008e-01 -8.21988404e-01
-1.75991841e-02 -3.27854335e-01 7.26458907e-01 7.48854458e-01
-6.08832717e-01 4.63126823e-02 -3.02931666e-01 -5.95097942e-03
6.36441171e-01 2.75367320e-01 7.36248672e-01 -9.77888286e-01
-1.39645383e-01 -5.13006330e-01 -3.24794471e-01 -1.42990685e+00
1.73491053e-02 -9.20822501e-01 2.57843137e-01 -1.28325701e+00
1.62352756e-01 -3.44327539e-01 -7.40782693e-02 4.90396082e-01
-1.24781035e-01 7.98460126e-01 6.22668982e-01 2.39839837e-01
-5.83014727e-01 5.07202327e-01 1.90526485e+00 -1.90185219e-01
-3.01606148e-01 -1.73519224e-01 -8.82519782e-01 8.90805066e-01
1.05530465e+00 -6.60793006e-01 -1.82690293e-01 -4.39891428e-01
-2.70106733e-01 -2.45365500e-01 5.69945276e-01 -9.58788574e-01
1.47834510e-01 -1.91725776e-01 3.20742249e-01 -5.98822594e-01
2.71058738e-01 -5.48965812e-01 -1.48872221e-02 7.80022264e-01
-1.33401349e-01 -2.05909014e-01 1.36119351e-01 6.85342669e-01
4.47520986e-02 -9.87756811e-03 1.01057959e+00 -9.68053117e-02
-7.53952861e-01 5.49471498e-01 -2.11555034e-01 1.44151270e-01
1.12456083e+00 -3.21105659e-01 -8.79665613e-01 -2.62083173e-01
-2.08918974e-01 4.49593991e-01 6.89660251e-01 4.22593325e-01
6.46945417e-01 -1.09155393e+00 -4.50688422e-01 9.94926170e-02
-1.42113850e-01 4.34937179e-01 1.82615131e-01 9.03025031e-01
-7.55642474e-01 2.48991191e-01 -1.93937719e-01 -4.57823873e-01
-1.03790665e+00 4.27477747e-01 2.79524058e-01 -1.15374945e-01
-7.40639508e-01 8.81835938e-01 3.61887723e-01 -3.23592037e-01
-1.35118654e-02 -4.74461287e-01 1.55883655e-01 -2.78958261e-01
4.66657311e-01 6.16198301e-01 -3.32055539e-01 -6.32737339e-01
-1.97257146e-01 2.55954891e-01 -1.86514467e-01 -1.18019432e-01
1.19506180e+00 5.97111974e-03 -1.86779201e-01 1.14289135e-01
1.07247806e+00 -1.81686819e-01 -1.48428631e+00 -1.65179253e-01
-2.93209910e-01 -7.24509239e-01 -4.20538150e-02 -7.48979211e-01
-1.67514646e+00 6.07785642e-01 4.43782240e-01 -6.70402199e-02
1.21429861e+00 1.07894267e-03 1.59685993e+00 1.00951426e-01
2.41769448e-01 -8.74957621e-01 7.64715448e-02 2.62774438e-01
5.52236021e-01 -8.70610356e-01 -7.23498911e-02 -8.29148829e-01
-4.07592326e-01 5.33279002e-01 7.79441893e-01 -6.46054685e-01
4.27514821e-01 2.12111443e-01 1.33144930e-01 -1.66192591e-01
-5.29819608e-01 -1.76196560e-01 1.65145248e-01 6.98491275e-01
-1.95147216e-01 6.41444698e-02 -4.05545115e-01 3.93195003e-01
-1.17139108e-01 1.78999960e-01 6.33997977e-01 9.66748059e-01
-4.94709611e-01 -8.38670135e-01 -4.50380474e-01 7.50779748e-01
-4.93171334e-01 -4.39224094e-02 -6.96719348e-01 6.39303446e-01
3.80758584e-01 9.53334272e-01 1.62636116e-01 -4.20601219e-01
2.34768629e-01 -3.20719123e-01 2.01059580e-01 -5.39338052e-01
-6.83960915e-01 1.18752688e-01 -1.25813246e-01 -8.60711038e-01
-2.87840933e-01 -4.89027113e-01 -1.41515028e+00 -3.40808958e-01
-2.60304391e-01 -1.77278757e-01 5.87932527e-01 8.18086386e-01
5.45255184e-01 5.05650461e-01 3.80425870e-01 -1.12221956e+00
-3.85376811e-02 -7.11386263e-01 -5.42037904e-01 4.07257318e-01
4.43944901e-01 -4.22936261e-01 -3.41217995e-01 1.73867524e-01] | [9.615591049194336, -0.13234348595142365] |
b9744d7b-c648-41d1-8f70-3abb73befb5d | unsupervised-document-embedding-via | 2103.14542 | null | https://arxiv.org/abs/2103.14542v1 | https://arxiv.org/pdf/2103.14542v1.pdf | Unsupervised Document Embedding via Contrastive Augmentation | We present a contrasting learning approach with data augmentation techniques to learn document representations in an unsupervised manner. Inspired by recent contrastive self-supervised learning algorithms used for image and NLP pretraining, we hypothesize that high-quality document embedding should be invariant to diverse paraphrases that preserve the semantics of the original document. With different backbones and contrastive learning frameworks, our study reveals the enormous benefits of contrastive augmentation for document representation learning with two additional insights: 1) including data augmentation in a contrastive way can substantially improve the embedding quality in unsupervised document representation learning, and 2) in general, stochastic augmentations generated by simple word-level manipulation work much better than sentence-level and document-level ones. We plug our method into a classifier and compare it with a broad range of baseline methods on six benchmark datasets. Our method can decrease the classification error rate by up to 6.4% over the SOTA approaches on the document classification task, matching or even surpassing fully-supervised methods. | ['Xiang Zhang', 'Haifeng Chen', 'Dongjin Song', 'Zhengzhang Chen', 'Yanchi Liu', 'Bo Zong', 'Xuchao Zhang', 'Wenchao Yu', 'Jingchao Ni', 'Wei Cheng', 'Dongsheng Luo'] | 2021-03-26 | null | null | null | null | ['document-embedding'] | ['methodology'] | [ 7.45299995e-01 1.29932821e-01 -6.81321979e-01 -3.88963103e-01
-8.64424586e-01 -7.74928808e-01 1.27412868e+00 5.15273511e-01
-5.45443714e-01 4.98536646e-01 7.83113956e-01 -2.21208036e-01
1.04929604e-01 -6.19172037e-01 -6.88656509e-01 -6.93436742e-01
7.86258280e-02 5.29532611e-01 -1.47644833e-01 -3.07397544e-01
5.13069630e-01 3.96665812e-01 -1.60005856e+00 5.91483653e-01
5.88363707e-01 8.19508851e-01 -5.34798913e-02 7.49681354e-01
-6.07683837e-01 9.41075027e-01 -5.99432528e-01 -3.44551861e-01
1.32354751e-01 -4.47819650e-01 -9.90030408e-01 4.02057409e-01
8.70590985e-01 -2.92577118e-01 -5.45942724e-01 8.51658523e-01
2.27429271e-01 -5.69045544e-03 1.13847387e+00 -9.22004104e-01
-1.35754669e+00 6.37161732e-01 -4.83367532e-01 1.14740394e-01
2.02315912e-01 -4.80591096e-02 1.35709858e+00 -1.13514161e+00
6.77813590e-01 1.13066351e+00 4.75865543e-01 6.84299707e-01
-1.70319319e+00 -3.71378869e-01 1.83293879e-01 -1.12299562e-01
-9.54592526e-01 -5.22302687e-01 8.95087421e-01 -3.50604773e-01
1.00195324e+00 1.24502361e-01 3.94245148e-01 1.54974091e+00
-1.62212864e-01 9.63465989e-01 1.21084785e+00 -9.71704960e-01
-3.83704714e-03 4.43996012e-01 4.65434343e-01 7.37425864e-01
4.44412082e-01 3.53155099e-02 -4.96511579e-01 -7.07258284e-02
3.70096773e-01 1.66793704e-01 -1.00489385e-01 -5.92903793e-01
-1.25344574e+00 1.15345359e+00 4.26360011e-01 5.51896513e-01
6.15962893e-02 3.24299693e-01 7.93214560e-01 5.35457492e-01
5.69020629e-01 8.16424131e-01 -3.54095548e-01 1.20745696e-01
-8.86953890e-01 4.66203839e-02 6.37353182e-01 7.85990417e-01
6.97458506e-01 2.41140023e-01 -5.49194574e-01 8.43714476e-01
4.41777222e-02 4.52616781e-01 1.00865567e+00 -5.40579617e-01
5.37643611e-01 5.32952011e-01 -2.28889480e-01 -9.35261548e-01
-5.36796860e-02 -5.70334315e-01 -1.12290859e+00 1.47113940e-02
8.34242329e-02 3.69789362e-01 -1.05980778e+00 1.56502402e+00
-4.49148744e-01 -2.67833471e-01 2.87178755e-01 2.48616561e-01
7.83236325e-01 8.96464765e-01 -6.61397427e-02 -2.04056904e-01
1.29894745e+00 -1.34138548e+00 -8.64254832e-01 -3.20537448e-01
1.05679691e+00 -6.22755408e-01 1.61870825e+00 2.80530840e-01
-8.39876950e-01 -6.61863089e-01 -1.30327415e+00 -1.40623048e-01
-7.07909107e-01 6.78093880e-02 8.44825387e-01 8.89401674e-01
-1.03336954e+00 4.36115026e-01 -4.67566997e-01 -4.20389652e-01
7.63659298e-01 1.05478227e-01 -5.85844398e-01 -3.20168316e-01
-7.95946479e-01 7.82989323e-01 6.55908942e-01 -6.78371072e-01
-8.40423107e-01 -7.01576054e-01 -9.88867640e-01 1.48653254e-01
1.54029518e-01 -5.80067337e-01 9.62956429e-01 -1.06546259e+00
-1.38961458e+00 1.16289032e+00 -1.20822929e-01 -6.69497490e-01
1.46533251e-01 -3.14649701e-01 -1.75999224e-01 2.48883784e-01
-2.10709907e-02 9.35323000e-01 1.13752508e+00 -1.45499313e+00
1.80596337e-02 -3.51265252e-01 -1.11567244e-01 2.49983579e-01
-1.40745175e+00 -1.09659061e-01 -4.07650083e-01 -1.18239534e+00
-2.84240432e-02 -7.96176374e-01 -1.44021243e-01 1.86921582e-01
-3.91776502e-01 -2.85676420e-02 9.49935019e-01 -2.86524981e-01
1.02491593e+00 -2.00245762e+00 1.98132232e-01 3.32114734e-02
1.32148221e-01 1.91429511e-01 -7.01256096e-01 6.90322578e-01
-3.13486725e-01 2.33682856e-01 -5.20047307e-01 -6.96189284e-01
3.42980810e-02 3.38658214e-01 -9.37808454e-01 2.10021526e-01
4.48950827e-01 1.38207173e+00 -8.53358567e-01 -4.40999418e-01
-5.20062074e-02 3.49628985e-01 -6.41930282e-01 1.47086963e-01
-2.92252630e-01 1.32043958e-01 -1.17210083e-01 5.63895702e-01
3.32241356e-01 -4.50681239e-01 1.02409385e-01 -1.52730152e-01
2.67041326e-01 4.91570503e-01 -6.45761967e-01 2.05276108e+00
-5.04315853e-01 9.26969469e-01 -4.86527979e-01 -1.52999687e+00
1.05512726e+00 1.25380084e-01 9.02111977e-02 -7.25497603e-01
-1.16029896e-01 2.38112416e-02 -3.69662464e-01 -1.84004933e-01
7.18143225e-01 -1.07129894e-01 -4.44942750e-02 7.88490295e-01
4.62693751e-01 -3.37063283e-01 2.43746042e-01 5.41966915e-01
1.14549792e+00 -9.06943679e-02 2.13644221e-01 -4.14938867e-01
4.87457186e-01 -7.84771144e-02 -2.56165534e-01 1.03133273e+00
2.09391654e-01 6.23490632e-01 4.40157980e-01 -4.03982431e-01
-1.17765677e+00 -8.98867607e-01 -1.86848521e-01 1.28408468e+00
-3.62621307e-01 -6.78410053e-01 -3.78492177e-01 -1.08707309e+00
1.51344821e-01 5.74484825e-01 -8.43542516e-01 -5.25367439e-01
-1.95009753e-01 -7.84276903e-01 6.56858265e-01 8.62091959e-01
4.31511164e-01 -9.88384068e-01 1.11306505e-02 -1.69651523e-01
1.63053781e-01 -1.15982103e+00 -3.95001441e-01 5.74041843e-01
-1.20814049e+00 -4.63040143e-01 -8.67837131e-01 -1.08335388e+00
9.57899630e-01 4.26026374e-01 1.24665797e+00 3.55132073e-01
-1.75406322e-01 6.47334158e-01 -5.24678886e-01 -2.50411123e-01
-7.00380147e-01 3.94491285e-01 1.16519760e-02 -2.42520690e-01
3.33323807e-01 -5.62424779e-01 -2.58410096e-01 -2.24874005e-01
-1.39487112e+00 -1.32383183e-01 7.45082796e-01 1.19724989e+00
3.36250067e-01 -1.17939241e-01 7.28681684e-01 -1.27084172e+00
9.84563410e-01 -1.37953624e-01 -1.93505645e-01 1.24021754e-01
-1.04739273e+00 5.35826266e-01 7.06480563e-01 -5.36897123e-01
-6.86877370e-01 -5.19415177e-02 -3.48581746e-03 -2.50733793e-01
-1.29724130e-01 5.33878505e-01 1.12811610e-01 4.45965566e-02
1.03715849e+00 7.56626487e-01 1.87537208e-01 -4.24950361e-01
8.05469751e-01 6.23769999e-01 5.01352131e-01 -6.69040442e-01
1.34525847e+00 7.56979764e-01 -8.14477950e-02 -7.41465867e-01
-1.22055531e+00 -2.56683022e-01 -1.00989020e+00 3.85085016e-01
4.53968734e-01 -9.37667847e-01 5.32132015e-03 8.91542658e-02
-1.14880502e+00 -4.55890819e-02 -8.47340465e-01 2.15785906e-01
-4.61115271e-01 7.19621181e-01 -4.95713562e-01 -4.95182276e-01
-4.32129383e-01 -7.35727072e-01 1.09185541e+00 -3.42709422e-01
-2.31038004e-01 -1.16260779e+00 3.11512262e-01 2.99583316e-01
4.04649436e-01 -1.00809410e-02 1.22144878e+00 -9.43945289e-01
-1.56916842e-01 -3.60690832e-01 -3.41041714e-01 8.49056780e-01
3.22531432e-01 -2.91639745e-01 -1.15931571e+00 -5.85726321e-01
-2.01248735e-01 -8.56983423e-01 1.43032146e+00 -9.95315146e-03
1.34272671e+00 -6.75500512e-01 -1.56717777e-01 4.80751574e-01
1.42946661e+00 -3.25017482e-01 8.32988441e-01 6.20529354e-01
6.04006052e-01 6.45818651e-01 1.53300822e-01 1.85570076e-01
-5.98387159e-02 5.03297687e-01 2.67601490e-01 -8.95593315e-02
-3.32932800e-01 -4.58910704e-01 4.92364705e-01 9.62353587e-01
2.44222790e-01 -3.25537890e-01 -7.48356640e-01 6.98921025e-01
-1.50462377e+00 -9.08836186e-01 2.47398615e-01 2.02626014e+00
1.15996778e+00 4.49231386e-01 3.23923528e-02 4.65888649e-01
3.01545382e-01 5.97775996e-01 -2.78355867e-01 -5.23701549e-01
-4.69980508e-01 4.31075007e-01 3.65449935e-01 2.41881937e-01
-1.19901276e+00 1.05980611e+00 7.01939249e+00 7.80300140e-01
-8.96908045e-01 -5.45557477e-02 5.85281849e-01 1.72895808e-02
-5.98161221e-01 -1.99054903e-03 -6.76284492e-01 2.16522306e-01
9.97148693e-01 -3.23368423e-02 2.65216708e-01 8.49716783e-01
-4.63761806e-01 5.04882872e-01 -1.37917578e+00 9.79626954e-01
6.75022602e-01 -1.86208165e+00 8.58530998e-01 1.03812553e-01
1.06235790e+00 -1.49889499e-01 5.23721218e-01 4.50816810e-01
1.59696430e-01 -1.26850677e+00 3.81338984e-01 2.81461000e-01
8.32161546e-01 -7.22270608e-01 5.75390279e-01 1.63874060e-01
-7.86144912e-01 -1.61301762e-01 -4.31086630e-01 -7.57451802e-02
-2.06451431e-01 4.52532202e-01 -6.91616118e-01 3.67204428e-01
4.25481766e-01 1.01846027e+00 -1.00917530e+00 3.72736096e-01
-3.39848012e-01 7.76512742e-01 3.73492055e-02 -6.70447946e-02
4.64600533e-01 -1.37501489e-03 4.01182503e-01 1.64126039e+00
-1.73325032e-01 -4.00101840e-01 -1.15560211e-01 7.67411470e-01
-5.84413707e-01 3.76356661e-01 -9.81124938e-01 -7.28201985e-01
1.84622914e-01 1.08088732e+00 -5.14041543e-01 -6.98679924e-01
-4.56756145e-01 1.31780851e+00 4.10397261e-01 3.20389092e-01
-3.40914696e-01 -5.40654719e-01 2.84203053e-01 1.82967819e-02
4.12802726e-01 -2.78253466e-01 -4.11558867e-01 -1.24094427e+00
2.03248158e-01 -9.09539342e-01 2.65566558e-01 -5.85581005e-01
-1.52769530e+00 6.70032144e-01 -1.96526468e-01 -1.24286759e+00
-4.25896436e-01 -8.66374135e-01 -4.74263072e-01 4.30191934e-01
-1.92370915e+00 -1.25654900e+00 -1.12670213e-01 4.82350230e-01
7.09570944e-01 -6.62253261e-01 1.45670843e+00 -5.90791889e-02
-2.18188643e-01 9.04329836e-01 5.87690890e-01 3.20170760e-01
6.68546617e-01 -1.28193724e+00 4.81205821e-01 5.63266098e-01
9.45834517e-01 7.05119848e-01 3.78321230e-01 -2.46466801e-01
-1.36929834e+00 -1.09946227e+00 8.07533085e-01 -8.10385227e-01
8.49544585e-01 -7.61153877e-01 -1.08749914e+00 7.32736051e-01
5.34079969e-01 1.97443753e-01 8.26832414e-01 1.98353976e-01
-1.17835593e+00 1.63464807e-02 -8.83668661e-01 6.53187633e-01
9.59978938e-01 -9.82072890e-01 -9.60204184e-01 7.10643113e-01
8.98094356e-01 2.12176263e-01 -5.46546578e-01 2.14908749e-01
4.99733239e-01 -4.65029180e-01 1.17164445e+00 -1.17831659e+00
1.02979350e+00 1.83509305e-01 -3.76643538e-01 -1.17594624e+00
-1.95303798e-01 -4.53737557e-01 -6.00631773e-01 1.29130435e+00
3.32661748e-01 -5.55256426e-01 8.70862305e-01 -1.42784163e-01
2.17234008e-02 -6.56454861e-01 -6.08024001e-01 -1.21151078e+00
5.02112746e-01 -2.49767765e-01 9.12678167e-02 1.21711159e+00
-7.44571770e-03 6.58433437e-01 -2.98343122e-01 -2.07665011e-01
6.15799546e-01 -1.05956476e-02 7.66119719e-01 -1.29787862e+00
-2.11091176e-01 -5.23328900e-01 -5.58677614e-01 -1.28303158e+00
6.44885957e-01 -1.49840152e+00 -2.60811031e-01 -1.45948386e+00
6.58446670e-01 -6.92219958e-02 -5.02320945e-01 5.97964108e-01
5.65176010e-02 4.89772826e-01 1.06262304e-01 4.40167516e-01
-4.90758777e-01 8.18385303e-01 9.11611080e-01 -7.00034201e-01
3.72376218e-02 -5.01739323e-01 -1.05250239e+00 5.76830089e-01
5.58221877e-01 -5.70587695e-01 -5.46805441e-01 -6.39290154e-01
2.20040038e-01 -7.65422165e-01 2.58346945e-01 -7.73805976e-01
8.16209465e-02 2.39876434e-01 4.34480488e-01 -3.97169292e-01
2.18351319e-01 -7.80175745e-01 -8.71471047e-01 5.38361669e-01
-9.79091108e-01 -1.17350802e-01 4.78733033e-01 8.29785824e-01
-3.84703606e-01 -6.00349307e-01 7.70582616e-01 -1.71807975e-01
-5.26115358e-01 6.73042014e-02 -5.28405428e-01 1.25948533e-01
5.22398472e-01 -3.28236967e-01 -4.66007560e-01 -5.18028557e-01
-3.96875203e-01 -2.72280216e-01 3.48391742e-01 6.42237782e-01
7.48418927e-01 -1.43750274e+00 -8.45471084e-01 2.73005813e-01
6.72519028e-01 -4.40020978e-01 -3.20823133e-01 3.95861506e-01
-1.55394509e-01 7.41232038e-01 -1.12217605e-01 -6.53477371e-01
-1.17289186e+00 5.85586309e-01 -1.36424035e-01 -5.59224725e-01
-6.33451939e-01 7.36320317e-01 1.54641941e-01 -5.24061918e-01
3.82310688e-01 -2.21470684e-01 -2.46404722e-01 2.22069472e-01
6.96090698e-01 -1.72146633e-01 2.04965279e-01 -3.77423525e-01
-8.79021063e-02 5.49132645e-01 -8.43981445e-01 -1.72898069e-01
1.53033876e+00 1.36123300e-01 -4.14952934e-02 5.32653809e-01
1.66144073e+00 3.84228714e-02 -7.47883499e-01 -6.28228605e-01
3.01979125e-01 -4.41732973e-01 2.59449720e-01 -6.72426283e-01
-7.96273708e-01 9.03767884e-01 5.25408149e-01 2.04308048e-01
9.98938024e-01 2.03695819e-01 4.32267100e-01 1.15101659e+00
-9.10467133e-02 -9.75520551e-01 9.24360812e-01 5.79426289e-01
1.08050835e+00 -1.44449925e+00 3.50018889e-01 -8.14162791e-02
-6.36803448e-01 1.16144156e+00 2.48191565e-01 -2.85058826e-01
4.62761521e-01 9.70624089e-02 -1.79236427e-01 -1.78171113e-01
-7.90399313e-01 -1.08237587e-01 5.57268083e-01 6.94612086e-01
6.69096708e-01 -5.14641821e-01 -2.66059581e-02 2.27981865e-01
-1.69573933e-01 -3.16295624e-01 4.94831383e-01 1.20881593e+00
-2.63507754e-01 -1.26998532e+00 -1.18011452e-01 5.58486998e-01
-2.06912905e-01 -4.30287838e-01 -5.85766435e-01 8.48056436e-01
-5.42082489e-01 5.82472563e-01 3.99100214e-01 -1.20261535e-01
1.51541978e-01 4.19434130e-01 6.07686222e-01 -1.08338404e+00
-3.30273569e-01 -3.95838559e-01 -5.76738939e-02 -2.13070959e-01
-5.85054755e-01 -4.94545311e-01 -1.04592037e+00 2.31657922e-01
-3.38711858e-01 7.69014284e-02 6.63235605e-01 8.83495271e-01
3.59521568e-01 4.12369370e-01 7.91246653e-01 -6.45084977e-01
-8.07254851e-01 -1.11956656e+00 -4.46737021e-01 6.43385291e-01
4.54933316e-01 -3.36728483e-01 -5.74463010e-01 3.93206269e-01] | [9.577845573425293, 2.6745615005493164] |
72403dc7-a0a0-42c0-bcb8-de250ac720b3 | neural-micro-planning-for-data-to-text | null | null | https://aclanthology.org/2020.dt4tp-1.2 | https://aclanthology.org/2020.dt4tp-1.2.pdf | Neural Micro-Planning for Data to Text Generation Produces more Cohesive Text | null | ['Michael Elhadad', 'Roy Eisenstadt'] | null | null | null | null | dt4tp-2020-12 | ['data-to-text-generation'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.400879859924316, 3.705815076828003] |
b7970e15-9351-42ee-b996-c6add9adea0a | ream-sharp-an-enhancement-approach-to-1 | null | null | https://aclanthology.org/2021.findings-acl.220 | https://aclanthology.org/2021.findings-acl.220.pdf | REAM\sharp: An Enhancement Approach to Reference-based Evaluation Metrics for Open-domain Dialog Generation | null | ['Shuming Shi', 'Ruifeng Xu', 'Wei Bi', 'Jun Gao'] | null | null | null | null | findings-acl-2021-8 | ['open-domain-dialog'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.357307434082031, 3.636298418045044] |
384b44d6-b1c7-4e22-8d98-fe8e4a1c9dbc | pseudo-session-based-recommendation-with | 2306.10029 | null | https://arxiv.org/abs/2306.10029v1 | https://arxiv.org/pdf/2306.10029v1.pdf | Pseudo session-based recommendation with hierarchical embedding and session attributes | Recently, electronic commerce (EC) websites have been unable to provide an identification number (user ID) for each transaction data entry because of privacy issues. Because most recommendation methods assume that all data are assigned a user ID, they cannot be applied to the data without user IDs. Recently, session-based recommendation (SBR) based on session information, which is short-term behavioral information of users, has been studied. A general SBR uses only information about the item of interest to make a recommendation (e.g., item ID for an EC site). Particularly in the case of EC sites, the data recorded include the name of the item being purchased, the price of the item, the category hierarchy, and the gender and region of the user. In this study, we define a pseudo--session for the purchase history data of an EC site without user IDs and session IDs. Finally, we propose an SBR with a co-guided heterogeneous hypergraph and globalgraph network plus, called CoHHGN+. The results show that our CoHHGN+ can recommend items with higher performance than other methods. | ['Satoshi Takahashi', 'Ryusei Numata', 'Yuta Sumiya'] | 2023-06-06 | null | null | null | null | ['session-based-recommendations'] | ['miscellaneous'] | [-2.75044620e-01 -2.15851292e-01 -7.66862988e-01 -4.26949203e-01
1.73793156e-02 -7.38758504e-01 2.33006924e-01 4.58018005e-01
-3.17892909e-01 4.13512409e-01 2.26397157e-01 -3.37369084e-01
-3.25681776e-01 -1.28412437e+00 -3.32257330e-01 -4.61272031e-01
-4.15611751e-02 5.54778457e-01 4.14100915e-01 -3.74123186e-01
5.27467787e-01 5.48275895e-02 -1.01096463e+00 2.90823013e-01
6.88183308e-01 1.01828301e+00 2.43740082e-01 -1.37788624e-01
-3.10382515e-01 4.25748862e-02 -2.91049182e-01 -9.13952053e-01
5.22102475e-01 -4.10976082e-01 -7.65785456e-01 -4.35911715e-02
-9.08426046e-02 -6.20253623e-01 -5.29509187e-01 9.67324495e-01
-2.37948615e-02 3.16746801e-01 6.56063676e-01 -1.22813773e+00
-7.19829023e-01 1.15782082e+00 -6.51219964e-01 -4.15093340e-02
3.96205842e-01 -6.64315224e-01 1.44971371e+00 -4.48364198e-01
9.82017756e-01 6.24023020e-01 2.46352136e-01 2.24791944e-01
-9.32205856e-01 -5.89956582e-01 3.99377525e-01 3.51955205e-01
-1.33222401e+00 2.46223181e-01 6.24942601e-01 -2.28264108e-01
4.04734761e-01 3.92817020e-01 5.82123220e-01 7.59338140e-01
-7.24880546e-02 3.22676599e-01 5.62594235e-01 -8.70992765e-02
3.70721847e-01 4.35605645e-01 8.15263450e-01 1.86754897e-01
7.58679152e-01 -3.53695929e-01 -2.67513782e-01 -4.66388345e-01
8.40557754e-01 5.34613907e-01 -2.45725051e-01 -4.85703170e-01
-9.34767962e-01 9.94600594e-01 5.01546621e-01 2.91182458e-01
-4.43046331e-01 -5.79347849e-01 3.65892977e-01 3.85951340e-01
2.24763200e-01 2.18056366e-01 -4.72202778e-01 3.72949809e-01
-5.04023075e-01 -9.18164700e-02 1.38712287e+00 1.27687633e+00
7.86610901e-01 -5.07852077e-01 1.62332445e-01 8.34404588e-01
6.26522720e-01 1.37926891e-01 3.84172112e-01 -5.32568038e-01
4.46048766e-01 6.65646911e-01 2.40974531e-01 -1.42020309e+00
-1.26679137e-01 -5.32349944e-01 -7.90541172e-01 -6.55137181e-01
4.41959947e-01 -1.99642822e-01 -5.18560886e-01 1.42765033e+00
3.21084142e-01 1.78274568e-02 -5.61813377e-02 1.01133311e+00
1.15042949e+00 6.92390084e-01 -6.41613603e-02 -5.53134978e-01
1.32936704e+00 -7.28661537e-01 -7.45471954e-01 2.46363685e-01
6.05399311e-01 -4.61812198e-01 3.44597578e-01 2.98253030e-01
-5.56862652e-01 -2.01633081e-01 -7.33684599e-01 1.85504809e-01
-6.76231563e-01 -1.67945996e-01 8.00227106e-01 7.35013604e-01
-7.93888509e-01 4.81286973e-01 -3.19223642e-01 -8.00887287e-01
-2.28199795e-01 3.84112805e-01 -3.18712115e-01 -3.43995243e-01
-1.37933505e+00 2.78357267e-01 3.39922011e-01 4.18328978e-02
-2.94470280e-01 -3.81479025e-01 -5.39303005e-01 3.52900416e-01
6.66525722e-01 -4.36688334e-01 6.96719646e-01 -6.48253679e-01
-1.09232998e+00 4.36259925e-01 9.03482959e-02 -4.06976342e-01
1.90247253e-01 2.96775758e-01 -1.00065136e+00 -2.19361931e-01
-8.87677539e-03 -1.59734085e-01 4.18946028e-01 -1.21789253e+00
-1.04079342e+00 -7.57135034e-01 4.03669626e-01 2.30584100e-01
-5.44265091e-01 -1.60528481e-01 -1.06423008e+00 -4.30061311e-01
4.66227084e-01 -9.79150891e-01 -2.59985536e-01 -6.57127917e-01
-4.47369158e-01 -3.64706606e-01 6.21919870e-01 -8.25239778e-01
1.81433666e+00 -2.32690287e+00 -1.29056230e-01 1.07247603e+00
2.44720295e-01 3.86457285e-03 6.70554349e-04 8.41577768e-01
4.72583711e-01 4.08941835e-01 4.15128499e-01 1.78469509e-01
4.95289080e-02 1.70254752e-01 -2.97800124e-01 3.62914830e-01
-1.13947332e+00 6.09682202e-01 -6.29363894e-01 -3.30540210e-01
-1.88264266e-01 1.98057368e-01 -7.53547549e-01 2.20025666e-02
-2.86181718e-02 2.94355098e-02 -7.15016842e-01 4.46759999e-01
8.57350051e-01 -4.13351327e-01 9.48191345e-01 -6.04072094e-01
-3.91428135e-02 2.98688531e-01 -1.31554079e+00 1.29287136e+00
-2.06234246e-01 -1.79824725e-01 2.54984647e-01 -5.07994294e-01
1.10322917e+00 2.28183255e-01 7.57359445e-01 -5.38671136e-01
-1.13540264e-02 2.64553398e-01 7.94132799e-02 -1.41608924e-01
7.03657389e-01 6.31132305e-01 -2.16617584e-01 7.60421395e-01
-2.43723273e-01 7.93226182e-01 3.50874722e-01 6.07573390e-01
1.01062691e+00 -3.01828921e-01 3.85025024e-01 -2.58409381e-01
5.14686108e-01 -6.37108535e-02 7.99378455e-01 9.04862940e-01
3.23966086e-01 3.31263304e-01 2.35598281e-01 -2.50367135e-01
-9.02947485e-01 -6.15949929e-01 -9.07149389e-02 1.14492011e+00
7.15708554e-01 -7.59397566e-01 -4.76961106e-01 -9.83331859e-01
4.42941427e-01 5.89499831e-01 -4.53203261e-01 4.00153883e-02
-4.28156197e-01 -4.50773239e-01 -3.19575459e-01 1.16284236e-01
6.56372309e-01 -8.42136323e-01 3.43838602e-01 4.52790678e-01
-4.09687847e-01 -8.80844116e-01 -1.05064976e+00 -3.69729817e-01
-7.93760478e-01 -9.39088523e-01 -5.20304263e-01 -9.41587746e-01
8.01469922e-01 4.85230476e-01 7.98547983e-01 4.06395674e-01
4.11358744e-01 2.97542006e-01 -8.67294073e-01 2.14936942e-01
1.21743701e-01 4.69399810e-01 -6.62598163e-02 5.24474740e-01
6.84470117e-01 -5.09556770e-01 -9.12232637e-01 1.15353572e+00
-8.25622320e-01 -3.21874142e-01 1.89857945e-01 4.01462942e-01
4.73260432e-01 2.19950184e-01 7.68809140e-01 -1.68142569e+00
6.20021999e-01 -1.02151632e+00 -3.26369971e-01 5.69797635e-01
-1.14844120e+00 -5.37306190e-01 5.80179572e-01 -2.24562287e-01
-1.05541420e+00 -3.01892579e-01 -5.29559031e-02 3.13100219e-01
-1.05555832e-01 1.05547953e+00 -4.05196756e-01 9.89270285e-02
6.72296435e-02 1.73291955e-02 -2.46888220e-01 -9.28622186e-01
6.03347495e-02 1.07256901e+00 2.11637199e-01 -1.98401764e-01
4.42436576e-01 2.84145057e-01 -2.33729616e-01 -4.68717277e-01
-5.89367509e-01 -1.08520806e+00 -5.93537152e-01 -5.02170511e-02
4.74035174e-01 -6.95365489e-01 -8.00713778e-01 2.88146496e-01
-6.03027523e-01 3.06296170e-01 1.97538942e-01 8.19761038e-01
9.03029516e-02 5.47094166e-01 -8.53264987e-01 -4.20972794e-01
-3.09180617e-01 -5.56814849e-01 5.29859029e-02 2.19464935e-02
-2.73138396e-02 -1.03053582e+00 -1.32317096e-01 5.11756837e-01
4.79534656e-01 -2.55028903e-01 1.01849234e+00 -1.29229164e+00
-6.49036288e-01 -5.24160862e-01 -1.65270656e-01 -1.71106875e-01
3.20030957e-01 -2.12655425e-01 7.79982358e-02 -4.76263493e-01
-2.64302075e-01 5.55242896e-01 5.41605294e-01 3.05528343e-01
9.69147801e-01 -7.05752552e-01 -7.57803202e-01 5.28535008e-01
1.62339056e+00 6.48453414e-01 6.45586789e-01 3.17248046e-01
7.65599847e-01 4.80056494e-01 7.19389379e-01 6.91447496e-01
5.51737010e-01 9.41711187e-01 3.23957533e-01 2.56981671e-01
5.08244991e-01 -5.39237857e-01 5.97967505e-02 9.53839839e-01
-3.97631139e-01 -8.44922960e-01 -2.02651724e-01 2.53309339e-01
-2.02688456e+00 -8.87887299e-01 -4.75477278e-01 2.47595477e+00
2.76534468e-01 -1.37707964e-01 2.98843324e-01 -2.31553838e-01
1.04878199e+00 -2.04573408e-01 -7.47988582e-01 -2.84589320e-01
2.27280200e-01 -4.39414382e-01 9.07323539e-01 2.05679744e-01
-6.21715724e-01 6.11145675e-01 5.30418730e+00 6.06619477e-01
-5.15728235e-01 1.80097215e-03 2.76259899e-01 2.71931022e-01
-7.18533039e-01 4.72987592e-02 -1.14121163e+00 7.24512577e-01
6.81320727e-01 -4.04288054e-01 7.19764948e-01 7.63167799e-01
2.08138302e-02 4.62423302e-02 -1.01306272e+00 7.62776077e-01
1.48485601e-01 -1.08077657e+00 1.36538744e-01 7.13072121e-01
6.90421939e-01 -4.34228092e-01 1.01134419e-01 3.77658367e-01
5.03761709e-01 -2.90542603e-01 1.31548077e-01 1.23583272e-01
4.32564080e-01 -7.90005624e-01 9.35490668e-01 3.11894208e-01
-1.28219175e+00 -3.33372384e-01 -5.79115748e-01 3.85053188e-01
2.76523560e-01 4.27157283e-01 -4.90832269e-01 9.48030770e-01
6.93558335e-01 9.71705616e-01 -3.45423251e-01 1.32924330e+00
1.49701566e-01 7.10893810e-01 -2.87181854e-01 -1.70865387e-01
-3.53230014e-02 -8.56116593e-01 3.28813851e-01 7.05911756e-01
6.84360564e-01 5.87154567e-01 1.98838919e-01 2.65217692e-01
-3.18847865e-01 8.37379098e-01 -4.35050845e-01 1.42757878e-01
6.87017143e-01 1.35295653e+00 -7.21509159e-01 -1.05639376e-01
-8.06192219e-01 6.91097140e-01 6.65441379e-02 4.22849715e-01
-3.55897933e-01 -4.06104773e-01 5.38092077e-01 5.84477603e-01
6.49749577e-01 -3.53958160e-02 -6.04800023e-02 -1.16869533e+00
-2.70348400e-01 -6.41108572e-01 9.71017540e-01 -2.88074046e-01
-1.50891733e+00 3.62148106e-01 -1.39229178e-01 -1.23529530e+00
-1.07626535e-01 -1.42729044e-01 -4.05182898e-01 6.41473651e-01
-7.97892570e-01 -8.41616988e-01 -1.20347917e-01 7.49490917e-01
-1.77163463e-02 -2.62801141e-01 6.99007809e-01 6.48017108e-01
-5.32872021e-01 8.56836796e-01 4.92603093e-01 4.03234780e-01
6.27808869e-01 -7.92068481e-01 -3.74956727e-02 3.91378582e-01
1.54655129e-01 1.09659266e+00 4.60655957e-01 -1.11563039e+00
-1.56348848e+00 -9.51718569e-01 9.41596568e-01 -2.05167890e-01
4.37610388e-01 -3.14512789e-01 -9.52004433e-01 1.04490399e+00
8.49678144e-02 -4.38551039e-01 9.65657175e-01 6.33924067e-01
-1.96823090e-01 -5.48131466e-01 -1.41994357e+00 4.28897470e-01
1.31023717e+00 -8.35859403e-02 -2.52249151e-01 1.48048073e-01
6.68583930e-01 7.56395385e-02 -1.26692426e+00 1.76084220e-01
5.97296119e-01 -6.29204988e-01 8.79492700e-01 -4.33470905e-01
-1.25121297e-02 -2.02845931e-01 -1.77468061e-01 -1.16240025e+00
-8.25531185e-01 -2.30857208e-01 -9.87009704e-02 1.63590884e+00
7.16143548e-01 -8.50390673e-01 8.05400968e-01 1.17961729e+00
3.40750188e-01 -2.36239985e-01 -4.81359452e-01 -5.55980802e-01
-4.98421729e-01 2.28074610e-01 1.15907609e+00 1.00733924e+00
4.25897032e-01 3.96631420e-01 -6.03588283e-01 2.93469012e-01
6.57508075e-01 7.24284232e-01 6.17990196e-01 -1.66712606e+00
-2.50340879e-01 -2.32565980e-02 -7.35487118e-02 -1.03168559e+00
-2.21699625e-01 -9.80105102e-01 -5.10837376e-01 -1.79064369e+00
4.35811281e-01 -6.46628916e-01 -6.84922218e-01 1.73074633e-01
4.09169436e-01 2.54615378e-02 2.18489617e-02 6.75225914e-01
-7.47802973e-01 -2.33489051e-01 1.12806308e+00 2.56336391e-01
-5.35725355e-01 5.13566256e-01 -7.58907497e-01 1.84656918e-01
7.78928459e-01 -3.62032384e-01 -3.69072556e-01 -1.11099832e-01
4.00291771e-01 4.21244830e-01 -2.52564102e-01 -2.61320978e-01
6.04934633e-01 -3.55166823e-01 7.71563081e-03 -8.42060626e-01
-9.43864360e-02 -1.31317699e+00 8.47632349e-01 1.07605532e-01
-7.19442248e-01 -8.76479447e-02 -7.84945905e-01 8.38934362e-01
-1.96867362e-01 -5.05459011e-01 -7.90156331e-03 -2.42240146e-01
-7.36295581e-01 9.84344363e-01 -7.51606598e-02 -4.06920642e-01
9.21282172e-01 -7.75705799e-02 -3.83842945e-01 -5.40892601e-01
-8.75840306e-01 3.69113594e-01 4.99978840e-01 5.11700869e-01
4.20811594e-01 -1.32361650e+00 -5.67079961e-01 9.74006504e-02
1.99848041e-01 -5.33408523e-01 3.69401395e-01 4.11836267e-01
5.86955249e-02 4.06911373e-01 -2.15181455e-01 7.79523328e-02
-1.39516485e+00 8.07565212e-01 -3.22972089e-01 -2.16662794e-01
-4.20307636e-01 2.29286864e-01 2.55807847e-01 -3.80651742e-01
3.01936090e-01 2.97652543e-01 -9.65053201e-01 4.53614891e-01
3.24360162e-01 5.26945710e-01 2.07394585e-02 -5.85064709e-01
-1.08052641e-01 3.38507146e-01 -6.40275657e-01 2.39647180e-02
1.36464870e+00 -6.93548918e-01 -3.42472017e-01 1.48808762e-01
1.24796951e+00 1.86988860e-01 -5.45701087e-01 -7.05263019e-01
-9.82045457e-02 -7.96951711e-01 -1.03710601e-02 -7.67067552e-01
-1.67053413e+00 9.58747193e-02 1.62959456e-01 7.77016461e-01
7.70382941e-01 -8.66642818e-02 9.24485028e-01 2.77442873e-01
1.02874553e+00 -1.17697489e+00 -5.29173672e-01 3.15086097e-01
4.85683382e-01 -9.05784249e-01 -1.17910661e-01 -7.90283084e-01
-7.14562595e-01 9.17757928e-01 7.23637521e-01 2.71476626e-01
1.38798296e+00 -4.16939646e-01 -1.66742831e-01 -7.13257790e-02
-4.83657628e-01 5.19354502e-03 2.04500496e-01 4.66927767e-01
1.18635878e-01 2.11095870e-01 -1.07469904e+00 9.85508561e-01
1.26626287e-02 -2.21259911e-02 6.37911856e-01 7.38974094e-01
-3.51320148e-01 -1.27644718e+00 1.78541183e-01 9.27611113e-01
-7.47935891e-01 -2.21576393e-02 -1.31623417e-01 4.25325632e-01
-1.56904638e-01 1.04564774e+00 1.45689711e-01 -7.41811931e-01
3.36345464e-01 -2.26190969e-01 -1.21872850e-01 -7.50935137e-01
-7.86420107e-01 1.17687769e-02 2.21094787e-01 -2.05158174e-01
-5.58815561e-02 -5.90007722e-01 -1.11310160e+00 -8.25722396e-01
-6.19163692e-01 7.76173234e-01 7.02736497e-01 4.69176173e-01
4.56429392e-01 -1.46224141e-01 1.12554348e+00 -1.12080477e-01
-3.38063329e-01 -6.57770216e-01 -1.44074368e+00 6.90282404e-01
-2.53121674e-01 -2.71842092e-01 -4.91154999e-01 -1.71252087e-01] | [10.10521125793457, 5.661802291870117] |
d43cc455-dab2-4785-8fd9-1744dc94fb71 | pose-forecasting-in-industrial-human-robot | 2208.07308 | null | https://arxiv.org/abs/2208.07308v1 | https://arxiv.org/pdf/2208.07308v1.pdf | Pose Forecasting in Industrial Human-Robot Collaboration | Pushing back the frontiers of collaborative robots in industrial environments, we propose a new Separable-Sparse Graph Convolutional Network (SeS-GCN) for pose forecasting. For the first time, SeS-GCN bottlenecks the interaction of the spatial, temporal and channel-wise dimensions in GCNs, and it learns sparse adjacency matrices by a teacher-student framework. Compared to the state-of-the-art, it only uses 1.72% of the parameters and it is ~4 times faster, while still performing comparably in forecasting accuracy on Human3.6M at 1 second in the future, which enables cobots to be aware of human operators. As a second contribution, we present a new benchmark of Cobots and Humans in Industrial COllaboration (CHICO). CHICO includes multi-view videos, 3D poses and trajectories of 20 human operators and cobots, engaging in 7 realistic industrial actions. Additionally, it reports 226 genuine collisions, taking place during the human-cobot interaction. We test SeS-GCN on CHICO for two important perception tasks in robotics: human pose forecasting, where it reaches an average error of 85.3 mm (MPJPE) at 1 sec in the future with a run time of 2.3 msec, and collision detection, by comparing the forecasted human motion with the known cobot motion, obtaining an F1-score of 0.64. | ['Fabio Galasso', 'Marco Cristani', 'Francesco Setti', 'Geri Skenderi', 'Federico Cunico', 'Andrea Avogaro', "Guido D'Amely", 'Alessio Sampieri'] | 2022-07-24 | null | null | null | null | ['human-pose-forecasting'] | ['computer-vision'] | [-1.74953938e-01 5.30882418e-01 5.58832228e-01 1.53612137e-01
-4.16413963e-01 -4.77295756e-01 8.00804347e-02 -1.63486853e-01
-3.71922761e-01 2.92029053e-01 -4.04752970e-01 -1.42709926e-01
-3.86435628e-01 -4.44356501e-01 -9.34505463e-01 -5.35199702e-01
-7.74111152e-01 9.47305679e-01 4.10177678e-01 -4.41795051e-01
1.85645912e-02 6.00022554e-01 -1.35682726e+00 1.87440217e-01
4.32191461e-01 1.05261588e+00 7.58347809e-01 9.14675593e-01
6.86135530e-01 8.94369423e-01 -7.92412400e-01 -2.94172913e-01
5.38193047e-01 7.08099306e-02 -7.53144443e-01 2.09223796e-02
3.04245025e-01 -2.27507517e-01 -8.15881193e-01 7.33097911e-01
5.23771584e-01 2.53099173e-01 4.64601785e-01 -1.64359307e+00
-2.03118816e-01 5.86377680e-01 -3.57360035e-01 -1.75009668e-01
4.04179782e-01 6.45403445e-01 4.93924141e-01 -8.63404274e-01
9.73861277e-01 1.60522127e+00 9.63855028e-01 1.62834033e-01
-6.92711174e-01 -6.68260455e-01 2.56556511e-01 3.02520871e-01
-1.21872687e+00 1.05515853e-01 4.59537715e-01 -5.35675764e-01
1.15320981e+00 -2.01408684e-01 8.07360709e-01 1.29674137e+00
8.07311475e-01 5.94904006e-01 3.49175662e-01 1.67679176e-01
1.39177769e-01 -5.39655149e-01 -3.85591805e-01 9.38808024e-01
2.95184553e-01 1.95243418e-01 -6.41197741e-01 1.10995665e-01
1.12827349e+00 1.46765679e-01 -2.03283839e-02 -6.71312988e-01
-1.82378709e+00 5.73680520e-01 9.43055868e-01 -1.52104199e-01
-5.85143387e-01 5.67970812e-01 5.93764305e-01 2.87806362e-01
-6.52117431e-02 5.36252558e-01 -4.96725082e-01 -3.40219617e-01
2.32415304e-01 5.15739381e-01 9.39960182e-01 1.70602930e+00
4.43528533e-01 -2.59073764e-01 2.16761887e-01 3.83467913e-01
6.33753389e-02 7.72778928e-01 -1.87027007e-01 -1.46758795e+00
7.58804977e-01 4.32322830e-01 3.42720807e-01 -1.41302693e+00
-8.78670216e-01 -2.35746801e-01 -8.90059590e-01 3.18433166e-01
5.82766831e-01 -4.01682317e-01 -6.13892913e-01 1.25666761e+00
4.35862958e-01 -1.65823877e-01 -7.98921436e-02 1.18591249e+00
5.96786201e-01 4.22809482e-01 -3.02831709e-01 1.12507790e-01
1.16045272e+00 -1.36766434e+00 -6.59932435e-01 -3.07063520e-01
6.84902847e-01 -7.32312799e-01 6.61082983e-01 8.28839242e-01
-7.55132377e-01 -9.66461897e-01 -8.90148878e-01 2.51764119e-01
-1.35254383e-01 3.44922394e-01 1.01164734e+00 -1.33226454e-01
-5.60266793e-01 9.48277235e-01 -9.97156560e-01 -4.87042367e-01
1.73767239e-01 5.10284185e-01 -5.33357322e-01 -2.46415555e-01
-8.35198164e-01 8.67169261e-01 1.31501943e-01 5.55205464e-01
-1.32813728e+00 -5.94336808e-01 -8.76782179e-01 -2.03932747e-01
9.21235204e-01 -6.31940007e-01 1.41382980e+00 4.00852524e-02
-1.51924276e+00 2.89797783e-01 5.42732298e-01 -4.81715083e-01
8.19388628e-01 -8.30955625e-01 -2.19059199e-01 1.30489081e-01
2.12889135e-01 7.74343610e-01 6.69601083e-01 -1.38348997e+00
-7.30653524e-01 -3.52941900e-01 4.42655295e-01 1.60902441e-01
3.41628551e-01 -4.77495939e-01 -5.61854005e-01 -2.87207991e-01
3.29797953e-01 -1.67414796e+00 -6.30377114e-01 6.82789236e-02
-5.14137864e-01 -2.73230433e-01 8.66552770e-01 -4.13658559e-01
4.45830762e-01 -2.14321995e+00 4.75153506e-01 -8.36408660e-02
3.50453079e-01 -8.03632364e-02 -2.19111055e-01 8.08878899e-01
3.20722222e-01 -7.27411032e-01 1.41388208e-01 -4.20112401e-01
2.32687712e-01 1.91397503e-01 4.87088040e-02 6.26926005e-01
-1.64389163e-01 1.05617225e+00 -1.17323387e+00 -2.34173983e-01
3.96806747e-01 2.44832411e-01 -5.23942769e-01 3.11632782e-01
-1.38549246e-02 6.19357646e-01 -1.66944802e-01 6.72138989e-01
6.45731449e-01 -1.65885314e-01 6.16329730e-01 -4.35908645e-01
1.04955165e-03 -9.89456326e-02 -1.17972648e+00 2.02016950e+00
-5.36056399e-01 4.94347841e-01 4.08330768e-01 -7.24613428e-01
1.08317602e+00 1.13104425e-01 9.04113352e-01 -3.23456705e-01
2.73078173e-01 1.96085557e-01 3.13410759e-02 -5.02079964e-01
4.25145328e-01 6.18303299e-01 -5.14166892e-01 9.33461860e-02
2.29053423e-01 -4.53614503e-01 2.12667286e-01 2.75138915e-01
1.57185805e+00 2.11272165e-01 -1.90201178e-01 -1.41744375e-01
8.53272825e-02 1.54244289e-01 4.02047962e-01 7.55733252e-01
-3.02166253e-01 4.19841409e-01 4.86920148e-01 -8.18714738e-01
-5.55655777e-01 -1.22479355e+00 6.85088933e-01 9.11489487e-01
7.48672307e-01 -5.28176010e-01 -5.56040704e-01 -9.29817855e-01
3.57594758e-01 2.92824090e-01 -3.25104445e-01 -2.27755755e-01
-1.01418638e+00 -4.28038612e-02 1.89239457e-01 8.01285446e-01
4.13159013e-01 -1.37625253e+00 -1.10385942e+00 3.88373733e-01
-2.19443485e-01 -1.60880947e+00 -3.25837493e-01 3.99770796e-01
-4.97417450e-01 -1.55155051e+00 -4.12630498e-01 -8.66642237e-01
5.26237488e-01 4.33020234e-01 1.11203265e+00 -1.79877579e-01
-6.02042317e-01 3.93183351e-01 -4.60018367e-01 -4.90494400e-01
-2.57207990e-01 2.54442394e-02 4.84493405e-01 -7.21780717e-01
-8.67422000e-02 -6.73776925e-01 -5.58230102e-01 6.71441793e-01
-1.05401300e-01 -7.40017593e-02 8.12252045e-01 7.25798070e-01
3.32646906e-01 3.60849760e-02 2.61264443e-01 -5.70828021e-01
3.29394370e-01 -3.08971614e-01 -7.07231760e-01 -1.71264708e-01
-1.51565537e-01 -5.54492235e-01 5.52348673e-01 -6.66600883e-01
-6.03443921e-01 6.97639525e-01 3.23709756e-01 -9.49993849e-01
-1.94361791e-01 2.38100708e-01 1.50163352e-01 -4.83877845e-02
6.13609612e-01 -3.65518004e-01 2.26864919e-01 -3.16675067e-01
4.19751853e-01 2.89716274e-01 9.74401057e-01 -3.96834224e-01
8.12367439e-01 2.55340278e-01 2.17299059e-01 -6.13731086e-01
-6.08035505e-01 -5.70360303e-01 -1.00718904e+00 -7.51899004e-01
9.33097839e-01 -1.05915165e+00 -1.79400253e+00 6.78997934e-01
-1.60512614e+00 -7.03288376e-01 -2.15329424e-01 7.45116472e-01
-1.03487396e+00 2.06482723e-01 -8.98207963e-01 -8.29210162e-01
4.54166234e-02 -1.20527112e+00 1.43871105e+00 -4.14210618e-01
-1.73958883e-01 -3.38187575e-01 -4.99109894e-01 4.43174422e-01
1.01973735e-01 4.28394437e-01 4.32948351e-01 -2.92815059e-01
-8.30098987e-01 -2.20524207e-01 -4.15551923e-02 -7.16933161e-02
5.05162738e-02 -4.91076469e-01 -5.05323172e-01 -5.67712724e-01
-2.93588996e-01 -4.11932737e-01 4.07368600e-01 2.95053810e-01
8.15237820e-01 8.06810334e-03 -6.05041802e-01 1.51422858e-01
7.67200053e-01 3.41547757e-01 3.40493321e-01 9.75790620e-02
1.09391069e+00 6.60260439e-01 1.42780626e+00 6.72779322e-01
4.84527439e-01 7.70091653e-01 1.28326476e+00 5.27891368e-02
1.45334035e-01 -1.83170393e-01 5.16184568e-01 1.11894333e+00
-4.98964190e-01 -2.66098768e-01 -9.26264465e-01 3.62018138e-01
-2.27900219e+00 -5.42522848e-01 -3.76498342e-01 1.86580467e+00
-1.19633488e-01 3.28630984e-01 2.63917029e-01 -9.21829417e-02
6.81042910e-01 -5.65154962e-02 -7.02252626e-01 5.00596315e-02
5.75278997e-01 -2.19910145e-01 7.81670570e-01 3.40430379e-01
-1.28662252e+00 9.51406896e-01 5.71433926e+00 3.31558436e-01
-6.37421131e-01 -1.13410316e-01 -4.85999547e-02 -1.30084217e-01
7.00301826e-01 -3.34184051e-01 -5.09952843e-01 9.52682197e-02
6.38855696e-01 2.45005682e-01 6.11909568e-01 1.30963850e+00
-1.68474749e-01 -1.23963647e-01 -1.21477532e+00 1.18744409e+00
-5.12518287e-02 -1.08028841e+00 -4.63144511e-01 2.17348605e-01
5.25648177e-01 2.25772321e-01 -1.34677887e-01 6.56472504e-01
5.27876556e-01 -1.00237715e+00 1.01492381e+00 4.28124875e-01
3.76060009e-01 -8.97776723e-01 1.25025392e+00 7.13641167e-01
-1.58675385e+00 -4.74192291e-01 -4.91137356e-01 -2.28097171e-01
5.98273754e-01 4.46684778e-01 -9.11550820e-01 9.18185771e-01
1.05952036e+00 6.95354342e-01 -2.22202986e-01 5.65809071e-01
-8.24905112e-02 -1.60467729e-01 -3.42455447e-01 -1.99696600e-01
2.75731206e-01 4.46624495e-03 5.58901489e-01 8.19351196e-01
5.08738220e-01 -1.18763767e-01 7.49759436e-01 7.04858363e-01
2.02801332e-01 -4.56564277e-01 -8.52686763e-01 5.64009286e-02
5.46023548e-01 1.27924097e+00 -7.15238452e-01 4.13493663e-02
1.27874911e-01 1.33786190e+00 2.76110291e-01 4.93289828e-02
-1.09132612e+00 -5.27943254e-01 7.11882353e-01 -1.91383839e-01
5.43654263e-01 -9.01184320e-01 2.60075897e-01 -5.04642963e-01
3.00064623e-01 -8.69018972e-01 -2.89175473e-02 -7.43578076e-01
-1.22785187e+00 7.14243293e-01 -6.76120147e-02 -1.45524001e+00
-4.16429549e-01 -1.10803175e+00 -5.56695908e-02 3.53854805e-01
-6.25102758e-01 -1.14691675e+00 -7.87298203e-01 3.31485212e-01
7.31674731e-01 -1.36967182e-01 8.50759923e-01 1.82116285e-01
-1.52127609e-01 2.61682630e-01 -4.25176471e-01 -3.58405188e-02
6.10743105e-01 -1.08277404e+00 1.04642546e+00 1.52626336e-01
9.53269750e-02 4.80616361e-01 7.04277158e-01 -8.26867521e-01
-1.92663193e+00 -1.07213843e+00 5.50018966e-01 -7.13121533e-01
6.28657639e-01 -7.61990130e-01 -4.11524355e-01 7.97720075e-01
-1.65280759e-01 3.70215267e-01 -2.53982812e-01 9.52693522e-02
3.57646607e-02 1.06664740e-01 -9.00741458e-01 4.35675949e-01
1.86954832e+00 -2.18273669e-01 -1.23447023e-01 8.85234952e-01
1.08794165e+00 -1.11947083e+00 -1.16069305e+00 7.64856219e-01
6.47456884e-01 -1.06128418e+00 1.03719461e+00 -3.88726413e-01
2.54599899e-01 -3.56817603e-01 -1.54979438e-01 -1.26375961e+00
-4.85854894e-01 -8.74484181e-01 -1.99632719e-01 4.66439247e-01
7.23657236e-02 -3.23268950e-01 8.38640809e-01 -1.50992945e-01
-5.75771272e-01 -9.70240712e-01 -8.98042679e-01 -1.22330666e+00
-4.53260690e-01 -5.21140695e-01 3.87601703e-01 5.95440030e-01
5.60150295e-02 3.50093156e-01 -5.25731027e-01 4.74694878e-01
3.85007232e-01 6.38141716e-03 1.40255868e+00 -1.33109152e+00
-3.17067593e-01 1.40246734e-01 -7.12697268e-01 -1.47792029e+00
1.15726285e-01 -5.54233253e-01 7.38014221e-01 -1.54834497e+00
-2.89322734e-01 -2.84306586e-01 1.78260505e-01 3.03066045e-01
3.28579992e-01 -6.25807941e-02 5.52038491e-01 4.36136648e-02
-9.88189042e-01 5.66777408e-01 1.53295457e+00 -2.30806857e-01
-2.34703701e-02 1.06378235e-01 7.88662434e-02 8.56078804e-01
6.39638186e-01 -3.28377217e-01 -4.09249157e-01 -3.40571463e-01
-1.60822440e-02 4.25845116e-01 7.15198994e-01 -1.64432299e+00
4.07457769e-01 -5.09355664e-02 7.65066221e-02 -9.47548747e-01
7.78470755e-01 -1.00439250e+00 3.80664945e-01 9.14249003e-01
3.73273180e-03 5.13422310e-01 2.04004988e-01 9.08570409e-01
1.16422229e-01 1.52687728e-01 1.52536318e-01 -3.83542985e-01
-9.21867371e-01 2.96076506e-01 -5.46129525e-01 -2.06042916e-01
1.14979303e+00 9.34066474e-02 -3.20995778e-01 -4.25514489e-01
-7.61356056e-01 2.51891613e-01 1.82634383e-01 7.74807215e-01
5.71957350e-01 -1.21405065e+00 -2.76604235e-01 -4.98726964e-02
2.07154810e-01 4.70487654e-01 4.85293269e-01 9.26380992e-01
-8.18905413e-01 3.58705878e-01 -2.17527732e-01 -9.00869131e-01
-1.03391302e+00 7.26147771e-01 3.31492871e-02 -2.20889360e-01
-1.06078351e+00 9.48417664e-01 2.05070853e-01 -9.34318125e-01
6.17598295e-01 -4.92783755e-01 2.03469217e-01 -4.92471963e-01
-4.39020544e-02 7.46383011e-01 4.18185741e-02 -5.85676610e-01
-4.99607712e-01 4.59789872e-01 3.34395349e-01 3.16045314e-01
1.36403656e+00 1.40021801e-01 1.13604441e-01 4.36305702e-01
7.97744215e-01 -2.30100214e-01 -1.51974952e+00 3.25183928e-01
-7.33347051e-03 -3.60067487e-01 -6.21483564e-01 -7.01352358e-01
-1.10274565e+00 6.65879905e-01 6.24156654e-01 1.55810863e-01
4.74547982e-01 2.72188544e-01 9.63152051e-01 1.05536413e+00
1.30769885e+00 -1.07203484e+00 6.18465960e-01 1.09148896e+00
1.31384003e+00 -1.21019113e+00 -1.02866143e-01 -1.00233018e+00
-7.61103928e-01 1.02975988e+00 1.02582383e+00 -3.52287114e-01
4.68651086e-01 4.90093440e-01 3.04737866e-01 -6.54146910e-01
-6.95556402e-01 1.30178452e-01 -5.87325878e-02 9.06803787e-01
-5.72709106e-02 2.71957994e-01 3.80803615e-01 5.25647461e-01
-4.67275351e-01 -3.48998487e-01 1.54058009e-01 1.04645324e+00
-3.21504384e-01 -4.11903232e-01 -1.17653452e-01 6.61071837e-02
3.09266597e-01 5.13795018e-01 -2.95880437e-01 1.09947014e+00
2.02387258e-01 1.12331450e+00 3.67503352e-02 -1.04168344e+00
1.02253008e+00 -5.31990767e-01 5.29143214e-01 -5.36666274e-01
-6.72020495e-01 -2.69694209e-01 2.57362843e-01 -1.47303128e+00
-2.27648631e-01 -5.66423416e-01 -1.45863712e+00 -4.12760139e-01
-3.68707120e-01 -9.74998847e-02 6.07653856e-01 7.06803083e-01
5.16181231e-01 9.69545603e-01 3.89009356e-01 -1.79890609e+00
-6.85014904e-01 -1.14578474e+00 -6.45750761e-01 3.15529853e-01
-8.10510963e-02 -1.17645597e+00 -3.58787477e-01 -2.27182761e-01] | [5.007174015045166, 0.4563662111759186] |
6279ee51-0926-470b-9ef7-45454dbfa923 | prin-pointwise-rotation-invariant-network | 1811.09361 | null | https://arxiv.org/abs/1811.09361v5 | https://arxiv.org/pdf/1811.09361v5.pdf | Pointwise Rotation-Invariant Network with Adaptive Sampling and 3D Spherical Voxel Convolution | Point cloud analysis without pose priors is very challenging in real applications, as the orientations of point clouds are often unknown. In this paper, we propose a brand new point-set learning framework PRIN, namely, Pointwise Rotation-Invariant Network, focusing on rotation-invariant feature extraction in point clouds analysis. We construct spherical signals by Density Aware Adaptive Sampling to deal with distorted point distributions in spherical space. In addition, we propose Spherical Voxel Convolution and Point Re-sampling to extract rotation-invariant features for each point. Our network can be applied to tasks ranging from object classification, part segmentation, to 3D feature matching and label alignment. We show that, on the dataset with randomly rotated point clouds, PRIN demonstrates better performance than state-of-the-art methods without any data augmentation. We also provide theoretical analysis for the rotation-invariance achieved by our methods. | ['Lizhuang Ma', 'Yu-Wing Tai', 'Yang You', 'Cewu Lu', 'Yujing Lou', 'Weiming Wang', 'Qi Liu'] | 2018-11-23 | null | null | null | null | ['3d-feature-matching'] | ['computer-vision'] | [ 1.66074052e-01 -2.74970420e-02 -5.45698181e-02 -4.43184346e-01
-6.98517025e-01 -5.12167156e-01 4.23382282e-01 -8.37912634e-02
-3.19325686e-01 1.86485335e-01 -2.77204126e-01 1.31156191e-01
-3.23661089e-01 -5.85757911e-01 -1.17055237e+00 -6.24747872e-01
1.43009543e-01 1.03238130e+00 1.29974619e-01 6.32707626e-02
2.03713387e-01 1.45961118e+00 -1.55587626e+00 -2.58517355e-01
5.96764326e-01 9.40912724e-01 1.61810547e-01 2.98033655e-01
-7.28166327e-02 -2.79146314e-01 -1.88152909e-01 -1.50914088e-01
3.93338174e-01 4.53718156e-01 -5.96945286e-01 2.34273732e-01
6.91246927e-01 -4.33708876e-02 -2.36434653e-01 1.04790413e+00
2.73209751e-01 1.96944550e-01 1.04299009e+00 -1.13289940e+00
-5.46013653e-01 3.28262061e-01 -7.39200652e-01 -1.54151186e-01
-1.38455331e-01 -2.61848480e-01 8.19084227e-01 -1.26318741e+00
5.46822309e-01 1.41007710e+00 7.65017450e-01 5.02357900e-01
-1.19983232e+00 -5.47362685e-01 9.00502130e-02 -4.06131670e-02
-1.48961687e+00 -1.25495687e-01 1.19063222e+00 -4.69181359e-01
6.25149190e-01 3.02367628e-01 5.45435071e-01 8.96788120e-01
-1.16497345e-01 6.54683173e-01 5.74103892e-01 -2.62420952e-01
3.85421574e-01 -1.50735304e-01 1.83531009e-02 3.31989378e-01
3.66308689e-01 -1.08269766e-01 -1.65089026e-01 -2.74892420e-01
1.48903739e+00 6.00229621e-01 -2.25410655e-01 -1.17757726e+00
-1.45185435e+00 7.76599586e-01 7.56507099e-01 2.15997741e-01
-5.25636435e-01 3.89759243e-01 1.29846364e-01 -1.37013510e-01
8.67284119e-01 3.99788529e-01 -7.30396330e-01 1.54134974e-01
-7.29899049e-01 3.29598427e-01 4.20481801e-01 1.42648733e+00
7.63702035e-01 -9.31025967e-02 -1.39031395e-01 8.66750062e-01
5.65292954e-01 8.95370126e-01 1.26893029e-01 -1.16191280e+00
1.21163420e-01 5.32873154e-01 6.78826645e-02 -8.61823678e-01
-6.33146703e-01 -5.22585511e-01 -1.01279032e+00 -8.27682437e-04
3.32400173e-01 2.76531607e-01 -1.05439675e+00 1.39591193e+00
4.98880416e-01 5.99438548e-01 -3.45937163e-01 9.03547764e-01
6.29962683e-01 1.78418294e-01 -4.92772192e-01 3.63083333e-02
1.26646125e+00 -3.75547022e-01 -1.52765229e-01 6.63948059e-02
6.05820194e-02 -6.77819073e-01 8.47254097e-01 1.96227670e-01
-9.88649547e-01 -3.91564548e-01 -7.97311366e-01 -9.83823761e-02
6.98037669e-02 1.44668091e-02 6.55157089e-01 4.03721690e-01
-8.79773915e-01 7.76585460e-01 -1.18710744e+00 -2.25791603e-01
9.03614461e-01 4.98650312e-01 -3.36629540e-01 -1.78310350e-01
-2.58329272e-01 4.86008734e-01 -2.08170220e-01 2.52512135e-02
-4.59617972e-01 -1.01800776e+00 -1.03348660e+00 4.62481938e-02
1.95174411e-01 -1.03744817e+00 1.31186426e+00 -3.91136944e-01
-1.43366945e+00 8.65981877e-01 -4.63191122e-01 -3.53010237e-01
5.06869137e-01 -3.39588344e-01 2.74320483e-01 2.05339342e-01
1.88565031e-01 6.08754456e-01 1.19543910e+00 -1.42017782e+00
-1.91914305e-01 -9.44258392e-01 -3.68164331e-01 3.32794547e-01
7.75218159e-02 -1.34382844e-01 -5.64482212e-01 -4.88868833e-01
8.76810730e-01 -1.07530463e+00 -4.39636886e-01 3.30804259e-01
-2.47224525e-01 -5.20962536e-01 7.34687805e-01 -1.84590351e-02
5.49696460e-02 -2.33397651e+00 1.74259081e-01 4.60882694e-01
3.39644074e-01 -2.24484667e-01 -1.33326516e-01 -2.20568582e-01
-3.14708233e-01 -3.97955514e-02 -3.24341685e-01 -5.20870805e-01
1.29917666e-01 1.43382668e-01 -3.69394690e-01 1.05511987e+00
4.02778983e-01 1.01167488e+00 -7.56017029e-01 -2.56758332e-01
5.85280299e-01 8.19514751e-01 -4.43535864e-01 -2.22045943e-01
-1.51356339e-01 5.54002285e-01 -5.55655539e-01 7.62744606e-01
1.25062227e+00 -4.05550897e-01 -5.08947611e-01 -3.70450169e-01
8.92865285e-02 -7.52863958e-02 -1.22083700e+00 2.00271964e+00
-4.11785603e-01 3.75681520e-01 1.69517189e-01 -9.13660884e-01
1.00827444e+00 1.71156704e-01 9.87767041e-01 -2.68508643e-02
2.13433206e-01 1.06098942e-01 -2.61377603e-01 7.41154328e-02
4.48675185e-01 -1.23021968e-01 1.93076819e-01 8.46205186e-03
1.38942376e-01 -7.76440680e-01 -3.53301018e-01 -1.84183955e-01
9.17266250e-01 6.10806867e-02 1.83330700e-02 -1.00562304e-01
3.04600686e-01 -3.33736390e-01 4.44249690e-01 5.45668602e-01
7.93830827e-02 1.15448010e+00 8.90282169e-02 -2.56098568e-01
-1.14418125e+00 -1.32177043e+00 -7.83846080e-01 5.82596958e-01
3.05246860e-01 1.95449069e-02 -5.71706951e-01 -4.98573214e-01
3.53153974e-01 5.02143621e-01 -2.47558162e-01 1.52216092e-01
-7.44021177e-01 -5.02287507e-01 1.16199419e-01 8.38572025e-01
1.30067453e-01 -7.79279351e-01 -3.49425137e-01 4.25302573e-02
1.77608535e-01 -1.26309204e+00 -4.55977350e-01 1.03145882e-01
-1.29445624e+00 -9.46145833e-01 -1.06151891e+00 -6.81472063e-01
8.51158738e-01 5.58876038e-01 1.11498034e+00 -4.16028678e-01
-1.72992885e-01 8.70622814e-01 -2.48374462e-01 -6.90093637e-01
1.59155682e-01 2.15851486e-01 1.89428881e-01 -1.06457874e-01
5.20095587e-01 -9.31694984e-01 -5.40742457e-01 6.43026531e-01
-6.77836955e-01 -3.45211297e-01 6.11547530e-01 5.94209969e-01
1.16087103e+00 -2.11473987e-01 1.83667064e-01 -5.16382337e-01
1.28421068e-01 -2.43728861e-01 -7.69777954e-01 -1.80865586e-01
-1.55688534e-02 8.59171674e-02 1.75926834e-01 -2.78130025e-01
-3.78037244e-01 5.13208568e-01 -7.77141973e-02 -1.24529099e+00
-4.48845834e-01 2.44234074e-02 -1.14496738e-01 -4.81054902e-01
6.52291596e-01 1.45351276e-01 1.99913546e-01 -5.91189265e-01
6.18671298e-01 3.91861409e-01 5.65831244e-01 -7.22361326e-01
1.25348246e+00 1.03614759e+00 4.83976603e-01 -1.04838705e+00
-4.98019129e-01 -9.50867593e-01 -1.10365450e+00 2.76678731e-03
7.88141370e-01 -9.89508092e-01 -9.18154657e-01 3.69946271e-01
-1.49358284e+00 2.61898693e-02 -7.09692299e-01 6.30536735e-01
-9.73745048e-01 4.07636017e-01 -2.60812044e-01 -7.04392254e-01
-3.85083944e-01 -1.13843846e+00 1.63684750e+00 5.71715273e-02
1.69385523e-01 -6.89962327e-01 3.46343257e-02 -1.74290240e-02
8.86673406e-02 1.06738828e-01 4.96033847e-01 -6.93356812e-01
-1.00536823e+00 -2.56008923e-01 -3.88235599e-01 3.23415548e-01
7.28318095e-02 -1.01850979e-01 -9.58066106e-01 -2.57930368e-01
2.53792286e-01 7.25794360e-02 6.39927328e-01 8.24732661e-01
1.75755894e+00 1.51128620e-01 -6.22880340e-01 1.08295524e+00
1.11581683e+00 -3.28518122e-01 3.73935819e-01 8.08434281e-03
1.00808668e+00 3.83017361e-01 6.27427042e-01 3.73557866e-01
2.43339360e-01 7.39571691e-01 7.96887100e-01 -1.59746886e-03
2.13186249e-01 -5.70661873e-02 -1.59238875e-01 5.56163311e-01
-4.04460698e-01 2.26808935e-01 -7.78617918e-01 5.06786764e-01
-1.88153291e+00 -6.55737698e-01 -2.46738195e-01 2.27415299e+00
3.07755768e-01 -1.43538207e-01 -5.42136990e-02 -6.86218292e-02
7.71655202e-01 -1.34162217e-01 -8.53498518e-01 3.02143753e-01
-4.59764898e-03 5.85543394e-01 8.73650908e-01 1.08955599e-01
-1.37517953e+00 7.55277872e-01 5.86595011e+00 7.78412580e-01
-1.12342322e+00 4.35057543e-02 1.70202985e-01 -4.79907878e-02
-3.47973108e-01 -3.72010112e-01 -7.18733490e-01 2.43698299e-01
3.43523681e-01 1.90961715e-02 3.64253879e-01 1.30549133e+00
-1.62643999e-01 4.98915583e-01 -1.33777046e+00 1.50105834e+00
1.94852650e-01 -1.37510824e+00 -9.73117650e-02 2.12493762e-01
5.48928201e-01 6.77569509e-01 1.85527802e-01 -5.85892349e-02
1.02092698e-01 -1.02294874e+00 5.75550854e-01 5.38441956e-01
8.17803323e-01 -9.21200454e-01 5.50506234e-01 3.21761698e-01
-1.13812602e+00 3.72845352e-01 -8.09278369e-01 3.01040769e-01
1.70934558e-01 8.47811818e-01 -1.06042933e+00 7.65994191e-01
8.31842959e-01 1.02930272e+00 -3.18163007e-01 1.39071321e+00
-5.25027476e-02 2.89348513e-01 -8.88075888e-01 9.29423794e-02
-2.37248391e-01 -4.33374941e-01 9.03369427e-01 7.53203094e-01
5.50699651e-01 -9.98325571e-02 -6.00497350e-02 1.22659826e+00
-1.64025873e-01 -8.65710229e-02 -7.81039536e-01 5.99401891e-01
5.74342072e-01 1.41027939e+00 -9.13735151e-01 4.83869538e-02
-2.50695765e-01 8.04207623e-01 1.74658105e-01 2.72579312e-01
-5.91956258e-01 -2.26652756e-01 9.13359523e-01 2.92495728e-01
6.62156284e-01 -6.66372657e-01 -4.07207519e-01 -1.22468185e+00
1.61599979e-01 -3.04911882e-01 -1.94251359e-01 -8.19489241e-01
-1.72792435e+00 2.87670106e-01 3.16733927e-01 -1.37898672e+00
-6.30030185e-02 -1.02217996e+00 -5.40515602e-01 8.10490549e-01
-1.50247574e+00 -1.20574534e+00 -2.48641044e-01 6.76757038e-01
5.45464873e-01 -6.24615178e-02 4.82078731e-01 1.89853758e-01
-1.95989795e-02 5.31554580e-01 2.06523433e-01 2.47713283e-01
5.56873500e-01 -1.36715710e+00 6.92135751e-01 4.47211355e-01
4.28326219e-01 7.86851406e-01 4.39198107e-01 -5.06423354e-01
-1.50728619e+00 -1.29994845e+00 9.42581743e-02 -7.70777643e-01
3.15511912e-01 -5.39857566e-01 -9.57920969e-01 8.13983083e-01
-5.16215563e-01 4.47261781e-01 3.93035620e-01 2.29157344e-01
-3.07323903e-01 4.15247269e-02 -1.40715766e+00 3.95899057e-01
1.21561790e+00 -2.94039965e-01 -5.53781867e-01 7.00655997e-01
1.02311313e+00 -7.88729608e-01 -8.70264888e-01 7.30084836e-01
1.95454910e-01 -4.15258616e-01 1.51656055e+00 -3.95297706e-01
1.15465457e-02 -3.87948215e-01 -3.15835446e-01 -1.32298565e+00
-5.98564148e-01 -3.87852579e-01 3.64262750e-03 8.78973782e-01
8.15080851e-02 -8.01604152e-01 1.11985838e+00 3.22420716e-01
-5.19528985e-01 -4.89435434e-01 -1.39134860e+00 -8.68128121e-01
3.46449405e-01 -7.87993670e-01 9.12200391e-01 8.89798582e-01
-4.47710276e-01 2.26374373e-01 3.44381213e-01 5.26236892e-01
9.32564676e-01 1.54930875e-01 9.47289288e-01 -1.76329088e+00
2.17024740e-02 -5.22999883e-01 -8.38073730e-01 -1.31962442e+00
5.21689117e-01 -1.09892559e+00 4.82166559e-02 -1.26280248e+00
2.76379660e-02 -6.88819706e-01 7.50715733e-02 1.75388396e-01
2.04171494e-01 1.76540598e-01 1.86069801e-01 2.75393397e-01
-3.26030672e-01 8.35862279e-01 1.39239872e+00 -1.72344685e-01
-1.02326587e-01 5.20249903e-01 -4.45653021e-01 1.07093334e+00
6.45547271e-01 -3.98206085e-01 -1.17900923e-01 -4.24624652e-01
9.49151143e-02 -2.14437053e-01 6.92505777e-01 -1.07222116e+00
1.68339849e-01 -1.77261829e-01 5.58367848e-01 -1.24253392e+00
7.15753138e-01 -1.17244112e+00 -8.44999477e-02 -1.53091192e-01
1.61582112e-01 -7.70558640e-02 1.20645739e-01 7.08738983e-01
1.65647075e-01 -2.08027676e-01 7.94421852e-01 -3.52211036e-02
-4.33831774e-02 8.35903406e-01 3.32764685e-01 -2.74605423e-01
9.24921870e-01 -1.57029822e-01 -3.67805287e-02 -5.87003417e-02
-6.98008716e-01 7.08035827e-02 8.23081732e-01 2.93832242e-01
9.38950300e-01 -1.46403050e+00 -6.64119244e-01 4.24901992e-01
1.76149160e-01 1.17443562e+00 1.83798075e-01 7.31415272e-01
-5.72251439e-01 4.02187318e-01 8.04832056e-02 -1.36148131e+00
-1.06942105e+00 5.62703192e-01 2.69100845e-01 2.51011133e-01
-8.40511441e-01 7.97295749e-01 2.63666511e-01 -9.68122542e-01
8.58548060e-02 -8.79757047e-01 -3.40269767e-02 -3.28236252e-01
1.07174985e-01 3.07568312e-01 2.52572864e-01 -6.72764122e-01
-4.59832102e-01 1.13670659e+00 -1.31003829e-02 -7.22602615e-03
1.50104797e+00 3.02689046e-01 -1.73110649e-01 5.61823368e-01
1.18197322e+00 -6.10128641e-02 -1.23872054e+00 -4.50423896e-01
-2.48873398e-01 -6.20357394e-01 -2.12689769e-02 -1.38086557e-01
-1.09421945e+00 8.01887751e-01 5.13234913e-01 -9.46660526e-03
4.62030381e-01 5.34529209e-01 3.30936760e-01 5.07514417e-01
5.32554090e-01 -5.68963170e-01 -2.30965927e-01 7.18527794e-01
1.15628350e+00 -1.18608177e+00 2.64437888e-02 -6.93916380e-01
-1.61235332e-01 1.05898416e+00 4.26859140e-01 -6.86069191e-01
1.06253684e+00 -5.30051813e-02 -3.61760050e-01 -3.49804968e-01
-2.11242542e-01 -8.46534744e-02 6.08591914e-01 9.20467317e-01
4.71661873e-02 1.72797233e-01 2.14278147e-01 3.49314630e-01
-5.04508376e-01 -1.36058956e-01 2.19410643e-01 6.67935789e-01
-2.28066280e-01 -7.98409283e-01 -8.28501105e-01 5.20033240e-01
-7.98516646e-02 1.85729325e-01 -2.49289662e-01 6.34753823e-01
-2.70534217e-01 2.80381113e-01 7.55399942e-01 2.36583408e-02
5.28675020e-01 -4.78164665e-02 8.04376006e-01 -7.96296060e-01
1.04904763e-01 2.55768776e-01 -7.10206091e-01 -5.18049061e-01
-5.20374179e-01 -1.02872264e+00 -1.29930925e+00 -5.75577654e-02
-2.45953962e-01 -1.40092954e-01 1.14813840e+00 6.59957945e-01
5.52165568e-01 4.56137300e-01 6.90037549e-01 -1.68697345e+00
-7.06378818e-01 -9.85520124e-01 -6.56681895e-01 2.31431767e-01
4.13908869e-01 -9.56984580e-01 -4.96676832e-01 -4.02622700e-01] | [7.961167812347412, -3.4737086296081543] |
8283d072-88b3-4a8a-91e2-de68a5eacb2f | cross-language-learning-with-adversarial | null | null | https://aclanthology.org/K17-1024 | https://aclanthology.org/K17-1024.pdf | Cross-language Learning with Adversarial Neural Networks | We address the problem of cross-language adaptation for question-question similarity reranking in community question answering, with the objective to port a system trained on one input language to another input language given labeled training data for the first language and only unlabeled data for the second language. In particular, we propose to use adversarial training of neural networks to learn high-level features that are discriminative for the main learning task, and at the same time are invariant across the input languages. The evaluation results show sizable improvements for our cross-language adversarial neural network (CLANN) model over a strong non-adversarial system. | ["Llu{\\'\\i}s M{\\`a}rquez", 'Preslav Nakov', 'Shafiq Joty', 'Israa Jaradat'] | 2017-08-01 | null | null | null | conll-2017-8 | ['question-similarity'] | ['natural-language-processing'] | [ 2.77895570e-01 1.91533178e-01 2.18656778e-01 -4.36876327e-01
-1.33171487e+00 -8.60007286e-01 7.12461710e-01 2.68528789e-01
-7.39540815e-01 4.53711450e-01 3.75353187e-01 -5.26295424e-01
2.31227770e-01 -7.49990940e-01 -5.94134450e-01 -1.47838384e-01
1.54947877e-01 8.34268332e-01 5.00905395e-01 -5.04260898e-01
-3.33523154e-01 2.76304901e-01 -8.79975736e-01 6.87609315e-01
8.22145641e-01 8.02486777e-01 -2.00767070e-01 1.18574333e+00
-2.77172983e-01 1.26503444e+00 -6.60868585e-01 -7.23720551e-01
1.63931087e-01 -5.28084874e-01 -1.55237460e+00 -2.44679332e-01
1.18027246e+00 -1.13763280e-01 -8.00879240e-01 8.94835591e-01
3.12148273e-01 1.54930919e-01 7.09892809e-01 -1.16569316e+00
-1.20313513e+00 2.50641167e-01 1.01711236e-01 4.79591966e-01
5.28645813e-01 1.12390473e-01 1.56170309e+00 -6.92681551e-01
3.66173625e-01 1.43078208e+00 3.34922105e-01 9.74492073e-01
-1.28725505e+00 -4.05119956e-01 1.86112836e-01 2.84810662e-01
-1.09167647e+00 -2.90045142e-01 6.76322818e-01 -4.44417417e-01
6.76887631e-01 4.43305254e-01 -3.06512296e-01 1.05361044e+00
-1.44619405e-01 7.53472388e-01 9.08488631e-01 -3.99730027e-01
6.36884868e-02 2.01096907e-01 5.57538629e-01 6.55397654e-01
-1.21171415e-01 -3.45107794e-01 -3.98295149e-02 -7.77639151e-01
-7.16337711e-02 -1.03101693e-01 -2.55184114e-01 -2.41736129e-01
-8.82798910e-01 1.03162801e+00 6.11150265e-01 7.19712973e-01
4.57589813e-02 8.07262808e-02 4.46605057e-01 8.26592863e-01
3.96001667e-01 8.73765349e-01 -6.87364459e-01 5.47931492e-01
-5.65207481e-01 4.39205766e-01 1.03380978e+00 5.08476198e-01
6.42551184e-01 -1.99863583e-01 -7.27254927e-01 9.89305615e-01
9.73053575e-02 6.94574952e-01 7.40471363e-01 -7.99086452e-01
7.59548068e-01 5.87869525e-01 -2.57149518e-01 -6.93129838e-01
-1.13864437e-01 -1.75309032e-01 -8.64793301e-01 -1.91073954e-01
6.79893196e-01 -1.01091817e-01 -5.29168546e-01 2.00023937e+00
-9.72556230e-03 3.98373827e-02 4.35856044e-01 5.48911214e-01
1.08041167e+00 5.85597157e-01 3.59777629e-01 4.04693455e-01
1.44164145e+00 -1.36778653e+00 -2.98658192e-01 -5.60731769e-01
5.46163559e-01 -6.06999040e-01 1.41521943e+00 -3.38860571e-01
-8.38654399e-01 -7.78479695e-01 -7.78264999e-01 -3.91130656e-01
-5.02563119e-01 -2.16414899e-01 9.61818993e-02 5.65701485e-01
-1.16861916e+00 2.53308773e-01 -2.40784064e-01 -5.21125555e-01
1.23454146e-01 1.55034572e-01 -3.22753817e-01 -3.06117296e-01
-1.46846807e+00 8.67411494e-01 3.73027593e-01 -5.24513125e-01
-9.15870786e-01 -6.82706714e-01 -7.90905118e-01 1.76461801e-01
5.73552214e-03 -7.45416164e-01 1.57097220e+00 -1.34918952e+00
-1.21344924e+00 1.12549603e+00 6.62609935e-02 -4.96614456e-01
2.47143313e-01 -5.03806630e-03 -4.93017703e-01 4.74901080e-01
3.08744550e-01 4.31322396e-01 8.03973079e-01 -1.09072435e+00
-2.77669966e-01 -2.57303953e-01 4.36190277e-01 -1.44142611e-02
-6.69958591e-01 1.34632289e-01 -3.58128309e-01 -7.14053094e-01
-3.92157346e-01 -6.92529440e-01 -1.98938221e-01 7.66972005e-02
-4.65290844e-02 -6.44301116e-01 7.51145661e-01 -1.19209790e+00
8.45677018e-01 -1.80183887e+00 4.95356321e-02 4.80564870e-02
3.67432982e-01 4.93975043e-01 -8.84371459e-01 2.72152603e-01
-3.40183109e-01 7.99020231e-02 -2.97696918e-01 -1.58937454e-01
-1.93450570e-01 2.67862201e-01 -4.75603074e-01 2.20285982e-01
5.80807686e-01 1.13710284e+00 -1.25990772e+00 -5.38098693e-01
-3.61730903e-01 -1.19496115e-01 -7.05401480e-01 1.01855147e+00
-4.32972550e-01 3.46111298e-01 -5.43095410e-01 1.20039813e-01
3.75150234e-01 -4.98145580e-01 1.19625963e-01 2.66394615e-02
8.55023623e-01 5.32930911e-01 -6.18596792e-01 1.29504490e+00
-8.27038765e-01 5.30995965e-01 1.07209638e-01 -9.96319294e-01
7.71717429e-01 4.73883718e-01 1.33008435e-01 -6.44965351e-01
-1.75089866e-01 2.57342458e-01 -8.90026987e-02 -6.02193356e-01
3.58499348e-01 -3.02172840e-01 -3.20635408e-01 6.73176289e-01
4.41084296e-01 -2.02080488e-01 -1.25409380e-01 5.39452434e-01
1.32452488e+00 -4.68021512e-01 3.87609154e-02 -1.98920295e-01
1.21433616e+00 -2.99601972e-01 -7.12951720e-02 8.49280238e-01
-2.79529393e-01 7.43976116e-01 2.97742963e-01 -2.18751371e-01
-1.01050818e+00 -1.12905097e+00 2.76066482e-01 1.78157127e+00
-1.53279722e-01 -1.10462874e-01 -6.93822443e-01 -1.51529455e+00
2.18869761e-01 7.80290782e-01 -7.49489605e-01 -4.92559791e-01
-8.73074889e-01 1.29080620e-02 1.13506854e+00 3.62133116e-01
3.67861837e-01 -9.14340317e-01 2.05688775e-01 -1.94529481e-02
-3.74189973e-01 -1.35393012e+00 -1.05108869e+00 -5.46454415e-02
-5.03157973e-01 -1.40695083e+00 -9.48677421e-01 -1.26713753e+00
6.57266855e-01 6.99492171e-02 1.65740895e+00 4.13000941e-01
2.12161198e-01 9.45419550e-01 -2.06632078e-01 1.73419178e-01
-9.67333198e-01 4.85286951e-01 -5.61291650e-02 2.63376176e-01
3.65336061e-01 -4.70221281e-01 -1.97945327e-01 2.08126843e-01
-1.21594214e+00 -4.72242892e-01 5.42408042e-02 1.02176630e+00
2.55245864e-01 -6.50361657e-01 8.26588631e-01 -9.30185735e-01
1.10207367e+00 -6.63816392e-01 -3.44017684e-01 1.00376749e+00
-2.21813247e-01 5.13983727e-01 1.17717481e+00 -6.42538428e-01
-9.06732321e-01 -3.65681857e-01 -4.80208606e-01 -2.98530757e-01
-6.29034340e-02 1.15044400e-01 -3.47175270e-01 -5.71719445e-02
1.03611755e+00 7.93789774e-02 -2.49285847e-01 -6.78450584e-01
7.48773217e-01 7.92625129e-01 8.15303087e-01 -7.81026483e-01
1.20793414e+00 2.28023455e-01 -4.68482643e-01 -5.57600915e-01
-1.10090303e+00 -7.07622409e-01 -6.85742617e-01 1.93919390e-01
1.08412099e+00 -7.39912808e-01 -6.28957152e-01 2.49524415e-01
-1.44736755e+00 -2.71586537e-01 -6.03512228e-01 -6.21259362e-02
-2.29746208e-01 6.09635890e-01 -5.47792852e-01 -3.46764326e-01
-6.15880132e-01 -7.63086915e-01 7.98021913e-01 7.16419220e-02
-1.75215542e-01 -1.53634751e+00 5.39644599e-01 9.94704127e-01
4.77000684e-01 -2.59732395e-01 1.29522896e+00 -1.82656515e+00
-2.91696012e-01 -3.26505601e-01 -1.83780953e-01 9.16156292e-01
2.25183651e-01 -6.71587884e-01 -9.58118200e-01 -5.30638814e-01
1.50858998e-01 -8.51009965e-01 6.75214767e-01 -6.12286925e-01
1.17392874e+00 -5.41764021e-01 1.60430282e-01 1.64474428e-01
9.92911220e-01 -5.82645953e-01 2.96617329e-01 1.70117050e-01
8.19817662e-01 6.18887186e-01 -7.06371590e-02 -5.31684458e-01
5.82340300e-01 6.98713720e-01 1.05227947e-01 -2.31738895e-01
-4.09336478e-01 -5.55930734e-01 3.03696990e-01 1.10501623e+00
7.01010823e-01 -3.30751002e-01 -1.02989984e+00 7.75841713e-01
-1.39420938e+00 -7.87023485e-01 8.50652605e-02 2.01441908e+00
1.01237857e+00 -2.61336118e-01 -8.61576200e-02 -3.85442436e-01
6.85232401e-01 4.38334912e-01 -4.61321294e-01 -4.07175571e-01
-2.24049479e-01 6.41629994e-01 2.22153038e-01 8.00235391e-01
-1.21308279e+00 7.25288332e-01 6.50149822e+00 7.35076368e-01
-7.70543277e-01 6.68663800e-01 4.76148546e-01 4.95395422e-01
-7.38356292e-01 -5.55914454e-03 -2.37989336e-01 1.63724214e-01
1.05326140e+00 -3.82772863e-01 3.39739650e-01 7.60489404e-01
-5.50055981e-01 4.93279070e-01 -1.39922738e+00 5.40388882e-01
4.54896390e-01 -1.08146906e+00 6.04426026e-01 -4.59979862e-01
8.57484579e-01 2.39522725e-01 -1.27927423e-01 7.09918201e-01
5.83299220e-01 -1.10909772e+00 1.92774206e-01 4.31959122e-01
6.34901047e-01 -3.47375691e-01 8.19852591e-01 4.31095958e-01
-9.40425515e-01 1.96864173e-01 -1.47950321e-01 8.59725177e-02
-1.72700733e-03 -8.34232867e-02 -8.22394609e-01 5.67467213e-01
4.57026482e-01 1.68153375e-01 -1.09755123e+00 5.76963544e-01
-2.26813674e-01 8.94955814e-01 -1.58504229e-02 -6.24832623e-02
3.01576972e-01 1.25097215e-01 4.50928032e-01 1.16785264e+00
-3.56460243e-01 -1.66402236e-01 4.45659339e-01 6.89063311e-01
-8.59418392e-01 4.12587285e-01 -5.47147155e-01 -4.65205610e-02
-4.89060283e-02 8.69681120e-01 1.82958305e-01 -5.81529558e-01
-5.49955010e-01 1.32327223e+00 7.77963758e-01 4.98666704e-01
-3.51132810e-01 -3.30206096e-01 6.21744752e-01 3.04009169e-02
2.93539409e-02 2.45297477e-02 1.18131451e-01 -1.52609968e+00
1.02581143e-01 -1.37826967e+00 9.56340551e-01 -5.82111776e-01
-2.19718862e+00 8.12582850e-01 -3.25752378e-01 -8.40884924e-01
-4.50609207e-01 -6.04627013e-01 -7.49331295e-01 1.28826213e+00
-1.59026408e+00 -1.33051372e+00 5.13305888e-02 9.50692892e-01
5.68057358e-01 -6.99407935e-01 1.14911771e+00 4.55312073e-01
2.44600102e-02 1.13457441e+00 4.24940258e-01 6.83097720e-01
8.90032649e-01 -1.34755039e+00 7.78470337e-01 8.70453835e-01
6.14401877e-01 4.79612172e-01 3.42942208e-01 -1.86741650e-01
-9.65794444e-01 -1.37958550e+00 1.39914250e+00 -9.17773247e-01
1.06606269e+00 -3.51673305e-01 -1.42893267e+00 6.77431345e-01
4.22821581e-01 2.39797816e-01 7.65519857e-01 5.52473329e-02
-1.09920025e+00 -2.68813297e-02 -1.16703391e+00 4.48505342e-01
2.76868433e-01 -1.52024162e+00 -1.24157023e+00 6.47934616e-01
1.12872601e+00 -1.28626190e-02 -8.05829644e-01 1.28175825e-01
1.56430781e-01 -3.58627915e-01 1.21670365e+00 -1.63735795e+00
2.31479004e-01 -6.37276471e-02 -2.96971262e-01 -8.42111170e-01
-3.43846083e-01 -3.30917358e-01 1.15096718e-01 1.36857677e+00
4.94135857e-01 -6.55500412e-01 5.41220903e-01 4.89372522e-01
4.26539630e-01 -3.55142921e-01 -1.17968845e+00 -6.80557013e-01
8.46211016e-01 -2.25735769e-01 3.92955542e-01 1.03798866e+00
-3.59443128e-01 1.12377095e+00 -2.01967880e-01 3.83338511e-01
3.20836037e-01 6.07976690e-02 6.49714291e-01 -1.00561881e+00
-5.37528217e-01 -1.67632788e-01 -1.52562633e-01 -1.26155376e+00
8.73507738e-01 -1.48989475e+00 -4.39271182e-02 -1.39628780e+00
3.42651546e-01 -2.33154774e-01 -4.72154319e-01 2.68399537e-01
-8.39382052e-01 1.22127622e-01 3.31644386e-01 1.76414609e-01
-1.00618184e+00 7.19233036e-01 1.05391777e+00 -7.49176502e-01
3.26167285e-01 3.51108819e-01 -7.00053096e-01 5.70191562e-01
5.96618056e-01 -6.91917896e-01 -4.00232673e-01 -7.03838766e-01
-1.00123778e-01 -5.44360802e-02 4.98943686e-01 -6.35823071e-01
1.72731951e-01 5.71148694e-02 -2.07447708e-01 -1.60270631e-02
2.24239156e-02 -8.58295918e-01 -6.73447311e-01 5.33567190e-01
-1.10859287e+00 1.15693673e-01 9.80392098e-02 6.50126338e-01
-3.12139392e-01 -6.54409885e-01 8.47677290e-01 3.01182494e-02
-2.63633013e-01 3.51405799e-01 -1.28328741e-01 1.14007831e+00
3.91788006e-01 6.10480070e-01 -4.04062808e-01 -7.61871457e-01
-5.89922905e-01 5.17097116e-01 1.96303576e-01 8.93440843e-01
2.22813591e-01 -1.45686781e+00 -1.22825694e+00 -1.68813124e-01
4.87198770e-01 -5.09637117e-01 1.63648594e-02 -5.84362075e-02
-4.21359897e-01 4.67843533e-01 1.65796876e-01 -3.17324132e-01
-1.26020503e+00 7.24097013e-01 9.18969393e-01 -1.02857757e+00
9.86214355e-02 9.85152900e-01 2.66235828e-01 -1.33147895e+00
1.13612533e-01 -2.40287427e-02 -2.72793978e-01 -2.93092966e-01
3.94195855e-01 1.14398852e-01 1.74019948e-01 -7.71157026e-01
-3.57070565e-01 2.59196758e-01 -1.32205710e-01 -8.18398502e-03
9.46740508e-01 1.04318768e-01 -3.40948880e-01 3.01447511e-01
1.76835132e+00 2.27621406e-01 -4.64297652e-01 -9.05581713e-01
3.28068793e-01 -5.09772152e-02 -4.91469920e-01 -7.05327392e-01
-8.18406343e-01 9.16087449e-01 7.18071222e-01 2.35101417e-01
7.27942467e-01 4.89304900e-01 8.65817785e-01 9.64123964e-01
-1.56840131e-01 -6.34112120e-01 5.88503420e-01 9.33077395e-01
1.14405036e+00 -1.25117159e+00 -3.74635398e-01 -1.81148201e-03
-4.59874719e-01 8.51299047e-01 6.99724972e-01 -2.73639739e-01
5.50156951e-01 -3.55999678e-01 3.23616326e-01 -8.70881453e-02
-6.79476976e-01 -2.00324625e-01 7.76183009e-01 6.72753692e-01
3.99061441e-01 -5.27352691e-02 5.45393527e-02 5.64106464e-01
-2.78488472e-02 -4.76117611e-01 -2.20645145e-02 4.40467805e-01
-5.20414785e-02 -1.48221040e+00 -3.35919797e-01 3.41063380e-01
-7.15039611e-01 -2.50667959e-01 -5.84192157e-01 4.78782594e-01
-2.79569030e-02 1.11602807e+00 -6.12035096e-02 -3.27102751e-01
6.16005182e-01 4.18670833e-01 1.50295466e-01 -8.34945500e-01
-1.26570582e+00 -8.74095500e-01 2.71513224e-01 -1.69435531e-01
-2.42469549e-01 -3.29567730e-01 -7.40696430e-01 2.59337634e-01
-1.53377146e-01 6.81579173e-01 5.73471561e-03 1.19381952e+00
1.73113257e-01 4.21516627e-01 8.77678990e-01 2.09998414e-02
-1.05843675e+00 -1.09118426e+00 -4.18416299e-02 1.08825397e+00
6.50364876e-01 9.01485160e-02 -4.78451371e-01 2.08915740e-01] | [11.301753044128418, 8.182782173156738] |
2227f9d7-935f-411f-b77f-4ae14431419b | a-framework-for-adapting-offline-algorithms | 2301.13326 | null | https://arxiv.org/abs/2301.13326v1 | https://arxiv.org/pdf/2301.13326v1.pdf | A Framework for Adapting Offline Algorithms to Solve Combinatorial Multi-Armed Bandit Problems with Bandit Feedback | We investigate the problem of stochastic, combinatorial multi-armed bandits where the learner only has access to bandit feedback and the reward function can be non-linear. We provide a general framework for adapting discrete offline approximation algorithms into sublinear $\alpha$-regret methods that only require bandit feedback, achieving $\mathcal{O}\left(T^\frac{2}{3}\log(T)^\frac{1}{3}\right)$ expected cumulative $\alpha$-regret dependence on the horizon $T$. The framework only requires the offline algorithms to be robust to small errors in function evaluation. The adaptation procedure does not even require explicit knowledge of the offline approximation algorithm -- the offline algorithm can be used as black box subroutine. To demonstrate the utility of the proposed framework, the proposed framework is applied to multiple problems in submodular maximization, adapting approximation algorithms for cardinality and for knapsack constraints. The new CMAB algorithms for knapsack constraints outperform a full-bandit method developed for the adversarial setting in experiments with real-world data. | ['Christopher John Quinn', 'Vaneet Aggarwal', 'Yanhui Zhu', 'Yididiya Y Nadew', 'Guanyu Nie'] | 2023-01-30 | null | null | null | null | ['multi-armed-bandits'] | ['miscellaneous'] | [ 1.01217486e-01 2.94511288e-01 -5.29397190e-01 -3.06909770e-01
-1.35678220e+00 -1.15979898e+00 -2.90413082e-01 1.47738039e-01
-7.63245106e-01 1.58503330e+00 -4.65583801e-01 -6.25095963e-01
-7.36808777e-01 -8.39024603e-01 -1.43436420e+00 -1.08874357e+00
-2.31798872e-01 6.09854519e-01 -3.47764105e-01 -1.15989439e-01
1.76949501e-01 3.77574712e-01 -1.09962916e+00 4.74650599e-02
9.24078882e-01 1.87505233e+00 -1.49486408e-01 7.42497146e-01
-8.41106400e-02 5.44530749e-01 -5.02410293e-01 -7.78434753e-01
9.56951022e-01 -3.02349806e-01 -6.55219793e-01 1.77404910e-01
4.48243707e-01 -6.36514485e-01 -2.56940246e-01 1.27704763e+00
2.38334000e-01 3.33162516e-01 2.92318106e-01 -1.01971078e+00
-3.00968617e-01 8.66114736e-01 -6.02330744e-01 3.37707162e-01
4.79819402e-02 -6.49846196e-02 1.20634460e+00 -7.34224319e-02
3.58679414e-01 8.80098164e-01 3.43920469e-01 3.83480936e-01
-1.06549847e+00 -7.08277225e-01 5.42830706e-01 9.46481079e-02
-9.86953318e-01 -1.10837035e-01 4.33162987e-01 -6.09100685e-02
4.97395009e-01 9.72780943e-01 7.25093007e-01 3.94271255e-01
-3.05812508e-01 1.07190728e+00 1.36744571e+00 -3.47955942e-01
5.36835611e-01 4.48149979e-01 2.16616362e-01 5.28717637e-01
6.05013907e-01 2.07374588e-01 -2.94908971e-01 -2.63497651e-01
4.39353704e-01 3.60427797e-02 -3.58306706e-01 -2.56371737e-01
-3.40233743e-01 1.00368464e+00 3.76038283e-01 -6.81682080e-02
-3.73869061e-01 6.04198396e-01 1.09497041e-01 7.38078177e-01
6.17647171e-01 -9.34191979e-03 -5.47085702e-01 -1.08358547e-01
-1.23563218e+00 5.86496294e-01 9.16204095e-01 1.48477650e+00
6.04939699e-01 2.17412621e-01 -3.09089214e-01 5.08597553e-01
-1.68984845e-01 7.10936666e-01 4.55004536e-02 -9.07853484e-01
1.08541417e+00 8.35025311e-02 9.68093872e-01 -5.82159281e-01
-1.19908653e-01 -8.50200653e-01 -4.82406020e-01 1.17652722e-01
6.11193895e-01 -4.79048818e-01 -6.58258319e-01 1.49967957e+00
5.90199828e-01 -5.10290980e-01 -8.32486451e-02 9.82602894e-01
2.06651501e-02 7.82286942e-01 -5.57907343e-01 -1.03128088e+00
7.39797115e-01 -9.36857283e-01 -6.79023325e-01 -1.04469255e-01
3.91946495e-01 -4.61475968e-01 5.54559886e-01 7.47336686e-01
-1.70523858e+00 2.37587109e-01 -1.02519476e+00 4.36725795e-01
-1.43148467e-01 -2.58792877e-01 8.39509368e-01 1.30291331e+00
-6.63206279e-01 5.49121976e-01 -5.20020723e-01 2.94418901e-01
5.87782085e-01 7.56860256e-01 -3.51181775e-02 -2.39311606e-01
-7.40914166e-01 4.11603808e-01 6.53664827e-01 3.76383394e-01
-1.14013362e+00 -6.18806362e-01 -3.64917696e-01 2.15120837e-01
8.56395304e-01 -4.06630397e-01 1.34136379e+00 -1.05000818e+00
-1.41466594e+00 4.43646073e-01 1.16380803e-01 -8.40082884e-01
1.09431708e+00 -2.04004511e-01 3.20498645e-01 9.92185771e-02
-3.82857442e-01 -2.16012195e-01 7.73245454e-01 -1.11204135e+00
-1.00796342e+00 -6.81688845e-01 6.16267622e-01 2.90884018e-01
-4.63207245e-01 -2.33219177e-01 4.87101860e-02 -5.64737380e-01
1.64659306e-01 -8.27446222e-01 -3.56192529e-01 -2.18571022e-01
-1.66548863e-01 3.05030614e-01 2.98971117e-01 -6.55848265e-01
1.18323505e+00 -1.64719772e+00 1.85919374e-01 5.61733007e-01
-5.60256720e-01 -1.14066377e-02 4.20196861e-01 4.52832311e-01
1.75181597e-01 1.94229007e-01 -1.45169303e-01 -1.92046732e-01
2.66666353e-01 2.60690928e-01 -5.45668185e-01 8.85984421e-01
-8.60026419e-01 4.46831197e-01 -5.61642349e-01 -1.47970617e-01
-1.77848339e-01 -5.42579472e-01 -5.58998644e-01 -2.11359002e-02
-7.39054143e-01 1.42855182e-01 -7.98395634e-01 7.85205007e-01
1.00498176e+00 2.05553509e-02 -2.02868041e-02 4.74838376e-01
-2.44553372e-01 -2.47924581e-01 -1.69611537e+00 1.51826262e+00
-5.11139035e-01 -1.11620210e-01 7.98310041e-01 -1.46429682e+00
5.37012517e-01 2.36367241e-01 4.52884138e-01 -4.39534426e-01
2.90530175e-01 4.51167405e-01 -5.10643780e-01 -2.63170183e-01
3.32344770e-01 -4.38732862e-01 -6.15871474e-02 5.96325338e-01
-1.03495516e-01 -7.79740736e-02 2.37146899e-01 -1.65799350e-01
9.42531228e-01 -1.17526434e-01 1.26290351e-01 -2.04650521e-01
4.57500726e-01 1.78143024e-01 7.41959274e-01 1.15802372e+00
8.48572254e-02 4.42792140e-02 5.14458358e-01 -5.90568006e-01
-9.75594640e-01 -6.94220364e-01 -2.51901764e-02 1.55212069e+00
1.37740076e-01 4.43047494e-01 -6.36939108e-01 -6.89547837e-01
4.85594451e-01 1.04537296e+00 -7.44131267e-01 4.07799810e-01
-1.92054853e-01 -8.48584950e-01 1.90961108e-01 1.67154118e-01
3.28428537e-01 -3.91652524e-01 -2.86971271e-01 2.92438120e-01
3.99236828e-02 -7.86447704e-01 -6.38245940e-01 5.21552980e-01
-9.82442915e-01 -6.49028838e-01 -8.21565032e-01 -2.19699740e-01
8.29785943e-01 1.03225611e-01 4.90045995e-01 -3.01889628e-01
-1.02658138e-01 6.79587483e-01 -3.66860211e-01 -8.55269730e-01
1.23196557e-01 -1.35421932e-01 -2.03457158e-02 2.40725547e-01
-2.15654507e-01 -6.63796902e-01 -7.92394578e-01 2.61558115e-01
-1.04170561e+00 -3.34907442e-01 3.62158656e-01 8.69928837e-01
1.02564263e+00 1.07507326e-03 5.18208206e-01 -1.09557331e+00
4.04742658e-01 -2.51655966e-01 -1.50961137e+00 4.94529992e-01
-6.69694066e-01 3.27821076e-02 8.81816208e-01 -3.94531310e-01
-9.12124455e-01 6.69874847e-02 3.23684573e-01 -4.29493994e-01
4.08602268e-01 4.11148489e-01 -8.33197311e-03 -5.11219084e-01
4.72793698e-01 6.07908487e-01 -4.25379395e-01 -4.80636209e-01
1.94922760e-01 5.12004614e-01 2.49685481e-01 -1.10185766e+00
8.27293158e-01 5.27792990e-01 4.42634940e-01 -2.20882148e-01
-1.36486816e+00 -2.90594369e-01 1.06178939e-01 -1.01211071e-01
1.36472076e-01 -7.31829822e-01 -1.30075049e+00 -7.83735290e-02
-8.92222047e-01 -1.93407815e-02 -4.83452052e-01 5.11264682e-01
-1.10003483e+00 2.01419100e-01 -2.53107190e-01 -1.60990858e+00
-6.14929378e-01 -7.27928758e-01 2.63784200e-01 3.86168569e-01
7.40750611e-01 -6.98450685e-01 -1.48907468e-01 1.06261456e+00
2.33423129e-01 3.74730200e-01 5.90855837e-01 -7.82948554e-01
-1.03399241e+00 -5.74318588e-01 -1.05757755e-03 5.85180938e-01
-2.48097986e-01 -7.49618530e-01 -8.09586644e-01 -8.03344011e-01
3.01657647e-01 -7.27097869e-01 6.72852337e-01 5.80126822e-01
1.68276179e+00 -1.31850505e+00 3.62461666e-03 9.17587399e-01
1.74916685e+00 3.36497724e-01 2.38540486e-01 4.67246383e-01
3.90040167e-02 1.19817145e-01 9.86001372e-01 1.08541155e+00
-7.46587664e-02 2.66279101e-01 9.66621757e-01 4.87429798e-01
8.97966206e-01 1.27913747e-02 1.97952628e-01 1.26372814e-01
-3.46260190e-01 -2.51006693e-01 -3.25449854e-01 7.63798594e-01
-1.98232663e+00 -1.06586993e+00 3.02320212e-01 2.78609705e+00
1.07134354e+00 1.58409536e-01 2.99382806e-01 2.56845299e-02
7.66074896e-01 -6.08455315e-02 -9.50997591e-01 -9.16815698e-01
-1.63926855e-02 3.78602445e-01 1.34342301e+00 6.31689548e-01
-8.01060975e-01 6.41934812e-01 4.56612349e+00 1.26330292e+00
-7.43141234e-01 2.93994784e-01 8.36467505e-01 -1.00895286e+00
-4.51253861e-01 5.87900802e-02 -8.63221288e-01 5.38395166e-01
6.47427380e-01 -5.59164584e-01 1.14304864e+00 1.12013268e+00
1.61654368e-01 -3.43328744e-01 -1.19138479e+00 8.65127802e-01
-3.00248832e-01 -1.49124134e+00 -3.58479261e-01 1.46228492e-01
1.10260057e+00 -3.84105265e-01 3.83320570e-01 2.25310862e-01
5.06027222e-01 -9.69532251e-01 8.20343554e-01 2.87324578e-01
7.79896677e-01 -1.12224662e+00 6.09148741e-01 8.16357076e-01
-7.98630416e-01 -7.17133880e-01 -4.48318034e-01 7.72255240e-04
-4.33697877e-03 3.89120162e-01 -7.81621993e-01 9.51445222e-01
7.69081652e-01 -5.14351666e-01 4.18195784e-01 1.37561619e+00
1.52076751e-01 4.72372174e-01 -9.17832077e-01 -4.26690727e-01
5.50332606e-01 -4.52529311e-01 6.05852723e-01 1.03963959e+00
4.41158414e-01 5.18666863e-01 2.47901857e-01 5.41579306e-01
-1.73437625e-01 4.48855489e-01 -1.41979098e-01 -1.50186971e-01
4.28111106e-01 9.63816166e-01 -4.07791525e-01 1.02985941e-01
-1.35249227e-01 7.23117828e-01 3.31002921e-01 4.40431982e-01
-9.99221146e-01 -3.42863292e-01 3.30228060e-01 -6.75514415e-02
8.89735639e-01 -6.74290881e-02 -4.80114490e-01 -6.75626576e-01
4.56721187e-01 -5.40446103e-01 8.69842768e-01 -1.61280006e-01
-1.06020319e+00 -2.47464404e-02 1.88795239e-01 -9.58327472e-01
-1.34421691e-01 -4.92064476e-01 -1.53443858e-01 8.51435304e-01
-1.36373270e+00 -9.22666073e-01 1.50399476e-01 8.01143885e-01
1.84735402e-01 -2.49398708e-01 4.91161019e-01 1.53843060e-01
-4.28085893e-01 1.12426174e+00 1.01951492e+00 -4.36054647e-01
2.28133854e-02 -1.19012201e+00 -7.11746633e-01 7.24302113e-01
-3.01950544e-01 1.77045986e-01 1.07642794e+00 -2.87310958e-01
-1.81987858e+00 -9.09509301e-01 1.15754887e-01 1.78894073e-01
8.31701994e-01 -2.12913930e-01 -4.17792052e-02 8.46527278e-01
-2.37303704e-01 3.22322398e-01 7.37666667e-01 -7.80655593e-02
-1.03288718e-01 -8.85739386e-01 -1.80239284e+00 1.07087955e-01
9.34343517e-01 6.69833869e-02 -5.13719302e-03 7.63939321e-01
7.70278931e-01 -9.26283956e-01 -9.60420430e-01 6.73966557e-02
5.33529162e-01 -7.70487964e-01 6.90544546e-01 -8.90574217e-01
-6.75483346e-02 2.19041765e-01 -5.72245955e-01 -8.66609931e-01
1.48921773e-01 -1.28738034e+00 -4.04127181e-01 8.97598803e-01
6.60173178e-01 -7.49960423e-01 1.20722353e+00 9.98212457e-01
2.18731966e-02 -9.83345807e-01 -1.77539456e+00 -8.74566317e-01
3.37370992e-01 -4.17124718e-01 4.98558402e-01 5.96564412e-01
-1.26451179e-01 -6.82044208e-01 -7.60224700e-01 5.23906171e-01
7.80037582e-01 5.84283769e-01 7.13131845e-01 -5.72890043e-01
-9.99025702e-01 -4.66351584e-02 1.76711768e-01 -1.13351035e+00
-2.64900923e-01 -6.75237417e-01 -2.00744018e-01 -1.12593365e+00
1.59400046e-01 -6.42411113e-01 -6.12792075e-01 4.99467283e-01
3.61017615e-01 -3.70025039e-01 3.87421846e-01 -3.28867614e-01
-7.63289630e-01 5.74642122e-01 1.30399108e+00 -3.34181696e-01
-1.63001731e-01 8.11878324e-01 -8.33962679e-01 3.67897689e-01
5.55176973e-01 -6.57533765e-01 -4.38755959e-01 -5.05024731e-01
5.05899847e-01 8.71235788e-01 7.89454766e-03 -6.55195951e-01
2.99435318e-01 -8.76852751e-01 4.96277511e-02 -5.91421902e-01
4.39035714e-01 -1.35155308e+00 1.65235564e-01 4.42701310e-01
-3.84959549e-01 -3.22373807e-01 1.92760974e-01 9.34308112e-01
2.47871116e-01 -9.65111554e-01 7.54019380e-01 -3.30591530e-01
3.44896585e-01 6.04318678e-01 1.21103153e-01 2.94090938e-02
1.47207081e+00 -1.78472295e-01 -3.62703264e-01 -8.18287969e-01
-8.60962808e-01 5.68102777e-01 1.91242453e-02 -5.09671450e-01
1.85271233e-01 -9.92603481e-01 -5.25310218e-01 -4.72744286e-01
-4.20358479e-01 2.35250279e-01 4.65367526e-01 6.20669484e-01
-5.07039011e-01 6.99937344e-01 4.27796766e-02 -7.43431523e-02
-1.01953995e+00 7.43280470e-01 4.57058817e-01 -5.52300394e-01
1.58789277e-01 1.24353027e+00 -4.32889491e-01 -1.33127382e-03
7.26553857e-01 -1.77075192e-01 6.11298501e-01 2.18654349e-02
3.18821073e-01 8.36748123e-01 1.02422729e-01 1.80098370e-01
1.28571063e-01 -4.83205169e-02 -4.95480508e-01 -4.37768817e-01
1.71984732e+00 -2.52257228e-01 -1.62975207e-01 2.66743422e-01
8.50667238e-01 -1.58610772e-02 -1.22815084e+00 -3.68138343e-01
-3.49145830e-01 -8.11546564e-01 6.21126927e-02 -1.00541615e+00
-1.08155072e+00 4.13303614e-01 7.11489201e-01 4.73490030e-01
1.24870479e+00 -4.14558977e-01 7.46830404e-01 8.18869829e-01
9.59837139e-01 -1.60487449e+00 -5.18045902e-01 1.57310486e-01
9.73054290e-01 -9.63693976e-01 4.53865469e-01 -2.94395704e-02
-3.62726957e-01 1.16924763e+00 3.20760399e-01 -1.77662060e-01
3.92774552e-01 -2.53937929e-03 -5.75134158e-01 3.53156835e-01
-6.56193376e-01 6.19920380e-02 -9.08428803e-03 3.90409604e-02
-1.52627781e-01 2.34545603e-01 -9.39780414e-01 1.04710853e+00
-4.20238048e-01 7.10648596e-02 4.87801313e-01 1.04648006e+00
-6.32743418e-01 -9.45465028e-01 -7.82636762e-01 5.41048646e-01
-1.09875166e+00 2.22210094e-01 -1.90294366e-02 6.75365210e-01
-5.04983682e-03 8.46688509e-01 -3.20193708e-01 2.19849110e-01
1.21118233e-01 -4.81274538e-02 7.72016346e-01 -6.13058284e-02
-6.14081442e-01 1.20513991e-01 2.13474095e-01 -4.15539414e-01
-1.12950370e-01 -7.63517618e-01 -9.52089489e-01 -4.22351450e-01
-3.55374694e-01 5.50086439e-01 8.30676913e-01 9.23011541e-01
-1.64201364e-01 -1.12899914e-01 1.17977619e+00 -5.10935962e-01
-1.39478695e+00 -7.27391481e-01 -8.48285079e-01 -1.81056429e-02
4.48321849e-01 -2.67294854e-01 -3.51391912e-01 -4.35185641e-01] | [4.576333522796631, 3.3663978576660156] |
cebd6b77-4912-4ce8-90ca-e4823ee6c28c | exploring-stroke-level-modifications-for | 2212.01982 | null | https://arxiv.org/abs/2212.01982v1 | https://arxiv.org/pdf/2212.01982v1.pdf | Exploring Stroke-Level Modifications for Scene Text Editing | Scene text editing (STE) aims to replace text with the desired one while preserving background and styles of the original text. However, due to the complicated background textures and various text styles, existing methods fall short in generating clear and legible edited text images. In this study, we attribute the poor editing performance to two problems: 1) Implicit decoupling structure. Previous methods of editing the whole image have to learn different translation rules of background and text regions simultaneously. 2) Domain gap. Due to the lack of edited real scene text images, the network can only be well trained on synthetic pairs and performs poorly on real-world images. To handle the above problems, we propose a novel network by MOdifying Scene Text image at strokE Level (MOSTEL). Firstly, we generate stroke guidance maps to explicitly indicate regions to be edited. Different from the implicit one by directly modifying all the pixels at image level, such explicit instructions filter out the distractions from background and guide the network to focus on editing rules of text regions. Secondly, we propose a Semi-supervised Hybrid Learning to train the network with both labeled synthetic images and unpaired real scene text images. Thus, the STE model is adapted to real-world datasets distributions. Moreover, two new datasets (Tamper-Syn2k and Tamper-Scene) are proposed to fill the blank of public evaluation datasets. Extensive experiments demonstrate that our MOSTEL outperforms previous methods both qualitatively and quantitatively. Datasets and code will be available at https://github.com/qqqyd/MOSTEL. | ['Yongdong Zhang', 'Yuxin Wang', 'Jianjun Xu', 'Hongtao Xie', 'Qingfeng Tan', 'Yadong Qu'] | 2022-12-05 | null | null | null | null | ['scene-text-editing'] | ['computer-vision'] | [ 6.96251035e-01 -2.36829564e-01 1.72576830e-01 -3.53440613e-01
-3.17512095e-01 -5.13570309e-01 5.77457249e-01 -5.53461730e-01
-4.44682539e-01 7.60610402e-01 5.73308654e-02 -1.86753497e-01
4.02934045e-01 -6.92586303e-01 -8.56953084e-01 -6.32187307e-01
7.97330797e-01 2.07390711e-01 4.16843504e-01 -2.46619523e-01
3.73898208e-01 2.79287517e-01 -1.19884741e+00 5.26872218e-01
1.33327627e+00 5.05783737e-01 6.16174161e-01 5.60775280e-01
-4.29374814e-01 1.00622821e+00 -7.39888072e-01 -4.40507114e-01
3.79284054e-01 -7.44270504e-01 -4.82131720e-01 4.32810009e-01
4.74182039e-01 -7.61636317e-01 -6.25274777e-01 1.31593430e+00
5.89727461e-01 8.24160948e-02 5.68015814e-01 -1.31550598e+00
-9.66980755e-01 6.21744812e-01 -8.39471400e-01 -2.20316634e-01
1.03406399e-01 2.49614462e-01 3.31251383e-01 -1.17046928e+00
7.50142515e-01 1.04924488e+00 4.84524429e-01 6.67151034e-01
-1.11022758e+00 -7.68872321e-01 1.25618488e-01 1.21214114e-01
-1.35066223e+00 -5.06913662e-01 9.91837680e-01 -2.49246657e-01
3.74970436e-01 3.70476305e-01 2.55869210e-01 1.57165313e+00
1.02202706e-01 9.60231423e-01 1.33177936e+00 -6.26578450e-01
-3.28243934e-02 4.52470183e-01 -3.24439138e-01 5.83923221e-01
9.57333148e-02 1.13201044e-01 -4.52450842e-01 3.20504725e-01
8.69715869e-01 -2.09865365e-02 -5.35488844e-01 -4.97772276e-01
-1.54942667e+00 4.77022111e-01 4.00291057e-03 2.14345276e-01
8.57132003e-02 3.01014502e-02 3.75789642e-01 4.11951333e-01
3.44886839e-01 3.35941344e-01 -2.81406134e-01 -7.54662529e-02
-9.72484231e-01 5.81315458e-02 5.14079511e-01 1.36414564e+00
8.05496931e-01 2.62405813e-01 -3.93675476e-01 1.06931353e+00
-2.72181034e-02 7.07239330e-01 4.42530632e-01 -8.37372124e-01
8.00741076e-01 3.69953960e-01 2.25722238e-01 -1.13605785e+00
1.92421064e-01 -7.86807537e-02 -1.24447083e+00 2.23329589e-01
4.85773683e-01 -2.71797210e-01 -9.85538065e-01 1.42561221e+00
6.38431460e-02 6.36801720e-02 -2.18338937e-01 7.43368745e-01
6.96622133e-01 7.45588899e-01 -2.90190309e-01 -1.36714941e-02
1.04912102e+00 -1.40364182e+00 -1.10141158e+00 -4.64237154e-01
5.44403613e-01 -1.12015069e+00 1.47945869e+00 3.61660272e-01
-9.28798318e-01 -6.51562333e-01 -9.77624655e-01 -1.10204116e-01
-5.55660903e-01 6.71372890e-01 1.80481151e-02 5.95020294e-01
-9.81271982e-01 3.98461670e-01 -4.67037410e-01 -3.07368547e-01
4.29754883e-01 -7.95186162e-02 -1.84344441e-01 -2.48387054e-01
-1.33516693e+00 7.14555860e-01 5.56981385e-01 3.02497119e-01
-7.26776302e-01 -5.01230896e-01 -7.35231876e-01 -1.17244072e-01
5.37338436e-01 -4.27074879e-01 1.12309051e+00 -1.53931904e+00
-1.75202060e+00 8.04378867e-01 -5.02100885e-02 6.04284462e-03
1.29129815e+00 -2.72886097e-01 -4.93730843e-01 7.77288377e-02
2.23148912e-01 8.21886063e-01 1.24175990e+00 -1.57812619e+00
-6.35590434e-01 8.16520378e-02 -2.01945752e-01 2.61282742e-01
-4.14123148e-01 4.77731414e-02 -9.06993926e-01 -1.31886077e+00
-2.69039124e-01 -7.24282205e-01 4.92421258e-03 3.07149053e-01
-6.59131706e-01 4.81035531e-01 1.27729666e+00 -8.90827775e-01
1.27659762e+00 -2.30033588e+00 -7.49071985e-02 -1.48585066e-02
2.32025370e-01 4.85486418e-01 -4.07719135e-01 4.30736005e-01
-1.11293979e-01 1.24982536e-01 -4.48365033e-01 -3.94748926e-01
-7.69224465e-02 1.69952318e-01 -5.72043061e-01 2.21612900e-01
4.68057245e-02 1.01949346e+00 -8.17173898e-01 -6.68577731e-01
4.87097979e-01 3.68386805e-01 -3.05107325e-01 1.45562753e-01
-2.39186779e-01 4.82507080e-01 -3.44131082e-01 4.92316186e-01
1.08336163e+00 -5.93025759e-02 -7.43272435e-03 -3.75232875e-01
-5.51639758e-02 -2.40905657e-01 -1.24498129e+00 1.63306081e+00
-2.73452908e-01 9.73703384e-01 5.34888804e-02 -8.44916880e-01
8.01200449e-01 -2.11504716e-02 1.40233755e-01 -1.00097835e+00
1.51647016e-01 1.89380899e-01 -3.68200481e-01 -5.59315205e-01
7.37502337e-01 3.21271211e-01 1.51927501e-01 5.47406614e-01
-4.25908476e-01 -2.61317253e-01 2.51061201e-01 3.19034904e-01
6.72823370e-01 3.78975362e-01 -7.93419331e-02 -8.64221379e-02
4.91408557e-01 -2.67589930e-02 4.98175472e-01 1.02229345e+00
-2.59260476e-01 9.01025951e-01 3.91148090e-01 -3.76940489e-01
-1.27615070e+00 -9.16891575e-01 1.14829034e-01 9.88199711e-01
6.44286513e-01 -2.15515628e-01 -1.00792468e+00 -7.43379176e-01
-4.27197427e-01 9.12720501e-01 -6.73287094e-01 -5.23568243e-02
-6.67510092e-01 -5.03673315e-01 8.39090466e-01 3.33225191e-01
1.14237905e+00 -1.26034594e+00 -3.75117123e-01 2.11185291e-02
-5.48337102e-01 -1.36765254e+00 -1.01753283e+00 -9.22791213e-02
-4.83464360e-01 -9.04397190e-01 -9.06621754e-01 -9.62314129e-01
9.54861104e-01 6.40077412e-01 8.30722392e-01 1.89886421e-01
-1.45445779e-01 9.22846720e-02 -5.19697070e-01 -3.19732875e-01
-7.76658654e-01 -9.22658518e-02 -2.44793192e-01 2.10064873e-01
-2.13176059e-03 -1.97951317e-01 -4.62912023e-01 7.28698611e-01
-1.38896418e+00 8.81701767e-01 6.94110990e-01 1.01938701e+00
4.93800551e-01 3.00835371e-01 1.26631364e-01 -1.11840618e+00
5.76413453e-01 5.25319390e-02 -5.64731658e-01 5.39752007e-01
-4.93171424e-01 -1.64627925e-01 8.17105711e-01 -5.92005253e-01
-1.51374245e+00 1.20994612e-03 2.03918934e-01 -4.38504159e-01
-3.44646335e-01 1.69873200e-02 -3.69482368e-01 -7.93860108e-02
5.48194110e-01 6.60449505e-01 -1.21658057e-01 -2.68271536e-01
3.16310853e-01 7.72302687e-01 6.74139440e-01 -6.55208409e-01
1.13442349e+00 6.01469100e-01 -5.83399951e-01 -7.58653641e-01
-5.93689442e-01 8.56282189e-02 -6.65656924e-01 -2.15499490e-01
7.40837634e-01 -7.89954901e-01 1.04046218e-01 1.10860085e+00
-1.15598500e+00 -9.13371801e-01 -2.16686040e-01 1.57217130e-01
-5.54284573e-01 8.57055545e-01 -6.06384814e-01 -2.53745437e-01
-1.51521876e-01 -1.22953534e+00 1.12575579e+00 1.70534447e-01
1.94060788e-01 -8.85904610e-01 -2.02715889e-01 2.89319575e-01
5.76110959e-01 1.78655952e-01 7.99612045e-01 -1.60087973e-01
-5.53208053e-01 -1.04963817e-02 -6.55553699e-01 4.72591668e-01
4.52206999e-01 2.83188164e-01 -8.54277909e-01 -2.41365552e-01
-9.54519287e-02 -1.84542000e-01 7.97065616e-01 1.61583483e-01
1.56763983e+00 -2.92668045e-01 -2.26901338e-01 8.43707860e-01
1.28528380e+00 3.01725417e-01 1.01655757e+00 5.40877879e-01
9.29186761e-01 5.11632144e-01 7.47936726e-01 3.88148725e-01
1.03073999e-01 6.62062943e-01 1.59461066e-01 -5.10966063e-01
-4.44894016e-01 -5.44452906e-01 5.56993723e-01 6.95887923e-01
3.43050987e-01 -6.84291363e-01 -7.34609663e-01 4.00799423e-01
-1.85350096e+00 -9.76246774e-01 -1.99496999e-01 2.03422260e+00
9.46461260e-01 1.08817980e-01 -3.49286735e-01 -6.53007179e-02
1.13426054e+00 3.32603872e-01 -5.46636105e-01 -1.62181109e-01
-6.13463938e-01 -2.40176991e-01 5.66233873e-01 3.68043542e-01
-1.07271147e+00 1.27729976e+00 5.34232378e+00 1.32815313e+00
-1.23904598e+00 -7.03387382e-03 7.72278488e-01 6.91976473e-02
-2.62594253e-01 -2.31483560e-02 -5.71373165e-01 8.31213236e-01
2.48695254e-01 -2.31017708e-03 7.73448527e-01 3.80440414e-01
4.99225259e-01 -5.19483089e-02 -8.99273276e-01 1.13870430e+00
1.84618056e-01 -1.32435298e+00 4.17757183e-01 -1.76665947e-01
1.12187660e+00 -1.61499187e-01 1.18759170e-01 1.86049685e-01
2.74366319e-01 -8.92053008e-01 9.73382652e-01 4.34822887e-01
1.26812494e+00 -3.88404489e-01 5.04205644e-01 4.68337953e-01
-9.85416770e-01 1.77251279e-01 -4.00524944e-01 3.00650418e-01
6.15777224e-02 5.27949333e-01 -4.91135269e-01 5.72540522e-01
5.99071324e-01 8.42467308e-01 -8.66904557e-01 6.83596313e-01
-3.72720391e-01 4.91238177e-01 -1.07503407e-01 1.89566433e-01
1.74872115e-01 -3.84605736e-01 3.58517081e-01 1.39834738e+00
4.17823583e-01 -1.95861667e-01 -6.00786135e-02 1.18631184e+00
-2.18694180e-01 1.00728095e-01 -7.12955892e-01 -1.41603768e-01
5.11149466e-01 1.13023174e+00 -8.53397846e-01 -4.97911096e-01
-4.11984771e-01 1.60836029e+00 -1.32604339e-03 7.91848481e-01
-1.22422290e+00 -8.70973825e-01 1.49958014e-01 -3.73388678e-02
3.14132452e-01 -1.09253734e-01 -4.36343402e-01 -1.49249721e+00
1.16120018e-01 -1.30932248e+00 -1.10850133e-01 -1.22429466e+00
-1.09309602e+00 4.67461586e-01 -1.79100782e-01 -1.42374897e+00
3.23635697e-01 -5.42818487e-01 -7.03584373e-01 7.96770394e-01
-1.50272202e+00 -1.23188150e+00 -8.24562848e-01 6.92022920e-01
9.79252517e-01 -1.86157286e-01 3.18569630e-01 4.60816711e-01
-7.74391055e-01 8.08317721e-01 5.45908153e-01 4.92943853e-01
1.33847320e+00 -1.00631881e+00 5.74623585e-01 1.12041402e+00
-3.77775244e-02 2.98449278e-01 6.04145288e-01 -7.60981619e-01
-1.04222584e+00 -1.35985899e+00 5.35346985e-01 -4.56596941e-01
4.31018949e-01 -6.08214676e-01 -9.62937236e-01 6.57145977e-01
5.32077193e-01 -1.84113979e-01 9.11032483e-02 -8.27867866e-01
-2.03495190e-01 -2.77383020e-03 -9.61053133e-01 1.03526425e+00
1.05320644e+00 -3.77920359e-01 -2.67943740e-01 2.83474803e-01
6.51387453e-01 -6.98241234e-01 -2.89840996e-01 2.31960624e-01
4.16980863e-01 -1.13449383e+00 7.34570682e-01 -1.99853227e-01
8.19256067e-01 -5.69201410e-01 -4.73190583e-02 -1.36889625e+00
-6.30954951e-02 -7.59848237e-01 1.99112371e-01 1.39288568e+00
1.78539544e-01 -6.49972081e-01 5.96087873e-01 2.91393667e-01
-1.10342233e-02 -3.92088532e-01 -3.79171729e-01 -5.87879300e-01
1.22100659e-01 -1.75152123e-01 7.09785223e-01 1.39686120e+00
-5.78543007e-01 1.50541365e-01 -8.60787988e-01 -1.66313667e-02
5.67295969e-01 -1.75892990e-02 1.09034228e+00 -5.63160777e-01
-1.82290897e-01 -5.70395768e-01 1.53109595e-01 -1.37274122e+00
6.91001490e-02 -5.02770364e-01 1.52281538e-01 -1.26013803e+00
2.72346228e-01 -4.30823177e-01 1.91682890e-01 4.26780641e-01
-3.10471088e-01 3.72748584e-01 1.98104277e-01 1.81521356e-01
-4.81574833e-01 7.04786599e-01 1.65882421e+00 -3.65740627e-01
-6.06039725e-02 -3.25823635e-01 -5.76295435e-01 6.82396591e-01
1.00526834e+00 -3.53708684e-01 -5.25440037e-01 -1.02930164e+00
1.81903526e-01 -1.16010576e-01 4.57257003e-01 -8.04530025e-01
1.82997882e-01 -4.51652735e-01 6.20383859e-01 -6.61603928e-01
-6.85192049e-02 -8.47458124e-01 1.30071014e-01 3.03964615e-01
-4.46336240e-01 -9.60982963e-02 1.45785525e-01 5.59051096e-01
-1.69684529e-01 -2.44656622e-01 1.01361084e+00 -5.69010712e-02
-8.59595060e-01 2.57236272e-01 -3.91564727e-01 1.39887318e-01
9.81855273e-01 -6.08259380e-01 -6.44506395e-01 -5.69988489e-01
-2.84252793e-01 2.66003191e-01 8.89468074e-01 5.84564030e-01
6.97262943e-01 -1.30446410e+00 -7.00862527e-01 3.91288549e-01
1.68530881e-01 1.20584786e-01 3.98837537e-01 7.78786004e-01
-9.87707853e-01 8.85513201e-02 -3.61625195e-01 -3.92542690e-01
-1.09262800e+00 5.25150537e-01 4.73282933e-01 -1.62237316e-01
-7.55225301e-01 5.33243954e-01 7.28285253e-01 -6.14685595e-01
2.73210555e-01 -2.38602415e-01 1.78688079e-01 -3.47870111e-01
5.47271550e-01 3.43062222e-01 -1.97846636e-01 -5.28733730e-01
1.15780897e-01 5.51671624e-01 -3.49308193e-01 9.57692135e-03
9.45248127e-01 -5.46437860e-01 -7.98162743e-02 8.62679929e-02
9.85675335e-01 3.23123842e-01 -1.48564315e+00 -4.41402972e-01
-3.45726490e-01 -8.17830384e-01 -2.09181666e-01 -9.91658866e-01
-1.14041388e+00 1.00980937e+00 6.26297414e-01 -1.62491381e-01
1.29238117e+00 -6.36231959e-01 8.69892538e-01 4.98361498e-01
7.53382519e-02 -1.47668612e+00 3.54879320e-01 5.45940340e-01
9.34516907e-01 -1.27420795e+00 -1.14916839e-01 -5.23777783e-01
-8.90198529e-01 1.08572066e+00 1.01694417e+00 1.60698608e-01
1.49893001e-01 3.31976354e-01 4.36237067e-01 2.11773559e-01
-3.62660855e-01 2.26996139e-01 -1.13284819e-01 7.10195601e-01
1.66634977e-01 -2.02561304e-01 -4.43488359e-02 2.02755034e-01
1.68667138e-02 6.63273875e-03 8.49470139e-01 8.59159708e-01
-1.85817301e-01 -1.10946465e+00 -5.36598980e-01 3.75438273e-01
-2.29653507e-01 -1.52613685e-01 -6.81210697e-01 8.37376237e-01
1.65131927e-01 7.86069989e-01 -1.26315847e-01 -2.59608150e-01
2.48580590e-01 -1.88175604e-01 2.60143816e-01 -3.05407315e-01
-2.48128682e-01 1.39867231e-01 -1.92909941e-01 -4.30123568e-01
-2.94670463e-01 -2.88752079e-01 -1.00839162e+00 -4.79623467e-01
-3.70175540e-01 -1.36281371e-01 3.89971942e-01 6.83411598e-01
3.21563959e-01 6.46162033e-01 7.51931787e-01 -8.11104238e-01
-4.61090475e-01 -8.73097658e-01 -6.50976241e-01 6.18207276e-01
1.65935099e-01 -3.98792475e-01 -3.41343462e-01 4.99830097e-01] | [11.516582489013672, -0.20081087946891785] |
76a43e22-e85f-407a-a129-60da27b873b0 | cherrypicker-semantic-skeletonization-and | 2304.04708 | null | https://arxiv.org/abs/2304.04708v1 | https://arxiv.org/pdf/2304.04708v1.pdf | CherryPicker: Semantic Skeletonization and Topological Reconstruction of Cherry Trees | In plant phenotyping, accurate trait extraction from 3D point clouds of trees is still an open problem. For automatic modeling and trait extraction of tree organs such as blossoms and fruits, the semantically segmented point cloud of a tree and the tree skeleton are necessary. Therefore, we present CherryPicker, an automatic pipeline that reconstructs photo-metric point clouds of trees, performs semantic segmentation and extracts their topological structure in form of a skeleton. Our system combines several state-of-the-art algorithms to enable automatic processing for further usage in 3D-plant phenotyping applications. Within this pipeline, we present a method to automatically estimate the scale factor of a monocular reconstruction to overcome scale ambiguity and obtain metrically correct point clouds. Furthermore, we propose a semantic skeletonization algorithm build up on Laplacian-based contraction. We also show by weighting different tree organs semantically, our approach can effectively remove artifacts induced by occlusion and structural size variations. CherryPicker obtains high-quality topology reconstructions of cherry trees with precise details. | ['Marc Stamminger', 'Oliver Scholz', 'Andreas Gilson', 'Lukas Meyer'] | 2023-04-10 | null | null | null | null | ['plant-phenotyping'] | ['computer-vision'] | [ 4.61207122e-01 2.11549178e-02 1.22153483e-01 -2.32226342e-01
-4.17427540e-01 -1.04518509e+00 -1.21019976e-02 5.65816164e-01
3.40816617e-01 1.75309896e-01 -5.95849633e-01 -3.68474036e-01
-2.98569351e-01 -1.00306177e+00 -5.12586534e-01 -3.08614612e-01
-9.04284976e-03 9.25805509e-01 6.53102636e-01 9.33894962e-02
2.86472380e-01 1.30859971e+00 -1.43076873e+00 -1.00030772e-01
9.65774834e-01 9.08037901e-01 6.51004553e-01 7.72085905e-01
-5.94783545e-01 -2.30341107e-01 -3.01942199e-01 -3.37857991e-01
3.55388820e-01 -9.40447673e-02 -7.11312294e-01 5.15735745e-01
5.11116624e-01 -1.88398704e-01 4.38795179e-01 8.79825294e-01
1.68387517e-01 -7.16701627e-01 5.82459390e-01 -1.10229075e+00
-2.52764821e-01 5.94376028e-01 -9.22219634e-01 -6.64604843e-01
4.26603779e-02 -2.14064363e-02 8.77191007e-01 -8.16608310e-01
9.24256802e-01 1.02392316e+00 8.91070247e-01 1.00717200e-02
-1.89119613e+00 -2.06255555e-01 7.56943375e-02 8.64790380e-02
-1.35140467e+00 -9.71641392e-02 9.36238050e-01 -4.40797508e-01
3.10984224e-01 6.06190622e-01 9.81187701e-01 3.14924598e-01
-6.20776229e-02 6.16855502e-01 6.65603757e-01 -3.52350473e-01
4.94460791e-01 -3.68745536e-01 -7.02939406e-02 6.94187045e-01
1.84012264e-01 -3.51614356e-01 -2.84326784e-02 -3.34104776e-01
1.21552360e+00 -5.13194203e-02 -2.06060097e-01 -1.24552381e+00
-1.11097407e+00 2.79491782e-01 5.58046639e-01 2.96723217e-01
-5.11819065e-01 1.68639556e-01 1.16918601e-01 -1.75682679e-01
4.36660111e-01 3.50549459e-01 -9.22842145e-01 1.40386328e-01
-1.23109829e+00 7.68613666e-02 7.22159028e-01 1.18513906e+00
8.68717730e-01 -2.94349790e-01 3.09678674e-01 6.58858061e-01
3.22371840e-01 5.25273085e-01 -3.04211795e-01 -1.45369446e+00
-1.33086860e-01 1.13840997e+00 -1.12179659e-01 -9.42796588e-01
-4.55923557e-01 -2.17801660e-01 -6.83502197e-01 4.46729451e-01
4.85653818e-01 6.10424340e-01 -8.88718367e-01 1.18048167e+00
7.97923446e-01 4.90726642e-02 -3.63635898e-01 5.27739584e-01
6.73842192e-01 9.15140361e-02 -1.32045830e-02 -6.05913624e-02
1.56091428e+00 -1.08550027e-01 -2.12448731e-01 1.07570663e-01
5.32572508e-01 -9.62095857e-01 8.89671206e-01 5.30584335e-01
-1.25608158e+00 -2.15720832e-01 -7.95043647e-01 -7.95330927e-02
-2.16576472e-01 6.26974642e-01 6.13847733e-01 5.88484824e-01
-7.72391021e-01 1.10384893e+00 -1.03186500e+00 -5.30772626e-01
8.49786103e-01 2.92555332e-01 -4.99723375e-01 1.60686508e-01
4.66810390e-02 7.28661299e-01 3.71635884e-01 2.17898451e-02
-6.72489643e-01 -9.42486405e-01 -4.29165661e-01 1.26471654e-01
4.66887087e-01 -7.98176408e-01 8.67163539e-01 -6.65973783e-01
-1.82144618e+00 1.44007254e+00 -9.16851014e-02 -2.38149673e-01
4.20031130e-01 2.88362075e-02 2.06446409e-01 3.87555331e-01
-1.96984746e-02 6.26855016e-01 9.24960017e-01 -1.51860034e+00
-2.42564410e-01 -1.15501988e+00 -5.23415983e-01 -1.64739117e-01
8.76943469e-02 -1.64500833e-01 -2.80562490e-01 -3.48036557e-01
1.29831147e+00 -9.18502867e-01 -2.38977164e-01 9.01319385e-01
-4.97091651e-01 3.40420812e-01 1.27551246e+00 -8.86899233e-01
4.36618626e-01 -2.04053116e+00 4.58201170e-01 2.32971221e-01
5.10750771e-01 8.58309716e-02 -2.15637714e-01 1.01394087e-01
-2.10877150e-01 3.21423918e-01 -7.45894492e-01 -2.35815138e-01
-3.45816344e-01 3.73376399e-01 -2.54157800e-02 5.41838169e-01
1.83005318e-01 6.51068032e-01 -6.26762092e-01 -6.44652724e-01
5.53231120e-01 3.79136592e-01 -1.09151110e-01 -4.30338308e-02
-4.39116448e-01 4.10213083e-01 -5.46246946e-01 1.28984869e+00
1.49004412e+00 7.41511807e-02 1.61935449e-01 -2.71312833e-01
-4.48829859e-01 -2.59469181e-01 -1.11850214e+00 1.98551357e+00
-4.75867093e-01 1.15527302e-01 1.02229190e+00 -7.95774400e-01
1.28176177e+00 1.12726621e-01 9.51761425e-01 3.76751989e-01
3.50366123e-02 2.90506482e-01 -4.73054826e-01 1.34587690e-01
2.19131440e-01 1.77337199e-01 5.42251289e-01 -3.89679037e-02
-8.94569457e-02 -1.20819104e+00 -8.55202004e-02 1.74003080e-01
1.04731369e+00 7.12914288e-01 1.89143300e-01 -4.04510289e-01
6.07345283e-01 3.01304668e-01 5.88538229e-01 -4.57525179e-02
-1.99392080e-01 8.26047182e-01 5.96224129e-01 -3.03002030e-01
-1.26764309e+00 -1.32519937e+00 -5.34240425e-01 4.87256944e-01
-5.13176695e-02 -6.98087633e-01 -9.12999809e-01 -6.06300890e-01
4.18258846e-01 7.03713834e-01 -4.59785312e-01 1.83048248e-01
-2.24458769e-01 -5.32479227e-01 1.92844078e-01 2.15708181e-01
3.33909541e-01 -7.66747177e-01 -1.01176333e+00 1.56632125e-01
3.90875489e-02 -1.24237001e+00 2.27877915e-01 1.18399844e-01
-1.48927319e+00 -1.25935149e+00 -5.92465222e-01 -4.57665443e-01
6.71063840e-01 4.70798612e-01 1.22189867e+00 1.41096503e-01
-7.67787814e-01 2.96041131e-01 -5.48819780e-01 -3.51061583e-01
-3.84811074e-01 2.19603673e-01 -5.25256574e-01 -2.46689335e-01
6.65689185e-02 -1.27198946e+00 -3.76488835e-01 4.64358598e-01
-6.71448350e-01 2.27928385e-01 4.01569903e-01 2.86419630e-01
1.24235654e+00 4.62845303e-02 -1.72480926e-01 -7.01373875e-01
-1.35069981e-01 1.26629639e-02 -1.23248279e+00 3.21359724e-01
-1.83950946e-01 -1.40796989e-01 5.02371550e-01 -6.60568625e-02
-6.87579930e-01 7.96831310e-01 5.86043634e-02 -3.58607382e-01
-5.69403052e-01 2.60070078e-02 -5.23558915e-01 -3.20641726e-01
5.10809898e-01 1.12672985e-01 8.35447386e-02 -8.67498755e-01
7.45559990e-01 3.69601548e-01 7.10162699e-01 -5.63999057e-01
1.05991602e+00 9.15805638e-01 6.63520753e-01 -1.34962392e+00
-4.20736909e-01 -4.08956617e-01 -1.56425190e+00 -2.65827239e-01
8.33899140e-01 -4.59852993e-01 -9.74817634e-01 5.28690994e-01
-1.36203074e+00 -1.42437190e-01 -7.46596515e-01 2.37237200e-01
-8.17944348e-01 6.21735573e-01 -2.47426927e-01 -5.68908691e-01
-3.28161508e-01 -9.67283726e-01 1.60110295e+00 1.98768862e-02
5.50647676e-02 -7.04669774e-01 7.22730532e-02 2.87690461e-01
1.33592427e-01 6.52284980e-01 8.36423874e-01 2.24639744e-01
-9.11336243e-01 -2.36110330e-01 -4.22087908e-01 3.59885767e-02
-2.18629520e-02 1.07682586e+00 -9.08939302e-01 1.67285725e-01
-1.56834662e-01 5.97733371e-02 4.71505314e-01 7.17502892e-01
1.44962454e+00 3.98040146e-01 -4.88530159e-01 9.75029349e-01
1.47570872e+00 -3.99157315e-01 6.38194799e-01 6.52481318e-02
8.37545574e-01 9.82057333e-01 6.02600753e-01 4.91479576e-01
1.11872010e-01 6.75157368e-01 1.09164608e+00 3.35966125e-02
-1.05233222e-01 -1.59010202e-01 -1.21764071e-01 4.83125001e-01
-2.63709635e-01 4.34216976e-01 -9.92944419e-01 4.87053752e-01
-1.48721254e+00 -7.16909230e-01 -1.10920441e+00 2.30173755e+00
5.48263252e-01 -1.55385002e-01 -5.72347715e-02 4.05086488e-01
7.18725145e-01 -2.49320298e-01 -8.00556064e-01 -2.79878050e-01
-3.68098855e-01 4.42596495e-01 9.40718353e-01 4.84668761e-01
-9.04233515e-01 1.43901253e+00 6.16312265e+00 5.53179264e-01
-7.69392371e-01 -8.75040144e-03 1.82284549e-01 4.45015162e-01
-3.14443737e-01 7.92072475e-01 -4.90210384e-01 -1.99883059e-01
5.52720785e-01 1.07814118e-01 2.21252665e-01 1.14548123e+00
2.71566182e-01 -3.20625901e-01 -5.84000230e-01 7.82496631e-01
-3.58623713e-01 -9.96545196e-01 -3.00240606e-01 2.88851827e-01
2.82665193e-01 -1.93675011e-01 -4.82632071e-01 -5.57209373e-01
5.15462577e-01 -7.38244772e-01 6.86929226e-01 2.73603588e-01
8.71153057e-01 -4.17239606e-01 1.32344559e-01 4.39293087e-01
-1.57773244e+00 2.77536124e-01 -7.92017341e-01 3.48102748e-01
1.13953128e-01 1.46055138e+00 -9.44142342e-01 9.44320023e-01
5.98943472e-01 8.22443128e-01 -1.14437616e+00 1.00565314e+00
-2.32001737e-01 5.67460656e-01 -8.20331395e-01 5.62206507e-01
-3.28274757e-01 -6.42224729e-01 5.84244967e-01 7.01297700e-01
5.79431653e-01 -6.46958947e-02 9.77022275e-02 1.33993423e+00
1.68698847e-01 3.82961005e-01 -4.02822286e-01 2.33081402e-03
5.12992561e-01 1.74273658e+00 -1.39792717e+00 -7.42877945e-02
3.27453941e-01 1.15754056e+00 1.14661604e-01 -3.66226494e-01
-4.32118058e-01 -3.62550206e-02 3.34924579e-01 2.51766533e-01
4.69017565e-01 -7.01599419e-01 -8.88638437e-01 -1.05133855e+00
1.72128659e-02 -3.07276845e-01 -1.30143091e-01 -1.15934861e+00
-8.16607893e-01 9.62399617e-02 -1.07145630e-01 -1.00602472e+00
4.12292689e-01 -6.91176414e-01 -4.75493848e-01 7.36491203e-01
-1.04566193e+00 -1.62819445e+00 -7.61311412e-01 2.27131873e-01
3.34975332e-01 3.37039262e-01 1.26835525e+00 -1.56550854e-01
-4.58727092e-01 -3.40698034e-01 1.61262617e-01 -5.26508629e-01
3.01719278e-01 -1.38086617e+00 4.99699950e-01 7.76938319e-01
1.33620381e-01 6.61075339e-02 5.03518522e-01 -9.08615112e-01
-1.40420198e+00 -8.71033311e-01 6.97302759e-01 -5.15017509e-01
4.34982806e-01 -2.30537921e-01 -9.53394055e-01 5.94910204e-01
-3.77044618e-01 -5.77177666e-02 3.85804355e-01 -1.18505307e-01
-1.00272924e-01 -8.32651705e-02 -1.44926298e+00 3.08684707e-01
1.26328242e+00 -5.52895740e-02 -1.34104773e-01 4.46350574e-01
2.64776081e-01 -2.19162673e-01 -1.21964836e+00 4.99102533e-01
6.94262207e-01 -1.13604975e+00 1.31311786e+00 6.70240670e-02
1.46144107e-01 -6.59587741e-01 -1.77228332e-01 -1.27886701e+00
-4.85106945e-01 -5.26700914e-01 1.92890495e-01 1.44163573e+00
-1.00880057e-01 -1.72910601e-01 1.04507720e+00 1.75009921e-01
-6.35493621e-02 -6.93670660e-02 -8.30949068e-01 -7.37569094e-01
-4.23099548e-02 -4.04853821e-01 6.56198502e-01 8.66872311e-01
-5.47439575e-01 -5.31941699e-03 3.52206796e-01 3.59313965e-01
1.07287204e+00 4.56412882e-01 9.97209311e-01 -2.14712691e+00
5.90276085e-02 -5.79321861e-01 -6.69700682e-01 -7.66541541e-01
2.68770158e-01 -9.45999563e-01 -2.42199540e-01 -1.45074952e+00
-1.56997547e-01 -6.86335742e-01 8.22895586e-01 3.38014841e-01
1.83384016e-01 -4.43547219e-02 2.30556518e-01 2.22186774e-01
2.78772116e-01 4.33802366e-01 1.10695732e+00 1.89047098e-01
-2.49513075e-01 2.18916848e-01 -2.03409716e-01 1.12495577e+00
9.97683227e-01 -4.88765925e-01 -1.40209466e-01 -3.83377522e-01
-7.39960298e-02 -8.22513029e-02 6.81358576e-01 -8.59353900e-01
-4.00287300e-01 -2.48215854e-01 4.96628910e-01 -1.20646346e+00
6.05206490e-01 -1.32153726e+00 5.94290793e-01 6.71612859e-01
9.91134271e-02 -8.63591880e-02 8.61797854e-02 2.34111622e-01
5.14655471e-01 -6.48146987e-01 1.13504243e+00 -3.41932595e-01
-1.09770015e-01 2.34523803e-01 -1.20969988e-01 -4.51060116e-01
1.30345595e+00 -4.07028586e-01 9.92086679e-02 5.31598143e-02
-8.16228807e-01 1.67378187e-02 1.10448349e+00 -2.38479838e-01
7.67767906e-01 -1.09715331e+00 -8.47788036e-01 1.30572930e-01
-7.02208234e-03 3.21940362e-01 -9.96762812e-02 6.50594294e-01
-1.50725186e+00 1.10774152e-01 -5.09302974e-01 -1.10773468e+00
-1.71202552e+00 3.48688394e-01 1.59117803e-01 2.75304765e-02
-6.88345075e-01 6.12408161e-01 -1.52873188e-01 -7.10280418e-01
-1.34618938e-01 -8.17013085e-01 2.21791714e-02 -6.05694503e-02
-8.39507133e-02 4.87968653e-01 1.58150882e-01 -7.77600408e-01
-5.20480156e-01 1.21510923e+00 5.68417490e-01 -2.09753513e-02
1.60584211e+00 -1.45878773e-02 -4.03797686e-01 4.12710756e-01
5.90134740e-01 1.26786783e-01 -1.24562263e+00 1.43183589e-01
1.93903819e-01 -9.47131157e-01 2.33375147e-01 -5.03503621e-01
-1.10110748e+00 1.05837274e+00 5.51084101e-01 6.08164310e-01
1.17578864e+00 5.79683408e-02 4.80146646e-01 -1.30220175e-01
3.95859063e-01 -9.54624712e-01 -5.76566696e-01 1.00247726e-01
1.04506099e+00 -7.88427532e-01 3.18286061e-01 -1.30480671e+00
-1.51871340e-02 1.44421268e+00 1.75811991e-01 -1.61210448e-01
8.12570751e-01 2.84936935e-01 -1.80217639e-01 -2.05250785e-01
-3.45076710e-01 -5.14309883e-01 -4.64106984e-02 1.06238437e+00
3.96312088e-01 3.30273807e-01 -2.94686884e-01 8.61097947e-02
-4.14076149e-01 -1.85403913e-01 4.53439325e-01 1.01759946e+00
-5.87614834e-01 -1.36736310e+00 -9.69733417e-01 5.55227362e-02
5.39516099e-02 1.85619920e-01 -9.82800901e-01 5.07949114e-01
1.21241286e-01 3.82725596e-01 -9.64603126e-02 -7.81793296e-02
5.49222291e-01 7.75848106e-02 8.36188912e-01 -6.36649132e-01
-3.27301681e-01 3.37413758e-01 -3.60231519e-01 -5.51911652e-01
-2.97756225e-01 -9.21385050e-01 -9.57713246e-01 -2.89189994e-01
-6.15750730e-01 -3.82096678e-01 1.48662651e+00 3.45179290e-01
4.53385025e-01 4.27317679e-01 6.26921356e-01 -9.98031855e-01
-2.95026839e-01 -5.92026114e-01 -8.63877654e-01 3.39656502e-01
-3.28338802e-01 -5.13847589e-01 -1.17095396e-01 3.58544499e-01] | [8.973226547241211, -1.7773329019546509] |
0deba036-9561-4b77-abc8-3f212980aec7 | a-natural-upper-bound-to-the-accuracy-of | 1809.10389 | null | http://arxiv.org/abs/1809.10389v1 | http://arxiv.org/pdf/1809.10389v1.pdf | A natural upper bound to the accuracy of predicting protein stability changes upon mutations | Accurate prediction of protein stability changes upon single-site variations
(DDG) is important for protein design, as well as our understanding of the
mechanism of genetic diseases. The performance of high-throughput computational
methods to this end is evaluated mostly based on the Pearson correlation
coefficient between predicted and observed data, assuming that the upper bound
would be 1 (perfect correlation). However, the performance of these predictors
can be limited by the distribution and noise of the experimental data. Here we
estimate, for the first time, a theoretical upper-bound to the DDG prediction
performances imposed by the intrinsic structure of currently available DDG
data. Given a set of measured DDG protein variations, the theoretically best
predictor is estimated based on its similarity to another set of experimentally
determined DDG values. We investigate the correlation between pairs of measured
DDG variations, where one is used as a predictor for the other. We analytically
derive an upper bound to the Pearson correlation as a function of the noise and
distribution of the DDG data. We also evaluate the available datasets to
highlight the effect of the noise in conjunction with DDG distribution. We
conclude that the upper bound is a function of both uncertainty and spread of
the DDG values, and that with current data the best performance should be
between 0.7-0.8, depending on the dataset used; higher Pearson correlations
might be indicative of overtraining. It also follows that comparisons of
predictors using different datasets are inherently misleading. | [] | 2018-09-27 | null | null | null | null | ['protein-design'] | ['medical'] | [ 2.31610447e-01 -1.20627537e-01 -6.18013255e-02 -4.15078908e-01
-6.80774093e-01 -6.22817218e-01 3.04274052e-01 7.47405171e-01
-4.18207377e-01 1.02160537e+00 -7.58548155e-02 -4.48389679e-01
-2.56545126e-01 -4.70225573e-01 -7.84272730e-01 -9.76025581e-01
-1.65559016e-02 4.97801185e-01 5.25454044e-01 -8.57064575e-02
1.33729115e-01 5.50277710e-01 -1.40238047e+00 -4.21613082e-03
1.02452707e+00 7.95970142e-01 2.65981913e-01 4.84634459e-01
-1.19614847e-01 -3.61723863e-02 -7.94279099e-01 -2.04474226e-01
1.00848684e-02 -6.04375064e-01 -4.27749008e-01 -2.91442662e-01
-7.80322999e-02 1.30468920e-01 3.13640743e-01 7.92782545e-01
5.83619714e-01 -7.74253756e-02 8.25415254e-01 -8.00906122e-01
-3.10892791e-01 1.12509467e-01 -3.72057021e-01 9.52182710e-02
3.66720110e-01 4.01270896e-01 8.66416693e-01 -3.62168700e-01
8.71450186e-01 7.41211951e-01 4.72555310e-01 1.18045442e-01
-1.53312683e+00 -2.16112554e-01 -2.29017884e-01 1.18780978e-01
-1.34256053e+00 -1.91227227e-01 2.82412380e-01 -4.80557382e-01
1.03653359e+00 8.52567255e-02 6.01230323e-01 5.67969024e-01
4.14116174e-01 -7.11168647e-02 9.07320738e-01 -5.86205721e-01
4.87937808e-01 2.49368902e-02 1.38533726e-01 3.38726014e-01
5.78160346e-01 2.58945078e-01 -3.84110481e-01 -4.70989764e-01
2.75042027e-01 -4.87332642e-01 -4.73242432e-01 -6.29190743e-01
-8.11109602e-01 6.46050990e-01 4.40390445e-02 2.79827893e-01
-2.84779549e-01 -2.95207351e-01 3.44221503e-01 2.13833734e-01
6.66263923e-02 6.20225728e-01 -9.49744821e-01 -3.00048113e-01
-6.57262087e-01 2.54997849e-01 7.84682155e-01 3.92248511e-01
6.56817317e-01 -3.61319244e-01 1.71109736e-01 8.20345223e-01
1.86185792e-01 4.39861298e-01 4.31219786e-01 -4.54070359e-01
-6.43548695e-03 5.05234241e-01 5.38046837e-01 -6.48336589e-01
-6.62731171e-01 -2.12344661e-01 -2.70392865e-01 1.52506232e-01
1.19640219e+00 -1.93372741e-01 -6.35488987e-01 1.89751506e+00
3.24174792e-01 -4.82332677e-01 3.06445081e-02 7.67279863e-01
3.36062044e-01 3.71266186e-01 2.09856465e-01 -6.52588665e-01
1.06141138e+00 -1.19675528e-02 -3.32962185e-01 5.01432978e-02
7.31775045e-01 -7.86607802e-01 8.62483621e-01 4.18645501e-01
-5.12834489e-01 -2.43969813e-01 -1.14089262e+00 2.60032356e-01
-2.22794250e-01 1.64107889e-01 2.04770401e-01 6.70484126e-01
-5.63064158e-01 9.62594390e-01 -8.79048645e-01 -6.52120769e-01
-2.53952384e-01 5.43035805e-01 -3.17125708e-01 2.95194060e-01
-1.07217395e+00 1.19463003e+00 5.59644580e-01 -3.22425552e-02
-2.10382417e-02 -7.40586340e-01 -2.93000668e-01 -3.39668803e-02
-2.24978328e-02 -1.82072550e-01 7.51062572e-01 -5.62484443e-01
-1.34612644e+00 7.42760360e-01 -2.12026704e-02 -4.72652912e-01
3.80076319e-01 6.98536336e-02 -4.41666573e-01 -1.44967124e-01
-2.60107666e-01 9.93387699e-02 1.43231705e-01 -8.38268101e-01
-3.00161988e-02 -5.11107862e-01 -3.39651316e-01 -2.23567456e-01
5.29951990e-01 -1.65726338e-02 9.67447981e-02 -8.53777826e-02
2.84821540e-01 -1.03152454e+00 -1.58575684e-01 -8.35678279e-02
-9.63600427e-02 6.89562485e-02 3.80771458e-01 -6.22335374e-01
1.10931480e+00 -2.04403663e+00 -1.36218429e-01 4.69071686e-01
-7.89023414e-02 4.11948681e-01 -7.19530582e-02 7.34880447e-01
-3.24066699e-01 -1.11300677e-01 -2.02012599e-01 6.29085481e-01
-2.45637864e-01 -2.27695536e-02 5.21681719e-02 8.75787616e-01
3.39297444e-01 5.48190355e-01 -8.02946746e-01 2.49250270e-02
7.68218562e-02 4.81020242e-01 -2.07981676e-01 1.87347904e-01
-3.70490223e-01 3.64721864e-01 -2.56409168e-01 1.73330307e-01
7.12676346e-01 -2.03516155e-01 8.45106125e-01 -3.30681562e-01
-2.47376814e-01 2.91961879e-01 -8.84023011e-01 9.92625475e-01
4.90089767e-02 3.39635044e-01 -2.85104126e-01 -7.71695673e-01
1.41814566e+00 5.36681227e-02 5.38023770e-01 -3.20704788e-01
1.08744055e-01 4.79046077e-01 8.54070663e-01 -2.02980191e-01
1.26099974e-01 -4.14960206e-01 2.72898495e-01 2.52124786e-01
-6.19335882e-02 2.30214715e-01 1.91282704e-01 -1.33703291e-01
1.02971542e+00 3.58991295e-01 5.68210304e-01 -6.54844284e-01
4.11274672e-01 -1.52042747e-01 6.03392363e-01 3.25095206e-01
-3.05156916e-01 5.81577122e-01 1.28365803e+00 -2.57589072e-01
-1.36333537e+00 -9.50052083e-01 -5.85630059e-01 5.37144244e-01
8.72884411e-03 -5.27089417e-01 -6.67845130e-01 -3.20352197e-01
2.95081586e-01 5.74083447e-01 -3.46503764e-01 -3.92262429e-01
-1.34910136e-01 -1.45049536e+00 3.80345464e-01 1.42884001e-01
-8.74427110e-02 -4.11450237e-01 -7.11820245e-01 3.24166566e-01
1.72405243e-01 -1.01719654e+00 -7.36240949e-03 5.99472106e-01
-6.06265306e-01 -1.46270132e+00 -3.22328180e-01 1.02946296e-01
3.61995190e-01 -1.28113315e-01 1.05101430e+00 1.85292393e-01
-2.24794194e-01 -3.93631719e-02 -3.03384691e-01 -2.89584279e-01
-7.46516168e-01 -1.18015416e-01 1.77320510e-01 -3.89223963e-01
6.99886858e-01 -5.81693590e-01 -6.28931463e-01 4.54040736e-01
-6.55050457e-01 -4.77674007e-01 5.57653666e-01 9.36585486e-01
5.96762240e-01 -1.07985146e-01 5.95873535e-01 -7.82967508e-01
3.59985262e-01 -3.43411088e-01 -9.08024311e-01 4.49911594e-01
-9.06940758e-01 5.95313907e-01 5.20502388e-01 -3.53062242e-01
-8.84260237e-01 3.04582536e-01 5.71066467e-03 -4.08339910e-02
-2.99614906e-01 5.35671651e-01 -3.56365323e-01 -6.80644661e-02
7.83982515e-01 1.48173356e-02 3.74613971e-01 -5.19441485e-01
-2.88554411e-02 5.40512145e-01 2.96970725e-01 -7.87374020e-01
9.02043656e-02 -1.61702856e-02 4.01484460e-01 -9.58063841e-01
-4.72550631e-01 -2.99550146e-01 -6.48461223e-01 6.21932698e-03
4.67204541e-01 -4.35491353e-01 -9.31629062e-01 3.29642326e-01
-7.82760799e-01 -2.59887904e-01 2.08854988e-01 5.64117432e-01
-6.55109584e-01 7.67051697e-01 -2.35356152e-01 -8.03558946e-01
-1.54178008e-01 -1.28566980e+00 8.30756247e-01 -1.91225171e-01
-4.11273003e-01 -7.99291611e-01 1.08948350e-01 1.67217210e-01
2.84151006e-02 5.41203916e-01 1.30641818e+00 -8.03578854e-01
-1.23157673e-01 -3.99711639e-01 -4.26128954e-02 1.53542504e-01
2.56385833e-01 4.11965966e-01 -5.85341334e-01 -2.23076522e-01
-2.14305520e-01 -2.64359117e-02 6.93913281e-01 5.73118627e-01
6.08623207e-01 2.20320776e-01 -3.08078915e-01 8.04618821e-02
1.60509741e+00 2.22382888e-01 6.66427433e-01 3.73710543e-01
1.33049607e-01 5.70697486e-01 7.91334331e-01 4.95735049e-01
-3.65061402e-01 1.05826914e+00 2.29134679e-01 2.64933169e-01
2.13581592e-01 -1.37626961e-01 2.28882223e-01 1.61716297e-01
-8.10381174e-02 -2.12366879e-01 -9.65172172e-01 6.37604222e-02
-1.78232837e+00 -7.42032647e-01 -5.73278666e-01 2.86665702e+00
7.84223676e-01 3.59964281e-01 5.09615123e-01 2.54469104e-02
6.99730873e-01 -1.87585577e-01 -6.27975643e-01 -6.17024899e-01
-3.41434956e-01 6.39117733e-02 7.46433556e-01 4.76083815e-01
-5.88547528e-01 3.46083462e-01 7.17185831e+00 5.31572759e-01
-1.11485362e+00 -5.00062644e-01 5.22610068e-01 -9.52010229e-02
-3.63492556e-02 1.51873022e-01 -8.23365629e-01 8.55345845e-01
1.39132285e+00 -4.46236104e-01 4.59720753e-02 4.76699829e-01
3.92444760e-01 -5.47756493e-01 -9.49287891e-01 5.32489538e-01
-4.51917827e-01 -8.53532791e-01 -3.78869653e-01 3.37778449e-01
4.80953902e-01 -6.29357761e-03 -3.03709298e-01 -2.39341289e-01
3.61131467e-02 -7.34110534e-01 1.82491049e-01 5.09122193e-01
5.79913437e-01 -8.04009736e-01 1.11634386e+00 4.36650783e-01
-7.38233745e-01 3.14661950e-01 -4.78097022e-01 -1.45867094e-01
-7.89480358e-02 1.06420505e+00 -1.09091175e+00 3.49841446e-01
1.51103184e-01 1.81709319e-01 -3.67613405e-01 9.72321093e-01
-4.97137085e-02 4.89196301e-01 -7.54498780e-01 -2.71882862e-01
-1.29863515e-01 -6.51430905e-01 1.54023707e-01 1.00822449e+00
2.07803801e-01 1.29779801e-01 -1.47587329e-01 7.85631716e-01
3.83306026e-01 2.92883068e-01 -6.98327497e-02 -2.26183563e-01
5.40889502e-01 7.81452775e-01 -5.95397353e-01 1.38147213e-02
-5.10299504e-01 4.53044534e-01 4.42384571e-01 2.24104747e-01
-5.91120720e-01 -2.22847864e-01 9.54853356e-01 2.70879805e-01
3.23765397e-01 -1.74536496e-01 -8.47445428e-02 -7.89183974e-01
1.69204786e-01 -7.10390568e-01 2.90497184e-01 -4.58644658e-01
-1.10746646e+00 2.04615712e-01 3.46694477e-02 -1.03376281e+00
-3.78603160e-01 -9.46689308e-01 -1.73441648e-01 1.22969306e+00
-8.89345527e-01 -3.41312617e-01 2.70420283e-01 -4.02230859e-01
-1.37278363e-01 2.04795688e-01 8.16773593e-01 -9.33399349e-02
-4.25744653e-01 3.91940325e-01 7.07406700e-01 -4.27962452e-01
7.89043665e-01 -1.14883459e+00 1.73602611e-01 5.52444160e-01
-3.07652205e-01 7.63096988e-01 1.37180173e+00 -7.95526385e-01
-1.06649160e+00 -5.54274619e-01 7.07273066e-01 -4.56129879e-01
7.30426252e-01 3.80167924e-02 -1.22489595e+00 2.44358316e-01
-5.12793839e-01 -1.26427948e-01 1.00714386e+00 4.48701054e-01
-3.81380737e-01 1.31459773e-01 -1.27763915e+00 1.00200981e-01
5.74877024e-01 -2.69880295e-01 -3.10284942e-01 2.04074293e-01
1.94543228e-01 -3.47609609e-01 -1.34360504e+00 3.96976888e-01
9.76232767e-01 -1.23682010e+00 7.23102748e-01 -6.71987593e-01
1.64397329e-01 -4.94520336e-01 -1.94106802e-01 -1.25251031e+00
-2.98086941e-01 -1.41461402e-01 2.97257543e-01 9.47722971e-01
6.78285539e-01 -5.12819052e-01 8.09927642e-01 1.08734155e+00
2.70359606e-01 -1.10584557e+00 -1.00429034e+00 -9.47959483e-01
1.45718828e-01 -1.64583936e-01 6.21705651e-01 4.92405564e-01
3.21370184e-01 2.14287177e-01 -2.42736369e-01 1.40574038e-01
4.63208497e-01 3.32864672e-02 5.65083683e-01 -1.25470281e+00
-5.71907103e-01 -6.29916787e-02 -7.61361539e-01 -6.13852620e-01
-1.57997906e-01 -3.39089513e-01 -5.29403575e-02 -8.13607097e-01
2.11130247e-01 -2.02746734e-01 -3.14356744e-01 3.25066112e-02
-2.15099797e-01 -2.18434542e-01 9.49811563e-03 5.65816127e-02
-2.61036474e-02 1.76248252e-01 7.35190570e-01 3.39672893e-01
-4.02703792e-01 -5.72629012e-02 -4.02452558e-01 4.43476588e-01
7.94533610e-01 -2.74865299e-01 -2.62455344e-01 2.63531357e-01
2.69115776e-01 5.23207262e-02 4.89143059e-02 -8.01323652e-01
-3.47507983e-01 -3.29883128e-01 5.31081259e-01 -4.33165163e-01
9.79890078e-02 -6.99587822e-01 6.86543941e-01 5.23918927e-01
-3.41094464e-01 -3.03387642e-01 1.21074274e-01 6.43186450e-01
2.29084417e-01 -3.39100093e-01 1.01690710e+00 -7.85994306e-02
-2.17301577e-01 -2.03436121e-01 -2.45993346e-01 -2.54974872e-01
6.27720952e-01 -1.89833909e-01 -2.98773289e-01 -1.87852859e-01
-7.74637461e-01 -2.00110063e-01 9.29134429e-01 -5.68559021e-02
8.28964934e-02 -6.77035928e-01 -4.16823328e-01 8.22190121e-02
3.05358857e-01 -6.91983163e-01 -6.39154064e-03 8.46976995e-01
-5.68150938e-01 6.71902895e-01 -1.69744253e-01 -5.02360463e-01
-1.49201846e+00 7.88107574e-01 4.58081514e-01 6.82932045e-03
-6.95181414e-02 1.87782183e-01 -5.45853414e-02 -1.19577691e-01
-3.38546216e-01 -1.97403461e-01 1.83596179e-01 -2.46722639e-01
3.05617213e-01 3.86022627e-01 4.43191379e-01 -5.39024591e-01
-4.64791209e-01 5.17164707e-01 -1.34529635e-01 4.35014427e-01
1.07972860e+00 -1.01154394e-01 -1.88043222e-01 4.84624088e-01
1.06220782e+00 -9.87596586e-02 -1.14151382e+00 1.21382080e-01
2.64456570e-01 -4.34352368e-01 -2.12323725e-01 -9.98632669e-01
-4.13039535e-01 3.35069090e-01 7.92279959e-01 1.92254931e-02
1.00035512e+00 9.65474471e-02 3.53611559e-01 3.03267241e-01
5.09697914e-01 -9.33194816e-01 -5.68265617e-01 8.85791257e-02
5.88222146e-01 -1.11352193e+00 4.57425535e-01 -6.25051260e-01
-4.61913645e-01 1.15956819e+00 3.78312588e-01 6.65787980e-02
4.03856099e-01 1.44757763e-01 9.35401544e-02 7.74105638e-02
-9.43057120e-01 -1.72702879e-01 1.27236068e-01 7.43451416e-01
8.49726856e-01 2.27911264e-01 -1.25293708e+00 4.74310905e-01
-1.54053986e-01 2.03005120e-01 4.58873957e-01 6.54442012e-01
-6.37576759e-01 -1.58805609e+00 -3.11909705e-01 5.01719773e-01
-4.14498210e-01 1.42527744e-01 -5.87259114e-01 7.34247804e-01
-7.98447896e-03 8.54175687e-01 -1.85542494e-01 -1.69270247e-01
4.12368447e-01 2.92329997e-01 5.73694944e-01 -1.78461716e-01
-1.83192387e-01 3.57647873e-02 3.79654855e-01 -2.66452909e-01
-3.30113947e-01 -7.32779860e-01 -1.26975214e+00 -3.55974376e-01
-6.63078964e-01 5.62656045e-01 6.48262322e-01 9.45370853e-01
4.53883678e-01 -8.53232518e-02 3.76120269e-01 -3.03200692e-01
-7.54570007e-01 -8.14466357e-01 -9.05299485e-01 4.17021155e-01
4.06854078e-02 -8.08191478e-01 -4.36650097e-01 -1.48997530e-01] | [4.88426399230957, 5.357780456542969] |
224f7291-3759-430b-99df-43bb24b1c473 | image-based-navigation-using-visual-features | 1812.03795 | null | https://arxiv.org/abs/1812.03795v2 | https://arxiv.org/pdf/1812.03795v2.pdf | Mapping, Localization and Path Planning for Image-based Navigation using Visual Features and Map | Building on progress in feature representations for image retrieval, image-based localization has seen a surge of research interest. Image-based localization has the advantage of being inexpensive and efficient, often avoiding the use of 3D metric maps altogether. That said, the need to maintain a large number of reference images as an effective support of localization in a scene, nonetheless calls for them to be organized in a map structure of some kind. The problem of localization often arises as part of a navigation process. We are, therefore, interested in summarizing the reference images as a set of landmarks, which meet the requirements for image-based navigation. A contribution of this paper is to formulate such a set of requirements for the two sub-tasks involved: map construction and self-localization. These requirements are then exploited for compact map representation and accurate self-localization, using the framework of a network flow problem. During this process, we formulate the map construction and self-localization problems as convex quadratic and second-order cone programs, respectively. We evaluate our methods on publicly available indoor and outdoor datasets, where they outperform existing methods significantly. | ['Luc van Gool', 'Danda Pani Paudel', 'Thomas Probst', 'Ajad Chhatkuli', 'Janine Thoma'] | 2018-12-10 | mapping-localization-and-path-planning-for | http://openaccess.thecvf.com/content_CVPR_2019/html/Thoma_Mapping_Localization_and_Path_Planning_for_Image-Based_Navigation_Using_Visual_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Thoma_Mapping_Localization_and_Path_Planning_for_Image-Based_Navigation_Using_Visual_CVPR_2019_paper.pdf | cvpr-2019-6 | ['image-based-localization'] | ['computer-vision'] | [ 3.93153988e-02 -2.43155196e-01 -2.41934940e-01 -5.05481958e-01
-9.29704666e-01 -7.97340214e-01 5.95744669e-01 4.25626785e-01
-5.76494932e-01 5.76312840e-01 -4.50977720e-02 -2.42893130e-01
-4.60613400e-01 -7.79747188e-01 -8.25356662e-01 -5.36944211e-01
-1.63227804e-02 1.94821924e-01 1.22128539e-01 -1.92856178e-01
6.65455222e-01 8.14467490e-01 -1.58598042e+00 -4.55426127e-01
9.74117100e-01 1.20415628e+00 4.63779718e-01 3.86155486e-01
-1.16766825e-01 6.19095266e-01 -1.45721301e-01 -1.14301413e-01
3.05982023e-01 -2.58625984e-01 -8.48066032e-01 2.95717508e-01
4.75460649e-01 4.10665125e-02 -2.65192628e-01 1.20980465e+00
4.86049920e-01 5.97442687e-01 5.26026905e-01 -1.27852535e+00
-3.34222466e-01 -1.98082346e-02 -4.87484753e-01 -7.64756575e-02
6.33064687e-01 -6.22091472e-01 9.36520398e-01 -1.14567637e+00
6.68945849e-01 6.45455241e-01 6.08946800e-01 1.10505838e-02
-1.06460905e+00 -7.43361190e-02 1.53374478e-01 1.37372911e-01
-1.82881856e+00 -6.40936494e-01 8.17769170e-01 -4.35506374e-01
4.21057075e-01 3.31592053e-01 5.61452389e-01 3.76270235e-01
3.89475115e-02 4.12741899e-01 8.92299891e-01 -5.55290937e-01
5.66173255e-01 3.67802441e-01 2.32452434e-02 8.97396505e-01
2.31775194e-01 -2.78461069e-01 -7.53818512e-01 -8.92811790e-02
8.69077682e-01 1.11767948e-01 -4.82286394e-01 -9.79602277e-01
-1.18280578e+00 8.61051917e-01 7.84875453e-01 3.26368332e-01
-3.46520662e-01 2.49334872e-01 7.43106455e-02 4.43005823e-02
4.51909930e-01 3.32004935e-01 1.00754283e-01 -1.23818956e-01
-1.01069558e+00 1.05935954e-01 8.89756978e-01 1.31887853e+00
1.26732528e+00 -4.78423178e-01 3.06290418e-01 8.14725876e-01
2.87816912e-01 5.75486422e-01 1.83096603e-01 -9.53561842e-01
6.07560754e-01 6.68393075e-01 4.08863127e-01 -1.71036077e+00
-2.81878620e-01 -5.74964046e-01 -8.35489333e-01 -6.58913404e-02
2.68904418e-01 3.71634096e-01 -5.05780637e-01 1.76325369e+00
3.63732994e-01 1.41381711e-01 -7.72561654e-02 8.45487714e-01
2.17623830e-01 6.07227266e-01 -5.03390312e-01 -7.12505579e-02
1.03059888e+00 -1.08522439e+00 -4.96340871e-01 -2.04036802e-01
6.00340009e-01 -6.45867050e-01 7.18430042e-01 -8.90796632e-02
-9.42251146e-01 -4.78205651e-01 -1.14515603e+00 -1.77966475e-01
-6.73112512e-01 1.78144723e-01 5.90606511e-01 5.71018040e-01
-1.48208165e+00 3.25630933e-01 -5.71722448e-01 -4.97757375e-01
9.52079147e-02 3.96947533e-01 -7.56541669e-01 -2.94896901e-01
-6.81192696e-01 8.25324774e-01 2.05829397e-01 4.76551503e-01
-3.61558914e-01 -3.09355170e-01 -1.24839032e+00 -9.99978036e-02
3.30964446e-01 -5.62758625e-01 7.98363388e-01 -5.66843390e-01
-9.63725984e-01 9.55570459e-01 -5.91202378e-01 -3.00350100e-01
6.10776126e-01 -1.21454298e-01 -5.58728762e-02 1.32123709e-01
7.20245659e-01 4.72367555e-01 5.77903986e-01 -1.33232737e+00
-8.71591866e-01 -5.17598033e-01 3.88627440e-01 4.22247440e-01
-3.02903317e-02 -3.76705199e-01 -9.28678095e-01 -1.59509808e-01
7.52696812e-01 -1.00290275e+00 -4.63304073e-01 2.49299362e-01
-4.00752455e-01 -1.04634017e-01 5.58756948e-01 -3.98449332e-01
1.09860957e+00 -2.10620332e+00 2.42914677e-01 6.09185159e-01
1.34479254e-01 -3.47601920e-01 5.63615970e-02 5.11208951e-01
2.04566047e-01 6.93850890e-02 -2.06602573e-01 -6.82127059e-01
-1.30125908e-02 2.48759925e-01 -3.17535162e-01 9.23250616e-01
-1.13483950e-01 7.28202939e-01 -1.12790775e+00 -6.85273528e-01
4.86764073e-01 3.89537781e-01 -4.16559130e-01 1.22510577e-02
2.80729681e-01 5.26780427e-01 -7.33252943e-01 7.97631383e-01
6.93388939e-01 -2.13552505e-01 -3.23923141e-01 -2.22557858e-01
-6.75929964e-01 -8.94505084e-02 -1.29329360e+00 2.30220556e+00
-6.73366189e-01 6.24819219e-01 1.15186512e-01 -1.09001946e+00
9.35482085e-01 -1.78489491e-01 8.51213694e-01 -7.94336140e-01
-9.90158245e-02 5.90944350e-01 -5.80816984e-01 -7.98273459e-02
7.78654218e-01 4.62111056e-01 -2.28586674e-01 3.91842633e-01
-2.24068239e-01 -1.69105396e-01 3.87118161e-01 2.48519853e-01
1.04006338e+00 8.45355764e-02 1.86172515e-01 -4.20592666e-01
8.07366192e-01 2.77158707e-01 2.02475190e-01 9.05287325e-01
-3.31404395e-02 5.97725153e-01 8.19124579e-02 -5.31571567e-01
-8.92094016e-01 -9.45536613e-01 -2.49221429e-01 6.91211700e-01
6.58978939e-01 -4.24191266e-01 -6.14325225e-01 -4.86740977e-01
-4.18851860e-02 1.04869835e-01 -5.23696423e-01 1.88762859e-01
-6.64724290e-01 -2.95709789e-01 8.52681249e-02 1.71158388e-01
7.11965263e-01 -5.05072534e-01 -8.77173185e-01 2.57754236e-01
-5.04047632e-01 -1.01651490e+00 -5.48107386e-01 4.01138306e-01
-5.56627214e-01 -1.24110556e+00 -9.70102608e-01 -9.78751302e-01
1.19165075e+00 8.68366241e-01 8.97291183e-01 1.21864907e-01
-2.47967124e-01 8.94627869e-01 -4.39267695e-01 -1.07860588e-01
2.47039825e-01 2.09403142e-01 2.17165098e-01 1.60378397e-01
-1.50574908e-01 -5.78418016e-01 -7.36978531e-01 5.39132357e-01
-7.89137304e-01 -4.75670807e-02 5.50207675e-01 7.63761520e-01
1.13628399e+00 -1.70046613e-02 1.92428797e-01 -4.66218174e-01
3.44589531e-01 -5.14462173e-01 -9.53058779e-01 4.54356641e-01
-4.19779748e-01 1.29368097e-01 3.66046995e-01 3.43383849e-02
-5.19054294e-01 6.07210994e-01 -7.48927053e-03 -9.61405858e-02
2.29874775e-01 6.75305843e-01 -1.38601542e-01 -7.58439183e-01
4.07778502e-01 7.31812000e-01 -7.32804686e-02 -2.17297137e-01
5.55078983e-01 5.93842745e-01 5.62957227e-01 -5.02390862e-01
9.76200819e-01 6.99744284e-01 3.40749025e-01 -1.01756334e+00
-9.73574519e-01 -1.02726126e+00 -1.02872705e+00 -3.14431548e-01
5.16430080e-01 -8.20310473e-01 -5.03986239e-01 9.75032803e-03
-1.14418316e+00 1.36754841e-01 -1.83331162e-01 2.44079500e-01
-8.73402953e-01 4.06724811e-01 1.67089552e-01 -8.10749590e-01
1.00588150e-01 -1.28360188e+00 1.10976422e+00 2.85789788e-01
1.63203642e-01 -1.06839490e+00 3.45128298e-01 1.45881385e-01
4.40512896e-01 3.87915939e-01 4.28044617e-01 -2.04550624e-01
-1.04131746e+00 -5.92198670e-01 -4.07270551e-01 8.04322734e-02
2.48751566e-01 -5.95782518e-01 -7.52631485e-01 -2.76569694e-01
6.09881282e-02 -6.45595118e-02 6.00793958e-01 3.39268774e-01
9.78410363e-01 -9.31315199e-02 -4.63960797e-01 9.78120685e-01
1.74005711e+00 1.03630893e-01 3.71483058e-01 5.28839469e-01
5.06462157e-01 7.35809386e-01 7.95792818e-01 2.38621324e-01
7.17907488e-01 8.69100928e-01 5.82549930e-01 -1.02034591e-01
2.76484758e-01 -4.52340037e-01 -1.95005476e-01 7.68384874e-01
-4.86986293e-03 -2.40547523e-01 -9.17553663e-01 5.69429994e-01
-2.27297926e+00 -7.41496384e-01 -7.61281177e-02 2.37899208e+00
2.44374201e-01 -4.90688890e-01 -2.56342173e-01 2.24388875e-02
6.43263698e-01 1.83324605e-01 -3.57867748e-01 9.63972658e-02
-7.51145557e-02 -3.28418136e-01 7.66223907e-01 5.38472354e-01
-1.37135398e+00 6.17003977e-01 5.60387945e+00 6.34730577e-01
-9.82213497e-01 -9.36667770e-02 4.93102342e-01 5.32297254e-01
-1.41754463e-01 1.75886467e-01 -8.14074397e-01 3.83429825e-01
6.46717012e-01 -2.54129887e-01 3.91094089e-01 9.97101665e-01
9.21878070e-02 -6.44797385e-01 -1.11424887e+00 1.41813087e+00
2.82850325e-01 -1.42490900e+00 -2.34173283e-01 2.39147484e-01
7.66611457e-01 1.00682133e-04 1.29548818e-01 -1.49918303e-01
-2.01292440e-01 -8.75078440e-01 1.01016057e+00 3.77666861e-01
8.38248968e-01 -7.59731531e-01 7.09767520e-01 3.71649027e-01
-1.58557999e+00 4.85275276e-02 -6.35156751e-01 1.04594484e-01
4.12375182e-01 5.69242954e-01 -3.79426986e-01 8.63999605e-01
4.67212349e-01 7.29128599e-01 -4.16244030e-01 1.58274293e+00
-3.79802808e-02 -1.70430988e-01 -4.79585528e-01 -1.30464166e-01
4.04673904e-01 -4.90390390e-01 2.46995792e-01 9.94630158e-01
5.13279438e-01 -1.03803918e-01 3.68204445e-01 6.61592245e-01
1.12473935e-01 3.50673229e-01 -9.10792828e-01 2.60314643e-01
5.78215659e-01 1.19672680e+00 -1.06531489e+00 7.31761605e-02
-3.42083156e-01 1.13852501e+00 4.94827956e-01 1.92492247e-01
-7.13444054e-01 -6.39372110e-01 4.31482583e-01 9.28010568e-02
-1.45546734e-01 -7.56031096e-01 -4.49924469e-02 -1.07033110e+00
3.74069691e-01 -1.60056651e-01 -8.19476321e-02 -4.75976437e-01
-7.05536723e-01 5.55486023e-01 -1.11842923e-01 -1.30970883e+00
-1.99746072e-01 -4.11071688e-01 -3.01939696e-01 8.95315766e-01
-1.82571423e+00 -1.13626182e+00 -8.51131201e-01 7.94496536e-01
3.63344818e-01 2.61714488e-01 8.28625798e-01 5.64470172e-01
-2.05673456e-01 2.39649177e-01 4.35735106e-01 1.10432170e-01
3.47777426e-01 -1.25551474e+00 8.68989751e-02 9.43314612e-01
5.42326331e-01 7.31915951e-01 4.21284825e-01 -4.67819065e-01
-1.70306492e+00 -1.11154747e+00 1.08555806e+00 -4.16688889e-01
5.63673675e-01 -5.22543252e-01 -3.01578194e-01 4.44906384e-01
-4.06978667e-01 3.01863760e-01 3.13882679e-01 -8.72739926e-02
-4.11465503e-02 -2.12249786e-01 -9.16854441e-01 4.49252695e-01
1.30397475e+00 -6.38637781e-01 -5.80004603e-02 5.66972613e-01
4.08049464e-01 -5.71505487e-01 -3.73886198e-01 -3.82256098e-02
3.67209435e-01 -9.49694574e-01 1.25272250e+00 2.12904379e-01
3.27452123e-02 -5.90464771e-01 -5.32711446e-01 -1.19119525e+00
-9.13622156e-02 -4.90498781e-01 3.52274299e-01 1.07417691e+00
1.95364252e-01 -6.28707230e-01 9.65272248e-01 5.16048431e-01
-1.10186234e-01 -6.19951963e-01 -1.17891014e+00 -8.28701973e-01
-6.14388645e-01 -4.36923474e-01 2.11469948e-01 5.94146430e-01
-1.02713563e-01 8.97539183e-02 -3.53082627e-01 3.17065209e-01
7.81263769e-01 1.55704856e-01 8.71357620e-01 -1.12255156e+00
3.71977836e-01 -1.83228076e-01 -8.77206564e-01 -1.49124980e+00
1.27342343e-01 -8.79247963e-01 4.03783679e-01 -1.82889986e+00
-6.82544634e-02 -1.02444935e+00 -2.04835728e-01 2.53545016e-01
3.68609667e-01 3.54769647e-01 6.87488839e-02 4.92611468e-01
-1.07622194e+00 5.97525537e-01 7.37575233e-01 -1.26843676e-01
-2.07915306e-01 2.95191884e-01 -5.64175427e-01 4.72152740e-01
5.27681291e-01 -3.31186146e-01 -5.52973449e-01 -5.42931855e-01
4.47111607e-01 7.36384690e-02 4.29344356e-01 -1.20238245e+00
9.26408589e-01 -4.79244441e-02 1.83800980e-01 -8.90301764e-01
6.60839975e-01 -1.15920985e+00 1.57907590e-01 3.65241230e-01
-1.11006141e-01 4.34252322e-01 -2.65473753e-01 8.91403258e-01
-5.65012753e-01 -4.24625456e-01 4.14606005e-01 -1.85620904e-01
-1.07850015e+00 4.57973391e-01 -4.09241207e-02 -1.52216569e-01
1.21323550e+00 -5.43719828e-01 1.05943874e-01 -4.53874856e-01
-4.71695572e-01 3.37027282e-01 6.40359104e-01 2.77651876e-01
8.06564927e-01 -1.33242810e+00 -2.70302743e-01 2.26572901e-01
3.54271859e-01 3.00924867e-01 8.66961032e-02 9.63655412e-01
-8.35826159e-01 7.56537080e-01 -9.23580155e-02 -8.48925948e-01
-8.69023025e-01 3.42023075e-01 1.97564065e-01 -4.18613292e-02
-5.31812191e-01 7.70550907e-01 8.46786723e-02 -4.44266140e-01
4.50015247e-01 -1.50514066e-01 -3.16470936e-02 3.46106365e-02
4.65292543e-01 4.07918215e-01 3.76199894e-02 -9.52166975e-01
-6.15689814e-01 1.03717530e+00 4.22644675e-01 -1.59622967e-01
1.32233405e+00 -7.10118115e-01 -3.88455898e-01 2.59221911e-01
1.34025562e+00 1.79365382e-01 -1.00320137e+00 -2.50458837e-01
3.45665067e-01 -6.91777527e-01 -2.67919954e-02 -3.04600626e-01
-7.11004734e-01 5.95146120e-01 4.94766325e-01 1.80986717e-01
9.57445920e-01 1.06655516e-01 3.43427032e-01 6.60428107e-01
1.14960074e+00 -9.64763939e-01 5.46990288e-03 5.87101877e-01
8.24680686e-01 -1.37689817e+00 -1.57927811e-01 -5.57599962e-01
-3.18670273e-02 1.16596842e+00 2.18473628e-01 -1.58994541e-01
7.13858485e-01 -2.14791410e-02 -1.83930531e-01 -1.24608725e-01
1.46617457e-01 -3.50685567e-01 3.90874594e-01 7.04957545e-01
1.02682181e-01 -1.75849304e-01 -1.70407802e-01 1.38192266e-01
-1.63210899e-01 -2.82064736e-01 1.28162786e-01 1.15651369e+00
-6.97764456e-01 -9.81789351e-01 -2.78490305e-01 1.08831741e-01
2.79272925e-02 9.30970162e-02 -1.67384073e-01 7.46454239e-01
6.19805343e-02 9.80805993e-01 -5.40984534e-02 -1.75675586e-01
3.86846781e-01 -3.91946048e-01 2.97725558e-01 -5.12813389e-01
1.26898274e-01 -2.97245651e-01 -3.42474699e-01 -8.55836153e-01
-6.08918905e-01 -3.86791736e-01 -8.92228186e-01 -4.85020094e-02
-4.49397504e-01 4.88991648e-01 1.26053989e+00 7.91442513e-01
3.42826605e-01 1.32482052e-01 8.52231979e-01 -1.09185624e+00
-4.37276453e-01 -2.62849897e-01 -6.49231791e-01 -5.55190630e-03
3.78978580e-01 -6.98628902e-01 -2.47434869e-01 -1.23337068e-01] | [7.662446975708008, -2.175915241241455] |
2dee3d9a-ae21-4686-92f1-6e8c5f1a8b7d | unsupervised-document-embedding-with-cnns | 1711.04168 | null | http://arxiv.org/abs/1711.04168v3 | http://arxiv.org/pdf/1711.04168v3.pdf | Unsupervised Document Embedding With CNNs | We propose a new model for unsupervised document embedding. Leading existing
approaches either require complex inference or use recurrent neural networks
(RNN) that are difficult to parallelize. We take a different route and develop
a convolutional neural network (CNN) embedding model. Our CNN architecture is
fully parallelizable resulting in over 10x speedup in inference time over RNN
models. Parallelizable architecture enables to train deeper models where each
successive layer has increasingly larger receptive field and models longer
range semantic structure within the document. We additionally propose a fully
unsupervised learning algorithm to train this model based on stochastic forward
prediction. Empirical results on two public benchmarks show that our approach
produces comparable to state-of-the-art accuracy at a fraction of computational
cost. | ['Maksims Volkovs', 'Shunan Zhao', 'Chundi Liu'] | 2017-11-11 | null | null | null | null | ['document-embedding'] | ['methodology'] | [ 1.63690254e-01 3.27446401e-01 -5.81220567e-01 -3.52740943e-01
-6.13153100e-01 -4.09859449e-01 7.14722216e-01 2.20789611e-01
-6.93065226e-01 3.15962285e-01 6.20310664e-01 -5.73340654e-01
1.97239712e-01 -9.87319529e-01 -7.55664885e-01 -4.41330224e-01
5.65156937e-02 4.22538012e-01 1.62560910e-01 -7.12877661e-02
4.19644922e-01 2.80197680e-01 -1.21597028e+00 5.12544096e-01
2.05544636e-01 6.92511022e-01 9.46782380e-02 1.19983244e+00
-2.63560504e-01 1.13253701e+00 -3.14877898e-01 -2.62722969e-01
1.31261885e-01 -4.13089618e-02 -1.02366197e+00 -3.45865548e-01
2.15785548e-01 -5.47595978e-01 -7.85437882e-01 8.58049810e-01
4.57761675e-01 2.75460601e-01 6.04470789e-01 -6.47228003e-01
-1.04915297e+00 9.43589330e-01 -2.69576292e-02 8.32853690e-02
-1.71569869e-01 -1.44317135e-01 1.48578858e+00 -1.05600083e+00
5.73526144e-01 1.03244257e+00 8.88901472e-01 6.78726673e-01
-1.41970074e+00 -2.57432491e-01 3.41418475e-01 -3.74294408e-02
-1.16946030e+00 -2.23299712e-01 5.60516894e-01 -5.32464571e-02
1.73675382e+00 1.21107861e-01 4.75612700e-01 1.29424024e+00
2.50931710e-01 8.84453893e-01 7.19270289e-01 -5.71304023e-01
2.48784006e-01 1.52601168e-01 4.27127630e-01 8.71446371e-01
2.10918039e-01 -4.29032594e-02 -3.50073427e-01 -4.07587178e-02
6.45758867e-01 5.29217303e-01 3.62143070e-01 4.75143157e-02
-1.04187036e+00 1.13110733e+00 6.69832647e-01 2.52622157e-01
-1.22158863e-01 7.36291051e-01 7.41194189e-01 3.24631840e-01
5.90587616e-01 3.18356484e-01 -6.21369660e-01 -2.72145391e-01
-9.92145061e-01 1.31612435e-01 8.22281003e-01 7.64479756e-01
6.53363287e-01 1.47403076e-01 -5.57962842e-02 9.24801528e-01
2.53501922e-01 1.98801011e-01 7.82168865e-01 -8.21186960e-01
3.44330370e-01 6.21948898e-01 -3.44092190e-01 -1.05382442e+00
-3.33552957e-01 -5.02324045e-01 -1.05399024e+00 -2.70141605e-02
-7.56106675e-02 1.26235217e-01 -9.82908070e-01 1.09409547e+00
-2.18584985e-01 8.88949707e-02 2.17056617e-01 4.41351384e-01
6.39470279e-01 1.12619865e+00 -8.68303999e-02 3.46856564e-01
1.17444241e+00 -1.66835296e+00 -4.67698663e-01 -2.42996797e-01
1.08298326e+00 -4.17368829e-01 9.77819800e-01 2.98252285e-01
-1.05070662e+00 -5.78558922e-01 -1.01472294e+00 -5.50308764e-01
-8.86964023e-01 3.46931934e-01 9.50753391e-01 6.47198379e-01
-1.37829149e+00 5.95893443e-01 -1.10631406e+00 -3.07527989e-01
4.61389720e-01 5.76727331e-01 -2.36047253e-01 1.77040383e-01
-9.61630821e-01 6.73068762e-01 6.62966967e-01 2.61587709e-01
-8.07302237e-01 -5.71250796e-01 -9.68212783e-01 3.02003056e-01
-1.73624024e-01 -6.54132128e-01 1.12694764e+00 -9.13300812e-01
-1.86761773e+00 5.66922009e-01 -4.96130228e-01 -9.85623896e-01
-3.33005376e-02 -2.57210881e-01 -2.12000370e-01 1.70409549e-02
-3.40870738e-01 8.69199157e-01 7.68072963e-01 -7.85838485e-01
-1.71451166e-01 -3.21676061e-02 1.50627002e-01 -2.09254950e-01
-1.03175187e+00 1.93144426e-01 -4.48974341e-01 -9.07603741e-01
1.08252212e-01 -1.08585656e+00 -5.55684447e-01 -1.38390372e-02
-5.17957449e-01 -4.17112023e-01 7.70736039e-01 -5.48962712e-01
1.46049726e+00 -1.97320509e+00 1.57401234e-01 1.49187371e-01
3.76429200e-01 2.66567558e-01 -1.75967231e-01 6.37855291e-01
8.75146687e-02 2.46242344e-01 -2.84826487e-01 -6.24529362e-01
2.35975742e-01 5.09146631e-01 -7.72842109e-01 4.58194502e-02
2.21400276e-01 1.36661828e+00 -8.25851321e-01 -2.78153360e-01
-9.24507454e-02 6.37610793e-01 -1.09731078e+00 6.98680654e-02
-3.44788373e-01 -1.66348442e-01 -1.75756857e-01 4.94140208e-01
3.40476006e-01 -7.44685054e-01 3.72527301e-01 2.58673821e-02
6.68979660e-02 8.62467825e-01 -6.82526469e-01 1.81206417e+00
-7.03137636e-01 1.01387036e+00 -6.98994458e-01 -1.28756225e+00
8.78205597e-01 1.18058234e-01 -5.31531610e-02 -3.40177000e-01
-8.68205428e-02 1.70034021e-01 -1.57090291e-01 -1.96729302e-01
8.96227062e-01 2.07713142e-01 2.88117751e-02 7.48684466e-01
3.31085443e-01 3.13333422e-01 -1.27757877e-01 3.68732035e-01
1.43294990e+00 -1.54074818e-01 -1.39790550e-01 -2.14363411e-01
3.59218985e-01 -2.35056937e-01 1.53466523e-01 1.00051045e+00
2.37927169e-01 5.29888093e-01 5.70308626e-01 -9.95414436e-01
-1.24112189e+00 -9.63285863e-01 7.57313147e-02 1.42010891e+00
-4.49269921e-01 -8.93062830e-01 -5.45867801e-01 -8.16824496e-01
-1.18947618e-01 4.06925946e-01 -9.19849038e-01 -1.50284678e-01
-7.31419683e-01 -7.83218265e-01 9.41761076e-01 1.18168604e+00
2.77423501e-01 -1.23294520e+00 -4.26432937e-01 2.69785255e-01
2.06252202e-01 -1.01741910e+00 -1.55072138e-01 5.17870963e-01
-1.38008416e+00 -4.18434560e-01 -5.80851734e-01 -9.71998632e-01
9.86713171e-01 -9.34955180e-02 1.15915310e+00 2.36965761e-01
-2.97324568e-01 2.11407873e-03 -1.84955209e-01 -2.30230257e-01
-1.96412235e-01 6.54717565e-01 -1.33537635e-01 -4.26818907e-01
5.35967708e-01 -4.86912549e-01 -5.95821083e-01 -3.11184406e-01
-1.06642497e+00 -1.07235856e-01 5.13688207e-01 1.29402101e+00
4.52267587e-01 6.12381957e-02 2.77629375e-01 -1.23158813e+00
6.55746698e-01 -4.28977638e-01 -5.22448301e-01 4.89820316e-02
-8.50489259e-01 4.44114566e-01 8.55839550e-01 -4.51296300e-01
-8.08057487e-01 -5.59634976e-02 -3.45594138e-01 -2.47822389e-01
1.07007511e-01 6.56816721e-01 6.42919064e-01 1.78439602e-01
4.85667020e-01 4.74896580e-01 -3.83348241e-02 -5.77533960e-01
4.72425997e-01 6.11449718e-01 2.38968730e-01 -5.13187647e-01
7.47975290e-01 7.44362295e-01 -2.39517525e-01 -6.68806732e-01
-8.66198421e-01 -1.98422208e-01 -5.91912687e-01 4.17697698e-01
6.83894098e-01 -9.66796577e-01 -7.61079013e-01 1.05810530e-01
-1.36989808e+00 -5.46538889e-01 -2.35115886e-01 4.66939688e-01
-2.49183238e-01 1.32755026e-01 -1.21853530e+00 -5.99474728e-01
-8.13844860e-01 -1.00967932e+00 1.07093275e+00 -1.08909048e-01
-2.89954364e-01 -1.44714832e+00 3.84137213e-01 -2.66340468e-02
7.48636246e-01 -3.68873060e-01 9.66342866e-01 -6.72906518e-01
-5.87263107e-01 -4.06115770e-01 -4.45811331e-01 6.12854242e-01
-1.80009797e-01 1.27094999e-01 -9.74669337e-01 -4.27213311e-01
-4.82376724e-01 -4.76898193e-01 1.33719981e+00 2.69323826e-01
1.70105839e+00 -5.65264404e-01 -3.73897880e-01 7.64397919e-01
1.62714732e+00 -3.78995597e-01 8.99855733e-01 5.83391964e-01
7.89422631e-01 3.29287708e-01 5.32045262e-03 3.29621524e-01
2.96700060e-01 3.83541852e-01 7.10926875e-02 -4.41338792e-02
4.82018143e-02 -5.12397587e-01 6.82690322e-01 1.50491667e+00
-2.37030923e-01 -1.96107969e-01 -8.94570529e-01 7.47485459e-01
-1.81926107e+00 -8.67492795e-01 1.62115693e-01 1.69710636e+00
7.68742621e-01 4.01244015e-01 -1.11936219e-01 8.34865198e-02
1.21078923e-01 3.72266322e-01 -2.84217715e-01 -1.04763877e+00
5.82106486e-02 8.18930507e-01 8.24540317e-01 5.41782975e-01
-9.87340748e-01 1.24593532e+00 7.56250668e+00 6.17929459e-01
-1.14214087e+00 1.34311602e-01 6.58039033e-01 -1.57382339e-01
-5.26680768e-01 1.01759285e-01 -1.18255258e+00 3.93929690e-01
1.48247671e+00 3.26826036e-01 4.21289772e-01 8.92608643e-01
-2.99047977e-01 4.49266732e-01 -9.55282986e-01 8.19575489e-01
7.83725232e-02 -1.95988357e+00 3.50410581e-01 1.55181199e-01
9.51752841e-01 5.90018511e-01 2.53482342e-01 4.37976480e-01
6.55317128e-01 -1.33533144e+00 1.94874719e-01 5.56781173e-01
6.48275495e-01 -9.55065370e-01 7.01817155e-01 2.52420634e-01
-1.10320914e+00 -2.29173526e-01 -8.91636550e-01 -2.41444841e-01
-2.65847832e-01 3.44586879e-01 -6.42740250e-01 8.75995830e-02
8.42525423e-01 9.60201085e-01 -6.37875497e-01 4.81471062e-01
-3.12675953e-01 8.41914058e-01 -2.43355945e-01 -4.75369662e-01
7.65171528e-01 1.69094056e-01 1.71408817e-01 1.62066495e+00
5.93107492e-02 -2.68208116e-01 -1.47516906e-01 7.48643458e-01
-5.46679378e-01 -9.57602356e-03 -7.44331241e-01 -3.01163018e-01
2.68080205e-01 1.14256227e+00 -7.80130565e-01 -5.39769709e-01
-4.78215635e-01 1.25156009e+00 7.89998591e-01 4.65964437e-01
-6.32906616e-01 -7.86599934e-01 4.99623030e-01 -1.60276070e-01
6.92299724e-01 -3.93178374e-01 -2.80272156e-01 -1.29472482e+00
4.48929258e-02 -6.58403456e-01 1.87231958e-01 -5.22020400e-01
-1.20981371e+00 7.21655011e-01 -4.18219477e-01 -8.31090748e-01
-3.82274151e-01 -1.11392200e+00 -5.76049805e-01 6.12609863e-01
-1.59339213e+00 -1.09880209e+00 1.68948442e-01 2.57073492e-01
6.15432501e-01 -2.85818815e-01 1.50725925e+00 1.71843335e-01
-4.69993740e-01 9.20963883e-01 6.01121902e-01 6.93798304e-01
2.64037877e-01 -1.18853176e+00 1.00014567e+00 5.15757322e-01
4.43376780e-01 1.09523368e+00 1.29625738e-01 -2.69098222e-01
-1.59003413e+00 -1.30160463e+00 1.39274240e+00 -6.21769667e-01
9.33796704e-01 -7.33485520e-01 -7.62542129e-01 1.06616569e+00
3.84939909e-01 4.06897724e-01 8.02112520e-01 4.96317297e-01
-8.99454415e-01 6.08162172e-02 -6.05280638e-01 5.81465125e-01
8.13872695e-01 -1.12382364e+00 -5.05065262e-01 2.97495425e-01
1.01413262e+00 -9.83100906e-02 -7.64837146e-01 2.33705729e-01
8.14238966e-01 -4.30812508e-01 1.12277865e+00 -9.81119454e-01
8.22902083e-01 4.97737080e-02 -8.05378854e-02 -8.96796763e-01
-3.72710019e-01 -5.35970688e-01 -7.42574930e-01 7.49938726e-01
6.77117109e-01 -8.43286932e-01 1.12190962e+00 2.29062691e-01
1.76968172e-01 -1.14671600e+00 -3.92517269e-01 -8.91459823e-01
3.08191121e-01 -6.18968368e-01 2.32724175e-01 8.68528366e-01
6.69049025e-02 3.22241157e-01 -2.92441994e-01 4.06580679e-02
3.10606748e-01 -3.71525511e-02 5.52124321e-01 -1.06501615e+00
-4.80003148e-01 -4.66608346e-01 -6.06924415e-01 -1.50584865e+00
5.30259550e-01 -1.28590763e+00 -1.73365936e-01 -1.65673459e+00
3.82450670e-01 -2.21880406e-01 -8.23226511e-01 6.47797048e-01
1.88301340e-01 5.25764108e-01 -1.33826226e-01 1.41941011e-01
-7.14854658e-01 4.93684441e-01 4.71381515e-01 -4.98209387e-01
-3.54139358e-02 -4.76824522e-01 -5.80751359e-01 7.63531625e-01
8.90823126e-01 -7.64827967e-01 -2.96048731e-01 -1.01577842e+00
6.92347229e-01 -5.35520971e-01 4.05438006e-01 -8.30090165e-01
5.23746550e-01 5.59361577e-01 6.02332652e-01 -7.22903192e-01
2.31756791e-01 -5.74128389e-01 -5.08708358e-01 6.37088597e-01
-8.89736414e-01 3.15921217e-01 3.29228520e-01 6.26650989e-01
-2.58937210e-01 -3.76810312e-01 3.76291335e-01 -9.30678844e-02
-3.66100997e-01 3.04695994e-01 -6.75360262e-01 -2.89353728e-01
3.16908747e-01 2.19982769e-03 -1.53950617e-01 -2.56702751e-01
-5.95345497e-01 -2.10733026e-01 1.65870085e-01 5.30382633e-01
8.98078203e-01 -1.23864973e+00 -5.34165025e-01 2.22998574e-01
5.96809536e-02 -2.55951077e-01 -1.29372478e-01 4.40452158e-01
-7.88642168e-01 8.95471990e-01 2.69134849e-01 -4.08049911e-01
-9.98152018e-01 4.65522140e-01 7.33035877e-02 -7.34721959e-01
-8.64677370e-01 1.11653280e+00 -6.93439841e-02 -8.93051147e-01
4.22778189e-01 -5.40161252e-01 2.58474518e-02 -3.36312920e-01
6.99678183e-01 2.54675865e-01 -4.71808575e-02 -2.90358484e-01
-7.79661685e-02 3.23017120e-01 -6.21678889e-01 -9.37905014e-02
1.52547336e+00 3.43526334e-01 -5.12776494e-01 5.00418663e-01
1.88474870e+00 -1.93595782e-01 -8.68746519e-01 -4.40219194e-01
1.72872826e-01 -2.75093727e-02 4.42423403e-01 -3.74212354e-01
-7.84073710e-01 1.15921438e+00 3.89382660e-01 8.02969113e-02
8.14360917e-01 -2.09744707e-01 1.08427072e+00 1.19979978e+00
3.25293057e-02 -1.34883189e+00 2.29526490e-01 1.03280127e+00
3.84714156e-01 -1.03873146e+00 5.46302721e-02 1.00364342e-01
-1.37638301e-01 1.46786666e+00 2.86864012e-01 -7.53327072e-01
9.44955885e-01 3.82559568e-01 -2.81582594e-01 -2.13734552e-01
-1.14532769e+00 9.04249027e-02 2.97812551e-01 7.39387572e-02
6.81297898e-01 -7.31694251e-02 5.16664945e-02 2.97214299e-01
-1.50186956e-01 -5.96656613e-02 2.40981560e-02 1.08970189e+00
-1.33531719e-01 -1.25929952e+00 2.80098200e-01 4.66079503e-01
-6.20021641e-01 -7.51619935e-01 -7.12862089e-02 4.36312020e-01
-3.84953916e-01 4.91440028e-01 4.25457090e-01 -3.99250954e-01
-1.56318069e-01 1.73107460e-01 3.55116278e-01 -7.69182086e-01
-6.67581499e-01 -2.35644668e-01 -2.08359569e-01 -6.09344244e-01
-2.59461194e-01 -1.29894674e-01 -1.35375369e+00 -3.77960652e-01
-2.62746632e-01 4.87127453e-02 8.16433966e-01 7.55353868e-01
6.71741128e-01 5.98425269e-01 5.77006757e-01 -6.59242272e-01
-5.09670913e-01 -9.84614313e-01 -1.82200551e-01 -2.35733241e-01
2.99721509e-01 -4.95833047e-02 -1.72947228e-01 1.67207137e-01] | [10.749000549316406, 7.705173015594482] |
d379ec2e-f548-4141-a71b-44e0c849fcfe | a-data-driven-strategy-to-combine-word | 2105.12788 | null | https://arxiv.org/abs/2105.12788v1 | https://arxiv.org/pdf/2105.12788v1.pdf | A data-driven strategy to combine word embeddings in information retrieval | Word embeddings are vital descriptors of words in unigram representations of documents for many tasks in natural language processing and information retrieval. The representation of queries has been one of the most critical challenges in this area because it consists of a few terms and has little descriptive capacity. Strategies such as average word embeddings can enrich the queries' descriptive capacity since they favor the identification of related terms from the continuous vector representations that characterize these approaches. We propose a data-driven strategy to combine word embeddings. We use Idf combinations of embeddings to represent queries, showing that these representations outperform the average word embeddings recently proposed in the literature. Experimental results on benchmark data show that our proposal performs well, suggesting that data-driven combinations of word embeddings are a promising line of research in ad-hoc information retrieval. | ['Marcelo Mendoza', 'Alfredo Silva'] | 2021-05-26 | null | null | null | null | ['ad-hoc-information-retrieval'] | ['natural-language-processing'] | [-2.47341082e-01 -1.79013520e-01 -6.83696389e-01 -1.47475764e-01
-6.38370514e-01 -5.20158529e-01 1.22032034e+00 8.74399960e-01
-9.16702569e-01 1.47220954e-01 5.77590406e-01 -1.72915593e-01
-6.22927248e-01 -8.61945868e-01 -3.83341424e-02 -4.49184775e-01
-2.31493950e-01 7.14027405e-01 4.35032398e-01 -8.05175126e-01
6.95111215e-01 7.84225345e-01 -1.80078280e+00 6.71214238e-02
3.33276480e-01 9.24476981e-01 2.59343535e-03 4.67114002e-01
-8.89187932e-01 1.26398981e-01 -5.70466280e-01 -1.47170216e-01
5.42193614e-02 2.17643246e-01 -8.57703149e-01 -3.87910098e-01
3.43794852e-01 -1.83621645e-01 -6.72239542e-01 9.17797327e-01
5.00928760e-01 5.58728278e-01 1.07534456e+00 -9.20800090e-01
-1.31007659e+00 3.50688130e-01 -2.96432436e-01 5.68205118e-01
6.11564636e-01 -6.47781491e-01 1.55548704e+00 -1.18645656e+00
5.33036530e-01 1.54160273e+00 3.40898395e-01 3.87863249e-01
-1.12650514e+00 -7.08395392e-02 1.40929610e-01 4.54612970e-01
-1.77123868e+00 -5.07913940e-02 6.20853961e-01 -4.23127443e-01
1.33575189e+00 5.25995076e-01 3.73097181e-01 1.03265870e+00
1.71332493e-01 7.19550669e-01 3.38916659e-01 -8.20488214e-01
2.81622242e-02 4.90196168e-01 9.61356461e-01 3.25566918e-01
7.21470237e-01 -1.34580284e-01 -1.74685776e-01 -6.69602394e-01
5.59215844e-01 4.58818227e-01 -2.20542252e-01 -4.69472706e-01
-1.01882076e+00 1.38570094e+00 4.92648810e-01 9.59558010e-01
-4.77214009e-01 3.49303484e-01 6.07825994e-01 3.04987937e-01
5.80699503e-01 1.14826524e+00 -2.49007329e-01 -4.74996641e-02
-5.75945020e-01 5.38902044e-01 4.30107057e-01 9.94184077e-01
7.50706732e-01 -2.66329318e-01 -5.78608394e-01 1.11359191e+00
4.49705780e-01 1.89792931e-01 1.18916821e+00 -3.27169538e-01
2.44287332e-03 8.75955641e-01 1.87594935e-01 -1.26912355e+00
-1.91648394e-01 -1.55667961e-01 -3.38022560e-01 -4.26441789e-01
-2.03346223e-01 3.03860605e-01 -8.92523527e-01 1.15976620e+00
-4.84861322e-02 -3.36815536e-01 -6.23349193e-03 7.18174934e-01
7.06877112e-01 9.87824082e-01 1.03066787e-01 -3.80896367e-02
1.76098239e+00 -6.81250811e-01 -1.03945696e+00 -1.25837162e-01
8.23904514e-01 -8.03197563e-01 9.18221653e-01 -1.46547789e-02
-8.20249796e-01 -6.09456420e-01 -9.86252606e-01 -4.52751100e-01
-1.35157728e+00 -2.78353184e-01 7.25824833e-01 6.84142053e-01
-1.12119913e+00 3.64717662e-01 -3.54499608e-01 -7.61970699e-01
4.02819701e-02 3.22839558e-01 -2.92364955e-01 -1.33669168e-01
-1.43522382e+00 1.09481895e+00 7.89896190e-01 -5.35041153e-01
-3.96541774e-01 -4.57065374e-01 -1.02717936e+00 3.58734310e-01
-4.30426486e-02 -3.50534856e-01 9.75409269e-01 -1.89517632e-01
-6.21033192e-01 8.18712294e-01 -2.96064645e-01 -3.39973241e-01
-3.10648888e-01 -2.78766185e-01 -6.42145336e-01 3.20058584e-01
-5.74768931e-02 6.69127524e-01 6.91946924e-01 -1.07562721e+00
-4.08669889e-01 -2.91906744e-01 2.06584468e-01 -3.11775636e-02
-1.31991601e+00 4.66105826e-02 -4.65013146e-01 -7.79710233e-01
-8.36263001e-02 -6.76571131e-01 -2.61466473e-01 3.68149690e-02
1.43452421e-01 -1.11376679e+00 7.40526617e-01 -4.59434167e-02
1.86247265e+00 -2.03162599e+00 1.20997047e-02 2.21201032e-01
3.06290984e-01 7.06519365e-01 -4.29961026e-01 1.21063626e+00
-1.55931368e-01 5.29613912e-01 2.92957902e-01 -1.85412377e-01
2.38701016e-01 3.52984339e-01 -6.36003137e-01 3.55798602e-01
2.79911250e-01 1.00012672e+00 -1.00929832e+00 -6.09651387e-01
1.64104864e-01 6.51634872e-01 -5.93934357e-01 4.12709147e-01
-1.61098048e-01 -5.86298645e-01 -9.69774604e-01 4.13928390e-01
4.19821054e-01 8.81244764e-02 -3.50898132e-02 -2.51811177e-01
4.90192398e-02 3.75518560e-01 -8.31452608e-01 1.53244233e+00
-6.41886592e-01 7.09649205e-01 -7.88487256e-01 -1.16976964e+00
1.09496570e+00 3.64336461e-01 4.13903117e-01 -6.73967421e-01
2.45174110e-01 1.17351070e-01 -2.30933055e-01 -8.40421736e-01
1.30723703e+00 -2.89991759e-02 -3.15658897e-01 5.13274550e-01
2.57163733e-01 -6.23271465e-02 4.49673742e-01 3.97757798e-01
9.58927691e-01 -5.94410717e-01 5.89597285e-01 -4.01884556e-01
5.90558946e-01 4.08813357e-02 -1.44887105e-01 8.77830625e-01
-5.18081561e-02 4.79092836e-01 1.37072682e-01 -4.53516871e-01
-1.00321245e+00 -8.36025238e-01 -4.38445449e-01 1.51397979e+00
4.62294333e-02 -9.08606350e-01 -6.68234453e-02 -3.43073159e-01
2.95633703e-01 7.45167434e-01 -9.09352958e-01 -4.82697904e-01
-3.77089471e-01 -4.42263961e-01 4.89080638e-01 7.23862171e-01
-5.37006617e-01 -7.56402612e-01 -2.58899570e-01 3.56793553e-01
3.65932733e-01 -7.93769419e-01 -4.41584408e-01 3.35981339e-01
-7.72255301e-01 -8.61422777e-01 -1.04748714e+00 -9.96783197e-01
5.62321484e-01 7.26886928e-01 1.12937689e+00 2.90485054e-01
-5.64525247e-01 9.80143070e-01 -9.02954876e-01 -4.10363793e-01
-1.00831196e-01 9.04898196e-02 2.75961161e-01 -7.20326155e-02
1.17604971e+00 -2.46551126e-01 -6.37693346e-01 1.75063573e-02
-1.53778362e+00 -1.15248656e+00 3.03041846e-01 1.07650936e+00
4.52324152e-01 -3.45484436e-01 5.12609303e-01 -7.49814928e-01
1.58234847e+00 -8.28365266e-01 -2.79331326e-01 5.21479487e-01
-9.10149038e-01 5.40876627e-01 4.02852148e-01 -7.06930399e-01
-4.28301275e-01 -6.54997170e-01 1.02069564e-01 -4.18161780e-01
6.85660541e-02 6.47924781e-01 3.93485278e-01 8.72569382e-02
7.42663920e-01 3.39821219e-01 -1.42338842e-01 -7.14940250e-01
7.74551570e-01 9.77627516e-01 -1.48866445e-01 -5.57465494e-01
5.95900893e-01 1.56288311e-01 -2.03924477e-01 -1.21803367e+00
-3.49007368e-01 -1.37672067e+00 -4.19952005e-01 2.43287295e-01
8.78136516e-01 -5.21994293e-01 -3.25871229e-01 -5.41081607e-01
-1.39927649e+00 8.93443704e-01 -5.36494076e-01 5.64315379e-01
-1.66850746e-01 5.30101299e-01 -1.81621283e-01 -8.24306786e-01
-1.57803252e-01 -9.24215913e-01 1.24426985e+00 2.15595469e-01
-5.32818615e-01 -1.37049091e+00 7.02315867e-01 -1.70070022e-01
7.51601398e-01 -3.12476248e-01 1.40969241e+00 -1.51238036e+00
-1.21563673e-01 -8.92934024e-01 -2.01471433e-01 3.53153437e-01
3.96739423e-01 -1.01673663e-01 -9.44228232e-01 -3.93647969e-01
-3.70491117e-01 -3.71353209e-01 1.16618764e+00 2.45133601e-02
1.23581362e+00 -1.54200479e-01 -6.52760386e-01 5.02034761e-02
1.50412047e+00 3.21726620e-01 5.54111183e-01 3.50613445e-01
2.55393296e-01 6.83005333e-01 5.84753811e-01 5.97885907e-01
-3.05994917e-02 6.94983304e-01 3.89227509e-01 2.23380730e-01
5.85645325e-02 -2.67920732e-01 -5.81609970e-03 9.73701239e-01
2.40267560e-01 -4.62705582e-01 -8.78777087e-01 9.99846637e-01
-1.49244404e+00 -8.50459516e-01 1.00839525e-01 2.01580000e+00
6.08211040e-01 -1.81876183e-01 -1.33738801e-01 3.38084519e-01
4.36001956e-01 6.28133297e-01 6.59821630e-02 -1.08560431e+00
1.90571532e-01 6.93437934e-01 3.20575446e-01 4.27265137e-01
-9.19826269e-01 8.35331917e-01 6.98789406e+00 9.66710269e-01
-7.32725859e-01 -8.74305740e-02 -1.25092819e-01 2.54169554e-01
-7.24018574e-01 -1.64287686e-01 -9.11293685e-01 1.43881604e-01
1.05907655e+00 -7.44430780e-01 -1.66823179e-01 1.03618217e+00
-1.38595358e-01 2.76613861e-01 -1.27755964e+00 1.04091740e+00
4.74802345e-01 -1.41752017e+00 9.09968853e-01 9.33352336e-02
4.40062106e-01 -2.54931539e-01 2.04066843e-01 5.83119869e-01
-7.31624290e-02 -1.06877470e+00 -1.51677102e-01 4.91349131e-01
4.56020027e-01 -5.73433042e-01 7.81439126e-01 4.77757901e-02
-1.09042990e+00 -6.03131168e-02 -1.05237949e+00 4.04598154e-02
-8.23258981e-02 2.55942643e-01 -9.38527107e-01 4.86153960e-01
2.31483117e-01 5.69392502e-01 -7.58923233e-01 1.05216992e+00
2.66511619e-01 4.09433171e-02 1.71375740e-02 -7.29051650e-01
6.89548910e-01 -4.93982844e-02 5.02174437e-01 1.59059525e+00
2.40858644e-01 -1.35791421e-01 6.15695119e-02 5.09003580e-01
-1.42679095e-01 5.51735461e-01 -1.12227750e+00 -7.13376880e-01
5.75104654e-01 1.05342329e+00 -3.33325624e-01 -5.14358938e-01
-4.01028007e-01 7.05108583e-01 2.44069964e-01 1.82077125e-01
-3.34706545e-01 -8.67934108e-01 1.14668000e+00 5.30497991e-02
4.09927458e-01 -3.68769616e-01 4.69272405e-01 -9.01141703e-01
-9.08504874e-02 -4.60766941e-01 5.75413525e-01 -4.14743453e-01
-1.64098382e+00 9.03005779e-01 5.36520720e-01 -1.49187529e+00
-5.53771079e-01 -1.02427125e+00 -4.23082680e-01 7.96830952e-01
-1.68527198e+00 -8.08983088e-01 -2.15937551e-02 5.31058431e-01
6.42544866e-01 -3.66612852e-01 1.46762693e+00 3.13087732e-01
-1.47231996e-01 6.84561491e-01 6.14247024e-01 1.01380132e-01
6.93070650e-01 -1.16356587e+00 2.03019008e-01 1.70075178e-01
6.60403728e-01 1.25365448e+00 6.08146548e-01 -1.78742021e-01
-1.68193376e+00 -7.75995076e-01 1.38432336e+00 -3.76013398e-01
9.82956529e-01 -1.87088266e-01 -1.15538740e+00 4.67069238e-01
4.37603414e-01 8.04517940e-02 1.20152140e+00 1.45478189e-01
-6.64655685e-01 -1.19317807e-02 -7.48201966e-01 5.21761000e-01
4.62536186e-01 -7.69542515e-01 -1.32081378e+00 5.88843286e-01
1.07796192e+00 3.98630559e-01 -8.41579378e-01 7.25238472e-02
4.18393165e-01 -4.72557008e-01 1.40383363e+00 -1.17491019e+00
3.40610184e-02 7.47821704e-02 -4.87065107e-01 -1.35916376e+00
-4.63053316e-01 -1.48706838e-01 -1.38782859e-01 1.25056326e+00
1.60112351e-01 -8.37920368e-01 2.66171843e-01 2.92005360e-01
1.74170613e-01 -7.81446695e-01 -8.49160850e-01 -1.24515975e+00
4.52242523e-01 -3.31558555e-01 4.61270630e-01 8.41188252e-01
2.13137046e-01 3.34976077e-01 7.57368878e-02 -2.97133714e-01
1.48545712e-01 -6.64635822e-02 4.73329961e-01 -1.45259750e+00
2.16117859e-01 -6.40285492e-01 -9.78292167e-01 -1.42230225e+00
3.51032764e-01 -1.08371782e+00 -3.71022403e-01 -1.64846122e+00
-6.75623864e-02 -3.43907058e-01 -7.01139927e-01 -1.73481591e-02
-1.08970791e-01 3.29978243e-02 -1.13207325e-01 1.50004789e-01
-4.66638952e-01 8.31042051e-01 7.42593646e-01 -5.42651117e-01
1.70961931e-01 -4.27205801e-01 -7.53915668e-01 3.61648619e-01
4.75121319e-01 -6.59398437e-01 -5.10158539e-01 -6.59468412e-01
2.30122969e-01 -5.34213245e-01 -1.00208752e-01 -5.14579713e-01
2.84135163e-01 -1.05251819e-01 4.24501039e-02 -5.10865808e-01
7.40887403e-01 -1.14653337e+00 -6.05370522e-01 3.33240598e-01
-7.60422468e-01 5.14088631e-01 -9.20862542e-04 7.84273863e-01
-7.20800579e-01 -7.47211933e-01 2.93639719e-01 -9.44062695e-02
-8.89116228e-01 2.88659811e-01 -5.68111658e-01 1.89102292e-01
8.95351827e-01 -1.29300088e-01 -1.39932588e-01 -2.95792997e-01
-2.31252506e-01 1.51353851e-01 1.04537876e-02 9.52250481e-01
8.98322940e-01 -1.60070956e+00 -5.05484402e-01 2.51289029e-02
9.35623825e-01 -5.32388270e-01 -4.47182328e-01 1.79512680e-01
-4.96064544e-01 1.14045215e+00 -4.49556187e-02 -3.00478965e-01
-1.12563300e+00 1.06605875e+00 -1.84602439e-01 -4.07325447e-01
-3.19023073e-01 8.07400286e-01 7.61634037e-02 5.42598702e-02
4.61299926e-01 -4.21244472e-01 -9.33660805e-01 8.06896567e-01
8.94336879e-01 3.38952005e-01 1.53803080e-01 -6.45338356e-01
-5.60167372e-01 7.21563935e-01 -5.67476094e-01 2.95252185e-02
1.43614542e+00 -1.14158159e-02 -2.67461777e-01 5.25258541e-01
1.93721116e+00 -1.97322994e-01 1.03634037e-01 -5.62231004e-01
6.07370257e-01 -8.22463512e-01 1.85667008e-01 -7.50660896e-02
-6.02331221e-01 1.14986932e+00 6.58814967e-01 8.26756358e-01
7.98644543e-01 2.60819852e-01 7.46575713e-01 8.46974254e-01
1.62845582e-01 -9.63165283e-01 3.51100087e-01 6.02522075e-01
1.06557679e+00 -1.07787240e+00 8.28756616e-02 7.74548575e-02
-1.45153254e-01 1.51767278e+00 2.45898232e-01 -5.49106598e-01
8.65688443e-01 -2.93090045e-01 -9.74670500e-02 -3.99136186e-01
-7.63305068e-01 -6.18546426e-01 7.37930894e-01 6.32058382e-01
6.43444061e-01 -2.37095967e-01 -1.07630789e+00 4.50347036e-01
2.69304693e-01 -5.73674977e-01 2.97387302e-01 1.19555616e+00
-7.20418513e-01 -1.60674095e+00 -3.43364477e-01 4.12716717e-01
-4.18784231e-01 -3.03324908e-01 -5.71710408e-01 7.10921705e-01
-2.81793058e-01 8.64887595e-01 2.51672357e-01 -3.52842420e-01
3.80672246e-01 3.64332408e-01 2.24742427e-01 -9.20650244e-01
-5.32207012e-01 -2.88975865e-01 -1.46874771e-01 -3.83216321e-01
-2.98429996e-01 -8.91329534e-03 -8.41073215e-01 4.10988331e-02
-5.04676878e-01 6.04641557e-01 7.47433484e-01 7.25525320e-01
4.56454724e-01 4.14210528e-01 7.89392054e-01 -7.65483141e-01
-8.43580186e-01 -1.13959825e+00 -5.56204855e-01 6.89641893e-01
3.26620787e-01 -9.37891483e-01 -4.30558026e-01 -3.28295022e-01] | [10.609394073486328, 8.505160331726074] |
75404176-0560-443d-a73b-2ea9b77e015b | automatic-personality-prediction-an-enhanced | 2007.04571 | null | https://arxiv.org/abs/2007.04571v3 | https://arxiv.org/pdf/2007.04571v3.pdf | Automatic Personality Prediction; an Enhanced Method Using Ensemble Modeling | Human personality is significantly represented by those words which he/she uses in his/her speech or writing. As a consequence of spreading the information infrastructures (specifically the Internet and social media), human communications have reformed notably from face to face communication. Generally, Automatic Personality Prediction (or Perception) (APP) is the automated forecasting of the personality on different types of human generated/exchanged contents (like text, speech, image, video, etc.). The major objective of this study is to enhance the accuracy of APP from the text. To this end, we suggest five new APP methods including term frequency vector-based, ontology-based, enriched ontology-based, latent semantic analysis (LSA)-based, and deep learning-based (BiLSTM) methods. These methods as the base ones, contribute to each other to enhance the APP accuracy through ensemble modeling (stacking) based on a hierarchical attention network (HAN) as the meta-model. The results show that ensemble modeling enhances the accuracy of APP. | ['Taymaz Rahkar-Farshi', 'Elnaz Zafarani-Moattar', 'Zoleikha Jahanbakhsh-Nagadeh', 'Mehrdad Ranjbar-Khadivi', 'Narjes Nikzad-Khasmakhi', 'Ali-Reza Feizi-Derakhshi', 'Meysam Asgari-Chenaghlu', 'Mohammad-Ali Balafar', 'Mohammad-Reza Feizi-Derakhshi', 'Majid Ramezani'] | 2020-07-09 | null | null | null | null | ['personality-trait-recognition'] | ['computer-vision'] | [-8.19532350e-02 3.74186337e-02 1.57399207e-01 -2.58947909e-01
7.57917613e-02 2.74145812e-01 8.39960217e-01 8.95268172e-02
-2.11148351e-01 7.80037045e-01 6.76533997e-01 3.33895773e-01
-3.31693888e-01 -7.83114791e-01 -2.14806735e-03 -6.35266900e-01
4.14758474e-02 2.64949709e-01 -1.13009505e-01 -4.25636858e-01
5.80766439e-01 -5.60432337e-02 -1.98015356e+00 5.01031041e-01
1.04073155e+00 1.21081340e+00 4.46796149e-01 3.73855174e-01
-5.30622423e-01 1.20114803e+00 -3.64441425e-01 -8.12046587e-01
-7.15632856e-01 -2.60029078e-01 -9.21592891e-01 1.64996728e-01
-2.98391700e-01 -1.85862184e-01 -3.91786218e-01 1.19529068e+00
6.16347730e-01 2.67414272e-01 7.86340833e-01 -1.17170179e+00
-1.20444798e+00 5.90701640e-01 -3.40199351e-01 -2.20820993e-01
5.69105685e-01 -2.49443084e-01 4.36484516e-01 -9.33351696e-01
3.40791702e-01 1.47494400e+00 7.55695283e-01 3.86185288e-01
-5.74664891e-01 -3.45055372e-01 -1.13162644e-01 7.20254779e-01
-1.27090526e+00 -3.14236760e-01 9.14592743e-01 -6.13275111e-01
7.37181246e-01 3.84704694e-02 6.27815127e-01 1.62425196e+00
5.27471125e-01 5.56111336e-01 1.40258849e+00 -4.20597911e-01
-2.60030836e-01 7.32608974e-01 1.82561830e-01 7.43933499e-01
-1.39702499e-01 -1.93959981e-01 -7.92054713e-01 -8.99800658e-02
2.75878459e-01 6.69650137e-02 1.25306070e-01 4.64894325e-01
-8.95927072e-01 6.06543541e-01 -3.15711536e-02 3.99885714e-01
-7.28320360e-01 -3.93521816e-01 5.98280847e-01 5.79262823e-02
6.69247806e-01 2.22672671e-01 -2.10376754e-01 -4.68268305e-01
-6.89563274e-01 1.47592857e-01 9.02200282e-01 8.87462020e-01
6.99461579e-01 8.26716870e-02 -3.81872714e-01 1.22582638e+00
4.09932137e-01 3.67269844e-01 1.18033648e+00 -7.24084616e-01
2.17792224e-02 7.62640297e-01 -5.60576841e-02 -1.64194846e+00
-3.63425940e-01 -5.88981509e-01 -1.06694448e+00 -4.82905656e-01
-3.18556547e-01 -1.56901792e-01 -3.25880438e-01 1.36466181e+00
6.66361749e-02 8.92589763e-02 1.05509497e-01 4.88305479e-01
1.15221059e+00 8.00716817e-01 4.63348240e-01 -4.66248602e-01
1.66235375e+00 -1.04345286e+00 -1.07658637e+00 3.10484841e-02
3.22288722e-01 -7.48989999e-01 8.81021500e-01 5.42861521e-01
-8.93264234e-01 -9.98987973e-01 -8.41114700e-01 -3.35265249e-02
-7.80763507e-01 3.30045849e-01 4.06197369e-01 5.43842077e-01
-9.54635859e-01 7.62024224e-01 -6.49160743e-02 -6.55201614e-01
3.29549879e-01 2.06290603e-01 -3.23493779e-01 2.79933333e-01
-1.58122146e+00 1.01459038e+00 5.49748361e-01 -5.55723719e-02
-4.35590059e-01 -2.04235688e-01 -6.13347054e-01 2.64939576e-01
-6.73164381e-03 -8.24379861e-01 9.44855392e-01 -1.32985377e+00
-1.74668145e+00 9.11791325e-01 -2.87563920e-01 -3.48968208e-01
4.31936160e-02 -1.87904730e-01 -9.42355931e-01 -2.31221225e-02
-8.48082080e-03 5.27398944e-01 8.81963611e-01 -1.21703184e+00
-6.51629150e-01 -6.17236674e-01 -3.56529355e-01 8.96649137e-02
-1.11987484e+00 3.04119825e-01 -2.19149828e-01 -4.11467046e-01
-2.40417421e-01 -7.18981922e-01 2.36100823e-01 -8.29061508e-01
-3.74593437e-01 -7.25794077e-01 9.94470000e-01 -1.29610455e+00
1.62652588e+00 -2.08791518e+00 2.69236624e-01 -5.70004731e-02
2.65328079e-01 1.83959708e-01 2.34796047e-01 6.61507130e-01
3.29414696e-01 1.48909956e-01 -2.57474650e-02 -6.77116632e-01
2.22401932e-01 1.92165941e-01 -2.41666898e-01 -2.08392814e-01
-1.18652180e-01 7.62939036e-01 -7.64306903e-01 -8.94554257e-01
3.03637367e-02 4.64477092e-01 -8.30506682e-02 1.96194813e-01
3.01720686e-02 5.83998442e-01 -4.31500673e-01 3.72566551e-01
2.88914949e-01 -2.65336663e-01 2.48906896e-01 -1.67924821e-01
-3.41391295e-01 2.49759555e-02 -5.20864189e-01 1.50646794e+00
-5.36577821e-01 3.76208365e-01 -2.53598392e-01 -8.35422516e-01
1.26254106e+00 5.29703498e-01 5.95610321e-01 -5.42113602e-01
3.41775090e-01 5.87366857e-02 -6.43106252e-02 -1.14555705e+00
7.36691654e-01 7.88817108e-02 -8.04367065e-02 1.40422925e-01
2.52985775e-01 6.64069355e-01 -7.86934197e-02 8.25223327e-02
5.57934225e-01 1.23713247e-01 4.52400535e-01 -2.45155364e-01
1.06246412e+00 -3.79820824e-01 5.31891704e-01 2.26989567e-01
-2.22842395e-01 -1.10949809e-03 4.39588070e-01 -2.24447027e-01
-1.06702423e+00 -5.00545084e-01 -2.08349764e-01 1.12338388e+00
-3.45015563e-02 -4.21367913e-01 -9.15796399e-01 -3.13759059e-01
-1.10892273e-01 7.29744196e-01 -5.23817122e-01 -3.37649107e-01
5.31313336e-03 -6.90434456e-01 4.04969066e-01 3.52045625e-01
1.15206885e+00 -1.57787037e+00 9.68997777e-02 4.06423926e-01
-4.67540801e-01 -1.10026300e+00 -2.52177864e-01 -4.39312041e-01
-6.15871072e-01 -3.75845939e-01 -6.75770938e-01 -6.85245991e-01
3.22387695e-01 -1.20429255e-01 8.02303314e-01 1.99234243e-02
2.65686303e-01 5.13955235e-01 -6.35906100e-01 -5.39379656e-01
-2.96799511e-01 8.10307823e-03 5.40979922e-01 5.87101936e-01
7.65364110e-01 -7.98157036e-01 -2.79064149e-01 -2.11553071e-02
-4.40870136e-01 7.55151957e-02 5.00167489e-01 7.71483541e-01
9.92102399e-02 5.36453307e-01 9.01625693e-01 -6.85494184e-01
1.05112481e+00 -1.00875902e+00 2.01658428e-01 1.85886309e-01
-8.68630588e-01 -3.57568920e-01 4.49386299e-01 -3.82694572e-01
-1.58468390e+00 -4.18873489e-01 -2.24304721e-01 -2.95022041e-01
-2.91968703e-01 6.80899799e-01 -2.67635763e-01 2.20996544e-01
2.45603591e-01 7.39406407e-01 2.50100583e-01 -5.11720657e-01
-8.39289371e-03 1.46016622e+00 4.52105016e-01 -5.49930871e-01
1.37114763e-01 1.42662823e-01 -1.96918175e-01 -1.07805347e+00
-8.67152095e-01 -3.19261342e-01 -4.99024838e-01 -6.82975411e-01
1.20385134e+00 -6.42647147e-01 -1.23201454e+00 6.93673193e-01
-1.31561697e+00 1.73850715e-01 3.71063977e-01 4.59416389e-01
-5.32277524e-01 6.55686975e-01 -6.79299057e-01 -1.22878766e+00
-4.90806699e-01 -9.67097461e-01 9.02619958e-01 3.49038005e-01
-5.58591962e-01 -1.16588652e+00 -2.22246662e-01 6.74852431e-01
3.86287272e-01 -1.72396153e-01 1.01526189e+00 -6.76448464e-01
-6.05531549e-03 2.85028648e-02 -3.06113034e-01 5.83895981e-01
1.24640092e-02 -6.07720017e-02 -1.12104285e+00 3.16905379e-01
1.80381998e-01 -6.04892485e-02 7.95784056e-01 3.86181533e-01
1.63250291e+00 -6.95546150e-01 -1.74367607e-01 1.40896559e-01
1.12166989e+00 4.11135763e-01 9.89339352e-01 2.40675479e-01
6.58993185e-01 1.23621726e+00 3.16910267e-01 7.86749363e-01
8.27451050e-01 6.28587067e-01 1.13531962e-01 5.11202097e-01
7.19125271e-02 -4.65477884e-01 5.85265577e-01 1.40099156e+00
-5.69724679e-01 -2.51918733e-01 -7.96726108e-01 2.77514189e-01
-2.02313137e+00 -1.37976074e+00 -2.89394647e-01 1.90574932e+00
5.32802403e-01 -9.32153687e-02 2.52245873e-01 1.54548481e-01
8.23964000e-01 -5.33276238e-02 -2.36493930e-01 -5.82175493e-01
-7.11357072e-02 -3.08668762e-01 1.06343804e-02 2.25948542e-01
-1.09459662e+00 8.23521256e-01 5.03339767e+00 9.24397469e-01
-9.49017823e-01 4.48835075e-01 7.25841403e-01 4.61842120e-01
-2.14241505e-01 -5.51373541e-01 -8.84017885e-01 1.07373250e+00
1.13462687e+00 -3.54783803e-01 5.11774302e-01 8.87927949e-01
3.90430063e-01 1.50279552e-01 -4.67978358e-01 1.32430518e+00
4.49574649e-01 -1.05803216e+00 1.37913615e-01 3.29402596e-01
5.84600389e-01 -6.00924969e-01 2.59878039e-01 8.24612975e-01
-1.58978745e-01 -7.66944528e-01 5.29788971e-01 1.22992861e+00
5.60435891e-01 -7.23759532e-01 1.06288660e+00 4.55252469e-01
-9.24978435e-01 -4.84522879e-01 -3.03389221e-01 -2.32459396e-01
1.18515864e-01 5.77339411e-01 -3.68520319e-01 7.00579584e-01
8.89882863e-01 8.89323831e-01 -5.01138687e-01 4.28316653e-01
-6.71443790e-02 4.83886421e-01 4.55940187e-01 -3.44181508e-01
-4.94897589e-02 -2.79782534e-01 4.96624053e-01 1.07689595e+00
6.49573624e-01 2.73292661e-01 7.24685863e-02 8.56994331e-01
-2.25796569e-02 4.63263035e-01 -7.17825413e-01 -3.90500128e-01
4.04209852e-01 1.49740648e+00 -3.16279501e-01 -5.47239542e-01
-4.95172113e-01 1.14478421e+00 2.27655008e-01 2.03499362e-01
-6.62347794e-01 -2.28872627e-01 4.20894504e-01 4.76172529e-02
-2.82192707e-01 -5.37901074e-02 -3.46172839e-01 -8.63598824e-01
-8.85690227e-02 -6.46781266e-01 1.57095209e-01 -1.05808747e+00
-1.71981025e+00 6.81189775e-01 -2.40773022e-01 -1.01707649e+00
2.79559027e-02 -4.93764251e-01 -4.61981654e-01 8.23562622e-01
-1.03983021e+00 -1.13666081e+00 -4.90820318e-01 3.79928738e-01
6.09119534e-01 -6.98130906e-01 9.41023231e-01 5.86775839e-01
-6.29922926e-01 2.70342857e-01 -1.06098287e-01 -7.21312943e-04
5.53720653e-01 -1.04338455e+00 -3.36780161e-01 2.76772499e-01
-3.94376904e-01 5.84096730e-01 6.07887089e-01 -9.23386455e-01
-1.06113291e+00 -9.95678425e-01 1.46165049e+00 -3.80961984e-01
5.95932126e-01 7.60503560e-02 -7.90894210e-01 5.56984186e-01
4.93580729e-01 -9.66367424e-01 7.43734002e-01 2.83194929e-01
9.05074179e-02 -2.98465550e-01 -1.08630955e+00 5.54638624e-01
9.73881841e-01 -5.61075568e-01 -5.80921292e-01 2.87654757e-01
7.75664628e-01 2.62693554e-01 -1.25626898e+00 3.23870182e-01
8.11373353e-01 -1.01494658e+00 8.65227878e-01 -5.24382234e-01
1.01411915e+00 1.14089854e-01 -2.72189826e-01 -1.29350984e+00
-7.56558597e-01 -1.90721333e-01 -3.91199291e-01 1.70673227e+00
-3.57380323e-02 -7.78409660e-01 3.74419361e-01 5.67427993e-01
-3.91859919e-01 -7.55609214e-01 -5.63592374e-01 -5.77010691e-01
-3.67689341e-01 -3.08230549e-01 8.12206268e-01 1.05103731e+00
5.32135963e-01 4.86104876e-01 -8.56455028e-01 -9.57645997e-02
5.03831744e-01 -4.31691796e-01 4.35103774e-01 -1.61616111e+00
-7.45376423e-02 -6.78971589e-01 -4.77512598e-01 -5.79422057e-01
4.33441043e-01 -7.03678012e-01 -4.46334690e-01 -1.45353401e+00
4.88098115e-01 -4.84792143e-02 -3.00937831e-01 1.97994918e-01
-1.43027589e-01 -1.27357543e-01 -3.37414667e-02 3.43661189e-01
-6.50904715e-01 1.03424561e+00 1.33203268e+00 -1.59734204e-01
2.83571333e-02 -9.22820270e-02 -7.74299979e-01 8.48994315e-01
9.35491204e-01 -1.39178216e-01 -2.42066294e-01 5.89824803e-02
3.51511568e-01 2.67725378e-01 4.99681383e-01 -1.04729021e+00
2.57932663e-01 1.14697896e-01 3.32608491e-01 -4.19698656e-01
5.46468735e-01 -6.52098119e-01 -6.69607669e-02 3.46602350e-01
-3.30207258e-01 7.31321350e-02 -7.94732720e-02 5.25277495e-01
-3.77902865e-01 -3.14602226e-01 4.76111025e-01 -2.05929428e-01
-9.19086397e-01 3.91514510e-01 -6.05289757e-01 -5.06513953e-01
8.55341494e-01 -2.34264284e-01 -3.57550383e-01 -6.43423498e-01
-1.07810140e+00 -2.25219224e-03 -1.50757805e-01 5.79064846e-01
5.68777204e-01 -1.46451640e+00 -6.29230440e-01 2.32233796e-02
6.27021790e-02 -8.92982066e-01 9.76495743e-01 1.16687453e+00
7.17446432e-02 6.85002029e-01 -4.51113224e-01 -7.46543333e-02
-1.31760716e+00 5.23469508e-01 1.17211275e-01 -2.82062024e-01
-2.28876337e-01 6.80532992e-01 2.95600127e-02 -4.65276450e-01
2.17920437e-01 5.01349270e-01 -1.20292735e+00 2.25476786e-01
5.57000399e-01 8.39066386e-01 -1.93765134e-01 -1.02762771e+00
5.03414832e-02 2.49544188e-01 9.96535793e-02 -6.86049089e-02
1.11099350e+00 -3.86684358e-01 -7.06731141e-01 7.56555080e-01
1.02300763e+00 -1.39304996e-01 -5.47396064e-01 -2.36950472e-01
1.73046743e-03 -1.29481196e-01 8.66304860e-02 -7.26266623e-01
-5.83715558e-01 9.53300118e-01 5.13124526e-01 4.92960870e-01
9.90269601e-01 -1.97152838e-01 9.99918818e-01 2.42125362e-01
5.40530086e-01 -1.62316835e+00 1.48513854e-01 5.25264144e-01
9.31647718e-01 -1.12553310e+00 -1.53104320e-01 -1.44706398e-01
-1.07882011e+00 1.11659384e+00 6.38221920e-01 3.62509161e-01
8.67900193e-01 -3.46453965e-01 -3.93326879e-01 -1.97763786e-01
-8.26841176e-01 -1.46540970e-01 4.26093012e-01 5.40110409e-01
6.00601912e-01 2.85052747e-01 -5.85035861e-01 1.36328983e+00
-4.75575775e-01 1.02110870e-01 2.31180385e-01 3.05035502e-01
-6.99715734e-01 -1.09315026e+00 -3.46326292e-01 6.86273038e-01
-4.57127661e-01 -1.80834159e-01 -3.27539861e-01 8.84773359e-02
5.99415600e-01 1.33333802e+00 -4.99501713e-02 -9.62044537e-01
-7.89130777e-02 6.59230232e-01 3.44764441e-01 -4.27740604e-01
-4.62314636e-01 -3.95187944e-01 3.68464440e-01 -7.64245316e-02
-5.74564576e-01 -7.85404384e-01 -9.79740441e-01 -6.24628901e-01
-9.99819953e-03 1.12924993e-01 6.38639450e-01 1.26373208e+00
5.40845454e-01 8.56671929e-01 7.05281079e-01 -7.49684274e-01
-2.57172793e-01 -1.46395063e+00 -7.88513422e-01 3.90920848e-01
-4.79842782e-01 -8.99069965e-01 -1.37495488e-01 2.70833284e-01] | [12.992971420288086, 5.909172058105469] |
8ab89246-1c91-4339-b7b1-18da02c868f4 | search-based-task-and-motion-planning-for | 2301.10384 | null | https://arxiv.org/abs/2301.10384v1 | https://arxiv.org/pdf/2301.10384v1.pdf | Search-Based Task and Motion Planning for Hybrid Systems: Agile Autonomous Vehicles | To achieve optimal robot behavior in dynamic scenarios we need to consider complex dynamics in a predictive manner. In the vehicle dynamics community, it is well know that to achieve time-optimal driving on low surface, the vehicle should utilize drifting. Hence many authors have devised rules to split circuits and employ drifting on some segments. These rules are suboptimal and do not generalize to arbitrary circuit shapes (e.g., S-like curves). So, the question "When to go into which mode and how to drive in it?" remains unanswered. To choose the suitable mode (discrete decision), the algorithm needs information about the feasibility of the continuous motion in that mode. This makes it a class of Task and Motion Planning (TAMP) problems, which are known to be hard to solve optimally in real-time. In the AI planning community, search methods are commonly used. However, they cannot be directly applied to TAMP problems due to the continuous component. Here, we present a search-based method that effectively solves this problem and efficiently searches in a highly dimensional state space with nonlinear and unstable dynamics. The space of the possible trajectories is explored by sampling different combinations of motion primitives guided by the search. Our approach allows to use multiple locally approximated models to generate motion primitives (e.g., learned models of drifting) and effectively simplify the problem without losing accuracy. The algorithm performance is evaluated in simulated driving on a mixed-track with segments of different curvatures (right and left). Our code is available at https://git.io/JenvB | ['Antonella Ferrara', 'Martin Horn', 'Hana Ćatić', 'Barys Shyrokau', 'Enrico Regolin', 'Zlatan Ajanović'] | 2023-01-25 | null | null | null | null | ['community-search', 'motion-planning'] | ['graphs', 'robots'] | [-9.18485224e-03 -6.01364952e-03 -5.57783306e-01 3.81107703e-02
-2.73850530e-01 -8.62157226e-01 5.42534888e-01 -2.80448496e-01
-2.00552255e-01 7.07782507e-01 -4.54633623e-01 -7.82372773e-01
-3.70634377e-01 -9.09609020e-01 -7.69897938e-01 -8.17367554e-01
6.03300817e-02 7.74054945e-01 5.99121928e-01 -8.64270866e-01
4.73786533e-01 7.74804175e-01 -1.67317724e+00 -1.96613029e-01
1.16184902e+00 4.58318591e-01 5.13543963e-01 5.89548469e-01
-3.38951685e-02 1.58061728e-01 -3.14254016e-02 3.19414809e-02
3.07508022e-01 -4.48043436e-01 -7.88778305e-01 -4.15531471e-02
-1.48678154e-01 -1.04287095e-01 -3.27479690e-01 9.33047533e-01
6.58096373e-02 2.80744374e-01 8.15066218e-01 -1.39062989e+00
2.66619533e-01 1.91221029e-01 -2.32485965e-01 -2.91374493e-02
1.77870870e-01 4.51680243e-01 5.56064188e-01 -6.43478215e-01
8.18662643e-01 1.16038597e+00 4.90665793e-01 6.70028210e-01
-1.30897069e+00 -5.12200713e-01 9.22870040e-02 4.43609387e-01
-1.44187379e+00 -3.49527240e-01 8.71642888e-01 -5.62855601e-01
7.49596834e-01 2.38756314e-01 9.09570634e-01 8.68945777e-01
7.04490840e-01 6.38444006e-01 7.73706853e-01 -1.33887693e-01
4.34580684e-01 1.18153140e-01 -9.95158106e-02 6.97960556e-01
3.92405242e-01 3.41966003e-01 -1.58550188e-01 -1.61314979e-02
5.77965021e-01 -2.94442564e-01 -1.11426882e-01 -8.54229927e-01
-9.69827414e-01 9.95974898e-01 2.66566932e-01 3.09364378e-01
-1.79186359e-01 2.83017337e-01 7.20696896e-02 1.82065010e-01
-2.04455882e-01 5.62353015e-01 -2.76497185e-01 -2.81199634e-01
-6.01644456e-01 1.01137257e+00 9.07324255e-01 1.02433527e+00
9.95024323e-01 -8.66839811e-02 2.71582842e-01 5.36907494e-01
1.76199868e-01 5.56183875e-01 -8.26994255e-02 -1.24997771e+00
4.39499617e-01 6.07259810e-01 5.41280031e-01 -9.89754975e-01
-7.05416203e-01 -1.82824172e-02 -7.00349450e-01 5.48362136e-01
6.11136317e-01 -4.12290484e-01 -8.23331177e-01 1.54615664e+00
4.30787504e-01 2.53760274e-02 5.74981011e-02 9.21652198e-01
1.73669476e-02 8.21213901e-01 -5.07133126e-01 -2.63582021e-01
9.09166217e-01 -7.62777865e-01 -7.66324162e-01 -4.40890014e-01
8.65525424e-01 -3.93430382e-01 9.88616168e-01 3.42159718e-01
-9.87647176e-01 -3.44800472e-01 -1.27227008e+00 2.89192766e-01
-3.73932451e-01 -1.70466378e-01 3.38664055e-01 2.93075770e-01
-9.97558773e-01 7.79734075e-01 -1.07613456e+00 -3.86595309e-01
1.05696209e-01 4.61808205e-01 -4.73841131e-02 -8.53545293e-02
-1.22670424e+00 1.33684862e+00 3.07964236e-01 2.36206010e-01
-8.40879440e-01 -2.27103114e-01 -6.51212990e-01 -2.98437476e-01
6.81242526e-01 -5.73734999e-01 1.30568242e+00 -7.36386955e-01
-1.78525269e+00 2.61401922e-01 -5.06861448e-01 -3.99042457e-01
8.63744497e-01 1.34735137e-01 2.39621419e-02 -1.20326877e-01
-1.91886816e-02 6.78498805e-01 6.35362625e-01 -1.28957283e+00
-5.42927325e-01 -4.88546155e-02 1.69892371e-01 3.92260075e-01
1.19299136e-01 -4.96143132e-01 -5.49347699e-01 -4.99405004e-02
2.14642122e-01 -1.62953329e+00 -7.18451679e-01 -1.58670582e-02
-2.35541344e-01 -2.38282725e-01 9.05745327e-01 -6.73907921e-02
1.25273347e+00 -1.83844566e+00 4.80920732e-01 4.09302086e-01
-3.04897398e-01 1.11101516e-01 2.07745671e-01 6.96136832e-01
5.72506785e-01 1.15886450e-01 -5.30013144e-01 1.72416657e-01
-1.66555807e-01 6.80755258e-01 -3.09697986e-01 5.22498548e-01
7.71505907e-02 7.61050940e-01 -1.02805221e+00 -3.62978309e-01
1.63552687e-01 2.00756907e-01 -7.50086248e-01 -1.77864522e-01
-3.63640130e-01 6.01008713e-01 -9.72774744e-01 3.74452144e-01
5.68432450e-01 2.18569174e-01 1.67956352e-01 2.42035434e-01
-6.92577600e-01 1.69048198e-02 -1.36383140e+00 1.34327435e+00
-4.00368482e-01 7.22612202e-01 3.41049194e-01 -1.01875734e+00
1.08410859e+00 -8.94355252e-02 4.80300903e-01 -4.10360008e-01
1.19367838e-01 5.57002962e-01 1.28310725e-01 -7.74595916e-01
6.59352303e-01 -8.47396106e-02 -1.84784219e-01 -6.02414906e-02
-6.49817348e-01 -8.13652873e-01 2.71174788e-01 -2.63490289e-01
1.20028841e+00 1.94495931e-01 -4.95104752e-02 -6.54541373e-01
4.77711320e-01 8.38881433e-01 6.69771492e-01 5.41465104e-01
-1.69019088e-01 2.41243586e-01 5.69989145e-01 -3.95207793e-01
-1.07000780e+00 -6.92420661e-01 -2.11122781e-01 3.31469864e-01
1.05076981e+00 -2.44095683e-01 -8.36285889e-01 -9.93045345e-02
1.01493239e-01 9.03397441e-01 -4.02213663e-01 -2.54950970e-01
-1.07757723e+00 -4.11659181e-01 1.96545810e-01 1.89401716e-01
3.41790378e-01 -7.82111466e-01 -1.10716581e+00 4.30075407e-01
-1.63926169e-01 -8.11631024e-01 -2.36697912e-01 1.34549782e-01
-9.48537707e-01 -1.07448602e+00 -3.48970801e-01 -6.99067116e-01
5.89260280e-01 3.61704051e-01 5.90250254e-01 7.97454789e-02
-1.27571076e-01 6.72559217e-02 -6.76092282e-02 -2.53132790e-01
-6.46263242e-01 2.95726269e-01 1.02662724e-02 -1.48177221e-01
-2.85084210e-02 -2.79389620e-01 -5.05230725e-01 9.57911909e-01
-5.72635055e-01 3.40693086e-01 4.34192598e-01 7.12791979e-01
6.69612169e-01 2.58368462e-01 4.42391872e-01 -3.92022908e-01
5.14591157e-01 -5.12833595e-01 -9.05053437e-01 -1.27166137e-01
-5.61848819e-01 3.04282516e-01 6.77784443e-01 -6.56735182e-01
-8.11472654e-01 5.05380154e-01 -3.01213469e-02 -4.13194090e-01
-1.61924779e-01 3.33121538e-01 -5.30453101e-02 -1.21454865e-01
5.61536908e-01 1.73234954e-01 3.13870668e-01 -8.46576095e-02
2.97992021e-01 4.05174166e-01 2.37273797e-01 -5.94140589e-01
7.88358152e-01 5.74985921e-01 3.73338968e-01 -9.59219396e-01
-2.45445743e-02 -2.97853380e-01 -6.13068879e-01 -6.15034640e-01
8.79881680e-01 -2.97551543e-01 -8.54155362e-01 3.10428500e-01
-1.14001620e+00 -7.74810433e-01 -9.05007310e-03 2.07658231e-01
-9.74343359e-01 -1.39976358e-02 -8.54569897e-02 -1.01208079e+00
3.28437179e-01 -1.51557076e+00 7.54601538e-01 3.08573246e-01
-2.75394022e-01 -6.69142067e-01 7.00517893e-02 -1.23541124e-01
5.02295554e-01 3.74933362e-01 8.20335925e-01 3.48625295e-02
-8.34237635e-01 -1.55755892e-01 4.22392190e-01 -3.28125656e-01
-1.24591641e-01 2.68777609e-01 -3.73358667e-01 -1.56579271e-01
-1.46414042e-01 2.40354314e-01 6.96305811e-01 4.98039663e-01
7.59787440e-01 -3.44677240e-01 -1.11652899e+00 3.53376150e-01
1.31667101e+00 6.53877616e-01 6.73266292e-01 4.63562965e-01
4.40460503e-01 1.06996500e+00 8.77124846e-01 1.23754255e-01
5.77175796e-01 1.06653857e+00 5.17329454e-01 3.82117510e-01
3.44433933e-01 -2.54186124e-01 5.32701492e-01 4.25533980e-01
-5.89097850e-02 -1.98618650e-01 -1.30973697e+00 7.09621906e-01
-2.04722857e+00 -9.91190851e-01 -3.49449813e-01 2.01243186e+00
5.08745193e-01 3.02561939e-01 1.29856512e-01 8.91145542e-02
7.11186767e-01 -2.01060176e-01 -7.87232876e-01 -6.29574955e-01
1.30765945e-01 -4.65821356e-01 8.61845613e-01 7.46530116e-01
-8.08692515e-01 9.83020067e-01 5.97549915e+00 7.99594641e-01
-1.19818246e+00 -1.97290272e-01 2.91372657e-01 2.21524332e-02
-4.72287804e-01 4.14765149e-01 -1.11529744e+00 4.53587890e-01
9.20671523e-01 -3.81304204e-01 6.04585528e-01 7.49362707e-01
6.78307354e-01 -3.12067777e-01 -9.09309924e-01 6.56051636e-01
-4.49227005e-01 -1.32184458e+00 -3.31153780e-01 6.41621500e-02
7.55260944e-01 -3.46221179e-02 -1.35257319e-01 1.92367405e-01
1.40057608e-01 -9.22015429e-01 9.52308238e-01 4.71388668e-01
4.07945722e-01 -7.78220356e-01 3.30526322e-01 9.46594357e-01
-1.24752510e+00 -3.90255481e-01 -2.35473439e-01 -3.63729112e-02
6.47185385e-01 1.21272340e-01 -8.08445334e-01 4.10879672e-01
4.96838868e-01 5.09008288e-01 -1.02377802e-01 1.06577647e+00
1.48473218e-01 3.43399465e-01 -7.20370650e-01 -5.84230483e-01
4.47988659e-01 -6.02091312e-01 8.98659825e-01 8.24282527e-01
5.92546940e-01 1.45300061e-01 1.75990060e-01 9.98774469e-01
6.56154633e-01 -1.71862081e-01 -1.12112105e+00 2.42411450e-01
4.56555188e-01 9.24971879e-01 -8.40859711e-01 6.79832399e-02
-7.56758004e-02 3.62701178e-01 -2.40386110e-02 4.65747654e-01
-9.27851677e-01 -3.90636742e-01 8.80761385e-01 7.38533616e-01
2.51625508e-01 -7.79241145e-01 -4.33423072e-01 -6.93499506e-01
-6.65476322e-02 -4.00264204e-01 -1.08557969e-01 -4.21491385e-01
-4.89442527e-01 4.96581078e-01 5.60692489e-01 -1.42359030e+00
-6.25451326e-01 -4.28255647e-01 -6.55433416e-01 5.79632580e-01
-1.40586805e+00 -6.13916934e-01 -2.03804210e-01 3.86177331e-01
7.48725176e-01 2.15077654e-01 2.39366397e-01 1.20617174e-01
-5.04168749e-01 1.24998741e-01 2.07353339e-01 -5.21610796e-01
1.96170628e-01 -7.96444714e-01 2.75718689e-01 6.91855133e-01
-5.62834561e-01 2.82000750e-01 1.13672256e+00 -7.70607054e-01
-2.00681949e+00 -1.00015688e+00 6.97494507e-01 -4.33577269e-01
5.92235386e-01 -2.70664990e-01 -1.01557934e+00 3.25610012e-01
-1.48150861e-01 -2.79968888e-01 -3.92764747e-01 -5.51218987e-01
5.69744647e-01 -8.53565149e-03 -9.22959149e-01 1.01906633e+00
1.18740869e+00 9.37863961e-02 -2.02484980e-01 3.86552438e-02
4.43796188e-01 -6.17724121e-01 -2.94288933e-01 6.42058909e-01
5.64315796e-01 -6.00502610e-01 6.57809258e-01 -1.53174788e-01
-3.59540246e-03 -5.81008434e-01 9.92835760e-02 -1.25683749e+00
-2.64893584e-02 -1.01069391e+00 1.68804713e-02 6.87256336e-01
5.98931611e-01 -8.50576103e-01 8.49192202e-01 5.70301354e-01
-4.40481752e-01 -1.17886281e+00 -1.16998541e+00 -9.40375686e-01
4.17959362e-01 -4.80273694e-01 5.12805223e-01 4.90766168e-01
8.35357755e-02 7.58642480e-02 -9.58081633e-02 2.34158635e-01
3.23546380e-01 1.69441029e-01 9.60919499e-01 -1.09986508e+00
-3.72812375e-02 -7.08434403e-01 -1.21042699e-01 -1.37185669e+00
3.25253069e-01 -7.36025512e-01 5.89520693e-01 -1.60049272e+00
-4.76096392e-01 -1.04576552e+00 5.86163342e-01 1.97967738e-01
2.88048923e-01 -4.34876293e-01 1.18170105e-01 4.01293039e-01
4.89722975e-02 6.13669336e-01 1.44267642e+00 -1.59147933e-01
-6.88917518e-01 3.10086876e-01 -1.87701404e-01 7.29168952e-01
1.09457660e+00 -3.74654084e-01 -5.44814944e-01 -1.84045941e-01
4.18179244e-01 4.49747592e-01 2.20440835e-01 -1.04472113e+00
4.17103559e-01 -8.13380361e-01 -4.16385353e-01 -7.42391527e-01
4.88723606e-01 -7.33310938e-01 5.15562654e-01 8.32436144e-01
-7.58884251e-02 2.71035105e-01 2.89453626e-01 6.29026890e-01
-2.96275504e-02 -3.89469892e-01 8.65981102e-01 9.80349332e-02
-7.33488500e-01 1.55806437e-01 -8.70612860e-01 -2.35979110e-01
1.39960015e+00 -4.59647268e-01 -9.10530537e-02 -2.98278064e-01
-4.95089054e-01 7.23376453e-01 7.60162830e-01 4.98156428e-01
5.62931478e-01 -1.21389794e+00 -4.36220437e-01 1.20268330e-01
-5.39903482e-03 3.05582613e-01 7.68853351e-02 8.61592352e-01
-6.63221776e-01 4.49156612e-01 -8.47582147e-02 -8.47547472e-01
-8.08326483e-01 5.01346588e-01 6.65933549e-01 9.37033594e-02
-6.72595382e-01 2.75925547e-01 2.31572427e-02 -3.16105664e-01
-1.77339837e-01 -5.34768522e-01 -3.55823457e-01 -1.26035899e-01
1.14601538e-01 5.92478573e-01 -9.35011283e-02 -7.51320004e-01
-4.15128142e-01 1.05235910e+00 4.31335211e-01 -3.37333590e-01
9.48366344e-01 -2.40817532e-01 1.75229535e-01 3.72212738e-01
8.77708018e-01 -2.09514827e-01 -1.48872674e+00 3.66340488e-01
1.46540746e-01 -2.14200020e-01 -1.19449377e-01 -1.08171925e-01
-7.77086377e-01 7.26457357e-01 3.40183765e-01 3.15130055e-01
7.71829903e-01 -4.27196324e-02 7.09546387e-01 5.34760773e-01
6.37580991e-01 -1.25954878e+00 -2.45939687e-01 8.41417789e-01
1.00190496e+00 -8.40418100e-01 -4.01593029e-01 -4.34317350e-01
-6.86349273e-01 1.26984990e+00 6.95026457e-01 -2.71784246e-01
6.62252545e-01 4.11610097e-01 -1.57703891e-01 -2.87585463e-02
-9.34826672e-01 -2.13838309e-01 -4.36754413e-02 3.23342890e-01
-2.67513692e-01 8.01848322e-02 -6.27182782e-01 1.75138637e-01
-2.19404370e-01 -9.24347490e-02 7.81320214e-01 1.06608188e+00
-7.02470183e-01 -1.14351058e+00 -4.56653804e-01 2.76915520e-01
2.58182108e-01 7.28799522e-01 -1.85613856e-01 8.57976139e-01
1.78316370e-01 1.09011304e+00 4.77161147e-02 -5.46996891e-01
6.33916080e-01 -5.72572835e-02 2.23908797e-01 -2.53638148e-01
-2.22050399e-01 -8.65800232e-02 2.40254283e-01 -7.00221956e-01
-1.84828699e-01 -9.45705354e-01 -1.67016304e+00 -4.27709788e-01
-3.35948437e-01 2.17199147e-01 7.99380302e-01 9.00526941e-01
4.40945268e-01 2.27606118e-01 5.77737570e-01 -1.19323492e+00
-6.07112050e-01 -3.71399671e-01 -1.07485659e-01 -5.48517257e-02
4.36463386e-01 -1.16174352e+00 -3.06808949e-01 -1.76711470e-01] | [5.155618190765381, 1.5857784748077393] |
3d17d01e-f516-41f5-99a5-1387b0705a2b | to-drop-or-not-to-drop-robustness-consistency | 1503.02031 | null | http://arxiv.org/abs/1503.02031v1 | http://arxiv.org/pdf/1503.02031v1.pdf | To Drop or Not to Drop: Robustness, Consistency and Differential Privacy Properties of Dropout | Training deep belief networks (DBNs) requires optimizing a non-convex
function with an extremely large number of parameters. Naturally, existing
gradient descent (GD) based methods are prone to arbitrarily poor local minima.
In this paper, we rigorously show that such local minima can be avoided (upto
an approximation error) by using the dropout technique, a widely used heuristic
in this domain. In particular, we show that by randomly dropping a few nodes of
a one-hidden layer neural network, the training objective function, up to a
certain approximation error, decreases by a multiplicative factor.
On the flip side, we show that for training convex empirical risk minimizers
(ERM), dropout in fact acts as a "stabilizer" or regularizer. That is, a simple
dropout based GD method for convex ERMs is stable in the face of arbitrary
changes to any one of the training points. Using the above assertion, we show
that dropout provides fast rates for generalization error in learning (convex)
generalized linear models (GLM). Moreover, using the above mentioned stability
properties of dropout, we design dropout based differentially private
algorithms for solving ERMs. The learned GLM thus, preserves privacy of each of
the individual training points while providing accurate predictions for new
test points. Finally, we empirically validate our stability assertions for
dropout in the context of convex ERMs and show that surprisingly, dropout
significantly outperforms (in terms of prediction accuracy) the L2
regularization based methods for several benchmark datasets. | ['Oliver Williams', 'Vivek Kulkarni', 'Abhradeep Thakurta', 'Prateek Jain'] | 2015-03-06 | null | null | null | null | ['l2-regularization'] | ['methodology'] | [-1.77604944e-01 4.63531494e-01 -2.70064592e-01 -3.91268015e-01
-1.05204666e+00 -3.98631752e-01 1.08756468e-01 3.00163746e-01
-6.85421288e-01 1.07930934e+00 -2.25967973e-01 -1.49597988e-01
-1.52626531e-02 -8.43885601e-01 -1.36541224e+00 -1.11196625e+00
-5.57730570e-02 3.78909707e-01 -1.81737259e-01 1.19564056e-01
-7.27544054e-02 6.66120172e-01 -1.05495203e+00 -3.42708752e-02
7.65039504e-01 9.64331925e-01 -4.90257174e-01 3.99137259e-01
3.44502717e-01 7.61102557e-01 -4.72101390e-01 -7.50528932e-01
3.73268902e-01 -2.97809124e-01 -7.74197698e-01 -1.86240926e-01
7.11151123e-01 -5.36747396e-01 -4.36205208e-01 1.36693406e+00
4.04091477e-01 2.24386036e-01 5.24276435e-01 -1.12134790e+00
-7.71074414e-01 6.44597709e-01 -3.22301179e-01 -1.56466916e-01
-2.16220021e-01 1.51436252e-03 9.47326839e-01 -7.86482155e-01
4.53303337e-01 9.54717994e-01 8.98698688e-01 9.10333991e-01
-1.68698204e+00 -6.50067449e-01 9.63538066e-02 -3.06452036e-01
-1.38166618e+00 -4.98047382e-01 5.17969489e-01 -3.06645602e-01
7.27850556e-01 2.98228741e-01 2.95230329e-01 1.13659179e+00
8.77153128e-02 8.52479756e-01 8.95084500e-01 -2.75635898e-01
5.59855044e-01 7.04945385e-01 5.30874133e-01 1.01760733e+00
4.82772887e-01 8.40577036e-02 -3.29834610e-01 -4.45681393e-01
3.48240405e-01 1.46860465e-01 -4.97409195e-01 -4.35034394e-01
-4.01515216e-01 1.12769663e+00 6.77213132e-01 2.36847594e-01
1.65818632e-02 5.00106990e-01 2.83842236e-01 6.44479573e-01
8.00336182e-01 2.00932652e-01 -4.06383127e-01 3.03530902e-01
-8.05712640e-01 2.11011171e-02 9.46129560e-01 7.97491729e-01
8.92381072e-01 1.35022461e-01 -1.74788818e-01 5.11298299e-01
1.04238629e-01 3.26892287e-01 5.25483668e-01 -6.83418334e-01
4.67312485e-01 2.19714388e-01 4.86562178e-02 -1.01214409e+00
-2.86736637e-01 -7.37247169e-01 -1.09254122e+00 5.18325567e-01
5.46195090e-01 -3.18063200e-01 -6.70198679e-01 2.25271344e+00
9.44854692e-02 5.78180104e-02 1.70640856e-01 6.55073762e-01
2.10920677e-01 5.53391278e-01 -9.91625115e-02 -2.55493850e-01
5.89916468e-01 -5.12195110e-01 -4.76825446e-01 2.09941454e-02
9.49749589e-01 8.98589641e-02 1.10622466e+00 4.22671527e-01
-1.12654471e+00 5.80765195e-02 -1.08526576e+00 -2.41405889e-01
-1.61392540e-01 1.30458146e-01 4.48300391e-01 9.41152632e-01
-1.22168303e+00 1.02848673e+00 -1.11071146e+00 1.04864523e-01
7.85183907e-01 7.48726606e-01 -5.66603959e-01 -1.77923664e-02
-9.93019879e-01 6.30638778e-01 8.56359303e-02 2.71953911e-01
-1.12338960e+00 -9.30776656e-01 -6.44812703e-01 2.52150476e-01
1.18933450e-02 -6.11159801e-01 8.43495667e-01 -1.14196706e+00
-1.55482066e+00 1.04883921e+00 -9.56729054e-02 -1.01125205e+00
1.10801029e+00 -3.04787278e-01 3.35013568e-01 -8.02518614e-03
-4.44528818e-01 3.52424234e-01 1.02030563e+00 -9.60695386e-01
-2.74624050e-01 -7.63532758e-01 9.10463184e-02 -1.51910186e-01
-7.16419995e-01 -3.18114579e-01 2.47908980e-01 -2.69902766e-01
-1.78710744e-01 -8.16110730e-01 -2.27580220e-01 2.40183800e-01
-5.55365682e-01 -3.06214616e-02 5.52566946e-01 -5.74131131e-01
1.01622510e+00 -2.49359345e+00 2.34456837e-01 3.80282789e-01
4.46262866e-01 1.55544728e-01 1.75854549e-01 -8.02767575e-02
-4.04716767e-02 3.56673390e-01 -4.76907432e-01 -1.01718605e+00
1.72392204e-01 1.52548775e-01 -4.34949160e-01 1.12085593e+00
-1.31772578e-01 8.14551234e-01 -6.68468356e-01 -5.63266911e-02
-4.10775468e-02 5.75010955e-01 -8.98215234e-01 -1.03315808e-01
1.16891311e-02 2.51070708e-01 -3.78378958e-01 6.89076167e-03
8.16133440e-01 -1.96136698e-01 -5.58711076e-03 2.25417212e-01
1.75522491e-01 2.24153558e-03 -9.44635451e-01 1.31389785e+00
-6.08727038e-01 6.64885163e-01 1.32081285e-01 -1.23347032e+00
8.02116156e-01 3.39553744e-01 1.33412480e-01 8.02446678e-02
3.04566264e-01 6.11837685e-01 -4.98871863e-01 -4.89851646e-02
-2.48811468e-02 -3.93655628e-01 2.40599409e-01 9.23959836e-02
1.85039043e-01 4.87370968e-01 -3.92767757e-01 -7.29486346e-02
1.09671187e+00 -6.14880264e-01 -1.78660274e-01 -1.94800422e-01
3.77568543e-01 -4.33587641e-01 5.51916361e-01 9.86214399e-01
-1.39770120e-01 6.28363431e-01 8.30461740e-01 -2.79063672e-01
-9.59322035e-01 -8.33132446e-01 -3.75852197e-01 7.99960196e-01
-3.15714896e-01 6.97823465e-02 -1.01284468e+00 -6.89877868e-01
2.56186038e-01 7.75655270e-01 -8.08240592e-01 -6.59505308e-01
-3.86924952e-01 -1.09374177e+00 7.12602615e-01 1.14891529e-01
7.05469549e-01 -7.44728327e-01 -1.94470540e-01 1.98644050e-03
3.31700176e-01 -6.96135461e-01 -3.74655575e-01 5.23352504e-01
-1.21012104e+00 -8.65343392e-01 -9.29215133e-01 -9.02534664e-01
9.13285911e-01 -1.14144646e-01 7.36366928e-01 1.64536998e-01
3.93101424e-02 1.25676453e-01 4.26876582e-02 -1.53335901e-02
-4.27005440e-01 3.13258976e-01 1.18129812e-02 3.70446295e-01
1.79876328e-01 -7.63359189e-01 -4.51925039e-01 -1.98189035e-01
-8.67414594e-01 -3.42825860e-01 2.28493690e-01 9.18693781e-01
6.91915989e-01 6.61813915e-02 3.85609835e-01 -1.24330151e+00
5.26269853e-01 -5.84802270e-01 -1.07066214e+00 2.43238732e-01
-6.06368363e-01 4.02773350e-01 9.76054370e-01 -4.06466812e-01
-6.13331497e-01 1.45631075e-01 -8.81627649e-02 -5.81623435e-01
1.81321755e-01 3.49354923e-01 -9.10539627e-02 -4.72343773e-01
8.34888935e-01 1.83554247e-01 1.41190309e-02 -6.68612778e-01
2.68519014e-01 5.46085596e-01 2.93306231e-01 -5.12440443e-01
7.21038818e-01 7.14691162e-01 2.15764910e-01 -7.65765905e-01
-9.73810792e-01 1.02507599e-01 -2.71459609e-01 2.02726945e-01
6.07102633e-01 -7.82452643e-01 -1.02837503e+00 6.34063542e-01
-8.97814095e-01 -5.29203713e-01 -5.55962622e-01 4.61357951e-01
-6.95509553e-01 1.39702573e-01 -8.19861293e-01 -1.03471398e+00
-3.60901564e-01 -1.01970088e+00 5.22334933e-01 -2.41980329e-02
1.89423561e-01 -1.18785024e+00 -3.28752846e-02 8.54133070e-02
2.52288222e-01 4.77564245e-01 9.71543014e-01 -8.06583464e-01
-4.05460507e-01 -5.53734422e-01 1.04685120e-01 9.05739963e-01
-2.33728170e-01 -2.18112215e-01 -1.06089580e+00 -6.01453245e-01
6.57455623e-01 -3.90558690e-01 1.15662646e+00 5.39713204e-01
1.30485725e+00 -8.47903788e-01 -5.12525998e-03 1.12978721e+00
1.59410906e+00 -4.41546142e-01 4.45892543e-01 1.97775304e-01
6.02210999e-01 3.58998418e-01 -2.41070837e-01 4.70544666e-01
-1.06793568e-01 3.22097421e-01 4.71572936e-01 5.23593985e-02
2.71580338e-01 -3.15688431e-01 7.22897828e-01 3.11552733e-01
2.75582939e-01 -1.90262850e-02 -3.98506075e-01 4.48880792e-01
-1.78161144e+00 -7.21387267e-01 -2.25263864e-01 2.83737946e+00
1.30140090e+00 8.23169574e-02 6.21189456e-03 -3.50425169e-02
6.64021015e-01 -1.11029424e-01 -8.15127254e-01 -5.21283388e-01
-3.71330410e-01 3.04279119e-01 1.08614695e+00 8.61930847e-01
-1.09437621e+00 7.73843288e-01 5.72730398e+00 7.84975827e-01
-1.07027698e+00 3.96432787e-01 1.07519639e+00 -4.62518930e-01
-3.04871619e-01 -1.85262367e-01 -8.26275885e-01 5.39245605e-01
1.10506713e+00 -2.84078568e-01 7.73750663e-01 1.16792905e+00
9.01682600e-02 1.77174687e-01 -1.36837757e+00 9.19114470e-01
-3.04559618e-01 -1.33860230e+00 -1.50311321e-01 2.47711822e-01
8.54129136e-01 1.41368538e-01 6.21953785e-01 2.09257066e-01
5.31997144e-01 -1.22243881e+00 6.71500564e-01 3.25659692e-01
5.63848317e-01 -1.12478197e+00 7.10114360e-01 4.33870524e-01
-2.40793481e-01 -1.29060373e-01 -7.61972010e-01 3.16793948e-01
-2.90028930e-01 1.04541612e+00 -1.62674040e-01 1.43229917e-01
4.28018600e-01 3.70607764e-01 -2.08157316e-01 9.83751893e-01
-3.19515586e-01 8.32963109e-01 -8.21627319e-01 8.13213959e-02
4.14511412e-01 -3.80427152e-01 6.30958974e-01 1.01260388e+00
1.53460771e-01 -2.15896919e-01 -2.65964895e-01 1.10437274e+00
-7.44046092e-01 8.61403197e-02 -6.64939702e-01 3.19782436e-01
1.47052571e-01 7.49947608e-01 -6.80303872e-02 -9.15294793e-03
-1.95909053e-01 1.15916121e+00 8.31357002e-01 5.86157620e-01
-6.52420998e-01 -2.92225748e-01 8.11969578e-01 -6.08644336e-02
3.13809276e-01 2.20806040e-02 -3.98233175e-01 -1.23618793e+00
4.64400470e-01 -6.66525841e-01 3.88726294e-01 -1.56477258e-01
-1.31193626e+00 3.71782154e-01 -4.75743264e-01 -6.41169369e-01
8.97250548e-02 -5.61542630e-01 -3.50126177e-01 8.12906981e-01
-1.55575573e+00 -6.31330609e-01 2.77203053e-01 6.91605985e-01
-1.34305567e-01 -4.81565334e-02 8.03936601e-01 3.23915660e-01
-9.11456883e-01 1.27120733e+00 9.25277531e-01 3.09013128e-01
3.26207787e-01 -1.42670465e+00 -9.94593278e-02 7.28240848e-01
2.64878303e-01 7.11702585e-01 7.05930352e-01 -2.55684733e-01
-1.26861405e+00 -1.25717306e+00 9.20440376e-01 -2.35184848e-01
6.49803162e-01 -5.06019235e-01 -1.13160706e+00 1.01181436e+00
-3.77880424e-01 3.08025539e-01 5.78110635e-01 1.34909287e-01
-4.31306124e-01 -2.17238173e-01 -1.67055070e+00 4.94677395e-01
6.37714088e-01 -6.32245004e-01 -1.01318337e-01 8.54981780e-01
7.01599002e-01 -3.95989239e-01 -7.55220711e-01 -1.30684882e-01
2.16968909e-01 -9.75803554e-01 7.80569851e-01 -1.18503976e+00
2.42890999e-01 1.40270039e-01 -2.46665880e-01 -1.11239755e+00
1.10427462e-01 -1.11235523e+00 -4.45149481e-01 1.04248142e+00
4.78600800e-01 -9.53072190e-01 1.11847711e+00 1.06273592e+00
1.22658648e-01 -8.17378342e-01 -1.43377495e+00 -1.04753411e+00
7.12714493e-01 -3.05263132e-01 1.29405037e-01 7.49912560e-01
-6.63001314e-02 -1.38644308e-01 -6.28232777e-01 3.33035707e-01
7.55482256e-01 -3.25057715e-01 3.94615918e-01 -9.58815217e-01
-4.92489189e-01 -4.57889855e-01 -6.33275092e-01 -9.98157859e-01
5.60398459e-01 -1.26909363e+00 -6.90180659e-02 -7.03187346e-01
1.84172660e-01 -4.43569958e-01 -5.77792406e-01 5.31507194e-01
-4.60718907e-02 9.59712341e-02 1.00168981e-01 9.88918394e-02
-2.97061741e-01 7.03164876e-01 7.31166482e-01 -5.81794046e-02
-3.00981760e-01 5.44731379e-01 -6.46213412e-01 6.89768910e-01
8.44402611e-01 -1.03647625e+00 -7.37766027e-02 -4.64416862e-01
3.66185904e-01 -2.84512430e-01 6.43218100e-01 -8.42494011e-01
2.31612131e-01 2.98146248e-01 1.28462343e-02 6.84580430e-02
1.88364819e-01 -7.81180203e-01 -3.65907960e-02 6.92454517e-01
-7.69410729e-01 -5.03049850e-01 -2.99463943e-02 6.76600575e-01
9.10864398e-02 -7.04177558e-01 1.18480647e+00 1.41419411e-01
2.38583654e-01 4.93286431e-01 -2.86849469e-01 3.50276291e-01
7.99209535e-01 6.30690455e-02 -1.47761563e-02 -4.70095485e-01
-9.33508039e-01 1.41916290e-01 4.54947740e-01 -3.96993816e-01
4.01323617e-01 -1.15030515e+00 -6.71437800e-01 6.47369549e-02
-3.88974249e-01 8.52767378e-02 1.84562467e-02 9.63118732e-01
-4.20543671e-01 1.70011699e-01 3.00405443e-01 -1.87405884e-01
-9.75942075e-01 4.72714931e-01 8.06097031e-01 -2.34056979e-01
-7.20469832e-01 1.14859045e+00 1.02848031e-01 -2.93286562e-01
7.62390435e-01 -4.21546340e-01 4.60030198e-01 -2.10922822e-01
4.92893308e-01 3.87834340e-01 2.20665723e-01 -1.85038224e-01
-1.39006674e-01 1.35585964e-01 -2.00811297e-01 -4.29007262e-02
1.65321589e+00 8.03675950e-02 -6.97968453e-02 4.90352243e-01
2.05817866e+00 -2.19894107e-02 -1.58823180e+00 -2.18790978e-01
9.50186700e-03 -2.77736872e-01 2.78851241e-01 -2.81751007e-01
-1.40740752e+00 9.52782393e-01 5.71077585e-01 2.05380946e-01
9.16886687e-01 -1.99132994e-01 6.68880343e-01 7.57171333e-01
3.97929639e-01 -8.43951941e-01 -4.96560514e-01 3.41931790e-01
6.62655294e-01 -1.16262543e+00 -2.46817365e-01 2.50917003e-02
-1.92492977e-01 1.10134745e+00 -2.31411699e-02 -5.46139598e-01
8.20200384e-01 1.24130413e-01 -2.98670888e-01 1.98568627e-01
-5.69351852e-01 3.77812922e-01 -4.49769869e-02 4.76782173e-01
4.48972546e-02 -1.58440024e-01 -1.05015844e-01 5.02371728e-01
-1.78651124e-01 -4.57975678e-02 5.10654211e-01 5.50453067e-01
-4.93191898e-01 -8.91391098e-01 -5.79562150e-02 4.12513912e-01
-7.60139942e-01 -2.20057189e-01 -3.24592471e-01 7.44059324e-01
-4.28105354e-01 6.00287497e-01 -1.45383894e-01 -5.20993248e-02
8.31637755e-02 9.20996517e-02 3.76685292e-01 -4.14565533e-01
-7.39592969e-01 -6.27851605e-01 -3.76911402e-01 -5.78128457e-01
5.55130281e-02 -6.54464662e-01 -1.15023160e+00 -7.98381329e-01
-3.51487011e-01 1.08530462e-01 5.37487209e-01 8.60048950e-01
3.57359916e-01 4.23833467e-02 8.07706356e-01 -2.64054418e-01
-1.37154293e+00 -6.03264630e-01 -8.03451359e-01 3.62821072e-01
7.40107775e-01 -2.49822542e-01 -1.16256273e+00 -2.79013366e-01] | [7.767922401428223, 3.9903976917266846] |
e1b3b444-8424-47a9-b45f-4a6e835914ac | deep-metric-learning-based-feature-embedding | null | null | https://doi.org/10.1109/TGRS.2019.2946318 | https://doi.org/10.1109/TGRS.2019.2946318 | Deep Metric Learning-Based Feature Embedding for Hyperspectral Image Classification | Learning from a limited number of labeled samples (pixels) remains a key challenge in the hyperspectral image (HSI) classification. To address this issue, we propose a deep metric learning-based feature embedding model, which can meet the tasks both for same- and cross-scene HSI classifications. In the first task, when only a few labeled samples are available, we employ ideas from metric learning based on deep embedding features and make a similarity learning between pairs of samples. In this case, the proposed model can learn well to compare whether two samples belong to the same class. In another task, when an HSI image (target scene) that needs to be classified is not labeled at all, the embedding model can learn from another similar HSI image (source scene) with sufficient labeled samples and then transfer to the target model by using an unsupervised domain adaptation technique, which not only employs the adversarial approach to make the embedding features from the source and target samples indistinguishable but also encourages the target scene's embeddings to form similar clusters with the source scene one. After the domain adaptation between the HSIs of the two scenes is finished, any traditional HSI classifier can be used. In a simple manner, the nearest neighbor (NN) algorithm is selected as the classifier for the classification tasks throughout this article. The experimental results from a series of popular HSIs demonstrate the advantages of the proposed model both in the same- and cross-scene classification tasks. | ['Daming Shi', 'Sen Jia', 'Bin Deng'] | 2019-10-30 | null | null | null | ieee-transactions-on-geoscience-and-remote-15 | ['metric-learning', 'scene-classification', 'few-shot-image-classification', 'metric-learning'] | ['computer-vision', 'computer-vision', 'computer-vision', 'methodology'] | [ 5.38271546e-01 -2.73802549e-01 7.01561049e-02 -5.75476408e-01
-5.53465068e-01 -4.98251051e-01 4.27359074e-01 6.00230731e-02
-3.54180098e-01 5.97981989e-01 -2.30735600e-01 5.56075089e-02
-1.94030792e-01 -1.06319642e+00 -3.53025615e-01 -1.20303035e+00
2.34709397e-01 8.06275085e-02 2.25321889e-01 -7.33678341e-02
1.38309911e-01 6.13007426e-01 -1.51552069e+00 1.37759537e-01
9.60372448e-01 1.30665469e+00 3.06621045e-01 3.20813924e-01
-2.70687908e-01 5.70409417e-01 -3.89223605e-01 1.02634996e-01
6.85173988e-01 -6.88515425e-01 -6.66140258e-01 2.50512540e-01
4.61841345e-01 -2.58206695e-01 -5.54670811e-01 1.43376732e+00
4.62308466e-01 4.69164461e-01 7.79866159e-01 -1.47205234e+00
-8.20029974e-01 8.59191045e-02 -5.77504337e-01 -9.10082087e-02
4.25464101e-02 5.13453484e-02 5.41867495e-01 -8.03277791e-01
1.92401931e-01 8.84577036e-01 4.98643547e-01 4.51632798e-01
-1.14856398e+00 -8.15615356e-01 -3.50079983e-02 4.08787936e-01
-1.62741995e+00 -3.03857237e-01 1.07835126e+00 -4.76644009e-01
9.18000471e-03 3.05909783e-01 5.01033127e-01 5.70504248e-01
-1.52566105e-01 3.48087132e-01 1.34216595e+00 -4.19800133e-01
4.33231771e-01 4.50473130e-01 9.85287055e-02 4.06081945e-01
-1.29909662e-03 2.10196584e-01 -7.39742890e-02 4.63014506e-02
1.63235381e-01 4.62033391e-01 -6.02949679e-01 -5.13715982e-01
-1.17494249e+00 9.85577941e-01 9.80655849e-01 4.67643648e-01
-2.40540132e-01 -4.79445100e-01 3.08860153e-01 3.77858639e-01
2.34383240e-01 3.57671291e-01 -1.22135423e-01 5.50852239e-01
-8.78470004e-01 -3.16981703e-01 5.27299702e-01 8.57412398e-01
1.28495455e+00 -6.25348464e-02 -1.54710427e-01 7.59997010e-01
1.89603567e-01 5.19078314e-01 6.99219704e-01 -4.87006277e-01
3.23109686e-01 8.32230330e-01 8.43180642e-02 -1.37805045e+00
-5.10313511e-02 -2.48441890e-01 -1.06658649e+00 4.79608089e-01
3.24489862e-01 1.17597006e-01 -8.00641954e-01 1.55554664e+00
5.47281265e-01 5.32324791e-01 3.93046498e-01 1.06455362e+00
8.56867373e-01 8.90784264e-01 -1.61075369e-01 -2.22983703e-01
9.20279860e-01 -8.11340868e-01 -4.13622856e-01 -2.32651472e-01
4.31039095e-01 -6.34184420e-01 1.05365324e+00 8.10669288e-02
-2.40406498e-01 -8.37539971e-01 -1.46554804e+00 3.91666926e-02
-7.90184557e-01 1.06281310e-01 2.73005158e-01 4.67803955e-01
-6.74645603e-01 6.24993801e-01 -4.83223349e-01 -5.99543393e-01
5.46807289e-01 1.78440198e-01 -5.89128554e-01 -7.15084434e-01
-1.04819834e+00 5.58464944e-01 6.78845704e-01 1.28653869e-01
-9.48104501e-01 -6.40509963e-01 -9.31620777e-01 -5.35780117e-02
1.19527772e-01 -1.06640525e-01 5.13519049e-01 -1.23454010e+00
-1.23939800e+00 8.60172510e-01 -4.48564887e-02 1.80519626e-01
3.99715245e-01 3.46771032e-01 -7.22906351e-01 1.55699581e-01
3.15707862e-01 4.39545959e-01 9.86218929e-01 -1.30089498e+00
-7.81578958e-01 -5.11922359e-01 -2.18541250e-02 3.28087956e-01
-6.42686903e-01 -3.17163706e-01 1.21758275e-01 -3.24065953e-01
5.10878146e-01 -8.12673807e-01 7.68084675e-02 5.48055351e-01
-5.15142918e-01 -7.26923794e-02 1.31576765e+00 -5.29709518e-01
5.53767562e-01 -2.73288608e+00 -3.32221724e-02 2.79305071e-01
6.69915928e-03 2.86950141e-01 -3.32281619e-01 2.04065263e-01
-3.91789407e-01 -4.29025330e-02 -6.45765901e-01 6.33215159e-02
-1.72413081e-01 1.31278709e-01 -1.60637408e-01 7.87405789e-01
2.11502254e-01 4.36260283e-01 -1.13650143e+00 -4.15138960e-01
4.55218792e-01 1.85163274e-01 1.42921163e-02 6.65807307e-01
2.40997478e-01 6.65535390e-01 -3.20976019e-01 4.54147130e-01
1.09766042e+00 1.00618705e-01 -1.37919709e-01 -7.39764988e-01
-3.60533521e-02 -2.18441635e-01 -1.38345385e+00 1.63544917e+00
-2.27823377e-01 4.87303883e-01 -1.19035266e-01 -1.40197682e+00
1.11402011e+00 7.98543617e-02 3.87538880e-01 -4.93760377e-01
-7.92634189e-02 1.84006870e-01 5.03255390e-02 -5.26102364e-01
-8.46267119e-02 -3.95046294e-01 2.14000508e-01 4.50609356e-01
3.92075442e-02 -3.28855246e-01 -1.83590725e-02 -1.88757554e-01
7.56229699e-01 -2.20906183e-01 2.87402034e-01 -2.58259803e-01
8.11626017e-01 6.93911090e-02 7.36515164e-01 4.60512370e-01
-4.73169833e-01 5.52068472e-01 -7.87228420e-02 -3.90558094e-01
-9.44438815e-01 -1.18412876e+00 -4.42501247e-01 8.05447221e-01
5.41508317e-01 3.74386966e-01 -3.34094524e-01 -9.58338320e-01
-6.08022884e-02 5.26194751e-01 -7.46567369e-01 -5.91342330e-01
7.70075917e-02 -5.95483840e-01 2.57206112e-01 2.78903961e-01
9.84419346e-01 -8.89463961e-01 -1.44114405e-01 -4.06022742e-02
-2.47296933e-02 -8.34179819e-01 -4.06926513e-01 4.01154578e-01
-5.78779280e-01 -1.24464035e+00 -5.94730139e-01 -1.14428389e+00
8.08442891e-01 6.68145895e-01 3.30443680e-01 -2.44298756e-01
-1.73618361e-01 -4.19248417e-02 -5.06697714e-01 -1.19617276e-01
-3.44648302e-01 -2.52618372e-01 2.47053891e-01 6.53993249e-01
5.74419141e-01 -5.80047369e-01 -5.82526982e-01 4.99496460e-01
-1.21491361e+00 -1.43178806e-01 4.07712370e-01 9.16931391e-01
4.89032358e-01 6.29473329e-01 5.22568226e-01 -7.27782965e-01
7.79191777e-02 -6.31273091e-01 -3.70697767e-01 3.33604753e-01
-3.74176145e-01 -2.20064625e-01 1.02789259e+00 -5.81784487e-01
-8.16458225e-01 1.28186598e-01 2.88821071e-01 -5.30597806e-01
-5.07981956e-01 5.55178761e-01 -6.42674387e-01 -3.86616647e-01
7.45629489e-01 6.65769994e-01 1.14269853e-01 -2.51839608e-01
2.61612892e-01 1.02799714e+00 5.81148684e-01 -1.41252100e-01
1.46207774e+00 5.85850894e-01 6.89616287e-03 -8.02502930e-01
-7.69851506e-01 -5.92539608e-01 -9.39022303e-01 -7.37785473e-02
9.32763100e-01 -8.20269883e-01 -1.74175113e-01 5.06993294e-01
-7.56204486e-01 -1.00488797e-01 -4.42997187e-01 7.24227190e-01
-2.24110827e-01 4.85901713e-01 -1.89789787e-01 -5.50057411e-01
8.67740810e-02 -1.00653851e+00 7.69328535e-01 6.70262754e-01
3.66936773e-01 -1.08169782e+00 -3.38020846e-02 1.90108130e-03
2.58239657e-01 4.52277482e-01 1.06158030e+00 -9.89341021e-01
-2.59725869e-01 -4.16758597e-01 -4.26648259e-01 7.84445822e-01
9.07390893e-01 -1.34032890e-01 -1.21151006e+00 -5.81663907e-01
4.60390568e-01 -4.25608903e-01 4.98552620e-01 -3.97599339e-02
1.21448660e+00 -2.02884108e-01 -2.07710773e-01 9.79980886e-01
1.81830871e+00 4.63586092e-01 5.52581251e-01 3.46296847e-01
8.35331082e-01 7.50753641e-01 8.90593827e-01 1.43451750e-01
-2.62492653e-02 2.69430190e-01 4.94127333e-01 -4.34663445e-01
1.44960312e-02 -2.98536152e-01 2.45464131e-01 7.20916033e-01
5.32934427e-01 1.11370869e-01 -6.03739321e-01 4.31715488e-01
-1.43724430e+00 -9.11475182e-01 5.86021729e-02 2.41522074e+00
6.67856932e-01 -2.79024243e-01 -2.60553002e-01 3.99658352e-01
1.11738884e+00 2.97143340e-01 -9.31934595e-01 -2.59282924e-02
-1.37754083e-01 5.61564080e-02 3.12415570e-01 1.13689147e-01
-1.36954570e+00 6.60740674e-01 5.19540691e+00 8.72834563e-01
-1.50133049e+00 -3.44470404e-02 5.99403262e-01 3.32288086e-01
-2.35766135e-02 9.82038602e-02 -2.33473465e-01 5.59616327e-01
4.13726926e-01 -2.66750246e-01 7.36947417e-01 8.03620517e-01
1.61270410e-01 -4.72320914e-02 -1.36112297e+00 1.13526475e+00
2.36859679e-01 -8.39320242e-01 6.51836693e-02 -6.43122047e-02
7.45297074e-01 -1.97491765e-01 9.18785036e-02 2.83892155e-01
-5.02053276e-02 -9.42249358e-01 2.34430179e-01 4.89398777e-01
1.03249562e+00 -6.97385609e-01 8.28107476e-01 5.14620662e-01
-1.34642744e+00 -2.48245105e-01 -9.61669683e-01 1.27215981e-01
-4.04983222e-01 5.99516153e-01 -6.15639031e-01 9.38371360e-01
7.12231040e-01 1.07551849e+00 -7.71111190e-01 1.09331882e+00
-3.77448350e-02 3.51591229e-01 -1.30935609e-01 3.15448463e-01
2.20462158e-01 -7.79466569e-01 3.28833044e-01 7.32918918e-01
5.10695755e-01 1.99702114e-01 5.29385269e-01 9.74317729e-01
-3.51598929e-03 2.11022526e-01 -1.04119980e+00 1.33486846e-02
5.14457285e-01 1.37797678e+00 -4.35621470e-01 -2.99011856e-01
-4.16792363e-01 1.31181026e+00 1.69537336e-01 5.41443467e-01
-7.16666043e-01 -8.92256141e-01 6.01072848e-01 -2.83010542e-01
1.34701997e-01 7.55820572e-02 2.53457762e-03 -1.02569807e+00
1.03311405e-01 -6.69742227e-01 3.44839990e-01 -9.76360619e-01
-1.69069660e+00 5.18214047e-01 -2.33561561e-01 -1.75325501e+00
3.02693129e-01 -6.76153481e-01 -9.49549675e-01 1.14590299e+00
-1.88159645e+00 -1.22270441e+00 -8.71841192e-01 1.09008884e+00
7.45877326e-02 -2.37324700e-01 9.10125196e-01 2.68100351e-01
-6.08183444e-01 5.92399895e-01 6.42731786e-01 5.56286454e-01
7.86955118e-01 -1.12329698e+00 -4.97835577e-01 9.87799108e-01
-6.87585622e-02 1.84245870e-01 2.49915794e-01 -2.86735952e-01
-1.09454393e+00 -1.65715969e+00 4.48385179e-01 1.38562590e-01
5.28165042e-01 -1.60449103e-01 -1.07096112e+00 3.51230979e-01
-7.26062059e-02 4.44668770e-01 1.07709372e+00 -4.27576512e-01
-4.40928578e-01 -7.00598955e-01 -1.54685748e+00 3.84186059e-01
7.32314527e-01 -8.62014592e-01 -5.71931303e-01 5.96150696e-01
6.27701044e-01 1.33048788e-01 -9.84857917e-01 3.50304425e-01
1.55579254e-01 -9.20360446e-01 7.04960823e-01 -5.58923662e-01
3.34260494e-01 -7.76137412e-01 -6.17037952e-01 -1.60500836e+00
-4.84161258e-01 2.84140967e-02 4.91035283e-01 1.43015444e+00
-7.53783584e-02 -9.05313551e-01 6.00241184e-01 4.28787082e-01
-1.11541197e-01 -1.88232794e-01 -8.58658195e-01 -9.96594071e-01
1.20581210e-01 -2.06202026e-02 8.23930681e-01 1.40764308e+00
-2.27632776e-01 7.57838935e-02 -9.29381028e-02 6.18661404e-01
7.37600446e-01 5.03409028e-01 6.00537360e-01 -1.34022093e+00
1.05379999e-01 -1.29622594e-01 -8.02784026e-01 -6.40594363e-01
3.41293842e-01 -1.27848983e+00 2.15343371e-01 -1.24706590e+00
3.04606467e-01 -6.12695813e-01 -5.83992302e-01 5.18301189e-01
-2.09598407e-01 3.59509945e-01 1.87191337e-01 3.48235518e-01
-9.39230323e-02 8.36787224e-01 1.21732426e+00 -6.83176875e-01
-7.58979172e-02 -2.76115071e-02 -6.99814856e-01 4.19560254e-01
8.08124244e-01 -5.89564025e-01 -4.77251559e-01 -2.54416078e-01
-5.78177035e-01 -2.03872800e-01 3.25092047e-01 -1.45114946e+00
2.69023329e-01 -4.86961275e-01 6.44212127e-01 -1.84263408e-01
1.52487695e-01 -1.31685388e+00 -4.42181826e-02 4.47498679e-01
-2.42555976e-01 -4.93954986e-01 -6.09726505e-03 7.00670719e-01
-3.12655598e-01 -4.34578210e-01 1.27954018e+00 1.64914787e-01
-8.73955309e-01 6.96266234e-01 -4.91439411e-03 -1.30310461e-01
1.38295543e+00 -5.74916422e-01 -2.16984659e-01 -2.03972518e-01
-7.49919534e-01 1.29845038e-01 6.67585671e-01 2.10370034e-01
7.58541107e-01 -1.80312884e+00 -7.58544266e-01 4.87528235e-01
6.92849994e-01 1.56872973e-01 1.83271945e-01 6.19011104e-01
-3.14358711e-01 -5.91361895e-02 -3.33865494e-01 -8.25508416e-01
-8.93574357e-01 9.24150288e-01 4.76769418e-01 2.24972442e-01
-2.53004134e-01 6.00375295e-01 3.67553353e-01 -8.12656641e-01
-6.13938980e-02 -2.90414188e-02 -2.02446252e-01 5.62682152e-02
6.90258622e-01 2.27085307e-01 -5.50649203e-02 -7.51343906e-01
-3.10034424e-01 6.58697069e-01 1.99105993e-01 1.57878548e-01
1.17873168e+00 -7.55460560e-02 -2.94470787e-01 7.06219971e-01
1.87210667e+00 -2.97236651e-01 -1.25890076e+00 -5.51174700e-01
-3.42466086e-01 -9.34026301e-01 1.16613775e-01 -6.09498143e-01
-1.24225616e+00 1.05290949e+00 1.32722926e+00 1.96562096e-01
1.30566740e+00 -2.06328347e-01 5.59585690e-01 4.07380790e-01
1.56799257e-01 -8.82285476e-01 2.22834591e-02 2.09396675e-01
6.66106105e-01 -1.62862074e+00 -2.67736316e-01 -1.72870412e-01
-4.96790588e-01 1.21713984e+00 6.23898387e-01 -1.71690434e-01
8.26074719e-01 -3.13901246e-01 1.82125181e-01 5.92445172e-02
1.24939255e-01 -1.90751761e-01 2.32278794e-01 1.11415243e+00
1.03990536e-03 9.41987857e-02 2.30997056e-01 1.90538168e-01
1.00401819e-01 -2.73673445e-01 3.55244964e-01 7.62998462e-01
-5.66550136e-01 -8.98522496e-01 -4.03653294e-01 5.23488343e-01
3.14187855e-01 -1.71894580e-02 -3.37212056e-01 5.71954012e-01
3.72665703e-01 1.16901147e+00 1.36098742e-01 -7.01193333e-01
2.40211561e-01 8.45852718e-02 8.69712159e-02 -7.87993968e-01
-6.05557971e-02 -3.95167559e-01 -6.83290184e-01 -3.43246549e-01
-5.38973510e-01 -4.83291477e-01 -9.89756644e-01 -2.78387934e-01
-3.31990331e-01 2.04298019e-01 5.57479858e-01 9.27074194e-01
-1.36620626e-02 1.47358298e-01 1.47016537e+00 -9.31946218e-01
-7.09497988e-01 -9.13386166e-01 -1.18493140e+00 7.18578279e-01
4.56186265e-01 -5.08415401e-01 -7.22502589e-01 6.00214042e-02] | [9.961955070495605, -1.4404453039169312] |
0c7ed5fd-44af-40f7-a3a2-3aad7d64de07 | a-little-pretraining-goes-a-long-way-a-case | 2102.06551 | null | https://arxiv.org/abs/2102.06551v2 | https://arxiv.org/pdf/2102.06551v2.pdf | A Little Pretraining Goes a Long Way: A Case Study on Dependency Parsing Task for Low-resource Morphologically Rich Languages | Neural dependency parsing has achieved remarkable performance for many domains and languages. The bottleneck of massive labeled data limits the effectiveness of these approaches for low resource languages. In this work, we focus on dependency parsing for morphological rich languages (MRLs) in a low-resource setting. Although morphological information is essential for the dependency parsing task, the morphological disambiguation and lack of powerful analyzers pose challenges to get this information for MRLs. To address these challenges, we propose simple auxiliary tasks for pretraining. We perform experiments on 10 MRLs in low-resource settings to measure the efficacy of our proposed pretraining method and observe an average absolute gain of 2 points (UAS) and 3.6 points (LAS). Code and data available at: https://github.com/jivnesh/LCM | ['Pawan Goyal', 'Laxmidhar Behera', 'Ashim Gupta', 'Amrith Krishna', 'Jivnesh Sandhan'] | 2021-02-12 | null | https://aclanthology.org/2021.eacl-srw.16 | https://aclanthology.org/2021.eacl-srw.16.pdf | eacl-2021-2 | ['morphological-disambiguation'] | ['natural-language-processing'] | [-1.26227960e-01 3.90903987e-02 -1.16597436e-01 -5.25928795e-01
-1.08604193e+00 -7.41107643e-01 3.04523915e-01 2.43775308e-01
-9.44435954e-01 7.44322062e-01 2.04429924e-01 -7.51479447e-01
4.33321029e-01 -5.49999952e-01 -6.45784616e-01 -2.95207977e-01
7.75329322e-02 3.10098737e-01 2.29330093e-01 -7.33013824e-02
-1.06402732e-01 4.08450991e-01 -8.56487870e-01 2.15414256e-01
9.59359586e-01 3.07436973e-01 4.86038148e-01 5.49184620e-01
-3.08767438e-01 4.75166231e-01 -5.75797319e-01 -6.37494981e-01
3.21153075e-01 -2.56538659e-01 -8.68807495e-01 -4.25983518e-01
1.32424846e-01 -2.41152793e-01 -8.95844325e-02 1.16999388e+00
6.48975492e-01 1.76799953e-01 2.33827889e-01 -7.97419012e-01
-6.13301218e-01 1.28071511e+00 -5.67670822e-01 6.23858869e-01
8.38511139e-02 7.71847665e-02 1.12713325e+00 -1.01312196e+00
5.90664089e-01 1.24260890e+00 2.58672327e-01 6.44385278e-01
-8.68608236e-01 -7.64109969e-01 3.50567818e-01 -1.91690307e-02
-1.22185743e+00 -8.52259338e-01 5.67561209e-01 -7.72472695e-02
1.26757431e+00 -8.34897161e-02 -8.62257257e-02 9.34140205e-01
-2.54084080e-01 7.79960811e-01 1.23552632e+00 -6.82068586e-01
2.25219666e-03 -5.26357628e-02 3.63889307e-01 7.47317016e-01
4.44738746e-01 -2.42815137e-01 -5.16374409e-01 2.24059120e-01
6.76679850e-01 -3.72097105e-01 1.70151878e-04 3.53679895e-01
-9.29123640e-01 8.54674935e-01 2.88633704e-01 4.42006946e-01
-3.53285670e-02 -1.86947942e-01 4.37912494e-01 2.40404606e-01
4.32993591e-01 3.60948682e-01 -8.62056851e-01 -1.20218895e-01
-4.94563222e-01 -3.98789644e-01 6.88627958e-01 1.19040883e+00
5.12102783e-01 1.73625425e-01 2.03759804e-01 1.18696940e+00
1.16410725e-01 4.28786725e-01 4.40824717e-01 -6.41553462e-01
9.88680363e-01 5.36091745e-01 -3.20304215e-01 -3.57845455e-01
-6.70159757e-01 -3.38309199e-01 -7.07624018e-01 -2.47998267e-01
8.07153881e-01 -5.08853495e-01 -9.89458859e-01 1.97040772e+00
4.27955717e-01 -2.34478772e-01 2.86134213e-01 7.14738309e-01
9.32758808e-01 6.44308627e-01 4.38641012e-01 -1.69047207e-01
1.56761777e+00 -9.44184244e-01 -4.90445375e-01 -5.84766269e-01
1.10115194e+00 -8.28206241e-01 1.54792714e+00 6.76254854e-02
-1.09980106e+00 -3.96661967e-01 -9.65462923e-01 -4.38259989e-01
-3.24844509e-01 4.11415964e-01 8.10185790e-01 5.06410420e-01
-8.84235442e-01 4.11335588e-01 -1.21465147e+00 -4.22689199e-01
3.64098549e-01 4.92665023e-01 -5.69351494e-01 -3.43936294e-01
-1.05728459e+00 6.07154489e-01 7.10220993e-01 9.98415872e-02
-4.77890968e-01 -4.19062614e-01 -9.36941683e-01 -3.19629465e-03
5.21570504e-01 -9.97283533e-02 1.28524065e+00 -5.02445936e-01
-1.43197727e+00 9.78035986e-01 -9.43331569e-02 -1.87028795e-01
2.55229086e-01 -5.71069479e-01 -3.34527135e-01 -1.09294146e-01
8.13816637e-02 6.37971699e-01 2.43227676e-01 -8.04865301e-01
-7.61907816e-01 -3.85105044e-01 3.49765986e-01 1.77487895e-01
-3.23907197e-01 6.69999242e-01 -8.28532875e-01 -6.81357980e-01
1.37610719e-01 -9.44778323e-01 -2.90107667e-01 -7.99040079e-01
-3.71087432e-01 -3.81493896e-01 2.30384871e-01 -1.02043653e+00
1.14245343e+00 -2.14998579e+00 -2.37173513e-01 -3.21664900e-01
-1.58471435e-01 5.26714206e-01 -2.49509588e-01 2.54689097e-01
5.84413111e-02 3.87026906e-01 -5.44176102e-01 -4.93970662e-01
-1.53508812e-01 3.40775192e-01 1.42264590e-01 3.07808101e-01
5.54127634e-01 8.44183981e-01 -6.76772058e-01 -6.18227005e-01
-1.25417069e-01 4.04229760e-01 -4.93876994e-01 3.55717480e-01
-8.20611790e-02 6.59272671e-01 -3.37338030e-01 8.63881230e-01
6.41559839e-01 1.71189960e-02 5.99855840e-01 -2.85158306e-02
-1.93414167e-01 7.72503197e-01 -1.04011917e+00 1.86678302e+00
-6.71695352e-01 1.77436054e-01 1.77871615e-01 -7.51218021e-01
7.98210442e-01 2.17448771e-01 -3.90254259e-02 -8.82263899e-01
3.13309759e-01 3.54786962e-01 3.48816097e-01 -9.83691588e-02
2.27389261e-01 -5.38187586e-02 -3.86166811e-01 5.75637162e-01
2.72793233e-01 3.68852228e-01 5.05688250e-01 2.48257920e-01
1.25652575e+00 1.56230047e-01 4.63385522e-01 -3.55930537e-01
4.74069655e-01 4.52393293e-02 1.03610837e+00 4.01255369e-01
-2.12637082e-01 2.59366632e-01 3.81488830e-01 -2.70034701e-01
-9.81319666e-01 -9.41382170e-01 -8.41548666e-02 1.60771787e+00
-2.37651691e-01 -5.06948173e-01 -7.25696445e-01 -9.27776933e-01
-4.47209239e-01 6.35621607e-01 -3.54536250e-02 2.36354873e-01
-1.31044710e+00 -1.15498209e+00 8.15221786e-01 8.66201758e-01
5.24035096e-01 -1.27640474e+00 -5.36867559e-01 1.68477967e-01
-1.79897666e-01 -1.66905010e+00 -3.15646291e-01 5.38864851e-01
-8.53249609e-01 -7.73802817e-01 -3.16918045e-01 -1.02584171e+00
7.98738420e-01 -5.34124151e-02 1.31531072e+00 2.42798284e-01
-3.69972736e-02 -4.05653059e-01 -6.01407230e-01 -3.66059095e-01
-5.41108310e-01 5.77420712e-01 2.90741414e-01 -5.72154045e-01
3.52330297e-01 -4.62112695e-01 -3.31605196e-01 1.01220921e-01
-7.67211258e-01 1.08648382e-01 8.39076757e-01 6.49704695e-01
5.80152512e-01 -2.45236337e-01 5.46749353e-01 -1.34092712e+00
4.25061226e-01 -3.43463451e-01 -8.99935067e-01 1.21186532e-01
-4.10049349e-01 3.21435481e-01 9.48516309e-01 -1.80355251e-01
-1.19229412e+00 2.77274191e-01 -5.23350000e-01 2.18127698e-01
-4.41609472e-01 5.39305508e-01 -4.82118994e-01 3.60044241e-01
4.18731123e-01 -2.93484449e-01 -6.48969769e-01 -1.05848849e+00
3.81018043e-01 6.59991741e-01 5.49745202e-01 -9.43957567e-01
7.57834435e-01 1.31914899e-01 -2.69602448e-01 -6.08875513e-01
-9.76332724e-01 -3.16724032e-01 -1.01340199e+00 5.20365953e-01
7.01956987e-01 -1.08980811e+00 -2.36017227e-01 1.23086862e-01
-1.09193766e+00 -5.93592167e-01 1.46483453e-02 5.80384851e-01
3.95293394e-03 4.10593987e-01 -1.08923686e+00 -5.31848431e-01
-6.91458762e-01 -1.03993511e+00 5.87787628e-01 1.89447060e-01
1.72240764e-03 -8.07919264e-01 -8.38731322e-03 3.41436446e-01
1.80144966e-01 -1.00420713e-01 1.11249006e+00 -1.09137166e+00
-3.26804847e-01 1.01657994e-01 -4.12386477e-01 4.11641777e-01
1.54166266e-01 -2.62011558e-01 -9.56019640e-01 -3.93835694e-01
-1.26210645e-01 -4.72895980e-01 6.90836608e-01 1.54689461e-01
7.43829846e-01 -1.24104112e-01 -8.36431012e-02 7.83731639e-01
1.37573361e+00 1.49981976e-01 1.72685415e-01 3.63473445e-01
9.15633023e-01 5.95031440e-01 7.52922952e-01 2.57057905e-01
5.04265368e-01 4.35035765e-01 -4.42412049e-02 -1.40750125e-01
-3.29706877e-01 -2.96267122e-01 5.65529168e-01 1.53200877e+00
-3.98485400e-02 -3.05655092e-01 -1.30110347e+00 6.30237699e-01
-1.57668924e+00 -1.84787750e-01 -2.51934901e-02 2.04009652e+00
1.09274352e+00 3.10183316e-01 1.02892667e-02 1.87832285e-02
7.99587369e-01 -1.33993765e-02 -4.41788137e-01 -6.71091557e-01
-1.05068900e-01 5.02191305e-01 5.11188805e-01 5.46785891e-01
-1.19082355e+00 1.67800605e+00 5.69509172e+00 7.42538929e-01
-1.00061607e+00 4.09956217e-01 6.22485399e-01 6.05090579e-04
4.28183787e-02 6.52760044e-02 -1.32953119e+00 1.79256648e-01
1.28398836e+00 1.33785203e-01 2.80239761e-01 7.31146216e-01
-2.65099835e-02 -3.05399429e-02 -9.55044568e-01 6.76563203e-01
-2.99702734e-01 -8.46264184e-01 -1.10363916e-01 -2.13488266e-01
3.27653080e-01 6.47019446e-01 -1.94017664e-01 4.82223421e-01
6.71467960e-01 -9.47434604e-01 4.23860669e-01 -4.54201102e-01
1.09314847e+00 -8.06252003e-01 7.79875994e-01 4.77753282e-01
-1.24317658e+00 1.15348011e-01 -6.62009180e-01 -2.22800314e-01
2.37789750e-01 5.49319983e-01 -1.01845837e+00 4.61122990e-01
6.36547208e-01 2.99477577e-01 -8.11310172e-01 5.82812548e-01
-7.86319494e-01 9.18807447e-01 -6.07909143e-01 7.06977025e-02
2.08688155e-01 -1.15459420e-01 2.77395129e-01 1.70029056e+00
1.21523373e-01 2.01081559e-01 2.50793129e-01 4.50694591e-01
-4.40159202e-01 5.53882182e-01 -3.23325932e-01 -2.20830783e-01
6.02195919e-01 1.54358816e+00 -1.06472695e+00 -1.21804319e-01
-5.48702955e-01 7.82535255e-01 9.26390707e-01 -1.82174072e-02
-8.45280647e-01 -3.63745868e-01 5.57298481e-01 -1.40342847e-01
1.81367353e-01 -6.49099469e-01 -3.06557268e-01 -1.35083365e+00
8.38829875e-02 -8.45904827e-01 8.35770845e-01 -1.34477600e-01
-1.06290758e+00 9.39154863e-01 1.67560615e-02 -6.19112492e-01
-1.47948861e-01 -7.97829151e-01 -3.90403420e-01 9.07138705e-01
-1.62534356e+00 -1.30616903e+00 9.14874673e-03 5.92994034e-01
7.81795084e-01 -8.08288082e-02 8.67759824e-01 5.11965394e-01
-1.01852107e+00 9.33547795e-01 -2.46608093e-01 7.09412754e-01
7.36554384e-01 -1.33468080e+00 1.00808775e+00 1.51724446e+00
4.67634737e-01 7.27063537e-01 2.94733137e-01 -6.46087229e-01
-1.30804682e+00 -1.11442518e+00 9.71517622e-01 -4.46577132e-01
6.15810335e-01 -6.38920784e-01 -9.93892610e-01 7.65877664e-01
9.73025411e-02 6.12724908e-02 7.61486113e-01 4.62073952e-01
-5.65601289e-01 1.72060743e-01 -9.17645991e-01 6.56705856e-01
1.35048413e+00 -3.38644624e-01 -4.38092321e-01 2.14977682e-01
8.64454865e-01 -4.86550748e-01 -9.33861077e-01 5.14822423e-01
2.28962108e-01 -5.97613633e-01 5.97567022e-01 -6.64646268e-01
3.31703693e-01 -1.31510556e-01 -3.51561040e-01 -1.21831667e+00
-2.80778170e-01 -5.06877124e-01 2.38444507e-01 1.59471297e+00
8.74332130e-01 -6.42356992e-01 6.47264600e-01 5.91196537e-01
-3.00464869e-01 -5.62915623e-01 -7.89309025e-01 -7.64193594e-01
3.78274947e-01 -5.01567721e-01 5.12281597e-01 1.03581297e+00
-8.80614743e-02 8.14395189e-01 -9.32703391e-02 3.78067255e-01
4.89406168e-01 1.16157420e-01 5.08280814e-01 -9.08243179e-01
-4.82247233e-01 -8.66360813e-02 4.14820500e-02 -6.95908368e-01
3.58220249e-01 -1.16440976e+00 4.77532633e-02 -1.41375887e+00
1.43850997e-01 -7.59156108e-01 -4.23486263e-01 9.29315686e-01
-4.01168317e-01 1.37576550e-01 3.99493217e-01 1.29533306e-01
-5.40369689e-01 4.33940887e-02 7.87792504e-01 3.01192790e-01
-2.87044019e-01 -2.20503986e-01 -8.12673569e-01 9.86591399e-01
1.35437953e+00 -8.12063813e-01 -1.52769819e-01 -1.15779924e+00
1.66425020e-01 -1.06774800e-01 -3.63097727e-01 -8.42695296e-01
8.21083188e-02 -5.43045290e-02 2.02586323e-01 -3.04180145e-01
1.10263847e-01 -3.58630270e-01 -3.13003898e-01 3.87618184e-01
-8.54993388e-02 5.56911945e-01 4.37947750e-01 -2.00346392e-02
-1.41043559e-01 -3.53944838e-01 8.83880436e-01 -4.05567616e-01
-7.96920598e-01 2.88305253e-01 -5.95887266e-02 5.12285173e-01
5.31883478e-01 2.62930393e-01 -4.23785061e-01 1.24814630e-01
-5.19675732e-01 1.27527580e-01 4.13508117e-01 3.85029197e-01
2.90240049e-01 -8.12412322e-01 -8.80240381e-01 2.61291265e-01
5.10940664e-02 2.83154786e-01 -1.71800509e-01 5.69772661e-01
-6.66330814e-01 4.41302568e-01 -3.26603323e-01 -7.06280917e-02
-1.33212984e+00 4.46027517e-01 -1.39829516e-01 -6.11660600e-01
-6.07897401e-01 1.05379868e+00 1.64519981e-01 -7.67728806e-01
1.51265571e-02 -4.22259927e-01 -2.19198853e-01 -1.93626449e-01
3.94542456e-01 1.15252987e-01 2.64694422e-01 -6.97759688e-01
-5.71443975e-01 2.22063646e-01 -3.65532219e-01 -1.39431804e-01
1.45986724e+00 -1.93314459e-02 -8.15086663e-02 1.54906645e-01
9.70896900e-01 4.09864664e-01 -9.24109221e-01 -3.97153914e-01
5.27767599e-01 -1.45243570e-01 -8.83725435e-02 -7.80162096e-01
-9.73856568e-01 1.01920724e+00 1.97097406e-01 -2.54421979e-01
1.17587709e+00 2.99176455e-01 1.03158867e+00 5.55822432e-01
3.82696390e-01 -1.08140171e+00 -4.45697010e-01 9.35596704e-01
4.59272087e-01 -1.37885118e+00 -8.55894685e-02 -7.17153490e-01
-7.35809982e-01 6.48411274e-01 6.73082471e-01 3.76780108e-02
5.04590690e-01 6.41664863e-01 3.80934358e-01 7.21511170e-02
-6.95256174e-01 -5.56321383e-01 -1.86379150e-01 4.80932623e-01
8.88438284e-01 2.15774536e-01 -5.16074121e-01 8.69619250e-01
-4.46970463e-01 -5.08840561e-01 3.80823106e-01 1.05752909e+00
-2.26833627e-01 -1.47223437e+00 -2.79979762e-02 1.85358092e-01
-1.11762285e+00 -5.02764404e-01 -1.67656526e-01 8.55314851e-01
-1.15804717e-01 9.96796846e-01 -1.17661409e-01 -1.02711782e-01
3.54316533e-01 2.29265943e-01 4.16110575e-01 -1.05606461e+00
-6.19496703e-01 2.00780660e-01 4.93285179e-01 -4.76751387e-01
-2.05697507e-01 -5.65038860e-01 -1.59210503e+00 -1.58076733e-01
-1.81605861e-01 6.46440014e-02 6.26165807e-01 7.67975986e-01
3.05434316e-01 2.77593642e-01 2.81283379e-01 -3.62296730e-01
-4.58021402e-01 -1.12585950e+00 -3.67242962e-01 3.85184109e-01
-1.88962653e-01 -3.74172807e-01 -1.67973638e-01 2.23050900e-02] | [10.467031478881836, 9.869367599487305] |
f1fb1dbe-8b45-42a0-a43f-6c0c07c45270 | schooling-to-exploit-foolish-contracts | 2304.10737 | null | https://arxiv.org/abs/2304.10737v1 | https://arxiv.org/pdf/2304.10737v1.pdf | Schooling to Exploit Foolish Contracts | We introduce SCooLS, our Smart Contract Learning (Semi-supervised) engine. SCooLS uses neural networks to analyze Ethereum contract bytecode and identifies specific vulnerable functions. SCooLS incorporates two key elements: semi-supervised learning and graph neural networks (GNNs). Semi-supervised learning produces more accurate models than unsupervised learning, while not requiring the large oracle-labeled training set that supervised learning requires. GNNs enable direct analysis of smart contract bytecode without any manual feature engineering, predefined patterns, or expert rules. SCooLS is the first application of semi-supervised learning to smart contract vulnerability analysis, as well as the first deep learning-based vulnerability analyzer to identify specific vulnerable functions. SCooLS's performance is better than existing tools, with an accuracy level of 98.4%, an F1 score of 90.5%, and an exceptionally low false positive rate of only 0.8%. Furthermore, SCooLS is fast, analyzing a typical function in 0.05 seconds. We leverage SCooLS's ability to identify specific vulnerable functions to build an exploit generator, which was successful in stealing Ether from 76.9% of the true positives. | ['Aquinas Hobor', 'Tamer Abdelaziz'] | 2023-04-21 | null | null | null | null | ['feature-engineering'] | ['methodology'] | [-3.87791246e-02 4.92891550e-01 -7.07489848e-01 -2.58142233e-01
-6.52046263e-01 -1.01532650e+00 4.67164487e-01 5.05869985e-02
-6.92717955e-02 3.84078741e-01 -1.09752730e-01 -1.45627248e+00
2.08442792e-01 -1.13712883e+00 -5.96197844e-01 -2.62428463e-01
-4.98479992e-01 6.63000345e-01 4.38453436e-01 -1.73239738e-01
3.49946499e-01 3.85870725e-01 -8.14834952e-01 3.72513115e-01
7.41766453e-01 7.92455792e-01 -7.09688365e-01 8.08496237e-01
-8.63501877e-02 1.22808504e+00 -7.59506345e-01 -8.70582759e-01
4.22290474e-01 -4.80163321e-02 -9.04705524e-01 -8.32074404e-01
-7.14817494e-02 -3.63341808e-01 -2.49609262e-01 1.15966487e+00
2.58755922e-01 -7.23890245e-01 3.52279216e-01 -1.56439185e+00
-2.28668615e-01 1.22881591e+00 -2.22264633e-01 1.85528472e-01
1.44548997e-01 6.84644520e-01 1.24580848e+00 -2.32746243e-01
7.70996928e-01 9.29151654e-01 8.40786457e-01 5.37392974e-01
-1.39881098e+00 -9.66621339e-01 -6.69869423e-01 -1.77589849e-01
-9.53402460e-01 -3.60511571e-01 8.86558414e-01 -4.82510179e-01
1.65176535e+00 -3.55909392e-02 4.27244484e-01 1.13760102e+00
5.21261632e-01 5.17671585e-01 1.06365597e+00 -2.02232033e-01
4.52001899e-01 6.38909489e-02 3.63147318e-01 8.56748819e-01
5.36832511e-01 7.46961117e-01 8.86033177e-02 -8.89829457e-01
2.76697695e-01 -1.31502068e-02 3.07162911e-01 -3.41233492e-01
-7.88950801e-01 1.30950308e+00 -9.96196270e-02 2.73197442e-01
-5.48211932e-02 3.52492511e-01 1.12854195e+00 7.85043955e-01
3.21512818e-02 7.84768701e-01 -1.05468607e+00 -6.74286187e-01
-7.59229183e-01 -2.24986479e-01 1.47627044e+00 7.60093331e-01
7.34539509e-01 6.46749616e-01 5.83673000e-01 -1.48367919e-02
2.05396041e-01 3.72828275e-01 1.76894262e-01 -1.08178341e+00
4.23330486e-01 8.95040512e-01 -4.81511205e-01 -8.04817080e-01
-4.00988966e-01 -5.47108471e-01 -2.33130962e-01 5.98282814e-01
5.20892262e-01 -5.09272397e-01 -6.59100890e-01 1.40851426e+00
-2.44021565e-01 -1.62115261e-01 1.78851083e-01 -8.70008245e-02
4.54349697e-01 1.50344253e-01 -2.40130857e-01 -4.31598872e-02
9.18764651e-01 -5.99745452e-01 -1.42168373e-01 2.86911670e-02
1.19732738e+00 -3.27117771e-01 8.86628032e-01 5.20506680e-01
-6.98657632e-01 2.64958292e-01 -1.02742708e+00 7.25877464e-01
-7.02653348e-01 -7.10884035e-01 1.21714318e+00 1.62940395e+00
-9.26549911e-01 6.66381478e-01 -7.11088479e-01 2.54522711e-01
7.08060801e-01 7.40434647e-01 -1.53588891e-01 3.53210181e-01
-1.13251996e+00 5.20357311e-01 6.46037400e-01 -7.28422463e-01
-1.23078060e+00 -6.38351977e-01 -1.00988078e+00 1.50387257e-01
7.12169528e-01 7.52731413e-02 1.28726447e+00 -9.98986304e-01
-1.32205999e+00 5.50233603e-01 3.75393569e-01 -7.69133389e-01
2.48695254e-01 4.08999115e-01 -7.70383537e-01 2.38994509e-01
-1.34181872e-01 -4.22223397e-02 6.89215302e-01 -1.06250811e+00
-2.74668008e-01 1.53991962e-02 2.54112512e-01 -1.01513803e+00
-1.35714576e-01 3.02222162e-01 2.21949011e-01 -4.20169234e-01
-4.54309344e-01 -8.82571161e-01 -1.13583051e-01 -1.01420331e+00
-5.07204711e-01 -1.55751169e-01 8.65054786e-01 -7.35859692e-01
1.02488351e+00 -1.87166083e+00 -6.59670055e-01 1.01155436e+00
4.11832064e-01 5.70847154e-01 -3.34661663e-01 5.21648586e-01
-4.01844561e-01 3.87031317e-01 -4.22462910e-01 2.87751049e-01
3.42064887e-01 1.36925206e-01 -3.59495461e-01 2.32088730e-01
9.64572579e-02 1.38090801e+00 -1.11938083e+00 -3.38871717e-01
-1.57179281e-01 9.23092514e-02 -8.21185529e-01 1.08254887e-01
-4.86168772e-01 8.65984857e-02 -1.38758585e-01 1.36789691e+00
2.56718695e-01 -1.01253062e-01 6.34190857e-01 1.74968168e-01
2.10779712e-01 5.92626274e-01 -5.48929989e-01 1.12943506e+00
-4.80104476e-01 5.82307518e-01 -1.53821498e-01 -1.10036707e+00
1.10420489e+00 2.98818946e-01 6.32466018e-01 -8.05114388e-01
4.00755852e-01 4.44832176e-01 2.65463144e-01 -4.72733885e-01
-8.90209675e-02 1.98492497e-01 -3.20282668e-01 1.29497647e+00
2.39116058e-01 2.77381718e-01 1.59410343e-01 3.71538907e-01
1.84586632e+00 -2.96184253e-02 2.17421234e-01 -2.88194835e-01
3.45074207e-01 6.39529005e-02 7.11987138e-01 8.97988439e-01
-2.14899912e-01 -1.33732706e-01 1.40759182e+00 -8.92955601e-01
-9.93918121e-01 -8.72253299e-01 2.60509938e-01 1.04505992e+00
-5.24670482e-01 -6.80434942e-01 -9.13294375e-01 -1.73077750e+00
2.67507583e-01 8.81272972e-01 -4.19390470e-01 -3.98857176e-01
-6.19055092e-01 -5.72375476e-01 1.30334377e+00 5.24555743e-01
3.28434765e-01 -1.47457600e+00 -4.27383453e-01 2.04763249e-01
1.96273759e-01 -8.73865902e-01 -2.39662409e-01 4.85657603e-01
-8.14932227e-01 -1.84018385e+00 2.26880237e-01 -7.59101212e-01
5.53428233e-01 -5.17046571e-01 1.40579045e+00 3.06891471e-01
-3.43455434e-01 1.18956342e-01 -1.04592748e-01 -1.86930284e-01
-1.14713585e+00 3.08686614e-01 -7.99806640e-02 -3.15089345e-01
8.63427281e-01 -9.32191312e-01 -7.61157125e-02 3.06212574e-01
-8.34624529e-01 -8.08542311e-01 5.78737915e-01 8.00426066e-01
-6.52898774e-02 2.69148409e-01 7.95029700e-01 -1.32636070e+00
6.47545874e-01 -5.94536781e-01 -1.10890889e+00 1.45329669e-01
-1.43911600e+00 1.67501777e-01 1.02623117e+00 -1.10534899e-01
-5.68027020e-01 2.38733478e-02 -2.37295404e-01 -1.13691963e-01
2.83336006e-02 3.49162847e-01 -1.79332182e-01 -4.87851769e-01
8.70967090e-01 1.16493598e-01 4.24205065e-01 -3.12553167e-01
1.54578045e-01 6.93347871e-01 4.55029607e-01 -5.13306916e-01
1.11528969e+00 1.10888213e-01 -1.11251950e-01 -1.41295642e-01
-1.07108511e-01 -9.66972858e-02 -3.74531895e-01 1.26844391e-01
3.32351178e-01 -9.99417529e-02 -1.41329944e+00 4.44410980e-01
-7.98860490e-01 -7.86707699e-01 -1.87117085e-01 3.18810791e-02
-4.29785192e-01 6.63010716e-01 -8.68376493e-01 -6.58235610e-01
-8.22250366e-01 -1.26228118e+00 2.59255826e-01 -2.19293460e-01
-4.50528800e-01 -1.22568190e+00 2.92237371e-01 3.86708707e-01
5.73476672e-01 5.65204203e-01 1.14129066e+00 -1.58293939e+00
-2.18980402e-01 -5.96859574e-01 -1.76331773e-01 5.22416353e-01
1.78655386e-01 2.85271909e-02 -7.89393723e-01 -5.09529591e-01
-1.79402292e-01 -5.45315027e-01 6.84938312e-01 -1.70224011e-01
9.92555201e-01 -7.08612859e-01 -7.96045735e-02 8.02895963e-01
1.17768991e+00 7.01376677e-01 6.64790273e-01 6.23247206e-01
5.53095639e-01 3.89499247e-01 1.79103941e-01 1.15822814e-01
9.91027951e-02 1.78604811e-01 7.55171120e-01 5.96703626e-02
4.10808086e-01 -1.41537875e-01 5.90554833e-01 3.55835319e-01
-1.10274471e-01 3.97295564e-01 -1.40356696e+00 2.28645295e-01
-1.46862924e+00 -1.10822272e+00 4.54704389e-02 2.15675378e+00
9.20830607e-01 7.33400464e-01 5.35367012e-01 2.47964069e-01
4.58090216e-01 1.04909353e-01 -4.97222066e-01 -9.62141097e-01
1.15062773e-01 8.11498046e-01 9.33671415e-01 5.80001652e-01
-1.09251881e+00 8.89641285e-01 6.79076004e+00 7.34324574e-01
-9.47012365e-01 -1.43698715e-02 4.32339936e-01 1.28333211e-01
-5.87138712e-01 5.08828461e-01 -4.91147518e-01 6.25694573e-01
1.39464664e+00 -6.11402728e-02 9.05711591e-01 1.27204835e+00
-4.33480084e-01 3.67296427e-01 -8.36161435e-01 5.37274003e-01
-1.51376456e-01 -1.67891967e+00 -4.71618205e-01 4.06587243e-01
5.18407106e-01 4.14102107e-01 -2.18723387e-01 6.18514776e-01
1.19802177e+00 -1.14660728e+00 2.21801758e-01 -9.09771174e-02
9.55662131e-01 -1.36569679e+00 1.16733265e+00 9.04159248e-02
-7.38732874e-01 -5.02440333e-01 1.16499014e-01 2.84370422e-01
-1.39052168e-01 5.19157529e-01 -1.24045491e+00 2.40898475e-01
4.72416401e-01 4.57994282e-01 -5.36624670e-01 4.64427054e-01
-3.25273603e-01 1.16972041e+00 -2.36580357e-01 8.91904011e-02
3.22618574e-01 1.15252092e-01 4.01590168e-01 1.50827658e+00
-2.71659106e-01 -3.69210392e-01 3.16776000e-02 7.60455430e-01
6.51183655e-04 -3.49373728e-01 -6.95640326e-01 -5.74569345e-01
4.79943573e-01 1.10934401e+00 -6.84051156e-01 -4.81313646e-01
-4.52486694e-01 1.97637588e-01 1.12618022e-02 2.18179315e-01
-7.10508883e-01 -1.15799165e+00 3.42794210e-01 2.99375840e-02
3.00861001e-01 -1.48674741e-01 -3.65458339e-01 -1.00583851e+00
-2.90691167e-01 -1.52761984e+00 7.18982637e-01 5.62422909e-02
-8.52042198e-01 6.06402755e-01 -2.84131050e-01 -9.92098153e-01
-7.17968941e-01 -7.14066863e-01 -1.02929628e+00 3.42537612e-01
-1.45347488e+00 -1.05141437e+00 2.80207932e-01 5.78083217e-01
-2.57619262e-01 -1.09354019e+00 1.09033811e+00 4.49956879e-02
-5.11393428e-01 9.50906992e-01 -8.48163515e-02 7.68284917e-01
2.66990036e-01 -1.39854097e+00 1.09498942e+00 6.76177740e-01
-4.41562608e-02 8.87531817e-01 3.47601563e-01 -1.09055531e+00
-1.62485266e+00 -8.59351039e-01 8.40662718e-01 -3.28229964e-01
1.15339887e+00 -4.66434479e-01 -7.49813199e-01 9.85165715e-01
-2.41601318e-02 1.02387726e-01 9.38666940e-01 5.28474292e-03
-1.27427113e+00 -1.33380845e-01 -1.47052085e+00 3.28813314e-01
9.14319873e-01 -8.31733882e-01 -4.51111048e-01 2.53681839e-01
7.03836858e-01 4.92315814e-02 -1.15526426e+00 3.51961076e-01
7.55059838e-01 -1.26478648e+00 8.68680954e-01 -8.94056439e-01
2.21683994e-01 1.83052346e-01 5.30791171e-02 -8.23475778e-01
-1.26576290e-01 -1.31527913e+00 -7.18780577e-01 1.03612554e+00
8.36156964e-01 -1.36737764e+00 1.20422733e+00 4.82433408e-01
1.03678569e-01 -6.64588869e-01 -6.65552318e-01 -1.15537012e+00
5.43832928e-02 -6.60337627e-01 1.05061948e+00 1.37376022e+00
6.57074690e-01 -2.21305326e-01 -5.29131219e-02 -1.19582385e-01
9.56285954e-01 1.68903753e-01 8.53609920e-01 -1.13572061e+00
-8.58291388e-01 -4.66931283e-01 -5.25703192e-01 -1.71081617e-03
4.98617709e-01 -1.08701432e+00 -5.60289383e-01 -7.31163442e-01
1.36175230e-02 -3.88289183e-01 -3.92645687e-01 1.30740726e+00
5.36884606e-01 2.32380792e-01 -3.24402481e-01 7.86631182e-02
-3.27883482e-01 -2.18668461e-01 3.94308954e-01 -4.93314058e-01
-3.43217760e-01 2.04276487e-01 -7.42082000e-01 7.43867099e-01
1.42030144e+00 -7.36723661e-01 -2.21849263e-01 2.19109342e-01
4.71392006e-01 -8.91471580e-02 1.91529348e-01 -5.04924178e-01
-6.64784908e-02 -2.61285812e-01 -9.56557170e-02 -3.76907945e-01
-3.68203312e-01 -5.76210976e-01 7.80305937e-02 1.12901413e+00
-1.29996568e-01 8.43479391e-03 8.76553059e-02 1.33310810e-01
1.55144051e-01 -4.06449348e-01 4.99193221e-01 -2.63539195e-01
-5.55693090e-01 1.06740721e-01 -4.79238868e-01 2.06154615e-01
9.27783787e-01 -1.43079281e-01 -6.54742718e-01 -3.50872815e-01
-3.84858996e-01 2.10431572e-02 6.76075399e-01 3.94239396e-01
4.62640435e-01 -8.47329199e-01 -2.99251527e-01 5.99838078e-01
1.12495005e-01 -5.40238917e-01 -4.89737123e-01 5.69611430e-01
-7.11677372e-01 4.23419118e-01 -2.69948810e-01 -1.75724864e-01
-1.18069685e+00 7.79263139e-01 2.74818540e-01 -5.13255596e-01
-6.47856534e-01 2.45719537e-01 -6.18598998e-01 -8.83640110e-01
2.35150501e-01 1.03717670e-01 -7.03694746e-02 -3.60926419e-01
4.90601957e-01 4.49799836e-01 8.31702352e-03 -1.70350596e-01
-5.88631332e-01 2.73517132e-01 -1.59993485e-01 -6.48026355e-03
1.46304321e+00 9.94873643e-01 -4.54211682e-01 -3.02151144e-01
1.39992023e+00 3.78326923e-01 -8.95008922e-01 -7.55365891e-03
6.34710908e-01 -2.66670763e-01 -3.53204682e-02 -1.00190771e+00
-1.33290660e+00 5.41429520e-01 6.92210793e-02 5.00265777e-01
6.83088660e-01 -5.75993843e-02 1.06279659e+00 7.77375162e-01
6.48236930e-01 -9.23964083e-01 9.11683738e-02 6.84415042e-01
1.45966634e-01 -9.67357576e-01 -2.30408981e-01 -1.59898371e-01
-3.18808079e-01 1.27797818e+00 3.34844500e-01 -2.42185935e-01
4.43290174e-01 9.69588518e-01 3.69477607e-02 -4.83812422e-01
-5.98820031e-01 3.06767464e-01 -2.92099625e-01 8.71177197e-01
8.26389529e-03 8.67410153e-02 1.38875633e-01 6.72300637e-01
-3.81034285e-01 4.36947756e-02 5.52256525e-01 1.18070817e+00
-2.84446061e-01 -1.34496891e+00 -1.94586307e-01 6.13078892e-01
-8.84610176e-01 -1.62172958e-01 -6.50006413e-01 7.85947561e-01
-2.72459745e-01 8.38835537e-01 -1.50798276e-01 -1.00508595e+00
4.28361027e-03 3.42582643e-01 1.09103508e-02 -4.25210923e-01
-1.02870095e+00 -3.37057799e-01 6.43648922e-01 -9.58321810e-01
2.83817321e-01 -4.05380785e-01 -1.30323422e+00 -7.05513000e-01
-1.77390218e-01 4.14085835e-01 5.54171562e-01 7.48689711e-01
3.50168943e-01 5.41602559e-02 1.13680255e+00 -2.36225784e-01
-6.61836147e-01 -5.58401287e-01 -3.53449017e-01 1.48812115e-01
1.48905933e-01 -1.97918147e-01 -6.71635926e-01 -3.27465773e-01] | [6.871028423309326, 7.426784515380859] |
30cec8ff-c663-481b-b376-f66dfd35856f | variational-autoencoding-molecular-graphs | 2307.00623 | null | https://arxiv.org/abs/2307.00623v1 | https://arxiv.org/pdf/2307.00623v1.pdf | Variational Autoencoding Molecular Graphs with Denoising Diffusion Probabilistic Model | In data-driven drug discovery, designing molecular descriptors is a very important task. Deep generative models such as variational autoencoders (VAEs) offer a potential solution by designing descriptors as probabilistic latent vectors derived from molecular structures. These models can be trained on large datasets, which have only molecular structures, and applied to transfer learning. Nevertheless, the approximate posterior distribution of the latent vectors of the usual VAE assumes a simple multivariate Gaussian distribution with zero covariance, which may limit the performance of representing the latent features. To overcome this limitation, we propose a novel molecular deep generative model that incorporates a hierarchical structure into the probabilistic latent vectors. We achieve this by a denoising diffusion probabilistic model (DDPM). We demonstrate that our model can design effective molecular latent vectors for molecular property prediction from some experiments by small datasets on physical properties and activity. The results highlight the superior prediction performance and robustness of our model compared to existing approaches. | ['Shigehiko Kanaya', 'Naoaki Ono', 'Daiki Koge'] | 2023-07-02 | null | null | null | null | ['drug-discovery', 'property-prediction', 'transfer-learning', 'molecular-property-prediction'] | ['medical', 'medical', 'miscellaneous', 'miscellaneous'] | [ 1.39519006e-01 -1.87535748e-01 -5.36832809e-01 -1.90040529e-01
-6.53798699e-01 -3.00929248e-01 6.29797995e-01 4.56414893e-02
-6.79032058e-02 9.94145572e-01 4.05929148e-01 -2.06402406e-01
-2.75018662e-01 -1.02294922e+00 -8.40115309e-01 -1.31018448e+00
6.75452352e-02 3.59357536e-01 8.00849274e-02 4.04111482e-02
1.80751026e-01 4.51632112e-01 -1.13109362e+00 2.27620482e-01
1.17783988e+00 6.16676331e-01 3.55721831e-01 1.71499044e-01
-9.30103362e-02 6.82990015e-01 -4.72019702e-01 -2.80756623e-01
-2.62191951e-01 -4.51308012e-01 -3.19947720e-01 -2.66705066e-01
-1.26449019e-01 -4.53277409e-01 -5.52311122e-01 1.01790214e+00
6.51366115e-01 1.86792254e-01 1.34041905e+00 -9.40451503e-01
-1.18682826e+00 4.10399079e-01 -1.14285193e-01 -2.96099007e-01
1.82229117e-01 1.53793946e-01 9.95609224e-01 -8.11558306e-01
6.59030557e-01 1.31992483e+00 6.62328064e-01 6.77445471e-01
-1.59081411e+00 -6.72676086e-01 -1.80633944e-02 3.03085506e-01
-1.64333022e+00 -4.04390067e-01 8.83571148e-01 -6.71192169e-01
1.00019574e+00 -7.86947235e-02 4.40046251e-01 1.62185764e+00
6.98209763e-01 7.61059403e-01 6.42185390e-01 1.52566627e-01
7.17493176e-01 -1.12241162e-02 -3.07181269e-01 5.73140979e-01
1.06580108e-01 1.45291209e-01 -5.03620625e-01 -7.39995956e-01
8.24734271e-01 5.43459535e-01 -2.97248483e-01 -5.14643669e-01
-9.07953382e-01 1.38630068e+00 4.80463505e-01 4.58895694e-03
-7.26635993e-01 3.40768367e-01 5.44413459e-03 -3.11374426e-01
4.59516555e-01 2.17211425e-01 -3.11196595e-01 -3.78958546e-02
-8.69923234e-01 2.60749042e-01 5.85721672e-01 7.97086298e-01
5.37891209e-01 2.12009370e-01 -3.01324606e-01 6.26530051e-01
9.35068011e-01 4.69659775e-01 3.52202028e-01 -7.38481104e-01
-8.24751239e-03 2.92216361e-01 2.64949389e-02 -8.58643711e-01
-4.47355695e-02 -3.63932341e-01 -9.79502916e-01 -1.22728594e-01
1.06837926e-02 4.93406951e-02 -9.84758615e-01 1.76970446e+00
3.20602089e-01 6.41835108e-02 2.89833635e-01 6.94631696e-01
1.00009406e+00 1.19185102e+00 6.54383123e-01 -5.37019074e-01
9.50170636e-01 -4.40066844e-01 -9.00040805e-01 4.53409255e-01
4.76076037e-01 -6.44398451e-01 6.44939542e-01 3.62094551e-01
-6.63060844e-01 -3.41178149e-01 -1.03218293e+00 -6.14040419e-02
-2.71962464e-01 -6.79944158e-02 1.07232392e+00 6.70972764e-01
-5.17694592e-01 1.08280993e+00 -1.19195402e+00 1.46804601e-01
6.52622700e-01 5.15422046e-01 -2.82561570e-01 -4.21673656e-02
-1.16087043e+00 3.91263664e-01 3.92967850e-01 1.10986814e-01
-1.28982270e+00 -7.46383846e-01 -7.62438059e-01 2.68089712e-01
-1.08377501e-01 -1.00596666e+00 7.36791551e-01 -4.60492373e-01
-2.00680423e+00 -7.88045675e-02 -3.62880141e-01 -2.40700871e-01
1.62781522e-01 3.31428163e-02 -1.36193365e-01 2.08783783e-02
-8.78708363e-02 6.41169667e-01 8.44685078e-01 -1.14065516e+00
5.57944477e-02 -1.73569918e-01 -2.82345086e-01 -1.81875378e-01
-4.16413128e-01 -3.55532497e-01 -4.39738818e-02 -6.66058064e-01
5.56275062e-02 -7.34263599e-01 -4.96643305e-01 -2.36528173e-01
-4.82385010e-01 -6.05350673e-01 6.51762664e-01 -4.26080287e-01
1.27907276e+00 -1.97723114e+00 4.72338438e-01 1.70983598e-01
4.41613138e-01 1.16591074e-01 -1.59671187e-01 7.77417600e-01
-6.06063455e-02 3.16448569e-01 -1.72722965e-01 -1.22492656e-01
2.21670032e-01 3.16999584e-01 -4.33262974e-01 4.62914407e-01
2.01122537e-01 9.76507366e-01 -8.95368040e-01 -3.13904375e-01
7.07936734e-02 1.14958262e+00 -8.94712806e-01 2.90815443e-01
-5.72180986e-01 3.58648390e-01 -1.06112397e+00 5.96385360e-01
7.70025611e-01 -4.61580902e-01 3.41685921e-01 -3.64512712e-01
7.88762197e-02 1.54348060e-01 -7.60681808e-01 1.57851255e+00
-4.65537570e-02 1.57548204e-01 -5.74896276e-01 -8.22642565e-01
9.70123291e-01 4.60290492e-01 6.85252666e-01 -3.33150715e-01
-9.73976590e-03 5.50529622e-02 1.53136075e-01 -3.46450925e-01
6.06476925e-02 -4.75454688e-01 2.49782294e-01 1.68043911e-01
2.76549131e-01 1.85513780e-01 -1.41735911e-01 1.78701743e-01
8.95251274e-01 3.57979000e-01 8.29974785e-02 -7.72265345e-02
3.63427699e-01 -2.68631428e-01 8.52673292e-01 5.31796992e-01
-1.86132416e-02 4.30200875e-01 7.16858566e-01 -3.51800114e-01
-1.12649846e+00 -1.24382687e+00 -5.50811708e-01 7.30822802e-01
-1.05839029e-01 -9.16785300e-01 -6.45908535e-01 -5.39404809e-01
2.68959273e-02 5.10818541e-01 -6.08337700e-01 -6.40352309e-01
3.47684883e-02 -1.34983897e+00 2.36704871e-01 5.15625179e-01
1.58283412e-02 -7.60948777e-01 3.47704947e-01 6.13962591e-01
2.36907840e-01 -6.39801681e-01 -2.33486831e-01 2.59610683e-01
-9.06363070e-01 -7.81443119e-01 -7.58958459e-01 -4.66480881e-01
5.21925330e-01 -1.56088382e-01 6.63579524e-01 -4.24469531e-01
-1.07626267e-01 -9.98888910e-02 -1.84421524e-01 -3.71474117e-01
-5.35713792e-01 9.83213354e-03 3.06765437e-01 9.64088086e-03
6.10312700e-01 -9.00657654e-01 -9.93312061e-01 1.27875537e-01
-1.00357938e+00 -7.75073543e-02 7.19265938e-01 1.00259805e+00
9.66976345e-01 3.90076488e-02 7.59815812e-01 -7.50559807e-01
8.22189808e-01 -7.67209768e-01 -5.92694044e-01 1.57166749e-01
-6.67150199e-01 5.44697940e-01 6.66851103e-01 -5.92228055e-01
-9.34636772e-01 1.71421856e-01 -6.44633651e-01 -5.06269515e-01
-1.69137701e-01 7.14010417e-01 -4.56857413e-01 -1.39026260e-02
4.96573001e-01 4.93545890e-01 -1.28775969e-01 -5.95630765e-01
4.31255966e-01 6.69701278e-01 -1.55300096e-01 -6.55467868e-01
4.89141166e-01 3.21740359e-01 3.61362815e-01 -8.92658532e-01
-5.98634601e-01 -1.27540365e-01 -5.69016337e-01 3.42974871e-01
9.97056961e-01 -9.32145894e-01 -1.08600342e+00 1.50517058e-02
-1.30237114e+00 6.04513623e-02 -2.86709226e-04 8.42381120e-01
-6.10806465e-01 5.49689591e-01 -8.29703033e-01 -6.44958973e-01
-4.56822217e-01 -1.52605772e+00 1.07078505e+00 1.73102066e-01
-9.76589769e-02 -1.11847484e+00 4.21669424e-01 6.70853108e-02
3.44300866e-01 3.58317316e-01 1.20705676e+00 -4.46649253e-01
-8.41075361e-01 -1.80870220e-01 8.54149088e-02 2.37013787e-01
3.38007122e-01 3.15454841e-01 -8.86285245e-01 -2.15663463e-01
-2.06243739e-01 -1.81957185e-01 1.07024717e+00 8.46850514e-01
1.25394773e+00 -2.91097224e-01 -6.36011004e-01 9.24501181e-01
1.33401227e+00 5.13100386e-01 7.47837007e-01 -1.20295584e-01
9.32421207e-01 1.59858420e-01 1.55804902e-01 6.05157971e-01
9.83211175e-02 5.04526556e-01 3.12606573e-01 1.00354431e-02
2.78520882e-01 -6.50772333e-01 4.93332416e-01 9.61795628e-01
-3.68893981e-01 -3.07350367e-01 -5.74770212e-01 5.40989712e-02
-1.90365994e+00 -1.15216553e+00 -2.23232433e-01 1.95998943e+00
1.01777041e+00 -1.25167072e-01 -1.21792242e-01 -3.22193176e-01
2.69759953e-01 1.20337076e-01 -7.90217757e-01 -1.99289575e-01
-8.17794800e-02 3.42446536e-01 2.60190278e-01 3.90315235e-01
-1.03257084e+00 1.04500258e+00 7.13261700e+00 1.27350521e+00
-1.05619216e+00 1.30649507e-01 5.58080375e-01 1.68594643e-01
-7.43360460e-01 2.17605364e-02 -9.83776450e-01 7.01112866e-01
1.11228800e+00 -7.28244185e-02 -3.67851965e-02 9.63930905e-01
4.62210923e-01 3.70974660e-01 -1.01959729e+00 1.07470751e+00
-2.58791476e-01 -1.71957529e+00 6.12615347e-01 5.82101703e-01
8.65648150e-01 -1.34839892e-01 3.60611260e-01 1.42072633e-01
2.83110291e-01 -1.37584269e+00 9.41833556e-02 7.91924775e-01
6.09437823e-01 -1.01189983e+00 5.12081206e-01 3.15733403e-01
-8.63872707e-01 3.14439267e-01 -9.20877099e-01 3.46035540e-01
-2.65245084e-02 6.09499335e-01 -7.73682892e-01 4.10902381e-01
1.44880012e-01 9.71763372e-01 -1.12505004e-01 9.90740895e-01
-2.47146606e-01 8.27228487e-01 -2.31268570e-01 -3.21646869e-01
1.45845354e-01 -6.29515171e-01 2.79185355e-01 9.57722068e-01
3.44834805e-01 1.11690396e-02 6.85409009e-02 1.47628915e+00
-1.28562927e-01 2.14462891e-01 -6.03804290e-01 -6.23668194e-01
2.39437819e-01 9.30456579e-01 -4.41252112e-01 -9.92258415e-02
-2.46133745e-01 9.06196654e-01 8.17202777e-02 7.45145738e-01
-9.83735144e-01 -3.29941571e-01 9.39122856e-01 -4.44625318e-02
3.86167437e-01 -4.89826977e-01 2.12480500e-01 -1.24260294e+00
-3.62747699e-01 -6.47867024e-01 -6.39527338e-03 -5.59562445e-01
-1.53196859e+00 5.37905931e-01 -1.82278216e-01 -1.11748266e+00
-7.11575672e-02 -8.93920422e-01 -4.37275380e-01 9.97195661e-01
-1.30954015e+00 -1.08303344e+00 2.67412245e-01 4.38424259e-01
2.34111741e-01 -2.89561868e-01 1.06111181e+00 3.38659465e-01
-8.17264199e-01 3.82217586e-01 8.87467980e-01 -2.55526364e-01
5.42365849e-01 -1.20072663e+00 4.65694964e-02 2.71689594e-01
9.82008502e-02 1.12825811e+00 7.94654489e-01 -8.06901157e-01
-1.52724659e+00 -1.00544608e+00 4.95984077e-01 -5.06773591e-01
5.82519770e-01 -4.43448305e-01 -1.05285704e+00 3.68469745e-01
-4.97133806e-02 -2.56278545e-01 1.51023293e+00 1.20938495e-01
-1.11642838e-01 2.62500972e-01 -7.87334561e-01 5.01047254e-01
8.27583253e-01 -6.23735547e-01 -3.11879605e-01 6.82256639e-01
7.85945117e-01 3.40111293e-02 -1.39744031e+00 3.65885496e-01
6.34732366e-01 -5.66175640e-01 1.23546064e+00 -1.00029075e+00
5.38325965e-01 -4.47629452e-01 -1.50787607e-01 -1.27995574e+00
-6.22572124e-01 -6.75538003e-01 -4.72418904e-01 1.07368398e+00
3.79885197e-01 -4.10187334e-01 1.04718947e+00 5.08626521e-01
2.07222123e-02 -9.00763094e-01 -7.87568510e-01 -5.65035164e-01
3.33947450e-01 -3.14227343e-01 6.76501930e-01 8.06332827e-01
-1.47935987e-01 5.04337132e-01 -5.82241416e-01 1.52700111e-01
7.32347488e-01 7.36199617e-02 4.09125865e-01 -1.27680147e+00
-3.87648702e-01 -3.16724718e-01 -4.85982925e-01 -1.32139051e+00
3.24653804e-01 -8.76147389e-01 -2.61496007e-01 -1.53304982e+00
5.25736928e-01 -1.61290482e-01 -4.46694911e-01 2.45731607e-01
-1.76604390e-01 -2.24484041e-01 -3.57254446e-01 3.89723986e-01
-4.68953192e-01 1.48089075e+00 1.42560375e+00 -3.56448829e-01
-2.99668759e-01 -1.65784694e-02 -5.73695362e-01 5.02295077e-01
5.76166153e-01 -6.90428853e-01 -5.68884969e-01 4.40942124e-02
2.76066869e-01 -1.37394024e-02 8.79754126e-02 -6.05489671e-01
1.59097195e-01 -4.75065619e-01 7.55984664e-01 -5.08285701e-01
4.84411180e-01 -4.44403112e-01 3.78976554e-01 4.43064630e-01
-1.11750491e-01 -5.80314994e-01 -7.38689080e-02 9.56159592e-01
-3.57709497e-01 -5.18720187e-02 5.23494899e-01 7.33531043e-02
-2.99314827e-01 9.07619596e-01 -5.32918215e-01 -6.40758812e-01
7.89156914e-01 -2.10986882e-01 -1.31248131e-01 -2.26923972e-01
-9.09023464e-01 -9.39098299e-02 2.89586484e-01 1.10997476e-01
1.00243962e+00 -1.61698377e+00 -4.46288943e-01 8.65820199e-02
1.09082848e-01 -8.20252895e-02 3.75492007e-01 6.10749245e-01
-5.75312972e-01 6.10549331e-01 -5.21599874e-02 -5.78939617e-01
-9.64436829e-01 9.11475539e-01 4.37968433e-01 -5.47575280e-02
-2.38206029e-01 7.53248155e-01 6.72432363e-01 -1.70044750e-01
-1.79207847e-02 -9.14895982e-02 -2.28292719e-01 -4.38369215e-02
3.98276538e-01 -9.17650536e-02 -3.34638655e-01 -5.64588845e-01
-2.53170699e-01 5.37210226e-01 -2.00367421e-01 4.16586876e-01
1.73488855e+00 3.04139107e-01 -2.21247389e-03 2.33196281e-02
1.43427002e+00 -1.12145804e-01 -1.49385345e+00 -1.10994264e-01
-2.48607352e-01 -2.05868348e-01 4.26601052e-01 -3.33940208e-01
-6.34274364e-01 1.13660681e+00 6.36710405e-01 -2.01066226e-01
5.37910104e-01 2.09634528e-02 8.16523731e-01 4.46639776e-01
1.08217075e-01 -8.51677537e-01 8.67849365e-02 2.68373877e-01
7.67507553e-01 -1.18134189e+00 1.06592499e-01 -4.00330871e-01
-4.46765929e-01 1.30029130e+00 1.45288229e-01 -3.20707262e-02
8.14796686e-01 -1.39577687e-01 -3.39339077e-01 -3.74810129e-01
-8.68555605e-01 -1.20899417e-01 5.08173883e-01 7.53012240e-01
7.38030195e-01 1.07590161e-01 -3.10878187e-01 8.93602371e-01
1.94172338e-01 -2.76737828e-02 3.84925827e-02 7.65969872e-01
-4.34122384e-01 -1.59374380e+00 -1.07149079e-01 1.66870430e-01
-4.40580100e-01 -3.53271842e-01 -2.96186417e-01 1.90078884e-01
1.27849087e-01 6.53749883e-01 -2.50135243e-01 -4.26435649e-01
7.81784505e-02 1.11653522e-01 4.30525750e-01 -5.00321865e-01
8.09224769e-02 4.52010930e-01 -4.64062303e-01 -3.48333120e-01
-3.34659219e-01 -4.15307909e-01 -9.88280416e-01 -4.55140561e-01
-4.16917950e-01 4.94916737e-01 5.84000826e-01 7.57087052e-01
6.72662318e-01 6.58417761e-01 6.51984274e-01 -7.16494083e-01
-6.21996105e-01 -7.97851741e-01 -7.65338719e-01 3.86376590e-01
1.56864017e-01 -8.40013266e-01 -1.29271477e-01 1.74807832e-01] | [5.129680156707764, 5.830612659454346] |
a692ac37-e873-4547-a222-df507dfbff66 | multi-view-human-body-mesh-translator | 2210.01886 | null | https://arxiv.org/abs/2210.01886v1 | https://arxiv.org/pdf/2210.01886v1.pdf | Multi-view Human Body Mesh Translator | Existing methods for human mesh recovery mainly focus on single-view frameworks, but they often fail to produce accurate results due to the ill-posed setup. Considering the maturity of the multi-view motion capture system, in this paper, we propose to solve the prior ill-posed problem by leveraging multiple images from different views, thus significantly enhancing the quality of recovered meshes. In particular, we present a novel \textbf{M}ulti-view human body \textbf{M}esh \textbf{T}ranslator (MMT) model for estimating human body mesh with the help of vision transformer. Specifically, MMT takes multi-view images as input and translates them to targeted meshes in a single-forward manner. MMT fuses features of different views in both encoding and decoding phases, leading to representations embedded with global information. Additionally, to ensure the tokens are intensively focused on the human pose and shape, MMT conducts cross-view alignment at the feature level by projecting 3D keypoint positions to each view and enforcing their consistency in geometry constraints. Comprehensive experiments demonstrate that MMT outperforms existing single or multi-view models by a large margin for human mesh recovery task, notably, 28.8\% improvement in MPVE over the current state-of-the-art method on the challenging HUMBI dataset. Qualitative evaluation also verifies the effectiveness of MMT in reconstructing high-quality human mesh. Codes will be made available upon acceptance. | ['Si Liu', 'Luoqi Liu', 'Zitian Wang', 'Xuecheng Nie', 'Xiangjian Jiang'] | 2022-10-04 | null | null | null | null | ['human-mesh-recovery'] | ['computer-vision'] | [ 1.85315624e-01 -8.55860412e-02 5.02624810e-02 4.51853499e-02
-8.13530743e-01 -3.93937886e-01 1.54448912e-01 -3.82859230e-01
-6.80822209e-02 3.59891683e-01 3.76946360e-01 5.47063589e-01
-4.01816033e-02 -6.61935389e-01 -7.18361855e-01 -5.18493474e-01
3.43978465e-01 6.39419377e-01 6.43559322e-02 -2.79837608e-01
-1.80432931e-01 3.47197980e-01 -1.60044265e+00 2.62016118e-01
5.27016938e-01 8.42565656e-01 2.70524532e-01 5.91567516e-01
5.65253079e-01 2.24969700e-01 -9.31805745e-02 -6.59620047e-01
3.80858183e-01 -2.44765118e-01 -5.94203591e-01 4.60066289e-01
8.07827055e-01 -6.48631692e-01 -5.59464693e-01 8.14287841e-01
7.54268944e-01 -6.69556782e-02 5.36973894e-01 -8.01183462e-01
-2.39293158e-01 -6.00100495e-02 -9.68284667e-01 -7.32479692e-02
8.77547324e-01 -5.56681044e-02 7.87565291e-01 -1.37294626e+00
9.94699538e-01 1.07856059e+00 8.29807937e-01 4.73189950e-01
-1.20152736e+00 -5.95362663e-01 2.37018373e-02 -2.39478927e-02
-1.46373177e+00 -5.78104675e-01 8.74997795e-01 -5.41571438e-01
6.58909202e-01 4.39035982e-01 1.02018106e+00 9.99883592e-01
2.76382238e-01 7.45793343e-01 9.53083277e-01 -2.31693879e-01
-3.08226168e-01 -3.98088455e-01 -3.33505124e-01 9.50932086e-01
3.11933547e-01 1.57829240e-01 -9.32300806e-01 -2.51724690e-01
1.05395472e+00 1.07844964e-01 -5.24181247e-01 -7.16141760e-01
-1.45291018e+00 4.32475626e-01 2.37850726e-01 -3.18338862e-03
-6.11466527e-01 4.66992781e-02 2.49336600e-01 -8.62197950e-02
6.18555546e-01 -8.91656578e-02 -1.07565582e-01 6.27478659e-02
-9.19749498e-01 4.56981689e-01 3.43465179e-01 9.97511685e-01
4.11530972e-01 2.62699760e-02 -1.64085813e-02 9.80593979e-01
5.51361263e-01 1.04605544e+00 2.08848622e-02 -1.04516840e+00
7.76117623e-01 4.98280108e-01 -2.12289635e-02 -1.43225062e+00
-3.95126581e-01 -2.99322784e-01 -1.15371609e+00 3.76770981e-02
2.74140239e-01 1.07138984e-01 -8.27890992e-01 1.34268975e+00
7.88472533e-01 -8.79386738e-02 -3.01708758e-01 1.29557025e+00
9.07496572e-01 3.60541850e-01 -3.51432055e-01 -1.95814610e-01
1.48253345e+00 -7.81283438e-01 -5.92819870e-01 -2.81818539e-01
3.74624990e-02 -8.84701848e-01 6.00263596e-01 6.98751628e-01
-1.45352113e+00 -6.38623297e-01 -8.76489520e-01 -1.55553281e-01
3.50046337e-01 2.72678286e-01 1.28469527e-01 3.52179527e-01
-7.04919100e-01 4.28259134e-01 -9.78764236e-01 -1.82574093e-01
1.23878337e-01 3.53355169e-01 -8.49307954e-01 -4.47675198e-01
-7.97110856e-01 7.97743261e-01 -7.82703981e-02 5.00811875e-01
-7.38464832e-01 -5.48564792e-01 -1.17300320e+00 -3.63454103e-01
3.81966203e-01 -1.25860035e+00 8.19045544e-01 -3.33720118e-01
-1.32488155e+00 1.10022140e+00 -2.70483464e-01 5.17410636e-02
7.07602620e-01 -5.23142040e-01 -1.17247552e-01 5.79106152e-01
1.89826831e-01 4.73697841e-01 1.05562103e+00 -1.50538695e+00
-3.77994299e-01 -7.62840033e-01 -1.75519675e-01 4.62513894e-01
-5.14238179e-02 -2.19869375e-01 -9.51947927e-01 -1.04754758e+00
6.94607675e-01 -1.18707943e+00 4.96008024e-02 2.79787481e-01
-4.16104794e-01 3.43523204e-01 5.21698296e-01 -1.29709768e+00
1.22190416e+00 -1.81865442e+00 1.00252044e+00 2.17475310e-01
4.34775740e-01 4.04493324e-02 3.40494663e-01 2.70559967e-01
6.77779466e-02 -4.75442141e-01 -2.92340100e-01 -6.89057589e-01
-1.85016245e-01 2.39025339e-01 6.96268156e-02 9.33286130e-01
-3.30637902e-01 8.05296659e-01 -6.63802385e-01 -6.99944794e-01
5.58868885e-01 6.24057829e-01 -6.82153642e-01 3.51690054e-01
2.84318775e-01 8.71422350e-01 -2.89017737e-01 9.11292076e-01
6.94757700e-01 -2.24361137e-01 2.14286804e-01 -6.73171163e-01
1.40136838e-01 -3.95836860e-01 -1.32916141e+00 2.17243981e+00
-4.04977620e-01 3.13970968e-02 4.78786796e-01 -6.67937934e-01
4.23389316e-01 5.91261566e-01 9.62797046e-01 -6.25370026e-01
1.25400469e-01 2.18133539e-01 -3.89462918e-01 -4.91689861e-01
2.45150760e-01 -2.25490645e-01 8.35169107e-03 1.06129386e-01
3.22557949e-02 -1.03301801e-01 -2.11160138e-01 -3.81214209e-02
5.95254302e-01 4.76568460e-01 2.31336027e-01 1.50753632e-01
6.20059907e-01 -2.76340812e-01 4.16938901e-01 3.45328674e-02
1.76106766e-01 1.34570479e+00 -1.88131139e-01 -2.40723714e-01
-1.05079019e+00 -1.04484534e+00 -1.32854506e-02 5.39450705e-01
2.34077245e-01 -5.79654753e-01 -7.11552858e-01 -4.84826595e-01
5.34625165e-02 8.28365907e-02 -6.25407219e-01 1.72436535e-02
-9.25391018e-01 -4.83202994e-01 3.70143980e-01 4.84229028e-01
4.65648383e-01 -6.82287216e-01 -9.16661978e-01 5.18276915e-02
-6.96475029e-01 -1.25422561e+00 -7.25295722e-01 -4.96645600e-01
-9.86455739e-01 -1.18551183e+00 -1.37732768e+00 -4.22546685e-01
7.74529219e-01 3.63939494e-01 9.27165866e-01 2.24975467e-01
-2.68703163e-01 7.00264096e-01 -2.34324753e-01 2.65651435e-01
-1.04593374e-01 -2.96339363e-01 2.15009674e-01 2.01028362e-01
-4.22211230e-01 -6.58834219e-01 -9.74313498e-01 5.16467512e-01
-6.07692063e-01 4.51311290e-01 4.78907645e-01 9.42216098e-01
9.20168936e-01 -2.84057111e-01 -7.16721267e-02 -4.94595468e-01
-1.76794276e-01 -3.57522726e-01 -9.65912268e-02 8.67137760e-02
-5.07052004e-01 -4.21097368e-01 2.85103649e-01 -1.85134023e-01
-9.41703975e-01 2.27483124e-01 -3.09451789e-01 -9.47194278e-01
4.99039628e-02 3.21814626e-01 -3.52671534e-01 -1.01306371e-01
1.44977093e-01 3.39952677e-01 2.47276708e-01 -6.65231049e-01
2.14402184e-01 3.19889337e-01 6.16074920e-01 -5.40241063e-01
7.82253623e-01 8.16833198e-01 1.01029947e-01 -7.35787868e-01
-7.27334738e-01 -5.81984580e-01 -9.17627096e-01 -6.76478148e-01
9.38078642e-01 -1.15181696e+00 -6.87172115e-01 6.67682588e-01
-9.94851232e-01 1.45940959e-01 3.28842178e-02 6.03192866e-01
-8.69009256e-01 7.82895982e-01 -7.46581435e-01 -5.79691410e-01
-6.14515901e-01 -1.14909613e+00 1.64113009e+00 -2.31864631e-01
-3.09508204e-01 -6.89982951e-01 -5.00838533e-02 9.88015890e-01
-1.13024481e-01 4.69354689e-01 5.71779549e-01 1.47872180e-01
-3.37809384e-01 -1.58237129e-01 6.10381067e-02 1.70700610e-01
6.96112774e-03 -3.45594376e-01 -7.90942788e-01 -5.30695558e-01
4.88242954e-02 -8.66297260e-02 4.33848083e-01 5.76296508e-01
6.43073976e-01 -1.51072830e-01 -3.55387419e-01 8.50732863e-01
1.20636356e+00 -3.19597244e-01 5.34356534e-01 2.63735324e-01
1.25897682e+00 6.62446916e-01 8.04665864e-01 6.51499391e-01
6.16034806e-01 1.26917398e+00 5.28584957e-01 -1.06947511e-01
-3.85320425e-01 -4.45405871e-01 1.74460456e-01 1.10457826e+00
-5.86249053e-01 -1.72407888e-02 -9.58840907e-01 4.64148104e-01
-1.82124257e+00 -7.97527730e-01 -3.74833882e-01 2.31493282e+00
5.68726480e-01 -1.63812354e-01 1.80154860e-01 1.93513930e-01
4.81734216e-01 1.07867435e-01 -4.69574273e-01 4.17263955e-01
1.79643705e-01 1.18293054e-01 2.89842695e-01 3.86543840e-01
-8.60876441e-01 4.84244227e-01 5.37348318e+00 6.36836052e-01
-9.51722085e-01 3.05316389e-01 1.30109593e-01 -1.81148529e-01
-9.54352841e-02 -2.50091046e-01 -6.31281495e-01 3.16011459e-01
3.58092815e-01 2.16691107e-01 2.11074099e-01 4.39151138e-01
1.78063270e-02 -2.01283187e-01 -8.76630843e-01 1.46475744e+00
4.48840261e-01 -1.00203252e+00 1.39155269e-01 3.91224593e-01
5.90996921e-01 -2.60997593e-01 -8.48551020e-02 -1.63242102e-01
-3.70323271e-01 -8.93200397e-01 1.15110612e+00 9.05656397e-01
1.03599644e+00 -7.09630251e-01 5.45215547e-01 5.40583313e-01
-1.52878618e+00 3.16241473e-01 -5.94268180e-02 1.85210437e-01
6.27568960e-01 6.09373808e-01 -2.48763993e-01 1.31179368e+00
8.64911020e-01 9.22824860e-01 -4.38047439e-01 6.97005033e-01
1.13735892e-01 2.38640234e-01 -2.98127234e-01 8.59116316e-01
-3.86756510e-01 -2.99916923e-01 7.85977185e-01 7.90693343e-01
4.87078577e-01 4.09536600e-01 3.80571306e-01 6.13800049e-01
2.98338085e-01 2.43910905e-02 -5.95415652e-01 4.91838664e-01
1.08902209e-01 1.15821123e+00 -6.23214483e-01 -2.91559845e-01
-3.13587308e-01 1.33741796e+00 9.30356681e-02 1.05961010e-01
-8.37233305e-01 3.60007942e-01 3.45083416e-01 6.06652439e-01
3.65146190e-01 -2.88616151e-01 -2.87847340e-01 -1.42745066e+00
3.68317127e-01 -1.03450274e+00 3.57241571e-01 -8.49557519e-01
-1.08381176e+00 4.93012995e-01 2.72939801e-01 -1.64386749e+00
-2.93177992e-01 -3.35804909e-01 1.67858407e-01 7.43231773e-01
-9.19261575e-01 -1.61915493e+00 -3.85621220e-01 7.80042529e-01
6.69525385e-01 1.19949289e-01 7.66568005e-01 5.73097050e-01
-3.21141034e-01 4.26618308e-01 -3.19666713e-01 7.21577927e-02
7.15497732e-01 -8.95286083e-01 2.80308843e-01 7.22758472e-01
1.21463574e-01 6.10130489e-01 6.81915700e-01 -1.03746057e+00
-1.94835234e+00 -8.12579751e-01 4.22079831e-01 -5.88284850e-01
-2.27940902e-02 -1.76581442e-01 -7.07381546e-01 6.20321631e-01
-1.61170363e-01 2.05767661e-01 4.13033575e-01 -2.80523419e-01
-9.07689035e-02 1.98110342e-01 -1.11682677e+00 3.20563644e-01
1.26972687e+00 -3.86297852e-01 -5.29830098e-01 1.34618059e-02
1.48478091e-01 -1.23942196e+00 -1.33962512e+00 7.81615078e-01
9.47328985e-01 -1.20374036e+00 1.45954061e+00 -5.52486777e-02
6.25293136e-01 -3.23348939e-01 -4.77797717e-01 -9.05785978e-01
-2.13937417e-01 -3.56149346e-01 -5.71730852e-01 1.02138937e+00
-3.16321045e-01 -2.74615467e-01 7.15558112e-01 3.85533333e-01
-1.10217996e-01 -9.52104390e-01 -1.19613767e+00 -3.47693294e-01
-2.63967335e-01 -3.07630688e-01 1.38636798e-01 7.39090621e-01
-2.55545467e-01 1.17630437e-01 -8.28804195e-01 4.35531914e-01
1.01150692e+00 2.70474583e-01 9.29677725e-01 -1.10706747e+00
-4.62267935e-01 8.01467896e-02 -3.66018474e-01 -1.17492700e+00
-1.20313458e-01 -7.24759758e-01 4.67148721e-02 -1.53999686e+00
2.70022422e-01 -1.85158476e-01 2.21458986e-01 1.13143593e-01
-1.63168073e-01 7.92169392e-01 3.27939332e-01 4.14931744e-01
-3.20674568e-01 6.74210429e-01 1.51738536e+00 3.98447104e-02
3.04372609e-01 -7.02046184e-03 -2.57775724e-01 1.02367115e+00
1.05479278e-01 -2.13313952e-01 -1.96539134e-01 -5.88407218e-01
1.80629149e-01 6.48051798e-01 8.17739725e-01 -1.07655489e+00
1.29628405e-01 8.90117437e-02 6.32375419e-01 -8.21502268e-01
8.18922639e-01 -9.37507272e-01 7.94655681e-01 4.02315974e-01
3.24806958e-01 3.22948813e-01 -1.46413734e-02 8.20954919e-01
-1.58376604e-01 7.59405196e-02 7.08305299e-01 -2.46775746e-01
-3.76695126e-01 5.63600063e-01 7.48500451e-02 6.19917363e-02
8.58062923e-01 -4.87086415e-01 3.43563974e-01 -2.92019725e-01
-1.06554914e+00 8.55041370e-02 8.69528234e-01 3.86904359e-01
1.01741719e+00 -1.43028712e+00 -7.20492959e-01 4.05889034e-01
1.01266712e-01 2.93687016e-01 7.66795695e-01 1.29391015e+00
-4.40698028e-01 1.31814733e-01 -1.66462481e-01 -9.50431585e-01
-1.66471231e+00 3.85831922e-01 3.93513292e-01 -2.21535936e-01
-1.23429370e+00 5.28110862e-01 3.17619026e-01 -4.97312725e-01
9.23785288e-03 6.94786087e-02 -9.37697664e-02 5.14768586e-02
2.94338793e-01 5.74995816e-01 2.04537690e-01 -1.43196297e+00
-3.10487658e-01 1.37496924e+00 2.16850311e-01 -2.41094470e-01
1.29980493e+00 -4.85201508e-01 8.03177245e-03 3.37369293e-01
1.01256013e+00 3.71494859e-01 -1.20698500e+00 -1.62011728e-01
-6.06333017e-01 -7.59916246e-01 -7.35431090e-02 -4.79815543e-01
-1.36853755e+00 8.49010348e-01 6.12007976e-01 -5.12877107e-01
1.01947832e+00 -5.59172854e-02 1.17883790e+00 -3.26343209e-01
9.62091327e-01 -7.58895397e-01 5.74782789e-02 7.52954483e-02
1.26898801e+00 -9.55398083e-01 5.97918451e-01 -8.35619628e-01
-5.81109166e-01 1.04161060e+00 5.04592299e-01 -1.01038285e-01
5.85523307e-01 -1.90822817e-02 -2.16501117e-01 -6.94101512e-01
-2.69619286e-01 2.16308627e-02 7.47684121e-01 4.55094397e-01
3.06671709e-01 1.76895335e-02 -2.28771999e-01 3.95615608e-01
-2.02940553e-01 -3.34746055e-02 -2.01398320e-03 9.37256336e-01
6.72231466e-02 -9.49260890e-01 -7.43623972e-01 2.38458753e-01
-4.45013732e-01 1.97833419e-01 -1.29705295e-02 6.90046787e-01
3.25559348e-01 7.63526857e-01 -4.03935373e-01 -4.93290693e-01
7.76992977e-01 -2.76355147e-01 9.26910281e-01 -5.79672277e-01
-5.56295633e-01 4.96757656e-01 1.15616776e-01 -8.39687347e-01
-6.72235727e-01 -8.71359229e-01 -1.11111844e+00 -3.21481556e-01
-3.10639352e-01 -3.58712286e-01 1.77398548e-01 9.46059704e-01
3.26830924e-01 3.69776309e-01 3.18749994e-01 -1.45304096e+00
-4.00411069e-01 -7.01529205e-01 -4.89723146e-01 6.31852210e-01
2.59860814e-01 -1.11248243e+00 -4.16636765e-02 2.62625098e-01] | [7.067468643188477, -1.1307491064071655] |
9e75e0e8-72c7-4013-8389-b70012a3836e | mdpgt-momentum-based-decentralized-policy | 2112.02813 | null | https://arxiv.org/abs/2112.02813v1 | https://arxiv.org/pdf/2112.02813v1.pdf | MDPGT: Momentum-based Decentralized Policy Gradient Tracking | We propose a novel policy gradient method for multi-agent reinforcement learning, which leverages two different variance-reduction techniques and does not require large batches over iterations. Specifically, we propose a momentum-based decentralized policy gradient tracking (MDPGT) where a new momentum-based variance reduction technique is used to approximate the local policy gradient surrogate with importance sampling, and an intermediate parameter is adopted to track two consecutive policy gradient surrogates. Moreover, MDPGT provably achieves the best available sample complexity of $\mathcal{O}(N^{-1}\epsilon^{-3})$ for converging to an $\epsilon$-stationary point of the global average of $N$ local performance functions (possibly nonconcave). This outperforms the state-of-the-art sample complexity in decentralized model-free reinforcement learning, and when initialized with a single trajectory, the sample complexity matches those obtained by the existing decentralized policy gradient methods. We further validate the theoretical claim for the Gaussian policy function. When the required error tolerance $\epsilon$ is small enough, MDPGT leads to a linear speed up, which has been previously established in decentralized stochastic optimization, but not for reinforcement learning. Lastly, we provide empirical results on a multi-agent reinforcement learning benchmark environment to support our theoretical findings. | ['Soumik Sarkar', 'Chinmay Hegde', 'Young M. Lee', 'Aditya Balu', 'Kai Liang Tan', 'Sin Yong Tan', 'Xian Yeow Lee', 'Zhanhong Jiang'] | 2021-12-06 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [-4.91341472e-01 3.41961533e-02 -3.37958664e-01 2.70521432e-01
-1.03023851e+00 -4.55903530e-01 3.98035586e-01 5.40833652e-01
-9.46619630e-01 1.41383934e+00 -2.48393923e-01 -5.75994432e-01
-5.43286502e-01 -6.28739119e-01 -9.14556086e-01 -9.38090980e-01
-6.99949861e-01 5.27843475e-01 1.07585669e-01 -3.29249293e-01
3.25317681e-01 1.53644815e-01 -1.20693195e+00 -5.56978762e-01
1.14866233e+00 1.11473811e+00 2.61290856e-02 8.61115217e-01
3.38708431e-01 9.15677309e-01 -5.92215300e-01 -6.28987476e-02
5.82669735e-01 -5.21975636e-01 -4.85345215e-01 6.65821228e-03
6.06525093e-02 -7.31072426e-01 -8.74101147e-02 1.23605871e+00
6.41473770e-01 4.21341658e-01 4.28534299e-01 -1.15623081e+00
-2.34227091e-01 6.49349988e-01 -7.64936030e-01 2.61231363e-01
1.32100759e-02 4.86400634e-01 8.58920038e-01 -2.77822793e-01
3.92419845e-01 1.22739708e+00 5.51768482e-01 4.60686177e-01
-1.14832926e+00 -5.23516357e-01 6.74300015e-01 -1.30092144e-01
-8.06392252e-01 8.00743103e-02 3.59704971e-01 -2.05260605e-01
7.69375861e-01 -2.07106080e-02 9.70978916e-01 5.91516733e-01
3.14152986e-01 8.28075528e-01 1.58079112e+00 -3.55320334e-01
7.62294650e-01 4.01564501e-02 -4.43005294e-01 9.80756521e-01
3.88927817e-01 5.82777858e-01 -2.71955490e-01 -4.30417180e-01
8.68555248e-01 -5.88351637e-02 -4.16300334e-02 -4.83135194e-01
-9.86777425e-01 1.06896091e+00 1.99766234e-01 -2.50996917e-01
-7.06672728e-01 7.18570709e-01 4.68234152e-01 7.60936201e-01
6.57508373e-01 4.06156123e-01 -6.84259176e-01 -6.06341422e-01
-7.72903860e-01 5.92877328e-01 1.06164885e+00 7.00166225e-01
7.06342697e-01 4.33692932e-01 -2.67881185e-01 3.84170026e-01
3.20549488e-01 8.71696889e-01 3.65172774e-01 -1.42351866e+00
6.03188932e-01 2.74917811e-01 8.48164976e-01 -5.25177002e-01
-3.15442652e-01 -5.40794373e-01 -5.28486311e-01 7.90468335e-01
6.92449093e-01 -9.44512248e-01 -2.83225238e-01 1.79290950e+00
6.81835175e-01 9.77790877e-02 2.14237049e-02 7.43868530e-01
-3.20828974e-01 4.26572442e-01 -2.24577010e-01 -8.32199097e-01
9.75518107e-01 -8.85066986e-01 -3.40185016e-01 -6.70893490e-02
5.79960525e-01 -4.21380550e-01 1.11149371e+00 3.29884887e-01
-1.23374808e+00 5.16392663e-02 -9.51665699e-01 8.88029575e-01
2.24338509e-02 -2.63846070e-02 5.99627852e-01 6.45712376e-01
-1.09887004e+00 1.08316576e+00 -9.99472737e-01 5.79634234e-02
3.34465861e-01 3.65982264e-01 3.51982981e-01 3.93422931e-01
-7.40726948e-01 7.29582191e-01 2.78658241e-01 -1.77796215e-01
-1.50213623e+00 -8.20311606e-01 -3.99473757e-01 -4.74517345e-02
9.02584553e-01 -6.33969069e-01 1.59547579e+00 -7.23728180e-01
-2.36117411e+00 -8.63318592e-02 1.09097108e-01 -8.46074462e-01
1.07956326e+00 -2.44437337e-01 1.99659973e-01 2.66129106e-01
9.35678408e-02 1.44485444e-01 1.00312841e+00 -1.16100335e+00
-1.01776040e+00 -1.63154960e-01 3.65643859e-01 4.71886396e-01
-3.92610610e-01 -9.93915498e-02 2.05968887e-01 -3.26754749e-01
-5.04639983e-01 -9.53106821e-01 -6.98858798e-01 -3.45692366e-01
1.09096309e-02 -2.01188132e-01 4.16811824e-01 -3.26272964e-01
1.17831540e+00 -1.63376629e+00 1.03144459e-01 4.01327670e-01
1.10701710e-01 1.58321247e-01 3.38333771e-02 6.30491495e-01
4.64719564e-01 6.77313730e-02 -1.28860757e-01 -2.19873950e-01
3.17532837e-01 5.92505038e-02 -2.56777346e-01 8.01991165e-01
-2.86844015e-01 7.02135980e-01 -1.36967003e+00 -1.97624981e-01
1.32419959e-01 -1.63574517e-01 -7.94950664e-01 2.46061608e-02
-5.01810491e-01 3.98947805e-01 -7.56304026e-01 3.80497575e-01
3.67858082e-01 -3.84106040e-01 4.63667274e-01 5.27818024e-01
-4.11086053e-01 -1.19272329e-01 -1.41184306e+00 1.15461850e+00
-4.64409888e-01 9.75417271e-02 5.31262994e-01 -1.20750809e+00
5.96433640e-01 3.05845588e-01 1.01845551e+00 -4.56321895e-01
1.32777482e-01 4.82673854e-01 -2.46524334e-01 -1.35337651e-01
3.81308556e-01 -1.34380385e-01 2.49962527e-02 8.03569376e-01
-1.45306081e-01 -2.59038329e-01 5.37966609e-01 4.28787954e-02
1.20259190e+00 1.99151680e-01 2.47609824e-01 -6.75888777e-01
3.56589377e-01 -5.99152073e-02 6.23026073e-01 1.14944863e+00
-5.40913343e-01 -5.47785759e-01 9.49478269e-01 -2.12609425e-01
-1.14352477e+00 -7.45285392e-01 3.71436834e-01 1.32722843e+00
1.38749331e-01 -2.51232386e-01 -6.88915849e-01 -9.56676602e-01
5.15995502e-01 4.34900939e-01 -7.89471865e-01 1.25479087e-01
-5.18891811e-01 -8.56811523e-01 2.43248269e-01 1.68342501e-01
5.71424067e-01 -9.13374841e-01 -8.98242354e-01 5.90991676e-01
4.77780879e-01 -6.19819224e-01 -5.55355310e-01 2.02900559e-01
-9.18549359e-01 -1.15920174e+00 -8.39695632e-01 -3.19279194e-01
6.99517787e-01 -3.63160041e-03 8.31188023e-01 -1.09562822e-01
3.40727478e-01 8.01589906e-01 -1.08336821e-01 -4.72461611e-01
-3.90758991e-01 3.24250609e-02 4.36961770e-01 -2.89383352e-01
-2.61815250e-01 -6.05383694e-01 -9.43665445e-01 1.86034590e-01
-6.00484788e-01 -4.10599858e-01 4.92493749e-01 1.07836139e+00
6.30231142e-01 -3.15598771e-02 7.96244264e-01 -4.73193318e-01
1.28068841e+00 -3.71724874e-01 -1.43518937e+00 1.45764932e-01
-1.03201902e+00 3.46652389e-01 1.20781732e+00 -6.59852624e-01
-6.92026317e-01 -2.79151559e-01 3.20713848e-01 -4.84676570e-01
2.72730470e-01 5.09315550e-01 8.37724090e-01 -2.23147228e-01
6.20391428e-01 4.13950801e-01 5.50400436e-01 -1.60082802e-01
4.00829524e-01 1.44278094e-01 -2.86426451e-02 -1.07621205e+00
5.47734082e-01 5.33679724e-01 2.19340220e-01 -4.24075842e-01
-5.56692541e-01 -1.07285082e-01 2.10721061e-01 -2.56491542e-01
2.21563727e-01 -7.71993041e-01 -1.78818512e+00 2.95514852e-01
-5.33433259e-01 -1.01171792e+00 -6.75226927e-01 6.45059943e-01
-1.13780844e+00 3.46754432e-01 -7.14175284e-01 -1.27913213e+00
-5.84636986e-01 -1.06409335e+00 4.33307052e-01 3.72287720e-01
4.43995744e-01 -1.04248476e+00 4.01304603e-01 -3.14545751e-01
6.82493687e-01 2.34425902e-01 4.84331220e-01 -3.68666887e-01
-3.40652138e-01 8.65105242e-02 8.68333802e-02 3.82985353e-01
-8.82944092e-03 -2.12128937e-01 -1.17777556e-01 -1.00115502e+00
-7.59545015e-04 -6.16756558e-01 5.38318038e-01 6.40848279e-01
7.19430029e-01 -9.16743279e-01 -8.72819275e-02 2.55896866e-01
1.64331222e+00 2.94924796e-01 -8.60022306e-02 7.04215169e-01
2.43280560e-01 4.50036414e-02 8.01799655e-01 1.31464124e+00
4.44235593e-01 2.29616016e-01 6.83158159e-01 1.64806053e-01
5.54802597e-01 -2.77862936e-01 6.75622880e-01 5.86285293e-01
-1.97786793e-01 2.80110151e-01 -5.77806890e-01 5.57190955e-01
-2.21821094e+00 -1.00000024e+00 5.22318304e-01 2.49322224e+00
1.18658447e+00 1.14870586e-01 6.16336107e-01 -3.57232243e-01
4.55368638e-01 -2.35932004e-02 -1.16857827e+00 -4.74808395e-01
2.35351697e-01 3.19028020e-01 1.04608583e+00 6.96389854e-01
-9.42871749e-01 1.00513053e+00 6.16654062e+00 9.98727262e-01
-1.00968969e+00 1.47444859e-01 5.06052256e-01 -4.14735496e-01
-5.67377172e-02 8.22216570e-02 -7.84419477e-01 6.37608588e-01
9.81345832e-01 -6.03134274e-01 1.00134361e+00 1.19068301e+00
8.31342399e-01 -3.77391607e-01 -6.04399025e-01 6.51791632e-01
-5.71056485e-01 -1.32987761e+00 -4.48431969e-01 3.19855392e-01
1.11514044e+00 1.57182053e-01 1.35290474e-01 7.00981438e-01
1.16915488e+00 -6.26433074e-01 5.34944415e-01 2.26242945e-01
3.40144753e-01 -1.23760450e+00 4.64794427e-01 5.49151838e-01
-1.16615021e+00 -6.16882682e-01 -3.84669095e-01 -1.79165766e-01
9.01950598e-02 4.51463789e-01 -7.21799552e-01 4.14975017e-01
6.53991938e-01 2.62405485e-01 2.97632497e-02 1.11333799e+00
-2.94020027e-01 5.88825643e-01 -6.37067020e-01 -6.48702681e-01
6.88564897e-01 -5.13470352e-01 7.17596769e-01 7.16568053e-01
2.22839728e-01 -1.80372834e-01 7.11774945e-01 5.02171099e-01
-2.18619294e-02 3.20213318e-01 -1.67095214e-01 2.06544939e-02
5.55081904e-01 1.14529979e+00 -6.81940675e-01 -4.93974328e-01
-4.94208224e-02 7.04844117e-01 5.79550624e-01 5.34607291e-01
-9.35104728e-01 -4.29008871e-01 8.61399531e-01 -3.12438846e-01
7.75482416e-01 -5.08745730e-01 2.27893218e-01 -9.73295331e-01
1.79708228e-01 -1.00968671e+00 2.94017822e-01 2.35520571e-01
-1.17929113e+00 1.99569389e-01 1.18390396e-02 -1.24914014e+00
-8.21956575e-01 -4.30811018e-01 -4.88030374e-01 4.49415326e-01
-1.66739893e+00 -4.72853154e-01 3.65300030e-01 6.72227442e-01
2.70604879e-01 -2.57153064e-01 6.38125539e-01 -3.32857743e-02
-5.74242413e-01 5.96349657e-01 9.51354563e-01 -2.21185312e-01
4.00465965e-01 -1.56384790e+00 -2.19511181e-01 7.29977250e-01
-5.20711541e-01 3.24786663e-01 9.77311194e-01 -6.14058971e-01
-1.66483617e+00 -9.19970632e-01 -1.73565298e-01 1.51787534e-01
1.09996045e+00 6.06944226e-02 -3.21289062e-01 4.56125915e-01
3.00039202e-01 8.75215754e-02 1.18007503e-01 -1.95648581e-01
2.47604594e-01 -2.91025192e-01 -1.18203831e+00 7.35479057e-01
5.99268556e-01 -1.20412312e-01 7.83195943e-02 4.97614294e-01
5.77883840e-01 -5.82252562e-01 -1.13884068e+00 1.60883710e-01
4.02486712e-01 -9.23040807e-01 5.90862393e-01 -7.43749797e-01
3.00734024e-02 -2.31491283e-01 6.84938999e-03 -1.77627039e+00
8.95256624e-02 -1.46927929e+00 -6.45569563e-01 5.95710337e-01
3.74608696e-01 -1.06244612e+00 9.63461995e-01 2.67240763e-01
8.97212923e-02 -1.17244458e+00 -1.20396602e+00 -1.32359743e+00
5.45873046e-01 -8.84755105e-02 3.01681638e-01 6.53035939e-01
1.68627083e-01 -1.72581881e-01 -5.37898719e-01 1.33369640e-01
1.06796157e+00 1.32994384e-01 7.96803534e-01 -5.58042049e-01
-8.44151437e-01 -7.34954476e-01 2.28668600e-01 -1.25800931e+00
5.17004989e-02 -3.17400247e-01 -4.14492935e-03 -1.35835814e+00
-2.66225152e-02 -5.97624958e-01 -4.27072465e-01 3.10789198e-01
-2.37568066e-01 -4.80736166e-01 3.35722864e-01 3.01982276e-03
-1.02104867e+00 1.10374677e+00 1.37744343e+00 6.30713031e-02
-6.33644462e-01 2.72014409e-01 -5.92562079e-01 6.63865268e-01
1.01210105e+00 -5.48362315e-01 -4.58209276e-01 -3.98362055e-02
2.74547935e-01 4.29167241e-01 3.63635086e-02 -8.45647752e-01
2.39493549e-01 -6.65575385e-01 -2.00289071e-01 -2.31031105e-01
3.02105229e-02 -5.11680424e-01 -4.89430666e-01 9.39141512e-01
-2.46161357e-01 3.52591246e-01 9.65994447e-02 8.30307126e-01
1.93072066e-01 -1.69890627e-01 7.42295623e-01 -2.58051723e-01
-1.62644237e-01 5.22808909e-01 -6.25154734e-01 3.86047304e-01
1.08294284e+00 3.23679149e-01 -3.67435664e-01 -6.40567124e-01
-3.94015729e-01 7.41684616e-01 2.55896151e-01 -2.65864909e-01
4.07705426e-01 -1.15064740e+00 -7.62545288e-01 -3.85757238e-01
-4.83381778e-01 -2.70954728e-01 -8.41447711e-03 1.05698586e+00
-2.65218318e-01 7.97560588e-02 1.16945669e-01 -4.69503701e-01
-6.46249294e-01 4.47226584e-01 5.62763095e-01 -8.39064240e-01
-3.82863790e-01 6.64383113e-01 -6.91911280e-01 -4.08562779e-01
2.00878844e-01 -4.63969648e-01 3.12320262e-01 -1.23438925e-01
4.67863083e-01 7.79452920e-01 -2.44986713e-01 2.45870709e-01
-7.43449330e-02 1.73972800e-01 -9.29851159e-02 -6.63180053e-01
1.41661239e+00 -5.33223264e-02 5.78490309e-02 1.60579368e-01
8.93549919e-01 4.99236304e-03 -2.05541396e+00 -1.92445114e-01
-2.12889314e-02 -3.04950088e-01 -7.48434067e-02 -5.84365547e-01
-9.93229449e-01 2.27585927e-01 7.24019527e-01 5.32580495e-01
7.38667071e-01 -4.52310264e-01 4.11282629e-01 5.96713901e-01
7.13170290e-01 -1.69083023e+00 2.99854785e-01 8.27067018e-01
6.99724734e-01 -1.19380498e+00 2.76739746e-01 4.32694763e-01
-7.49497235e-01 8.92174661e-01 5.28561354e-01 -6.18209600e-01
6.30576611e-01 1.70336977e-01 -1.95365652e-01 4.88036722e-02
-9.76296306e-01 -2.58902490e-01 -3.77233028e-01 2.47985065e-01
-2.14770380e-02 2.42770344e-01 -6.65844202e-01 1.14073768e-01
-5.61182685e-02 2.47184071e-03 4.79692012e-01 1.33838892e+00
-6.55912757e-01 -1.06662750e+00 -2.17591226e-01 5.03341794e-01
-5.49538195e-01 1.15797855e-02 2.24757060e-01 9.46846724e-01
-6.00433052e-01 1.04419684e+00 -2.10904896e-01 1.16024189e-01
1.68397829e-01 -3.00488502e-01 6.09973133e-01 -4.20005210e-02
-6.79425120e-01 2.33001530e-01 -1.43293276e-01 -7.87329316e-01
-3.92846793e-01 -5.99185407e-01 -1.34736872e+00 -5.86366415e-01
-1.71849385e-01 4.21923935e-01 6.20332479e-01 1.03390610e+00
4.18733656e-01 2.35214815e-01 1.09270060e+00 -8.75519216e-01
-1.48238480e+00 -8.84746969e-01 -5.92390180e-01 4.65872884e-02
3.57275397e-01 -7.63732374e-01 -5.77325761e-01 -7.08789706e-01] | [4.126123905181885, 2.5302019119262695] |
389b4b1d-1b10-40c2-942b-3a5e461b902b | improving-scheduled-sampling-for-neural | 2305.15958 | null | https://arxiv.org/abs/2305.15958v1 | https://arxiv.org/pdf/2305.15958v1.pdf | Improving Scheduled Sampling for Neural Transducer-based ASR | The recurrent neural network-transducer (RNNT) is a promising approach for automatic speech recognition (ASR) with the introduction of a prediction network that autoregressively considers linguistic aspects. To train the autoregressive part, the ground-truth tokens are used as substitutions for the previous output token, which leads to insufficient robustness to incorrect past tokens; a recognition error in the decoding leads to further errors. Scheduled sampling (SS) is a technique to train autoregressive model robustly to past errors by randomly replacing some ground-truth tokens with actual outputs generated from a model. SS mitigates the gaps between training and decoding steps, known as exposure bias, and it is often used for attentional encoder-decoder training. However SS has not been fully examined for RNNT because of the difficulty in applying SS to RNNT due to the complicated RNNT output form. In this paper we propose SS approaches suited for RNNT. Our SS approaches sample the tokens generated from the distiribution of RNNT itself, i.e. internal language model or RNNT outputs. Experiments in three datasets confirm that RNNT trained with our SS approach achieves the best ASR performance. In particular, on a Japanese ASR task, our best system outperforms the previous state-of-the-art alternative. | ['Ryo Masumura', 'Tomohiro Tanaka', 'Kohei Matsuura', 'Hiroshi Sato', 'Takanori Ashihara', 'Takafumi Moriya'] | 2023-05-25 | null | null | null | null | ['automatic-speech-recognition'] | ['speech'] | [ 4.56083894e-01 3.68529767e-01 6.22501448e-02 -1.80353269e-01
-7.51917243e-01 -1.36931702e-01 6.55388653e-01 -2.65267700e-01
-4.39181298e-01 6.51019752e-01 4.23206329e-01 -4.45742100e-01
6.27146125e-01 -4.23489690e-01 -9.21495914e-01 -6.85093462e-01
3.19349915e-01 4.16902214e-01 1.32072836e-01 -2.89833933e-01
5.03759272e-02 4.75303799e-01 -1.57203031e+00 4.29935753e-01
7.80712605e-01 7.89129794e-01 6.90782309e-01 7.16287017e-01
-4.18667644e-01 1.16510272e+00 -9.67591345e-01 -2.72138536e-01
3.01412633e-03 -6.87482417e-01 -5.04844546e-01 -2.00925276e-01
-2.72916742e-02 -1.82881385e-01 -4.64476347e-01 1.07274556e+00
4.36809242e-01 1.92671821e-01 4.80590522e-01 -5.82857788e-01
-6.91771686e-01 1.17383170e+00 1.75574813e-02 3.40775728e-01
1.75512508e-01 -8.72788765e-03 6.25892162e-01 -1.33217609e+00
3.56785506e-01 1.32913804e+00 4.88707811e-01 8.63078773e-01
-1.14147961e+00 -4.15248871e-01 2.25265935e-01 6.05622232e-02
-1.13575387e+00 -1.04985428e+00 8.30776393e-01 -1.57523304e-01
1.47667944e+00 3.20618331e-01 3.61745447e-01 1.34957671e+00
1.30349904e-01 1.03933847e+00 9.15604651e-01 -9.05263543e-01
2.69659907e-01 3.18319738e-01 2.68756300e-01 3.06164414e-01
-3.58230978e-01 3.51624280e-01 -6.41255975e-01 1.38838425e-01
7.27981627e-01 -1.59745917e-01 -2.08361343e-01 4.03223872e-01
-7.90566862e-01 5.64854324e-01 2.71612972e-01 5.65883398e-01
-9.61687028e-01 1.16103582e-01 4.37335402e-01 4.21522886e-01
5.86187005e-01 2.80839771e-01 -4.64738220e-01 -3.22486490e-01
-1.13290811e+00 -3.05974245e-01 5.37705362e-01 8.46696675e-01
4.00812924e-01 1.04746890e+00 -2.88394898e-01 1.15021431e+00
3.07089269e-01 4.82597411e-01 1.05420995e+00 -4.20288801e-01
5.13122022e-01 2.97687083e-01 -2.08146811e-01 -4.05697227e-01
1.51176929e-01 -6.49505377e-01 -8.05257797e-01 3.12401971e-04
1.74007252e-01 -5.04116295e-04 -1.33061254e+00 1.54243624e+00
-1.18238635e-01 3.06624830e-01 4.94351417e-01 5.75147212e-01
6.84989810e-01 1.15914714e+00 4.76256087e-02 -5.77283382e-01
9.97242332e-01 -1.05730069e+00 -1.13996911e+00 -5.19315064e-01
8.05920899e-01 -7.04474866e-01 9.82100487e-01 5.43582380e-01
-1.23063457e+00 -7.11695671e-01 -7.16701210e-01 6.75925910e-02
-3.90818805e-01 5.90197325e-01 -2.76781283e-02 5.79356909e-01
-1.15742540e+00 5.52761555e-01 -7.42956460e-01 2.46856571e-03
-1.05939224e-01 1.69755533e-01 -2.65801530e-02 1.42857268e-01
-1.56237388e+00 1.16027546e+00 4.16525424e-01 6.72716796e-01
-9.69109058e-01 -4.72389251e-01 -7.86347628e-01 8.20623040e-02
2.87176430e-01 -1.12353712e-01 1.53295410e+00 -1.36423051e+00
-2.06413412e+00 5.13400972e-01 -8.00409377e-01 -1.01212013e+00
2.71124870e-01 -3.27837855e-01 -6.54286027e-01 -3.09355110e-01
-4.27637130e-01 2.32467383e-01 1.17048323e+00 -9.66999590e-01
-8.74950364e-02 -1.92949161e-01 -4.72782731e-01 9.18895900e-02
-8.38356763e-02 2.43192300e-01 -2.40876693e-02 -8.85349751e-01
2.44792253e-01 -7.91503310e-01 -2.31217116e-01 -9.50708330e-01
-4.24794912e-01 -5.59287190e-01 5.06348014e-01 -9.55929995e-01
1.70105636e+00 -2.21985984e+00 7.66507164e-02 -1.12755168e-02
-3.89914364e-01 9.67208266e-01 -5.61696775e-02 5.45404971e-01
-4.59664881e-01 1.50224417e-01 4.63958681e-02 -6.77617908e-01
-1.95373803e-01 4.28880841e-01 -9.33020473e-01 4.09944393e-02
3.48735809e-01 8.22884679e-01 -8.27930391e-01 -1.31029919e-01
2.23489687e-01 6.88853562e-01 -5.25886007e-02 5.42561054e-01
-3.01717877e-01 1.51444420e-01 -1.72922518e-02 3.00256342e-01
3.04001510e-01 3.05696905e-01 -1.86605901e-01 7.59242103e-02
-1.88049629e-01 1.25357389e+00 -7.76794434e-01 1.21082664e+00
-6.83480024e-01 6.88200712e-01 -4.09481138e-01 -9.92176175e-01
1.40084529e+00 9.28607047e-01 -4.40151602e-01 -5.85211337e-01
8.05218220e-02 5.69007576e-01 9.74148586e-02 -3.19956005e-01
7.58962512e-01 -2.19092682e-01 2.56473511e-01 1.90159157e-01
-6.40498986e-03 -2.72414088e-02 -2.16800168e-01 7.74091259e-02
8.08937013e-01 1.07772455e-01 3.03322256e-01 2.73095131e-01
6.85384393e-01 -3.50501329e-01 5.18483102e-01 8.59654605e-01
1.34610817e-01 7.75258660e-01 3.30368191e-01 -2.15754658e-01
-1.11189651e+00 -8.66484165e-01 1.10812865e-01 1.08020449e+00
-8.09462368e-01 -3.01998466e-01 -6.99844599e-01 -5.24580896e-01
-5.62256873e-01 1.44068229e+00 -3.66732121e-01 -2.66283780e-01
-8.89107287e-01 -2.66352296e-01 7.33389497e-01 5.84950745e-01
6.33324385e-02 -1.60549498e+00 -1.31432667e-01 6.10447466e-01
-1.27870739e-01 -9.51436877e-01 -3.26392978e-01 4.72209394e-01
-9.70849812e-01 -2.37383842e-01 -9.34328079e-01 -6.27485931e-01
6.07194602e-01 -3.97322662e-02 9.62194085e-01 -3.52505706e-02
3.78546029e-01 -1.22198798e-01 -4.90319937e-01 -5.23751438e-01
-1.30261517e+00 8.13805386e-02 2.81512022e-01 1.29095227e-01
4.66327697e-01 -3.97479385e-01 2.03181598e-02 -8.67539719e-02
-7.75534511e-01 -2.05796529e-02 7.63020515e-01 9.32684720e-01
5.37307620e-01 -2.37370968e-01 7.44538784e-01 -1.09361219e+00
6.74328327e-01 -2.94454038e-01 -5.60661077e-01 2.63434678e-01
-4.99314368e-01 3.50942761e-01 9.60436046e-01 -6.82042658e-01
-1.12424314e+00 -5.36543503e-02 -4.44511205e-01 -8.87419462e-01
-1.27522796e-01 6.97902858e-01 9.72239971e-02 5.68387389e-01
6.92229986e-01 7.58926213e-01 -4.05958183e-02 -5.75492918e-01
1.86082840e-01 1.00412250e+00 3.55365962e-01 -1.05593845e-01
3.94132227e-01 -2.51247674e-01 -5.54324925e-01 -1.06794310e+00
-7.09151745e-01 -1.77339807e-01 -3.93256038e-01 -1.98573351e-01
3.35287571e-01 -7.97788799e-01 -9.62549299e-02 5.59758306e-01
-1.70201075e+00 -2.80249476e-01 -5.32261670e-01 5.73362589e-01
-4.01708037e-01 1.74594313e-01 -7.22846568e-01 -1.63439798e+00
-4.15685654e-01 -1.19352579e+00 8.31273437e-01 -1.89938173e-02
-3.15876186e-01 -8.04795861e-01 -4.37309556e-02 -3.84039506e-02
5.58366835e-01 -5.90328515e-01 6.61158621e-01 -1.03031945e+00
-4.19179499e-01 -3.26221496e-01 3.54346871e-01 9.52627301e-01
-1.21150561e-01 1.81651022e-03 -1.39641178e+00 -6.55998215e-02
3.07487190e-01 -1.50606379e-01 9.73616779e-01 4.39371794e-01
9.94897425e-01 -5.80047727e-01 -1.34022400e-01 1.49362609e-02
1.04063547e+00 5.24080098e-01 1.11668181e+00 7.51156546e-03
4.56282705e-01 4.82147813e-01 5.22935331e-01 5.16045503e-02
-1.70515433e-01 5.80125153e-01 7.62640610e-02 6.94217533e-02
-2.46544495e-01 -5.03447294e-01 9.97376919e-01 1.46499228e+00
4.98438664e-02 -4.51905817e-01 -7.67808199e-01 5.86873055e-01
-1.63941157e+00 -9.82468903e-01 -2.21424446e-01 2.47225571e+00
9.69763815e-01 4.47252214e-01 -2.38183320e-01 3.44128579e-01
7.70200193e-01 1.97712004e-01 -3.19178551e-01 -9.14104342e-01
-1.55128717e-01 2.94603378e-01 2.91162670e-01 6.81164384e-01
-4.60288078e-01 1.16614854e+00 6.13555288e+00 9.78159189e-01
-1.27193701e+00 2.05773756e-01 5.04195750e-01 4.90607843e-02
-2.12207705e-01 -1.08918682e-01 -1.15945041e+00 5.33948243e-01
1.63386321e+00 1.16484165e-01 5.36128163e-01 8.75400543e-01
5.55128634e-01 3.62293571e-02 -1.12168825e+00 9.38997746e-01
2.04661861e-01 -9.63725150e-01 1.22857444e-01 -2.05149740e-01
3.00656974e-01 1.22744203e-01 2.50400513e-01 6.29397690e-01
2.00556934e-01 -1.10303307e+00 8.28417957e-01 7.81713903e-01
6.33619070e-01 -5.46705246e-01 8.76666486e-01 5.61518312e-01
-8.61379385e-01 7.48680308e-02 -4.97911006e-01 -7.03217983e-02
2.52837181e-01 7.90317714e-01 -1.45111954e+00 2.63199031e-01
2.67032623e-01 4.10844475e-01 -2.03068182e-01 7.20229506e-01
-5.01984477e-01 1.26723146e+00 -2.86301434e-01 -2.79515564e-01
1.49871379e-01 -3.26619633e-02 8.65735352e-01 1.28697300e+00
4.18137193e-01 -1.85206532e-01 -4.27290022e-01 8.56288075e-01
8.71376600e-03 8.57675001e-02 -7.47646749e-01 -2.93613851e-01
7.52830863e-01 5.70459008e-01 -1.28084242e-01 -5.91666281e-01
1.85490362e-02 1.02694726e+00 4.90022689e-01 5.29271781e-01
-5.93616486e-01 -2.51982749e-01 3.17133456e-01 3.36481258e-02
4.44555908e-01 -9.76357013e-02 -9.63214487e-02 -9.90859807e-01
7.47727603e-02 -9.19221401e-01 -1.10832818e-01 -1.13952208e+00
-1.02549171e+00 1.21599674e+00 -3.29078764e-01 -1.05475724e+00
-9.41343069e-01 -4.25720900e-01 -4.91103798e-01 1.33677721e+00
-1.52706683e+00 -7.50218511e-01 5.02076507e-01 1.18461974e-01
1.16573584e+00 -3.94188493e-01 1.00580633e+00 -3.06354975e-03
-6.17735267e-01 5.78911722e-01 -2.28792354e-02 8.10843930e-02
4.03677315e-01 -1.08528912e+00 6.12428665e-01 1.21104968e+00
2.35200509e-01 6.95377111e-01 8.91142368e-01 -7.20618367e-01
-1.05600619e+00 -1.01084709e+00 1.68283117e+00 -3.09184343e-01
5.22276223e-01 -4.09172148e-01 -1.45545375e+00 9.72481728e-01
2.17962354e-01 -2.61989713e-01 2.35805005e-01 -9.79997441e-02
-1.35360181e-01 -1.50520712e-01 -6.17178917e-01 7.15643525e-01
6.87728167e-01 -8.21781337e-01 -1.02357161e+00 8.09088200e-02
8.79863560e-01 -4.96404469e-01 -4.79458362e-01 1.03683528e-02
1.95300072e-01 -8.58578265e-01 6.19260669e-01 -4.14447904e-01
2.69248217e-01 -2.66182989e-01 -5.42657636e-02 -1.60495365e+00
-7.37864226e-02 -7.77081668e-01 -3.45806301e-01 1.46972418e+00
7.39029586e-01 -7.62782812e-01 3.66712838e-01 4.41011995e-01
-5.05204380e-01 -5.28870225e-01 -1.12663972e+00 -8.88952076e-01
-2.24854663e-01 -8.21243584e-01 5.78153670e-01 5.54945290e-01
-1.79017738e-01 4.03794408e-01 -5.73036373e-01 1.40771374e-01
1.23293482e-01 -6.10537827e-01 3.98638636e-01 -8.09941113e-01
-2.39756376e-01 -3.50239635e-01 -1.48864254e-01 -1.37286878e+00
3.11704278e-01 -7.13319778e-01 4.11602437e-01 -1.20911467e+00
-6.16472304e-01 -4.70163912e-01 -2.90029913e-01 4.06829536e-01
-6.57835901e-02 -1.68603674e-01 2.97418475e-01 1.76445574e-01
-2.61001363e-02 7.57559597e-01 7.96379149e-01 2.22125605e-01
-4.34172988e-01 4.39104289e-01 -1.84722275e-01 5.55638015e-01
8.28776360e-01 -7.40131617e-01 -3.02170902e-01 -2.89866090e-01
7.94894248e-02 2.91212469e-01 1.15516871e-01 -8.57535481e-01
3.80298048e-01 3.28527153e-01 1.59723878e-01 -9.29705441e-01
5.93834639e-01 -7.16259837e-01 2.60364056e-01 4.90835249e-01
-5.75251877e-01 2.22119898e-01 6.22660071e-02 4.43877339e-01
-3.12097937e-01 -8.05574477e-01 5.60150981e-01 -2.24722385e-01
-5.07211924e-01 -3.00080597e-01 -8.95971060e-01 -3.69673491e-01
2.95947313e-01 -3.14484298e-01 8.54291096e-02 -3.68537039e-01
-8.90028179e-01 -2.93439537e-01 -1.05515249e-01 4.57663506e-01
7.79028237e-01 -1.10292768e+00 -7.26501346e-01 5.81771255e-01
-2.64821649e-01 -2.71381829e-02 -9.29064602e-02 8.76113236e-01
8.74636546e-02 6.93046331e-01 5.27909398e-01 -4.50225770e-01
-1.27842569e+00 5.12943149e-01 5.68624377e-01 -4.02286708e-01
-4.53945637e-01 7.52776206e-01 -9.73288044e-02 -3.03803623e-01
4.94747043e-01 -5.54578781e-01 -3.40427041e-01 -5.42284362e-02
5.98969817e-01 3.48771513e-01 4.68818396e-01 -6.17743552e-01
-3.41544263e-02 -1.86859127e-02 -3.59483510e-01 -3.79007161e-01
1.18825698e+00 -2.17505321e-01 -5.55079132e-02 1.19194031e+00
8.70847583e-01 -6.85538873e-02 -8.53752255e-01 -4.76067752e-01
5.04570603e-02 -1.19499750e-01 2.22668260e-01 -7.83886194e-01
-7.51767099e-01 1.04947925e+00 3.28488529e-01 4.88906741e-01
8.59379828e-01 -1.10269688e-01 6.47518694e-01 2.17028499e-01
1.96994603e-01 -1.15845180e+00 -1.53558210e-01 9.81121600e-01
1.15433919e+00 -9.01474178e-01 -6.56143069e-01 -1.57884300e-01
-8.13585103e-01 1.29961145e+00 4.41189796e-01 -1.96507186e-01
4.88591433e-01 1.60882249e-01 1.90988958e-01 2.97867715e-01
-1.07862771e+00 -9.56680179e-02 1.03459731e-01 4.60119694e-01
7.45896757e-01 4.39311191e-02 -1.18536018e-01 5.21979630e-01
-4.78590965e-01 8.41112882e-02 6.23795688e-01 5.20210624e-01
-4.13609982e-01 -1.03953278e+00 -5.58101714e-01 3.03647399e-01
-4.04624254e-01 -5.44933200e-01 -1.05129652e-01 3.95650446e-01
-2.81389385e-01 9.11767483e-01 1.84394047e-01 -2.14807406e-01
4.09626603e-01 6.25297010e-01 1.16397642e-01 -6.56024933e-01
-8.41461480e-01 3.74955118e-01 2.39302531e-01 -1.83686003e-01
-1.93266436e-01 -7.21693397e-01 -1.18017232e+00 2.11612567e-01
-6.13509834e-01 2.97506571e-01 8.94226789e-01 1.06553745e+00
2.38408804e-01 7.61157036e-01 7.42200434e-01 -7.14289844e-01
-8.61549795e-01 -1.43099880e+00 -4.41843271e-01 2.09035370e-02
6.42226756e-01 -2.59031028e-01 -5.68856418e-01 2.73820814e-02] | [14.465428352355957, 6.774567127227783] |
dee54fff-dc60-4e82-9fc6-4d166a13583f | designing-rotationally-invariant-neural | 2108.13993 | null | https://arxiv.org/abs/2108.13993v2 | https://arxiv.org/pdf/2108.13993v2.pdf | Designing Rotationally Invariant Neural Networks from PDEs and Variational Methods | Partial differential equation (PDE) models and their associated variational energy formulations are often rotationally invariant by design. This ensures that a rotation of the input results in a corresponding rotation of the output, which is desirable in applications such as image analysis. Convolutional neural networks (CNNs) do not share this property, and existing remedies are often complex. The goal of our paper is to investigate how diffusion and variational models achieve rotation invariance and transfer these ideas to neural networks. As a core novelty we propose activation functions which couple network channels by combining information from several oriented filters. This guarantees rotation invariance within the basic building blocks of the networks while still allowing for directional filtering. The resulting neural architectures are inherently rotationally invariant. With only a few small filters, they can achieve the same invariance as existing techniques which require a fine-grained sampling of orientations. Our findings help to translate diffusion and variational models into mathematically well-founded network architectures, and provide novel concepts for model-based CNN design. | ['Matthias Augustin', 'Pascal Peter', 'Joachim Weickert', 'Karl Schrader', 'Tobias Alt'] | 2021-08-31 | null | null | null | null | ['novel-concepts'] | ['reasoning'] | [ 8.12792182e-02 -1.48995236e-01 -2.34744355e-01 -2.13575602e-01
3.43422025e-01 -7.53626049e-01 7.51384020e-01 -3.95513654e-01
-4.96513993e-01 4.60920095e-01 1.23625405e-01 -1.82391763e-01
-4.27743971e-01 -9.28082049e-01 -6.73586130e-01 -9.15800929e-01
1.48895010e-01 -2.53893286e-01 3.21254045e-01 -3.55243087e-01
2.34637275e-01 1.24838901e+00 -1.34972811e+00 1.46337926e-01
5.46755910e-01 9.10585046e-01 3.59318429e-03 5.86517513e-01
1.20374955e-01 5.49413323e-01 -1.79339036e-01 -7.68795013e-02
3.28382164e-01 -3.67344499e-01 -7.30556190e-01 -7.06126913e-02
4.76702362e-01 -2.05296680e-01 -6.04856730e-01 1.02207673e+00
2.81302720e-01 4.10672605e-01 8.64128709e-01 -9.79704440e-01
-1.11204720e+00 2.51229167e-01 -3.64159465e-01 1.89427152e-01
-3.70262533e-01 -1.70423388e-01 1.05922854e+00 -7.63641894e-01
7.31217265e-01 1.15416932e+00 6.78455770e-01 9.29779232e-01
-1.44717121e+00 -7.04376698e-02 3.04276347e-01 -4.94774012e-03
-1.30789149e+00 -5.03689647e-01 1.13666308e+00 -3.76105994e-01
9.83320236e-01 3.24789226e-01 8.90444517e-01 7.81244338e-01
4.47131693e-01 5.45468330e-01 7.49698818e-01 -3.09157044e-01
2.90937543e-01 1.14559717e-02 3.01906951e-02 7.51999199e-01
3.50610316e-01 -5.18841147e-02 -4.28641915e-01 2.01841727e-01
1.58474231e+00 2.41755113e-01 -4.94569451e-01 -7.97151744e-01
-1.24620950e+00 8.71427059e-01 8.26668203e-01 5.40396988e-01
-3.81184578e-01 4.00810212e-01 -7.32130720e-04 2.56667644e-01
3.96079332e-01 6.20548129e-01 -3.06263328e-01 4.19257700e-01
-9.43949461e-01 3.90729755e-01 7.05609977e-01 6.45231366e-01
7.82036543e-01 3.29202920e-01 -6.09294884e-02 7.02114344e-01
5.10707438e-01 3.74078363e-01 3.12952936e-01 -1.24897897e+00
-1.98440060e-01 4.26247209e-01 -1.23286538e-01 -1.24421680e+00
-5.65576792e-01 -7.54389882e-01 -1.25921452e+00 3.93877298e-01
2.90921956e-01 -5.44612333e-02 -8.66226554e-01 1.99343514e+00
1.62681460e-01 -2.35324189e-01 -1.62692353e-01 9.56820667e-01
3.96321297e-01 5.23542345e-01 -2.89015561e-01 -2.04404593e-01
1.23288000e+00 -6.38907313e-01 -5.91136932e-01 1.06665827e-01
6.24907434e-01 -6.74385011e-01 7.32035100e-01 1.28648594e-01
-1.24902201e+00 -4.35015082e-01 -1.19911468e+00 -3.70183289e-01
-3.52114081e-01 3.77625860e-02 6.98272169e-01 6.02420032e-01
-1.48983347e+00 8.31174254e-01 -1.16306412e+00 -3.24549317e-01
3.59397918e-01 3.86206627e-01 -2.66963243e-01 3.35382521e-01
-1.00297081e+00 9.02446628e-01 -1.79572567e-01 4.69269425e-01
-3.67439836e-01 -8.18252027e-01 -7.49432445e-01 1.02857962e-01
-2.59600133e-01 -9.36541855e-01 1.10452306e+00 -1.15381467e+00
-1.67956173e+00 6.98022306e-01 -4.20749635e-01 -4.20647025e-01
6.16567016e-01 1.41841397e-01 -4.94025238e-02 1.91511646e-01
-3.00112873e-01 9.28215921e-01 1.04457724e+00 -1.09192657e+00
-2.49409914e-01 -3.20108384e-01 1.61974773e-01 1.47887086e-02
-7.20936835e-01 -2.96067119e-01 -2.37512410e-01 -8.15387189e-01
5.49943447e-01 -8.87901604e-01 -4.37415630e-01 4.84179497e-01
-2.54929930e-01 -9.92256626e-02 9.94143367e-01 -1.69523582e-01
1.03878748e+00 -2.04809546e+00 5.49971640e-01 2.77511388e-01
5.23731411e-01 1.81656033e-01 -2.27720410e-01 1.83399990e-01
-2.37664163e-01 2.11023763e-01 -1.46104187e-01 -3.05462211e-01
-3.92320752e-02 2.53418148e-01 -3.11410338e-01 8.71714175e-01
5.17756939e-01 1.09367919e+00 -4.85643089e-01 -1.16267316e-01
3.03765655e-01 1.07821965e+00 -1.06253946e+00 -2.66209275e-01
2.61666514e-02 3.68024498e-01 -3.71591657e-01 9.66119394e-02
8.45946252e-01 -8.71544778e-02 7.73517564e-02 -5.97815633e-01
-4.53301966e-01 1.46729037e-01 -1.24777079e+00 1.46350348e+00
-3.95402879e-01 1.05281985e+00 3.48701656e-01 -1.47871435e+00
8.66324902e-01 1.33733541e-01 7.70687878e-01 -6.43190026e-01
1.82867303e-01 1.07125044e-01 1.43161774e-01 -1.81347147e-01
5.37751198e-01 -1.44527599e-01 3.72943133e-01 3.30414861e-01
2.12671869e-02 -6.97001517e-02 1.07871443e-01 1.55732268e-02
8.00546288e-01 -1.08897071e-02 -1.61377281e-01 -7.56747603e-01
3.72342736e-01 -2.51788408e-01 5.18001616e-01 7.15118229e-01
-1.80347845e-01 7.17343867e-01 5.02592087e-01 -7.07247913e-01
-1.17350626e+00 -9.10628974e-01 -3.17686886e-01 6.62455797e-01
5.42305410e-02 -1.16069820e-02 -8.70534480e-01 -8.95568207e-02
-9.60830972e-02 4.59183753e-02 -8.69244277e-01 -2.34923601e-01
-7.35789001e-01 -7.11824358e-01 5.73819339e-01 6.86574042e-01
5.99671245e-01 -7.58666098e-01 -7.47706711e-01 2.98272222e-01
2.19257623e-01 -8.74332130e-01 -2.65988976e-01 2.39015743e-01
-1.11906672e+00 -8.60827088e-01 -1.18049121e+00 -7.82675803e-01
8.66942585e-01 4.49955463e-01 7.84723639e-01 -5.99188507e-02
-1.74072012e-01 3.21392626e-01 7.44425133e-02 -2.36515105e-02
-1.50707170e-01 3.22037309e-01 1.79986671e-01 1.95968360e-01
1.05383888e-01 -8.35400462e-01 -7.27354348e-01 3.22595417e-01
-1.19305623e+00 4.88406718e-02 4.73924220e-01 7.22384334e-01
5.95452428e-01 4.90219221e-02 2.81989485e-01 -5.02008200e-01
7.03732431e-01 1.52715687e-02 -7.00209260e-01 -6.42249212e-02
-4.38109636e-01 3.86780173e-01 6.02420866e-01 -4.74129409e-01
-1.02152979e+00 4.35251789e-03 -2.59150058e-01 -3.68719786e-01
2.59425561e-03 4.20137495e-01 1.03955477e-01 -4.07745332e-01
7.90506780e-01 8.68127644e-02 1.90006614e-01 -4.11544681e-01
4.87305909e-01 1.24832824e-01 2.89088458e-01 -4.24259365e-01
7.83156753e-01 9.63504016e-01 4.57263052e-01 -1.30493844e+00
-3.78151774e-01 -2.18339533e-01 -9.02831256e-01 -2.20784858e-01
9.34646785e-01 -5.20343363e-01 -8.15859556e-01 6.37904942e-01
-1.37879705e+00 -3.05557430e-01 -4.66219008e-01 4.85933393e-01
-5.68871379e-01 7.95860887e-02 -6.26678705e-01 -6.42196596e-01
6.15940765e-02 -1.21302724e+00 7.13403463e-01 3.38199794e-01
-2.45559573e-01 -1.43647742e+00 -6.37762621e-02 -4.19401288e-01
7.93912828e-01 3.09761446e-02 9.09382582e-01 -4.19647880e-02
-5.70064127e-01 1.41048720e-02 -3.17904919e-01 4.69775259e-01
3.03585399e-02 4.92344230e-01 -8.40197682e-01 -3.11765254e-01
1.23442575e-01 1.59169540e-01 1.31655991e+00 9.14301872e-01
9.16545331e-01 -1.83706075e-01 -2.30923980e-01 8.98856997e-01
1.33431804e+00 -7.08394125e-02 4.93916690e-01 1.30266696e-01
8.98993194e-01 6.68213785e-01 -3.96260619e-01 2.84615047e-02
-3.75700891e-02 5.73026359e-01 4.48096424e-01 -4.13713545e-01
-2.03554913e-01 9.96840373e-02 1.72800437e-01 9.02772307e-01
-4.98439044e-01 1.03889160e-01 -5.63574553e-01 5.68598807e-01
-1.78793931e+00 -1.11851919e+00 -3.38872164e-01 1.92737770e+00
6.76835001e-01 -1.01641901e-01 -1.47541508e-01 3.30159903e-01
4.38550264e-01 4.79919285e-01 -5.15938044e-01 -5.78330040e-01
-2.79239625e-01 2.48050287e-01 6.70810163e-01 5.83852232e-01
-1.08226132e+00 7.23193169e-01 7.06209040e+00 4.65988696e-01
-1.47165549e+00 -3.24925259e-02 3.90578657e-01 -1.49933547e-01
-5.23169339e-01 -1.84772313e-01 -7.58965552e-01 -1.48514733e-01
5.34696579e-01 1.25788480e-01 4.71879274e-01 6.65614128e-01
5.16277075e-01 2.32706815e-01 -9.83918488e-01 6.91539228e-01
-2.23392770e-01 -1.69375348e+00 1.72713891e-01 4.09981310e-01
1.01503396e+00 1.08071886e-01 4.08209115e-01 -2.77299434e-01
-2.24284288e-02 -1.05427170e+00 7.67416418e-01 6.37004435e-01
5.52874148e-01 -7.78923333e-01 2.45963737e-01 1.31798759e-01
-1.12140214e+00 1.54414335e-02 -6.44409955e-01 -2.43267789e-01
4.34934087e-02 8.42379212e-01 -2.57377271e-02 2.30924115e-01
6.20378911e-01 9.71566796e-01 -5.78369461e-02 8.82197261e-01
-4.46649380e-02 2.74521917e-01 -4.50530022e-01 -2.00506002e-01
3.35249931e-01 -2.74307549e-01 6.81455910e-01 1.16919541e+00
2.07021236e-01 -2.10858062e-01 -4.10481274e-01 1.33007812e+00
-1.67503312e-01 -6.10898733e-02 -6.79896653e-01 1.87111795e-02
2.31382623e-01 1.16819906e+00 -9.80365753e-01 -2.41182148e-02
-4.30102199e-01 9.97595549e-01 3.55863184e-01 8.21380794e-01
-5.67500055e-01 -4.91968036e-01 1.15143859e+00 7.14737922e-02
3.21343005e-01 -7.07111657e-01 -3.37167799e-01 -1.29991853e+00
2.25996133e-02 -4.35500532e-01 -2.93865025e-01 -3.34992528e-01
-9.40729320e-01 2.92051762e-01 -1.99066430e-01 -9.24729228e-01
-2.58351155e-02 -1.17340577e+00 -4.89231676e-01 9.23075795e-01
-1.65904844e+00 -1.07798600e+00 -2.12061822e-01 6.51257932e-01
1.90844327e-01 2.78334320e-01 6.75167680e-01 3.58261228e-01
-4.94421571e-01 3.37808728e-01 2.34175444e-01 3.45606327e-01
3.43702614e-01 -1.14204824e+00 3.61218244e-01 9.64489639e-01
2.67990798e-01 1.07337379e+00 7.34554589e-01 -7.18622953e-02
-1.70123792e+00 -8.90106201e-01 7.62881398e-01 -2.09787533e-01
5.24184167e-01 -3.07652265e-01 -9.45391655e-01 4.61559057e-01
8.33194256e-02 2.09006548e-01 2.89078057e-01 4.61564176e-02
-4.48516309e-01 -8.88205618e-02 -8.15176487e-01 7.27135122e-01
1.17569351e+00 -6.22998059e-01 -2.63840526e-01 8.62455294e-02
3.78586739e-01 -2.33347967e-01 -6.96944475e-01 2.86995053e-01
9.69659925e-01 -1.13455355e+00 1.07318723e+00 -7.73319542e-01
4.00036365e-01 -2.50547081e-01 -1.18393393e-04 -1.37141991e+00
-7.04532623e-01 -6.38202071e-01 -3.45175527e-02 7.80070543e-01
3.67187500e-01 -9.65406895e-01 5.83564997e-01 5.37133753e-01
-4.68193181e-02 -6.93937063e-01 -9.24253166e-01 -8.40678155e-01
5.56089520e-01 -3.48383188e-01 1.56917751e-01 1.07677829e+00
-2.35013530e-01 -1.02514811e-01 -1.90813225e-02 5.08669280e-02
4.99733359e-01 -1.61130190e-01 3.44380915e-01 -1.30449700e+00
-5.34216911e-02 -1.16728735e+00 -5.79830468e-01 -1.46264744e+00
1.06830038e-01 -7.48753965e-01 -2.83887178e-01 -1.54555130e+00
-1.69687495e-01 -3.45820278e-01 -3.70721877e-01 3.72789502e-01
4.15466696e-01 5.95822036e-01 -1.12783000e-01 3.74493241e-01
9.56437737e-02 5.37449896e-01 1.51423728e+00 4.32197303e-02
-4.45572734e-01 -2.10715295e-03 -6.42681539e-01 7.28376448e-01
7.63467133e-01 -1.53716668e-01 -6.14270806e-01 -8.83073807e-01
5.43146491e-01 -6.35645509e-01 6.40543520e-01 -7.40532041e-01
3.59975666e-01 -1.42157763e-01 4.87973660e-01 -8.58026594e-02
2.24517643e-01 -7.32831359e-01 2.47443579e-02 6.10911906e-01
-4.01925504e-01 1.32795662e-01 6.22174628e-02 3.89060795e-01
-2.36686707e-01 -2.22439785e-02 1.14176905e+00 -2.53023915e-02
-4.35772151e-01 4.17497575e-01 -5.22300363e-01 -2.66151696e-01
7.69245982e-01 -3.37482154e-01 -1.21546850e-01 -3.30468476e-01
-7.42099345e-01 -2.94208020e-01 5.74996293e-01 2.36472324e-01
4.58639085e-01 -1.36777341e+00 -3.27398866e-01 5.60002148e-01
-5.36109865e-01 1.42496318e-01 2.52295792e-01 1.12625110e+00
-7.68027663e-01 6.18490160e-01 -4.33067471e-01 -6.84738398e-01
-8.91256809e-01 2.56524116e-01 8.44241023e-01 8.58617201e-02
-5.69806874e-01 9.65243638e-01 3.25511843e-01 -3.21509153e-01
-4.02546078e-02 -6.47671521e-01 -1.42523125e-01 2.20142499e-01
3.75796586e-01 1.70557141e-01 8.67645219e-02 -7.06027389e-01
-3.30287695e-01 7.71725476e-01 9.12774354e-02 -3.21031600e-01
1.40516460e+00 4.21177074e-02 -2.57981926e-01 8.47388059e-02
1.43173540e+00 1.52783990e-02 -1.46963501e+00 -2.12066919e-01
-4.38362777e-01 -1.32746950e-01 4.72489834e-01 -3.63243595e-02
-1.44656229e+00 1.28420734e+00 2.26841837e-01 7.48695254e-01
1.06688166e+00 -2.26042584e-01 3.28686506e-01 3.73554081e-01
-1.47902369e-01 -9.54824030e-01 -1.31632060e-01 6.72079563e-01
8.13044965e-01 -6.00919843e-01 -1.30069852e-01 -2.51447827e-01
6.11131862e-02 1.30493391e+00 2.76002765e-01 -4.95814562e-01
9.40782428e-01 2.79960275e-01 -5.74695244e-02 -1.55006737e-01
-4.37255055e-01 -1.61093011e-01 4.15555596e-01 6.20477140e-01
6.37257516e-01 -2.09550023e-01 -2.95682549e-01 5.85418046e-02
-3.70795205e-02 -2.77588576e-01 2.93643594e-01 9.76934910e-01
-4.61318046e-01 -9.84700561e-01 -5.55677563e-02 3.92426625e-02
-3.80837172e-01 5.27213290e-02 -3.52976590e-01 8.67273211e-01
-2.00058639e-01 5.45100152e-01 4.25359905e-01 1.59276614e-03
3.70661914e-01 -9.67398509e-02 8.09299350e-01 -2.31866062e-01
-7.74100572e-02 1.69092759e-01 -4.16559964e-01 -6.36883378e-01
-7.73032963e-01 -5.82649529e-01 -1.04798746e+00 -4.83740926e-01
-1.40510827e-01 -1.66334629e-01 7.70853996e-01 8.72062206e-01
5.21504223e-01 7.62298584e-01 4.44385886e-01 -1.11450601e+00
-4.80564624e-01 -6.74231350e-01 -5.77721000e-01 1.45377547e-01
6.66662931e-01 -6.60427630e-01 -3.59387666e-01 2.16326401e-01] | [9.063260078430176, 2.3216493129730225] |
ff99f73f-ccc4-4bd5-8678-19ded7a07f91 | creativity-of-ai-automatic-symbolic-option | 2112.09836 | null | https://arxiv.org/abs/2112.09836v2 | https://arxiv.org/pdf/2112.09836v2.pdf | Creativity of AI: Hierarchical Planning Model Learning for Facilitating Deep Reinforcement Learning | Despite of achieving great success in real-world applications, Deep Reinforcement Learning (DRL) is still suffering from three critical issues, i.e., data efficiency, lack of the interpretability and transferability. Recent research shows that embedding symbolic knowledge into DRL is promising in addressing those challenges. Inspired by this, we introduce a novel deep reinforcement learning framework with symbolic options. Our framework features a loop training procedure, which enables guiding the improvement of policy by planning with planning models (including action models and hierarchical task network models) and symbolic options learned from interactive trajectories automatically. The learned symbolic options alleviate the dense requirement of expert domain knowledge and provide inherent interpretability of policies. Moreover, the transferability and data efficiency can be further improved by planning with the symbolic planning models. To validate the effectiveness of our framework, we conduct experiments on two domains, Montezuma's Revenge and Office World, respectively. The results demonstrate the comparable performance, improved data efficiency, interpretability and transferability. | ['Shuting Deng', 'Hankz Hankui Zhuo', 'Chao Yu', 'Chen Chen', 'Kebing Jin', 'Zhihao Ma', 'Mu Jin'] | 2021-12-18 | null | null | null | null | ['montezumas-revenge'] | ['playing-games'] | [-6.54291585e-02 3.53548974e-01 -5.33676624e-01 -3.28296751e-01
-2.15408102e-01 -4.79597241e-01 5.50612807e-01 -1.31491557e-01
-3.06021243e-01 1.10614657e+00 4.01977152e-01 -5.64250350e-01
-6.54447675e-01 -8.14800978e-01 -6.66523755e-01 -3.80748361e-01
-2.43973851e-01 5.08828163e-01 1.40254453e-01 -2.76389688e-01
3.15205216e-01 4.08671081e-01 -1.34041798e+00 2.68246800e-01
1.47906530e+00 8.50669444e-01 7.46914595e-02 1.01177044e-01
-1.84708998e-01 1.11877537e+00 -3.72319490e-01 1.43960463e-02
2.21444950e-01 -1.54705286e-01 -1.10239863e+00 7.05975592e-02
-3.57382715e-01 -8.82693708e-01 -3.22038054e-01 8.90674770e-01
1.53932348e-01 3.37465882e-01 3.64457428e-01 -1.66206217e+00
-8.53131056e-01 5.07568359e-01 -1.92442343e-01 -2.80331522e-01
3.40799332e-01 6.56405270e-01 9.90194857e-01 -5.26289940e-01
5.14925122e-01 1.53780901e+00 4.60142434e-01 5.53140819e-01
-1.01443958e+00 -4.82317239e-01 5.06249428e-01 5.13411045e-01
-7.79296279e-01 -7.90187865e-02 5.71524620e-01 -4.25912231e-01
8.67659867e-01 -8.19260720e-03 6.36815727e-01 1.28988886e+00
3.30715030e-02 8.97917211e-01 1.24515855e+00 -1.31726518e-01
3.62167686e-01 2.53141552e-01 -8.38315487e-02 8.10952723e-01
9.25894454e-02 6.17615819e-01 -2.62508512e-01 8.17766339e-02
1.21202075e+00 3.56057674e-01 -7.04184473e-02 -4.47084934e-01
-1.27782476e+00 8.63555908e-01 6.38261139e-01 1.08104218e-02
-3.61265957e-01 2.44282946e-01 6.15498483e-01 2.99047142e-01
2.22134106e-02 6.88188136e-01 -4.52120602e-01 -2.90112436e-01
-3.14762503e-01 2.60446191e-01 5.77689946e-01 1.16531563e+00
7.07511008e-01 1.71421334e-01 -3.08097363e-01 4.36911285e-01
3.21110755e-01 5.05362749e-01 5.73885083e-01 -1.27323997e+00
5.96387982e-01 1.06925166e+00 5.13140738e-01 -9.72112060e-01
-6.16102397e-01 -1.27490655e-01 -7.18528867e-01 4.81782824e-01
2.01812729e-01 -3.87311399e-01 -7.58338511e-01 1.74216747e+00
3.02546114e-01 1.70639858e-01 2.46349916e-01 9.78495419e-01
5.20022988e-01 6.94149077e-01 3.27962041e-01 -3.00799143e-02
1.09905243e+00 -1.23182642e+00 -8.63676190e-01 -4.36941907e-02
7.59917259e-01 -9.40608159e-02 1.48987174e+00 3.61429453e-01
-7.75267661e-01 -8.59144390e-01 -8.30959022e-01 2.44786497e-02
-3.57775599e-01 2.46611640e-01 9.06493962e-01 1.14206642e-01
-7.39684582e-01 7.83004522e-01 -9.26272988e-01 -2.20409513e-01
5.30362785e-01 5.65209389e-01 -2.44466245e-01 1.07647784e-01
-1.48363447e+00 9.43540573e-01 8.46048832e-01 2.03207694e-02
-1.15423405e+00 -3.55834335e-01 -7.39912033e-01 1.76677823e-01
7.69624710e-01 -4.64260608e-01 1.41381383e+00 -1.09952319e+00
-1.75261760e+00 7.98357744e-03 3.50502968e-01 -4.69210476e-01
6.63312197e-01 -5.30887961e-01 -4.33878720e-01 -6.12350255e-02
1.16884843e-01 5.39436877e-01 4.15050417e-01 -9.73879516e-01
-9.26169872e-01 -1.68522239e-01 6.77648664e-01 3.00793082e-01
-3.47561806e-01 -4.86295581e-01 -3.11398804e-02 -3.48267138e-01
-3.58502567e-01 -1.00725186e+00 -4.67517078e-01 -4.27406579e-02
-2.18949065e-01 -5.15986919e-01 7.88716793e-01 -7.73490191e-01
1.33528388e+00 -2.01911402e+00 2.12921485e-01 2.61601686e-01
1.59968242e-01 5.52407503e-01 -1.59037605e-01 5.97598553e-01
1.39822304e-01 2.31199667e-01 -1.30445719e-01 2.86767662e-01
2.88718194e-01 6.09531939e-01 -5.53486288e-01 -6.32541925e-02
2.98709512e-01 1.02345753e+00 -1.18702519e+00 -2.58351713e-01
3.65182817e-01 -1.40328050e-01 -6.81934297e-01 3.43493521e-01
-8.05173695e-01 8.84483457e-01 -1.01426995e+00 4.92598921e-01
3.46451193e-01 -3.83992136e-01 4.49516743e-01 2.73236763e-02
-4.69063781e-02 1.73154235e-01 -1.17677712e+00 1.75597346e+00
-4.17993844e-01 1.95387647e-01 -3.03756863e-01 -9.53892767e-01
9.33490634e-01 3.02910596e-01 3.08061004e-01 -1.04118276e+00
-8.98291990e-02 2.15363607e-01 4.67191003e-02 -1.13532805e+00
4.10800427e-01 1.41071588e-01 -8.31396356e-02 3.80014658e-01
-1.49345160e-01 1.88043267e-01 -5.89524545e-02 -2.66147312e-03
8.71819615e-01 8.18211138e-01 4.42152053e-01 -2.12387502e-01
5.82145512e-01 3.09608370e-01 6.50411665e-01 5.03855467e-01
-2.17349648e-01 -2.60838389e-01 7.35991776e-01 -7.39538908e-01
-9.97724235e-01 -8.31947386e-01 3.02418381e-01 1.02150226e+00
4.28363740e-01 -2.53875196e-01 -7.02115595e-01 -1.05097950e+00
1.41848624e-01 9.05280173e-01 -5.66711366e-01 -3.08848709e-01
-7.82821119e-01 -1.38265535e-01 4.42678481e-01 7.56262541e-01
9.42381859e-01 -1.41909575e+00 -7.55373001e-01 3.00406992e-01
-1.41580239e-01 -1.01750076e+00 -6.75869361e-02 -2.81312943e-01
-9.79133248e-01 -1.04617929e+00 -1.26273945e-01 -4.71986502e-01
5.73673785e-01 5.04187047e-02 8.95191908e-01 1.09753743e-01
1.32767826e-01 2.81028152e-01 -4.20434803e-01 -2.92527974e-01
-3.19182515e-01 1.52145430e-01 1.79711372e-01 -3.43437493e-01
1.37706518e-01 -5.49839497e-01 -6.92687988e-01 4.57328230e-01
-8.05670440e-01 4.98053074e-01 7.27822900e-01 1.00033772e+00
3.49611491e-01 -5.26264869e-02 1.06042004e+00 -9.10229146e-01
9.16458786e-01 -6.31556153e-01 -7.43713200e-01 3.85351539e-01
-9.54236209e-01 3.68289679e-01 9.94074345e-01 -2.68113077e-01
-1.34555626e+00 -6.58566728e-02 5.71143627e-02 -2.10683644e-01
-3.79149616e-01 5.74128449e-01 -1.08529888e-01 2.37165213e-01
7.22866595e-01 1.39544100e-01 1.40172035e-01 -3.47252071e-01
5.29445469e-01 6.88080907e-01 3.55860442e-01 -9.41764891e-01
6.97153568e-01 3.60395610e-01 -1.07149564e-01 -1.86251462e-01
-7.19276011e-01 1.00195028e-01 -4.31633472e-01 -1.60444364e-01
9.14465189e-01 -6.73921108e-01 -1.16342354e+00 5.04695959e-02
-1.02627778e+00 -4.65126514e-01 -3.43864322e-01 4.23432708e-01
-8.57128859e-01 2.21043184e-01 -2.78192014e-01 -7.93608546e-01
-1.29329458e-01 -1.33446538e+00 6.01805985e-01 3.74362141e-01
-1.73734963e-01 -1.13062656e+00 -1.32966608e-01 1.52743191e-01
3.60141397e-01 4.83647019e-01 1.18757820e+00 -6.84061408e-01
-8.77143919e-01 2.78929532e-01 -3.79808396e-01 2.30342925e-01
2.49539286e-01 -3.82703424e-01 -7.18127191e-01 -2.32578903e-01
-2.38389283e-01 -5.24310470e-01 2.30378643e-01 -9.46035087e-02
1.53888416e+00 -7.51780391e-01 -4.06319231e-01 3.96356583e-01
1.28103769e+00 7.07129955e-01 5.38213789e-01 4.94120717e-01
5.48832178e-01 4.95070815e-01 1.19520414e+00 5.34401059e-01
4.22000557e-01 5.68970025e-01 6.68835700e-01 -1.38883099e-01
2.67563671e-01 -6.45131528e-01 3.69087398e-01 3.84481996e-01
-5.46150267e-01 3.94652737e-03 -1.00000548e+00 4.37517732e-01
-2.48272967e+00 -1.02037060e+00 2.83689559e-01 1.84256721e+00
7.87401021e-01 1.07692145e-01 2.03811646e-01 -2.02431977e-01
3.93337041e-01 -5.27649075e-02 -1.06614816e+00 -3.74366075e-01
5.54713488e-01 -8.98679942e-02 2.33503729e-01 4.17996109e-01
-9.13879335e-01 1.11049640e+00 5.98205614e+00 8.52030933e-01
-9.65908706e-01 -1.39830053e-01 4.99815911e-01 2.17045769e-01
-3.45903218e-01 -1.37976035e-02 -6.53192163e-01 5.65523863e-01
7.82104492e-01 -1.18897155e-01 7.94718504e-01 1.06187940e+00
5.14034569e-01 9.86003578e-02 -1.28155792e+00 6.43622458e-01
-5.43435931e-01 -1.42201471e+00 2.32203156e-01 -2.93600708e-02
5.74780345e-01 -3.82488281e-01 4.41462807e-02 9.06185031e-01
7.05432475e-01 -1.22108853e+00 5.45563459e-01 6.64107919e-01
7.80032277e-01 -9.50219512e-01 6.06734514e-01 7.73387611e-01
-8.29180181e-01 -5.56950986e-01 -2.70294785e-01 -3.67821693e-01
-1.08222067e-01 -7.81017467e-02 -1.03877473e+00 7.93771386e-01
4.22091901e-01 9.02940094e-01 -3.07338238e-01 4.93108302e-01
-6.01769447e-01 3.42240393e-01 1.65201679e-01 -2.31951430e-01
6.85368896e-01 -5.22278309e-01 2.23820642e-01 8.66443455e-01
1.93331257e-01 2.25492656e-01 4.98882771e-01 9.75780785e-01
2.48109922e-01 -2.09073260e-01 -8.92003536e-01 -1.90614522e-01
5.73642313e-01 1.03019667e+00 -1.99803770e-01 -3.81859541e-01
-3.73094469e-01 6.61560059e-01 5.39841771e-01 6.18356764e-01
-1.17296338e+00 -3.30216169e-01 5.93560159e-01 1.73117022e-03
-5.05563281e-02 -4.05069649e-01 -3.43090415e-01 -9.60591733e-01
-6.05400428e-02 -1.15736639e+00 9.88459066e-02 -7.28786647e-01
-9.81164336e-01 4.64392066e-01 1.77853271e-01 -1.36758888e+00
-4.13175672e-01 -7.11329877e-01 -4.52663541e-01 6.99448884e-01
-1.58962333e+00 -1.10841954e+00 -5.05804062e-01 6.72708750e-01
5.49390435e-01 -4.09919590e-01 8.00060034e-01 2.83134151e-02
-4.99711066e-01 3.85623574e-01 1.07230961e-01 4.23362516e-02
3.26634556e-01 -1.22242463e+00 2.49744728e-01 3.44945490e-01
-4.37646806e-01 6.67443335e-01 3.72821540e-01 -4.77880478e-01
-1.34067690e+00 -1.27761912e+00 2.60398984e-01 -3.14994603e-01
6.75225973e-01 -6.57888129e-02 -7.56572425e-01 9.87758219e-01
2.49476448e-01 -4.60582495e-01 4.22582626e-01 7.54477084e-02
4.67295535e-02 -1.54863223e-01 -1.10924828e+00 8.96399796e-01
1.35380244e+00 -2.28395954e-01 -7.24361181e-01 4.66281950e-01
1.12748396e+00 -4.98524785e-01 -8.84038627e-01 4.55199599e-01
3.83448362e-01 -9.91552353e-01 8.76009643e-01 -1.20626700e+00
7.58756995e-01 -3.88439000e-01 1.90729916e-01 -1.34963810e+00
-4.92522180e-01 -4.75265354e-01 -1.50637731e-01 1.00737834e+00
2.20535398e-01 -8.00559819e-01 4.71820623e-01 8.26023638e-01
-2.64025956e-01 -9.03295636e-01 -6.34530067e-01 -8.92685771e-01
-2.11053371e-01 -1.26536071e-01 1.05809093e+00 9.08712029e-01
1.99153304e-01 3.24498534e-01 -6.53210104e-01 2.74181925e-02
1.15084417e-01 3.06457430e-01 8.95857632e-01 -1.08900583e+00
-4.80933905e-01 -1.07467316e-01 -6.67719319e-02 -1.27177167e+00
2.81941444e-01 -7.39886642e-01 -1.42522752e-01 -1.74266124e+00
-1.65000632e-01 -6.92028165e-01 -3.30736279e-01 9.53189969e-01
1.98588576e-02 -7.40430772e-01 1.87217981e-01 2.59751946e-01
-8.41609240e-01 8.29678595e-01 1.80362284e+00 8.20726231e-02
-3.69937539e-01 -2.83348025e-03 -5.70632041e-01 1.00453424e+00
1.02471161e+00 -2.07538098e-01 -9.59996998e-01 -6.41190648e-01
9.03431401e-02 4.87880379e-01 4.13893908e-01 -9.08188403e-01
3.01431827e-02 -8.57760847e-01 1.71719223e-01 -1.40946567e-01
1.84128687e-01 -1.14960170e+00 2.25473568e-03 7.27483273e-01
-4.08455014e-01 2.38321960e-01 3.19627613e-01 7.52951920e-01
-1.43342033e-01 -8.79508927e-02 5.06057680e-01 -1.73965007e-01
-1.18843234e+00 3.86089265e-01 -2.76388764e-01 -6.50292039e-02
1.39761221e+00 -1.02512076e-01 -4.11401153e-01 -3.65746647e-01
-8.37071478e-01 7.39477277e-01 6.05929196e-02 4.76430446e-01
5.49973965e-01 -1.56922650e+00 -2.60921121e-01 3.10334414e-01
-1.11325704e-01 1.79055840e-01 1.10916570e-01 7.11697876e-01
-5.27434111e-01 5.33876836e-01 -6.39142871e-01 -3.67732823e-01
-7.06997514e-01 7.21990705e-01 3.14677924e-01 -6.44411802e-01
-5.69254398e-01 1.52222663e-01 2.47170344e-01 -8.11141312e-01
3.23948085e-01 -5.98532081e-01 -3.69222671e-01 -3.84361356e-01
2.42926612e-01 6.41599476e-01 -4.50262547e-01 1.90376133e-01
-1.31669492e-01 2.99179196e-01 -7.05952151e-03 -6.29140660e-02
1.32269239e+00 7.14523718e-02 1.69302613e-01 2.28382349e-01
5.31056762e-01 -4.79141235e-01 -1.61331224e+00 -2.43952557e-01
3.27286094e-01 -4.12529677e-01 -4.50539500e-01 -1.16970527e+00
-6.62935376e-01 9.63695407e-01 4.09128606e-01 2.96939701e-01
1.08875084e+00 -4.64898705e-01 6.49474204e-01 7.74146616e-01
5.61340511e-01 -1.27386355e+00 2.92026311e-01 6.60972714e-01
1.08827925e+00 -1.27981734e+00 -1.98118657e-01 -1.02185801e-01
-9.29740012e-01 1.15420628e+00 1.06000698e+00 -2.73717523e-01
3.09811532e-01 -1.95615679e-01 -2.16552734e-01 -9.48472321e-02
-6.65268242e-01 5.51018417e-02 8.50642920e-02 7.51961946e-01
6.52413443e-02 1.62481979e-01 -1.34272218e-01 7.38766730e-01
7.38286152e-02 3.97277921e-01 2.05381349e-01 7.72499442e-01
-6.66767299e-01 -1.04931164e+00 -4.52668704e-02 1.62614048e-01
8.01866576e-02 2.12695077e-01 -7.52976313e-02 1.13454175e+00
2.16369972e-01 7.33572245e-01 -1.49145007e-01 -4.58197206e-01
5.50494432e-01 -4.59948331e-02 2.07257956e-01 -4.77000654e-01
-4.61453736e-01 -2.76899129e-01 1.46432996e-01 -8.69976401e-01
-2.85940528e-01 -2.57261038e-01 -1.73946726e+00 -3.61366451e-01
1.68079495e-01 1.76470473e-01 2.70695388e-01 1.20518076e+00
7.14917123e-01 8.47011745e-01 5.98923564e-01 -3.77920508e-01
-1.17628896e+00 -8.39022040e-01 -3.99105608e-01 5.28190792e-01
2.70700961e-01 -8.62617970e-01 1.44761190e-01 -1.09857256e-02] | [4.2476959228515625, 1.6844491958618164] |
ad1f1356-7f25-4668-ae09-808c471deba0 | morphological-change-forecasting-for-prostate | 2101.06425 | null | https://arxiv.org/abs/2101.06425v1 | https://arxiv.org/pdf/2101.06425v1.pdf | Morphological Change Forecasting for Prostate Glands using Feature-based Registration and Kernel Density Extrapolation | Organ morphology is a key indicator for prostate disease diagnosis and prognosis. For instance, In longitudinal study of prostate cancer patients under active surveillance, the volume, boundary smoothness and their changes are closely monitored on time-series MR image data. In this paper, we describe a new framework for forecasting prostate morphological changes, as the ability to detect such changes earlier than what is currently possible may enable timely treatment or avoiding unnecessary confirmatory biopsies. In this work, an efficient feature-based MR image registration is first developed to align delineated prostate gland capsules to quantify the morphological changes using the inferred dense displacement fields (DDFs). We then propose to use kernel density estimation (KDE) of the probability density of the DDF-represented \textit{future morphology changes}, between current and future time points, before the future data become available. The KDE utilises a novel distance function that takes into account morphology, stage-of-progression and duration-of-change, which are considered factors in such subject-specific forecasting. We validate the proposed approach on image masks unseen to registration network training, without using any data acquired at the future target time points. The experiment results are presented on a longitudinal data set with 331 images from 73 patients, yielding an average Dice score of 0.865 on a holdout set, between the ground-truth and the image masks warped by the KDE-predicted-DDFs. | ['Yipeng Hu', 'Dean Barratt', 'Matt Clarkson', 'Caroline Moore', 'Vasilis Stavrinides', 'Nooshin Ghavami', 'Francesco Giganti', 'Yunguan Fu', 'Tom Vercauteren', 'Qianye Yang'] | 2021-01-16 | null | null | null | null | ['holdout-set'] | ['computer-vision'] | [ 3.83249074e-01 4.19321865e-01 -8.78812000e-02 -7.51452565e-01
-6.48561239e-01 -5.14657080e-01 8.29302609e-01 6.01505041e-01
-7.32592165e-01 8.15860808e-01 1.84115842e-01 1.65640548e-01
-7.39790797e-01 -7.98179865e-01 -4.09800142e-01 -8.74783516e-01
-9.99117494e-01 9.03421104e-01 1.75851434e-01 2.61628747e-01
2.89292693e-01 9.32992995e-01 -7.62441576e-01 -2.14892507e-01
9.15492713e-01 7.07237661e-01 4.84443605e-01 6.81123972e-01
1.13893285e-01 1.15815483e-01 -2.36560374e-01 -3.75173151e-01
1.54115930e-01 -6.24205098e-02 -4.92157400e-01 1.46714330e-01
4.95626628e-01 -7.64509961e-02 -2.90165067e-01 1.11248887e+00
2.52877057e-01 -4.92607430e-02 9.11963522e-01 -6.81956291e-01
-3.65029007e-01 3.21767449e-01 -3.51926923e-01 6.82369173e-01
-1.92724839e-01 2.41867825e-01 1.84794478e-02 -5.35018206e-01
1.10357285e+00 8.52178991e-01 8.33885849e-01 6.08588278e-01
-1.39859414e+00 -2.18426198e-01 -2.82890052e-01 2.51406548e-03
-1.16252875e+00 -1.27107307e-01 7.04344153e-01 -7.70565510e-01
2.40997210e-01 5.99563599e-01 6.29371881e-01 6.03759110e-01
1.22224557e+00 1.98458672e-01 1.33247983e+00 -7.21600801e-02
2.88977653e-01 -5.09081446e-02 1.04872510e-01 5.41151822e-01
2.43594840e-01 3.40390921e-01 2.53784537e-01 -2.57975340e-01
1.05698562e+00 2.12595209e-01 -3.35773259e-01 -3.89414489e-01
-1.21434224e+00 6.52693331e-01 6.94829404e-01 7.87580311e-01
-5.85336924e-01 -8.31448808e-02 2.87509054e-01 -1.49718732e-01
5.69801271e-01 4.37509976e-02 -1.45992509e-03 2.29564160e-01
-1.08703756e+00 1.97662070e-01 4.75459725e-01 4.02485132e-01
3.54765177e-01 -3.91663522e-01 -4.74473298e-01 4.11674172e-01
4.80521321e-01 5.81090987e-01 7.75965214e-01 -7.23296881e-01
3.79050449e-02 7.25998998e-01 -7.62042850e-02 -1.21855152e+00
-7.68553972e-01 -7.63433874e-01 -1.30607843e+00 3.79490703e-02
5.69457591e-01 2.03961477e-01 -1.14410329e+00 1.29797614e+00
4.56526041e-01 2.18062833e-01 -9.65113565e-02 8.11137140e-01
3.03513020e-01 2.79701769e-01 1.51180580e-01 -6.37798846e-01
1.36719251e+00 -2.51692206e-01 -7.61865437e-01 -1.61920592e-01
5.15266478e-01 -3.10714394e-01 2.95129716e-01 1.57823473e-01
-9.82653499e-01 -3.88240635e-01 -7.89412200e-01 4.86419320e-01
-9.10697654e-02 4.01533484e-01 6.77709937e-01 5.73031843e-01
-1.10923362e+00 1.09737194e+00 -1.57308257e+00 -4.06171262e-01
3.81432623e-01 6.64324164e-01 -4.65820521e-01 -6.67569414e-02
-7.43493080e-01 9.94365811e-01 2.95920759e-01 6.64106369e-01
-9.66877460e-01 -7.34145164e-01 -7.04073429e-01 -2.29678482e-01
-2.59770453e-01 -4.01017219e-01 6.61862731e-01 -4.31706399e-01
-9.60353553e-01 1.00464201e+00 -1.34100303e-01 -7.39903867e-01
8.72799575e-01 4.73509699e-01 -4.71681625e-01 3.14820185e-02
3.42608280e-02 4.54989582e-01 4.40892398e-01 -1.27351832e+00
-1.80588320e-01 -1.11121368e+00 -6.89572394e-01 2.42704645e-01
-1.57157734e-01 -4.21540409e-01 -2.76039332e-01 -3.22714418e-01
3.83373588e-01 -1.16577554e+00 -7.89626420e-01 1.57825917e-01
-3.24876249e-01 6.91514537e-02 5.90692580e-01 -1.05798590e+00
8.85277271e-01 -2.11014175e+00 2.12027639e-01 3.99667412e-01
3.08477998e-01 -9.05266628e-02 3.77163708e-01 -2.46061414e-01
9.61380675e-02 -1.06744401e-01 -8.60451937e-01 -2.92632669e-01
-4.72458154e-01 4.15092468e-01 2.96442598e-01 9.98306036e-01
9.31498036e-02 7.94016242e-01 -8.80492449e-01 -5.94172060e-01
2.87995160e-01 5.31536579e-01 1.12300031e-01 5.58685027e-02
2.60296226e-01 1.02401090e+00 -5.47332287e-01 4.25283194e-01
9.75481570e-01 1.86836254e-02 2.85335422e-01 -3.08286279e-01
-3.23792130e-01 -5.47090650e-01 -7.55117893e-01 1.79631937e+00
-6.72538206e-02 4.44223136e-01 1.73363343e-01 -7.68782794e-01
1.31992459e+00 2.43483186e-01 9.16305840e-01 -6.12814248e-01
-4.55311313e-02 4.05311912e-01 2.30010957e-01 -3.09210509e-01
2.27989540e-01 -3.59390050e-01 4.91266586e-02 -1.53026074e-01
-2.23560259e-01 -5.48082292e-02 2.67668843e-01 -1.77137658e-01
1.23860681e+00 -1.98750407e-01 1.11539289e-01 -6.78531110e-01
8.50936532e-01 1.66269168e-01 5.51453590e-01 2.94687390e-01
-5.94907701e-01 4.73518640e-01 3.86139184e-01 -5.73012888e-01
-9.69493389e-01 -1.28293300e+00 -8.64719808e-01 -2.12424532e-01
1.20682381e-01 3.33998710e-01 -4.86569256e-01 -6.34603977e-01
1.51838243e-01 6.05768979e-01 -8.25041175e-01 -6.54714257e-02
-8.48046660e-01 -1.21982694e+00 2.12392151e-01 2.27285922e-01
2.05283940e-01 -9.16286409e-01 -5.55086553e-01 4.43337172e-01
3.25730562e-01 -6.79464102e-01 -2.83888549e-01 2.51579154e-02
-1.74469137e+00 -9.42895949e-01 -1.13964880e+00 -6.16031647e-01
1.07998526e+00 -5.56934237e-01 8.49391520e-01 -1.62769139e-01
-6.44713819e-01 3.15392077e-01 3.52997363e-01 2.79385954e-01
-5.83263338e-01 -4.03225809e-01 -1.04565680e-01 -2.81341020e-02
-8.98060426e-02 -7.32555568e-01 -8.10134351e-01 1.74442396e-01
-9.36960280e-01 -1.42233968e-01 8.81024301e-01 5.70489943e-01
9.75182176e-01 3.09275109e-02 2.16328904e-01 -7.53378034e-01
5.02413630e-01 -3.98435354e-01 -6.77575052e-01 3.20546031e-01
-8.02261472e-01 -2.57294104e-02 1.35030687e-01 -4.52757746e-01
-9.87766683e-01 3.37513059e-01 1.38474539e-01 -3.38536084e-01
-2.12946936e-01 5.34396589e-01 2.26908565e-01 -1.96320474e-01
5.18662751e-01 5.70647955e-01 2.29676768e-01 -3.49223703e-01
-4.82743755e-02 2.98460245e-01 8.92430484e-01 -4.15584892e-01
5.42446494e-01 7.17874825e-01 5.39933085e-01 -6.53530061e-01
2.67144553e-02 -4.03343260e-01 -1.03729498e+00 -5.13636708e-01
9.69632924e-01 -4.60593432e-01 -6.19440675e-01 3.71207684e-01
-9.38956976e-01 -1.74391806e-01 -3.30853701e-01 8.04619610e-01
-5.63274860e-01 4.08562362e-01 -6.40837729e-01 -7.88736582e-01
-5.19804299e-01 -1.21508646e+00 9.57270026e-01 5.58988273e-01
1.41618714e-01 -1.52186358e+00 3.21799606e-01 1.71061903e-02
6.06320202e-01 8.31132114e-01 7.98811376e-01 -6.43153906e-01
-4.51179445e-01 -4.09710258e-01 -5.16961738e-02 -4.33558002e-02
2.97946751e-01 -3.12758923e-01 -4.29879248e-01 -3.93506169e-01
3.55201513e-01 4.54784840e-01 6.59938753e-01 1.01793671e+00
8.89776707e-01 -7.11090490e-02 -7.18985498e-01 4.67386812e-01
1.76092124e+00 6.18514359e-01 7.83955872e-01 2.22728044e-01
4.71601933e-01 4.33433115e-01 6.07050478e-01 3.26832056e-01
3.52634460e-01 6.78099215e-01 5.60801029e-01 -7.16648921e-02
-7.99159855e-02 3.18528444e-01 4.09984589e-02 5.83571851e-01
-4.10160512e-01 2.35037491e-01 -1.20751023e+00 7.25700736e-01
-1.45652926e+00 -6.48223281e-01 -6.44943357e-01 2.57794309e+00
7.09053934e-01 2.62085795e-01 -2.95360863e-01 -4.06948060e-01
9.49005783e-01 -2.44795792e-02 -4.50902641e-01 -9.87681150e-02
2.58569717e-01 -4.44762073e-02 8.78877521e-01 8.34650278e-01
-1.14610088e+00 5.90618141e-02 5.46911049e+00 4.89494741e-01
-1.16449487e+00 2.77420938e-01 1.03246486e+00 2.45404184e-01
-2.61371434e-01 5.26297949e-02 -7.42635369e-01 4.57732111e-01
1.04479218e+00 -3.39687020e-01 5.08888848e-02 5.15313566e-01
3.91668528e-01 -4.37863439e-01 -9.54230249e-01 4.26931351e-01
2.60148868e-02 -1.20186722e+00 -5.65527260e-01 5.08717418e-01
4.61854339e-01 -1.53303340e-01 -8.48986059e-02 6.41962960e-02
-1.13708325e-01 -1.08058035e+00 6.35902658e-02 1.35026419e+00
7.19309628e-01 -3.36301118e-01 9.35029447e-01 2.79528052e-01
-1.15418959e+00 2.52636701e-01 -1.50686517e-01 4.57154274e-01
1.67490095e-01 8.72149646e-01 -1.37727261e+00 7.70922720e-01
1.19508296e-01 6.57727361e-01 -8.35607588e-01 1.23233414e+00
3.89144778e-01 2.63310552e-01 -3.14290553e-01 1.23638391e-01
-7.45607913e-02 -3.38985741e-01 7.90090621e-01 1.06929410e+00
5.98030388e-01 -4.04079519e-02 1.53853685e-01 1.04219759e+00
5.43838978e-01 -8.34601969e-02 -3.59086990e-01 7.21728057e-02
5.41272908e-02 1.52453995e+00 -1.34991598e+00 -8.26914683e-02
1.19535483e-01 9.84308064e-01 3.41914175e-03 -7.51806470e-03
-4.27094251e-01 3.40283155e-01 -3.70940492e-02 3.33349198e-01
-1.03453144e-01 -3.63923252e-01 -1.91593647e-01 -7.20561266e-01
1.24393702e-01 -9.33172852e-02 3.86428565e-01 -2.11149976e-01
-1.41610956e+00 6.90605402e-01 1.68318465e-01 -9.14058208e-01
-3.42134774e-01 -4.96875435e-01 -6.55151665e-01 9.66089725e-01
-1.31572688e+00 -9.74149942e-01 -2.08034128e-01 2.71636963e-01
3.31480503e-01 2.86003858e-01 9.35202658e-01 3.06320518e-01
-2.25833684e-01 -1.98729392e-02 2.76080251e-01 1.54507115e-01
4.41681325e-01 -1.55210471e+00 -1.15811206e-01 5.55370450e-01
-5.53692937e-01 4.62724239e-01 7.55796790e-01 -1.28665102e+00
-1.34726083e+00 -1.08923185e+00 7.82669187e-01 -2.96600401e-01
6.08203530e-01 9.48194712e-02 -9.14472938e-01 6.21521533e-01
-2.90673792e-01 6.25799537e-01 2.94183552e-01 -4.29489464e-01
8.03343654e-01 -2.21818775e-01 -1.59194458e+00 4.97991107e-02
5.65363288e-01 1.01890661e-01 -3.63476336e-01 4.09569293e-01
-6.27793954e-04 -7.46930897e-01 -1.72498608e+00 7.73601651e-01
4.20437366e-01 -8.47885013e-01 9.61761057e-01 4.07286035e-03
1.13542832e-01 -2.24473625e-01 1.34783909e-02 -1.17839181e+00
-3.09852749e-01 -8.94624740e-02 9.13175344e-02 1.14854121e+00
1.02674864e-01 -2.18338624e-01 1.12052476e+00 9.25211966e-01
-2.95335174e-01 -7.61924982e-01 -1.55524099e+00 -6.43573284e-01
8.09974745e-02 2.94352043e-02 4.67038006e-02 9.87374783e-01
-4.86631453e-01 -5.83372533e-01 1.76331609e-01 4.13964331e-01
9.80302811e-01 -1.49419084e-01 2.09887773e-01 -1.57354105e+00
-8.01729038e-02 -2.20810652e-01 -9.91236389e-01 -1.01223834e-01
-2.80816495e-01 -1.19689286e+00 -5.55646904e-02 -1.66903555e+00
4.77853179e-01 -6.14822507e-01 -2.37929881e-01 2.68431753e-02
-2.00459063e-02 4.88285162e-02 6.90379292e-02 4.60077018e-01
6.05001114e-02 4.90799576e-01 1.88103950e+00 -1.36369646e-01
-3.51925135e-01 2.44151786e-01 1.95107982e-01 6.14558041e-01
3.71202379e-01 -5.92926323e-01 -1.64677933e-01 2.60472804e-01
-5.18418014e-01 6.78650975e-01 6.09991789e-01 -1.28099179e+00
3.19286019e-01 -1.02607548e-01 7.92864025e-01 -7.05608189e-01
4.04154956e-01 -1.11316681e+00 7.85354316e-01 1.19567084e+00
-2.00359672e-01 3.40504944e-02 1.82785556e-01 7.58617461e-01
-3.90926898e-02 -5.21340072e-01 5.16895890e-01 -1.88459232e-01
-5.22212803e-01 4.37275559e-01 -1.46205276e-01 -5.44213831e-01
1.22526860e+00 -1.89020723e-01 -6.98721558e-02 1.27084479e-01
-1.37273324e+00 1.71821695e-02 2.88952887e-01 -2.95751672e-02
6.11462593e-01 -1.44318342e+00 -8.53946447e-01 8.49915855e-03
-2.26874933e-01 9.48114321e-02 5.81907392e-01 1.41078770e+00
-5.50942242e-01 3.38106394e-01 -2.39541411e-01 -1.06973433e+00
-1.24139118e+00 3.22593570e-01 6.45129263e-01 -1.13530219e+00
-7.59140730e-01 5.16478002e-01 -2.07655225e-02 -4.46490049e-01
-5.86177818e-02 -4.37640429e-01 -5.73940217e-01 -1.05014972e-01
2.31077000e-01 6.33788258e-02 2.16429368e-01 -7.42089212e-01
-4.33444232e-01 4.81003433e-01 -9.33927074e-02 -5.48534989e-02
1.45827472e+00 -1.77765101e-01 -2.56824464e-01 3.64992082e-01
1.00965333e+00 -9.80285257e-02 -1.42084968e+00 -4.54125702e-02
4.18330610e-01 -4.78307277e-01 3.69202971e-01 -9.12089169e-01
-1.08423233e+00 3.74648362e-01 1.66714382e+00 2.26677358e-01
1.01091778e+00 4.16758209e-02 3.75713140e-01 -2.21781656e-01
2.75757164e-01 -4.55428123e-01 -5.11778235e-01 2.03415081e-02
9.66648817e-01 -1.15159297e+00 2.22475275e-01 -5.02490103e-01
-5.36029160e-01 1.32567692e+00 2.88276196e-01 -4.66750473e-01
5.43185949e-01 3.32757145e-01 -1.63460895e-01 -4.75816727e-01
-2.69159019e-01 3.95192385e-01 4.87363100e-01 6.96146071e-01
4.72349405e-01 4.08274651e-01 -8.34279478e-01 4.55053300e-01
5.32562006e-03 1.98425770e-01 3.91594976e-01 8.98373842e-01
-2.91232109e-01 -1.03232527e+00 -6.59113348e-01 5.88184893e-01
-2.52910525e-01 3.85549158e-01 -2.69472245e-02 9.40832496e-01
1.36848971e-01 1.24655046e-01 4.48468029e-01 1.38708174e-01
2.21297845e-01 -1.35236144e-01 7.26513267e-01 -2.96549797e-01
-4.12671685e-01 1.42408788e-01 -1.19598784e-01 -2.78504193e-01
-5.54553390e-01 -1.02176154e+00 -1.43488646e+00 1.58877611e-01
-9.91064012e-02 -2.75660418e-02 1.16997766e+00 9.16743398e-01
3.54102030e-02 7.11866260e-01 6.42546833e-01 -1.01303756e+00
-3.91883403e-01 -9.51240301e-01 -7.57412851e-01 2.84029216e-01
1.61717087e-01 -6.49408460e-01 -2.63368160e-01 7.64667243e-02] | [14.41563892364502, -2.530344009399414] |
4f5205e8-6812-45a7-a6c7-ec76fa6a0a7e | correspondence-learning-via-linearly | 2010.13136 | null | https://arxiv.org/abs/2010.13136v1 | https://arxiv.org/pdf/2010.13136v1.pdf | Correspondence Learning via Linearly-invariant Embedding | In this paper, we propose a fully differentiable pipeline for estimating accurate dense correspondences between 3D point clouds. The proposed pipeline is an extension and a generalization of the functional maps framework. However, instead of using the Laplace-Beltrami eigenfunctions as done in virtually all previous works in this domain, we demonstrate that learning the basis from data can both improve robustness and lead to better accuracy in challenging settings. We interpret the basis as a learned embedding into a higher dimensional space. Following the functional map paradigm the optimal transformation in this embedding space must be linear and we propose a separate architecture aimed at estimating the transformation by learning optimal descriptor functions. This leads to the first end-to-end trainable functional map-based correspondence approach in which both the basis and the descriptors are learned from data. Interestingly, we also observe that learning a \emph{canonical} embedding leads to worse results, suggesting that leaving an extra linear degree of freedom to the embedding network gives it more robustness, thereby also shedding light onto the success of previous methods. Finally, we demonstrate that our approach achieves state-of-the-art results in challenging non-rigid 3D point cloud correspondence applications. | ['Maks Ovsjanikov', 'Simone Melzi', 'Marie-Julie Rakotosaona', 'Riccardo Marin'] | 2020-10-25 | null | http://proceedings.neurips.cc/paper/2020/hash/11953163dd7fb12669b41a48f78a29b6-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/11953163dd7fb12669b41a48f78a29b6-Paper.pdf | neurips-2020-12 | ['3d-dense-shape-correspondence'] | ['computer-vision'] | [-2.86419243e-02 2.27791831e-01 1.19630180e-01 -4.36923862e-01
-8.02316308e-01 -6.93443954e-01 8.89388740e-01 8.29206686e-03
-3.76513392e-01 1.95734933e-01 1.28492326e-01 -2.79671163e-03
-2.02345520e-01 -6.16225481e-01 -1.04583371e+00 -5.93563259e-01
-6.78089112e-02 7.74284899e-01 1.34759665e-01 -1.51485860e-01
3.17940891e-01 9.48844492e-01 -1.36108506e+00 -1.00745052e-01
4.70295161e-01 8.25456858e-01 -3.31374705e-02 4.01853621e-01
3.19929570e-02 1.43576920e-01 -1.32058024e-01 -3.02668631e-01
7.63958931e-01 -5.17647862e-02 -8.40980530e-01 6.51179478e-02
9.21696365e-01 -3.16234738e-01 -3.83510351e-01 7.50040412e-01
3.48389775e-01 3.83766666e-02 7.87617028e-01 -1.13344359e+00
-7.70559251e-01 -7.49842450e-02 -2.92789221e-01 -1.70592099e-01
4.84965622e-01 8.71220380e-02 1.25849330e+00 -1.19702542e+00
8.20249140e-01 1.33746886e+00 9.67321634e-01 2.34666750e-01
-1.61160612e+00 -3.05086613e-01 1.83443706e-02 -8.01013261e-02
-1.32005441e+00 -3.97686392e-01 7.95913160e-01 -7.40728915e-01
1.12665129e+00 1.52955383e-01 7.31070757e-01 7.93063045e-01
2.57338099e-02 4.11118835e-01 9.08592999e-01 -4.29323226e-01
3.35888229e-02 9.26180705e-02 -7.49204308e-02 6.26270533e-01
7.07229301e-02 3.14794958e-01 -3.87903839e-01 -1.53949797e-01
9.88606811e-01 1.62729397e-01 -2.07696036e-01 -1.16582048e+00
-1.38620174e+00 8.84060502e-01 7.25873411e-01 2.81706363e-01
-3.19821030e-01 3.02289575e-01 5.20539694e-02 3.20672303e-01
5.77142596e-01 5.26032567e-01 -2.85979062e-01 -9.92458463e-02
-7.89739430e-01 3.44741553e-01 8.86361659e-01 9.17696178e-01
1.13885248e+00 -3.78346175e-01 7.76503757e-02 4.80506599e-01
3.88739049e-01 3.09122711e-01 7.40449084e-03 -1.02117813e+00
3.18552166e-01 8.14562917e-01 1.14985779e-01 -1.03233325e+00
-4.63601708e-01 -3.92092019e-01 -4.80199009e-01 4.42675501e-01
4.42821860e-01 4.92014945e-01 -7.83109248e-01 1.64657891e+00
3.02995771e-01 4.90509868e-01 -2.74492949e-01 9.54511821e-01
2.16360703e-01 3.25249493e-01 -4.74661797e-01 2.66476095e-01
1.06260526e+00 -6.82897270e-01 -3.03860426e-01 5.07049896e-02
4.22548264e-01 -8.49463761e-01 1.24017894e+00 6.85028955e-02
-1.07676017e+00 -3.63961041e-01 -1.08228731e+00 -4.55934256e-01
-3.43622953e-01 -8.47148597e-02 4.75437075e-01 1.31772146e-01
-1.39328039e+00 9.06651139e-01 -1.11594498e+00 -6.47681355e-01
3.50062221e-01 4.70611244e-01 -7.62148559e-01 -2.98318733e-03
-7.12152123e-01 1.15608048e+00 4.72050831e-02 -2.67111603e-02
-4.41298068e-01 -9.27675486e-01 -8.06815982e-01 -8.83884951e-02
5.04250266e-02 -9.19896007e-01 8.59339893e-01 -3.57970715e-01
-1.44482219e+00 1.20694208e+00 -2.04685643e-01 -2.94421643e-01
7.22173035e-01 -3.12498629e-01 3.37912768e-01 6.19010627e-02
-5.10507682e-03 7.57505536e-01 8.36116672e-01 -1.24580598e+00
-1.89506412e-01 -4.72912073e-01 1.47286072e-01 5.76115735e-02
-3.12131733e-01 -3.07371438e-01 -3.88878345e-01 -3.41607958e-01
4.30070996e-01 -1.29404902e+00 4.81745154e-02 5.05355299e-01
-1.30722210e-01 -3.21873695e-01 7.36597478e-01 -4.57883954e-01
7.73859978e-01 -2.18069363e+00 5.84312975e-01 4.23417240e-01
3.94577026e-01 -1.58788487e-01 -1.27206296e-01 6.38713539e-01
-1.74816400e-01 1.20144345e-01 -4.45751935e-01 -5.03149092e-01
3.58488172e-01 3.11484128e-01 -3.20288002e-01 8.96731794e-01
6.49902284e-01 9.48696852e-01 -7.86486328e-01 -9.11286920e-02
4.91779268e-01 8.91731501e-01 -7.52081811e-01 1.52570859e-01
5.77683449e-02 6.23453856e-01 -2.87999868e-01 3.47293913e-01
7.73037255e-01 -1.71858221e-01 -3.43111962e-01 -4.15264726e-01
-2.26378679e-01 4.08995420e-01 -1.17788780e+00 2.15629649e+00
-6.13093495e-01 5.88690519e-01 -6.72841221e-02 -1.11744094e+00
9.74908710e-01 1.26932755e-01 8.00995350e-01 -3.96117270e-01
-3.89543399e-02 4.26038325e-01 -1.14673674e-01 -2.15890363e-01
2.72223592e-01 -2.61837512e-01 1.54054135e-01 3.35994929e-01
2.64104068e-01 -4.76294428e-01 -2.10258424e-01 -6.67343289e-02
1.04876614e+00 5.87728381e-01 7.82640278e-02 -4.49938297e-01
5.54024160e-01 -7.52448663e-02 1.64178059e-01 3.85312587e-01
9.68027338e-02 8.82569790e-01 3.84096384e-01 -5.73874116e-01
-1.32182980e+00 -1.21456575e+00 -4.92457896e-01 5.95370412e-01
5.72378188e-02 -4.37663376e-01 -4.99500602e-01 -5.49045146e-01
5.02446413e-01 4.17515904e-01 -7.25966573e-01 -1.71698079e-01
-7.87793934e-01 -2.33054504e-01 3.96381527e-01 4.46893603e-01
1.34609025e-02 -5.49409330e-01 -3.50228757e-01 7.41206929e-02
2.17738688e-01 -1.24088430e+00 -5.18502772e-01 1.45759195e-01
-1.01745856e+00 -9.71357822e-01 -4.74365562e-01 -6.31157815e-01
7.47278094e-01 2.08762169e-01 1.15988886e+00 1.64441988e-02
-1.98217332e-01 6.90373302e-01 -7.50334933e-02 -2.06640407e-01
-3.46904129e-01 1.66195333e-01 7.05825537e-02 -9.14450884e-02
4.84852821e-01 -9.76987422e-01 -5.88585436e-01 2.50679761e-01
-8.81945789e-01 -1.84633821e-01 4.75718170e-01 7.24361718e-01
6.67499721e-01 -5.84403813e-01 -8.11262056e-02 -5.69228590e-01
4.78584617e-01 -2.59994268e-01 -6.67620063e-01 -6.80671036e-02
-5.75217962e-01 5.00140667e-01 5.35043895e-01 -2.18207970e-01
-2.58393437e-01 3.73887271e-01 -3.02716523e-01 -7.82166839e-01
-1.20369508e-03 1.55118838e-01 1.73087075e-01 -4.79316205e-01
6.60680115e-01 7.31095150e-02 3.49866718e-01 -7.55859375e-01
5.06418943e-01 2.48806253e-01 4.27271158e-01 -6.37667358e-01
1.25994837e+00 7.98682511e-01 4.19228613e-01 -5.75785756e-01
-5.41488886e-01 -5.66148281e-01 -1.25676072e+00 1.06320478e-01
8.85887206e-01 -9.08640325e-01 -7.88332820e-01 1.83796193e-02
-1.34300017e+00 -5.22023626e-02 -5.17361999e-01 4.10824031e-01
-9.55898523e-01 2.68875480e-01 -2.30272934e-01 -3.75463516e-01
-3.94929387e-02 -1.35126328e+00 1.65913057e+00 -3.77020866e-01
-2.83754438e-01 -1.09604204e+00 3.41908216e-01 -6.51405230e-02
4.25784588e-01 4.51531172e-01 8.55880201e-01 -5.00068605e-01
-9.04091954e-01 -4.20338005e-01 -1.62810668e-01 3.28100502e-01
1.36953920e-01 -1.80134773e-02 -1.07172632e+00 -4.47981209e-01
1.34303302e-01 -2.03780495e-02 9.22809899e-01 1.68860238e-02
7.93525517e-01 -1.21416740e-01 -1.74275994e-01 1.06260431e+00
1.49218857e+00 -4.92696613e-01 5.19402385e-01 5.00154495e-01
8.99552882e-01 5.88064373e-01 3.14459771e-01 1.46569699e-01
5.00786364e-01 1.09206045e+00 5.18310726e-01 -1.05639897e-01
-1.93938971e-01 -4.90228802e-01 3.34300488e-01 6.91871941e-01
-1.62619710e-01 4.23686326e-01 -1.05799389e+00 3.88002753e-01
-1.84187829e+00 -6.70015693e-01 9.71109699e-03 2.43079972e+00
7.17784405e-01 1.47565939e-02 2.12287292e-01 -1.52143883e-02
3.69376689e-01 7.96847567e-02 -5.10024428e-01 -2.89478093e-01
1.00149125e-01 4.31263566e-01 4.95599359e-01 8.93731833e-01
-9.60498512e-01 9.17771935e-01 6.23737288e+00 3.11660111e-01
-1.42070365e+00 8.79540518e-02 -1.80552937e-02 -5.93901910e-02
-4.94823098e-01 2.80016989e-01 -6.08981490e-01 2.28610411e-01
6.19264483e-01 -4.60309498e-02 5.18493056e-01 7.20621645e-01
2.49080658e-02 3.18160504e-01 -1.60977089e+00 9.53237236e-01
3.79493879e-03 -1.37688649e+00 7.17437044e-02 4.56169784e-01
3.80554795e-01 2.17420831e-01 -4.46040705e-02 -4.60551586e-03
9.57321972e-02 -1.05191922e+00 8.01117063e-01 6.18353486e-01
7.57312000e-01 -4.37392533e-01 4.60882813e-01 3.52890790e-01
-1.08263981e+00 2.69857287e-01 -6.25434041e-01 -2.23065317e-02
2.96927430e-02 5.75358391e-01 -1.00346184e+00 5.28456271e-01
5.86461067e-01 1.00940144e+00 -6.28119409e-01 8.78987968e-01
-9.65956822e-02 9.27310362e-02 -6.33810520e-01 4.08514082e-01
2.04560906e-01 -5.64545512e-01 7.74610698e-01 1.14835942e+00
3.63160610e-01 -2.52471536e-01 2.21107289e-01 1.13355124e+00
-9.39788520e-02 -7.35052675e-02 -8.89219761e-01 1.44389272e-01
2.70382166e-01 1.31335640e+00 -4.53085005e-01 8.41168240e-02
-3.20513219e-01 1.00221121e+00 6.39731050e-01 2.43993789e-01
-4.97843534e-01 -8.97535309e-02 8.84157598e-01 3.89744312e-01
4.07302737e-01 -7.25033283e-01 -3.72537464e-01 -1.23024929e+00
3.57401282e-01 -4.90985960e-01 -9.74938273e-02 -4.86011535e-01
-1.31916845e+00 4.18705821e-01 3.82246003e-02 -1.43553030e+00
-3.83816063e-01 -8.55752468e-01 -5.45949996e-01 1.15849149e+00
-1.70854199e+00 -1.28267884e+00 -3.45847607e-01 4.83256638e-01
1.36963397e-01 8.19794834e-02 8.84958088e-01 1.96781456e-01
-5.44691235e-02 4.12385315e-01 -2.79765413e-03 -6.71708286e-02
8.51937175e-01 -1.45918918e+00 6.94047332e-01 6.22356415e-01
3.64322871e-01 9.35609460e-01 6.94204450e-01 -1.84687629e-01
-1.84779620e+00 -7.55288243e-01 8.34785223e-01 -9.29956913e-01
7.85803258e-01 -7.22553074e-01 -1.05145681e+00 7.24934697e-01
-5.17859720e-02 3.77639621e-01 4.40336257e-01 3.66997086e-02
-5.03551364e-01 -2.84654405e-02 -1.13007188e+00 3.13377231e-01
1.35660613e+00 -8.49078894e-01 -5.37336349e-01 3.40834588e-01
7.42176592e-01 -4.91377205e-01 -1.13837230e+00 3.13663095e-01
5.63133121e-01 -9.92822289e-01 1.21096027e+00 -6.14757299e-01
2.93712616e-01 -4.30420220e-01 -2.76098162e-01 -1.37071168e+00
-4.81123269e-01 -5.77118516e-01 -1.00388460e-01 9.09637332e-01
1.65006757e-01 -6.92869902e-01 6.86737239e-01 6.72344506e-01
-3.03699970e-01 -8.60224783e-01 -1.10998809e+00 -7.74282932e-01
4.25187051e-01 -3.46424371e-01 5.96315622e-01 1.02860498e+00
-2.91075110e-01 2.48296022e-01 -4.19237241e-02 2.57890463e-01
6.76890075e-01 9.84647945e-02 9.68751729e-01 -1.43865013e+00
-2.42766812e-01 -5.37319362e-01 -8.91791105e-01 -1.26702619e+00
4.82141584e-01 -1.36868572e+00 -7.47234225e-02 -1.40272546e+00
-1.50777891e-01 -5.20547450e-01 -1.18521221e-01 4.59517896e-01
8.82238448e-02 3.33821118e-01 4.34171706e-01 5.13373852e-01
-1.58964962e-01 5.73007345e-01 1.11401677e+00 7.49981478e-02
-1.30470738e-01 -1.25425398e-01 -5.46054363e-01 5.83584726e-01
3.53262872e-01 -4.05334741e-01 -1.08830146e-01 -7.24398911e-01
2.95986414e-01 -3.78409177e-01 7.41107225e-01 -6.53713703e-01
2.84830809e-01 3.30644585e-02 2.59220660e-01 -3.25378299e-01
5.41820228e-01 -1.15699434e+00 1.04924962e-01 2.73767114e-01
-1.23318695e-01 3.24269891e-01 1.49830997e-01 4.59535897e-01
-1.50005117e-01 8.88874009e-02 6.97337031e-01 8.62820819e-03
-3.97816867e-01 4.06349897e-01 3.88176501e-01 -2.32917160e-01
7.74742782e-01 -5.01743793e-01 -5.19486964e-02 -1.89147457e-01
-6.57072008e-01 -5.24608884e-03 1.02500057e+00 5.11794209e-01
5.97657561e-01 -1.51741910e+00 -7.66752660e-01 5.64908504e-01
2.62378931e-01 1.69524848e-01 -3.29724789e-01 1.08278799e+00
-6.29189372e-01 4.00310725e-01 -2.12856829e-01 -1.16250348e+00
-7.85954773e-01 2.74292260e-01 4.23395306e-01 -3.28482613e-02
-8.05971861e-01 6.82437837e-01 1.34524211e-01 -8.96989703e-01
5.79431653e-02 -4.17485356e-01 2.25387901e-01 -5.43409251e-02
-4.65358905e-02 2.87965983e-01 3.84167045e-01 -8.43643725e-01
-5.19559920e-01 1.12946272e+00 3.46822478e-02 -1.09944768e-01
1.68344450e+00 7.10837469e-02 -2.20791355e-01 4.85092998e-01
1.83083773e+00 1.19803131e-01 -1.42317605e+00 -2.82405078e-01
1.61537975e-01 -7.42962480e-01 3.95729914e-02 -2.59967327e-01
-7.78308034e-01 1.04023004e+00 6.01153135e-01 1.26740158e-01
6.71157062e-01 1.59035221e-01 4.79979247e-01 3.31560433e-01
3.82421970e-01 -5.30208886e-01 -4.90975380e-02 6.20771885e-01
1.21781051e+00 -1.32940459e+00 1.86106369e-01 -4.03722793e-01
-1.56950682e-01 1.27747333e+00 2.01199025e-01 -7.87493050e-01
7.55336940e-01 9.01125371e-02 -7.97860920e-02 -2.99106270e-01
-4.69636828e-01 -2.03478962e-01 6.72969878e-01 5.43320000e-01
4.58726257e-01 -2.60807753e-01 -6.56567663e-02 -9.59008634e-02
-5.46669185e-01 -1.94282427e-01 2.02385724e-01 7.71628439e-01
-2.51624316e-01 -1.45600009e+00 -3.61458659e-01 2.30707720e-01
-5.39256930e-02 5.75460084e-02 -5.58999777e-01 9.87019420e-01
-1.18936345e-01 3.38612646e-01 2.32060164e-01 -3.80633265e-01
7.85554528e-01 1.84828132e-01 7.78915942e-01 -6.82832301e-01
-4.25736606e-01 -9.21023116e-02 -4.22130883e-01 -8.41752827e-01
-4.54146236e-01 -6.29652262e-01 -1.16660583e+00 -2.02546000e-01
-1.37962997e-01 -9.48815644e-02 9.56897736e-01 7.53513038e-01
6.59671247e-01 9.34675336e-02 6.58904612e-01 -1.33401775e+00
-8.15471172e-01 -7.54033327e-01 -2.47098997e-01 7.91340351e-01
6.17159426e-01 -9.03144419e-01 -4.95533884e-01 1.09911012e-03] | [8.214181900024414, -3.181184768676758] |
666d2faf-24d0-4791-b325-a8268c35deda | kprnet-improving-projection-based-lidar | 2007.12668 | null | https://arxiv.org/abs/2007.12668v2 | https://arxiv.org/pdf/2007.12668v2.pdf | KPRNet: Improving projection-based LiDAR semantic segmentation | Semantic segmentation is an important component in the perception systems of autonomous vehicles. In this work, we adopt recent advances in both image and point cloud segmentation to achieve a better accuracy in the task of segmenting LiDAR scans. KPRNet improves the convolutional neural network architecture of 2D projection methods and utilizes KPConv to replace the commonly used post-processing techniques with a learnable point-wise component which allows us to obtain more accurate 3D labels. With these improvements our model outperforms the current best method on the SemanticKITTI benchmark, reaching an mIoU of 63.1. | ['Olaf Booij', 'Deyvid Kochanov', 'Fatemeh Karimi Nejadasl'] | 2020-07-24 | null | null | null | null | ['lidar-semantic-segmentation'] | ['computer-vision'] | [ 3.48540731e-02 1.33775681e-01 -3.23275238e-01 -8.12684774e-01
-4.06600863e-01 -4.31072235e-01 6.10437572e-01 -2.26464495e-01
-6.74643993e-01 8.08568522e-02 -5.62485158e-01 -5.51137269e-01
2.54440904e-01 -8.97241116e-01 -1.08623910e+00 -2.46214405e-01
2.52891779e-01 1.13685346e+00 8.60233009e-01 -2.61044592e-01
1.77120402e-01 1.12266278e+00 -1.62448895e+00 -2.48273369e-02
7.31313050e-01 1.31659067e+00 1.93368867e-01 3.35990489e-01
-6.36671543e-01 2.16160849e-01 6.05181083e-02 -2.99147785e-01
6.89439178e-01 4.77637172e-01 -7.18010664e-01 -4.28247526e-02
8.75013828e-01 -3.53526920e-01 -5.16131371e-02 1.05906093e+00
1.30591378e-01 1.04497653e-03 7.74341643e-01 -1.25320399e+00
-1.37391075e-01 4.70860809e-01 -8.23648810e-01 -1.40319645e-01
-2.37281814e-01 -9.35289711e-02 5.49477041e-01 -1.11756408e+00
4.42362815e-01 1.53468239e+00 8.75793874e-01 3.61417323e-01
-9.47645903e-01 -8.78466964e-01 2.18414776e-02 5.14201999e-01
-1.55237877e+00 -1.77276403e-01 7.91699111e-01 -4.53970969e-01
1.16674984e+00 -9.26479399e-02 7.37067699e-01 6.84926867e-01
5.69809750e-02 7.16333151e-01 9.73535120e-01 -6.09312952e-02
2.52522439e-01 -3.28182764e-02 3.41210723e-01 6.82157755e-01
2.28113860e-01 4.86147031e-02 9.81509499e-03 2.30842605e-01
6.63983464e-01 1.31392866e-01 4.12865728e-01 -7.07869112e-01
-8.84289384e-01 9.34987187e-01 9.73135352e-01 -6.54251575e-02
-4.33329791e-01 5.29259861e-01 7.89634734e-02 -1.49180278e-01
4.93636787e-01 1.43210977e-01 -5.49639046e-01 1.20700166e-01
-9.07313764e-01 6.28082901e-02 5.37464082e-01 1.04903817e+00
1.06138766e+00 1.33392245e-01 1.73775017e-01 7.36859977e-01
4.84914422e-01 9.42986488e-01 -1.51987057e-02 -1.45794857e+00
3.40474844e-01 8.77574503e-01 -8.75809491e-02 -5.23452818e-01
-6.20323300e-01 -3.73238355e-01 -4.48348016e-01 6.66528106e-01
1.07775360e-01 2.38180235e-01 -1.72734511e+00 9.01647270e-01
4.36897129e-01 4.62962449e-01 -9.04044807e-02 7.58415639e-01
8.36628675e-01 5.14012754e-01 2.71268368e-01 6.43814206e-01
1.25635707e+00 -1.09150755e+00 -1.76881149e-01 -5.25628626e-01
3.94864500e-01 -5.45466602e-01 7.63145864e-01 4.11733866e-01
-6.82900012e-01 -7.83483088e-01 -1.26625657e+00 -3.51972073e-01
-6.77126348e-01 1.93221476e-02 6.17743194e-01 7.01725125e-01
-1.11888564e+00 4.74857539e-01 -1.04886270e+00 -3.25828880e-01
1.00697088e+00 5.45539737e-01 -4.37860340e-01 -2.39561602e-01
-6.48328483e-01 1.15294898e+00 7.29768693e-01 3.56892049e-02
-6.49424493e-01 -7.61781394e-01 -7.63611019e-01 -1.75628126e-01
6.31694615e-01 -6.15714312e-01 1.35343969e+00 -3.52774203e-01
-1.49405730e+00 1.08413351e+00 -6.14677854e-02 -9.70874012e-01
7.11153448e-01 -5.34613848e-01 3.88261192e-02 2.44617298e-01
4.11810493e-03 1.58184898e+00 6.91951990e-01 -1.34222925e+00
-8.22994053e-01 -5.29164314e-01 -1.14616789e-01 2.53739983e-01
1.21594839e-01 -4.88853872e-01 -9.66306150e-01 8.31896737e-02
5.14953256e-01 -1.29974413e+00 -5.03610790e-01 8.79524127e-02
-4.07005757e-01 -4.85526055e-01 1.32809186e+00 -3.64741176e-01
1.03563838e-01 -2.01087475e+00 -1.10099293e-01 1.68593198e-01
3.09231341e-01 6.40582204e-01 -4.40225042e-02 -1.68758154e-01
1.19998917e-01 1.78476498e-01 -4.62565958e-01 -7.49949694e-01
4.99209762e-02 6.18515790e-01 -1.38493359e-01 2.66162127e-01
3.68733943e-01 1.23200977e+00 -4.34974551e-01 -5.75347126e-01
7.15308726e-01 5.64126372e-01 -4.72827405e-01 -6.80935755e-02
-6.02369428e-01 3.33436221e-01 -3.74686092e-01 5.55630803e-01
1.21189916e+00 1.56630829e-01 -4.98637855e-01 -5.47230728e-02
-3.08400512e-01 9.06409919e-02 -8.52395535e-01 1.90804589e+00
-2.13998914e-01 5.85202694e-01 -6.98729232e-02 -9.82458234e-01
1.03785813e+00 -1.81708604e-01 6.58935666e-01 -6.42692745e-01
2.92615920e-01 1.84994087e-01 -1.37760624e-01 -2.23668277e-01
5.65674722e-01 1.67050585e-01 8.82008448e-02 -9.83229950e-02
-5.51402383e-02 -8.33978117e-01 -3.29041220e-02 1.41844779e-01
5.81816852e-01 2.57729799e-01 -1.31938681e-01 -2.48585969e-01
3.98799449e-01 5.30426323e-01 3.69025856e-01 6.24306500e-01
-1.01515874e-01 7.29743302e-01 2.27004007e-01 -4.52706605e-01
-1.07767963e+00 -1.03676450e+00 -4.68672246e-01 7.63153493e-01
3.93913269e-01 6.67308271e-02 -8.46652389e-01 -8.75217021e-01
3.86102796e-01 1.02857792e+00 -4.58871037e-01 -3.25299916e-03
-7.18972206e-01 -3.98172915e-01 6.18446231e-01 9.40929115e-01
8.40016067e-01 -7.39915609e-01 -6.32721841e-01 -1.74592820e-03
1.27730146e-01 -1.64763224e+00 1.32215440e-01 3.61128479e-01
-9.51771498e-01 -9.75554883e-01 -2.31066912e-01 -6.60999238e-01
3.54685754e-01 5.17959177e-01 9.50671315e-01 -1.12379804e-01
-8.23508948e-02 1.02916144e-01 -2.51084805e-01 -1.04646266e+00
-2.04493254e-01 4.67055500e-01 -2.10534468e-01 -4.15708512e-01
7.36571848e-01 -4.17544186e-01 -3.83292645e-01 3.12277973e-01
-7.58224189e-01 1.18552305e-01 6.08947814e-01 2.39492089e-01
9.34241593e-01 -2.12458596e-01 2.21577808e-02 -9.38854933e-01
-1.28270492e-01 -2.24714905e-01 -9.40057933e-01 -3.20515007e-01
-8.22467327e-01 2.22753268e-03 1.75433338e-01 -2.00251434e-02
-7.55266368e-01 6.21861875e-01 -6.95081174e-01 -7.06845045e-01
-6.50542736e-01 1.24122120e-01 -5.72559573e-02 -4.04286146e-01
4.50967103e-01 -1.67606711e-01 8.26703236e-02 -7.61595249e-01
7.52561510e-01 4.78950113e-01 5.75936615e-01 -1.13700204e-01
9.33726788e-01 8.02017093e-01 3.52877855e-01 -8.72196257e-01
-7.81205535e-01 -7.01050043e-01 -1.17500794e+00 -4.26792577e-02
1.16107130e+00 -1.10527766e+00 -4.28673118e-01 4.67404842e-01
-1.21491241e+00 -3.07608783e-01 -3.56724054e-01 3.99573386e-01
-5.30434370e-01 1.87186763e-01 -3.07482898e-01 -3.08168828e-01
-2.57910311e-01 -1.44080746e+00 1.44101906e+00 3.38153392e-01
3.26587081e-01 -4.82750058e-01 -1.41555965e-01 5.61039090e-01
2.54767716e-01 6.57704920e-02 7.21658289e-01 -7.77076006e-01
-8.38381052e-01 -1.83856368e-01 -5.89922965e-01 6.13082230e-01
-3.92363399e-01 1.37242928e-01 -1.11382139e+00 2.80781865e-01
-1.60252333e-01 3.20885926e-02 1.25852990e+00 4.95463878e-01
1.24823213e+00 4.71650541e-01 -6.56806827e-01 1.01125753e+00
1.42661762e+00 1.76287770e-01 6.93340361e-01 2.31042713e-01
1.04026270e+00 4.88285065e-01 6.11940682e-01 -1.92352906e-01
7.27575421e-01 7.03107715e-01 1.06155097e+00 -2.39865631e-01
-3.31390440e-01 -2.22076833e-01 -2.22040564e-01 5.06047726e-01
-1.16099447e-01 7.00178966e-02 -1.43578351e+00 3.52706552e-01
-1.88536406e+00 -4.47082400e-01 -5.83018661e-01 1.74234045e+00
2.01534957e-01 4.46696401e-01 -1.65146902e-01 -1.12110786e-01
4.10927802e-01 -8.39702711e-02 -6.77119613e-01 -3.46416563e-01
3.75614762e-02 4.09903020e-01 1.19400632e+00 5.37400842e-01
-1.44646740e+00 1.61622155e+00 6.49858522e+00 7.50110328e-01
-1.14721680e+00 1.22302473e-01 2.41024509e-01 1.54313207e-01
5.11972047e-02 -1.53155774e-01 -1.02164125e+00 1.52434528e-01
8.61144602e-01 3.91299903e-01 7.43179843e-02 1.33104432e+00
2.67206170e-02 -3.25841039e-01 -9.68610704e-01 9.80683386e-01
-1.10607026e-02 -1.33740604e+00 -4.39519202e-03 2.10421637e-01
7.24590778e-01 1.04777551e+00 4.53736708e-02 3.13095570e-01
4.05439854e-01 -1.09925127e+00 8.27925682e-01 3.31039011e-01
5.52657425e-01 -8.97697687e-01 9.09146070e-01 5.03554583e-01
-1.12056231e+00 2.69451402e-02 -6.14214242e-01 6.50626943e-02
3.06055307e-01 6.20660841e-01 -1.41452444e+00 4.46240455e-01
9.15140688e-01 6.70975327e-01 -8.00216436e-01 1.15966690e+00
-3.45135778e-01 5.81525147e-01 -8.97080481e-01 2.70071507e-01
5.28719783e-01 -4.55722302e-01 5.29456377e-01 1.12116718e+00
2.07215920e-01 -1.10447675e-01 4.50585574e-01 7.94645429e-01
-1.26669765e-01 -1.53510243e-01 -7.03812599e-01 3.42694104e-01
3.69200051e-01 1.38292909e+00 -1.17035484e+00 -4.36441451e-01
-8.22549760e-02 6.44926548e-01 1.27838552e-01 7.24816099e-02
-9.22896326e-01 -9.82963890e-02 8.51805210e-01 6.71275184e-02
6.49291277e-01 -8.31793308e-01 -6.92848623e-01 -6.01489544e-01
-2.09830478e-01 -1.30013928e-01 -1.06713049e-01 -8.74229968e-01
-6.29703641e-01 5.79074681e-01 3.64786834e-01 -8.95736814e-01
-1.97585031e-01 -8.24949086e-01 -3.45101386e-01 7.01791346e-01
-2.04415822e+00 -1.45288146e+00 -5.47525167e-01 4.12566632e-01
6.32882714e-01 1.45460054e-01 3.95371079e-01 2.63745189e-01
-3.60357791e-01 -9.12992191e-03 -1.18671969e-01 1.08700946e-01
2.79605150e-01 -1.28970432e+00 9.42751765e-01 7.85906911e-01
2.29096174e-01 -6.50566490e-03 3.48434240e-01 -7.05512464e-01
-1.09575462e+00 -1.42592525e+00 4.18260336e-01 -7.50867188e-01
3.00822169e-01 -3.23777020e-01 -8.65083098e-01 6.73228621e-01
-7.58176446e-02 1.29734045e-02 1.55123115e-01 -3.02286953e-01
-3.95595491e-01 -1.08799040e-01 -1.11666679e+00 3.22453827e-01
1.27590680e+00 -2.49034330e-01 -6.60796046e-01 1.84438407e-01
1.18533051e+00 -6.37285352e-01 -5.09258032e-01 9.56555009e-01
3.56197953e-01 -8.00423265e-01 1.28910279e+00 -2.83533633e-01
6.70801252e-02 -6.25799835e-01 -2.06061259e-01 -9.87304211e-01
-6.97167441e-02 1.93569675e-01 1.01562209e-01 8.65108788e-01
3.29411715e-01 -7.00895905e-01 1.22567487e+00 2.93680042e-01
-5.67662954e-01 -5.49403250e-01 -1.05952132e+00 -6.78097367e-01
1.70771062e-01 -9.77382123e-01 7.28700519e-01 4.72282737e-01
-9.78596151e-01 2.80994356e-01 2.02886134e-01 2.98888475e-01
6.86618030e-01 8.80910978e-02 9.23343480e-01 -1.76852214e+00
5.37023783e-01 -5.85528851e-01 -6.89340770e-01 -1.22522616e+00
3.20627332e-01 -1.05215669e+00 2.41818056e-01 -1.68475938e+00
-2.19206855e-01 -5.86726248e-01 -9.43075120e-02 5.94338357e-01
2.11139634e-01 8.44935179e-01 3.19437414e-01 2.00315908e-01
-4.98206079e-01 5.25306463e-01 9.57014978e-01 -3.27108413e-01
-1.56657591e-01 2.27376387e-01 -3.14403266e-01 1.10385239e+00
9.95157301e-01 -5.76850295e-01 -2.80354708e-01 -6.86477125e-01
-2.05858909e-02 -6.50758624e-01 4.75097895e-01 -1.40418458e+00
1.57304212e-01 -1.37418164e-02 3.34258258e-01 -1.54517174e+00
7.91161060e-01 -1.13843834e+00 1.09482974e-01 5.25172710e-01
1.70658067e-01 -6.30913824e-02 3.71961355e-01 4.47163343e-01
-1.18045798e-02 -1.52085930e-01 8.38899791e-01 -1.07882097e-01
-1.17967367e+00 5.33968031e-01 -1.45655483e-01 -5.15381038e-01
1.07768357e+00 -5.88401914e-01 -1.15540355e-01 1.18839033e-01
-5.03310859e-01 5.08576512e-01 7.04215884e-01 6.23924017e-01
6.52868986e-01 -1.04467833e+00 -4.90103215e-01 2.67495215e-01
1.08685091e-01 7.87984133e-01 -6.67167315e-03 8.67273569e-01
-1.12576938e+00 5.99286973e-01 -2.41211697e-01 -1.34100258e+00
-1.05078757e+00 1.54513597e-01 2.36444205e-01 3.08282137e-01
-7.56178379e-01 8.53695393e-01 -7.61437044e-02 -7.96559811e-01
2.84446210e-01 -6.96026325e-01 -4.25996155e-01 -7.80055001e-02
-4.48852405e-02 4.22583818e-01 4.48717445e-01 -8.61576855e-01
-4.39689428e-01 1.03592014e+00 -3.27236988e-02 -2.57888087e-03
1.33522511e+00 4.14798185e-02 9.22891870e-03 2.74919420e-01
1.13637912e+00 -3.20235103e-01 -1.53909945e+00 -1.71778858e-01
1.73297331e-01 -2.11585641e-01 3.39698762e-01 -6.39168799e-01
-1.13179517e+00 1.15126789e+00 8.61207485e-01 -2.84430496e-02
6.64734542e-01 2.84961373e-01 9.04169798e-01 6.21130168e-01
5.26735127e-01 -1.13323927e+00 -3.37684512e-01 8.98663521e-01
5.94385445e-01 -1.49216366e+00 1.61061645e-01 -7.89491475e-01
-4.82068598e-01 9.77294922e-01 7.04649091e-01 -2.85426050e-01
8.07648897e-01 1.51925981e-01 3.87381822e-01 -3.11627835e-01
-1.09739669e-01 -5.31945944e-01 4.80174899e-01 7.86059737e-01
-2.37186700e-01 3.84329081e-01 -1.54347578e-02 9.02680159e-02
-3.87477368e-01 -1.83879495e-01 1.77112803e-01 6.34967506e-01
-7.08909273e-01 -1.05644166e+00 -3.33229184e-01 3.30742776e-01
3.91600002e-03 2.25884363e-01 -4.03548360e-01 9.63584006e-01
4.41043854e-01 5.96046507e-01 2.53562897e-01 -5.53052247e-01
3.94714952e-01 2.33015612e-01 2.68696636e-01 -6.74892485e-01
-1.13019153e-01 -9.13861617e-02 -1.13377340e-01 -8.75539184e-01
-6.19679570e-01 -5.68248212e-01 -1.56112778e+00 -1.39338225e-01
-3.32507640e-01 -2.88876176e-01 1.43341768e+00 1.04213715e+00
3.83746177e-01 4.09048021e-01 3.42398494e-01 -1.44432032e+00
-1.98911369e-01 -8.24692607e-01 -9.24161673e-02 1.06905080e-01
-4.01915936e-03 -8.85328114e-01 1.60682261e-01 -1.28000215e-01] | [8.07308578491211, -2.830575704574585] |
94d0ffed-0a21-403b-a3b7-c4dd0cb1b88a | a-new-pattern-recognition-method-for | null | null | http://dx.doi.org/10.4236/jbise.2014.710081 | https://pdfs.semanticscholar.org/7fcb/6e4f06394bfc671165946dadb5b9b80add38.pdf | A New Pattern Recognition Method for Detection and Localization of Myocardial Infarction Using T-Wave Integral and Total Integral as Extracted Features from One Cycle of ECG Signal | In this paper we used two new features i.e. T-wave integral and total integral as extracted feature from one cycle of normal and patient ECG signals to detection and localization of myocardial infarction (MI) in left ventricle of heart. In our previous work we used some features of body surface potential map data for this aim. But we know the standard ECG is more popular, so we focused our detection and localization of MI on standard ECG. We use the T-wave integral because this feature is important impression of T-wave in MI. The second feature in this research is total integral of one ECG cycle, because we believe that the MI affects the morphology of the ECG signal which leads to total integral changes. We used some pattern recognition method such as Artificial Neural Network (ANN) to detect and localize the MI, because this method has very good accuracy for classification of normal signal and abnormal signal. We used one type of Radial Basis Function (RBF) that called Probabilistic Neural Network (PNN) because of its nonlinearity property, and used other classifier such as k-Nearest Neighbors (KNN), Multilayer Perceptron (MLP) and Naive Bayes Classification. We used PhysioNet database as our training and test data. We reached over 76% for accuracy in test data for localization and over 94% for detection of MI. Main advantages of our method are simplicity and its good accuracy. Also we can improve the accuracy of classification by adding more features in this method. A simple method based on using only two features which were extracted from standard ECG is presented and has good accuracy in MI localization. | ['Naser Safdarian', 'Gholamreza Attarodi', 'Nader Jafarnia Dabanloo'] | 2014-08-01 | null | null | null | jbise-vol7-no10-august-2014-2014-8 | ['myocardial-infarction-detection'] | ['medical'] | [ 6.55721277e-02 -2.55280674e-01 2.44057804e-01 -3.48545939e-01
-2.03852698e-01 -3.41281891e-01 -1.59226239e-01 2.78132021e-01
-6.59910440e-01 8.60447049e-01 -1.09506086e-01 -3.76693040e-01
-5.47198892e-01 -1.01170611e+00 -2.21869752e-01 -5.95080972e-01
-5.21038532e-01 4.01334435e-01 5.41022420e-01 -5.17958729e-03
4.81127918e-01 7.63440013e-01 -1.24917614e+00 2.99523056e-01
7.14766085e-01 7.68978238e-01 1.33113772e-01 8.25675189e-01
1.29067287e-01 6.84079349e-01 -6.56536222e-01 4.37863976e-01
6.45043701e-02 -7.68433928e-01 -9.14950490e-01 -4.16537434e-01
-6.85063660e-01 -3.19408663e-02 7.28969127e-02 6.17370427e-01
9.37946916e-01 -1.03283562e-01 1.17793441e+00 -1.00833595e+00
-2.80472450e-02 5.53567290e-01 -5.38342118e-01 6.63644493e-01
2.44328812e-01 -2.38130361e-01 -4.01403233e-02 -6.96889997e-01
3.53845417e-01 6.63903475e-01 1.16454995e+00 2.13906527e-01
-8.83389175e-01 -5.24355888e-01 -8.64598751e-01 7.74055123e-01
-1.66244769e+00 1.47350401e-01 8.59616578e-01 -5.32497883e-01
6.41916752e-01 3.64200234e-01 7.75293529e-01 9.26495418e-02
6.46330714e-01 2.34795779e-01 1.53539062e+00 -8.58130097e-01
-1.40431479e-01 3.87113035e-01 4.55738008e-01 4.09677327e-01
3.41140091e-01 5.64660951e-02 2.12292820e-01 -3.02782714e-01
8.10532272e-01 2.08682463e-01 -4.00163472e-01 2.59227574e-01
-1.23722517e+00 5.36749482e-01 1.20753706e-01 1.05958915e+00
-6.93344712e-01 -3.09268888e-02 3.18275630e-01 2.50332057e-01
-2.78049290e-01 2.77790517e-01 -6.78087413e-01 -4.50948656e-01
-6.39515877e-01 -2.70110611e-02 8.94249439e-01 3.63631956e-02
4.46466357e-01 -1.24226376e-01 -2.28692725e-01 7.06211209e-01
3.28854263e-01 7.54128754e-01 8.39165509e-01 -6.04025245e-01
3.91600206e-02 7.23999977e-01 -2.22137928e-01 -1.27709270e+00
-6.44871771e-01 -4.06284988e-01 -9.73110974e-01 5.35001993e-01
4.58209217e-01 -3.25476825e-01 -7.03622162e-01 1.19279063e+00
5.24588227e-02 1.16156116e-01 1.89998269e-01 8.27953637e-01
8.68999839e-01 7.00325370e-01 -5.04640453e-02 -4.39638048e-01
1.42839932e+00 1.26782879e-02 -6.93747044e-01 5.39543092e-01
6.15728498e-01 -8.60091507e-01 5.56871057e-01 6.66879654e-01
-5.66210628e-01 -5.12506843e-01 -9.08249259e-01 6.98497295e-01
-3.72829437e-01 2.54857302e-01 2.93899238e-01 8.87544990e-01
-5.91683686e-01 1.09304392e+00 -7.99867451e-01 -4.94943380e-01
4.38030846e-02 4.51670080e-01 -5.09618700e-01 2.76234925e-01
-1.26513302e+00 1.15919113e+00 4.42927748e-01 1.39211565e-01
2.63355281e-02 -8.38810578e-02 -4.90017951e-01 -1.34294227e-01
-1.89123258e-01 -4.18642610e-01 6.40795112e-01 -8.34592462e-01
-1.22001183e+00 5.61797500e-01 -7.15600848e-02 -2.37081811e-01
1.11244038e-01 2.17518374e-01 -5.00634551e-01 4.14557070e-01
-1.54257447e-01 8.19118246e-02 3.55308384e-01 -7.01682210e-01
-2.82073677e-01 -6.21793389e-01 -5.10363758e-01 -4.17940691e-02
-8.76058787e-02 6.72111660e-02 4.22550768e-01 -4.12414074e-01
7.04081833e-01 -6.93981886e-01 2.66522504e-02 -6.70917809e-01
-2.47176103e-02 -3.33409607e-01 7.42012620e-01 -1.00897133e+00
1.41039681e+00 -1.99069691e+00 -4.23879713e-01 7.20513582e-01
-2.00168341e-02 4.79276359e-01 5.22581697e-01 3.98759156e-01
-2.24932671e-01 3.04891467e-01 -2.97987700e-01 7.49983013e-01
-5.31273067e-01 3.46274376e-01 2.20582396e-01 5.02195239e-01
-6.53423592e-02 5.24646163e-01 -3.58109802e-01 -8.91666055e-01
1.79293811e-01 6.47684991e-01 1.05527565e-02 -1.34577435e-02
7.72120118e-01 5.21944642e-01 -3.42593521e-01 4.15963709e-01
7.03831553e-01 2.39499524e-01 -1.47176415e-01 -5.66422760e-01
1.74917486e-02 -1.62545949e-01 -1.40303075e+00 9.47332740e-01
9.05535929e-03 3.93802434e-01 -6.27155125e-01 -1.38732338e+00
1.38318884e+00 8.88492584e-01 5.20087123e-01 -4.64907736e-01
2.79437721e-01 2.66723037e-01 3.54526848e-01 -1.08237433e+00
-5.41322589e-01 -2.54383445e-01 5.30743957e-01 3.53451669e-01
-2.31326818e-01 1.91405028e-01 1.16978623e-01 -3.55096519e-01
9.00850117e-01 1.82938665e-01 6.81651354e-01 -5.06708503e-01
7.39989758e-01 -7.32834563e-02 7.74415612e-01 6.25202477e-01
-2.22170070e-01 6.96755588e-01 7.05778122e-01 -7.02016473e-01
-6.63136482e-01 -7.19295859e-01 -5.93214631e-01 8.06022808e-02
-1.12917326e-01 1.26938626e-01 -3.74566257e-01 -4.34870720e-01
-1.70783013e-01 3.33329678e-01 -2.03213066e-01 8.01071227e-02
-8.67864430e-01 -1.24490178e+00 7.20737457e-01 4.16200876e-01
6.15998387e-01 -1.23911452e+00 -8.57108951e-01 3.75853598e-01
-2.42459197e-02 -3.41377765e-01 4.36684310e-01 9.08464417e-02
-1.53554630e+00 -1.29998386e+00 -7.93345034e-01 -6.93335354e-01
4.59356487e-01 -4.02067900e-01 8.56248975e-01 2.05453783e-01
-7.11440861e-01 1.42378017e-01 -4.22821134e-01 -6.29298091e-01
-5.47192574e-01 -2.42825344e-01 -9.59100202e-02 8.27614404e-03
2.24263147e-01 -8.69892001e-01 -7.13203728e-01 3.24896425e-01
-6.06940448e-01 -5.82794964e-01 9.56078649e-01 3.54070336e-01
3.31299931e-01 3.96626532e-01 8.97879899e-01 -8.96508992e-01
6.27924323e-01 -3.02895755e-01 -1.96588188e-01 -8.08936507e-02
-9.77654815e-01 -1.22731283e-01 2.94228643e-01 -1.85405508e-01
-5.25122523e-01 -1.53096840e-01 -3.77762139e-01 3.57402079e-02
-4.35817182e-01 6.16677761e-01 8.39562789e-02 -2.07861826e-01
1.10074687e+00 2.37646341e-01 3.96942496e-01 -5.14258385e-01
-6.10304713e-01 9.35537338e-01 2.92708516e-01 -2.65358746e-01
3.41168255e-01 4.52925377e-02 6.19137108e-01 -8.50574374e-01
1.00916527e-01 -5.45533180e-01 -5.87695420e-01 -1.77717194e-01
9.09991264e-01 -2.84346521e-01 -1.09035134e+00 4.86641407e-01
-1.15893078e+00 2.60907948e-01 8.77960995e-02 1.20911312e+00
-1.78426266e-01 6.77935004e-01 -5.56873977e-01 -1.17574751e+00
-7.43455708e-01 -6.98504031e-01 6.84600919e-02 2.41061062e-01
-1.25622705e-01 -8.82812321e-01 2.03357026e-01 -2.18432993e-01
6.42710507e-01 7.49650359e-01 9.80722725e-01 -7.03530490e-01
3.16496752e-02 -7.78438807e-01 -9.84491855e-02 7.27717698e-01
1.67741373e-01 -6.41599298e-04 -6.81387722e-01 1.96099430e-01
4.50768560e-01 4.58961010e-01 7.15850234e-01 6.26328766e-01
1.01983869e+00 7.88607225e-02 -5.46735883e-01 2.74227381e-01
1.68612742e+00 9.14868534e-01 1.11579454e+00 2.81773508e-01
4.25631762e-01 2.30704993e-01 6.18212283e-01 1.98265463e-01
1.37203172e-01 4.48052168e-01 8.88336450e-02 -2.87110269e-01
-4.79184696e-03 4.21212137e-01 1.85377803e-02 7.10095227e-01
-8.66813064e-01 2.57220984e-01 -1.16618419e+00 2.84569085e-01
-1.83527410e+00 -1.14291549e+00 -8.38202357e-01 2.57246757e+00
8.16045165e-01 2.27657169e-01 1.87891871e-01 9.75021541e-01
8.56536269e-01 -7.58289456e-01 2.05561638e-01 -4.92393970e-01
4.18429263e-02 5.13119578e-01 4.36407804e-01 3.58890831e-01
-9.76820290e-01 -1.62044257e-01 5.73843288e+00 6.04729474e-01
-1.45414042e+00 9.94334891e-02 6.58820271e-01 5.91460824e-01
3.52631778e-01 -2.10917927e-02 -6.00688100e-01 5.79198480e-01
9.05924499e-01 1.18770853e-01 -6.72298819e-02 6.33630097e-01
2.43918866e-01 -4.02283490e-01 -5.86222649e-01 1.16850066e+00
4.07528989e-02 -9.55149055e-01 -5.23935556e-01 -2.25942522e-01
9.18814987e-02 -2.78037548e-01 -5.92984200e-01 2.16827542e-01
-9.95727241e-01 -1.06939197e+00 -1.26798481e-01 1.03933704e+00
6.40814424e-01 -8.31015885e-01 1.55560708e+00 3.89328182e-01
-1.02565157e+00 5.19498475e-02 -3.30554426e-01 -4.54950035e-01
-2.00455621e-01 8.54753852e-01 -1.17492151e+00 5.29945493e-01
6.59905791e-01 3.96455199e-01 -4.49946880e-01 1.58116400e+00
1.01255640e-01 9.62420166e-01 -7.26197243e-01 -1.39333561e-01
-2.80513942e-01 -9.75982398e-02 5.76404452e-01 9.74142194e-01
3.77332896e-01 2.56828874e-01 -3.70321423e-02 8.00812721e-01
8.00033927e-01 5.32301128e-01 -7.52667427e-01 5.44834256e-01
3.19490731e-01 1.22258353e+00 -1.10541093e+00 -3.36126566e-01
-4.70705936e-03 4.42024142e-01 -5.20681024e-01 5.38170859e-02
-5.61930776e-01 -1.16383004e+00 -4.65176165e-01 4.46945101e-01
-1.77893847e-01 1.95354059e-01 -4.67981011e-01 -6.28361881e-01
6.04503527e-02 -4.77383524e-01 4.34979349e-01 -5.02088606e-01
-9.23935533e-01 7.86143363e-01 3.31601709e-01 -1.20046628e+00
-3.17458391e-01 -7.07033515e-01 -8.81917775e-01 1.35120010e+00
-9.56719995e-01 -7.64517903e-01 -2.04263687e-01 5.54953218e-01
-1.94361359e-01 -2.04178885e-01 1.06392992e+00 4.14757729e-01
-1.39724046e-01 1.77306563e-01 -2.36468166e-01 3.97414446e-01
6.98631048e-01 -1.41534543e+00 -5.42894721e-01 5.88953137e-01
-2.15200037e-01 3.61326724e-01 5.28343678e-01 -6.20440185e-01
-7.23238587e-01 -3.96525472e-01 1.25080216e+00 -2.80311733e-01
-3.29709724e-02 3.33951384e-01 -7.35092461e-01 1.70394450e-01
-7.22006410e-02 8.62001553e-02 6.30973339e-01 -1.68275818e-01
4.67771113e-01 -5.02651334e-01 -1.42553639e+00 2.61636041e-02
1.94739908e-01 2.76826657e-02 -9.53548729e-01 1.75155848e-01
-6.37943298e-02 -3.64731438e-02 -1.22973680e+00 1.01829493e+00
9.47798133e-01 -1.21645534e+00 9.01516318e-01 -3.95017862e-02
-1.80006295e-01 -5.07859230e-01 1.93324029e-01 -8.69924366e-01
-4.01291132e-01 -3.26944292e-01 4.77538526e-01 1.05869603e+00
4.60285813e-01 -1.06346691e+00 5.45110822e-01 2.11876869e-01
1.45721897e-01 -9.73557830e-01 -6.74694419e-01 -4.67450052e-01
-2.17544883e-01 -1.78365588e-01 1.08477294e-01 8.38701069e-01
2.45025918e-01 2.78266907e-01 -2.29854565e-02 -6.40906245e-02
5.27023673e-01 -1.98374428e-02 2.02573135e-01 -1.54664302e+00
-4.21339124e-01 -1.15182139e-01 -1.02567935e+00 1.69642910e-01
-5.60912549e-01 -7.70407438e-01 -4.19594109e-01 -1.68684232e+00
1.25677228e-01 -7.47759938e-01 -6.86230540e-01 4.09694344e-01
-1.73671722e-01 6.65955722e-01 -7.65656605e-02 1.52722925e-01
3.29169601e-01 -2.89160758e-01 9.99632061e-01 3.73720407e-01
-6.36632919e-01 6.34544373e-01 -1.75501928e-01 8.93441081e-01
9.89587963e-01 -9.78991330e-01 -2.29844064e-01 3.16046357e-01
2.27874607e-01 4.05835152e-01 3.46414000e-01 -1.44954813e+00
5.35158813e-03 2.19806865e-01 9.44864094e-01 -7.42350161e-01
4.48814742e-02 -9.88388658e-01 3.23326796e-01 1.02040327e+00
2.35508949e-01 1.55366346e-01 2.89415009e-02 3.79701629e-02
-2.44298026e-01 -7.88146138e-01 6.38933480e-01 -2.99187660e-01
-4.76716280e-01 -1.41441718e-01 -5.86250186e-01 -3.58014643e-01
1.04368019e+00 -5.92213631e-01 7.76280165e-02 -1.26006842e-01
-1.13870037e+00 -3.64260972e-01 -1.37382597e-01 -2.91326374e-01
6.14914775e-01 -1.08353138e+00 -8.45379710e-01 1.82113126e-01
-1.04094885e-01 -4.25959349e-01 1.66542307e-01 1.75801730e+00
-1.27299893e+00 1.94900870e-01 -6.36265695e-01 -7.55389571e-01
-1.48541474e+00 2.61864245e-01 5.72237134e-01 -1.10610723e-01
-6.98428750e-01 4.43688959e-01 -5.33616424e-01 -1.94215905e-02
2.92636314e-03 -3.32371384e-01 -1.11633086e+00 -1.21245004e-01
4.61783230e-01 7.83401430e-01 6.04539588e-02 -3.92701864e-01
-7.09226251e-01 1.12675786e+00 4.81090069e-01 -8.47653821e-02
9.93903279e-01 1.68936223e-01 -6.52149677e-01 6.98844910e-01
1.10597444e+00 5.06722555e-02 -1.65043294e-01 3.37263256e-01
-6.91792369e-02 -1.64899349e-01 1.16084531e-01 -1.08405805e+00
-6.86800957e-01 8.82797062e-01 1.22938728e+00 4.31742191e-01
1.29458535e+00 -4.37194437e-01 4.90480036e-01 4.67900842e-01
3.70324045e-01 -6.95123494e-01 -5.34992278e-01 2.78675053e-02
5.24348080e-01 -9.32390869e-01 5.94293699e-03 -1.98878318e-01
-5.69231689e-01 1.68796766e+00 2.72215549e-02 -4.55283970e-01
1.25897527e+00 2.66864628e-01 2.36453056e-01 -5.63823059e-03
-2.47144848e-01 -6.84895664e-02 1.93137676e-01 7.35285997e-01
7.15584517e-01 2.20296711e-01 -1.19422877e+00 6.75790429e-01
1.56436756e-01 7.08847642e-01 4.75261748e-01 1.14080429e+00
-8.40452075e-01 -1.21449244e+00 -8.01464260e-01 7.52559483e-01
-1.01786196e+00 7.47014284e-02 2.67830849e-01 7.89106488e-01
6.22884870e-01 9.60365653e-01 -2.75078326e-01 -3.94490302e-01
2.20471933e-01 4.52494860e-01 6.55379474e-01 -1.61772937e-01
-5.79808235e-01 -1.02350931e-03 -7.52545148e-02 -2.10143849e-01
-4.38892752e-01 -3.92981946e-01 -1.58368552e+00 -3.05146687e-02
-3.54833931e-01 6.39629006e-01 1.09530616e+00 9.69806015e-01
1.09726332e-01 5.12362599e-01 5.73283911e-01 -4.25620079e-01
-3.45163792e-01 -1.50960946e+00 -1.02464986e+00 2.14617565e-01
3.61716822e-02 -5.31200767e-01 -3.84660840e-01 2.84708798e-01] | [14.186064720153809, 3.221101760864258] |
5d1c9287-cf14-4311-b9ce-e8ba585b3bbb | microstructural-segmentation-using-a-union-of | null | null | https://www.nature.com/articles/s41598-023-32318-9#Abs1 | https://www.nature.com/articles/s41598-023-32318-9 | Microstructural segmentation using a union of attention guided U-Net models with different color transformed images | Metallographic images or often called the microstructures contain important information about metals, such as strength, toughness, ductility, corrosion resistance, which are used to choose the proper materials for various engineering applications. Thus by understanding the microstructures, one can determine the behaviour of a component made of a particular metal, and can predict the failure of that component in certain conditions. Image segmentation is a powerful technique for determination of morphological features of the microstructure like volume fraction, inclusion morphology, void, and crystal orientations. These are some key factors for determining the physical properties of metal. Therefore, automatic micro-structure characterization using image processing is useful for industrial applications which currently adopts deep learning-based segmentation models. In this paper, we propose a metallographic image segmentation method using an ensemble of modified U-Nets. Three U-Net models having the same architecture are separately fed with color transformed imaged (RGB, HSV and YUV). We improvise the U-Net with dilated convolutions and attention mechanisms to get finer grained features. Then we apply the sum-rule-based ensemble method on the outcomes of U-Net models to get the final prediction mask. We achieve the mean intersection over union (IoU) score of 0.677 on a publicly available standard dataset, namely MetalDAM. We also show that the proposed method obtains results comparable to state-of-the-art methods with fewer number of model parameters. The source code of the proposed work can be found at https://github.com/mb16biswas/attention-unet. | ['Ram Sarkar', 'Dmitry Kaplun', 'Aleksandr Sinitca', 'Shibaprasad Sen', 'Rishav Pramanik', 'Momojit Biswas'] | 2023-04-07 | null | null | null | scientific-reports-2023-4 | ['2d-semantic-segmentation'] | ['computer-vision'] | [ 2.03390837e-01 -2.65111536e-01 5.30458391e-02 -2.95053601e-01
-4.33540314e-01 -1.51832789e-01 1.77967384e-01 1.00783324e-02
-1.22165881e-01 5.57207882e-01 -5.03362477e-01 -2.92926013e-01
-2.17246369e-01 -1.18602490e+00 -8.71001124e-01 -1.02714717e+00
2.89856344e-01 4.64798123e-01 3.15296888e-01 -1.37560278e-01
7.01090872e-01 5.78791559e-01 -1.67392087e+00 4.04039800e-01
8.05628717e-01 1.69823039e+00 5.06994486e-01 5.20842552e-01
6.02331683e-02 5.27306318e-01 -1.02498181e-01 -2.20362307e-03
1.14543870e-01 -4.08181511e-02 -8.53581607e-01 2.16658428e-01
-1.06110021e-01 -1.53943911e-01 -1.35411441e-01 1.04081357e+00
1.21862501e-01 2.97507763e-01 1.12602031e+00 -8.67573023e-01
-9.39637780e-01 5.77407479e-01 -5.66488266e-01 2.74260100e-02
-3.24929990e-02 1.00679390e-01 8.93932998e-01 -8.00074100e-01
4.59629536e-01 1.12623918e+00 3.27199608e-01 2.89081246e-01
-9.23757792e-01 -6.22934341e-01 -2.47475840e-02 6.51364267e-01
-1.16405523e+00 -1.72389597e-01 7.81017244e-01 -4.94218290e-01
7.74817467e-01 2.73472786e-01 2.49786869e-01 5.70609331e-01
7.39397585e-01 6.28734112e-01 1.11815739e+00 -2.87401170e-01
1.86775610e-01 7.68154114e-03 3.91420633e-01 7.81888425e-01
7.63861416e-03 -1.54493883e-01 1.32839471e-01 2.34500825e-01
9.48444009e-01 1.97897822e-01 -2.78075397e-01 2.10742056e-01
-7.89691985e-01 5.50467432e-01 5.88240683e-01 4.60069805e-01
-4.02853101e-01 1.60073280e-01 1.06411003e-01 2.29536444e-01
5.09723246e-01 4.87775087e-01 -6.32113814e-01 8.69451016e-02
-4.82378244e-01 8.93263593e-02 5.90392649e-01 5.30176401e-01
1.15758157e+00 -4.23401408e-03 1.05521217e-01 1.01727855e+00
3.08148593e-01 4.35747296e-01 5.77276886e-01 -8.70522261e-01
7.04159541e-03 8.00330877e-01 -2.67814577e-01 -7.92824984e-01
-3.56329799e-01 7.05699101e-02 -8.64464045e-01 5.06649137e-01
-6.37095794e-03 -8.00771415e-02 -1.30866122e+00 1.03888321e+00
4.44522947e-01 4.78728712e-02 -1.74182221e-01 7.96761453e-01
9.56420243e-01 8.27849507e-01 -2.19592378e-01 2.26266220e-01
1.28979516e+00 -7.63173342e-01 -4.16255772e-01 -5.45725822e-02
1.78117454e-01 -8.08238447e-01 1.00619841e+00 5.54284573e-01
-7.45370805e-01 -6.99927866e-01 -1.39328039e+00 1.45170644e-01
-5.02264619e-01 1.91529036e-01 4.99269009e-01 5.03538728e-01
-6.03966057e-01 1.16382980e+00 -9.97135520e-01 -2.07280349e-02
4.77707535e-01 5.78487098e-01 -2.29952097e-01 -3.37774828e-02
-1.09690475e+00 6.63771808e-01 3.39209288e-01 3.48568708e-01
-8.67182553e-01 -2.89932072e-01 -7.77057111e-01 -1.89182475e-01
4.65152740e-01 -1.10487789e-01 1.07665122e+00 -6.22217596e-01
-1.53566265e+00 4.54862803e-01 8.44058096e-02 -4.92545739e-02
2.30582580e-01 -1.65812954e-01 -1.40990883e-01 -2.30918117e-02
5.33051509e-03 2.88781762e-01 7.73429513e-01 -1.51098990e+00
-5.81971467e-01 -3.60355169e-01 1.38902083e-01 -1.22832216e-01
-1.28716022e-01 -1.48804054e-01 -6.14193797e-01 -4.64199692e-01
4.83256221e-01 -7.65590131e-01 -2.23321840e-01 -4.20376211e-01
-8.55157197e-01 -4.05694544e-01 1.02027333e+00 -7.48824954e-01
8.12955379e-01 -1.90230763e+00 7.57207200e-02 4.36901391e-01
1.17043115e-01 -6.89241514e-02 1.31554991e-01 8.32674131e-02
-7.34747350e-02 2.22383261e-01 -6.66787863e-01 6.74633831e-02
-1.51295274e-01 -1.27993328e-02 3.09797257e-01 4.38361138e-01
1.15943208e-01 6.99443042e-01 -3.67134780e-01 -3.69249344e-01
3.76816899e-01 2.44281784e-01 -2.33778700e-01 2.25836068e-01
-3.92747283e-01 4.60264117e-01 -6.92719460e-01 1.00892425e+00
6.68571711e-01 -1.92558646e-01 -1.02157213e-01 -5.19415379e-01
-1.30560026e-01 -2.12746337e-01 -1.28280103e+00 1.21079350e+00
-5.42946756e-01 3.55619013e-01 1.58198103e-01 -1.09696352e+00
1.03747463e+00 2.37914756e-01 6.38573170e-01 -7.07029939e-01
7.16748834e-01 3.04893821e-01 1.04465649e-01 -7.48206019e-01
2.83003062e-01 5.79877533e-02 4.23508212e-02 5.63685656e-01
-1.46527678e-01 -3.76816779e-01 3.08549881e-01 -2.14053094e-01
8.53072047e-01 1.64594650e-01 -2.41839498e-01 -2.61400759e-01
6.60945117e-01 -1.83977336e-01 4.53727692e-01 4.51654255e-01
1.74208358e-01 7.15519726e-01 3.43774050e-01 -4.36319023e-01
-1.07363594e+00 -8.20784926e-01 -4.78177845e-01 6.96276963e-01
3.91667366e-01 3.08796525e-01 -9.99792755e-01 -3.29189360e-01
3.80790643e-02 3.93140048e-01 -8.04346561e-01 -1.71555907e-01
-5.89703679e-01 -9.41860616e-01 4.26164046e-02 6.42937005e-01
7.22908676e-01 -1.45640385e+00 -4.96458292e-01 3.20889324e-01
2.23643407e-02 -1.05503428e+00 -9.83170867e-02 3.41817528e-01
-8.53372753e-01 -1.42456007e+00 -4.33838755e-01 -8.98875177e-01
5.70452988e-01 -1.36175498e-01 8.53114247e-01 4.20859277e-01
-4.23789322e-01 9.05131549e-03 -3.96423995e-01 -3.44462395e-01
-5.01942217e-01 4.28295583e-02 1.19639486e-02 1.66939422e-01
7.51534924e-02 -5.07053494e-01 -6.96110904e-01 4.35057133e-01
-1.02781820e+00 4.85519990e-02 6.35490656e-01 5.69919288e-01
9.81795371e-01 5.11959732e-01 4.32866096e-01 -1.11839986e+00
5.02985477e-01 -3.83454621e-01 -3.91664028e-01 2.62937576e-01
-6.80948138e-01 -1.01676963e-01 6.65419936e-01 -2.41413623e-01
-1.08062422e+00 -1.44684717e-01 -2.80882180e-01 -4.17235553e-01
-4.34947133e-01 5.58120489e-01 -3.50828290e-01 -1.27028942e-01
2.44427070e-01 1.15390502e-01 -1.90845490e-01 -4.99150634e-01
6.12774789e-02 9.95660424e-01 4.37708408e-01 -8.72659862e-01
3.82802129e-01 3.44052941e-01 -1.23732306e-01 -9.60723579e-01
-4.04462814e-01 -1.72658771e-01 -6.32115424e-01 -4.32841361e-01
9.74666893e-01 -3.13419372e-01 -7.61273146e-01 9.78007019e-01
-9.43341494e-01 -4.24656034e-01 1.73933804e-01 2.14703307e-01
-4.39546376e-01 2.69667149e-01 -1.13117611e+00 -6.77608252e-01
-5.99187911e-01 -1.64482141e+00 7.82140613e-01 3.87284786e-01
8.61912817e-02 -7.14726210e-01 -3.48526567e-01 5.18855989e-01
7.68757910e-02 2.52715170e-01 1.29400694e+00 -4.79452342e-01
-5.54639697e-01 -3.03182930e-01 -3.49669278e-01 7.02902913e-01
3.84346843e-01 2.68689632e-01 -9.22508955e-01 1.22846231e-01
1.56826064e-01 -2.43594915e-01 1.08804345e+00 7.44780660e-01
1.64286256e+00 1.46564066e-01 -2.20895931e-01 4.79212105e-01
1.58132720e+00 7.14370608e-01 7.82679975e-01 4.60551322e-01
1.07954216e+00 5.12816370e-01 4.69938993e-01 3.42772454e-01
1.01564080e-01 1.78670034e-01 6.66314363e-01 4.50932160e-02
-6.35996182e-03 4.91870642e-01 9.69226956e-02 7.36048400e-01
-4.56300348e-01 -1.33879706e-01 -9.40230966e-01 5.08239388e-01
-1.58957314e+00 -4.96461242e-01 -1.48479089e-01 1.75881183e+00
6.20895982e-01 3.72378200e-01 -3.28312010e-01 4.67232764e-01
9.33972239e-01 -1.54752374e-01 -8.80482912e-01 -4.21782911e-01
8.61421376e-02 3.56354713e-01 5.73810399e-01 3.09113145e-01
-1.15826392e+00 7.16839194e-01 5.32957745e+00 1.07759535e+00
-1.30577636e+00 -3.91074456e-02 1.25683475e+00 3.22443485e-01
-1.44358650e-01 -2.99438536e-01 -5.28783619e-01 4.39754367e-01
5.52263558e-01 3.64748567e-01 5.03032446e-01 5.57400703e-01
1.10257849e-01 -4.54291224e-01 -1.00895047e+00 7.43754089e-01
-1.80861115e-01 -1.41853118e+00 -3.20948809e-01 2.42875274e-02
8.61935794e-01 5.18766455e-02 1.17408633e-01 -1.20091759e-01
2.01949164e-01 -9.17794585e-01 8.30390155e-01 4.49833095e-01
8.44346642e-01 -7.26857364e-01 9.74485040e-01 -6.76842853e-02
-1.19969738e+00 -1.23139471e-01 -4.28184062e-01 1.67142376e-01
-1.84389427e-02 8.34097624e-01 -5.24504125e-01 6.25105739e-01
9.75007355e-01 5.85461497e-01 -1.44103914e-01 7.31871307e-01
-8.75073820e-02 6.29734755e-01 -4.82204735e-01 -3.54478545e-02
2.50732809e-01 -5.51060557e-01 1.85172677e-01 6.66486025e-01
3.87953639e-01 5.23234531e-02 1.68048128e-01 1.03951430e+00
-2.04091862e-01 1.00244053e-01 -1.93790033e-01 9.06027406e-02
3.41738701e-01 1.39723885e+00 -1.32830238e+00 -2.55587608e-01
-2.10473225e-01 6.98646426e-01 6.76627178e-03 2.77603209e-01
-7.18815148e-01 -4.67049718e-01 5.05384386e-01 3.66718657e-02
3.50198269e-01 -2.49721017e-02 -8.02676380e-01 -6.33138895e-01
-9.24835429e-02 -6.48007512e-01 3.95489670e-02 -7.72228718e-01
-1.27129650e+00 7.52330482e-01 -1.97931707e-01 -8.79848599e-01
3.06124479e-01 -1.04992938e+00 -9.35366809e-01 7.86901176e-01
-1.12246346e+00 -1.06824684e+00 -1.87948182e-01 1.61064535e-01
7.84871995e-01 1.92561168e-02 4.09983367e-01 1.94536567e-01
-1.03881991e+00 6.84659109e-02 3.66269171e-01 1.22464024e-01
3.52582991e-01 -1.17668569e+00 3.03234488e-01 5.90474725e-01
-1.69254333e-01 1.41852722e-01 6.96532488e-01 -8.62497747e-01
-1.27279818e+00 -1.09601176e+00 1.38505146e-01 -1.30865872e-01
5.32011986e-01 1.25266220e-02 -1.05583704e+00 3.57433856e-01
4.80206050e-02 -2.49077633e-01 4.13389295e-01 -1.01965904e-01
2.69971579e-01 -9.80343595e-02 -1.21189594e+00 3.32622290e-01
5.31020522e-01 -1.52790263e-01 -1.63091138e-01 2.47255743e-01
6.70517385e-01 -3.80466729e-01 -1.17797208e+00 5.26386321e-01
7.36741006e-01 -1.01050615e+00 7.75413156e-01 -8.38517919e-02
8.67104292e-01 -9.69098061e-02 -3.63904566e-01 -1.09303617e+00
-5.06792605e-01 2.76349276e-01 1.85455889e-01 1.13545418e+00
5.59634268e-01 -5.79989731e-01 9.26879883e-01 6.47143960e-01
-4.67538625e-01 -1.35726011e+00 -7.03706443e-01 -3.95468950e-01
1.60508499e-01 -7.47133136e-01 8.80788267e-01 6.07973635e-01
-5.46841860e-01 -8.99532437e-02 4.34321240e-02 1.82952434e-01
4.52652395e-01 2.80604362e-01 1.72050208e-01 -1.31127667e+00
1.08190654e-02 -4.29635167e-01 -2.43552759e-01 -4.92329061e-01
2.23125666e-01 -7.07295001e-01 1.53641775e-01 -1.65201557e+00
1.04230836e-01 -6.68450236e-01 -6.21676266e-01 5.45724392e-01
-1.01506196e-01 4.74299341e-01 -9.64404643e-03 3.12459499e-01
-1.79756284e-01 4.72129285e-01 1.57205856e+00 -5.57147563e-01
-1.71298280e-01 -5.11269979e-02 -4.62356746e-01 7.06254244e-01
1.10528851e+00 -1.36910960e-01 -1.68702543e-01 -4.48111266e-01
-1.88604277e-02 -1.10510066e-01 2.19448999e-01 -1.29227364e+00
1.85019244e-02 -2.55456060e-01 5.61961114e-01 -5.83666861e-01
1.32514358e-01 -8.36781323e-01 1.65552199e-01 4.89528477e-01
-4.00180026e-04 -1.39359832e-01 5.87792918e-02 2.64820188e-01
-1.68249562e-01 -5.12423813e-01 7.57643819e-01 -4.31636781e-01
-9.18188930e-01 6.16759479e-01 -4.19483036e-01 -6.01362348e-01
1.02218640e+00 -4.84756291e-01 -1.43753998e-02 3.10828518e-02
-6.54087782e-01 1.76998630e-01 5.67685902e-01 2.67711073e-01
7.93189526e-01 -1.18174243e+00 -5.50647378e-01 3.37739646e-01
-2.11851299e-01 2.64776319e-01 6.52082801e-01 5.26288331e-01
-8.75042677e-01 -7.52455788e-03 -3.29777658e-01 -5.45812666e-01
-9.40233171e-01 1.13025188e-01 3.75583202e-01 -1.77666247e-01
-4.44917828e-01 9.56027806e-01 2.18601122e-01 -3.25451761e-01
-3.49865705e-01 -4.18537855e-01 -4.93487298e-01 -1.04251817e-01
2.27133319e-01 6.75657511e-01 5.09502590e-01 -6.45040452e-01
-2.77874589e-01 8.55823100e-01 -1.37684539e-01 3.17337453e-01
1.60617566e+00 -5.37921824e-02 -5.59531689e-01 4.85951364e-01
1.19066799e+00 -5.82598329e-01 -1.27405739e+00 3.98850255e-02
-2.61662811e-01 4.55375090e-02 3.02403927e-01 -5.85330606e-01
-1.76031661e+00 8.24782908e-01 7.91242540e-01 2.91154623e-01
1.21273637e+00 9.84668881e-02 1.10675538e+00 2.74200976e-01
2.50473261e-01 -1.23942363e+00 -1.35091901e-01 5.09047270e-01
7.34179795e-01 -1.25726712e+00 6.98972493e-02 -5.32991230e-01
-3.67516071e-01 1.16145813e+00 9.18546677e-01 -3.41194302e-01
1.04895723e+00 3.57462376e-01 2.81216919e-01 -4.78097171e-01
-4.38718319e-01 -2.56972089e-02 2.72390336e-01 4.00138915e-01
4.77399677e-01 2.48774692e-01 2.04548961e-03 8.36130798e-01
-9.09047797e-02 -3.39018643e-01 3.77909899e-01 9.43652570e-01
-6.16317153e-01 -9.84222114e-01 -6.56851888e-01 9.53529835e-01
-6.55626118e-01 5.61288856e-02 -3.37878317e-02 3.82308781e-01
3.37498039e-01 9.78528678e-01 2.17788413e-01 -8.20178628e-01
2.44813129e-01 -1.24486096e-01 4.98006314e-01 -4.92031932e-01
-3.93051922e-01 3.43323499e-02 -1.39134437e-01 -4.14906830e-01
-3.51364195e-01 -4.46370691e-01 -1.40647566e+00 -2.46873945e-01
-5.21253228e-01 -2.49251910e-03 7.02614248e-01 1.10284162e+00
-3.95343155e-02 9.07603800e-01 7.58491218e-01 -1.33931720e+00
-1.13263018e-01 -1.08167303e+00 -9.51834321e-01 5.84516764e-01
-3.67612056e-02 -9.58641469e-01 -9.98909622e-02 1.55703858e-01] | [7.502292156219482, 1.8192315101623535] |
b7635af1-d4f7-4160-aac0-ce643d836bcb | mimetics-towards-understanding-human-actions | 1912.07249 | null | https://arxiv.org/abs/1912.07249v3 | https://arxiv.org/pdf/1912.07249v3.pdf | Mimetics: Towards Understanding Human Actions Out of Context | Recent methods for video action recognition have reached outstanding performances on existing benchmarks. However, they tend to leverage context such as scenes or objects instead of focusing on understanding the human action itself. For instance, a tennis field leads to the prediction playing tennis irrespectively of the actions performed in the video. In contrast, humans have a more complete understanding of actions and can recognize them without context. The best example of out-of-context actions are mimes, that people can typically recognize despite missing relevant objects and scenes. In this paper, we propose to benchmark action recognition methods in such absence of context and introduce a novel dataset, Mimetics, consisting of mimed actions for a subset of 50 classes from the Kinetics benchmark. Our experiments show that (a) state-of-the-art 3D convolutional neural networks obtain disappointing results on such videos, highlighting the lack of true understanding of the human actions and (b) models leveraging body language via human pose are less prone to context biases. In particular, we show that applying a shallow neural network with a single temporal convolution over body pose features transferred to the action recognition problem performs surprisingly well compared to 3D action recognition methods. | ['Grégory Rogez', 'Philippe Weinzaepfel'] | 2019-12-16 | null | null | null | null | ['3d-human-action-recognition'] | ['computer-vision'] | [ 4.38543737e-01 -7.89155886e-02 -2.08840936e-01 -3.56173426e-01
-2.43627846e-01 -4.10530329e-01 9.35272932e-01 -2.97505230e-01
-5.65896332e-01 4.44338918e-01 6.93892002e-01 8.17118436e-02
1.83297560e-01 -4.13481086e-01 -1.05066180e+00 -5.56052446e-01
-2.08017126e-01 3.74696285e-01 2.50054419e-01 -4.43145186e-01
2.40434200e-01 3.16185176e-01 -1.75546169e+00 7.53210068e-01
-6.29007816e-02 1.14254272e+00 -2.35549748e-01 7.31428683e-01
2.76178509e-01 1.39907241e+00 -5.06013036e-01 -2.04250768e-01
5.13375580e-01 -5.82158267e-01 -1.01305735e+00 3.76487821e-01
1.08381009e+00 -8.07211459e-01 -8.55131269e-01 6.57833457e-01
2.72310764e-01 2.40062192e-01 6.45092964e-01 -1.18042123e+00
-5.92091799e-01 4.35242027e-01 -2.00591564e-01 4.20580208e-01
7.91029036e-01 6.05248928e-01 1.01690423e+00 -6.64485991e-01
7.35001802e-01 1.38613951e+00 6.64557099e-01 7.88998544e-01
-1.10584605e+00 -3.05934608e-01 5.25191009e-01 4.50414181e-01
-1.00525463e+00 -5.82470000e-01 5.49198091e-01 -5.99803925e-01
1.34196901e+00 2.16861442e-01 1.00260282e+00 1.80754435e+00
3.44292372e-01 1.24464381e+00 9.08488631e-01 -1.03785180e-01
1.81304559e-01 -6.27444685e-01 -1.13354333e-01 5.02277315e-01
-7.30432346e-02 3.09120148e-01 -9.38770354e-01 3.44922006e-01
8.10144901e-01 1.55179515e-01 -1.53527692e-01 -6.67841971e-01
-1.67294121e+00 3.77404869e-01 3.81125242e-01 2.33195901e-01
-5.26193202e-01 5.33193588e-01 7.22678900e-01 2.84689754e-01
2.88981646e-01 2.30934456e-01 -5.13003767e-01 -4.74458307e-01
-7.35431254e-01 6.08140230e-01 7.84724474e-01 7.35456169e-01
2.81942546e-01 4.39140899e-03 -5.58102548e-01 2.12936118e-01
-6.73560649e-02 4.62272167e-01 5.98519742e-01 -1.06314027e+00
5.48461497e-01 5.90330243e-01 2.21700042e-01 -9.62072849e-01
-5.14900148e-01 -4.79155406e-02 -4.53175962e-01 3.15019101e-01
8.78546596e-01 2.02335879e-01 -1.08136880e+00 1.78541875e+00
2.43801519e-01 1.77129343e-01 2.52071768e-02 1.34794271e+00
7.93776870e-01 1.00724146e-01 2.67269164e-01 3.60820591e-01
1.31097996e+00 -9.95778620e-01 -3.86339426e-01 -6.11792743e-01
7.73179352e-01 -2.90663451e-01 1.05134594e+00 4.19250429e-01
-8.93177092e-01 -8.45195591e-01 -9.21491385e-01 -2.36341104e-01
-4.13683176e-01 1.20969020e-01 7.46049583e-01 4.19114411e-01
-8.49787831e-01 1.00597858e+00 -1.04174125e+00 -6.84664965e-01
5.41636348e-01 2.97522038e-01 -7.91601479e-01 2.37381533e-02
-1.03674066e+00 1.12598383e+00 4.15605664e-01 1.02022171e-01
-1.35298908e+00 -5.25076210e-01 -1.11348593e+00 -1.92312062e-01
7.40235507e-01 -6.34795249e-01 1.41559899e+00 -1.52041888e+00
-1.31530488e+00 1.26634383e+00 2.46064752e-01 -7.36084223e-01
8.96926939e-01 -7.88114965e-01 -1.76008120e-01 3.47995579e-01
6.87433360e-03 8.47089052e-01 9.19937432e-01 -7.15565979e-01
-4.61284161e-01 -4.76168841e-01 4.92125630e-01 2.14212567e-01
7.54503086e-02 -4.69832756e-02 -2.46858388e-01 -6.99053705e-01
-6.67519576e-04 -1.06149924e+00 -1.13914637e-02 3.92944306e-01
-1.82812870e-01 -1.53584570e-01 7.65346944e-01 -6.23942792e-01
7.83669055e-01 -2.05286098e+00 4.71157163e-01 -3.78418982e-01
9.61551145e-02 1.73575476e-01 -1.98278978e-01 4.22324717e-01
-3.46405476e-01 -1.93728074e-01 4.08262126e-02 -1.81352317e-01
3.22758615e-01 5.16460359e-01 -2.87641913e-01 7.29350984e-01
2.69026339e-01 1.18989027e+00 -1.03351724e+00 -2.71978468e-01
4.99434352e-01 3.77585471e-01 -6.30898058e-01 7.57135600e-02
-4.49941695e-01 6.60073638e-01 -3.91970932e-01 7.09720492e-01
-2.35888790e-02 -2.46720314e-01 1.98432073e-01 -2.20198154e-01
1.23038583e-01 3.30852777e-01 -9.52391803e-01 1.99942005e+00
-9.75588784e-02 5.55975437e-01 -7.31517002e-02 -1.20357120e+00
3.76705229e-01 3.26759756e-01 7.96218991e-01 -7.72697389e-01
2.10562989e-01 -4.58863527e-02 1.79925486e-01 -7.13210106e-01
2.49050751e-01 -2.06479967e-01 -2.53396273e-01 3.30738008e-01
2.22292155e-01 1.99703630e-02 2.15223491e-01 6.76054209e-02
1.41052878e+00 8.59644830e-01 4.20530468e-01 -4.17132303e-02
4.50167149e-01 3.86571847e-02 4.07575160e-01 9.21353519e-01
-6.62622094e-01 7.63310432e-01 5.49512744e-01 -9.67573404e-01
-9.61173654e-01 -9.07671452e-01 3.90896261e-01 1.33066082e+00
-2.59388406e-02 -4.91892964e-01 -5.78303695e-01 -8.61617327e-01
1.43513024e-01 4.70399290e-01 -1.19927549e+00 -2.50904560e-01
-7.66749024e-01 -2.14751124e-01 6.09151781e-01 1.09386456e+00
6.51406050e-01 -1.26273024e+00 -1.19209719e+00 6.34488761e-02
-7.25377873e-02 -1.32659662e+00 -4.34192210e-01 1.18613318e-02
-8.16882908e-01 -1.36641347e+00 -7.23559737e-01 -3.38925779e-01
3.29801351e-01 -5.12742847e-02 1.31122315e+00 7.14612603e-02
-3.36180508e-01 9.35008585e-01 -5.56340218e-01 -2.26507291e-01
-3.09525728e-01 -2.58735806e-01 3.16734791e-01 4.49743383e-02
6.36616707e-01 -5.96156418e-01 -7.10883856e-01 4.30741936e-01
-8.06049705e-01 1.17443338e-01 7.78567612e-01 7.51193821e-01
3.50311339e-01 -6.09037280e-01 7.80658722e-02 -4.46996421e-01
-1.32997751e-01 -2.66039103e-01 -2.33982261e-02 9.95684192e-02
1.28929198e-01 -5.80628105e-02 3.40384394e-01 -5.43032110e-01
-8.05656731e-01 4.28970754e-01 -3.43853352e-03 -6.21985972e-01
-6.33408308e-01 1.27413943e-01 -9.85123739e-02 3.86310108e-02
7.47256458e-01 3.20104450e-01 5.06810769e-02 -5.51953733e-01
2.37422645e-01 3.00431978e-02 6.50751293e-01 -7.35957444e-01
3.91395688e-01 7.42258012e-01 2.24770531e-01 -6.72792554e-01
-1.02565169e+00 -4.26341087e-01 -1.17034483e+00 -5.79944730e-01
1.25253332e+00 -9.62641060e-01 -7.72987008e-01 6.14956319e-01
-8.83488178e-01 -7.13871598e-01 -3.74211550e-01 6.06987834e-01
-1.09969854e+00 4.96931344e-01 -7.14910209e-01 -5.91295719e-01
2.28850111e-01 -9.16367054e-01 1.41286671e+00 -2.39921913e-01
-5.86739182e-01 -6.14744306e-01 -1.23675540e-01 5.30241609e-01
1.68084651e-01 5.07091165e-01 4.84133929e-01 -7.11784184e-01
-4.32563722e-01 -3.39094937e-01 9.11268145e-02 3.99975151e-01
1.36667788e-01 -3.76001179e-01 -9.41630125e-01 -5.44699915e-02
-1.23121805e-01 -8.33019078e-01 9.59666550e-01 1.55744433e-01
1.13194132e+00 -2.21146047e-01 -2.26588741e-01 3.22362483e-01
8.48145425e-01 -7.07649142e-02 7.01033473e-01 1.90589681e-01
6.87293172e-01 6.18327081e-01 6.01318955e-01 4.06164169e-01
6.31454661e-02 1.03605103e+00 5.15757859e-01 2.05332190e-01
-2.64354318e-01 -3.72929811e-01 6.66974306e-01 1.29406154e-01
-5.92222273e-01 2.83497050e-02 -9.25963938e-01 3.96608800e-01
-2.01653171e+00 -1.29987204e+00 1.33206725e-01 2.01881957e+00
6.88042998e-01 3.01169604e-01 4.05769289e-01 9.22722071e-02
3.14639598e-01 4.77984548e-01 -7.08122790e-01 -9.70737860e-02
-3.52245495e-02 9.14721638e-02 3.95370930e-01 1.94359571e-02
-1.62486839e+00 9.23306406e-01 6.57209635e+00 4.26309168e-01
-1.01746440e+00 -5.89823313e-02 3.43408227e-01 -5.10786176e-01
4.70049590e-01 -2.47918308e-01 -5.42746723e-01 1.74140945e-01
6.99997723e-01 4.21197057e-01 1.06883496e-01 9.34539318e-01
1.15528822e-01 -2.44742393e-01 -1.86712766e+00 1.01480043e+00
3.73557001e-01 -9.96972382e-01 1.96063951e-01 -9.93687883e-02
4.47136551e-01 -1.48763377e-02 -8.25212598e-02 6.42297268e-01
7.68933296e-02 -1.41137218e+00 1.10763288e+00 7.53762007e-01
4.07183021e-01 -7.60393590e-02 3.69963974e-01 4.42778558e-01
-9.12777483e-01 -2.51157224e-01 -1.48834348e-01 -7.27824628e-01
1.30472869e-01 -1.38598442e-01 -4.31515187e-01 3.55365217e-01
8.40974092e-01 1.12050605e+00 -7.59593546e-01 6.84007943e-01
-2.75961459e-01 3.88112605e-01 -2.03205928e-01 7.74360597e-02
5.24089694e-01 1.58587828e-01 5.17368376e-01 1.06573081e+00
1.11150695e-02 3.81210417e-01 4.18425560e-01 6.45065784e-01
1.47409499e-01 -1.15982778e-01 -7.92027116e-01 -2.77436495e-01
-4.48374212e-01 5.88105619e-01 -4.99966711e-01 -4.79226470e-01
-7.10118413e-01 1.14727926e+00 1.93451330e-01 2.41078258e-01
-8.23715150e-01 4.16646749e-01 9.96970773e-01 2.49130189e-01
5.82960963e-01 -4.10405517e-01 1.59675255e-02 -1.30640113e+00
2.43222982e-01 -1.19788086e+00 4.84725714e-01 -9.41971123e-01
-1.09232235e+00 6.41365647e-02 1.38275355e-01 -1.29210889e+00
-2.98097849e-01 -1.21231520e+00 -3.42921257e-01 3.01621616e-01
-8.64846110e-01 -1.26728737e+00 -3.12702149e-01 7.01213539e-01
8.63664329e-01 3.07821129e-02 6.95752800e-01 1.05836667e-01
-2.60932058e-01 2.21433461e-01 -4.55869973e-01 4.98881221e-01
6.79885149e-01 -1.05543780e+00 5.06593525e-01 5.17876148e-01
4.49863225e-01 4.84537452e-01 8.75518680e-01 -7.21246123e-01
-1.61444759e+00 -6.87609673e-01 6.16460800e-01 -1.07933974e+00
6.25034750e-01 -3.87818635e-01 -5.83442032e-01 1.13225794e+00
1.12263588e-02 2.85145849e-01 3.86540085e-01 1.10914722e-01
-4.87414151e-01 1.52960911e-01 -7.33789861e-01 7.90213287e-01
1.73700249e+00 -6.37211740e-01 -1.02175677e+00 4.82325017e-01
2.69578546e-01 -6.17021322e-01 -6.40976369e-01 6.28686249e-01
1.00861037e+00 -1.27615964e+00 1.23124909e+00 -1.37486386e+00
8.99575174e-01 -1.90728411e-01 -4.91775036e-01 -9.80270684e-01
-1.51993781e-01 -1.61316454e-01 -3.65906686e-01 5.59775054e-01
-7.57559910e-02 2.09391266e-02 9.64201927e-01 7.17446923e-01
-6.56968132e-02 -6.41094804e-01 -1.05966353e+00 -9.66469526e-01
4.00970504e-02 -6.48296416e-01 2.86289603e-01 6.93677843e-01
4.53537069e-02 5.84955327e-03 -6.51039064e-01 -2.90129155e-01
3.29433650e-01 1.22099563e-01 1.03498399e+00 -8.22596133e-01
-5.19487798e-01 -5.22228539e-01 -1.04497874e+00 -1.26250839e+00
3.49568158e-01 -5.47757804e-01 1.24966159e-01 -1.16616046e+00
2.61085898e-01 3.08896393e-01 -2.61681497e-01 6.72160029e-01
-3.61435139e-03 4.47928876e-01 3.86419594e-01 6.89393803e-02
-9.93844032e-01 5.19525290e-01 1.39223003e+00 -3.55806768e-01
2.90606380e-01 -7.63466880e-02 -1.26891762e-01 9.15452123e-01
5.21999538e-01 -1.66205332e-01 -2.99556762e-01 -4.54135180e-01
7.26376325e-02 8.43571723e-02 1.13407409e+00 -1.37901413e+00
-7.05296127e-03 -1.83094040e-01 7.79835820e-01 -4.16894197e-01
6.55665994e-01 -6.71346426e-01 3.54215987e-02 5.77777803e-01
-5.30438125e-01 -1.97438955e-01 2.05315441e-01 8.33163500e-01
-1.75334975e-01 1.81611598e-01 3.54237467e-01 -5.97221255e-01
-1.33040333e+00 2.64794022e-01 -4.52028126e-01 1.81146398e-01
9.41809356e-01 -4.94305313e-01 -1.20594203e-01 -5.35238981e-01
-1.12706149e+00 -1.59175210e-02 5.06397963e-01 7.44324028e-01
3.66031468e-01 -1.34255123e+00 -5.81662118e-01 1.18334517e-02
3.47946137e-01 -4.14678931e-01 5.06700754e-01 1.07258344e+00
-3.50687802e-01 5.86602449e-01 -5.89484096e-01 -7.46650040e-01
-1.25453699e+00 6.11230433e-01 5.72740495e-01 6.49648458e-02
-1.02945876e+00 7.82265306e-01 5.90718985e-01 -2.59041488e-01
3.32798183e-01 -5.75127244e-01 -3.85744497e-02 -4.13784534e-02
5.88939190e-01 8.55866745e-02 -3.31407227e-02 -1.02281690e+00
-6.15555882e-01 4.59292531e-01 1.81084380e-01 1.39436275e-01
1.21278119e+00 2.07056880e-01 3.81465673e-01 5.40394902e-01
9.02792752e-01 -5.47273695e-01 -1.80835521e+00 -1.44034907e-01
-2.45525707e-02 -6.46293044e-01 -3.41205567e-01 -9.36390758e-01
-1.07665777e+00 9.67916012e-01 3.82454634e-01 -2.05322146e-01
8.42663884e-01 4.56882343e-02 6.34535968e-01 7.12249160e-01
6.22396529e-01 -1.28749049e+00 4.44604069e-01 6.55525446e-01
1.14052904e+00 -1.42919016e+00 1.47395909e-01 -1.07178418e-02
-8.85332823e-01 1.14259684e+00 8.23257565e-01 -2.74422497e-01
4.29284632e-01 -1.55512273e-01 2.10782494e-02 -4.18140471e-01
-7.66032755e-01 -3.89433056e-01 4.77494478e-01 5.78261912e-01
4.22319740e-01 2.14560394e-04 -1.28449962e-01 6.73691094e-01
-1.92863028e-02 1.32382855e-01 2.39355445e-01 1.14814699e+00
-2.17499405e-01 -6.93270862e-01 -2.35940084e-01 3.11029255e-01
-4.98429835e-01 2.17388302e-01 -7.97503352e-01 1.03447104e+00
2.21800655e-01 5.75603068e-01 1.57238226e-02 -4.01544780e-01
5.68590581e-01 5.04526258e-01 9.76458490e-01 -6.36335075e-01
-5.94414830e-01 -4.23561692e-01 3.15122485e-01 -1.33918834e+00
-7.94484735e-01 -9.91881609e-01 -8.79749358e-01 -7.52483234e-02
3.76400411e-01 -5.05587041e-01 1.65657967e-01 1.31556845e+00
1.25355229e-01 3.65564823e-01 -2.05188707e-01 -1.39003944e+00
-7.18820572e-01 -9.74183321e-01 -6.03154302e-01 1.04045701e+00
2.97879308e-01 -1.12449682e+00 -2.53042400e-01 2.13947758e-01] | [8.163435935974121, 0.4954693019390106] |
d113bae8-59a9-4834-880a-c7fd0e1f42c6 | compressing-deep-neural-networks-via-layer | 2007.14917 | null | https://arxiv.org/abs/2007.14917v1 | https://arxiv.org/pdf/2007.14917v1.pdf | Compressing Deep Neural Networks via Layer Fusion | This paper proposes \textit{layer fusion} - a model compression technique that discovers which weights to combine and then fuses weights of similar fully-connected, convolutional and attention layers. Layer fusion can significantly reduce the number of layers of the original network with little additional computation overhead, while maintaining competitive performance. From experiments on CIFAR-10, we find that various deep convolution neural networks can remain within 2\% accuracy points of the original networks up to a compression ratio of 3.33 when iteratively retrained with layer fusion. For experiments on the WikiText-2 language modelling dataset where pretrained transformer models are used, we achieve compression that leads to a network that is 20\% of its original size while being within 5 perplexity points of the original network. We also find that other well-established compression techniques can achieve competitive performance when compared to their original networks given a sufficient number of retraining steps. Generally, we observe a clear inflection point in performance as the amount of compression increases, suggesting a bound on the amount of compression that can be achieved before an exponential degradation in performance. | ["James O' Neill", 'Aram Galstyan', 'Greg Ver Steeg'] | 2020-07-29 | null | null | null | null | ['exponential-degradation'] | ['time-series'] | [ 3.20275664e-01 4.27985489e-01 -1.04076982e-01 -4.75270331e-01
-5.43353260e-01 -2.36326724e-01 4.50798273e-01 1.27175033e-01
-9.92651463e-01 4.98409152e-01 6.97327703e-02 -6.26107931e-01
-1.52512670e-01 -4.92804706e-01 -9.43269432e-01 -4.02168065e-01
-1.10094972e-01 5.33449531e-01 2.38865048e-01 3.02584339e-02
-1.53312728e-01 4.47321713e-01 -1.43130219e+00 5.94960928e-01
5.50117791e-01 1.23934245e+00 3.40574205e-01 8.35051060e-01
-1.23672076e-01 1.01077151e+00 -4.51630533e-01 -7.44426131e-01
4.54939812e-01 -9.35009681e-03 -9.21520531e-01 -1.70499608e-01
6.67150497e-01 -3.63279790e-01 -8.85386705e-01 1.13642132e+00
2.14883745e-01 -2.49601435e-02 3.97021413e-01 -1.00505817e+00
-3.62817883e-01 1.48424780e+00 -5.06189585e-01 5.33354282e-01
-4.97101963e-01 -8.98838714e-02 1.06567729e+00 -6.11266792e-01
2.09452510e-01 1.26646376e+00 8.33177745e-01 4.61883038e-01
-1.42167318e+00 -1.00035632e+00 1.84347734e-01 -5.40782921e-02
-1.41277170e+00 -7.16040850e-01 2.90072829e-01 -8.13652351e-02
1.49149942e+00 2.07634613e-01 4.78848845e-01 7.55416811e-01
8.91624317e-02 5.25677621e-01 4.98384029e-01 -4.28853273e-01
1.43091585e-02 -2.21058098e-03 2.60007888e-01 8.97199273e-01
5.13405919e-01 -4.27496403e-01 -4.10158277e-01 -3.82615663e-02
6.17951512e-01 8.73896182e-02 -1.49426520e-01 1.71051025e-01
-1.00635040e+00 7.10536301e-01 7.59647369e-01 4.31815803e-01
-4.45020527e-01 6.99312210e-01 3.77999008e-01 5.83314478e-01
4.85747874e-01 5.16558707e-01 -7.08572805e-01 -1.26477256e-01
-1.34025002e+00 2.19432041e-01 7.08939612e-01 9.02304411e-01
5.58420539e-01 2.44849831e-01 1.01212889e-01 9.16471243e-01
-1.10213891e-01 1.77644506e-01 7.41301000e-01 -1.20704460e+00
7.99815536e-01 6.88280046e-01 -5.31219661e-01 -5.57489574e-01
-1.77527308e-01 -8.93605232e-01 -1.19279230e+00 3.93649451e-02
2.94541150e-01 -5.02040565e-01 -9.56385374e-01 1.93768120e+00
-4.28493530e-01 1.24303922e-01 1.71539754e-01 2.22872347e-01
4.00960028e-01 5.85633337e-01 8.28229040e-02 -8.17386806e-02
1.31069171e+00 -9.44958627e-01 -3.14289004e-01 -7.04464793e-01
7.79846787e-01 -6.08020842e-01 9.11070585e-01 4.15386736e-01
-1.65140140e+00 -4.02900875e-01 -1.39248979e+00 -3.47486921e-02
-1.22176811e-01 -1.30326152e-01 5.15168548e-01 3.17410588e-01
-1.34085298e+00 1.05063891e+00 -1.11485612e+00 4.58680205e-02
6.15468740e-01 6.84815288e-01 -5.01550853e-01 -2.56873697e-01
-7.76327610e-01 9.95569706e-01 9.99388695e-01 -2.62267470e-01
-6.99867487e-01 -1.08557308e+00 -6.24392092e-01 6.97907269e-01
-3.59769613e-02 -5.20917773e-01 1.52335978e+00 -9.26842928e-01
-8.19654703e-01 4.45204139e-01 -2.00160250e-01 -1.31617725e+00
2.35715508e-01 -1.21825494e-01 -3.07386607e-01 2.21829206e-01
-4.60640669e-01 1.06594741e+00 5.08392155e-01 -8.09035242e-01
-9.60774481e-01 -3.04056942e-01 1.31381765e-01 1.48439139e-01
-9.00133491e-01 -3.31838690e-02 -6.73022330e-01 -5.54609358e-01
1.61132470e-01 -9.06998158e-01 -3.21566850e-01 -1.77101381e-02
-1.92978442e-01 -1.47583947e-01 8.04854214e-01 -7.01733410e-01
1.34633768e+00 -2.06082129e+00 8.04321617e-02 9.55126882e-02
7.26724446e-01 5.59556842e-01 -2.71548808e-01 1.75802410e-02
-3.01267982e-01 3.77196431e-01 -2.94258654e-01 -7.95713961e-01
-5.64990304e-02 3.41610819e-01 -8.34338740e-02 1.81346491e-01
9.80836973e-02 8.71310115e-01 -3.65987867e-01 -1.99814543e-01
-1.85329452e-01 6.96580350e-01 -8.71786237e-01 -8.43615830e-02
6.19475655e-02 -5.81962347e-01 2.16356337e-01 6.89042732e-02
5.74179173e-01 -4.76578593e-01 8.08502361e-02 -2.39514843e-01
2.23890901e-01 4.03328896e-01 -7.58675337e-01 1.52769697e+00
-4.88255650e-01 1.05788076e+00 2.10107610e-01 -1.05865502e+00
7.89220452e-01 3.08429658e-01 3.23212832e-01 -6.36838019e-01
1.83168337e-01 2.07843050e-01 4.66225743e-01 1.73820511e-01
4.29253668e-01 -5.30433469e-02 2.18088105e-01 5.69093227e-01
4.68405157e-01 3.20405781e-01 3.80176693e-01 3.39154541e-01
1.43324864e+00 -7.98202515e-01 -2.49360889e-01 -3.66622716e-01
5.23818806e-02 -3.30399036e-01 2.36272007e-01 6.69158280e-01
1.06129266e-01 2.86031425e-01 6.92420661e-01 -4.73701179e-01
-1.54358709e+00 -7.00778186e-01 -5.90090230e-02 1.15787601e+00
-5.42136490e-01 -7.36428201e-01 -1.10810542e+00 -4.05945987e-01
-9.78327990e-02 6.84833050e-01 -6.33701622e-01 -4.61033195e-01
-5.78637421e-01 -8.38507235e-01 1.13744926e+00 7.94770420e-01
8.58561993e-01 -7.53098547e-01 -5.67903638e-01 1.81366101e-01
-5.06308824e-02 -1.28203499e+00 -1.51456907e-01 6.57541752e-01
-1.19080806e+00 -8.66509795e-01 -7.40171909e-01 -9.29845810e-01
7.30148792e-01 1.69677630e-01 1.19992197e+00 4.37951267e-01
1.47712424e-01 -4.28448886e-01 9.13335308e-02 -4.00898010e-01
-4.43483144e-01 5.35904169e-01 -2.65466198e-02 -4.87793118e-01
3.15311998e-01 -7.76661396e-01 -4.93817955e-01 -2.24976584e-01
-9.57716823e-01 3.31311375e-01 7.35788047e-01 5.97046375e-01
3.61936182e-01 4.28899556e-01 2.22205818e-01 -8.10020864e-01
7.59399593e-01 -2.13929847e-01 -3.12795758e-01 2.71903485e-01
-9.25331235e-01 4.85060185e-01 5.47577083e-01 -5.89422405e-01
-7.38029659e-01 -1.03428856e-01 -4.13327403e-02 -8.39401603e-01
1.79872498e-01 6.36168599e-01 3.38893354e-01 4.35512625e-02
6.38255298e-01 1.18412934e-01 1.94711372e-01 -6.57305896e-01
3.57797265e-01 5.85209668e-01 7.85213292e-01 -2.16402978e-01
4.72835481e-01 1.40504852e-01 -2.94555932e-01 -5.06376863e-01
-7.65134394e-01 -1.65470272e-01 -5.34171939e-01 3.35330933e-01
4.42748845e-01 -1.03801322e+00 -6.00833476e-01 2.31397092e-01
-1.02909327e+00 -4.93797749e-01 -3.14951181e-01 4.99183655e-01
-1.92460805e-01 6.53717145e-02 -1.00588310e+00 -4.54870254e-01
-7.56528974e-01 -1.08385682e+00 5.58071554e-01 -8.29205383e-03
-3.32877785e-02 -1.02542293e+00 -3.90877336e-01 6.27994165e-02
6.57865405e-01 -1.30833551e-01 1.32646644e+00 -7.51444101e-01
-4.38306853e-02 -4.16199207e-01 -4.61424559e-01 8.16119373e-01
-2.33152837e-01 -3.52222919e-01 -8.51154625e-01 -5.57867050e-01
-1.86229214e-01 -1.94550171e-01 1.35776281e+00 2.86819726e-01
1.36940670e+00 -7.97746778e-01 -1.71423957e-01 7.15764582e-01
1.43369794e+00 1.45361662e-01 5.21593034e-01 -4.25337628e-03
5.56034625e-01 2.07112208e-02 -4.53514934e-01 2.59013206e-01
5.09585105e-02 1.99818313e-01 3.84949476e-01 -2.62566298e-01
-3.26291084e-01 -3.16446215e-01 1.18779711e-01 1.10137725e+00
-1.25829414e-01 -2.20541388e-01 -9.76971209e-01 5.18377542e-01
-1.57129765e+00 -9.74209785e-01 4.12437528e-01 2.17673492e+00
1.05254757e+00 7.30900109e-01 -1.65536255e-01 5.35400689e-01
4.76072013e-01 2.58918535e-02 -5.31646609e-01 -6.56131446e-01
-2.88049523e-02 5.03310025e-01 1.04303789e+00 6.15860164e-01
-8.95305455e-01 8.89084160e-01 7.19125128e+00 7.80299246e-01
-1.17370880e+00 1.49375945e-01 1.04276574e+00 -5.62713444e-01
-1.36091337e-01 -3.10446411e-01 -8.13077092e-01 2.38294199e-01
1.57841384e+00 -4.12629575e-01 6.69943452e-01 7.98330307e-01
-2.13139445e-01 2.59124786e-01 -1.11389816e+00 1.03917825e+00
-7.49584809e-02 -1.54761457e+00 2.08641306e-01 2.78490841e-01
7.53544092e-01 7.09029078e-01 6.42420501e-02 5.09673953e-01
5.33518314e-01 -1.19184732e+00 5.86160541e-01 3.53679001e-01
9.46057558e-01 -1.24106634e+00 7.81541348e-01 6.51113153e-01
-9.91234720e-01 -3.11390340e-01 -8.56910825e-01 -1.22154215e-02
-1.44168019e-01 4.83483642e-01 -9.70304370e-01 1.10370042e-02
7.89203763e-01 3.73811305e-01 -7.74450839e-01 1.01073205e+00
3.06357622e-01 7.51702130e-01 -6.43923759e-01 1.91325694e-01
5.07429540e-01 4.02641535e-01 -3.38202417e-02 1.36151505e+00
3.11892152e-01 -8.56588259e-02 -3.52675825e-01 5.55885971e-01
-8.39579940e-01 -2.44483754e-01 -4.31219816e-01 -1.63614035e-01
6.30906284e-01 9.38809454e-01 -5.75556934e-01 -7.35056579e-01
-2.24482611e-01 9.41094160e-01 6.89886749e-01 2.78785646e-01
-7.08716094e-01 -5.87172449e-01 6.71093047e-01 2.22421050e-01
5.97494185e-01 -2.60257870e-01 -2.08083019e-01 -8.49303126e-01
-4.23451327e-02 -7.40835488e-01 1.46831512e-01 -5.25147319e-01
-6.47383690e-01 9.93726790e-01 1.75089076e-01 -5.42500496e-01
-2.62842447e-01 -5.41850150e-01 -4.93413806e-01 7.35600471e-01
-1.23112679e+00 -8.09280932e-01 -9.18165073e-02 5.39964795e-01
6.17368639e-01 -4.92580295e-01 9.26009238e-01 6.42804384e-01
-5.00530601e-01 1.14165699e+00 2.71504819e-01 3.89685839e-01
2.95858681e-02 -9.99087393e-01 6.15281284e-01 8.40550959e-01
2.99287438e-01 6.45355582e-01 5.07053375e-01 -2.32002094e-01
-1.01900637e+00 -1.24021101e+00 1.24855328e+00 1.76091790e-02
4.86385256e-01 -3.05257559e-01 -1.03994095e+00 1.09065795e+00
4.18972909e-01 8.47334787e-03 3.96877944e-01 6.39408976e-02
-6.00313902e-01 -2.31572166e-01 -1.01442695e+00 6.54912949e-01
1.05157101e+00 -4.18234646e-01 -4.41707879e-01 2.49268830e-01
1.14053476e+00 -3.02699924e-01 -1.03038704e+00 3.91307026e-01
5.13798714e-01 -7.92483270e-01 8.93738866e-01 -8.46572936e-01
4.82515007e-01 3.92295867e-01 -2.66961724e-01 -1.13137507e+00
-5.39377153e-01 -4.93155390e-01 -4.56184536e-01 9.26472366e-01
8.07127059e-01 -4.15207803e-01 8.99883628e-01 8.31709266e-01
2.08789594e-02 -9.57191825e-01 -9.45683122e-01 -6.83370471e-01
3.49439144e-01 -6.24835491e-01 6.79443955e-01 6.63894594e-01
-4.95519415e-02 3.36274743e-01 3.32602262e-02 -2.60124028e-01
5.60339868e-01 -6.52990699e-01 2.33058408e-01 -1.32136714e+00
-3.16743374e-01 -9.29326415e-01 -3.37087482e-01 -1.05708098e+00
2.50317663e-01 -1.24653733e+00 -3.85914505e-01 -1.39389026e+00
3.52720231e-01 -4.09355253e-01 -6.61414266e-01 9.12670076e-01
2.85525858e-01 4.89324659e-01 4.24039483e-01 2.69225985e-01
-4.49235260e-01 1.60516441e-01 8.71664762e-01 -3.41539294e-01
-1.95941934e-03 -1.47282749e-01 -1.15462315e+00 8.35222483e-01
9.64837074e-01 -4.41232830e-01 -6.10431790e-01 -1.06137407e+00
1.09943628e-01 -1.14340022e-01 3.52765471e-02 -1.42466509e+00
4.80005503e-01 4.61613983e-01 4.84639287e-01 -5.17172575e-01
3.94416749e-01 -8.47858548e-01 1.31119683e-01 7.03157127e-01
-8.76527905e-01 5.31845450e-01 5.19294441e-01 2.20562831e-01
-1.04500748e-01 -3.40026706e-01 9.29773569e-01 -2.19214171e-01
-2.48215973e-01 3.18216205e-01 -3.40655118e-01 -1.09062858e-01
6.48404598e-01 -5.78360446e-02 -2.36274272e-01 -2.36838609e-01
-7.01198578e-01 7.37502053e-02 7.32623115e-02 3.99512410e-01
5.63086927e-01 -1.31074297e+00 -7.90834844e-01 2.82118350e-01
-3.46956223e-01 7.03397617e-02 -3.86819839e-02 5.71838439e-01
-7.07795322e-01 6.54507875e-01 -6.00211509e-02 -3.44657928e-01
-1.41278410e+00 2.40241960e-01 5.97861528e-01 -3.99544209e-01
-7.37938225e-01 1.10524964e+00 -4.98742498e-02 1.80282500e-02
7.32651532e-01 -7.32574761e-01 1.10494971e-01 -1.65471941e-01
7.54266083e-01 2.03659296e-01 4.82324004e-01 -2.94370085e-01
-5.50124943e-02 -4.42379005e-02 -5.52966833e-01 -2.06733748e-01
1.40807974e+00 1.94775701e-01 9.74195376e-02 6.09313101e-02
1.60422218e+00 -5.87876856e-01 -1.38095307e+00 -5.63853264e-01
-4.27283719e-02 7.94921890e-02 6.63829744e-01 -6.18661582e-01
-1.64761662e+00 9.99416709e-01 6.11745179e-01 1.26001999e-01
1.11723208e+00 -7.90297706e-03 8.89930785e-01 8.93012762e-01
-7.25997314e-02 -9.69241798e-01 -1.62125602e-01 6.48856461e-01
8.08744192e-01 -7.04435825e-01 3.19770910e-03 2.14226305e-01
-3.63423467e-01 8.95852923e-01 4.88720298e-01 -3.11369210e-01
8.69246602e-01 6.97756946e-01 -3.89401257e-01 6.50940323e-03
-1.27953398e+00 2.12978333e-01 1.29738435e-01 3.13820392e-01
4.16349500e-01 5.24189731e-04 1.70784593e-01 6.62221551e-01
-7.39731610e-01 -1.12103797e-01 1.14210136e-01 7.19776690e-01
-8.24004710e-01 -7.64777839e-01 1.59600586e-01 8.66194904e-01
-8.45234454e-01 -5.31797409e-01 -1.19362362e-01 7.66017139e-01
3.14672180e-02 7.23675311e-01 7.25527227e-01 -6.70652807e-01
2.80108303e-01 2.61109322e-01 4.96656269e-01 -3.25238317e-01
-6.91543639e-01 8.85355994e-02 -1.25655150e-02 -3.92688215e-01
-1.01758815e-01 -2.40364596e-01 -1.29672956e+00 -1.01873481e+00
-3.95567119e-01 1.30814806e-01 6.45700157e-01 8.10268164e-01
3.42444062e-01 5.83887041e-01 1.25034630e-01 -6.68458819e-01
-5.67231119e-01 -1.30484593e+00 -3.85340840e-01 1.62710115e-01
3.15110534e-01 -8.07338282e-02 -3.48853618e-01 5.99224716e-02] | [8.597057342529297, 3.2147228717803955] |
d30b7c6e-0647-4a2a-89a2-0ab6a8873afd | on-games-and-simulators-as-a-platform-for | 2110.11305 | null | https://arxiv.org/abs/2110.11305v1 | https://arxiv.org/pdf/2110.11305v1.pdf | On games and simulators as a platform for development of artificial intelligence for command and control | Games and simulators can be a valuable platform to execute complex multi-agent, multiplayer, imperfect information scenarios with significant parallels to military applications: multiple participants manage resources and make decisions that command assets to secure specific areas of a map or neutralize opposing forces. These characteristics have attracted the artificial intelligence (AI) community by supporting development of algorithms with complex benchmarks and the capability to rapidly iterate over new ideas. The success of artificial intelligence algorithms in real-time strategy games such as StarCraft II have also attracted the attention of the military research community aiming to explore similar techniques in military counterpart scenarios. Aiming to bridge the connection between games and military applications, this work discusses past and current efforts on how games and simulators, together with the artificial intelligence algorithms, have been adapted to simulate certain aspects of military missions and how they might impact the future battlefield. This paper also investigates how advances in virtual reality and visual augmentation systems open new possibilities in human interfaces with gaming platforms and their military parallels. | ['Alexander Kott', 'Priya Narayanan', 'Theron Trout', 'Mark Dennison', 'Anne Logie', 'Manuel Vindiola', 'John Richardson', 'Mark Mittrick', 'Song Jun Park', 'Derrik E. Asher', 'Nicholas Waytowich', 'Vinicius G. Goecks'] | 2021-10-21 | null | null | null | null | ['real-time-strategy-games'] | ['playing-games'] | [-2.66152948e-01 6.91155493e-02 2.97740191e-01 1.31650254e-01
4.26652789e-01 -8.37274194e-01 7.50069141e-01 -7.38030300e-02
-7.88581014e-01 9.50081050e-01 -5.38292043e-02 -5.60396254e-01
-4.00305480e-01 -1.04861987e+00 6.91108778e-02 -1.73105702e-01
-8.29343796e-01 1.07583547e+00 3.16089511e-01 -1.50404215e+00
4.30852741e-01 7.68430710e-01 -1.83344817e+00 -3.43168452e-02
2.62481600e-01 4.42875504e-01 1.31201074e-01 8.41836929e-01
1.36486918e-01 6.18447363e-01 -8.80492866e-01 -4.35008854e-01
5.84892631e-01 -4.34356406e-02 -4.10696447e-01 -4.19012368e-01
-6.17957473e-01 -8.43352154e-02 -2.16998056e-01 7.08610237e-01
7.90167987e-01 3.95055652e-01 2.48247057e-01 -1.36905992e+00
3.53141367e-01 2.48695672e-01 -4.07538146e-01 5.46403110e-01
8.23455036e-01 3.91475588e-01 3.58021230e-01 -3.29456210e-01
6.54885173e-01 9.97536480e-01 4.94706273e-01 4.93572474e-01
-6.44677043e-01 -5.55928409e-01 3.97604033e-02 3.06935340e-01
-1.21554291e+00 1.28994480e-01 4.64505494e-01 -2.83614278e-01
1.18398595e+00 8.72802675e-01 1.25638568e+00 7.44778872e-01
3.82898718e-01 3.83703619e-01 1.12766063e+00 -1.33108526e-01
4.32811439e-01 2.25025639e-01 -2.62575507e-01 3.54394317e-01
3.84883583e-01 7.56813347e-01 -2.52971888e-01 -1.71663716e-01
8.86249542e-01 -3.58820826e-01 -1.27949715e-01 -2.46372834e-01
-1.13684523e+00 8.90969098e-01 4.08317029e-01 5.80971479e-01
-8.67379308e-01 -6.30987138e-02 4.87966835e-01 4.65352297e-01
-3.39830555e-02 1.19029224e+00 -5.12811281e-02 -6.43864214e-01
-4.87440944e-01 9.80874121e-01 1.22880650e+00 2.22587585e-01
4.62903202e-01 4.06654716e-01 4.23836291e-01 1.05500132e-01
6.01071641e-02 2.18601778e-01 2.14245647e-01 -7.48923779e-01
2.81911969e-01 6.72553957e-01 3.60514611e-01 -1.26299274e+00
-9.44935203e-01 -6.21666789e-01 -5.17607749e-01 1.21164250e+00
1.99251294e-01 -5.66840410e-01 -2.45832533e-01 1.13316488e+00
4.99360800e-01 9.81482342e-02 1.21165894e-01 1.24580884e+00
6.05299711e-01 4.39621568e-01 -1.17389269e-01 2.32330099e-01
1.23827493e+00 -2.47279286e-01 -4.86884803e-01 -4.58171815e-01
4.13269490e-01 -5.49917400e-01 5.63336372e-01 6.46933734e-01
-1.24922001e+00 -2.19823405e-01 -1.19040573e+00 1.00718319e+00
-6.03032351e-01 -6.86951458e-01 1.02148044e+00 1.26475871e+00
-1.10666370e+00 4.04189229e-01 -6.45980775e-01 -1.10180795e-01
2.52849460e-02 7.36327350e-01 -2.25266606e-01 2.75559455e-01
-1.56913853e+00 1.43376267e+00 5.68929911e-01 -1.16492817e-02
-9.09797668e-01 -4.63302225e-01 -6.52421892e-01 -5.32813072e-02
5.06470025e-01 -8.54478359e-01 5.61023116e-01 -8.30245972e-01
-1.57486570e+00 9.76483583e-01 9.42101300e-01 -6.33252263e-01
6.51710033e-01 1.89011425e-01 -3.12675387e-01 -4.26361769e-01
-1.21384650e-01 2.83897579e-01 7.99526721e-02 -9.22727227e-01
-7.19064355e-01 -4.57822412e-01 8.75655830e-01 8.55798483e-01
1.55846952e-02 3.73095602e-01 6.31555989e-02 -4.70511466e-01
-3.96415532e-01 -9.92196500e-01 -8.96434188e-01 -5.63692272e-01
7.13527873e-02 3.94509882e-01 5.67643940e-01 -2.14773178e-01
1.01471591e+00 -1.66310239e+00 8.41758132e-01 4.15002972e-01
1.31301969e-01 4.40663159e-01 1.12822086e-01 8.80189657e-01
5.02531528e-02 -8.39752406e-02 3.30421299e-01 2.84650892e-01
1.41162008e-01 2.70074815e-01 6.26312792e-02 2.05615699e-01
-4.39359844e-01 6.44065082e-01 -7.09741831e-01 4.19776328e-02
4.60101247e-01 1.57588482e-01 -5.65602005e-01 1.50497044e-02
1.38102937e-02 5.53708196e-01 -5.20328999e-01 4.83878911e-01
5.75668097e-01 6.01063251e-01 3.24109048e-01 4.50640500e-01
-6.28143370e-01 -1.87821954e-01 -1.49793851e+00 1.48375046e+00
-3.41339529e-01 5.25685906e-01 6.55572832e-01 -9.24825490e-01
7.77515769e-01 1.42605767e-01 4.78667766e-01 -9.30062175e-01
6.16987050e-01 -4.09295782e-02 3.84090424e-01 -3.65174174e-01
9.11584795e-01 -5.82929909e-01 -2.58128256e-01 5.95058680e-01
-4.12487119e-01 -5.53831398e-01 1.86715692e-01 7.38547742e-02
1.12151694e+00 -5.09383008e-02 2.88522601e-01 -3.73912930e-01
5.02893507e-01 6.36125565e-01 1.63638815e-01 7.03539252e-01
-2.37651169e-01 8.30667019e-02 2.21342653e-01 -8.73136342e-01
-6.47856593e-01 -9.71883833e-01 4.93863016e-01 1.02657425e+00
6.29360020e-01 -3.56932640e-01 -4.77555513e-01 -2.06052884e-02
-2.40932852e-01 7.40052819e-01 -5.74711204e-01 -7.01528937e-02
-6.37153625e-01 -9.12383556e-01 4.91046369e-01 1.19737431e-01
2.97027677e-01 -1.04100895e+00 -1.58543324e+00 4.91849989e-01
2.00075824e-02 -7.68723547e-01 6.30315602e-01 -1.67827278e-01
-6.61248863e-01 -1.19313323e+00 -4.55006629e-01 -1.70385823e-01
2.12306976e-02 2.60261029e-01 1.19757783e+00 4.28971589e-01
-4.63337988e-01 7.09380388e-01 -5.44468403e-01 -1.03812027e+00
-3.75231326e-01 -6.44649118e-02 3.00169021e-01 -4.65637565e-01
-4.47477922e-02 -6.82186306e-01 -5.58876514e-01 6.61272347e-01
-9.26349938e-01 1.44207999e-01 2.75747091e-01 5.47783911e-01
-2.31763959e-01 1.32793307e-01 1.19935989e-01 -4.52769458e-01
1.25434065e+00 -5.48858941e-01 -7.14351654e-01 -1.22131787e-01
-3.86800244e-02 -3.20582539e-01 3.01878393e-01 -2.83779502e-01
-6.56510592e-01 -4.40017819e-01 -2.11525559e-01 1.36008095e-02
4.37203832e-02 8.66986871e-01 -1.11775115e-01 -5.72769165e-01
8.56494009e-01 -5.43623529e-02 3.76217246e-01 1.40256524e-01
1.21156074e-01 2.31908619e-01 2.49952540e-01 -5.61078548e-01
8.21423292e-01 5.42346656e-01 2.10441977e-01 -7.17844725e-01
4.45822984e-01 -1.70162261e-01 -2.36636966e-01 -9.38870788e-01
5.69376588e-01 -6.72891259e-01 -1.14531672e+00 3.28404248e-01
-8.78828228e-01 -2.14427829e-01 -3.29518348e-01 6.82764828e-01
-5.13170540e-01 8.45482349e-02 -1.30992606e-01 -1.09390402e+00
-7.08665028e-02 -1.22657883e+00 4.06031728e-01 5.41761875e-01
-3.87389779e-01 -1.07136989e+00 6.07737839e-01 3.90373826e-01
9.50458348e-01 5.10901093e-01 2.22733304e-01 -6.75165355e-01
-4.45812464e-01 -5.85197628e-01 4.38401252e-01 -3.61011147e-01
-3.37133944e-01 -4.16870803e-01 -5.39991379e-01 -3.86977226e-01
3.15807946e-02 -2.12486386e-01 -5.75768650e-02 1.66933149e-01
3.18200216e-02 2.11832240e-01 -4.52677995e-01 2.64139354e-01
1.25057840e+00 6.07840478e-01 1.14551115e+00 1.01144445e+00
-3.37256193e-02 8.57563376e-01 1.07154560e+00 9.18617785e-01
1.77452862e-01 1.13125014e+00 1.02269030e+00 -1.76101759e-01
4.74871457e-01 4.87453908e-01 9.50397551e-02 2.66077399e-01
-1.08900654e+00 -5.46702854e-02 -1.15804839e+00 3.79658267e-02
-1.66331112e+00 -1.23881769e+00 8.34080502e-02 2.19373250e+00
1.08707912e-01 4.94534045e-01 3.51001650e-01 8.96252543e-02
5.06660700e-01 1.63610607e-01 -1.11829340e-01 -6.53519750e-01
-6.83764666e-02 3.57564390e-01 5.50791502e-01 5.33963263e-01
-7.23413706e-01 9.04829919e-01 6.55876112e+00 7.33801425e-01
-1.06871080e+00 1.18141346e-01 1.86658099e-01 -2.81251073e-01
-1.96886048e-01 2.03992613e-02 -2.61187941e-01 -9.92399734e-03
9.69993412e-01 -5.77576399e-01 7.46861935e-01 7.03529954e-01
3.64406019e-01 -4.29465860e-01 -3.58805090e-01 7.93318808e-01
-3.41660939e-02 -1.67067182e+00 -2.04792857e-01 4.36427116e-01
2.55766153e-01 1.32475682e-02 1.44650266e-01 5.23086846e-01
5.60957134e-01 -1.12347364e+00 6.35563731e-01 4.36814457e-01
2.94288397e-01 -1.00568378e+00 9.54506278e-01 5.39364219e-01
-1.09081149e+00 -1.54858813e-01 -1.29211739e-01 -1.14050496e+00
4.76616204e-01 -4.33832020e-01 -1.07155871e+00 8.90502632e-01
5.28912127e-01 -2.43103758e-01 -1.36838630e-01 1.21938097e+00
3.58956248e-01 -1.31670758e-01 -5.02351761e-01 -6.19236648e-01
6.26650989e-01 -5.32720923e-01 1.05856550e+00 6.89946890e-01
5.99086247e-02 4.51704472e-01 1.65622562e-01 4.77158725e-01
9.79292631e-01 3.02627068e-02 -1.04990876e+00 5.51636405e-02
1.67281717e-01 1.21763086e+00 -1.10379612e+00 9.62233469e-02
-1.70489669e-01 6.56552136e-01 -2.40684330e-01 -1.00920442e-02
-1.04972553e+00 -3.36685210e-01 1.18069625e+00 4.34050530e-01
-3.55264068e-01 -5.07008076e-01 -3.43533248e-01 -7.72842884e-01
-5.15629292e-01 -1.29654098e+00 2.54513294e-01 -7.62542427e-01
-4.87972945e-01 1.08124483e+00 2.02529281e-01 -1.31229186e+00
-6.91991270e-01 -7.55386651e-01 -8.32748413e-01 9.04432654e-01
-7.64245629e-01 -1.01495576e+00 -1.81974605e-01 6.36064172e-01
4.01495576e-01 -8.62881899e-01 1.06592989e+00 1.00206718e-01
-1.76113427e-01 -3.10456986e-03 -2.39095941e-01 -2.56391078e-01
-1.16276473e-01 -8.65785539e-01 5.99480808e-01 6.51519835e-01
-1.19823910e-01 2.65607238e-01 1.23871648e+00 -7.27408886e-01
-1.66856921e+00 1.56953279e-03 -9.19447690e-02 -4.40452605e-01
6.95479751e-01 -3.00386876e-01 -1.92224622e-01 3.55807900e-01
3.22926074e-01 -6.97898686e-01 6.71989620e-01 2.67648488e-01
4.91139948e-01 2.93459177e-01 -1.15559995e+00 1.09856439e+00
7.96510994e-01 -5.45977708e-03 -3.94724607e-01 1.00943565e-01
9.96102020e-02 -7.21411526e-01 -2.37457797e-01 2.23091424e-01
6.48716271e-01 -1.51066196e+00 1.31901670e+00 -8.82044315e-01
-1.75555855e-01 -1.62518993e-01 -3.15043852e-02 -1.47384524e+00
-2.29890212e-01 -8.56272817e-01 5.74747562e-01 2.95361519e-01
3.88513476e-01 -8.31414342e-01 1.22608531e+00 8.73563170e-01
-7.51114562e-02 -3.19355130e-01 -1.23027861e+00 -5.61248958e-01
-1.09362774e-01 -5.09908736e-01 5.21494925e-01 7.43708730e-01
3.92975152e-01 2.13478561e-02 -5.65511405e-01 1.76907197e-01
2.47316569e-01 -2.75154293e-01 1.11541736e+00 -1.50576389e+00
-5.96814036e-01 -6.28813982e-01 -1.26270771e+00 -2.04819143e-01
-1.82708800e-01 -4.97068942e-01 -5.44960797e-01 -1.21556735e+00
-4.22939390e-01 -5.78567028e-01 2.57190038e-03 -5.71424663e-02
2.10962489e-01 1.74220994e-01 5.73328197e-01 -1.42280534e-01
-4.19613689e-01 3.16711068e-01 1.11082637e+00 3.34896855e-02
-3.59577477e-01 3.63531709e-01 -6.04079723e-01 5.99417210e-01
7.91481614e-01 -1.46931022e-01 -3.97926748e-01 -6.02151155e-02
7.74498165e-01 6.13104224e-01 2.05527872e-01 -1.54417384e+00
5.31848609e-01 -3.99520636e-01 1.32996067e-01 -1.25816569e-01
7.38620162e-01 -1.00785232e+00 8.61150205e-01 1.07604265e+00
4.54550028e-01 7.78350830e-01 7.18110800e-01 3.66254419e-01
-2.62948155e-01 -2.52023876e-01 2.76126564e-01 -5.20490289e-01
-1.09460032e+00 -1.09389707e-01 -1.14199305e+00 -2.26728976e-01
1.80167019e+00 -7.18912721e-01 -1.58189461e-01 -6.65103376e-01
-9.69885409e-01 4.02541280e-01 5.67808211e-01 4.27468270e-01
6.58276498e-01 -8.86310875e-01 -8.42592001e-01 2.11031422e-01
-3.30668181e-01 -8.08596492e-01 4.25302893e-01 5.87226331e-01
-1.36638045e+00 1.87146530e-01 -1.29757810e+00 1.42597035e-02
-1.47320616e+00 4.93007898e-01 5.96734226e-01 -4.23730314e-01
-3.18056405e-01 8.13149273e-01 1.82685927e-02 -4.15794909e-01
5.83140086e-03 5.02630293e-01 -6.00522399e-01 3.38449068e-02
8.82326365e-01 3.84514391e-01 -3.31065065e-04 -5.54054022e-01
-6.66360915e-01 1.66330427e-01 2.15400562e-01 -8.65424514e-01
1.47148442e+00 1.66621968e-01 1.16674706e-01 2.87032295e-02
1.01340413e-02 5.25237173e-02 -6.60043657e-01 5.34481823e-01
-1.42967254e-01 -6.47023916e-01 1.12610929e-01 -1.02732182e+00
-9.24378216e-01 5.26826620e-01 6.47433460e-01 5.27556241e-01
1.05797410e+00 -5.28135300e-01 1.07420690e-01 2.90181369e-01
1.10799611e+00 -9.68718171e-01 -1.15177326e-01 6.55446768e-01
1.05168903e+00 -8.16273808e-01 4.60455567e-02 -3.88513714e-01
-1.10475719e+00 1.21317303e+00 5.87291181e-01 -3.07835937e-01
4.69563514e-01 8.11342180e-01 2.48810917e-01 -7.86240518e-01
-6.16899252e-01 -5.43510497e-01 -1.75192475e-01 1.17582929e+00
1.21332109e-01 1.23110838e-01 -5.99404812e-01 5.91592848e-01
-3.62894744e-01 -1.20093487e-02 7.80271769e-01 1.22268152e+00
-5.03169954e-01 -1.24992299e+00 -7.22857594e-01 8.12704787e-02
-2.31596515e-01 1.21069014e-01 -6.88557506e-01 1.39000630e+00
-7.75028020e-02 7.83584118e-01 3.62695791e-02 -9.04416442e-01
6.86116457e-01 -6.15450263e-01 5.46415329e-01 -3.98451000e-01
-1.47506833e+00 -4.59618449e-01 4.20773149e-01 -6.48391187e-01
-1.00195102e-01 -5.58715224e-01 -9.53630269e-01 -7.26444542e-01
-5.06041721e-02 6.16399825e-01 1.00081289e+00 7.15411961e-01
1.99645504e-01 5.71612895e-01 2.72058189e-01 -1.48620427e+00
-2.57078409e-01 -6.80930197e-01 -6.74719572e-01 -1.27999365e-01
-5.18988490e-01 -9.43439841e-01 2.25946307e-02 -9.25915599e-01] | [3.5176894664764404, 1.5237020254135132] |
4f75808b-90cc-4c76-92a4-8b9b1ffc261e | rethinking-range-view-representation-for | 2303.05367 | null | https://arxiv.org/abs/2303.05367v2 | https://arxiv.org/pdf/2303.05367v2.pdf | Rethinking Range View Representation for LiDAR Segmentation | LiDAR segmentation is crucial for autonomous driving perception. Recent trends favor point- or voxel-based methods as they often yield better performance than the traditional range view representation. In this work, we unveil several key factors in building powerful range view models. We observe that the "many-to-one" mapping, semantic incoherence, and shape deformation are possible impediments against effective learning from range view projections. We present RangeFormer -- a full-cycle framework comprising novel designs across network architecture, data augmentation, and post-processing -- that better handles the learning and processing of LiDAR point clouds from the range view. We further introduce a Scalable Training from Range view (STR) strategy that trains on arbitrary low-resolution 2D range images, while still maintaining satisfactory 3D segmentation accuracy. We show that, for the first time, a range view method is able to surpass the point, voxel, and multi-view fusion counterparts in the competing LiDAR semantic and panoptic segmentation benchmarks, i.e., SemanticKITTI, nuScenes, and ScribbleKITTI. | ['Ziwei Liu', 'Yu Qiao', 'Yuenan Hou', 'Yikang Li', 'Xinge Zhu', 'Yuexin Ma', 'Runnan Chen', 'Youquan Liu', 'Lingdong Kong'] | 2023-03-09 | null | null | null | null | ['panoptic-segmentation', 'lidar-semantic-segmentation'] | ['computer-vision', 'computer-vision'] | [ 2.94177622e-01 -1.96969613e-01 -1.31872728e-01 -9.79639709e-01
-7.71997273e-01 -8.95085812e-01 8.44220757e-01 -1.94300830e-01
-3.14300239e-01 3.37853372e-01 -2.33450159e-01 -6.21231735e-01
-3.07196259e-01 -9.82658565e-01 -8.69990468e-01 -4.78729874e-01
3.92076582e-01 1.03848481e+00 4.07146037e-01 -5.29755235e-01
4.21500504e-01 9.87004220e-01 -2.02641892e+00 3.65205202e-03
7.76205540e-01 1.02318203e+00 1.65811375e-01 6.18809283e-01
-2.78247535e-01 1.04333140e-01 9.51341167e-02 -1.54345319e-01
7.25768745e-01 4.00766641e-01 -1.83431849e-01 -4.76127602e-02
1.40942836e+00 -4.36025023e-01 -1.15881488e-01 9.12108719e-01
4.68008727e-01 1.06820039e-01 6.03682578e-01 -1.22363758e+00
-2.77046144e-01 3.24511975e-01 -8.95235598e-01 1.21356256e-01
-1.72227383e-01 2.11381659e-01 1.08287537e+00 -1.25996292e+00
6.16725087e-01 1.49209630e+00 9.10374045e-01 3.67633224e-01
-1.25812674e+00 -8.32842231e-01 3.73220265e-01 -6.00636676e-02
-1.11378407e+00 -3.98800492e-01 8.89604926e-01 -4.85486895e-01
1.19846058e+00 2.96500504e-01 8.25214684e-01 7.87040055e-01
3.06514829e-01 4.96069521e-01 1.49708784e+00 8.05404186e-02
8.00864547e-02 -6.14643022e-02 2.89435744e-01 4.24087107e-01
3.53767425e-01 6.83576643e-01 -6.98769748e-01 1.47996917e-01
6.07754648e-01 1.13859907e-01 -8.07839260e-02 -1.02755892e+00
-8.74366164e-01 9.01945353e-01 7.11410582e-01 -3.52944970e-01
1.35928311e-03 4.06166703e-01 2.09231704e-01 1.20946370e-01
8.10743511e-01 2.84989536e-01 -5.48961699e-01 9.43656191e-02
-1.17762268e+00 2.85066783e-01 5.14806449e-01 1.04042733e+00
1.18758810e+00 1.22389451e-01 5.34100413e-01 5.17033815e-01
4.33428258e-01 9.85470235e-01 -2.47518390e-01 -1.38005865e+00
6.30772352e-01 5.89905798e-01 -1.63967475e-01 -4.90680307e-01
-6.79707766e-01 -5.58384180e-01 -5.90412915e-01 7.62742460e-01
7.88259432e-02 3.38072300e-01 -1.51072073e+00 1.29366338e+00
2.82149643e-01 1.57021843e-02 -5.22731687e-04 8.55717838e-01
9.17469442e-01 4.13475215e-01 -1.48383275e-01 2.39759475e-01
1.00406718e+00 -7.70195663e-01 -2.24469766e-01 -8.39271903e-01
3.98060262e-01 -6.26655459e-01 1.18426347e+00 4.95961726e-01
-1.10473204e+00 -7.98358381e-01 -1.27011502e+00 -2.28706941e-01
-7.12205410e-01 -3.56459796e-01 6.65423214e-01 5.59049845e-01
-1.16045189e+00 5.15215695e-01 -7.73481071e-01 -2.55526394e-01
4.58146930e-01 3.29925328e-01 -4.18436408e-01 -4.12038416e-01
-6.02383077e-01 9.58520174e-01 2.88770676e-01 2.36199982e-02
-5.98912239e-01 -1.22606432e+00 -8.90911698e-01 -2.20529974e-01
5.07303894e-01 -9.38833594e-01 1.07452023e+00 -1.89392883e-02
-9.15595114e-01 1.15864074e+00 -2.02376261e-01 -5.65266013e-01
4.83678699e-01 -5.77994585e-01 -8.00633729e-02 1.88106857e-02
3.10743630e-01 1.35490859e+00 6.00046217e-01 -1.94745636e+00
-8.45356345e-01 -8.75984907e-01 -1.93614304e-01 4.14313555e-01
4.21780854e-01 -6.85738325e-01 -1.91426367e-01 7.04398677e-02
8.17203104e-01 -9.96645212e-01 -2.48874933e-01 6.31126687e-02
-2.83295393e-01 -7.83114433e-02 1.13141489e+00 -8.71331841e-02
2.92874873e-01 -2.16222143e+00 -9.59392786e-02 1.84659526e-01
3.70182455e-01 7.84862414e-02 1.34968951e-01 2.37596929e-01
-1.07200705e-01 1.47476703e-01 -3.95970106e-01 -3.95946205e-01
3.35877761e-02 7.47180343e-01 -5.12121022e-01 4.52517599e-01
8.00622255e-03 1.11419177e+00 -5.20945907e-01 -3.09208065e-01
7.58665323e-01 4.93663579e-01 -5.59922934e-01 -1.95892096e-01
-4.82025981e-01 5.22579730e-01 -2.00666800e-01 6.19035304e-01
1.21924973e+00 -9.84090846e-03 -4.88443017e-01 -2.61389971e-01
-5.60939193e-01 2.50259519e-01 -8.64982247e-01 1.94003475e+00
-4.33900774e-01 5.95574081e-01 4.44645703e-01 -6.13556862e-01
1.08216810e+00 -2.38163784e-01 3.83537710e-01 -8.14029396e-01
-5.67053668e-02 4.43630338e-01 -1.46498263e-01 -2.78109699e-01
7.95439601e-01 -3.07395339e-01 3.39160152e-02 6.94417581e-02
-8.54882747e-02 -9.83808875e-01 1.25392647e-02 1.00871235e-01
3.85208815e-01 4.88347858e-01 -1.80374663e-02 -1.27658680e-01
3.36368233e-01 4.24536586e-01 3.94683003e-01 8.38781118e-01
-9.87787545e-03 7.70454288e-01 1.52924612e-01 -6.48230195e-01
-1.09558773e+00 -1.55605221e+00 -2.99191296e-01 1.04559398e+00
3.47387344e-01 8.78439918e-02 -2.13691711e-01 -4.25900251e-01
4.99036789e-01 1.06735623e+00 -2.16353893e-01 5.42483144e-02
-8.91332746e-01 -3.34166646e-01 3.76714736e-01 8.89963925e-01
5.88289559e-01 -4.64819521e-01 -8.20902050e-01 3.19426656e-02
9.53195542e-02 -1.49979186e+00 -6.37330813e-04 5.96865535e-01
-1.28130245e+00 -8.10434222e-01 -1.02603629e-01 -3.51906598e-01
1.83590978e-01 8.50563824e-01 1.47423005e+00 -3.63973707e-01
-7.52115324e-02 3.54571640e-01 1.12317704e-01 -6.36649013e-01
-8.68668631e-02 1.00368403e-01 -1.80629432e-01 -6.27789974e-01
5.81992507e-01 -9.82063174e-01 -6.35640919e-01 2.46872306e-01
-4.88003582e-01 1.45724326e-01 8.21319044e-01 4.15449679e-01
1.11910951e+00 -4.80917484e-01 7.95638785e-02 -9.50000584e-01
3.09648551e-02 1.26800444e-02 -9.13511992e-01 -1.93724990e-01
-8.41092229e-01 -1.58302724e-01 2.04876766e-01 2.59936154e-01
-9.22472060e-01 3.61060709e-01 -5.14257193e-01 -7.95443833e-01
-6.91532850e-01 2.51496315e-01 -1.17849343e-01 -3.81902486e-01
6.95084631e-01 3.02634269e-01 1.24263488e-01 -4.13738102e-01
1.05908728e+00 2.88641959e-01 7.71497309e-01 -4.46298450e-01
9.47355390e-01 1.28844547e+00 3.72663110e-01 -8.68740857e-01
-1.14948213e+00 -9.14518774e-01 -1.12323165e+00 -1.65162742e-01
1.08739376e+00 -1.36508799e+00 -3.35062623e-01 7.56986141e-02
-1.15056252e+00 -1.01726778e-01 -5.04794657e-01 4.03395295e-01
-7.44780540e-01 6.99991360e-02 -1.14906132e-01 -6.45067453e-01
-1.49388775e-01 -1.19226813e+00 1.57503808e+00 1.99830875e-01
3.19253147e-01 -8.12494636e-01 -1.82253029e-02 5.73428571e-01
3.17089438e-01 2.65444249e-01 9.18841541e-01 -5.84490657e-01
-9.83391166e-01 1.22384861e-01 -6.03348374e-01 1.91644013e-01
-3.58951628e-01 -4.29647453e-02 -1.45136595e+00 -2.12765887e-01
1.90766230e-01 -2.16450617e-01 1.41534722e+00 5.56032836e-01
8.02489877e-01 4.31678057e-01 -5.53587973e-01 1.17411506e+00
1.63363767e+00 5.90710379e-02 3.27266812e-01 2.41618514e-01
1.05354047e+00 7.64473557e-01 5.35306513e-01 -1.20629847e-01
6.61943913e-01 3.96878183e-01 1.08522320e+00 -1.99953735e-01
-1.95806324e-01 -2.79058963e-01 -2.09007263e-01 6.59650326e-01
1.34820808e-02 -1.35802731e-01 -1.11717963e+00 4.55994099e-01
-1.57144344e+00 -7.05056369e-01 -5.49203813e-01 1.95027745e+00
2.20228638e-02 5.27911127e-01 -1.36574268e-01 -1.24362722e-01
1.45671964e-01 5.00617385e-01 -9.47702169e-01 -4.92995799e-01
-4.44798134e-02 2.78568864e-01 1.01188397e+00 5.97883165e-01
-1.07277298e+00 1.22848630e+00 6.00672531e+00 4.67922509e-01
-1.12685108e+00 1.01765677e-01 2.51913369e-01 -6.80940673e-02
-6.50347531e-01 4.71711764e-03 -1.32239008e+00 -2.81683862e-01
7.85222530e-01 4.85288918e-01 1.57473445e-01 9.64254856e-01
1.90808311e-01 -1.66900054e-01 -1.08877635e+00 9.48616922e-01
2.08327889e-01 -1.40413487e+00 1.19595215e-01 4.02277023e-01
5.77451944e-01 9.07739341e-01 2.33005598e-01 6.42870218e-02
3.37083310e-01 -1.03543282e+00 6.77626491e-01 3.08948636e-01
7.14236557e-01 -7.14003503e-01 4.24074590e-01 6.61252081e-01
-1.41284466e+00 -9.35953110e-02 -4.45695043e-01 1.15977645e-01
4.80776221e-01 5.85839629e-01 -5.40274501e-01 9.11859632e-01
6.93259656e-01 8.40783596e-01 -6.51745141e-01 7.80695260e-01
-4.76745032e-02 1.85209543e-01 -6.09916568e-01 5.42493045e-01
7.11163878e-01 -5.08692801e-01 7.32734621e-01 8.76075149e-01
3.87301952e-01 -1.88245863e-01 3.76147836e-01 1.17297113e+00
1.87075272e-01 -4.70521688e-01 -1.32128632e+00 5.04492283e-01
5.02545655e-01 1.35632432e+00 -8.73292863e-01 -2.88712561e-01
-4.90263462e-01 2.18854263e-01 3.91905829e-02 1.51631922e-01
-7.11708844e-01 1.06656626e-01 6.92404449e-01 4.23451662e-01
4.37577546e-01 -6.17001772e-01 -8.91746640e-01 -6.80317283e-01
-6.87550455e-02 -3.70617121e-01 3.53524946e-02 -1.05890095e+00
-1.18412530e+00 3.77545834e-01 4.34035271e-01 -1.12412000e+00
-2.19808862e-01 -8.49028051e-01 -4.06026453e-01 8.85719895e-01
-2.10837078e+00 -1.60496104e+00 -6.49200380e-01 3.33524853e-01
7.66367078e-01 6.77089542e-02 2.85892785e-01 1.50449917e-01
1.19581245e-01 -1.27397314e-01 -1.38077557e-01 -3.95348787e-01
5.69480538e-01 -1.51994359e+00 8.61374080e-01 6.38245881e-01
2.91800797e-01 3.16060603e-01 4.95344400e-01 -6.52254045e-01
-1.36531520e+00 -1.28812647e+00 5.60621917e-01 -7.95614362e-01
4.59693015e-01 -4.34624463e-01 -8.71761918e-01 8.77603829e-01
1.00988775e-01 -2.36705597e-02 3.22155297e-01 -1.41040543e-02
-5.16258419e-01 -2.02630162e-01 -9.17420626e-01 3.29192162e-01
1.42876482e+00 -5.56166768e-01 -6.75025284e-01 1.97059393e-01
8.00871551e-01 -6.60319865e-01 -5.79007387e-01 8.38369668e-01
5.19853234e-01 -1.55457366e+00 1.52564824e+00 -1.90753981e-01
2.97493756e-01 -2.69598156e-01 -5.09244442e-01 -1.29113019e+00
-9.62722301e-02 -2.97420233e-01 1.33156955e-01 8.66844058e-01
3.69770199e-01 -8.93261850e-01 1.12144494e+00 2.83555929e-02
-7.64285326e-01 -5.30136287e-01 -1.24152422e+00 -8.42913628e-01
5.06297529e-01 -7.70629227e-01 5.89788735e-01 5.65000236e-01
-1.00447512e+00 6.84247732e-01 7.86481798e-02 3.70674372e-01
9.65920210e-01 4.52344447e-01 8.98536146e-01 -1.81967402e+00
1.67988494e-01 -5.08274436e-01 -2.33926356e-01 -1.33130431e+00
4.04062755e-02 -9.90566730e-01 2.24431470e-01 -1.82502055e+00
-2.53669143e-01 -5.33340037e-01 2.27372721e-01 -3.64702120e-02
2.56612390e-01 3.35879833e-01 2.56252944e-01 2.96259165e-01
-1.53683931e-01 2.87606269e-01 1.51968265e+00 5.88349737e-02
-1.38631627e-01 3.32807712e-02 -6.16652906e-01 1.15246773e+00
8.87622535e-01 -3.43333781e-01 -6.45224869e-01 -8.31448615e-01
3.74581486e-01 -3.09828389e-02 7.02726603e-01 -1.22500265e+00
3.58182311e-01 -3.39498906e-03 2.84563631e-01 -1.68360388e+00
1.02048361e+00 -9.34520006e-01 -1.80897430e-01 2.36326993e-01
4.91522551e-02 3.83454442e-01 2.48087227e-01 7.52703726e-01
7.12557137e-02 1.42651513e-01 8.33216369e-01 -6.25604928e-01
-9.47130144e-01 2.86201298e-01 1.44873485e-01 1.02227263e-01
7.08892286e-01 -7.58763254e-01 -6.42710447e-01 -8.86797234e-02
-6.64308846e-01 4.81075406e-01 5.17320037e-01 3.20152342e-01
7.32162118e-01 -9.57291901e-01 -5.57622313e-01 5.97280860e-01
2.15927914e-01 7.05340743e-01 8.71127564e-03 8.80207956e-01
-5.45646071e-01 5.81062376e-01 -1.95765555e-01 -1.32953262e+00
-1.20329034e+00 2.38342881e-01 4.20874685e-01 -4.67293374e-02
-9.62425292e-01 9.24568295e-01 4.46530759e-01 -1.09464049e+00
3.32139735e-03 -5.69342256e-01 2.24665590e-02 1.23881146e-01
-1.68819308e-01 3.82591635e-01 2.64208555e-01 -8.11991394e-01
-2.35497192e-01 1.20099878e+00 -4.57599908e-02 -3.30265611e-01
1.28240752e+00 -2.75981963e-01 -3.30653265e-02 7.34723210e-01
9.26370144e-01 -3.24087173e-01 -1.40065658e+00 -5.29858842e-02
-2.27957100e-01 -3.85288805e-01 1.76371902e-01 -6.94296956e-01
-8.69235992e-01 1.63558114e+00 9.02882814e-01 1.42318070e-01
5.30327439e-01 1.15286432e-01 8.11516404e-01 4.58282530e-01
3.15166861e-01 -1.00748348e+00 -2.52118498e-01 7.75669634e-01
5.88685274e-01 -1.37173831e+00 2.12585896e-01 -7.30979025e-01
-3.15291375e-01 9.86620963e-01 8.01650405e-01 -2.21704230e-01
7.72587061e-01 3.79969716e-01 2.51073271e-01 -7.84307957e-01
-5.96730113e-01 -4.15198326e-01 3.54894489e-01 7.42076635e-01
7.00904280e-02 3.02588120e-02 4.00132239e-01 -3.35764557e-01
-6.18696153e-01 -2.95553297e-01 2.42409915e-01 9.58979309e-01
-9.32794809e-01 -7.22600102e-01 -3.53229940e-01 6.00549519e-01
1.91776291e-01 -1.15641477e-02 -5.59107780e-01 1.34114695e+00
3.03083539e-01 6.87494040e-01 6.59057915e-01 -4.19609278e-01
6.84853435e-01 2.70860255e-01 4.55720514e-01 -6.25714481e-01
-4.13687557e-01 3.65072608e-01 1.68475974e-02 -7.50872850e-01
-4.82929170e-01 -5.37217200e-01 -1.11437893e+00 -3.70621562e-01
-3.45774829e-01 -2.48251975e-01 9.09958243e-01 1.02311373e+00
3.35087538e-01 3.98513049e-01 4.14605498e-01 -1.36785090e+00
-5.18910408e-01 -5.62429667e-01 -2.93674052e-01 -1.48105267e-02
3.05821240e-01 -7.84875929e-01 -4.41254407e-01 -2.89924681e-01] | [8.184155464172363, -2.814225196838379] |
f844b8fe-eac5-4f01-b145-ea5b082d95bb | what-if-we-only-use-real-datasets-for-scene | 2103.04400 | null | https://arxiv.org/abs/2103.04400v2 | https://arxiv.org/pdf/2103.04400v2.pdf | What If We Only Use Real Datasets for Scene Text Recognition? Toward Scene Text Recognition With Fewer Labels | Scene text recognition (STR) task has a common practice: All state-of-the-art STR models are trained on large synthetic data. In contrast to this practice, training STR models only on fewer real labels (STR with fewer labels) is important when we have to train STR models without synthetic data: for handwritten or artistic texts that are difficult to generate synthetically and for languages other than English for which we do not always have synthetic data. However, there has been implicit common knowledge that training STR models on real data is nearly impossible because real data is insufficient. We consider that this common knowledge has obstructed the study of STR with fewer labels. In this work, we would like to reactivate STR with fewer labels by disproving the common knowledge. We consolidate recently accumulated public real data and show that we can train STR models satisfactorily only with real labeled data. Subsequently, we find simple data augmentation to fully exploit real data. Furthermore, we improve the models by collecting unlabeled data and introducing semi- and self-supervised methods. As a result, we obtain a competitive model to state-of-the-art methods. To the best of our knowledge, this is the first study that 1) shows sufficient performance by only using real labels and 2) introduces semi- and self-supervised methods into STR with fewer labels. Our code and data are available: https://github.com/ku21fan/STR-Fewer-Labels | ['Kiyoharu Aizawa', 'Yusuke Matsui', 'Jeonghun Baek'] | 2021-03-07 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Baek_What_if_We_Only_Use_Real_Datasets_for_Scene_Text_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Baek_What_if_We_Only_Use_Real_Datasets_for_Scene_Text_CVPR_2021_paper.pdf | cvpr-2021-1 | ['scene-text-recognition'] | ['computer-vision'] | [ 3.19696039e-01 2.92092025e-01 -1.51928738e-01 -3.15247029e-01
-7.22201765e-01 -7.03183115e-01 5.62284231e-01 -5.88553250e-02
-4.05785650e-01 8.94841433e-01 -7.71969333e-02 -4.86265779e-01
4.23745990e-01 -7.05002666e-01 -1.02105069e+00 -2.91701853e-01
5.54082334e-01 7.17746019e-01 2.30340824e-01 -1.74685568e-01
6.69444799e-02 3.08808923e-01 -1.54149532e+00 5.80442786e-01
9.70769227e-01 4.69929308e-01 1.34804264e-01 4.27822679e-01
-1.02105767e-01 9.38512683e-01 -4.85830188e-01 -3.23348820e-01
3.77575010e-01 -4.59879875e-01 -8.47126544e-01 4.35951650e-01
4.33964640e-01 -3.31176162e-01 -3.61427307e-01 8.15660417e-01
2.05788136e-01 -2.14548931e-01 6.80690110e-01 -1.14540434e+00
-7.32427061e-01 6.24471247e-01 -5.50642192e-01 -3.69592130e-01
3.67527455e-01 4.23225582e-01 8.47972810e-01 -9.78614271e-01
8.09664130e-01 9.57079887e-01 6.16391540e-01 7.28557706e-01
-1.45938706e+00 -5.45055151e-01 -9.30557027e-02 -1.09516345e-01
-1.45721698e+00 -4.38085645e-01 8.12227726e-01 -4.87285018e-01
6.90994918e-01 2.78200418e-01 4.02593404e-01 1.79848778e+00
-4.98384714e-01 1.11219060e+00 1.28623998e+00 -9.31285203e-01
1.70303300e-01 4.01562899e-01 -1.88848767e-02 5.74956238e-01
3.94518137e-01 1.81482863e-02 -4.08553958e-01 2.14333594e-01
6.65566504e-01 -1.44246757e-01 -2.03066260e-01 -2.32553914e-01
-1.40852416e+00 8.13297093e-01 -2.21788753e-02 4.67347622e-01
1.57069325e-01 -9.07267854e-02 3.21330488e-01 3.27454895e-01
3.47011715e-01 6.38904989e-01 -4.62915033e-01 -1.99918121e-01
-1.05580664e+00 -4.21671644e-02 9.21256721e-01 1.06912899e+00
8.87445688e-01 2.67385125e-01 2.68988043e-01 9.80267763e-01
5.06944731e-02 5.89019239e-01 6.37452066e-01 -8.46712351e-01
5.44340670e-01 6.70597732e-01 1.92709833e-01 -6.74589455e-01
-1.68654665e-01 -2.54193991e-01 -7.18472421e-01 7.51116723e-02
8.56268942e-01 -1.06025748e-01 -9.83563423e-01 1.60050857e+00
-2.81406716e-02 3.10118854e-01 2.12675110e-01 7.02332854e-01
8.13254416e-01 3.93389523e-01 -2.71383673e-01 4.21729051e-02
9.23289359e-01 -1.01080477e+00 -5.76224864e-01 -1.81065470e-01
1.33961082e+00 -8.28866959e-01 1.61080039e+00 6.71308219e-01
-5.47836542e-01 -5.79844534e-01 -1.09880054e+00 1.35216773e-01
-5.12744904e-01 3.93576860e-01 5.24025202e-01 9.61974502e-01
-8.53758335e-01 4.92208958e-01 -8.65043938e-01 -6.43376827e-01
4.34509873e-01 5.58456182e-02 -5.23426950e-01 -3.03482145e-01
-1.11856103e+00 7.42008746e-01 4.01385456e-01 1.06969222e-01
-7.47851133e-01 -3.51496875e-01 -1.03053427e+00 -5.19328415e-01
6.61186874e-01 -3.55295271e-01 1.41835260e+00 -1.06622231e+00
-1.45216477e+00 1.00774264e+00 -6.04796223e-02 -4.80602205e-01
8.60370457e-01 -1.84887186e-01 -4.53900039e-01 6.58514276e-02
-2.03126948e-02 7.26122677e-01 6.65080667e-01 -1.45553553e+00
-2.01217175e-01 1.03446469e-02 -1.94648117e-01 -9.45165753e-02
-3.28822166e-01 -1.23138785e-01 -1.27692923e-01 -6.12188697e-01
7.35171437e-02 -8.79499435e-01 -9.67247933e-02 -1.21643655e-01
-6.86981738e-01 -6.32122159e-02 1.00124383e+00 -4.28738236e-01
7.84783840e-01 -2.22512650e+00 -4.17460263e-01 -1.30727813e-01
2.06668720e-01 5.61202168e-01 -2.19322711e-01 5.71668506e-01
-7.05756322e-02 4.47661847e-01 -3.82513821e-01 -4.50912893e-01
2.34579314e-02 2.93605238e-01 -5.78210711e-01 2.95373023e-01
3.16135854e-01 9.88662899e-01 -9.68499243e-01 -5.84051251e-01
3.51140916e-01 2.27668226e-01 -3.83072138e-01 1.02752626e-01
-5.37694752e-01 6.06667817e-01 -4.56123799e-01 5.17440975e-01
6.85725927e-01 -4.34262276e-01 4.60362956e-02 -1.01688607e-02
-1.39701828e-01 1.87629342e-01 -1.35160518e+00 1.52248418e+00
-3.66354227e-01 7.55889237e-01 -5.70766628e-01 -1.09198916e+00
1.10727394e+00 2.19988450e-01 3.16586047e-01 -5.60866535e-01
2.58894265e-02 4.57098335e-01 -2.08212212e-01 -6.47813141e-01
4.61103916e-01 -1.06090218e-01 -6.80854404e-03 6.68281555e-01
-2.00918540e-02 -4.93045658e-01 4.16244715e-01 2.20279336e-01
1.19337499e+00 3.58134478e-01 -2.94225626e-02 1.56109825e-01
2.46231511e-01 2.30540767e-01 4.13233846e-01 8.10669661e-01
-6.81215823e-02 1.01757014e+00 4.41199332e-01 -4.43303823e-01
-1.35840642e+00 -8.77982974e-01 -1.67853266e-01 5.21983206e-01
-1.55697212e-01 -4.25383836e-01 -7.13188887e-01 -8.26179326e-01
-1.73895493e-01 1.09917068e+00 -5.76193869e-01 1.30477384e-01
-4.19065058e-01 -5.15149534e-01 9.74150598e-01 3.96608472e-01
5.32492340e-01 -1.24061036e+00 -5.33982754e-01 4.97876108e-02
-6.49922788e-02 -1.56794858e+00 -1.62780702e-01 1.46852538e-01
-6.62996173e-01 -1.02001202e+00 -5.24691820e-01 -5.80387652e-01
7.83644855e-01 1.32680908e-01 8.98721337e-01 1.58695117e-01
-1.91941738e-01 1.99223652e-01 -7.10736632e-01 -3.97126317e-01
-7.06807733e-01 7.70537034e-02 -6.16765097e-02 1.22932002e-01
3.25657606e-01 -3.73971969e-01 -1.59914047e-01 4.76047605e-01
-1.13563454e+00 5.48647821e-01 5.39521873e-01 8.98320138e-01
4.29025918e-01 -5.54313548e-02 5.30902803e-01 -1.39650011e+00
2.47352645e-01 -1.48063287e-01 -5.05752027e-01 2.48822585e-01
-3.41987461e-01 1.04210317e-01 8.31565678e-01 -5.87311387e-01
-1.10493076e+00 1.99716523e-01 -1.16850436e-01 -5.20156145e-01
-7.19859123e-01 5.06443202e-01 9.53068659e-02 1.77304104e-01
1.02193141e+00 3.37323189e-01 -2.24728450e-01 -5.06683886e-01
3.34541023e-01 9.16548073e-01 2.19253495e-01 -7.91395843e-01
7.32666135e-01 6.69709146e-01 -1.21999092e-01 -1.04235816e+00
-1.14776039e+00 -2.77298689e-01 -7.83175111e-01 -1.06460422e-01
5.37512422e-01 -7.47123003e-01 -3.00325215e-01 6.85917199e-01
-1.00477958e+00 -9.47219253e-01 -6.06664002e-01 5.48242629e-01
-6.77902758e-01 7.33996093e-01 -6.11143947e-01 -8.78620088e-01
8.02111551e-02 -9.71962273e-01 1.19614303e+00 -9.72256437e-02
-2.32203573e-01 -8.51346791e-01 1.64015725e-01 4.64007378e-01
1.46456227e-01 3.95727783e-01 2.99808294e-01 -8.65969718e-01
-5.55836499e-01 -2.21502945e-01 -2.44803265e-01 4.68661904e-01
2.38998413e-01 3.20521742e-01 -1.14299417e+00 -1.63536891e-01
-8.44956785e-02 -9.74649668e-01 8.34502339e-01 -1.71889603e-01
1.14409649e+00 -9.61380377e-02 -1.11644730e-01 3.79664987e-01
1.27845907e+00 -1.46582991e-01 8.59448671e-01 3.87455255e-01
8.56763184e-01 6.91278219e-01 6.59832954e-01 2.67319441e-01
3.31183940e-01 5.98349929e-01 1.39189810e-01 -3.87363493e-01
-2.19241157e-01 -5.80247819e-01 3.29561651e-01 8.33252907e-01
2.36268595e-01 -4.61614579e-01 -1.23745382e+00 5.39496183e-01
-1.93433630e+00 -7.81655908e-01 -5.37635982e-01 2.28820729e+00
1.02425838e+00 2.92340368e-01 -2.82400139e-02 3.44155550e-01
6.19174302e-01 -7.90918842e-02 -3.30351770e-01 -1.16493568e-01
-4.73026782e-01 1.46956265e-01 4.47277218e-01 3.49663734e-01
-1.13441956e+00 1.16806352e+00 5.78699064e+00 9.12268579e-01
-1.32052004e+00 3.86290625e-02 6.40123725e-01 1.24934234e-01
-2.96097606e-01 2.26876274e-01 -8.49279702e-01 4.25617367e-01
9.01661575e-01 2.39052862e-01 1.78083643e-01 7.63325512e-01
1.64315313e-01 -1.06384099e-01 -1.36537147e+00 9.71071780e-01
1.37563616e-01 -1.23546553e+00 2.76649091e-02 3.53978574e-02
7.71735311e-01 1.69633046e-01 -9.54791158e-02 3.13699514e-01
5.68374574e-01 -1.18777525e+00 7.20364511e-01 3.69340301e-01
9.82656538e-01 -2.29195178e-01 6.70524061e-01 6.68717027e-01
-7.64375925e-01 3.08494717e-01 -3.43538970e-01 1.91631317e-02
7.83125684e-02 7.71149218e-01 -1.09413671e+00 6.53024912e-01
3.86264622e-01 9.19555485e-01 -8.43590975e-01 8.59895349e-01
-5.12449145e-01 9.65160847e-01 -4.30106312e-01 6.86520338e-02
1.02497719e-01 -1.05602801e-01 1.98900312e-01 1.29450965e+00
3.94378036e-01 -1.40240103e-01 2.37333193e-01 9.20990109e-01
7.72012547e-02 1.90066621e-01 -1.11754918e+00 -1.39397457e-01
2.19285563e-01 1.00761044e+00 -7.10054159e-01 -2.56448120e-01
-6.18677914e-01 7.39709437e-01 3.64623934e-01 4.42185163e-01
-6.15525842e-01 -8.54027271e-02 -2.47815296e-01 3.49904150e-01
1.60420358e-01 -3.25264394e-01 -6.07351065e-01 -1.51751053e+00
4.22724783e-02 -9.63754296e-01 4.27544862e-01 -1.01744831e+00
-1.37128031e+00 6.29663050e-01 -9.46439505e-02 -1.47019577e+00
-3.56673747e-01 -6.59847677e-01 -1.52529091e-01 6.66232288e-01
-1.45623362e+00 -1.32548428e+00 -2.38806412e-01 4.96423483e-01
6.21241868e-01 -2.44366705e-01 8.18524659e-01 7.32912570e-02
-6.64087355e-01 7.29091108e-01 2.15347424e-01 3.23620617e-01
1.01379550e+00 -9.58507478e-01 4.02807444e-01 8.51589799e-01
4.48848158e-01 3.54519546e-01 7.74674773e-01 -7.03632891e-01
-1.15250432e+00 -9.76518154e-01 8.37915063e-01 -6.51747763e-01
9.07864809e-01 -7.41011202e-01 -1.14068186e+00 8.39214504e-01
-2.51078326e-02 8.24911296e-02 6.01446211e-01 -6.32098168e-02
-6.03021979e-01 1.93830878e-01 -1.16185868e+00 8.23577642e-01
1.05780947e+00 -4.24489051e-01 -4.61651117e-01 5.94001651e-01
6.17805302e-01 -2.46047467e-01 -4.83082265e-01 4.86059815e-01
1.99089780e-01 -8.93576980e-01 5.24903238e-01 -3.92390668e-01
5.65371990e-01 -4.71294254e-01 -3.03461313e-01 -1.03394616e+00
3.10333192e-01 -4.41498101e-01 1.04677185e-01 1.26509738e+00
6.84888840e-01 -8.12211394e-01 1.04300249e+00 5.42048156e-01
-1.28949106e-01 -4.74451751e-01 -7.21310139e-01 -1.17151415e+00
1.38585895e-01 -6.27173185e-01 3.02759498e-01 1.43958986e+00
-4.54536127e-03 2.79191792e-01 -4.07865763e-01 -1.47391871e-01
5.56261420e-01 9.65230260e-03 1.16726804e+00 -1.13998377e+00
-4.44486737e-01 -1.38036266e-01 -1.89207077e-01 -9.37179863e-01
3.27521324e-01 -8.50304365e-01 1.29730469e-02 -1.31509244e+00
3.45593661e-01 -6.62748754e-01 1.75831661e-01 9.34410930e-01
1.47277683e-01 3.98138821e-01 5.00095673e-02 3.71702373e-01
-5.34745395e-01 5.78146338e-01 1.35248232e+00 -3.94350290e-02
-6.26762435e-02 -3.10286433e-01 -6.07924521e-01 6.30752981e-01
1.10244513e+00 -5.25912344e-01 -3.58013898e-01 -3.22868139e-01
2.10910648e-01 -2.68051863e-01 3.63800138e-01 -7.64202416e-01
8.46550614e-02 -3.22724193e-01 1.05935507e-01 -4.70660716e-01
2.40945607e-01 -6.03317678e-01 2.07376629e-01 2.02256337e-01
-2.20765665e-01 -4.48134512e-01 1.59771234e-01 2.13208482e-01
-1.97395459e-01 -5.06118655e-01 6.66785181e-01 -2.00017422e-01
-5.20776689e-01 -2.33293027e-02 -3.90913904e-01 9.21657979e-02
9.29733694e-01 -4.83849078e-01 -5.60288310e-01 -4.69335109e-01
-5.32818913e-01 3.27966139e-02 8.44945908e-01 2.28604227e-01
5.87554753e-01 -1.09995663e+00 -1.02227724e+00 3.20128471e-01
3.44303489e-01 3.26378465e-01 8.45101178e-02 6.75786614e-01
-7.24607170e-01 3.50836426e-01 -4.99546528e-04 -7.44921982e-01
-1.03925860e+00 7.13062108e-01 8.08132589e-02 -4.83642429e-01
-6.48255408e-01 3.60717982e-01 2.14420885e-01 -8.55548322e-01
1.78888422e-02 -7.72005171e-02 2.76611261e-02 -2.91577429e-01
1.63718358e-01 -1.68597815e-03 2.16067538e-01 -6.24021590e-01
5.55865141e-03 4.05352324e-01 -2.01320142e-01 -3.85610312e-01
1.38231325e+00 2.41699040e-01 2.35469311e-01 6.90636873e-01
9.11481738e-01 3.62367220e-02 -1.23079205e+00 -3.07844937e-01
-8.57818930e-04 -6.05287492e-01 -3.73815268e-01 -6.40750647e-01
-7.84523726e-01 9.39984441e-01 1.85915962e-01 1.67998686e-01
9.47542548e-01 -2.57714968e-02 4.94579107e-01 8.01273644e-01
5.90957820e-01 -1.11931992e+00 4.67091382e-01 6.35702014e-01
8.58573794e-01 -1.32891822e+00 -1.09558716e-01 -6.62278533e-01
-7.88982034e-01 9.32771087e-01 6.96896672e-01 1.51925655e-02
3.87688398e-01 4.30830747e-01 2.37726003e-01 2.26402916e-02
-7.65626132e-01 -2.16355816e-01 -1.58467025e-01 6.19263828e-01
5.00076175e-01 3.76822017e-02 -1.74808919e-01 2.38911584e-01
-3.10409963e-01 3.32503676e-01 9.24970388e-01 1.12131536e+00
-1.21819891e-01 -1.61831212e+00 -4.59615976e-01 5.15002251e-01
-2.46160612e-01 -1.74730688e-01 -6.08823538e-01 9.33585286e-01
-1.11720838e-01 1.00844872e+00 -3.16353261e-01 -3.66480023e-01
3.51585835e-01 1.47021711e-01 5.79687834e-01 -8.64751577e-01
-2.19214395e-01 9.07811336e-03 3.50613803e-01 -6.10379539e-02
-2.85189122e-01 -8.03350270e-01 -1.12457287e+00 -1.85066029e-01
-3.14258844e-01 -6.04854375e-02 5.60506999e-01 8.55223417e-01
1.96783885e-01 7.37510026e-02 6.75158918e-01 -6.41027212e-01
-4.74725127e-01 -1.07430315e+00 -5.26234031e-01 5.89873374e-01
3.00234165e-02 -4.67411548e-01 -4.80883688e-01 4.64688301e-01] | [11.646038055419922, 1.9566041231155396] |
c5db257a-37be-478b-babe-afca75494523 | vandermonde-trajectory-bounds-for-linear | 2302.10995 | null | https://arxiv.org/abs/2302.10995v1 | https://arxiv.org/pdf/2302.10995v1.pdf | Vandermonde Trajectory Bounds for Linear Companion Systems | Fast and accurate safety assessment and collision checking are essential for motion planning and control of highly dynamic autonomous robotic systems. Informative, intuitive, and explicit motion trajectory bounds enable explainable and time-critical safety verification of autonomous robot motion. In this paper, we consider feedback linearization of nonlinear systems in the form of proportional-and-higher-order-derivative (PhD) control corresponding to companion dynamics. We introduce a novel analytic convex trajectory bound, called $\textit{Vandermonde simplex}$, for high-order companion systems, that is given by the convex hull of a finite weighted combination of system position, velocity, and other relevant higher-order state variables. Our construction of Vandermonde simplexes is built based on expressing the solution trajectory of companion dynamics in a newly introduced family of $\textit{Vandermonde basis functions}$ that offer new insights for understanding companion system motion compared to the classical exponential basis functions. In numerical simulations, we demonstrate that Vandermonde simplexes offer significantly more accurate motion prediction (e.g., at least an order of magnitude improvement in estimated motion volume) for describing the motion trajectory of companion systems compared to the standard invariant Lyapunov ellipsoids as well as exponential simplexes built based on exponential basis functions. | ['Aykut İşleyen', 'Ömür Arslan'] | 2023-02-21 | null | null | null | null | ['motion-prediction', 'motion-planning'] | ['computer-vision', 'robots'] | [-3.88464898e-01 6.12710059e-01 -1.29576385e-01 3.56848031e-01
-4.34961051e-01 -7.57565618e-01 5.39255619e-01 1.30273044e-01
-4.54844058e-01 9.08868790e-01 -3.30477387e-01 -6.45416915e-01
-5.22031546e-01 -2.63658553e-01 -9.78431761e-01 -9.49247003e-01
-6.23961687e-01 3.29683870e-01 1.24437429e-01 -6.75287485e-01
-8.81971717e-02 5.89189410e-01 -1.02754581e+00 -7.30912268e-01
9.59200084e-01 8.28655064e-01 5.68148047e-02 5.28004050e-01
8.10550928e-01 1.06528230e-01 -5.12769073e-02 1.20507754e-01
3.45782220e-01 5.30799627e-02 -3.52288693e-01 -2.02881888e-01
-3.71745765e-01 -3.15017045e-01 -4.70809281e-01 1.01593363e+00
5.60253263e-02 4.84335542e-01 8.77358079e-01 -1.82290304e+00
-1.14750311e-01 2.75127590e-01 -2.09133133e-01 -2.04635382e-01
-4.44370732e-02 4.31623638e-01 3.63367170e-01 -8.05053890e-01
5.38148582e-01 1.12533915e+00 7.85286963e-01 5.13876438e-01
-1.06242180e+00 -4.37354118e-01 -6.95792139e-02 1.12797819e-01
-1.62306452e+00 -1.71844125e-01 3.28242749e-01 -7.60750234e-01
7.34579325e-01 6.07606173e-01 6.24560416e-01 5.33183694e-01
8.86972010e-01 4.25787538e-01 3.26791197e-01 -1.45556713e-02
1.11592794e-02 1.28136814e-01 2.40373492e-01 1.03780830e+00
6.55030727e-01 5.03968298e-01 3.31437856e-01 -2.94689953e-01
7.51531303e-01 -2.36506999e-01 -5.62760532e-01 -8.30960572e-01
-1.38485181e+00 8.06356370e-01 3.08823586e-01 -3.70907426e-01
-3.81512582e-01 3.60557467e-01 3.24849099e-01 -1.19540676e-01
-4.61373404e-02 4.01277274e-01 -9.87945497e-02 -1.00884110e-01
-8.80246237e-02 4.60410446e-01 6.65541112e-01 1.24202085e+00
3.85826349e-01 3.96876991e-01 -3.66029404e-02 -3.10656000e-02
4.72221285e-01 8.51343393e-01 -2.15870127e-01 -1.18284631e+00
2.48804703e-01 5.61850488e-01 7.49150455e-01 -8.45522404e-01
-8.31782699e-01 -7.51004666e-02 -1.04622781e+00 6.39099836e-01
3.80654544e-01 -2.15542808e-01 -3.20529431e-01 1.70022309e+00
5.72577536e-01 -2.49260113e-01 3.36956263e-01 8.94324005e-01
-1.29416615e-01 8.01932216e-01 -3.68937969e-01 -4.87307310e-01
1.06643009e+00 -2.88973749e-01 -5.70507586e-01 2.70552635e-01
7.16300547e-01 -5.72385080e-02 8.61253798e-01 2.42002025e-01
-1.04548621e+00 -7.08011985e-02 -1.34604347e+00 3.50331783e-01
1.95182547e-01 2.70450532e-01 -1.85877308e-01 1.31171546e-03
-7.84389257e-01 7.47402310e-01 -1.07231236e+00 3.80166370e-04
-2.19818637e-01 4.60909963e-01 -4.86073285e-01 6.21049464e-01
-1.10263765e+00 1.36190569e+00 3.20288986e-01 3.17136556e-01
-1.00698245e+00 -8.74581039e-01 -1.11519468e+00 -2.37022847e-01
5.64232588e-01 -3.16814929e-01 1.08096874e+00 2.05662817e-01
-1.46882379e+00 1.73051327e-01 -5.37120700e-02 -5.12410164e-01
6.19432509e-01 -1.88119233e-01 9.57946107e-03 1.42400816e-01
-1.10635564e-01 3.05350333e-01 7.85618961e-01 -1.31420362e+00
-1.90798506e-01 -2.32657909e-01 -6.97170123e-02 1.16795450e-01
-2.66719222e-01 -5.32264233e-01 -1.38806878e-02 -1.88952073e-01
-1.61761120e-01 -1.45941377e+00 -5.55957079e-01 3.13634843e-01
-3.56804997e-01 -2.77074844e-01 9.85225022e-01 -5.38137853e-01
1.39721489e+00 -2.04973745e+00 5.52775502e-01 3.39546233e-01
3.09301555e-01 4.29158032e-01 3.44664454e-01 5.77330112e-01
1.39077291e-01 4.80576307e-02 -2.77310878e-01 1.89936176e-01
1.39409885e-01 1.47168159e-01 -5.19680142e-01 1.03956711e+00
2.58713692e-01 5.84174871e-01 -8.81812334e-01 -8.53244439e-02
3.26159596e-01 3.78775150e-01 -2.44693592e-01 -2.44916186e-01
-3.65147064e-03 4.56660658e-01 -7.46967852e-01 6.72111213e-02
3.82727861e-01 1.60540849e-01 -2.75596231e-01 -1.68220311e-01
-6.99017823e-01 -3.38036209e-01 -9.84384060e-01 7.41427422e-01
-2.25955695e-01 7.47136176e-01 7.25573599e-01 -8.81798148e-01
8.76183152e-01 2.22543150e-01 6.01058185e-01 -2.62921434e-02
5.93220055e-01 2.94219047e-01 -6.92757443e-02 -2.19006255e-01
6.45824611e-01 -1.28724799e-02 -2.28439867e-01 6.58940449e-02
-6.76377535e-01 -5.44912815e-01 6.13877028e-02 1.97486579e-01
8.84476185e-01 -2.29338929e-01 4.93012786e-01 -8.19297671e-01
9.94908512e-01 2.92745978e-01 4.53105807e-01 -4.14774865e-02
-5.39163649e-01 -9.61911231e-02 5.95888078e-01 1.26059264e-01
-1.53392267e+00 -8.50397944e-01 -3.47310811e-01 2.08251119e-01
6.38757586e-01 -4.09789294e-01 -8.46846819e-01 1.03230081e-01
3.81234854e-01 9.29906964e-01 -5.10490656e-01 -8.77801359e-01
-8.05795491e-01 -1.57979175e-01 5.75631678e-01 6.06289446e-01
1.87526926e-01 -2.16098905e-01 -1.03931344e+00 1.45259693e-01
1.40955687e-01 -1.02106285e+00 -6.95944011e-01 -2.43562497e-02
-5.42730391e-01 -1.21923375e+00 -4.93745297e-01 -5.55143476e-01
6.97531104e-01 8.37093219e-02 2.01110408e-01 -1.83219418e-01
-2.38220841e-01 7.26616263e-01 6.38555959e-02 -5.18835008e-01
-7.74869800e-01 -6.60234988e-01 7.86599338e-01 -1.66123852e-01
-5.34370899e-01 5.78749478e-02 -4.69200104e-01 8.76608849e-01
-6.98278904e-01 -1.11820042e-01 1.54200718e-01 7.96505570e-01
7.65872717e-01 1.72258355e-02 4.18002367e-01 1.98214307e-01
7.04285860e-01 -4.41316038e-01 -1.36559308e+00 9.20360312e-02
-6.77375734e-01 4.11743999e-01 9.25391912e-01 -4.47351903e-01
-5.58706701e-01 2.71067888e-01 3.26387584e-01 -7.39666879e-01
5.23763239e-01 4.14585412e-01 2.89846271e-01 -3.29999834e-01
5.55645168e-01 2.15160891e-01 6.53558493e-01 9.40966755e-02
5.28013587e-01 3.96796703e-01 8.22148919e-01 -7.03992426e-01
9.50393736e-01 5.85469127e-01 8.94832730e-01 -9.00708616e-01
2.60819644e-01 -4.87397283e-01 -5.56657851e-01 -5.58999300e-01
7.48403907e-01 -3.72774959e-01 -1.61264408e+00 3.10184330e-01
-1.13445055e+00 -4.10994232e-01 -3.33557665e-01 5.59331656e-01
-1.13593245e+00 4.77573037e-01 -3.98244411e-01 -1.52795696e+00
-2.42218494e-01 -1.26507759e+00 1.04724753e+00 -5.78251779e-02
-2.33321190e-01 -7.99496770e-01 7.49872029e-02 -3.35669935e-01
1.75954178e-01 7.79428899e-01 9.50045347e-01 1.31613448e-01
-3.33114147e-01 -5.28578937e-01 2.53394574e-01 1.16350114e-01
-3.80350798e-01 3.30028892e-01 -5.68749979e-02 -7.06181586e-01
2.79499680e-01 -6.04783967e-02 3.39857638e-01 4.27131534e-01
3.08202356e-01 -5.88653982e-01 -8.63812983e-01 2.97186404e-01
1.22824681e+00 4.10110176e-01 3.58032733e-01 2.64380515e-01
5.22649109e-01 7.41875827e-01 1.01563096e+00 5.87707818e-01
4.71764654e-01 6.92641914e-01 8.23550344e-01 5.74718893e-01
5.88291049e-01 6.91519827e-02 7.38051176e-01 3.05523396e-01
-6.75208569e-02 6.29485622e-02 -9.65909362e-01 6.67731762e-01
-1.89496160e+00 -6.25648558e-01 -7.06219137e-01 2.58653760e+00
4.98336256e-01 2.11454369e-02 3.02280068e-01 1.63539246e-01
9.51991439e-01 -4.80903685e-01 -9.17388320e-01 -4.46256727e-01
2.00837657e-01 -4.22068447e-01 9.93493795e-01 9.37797368e-01
-8.50989521e-01 2.80489057e-01 5.30270863e+00 7.45674551e-01
-9.78409588e-01 -3.27252984e-01 -9.11064595e-02 2.09564582e-01
-9.64245945e-02 2.84776222e-02 -8.90139341e-01 2.25904569e-01
1.12104344e+00 -9.33408618e-01 3.10405433e-01 1.00151205e+00
6.45108104e-01 -1.05890416e-01 -9.63605940e-01 5.53691328e-01
-5.36786616e-01 -1.14001060e+00 -4.02055651e-01 3.06800216e-01
5.59842110e-01 -2.30062291e-01 2.86284268e-01 -2.49846261e-02
1.49055153e-01 -7.73634195e-01 1.14739656e+00 4.21124309e-01
8.21615160e-01 -8.99106205e-01 4.03180629e-01 7.50344813e-01
-1.59858489e+00 -3.61904621e-01 -1.72950983e-01 2.40561161e-02
6.96763873e-01 -2.17207484e-02 -8.14521492e-01 6.57923698e-01
1.40586033e-01 5.58766186e-01 2.02844352e-01 9.44395185e-01
3.02423328e-01 1.63885534e-01 -6.82674885e-01 -5.60706794e-01
3.94240379e-01 -5.91108263e-01 1.33389103e+00 5.59504926e-01
4.36627567e-01 5.05262554e-01 1.37041107e-01 9.01844203e-01
7.19749689e-01 -3.04758489e-01 -7.48626769e-01 1.47584053e-02
3.47128361e-01 1.07779133e+00 -4.08736914e-01 -4.89168335e-03
1.90985397e-01 4.30251002e-01 -5.24260513e-02 2.92608410e-01
-1.14468908e+00 -5.80051422e-01 1.09929109e+00 1.24592483e-01
4.49909940e-02 -1.05703449e+00 -2.52630562e-01 -6.99294686e-01
2.90051818e-01 -1.33549199e-01 -6.20202646e-02 -5.08519888e-01
-6.06136382e-01 4.89566475e-01 5.83453238e-01 -1.69045103e+00
-8.13691020e-01 -8.73252034e-01 -6.10418081e-01 7.33992994e-01
-8.38107169e-01 -4.58741158e-01 1.04371423e-03 7.32098997e-01
6.03201985e-02 2.66494393e-01 6.92173064e-01 -1.37006968e-01
-6.84819281e-01 3.90999973e-01 6.98266745e-01 -4.53252226e-01
1.25019059e-01 -9.63888168e-01 2.01958328e-01 6.83084071e-01
-8.88249397e-01 5.80650568e-01 1.23037720e+00 -6.46121383e-01
-1.99856961e+00 -1.19142365e+00 1.86535314e-01 -4.81569409e-01
9.94452596e-01 -5.69970766e-03 -8.36132467e-01 7.08364785e-01
-3.33505690e-01 -9.84913260e-02 -2.96355486e-01 -9.87518728e-01
3.68698895e-01 1.27473185e-02 -1.15894294e+00 8.42305005e-01
6.73193932e-01 4.03697863e-02 -1.48312554e-01 5.70418201e-02
9.47789669e-01 -5.44430554e-01 -9.53704774e-01 8.77602398e-01
5.69770455e-01 -1.72959954e-01 7.79317319e-01 -5.67534208e-01
-2.22091258e-01 -7.18234658e-01 -9.42984074e-02 -1.26090658e+00
-9.44844931e-02 -1.24967277e+00 -3.24981630e-01 4.66610610e-01
2.83682793e-01 -7.63775170e-01 5.63115597e-01 6.96937621e-01
-7.04858124e-01 -1.22170019e+00 -1.24127913e+00 -1.38437235e+00
3.67859840e-01 -3.93544406e-01 -8.85502994e-02 3.31191957e-01
7.16599584e-01 -8.80836844e-02 -2.32029274e-01 6.08165324e-01
8.02851140e-01 -4.14020240e-01 7.23285496e-01 -8.85973334e-01
1.68565571e-01 -5.68808794e-01 -5.44867575e-01 -1.01140976e+00
2.34493598e-01 -6.01210892e-01 4.77784395e-01 -1.40766811e+00
-4.04487342e-01 -5.29047489e-01 2.97983795e-01 -3.44939455e-02
1.23272076e-01 -1.01105243e-01 8.34365413e-02 1.89561456e-01
3.34437937e-02 9.93979633e-01 1.19850910e+00 -4.50168438e-02
-3.58401060e-01 3.43964010e-01 1.04024217e-01 8.73352885e-01
6.07398033e-01 3.02688628e-02 -6.90707684e-01 9.73574743e-02
-1.04498737e-01 6.17707372e-01 6.34830594e-01 -9.93550658e-01
3.56583208e-01 -4.84341413e-01 -5.47252178e-01 -7.82177329e-01
5.19719958e-01 -8.11802268e-01 5.92203140e-01 1.00907409e+00
-1.29102007e-01 1.06333815e-01 3.95814180e-01 8.41158748e-01
1.38605796e-02 -1.13151029e-01 8.68788064e-01 5.84358990e-01
-5.66766441e-01 1.92950189e-01 -7.79927194e-01 -3.85682642e-01
1.62796760e+00 -2.73331046e-01 -3.08204561e-01 -5.09422541e-01
-5.68253815e-01 6.70574248e-01 3.97777140e-01 1.63290769e-01
7.15342939e-01 -1.40608656e+00 -5.96729934e-01 4.94904779e-02
-3.10280733e-03 6.70173690e-02 1.00241147e-01 1.19161260e+00
-5.26438653e-01 9.40141380e-01 -2.48515591e-01 -7.32415915e-01
-9.53603029e-01 4.85657781e-01 3.51814777e-01 1.24061853e-01
-5.91308832e-01 5.80048203e-01 4.25608158e-01 -1.56020850e-01
1.22286426e-02 -8.58765006e-01 3.41403522e-02 -4.47123080e-01
5.12784384e-02 9.83989179e-01 -2.35486165e-01 -1.03572261e+00
-4.37897742e-01 7.37867832e-01 3.98138374e-01 -2.66618371e-01
8.28984737e-01 -2.99310356e-01 -7.25817382e-02 2.67799288e-01
1.13070929e+00 -3.19579899e-01 -1.68420362e+00 4.86203998e-01
-1.82959706e-01 1.04236931e-01 -3.63692760e-01 -1.95694357e-01
-4.38811243e-01 3.91454250e-01 2.09217474e-01 2.91868865e-01
5.82415581e-01 -9.62498188e-02 7.61751413e-01 4.18460011e-01
7.69199848e-01 -8.07022274e-01 -3.21654707e-01 9.26977098e-01
1.38998830e+00 -6.66276872e-01 -1.28989249e-01 -7.13859260e-01
-4.63746548e-01 1.22507894e+00 6.73058391e-01 -4.70872700e-01
7.89751351e-01 3.24847221e-01 -3.51240218e-01 3.50233912e-01
-6.71467662e-01 1.75855383e-01 5.06680965e-01 5.35456359e-01
-1.85447752e-01 1.84523940e-01 -7.95747399e-01 7.43948221e-01
-8.33631530e-02 -2.90416569e-01 8.22942019e-01 8.33722651e-01
-7.88424194e-01 -5.08328378e-01 -5.66504002e-01 6.08738475e-02
3.33720893e-01 4.66900885e-01 7.60520622e-02 1.15739775e+00
-4.56686288e-01 9.35035229e-01 4.79204357e-02 -5.48297107e-01
6.90461159e-01 -3.67525816e-01 3.38576376e-01 -1.21032089e-01
1.58299580e-01 -2.89099753e-01 4.84208278e-02 -6.94199264e-01
3.37557495e-01 -6.36107683e-01 -2.01049638e+00 -4.09539551e-01
-5.74929953e-01 3.33183408e-01 7.69883811e-01 8.89746070e-01
2.72589862e-01 7.54969940e-02 5.69304883e-01 -1.20397019e+00
-1.31833160e+00 -6.33190036e-01 -4.06190217e-01 -1.80085942e-01
7.89434850e-01 -1.06088543e+00 -4.87764329e-01 -2.05179110e-01] | [5.140517711639404, 2.1776010990142822] |
4c4fdb7e-1f7b-44a2-ac56-c0c4f75930e9 | on-the-global-convergence-of-risk-averse | 2301.10932 | null | https://arxiv.org/abs/2301.10932v2 | https://arxiv.org/pdf/2301.10932v2.pdf | On the Global Convergence of Risk-Averse Policy Gradient Methods with Expected Conditional Risk Measures | Risk-sensitive reinforcement learning (RL) has become a popular tool to control the risk of uncertain outcomes and ensure reliable performance in various sequential decision-making problems. While policy gradient methods have been developed for risk-sensitive RL, it remains unclear if these methods enjoy the same global convergence guarantees as in the risk-neutral case. In this paper, we consider a class of dynamic time-consistent risk measures, called Expected Conditional Risk Measures (ECRMs), and derive policy gradient updates for ECRM-based objective functions. Under both constrained direct parameterization and unconstrained softmax parameterization, we provide global convergence and iteration complexities of the corresponding risk-averse policy gradient algorithms. We further test risk-averse variants of REINFORCE and actor-critic algorithms to demonstrate the efficacy of our method and the importance of risk control. | ['Lei Ying', 'Xian Yu'] | 2023-01-26 | null | null | null | null | ['policy-gradient-methods'] | ['methodology'] | [ 1.25352162e-04 2.75936127e-01 -4.67362911e-01 -2.59092033e-01
-1.20716882e+00 -4.52192843e-01 4.38188016e-01 2.67170668e-01
-8.91505957e-01 1.30232751e+00 2.00131044e-01 -5.92792869e-01
-5.78993797e-01 -6.37493968e-01 -4.34909672e-01 -7.69442379e-01
-5.87728500e-01 1.83240488e-01 -7.95981511e-02 -1.03152588e-01
3.97015214e-01 3.68504614e-01 -1.10694385e+00 -6.40770078e-01
1.20324087e+00 1.14319348e+00 -1.89510733e-01 6.27011240e-01
4.72331256e-01 9.04348373e-01 -5.83057284e-01 -4.34536129e-01
3.35719496e-01 -4.33191538e-01 -4.56507117e-01 -4.03532088e-01
-2.02717379e-01 -6.52810037e-01 2.15099156e-01 1.19527614e+00
6.95011377e-01 5.49914479e-01 6.33538306e-01 -1.28955758e+00
-1.31771713e-01 6.52836919e-01 -7.37221956e-01 2.68056840e-01
1.62649602e-01 1.66343823e-01 9.49338794e-01 -3.82869631e-01
2.46907845e-01 1.43348181e+00 4.02086318e-01 7.68072307e-01
-1.04734862e+00 -6.59726322e-01 5.57355702e-01 3.51076312e-02
-7.74022102e-01 -3.94108780e-02 3.49700302e-01 -2.12758556e-01
6.42755151e-01 -1.99697874e-02 6.63029492e-01 1.03384066e+00
6.77846014e-01 8.34864676e-01 1.44506788e+00 -3.50550860e-01
8.21872413e-01 -4.96566743e-02 -2.65947044e-01 4.88572180e-01
4.42084074e-01 8.75337541e-01 -2.46093795e-01 -4.89748627e-01
5.24191141e-01 -3.17084521e-01 5.40676005e-02 -4.61501122e-01
-7.34983087e-01 1.13376641e+00 1.09434769e-01 -3.18709135e-01
-6.84578300e-01 5.55065513e-01 3.19362402e-01 4.58882123e-01
5.32270432e-01 5.80041647e-01 -2.58881956e-01 -3.00134510e-01
-4.44159180e-01 6.37420535e-01 7.50176311e-01 5.38731277e-01
8.63178894e-02 5.30892730e-01 -6.15370154e-01 4.59207386e-01
4.36750174e-01 5.81788421e-01 1.94488242e-01 -1.29458141e+00
6.52823508e-01 -6.18917979e-02 8.48421097e-01 -6.69301689e-01
-4.09119010e-01 -4.93933529e-01 -3.02300751e-01 8.82926822e-01
5.02456784e-01 -7.51531601e-01 -5.39744258e-01 2.08938241e+00
4.61974174e-01 -1.52003884e-01 9.63451341e-02 6.93303049e-01
-5.46821415e-01 5.04855990e-01 4.03085172e-01 -8.94312561e-01
6.95433676e-01 -4.04047251e-01 -8.97278130e-01 -1.31414339e-01
2.80163556e-01 -2.09380165e-01 1.08077419e+00 3.98191988e-01
-1.37550962e+00 2.74805427e-01 -1.16620994e+00 6.06393456e-01
1.96157455e-01 -6.55194402e-01 4.55099255e-01 8.67391825e-01
-7.15600550e-01 9.06869650e-01 -8.92450929e-01 1.70160055e-01
4.45515841e-01 1.35339037e-01 4.47723031e-01 3.84640753e-01
-1.36920023e+00 1.17398453e+00 1.72875285e-01 -2.71911044e-02
-1.64395988e+00 -7.60016203e-01 -5.44116855e-01 4.15326394e-02
9.94098902e-01 -2.89621115e-01 1.69252956e+00 -6.18941784e-01
-2.11258554e+00 1.89662978e-01 5.07125020e-01 -6.94653392e-01
1.14640510e+00 -3.89020622e-01 -2.53759399e-02 8.99099410e-02
-3.64992097e-02 2.92450096e-02 1.01774418e+00 -9.90540266e-01
-7.81018198e-01 -1.45012230e-01 1.44406036e-01 7.12840676e-01
-2.25702077e-01 2.68526375e-01 5.51943004e-01 -6.88296974e-01
-6.10445738e-01 -8.71135473e-01 -7.32773602e-01 7.24261776e-02
-2.34971285e-01 -1.37029439e-02 1.04700752e-01 -5.53769886e-01
1.17193198e+00 -1.83297956e+00 -3.62419933e-02 4.68184590e-01
-3.92372310e-01 6.13500178e-02 3.49494033e-02 2.62731075e-01
2.33647913e-01 2.23926038e-01 -4.84036088e-01 -2.97690276e-02
3.44979227e-01 4.99976687e-02 -5.47305167e-01 5.50830364e-01
1.15085408e-01 5.57111263e-01 -1.06563795e+00 -2.91082501e-01
1.06897756e-01 3.32078934e-02 -4.60273683e-01 3.76428902e-01
-3.14944476e-01 2.41589919e-01 -8.47500920e-01 4.56335962e-01
1.33538529e-01 3.17614198e-01 2.67968893e-01 5.42427361e-01
-1.41474277e-01 -6.88481256e-02 -1.11585867e+00 8.70936871e-01
-6.16238356e-01 -8.02244693e-02 3.56810123e-01 -8.94034624e-01
6.00914180e-01 2.62233168e-01 4.34403241e-01 -6.33461714e-01
1.55044124e-01 9.03586224e-02 -3.43149066e-01 -1.06297955e-01
4.59424585e-01 -5.97794056e-01 -2.41117448e-01 7.31898546e-01
-4.81934875e-01 -3.36096257e-01 -5.47470599e-02 -3.70369069e-02
8.83231819e-01 2.71508604e-01 5.51900029e-01 -5.46807885e-01
9.21896622e-02 -3.69548172e-01 9.17760074e-01 9.59784091e-01
-7.16557145e-01 8.22518468e-02 9.72792029e-01 -7.12299049e-02
-6.97531879e-01 -1.15817511e+00 -6.76474050e-02 1.08795059e+00
4.77898195e-02 1.54465869e-01 -4.25704598e-01 -9.20614064e-01
5.26357651e-01 1.25219047e+00 -7.21519470e-01 -5.08192718e-01
-3.82693976e-01 -1.11367691e+00 2.71333873e-01 4.98327702e-01
2.70390004e-01 -8.06433618e-01 -1.03066885e+00 4.04170215e-01
3.66671026e-01 -4.51597691e-01 -5.74511766e-01 1.90872118e-01
-7.38906562e-01 -1.00574744e+00 -7.72057593e-01 9.35213566e-02
4.67406034e-01 -3.67768109e-01 8.32605958e-01 -5.84684491e-01
2.29136162e-02 7.69631326e-01 1.53232897e-02 -7.85518348e-01
-4.44966048e-01 -4.76428211e-01 3.04234713e-01 -6.58949688e-02
-5.02259314e-01 -2.92689592e-01 -8.29401195e-01 4.22279179e-01
-6.87055409e-01 -5.13472140e-01 8.56361166e-02 7.76993454e-01
6.99064076e-01 -4.60685380e-02 1.07068384e+00 -7.03372657e-01
1.38416338e+00 -5.46057343e-01 -1.25259125e+00 5.26682019e-01
-1.35305476e+00 3.84505004e-01 5.84997714e-01 -6.03940129e-01
-1.45766544e+00 -3.10389489e-01 2.26458535e-01 -9.18862075e-02
6.94874525e-01 5.38334250e-01 7.00414404e-02 -3.90204005e-02
5.66854835e-01 -2.72522449e-01 7.23724440e-02 -5.91376536e-02
3.79847318e-01 2.23053455e-01 8.79555121e-02 -1.05073845e+00
5.43983221e-01 1.79459453e-01 2.27750748e-01 -3.91700342e-02
-7.79084802e-01 3.81178975e-01 4.33101594e-01 -4.86928463e-01
6.32396460e-01 -6.93756461e-01 -1.13528240e+00 2.84605086e-01
-3.84235442e-01 -7.08661735e-01 -7.41827846e-01 6.14002347e-01
-1.22420967e+00 2.11414978e-01 -4.82480347e-01 -1.49312305e+00
-3.27249944e-01 -9.05852556e-01 3.23431730e-01 3.56273204e-01
1.10322036e-01 -1.06759465e+00 3.10203493e-01 -2.59830087e-01
5.19466698e-01 5.92957437e-01 7.85540104e-01 -2.91956991e-01
-1.80214703e-01 -7.21873641e-02 2.53767639e-01 3.68536741e-01
-1.25376299e-01 -3.32591683e-02 -4.51545984e-01 -6.83734417e-01
1.78035572e-01 -7.58749783e-01 6.65902078e-01 7.14390278e-01
9.79209602e-01 -7.53578305e-01 8.54814872e-02 3.74752522e-01
1.39606357e+00 7.20086873e-01 2.43327677e-01 6.37694597e-01
-1.29517016e-03 7.32851803e-01 1.08978128e+00 1.05387568e+00
1.93952560e-01 4.52943653e-01 6.74124599e-01 3.60355288e-01
7.78518319e-01 -4.65094328e-01 7.37072706e-01 5.71138822e-02
9.05997772e-03 -1.69601262e-01 -6.70166194e-01 2.87074119e-01
-2.07879686e+00 -9.56157982e-01 7.23156691e-01 2.63972521e+00
1.19252217e+00 2.48328388e-01 4.09361839e-01 -1.69244394e-01
8.33177328e-01 2.52759513e-02 -1.26405489e+00 -7.32024610e-01
4.37540449e-02 -2.68439837e-02 8.56971085e-01 8.69770467e-01
-8.49862516e-01 6.18253350e-01 7.45046902e+00 8.38197827e-01
-7.16065288e-01 -7.92986602e-02 1.07032752e+00 -5.61228156e-01
-4.15096521e-01 -7.49630406e-02 -6.69773877e-01 4.09340471e-01
1.07152665e+00 -8.38404357e-01 5.50914586e-01 1.09461427e+00
5.63612580e-01 -3.27151924e-01 -8.37444186e-01 3.75012100e-01
-6.33873701e-01 -7.96381950e-01 -5.88478208e-01 -6.92460760e-02
9.17964876e-01 -4.29711223e-01 3.85976315e-01 4.75501776e-01
1.11405802e+00 -9.10565495e-01 9.72310066e-01 5.62280834e-01
6.26637101e-01 -1.38337612e+00 4.82718855e-01 1.80370715e-02
-6.28420889e-01 -7.25074768e-01 -2.58355618e-01 1.29560963e-03
3.33527714e-01 6.00476027e-01 -4.97996092e-01 1.49187312e-01
4.49104965e-01 2.35169739e-01 -1.37095782e-03 8.79575908e-01
-5.61420560e-01 6.47824645e-01 -3.35252523e-01 -3.12025130e-01
4.11219180e-01 -2.47946084e-01 6.66131616e-01 6.03788972e-01
2.40254372e-01 -1.00650214e-01 1.09804168e-01 7.15436459e-01
2.32654184e-01 5.01966029e-02 -2.71785945e-01 -9.99495983e-02
5.94794035e-01 9.94419336e-01 -5.61275005e-01 6.65084124e-02
9.77873057e-02 3.26631308e-01 4.56108361e-01 4.31829810e-01
-7.68270433e-01 -3.31242800e-01 9.10276473e-01 -3.31978381e-01
7.58622810e-02 -1.64388746e-01 -1.01136126e-01 -8.76255810e-01
-3.87262553e-02 -7.82137036e-01 8.65916491e-01 -2.66690344e-01
-1.45362186e+00 1.60731487e-02 2.72081226e-01 -9.73632812e-01
-7.10892737e-01 -3.68579924e-01 -7.16681957e-01 6.55779064e-01
-1.61324537e+00 -2.59162068e-01 5.93483150e-01 5.96726239e-01
5.04054092e-02 1.65604930e-02 4.71282452e-01 -2.92970002e-01
-8.39204371e-01 7.00527668e-01 7.78392673e-01 -5.56657910e-01
6.14465535e-01 -1.47134781e+00 -6.45850077e-02 7.99618840e-01
-7.57847428e-01 3.07074815e-01 9.11861479e-01 -7.61067450e-01
-1.33990049e+00 -9.41518962e-01 3.62855829e-02 -9.84344855e-02
7.75142252e-01 3.01629603e-02 -4.15583879e-01 3.54205519e-01
-1.42897934e-01 -2.03983471e-01 1.41125679e-01 -8.01039338e-02
-9.33963805e-02 -2.94633567e-01 -1.56760263e+00 8.93615723e-01
8.07745576e-01 -1.47858605e-01 -3.84412408e-01 2.53784031e-01
7.18140781e-01 -2.51831979e-01 -8.54262948e-01 3.64116997e-01
4.78598505e-01 -8.23897600e-01 8.60774755e-01 -8.76990557e-01
1.30607625e-02 1.19054310e-01 -1.72878727e-02 -1.62771142e+00
-7.17359111e-02 -1.24212515e+00 -2.46681765e-01 9.03830588e-01
3.31613302e-01 -9.66795981e-01 5.79715669e-01 1.01149690e+00
1.05116554e-01 -9.65923548e-01 -1.37232137e+00 -1.16352940e+00
5.37061036e-01 -4.31343526e-01 3.41936558e-01 6.59081221e-01
4.01441962e-01 -3.56135637e-01 -6.32249773e-01 -5.58779724e-02
1.08064127e+00 -1.22687764e-01 3.39283422e-02 -5.43981135e-01
-4.21181172e-01 -6.77912772e-01 2.78962225e-01 -2.69300491e-01
3.40322793e-01 -3.26674223e-01 2.39281505e-01 -1.24239850e+00
-4.56748158e-02 -4.47440416e-01 -7.43594706e-01 4.96888131e-01
-6.61267459e-01 -5.89075327e-01 2.73422182e-01 -1.89303398e-01
-6.98394299e-01 1.12560129e+00 1.13215327e+00 6.01381622e-02
-4.87167001e-01 4.47145849e-01 -7.54297197e-01 4.98715967e-01
1.08555222e+00 -6.70711875e-01 -8.44787598e-01 1.44255057e-01
6.20107710e-01 5.85248649e-01 -1.10016055e-02 -5.96405745e-01
-1.94139823e-01 -9.76623535e-01 1.48033639e-02 -9.51088592e-02
-1.08990476e-01 -4.81564403e-01 -1.24890558e-01 9.47182715e-01
-9.69578207e-01 3.07187080e-01 1.49666807e-02 1.02639651e+00
1.86474979e-01 -2.69505054e-01 1.06622684e+00 -7.79393688e-02
-6.48112148e-02 3.70546490e-01 -8.29703152e-01 6.12545133e-01
1.15724659e+00 2.69424289e-01 -1.14536852e-01 -9.33676481e-01
-4.02678698e-01 7.28189886e-01 1.18811972e-01 1.18968874e-01
7.14756072e-01 -1.10486555e+00 -6.22028947e-01 -4.20524865e-01
-2.20205382e-01 -3.70554537e-01 3.81312892e-03 5.44833243e-01
-6.00366369e-02 6.78723380e-02 -1.63900688e-01 1.89278454e-01
-5.94824374e-01 4.91368175e-01 7.33950853e-01 -5.90900123e-01
-2.63718396e-01 6.69670582e-01 5.11809960e-02 1.89337917e-02
5.79009593e-01 -1.64160520e-01 7.96355382e-02 1.19952686e-01
7.52330422e-01 8.83886158e-01 -3.26217502e-01 3.29759449e-01
-1.84231594e-01 2.13085294e-01 3.80049311e-02 -8.70227039e-01
1.18922627e+00 -2.63578713e-01 4.44904268e-01 3.19621712e-01
5.76219380e-01 -2.82655329e-01 -2.10342503e+00 3.30878124e-02
4.28365320e-01 -3.85679901e-01 2.00315103e-01 -9.75021839e-01
-8.34711194e-01 5.91216981e-01 7.20140278e-01 5.98810390e-02
9.45130587e-01 -7.26405978e-01 2.22975671e-01 3.41309428e-01
5.38093448e-01 -1.61242568e+00 4.59923968e-02 3.03654075e-01
9.84043181e-01 -9.73166764e-01 2.96634257e-01 2.41374791e-01
-1.13542640e+00 7.03325689e-01 5.01813352e-01 -2.08343461e-01
5.99896789e-01 1.64784834e-01 -2.06240104e-03 4.74581510e-01
-9.94649291e-01 -6.42286465e-02 -1.97444737e-01 6.30403817e-01
-1.99156404e-01 2.30865315e-01 -6.73678875e-01 4.08534855e-01
1.74425036e-01 -4.66521047e-02 6.35854959e-01 1.28158164e+00
-3.65876615e-01 -9.58828926e-01 -2.57358879e-01 4.04254854e-01
-7.87539542e-01 1.25276238e-01 6.94581345e-02 6.79139078e-01
-7.90766358e-01 1.03870821e+00 -2.54722834e-01 2.56334040e-02
2.45178133e-01 -2.21772175e-02 4.98038143e-01 -8.70374888e-02
-5.85088074e-01 -8.49507377e-02 2.91094273e-01 -8.38985384e-01
-8.90796557e-02 -7.61822224e-01 -1.23487115e+00 -2.28950247e-01
-1.42755136e-01 2.46348307e-01 6.78667903e-01 8.10628295e-01
1.79094642e-01 1.69599861e-01 1.12956178e+00 -4.83805865e-01
-1.79407716e+00 -5.30606568e-01 -7.88430631e-01 1.33260399e-01
3.91861022e-01 -8.31132472e-01 -7.54205763e-01 -8.22808623e-01] | [4.244218826293945, 2.5323643684387207] |
56be8afb-5cdc-42f7-85d3-b0c4565dce01 | open-problems-in-applied-deep-learning | 2301.11316 | null | https://arxiv.org/abs/2301.11316v1 | https://arxiv.org/pdf/2301.11316v1.pdf | Open Problems in Applied Deep Learning | This work formulates the machine learning mechanism as a bi-level optimization problem. The inner level optimization loop entails minimizing a properly chosen loss function evaluated on the training data. This is nothing but the well-studied training process in pursuit of optimal model parameters. The outer level optimization loop is less well-studied and involves maximizing a properly chosen performance metric evaluated on the validation data. This is what we call the "iteration process", pursuing optimal model hyper-parameters. Among many other degrees of freedom, this process entails model engineering (e.g., neural network architecture design) and management, experiment tracking, dataset versioning and augmentation. The iteration process could be automated via Automatic Machine Learning (AutoML) or left to the intuitions of machine learning students, engineers, and researchers. Regardless of the route we take, there is a need to reduce the computational cost of the iteration step and as a direct consequence reduce the carbon footprint of developing artificial intelligence algorithms. Despite the clean and unified mathematical formulation of the iteration step as a bi-level optimization problem, its solutions are case specific and complex. This work will consider such cases while increasing the level of complexity from supervised learning to semi-supervised, self-supervised, unsupervised, few-shot, federated, reinforcement, and physics-informed learning. As a consequence of this exercise, this proposal surfaces a plethora of open problems in the field, many of which can be addressed in parallel. | ['Maziar Raissi'] | 2023-01-26 | null | null | null | null | ['automl'] | ['methodology'] | [ 4.01752681e-01 2.05334529e-01 -2.43006662e-01 -1.60902098e-01
-2.65789330e-01 -5.84809482e-01 5.54385900e-01 3.51797491e-01
-5.33926249e-01 7.09767580e-01 -4.80104357e-01 -4.21531230e-01
-6.01047516e-01 -7.59469211e-01 -7.41243660e-01 -8.59773397e-01
1.56306416e-01 7.00476527e-01 -2.05717623e-01 -9.70077738e-02
6.61001444e-01 7.53869712e-01 -1.58158565e+00 -3.72541457e-01
8.63521338e-01 9.42175031e-01 9.61174890e-02 5.52293956e-01
-1.94192305e-01 4.60050285e-01 -1.03848197e-01 -2.11926132e-01
2.25580037e-01 -4.84555960e-01 -9.99136806e-01 4.46868569e-01
-1.90458402e-01 4.40926164e-01 4.94339615e-01 1.08793426e+00
3.81096095e-01 7.85256922e-02 6.27648592e-01 -1.28042984e+00
-1.01100795e-01 2.58526504e-01 -2.03537300e-01 -1.84513673e-01
-6.82438388e-02 3.87346268e-01 7.52470851e-01 -6.07920527e-01
3.77547950e-01 8.48795950e-01 4.70230311e-01 1.69568211e-01
-1.45114350e+00 -2.63785154e-01 -1.04331784e-01 1.97494507e-01
-1.12130296e+00 -1.33375227e-01 7.78931737e-01 -8.18874180e-01
6.33240759e-01 2.64473617e-01 7.70004392e-01 6.62389874e-01
1.05821326e-01 3.21055681e-01 1.18141580e+00 -8.78575027e-01
7.62794256e-01 7.48519778e-01 3.71866763e-01 7.87078202e-01
3.67599756e-01 4.94246602e-01 -2.41711318e-01 8.60766247e-02
4.51743096e-01 -2.01727539e-01 7.68973827e-02 -7.51949012e-01
-7.52858400e-01 9.05431867e-01 8.54293779e-02 4.53410685e-01
-5.65292120e-01 -1.96579397e-01 1.80491030e-01 3.95869434e-01
1.82517380e-01 1.15989232e+00 -7.17474103e-01 -4.70799990e-02
-9.64024663e-01 1.30092680e-01 1.02248263e+00 3.72165143e-01
1.21368551e+00 -6.64606392e-02 1.71927735e-01 5.15800714e-01
4.12922680e-01 -3.84151042e-02 3.07861358e-01 -9.56531286e-01
-5.11271432e-02 1.02608621e+00 2.66677797e-01 -7.24181175e-01
-5.55865586e-01 -8.30696821e-01 -7.45597124e-01 7.07382441e-01
4.91590887e-01 -3.37227017e-01 -5.45443296e-01 1.49914241e+00
5.45659959e-01 -7.80626312e-02 -4.78824787e-02 6.93905473e-01
1.96130320e-01 6.17227554e-01 2.08117262e-01 -5.09842336e-01
1.25792158e+00 -1.10782182e+00 -5.50350726e-01 -1.53848067e-01
6.15866482e-01 -6.06264591e-01 1.13971758e+00 7.69193709e-01
-1.08571720e+00 -4.98540342e-01 -1.05325449e+00 2.80566007e-01
-8.25649321e-01 1.62207022e-01 6.75107777e-01 8.30602586e-01
-6.36387229e-01 9.96816337e-01 -6.47415996e-01 -3.99572998e-01
7.10215569e-02 5.32164156e-01 -1.58045501e-01 1.52033478e-01
-9.76052463e-01 1.19687188e+00 6.72627628e-01 2.38108724e-01
-6.06016874e-01 -7.74318635e-01 -5.82821250e-01 1.15711451e-01
7.22718239e-01 -7.04383910e-01 9.40017581e-01 -1.10839403e+00
-1.59316134e+00 9.41321790e-01 9.17089954e-02 -4.69411552e-01
5.96873164e-01 -1.33503765e-01 -1.16972543e-01 -3.49417061e-01
-3.38977844e-01 2.56732732e-01 7.35823452e-01 -1.33914042e+00
-3.27942580e-01 -4.44741398e-01 -8.38992279e-03 7.74901658e-02
-2.52810270e-01 -6.61542118e-02 -9.56322178e-02 -3.36924762e-01
-9.92083773e-02 -9.92199004e-01 -5.32534063e-01 -7.69528300e-02
-2.40672797e-01 -8.07153061e-02 6.41223550e-01 -4.32574242e-01
1.38115156e+00 -1.83853531e+00 3.36893380e-01 2.91468531e-01
-1.38063967e-01 4.18510318e-01 1.58901006e-01 5.93053758e-01
-2.15499118e-01 2.39457324e-01 -4.79474843e-01 3.03606912e-02
1.77067108e-02 -6.99096024e-02 1.53006658e-01 4.34060216e-01
3.55801761e-01 7.28971839e-01 -7.79892445e-01 -3.26012552e-01
4.58012670e-01 2.39547417e-01 -5.07059395e-01 3.41902316e-01
-5.48881650e-01 5.59062719e-01 -5.07146955e-01 3.71680379e-01
2.89936572e-01 -4.75920469e-01 1.80484906e-01 -3.16805035e-01
-4.18283015e-01 -5.19943796e-02 -1.43146670e+00 1.49703276e+00
-4.77422416e-01 2.87412554e-01 3.37264198e-03 -1.74154532e+00
1.02251995e+00 7.80297443e-02 7.47350931e-01 -4.77372408e-01
2.03768611e-01 2.06432492e-01 -2.73842085e-02 -7.19906330e-01
3.19685847e-01 -2.43331432e-01 1.22717008e-01 4.42516059e-01
1.40843853e-01 -1.47483811e-01 3.05416644e-01 -3.61523569e-01
8.00978661e-01 5.10103166e-01 5.98754227e-01 -3.07919562e-01
7.50337660e-01 4.80833858e-01 3.80317837e-01 5.47043800e-01
-8.15962479e-02 9.55406502e-02 5.15447199e-01 -4.21876580e-01
-1.09492052e+00 -6.77502811e-01 -1.32751584e-01 8.14754486e-01
3.99803482e-02 -1.82160214e-01 -9.18144763e-01 -4.96252596e-01
-1.34034008e-01 7.13237405e-01 -4.75214869e-01 -3.16694796e-01
-4.39847648e-01 -9.74380791e-01 -7.82363042e-02 -1.57108366e-01
3.44755292e-01 -1.03490269e+00 -7.94656575e-01 3.87824148e-01
3.25323820e-01 -7.61949539e-01 1.70386463e-01 6.51154816e-01
-1.11711955e+00 -1.20006573e+00 -3.12865376e-01 -4.55804229e-01
5.43343663e-01 -3.76480788e-01 1.22548079e+00 4.41798680e-02
-5.38734853e-01 2.78527200e-01 -3.55235934e-01 -6.79233074e-01
-4.61880416e-01 1.33014396e-01 -1.43371418e-01 9.54724923e-02
1.75397873e-01 -5.53656220e-01 -4.70082223e-01 2.02499121e-01
-8.30604136e-01 -1.03412429e-04 7.90585995e-01 8.12798321e-01
8.38230789e-01 3.96524638e-01 5.24103165e-01 -8.79187822e-01
4.76788104e-01 -3.42505664e-01 -9.44565535e-01 5.54145277e-01
-1.24404013e+00 3.81640404e-01 7.11785734e-01 -3.75250548e-01
-7.95394659e-01 3.47716987e-01 1.15642115e-01 -1.55332506e-01
-4.04417098e-01 5.45195162e-01 -3.52121234e-01 -1.75714538e-01
6.14261866e-01 1.91601068e-01 3.33938301e-01 -6.09909594e-01
3.21168035e-01 3.32555562e-01 1.53503552e-01 -5.97216785e-01
9.41373825e-01 6.75470531e-02 3.81710231e-01 -8.85018051e-01
-8.33816767e-01 -1.65743068e-01 -7.21495748e-01 -4.98467535e-01
9.28952157e-01 -3.20647746e-01 -7.23311663e-01 3.39146614e-01
-7.78192759e-01 -3.60317767e-01 -6.03165865e-01 4.02648807e-01
-6.58176601e-01 8.54747295e-02 7.83255976e-03 -1.02577853e+00
-2.80817747e-01 -1.24367166e+00 5.97966015e-01 4.91143942e-01
-1.70235753e-01 -1.16909230e+00 2.60987207e-02 5.46370924e-01
3.65323991e-01 4.24383491e-01 1.17926431e+00 -6.81355536e-01
-5.88625133e-01 -3.35595429e-01 1.57582283e-01 5.60333014e-01
3.67844626e-02 1.61153764e-01 -7.17888951e-01 -1.60023570e-01
1.49892807e-01 -4.35166508e-01 4.55756515e-01 3.79379392e-01
1.32297230e+00 -2.32975110e-01 -4.38536704e-03 4.41138238e-01
1.75254476e+00 3.86158139e-01 4.49485362e-01 5.12921572e-01
2.04916686e-01 8.68254960e-01 7.13607550e-01 2.85306007e-01
2.46422011e-02 6.29995286e-01 5.85940003e-01 -1.57977402e-01
2.53643364e-01 -3.53822969e-02 7.95094892e-02 5.73261380e-01
-1.48956537e-01 8.37391466e-02 -1.00625038e+00 8.84509906e-02
-1.73479867e+00 -9.41679657e-01 -8.76037031e-02 2.27774906e+00
4.76879448e-01 2.68163562e-01 2.58968800e-01 4.88687903e-01
6.36039019e-01 -2.25152582e-01 -6.84698761e-01 -7.00681090e-01
1.02917217e-01 2.21240684e-01 3.31703037e-01 3.69308800e-01
-9.32534337e-01 6.35601282e-01 5.81227827e+00 8.37378919e-01
-1.24555659e+00 -1.77557379e-01 8.57439458e-01 2.24928901e-01
-4.08673286e-02 2.28812173e-01 -6.22248113e-01 4.62026089e-01
9.31473076e-01 -1.51848555e-01 9.14070010e-01 8.03562880e-01
5.85665047e-01 -4.07588273e-01 -1.13126671e+00 7.16028214e-01
-3.10896188e-01 -1.43747926e+00 -3.33680779e-01 2.46321455e-01
6.34646475e-01 -1.84395283e-01 -2.70593822e-01 3.30772251e-01
2.45620310e-02 -1.11387932e+00 6.04150832e-01 7.84698486e-01
2.45072693e-01 -6.23719692e-01 4.19562370e-01 7.03268647e-01
-8.05490255e-01 -2.76707023e-01 -1.51015795e-03 -1.57317519e-01
2.18088794e-02 9.28180039e-01 -4.12102044e-01 7.23969579e-01
2.95331150e-01 2.14078426e-01 -6.19426966e-01 1.26536751e+00
4.53459062e-02 6.32768214e-01 -2.57932752e-01 -2.49717996e-01
2.67152131e-01 -7.83398688e-01 4.10927266e-01 1.04605138e+00
9.85048115e-02 -1.03610918e-01 1.37485430e-01 1.07163906e+00
2.60169357e-01 2.10273579e-01 -4.19980168e-01 -2.80597091e-01
2.11905077e-01 1.39680219e+00 -7.10578918e-01 -4.77179773e-02
-1.02210239e-01 5.06629407e-01 2.63869256e-01 2.10745037e-01
-5.98123848e-01 -2.02671528e-01 2.62463510e-01 1.59145042e-01
-1.19058527e-01 -3.87904160e-02 -6.74720109e-01 -8.32339883e-01
-8.77904668e-02 -9.03587580e-01 2.29390413e-01 -5.32693028e-01
-1.06098366e+00 1.96446061e-01 2.76056454e-02 -9.42801297e-01
-2.30817556e-01 -7.58013487e-01 -7.85200119e-01 7.35934556e-01
-1.44558227e+00 -6.40719473e-01 -2.32921764e-01 1.96683675e-01
2.89916039e-01 -1.91393375e-01 7.87809908e-01 2.46891916e-01
-9.56770599e-01 1.54886782e-01 1.47212669e-01 -4.08458382e-01
1.48435324e-01 -1.27213454e+00 -4.16707367e-01 5.71555257e-01
-1.11428417e-01 2.86133140e-01 9.31666434e-01 -2.95274496e-01
-1.65016043e+00 -1.02462935e+00 8.89730394e-01 6.58376236e-03
5.80713093e-01 -9.90626067e-02 -8.28037381e-01 -4.86709848e-02
9.94434085e-05 -2.45057091e-01 6.37124598e-01 4.54338901e-02
1.77022651e-01 -2.48783141e-01 -1.21213341e+00 3.34368497e-01
5.68510592e-01 -2.65963405e-01 -3.52861524e-01 5.18091619e-01
3.56141508e-01 7.58553073e-02 -1.17269051e+00 5.06200373e-01
4.04881209e-01 -7.84239471e-01 8.93406689e-01 -8.72704089e-01
3.79049629e-01 -2.08096072e-01 1.56947181e-01 -1.17462802e+00
-4.24648784e-02 -7.88250864e-01 -9.05986950e-02 1.23559415e+00
7.13254750e-01 -4.51545984e-01 8.97965312e-01 6.23754501e-01
1.11927815e-01 -1.27819109e+00 -5.61092854e-01 -8.94544005e-01
1.84556887e-01 -3.94106686e-01 3.21145952e-01 9.61991012e-01
-1.17487401e-01 5.09324193e-01 -1.34208202e-01 -4.40688385e-03
8.23358715e-01 2.14329630e-01 7.38865376e-01 -1.49391174e+00
-4.57106084e-01 -6.81147218e-01 -1.34742841e-01 -3.14059585e-01
-1.19103402e-01 -7.40719497e-01 -1.20057851e-01 -1.33128798e+00
-4.16585244e-02 -4.43892747e-01 -1.58393696e-01 1.00420661e-01
1.25331618e-02 -3.60118806e-01 1.20263949e-01 8.43718946e-02
-3.99719536e-01 3.23647261e-01 1.17311227e+00 1.35806367e-01
-3.44634533e-01 1.75603181e-01 -5.22185743e-01 7.86425412e-01
1.13019633e+00 -4.19416249e-01 -3.54053140e-01 3.94827612e-02
4.27090585e-01 1.08132452e-01 4.83039230e-01 -9.73464668e-01
1.93334371e-01 -5.74325025e-01 7.72108734e-02 -2.45267272e-01
1.06287397e-01 -9.83823478e-01 4.37445194e-01 7.21423030e-01
-3.69784355e-01 -1.16307311e-01 -1.09263785e-01 3.18360150e-01
-1.05372585e-01 -7.13373303e-01 1.10260010e+00 -3.61538261e-01
-4.73342359e-01 2.14058205e-01 -2.77697444e-01 -9.99093279e-02
1.27305591e+00 -4.28529710e-01 3.84740606e-02 8.98761377e-02
-1.01833057e+00 2.71083146e-01 3.47968102e-01 9.86916125e-02
-1.70838125e-02 -7.49656439e-01 -3.77862751e-01 -5.15436977e-02
-1.19442888e-01 -5.18855313e-03 1.15935346e-02 8.60140741e-01
-2.94420332e-01 3.85060817e-01 -1.60634130e-01 -3.77877355e-01
-9.35918748e-01 6.48642540e-01 5.37248433e-01 -8.13933432e-01
-7.86053240e-02 3.87832463e-01 -3.77474934e-01 -5.71397543e-01
3.31056416e-01 1.63971424e-01 -2.49683723e-01 2.27574602e-01
5.47505654e-02 6.92385197e-01 2.33899698e-01 -9.98685509e-02
-1.48553327e-01 5.66301405e-01 3.87495786e-01 7.07372800e-02
1.51598406e+00 6.87180972e-03 -1.99893311e-01 4.75515813e-01
1.11006761e+00 -5.69759130e-01 -1.02186525e+00 -6.04024716e-02
4.87761468e-01 9.47963744e-02 2.91357547e-01 -1.06712878e+00
-8.35603833e-01 8.14520001e-01 7.46629775e-01 3.83249998e-01
1.19005752e+00 -2.66140401e-01 1.19093895e-01 5.91716528e-01
2.70300001e-01 -1.58909655e+00 1.90058291e-01 3.59090120e-01
8.14187050e-01 -1.33686054e+00 7.73209780e-02 -2.00624928e-01
-3.48542035e-01 1.14658141e+00 6.97918713e-01 -1.37352690e-01
8.01208198e-01 2.55521327e-01 -3.80560726e-01 -2.53606290e-01
-6.50987327e-01 -2.51657367e-01 2.68920422e-01 3.55581164e-01
4.08128947e-01 -2.79192850e-02 -7.17442811e-01 5.63669205e-01
-8.39747488e-02 3.44217598e-01 1.70194253e-01 8.74381363e-01
-7.31207430e-01 -1.36012912e+00 -4.44619596e-01 4.78282452e-01
-2.70961344e-01 2.00084791e-01 -3.78059059e-01 7.80515015e-01
4.69938874e-01 9.67919946e-01 -2.96858579e-01 -2.00135380e-01
3.74265194e-01 3.48628491e-01 4.01078165e-01 -4.60387826e-01
-8.18739772e-01 -2.04720542e-01 1.12206653e-01 -3.10767114e-01
-4.55789387e-01 -5.63270867e-01 -9.12058353e-01 -1.97236672e-01
-5.94684482e-01 3.19296718e-01 1.20299506e+00 1.20575309e+00
1.06883392e-01 6.04719579e-01 8.37528825e-01 -8.17394614e-01
-8.28862727e-01 -6.42549574e-01 -2.93675005e-01 1.64759263e-01
2.15524901e-02 -6.20453775e-01 -2.86986738e-01 3.08992248e-02] | [6.23753547668457, 3.8093063831329346] |
c6612749-4e19-4e27-a981-55f7809cf4b9 | scene-change-detection-using-multiscale | 2212.10417 | null | https://arxiv.org/abs/2212.10417v1 | https://arxiv.org/pdf/2212.10417v1.pdf | Scene Change Detection Using Multiscale Cascade Residual Convolutional Neural Networks | Scene change detection is an image processing problem related to partitioning pixels of a digital image into foreground and background regions. Mostly, visual knowledge-based computer intelligent systems, like traffic monitoring, video surveillance, and anomaly detection, need to use change detection techniques. Amongst the most prominent detection methods, there are the learning-based ones, which besides sharing similar training and testing protocols, differ from each other in terms of their architecture design strategies. Such architecture design directly impacts on the quality of the detection results, and also in the device resources capacity, like memory. In this work, we propose a novel Multiscale Cascade Residual Convolutional Neural Network that integrates multiscale processing strategy through a Residual Processing Module, with a Segmentation Convolutional Neural Network. Experiments conducted on two different datasets support the effectiveness of the proposed approach, achieving average overall $\boldsymbol{F\text{-}measure}$ results of $\boldsymbol{0.9622}$ and $\boldsymbol{0.9664}$ over Change Detection 2014 and PetrobrasROUTES datasets respectively, besides comprising approximately eight times fewer parameters. Such obtained results place the proposed technique amongst the top four state-of-the-art scene change detection methods. | ['João P. Papa', 'Danilo Colombo', 'Rafael G. Pires', 'Daniel F. S. Santos'] | 2022-12-20 | null | null | null | null | ['scene-change-detection', 'change-detection'] | ['computer-vision', 'computer-vision'] | [ 0.5428773 -0.4535503 0.15348406 -0.2696291 -0.22159994 -0.31756383
0.41812146 0.367448 -0.6485766 0.62974524 -0.44530663 -0.4497982
-0.08358749 -0.94411564 -0.69691324 -0.69663334 0.01026187 -0.20699242
0.9229331 -0.09023795 0.4998962 0.5511075 -1.8537141 0.01286676
0.82279176 1.3001313 0.03114631 0.60547316 -0.14380386 0.7700485
-0.4245902 -0.38895702 0.26376164 -0.16691181 -0.33511797 0.14936578
0.42904472 -0.2735951 -0.24763842 1.1579996 0.38129118 0.24689396
0.57926977 -0.9722316 -0.06372441 0.05033924 -0.97608334 0.9401391
-0.07099519 0.27479565 0.46050492 -0.7020772 0.4247059 0.91228795
0.67863053 0.00765142 -0.8612659 -0.7809631 0.2608968 0.4503391
-1.2495866 -0.22086945 0.82902646 -0.6054175 0.8453775 0.39222053
0.40392634 0.41059622 0.26849335 0.44702864 1.1082301 -0.33538154
0.2229219 0.2060658 0.2955668 0.7490664 0.64879024 -0.3447824
-0.02430837 0.32959485 0.5566069 0.2582977 -0.18222715 -0.04244959
-0.835471 0.4939244 0.42969692 0.4041394 -0.28243193 0.11490517
0.42654634 0.07699722 0.21592735 -0.22537811 -0.22668038 -0.05056111
-0.9426038 -0.01625794 0.276074 0.73726726 0.8943414 0.22968368
-0.20046787 0.61715907 0.16103524 0.5180428 0.35720155 -0.388168
0.30428663 1.0131875 0.08072232 -1.5270994 -0.6359139 -0.47455797
-1.0400515 0.29147616 0.23580815 -0.19122925 -0.9252045 1.3120387
0.584891 0.40513766 -0.18320616 0.6871513 0.75747705 0.66523975
0.18759081 -0.2636121 1.3142399 -0.72263265 -0.36646932 -0.16375528
0.1794862 -0.90751123 0.8460543 0.2973488 -0.9106381 -0.9074049
-1.2909423 0.27609774 -0.37698156 0.31775507 0.465724 0.7176138
-0.7930594 0.47790232 -0.75907916 -0.52085364 0.4978049 0.32364485
-0.07577982 0.08893096 -0.7823103 0.4623812 0.50835246 0.26199275
-0.5750126 -0.41020238 -0.453473 0.06726314 0.38245794 -0.2362547
0.8851133 -0.9581613 -1.2530953 0.74930996 0.06387381 -0.57998145
0.5569122 0.01840928 -0.8346921 0.3223225 -0.06550483 0.3201381
0.9817211 -1.0126976 -1.1124359 -0.20648286 0.0598661 -0.05099253
-0.28117624 0.10000539 -0.51056206 -0.5702852 0.15700525 -0.71593165
-0.16084987 0.0733002 -0.12568027 -0.06360151 1.0896535 -0.62593997
1.399786 -2.1007972 -0.41247806 0.34973755 0.01448959 0.6411922
0.36018494 0.03272983 0.03827785 -0.04011406 -0.41654626 0.11349963
-0.21664894 -0.15423235 0.10875314 0.7349948 0.34779462 0.47404146
-0.48562473 -0.5346995 0.58736205 0.32786548 -0.33381635 -0.12974465
-0.04884138 0.35534018 -0.43610206 0.7019462 0.8653855 -0.05364348
-0.15031487 -0.27802837 -0.49194354 -0.2757199 -1.7359102 1.1366651
-0.14172785 0.69774437 0.08618571 -1.2718443 1.0019377 -0.06281549
0.54303324 -0.883771 0.41424587 0.09244338 0.06107108 -0.57544756
0.5502222 0.08052678 0.02974049 0.06490341 -0.3066618 0.26353663
0.46048105 -0.16378333 1.1927854 -0.03001083 0.22948968 -0.24666035
0.929704 -0.07187402 0.6513446 0.6479971 -0.4499039 0.33283255
0.32822132 -0.38596755 -0.7776274 -0.8116301 -0.32131273 0.73508745
0.5430399 0.20567684 -0.8055741 -0.37200892 -0.17086545 0.4541611
-0.3222009 -0.11925211 -0.75520307 -0.73598796 0.582087 0.34652436
1.0671636 -1.1064982 -1.2775656 0.23463063 0.260375 -1.2290546
-0.0752929 0.00542944 -0.7998877 -1.1013861 -0.39077872 -0.86950773
0.6734237 0.47305298 0.7441055 0.25568962 -0.6100249 0.16882604
-0.4099511 -0.3437652 -0.02170188 -0.0618434 -0.08918723 0.45849177
0.34435156 -0.625127 -0.8673501 0.26701993 -1.1381427 -0.03180991
0.82414514 0.4295529 0.5731113 0.37328553 0.43655792 -0.94260865
0.14670885 -0.30686697 -0.902404 0.13967766 -0.603992 -0.35133424
0.65485746 -0.18972519 -1.1217415 -0.0333685 0.05346051 -0.21636225
-0.50845665 0.18082882 -0.16970159 -0.02433131 0.46110603 0.40740052
-0.5070788 -0.45610535 0.14262922 0.6284952 0.77747506 -0.11798555
0.81525534 0.6504006 0.11975095 -0.99772894 -0.05594245 -0.65249646
-0.556355 -0.5416381 1.0296811 -0.7397991 -0.7485924 0.765773
-0.9555674 0.13421522 0.05680419 0.3306217 -0.14535068 0.42770523
-0.25231126 -0.9201857 -0.44221076 -1.1328123 0.7591208 0.6711092
0.11969462 -0.57842946 -0.33996993 0.3822926 0.6428167 0.6592187
1.0724546 -0.42533258 -0.7544536 -0.31368273 -0.5339207 0.37177634
0.2974591 0.00591622 -0.9221897 -0.20274921 -0.04825785 0.28139764
0.97887355 0.45016083 1.1157846 -0.09134578 -0.45126325 0.35897252
1.705442 0.620016 0.7225809 0.39171892 0.62285423 0.3036277
0.45886996 0.6045255 0.10409725 0.5548505 0.5307409 -0.29913065
-0.2605744 0.1989903 0.27869728 0.5668092 0.16711868 -0.19724348
-0.92986387 0.58151156 -1.6478176 -1.0651207 -0.42938048 2.118049
0.53530645 0.56075686 0.01465152 0.48996097 1.093932 0.1683594
-0.7081411 -0.44677243 -0.14953852 0.5500507 0.6192007 0.09850916
-1.4061106 0.83227044 4.1076097 0.82331115 -1.406073 0.06900782
0.6452402 0.03294448 0.21774042 -0.04363452 -0.7288564 0.73127127
0.751313 0.05916675 0.07309028 0.6835995 0.3795664 -0.6030352
-0.6047971 0.9816534 0.10573367 -1.1683152 0.05215155 -0.38340396
0.8098372 -0.2006968 0.0481003 0.1601207 -0.24316047 -0.69859153
0.6811991 0.56870353 0.37832025 -0.84150636 0.7843635 0.08339588
-1.4440038 -0.1779379 -0.17088065 -0.12794593 0.01735182 0.7952507
-0.5240615 0.53808147 1.0551871 0.47104964 -0.79804945 1.1801184
0.17140956 0.82419294 -0.31655198 -0.1174615 0.11111661 -0.19538008
0.55487216 1.4854797 0.20007531 0.13216098 0.23484197 0.60461164
-0.11172196 0.35656837 -0.3062717 0.39447132 0.3525215 1.3730706
-1.2700363 -0.4137961 -0.519543 0.84060943 -0.17856266 0.19092333
-1.0843968 -0.7748331 0.5086491 0.33021742 0.73817486 -0.2730239
-0.41470858 -0.7410549 0.3669409 -0.6785797 0.21736886 -0.35215247
-0.7591499 0.43895346 0.03767737 -1.410152 0.1963226 -0.58894455
-0.65782493 0.521183 -1.5715588 -1.0114837 -0.67767626 0.55682766
0.72029227 -0.10658871 0.05898645 0.718452 -1.0380968 0.47218445
0.19723125 0.10096872 0.390459 -0.796584 0.22815646 1.44275
-0.15747093 0.2675447 0.4358213 -0.5505292 -1.2988975 -1.3818452
0.11424811 0.15057738 0.34551352 -0.07661489 -0.8956882 0.16402976
0.24294125 0.09577882 0.18747786 -0.5796554 0.07105049 -0.44981414
-1.3079782 0.6135701 0.87368774 -0.02333852 -0.23723692 -0.12949882
0.42978194 -0.27767962 -0.5818169 0.5801204 0.38943753 -1.1636168
0.6702823 -0.20925295 0.16867465 -0.8409484 -0.21628742 -0.50682306
-0.33356777 -0.3274557 0.0322364 1.4581332 0.2341956 -0.5704437
0.6764734 0.42629084 -0.08314103 -0.5294125 -1.0904133 -0.516782
-0.27920854 -0.40379086 0.3544723 0.64118963 -0.76396817 0.14922535
-0.07256466 0.37521258 0.390726 -0.04609876 0.72763056 -1.2775851
-0.16583614 -0.59979266 -0.90528387 -0.66847396 -0.29142216 -0.45221126
-0.07691039 -1.3854318 0.05093861 -0.47902757 -0.5267802 0.44245806
-0.23303516 0.42052063 -0.01876882 0.04425739 -0.672055 0.20044826
0.67098 -0.24069943 -0.276305 -0.02708693 -0.37361962 0.89665264
0.92026883 -0.36505818 -0.2766496 -0.399234 -0.03074776 -0.3362995
0.41295177 -1.574285 0.3851304 -0.24289349 0.49288645 -0.61420923
0.01805888 -0.9098613 0.20111689 0.725923 0.18671587 0.2950986
0.60400665 0.6968589 -0.08268055 -0.32806385 1.1802652 -0.00918318
-1.2239116 -0.00430512 -0.6107387 -0.07078803 1.479401 -0.6724582
-0.28964266 0.00905038 -0.11773552 -0.04996574 0.15291314 0.45439047
0.41838914 -0.91146106 -0.60804987 0.23308659 0.07098869 0.02004368
0.55747706 0.8871718 -0.81094056 0.3068245 -0.23728299 -0.68150336
-1.3105729 0.36883655 0.2789115 -0.04238176 -0.53177977 0.6057353
-0.15963973 0.11088081 0.17806503 -0.5486262 -0.28316993 0.06421143
0.45686322 0.8901069 0.37163374 -0.6944191 -0.57570165 0.7313027
-0.11735554 0.17041704 1.0961835 -0.2001986 -0.01370592 0.25527406
1.0033988 -0.02004505 -1.1734681 -0.27393797 0.03582775 -0.3523891
0.13853624 -0.68257 -1.0625916 0.73597366 1.4332609 0.15900557
1.4099001 -0.39328876 0.77229404 0.2791758 0.25156704 -1.24155
0.07049125 0.23655565 0.4359947 -1.275581 0.16275868 -0.3599353
-0.30491182 0.95772344 0.8105575 -0.19315465 0.6964251 0.03332013
-0.03611846 -0.05151233 -0.2552504 -0.2958578 0.16875848 0.2972283
0.15627375 0.01611245 -0.53601 0.4025864 0.20536323 -0.06213208
0.37773472 1.077719 -0.6805106 -0.67980933 -0.42057177 0.5698575
-0.5125059 0.14183308 -0.08834513 0.8121318 0.64163065 1.2088648
0.2915282 -0.33887768 0.52792966 -0.12628037 0.15824912 -0.12231211
-0.6291976 -0.23570392 -0.28051487 -0.29588693 -0.56796217 -0.7259342
-1.3433126 -0.36018476 -0.21744914 -0.20994876 0.64451665 0.9437015
0.3063902 0.7230875 0.73502094 -0.5503988 -0.05426657 -0.9446348
-0.3510723 0.3710554 0.19957805 -0.6688291 0.0391951 0.19671302] | [8.75394344329834, -0.6865972280502319] |
4a5d1c3f-7055-4ea0-9b1c-deaa8eeabcf7 | atlantanet-inferring-the-3d-indoor-layout | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/604_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123530426.pdf | AtlantaNet: Inferring the 3D Indoor Layout from a Single 360(∘) Image beyond the Manhattan World Assumption | We introduce a novel end-to-end approach to predict a 3D room layout from a single panoramic image. Compared to recent state-of-the-art works, our method is not limited to Manhattan World environments, and can reconstruct rooms bounded by vertical walls that do not form right angles or are curved -- i.e., Atlanta World models. In our approach, we project the original gravity-aligned panoramic image on two horizontal planes, one above and one below the camera. This representation encodes all the information needed to recover the Atlanta World 3D bounding surfaces of the room in the form of a 2D room footprint on the floor plan and a room height. To predict the 3D layout, we propose an encoder-decoder neural network architecture, leveraging Recurrent Neural Networks (RNNs) to capture long-range geometric patterns, and exploiting a customized training strategy based on domain-specific knowledge. The experimental results demonstrate that our method outperforms state-of-the-art solutions in prediction accuracy, in particular in cases of complex wall layouts or curved wall footprints. | ['Marco Agus', 'Giovanni Pintore', 'Enrico Gobbetti'] | null | null | null | null | eccv-2020-8 | ['3d-room-layouts-from-a-single-rgb-panorama'] | ['computer-vision'] | [ 4.46456701e-01 2.22674951e-01 3.71894300e-01 -5.78448772e-01
-4.56734687e-01 -3.66421163e-01 4.20607358e-01 -2.09985718e-01
-1.44647202e-02 1.35636613e-01 5.11385441e-01 -5.01971304e-01
-1.53620154e-01 -1.07222342e+00 -1.03237021e+00 -3.33019555e-01
1.00190975e-01 4.89694566e-01 6.21222937e-03 -2.13979647e-01
3.35826367e-01 5.03542125e-01 -1.52970099e+00 2.45607838e-01
5.83920896e-01 9.99518991e-01 5.01665652e-01 9.06196475e-01
-3.04409638e-02 6.19826555e-01 -2.43636101e-01 -1.67948380e-01
3.94361317e-01 -1.35487124e-01 -3.74836415e-01 3.84965599e-01
7.33503580e-01 -4.24222797e-01 -5.28808713e-01 4.74526048e-01
2.66431242e-01 6.56062514e-02 5.49341559e-01 -4.10841465e-01
-5.75413108e-01 -5.64800948e-02 -2.99221516e-01 -5.45162022e-01
6.06756449e-01 -2.42777392e-01 8.43954384e-01 -9.45837379e-01
4.74328846e-01 1.04455149e+00 7.56219983e-01 2.11356685e-01
-9.61904764e-01 -1.31429091e-01 7.71606624e-01 -1.76037699e-01
-1.34278536e+00 -2.34792739e-01 1.09832275e+00 -4.12103832e-01
9.21933711e-01 3.19274098e-01 6.20342433e-01 1.22633791e+00
2.04978228e-01 5.23281991e-01 9.73006070e-01 -5.08056760e-01
1.99209496e-01 -1.72697216e-01 -3.21702391e-01 1.10323715e+00
1.00261383e-01 -1.59734517e-01 -2.14817181e-01 3.37819815e-01
1.03457856e+00 4.56742942e-01 -5.11332154e-01 -1.01240671e+00
-1.25899184e+00 3.71867418e-01 8.47277880e-01 1.55505031e-01
-2.10468709e-01 -9.41472873e-02 -1.25330776e-01 -2.23239809e-01
3.94423783e-01 4.19238597e-01 -3.65078628e-01 1.38028666e-01
-8.69291842e-01 2.06739590e-01 8.08847606e-01 1.11608601e+00
6.46380603e-01 -1.29054621e-01 2.60815144e-01 7.16974020e-01
5.70076466e-01 6.25465035e-01 9.22878012e-02 -7.55332828e-01
1.11625457e+00 6.19480371e-01 4.36360449e-01 -9.32243884e-01
-7.13080287e-01 -2.91126847e-01 -9.64587569e-01 -1.49283990e-01
2.78446436e-01 1.29814446e-01 -1.11641276e+00 1.25723982e+00
1.58486739e-01 3.68245952e-02 -1.28826380e-01 9.28173363e-01
6.23364091e-01 8.29834819e-01 -6.14743590e-01 3.37014288e-01
1.22813272e+00 -1.39449883e+00 -3.48658323e-01 -8.07070434e-01
2.46554181e-01 -7.29751706e-01 1.24192119e+00 3.53634179e-01
-8.68205547e-01 -5.51808298e-01 -1.14552534e+00 -1.77888989e-01
-4.27829981e-01 4.03890431e-01 1.80508375e-01 6.26161277e-01
-8.18699121e-01 4.33102429e-01 -8.66631687e-01 -2.96182334e-01
6.40707836e-02 8.17133486e-02 -3.72018337e-01 -4.26027477e-01
-4.72732008e-01 8.31624269e-01 -2.15068430e-01 5.00146985e-01
-6.74961984e-01 -4.68104452e-01 -1.15741026e+00 2.81042397e-01
4.25673366e-01 -9.54263628e-01 1.18698764e+00 -3.80346954e-01
-1.69277132e+00 6.53679550e-01 -3.79250169e-01 -1.05058707e-01
5.92549026e-01 -6.48344338e-01 -3.13833535e-01 -1.69055358e-01
-1.76371276e-01 1.19148947e-01 6.70417845e-01 -1.84943342e+00
-1.96022496e-01 -6.25105321e-01 -8.05391893e-02 4.17199373e-01
-2.38010027e-02 -6.17931128e-01 -7.34804213e-01 -3.29423666e-01
8.63146484e-01 -9.70902562e-01 -4.98481959e-01 -4.95691597e-02
-9.90899026e-01 6.20633602e-01 5.10455787e-01 -9.01040792e-01
1.09688926e+00 -1.98723340e+00 8.81946608e-02 5.73603392e-01
-1.70868725e-01 2.44093835e-02 1.92879528e-01 5.61416328e-01
2.24710740e-02 -1.12032443e-01 -3.32405865e-01 -8.59978199e-01
1.63695216e-01 4.36393023e-02 -4.76888418e-01 3.69286656e-01
-2.52593786e-01 4.20094132e-01 -6.31223738e-01 1.60683487e-02
6.77851319e-01 7.58665919e-01 -4.51986372e-01 5.43265224e-01
-2.16809899e-01 4.33367074e-01 -4.84983027e-01 5.28768957e-01
8.47512424e-01 -3.08567017e-01 4.76529866e-01 6.22894019e-02
-3.41494948e-01 6.74948573e-01 -1.11380994e+00 2.06591368e+00
-1.14037061e+00 3.75722080e-01 -8.60649720e-02 -4.82075274e-01
1.34997642e+00 7.30353594e-02 -4.26225290e-02 -6.41418159e-01
-1.33160368e-01 2.12298229e-01 -7.03281462e-01 -4.41224098e-01
6.93859398e-01 1.52705356e-01 -1.16040863e-01 1.16081811e-01
-4.92749810e-01 -5.50083041e-01 -4.61353391e-01 -4.45518643e-01
8.69039476e-01 5.73219895e-01 1.96681708e-01 2.04596341e-01
4.23783600e-01 -5.13423085e-01 1.78156659e-01 6.08574569e-01
4.14899439e-01 1.29459596e+00 2.62722850e-01 -1.02616203e+00
-1.38112020e+00 -1.21099317e+00 1.78394049e-01 6.76593423e-01
1.95849493e-01 -4.51154590e-01 -8.85231137e-01 -3.64850938e-01
-3.70574474e-01 7.95872569e-01 -6.02313757e-01 2.62789041e-01
-8.43966186e-01 -3.82690638e-01 -4.95286435e-02 6.04696214e-01
6.39397562e-01 -6.69864476e-01 -1.03758776e+00 3.87508720e-02
-3.31236541e-01 -1.48701036e+00 -3.44542295e-01 3.11754078e-01
-7.79657960e-01 -1.07439959e+00 -7.03398585e-01 -9.07837510e-01
9.05021191e-01 6.46522343e-01 1.19385338e+00 -3.01864028e-01
-1.09772697e-01 3.11396718e-01 -6.51679039e-02 -1.36654139e-01
1.12204768e-01 1.47996694e-01 -2.36039311e-01 1.65737458e-02
-3.03882621e-02 -7.04856396e-01 -7.87474871e-01 4.19118822e-01
-3.07739496e-01 6.86437130e-01 5.68387866e-01 5.36477864e-01
7.35957444e-01 -2.52507627e-01 -3.80824268e-01 -8.21789563e-01
4.75251004e-02 -1.92300811e-01 -8.93639684e-01 4.04688329e-01
-3.44263196e-01 2.97676772e-02 1.07694769e+00 1.41975507e-01
-1.21670878e+00 3.91183436e-01 -1.49452597e-01 -5.34233868e-01
-4.53730553e-01 9.50808376e-02 -4.47708607e-01 4.14268702e-01
3.64204228e-01 2.29097009e-01 -6.74595118e-01 -6.79570079e-01
8.81394744e-02 4.28589433e-01 4.39435840e-01 -3.59794915e-01
8.91466200e-01 6.85786963e-01 1.65800557e-01 -9.27765310e-01
-1.14110708e+00 -4.71442461e-01 -1.06834102e+00 -1.22670636e-01
1.04718530e+00 -1.04126740e+00 -5.37963569e-01 3.71788979e-01
-1.25015092e+00 -5.43043733e-01 3.09610870e-02 5.22923172e-01
-6.76675379e-01 -8.73946100e-02 -3.76736909e-01 -9.79391992e-01
-1.29005045e-01 -9.62009847e-01 1.47199190e+00 1.29488096e-01
2.59822533e-02 -7.92203307e-01 1.54967099e-01 5.58366477e-01
1.42230138e-01 3.76459450e-01 9.93104637e-01 6.93401024e-02
-1.16941643e+00 -2.96491027e-01 -6.25963584e-02 -4.85858917e-02
-6.66563436e-02 -6.17483780e-02 -1.28258705e+00 1.34861782e-01
-1.37213975e-01 9.63593349e-02 7.71435499e-01 4.48154211e-01
1.48915648e+00 -3.42828065e-01 -3.50351483e-01 9.77098465e-01
1.52287519e+00 2.62313187e-02 7.21935391e-01 5.71578979e-01
9.80920851e-01 7.07458794e-01 3.81974220e-01 6.03281200e-01
8.80297720e-01 7.31561720e-01 8.87203395e-01 -1.02150857e-01
2.10286215e-01 -1.04917192e+00 1.99547648e-01 7.89540827e-01
-3.63804519e-01 -3.81896049e-01 -1.06070232e+00 4.70506638e-01
-1.69101381e+00 -7.03814507e-01 1.58126019e-02 2.48852158e+00
-1.02741778e-01 1.80601120e-01 -4.34200138e-01 -7.94886798e-03
3.44838858e-01 6.83438718e-01 -4.27943021e-01 -5.67969561e-01
3.04508097e-02 1.47237396e-03 6.56644821e-01 6.16266787e-01
-1.18607867e+00 8.98057103e-01 5.73342943e+00 -1.76685788e-02
-9.69861031e-01 -2.65563756e-01 6.85721755e-01 -9.87685546e-02
-3.67868751e-01 -1.59963384e-01 -7.42532551e-01 -1.22274188e-02
8.59963596e-01 8.11152577e-01 6.59769773e-01 1.11357939e+00
2.22258881e-01 1.44369051e-01 -9.14129198e-01 9.24470603e-01
3.92465889e-01 -1.26943564e+00 -2.58110225e-01 2.05562085e-01
7.96804965e-01 8.45118165e-02 1.79371938e-01 9.67406780e-02
1.80762023e-01 -1.20378554e+00 9.26000059e-01 6.45872712e-01
8.51152897e-01 -4.87447202e-01 6.44501388e-01 5.38808644e-01
-1.43083155e+00 -2.21464127e-01 -3.13769639e-01 -3.86145636e-02
2.35019520e-01 4.11587179e-01 -1.12865436e+00 7.00408518e-01
9.04177785e-01 4.37001884e-01 -4.10272002e-01 8.14900219e-01
-5.82504153e-01 2.24908456e-01 -4.64263171e-01 6.49083853e-02
3.43031824e-01 -3.59969378e-01 8.32817629e-02 1.02373517e+00
8.27811837e-01 -9.42206085e-02 1.18389465e-01 9.26183641e-01
-1.28881752e-01 -1.75988898e-01 -1.08922637e+00 6.43227577e-01
3.93346727e-01 1.04409099e+00 -6.67950273e-01 1.16236329e-01
-4.54599410e-01 9.97892559e-01 3.72912705e-01 4.78633642e-01
-6.54788375e-01 -2.62553662e-01 2.39679247e-01 4.88904238e-01
7.97115743e-01 -5.97576082e-01 -5.32006681e-01 -1.14479268e+00
4.48871702e-01 -3.37863177e-01 -1.61871970e-01 -1.14981651e+00
-8.17673266e-01 7.02724576e-01 -2.55620033e-01 -1.23368454e+00
-1.21856876e-01 -9.96101022e-01 -6.82893515e-01 7.10199833e-01
-1.56783521e+00 -1.60509825e+00 -6.54661894e-01 3.76961678e-01
4.93614256e-01 2.02688038e-01 1.23565543e+00 -1.99183961e-03
-4.93001699e-01 1.43244579e-01 3.00475121e-01 1.30383894e-01
3.13136727e-01 -1.27784908e+00 8.01239371e-01 8.44911575e-01
1.77811071e-01 6.03908598e-01 6.06505275e-01 -2.35744074e-01
-1.53800976e+00 -1.27502298e+00 1.08551538e+00 -3.85242760e-01
-1.83822531e-02 -1.04053330e+00 -5.57980895e-01 9.74439800e-01
-1.51250049e-01 -9.37408861e-03 6.15347505e-01 4.60376084e-01
-5.62935233e-01 -2.09539399e-01 -7.53931582e-01 8.46166015e-01
1.32011449e+00 -4.81908739e-01 -4.41275716e-01 1.94354668e-01
5.44903517e-01 -7.95590878e-01 -4.55756128e-01 3.02095026e-01
6.68953657e-01 -1.36273217e+00 1.31943667e+00 9.50500667e-02
7.00128734e-01 -2.86739528e-01 -5.48328102e-01 -1.33826768e+00
-1.23280704e-01 -5.07161677e-01 -1.37024343e-01 4.98529553e-01
4.12811577e-01 -2.07289740e-01 9.38721120e-01 5.70354164e-01
-5.89342892e-01 -9.20776486e-01 -8.72642219e-01 -3.30170959e-01
-3.28529418e-01 -4.30274963e-01 8.95231187e-01 4.14413929e-01
-3.73931199e-01 3.69735450e-01 -5.74639738e-01 5.59850156e-01
2.98415154e-01 6.94331467e-01 1.11584973e+00 -1.10051250e+00
-1.19374029e-01 -2.29426458e-01 -1.06221095e-01 -1.68381417e+00
3.69997844e-02 -3.92290235e-01 2.61022717e-01 -1.94865406e+00
-2.93708324e-01 -3.34078819e-01 9.43373740e-02 1.56620279e-01
2.39230320e-01 -1.08934030e-01 1.88495800e-01 -1.30671963e-01
-4.89134371e-01 8.50369632e-01 1.31927788e+00 8.34246725e-02
-3.06690514e-01 1.56508461e-01 -4.19117332e-01 1.07959497e+00
6.26787364e-01 3.99937034e-02 -3.24894160e-01 -8.09551179e-01
3.78555208e-01 6.55488968e-02 3.46348733e-01 -1.24311709e+00
1.93429850e-02 -1.08041624e-02 7.06503510e-01 -9.50017810e-01
9.37221169e-01 -1.08680582e+00 -1.86794251e-01 2.02706411e-01
-9.38471928e-02 1.30767345e-01 -9.50643234e-03 7.14399159e-01
2.71367263e-02 6.54444173e-02 3.47958446e-01 -3.67589980e-01
-2.79423207e-01 2.12206632e-01 -1.52262717e-01 -3.78485829e-01
7.10924268e-01 -4.08210218e-01 -2.53746569e-01 -4.03237820e-01
-3.91755223e-01 -9.26991850e-02 7.32878447e-01 4.82789636e-01
1.03892899e+00 -9.96993899e-01 -2.74414271e-01 6.70129895e-01
1.41923249e-01 7.48070359e-01 4.80023682e-01 2.46347591e-01
-9.53513801e-01 8.08920920e-01 9.68229547e-02 -5.72679162e-01
-1.02246606e+00 5.78892410e-01 4.26664442e-01 -6.40052140e-01
-1.10811830e+00 6.89682126e-01 5.26749849e-01 -1.11453998e+00
2.12989286e-01 -8.30892801e-01 -1.15585411e-02 -5.11648953e-01
2.38097206e-01 -5.92024140e-02 1.93357751e-01 -5.02219498e-01
-1.65672272e-01 1.10145807e+00 3.10655743e-01 -1.44156873e-01
1.57873392e+00 -2.66929686e-01 4.21173573e-01 7.87827909e-01
8.75258565e-01 2.50479281e-01 -1.66493261e+00 -1.22237667e-01
-2.56313413e-01 -7.84916162e-01 -1.60808772e-01 -8.04681778e-01
-7.17439234e-01 1.31579244e+00 4.80912477e-01 -1.93413422e-01
9.11192477e-01 -3.16566944e-01 7.49521554e-01 6.21021509e-01
5.65195024e-01 -1.07507420e+00 9.67596546e-02 7.86824405e-01
9.57065701e-01 -1.08015037e+00 1.68795362e-02 -5.55372834e-01
-5.96931100e-01 1.28037739e+00 6.30314469e-01 -1.90950558e-01
5.57238877e-01 6.49545789e-02 -1.33255780e-01 -1.39234230e-01
-4.77780402e-01 1.31832033e-01 3.78839970e-01 4.52637255e-01
3.89478892e-01 3.63486171e-01 1.01539481e+00 4.48324561e-01
-6.71118259e-01 -4.07633364e-01 3.62935096e-01 8.30489039e-01
-3.81392717e-01 -6.64439797e-01 -7.49606729e-01 4.64692153e-02
1.67600811e-01 -6.14941977e-02 -1.51331946e-01 7.77107418e-01
1.35607556e-01 6.15624845e-01 3.54211867e-01 -3.52568895e-01
6.86085224e-01 6.06461838e-02 6.16826952e-01 -5.63732862e-01
-2.45661065e-01 6.50425702e-02 1.21686369e-01 -5.78477442e-01
-3.98411192e-02 -4.00681078e-01 -8.83115113e-01 -2.19405577e-01
-5.25976047e-02 -2.64962107e-01 9.79697704e-01 6.62906229e-01
3.94876391e-01 6.55351460e-01 7.91640639e-01 -1.30949903e+00
-7.77293891e-02 -8.41281652e-01 -4.48894024e-01 1.15930445e-01
5.85412621e-01 -5.25000453e-01 2.17432696e-02 9.73986462e-03] | [8.719344139099121, -2.866748809814453] |
7852e464-8891-4e16-8227-9dd42e4a38ac | amrita-lt-edi-eacl2021-hope-speech-detection | null | null | https://aclanthology.org/2021.ltedi-1.22 | https://aclanthology.org/2021.ltedi-1.22.pdf | Amrita@LT-EDI-EACL2021: Hope Speech Detection on Multilingual Text | Analysis and deciphering code-mixed data is imperative in academia and industry, in a multilingual country like India, in order to solve problems apropos Natural Language Processing. This paper proposes a bidirectional long short-term memory (BiLSTM) with the attention-based approach, in solving the hope speech detection problem. Using this approach an F1-score of 0.73 (9thrank) in the Malayalam-English data set was achieved from a total of 31 teams who participated in the competition. | ['Kothamasu Sai rahul', 'Ravi teja Tasubilli', 'Thara S'] | null | null | null | null | eacl-ltedi-2021-4 | ['hope-speech-detection'] | ['natural-language-processing'] | [-1.56934097e-01 -3.76702882e-02 2.79130518e-01 -7.81238005e-02
-1.03071308e+00 -2.96178192e-01 4.37822312e-01 1.68187737e-01
-7.48201907e-01 6.85962617e-01 4.10311967e-01 -5.72814107e-01
1.24688484e-01 -9.09815878e-02 -5.59762359e-01 -3.68259043e-01
-1.49780124e-01 5.11002019e-02 -5.90799041e-02 -3.40912670e-01
7.78317630e-01 5.38623631e-02 -1.09508109e+00 9.18124199e-01
9.77666318e-01 6.69136882e-01 6.69808865e-01 1.34600401e+00
-1.31558776e-01 1.42111671e+00 -5.25287569e-01 -8.00207376e-01
-2.66091675e-01 -4.16304111e-01 -1.16535270e+00 -2.81255335e-01
1.28378183e-01 1.10013239e-01 -1.42790347e-01 1.18667674e+00
5.16395330e-01 6.15188889e-02 1.94412485e-01 -6.15257561e-01
-1.10646319e+00 9.74310875e-01 -4.72582161e-01 7.98529506e-01
5.10661006e-01 -1.23966776e-01 1.08151329e+00 -1.33204162e+00
2.30554432e-01 1.08800983e+00 5.40331721e-01 4.30799454e-01
-8.17770600e-01 -6.79238021e-01 -8.89030471e-02 4.62257266e-01
-1.50554383e+00 -6.67185605e-01 5.11551023e-01 -3.63133758e-01
1.79679942e+00 -3.40318354e-03 3.84393305e-01 1.00240099e+00
6.33469880e-01 7.40036130e-01 9.74038303e-01 -8.99488211e-01
-2.32986704e-01 8.36819503e-03 3.66278142e-02 7.37533867e-01
-1.64365560e-01 1.93536859e-02 -9.26081359e-01 3.97000551e-01
2.12423578e-01 -4.96629953e-01 2.98229847e-02 4.63469297e-01
-1.41206408e+00 8.30835879e-01 1.95757017e-01 8.11270952e-01
-3.76239836e-01 6.64609596e-02 6.68872952e-01 7.02037454e-01
3.72466892e-01 3.46011132e-01 -4.51597899e-01 -7.18365014e-01
-7.47724235e-01 -1.93932086e-01 5.36383390e-01 7.78875589e-01
-5.54630235e-02 5.96226692e-01 2.02259243e-01 9.68109250e-01
6.02147996e-01 4.10131246e-01 1.16881120e+00 -5.88985741e-01
9.91845787e-01 3.29231828e-01 -3.53657633e-01 -1.10102713e+00
-1.93549752e-01 -4.44686562e-01 -8.74091685e-01 4.65397835e-02
2.01949030e-01 -2.44848579e-01 -9.11280334e-01 1.44624519e+00
-4.02927607e-01 -6.78533763e-02 5.82389534e-01 6.30550146e-01
8.84622514e-01 9.65876043e-01 -9.93948728e-02 7.36420676e-02
1.41059279e+00 -1.21937203e+00 -8.10654700e-01 -4.83696789e-01
6.50939345e-01 -1.25540304e+00 1.12241423e+00 6.85248733e-01
-1.16500425e+00 -6.94871724e-01 -1.25847697e+00 -2.46511772e-01
-3.69470268e-01 2.01472089e-01 4.22467776e-02 6.95944011e-01
-1.25210238e+00 1.85354352e-01 -5.81550241e-01 -3.67788941e-01
1.77456155e-01 3.95113826e-01 -5.04661262e-01 3.33510697e-01
-1.23510420e+00 1.07522833e+00 4.77733403e-01 3.32196444e-01
-9.61788237e-01 -1.06261946e-01 -6.33046687e-01 -6.70898557e-02
-8.87930915e-02 6.19926080e-02 1.30215311e+00 -1.16889083e+00
-1.51323426e+00 1.14673281e+00 -4.11817543e-02 -9.59785879e-01
3.48533571e-01 -3.62346351e-01 -7.32961059e-01 -1.40092805e-01
-1.24744460e-01 4.04089332e-01 7.20964193e-01 -5.54921448e-01
-7.50503957e-01 -2.10193619e-01 -1.32964104e-01 1.09740727e-01
-4.06556159e-01 6.68174446e-01 6.31576777e-02 -9.74003196e-01
-1.19156010e-01 -9.51997519e-01 8.50603357e-02 -8.27647924e-01
-4.12441969e-01 -7.18138069e-02 4.87663925e-01 -1.47781706e+00
1.31288862e+00 -2.41133332e+00 1.62585348e-01 -1.40716344e-01
-6.65587932e-02 5.70800126e-01 -1.38761312e-01 3.59492749e-01
-1.30476579e-01 2.16224998e-01 -2.81177133e-01 -2.42791161e-01
-3.68023068e-01 -3.61127453e-03 -2.55209506e-01 3.95784140e-01
2.91791528e-01 7.01502323e-01 -7.59739876e-01 -4.83639270e-01
-1.64112542e-02 3.94462556e-01 -2.62098253e-01 2.81140298e-01
3.32599849e-01 3.23555201e-01 -6.27812594e-02 4.96116579e-01
3.25327307e-01 1.79696783e-01 -2.50389930e-02 5.41671693e-01
-5.62047958e-01 3.33828986e-01 -7.44951367e-01 2.06365037e+00
-6.69322193e-01 1.27990258e+00 2.46786907e-01 -9.99441206e-01
8.69377017e-01 8.86274934e-01 -2.40806147e-01 -8.42455864e-01
3.72422367e-01 5.49063146e-01 5.06737649e-01 -6.77530110e-01
5.71793258e-01 -1.51993558e-01 -7.44122043e-02 2.24724188e-01
1.60888389e-01 9.11494195e-02 1.23221830e-01 1.89353615e-01
7.60344625e-01 -3.91598165e-01 2.59224743e-01 -4.60410506e-01
9.53264177e-01 -2.05849662e-01 1.76359713e-01 4.73268837e-01
-3.91147077e-01 6.12581193e-01 3.23644400e-01 -5.22569537e-01
-1.24140239e+00 -5.27676463e-01 8.57802629e-02 1.13275301e+00
-6.42943323e-01 -2.41977751e-01 -8.00297856e-01 -2.85478920e-01
-5.05695522e-01 6.33042693e-01 -2.92814732e-01 -2.41104767e-01
-6.52918100e-01 -3.85705233e-01 1.12896931e+00 3.55793357e-01
4.93675172e-01 -1.48949993e+00 -5.97499311e-01 4.35373485e-01
-5.17653942e-01 -1.13734591e+00 -5.03406882e-01 4.85350937e-01
-5.51682174e-01 -6.62310839e-01 -7.22006321e-01 -1.30278015e+00
1.13993227e-01 -1.38433129e-01 1.02067125e+00 6.68428168e-02
-9.50057358e-02 -3.64604205e-01 -4.28569645e-01 -5.05239785e-01
-7.66531944e-01 2.72838891e-01 -7.66995028e-02 6.24625236e-02
5.63431382e-01 -2.14755863e-01 8.63224268e-02 -2.20548093e-01
-3.96089762e-01 8.70193839e-02 8.83207202e-01 8.47915351e-01
1.45805103e-03 -1.74879581e-01 7.13899553e-01 -4.00791556e-01
9.31641936e-01 -5.30112445e-01 -5.79984009e-01 2.09892422e-01
-5.21059334e-01 1.90822750e-01 8.74450982e-01 -3.14472407e-01
-8.12986970e-01 -1.39241993e-01 -4.72038209e-01 5.58835082e-02
6.77698851e-02 8.61916065e-01 4.83386703e-02 -7.93904737e-02
5.69616199e-01 5.38186729e-01 -5.39235882e-02 -4.84395683e-01
1.51993958e-02 1.23017251e+00 6.58393085e-01 -3.39057416e-01
2.15965554e-01 -2.67697811e-01 -7.08296120e-01 -1.00758147e+00
-3.05294454e-01 -1.75185114e-01 -5.25959134e-01 -1.69501588e-01
1.14230597e+00 -1.22046399e+00 -5.41451514e-01 7.33036458e-01
-1.67701447e+00 -1.08196057e-01 3.08095962e-01 7.91271448e-01
-2.67849684e-01 2.27446720e-01 -8.50049794e-01 -8.23269367e-01
-4.53475833e-01 -1.31229734e+00 4.35377240e-01 -6.11457825e-02
-3.51341546e-01 -9.27331626e-01 7.75065497e-02 5.69408536e-01
4.57378596e-01 -9.82258096e-02 8.20013344e-01 -7.51121104e-01
-1.76260933e-01 -1.65581256e-01 -1.68578178e-01 7.77007639e-01
-1.91192031e-01 -1.21735288e-02 -1.07975411e+00 -3.42282206e-01
1.28658697e-01 -5.23459673e-01 7.77441144e-01 2.07100630e-01
8.46281826e-01 -1.90155700e-01 4.32752281e-01 8.72857273e-02
1.16144967e+00 5.77604413e-01 4.21543181e-01 2.52609372e-01
6.81642473e-01 4.37650323e-01 2.02567726e-01 2.86014020e-01
4.36196536e-01 3.71303529e-01 2.44418025e-01 1.50134429e-01
-2.21332088e-01 -4.29320857e-02 8.92489433e-01 1.81195581e+00
2.68412870e-03 -3.87193441e-01 -1.48009002e+00 8.11320662e-01
-1.49031997e+00 -7.47802317e-01 -4.85002190e-01 1.95458031e+00
8.36322844e-01 2.73540705e-01 -8.00590813e-02 4.32833433e-01
9.12791967e-01 4.30441685e-02 -1.84149444e-02 -1.40313399e+00
-1.63714468e-01 7.61397034e-02 2.91670024e-01 7.84242392e-01
-9.31824446e-01 9.96209800e-01 6.67604399e+00 9.25580204e-01
-1.27364230e+00 3.37918192e-01 5.09054542e-01 2.47182384e-01
-4.18696553e-03 -1.82355776e-01 -4.98758942e-01 5.30727327e-01
1.99148476e+00 -2.95684963e-01 7.49740124e-01 5.67023635e-01
1.74113154e-01 1.79268286e-01 -5.44682384e-01 1.08162045e+00
3.23791116e-01 -9.43992198e-01 -1.40105829e-01 -5.71641624e-02
6.16913974e-01 4.28677171e-01 3.17296237e-01 3.51127833e-01
4.39840555e-01 -1.40194511e+00 1.14891827e+00 3.18823606e-01
5.90119362e-01 -1.08393478e+00 9.22065735e-01 4.89537776e-01
-1.14596570e+00 -3.34860414e-01 -2.48843506e-01 -3.74872923e-01
-1.18934453e-01 2.98538417e-01 -8.98185372e-01 2.30447739e-01
7.03170180e-01 5.51413536e-01 -5.57333589e-01 7.75640666e-01
-3.15768033e-01 6.69699252e-01 1.48534894e-01 -2.65905052e-01
5.16930223e-01 2.42104843e-01 5.21111250e-01 1.59545839e+00
5.94929576e-01 -2.78130651e-01 -3.67313437e-02 5.02925277e-01
-3.76467735e-01 3.70412201e-01 -5.97917855e-01 -5.09823263e-01
1.94439694e-01 8.29075456e-01 -5.53067863e-01 -1.00680768e-01
-6.03522599e-01 8.92297328e-01 4.41591680e-01 1.36340395e-01
-8.22950780e-01 -8.25991511e-01 1.37581006e-01 -3.98153692e-01
2.28520632e-01 -4.20931727e-01 -2.24266231e-01 -9.17894185e-01
3.74163277e-02 -1.31583333e+00 3.43189180e-01 -8.03855181e-01
-8.55297804e-01 1.14648187e+00 -4.67781365e-01 -1.10788119e+00
-3.16333890e-01 -6.10965073e-01 -4.28517938e-01 9.45523739e-01
-1.51379943e+00 -1.17505109e+00 1.73110157e-01 5.42331040e-01
1.23597288e+00 -9.18855488e-01 9.30145204e-01 6.81765556e-01
-5.41585624e-01 5.78004479e-01 1.05080158e-01 4.61399943e-01
4.81770843e-01 -1.29451454e+00 5.39901495e-01 1.43591940e+00
5.93831480e-01 5.78648984e-01 6.43791199e-01 -3.90806377e-01
-1.24047160e+00 -8.29668581e-01 1.57292891e+00 -2.81805843e-01
8.51459980e-01 -3.84492725e-01 -6.77293181e-01 5.84756196e-01
9.59975958e-01 -5.80549007e-03 7.07390547e-01 -2.44838476e-01
-3.66285712e-01 7.74555579e-02 -8.38448346e-01 2.35079572e-01
5.00101209e-01 -1.08645880e+00 -7.89362490e-01 6.25752807e-02
7.62615263e-01 -3.09367061e-01 -6.50841594e-01 -2.32198387e-01
5.23924053e-01 -8.44928145e-01 7.13158906e-01 -5.92145026e-01
6.66873336e-01 -6.62365034e-02 -3.65924507e-01 -1.35775924e+00
-1.04512818e-01 -5.79187274e-01 4.87960242e-02 1.08283329e+00
7.69937456e-01 -3.70517671e-01 3.00503939e-01 -4.62166127e-03
-5.26957512e-01 -3.48605633e-01 -1.22463560e+00 -6.41415000e-01
4.19398010e-01 -7.00569630e-01 1.16558902e-01 7.36093760e-01
6.35479316e-02 4.46752489e-01 -6.23189092e-01 -4.89401370e-02
3.69958505e-02 -4.20245081e-01 2.62750655e-01 -8.88311803e-01
-6.95233494e-02 -3.92718047e-01 -4.90752161e-01 -4.60198373e-01
4.76572216e-01 -1.06517112e+00 1.23899840e-01 -1.31071305e+00
2.11261902e-02 1.84675053e-01 -6.42140210e-01 4.39680487e-01
1.93209663e-01 3.79513204e-01 3.12758058e-01 3.18471044e-02
-5.87593496e-01 3.00555795e-01 6.62498772e-01 -3.27090561e-01
1.84357703e-01 -2.58225530e-01 -6.34152770e-01 5.18565297e-01
8.54170322e-01 -7.06329226e-01 -2.31644794e-01 -8.12493503e-01
4.04013842e-01 4.91970889e-02 7.34727308e-02 -1.25458717e+00
3.46862078e-01 7.48139843e-02 8.45935196e-02 -5.26650727e-01
1.44887760e-01 -6.10897005e-01 -1.30405694e-01 8.78247678e-01
-6.01327419e-01 7.98362017e-01 2.71353930e-01 2.00906962e-01
-7.93618262e-01 -3.44751656e-01 8.03092539e-01 -3.71733874e-01
-8.53635371e-01 -3.29865724e-01 -9.09628689e-01 3.89242291e-01
8.83661211e-01 -2.81756166e-02 -1.90993920e-01 -2.78766513e-01
-5.30081511e-01 -3.01096570e-02 -1.13546968e-01 7.86854029e-01
8.33786607e-01 -1.27769494e+00 -1.22506082e+00 4.19054210e-01
1.13784187e-01 -5.64891636e-01 1.15004681e-01 1.04969382e+00
-1.00337827e+00 6.72392964e-01 -3.87730271e-01 -1.68789327e-01
-1.50094283e+00 2.17610523e-01 2.03467533e-01 -1.88766167e-01
-2.99173176e-01 1.33779442e+00 -6.78309977e-01 -4.37859386e-01
1.43980816e-01 -3.23176444e-01 -6.58226430e-01 2.17165858e-01
6.79962158e-01 2.99567044e-01 5.38990676e-01 -1.00657165e+00
-6.72105312e-01 3.51351827e-01 -1.48012027e-01 -4.75025535e-01
1.15273166e+00 -2.04127505e-01 -2.39671126e-01 7.92699993e-01
1.44723392e+00 3.14745277e-01 -3.96485031e-01 4.70972657e-02
5.38137317e-01 -3.66728872e-01 2.54983485e-01 -7.32354045e-01
-9.29139614e-01 1.34392190e+00 8.19708765e-01 3.07710022e-01
7.81940103e-01 -4.62049603e-01 8.82367313e-01 5.18498123e-01
3.61545682e-02 -1.47897351e+00 1.05246060e-01 1.17280209e+00
1.00052905e+00 -1.32376266e+00 -6.13746405e-01 4.59450096e-01
-7.64443278e-01 1.47020471e+00 3.25637311e-01 -6.46922290e-02
4.81238604e-01 3.68329167e-01 2.52677679e-01 1.69161052e-01
-9.39255595e-01 1.61603242e-01 1.64025322e-01 4.73491788e-01
1.06416440e+00 5.44919400e-03 -3.68063152e-01 5.56216598e-01
-4.22644943e-01 -2.51192153e-01 7.81835079e-01 9.68142211e-01
-5.53038001e-01 -9.28584337e-01 -4.80857313e-01 1.29106805e-01
-1.17491055e+00 -5.57392597e-01 -3.73577625e-01 3.60675156e-01
1.39689803e-01 1.38422537e+00 5.63512482e-02 -6.67918742e-01
5.92341013e-02 8.51568133e-02 -6.35799989e-02 -4.96204942e-01
-9.75453794e-01 -5.05663902e-02 1.48693293e-01 -1.84307158e-01
-3.46706808e-01 -5.06583333e-01 -1.43896747e+00 -4.41776454e-01
-3.14250976e-01 3.83537114e-01 9.45008099e-01 8.81842971e-01
2.71034896e-01 7.31710136e-01 4.87183362e-01 -3.31241369e-01
-3.25317949e-01 -1.25825584e+00 -2.39323929e-01 -2.77811270e-02
8.24749053e-01 -1.76661402e-01 -3.10449719e-01 1.64237276e-01] | [14.172866821289062, 6.728785037994385] |
b252ae6d-758d-402a-921c-cc3c934795c5 | diverse-audio-captioning-via-adversarial | 2110.06691 | null | https://arxiv.org/abs/2110.06691v2 | https://arxiv.org/pdf/2110.06691v2.pdf | Diverse Audio Captioning via Adversarial Training | Audio captioning aims at generating natural language descriptions for audio clips automatically. Existing audio captioning models have shown promising improvement in recent years. However, these models are mostly trained via maximum likelihood estimation (MLE),which tends to make captions generic, simple and deterministic. As different people may describe an audio clip from different aspects using distinct words and grammars, we argue that an audio captioning system should have the ability to generate diverse captions for a fixed audio clip and across similar audio clips. To address this problem, we propose an adversarial training framework for audio captioning based on a conditional generative adversarial network (C-GAN), which aims at improving the naturalness and diversity of generated captions. Unlike processing data of continuous values in a classical GAN, a sentence is composed of discrete tokens and the discrete sampling process is non-differentiable. To address this issue, policy gradient, a reinforcement learning technique, is used to back-propagate the reward to the generator. The results show that our proposed model can generate more diverse captions, as compared to state-of-the-art methods. | ['Wenwu Wang', 'Mark D. Plumbley', 'Jianyuan Sun', 'Xubo Liu', 'Xinhao Mei'] | 2021-10-13 | null | null | null | null | ['audio-captioning'] | ['audio'] | [ 4.12980795e-01 4.88545656e-01 1.04319617e-01 -3.30208331e-01
-1.45649314e+00 -6.10788763e-01 5.23082793e-01 -2.71651030e-01
2.17709064e-01 1.19472432e+00 5.53559065e-01 2.46498257e-01
5.20249009e-01 -6.52188480e-01 -9.48343098e-01 -5.42086899e-01
5.44608496e-02 5.77069938e-01 -2.25530133e-01 -9.39466283e-02
-1.49427295e-01 -3.08200670e-03 -1.38011551e+00 4.57937419e-01
8.07451606e-01 8.07016432e-01 2.84008890e-01 9.36870456e-01
-2.34426156e-01 7.97912598e-01 -1.14447737e+00 -4.71864283e-01
-4.54418063e-02 -1.01823330e+00 -6.71785831e-01 -7.06594214e-02
3.48152846e-01 -2.77649283e-01 -1.15674898e-01 1.06026888e+00
7.68224776e-01 -8.33350420e-02 7.48057127e-01 -1.51453292e+00
-9.60168362e-01 1.02550089e+00 -1.69488311e-01 -2.97778666e-01
7.53674507e-01 1.69499204e-01 1.17647743e+00 -4.89904672e-01
4.35458243e-01 1.25291502e+00 5.88893831e-01 1.19137537e+00
-1.14362800e+00 -8.65942299e-01 1.92730859e-01 -4.71718200e-02
-1.37723482e+00 -3.44605505e-01 9.94825006e-01 -3.81085992e-01
2.11991921e-01 2.69209683e-01 7.63925195e-01 1.70729136e+00
1.09342203e-01 1.06334066e+00 7.46001959e-01 -3.62172335e-01
5.12996078e-01 1.84513345e-01 -8.44843626e-01 -6.74287900e-02
-3.07862192e-01 -4.28042468e-03 -6.18942857e-01 -2.59469628e-01
7.16562569e-01 -3.82093072e-01 -2.71862119e-01 1.04931118e-02
-1.19109786e+00 9.58362937e-01 1.89723730e-01 -3.27004725e-03
-6.87275171e-01 5.26544988e-01 4.04033870e-01 1.57270476e-01
2.71737903e-01 6.65784419e-01 5.83173595e-02 -4.80722696e-01
-1.03631496e+00 5.50027728e-01 7.32553124e-01 1.07639301e+00
2.98509419e-01 5.42084634e-01 -7.55644083e-01 7.80639231e-01
2.97564238e-01 8.01922023e-01 6.55370295e-01 -9.63153481e-01
5.57921886e-01 -1.80428162e-01 2.80858010e-01 -5.76080859e-01
3.04244608e-01 -3.80526602e-01 -9.39703584e-01 1.77433286e-02
4.79969755e-02 -6.67280376e-01 -8.71491075e-01 2.06162906e+00
-8.19784626e-02 4.30425107e-01 3.16420764e-01 9.89931703e-01
6.37623191e-01 1.23540199e+00 3.32327366e-01 -2.72415012e-01
9.09757793e-01 -9.88409042e-01 -9.88195956e-01 -3.51715446e-01
-1.51837632e-01 -8.03163469e-01 1.23831296e+00 2.85348654e-01
-1.28288877e+00 -6.73918128e-01 -9.51777339e-01 4.55561131e-01
6.49373159e-02 -2.44073629e-01 2.84424484e-01 6.82082117e-01
-9.08508956e-01 1.61502361e-01 -5.00067055e-01 1.57196611e-01
3.29659611e-01 -2.20935848e-02 -1.35527756e-02 9.93195772e-02
-1.59423876e+00 4.07337159e-01 3.62524360e-01 -5.77792041e-02
-1.41025198e+00 -6.60557330e-01 -8.59684944e-01 2.46276245e-01
4.80479263e-02 -7.84671009e-01 1.67748070e+00 -1.42104769e+00
-2.00159502e+00 1.65830374e-01 -3.82218845e-02 -6.47831440e-01
5.72022080e-01 -2.66962469e-01 -4.99723613e-01 3.14016759e-01
1.67003334e-01 1.24494421e+00 1.21673667e+00 -1.51627123e+00
-4.12761390e-01 4.27321374e-01 3.32733914e-02 2.75695741e-01
-1.05255902e-01 -7.11900666e-02 -1.73594113e-02 -1.04744637e+00
-5.08181334e-01 -9.34681296e-01 -1.18431374e-01 -3.47415984e-01
-4.94892895e-01 3.53324562e-02 5.51178455e-01 -6.60561442e-01
1.05899501e+00 -2.12121487e+00 4.46978584e-02 -1.07289128e-01
-3.75389069e-01 2.28198260e-01 -2.68306762e-01 6.33170485e-01
4.36667539e-03 2.88380444e-01 -3.69715780e-01 -5.07879436e-01
2.48748094e-01 9.68601555e-02 -8.47665191e-01 -5.88113666e-02
5.95328271e-01 8.55721533e-01 -1.20617223e+00 -5.84831953e-01
-1.35179251e-01 7.56284475e-01 -5.79989374e-01 6.74051344e-01
-7.48648167e-01 6.62017047e-01 -4.28262949e-01 5.29622495e-01
4.74655628e-01 -6.96831988e-03 -5.33050038e-02 3.88484210e-01
3.10144663e-01 2.36960202e-01 -1.02775145e+00 1.85226166e+00
-6.65758431e-01 6.78693891e-01 -1.16738491e-01 -7.51008987e-01
1.16894722e+00 1.00805366e+00 1.88040018e-01 -3.90245795e-01
-7.76651576e-02 1.03960775e-01 -2.44989708e-01 -3.17717314e-01
4.56772238e-01 -3.92753303e-01 -3.91517609e-01 5.48003852e-01
-3.95875275e-02 -6.00648463e-01 1.13434042e-03 5.11647202e-02
8.04334462e-01 1.97345287e-01 -5.84508367e-02 3.75473261e-01
4.06855822e-01 -2.13981673e-01 4.64617580e-01 9.60965574e-01
-3.53127941e-02 1.28183854e+00 4.23410416e-01 6.27945662e-02
-1.19082010e+00 -1.27733946e+00 2.54161179e-01 8.61502826e-01
-8.27448517e-02 -1.87036455e-01 -1.01776266e+00 -4.15656090e-01
-4.61728066e-01 1.01470280e+00 -4.66863006e-01 -2.75431812e-01
-4.18900490e-01 -1.97571933e-01 6.92844391e-01 3.82730067e-01
4.68079358e-01 -1.70954764e+00 -3.35891545e-01 7.03431845e-01
-5.12957871e-01 -1.02461839e+00 -7.25020289e-01 -2.04382211e-01
-4.67863470e-01 -2.16234267e-01 -1.24255335e+00 -9.87590134e-01
3.76781136e-01 -2.83945143e-01 1.26862824e+00 -5.78611135e-01
8.97901133e-02 4.47584867e-01 -7.28736043e-01 -8.32111716e-01
-1.01701963e+00 2.48420671e-01 -7.60124326e-02 3.11402887e-01
-2.57525593e-02 -8.05459261e-01 -4.23808604e-01 -1.43057015e-02
-1.23258996e+00 2.31395289e-01 8.14350903e-01 9.45050359e-01
6.57676458e-01 -2.31509835e-01 1.29008281e+00 -7.03027427e-01
1.12888038e+00 -6.36558056e-01 -1.53726533e-01 2.49254018e-01
-1.87177971e-01 7.57555738e-02 1.25129938e+00 -8.77769351e-01
-1.00115836e+00 1.51926056e-01 -2.68265784e-01 -6.03996575e-01
-4.60096329e-01 4.32529390e-01 -1.93670452e-01 5.24689317e-01
5.93304634e-01 6.14151180e-01 -6.00512896e-04 -1.26266494e-01
4.79979903e-01 9.57148314e-01 7.68042564e-01 -6.64147973e-01
9.26288128e-01 1.17722906e-01 -4.15853828e-01 -5.38337827e-01
-8.59331369e-01 6.75031310e-03 -5.96603677e-02 -3.86977881e-01
8.98863137e-01 -1.10620391e+00 -3.75263661e-01 2.82034367e-01
-1.38842404e+00 -2.72055745e-01 -5.40785134e-01 5.04642963e-01
-1.08902669e+00 -1.00007810e-01 -4.27577555e-01 -1.12502360e+00
-5.53443074e-01 -9.58871484e-01 1.29129374e+00 4.59387392e-01
-5.05186737e-01 -7.18302906e-01 3.49771917e-01 5.74077293e-02
5.33728898e-01 5.74474573e-01 6.27551138e-01 -6.43315852e-01
-4.69326496e-01 -3.30428690e-01 2.94437438e-01 5.46918392e-01
1.07479975e-01 -2.34334633e-01 -1.08559418e+00 -1.41476586e-01
-1.34807646e-01 -6.37000084e-01 2.33630240e-01 2.78482676e-01
1.09460104e+00 -7.11876333e-01 9.49569121e-02 2.13738009e-01
1.17274332e+00 3.51733655e-01 9.04319644e-01 1.62706897e-01
3.80365551e-01 4.07712847e-01 6.60824716e-01 5.47185898e-01
3.73102203e-02 5.72281599e-01 4.95400786e-01 -6.45112917e-02
-1.26686618e-01 -9.51489687e-01 6.74968421e-01 7.78087854e-01
3.90832901e-01 -5.45921981e-01 -5.79946458e-01 7.79418230e-01
-1.68647981e+00 -1.27131724e+00 4.75287527e-01 1.95750606e+00
1.11629641e+00 1.33972093e-01 2.69276649e-01 1.88589007e-01
1.10688567e+00 1.36141360e-01 -5.73727548e-01 -6.41792238e-01
3.71214487e-02 2.71496356e-01 -5.49187511e-03 3.26620698e-01
-7.07138658e-01 7.10133791e-01 6.84843922e+00 8.33526015e-01
-1.23432839e+00 -7.84688890e-02 5.88991940e-01 -1.17589645e-01
-6.11852050e-01 -2.02557877e-01 -4.47090238e-01 8.54268789e-01
1.19254410e+00 -5.26992202e-01 5.51185250e-01 8.82581770e-01
3.71023089e-01 4.47531015e-01 -1.16437411e+00 9.56788719e-01
1.28224179e-01 -1.21449745e+00 5.08290827e-01 -2.94044048e-01
9.76614296e-01 -4.62148249e-01 3.57492924e-01 5.19881964e-01
2.91352332e-01 -1.25894821e+00 1.10151041e+00 6.56289041e-01
7.96117544e-01 -1.01322234e+00 7.48695731e-01 2.91921169e-01
-8.90771568e-01 9.33148563e-02 -1.97673917e-01 7.07174093e-02
7.30933070e-01 3.60948116e-01 -1.28105915e+00 4.10265326e-01
4.29309040e-01 1.87083751e-01 -1.01601578e-01 1.08168733e+00
-4.14376676e-01 1.03057384e+00 6.39382452e-02 -3.19172204e-01
2.99610555e-01 2.54690032e-02 7.31665134e-01 1.20080161e+00
6.24046504e-01 -3.23135376e-01 5.89077137e-02 1.14370871e+00
-2.16325015e-01 1.61540806e-02 -5.59726655e-01 -4.18972373e-01
6.30661368e-01 8.44619393e-01 -1.87379077e-01 -2.70446897e-01
-1.70871824e-01 1.00208676e+00 -2.05851361e-01 4.63287592e-01
-1.23034716e+00 -6.83081985e-01 3.50858271e-01 1.07317872e-01
3.85631263e-01 -4.27547796e-03 -9.09852143e-03 -7.88452148e-01
9.87016782e-02 -1.14644682e+00 2.06918288e-02 -1.21109927e+00
-1.44477475e+00 9.04843390e-01 3.52319367e-02 -1.61311185e+00
-1.01677239e+00 1.92711502e-01 -7.93368220e-01 8.81739974e-01
-1.46626163e+00 -1.13738692e+00 -9.78498608e-02 3.74974668e-01
8.32664311e-01 -3.93973380e-01 1.01553667e+00 1.03937060e-01
-3.72525789e-02 7.66950965e-01 5.31096831e-02 1.45817697e-01
8.30133319e-01 -1.24401641e+00 5.44735491e-01 6.38223469e-01
2.65758365e-01 1.10880539e-01 1.19686663e+00 -4.14622664e-01
-9.14221764e-01 -1.29248464e+00 7.17996001e-01 8.71396344e-03
4.94025946e-01 -5.39490640e-01 -8.59680057e-01 5.13156414e-01
5.67479312e-01 -3.23208928e-01 7.45407343e-01 -4.19535875e-01
-3.85954604e-02 -3.67924124e-02 -1.11791730e+00 6.33453250e-01
6.31997943e-01 -5.20374775e-01 -6.47619307e-01 3.19146633e-01
1.02473950e+00 -3.24770540e-01 -6.36598945e-01 1.22239456e-01
4.02915001e-01 -6.95784271e-01 6.78669333e-01 -5.31180561e-01
7.60816932e-01 -3.09933037e-01 -8.23216215e-02 -1.69871891e+00
-9.80881602e-02 -1.20113409e+00 -9.27210599e-02 1.71820188e+00
5.84190011e-01 -2.49857381e-01 7.69762754e-01 3.59228939e-01
-2.82516092e-01 -3.23516369e-01 -8.55668485e-01 -9.41922069e-01
1.60702750e-01 -3.63209426e-01 9.95759904e-01 5.20663321e-01
-7.73788914e-02 5.34805417e-01 -8.99703205e-01 2.17389390e-01
3.72221589e-01 3.97063009e-02 7.31442392e-01 -9.28713441e-01
-5.61404407e-01 -2.05809519e-01 -2.52389908e-01 -9.62444305e-01
3.92152369e-01 -6.40227199e-01 5.84680557e-01 -1.51644063e+00
-6.17656708e-02 -2.52895802e-01 3.58323082e-02 2.18817770e-01
-1.77907616e-01 3.17637682e-01 3.14357162e-01 4.91501503e-02
-5.58587492e-01 9.39697683e-01 1.41922474e+00 -3.86948258e-01
-2.77033001e-01 2.02086776e-01 -7.85345078e-01 2.38585860e-01
9.01497543e-01 -7.11212695e-01 -6.09610736e-01 -2.77145714e-01
1.90085649e-01 4.51831400e-01 1.62291065e-01 -1.23425734e+00
5.54857366e-02 -2.22305059e-01 1.66583702e-01 -2.36490503e-01
5.98392248e-01 -6.17553174e-01 3.97959292e-01 1.87286705e-01
-8.34664702e-01 -1.21485695e-01 5.03453799e-02 6.14195585e-01
-7.22638369e-01 -4.78699178e-01 5.65982044e-01 -2.35995397e-01
-7.81745389e-02 2.91463614e-01 -6.10844612e-01 3.43332589e-01
9.35770512e-01 5.65386973e-02 3.39741632e-02 -1.19243038e+00
-6.39648974e-01 4.54795025e-02 1.58992156e-01 5.16979694e-01
6.29618704e-01 -1.84459281e+00 -1.18248117e+00 -3.71441767e-02
3.75601090e-02 2.36682102e-01 1.72069997e-01 -7.96008762e-03
-2.52061695e-01 2.74944574e-01 -2.95840681e-01 -5.19802272e-01
-7.38754034e-01 4.48553920e-01 1.58885688e-01 -1.69191033e-01
-2.77426094e-01 5.69195986e-01 8.31398442e-02 -1.13516916e-02
3.60030472e-01 3.23806368e-02 -2.15907767e-01 -8.48890990e-02
5.92649460e-01 -1.28646120e-01 -3.30480874e-01 -4.26933110e-01
1.42842531e-01 2.30251595e-01 2.27534515e-03 -7.22181857e-01
1.18259454e+00 6.02182373e-03 4.07505959e-01 5.80457747e-01
1.03477407e+00 4.15538289e-02 -1.41487622e+00 2.41001397e-01
-5.74454010e-01 -3.31875503e-01 -3.19467455e-01 -7.41521239e-01
-7.20800996e-01 7.90440083e-01 4.16751951e-01 6.82873070e-01
1.06676638e+00 -4.40218486e-02 1.05265677e+00 -8.20873529e-02
2.11701736e-01 -9.50547814e-01 3.99019599e-01 3.13215971e-01
1.19447911e+00 -9.48267400e-01 -6.58911884e-01 -5.67287132e-02
-1.16881454e+00 9.14192975e-01 5.90969980e-01 -1.19020261e-01
5.24871014e-02 1.19316556e-01 2.31372252e-01 4.69642073e-01
-8.03341448e-01 7.12829307e-02 1.23917051e-01 1.06629479e+00
3.42047036e-01 7.58830607e-02 4.17863093e-02 7.23125875e-01
-6.87304616e-01 7.64773116e-02 7.87466228e-01 6.72356844e-01
-2.41585717e-01 -1.36368501e+00 -5.55308342e-01 8.87463018e-02
-5.87566674e-01 -6.36095628e-02 -3.41201037e-01 4.32624012e-01
-1.61132067e-01 1.04964328e+00 1.55538931e-01 -2.13356122e-01
7.83163905e-02 2.95943320e-01 1.11353494e-01 -7.05721438e-01
-3.62073928e-01 -1.12339063e-02 -1.04850754e-01 6.48212880e-02
-4.47035074e-01 -5.57888210e-01 -1.09071624e+00 1.41304046e-01
-1.58198729e-01 7.28717327e-01 6.67915404e-01 6.38580739e-01
3.16118538e-01 7.63564289e-01 1.00089526e+00 -8.98868322e-01
-7.42271304e-01 -1.03069496e+00 -5.68393767e-01 5.40177941e-01
3.86731088e-01 -1.48788601e-01 -5.13677716e-01 4.24779356e-01] | [15.271735191345215, 4.922688961029053] |
9eaf10ac-5630-47c2-91c8-a01261371c87 | deep-lifelong-cross-modal-hashing | 2304.13357 | null | https://arxiv.org/abs/2304.13357v1 | https://arxiv.org/pdf/2304.13357v1.pdf | Deep Lifelong Cross-modal Hashing | Hashing methods have made significant progress in cross-modal retrieval tasks with fast query speed and low storage cost. Among them, deep learning-based hashing achieves better performance on large-scale data due to its excellent extraction and representation ability for nonlinear heterogeneous features. However, there are still two main challenges in catastrophic forgetting when data with new categories arrive continuously, and time-consuming for non-continuous hashing retrieval to retrain for updating. To this end, we, in this paper, propose a novel deep lifelong cross-modal hashing to achieve lifelong hashing retrieval instead of re-training hash function repeatedly when new data arrive. Specifically, we design lifelong learning strategy to update hash functions by directly training the incremental data instead of retraining new hash functions using all the accumulated data, which significantly reduce training time. Then, we propose lifelong hashing loss to enable original hash codes participate in lifelong learning but remain invariant, and further preserve the similarity and dis-similarity among original and incremental hash codes to maintain performance. Additionally, considering distribution heterogeneity when new data arriving continuously, we introduce multi-label semantic similarity to supervise hash learning, and it has been proven that the similarity improves performance with detailed analysis. Experimental results on benchmark datasets show that the proposed methods achieves comparative performance comparing with recent state-of-the-art cross-modal hashing methods, and it yields substantial average increments over 20\% in retrieval accuracy and almost reduces over 80\% training time when new data arrives continuously. | ['Jiancheng Lv', 'Weisheng Li', 'Bochuan Zheng', 'Hanqi Li', 'Liming Xu'] | 2023-04-26 | null | null | null | null | ['semantic-textual-similarity', 'semantic-similarity'] | ['natural-language-processing', 'natural-language-processing'] | [-4.23990726e-01 -6.40046358e-01 -4.18943375e-01 -2.40409374e-01
-1.31076491e+00 -4.92314160e-01 2.55606174e-01 5.39807916e-01
-6.58234358e-01 6.52608097e-01 1.65182009e-01 2.86550611e-01
-3.26771349e-01 -9.34271872e-01 -6.69760823e-01 -9.50062692e-01
-2.39825130e-01 7.24052429e-01 5.93847990e-01 -1.72886893e-01
2.94581831e-01 2.99325764e-01 -1.80597639e+00 5.54780215e-02
5.67024231e-01 1.00933790e+00 1.33281097e-01 1.92113310e-01
9.55395177e-02 3.11424822e-01 -3.28253835e-01 -2.25070417e-01
1.38805464e-01 3.84297743e-02 -5.75353682e-01 -5.78006506e-01
3.45862567e-01 -6.67571485e-01 -1.01921427e+00 7.84889340e-01
1.00152183e+00 2.79088259e-01 4.89626467e-01 -1.26761675e+00
-1.00662220e+00 5.64028859e-01 -5.72654188e-01 2.96035767e-01
4.35737967e-01 1.53361047e-02 1.19359720e+00 -1.24750173e+00
2.21277371e-01 1.14333904e+00 9.49434280e-01 3.73250812e-01
-8.79864693e-01 -1.15939176e+00 -2.26725161e-01 5.79563320e-01
-2.19820189e+00 -2.47475490e-01 4.64140981e-01 1.00837380e-01
8.85967076e-01 1.37944268e-02 5.17272413e-01 5.67296207e-01
2.01663330e-01 1.05946219e+00 5.72835505e-01 -2.98766270e-02
3.63547504e-02 2.57393196e-02 1.43727288e-01 7.54846811e-01
2.19988123e-01 1.00173742e-01 -6.32496953e-01 -4.67953891e-01
2.88841903e-01 5.28765142e-01 -2.26498768e-01 -2.88293421e-01
-1.18569696e+00 9.92830515e-01 8.09664130e-01 3.24617833e-01
-2.73633480e-01 3.28279883e-01 8.26045573e-01 4.35580194e-01
-5.52094206e-02 -8.27167928e-02 -1.61540836e-01 1.98737420e-02
-1.10779583e+00 4.53506976e-01 4.48621988e-01 1.10104609e+00
1.20925009e+00 -2.82766610e-01 -3.03760618e-01 9.57224250e-01
2.21778139e-01 1.00138509e+00 1.05035102e+00 -5.05153418e-01
3.01375210e-01 4.20898318e-01 -2.40652010e-01 -1.23311973e+00
-3.51362675e-01 -3.07081342e-01 -1.15510380e+00 -6.83565855e-01
-2.31260449e-01 5.66423833e-01 -8.27904284e-01 1.66930580e+00
3.66269588e-01 2.68382549e-01 3.19523700e-02 7.29545414e-01
7.67305613e-01 1.07514489e+00 3.03873252e-02 -3.58301252e-01
1.36435390e+00 -8.95242691e-01 -6.97764337e-01 2.86889493e-01
6.80852890e-01 -8.45148325e-01 1.23595619e+00 2.49003708e-01
-1.04208171e+00 -8.53078246e-01 -1.20434856e+00 -3.54371905e-01
-5.59189200e-01 -3.85701507e-01 5.29404879e-01 2.62402445e-01
-1.01115704e+00 2.95316607e-01 -5.83354414e-01 -7.88903609e-02
9.99508277e-02 6.14495933e-01 -2.45198488e-01 -5.44805884e-01
-1.88966608e+00 4.19205159e-01 7.34026790e-01 -1.45861119e-01
-1.00776243e+00 -7.52678037e-01 -8.82908642e-01 2.21971482e-01
-4.15667780e-02 -4.52921242e-01 1.07092869e+00 1.60052598e-01
-8.25227559e-01 5.63571870e-01 -4.80386764e-02 -3.17058295e-01
9.38585773e-02 -1.80629566e-01 -5.62901199e-01 2.35302329e-01
2.55245924e-01 8.53747129e-01 8.32185268e-01 -1.02773786e+00
-8.01447928e-01 -4.21939403e-01 -2.52123654e-01 3.49911690e-01
-1.02459729e+00 -5.29854059e-01 -8.74835014e-01 -6.75045371e-01
1.73397928e-01 -9.31997657e-01 2.93182492e-01 3.88767719e-02
1.01582691e-01 -6.13449693e-01 1.17785823e+00 -2.03397065e-01
1.80508387e+00 -2.44482899e+00 -5.99864346e-04 1.36443451e-01
8.95180628e-02 2.52561808e-01 -1.01154909e-01 5.86576521e-01
4.24855232e-01 -2.91176051e-01 -2.38807067e-01 -4.38367218e-01
2.52022713e-01 3.79342318e-01 -5.91273189e-01 5.24471939e-01
-3.46451253e-01 8.43457162e-01 -8.91833365e-01 -9.62129712e-01
4.75818320e-04 7.29203403e-01 -5.46200812e-01 1.30346656e-01
2.35749722e-01 -3.08680266e-01 -2.41828308e-01 8.98617029e-01
9.69333708e-01 -3.59200448e-01 -2.23482817e-01 -3.77851695e-01
2.34304607e-01 7.32499361e-02 -1.11068118e+00 1.88263142e+00
-5.12255728e-01 9.53186080e-02 -2.96335429e-01 -8.76774073e-01
9.01790142e-01 1.65508881e-01 4.43675727e-01 -1.15349805e+00
-1.75411507e-01 5.39464593e-01 -6.10433340e-01 -1.85152844e-01
8.51109982e-01 -3.29270899e-01 -5.76911390e-01 6.38873518e-01
9.31072515e-03 -5.20611219e-02 1.22153349e-01 2.37581789e-01
8.28569829e-01 -5.62349617e-01 -6.13170378e-02 4.71172892e-02
8.61118555e-01 -2.22264662e-01 5.31907022e-01 7.50245869e-01
-2.47153580e-01 6.03666067e-01 -3.72563988e-01 -5.27026594e-01
-9.85423088e-01 -1.22915876e+00 -3.72998416e-01 1.29944968e+00
6.99262261e-01 -4.31561679e-01 -2.50429899e-01 -6.37061477e-01
5.83257437e-01 1.27973884e-01 -4.73809212e-01 -8.56159389e-01
-7.83553243e-01 -8.10516715e-01 8.97334576e-01 5.10627568e-01
7.36986160e-01 -1.20376277e+00 -2.29986519e-01 3.36998969e-01
-2.18393102e-01 -6.53155327e-01 -9.47059810e-01 7.53636807e-02
-7.29163051e-01 -7.24366963e-01 -9.18644607e-01 -1.21534812e+00
2.71253943e-01 7.05727279e-01 7.84297705e-01 4.13855940e-01
-5.36719382e-01 3.90796304e-01 -4.21238899e-01 3.07629406e-01
7.88740441e-02 6.00166619e-01 2.95138776e-01 -3.49901915e-01
5.28691590e-01 -4.49897110e-01 -1.08886337e+00 4.63433444e-01
-1.37085354e+00 -7.12193489e-01 7.75626838e-01 1.20958662e+00
8.18168640e-01 3.93410504e-01 7.36412883e-01 -4.01283622e-01
4.25097704e-01 -7.50658751e-01 -2.76676744e-01 3.96810263e-01
-1.08050430e+00 1.42183751e-01 7.37899065e-01 -4.11570340e-01
-5.46002209e-01 -3.84883463e-01 -4.04618718e-02 -5.53215921e-01
5.26063561e-01 4.80663568e-01 1.70279175e-01 -6.42439872e-02
2.68373817e-01 9.54843938e-01 8.37317407e-02 -3.22644740e-01
2.86047071e-01 7.92551756e-01 5.02528667e-01 -5.96516550e-01
1.04024553e+00 5.66220999e-01 -1.87229514e-01 -1.47979990e-01
-7.22896039e-01 -8.43204618e-01 -2.34071076e-01 2.43310139e-01
4.42884624e-01 -1.16957653e+00 -8.85738552e-01 6.44799769e-01
-6.68873310e-01 4.28919569e-02 -1.98328346e-01 4.44308966e-01
-3.71115535e-01 6.29191637e-01 -9.54989851e-01 -3.73666674e-01
-8.55075896e-01 -1.05343747e+00 1.39765406e+00 3.21175694e-01
2.94605345e-01 -9.23000872e-01 2.90299922e-01 8.24453458e-02
5.51365554e-01 -4.36458081e-01 9.81053054e-01 -3.19753855e-01
-7.22737730e-01 -4.65873420e-01 -4.21210438e-01 1.36492506e-01
1.54276177e-01 -4.75122780e-01 -7.20480680e-01 -1.11658752e+00
-3.34922105e-01 -8.19620550e-01 1.01965559e+00 -2.67467529e-01
1.25497580e+00 -2.79247582e-01 -4.70681369e-01 6.55018866e-01
1.52825630e+00 2.03575671e-01 5.44397593e-01 4.77424711e-01
5.40249050e-01 2.31309995e-01 1.10289383e+00 7.33282089e-01
7.67839611e-01 7.30618656e-01 1.77144542e-01 7.63271898e-02
-1.45669475e-01 -6.14062309e-01 1.73982844e-01 1.35666502e+00
6.83955789e-01 -8.37696269e-02 -7.06284404e-01 8.60169768e-01
-1.79414058e+00 -9.16543067e-01 4.79637206e-01 2.39033365e+00
1.18699193e+00 7.51721710e-02 1.40009755e-02 3.02441150e-01
7.26342380e-01 2.04524562e-01 -8.50153625e-01 1.64856091e-01
-1.04746409e-01 2.52072692e-01 5.39960682e-01 4.58231091e-01
-1.11033320e+00 9.14461672e-01 5.89128590e+00 1.27971232e+00
-1.24742162e+00 1.70206338e-01 2.27862194e-01 -6.80150241e-02
-4.21039373e-01 -1.31433710e-01 -1.36250961e+00 7.50612795e-01
7.42437720e-01 -3.45386147e-01 3.21301639e-01 8.02377820e-01
-5.41597307e-01 2.26357624e-01 -8.23724389e-01 1.32542133e+00
3.82611930e-01 -1.00477946e+00 5.38501203e-01 -1.68499932e-01
4.56340939e-01 -4.14729454e-02 4.97160733e-01 9.43602920e-01
-6.79214001e-02 -6.39762044e-01 3.47759634e-01 3.76497388e-01
8.95766377e-01 -1.05910242e+00 8.54192734e-01 1.06166124e-01
-1.70091271e+00 -3.06634665e-01 -7.10210681e-01 4.49951738e-01
6.53458536e-02 4.78692740e-01 -7.17534721e-01 4.91104543e-01
9.76068318e-01 7.85794973e-01 -6.84463501e-01 1.07718158e+00
4.14759308e-01 1.13291398e-01 -3.61157835e-01 2.40774695e-02
3.49400878e-01 3.47409934e-01 6.58225361e-03 1.17810559e+00
5.88057637e-01 2.83339825e-02 4.05180812e-01 2.48174481e-02
-1.69308007e-01 1.19265877e-01 -5.22175372e-01 2.79696465e-01
1.02537942e+00 9.95850623e-01 -2.57614911e-01 -5.04183412e-01
-3.02783877e-01 1.13223910e+00 3.22947025e-01 7.97495693e-02
-9.49459136e-01 -9.18062150e-01 4.41783041e-01 3.99561860e-02
4.62498873e-01 -1.09196536e-01 4.00066644e-01 -9.66637373e-01
1.19345702e-01 -5.41964233e-01 1.08119404e+00 -5.35232365e-01
-1.48370588e+00 4.63549972e-01 -9.98752564e-02 -1.40618408e+00
-6.67975545e-02 6.10413775e-03 -2.59200279e-02 3.44399899e-01
-1.93926716e+00 -1.15327203e+00 -4.87585455e-01 9.36121643e-01
4.39795017e-01 -1.54969230e-01 7.63930798e-01 9.67040181e-01
-2.48370364e-01 1.40670657e+00 5.32240629e-01 -1.34703770e-01
1.24823236e+00 -1.01728165e+00 5.89299537e-02 1.46791011e-01
-1.68221936e-01 9.18399990e-01 2.59681553e-01 -5.06412148e-01
-1.79878211e+00 -1.06581438e+00 8.90946448e-01 1.09322079e-01
5.35228014e-01 -3.31193238e-01 -1.35789084e+00 2.87447810e-01
-1.64381683e-01 3.45357984e-01 7.85272717e-01 -1.88611314e-01
-7.99825311e-01 -8.05153430e-01 -1.13313198e+00 1.09820709e-01
9.13846850e-01 -9.73258436e-01 -6.60527110e-01 3.39529753e-01
1.03617775e+00 -2.60780603e-01 -1.12217462e+00 7.17278421e-01
8.26309204e-01 -7.22593188e-01 1.19877243e+00 2.91347001e-02
-2.08433151e-01 -6.43986523e-01 -4.39137161e-01 -7.38706470e-01
-5.75268924e-01 -3.49717259e-01 -3.30690742e-01 1.14076889e+00
6.49146810e-02 -6.50289536e-01 7.82272160e-01 1.02613211e-01
-9.28009152e-02 -7.30098546e-01 -1.06816995e+00 -8.33594382e-01
6.71259090e-02 2.74779826e-01 7.53213286e-01 8.92279029e-01
-1.43339336e-01 1.10244550e-01 -5.20749688e-01 1.21990293e-01
7.15529740e-01 4.66472566e-01 6.31469548e-01 -1.04949355e+00
-1.54008314e-01 -6.09643161e-02 -5.14953613e-01 -1.30535436e+00
1.10687181e-01 -9.18298423e-01 4.51611131e-02 -1.11452293e+00
6.43737376e-01 -9.96170044e-01 -8.58195901e-01 6.40135288e-01
-4.27912593e-01 6.38603032e-01 -4.53074910e-02 8.00183833e-01
-1.15201986e+00 1.12689984e+00 1.09874308e+00 -4.82124835e-01
8.59360676e-03 -3.53946596e-01 -4.67267394e-01 -1.12101734e-01
3.51692200e-01 -6.44956470e-01 -6.03608787e-01 -3.87781888e-01
2.01904580e-01 5.05542532e-02 4.47222702e-02 -1.25153136e+00
6.89026177e-01 3.52780253e-01 2.66559452e-01 -1.04701543e+00
4.43799347e-01 -8.65161538e-01 -6.87435688e-03 8.68903399e-01
-2.36569166e-01 3.73077601e-01 4.86465469e-02 6.84631526e-01
-6.76569998e-01 -1.48059651e-01 7.10113466e-01 2.19471380e-01
-7.09240556e-01 7.72134542e-01 1.06858052e-01 4.57958207e-02
9.82217610e-01 -1.46408956e-02 -2.45270237e-01 -2.93811411e-01
-4.63712156e-01 5.65348029e-01 5.30335665e-01 5.19828022e-01
9.15121257e-01 -1.84870732e+00 -4.61254984e-01 3.51017445e-01
4.33685958e-01 8.41545239e-02 6.29398406e-01 4.91835028e-01
-4.29010421e-01 6.45954788e-01 1.64025184e-02 -6.82278454e-01
-1.03551173e+00 9.64095175e-01 -1.77754700e-01 -3.58988971e-01
-4.87239569e-01 8.22586656e-01 9.08960849e-02 -4.17297423e-01
4.83850032e-01 1.50783867e-01 2.02255458e-01 3.17276090e-01
7.14727879e-01 3.16555232e-01 2.06887096e-01 -4.79103506e-01
-3.87970090e-01 9.62972403e-01 -7.80544579e-01 1.01584829e-01
1.07705784e+00 -3.97441626e-01 -1.87760085e-01 4.70745742e-01
1.95849991e+00 -3.20269972e-01 -8.23504090e-01 -7.27935553e-01
-2.29605407e-01 -5.42711854e-01 -3.59270275e-02 -3.03151816e-01
-1.04336226e+00 8.16146374e-01 1.01407039e+00 6.97163418e-02
1.26234865e+00 -1.44928293e-02 1.91895366e+00 7.53922343e-01
6.71678245e-01 -1.06520534e+00 4.78092402e-01 4.66582179e-01
5.71374893e-01 -1.21394753e+00 1.92517236e-01 1.67145357e-01
-2.79012054e-01 7.92965055e-01 6.87710881e-01 -4.92638461e-02
8.63340676e-01 -1.21255241e-01 -1.16009749e-01 -2.32366338e-01
-6.37855828e-01 6.90579191e-02 9.40064266e-02 2.50304610e-01
1.92931723e-02 -1.62716240e-01 -3.48987281e-01 1.37137875e-01
-4.21908982e-02 -2.04034477e-01 -2.15952918e-01 1.15637207e+00
-7.13082135e-01 -1.15809751e+00 -2.55171418e-01 2.81772822e-01
-1.47399127e-01 -2.62991667e-01 2.86382705e-01 8.03184986e-01
-1.09487874e-02 3.36986363e-01 1.40389144e-01 -4.72013444e-01
2.28060246e-01 4.43182662e-02 2.04528615e-01 -3.35129946e-01
-3.51345927e-01 -1.00723460e-01 -7.08313763e-01 -3.22986811e-01
-4.86896001e-02 -5.81582308e-01 -1.58060157e+00 -4.08414155e-01
-4.91742551e-01 7.08137155e-01 4.17295337e-01 4.44886774e-01
4.62592721e-01 -1.00896880e-01 1.18179965e+00 -5.06002545e-01
-9.12102044e-01 -7.29900599e-01 -5.97768664e-01 5.08353710e-01
5.51055372e-01 -8.76774907e-01 -6.74300313e-01 -3.30116779e-01] | [11.299572944641113, 0.9745357036590576] |
b7519900-b823-4dc8-9f1b-271d3917a394 | knowledge-extraction-from-texts-based-on | null | null | https://aclanthology.org/2022.naacl-industry.33 | https://aclanthology.org/2022.naacl-industry.33.pdf | Knowledge Extraction From Texts Based on Wikidata | This paper presents an effort within our company of developing knowledge extraction pipeline for English, which can be further used for constructing an entreprise-specific knowledge base. We present a system consisting of entity detection and linking, coreference resolution, and relation extraction based on the Wikidata schema. We highlight existing challenges of knowledge extraction by evaluating the deployed pipeline on real-world data. We also make available a database, which can serve as a new resource for sentential relation extraction, and we underline the importance of having balanced data for training classification models. | ['Frédéric Herledan', 'Johannes Heinecke', 'Anastasia Shimorina'] | null | null | null | null | naacl-acl-2022-7 | ['coreference-resolution'] | ['natural-language-processing'] | [-9.01236087e-02 9.78891850e-01 -4.67582643e-01 -3.14482033e-01
-5.80431104e-01 -6.48800194e-01 7.71589160e-01 8.08146119e-01
-7.07485437e-01 1.18573105e+00 5.58863997e-01 -2.34542847e-01
-5.40228248e-01 -9.44471836e-01 -4.16186601e-01 1.27868712e-01
-2.16166690e-01 1.14539909e+00 5.94109654e-01 -7.03235865e-01
-2.94832438e-02 4.97365773e-01 -1.35483718e+00 4.90194738e-01
5.64875603e-01 6.35070264e-01 -1.31233841e-01 3.40192467e-01
-4.33388352e-01 1.09714329e+00 -6.33197665e-01 -1.00211370e+00
-1.97013855e-01 4.55420241e-02 -1.72817457e+00 -8.24052274e-01
1.47038490e-01 1.79775000e-01 -4.24386293e-01 6.64962232e-01
4.80803579e-01 -8.48125890e-02 3.15615892e-01 -1.21502149e+00
-1.94665372e-01 1.40606499e+00 2.12525606e-01 4.94419634e-01
7.76738942e-01 -6.63614094e-01 1.28791070e+00 -3.90876591e-01
1.50703144e+00 7.89068520e-01 6.25873506e-01 3.82029951e-01
-6.97515309e-01 -6.47727609e-01 -3.64093930e-01 6.80251896e-01
-1.42296350e+00 -8.70959401e-01 3.16502154e-01 -2.40917236e-01
1.82691646e+00 4.42766994e-01 5.40728629e-01 9.11911607e-01
-2.93627501e-01 5.40757120e-01 6.65301800e-01 -9.56994951e-01
-3.57080102e-01 4.91010487e-01 4.81424183e-01 5.93177438e-01
6.71031117e-01 6.13447167e-02 -7.59413242e-01 -2.43407294e-01
4.89639014e-01 -9.32294846e-01 -4.44455355e-01 -2.92886317e-01
-9.85248685e-01 3.47506523e-01 1.72137946e-01 6.26693845e-01
-4.74548012e-01 -5.13405263e-01 5.40647566e-01 3.84965211e-01
1.26567751e-01 8.57793331e-01 -9.68890727e-01 -2.37199157e-01
-6.28358543e-01 4.21665132e-01 1.65669060e+00 1.32995129e+00
4.98391062e-01 -8.05388093e-01 2.59138793e-01 8.38402331e-01
1.93317875e-01 -2.89849509e-02 2.83781409e-01 -9.36976910e-01
6.75791264e-01 9.30098832e-01 2.86578029e-01 -8.63623857e-01
-6.05894864e-01 -1.12571701e-01 -4.40008380e-02 -3.87389421e-01
4.46367919e-01 -1.97234631e-01 -6.06675111e-02 1.38555503e+00
5.42642653e-01 1.35580182e-01 8.05243015e-01 4.47808176e-01
1.70287871e+00 4.43629315e-03 3.01684052e-01 -4.95163232e-01
1.56655312e+00 -6.72526240e-01 -1.34057343e+00 1.77110657e-01
1.02293599e+00 -8.19460988e-01 -3.87872681e-02 1.54669777e-01
-1.18505418e+00 1.02217426e-03 -1.00956464e+00 -2.24303365e-01
-1.09444296e+00 -2.17577815e-01 1.21467090e+00 4.55927432e-01
-6.35373116e-01 7.74940670e-01 -7.28411913e-01 -9.39829826e-01
8.50474089e-02 3.87891173e-01 -9.84216630e-01 2.37582117e-01
-1.73335576e+00 1.65160155e+00 1.44839144e+00 -1.69913307e-01
-8.65654573e-02 -6.64004385e-01 -9.89015520e-01 -2.45738372e-01
7.82625914e-01 -5.18796682e-01 1.12405872e+00 -3.38896886e-02
-1.04335320e+00 1.28023720e+00 2.58400626e-02 -7.71828651e-01
1.75896939e-02 -2.65436828e-01 -1.32059872e+00 1.44977227e-01
2.36384585e-01 2.27099255e-01 -2.48657122e-01 -8.32306027e-01
-1.03812814e+00 -3.18791419e-01 1.50758564e-01 2.28976309e-01
-2.32992530e-01 6.82726979e-01 -5.89650273e-01 -2.32041806e-01
-2.21851557e-01 -4.26507771e-01 2.14716420e-01 -7.99091578e-01
-5.63051105e-01 -5.13143420e-01 5.22661030e-01 -1.05399692e+00
1.73623157e+00 -1.55430627e+00 1.00642882e-01 2.82124281e-01
1.58714592e-01 5.10618389e-01 2.64888465e-01 9.06425834e-01
-3.34245473e-01 1.63437366e-01 2.64861256e-01 2.65535951e-01
-8.20944011e-02 4.65244889e-01 -1.29063964e-01 -2.21300259e-01
3.65851283e-01 1.04485738e+00 -1.07898021e+00 -8.42418313e-01
-7.09457248e-02 2.09222883e-01 -1.10384002e-01 5.25206253e-02
-2.80675124e-02 -1.16354413e-01 -5.00760257e-01 4.07834053e-01
1.55500427e-01 8.58930573e-02 9.54081595e-01 -3.53704661e-01
-2.95237452e-01 8.92570436e-01 -1.32222581e+00 1.58021069e+00
-1.42242476e-01 4.38234121e-01 1.76565647e-02 -7.68138051e-01
8.92247677e-01 7.22584903e-01 3.07103992e-01 -3.06892306e-01
7.79017732e-02 4.18051243e-01 -8.33342373e-02 -7.53020048e-01
9.00436580e-01 2.60848522e-01 -8.43628347e-02 1.01943381e-01
6.31488085e-01 2.05856070e-01 6.27354980e-01 5.36642313e-01
1.29043114e+00 3.98154348e-01 9.87214804e-01 -1.05963819e-01
5.38344562e-01 5.38832903e-01 7.00180590e-01 1.28851175e-01
2.49802351e-01 -2.39138395e-01 4.51159209e-01 -3.29522520e-01
-7.01241136e-01 -7.15639114e-01 -5.30526876e-01 6.34056985e-01
-1.89754695e-01 -1.05777872e+00 -3.12213540e-01 -9.93948519e-01
-2.27035247e-02 8.23833168e-01 -5.58301151e-01 1.95365325e-01
-6.23372436e-01 -6.54979765e-01 1.11980116e+00 4.02193606e-01
2.45555103e-01 -1.12167656e+00 -5.28207183e-01 3.84369433e-01
-6.37814403e-01 -1.42890358e+00 3.78103554e-01 3.97629589e-01
-2.84663975e-01 -1.79944992e+00 2.74471939e-01 -8.58001709e-01
-4.52764751e-03 -4.55847144e-01 1.54701805e+00 1.43622190e-01
-1.83499366e-01 3.09755474e-01 -5.05684555e-01 -6.25475705e-01
-4.34334904e-01 6.37967110e-01 -1.29718497e-01 -8.67431939e-01
1.17742622e+00 -4.20509279e-01 7.20806792e-02 7.65280128e-02
-6.00570083e-01 -2.53626227e-01 1.70216411e-01 5.02186835e-01
1.90482289e-01 7.43166059e-02 7.06237912e-01 -1.41268528e+00
6.80841386e-01 -7.10454941e-01 -3.14899683e-01 6.13503695e-01
-6.86920047e-01 4.30694222e-02 -1.69394508e-01 2.50127107e-01
-1.48855579e+00 -4.23075370e-02 -3.70826870e-01 4.43862945e-01
-3.29601347e-01 9.25801396e-01 -3.52619380e-01 7.60005713e-02
8.23615074e-01 -4.43442613e-01 -2.30717212e-01 -7.73947775e-01
6.62557364e-01 8.60259891e-01 8.63087356e-01 -7.13436961e-01
2.69395143e-01 1.43469825e-01 -1.23381883e-01 -6.71756506e-01
-7.59600997e-01 -7.94330239e-01 -1.16147542e+00 2.94867873e-01
5.02674937e-01 -9.36927736e-01 -7.00314760e-01 -7.18336254e-02
-1.18856478e+00 5.75715713e-02 -4.42967653e-01 4.04374242e-01
-2.55631089e-01 3.30581069e-01 -6.45189524e-01 -3.41858119e-01
-4.21427906e-01 -2.31777906e-01 3.66212994e-01 2.67141312e-01
-7.79694796e-01 -1.31177580e+00 4.46667671e-01 5.78273058e-01
-4.42950428e-02 2.58334786e-01 6.72797978e-01 -1.37138569e+00
-1.04941003e-01 -2.90807784e-01 -1.65481806e-01 -3.30927074e-01
-4.32054736e-02 2.15161353e-01 -5.96665084e-01 2.45062113e-01
-8.65779161e-01 -4.50859278e-01 5.16369462e-01 -3.72590184e-01
1.81458652e-01 -2.35735044e-01 -9.82012093e-01 3.96472603e-01
1.31616890e+00 5.65493330e-02 8.00220907e-01 1.08396327e+00
3.32648814e-01 9.82546568e-01 8.66175473e-01 9.45840478e-02
9.24290478e-01 9.29097354e-01 -1.36772335e-01 2.84321040e-01
-2.24099860e-01 -3.04021448e-01 -4.09698009e-01 6.49532676e-01
-5.45843124e-01 1.57899365e-01 -1.20362020e+00 8.35688412e-01
-1.95875764e+00 -1.02227843e+00 -4.87309515e-01 1.70624804e+00
1.47368133e+00 1.13361310e-02 4.81744446e-02 1.00369394e-01
4.02358204e-01 -4.17841762e-01 2.02827081e-01 -2.77512193e-01
-2.43041664e-01 7.09521413e-01 5.32370448e-01 6.23423100e-01
-1.20197344e+00 1.38357484e+00 7.16999054e+00 4.77485418e-01
-5.84032059e-01 5.15636839e-02 -5.00110745e-01 2.03639850e-01
7.60549977e-02 2.53277719e-01 -1.37676656e+00 -3.78808822e-03
1.31032348e+00 -4.60354835e-01 1.69427395e-02 5.92349827e-01
-3.72900486e-01 -7.11340681e-02 -1.06017900e+00 4.63402599e-01
-2.59798735e-01 -1.70992422e+00 -2.28904322e-01 -8.74788761e-02
5.49218245e-02 1.90333202e-01 -9.85823035e-01 4.61915404e-01
7.68760622e-01 -8.28976810e-01 8.94025937e-02 4.92030054e-01
5.75990915e-01 -6.39692485e-01 1.11678290e+00 1.19572021e-01
-1.17670321e+00 3.70056927e-01 -1.80194005e-01 7.96512067e-02
2.42716894e-01 3.23956460e-01 -1.13397491e+00 1.39131236e+00
7.98396707e-01 8.03240240e-01 -5.75456798e-01 9.86210346e-01
-6.02808952e-01 2.21067891e-01 -4.05007899e-01 8.86221156e-02
-4.54066664e-01 2.48585686e-01 5.25983572e-01 1.68295121e+00
-1.01209052e-01 4.82798576e-01 -6.70477152e-02 3.99223924e-01
-3.39269675e-02 5.47601104e-01 -5.57192445e-01 -1.74090669e-01
1.13816345e+00 1.45283008e+00 -3.65643263e-01 -4.68455911e-01
-4.72540855e-01 6.15362763e-01 7.86997795e-01 5.25679030e-02
-3.18599135e-01 -9.46653545e-01 4.13565844e-01 -1.15280353e-01
3.29326659e-01 1.35251597e-01 2.25870118e-01 -1.22901285e+00
-1.57238618e-01 -7.84587979e-01 1.14749742e+00 -5.19160628e-01
-9.04334009e-01 7.02484488e-01 4.49983150e-01 -5.32729566e-01
-6.34011984e-01 -5.27039528e-01 -1.85711518e-01 7.87665188e-01
-1.59509850e+00 -1.44727564e+00 -8.15975368e-02 5.49194515e-01
-3.90222967e-01 -1.90167055e-01 1.44170284e+00 6.82232499e-01
-6.82598412e-01 5.67461610e-01 -5.24959862e-01 7.31361449e-01
9.47866440e-01 -1.19986892e+00 3.36470485e-01 4.84866589e-01
2.80446947e-01 1.02936637e+00 6.72907710e-01 -8.97468388e-01
-9.38180923e-01 -9.19496655e-01 1.79831553e+00 -9.16379154e-01
1.02311110e+00 4.58035171e-02 -9.68819141e-01 1.21747351e+00
5.00579298e-01 -3.52302909e-01 1.09586370e+00 8.03495228e-01
-5.16468942e-01 2.44796827e-01 -1.23528218e+00 4.06000763e-02
1.28695869e+00 -3.40598583e-01 -1.45458281e+00 1.78884178e-01
4.47913557e-01 -6.99747860e-01 -1.68308687e+00 6.12332165e-01
3.96117359e-01 -4.27813023e-01 8.35850120e-01 -1.02252781e+00
2.14066669e-01 -1.33007005e-01 8.84535909e-02 -1.18681073e+00
-1.31242678e-01 -5.53816259e-01 -6.14531517e-01 1.83740473e+00
9.73571420e-01 -6.50549412e-01 5.08742213e-01 7.22472191e-01
1.12860046e-01 -1.53349549e-01 -9.56324220e-01 -6.08609736e-01
-8.46443698e-02 -3.90109986e-01 5.78985989e-01 1.43174958e+00
1.11693943e+00 9.63767290e-01 7.82384500e-02 2.54516244e-01
2.80365527e-01 -9.01298150e-02 5.70842505e-01 -1.58035517e+00
-2.10664332e-01 -5.81307337e-02 -5.40072739e-01 -3.52734536e-01
2.59053946e-01 -1.00221312e+00 -4.43728805e-01 -1.68697023e+00
6.53901398e-02 -5.71847677e-01 -1.47417054e-01 8.44275117e-01
1.03599906e-01 9.57697332e-02 -2.41372943e-01 1.60972849e-01
-6.61125302e-01 -7.45085329e-02 5.82728744e-01 3.27824146e-01
-1.68155506e-01 -5.05535305e-01 -9.42633867e-01 6.56921685e-01
5.57736814e-01 -4.80060279e-01 1.63874514e-02 -8.00042748e-02
5.68752944e-01 -1.34756371e-01 -1.67283654e-01 -6.93807125e-01
5.39165795e-01 -1.08573744e-02 6.61677569e-02 -5.42023659e-01
2.48535901e-01 -7.46155739e-01 3.84002894e-01 1.73521027e-01
-3.00696909e-01 -3.06552052e-01 4.19798315e-01 9.38025638e-02
-4.52905506e-01 -3.91379625e-01 1.84873790e-01 -2.57111132e-01
-1.05655551e+00 -2.37300128e-01 4.96688634e-02 2.92525649e-01
1.08645093e+00 2.06935778e-01 -7.03652740e-01 6.81033581e-02
-1.00274003e+00 4.90914911e-01 2.68180698e-01 5.73787868e-01
2.79267162e-01 -1.18137658e+00 -8.38268220e-01 -1.57482490e-01
4.48725998e-01 -3.96681041e-01 -2.48314306e-01 6.39251292e-01
-4.47214723e-01 7.88862765e-01 -3.84838849e-01 2.85663784e-01
-1.72754896e+00 5.24460912e-01 1.33714765e-01 -6.20424330e-01
-5.82581460e-01 7.19849169e-01 -9.64870334e-01 -6.94347978e-01
2.17944637e-01 1.33183315e-01 -1.10721946e+00 4.76456106e-01
7.57141173e-01 2.54311353e-01 5.12433350e-01 -5.68869114e-01
-8.10063243e-01 -6.79648370e-02 -3.01410258e-01 -2.55822074e-02
1.63096797e+00 -1.70181572e-01 -6.81454837e-01 3.49035889e-01
6.08445227e-01 3.20990026e-01 -8.73830020e-02 -4.06399429e-01
8.24303806e-01 5.63161150e-02 -5.10626435e-02 -1.14151371e+00
-5.84361672e-01 1.20895691e-01 1.46487458e-02 4.39574182e-01
6.65848911e-01 3.65657836e-01 4.72655982e-01 6.67328715e-01
6.70301020e-01 -1.27556181e+00 -9.46402609e-01 7.34077513e-01
7.61693120e-01 -9.87413406e-01 1.44696906e-01 -1.14166999e+00
-4.66584951e-01 1.06664574e+00 4.61057961e-01 3.06334615e-01
8.95745635e-01 6.78851366e-01 1.89360827e-01 -6.60159349e-01
-1.01448917e+00 -7.26972878e-01 3.22338015e-01 9.73743379e-01
9.19815719e-01 -3.35866660e-02 -8.74229968e-01 9.50714529e-01
-4.55428541e-01 2.66485006e-01 3.19296986e-01 1.08204353e+00
-1.14471428e-01 -1.77991962e+00 7.96175674e-02 3.19149375e-01
-8.29924822e-01 -4.91165936e-01 -9.04659867e-01 1.14672554e+00
3.13698560e-01 8.80878806e-01 -1.07258044e-01 -2.87915289e-01
7.35742390e-01 4.35298413e-01 6.63285375e-01 -9.37508941e-01
-7.53228247e-01 -6.28243327e-01 1.49753964e+00 -4.55810815e-01
-8.69582474e-01 -7.51100719e-01 -1.34928560e+00 -3.42294157e-01
-4.51494396e-01 7.56481349e-01 2.35285193e-01 1.03525198e+00
4.61776674e-01 4.10946667e-01 -2.23529488e-01 -7.91646726e-03
1.94624186e-01 -1.02447891e+00 -3.92388999e-01 4.39650059e-01
-3.82482141e-01 -7.59950221e-01 1.36398703e-01 7.97377154e-02] | [9.345810890197754, 8.725204467773438] |
23b10eeb-cfe7-4efc-92b6-3574522f0cfa | graphsha-synthesizing-harder-samples-for | 2306.09612 | null | https://arxiv.org/abs/2306.09612v1 | https://arxiv.org/pdf/2306.09612v1.pdf | GraphSHA: Synthesizing Harder Samples for Class-Imbalanced Node Classification | Class imbalance is the phenomenon that some classes have much fewer instances than others, which is ubiquitous in real-world graph-structured scenarios. Recent studies find that off-the-shelf Graph Neural Networks (GNNs) would under-represent minor class samples. We investigate this phenomenon and discover that the subspaces of minor classes being squeezed by those of the major ones in the latent space is the main cause of this failure. We are naturally inspired to enlarge the decision boundaries of minor classes and propose a general framework GraphSHA by Synthesizing HArder minor samples. Furthermore, to avoid the enlarged minor boundary violating the subspaces of neighbor classes, we also propose a module called SemiMixup to transmit enlarged boundary information to the interior of the minor classes while blocking information propagation from minor classes to neighbor classes. Empirically, GraphSHA shows its effectiveness in enlarging the decision boundaries of minor classes, as it outperforms various baseline methods in class-imbalanced node classification with different GNN backbone encoders over seven public benchmark datasets. Code is avilable at https://github.com/wenzhilics/GraphSHA. | ['Jian-Huang Lai', 'Hui Xiong', 'Chang-Dong Wang', 'Wen-Zhi Li'] | 2023-06-16 | null | null | null | null | ['node-classification', 'blocking'] | ['graphs', 'natural-language-processing'] | [ 6.77925125e-02 5.61903596e-01 -6.84663117e-01 -3.27738047e-01
6.28616586e-02 -5.63089132e-01 3.00512075e-01 1.76538289e-01
5.61941974e-02 7.93303907e-01 1.40917644e-01 -5.17458260e-01
-1.16038002e-01 -1.26168251e+00 -7.37973154e-01 -8.08822453e-01
-4.11180332e-02 6.06001258e-01 1.33931026e-01 -2.25528955e-01
3.91854346e-02 3.79791141e-01 -1.20731604e+00 3.25697571e-01
9.54760075e-01 6.00740910e-01 -4.45700675e-01 3.09358686e-01
-2.40444720e-01 5.46498120e-01 -8.07970703e-01 -4.80226129e-01
4.07897681e-01 -3.77960980e-01 -6.82481885e-01 2.52404094e-01
6.23202384e-01 -2.12490380e-01 -7.53780305e-01 1.46222126e+00
4.92190570e-01 -2.61180878e-01 5.02790332e-01 -1.84168255e+00
-7.18297720e-01 9.69101071e-01 -9.77602661e-01 2.81571627e-01
-6.13372810e-02 -7.69405365e-02 1.08434069e+00 -4.13933456e-01
7.97306120e-01 1.25170612e+00 7.44563401e-01 4.80601549e-01
-1.16473103e+00 -1.04945266e+00 4.05876160e-01 3.16095382e-01
-1.55039334e+00 -4.82359499e-01 1.24614942e+00 -1.80064306e-01
6.11314714e-01 5.55107772e-01 5.56484699e-01 1.22032881e+00
1.59543574e-01 7.67052710e-01 6.01652980e-01 -8.07719305e-02
1.68731093e-01 1.78306207e-01 4.05213058e-01 7.04169869e-01
7.27278709e-01 -2.91086018e-01 -4.60892797e-01 -2.66411364e-01
5.30842066e-01 2.24304467e-01 -5.86272717e-01 -8.54767501e-01
-9.17314112e-01 1.03279614e+00 7.66675711e-01 3.05183858e-01
-5.78960031e-02 -7.23601058e-02 4.35268342e-01 4.13780153e-01
7.78014183e-01 1.15621239e-01 -2.39148155e-01 2.99181968e-01
-8.54303956e-01 -1.80806056e-01 8.61559093e-01 1.03569281e+00
7.72709489e-01 -9.63698253e-02 -1.43472910e-01 5.99380076e-01
1.52672350e-01 3.57959829e-02 3.55070412e-01 -6.25383496e-01
7.83271909e-01 1.10634875e+00 -4.01935518e-01 -1.41426754e+00
-4.24354613e-01 -1.10635829e+00 -1.45975184e+00 -1.77598014e-01
4.72856224e-01 4.41986807e-02 -9.48400378e-01 1.72005093e+00
5.59120655e-01 -2.47280113e-02 -1.04631364e-01 8.29038441e-01
9.58923578e-01 5.71209967e-01 -2.80846000e-01 -1.39596701e-01
1.01448536e+00 -1.14129651e+00 -6.26899838e-01 -2.83214569e-01
8.95087659e-01 -2.59693086e-01 9.58529949e-01 3.22152317e-01
-6.53694391e-01 -3.94197494e-01 -1.23209023e+00 1.17689386e-01
-3.86416912e-01 5.24615459e-02 9.07540023e-01 7.79853046e-01
-1.01698422e+00 7.28695095e-01 -7.35912919e-01 -2.04936430e-01
8.19024622e-01 1.81767628e-01 -5.14924169e-01 -1.20122164e-01
-1.19020855e+00 1.82040408e-01 3.98921907e-01 3.59210879e-01
-7.45394111e-01 -6.60383284e-01 -8.82675827e-01 8.71604756e-02
5.27360082e-01 -3.35366726e-01 3.22850645e-01 -1.11705959e+00
-7.41446495e-01 7.51058757e-01 2.19527140e-01 -2.38018617e-01
5.48367143e-01 3.19146037e-01 -4.07314211e-01 -9.49157923e-02
5.35097532e-02 5.58319390e-01 7.55204737e-01 -1.12180328e+00
-1.92281500e-01 -6.55047297e-01 -3.24682402e-03 1.23550802e-01
-8.51441205e-01 -6.59763336e-01 -4.05008137e-01 -6.28849745e-01
5.26265740e-01 -7.48319924e-01 -1.45200357e-01 -4.49333563e-02
-1.14611077e+00 -3.53840947e-01 9.42727149e-01 -4.72842664e-01
1.46886361e+00 -2.16849494e+00 2.42422909e-01 3.93503010e-01
1.10549951e+00 1.42937854e-01 -3.21713358e-01 4.08023030e-01
-3.81495863e-01 2.62660533e-01 -1.56272694e-01 -8.94561410e-02
6.81745112e-02 3.34509611e-01 -2.29942694e-01 8.80337298e-01
-6.90011233e-02 5.91609061e-01 -8.10986996e-01 -1.78900033e-01
-9.88601893e-02 4.21816379e-01 -6.53814733e-01 -1.35571808e-02
4.95428406e-02 1.84518173e-01 -2.93956637e-01 6.11330450e-01
1.23419702e+00 -4.88032907e-01 3.63501489e-01 -2.65472442e-01
3.88672352e-01 4.13229078e-01 -1.29284513e+00 1.44888818e+00
2.96484083e-01 6.24171317e-01 1.95350014e-02 -1.35158885e+00
9.24174249e-01 -9.48573798e-02 3.46448511e-01 -4.15620834e-01
5.46040833e-02 1.69397324e-01 2.43082508e-01 -1.54959664e-01
2.16109395e-01 1.34739012e-01 2.42233858e-01 2.57382989e-01
1.25042751e-01 3.70985419e-01 4.07468766e-01 5.10710418e-01
1.27737343e+00 -3.58420193e-01 1.47358263e-02 -3.99207473e-01
2.83420384e-01 -3.11019152e-01 8.98564398e-01 7.19710946e-01
-5.34022093e-01 4.43748087e-01 1.12886262e+00 -7.35980093e-01
-8.22687089e-01 -9.72767889e-01 -1.10396512e-01 8.61591876e-01
2.79753923e-01 -6.28840387e-01 -7.22798824e-01 -1.19075656e+00
1.21377878e-01 4.89284307e-01 -7.55676866e-01 -5.32495975e-01
-3.16342592e-01 -1.14886737e+00 6.02475882e-01 2.75035679e-01
4.02082086e-01 -8.34770203e-01 4.70248237e-02 -9.23165977e-02
-2.82331824e-01 -7.81266510e-01 -3.41422647e-01 4.15765047e-01
-8.20193529e-01 -1.34157586e+00 -6.26232743e-01 -6.78358853e-01
1.15522611e+00 2.86969125e-01 1.33751476e+00 3.76571685e-01
-2.13011354e-01 -2.55116433e-01 -3.47109944e-01 -1.09735675e-01
-4.62402433e-01 5.29785514e-01 -1.18458480e-01 -4.76343604e-03
6.32575035e-01 -6.94048166e-01 -7.28314638e-01 3.64426941e-01
-9.59532738e-01 1.93563133e-01 3.29365641e-01 5.50783753e-01
3.71943176e-01 5.47201991e-01 5.50599158e-01 -1.17540491e+00
4.40518528e-01 -8.40675712e-01 -3.71730000e-01 9.51492041e-02
-6.00754976e-01 -5.36359996e-02 5.59094369e-01 -3.02334607e-01
-2.86609560e-01 -4.51157093e-01 -1.16844639e-01 -4.16055441e-01
-1.20883942e-01 1.86167568e-01 -5.32293499e-01 1.25492990e-01
6.14083409e-01 1.08609181e-02 -9.32282060e-02 -3.68522257e-01
4.20524133e-03 7.38739669e-01 -9.09226909e-02 -2.45045096e-01
8.54241848e-01 4.97249335e-01 -1.02357477e-01 -7.33862758e-01
-7.64886439e-01 -3.04733932e-01 -4.04810041e-01 -2.20498100e-01
4.43208456e-01 -8.54919553e-01 -3.42787027e-01 4.49873149e-01
-9.97703791e-01 -4.09229249e-01 -2.74744242e-01 -3.29939462e-02
7.72074834e-02 5.17519236e-01 -7.82310605e-01 -3.41973126e-01
-3.77404600e-01 -1.16203499e+00 9.43924367e-01 2.21831471e-01
-4.76195700e-02 -8.30060720e-01 -1.77220050e-02 3.56417269e-01
2.34330177e-01 1.50901154e-01 1.19299126e+00 -8.78581405e-01
-3.80828470e-01 -2.53323525e-01 -3.22346807e-01 2.35633686e-01
2.17225567e-01 3.17803547e-02 -8.35084856e-01 -9.12800312e-01
-2.93030500e-01 -6.60620630e-02 1.08334005e+00 2.91448683e-01
1.37210941e+00 -4.50069904e-01 -7.42124319e-01 9.07855570e-01
1.32828319e+00 -1.24156490e-01 6.90650165e-01 1.47637948e-01
1.05041587e+00 6.25072837e-01 -7.66319335e-02 1.56795204e-01
4.76273388e-01 4.26080525e-01 7.08471477e-01 -3.69356662e-01
-2.21835971e-01 -2.16352493e-01 2.97215939e-01 9.81292069e-01
3.48271728e-01 -8.39520216e-01 -8.59514713e-01 5.72497308e-01
-1.59372914e+00 -5.23122132e-01 -2.61255920e-01 2.09114695e+00
6.97776914e-01 4.63944465e-01 -2.72022299e-02 4.33473408e-01
1.03372288e+00 4.55502331e-01 -6.69015288e-01 -9.00059268e-02
-2.41856515e-01 -2.95309901e-01 5.09685576e-01 2.66843915e-01
-1.13487220e+00 6.51862681e-01 4.81652403e+00 1.07455444e+00
-8.80998850e-01 -4.84986156e-02 1.07697415e+00 9.21019837e-02
-6.43510342e-01 -8.81083980e-02 -8.99039268e-01 4.47868288e-01
5.75770080e-01 2.83454627e-01 2.78051317e-01 5.66363633e-01
-2.12065399e-01 3.05817753e-01 -1.11036372e+00 6.27867699e-01
6.63428977e-02 -1.27054393e+00 2.50799388e-01 2.42308989e-01
7.34911382e-01 2.06761658e-01 -1.46012664e-01 4.14145350e-01
2.20015407e-01 -9.10222411e-01 5.09823799e-01 5.24439253e-02
6.97818041e-01 -7.84113526e-01 7.67164111e-01 3.65131378e-01
-9.27671194e-01 1.35040078e-02 -7.97982216e-01 -7.42057525e-03
-4.25663978e-01 1.20490575e+00 -6.31787121e-01 4.81152922e-01
6.66905046e-01 7.69077897e-01 -8.70948136e-01 8.96450758e-01
-1.30741954e-01 6.54198349e-01 -2.67271399e-01 2.78841943e-01
7.06553236e-02 -3.75169843e-01 8.18511605e-01 7.69196868e-01
1.84301615e-01 -4.39504713e-01 1.41866952e-01 9.58730578e-01
-4.35965300e-01 -1.53413806e-02 -7.48516858e-01 -1.73759922e-01
3.86156440e-01 1.23008728e+00 -1.15337574e+00 -3.05798322e-01
-2.75774926e-01 9.16248858e-01 4.62712139e-01 4.70971286e-01
-9.33036804e-01 -7.94564843e-01 6.03582978e-01 2.38888681e-01
1.06649242e-01 2.02294037e-01 -2.02775355e-02 -1.27598464e+00
1.85038343e-01 -1.14378369e+00 6.36828840e-01 -3.46369117e-01
-1.32798088e+00 6.98649704e-01 -4.59169835e-01 -1.20875955e+00
4.38383043e-01 -5.35000265e-01 -5.35536766e-01 3.94496650e-01
-1.31391346e+00 -8.48704934e-01 -5.24680972e-01 4.15291041e-01
3.22467804e-01 -1.19721346e-01 6.05027735e-01 6.69781625e-01
-9.60660815e-01 8.44405055e-01 7.14864805e-02 4.80836838e-01
5.87739110e-01 -1.15295148e+00 3.12791914e-01 8.36751044e-01
3.20135742e-01 4.40410227e-01 6.48454785e-01 -8.68461549e-01
-1.12432134e+00 -1.16175818e+00 7.24494517e-01 -4.82602939e-02
5.35447299e-01 -8.07678342e-01 -9.04099584e-01 6.44274473e-01
1.32090062e-01 2.96964347e-01 7.19084680e-01 7.93098137e-02
-5.84314108e-01 -3.63883436e-01 -1.00722313e+00 5.33027768e-01
1.29921091e+00 -3.45767766e-01 1.42555051e-02 5.39986253e-01
7.28216767e-01 -2.42547557e-01 -5.41647613e-01 7.37799942e-01
2.38030970e-01 -1.12784457e+00 8.78503084e-01 -7.02591896e-01
4.97185349e-01 -3.37029755e-01 6.58906028e-02 -1.49136209e+00
-2.45868996e-01 -4.08090413e-01 -3.51709425e-01 1.22153080e+00
4.01664734e-01 -7.51998305e-01 1.36949706e+00 -1.49385691e-01
-6.20381348e-02 -6.11508131e-01 -9.60184276e-01 -4.85640079e-01
2.02604264e-01 5.71827637e-03 7.98555076e-01 1.34330082e+00
-1.73060298e-01 3.61941755e-01 -1.43968865e-01 1.39476016e-01
8.69455278e-01 1.67814210e-01 6.79790914e-01 -1.32730198e+00
-2.07145557e-01 -4.99411613e-01 -5.84284246e-01 -9.53110635e-01
1.69935182e-01 -1.28734505e+00 -3.16252768e-01 -1.44785106e+00
4.71458644e-01 -7.58363962e-01 -5.12040138e-01 6.76932216e-01
3.17733013e-03 4.65829670e-01 -4.93020862e-02 1.28569081e-01
-6.00510955e-01 6.45942450e-01 1.50556004e+00 -5.32492220e-01
-1.47551671e-01 7.52933547e-02 -9.30317879e-01 6.65234864e-01
8.84862781e-01 -6.67410493e-01 -5.84232807e-01 -4.14977759e-01
5.78692794e-01 -1.08762696e-01 2.71964967e-01 -1.04184496e+00
1.96625199e-02 1.60361573e-01 4.79928881e-01 -7.21247375e-01
-2.10878506e-01 -7.75206268e-01 2.59266496e-01 6.24684036e-01
-4.12722468e-01 -1.76303357e-01 -4.53894492e-03 6.48414969e-01
-1.21487185e-01 -4.30364981e-02 5.72075903e-01 4.47678156e-02
-2.01335177e-01 4.88898933e-01 -1.06628768e-01 2.64848411e-01
9.77887273e-01 -1.65477127e-01 -8.35019946e-01 -3.26527536e-01
-5.03490865e-01 2.93205649e-01 3.65172774e-01 5.28104842e-01
3.62289429e-01 -1.33650613e+00 -7.09269047e-01 7.20331430e-01
1.03021473e-01 1.40070423e-01 3.70811075e-01 8.70195448e-01
-6.52387679e-01 2.85763890e-01 -1.61552161e-01 -6.30380154e-01
-1.04471266e+00 5.20673871e-01 5.74724674e-01 -3.26070428e-01
-6.31923497e-01 1.14587915e+00 3.55186075e-01 -9.46721613e-01
4.92154270e-01 -1.33700639e-01 -2.36434564e-01 2.35233858e-01
1.68158919e-01 4.81695056e-01 2.24328160e-01 -5.57367265e-01
-4.39151108e-01 9.05361399e-03 -2.32642040e-01 8.05177629e-01
1.36600459e+00 -7.58616403e-02 -5.11844695e-01 1.74130157e-01
1.41810524e+00 2.17106596e-01 -9.43136513e-01 1.42105278e-02
-4.30567414e-02 -3.06689829e-01 7.52204563e-03 -3.17033261e-01
-1.69917107e+00 8.38939607e-01 5.61111808e-01 6.63651526e-01
1.16907728e+00 -3.76065560e-02 6.75878286e-01 1.24325015e-01
2.63766944e-01 -9.46486115e-01 -1.05881490e-01 1.93719596e-01
5.89759707e-01 -1.13925326e+00 7.37153515e-02 -8.44846487e-01
-1.46273494e-01 1.08389425e+00 8.23913693e-01 -1.54765546e-01
7.16005743e-01 1.71947330e-01 -2.09066141e-02 -5.51299274e-01
-6.94234550e-01 2.91893274e-01 1.09411091e-01 3.77580911e-01
3.41448992e-01 3.27923149e-01 -2.35117897e-01 4.55209792e-01
-2.69165635e-01 -4.05315131e-01 6.70575857e-01 5.43405414e-01
-6.38186708e-02 -1.12056315e+00 -1.13948464e-01 7.71963000e-01
-3.20180923e-01 -1.20398238e-01 -4.36868757e-01 6.68693185e-01
3.84704381e-01 7.10990250e-01 4.33269978e-01 -5.55510581e-01
-1.96065772e-02 -7.39615411e-02 1.94631174e-01 -4.65034217e-01
-2.42209926e-01 1.25552099e-02 8.07606354e-02 -6.23255670e-01
-1.91605791e-01 -1.79878101e-01 -8.49299729e-01 -5.26859760e-01
-5.24880707e-01 3.20006758e-01 1.92372322e-01 5.78620672e-01
4.73636299e-01 7.81014979e-01 6.01823747e-01 -3.63511950e-01
-5.19398510e-01 -9.65782821e-01 -7.47030854e-01 3.37506741e-01
3.71681094e-01 -5.30264735e-01 -8.10519159e-01 -5.33256114e-01] | [7.315227508544922, 5.999572277069092] |
15993325-1032-4f7f-942c-5df6f3e50893 | ai-bind-improving-binding-predictions-for | 2112.13168 | null | https://arxiv.org/abs/2112.13168v5 | https://arxiv.org/pdf/2112.13168v5.pdf | AI-Bind: Improving Binding Predictions for Novel Protein Targets and Ligands | Identifying novel drug-target interactions (DTI) is a critical and rate limiting step in drug discovery. While deep learning models have been proposed to accelerate the identification process, we show that state-of-the-art models fail to generalize to novel (i.e., never-before-seen) structures. We first unveil the mechanisms responsible for this shortcoming, demonstrating how models rely on shortcuts that leverage the topology of the protein-ligand bipartite network, rather than learning the node features. Then, we introduce AI-Bind, a pipeline that combines network-based sampling strategies with unsupervised pre-training, allowing us to limit the annotation imbalance and improve binding predictions for novel proteins and ligands. We illustrate the value of AI-Bind by predicting drugs and natural compounds with binding affinity to SARS-CoV-2 viral proteins and the associated human proteins. We also validate these predictions via docking simulations and comparison with recent experimental evidence, and step up the process of interpreting machine learning prediction of protein-ligand binding by identifying potential active binding sites on the amino acid sequence. Overall, AI-Bind offers a powerful high-throughput approach to identify drug-target combinations, with the potential of becoming a powerful tool in drug discovery. | ['Michael Sebek', 'Omair Shafi Ahmed', 'Zohair Shafi', 'Robin Walters', 'Giulia Menichetti', 'Albert-László Barabási', 'Tina Eliassi-Rad', 'Rose Yu', 'Deisy Gysi', 'Ayan Chatterjee'] | 2021-12-25 | null | null | null | null | ['unsupervised-pre-training'] | ['methodology'] | [ 5.46483636e-01 -4.21934538e-02 -6.42839909e-01 -2.83915550e-01
-5.73412418e-01 -8.63936841e-01 2.60918498e-01 4.19897646e-01
-2.40873516e-01 1.32498443e+00 -1.82481632e-02 -8.70142341e-01
-3.71770680e-01 -3.62436175e-01 -8.49062979e-01 -7.80266523e-01
-4.74271566e-01 9.60138440e-01 3.59751917e-02 -4.09491025e-02
3.32632773e-02 9.33092713e-01 -8.02596927e-01 4.92247164e-01
8.28316033e-01 1.74556419e-01 7.60048851e-02 2.64110416e-01
1.70780689e-01 2.20710352e-01 -2.39618644e-01 -2.79021502e-01
-6.40176563e-03 -6.14767969e-01 -6.91156328e-01 -5.25661886e-01
1.16307527e-01 -1.11517034e-01 -3.62488404e-02 5.92397034e-01
7.23356009e-01 -2.28718728e-01 8.38297307e-01 -7.67919540e-01
-3.35947156e-01 1.62204891e-01 -4.09177572e-01 3.07222277e-01
3.17694485e-01 4.48950887e-01 1.26923931e+00 -1.14100170e+00
1.16734278e+00 1.00464964e+00 8.18989038e-01 8.26442778e-01
-1.76512015e+00 -7.11691320e-01 -2.35748112e-01 2.90983677e-01
-1.34966338e+00 -3.89917493e-01 2.96394974e-01 -7.44892418e-01
1.56779158e+00 1.81488872e-01 5.89823842e-01 1.22027123e+00
4.22874868e-01 5.37428141e-01 7.10076571e-01 -4.12944555e-02
3.43703002e-01 -3.32128882e-01 -1.06904961e-01 7.23966241e-01
3.68264049e-01 2.60276407e-01 -6.19204938e-01 -1.00700331e+00
4.44457769e-01 2.76574999e-01 -3.31327111e-01 -6.13164544e-01
-8.90538812e-01 9.17873800e-01 2.88922518e-01 4.25065756e-02
-5.58290541e-01 -1.44451216e-01 3.99203211e-01 -8.75657648e-02
2.89481550e-01 8.48536015e-01 -1.13697696e+00 1.57407194e-01
-7.95816123e-01 2.66607940e-01 7.59993911e-01 1.94543600e-01
6.34555161e-01 -3.10145587e-01 6.61867931e-02 5.90453446e-01
3.32270086e-01 3.96539718e-02 -6.30400702e-02 -2.69717991e-01
-2.43540198e-01 4.54964429e-01 1.07574530e-01 -5.55357635e-01
-7.99074709e-01 -4.17777687e-01 -5.56805491e-01 1.60661101e-01
4.87688541e-01 -1.65400103e-01 -8.37340474e-01 1.88524306e+00
3.69190097e-01 2.75961787e-01 5.67525215e-02 6.71646535e-01
5.85726321e-01 4.27021593e-01 5.51023006e-01 -6.03761911e-01
1.17672217e+00 -5.82307220e-01 -2.70149231e-01 1.79965466e-01
8.43096077e-01 -4.87985820e-01 6.27610862e-01 4.14614499e-01
-7.38309324e-01 4.85747755e-02 -9.19328153e-01 4.66336995e-01
-3.77534837e-01 -2.92281985e-01 8.83086443e-01 4.61162239e-01
-6.14930093e-01 9.62029219e-01 -8.97174716e-01 -4.72493857e-01
7.42374957e-01 9.86481071e-01 -4.88652587e-01 5.83053827e-02
-1.15816391e+00 1.07334995e+00 4.23395336e-01 -1.62352055e-01
-1.10983825e+00 -1.04414737e+00 -2.68623620e-01 -5.56909777e-02
2.80652285e-01 -8.84942949e-01 8.90056849e-01 -7.25016296e-01
-1.17260325e+00 6.86991036e-01 -4.03893828e-01 -5.25451064e-01
-1.99797992e-02 1.22102894e-01 -2.36552879e-01 7.36357570e-02
-7.35202581e-02 5.85543811e-01 1.76226944e-01 -8.80377889e-01
-3.28725636e-01 -4.58037913e-01 -1.80398762e-01 1.22366123e-01
1.92808628e-01 9.82587561e-02 1.35255635e-01 -2.75824010e-01
-3.84156436e-01 -9.14752901e-01 -6.84435725e-01 -2.14211792e-01
-4.16996241e-01 -3.19318324e-01 5.94228745e-01 -3.36217314e-01
8.88720155e-01 -1.65107572e+00 2.61939585e-01 6.49981141e-01
6.18486643e-01 6.79973364e-01 -3.78525704e-01 9.92578745e-01
-5.34883320e-01 1.96188092e-01 -7.22792074e-02 4.43169087e-01
-4.85473037e-01 -5.36915788e-04 -1.45342797e-01 6.86554909e-01
2.76124716e-01 1.14091778e+00 -1.01081693e+00 5.52335568e-02
5.43084703e-02 6.48956716e-01 -7.62611032e-01 -6.12502918e-02
-8.21557164e-01 9.87090528e-01 -6.72599137e-01 6.80559874e-01
6.18130863e-01 -6.22050107e-01 9.69121933e-01 -1.18321404e-01
9.37834010e-02 5.55617929e-01 -4.55721229e-01 1.33053243e+00
1.84716150e-01 2.11754024e-01 -2.67918110e-01 -8.92180920e-01
6.42763913e-01 4.83383685e-01 7.87172437e-01 -3.10416937e-01
-4.46380451e-02 3.98449302e-01 6.61923051e-01 -2.82625377e-01
-5.52388847e-01 -2.12976262e-01 5.99810481e-01 2.44531497e-01
5.47748357e-02 5.14176786e-01 9.55757275e-02 9.01528224e-02
1.27823544e+00 2.17239529e-01 5.55495620e-01 -3.10533106e-01
3.63711357e-01 2.44542912e-01 7.46120632e-01 5.07401586e-01
-1.11181373e-02 1.13273725e-01 6.41398370e-01 -7.99301744e-01
-1.21842742e+00 -7.93005228e-01 -3.63560766e-01 1.09491551e+00
-2.82356381e-01 -6.35357559e-01 -5.71568549e-01 -8.17142189e-01
1.55278787e-01 3.10389131e-01 -6.81181431e-01 -1.38366595e-01
-4.50567871e-01 -1.21924937e+00 5.53278089e-01 3.73422988e-02
-3.94086540e-01 -9.30211484e-01 -1.25983492e-01 5.11217833e-01
2.63242006e-01 -5.45825481e-01 -1.50164932e-01 7.27343142e-01
-6.76779687e-01 -1.73907542e+00 -5.36590219e-01 -7.11723030e-01
5.23476541e-01 -1.79072574e-01 1.00842154e+00 1.69517219e-01
-5.21630406e-01 -3.12149405e-01 9.27168205e-02 -4.18411493e-01
-4.20985341e-01 5.02671488e-02 4.25624818e-01 -4.83066201e-01
7.97653913e-01 -7.74026394e-01 -9.36996520e-01 3.79885435e-01
-6.28422081e-01 1.02956839e-01 6.49707496e-01 1.26970768e+00
7.62199998e-01 -4.92721021e-01 9.67188597e-01 -1.24858141e+00
6.31937087e-01 -6.88941300e-01 -6.53786957e-01 3.04782391e-01
-7.86735237e-01 1.03839949e-01 6.28990352e-01 -3.83329660e-01
-5.85968137e-01 6.60660267e-01 -5.92821538e-01 -3.07137389e-02
-7.68046454e-02 7.79434800e-01 -1.17795408e-01 -3.10664266e-01
8.94286275e-01 1.40904844e-01 1.45545587e-01 -6.04886413e-01
2.18432769e-01 3.64908159e-01 -4.89930287e-02 -4.87120658e-01
4.05889243e-01 3.02140743e-01 4.28178489e-01 -6.34788215e-01
-6.76586151e-01 -5.24953544e-01 -5.67010880e-01 3.10459226e-01
8.19546103e-01 -6.94199383e-01 -1.47945571e+00 4.71583344e-02
-1.21798623e+00 -2.23521560e-01 2.35220283e-01 5.93343198e-01
-4.58309561e-01 3.20866644e-01 -7.78030694e-01 -3.53507519e-01
-4.13257301e-01 -1.46258056e+00 8.20309341e-01 -3.41039971e-02
-5.10908782e-01 -9.83017862e-01 5.99321365e-01 2.22600445e-01
1.21368542e-01 2.64889747e-01 1.44326985e+00 -1.56535387e+00
-6.98244631e-01 1.80411890e-01 -1.57467291e-01 -1.46605447e-01
1.58649668e-01 8.60956684e-03 -6.83582187e-01 -2.81629682e-01
-5.92067182e-01 -3.70841771e-01 9.04845774e-01 5.55827618e-01
8.39365900e-01 -4.66812879e-01 -9.78864193e-01 6.57125771e-01
1.31078196e+00 5.90936065e-01 5.65607548e-01 6.73298985e-02
6.76456511e-01 3.43391091e-01 3.12504500e-01 3.45361084e-01
-1.67822108e-01 8.51085305e-01 4.38331515e-01 -4.42483425e-01
2.98728645e-01 -3.26384902e-01 -7.25311711e-02 3.63796987e-02
-8.88626128e-02 -2.14680135e-01 -1.19739032e+00 1.70359328e-01
-1.81165540e+00 -8.06553423e-01 -1.66466519e-01 2.11462402e+00
1.46333754e+00 2.25573983e-02 3.54361355e-01 -5.78913629e-01
4.81484890e-01 -3.85579795e-01 -9.67810929e-01 -2.46173516e-01
-2.80750662e-01 7.75839746e-01 4.82403517e-01 6.95699871e-01
-8.72484386e-01 1.09694886e+00 7.36822271e+00 9.95604753e-01
-1.11330581e+00 -2.41921425e-01 8.18920076e-01 4.88655418e-02
-3.04406971e-01 1.90535694e-01 -7.97546029e-01 1.55648515e-01
1.02103865e+00 4.79775034e-02 4.32205021e-01 6.70106709e-01
4.40586120e-01 2.34639958e-01 -1.52577007e+00 5.95830917e-01
-3.43153566e-01 -2.16745543e+00 7.34820887e-02 3.15005809e-01
6.54606283e-01 3.43303531e-01 -1.28257185e-01 -3.06589246e-01
5.55451512e-01 -1.41861951e+00 -3.50939959e-01 4.27290767e-01
8.64364386e-01 -7.98826993e-01 5.66466570e-01 2.25215048e-01
-7.14729965e-01 1.59471899e-01 -3.63797814e-01 1.34802591e-02
-6.48316070e-02 4.72942531e-01 -1.53635716e+00 1.78158358e-01
2.25723386e-01 7.89821744e-01 -1.54273659e-01 1.16289771e+00
-5.21445051e-02 5.22391617e-01 -2.82420486e-01 -1.09805562e-01
2.00184539e-01 -9.99668315e-02 4.84441936e-01 1.16469026e+00
-2.72696763e-01 2.53215134e-01 3.91440928e-01 9.16157663e-01
-2.07689166e-01 2.62724668e-01 -6.26768291e-01 -3.46164048e-01
3.76199663e-01 9.50273752e-01 -5.26678145e-01 -1.99167162e-01
-2.64011115e-01 8.11031401e-01 2.25851402e-01 5.44928193e-01
-4.69938785e-01 -1.59781918e-01 1.06939507e+00 1.42811924e-01
2.54549295e-01 2.88605094e-02 1.69769838e-01 -6.76054597e-01
-5.75564682e-01 -1.21822000e+00 3.15153807e-01 -3.56497377e-01
-1.37454176e+00 2.44971886e-01 -3.09913903e-01 -8.24969292e-01
-1.34508491e-01 -8.13729227e-01 -3.33266169e-01 1.04275358e+00
-1.22857726e+00 -1.06860518e+00 6.57061636e-01 2.17443764e-01
3.56943756e-01 -2.67450064e-01 1.25648355e+00 3.45364034e-01
-5.87181866e-01 4.14091855e-01 7.09208071e-01 -3.58747452e-01
6.91055119e-01 -7.73218632e-01 6.14154100e-01 3.75181250e-02
3.77994254e-02 1.15798771e+00 9.07113016e-01 -9.47017372e-01
-1.25037956e+00 -8.76618564e-01 7.27019191e-01 -6.17711425e-01
7.14445293e-01 -5.49965024e-01 -8.85216296e-01 5.52226126e-01
-4.18424122e-02 -2.61014849e-01 1.38114786e+00 3.67852390e-01
-5.23028076e-01 4.53097343e-01 -1.14933574e+00 7.00356543e-01
9.40119267e-01 -4.42144573e-01 -2.21940771e-01 8.52739930e-01
4.71699387e-01 -2.15628847e-01 -6.50377810e-01 6.77419007e-01
7.12933302e-01 -8.68689537e-01 1.25246060e+00 -1.41744840e+00
3.94781381e-02 -3.98412645e-01 2.06718907e-01 -1.11859357e+00
-5.49883485e-01 -7.56569326e-01 6.26903698e-02 3.83698672e-01
1.02778256e+00 -4.94480610e-01 1.14420593e+00 1.98010355e-01
8.52952302e-02 -1.30151474e+00 -9.90109503e-01 -5.68264723e-01
5.71035333e-02 9.85420644e-02 3.19342136e-01 1.07893407e+00
3.88999492e-01 8.77474248e-01 -5.54111183e-01 -9.17535722e-02
2.79507518e-01 -2.01093033e-01 6.46165729e-01 -1.52580011e+00
-6.86806500e-01 -3.99478942e-01 -2.68620461e-01 -9.52958047e-01
1.07725173e-01 -9.81034338e-01 -3.64377439e-01 -1.40331244e+00
7.29091942e-01 -3.29308808e-01 -3.39595586e-01 6.69219673e-01
6.13796189e-02 2.40297735e-01 -3.94143224e-01 5.40400088e-01
-6.08941674e-01 2.05804080e-01 1.02947950e+00 -1.61328703e-01
-3.59796375e-01 4.84655015e-02 -5.60940802e-01 7.10501313e-01
6.94468856e-01 -6.92322195e-01 -2.97663093e-01 3.70647609e-01
5.04715443e-01 -1.32802933e-01 2.32824862e-01 -2.05229193e-01
-3.87571864e-02 -5.69312572e-01 6.64485514e-01 -6.09440863e-01
2.80389011e-01 -7.27810681e-01 5.45059443e-01 8.60510945e-01
-2.89209247e-01 -2.75320888e-01 4.21476126e-01 6.46572948e-01
3.19244981e-01 3.03525813e-02 8.39841902e-01 -1.75397024e-01
-2.46130779e-01 4.93308693e-01 -6.57978773e-01 -2.98078984e-01
9.97638941e-01 -2.60110646e-01 -3.19300950e-01 -2.31398810e-02
-1.14390492e+00 -1.00591742e-01 4.00013894e-01 2.82302629e-02
5.27778208e-01 -7.86593974e-01 -6.07842982e-01 1.11065775e-01
3.72397125e-01 -5.93134880e-01 1.01785166e-02 7.26424754e-01
-7.50921607e-01 8.31273675e-01 -3.26685421e-02 -5.43037176e-01
-1.52101338e+00 7.44359493e-01 7.28983581e-01 -3.84332538e-01
-2.80958861e-01 9.10939395e-01 5.61973274e-01 -3.83463651e-01
2.02138707e-01 4.32583928e-01 -5.66475727e-02 -2.71259695e-01
3.53275061e-01 -1.03845388e-01 1.31727308e-02 -4.82621700e-01
-8.21175098e-01 2.32658640e-01 -5.23755014e-01 5.48096657e-01
1.55793798e+00 3.75123769e-01 -3.97615194e-01 -1.14782959e-01
1.11247706e+00 -6.57710135e-02 -8.75299752e-01 -3.45891118e-01
3.36766720e-01 -2.74111837e-01 -5.63552082e-01 -1.27468669e+00
-4.71869349e-01 5.60189188e-01 7.36186743e-01 -4.88198549e-01
6.08338594e-01 3.00047487e-01 6.53956354e-01 8.10482562e-01
2.68109292e-01 -5.97286761e-01 -5.66995181e-02 4.42641258e-01
5.07108331e-01 -1.06650257e+00 2.15477034e-01 -3.49778324e-01
-2.37361595e-01 9.67816412e-01 3.95686120e-01 2.75031298e-01
5.16400814e-01 1.14916265e-01 -1.68178841e-01 -7.05950737e-01
-1.05110991e+00 4.32202481e-02 3.59080821e-01 7.23618209e-01
8.69141519e-01 -1.07641958e-01 -6.05537713e-01 3.75163138e-01
4.90660995e-01 1.31756574e-01 5.20082079e-02 6.97063982e-01
-4.36615229e-01 -1.84086382e+00 5.08646928e-02 4.74594802e-01
-8.76127183e-01 -7.17021286e-01 -1.05595541e+00 5.67582190e-01
6.82943761e-02 5.55604875e-01 -4.83812571e-01 -3.16507101e-01
1.30199015e-01 2.25263238e-01 4.88832444e-01 -6.70357585e-01
-5.89984775e-01 4.54791188e-01 1.42074943e-01 -2.54355222e-01
-2.31003687e-01 -3.23156893e-01 -1.30490232e+00 -3.14594597e-01
-6.58119321e-01 4.50193375e-01 4.19698924e-01 8.81487191e-01
1.00897181e+00 2.67883420e-01 3.89162898e-01 -6.24366283e-01
-2.33644873e-01 -6.20659828e-01 -3.83817106e-01 1.23105220e-01
3.25606674e-01 -5.48364282e-01 5.57892546e-02 -5.23670809e-03] | [4.949052810668945, 5.664579391479492] |
7d11fbb5-8350-4fe1-8b3c-701c392d2a3a | a-two-stage-method-for-text-line-detection-in | 1802.03345 | null | https://arxiv.org/abs/1802.03345v2 | https://arxiv.org/pdf/1802.03345v2.pdf | A Two-Stage Method for Text Line Detection in Historical Documents | This work presents a two-stage text line detection method for historical documents. Each detected text line is represented by its baseline. In a first stage, a deep neural network called ARU-Net labels pixels to belong to one of the three classes: baseline, separator or other. The separator class marks beginning and end of each text line. The ARU-Net is trainable from scratch with manageably few manually annotated example images (less than 50). This is achieved by utilizing data augmentation strategies. The network predictions are used as input for the second stage which performs a bottom-up clustering to build baselines. The developed method is capable of handling complex layouts as well as curved and arbitrarily oriented text lines. It substantially outperforms current state-of-the-art approaches. For example, for the complex track of the cBAD: ICDAR2017 Competition on Baseline Detection the F-value is increased from 0.859 to 0.922. The framework to train and run the ARU-Net is open source. | ['Tobias Strauß', 'Gundram Leifert', 'Tobias Grüning', 'Johannes Michael', 'Roger Labahn'] | 2018-02-09 | null | null | null | null | ['line-detection'] | ['computer-vision'] | [ 4.13599074e-01 1.02175340e-01 -1.04141608e-01 -2.72951901e-01
-7.89000154e-01 -6.66021645e-01 7.69009888e-01 4.98339742e-01
-3.68073195e-01 4.40138727e-01 4.14175950e-02 -5.02155066e-01
3.47124845e-01 -6.55551374e-01 -7.59954154e-01 -4.44669545e-01
2.79521465e-01 7.05433786e-01 4.72229570e-01 1.54051691e-01
4.67159837e-01 6.55453801e-01 -1.21157503e+00 6.74240530e-01
6.94428325e-01 9.87757444e-01 1.52336378e-02 1.06114459e+00
-3.75280619e-01 6.95521533e-01 -8.27871680e-01 -3.32277536e-01
1.24915309e-01 -3.63514572e-01 -6.16873860e-01 2.80891806e-01
9.58074391e-01 -3.90424699e-01 -3.06091964e-01 9.64624882e-01
4.76986587e-01 -3.59572135e-02 9.59201455e-01 -1.02078092e+00
-4.11095530e-01 9.89230216e-01 -9.68928456e-01 -2.40173656e-02
1.57583728e-01 -2.08298117e-01 1.12938797e+00 -9.74840343e-01
8.34797680e-01 1.03660214e+00 9.27625418e-01 3.71648788e-01
-1.11091065e+00 -2.07549408e-01 1.65664971e-01 -7.95482695e-02
-1.06212723e+00 -2.13714600e-01 6.14857972e-01 -7.29611397e-01
7.57946730e-01 1.78143427e-01 3.04031014e-01 9.67410922e-01
-2.19046362e-02 1.43197727e+00 6.29633069e-01 -8.84529650e-01
1.41568899e-01 1.34798020e-01 5.77870309e-01 7.98718691e-01
2.78696597e-01 -5.76728582e-01 -2.29126811e-01 1.58268407e-01
5.78227639e-01 -1.72647938e-01 -1.50276393e-01 -2.40952507e-01
-1.16413820e+00 6.63412869e-01 3.69671613e-01 4.82027054e-01
-1.04450479e-01 -1.35019124e-01 6.78650498e-01 -1.37518317e-01
2.26419032e-01 4.59292859e-01 -1.83145687e-01 8.65953863e-02
-1.47834122e+00 2.24139363e-01 6.34102702e-01 1.16577029e+00
4.38968211e-01 -1.08674869e-01 -5.56069314e-01 1.11475277e+00
1.77206010e-01 1.55860022e-01 4.86219734e-01 -4.06304508e-01
8.81204009e-01 7.53326118e-01 1.30424321e-01 -7.31324792e-01
-7.21802056e-01 -5.35958230e-01 -8.92367959e-01 2.69415736e-01
6.63561046e-01 -2.11617917e-01 -1.27629292e+00 9.62978244e-01
8.60852897e-02 -3.80480707e-01 -1.85930997e-01 4.83414561e-01
7.68052936e-01 8.72518361e-01 -3.33527148e-01 1.68649897e-01
1.33880258e+00 -1.65674412e+00 -8.48762035e-01 -3.00402224e-01
8.64822447e-01 -1.12117136e+00 1.20340014e+00 5.71660697e-01
-1.01290727e+00 -7.01046705e-01 -1.44868863e+00 -1.45121261e-01
-6.27130985e-01 1.02582610e+00 2.51598656e-01 6.68909192e-01
-9.83180821e-01 4.54331696e-01 -8.09765458e-01 -4.98467922e-01
5.83206952e-01 2.45260641e-01 -1.50016785e-01 -1.31861365e-03
-7.39641488e-01 4.79237229e-01 6.01886570e-01 1.78171977e-01
-5.03397882e-01 -3.30792695e-01 -6.13692999e-01 1.69429302e-01
1.83721483e-01 -4.18326080e-01 1.30337548e+00 -9.37310934e-01
-1.26818275e+00 9.99631166e-01 -2.92553827e-02 -6.77757442e-01
1.25268316e+00 -6.64482892e-01 -3.91708583e-01 1.62463844e-01
1.09881096e-01 8.71244133e-01 6.27542496e-01 -1.45764148e+00
-9.48698044e-01 -1.42208457e-01 -4.45679724e-01 7.32268617e-02
-3.74353260e-01 4.01422121e-02 -1.02479959e+00 -8.82460713e-01
4.23411280e-02 -7.49258280e-01 3.41088064e-02 -4.59976196e-02
-1.22767234e+00 -2.31777877e-01 9.47873235e-01 -7.28052557e-01
1.41172457e+00 -1.72032118e+00 -4.80360866e-01 2.85969496e-01
-5.50606567e-03 3.21686566e-01 7.13846507e-03 4.56152469e-01
-1.46652013e-01 1.31955966e-01 -1.42314613e-01 -5.39869130e-01
3.18647251e-02 -5.13410807e-01 -2.58248389e-01 4.04365152e-01
1.11011058e-01 6.12445354e-01 -5.70733130e-01 -4.46185261e-01
3.06083500e-01 1.99312240e-01 -2.53754914e-01 1.80107176e-01
-2.62326509e-01 -6.87456504e-02 -4.41733114e-02 5.28525531e-01
7.94547021e-01 -1.68627098e-01 6.53345585e-02 -2.63603061e-01
-4.47399110e-01 1.23271041e-01 -1.25163877e+00 1.54188597e+00
-1.07476860e-01 1.25738788e+00 -3.23203355e-01 -6.51928186e-01
1.06527913e+00 7.06418604e-02 2.42449418e-01 -3.80442709e-01
3.05782050e-01 1.65370256e-01 -2.72123585e-03 -2.70128906e-01
7.94320881e-01 7.13898003e-01 3.94039452e-02 2.87575454e-01
2.49900371e-02 3.21941413e-02 7.36884654e-01 2.85896868e-01
1.10137665e+00 2.98934758e-01 7.17381835e-02 -2.60762006e-01
6.92353785e-01 1.50826976e-01 2.62577653e-01 1.16449428e+00
2.83762570e-02 1.00908446e+00 6.50557101e-01 -5.39384305e-01
-1.41539681e+00 -9.25250828e-01 -1.55993760e-01 1.02606940e+00
-1.53133363e-01 -4.86909360e-01 -1.38501108e+00 -7.34466255e-01
-2.92786688e-01 9.11905110e-01 -6.83469117e-01 1.55907601e-01
-6.49151504e-01 -6.22524381e-01 6.76510692e-01 7.63319850e-01
6.67954445e-01 -1.18045163e+00 -4.12036300e-01 4.67456542e-02
8.25348049e-02 -1.04711545e+00 -5.03867686e-01 4.85588908e-01
-6.78840339e-01 -9.29130018e-01 -9.85235810e-01 -1.07874012e+00
1.03333926e+00 -1.09519973e-01 9.44525480e-01 -1.08308427e-01
-4.30921882e-01 -7.32432231e-02 -3.33168626e-01 -5.73957741e-01
-5.62711537e-01 5.03202558e-01 -3.68616343e-01 5.57873473e-02
4.10943657e-01 2.73556173e-01 -5.72542727e-01 2.73773354e-02
-7.44984806e-01 2.39040166e-01 5.69348752e-01 8.52995098e-01
5.37809074e-01 8.65337178e-02 1.38972476e-01 -1.26087368e+00
6.56885624e-01 8.84655267e-02 -8.16599667e-01 4.51830447e-01
-4.52059686e-01 -2.25117370e-01 7.81779230e-01 -3.60722065e-01
-1.00300348e+00 6.33206367e-01 -1.87854424e-01 -5.30026406e-02
-5.64005196e-01 2.99704224e-01 -5.87865412e-02 4.89727110e-01
7.20535576e-01 3.69696468e-02 -6.75224066e-01 -5.13870895e-01
4.03528422e-01 9.23490286e-01 9.66713011e-01 -1.71293125e-01
7.94189155e-01 3.61154079e-01 -3.21393430e-01 -1.02481973e+00
-1.03034389e+00 -6.04131699e-01 -1.26856291e+00 -3.39511275e-01
8.97027373e-01 -5.05703330e-01 -2.09419698e-01 1.00687027e+00
-1.10034013e+00 -5.55067480e-01 -6.97206631e-02 -6.63612271e-03
-2.51569659e-01 3.74591172e-01 -8.67533147e-01 -8.84431601e-01
-5.51003516e-01 -9.61044014e-01 1.17963171e+00 4.58635271e-01
-3.14581782e-01 -8.82246852e-01 -9.14189219e-02 3.73155355e-01
-1.81934446e-01 3.54448915e-01 9.54480946e-01 -1.06912816e+00
-2.31221020e-01 -7.57789493e-01 -2.91253000e-01 1.40085652e-01
-9.02761891e-02 6.54718697e-01 -1.12663007e+00 -1.66281968e-01
-5.63010275e-01 -3.59885573e-01 1.08236635e+00 5.23803830e-01
1.28603482e+00 -1.18538924e-01 -5.32514632e-01 3.73139143e-01
1.36026835e+00 4.88792956e-01 7.11974978e-01 7.93959498e-01
8.15676987e-01 4.84546006e-01 5.89259028e-01 3.32212061e-01
4.60847728e-02 4.43220764e-01 2.12866381e-01 -5.94154775e-01
-1.87141210e-01 -3.07060182e-01 7.20587447e-02 3.92379552e-01
5.64203918e-01 -7.24743128e-01 -1.32018983e+00 5.05630553e-01
-1.82025373e+00 -9.15955961e-01 -7.09652305e-01 2.05177522e+00
6.31602407e-01 9.47026491e-01 2.18503565e-01 6.00971699e-01
1.02370536e+00 -2.51445509e-02 -3.59754324e-01 -4.70197648e-01
-3.92435402e-01 -1.89319491e-01 5.92930555e-01 2.71183282e-01
-1.66534054e+00 1.16281998e+00 5.86720991e+00 8.03183556e-01
-1.08437812e+00 -3.35000038e-01 1.00441146e+00 3.39320332e-01
2.21074715e-01 -3.46487552e-01 -1.16366458e+00 2.32983187e-01
5.90140164e-01 4.43327278e-01 -1.77979589e-01 8.21317315e-01
1.57841250e-01 -3.46439779e-01 -1.09305680e+00 7.52808690e-01
2.75401175e-01 -1.57645428e+00 1.43107653e-01 -2.10793778e-01
1.08692181e+00 1.68997329e-02 -3.51967216e-02 1.40180483e-01
2.89081931e-01 -8.61211061e-01 9.65326309e-01 4.04369652e-01
7.78892279e-01 -8.67352843e-01 7.67513633e-01 3.49754363e-01
-1.04703748e+00 -2.55489349e-03 -3.43069851e-01 4.48886126e-01
2.52653696e-02 5.78527451e-01 -1.12364805e+00 3.31868500e-01
5.10774255e-01 4.32299107e-01 -8.52996528e-01 1.39838684e+00
-2.09675431e-01 7.80833781e-01 -3.25396687e-01 -1.31054665e-03
5.31525075e-01 -2.26545706e-01 5.18313125e-02 1.68469954e+00
2.14151010e-01 -6.14006519e-01 2.74917185e-01 6.75049484e-01
-3.88484478e-01 3.48855555e-01 -4.02237624e-01 -5.31261712e-02
4.70231831e-01 1.48130739e+00 -1.42971277e+00 -4.73799080e-01
-3.76899004e-01 1.21608365e+00 2.80655503e-01 2.59417623e-01
-6.80623531e-01 -8.89745712e-01 -3.29731017e-01 -3.23481038e-02
4.40829009e-01 4.47394215e-02 -7.09463120e-01 -8.43585789e-01
-5.18074334e-02 -6.09606802e-01 4.18855488e-01 -8.33567679e-01
-9.19484794e-01 7.50584185e-01 -4.95839089e-01 -1.08156836e+00
-4.53427099e-02 -9.55680013e-01 -1.01652515e+00 6.55247808e-01
-1.09153366e+00 -1.38016796e+00 -4.42055315e-01 2.10801825e-01
9.48185027e-01 -3.89545500e-01 5.63666165e-01 1.16426818e-01
-1.06229162e+00 8.19080830e-01 7.16151297e-01 7.49077022e-01
8.30262363e-01 -1.75551093e+00 6.89826846e-01 1.00700557e+00
3.09874773e-01 3.79455090e-01 7.84724176e-01 -7.12955892e-01
-6.25778139e-01 -1.29064226e+00 7.98237383e-01 -2.59094536e-01
6.54127121e-01 -7.06560254e-01 -7.48815894e-01 6.78912997e-01
3.81246209e-01 -2.49799296e-01 5.34465551e-01 -1.18534826e-01
-2.34978586e-01 3.55956471e-03 -7.99773872e-01 7.86551118e-01
4.32460457e-01 -1.63935080e-01 -4.24795657e-01 5.97843230e-01
1.95424885e-01 -4.99284267e-01 -5.81003547e-01 1.12204626e-01
5.61091363e-01 -1.03923929e+00 6.56946898e-01 -2.80896097e-01
7.05321670e-01 -2.09205881e-01 3.09013039e-01 -9.08437788e-01
-1.47218570e-01 -5.80430508e-01 -9.61717963e-02 1.58862650e+00
7.93669999e-01 5.91211691e-02 1.02534533e+00 -3.83984074e-02
-2.93823749e-01 -6.43351495e-01 -3.28054100e-01 -5.10769784e-01
1.34467974e-01 -2.13979140e-01 9.34719816e-02 8.88984025e-01
-2.15746909e-01 5.44597983e-01 -1.97934493e-01 -1.58090722e-02
4.36642259e-01 2.67466959e-02 9.48774040e-01 -1.28036833e+00
1.00322992e-01 -8.30235600e-01 -1.36259124e-01 -1.27501810e+00
-1.03521742e-01 -7.17713237e-01 1.75386816e-01 -1.86497176e+00
7.05443695e-02 -1.81649879e-01 -4.39577550e-02 2.80600190e-01
-1.93091393e-01 2.91333020e-01 2.11299881e-01 2.02751517e-01
-4.58862424e-01 8.20036530e-02 8.37714314e-01 -3.96369576e-01
-2.97753692e-01 2.01407567e-01 -7.79280886e-02 1.00428009e+00
8.98994267e-01 -1.25989750e-01 -1.83726355e-01 -3.27602148e-01
4.72267233e-02 -2.41958976e-01 -2.49262676e-01 -1.40402985e+00
3.41396809e-01 3.98889095e-01 9.47725892e-01 -1.57759297e+00
-1.26538649e-01 -6.71072423e-01 -5.36471963e-01 4.71398503e-01
-8.34303975e-01 1.37728408e-01 2.43332341e-01 2.85581857e-01
-1.53644765e-02 -7.37275839e-01 8.86961162e-01 1.35473877e-01
-5.20360053e-01 -1.43447310e-01 -6.31306708e-01 -1.35336533e-01
9.93747890e-01 -4.88634437e-01 -5.08876264e-01 -2.35439554e-01
-6.51760042e-01 2.63001591e-01 3.41584831e-01 4.63977575e-01
3.05953085e-01 -1.06539500e+00 -4.88666505e-01 6.48699701e-02
1.79705366e-01 1.06055871e-01 1.61583368e-02 4.92164582e-01
-1.26285911e+00 5.97416103e-01 -2.05692351e-01 -7.04217672e-01
-1.35670948e+00 5.32269239e-01 4.17825758e-01 -5.76310873e-01
-7.08715081e-01 7.68308401e-01 1.51500940e-01 -2.92798728e-01
7.14866221e-01 -2.84521103e-01 -5.85179150e-01 2.06522390e-01
5.58347166e-01 4.69149619e-01 2.42894784e-01 -4.86318588e-01
2.88856663e-02 5.30542970e-01 -4.71034229e-01 -1.05208948e-01
1.20547414e+00 5.78512810e-02 1.57231018e-01 6.80808604e-01
1.02890015e+00 1.61199868e-01 -1.30530071e+00 8.47386643e-02
5.16709685e-01 1.01667613e-01 2.12671962e-02 -1.04065251e+00
-8.68500531e-01 1.03258908e+00 6.71630800e-01 2.72222430e-01
9.21462178e-01 -4.56426084e-01 5.62569737e-01 6.86662972e-01
-2.86949694e-01 -1.54912591e+00 7.71539658e-02 6.30117714e-01
8.04725766e-01 -1.07952976e+00 2.15086173e-02 -2.27663115e-01
-5.35735428e-01 1.69423854e+00 9.23040092e-01 -1.49282977e-01
8.23138207e-02 5.04839242e-01 2.05718935e-01 1.62507892e-02
-3.54343683e-01 7.82140121e-02 5.43557703e-01 3.64313543e-01
7.79346764e-01 -2.72171825e-01 -1.67813465e-01 1.29216969e-01
-3.08188617e-01 -2.89208740e-01 7.19341516e-01 8.40352595e-01
-5.55862248e-01 -8.67887974e-01 -5.78254819e-01 6.86087728e-01
-5.59859395e-01 -8.26656744e-02 -8.06855917e-01 9.60861981e-01
-1.15222223e-01 6.23834550e-01 4.57405031e-01 -9.81115922e-02
2.95199037e-01 1.72858655e-01 1.72269225e-01 -4.51937139e-01
-7.87044346e-01 6.21630073e-01 1.54780820e-01 -8.99926126e-02
-8.03440586e-02 -7.67821252e-01 -1.28813553e+00 9.04613882e-02
-5.34994960e-01 -1.49300098e-01 6.59649611e-01 6.56987369e-01
-1.54474244e-01 7.87212253e-01 3.25236708e-01 -7.89816678e-01
-2.33151451e-01 -1.18104303e+00 -3.49545717e-01 3.31467390e-01
2.80938953e-01 -1.28707796e-01 -2.31065061e-02 4.64337707e-01] | [11.825384140014648, 2.5458343029022217] |
f25b57fc-cbb7-4f0e-9666-14ba4d8ac517 | clearing-the-skies-a-deep-network | 1609.02087 | null | http://arxiv.org/abs/1609.02087v2 | http://arxiv.org/pdf/1609.02087v2.pdf | Clearing the Skies: A deep network architecture for single-image rain removal | We introduce a deep network architecture called DerainNet for removing rain
streaks from an image. Based on the deep convolutional neural network (CNN), we
directly learn the mapping relationship between rainy and clean image detail
layers from data. Because we do not possess the ground truth corresponding to
real-world rainy images, we synthesize images with rain for training. In
contrast to other common strategies that increase depth or breadth of the
network, we use image processing domain knowledge to modify the objective
function and improve deraining with a modestly-sized CNN. Specifically, we
train our DerainNet on the detail (high-pass) layer rather than in the image
domain. Though DerainNet is trained on synthetic data, we find that the learned
network translates very effectively to real-world images for testing. Moreover,
we augment the CNN framework with image enhancement to improve the visual
results. Compared with state-of-the-art single image de-raining methods, our
method has improved rain removal and much faster computation time after network
training. | ['Jia-Bin Huang', 'Xueyang Fu', 'Xinghao Ding', 'Yinghao Liao', 'John Paisley'] | 2016-09-07 | null | null | null | null | ['single-image-deraining'] | ['computer-vision'] | [ 1.79803863e-01 1.44984499e-01 6.63147628e-01 -6.43019259e-01
-4.11230952e-01 -3.65241259e-01 8.14413652e-02 -5.72738111e-01
-4.14465368e-01 8.95062029e-01 -5.84599189e-02 -3.56250137e-01
4.86814886e-01 -1.14732897e+00 -1.03768337e+00 -8.21602046e-01
2.88967658e-02 -1.27564758e-01 8.73069316e-02 -3.98848861e-01
-1.81476787e-01 6.00122154e-01 -1.44263422e+00 3.06772381e-01
1.24866903e+00 6.37428164e-01 3.71430516e-01 9.60408270e-01
5.32030240e-02 1.09785998e+00 -7.47290015e-01 -4.68497463e-02
5.47894835e-01 -6.46615326e-01 -4.21393543e-01 1.37945145e-01
1.14975166e+00 -8.44302475e-01 -5.86389720e-01 1.17366207e+00
3.77956331e-01 -7.20875338e-02 2.03918532e-01 -6.59525216e-01
-7.65773892e-01 8.87190849e-02 -5.04526556e-01 3.19846272e-01
-1.01101376e-01 4.09544319e-01 5.43985963e-01 -1.19926500e+00
5.28424382e-01 1.24626386e+00 8.66972566e-01 3.93357903e-01
-9.69335854e-01 -7.32154965e-01 3.22992623e-01 2.94735674e-02
-1.03489673e+00 -2.76374698e-01 4.09116328e-01 -9.71241593e-02
6.46815896e-01 2.20604166e-01 9.36775148e-01 8.33389759e-01
2.11907268e-01 6.19208753e-01 1.36986494e+00 -3.02913368e-01
1.43720746e-01 -2.05925167e-01 1.89881533e-01 6.57646418e-01
6.44765913e-01 4.50687498e-01 -1.94213837e-01 5.56572199e-01
8.91706944e-01 2.70370543e-01 -6.72700584e-01 1.47435993e-01
-8.35877180e-01 7.88525283e-01 1.05848420e+00 -6.93091974e-02
-5.22373736e-01 1.80337355e-01 -2.73985267e-01 6.62498951e-01
8.25327814e-01 5.26612461e-01 -4.14978236e-01 5.78241646e-01
-1.45959044e+00 5.91872334e-01 7.97452688e-01 5.88911295e-01
1.23647785e+00 6.27472699e-01 -1.21458590e-01 5.13616145e-01
2.94859018e-02 1.21137178e+00 -2.66227573e-01 -1.11577809e+00
3.74700308e-01 1.07579328e-01 3.05824161e-01 -6.65997624e-01
-3.44393462e-01 -5.39999485e-01 -1.02350318e+00 9.97147560e-01
4.14090693e-01 -4.43066388e-01 -1.65468252e+00 1.27588999e+00
-8.01941473e-03 3.70908469e-01 3.38154972e-01 1.35458291e+00
9.44938064e-01 9.12976027e-01 -1.80629894e-01 -4.69592027e-02
9.76611376e-01 -1.15655291e+00 -7.69468486e-01 -7.29740798e-01
1.34944811e-01 -5.91298521e-01 1.11573339e+00 4.54530954e-01
-1.10177219e+00 -6.03703082e-01 -1.28370988e+00 -1.55899972e-01
-3.54145646e-01 -1.93960760e-02 6.68407559e-01 3.67277592e-01
-1.05127776e+00 7.93371379e-01 -7.64200926e-01 -3.31032485e-01
5.15721142e-01 -1.79285258e-01 -1.94004804e-01 -7.29750633e-01
-1.27752304e+00 1.10445571e+00 1.72265485e-01 8.03541422e-01
-1.35330296e+00 -9.72546935e-01 -9.97733533e-01 5.33301122e-02
-9.03941765e-02 -7.30872154e-01 1.00690758e+00 -1.42568505e+00
-1.37524283e+00 7.30021060e-01 -9.49290842e-02 -6.33739889e-01
2.44082570e-01 -7.21969664e-01 -4.50646579e-01 2.99990833e-01
-2.19087839e-01 6.29642308e-01 1.35508227e+00 -1.64237058e+00
-6.98563695e-01 7.80171156e-02 3.60818297e-01 2.24063173e-01
3.14409047e-01 -3.43164235e-01 -2.55294085e-01 -7.74465561e-01
4.10454459e-02 -8.18877459e-01 -3.11909527e-01 2.90930063e-01
-1.94849595e-01 8.44376624e-01 1.05578542e+00 -1.05862796e+00
8.10797870e-01 -2.07217097e+00 -7.20139891e-02 1.08117327e-01
5.17432988e-01 6.53180897e-01 -4.49513912e-01 8.11594352e-02
-2.35134304e-01 -1.54865682e-01 -8.48655581e-01 -2.86640733e-01
-5.15117407e-01 6.20346963e-01 -6.00619733e-01 6.97816730e-01
5.65191746e-01 8.48208547e-01 -7.48887062e-01 -7.45605305e-02
3.09692204e-01 6.94924235e-01 -5.64399600e-01 7.12879419e-01
-2.56720632e-01 5.08445323e-01 2.69359481e-02 6.96425915e-01
1.41400290e+00 4.78883348e-02 3.17108817e-02 -2.88021743e-01
-2.52814908e-02 7.18078539e-02 -8.92850995e-01 1.19673550e+00
-8.09212208e-01 9.80477989e-01 3.93419713e-01 -7.25035608e-01
7.95420587e-01 -2.98362458e-03 -2.66045630e-01 -9.94208038e-01
-1.85153514e-01 2.80174445e-02 -1.20692790e-01 -6.93267345e-01
4.14756835e-01 -3.47010583e-01 5.38553178e-01 1.98661983e-01
5.51771233e-03 -5.34630299e-01 -4.74891588e-02 3.13756727e-02
1.04133308e+00 2.06521124e-01 -1.01294674e-01 -1.71518981e-01
-2.89400318e-03 7.84523934e-02 4.87042695e-01 9.59898651e-01
3.94954492e-04 1.21843791e+00 5.63678294e-02 -7.35013127e-01
-1.10758388e+00 -1.38318610e+00 -1.08749993e-01 8.21100116e-01
2.34794110e-01 1.74450144e-01 -8.94664466e-01 -4.17034715e-01
-1.05604015e-01 7.11922467e-01 -8.82474482e-01 1.65863067e-01
-8.84640276e-01 -1.02843249e+00 4.47490096e-01 3.53315562e-01
1.00435817e+00 -1.41159213e+00 -5.10190606e-01 9.18862596e-02
-6.00277744e-02 -1.24372506e+00 -2.31081530e-01 2.74860770e-01
-9.30425107e-01 -1.22390735e+00 -5.76931775e-01 -7.69834399e-01
7.43389666e-01 6.64606631e-01 1.75273645e+00 3.77534330e-01
-5.13656676e-01 -8.62769708e-02 -5.32454431e-01 -5.78440070e-01
-1.34012297e-01 -2.53017694e-01 -5.88976741e-01 -2.32280537e-01
1.09412082e-01 -7.53627956e-01 -1.12043941e+00 -3.34721059e-01
-1.29342782e+00 2.46133958e-03 7.65198827e-01 7.51157641e-01
3.71293426e-01 2.15411503e-02 1.55979067e-01 -1.13244402e+00
3.03783119e-01 -3.04106653e-01 -7.09514380e-01 -2.22512223e-02
-5.41263163e-01 2.48916447e-02 4.37436372e-01 -7.88125675e-03
-1.32459831e+00 8.43442827e-02 -1.95454225e-01 -5.02166271e-01
-1.42857686e-01 4.12409425e-01 7.20565543e-02 -2.42180526e-01
8.34257245e-01 5.40919751e-02 -9.61889848e-02 -3.43835831e-01
5.72821975e-01 1.44189090e-01 8.12501252e-01 -6.05212711e-02
1.33322203e+00 8.22730720e-01 -2.93009251e-01 -7.46498227e-01
-1.45401466e+00 -7.27469176e-02 -3.98521602e-01 -1.02242619e-01
1.12834609e+00 -1.49195373e+00 -3.87682915e-01 5.90998888e-01
-1.15079081e+00 -8.36197853e-01 -2.82590002e-01 2.83490330e-01
1.41779762e-02 8.93799812e-02 -8.13654900e-01 -7.39181519e-01
-6.87579691e-01 -6.75777256e-01 9.89886701e-01 3.40226799e-01
3.44552934e-01 -8.71490598e-01 6.48298934e-02 -5.60637452e-02
7.73536265e-01 4.67077076e-01 4.57856297e-01 4.85019892e-01
-9.36491668e-01 7.35257640e-02 -6.37931764e-01 8.62092018e-01
1.38951361e-01 7.78931677e-02 -1.27953517e+00 -4.83496934e-01
2.45832875e-01 -2.38030955e-01 1.48092246e+00 5.53149343e-01
1.03940165e+00 -3.18248808e-01 2.51732826e-01 1.15323865e+00
1.74915099e+00 -1.88171864e-01 1.34881294e+00 4.91253048e-01
8.39033723e-01 4.16497946e-01 6.72123492e-01 9.52941403e-02
2.28711799e-01 1.95899848e-02 8.56017053e-01 -1.04525256e+00
-5.82292974e-01 1.11660399e-01 4.70905095e-01 4.40257043e-01
-2.72241384e-01 -2.42814466e-01 -5.57517946e-01 6.18741930e-01
-1.59336472e+00 -1.07204473e+00 -5.06416336e-02 1.90527487e+00
8.45989525e-01 7.79673830e-02 -4.95401472e-01 -1.67435780e-01
2.90809095e-01 5.33196390e-01 -4.72730815e-01 -2.49416888e-01
-4.86983746e-01 8.89342248e-01 9.89017367e-01 1.00135756e+00
-1.14017200e+00 1.14438570e+00 6.99548388e+00 1.07867524e-01
-1.49514699e+00 1.46062048e-02 4.66441214e-01 -2.27699026e-01
-1.51971683e-01 -1.17045045e-01 -5.48574269e-01 2.81913877e-01
7.15266347e-01 4.91595745e-01 8.13470840e-01 5.83586276e-01
4.93646979e-01 -1.68991193e-01 -6.98245108e-01 7.85323977e-01
8.63352194e-02 -1.27034223e+00 -2.98395678e-02 -4.14445609e-01
9.17587996e-01 4.20722991e-01 -6.20544329e-03 3.42210174e-01
6.72309041e-01 -1.46579921e+00 5.96357584e-01 8.25779498e-01
6.80485129e-01 -5.31255186e-01 9.38428700e-01 -6.20780699e-02
-8.19351971e-01 1.70959130e-01 -6.24498606e-01 -1.83249637e-01
-1.65391210e-02 1.30455244e+00 -6.19971693e-01 4.63830948e-01
1.18770897e+00 7.05226004e-01 -5.70514798e-01 8.32473099e-01
-7.96858370e-01 9.54490066e-01 -2.31045857e-01 8.44334543e-01
1.77245155e-01 -3.91728610e-01 2.50395298e-01 1.48401308e+00
2.78962255e-01 2.03416035e-01 -6.90965801e-02 9.12252784e-01
-1.69248760e-01 -7.53725231e-01 -7.54376888e-01 2.42222741e-01
1.22511432e-01 1.29220223e+00 -2.26855993e-01 -4.71293747e-01
-4.43548054e-01 1.30603909e+00 1.82357505e-01 9.99089241e-01
-8.84799898e-01 -6.25785053e-01 1.07764077e+00 1.28497168e-01
6.76003397e-01 -3.83768171e-01 -5.40868461e-01 -1.23062360e+00
-8.42919275e-02 -1.14186597e+00 -2.31850594e-01 -1.15656817e+00
-1.29384005e+00 9.20943916e-01 -3.25034976e-01 -1.20465887e+00
1.23876661e-01 -6.54974043e-01 -9.89168286e-01 1.23632348e+00
-2.37772632e+00 -1.12028944e+00 -1.11693311e+00 5.68261266e-01
4.73157316e-01 4.35260832e-01 5.86959541e-01 3.24197859e-01
-3.26277614e-01 1.84160352e-01 9.68582630e-02 1.71179220e-01
9.69237626e-01 -1.33553803e+00 7.48822153e-01 1.43268692e+00
-1.94064647e-01 4.14082974e-01 1.09365785e+00 -4.68337506e-01
-1.10115147e+00 -1.70554137e+00 5.67705393e-01 -1.93738669e-01
2.65373051e-01 -4.19510514e-01 -1.36804605e+00 7.63106227e-01
4.38704133e-01 4.94261533e-01 -1.17903069e-01 -1.57189816e-02
-5.95073998e-01 -4.27394271e-01 -1.16706455e+00 5.73039293e-01
9.42408979e-01 -5.04136562e-01 -5.07462561e-01 3.86128426e-01
9.08885062e-01 -7.57203519e-01 -4.47853476e-01 5.40763915e-01
2.58920103e-01 -1.19432044e+00 9.88152504e-01 -3.60735416e-01
8.89552772e-01 -6.46367967e-01 -1.56144053e-01 -1.74172819e+00
-2.73329735e-01 -1.67156711e-01 -1.92306265e-01 5.75005949e-01
4.81480181e-01 -5.57761490e-01 7.58575439e-01 9.71453711e-02
-2.97729611e-01 -3.55581552e-01 -3.95255744e-01 -5.09303927e-01
2.06037655e-01 -1.00668140e-01 5.14758945e-01 9.63298202e-01
-8.96635830e-01 1.77903309e-01 -6.36566460e-01 7.80341268e-01
8.88302505e-01 5.54937840e-01 7.54791975e-01 -9.29322720e-01
-2.11803377e-01 2.36044034e-01 1.26648188e-01 -9.94860291e-01
3.35448198e-02 -4.02366757e-01 6.99669123e-01 -1.73231339e+00
-2.41353363e-02 -1.07619122e-01 -2.11332545e-01 7.33996809e-01
-6.17015898e-01 6.97020411e-01 2.72031963e-01 8.99822786e-02
-3.10417980e-01 5.57145834e-01 1.71217561e+00 -2.17076510e-01
-2.02351868e-01 -3.16445112e-01 -5.35379112e-01 6.76600456e-01
9.72656846e-01 -5.23127556e-01 -2.13479564e-01 -9.08548057e-01
1.98362842e-01 -2.02472568e-01 5.87771058e-01 -1.13563120e+00
-1.77069739e-01 -1.32780001e-01 6.87008739e-01 -3.78378302e-01
2.32184529e-01 -6.78315222e-01 -1.74615160e-02 4.44165349e-01
-3.51389498e-02 -2.12669268e-01 2.23268017e-01 2.77548492e-01
-4.61833537e-01 7.57037774e-02 1.27399957e+00 -2.94910043e-01
-6.98431671e-01 3.54595244e-01 -3.80586743e-01 -1.89394057e-01
3.30237985e-01 -9.26232710e-02 -8.27234507e-01 -5.82298160e-01
-8.05111051e-01 1.17606170e-01 5.69659352e-01 1.26688376e-01
7.83825338e-01 -7.37584412e-01 -9.06458676e-01 3.17412317e-01
-2.33483374e-01 3.92178625e-01 3.18120956e-01 4.85997200e-01
-1.16009653e+00 -1.77038431e-01 -3.65621567e-01 -2.63499647e-01
-9.69679415e-01 4.03540850e-01 7.14652121e-01 -1.24883346e-01
-1.06274569e+00 5.52380264e-01 7.03962445e-01 -3.79766881e-01
7.32878968e-02 -5.80166340e-01 1.09068483e-01 -4.53930020e-01
7.24181533e-01 -1.09503053e-01 3.54762495e-01 7.97148794e-02
4.61529158e-02 2.81118572e-01 -5.63963652e-02 1.94843784e-01
1.65545595e+00 -8.17550570e-02 -2.98848569e-01 9.15579870e-02
9.37937200e-01 -5.10303341e-02 -1.77295411e+00 -1.48304567e-01
-6.10973656e-01 -5.32808304e-01 3.91777098e-01 -8.95187855e-01
-1.77825093e+00 9.26967084e-01 1.09182394e+00 -5.47221042e-02
1.45341265e+00 -5.08131504e-01 7.37262309e-01 7.73379982e-01
-1.06418319e-01 -7.31687725e-01 -7.24062994e-02 8.85610878e-01
1.06909811e+00 -1.45659244e+00 2.32948184e-01 -4.18538183e-01
-4.99608397e-01 1.05388939e+00 7.45868623e-01 -7.94551611e-01
8.27447414e-01 6.80710554e-01 7.31393456e-01 -3.74430656e-01
-3.82631868e-01 -5.04138649e-01 -2.06756577e-01 7.30045319e-01
2.99668223e-01 -1.19254798e-01 1.03619382e-01 -4.36572395e-02
-1.96408167e-01 1.93439111e-01 7.66475022e-01 8.99623394e-01
-9.16104019e-01 -5.97621799e-01 -6.93999827e-01 3.22357744e-01
-3.12201291e-01 -7.36557066e-01 6.11408725e-02 8.71245921e-01
1.94102719e-01 8.16262007e-01 1.86856270e-01 -6.94058836e-02
5.03691792e-01 -4.67976660e-01 5.93069077e-01 -5.25835276e-01
-4.73284096e-01 -3.68652493e-01 -5.81935607e-02 -6.66955292e-01
-6.53113604e-01 -1.63446888e-02 -1.00887525e+00 -5.24912775e-01
1.00208141e-01 -1.33814499e-01 5.07690012e-01 8.94747913e-01
1.13827571e-01 8.15035105e-01 6.60412312e-01 -1.34032249e+00
1.15032360e-01 -1.10291255e+00 -9.05223608e-01 2.77947605e-01
1.04202867e+00 -1.99766994e-01 -4.94035333e-01 3.72648478e-01] | [10.918107986450195, -3.2303073406219482] |
ef65e7c5-41db-4e0f-b1ed-be3e8dd3a36c | increasing-performance-and-sample-efficiency | 2306.16431 | null | https://arxiv.org/abs/2306.16431v1 | https://arxiv.org/pdf/2306.16431v1.pdf | Increasing Performance And Sample Efficiency With Model-agnostic Interactive Feature Attributions | Model-agnostic feature attributions can provide local insights in complex ML models. If the explanation is correct, a domain expert can validate and trust the model's decision. However, if it contradicts the expert's knowledge, related work only corrects irrelevant features to improve the model. To allow for unlimited interaction, in this paper we provide model-agnostic implementations for two popular explanation methods (Occlusion and Shapley values) to enforce entirely different attributions in the complex model. For a particular set of samples, we use the corrected feature attributions to generate extra local data, which is used to retrain the model to have the right explanation for the samples. Through simulated and real data experiments on a variety of models we show how our proposed approach can significantly improve the model's performance only by augmenting its training dataset based on corrected explanations. Adding our interactive explanations to active learning settings increases the sample efficiency significantly and outperforms existing explanatory interactive strategies. Additionally we explore how a domain expert can provide feature attributions which are sufficiently correct to improve the model. | ['Johan Suykens', 'Maarten De Vos', 'Joran Michiels'] | 2023-06-28 | null | null | null | null | ['active-learning', 'active-learning'] | ['methodology', 'natural-language-processing'] | [ 2.48058155e-01 9.00096059e-01 -5.75651646e-01 -7.34405160e-01
-5.53028166e-01 -5.47342420e-01 3.94742310e-01 2.90016413e-01
-1.22557558e-01 9.71635044e-01 -3.38827483e-02 -4.75073934e-01
-2.25996897e-01 -4.74246651e-01 -6.81115746e-01 -5.09657264e-01
1.05295755e-01 7.80930281e-01 2.72340745e-01 8.86378959e-02
5.73954940e-01 4.08247113e-01 -1.65197921e+00 3.66854757e-01
1.14358997e+00 5.20429313e-01 1.27556056e-01 6.45763934e-01
-2.04535082e-01 7.46582091e-01 -6.00768924e-01 -3.82770568e-01
4.01020944e-01 -2.61494398e-01 -7.81662643e-01 2.74214447e-01
5.68391562e-01 -5.03620327e-01 1.92363918e-01 6.62773907e-01
5.80187105e-02 2.44456708e-01 5.84535897e-01 -1.80285513e+00
-6.30648613e-01 8.16069007e-01 -2.44934902e-01 4.89025982e-03
3.04195076e-01 3.71659040e-01 1.12677753e+00 -7.76102960e-01
5.95130861e-01 1.25324738e+00 3.76611859e-01 5.91583848e-01
-1.48228157e+00 -8.96411479e-01 8.27004135e-01 3.48475009e-01
-9.28824067e-01 -3.53121817e-01 8.97993088e-01 -1.95197821e-01
7.61707067e-01 4.94544595e-01 6.77503347e-01 9.95563805e-01
-1.63860515e-01 7.22471714e-01 1.05552340e+00 -5.29980838e-01
6.14602506e-01 7.96191931e-01 6.07108593e-01 8.13371599e-01
5.31994522e-01 1.40605465e-01 -7.63766110e-01 -6.45421386e-01
5.32102585e-01 8.65636021e-02 -1.88483641e-01 -8.24096441e-01
-8.39346290e-01 1.07712352e+00 4.70193386e-01 -2.85176367e-01
-4.13606584e-01 1.45298585e-01 -2.69826859e-01 5.25800407e-01
4.58810121e-01 9.94246364e-01 -9.03439701e-01 1.66665956e-01
-5.85007906e-01 2.58457541e-01 9.02789593e-01 7.23327637e-01
1.11782312e+00 -2.20182210e-01 -4.41486873e-02 5.46840787e-01
5.40578544e-01 6.56460300e-02 2.32978791e-01 -1.29866552e+00
1.44222319e-01 1.13834536e+00 3.99678230e-01 -8.29648614e-01
-1.34838492e-01 -6.77636981e-01 -1.11827977e-01 6.04005396e-01
4.68283385e-01 -1.06095128e-01 -8.07982922e-01 1.68533695e+00
6.37621582e-01 2.70829409e-01 2.51182411e-02 8.43106925e-01
5.52745223e-01 1.74386606e-01 1.24850743e-01 -7.79429972e-02
8.12152684e-01 -1.07035983e+00 -4.27326947e-01 -6.07706070e-01
7.76925862e-01 -5.10658264e-01 1.24828422e+00 4.29244488e-01
-7.65886605e-01 -3.88855845e-01 -1.08219600e+00 1.54650450e-01
-3.08365554e-01 6.73552603e-02 8.86914790e-01 5.98568499e-01
-6.79434657e-01 7.41328180e-01 -8.34725618e-01 -4.28777397e-01
5.99437952e-01 6.31768048e-01 -2.30166093e-01 -3.04928031e-02
-9.64105189e-01 9.31611538e-01 1.40653431e-01 -2.30660424e-01
-7.88225830e-01 -9.90436435e-01 -5.93460500e-01 2.42812559e-01
6.78962648e-01 -8.38229895e-01 1.28771079e+00 -1.27846360e+00
-1.24751890e+00 2.93882251e-01 -2.90518433e-01 -4.82675284e-01
7.68189073e-01 -1.85640424e-01 5.52550331e-02 -7.70433247e-02
6.49602115e-02 9.85258162e-01 8.61901700e-01 -1.57230687e+00
-6.32551670e-01 1.53720565e-02 5.90322852e-01 2.57657349e-01
-4.30014640e-01 -4.31246549e-01 -1.53907478e-01 -2.21760124e-01
1.69486552e-01 -1.08046877e+00 -5.20223737e-01 5.46749949e-01
-5.63594759e-01 -7.98591152e-02 9.47292268e-01 -4.10722643e-02
8.70405436e-01 -1.90866959e+00 -1.51171148e-01 4.46604013e-01
5.51128387e-01 -4.66954149e-02 -1.46373987e-01 2.26157293e-01
-1.08682871e-01 3.93613577e-01 -1.86084107e-01 -5.16832650e-01
4.35603932e-02 5.27478874e-01 -2.57030010e-01 3.41300040e-01
2.20199689e-01 5.41428506e-01 -6.57313704e-01 -4.69975948e-01
2.37542167e-01 2.03917965e-01 -9.68194902e-01 3.66896361e-01
-4.38351363e-01 3.74547124e-01 -4.45340872e-01 2.48697087e-01
6.43508077e-01 -6.60029113e-01 3.18385750e-01 1.39556259e-01
2.26356760e-01 2.36632451e-01 -1.49267089e+00 1.12707675e+00
-4.03830767e-01 5.73383033e-01 -1.39016509e-01 -6.87854111e-01
8.56075585e-01 6.03914820e-03 7.04658553e-02 -2.58091956e-01
-2.08087936e-01 7.59474337e-02 2.02071965e-01 -2.57287174e-01
2.14063942e-01 1.68157637e-01 3.43397468e-01 9.45503712e-01
-1.75187185e-01 3.18850011e-01 -1.51435927e-01 5.05107939e-01
1.08164179e+00 7.40673617e-02 5.27729154e-01 -2.51123667e-01
3.07008058e-01 2.27895215e-01 5.32583952e-01 1.18438077e+00
-3.28769460e-02 2.82884777e-01 4.82474685e-01 -5.22864103e-01
-7.60613799e-01 -8.81465852e-01 -9.66890622e-03 1.29997182e+00
1.58657700e-01 -4.99655396e-01 -4.78848606e-01 -1.22921467e+00
2.62533635e-01 1.08840656e+00 -1.03746080e+00 -2.98043221e-01
-1.60702839e-01 -4.27678406e-01 -4.63043340e-02 5.28166056e-01
5.08844972e-01 -8.65155101e-01 -8.08918595e-01 -9.87483654e-03
-1.72206312e-02 -3.72112572e-01 -1.09763421e-01 4.58827734e-01
-8.34866583e-01 -1.32208240e+00 -1.71291381e-01 -1.56215057e-01
1.26066017e+00 3.30022365e-01 8.03142786e-01 7.49668598e-01
1.36006370e-01 2.57761747e-01 -2.82848507e-01 -3.81832421e-01
-5.29925704e-01 2.18333587e-01 -1.32706612e-01 -2.25383908e-01
2.29451075e-01 -4.05516803e-01 -4.07989532e-01 5.45694351e-01
-6.32449865e-01 2.72350848e-01 4.73616838e-01 8.20612967e-01
2.66002089e-01 -9.86091644e-02 4.91787672e-01 -1.65832126e+00
6.09739065e-01 -4.81974036e-01 -5.07412493e-01 3.22734475e-01
-1.32560098e+00 5.37441134e-01 5.77062368e-01 -7.84277141e-01
-1.27080274e+00 1.38768643e-01 4.98164415e-01 -4.30718899e-01
-1.89618617e-01 4.12038237e-01 -1.24146782e-01 -1.47202194e-01
9.81529295e-01 -3.09260726e-01 -1.98864192e-01 -4.96143013e-01
2.65725523e-01 4.08556581e-01 1.01924255e-01 -4.69025582e-01
9.92516279e-01 2.07534492e-01 -1.57587200e-01 -1.37717053e-01
-1.02261996e+00 -1.22503862e-01 -7.65765905e-01 -5.55940866e-02
5.53651899e-02 -6.26547694e-01 -5.75998902e-01 -2.40338907e-01
-1.00088966e+00 -4.59191799e-01 -2.78440058e-01 3.78807068e-01
-2.99416810e-01 4.65456620e-02 7.26998150e-02 -8.27294588e-01
1.66622415e-01 -1.17417109e+00 6.33655846e-01 3.15925747e-01
-7.13025272e-01 -1.23677325e+00 4.33992855e-02 3.57583553e-01
4.02727634e-01 -6.29714504e-02 1.11180902e+00 -1.20649791e+00
-8.07999790e-01 -1.53157100e-01 4.05156799e-02 -1.38776273e-01
1.40812337e-01 2.82221884e-01 -1.34427679e+00 -1.53923765e-01
-3.83866459e-01 -1.77045062e-01 7.55517721e-01 4.53792363e-02
1.13890314e+00 -7.74683535e-01 -6.16886914e-01 2.43916124e-01
9.13651347e-01 3.11015435e-02 3.34959716e-01 5.65600395e-01
4.62940782e-01 5.40640116e-01 9.13062274e-01 5.65804899e-01
2.48567656e-01 4.79746729e-01 5.84004581e-01 -4.01710451e-01
-8.57590362e-02 -3.90503138e-01 5.57144396e-02 3.02930456e-02
2.87348777e-01 -2.40550980e-01 -6.02337122e-01 3.06632698e-01
-2.21639562e+00 -7.77628124e-01 -1.57374665e-01 2.31762409e+00
7.65440345e-01 2.90676266e-01 -1.53030574e-01 1.01428360e-01
5.29082298e-01 -1.64425135e-01 -1.00790524e+00 -4.44529206e-01
1.63941532e-01 -3.97109985e-02 3.33286136e-01 8.62366736e-01
-7.19473064e-01 8.04701030e-01 6.67851925e+00 3.98185551e-01
-9.43937361e-01 3.39193568e-02 5.27885735e-01 -2.09986642e-01
-7.77285397e-01 6.91011071e-01 -6.74663007e-01 1.77856684e-01
7.45752335e-01 -3.33460450e-01 3.63984734e-01 1.25205469e+00
1.48775905e-01 -3.15847814e-01 -1.58856881e+00 4.21397001e-01
-5.01079150e-02 -1.43554091e+00 1.26494810e-01 2.17406869e-01
7.03965783e-01 -4.73926008e-01 9.13401619e-02 3.92343551e-01
7.62725949e-01 -8.20017874e-01 6.01472139e-01 4.76402491e-01
2.20776677e-01 -7.30994225e-01 6.81415439e-01 6.57831073e-01
-6.77025914e-01 -2.99362779e-01 -3.17241490e-01 -3.29240829e-01
-3.13083887e-01 2.73281693e-01 -1.55817258e+00 1.29677176e-01
3.65548134e-01 5.05195320e-01 -1.15704584e+00 9.68478680e-01
-5.84214807e-01 7.64388740e-01 -3.37806940e-01 -2.70158220e-02
-1.20124698e-01 2.63436943e-01 3.89235944e-01 6.62942469e-01
-7.82004222e-02 3.12011614e-02 2.35531002e-01 1.08514118e+00
3.37006189e-02 -4.77874428e-02 -4.49066460e-01 2.06335828e-01
7.90330827e-01 1.24914265e+00 -5.94714522e-01 -5.96021235e-01
-1.10235482e-01 8.69454801e-01 5.94183981e-01 5.17228723e-01
-8.84795189e-01 5.46965562e-02 5.53113222e-01 2.66008466e-01
1.50410697e-01 3.47799063e-01 -4.73335147e-01 -1.12394524e+00
-3.47401232e-01 -8.32629919e-01 4.65691537e-01 -1.03970945e+00
-1.06248200e+00 2.77898222e-01 2.41624042e-01 -9.86716986e-01
-4.62034136e-01 -2.49703363e-01 -8.08814585e-01 7.57465780e-01
-1.35842776e+00 -8.49505544e-01 -5.26459396e-01 4.77821916e-01
4.71234620e-01 -4.55987127e-03 8.13572407e-01 -2.18251213e-01
-5.33827722e-01 7.19998062e-01 -2.09940627e-01 -3.48233581e-01
9.39917147e-01 -1.42396045e+00 2.65545160e-01 7.79086292e-01
1.50596336e-01 1.07386124e+00 1.16292584e+00 -7.99449027e-01
-1.03088570e+00 -9.90660191e-01 7.90112913e-01 -4.95139301e-01
4.08483088e-01 -2.41086334e-01 -1.16311955e+00 1.10936558e+00
3.13869491e-03 6.99519273e-03 8.82326424e-01 3.86885762e-01
-3.56247813e-01 5.89633286e-02 -1.56045902e+00 7.63703585e-01
9.05270159e-01 -1.41223103e-01 -5.42788804e-01 4.48429257e-01
5.84073961e-01 -7.30727687e-02 -4.79090750e-01 1.89965010e-01
5.50861537e-01 -9.55502689e-01 7.89288402e-01 -1.00143075e+00
1.19302057e-01 -3.40124786e-01 -5.38441315e-02 -1.36175811e+00
-5.30889332e-01 -4.37633187e-01 -1.73134312e-01 1.06621182e+00
8.65713298e-01 -1.02237749e+00 9.38742518e-01 1.29066443e+00
1.23268247e-01 -6.75911188e-01 -5.76280057e-01 -3.26453745e-01
-1.82257116e-01 -3.34416538e-01 8.84201705e-01 9.87832248e-01
-7.39686517e-03 3.10153306e-01 -1.47807822e-01 4.86724705e-01
6.83095574e-01 3.78001422e-01 1.17201734e+00 -1.63558221e+00
-5.80646932e-01 -6.92849010e-02 1.15535162e-01 -8.91865671e-01
3.08058828e-01 -6.02877378e-01 -2.34545454e-01 -1.51685512e+00
4.56782758e-01 -7.34121442e-01 -1.51387200e-01 1.09336984e+00
-4.03759152e-01 7.06436709e-02 2.20906585e-01 3.87295485e-01
-5.43421745e-01 1.36529431e-01 1.04410541e+00 8.76273215e-03
-3.75831962e-01 9.95175317e-02 -1.09731710e+00 8.41263950e-01
7.63959348e-01 -8.70746493e-01 -7.65859306e-01 5.15217297e-02
2.19135016e-01 -1.06869318e-01 6.65079474e-01 -6.43582404e-01
3.45152378e-01 -5.53568423e-01 7.65613437e-01 -2.38563880e-01
3.77417445e-01 -9.83654439e-01 3.39030743e-01 5.45063913e-01
-1.06819499e+00 -2.25212842e-01 8.23774263e-02 7.13750780e-01
3.94935727e-01 -3.44333261e-01 6.08208179e-01 -1.17480792e-01
-4.29858327e-01 2.01613680e-02 -3.54159832e-01 -4.27694887e-01
1.11105061e+00 -4.39191878e-01 -6.56291485e-01 -6.10443115e-01
-9.06578302e-01 2.99755961e-01 8.16968143e-01 1.41126364e-01
4.77774858e-01 -1.01333606e+00 -3.27818215e-01 3.38570416e-01
3.21873099e-01 -2.74629116e-01 -1.10296354e-01 6.99910939e-01
7.25694671e-02 2.59472042e-01 -1.11491516e-01 -5.58669329e-01
-1.47117281e+00 5.88796377e-01 2.90070087e-01 -3.97745930e-02
-4.42048967e-01 5.75062335e-01 3.44117492e-01 -7.12484419e-01
2.48664111e-01 -2.93326050e-01 -1.08977355e-01 -1.87072679e-01
3.49560022e-01 2.71590978e-01 -5.43364510e-02 4.02957126e-02
-3.93011212e-01 7.27490932e-02 -4.86800611e-01 -1.68863371e-01
1.22695994e+00 -1.73954651e-01 5.53048730e-01 5.23657143e-01
5.90158224e-01 1.54150188e-01 -1.74464178e+00 -1.22691303e-01
-3.41703743e-01 -8.87915969e-01 1.48203552e-01 -1.38586223e+00
-1.03984213e+00 6.73983216e-01 3.74362200e-01 1.81366846e-01
7.97895789e-01 1.99880730e-02 -1.26167551e-01 7.13392317e-01
2.32598782e-01 -7.24227130e-01 1.92443386e-01 -1.47782028e-01
7.79223084e-01 -1.32741082e+00 4.42217678e-01 -5.36321461e-01
-8.41818810e-01 1.11419797e+00 9.90470529e-01 -8.58703628e-02
3.71727854e-01 6.68956563e-02 1.30533874e-01 -1.05128385e-01
-1.42365253e+00 1.31923184e-01 2.12462351e-01 6.02455795e-01
2.35501334e-01 -8.13054442e-02 9.77157243e-03 5.54347813e-01
-1.99642196e-01 -1.18408024e-01 6.75051391e-01 8.70347500e-01
-7.06004620e-01 -1.21432781e+00 -4.05013740e-01 6.87077880e-01
1.27574563e-01 5.57685830e-02 -9.43749130e-01 1.06384945e+00
1.40957590e-02 1.14973760e+00 3.37703712e-02 -1.58257470e-01
2.52357777e-02 1.17660031e-01 2.88284510e-01 -7.84446418e-01
-7.55543351e-01 -1.29733682e-01 1.61920637e-01 -5.63325822e-01
-1.85946241e-01 -5.42357385e-01 -1.47914338e+00 -4.19303566e-01
-6.65020943e-01 3.29339564e-01 4.71652091e-01 1.10715330e+00
6.09224081e-01 4.05148029e-01 6.76072717e-01 -4.15424138e-01
-5.52425623e-01 -8.05407226e-01 -1.96137190e-01 2.88973331e-01
3.11642230e-01 -1.05309248e+00 -8.42671454e-01 3.74466181e-03] | [8.83051586151123, 5.723832607269287] |
cf6fbabf-7d37-4a80-a864-57950c9a7701 | more-behind-your-electricity-bill-a-dual-dnn | 2106.00297 | null | https://arxiv.org/abs/2106.00297v1 | https://arxiv.org/pdf/2106.00297v1.pdf | More Behind Your Electricity Bill: a Dual-DNN Approach to Non-Intrusive Load Monitoring | Non-intrusive load monitoring (NILM) is a well-known single-channel blind source separation problem that aims to decompose the household energy consumption into itemised energy usage of individual appliances. In this way, considerable energy savings could be achieved by enhancing household's awareness of energy usage. Recent investigations have shown that deep neural networks (DNNs) based approaches are promising for the NILM task. Nevertheless, they normally ignore the inherent properties of appliance operations in the network design, potentially leading to implausible results. We are thus motivated to develop the dual Deep Neural Networks (dual-DNN), which aims to i) take advantage of DNNs' learning capability of latent features and ii) empower the DNN architecture with identification ability of universal properties. Specifically in the design of dual-DNN, we adopt one subnetwork to measure power ratings of different appliances' operation states, and the other subnetwork to identify the running states of target appliances. The final result is then obtained by multiplying these two network outputs and meanwhile considering the multi-state property of household appliances. To enforce the sparsity property in appliance's state operating, we employ median filtering and hard gating mechanisms to the subnetwork for state identification. Compared with the state-of-the-art NILM methods, our dual-DNN approach demonstrates a 21.67% performance improvement in average on two public benchmark datasets. | ['Hong Xu', 'Yi Wang', 'Qianyi Huang', 'Guoming Tang', 'Yu Zhang'] | 2021-06-01 | null | null | null | null | ['non-intrusive-load-monitoring', 'non-intrusive-load-monitoring', 'non-intrusive-load-monitoring'] | ['knowledge-base', 'miscellaneous', 'time-series'] | [ 1.46488979e-01 -1.83452234e-01 -4.17681426e-01 -4.47329134e-01
-4.04692799e-01 -3.96891594e-01 3.98962110e-01 -3.85680139e-01
-6.08261824e-02 5.18827379e-01 2.61899054e-01 -2.99051106e-01
-8.93670842e-02 -7.78864920e-01 -4.77749825e-01 -1.28589213e+00
8.72308686e-02 9.99508500e-02 -5.88107288e-01 2.78777957e-01
-3.04739058e-01 3.44948381e-01 -1.45214522e+00 -1.23089086e-02
7.69624412e-01 1.42215455e+00 1.35136656e-02 2.57850796e-01
1.12128809e-01 9.82710958e-01 -6.27086818e-01 8.23308751e-02
7.43971244e-02 -3.30753654e-01 -4.74462241e-01 1.00426905e-01
-2.01408088e-01 -7.72664011e-01 -5.59042454e-01 1.41266799e+00
5.99759042e-01 3.44377309e-02 4.76175517e-01 -1.52362311e+00
-5.95195889e-01 9.53051805e-01 -5.72403312e-01 2.15335503e-01
2.21306056e-01 2.60600090e-01 9.33874130e-01 -9.65495259e-02
-3.70219260e-01 9.14597571e-01 4.68413115e-01 4.73375082e-01
-1.35976505e+00 -1.01848173e+00 3.53317529e-01 4.22935635e-01
-1.32374108e+00 -7.01745808e-01 1.25930989e+00 -7.88555741e-02
1.20381474e+00 5.46800673e-01 4.53218400e-01 1.59875834e+00
-2.05363065e-01 1.04853439e+00 1.15401351e+00 -1.55915380e-01
5.78679502e-01 1.76696137e-01 2.81197816e-01 3.26386482e-01
2.39756986e-01 -3.56345735e-02 -3.37531835e-01 -2.45149165e-01
3.74458104e-01 2.40741208e-01 -3.95029128e-01 8.81499574e-02
-9.00072157e-01 5.79386234e-01 2.17183068e-01 5.20599663e-01
-6.38787687e-01 1.82045788e-01 5.44807076e-01 -4.66688313e-02
3.72370154e-01 -1.15185969e-01 -5.62509894e-01 -4.67462242e-02
-1.05574262e+00 -3.91621202e-01 8.37654352e-01 6.77024364e-01
6.07743740e-01 4.36967015e-01 -1.27916828e-01 5.14300108e-01
4.99577940e-01 5.09923577e-01 8.27384949e-01 -9.00206327e-01
3.22428823e-01 8.52751374e-01 -1.38479650e-01 -5.91332018e-01
-6.85591102e-01 -3.00000399e-01 -1.56375718e+00 1.58054512e-02
-3.44738625e-02 -2.53320336e-01 -8.10419321e-01 2.17516017e+00
1.09605789e-01 4.17308897e-01 2.04006344e-01 6.50586963e-01
7.59773791e-01 7.49348819e-01 2.61785418e-01 -5.73443294e-01
1.42345309e+00 -6.40500128e-01 -1.02810311e+00 -4.44235690e-02
2.38073766e-01 -2.49055922e-01 6.34152591e-01 4.36756104e-01
-7.37615347e-01 -4.41520751e-01 -1.12998366e+00 2.91477859e-01
-4.47923809e-01 4.33726579e-01 8.12692761e-01 7.55272627e-01
-6.71884120e-01 5.67833662e-01 -1.19029033e+00 -2.16891721e-01
5.29182315e-01 6.89747274e-01 2.99083591e-02 4.11076128e-01
-1.21051085e+00 5.89398146e-01 4.78253424e-01 4.97201830e-01
-1.08990884e+00 -3.96270722e-01 -7.69956708e-01 3.22776794e-01
1.85895637e-01 -5.60209692e-01 1.18655646e+00 -9.49306071e-01
-1.65382051e+00 3.27039689e-01 -1.35862663e-01 -4.46919799e-01
2.22996280e-01 2.18384620e-02 -9.05485332e-01 -1.77192013e-03
-1.86379962e-02 1.82193846e-01 8.95693958e-01 -1.13867474e+00
-5.96919179e-01 -5.31426668e-01 2.31871046e-02 -8.76468420e-02
-8.20253968e-01 -2.76661605e-01 -1.68572232e-01 -4.76399034e-01
6.33894950e-02 -7.12110639e-01 9.84417200e-02 -5.13048828e-01
-9.77071047e-01 -5.42062461e-01 1.01673079e+00 -8.11188102e-01
1.38809776e+00 -2.18309450e+00 -3.02614987e-01 1.74425989e-01
2.23656386e-01 3.48236501e-01 -8.25855788e-03 4.38933037e-02
-4.20758247e-01 -2.96028048e-01 -4.14300570e-03 -6.77771986e-01
5.48353136e-01 3.51234823e-01 -2.49902621e-01 8.02833021e-01
-1.74307674e-01 9.91202652e-01 -6.72409117e-01 -1.88877821e-01
7.56521463e-01 5.36744595e-01 -5.95324226e-02 2.49520123e-01
-2.58962270e-02 2.73622453e-01 -3.91039401e-01 8.55413795e-01
6.82707965e-01 -3.23543310e-01 5.24315178e-01 -7.97815740e-01
2.40411144e-02 6.22961283e-01 -1.28568077e+00 1.50099897e+00
-6.51909471e-01 4.53476965e-01 2.54804194e-01 -1.23254383e+00
5.28437734e-01 8.66360188e-01 8.26316833e-01 -7.54265547e-01
5.92229605e-01 -1.64940909e-01 -2.05691993e-01 -3.71723145e-01
-3.31237316e-02 -2.50936002e-02 -1.85231194e-01 5.47297478e-01
1.47831649e-01 6.19712293e-01 -6.09090552e-02 -2.26336151e-01
1.09678268e+00 -2.52770960e-01 3.38180602e-01 -3.25957745e-01
4.31063592e-01 -7.38125563e-01 8.33658159e-01 4.22595680e-01
-4.69272137e-01 -1.43513441e-01 2.92891711e-01 -3.08956742e-01
-4.95280117e-01 -1.20635128e+00 2.63106339e-02 9.49231029e-01
-2.31706016e-02 6.14513755e-02 -6.77967489e-01 -6.27938747e-01
-1.27772942e-01 1.10179949e+00 -6.17967784e-01 -5.01107752e-01
-3.30925167e-01 -1.32710576e+00 6.04572237e-01 7.87201405e-01
8.85097086e-01 -9.17632520e-01 -6.65569425e-01 2.48403460e-01
-3.35614651e-01 -1.09031689e+00 -3.79977554e-01 9.48336661e-01
-5.39974391e-01 -9.03160214e-01 -2.86493897e-01 -7.20308363e-01
5.60312569e-01 1.33555815e-01 9.75526750e-01 -4.86516833e-01
-1.57675091e-02 8.29362422e-02 8.58526304e-02 -1.97461262e-01
-3.78999919e-01 2.63664931e-01 6.00070357e-01 2.66071439e-01
9.76862490e-01 -1.38698697e+00 -7.64195085e-01 1.66320100e-01
-7.77251542e-01 -1.36114329e-01 6.09626234e-01 2.98096508e-01
3.47310752e-01 8.17851782e-01 7.49893248e-01 -3.96809876e-01
5.35839498e-01 -6.63748920e-01 -7.53685951e-01 1.95960402e-01
-8.97911787e-01 -6.72324896e-02 8.68962705e-01 -7.60528862e-01
-1.05535507e+00 2.56169796e-01 -9.36439857e-02 -5.51541269e-01
-4.58389342e-01 -1.04766376e-01 -8.26540887e-01 4.68574613e-01
-7.19414875e-02 4.44396943e-01 -5.22742510e-01 -6.23419523e-01
1.89059183e-01 8.97952974e-01 8.06847751e-01 -3.91101629e-01
5.90498447e-01 4.99851227e-01 -1.51091874e-01 -5.52242219e-01
-6.86652720e-01 -4.64974612e-01 -2.95566410e-01 9.01324674e-02
1.11837244e+00 -1.22968507e+00 -1.51753628e+00 7.97851562e-01
-1.19374084e+00 -1.83840483e-01 -3.23017180e-01 3.09113562e-01
-2.86214918e-01 2.86262065e-01 -7.42632449e-01 -1.16364825e+00
-5.78842461e-01 -1.08257365e+00 9.68842089e-01 4.23691452e-01
-3.75343561e-01 -1.03300488e+00 -1.50153682e-01 2.28630468e-01
2.68406242e-01 2.34217942e-01 9.88956332e-01 -7.33611524e-01
-2.63744712e-01 -1.17004365e-01 -1.44617647e-01 8.34167063e-01
5.86096883e-01 -4.40002322e-01 -1.60671401e+00 -3.43128324e-01
6.40293539e-01 4.79994901e-03 7.41532087e-01 6.22206271e-01
1.19111574e+00 -5.03932118e-01 -2.74645746e-01 6.08848095e-01
1.46351504e+00 3.86055768e-01 4.81054127e-01 1.37696788e-02
7.55217671e-01 2.29380503e-01 -3.74374747e-01 6.14438176e-01
5.28882325e-01 2.78401256e-01 7.00097620e-01 -2.22492054e-01
1.39199898e-01 -3.04097325e-01 6.85710847e-01 1.02150035e+00
1.50428072e-01 -5.06523669e-01 -1.96423113e-01 6.56770229e-01
-1.88034439e+00 -8.53486121e-01 9.72286612e-02 1.80239618e+00
6.05096579e-01 7.65827969e-02 1.27205327e-01 5.61133802e-01
6.81938171e-01 4.88935053e-01 -1.18125689e+00 -1.96419239e-01
-1.28300175e-01 -4.38059233e-02 6.99250102e-01 -1.08211458e-01
-1.17526913e+00 7.64393657e-02 5.63435078e+00 8.96020114e-01
-8.48995805e-01 4.84859735e-01 5.21288455e-01 -3.15695494e-01
-4.36194651e-02 -5.39308965e-01 -6.74432874e-01 8.84061694e-01
1.24409008e+00 2.19490349e-01 7.95030117e-01 7.71422267e-01
5.56524158e-01 -1.20246269e-01 -1.38034439e+00 1.21283281e+00
-1.98198184e-01 -6.42615438e-01 -3.09524804e-01 2.77202338e-01
7.81333804e-01 1.58809617e-01 -1.21253850e-02 2.12355301e-01
4.73151803e-01 -7.74421573e-01 5.18430412e-01 3.77194464e-01
5.06750584e-01 -7.57515192e-01 7.75593579e-01 4.84912246e-01
-1.44484711e+00 -4.14506197e-01 5.77003621e-02 -1.35063663e-01
2.65988797e-01 1.03853846e+00 -2.87543416e-01 4.50248808e-01
9.01955128e-01 4.84829754e-01 -1.18505850e-01 3.37680221e-01
-3.01848650e-01 8.99972260e-01 -6.13124549e-01 2.05179065e-01
1.26830237e-02 -2.59568274e-01 1.91521466e-01 9.19868410e-01
1.31489366e-01 -1.26667053e-01 3.82238231e-03 1.26461613e+00
-3.14759552e-01 -6.01076663e-01 -4.20571595e-01 -9.40371975e-02
5.21002591e-01 1.45788515e+00 -6.46618903e-01 -4.36914772e-01
-5.48362076e-01 1.36535382e+00 -1.12851910e-01 5.16780257e-01
-1.06609213e+00 -1.05048835e-01 1.12019956e+00 -4.57338035e-01
3.24794859e-01 1.06805205e-01 -2.12682515e-01 -1.18016481e+00
1.33854032e-01 -6.92597866e-01 1.60767496e-01 -4.90143150e-01
-1.42716646e+00 3.45148891e-01 -1.38245404e-01 -9.05958414e-01
-2.45022506e-01 -3.73088062e-01 -8.03554237e-01 8.19402814e-01
-1.61685288e+00 -1.20180428e+00 -1.59034416e-01 7.75690615e-01
5.07762969e-01 5.70064858e-02 1.02453387e+00 7.71334350e-01
-1.20983589e+00 6.16496265e-01 4.26123202e-01 3.30545485e-01
1.65178865e-01 -1.28151047e+00 2.68219978e-01 8.92898738e-01
-4.40984815e-02 4.80313241e-01 5.69624484e-01 -2.70244151e-01
-1.38974035e+00 -1.29969597e+00 4.99917030e-01 -3.63822371e-01
6.39495313e-01 -4.94715333e-01 -6.19173527e-01 7.31781840e-01
3.86245161e-01 -8.58782679e-02 8.39993775e-01 -1.52172044e-01
-2.11823061e-01 -3.30061108e-01 -1.26736307e+00 1.43725455e-01
7.16774046e-01 -9.14508522e-01 -4.63678151e-01 2.58615673e-01
5.33236206e-01 1.67407617e-01 -9.13043261e-01 3.29772711e-01
4.65743363e-01 -9.69566524e-01 8.48113716e-01 -2.57035345e-01
-5.32473624e-02 -4.18554544e-01 -6.47761822e-01 -1.15976071e+00
-4.10728872e-01 -7.12693453e-01 -9.33984756e-01 1.94803822e+00
-1.21164113e-01 -8.45588684e-01 6.53257668e-01 6.34445667e-01
1.27415359e-01 -5.29589295e-01 -1.05798626e+00 -8.35116208e-01
-3.43575805e-01 -6.16992712e-01 1.02750540e+00 8.27753365e-01
-1.05294421e-01 4.12522554e-01 -4.50699419e-01 6.76461220e-01
6.81561887e-01 1.69361219e-01 8.88150483e-02 -1.14184737e+00
-3.32257062e-01 -5.10208070e-01 -5.24060093e-02 -1.07539666e+00
5.49987018e-01 -5.93682289e-01 2.51113307e-02 -1.29143655e+00
3.81565541e-01 6.64354488e-02 -9.48382080e-01 6.97027087e-01
2.19669491e-01 3.62627715e-01 -2.46820878e-02 2.76298076e-02
-6.32710397e-01 5.55118442e-01 4.25085604e-01 -4.68070835e-01
-2.14319125e-01 2.87942469e-01 -9.31327164e-01 9.78912115e-01
8.82858336e-01 -2.32973829e-01 -4.90786284e-01 -3.54272217e-01
-9.87845138e-02 -3.41468841e-01 6.31788969e-01 -9.96599913e-01
1.43001094e-01 1.12007104e-01 4.38431621e-01 -7.57905662e-01
3.13477695e-01 -1.25724065e+00 3.89327496e-01 4.89888638e-01
3.63068245e-02 -2.90945500e-01 1.29929915e-01 5.52744567e-01
1.00361392e-01 2.49093063e-02 5.33684850e-01 -1.44140068e-02
-6.51128709e-01 1.69448927e-01 -5.16072392e-01 -5.70641637e-01
9.06492651e-01 6.74127489e-02 -2.79298425e-01 -2.95578986e-01
-5.55443168e-01 2.89781898e-01 4.43537124e-02 3.86078656e-01
-5.71461990e-02 -1.56960058e+00 -2.20496915e-02 3.01407397e-01
-2.41999805e-01 -3.69317271e-02 3.48885983e-01 7.90489435e-01
3.93184423e-01 5.20909309e-01 2.53008604e-01 -4.97644633e-01
-9.22936976e-01 7.99913287e-01 3.68550122e-01 -4.13476318e-01
-2.47748867e-01 4.08584028e-01 2.51811713e-01 -1.88557357e-01
5.09769738e-01 -7.39562333e-01 -8.27526767e-03 4.78815943e-01
4.53005940e-01 8.63830209e-01 2.15622380e-01 -5.90361416e-01
-6.32682741e-01 3.36086839e-01 2.85728544e-01 4.29405987e-01
1.37083101e+00 -5.02035677e-01 -1.81053683e-01 5.62053263e-01
1.58922374e+00 -5.09875953e-01 -1.15629315e+00 -9.72348079e-02
-2.82240093e-01 3.96466434e-01 5.83181381e-01 -8.97704244e-01
-1.49836326e+00 7.43982196e-01 1.13116050e+00 5.99375069e-01
1.80358732e+00 4.20803055e-02 1.21095800e+00 1.46234229e-01
3.21312040e-01 -1.21610248e+00 -4.75479782e-01 -1.92334861e-01
4.55122665e-02 -1.31362212e+00 -1.92906067e-01 1.00047186e-01
7.64443502e-02 9.40515518e-01 3.19184750e-01 2.61577487e-01
7.81096101e-01 4.02517647e-01 -7.29308799e-02 -1.62855640e-01
-5.27034163e-01 -2.14712083e-01 7.17280060e-02 5.30651510e-01
-7.58867711e-02 4.98580068e-01 3.63369077e-01 1.12597787e+00
6.88909665e-02 1.28923893e-01 -3.42681594e-02 6.52756274e-01
3.16394866e-03 -5.92583537e-01 -2.36622006e-01 6.59135520e-01
-4.65383261e-01 4.84880693e-02 1.44118801e-01 3.20788234e-01
4.48409200e-01 1.37515402e+00 8.77890065e-02 -5.05845666e-01
3.79074216e-01 1.77710861e-01 2.86567472e-02 -1.55624792e-01
-4.07478184e-01 8.77408311e-02 -2.10999876e-01 -7.39163339e-01
-7.08873093e-01 -5.99892557e-01 -1.01957607e+00 -4.50934649e-01
-5.51113248e-01 -2.11478934e-01 6.10484540e-01 1.15601826e+00
2.39875033e-01 1.03075051e+00 9.57746863e-01 -1.04148233e+00
-7.38891661e-01 -1.07687151e+00 -9.96005297e-01 3.30562502e-01
7.00609088e-01 -4.58971590e-01 -7.04289973e-01 3.52218049e-03] | [16.066070556640625, 7.580445766448975] |
366b1dd5-0966-422b-9f5f-78c27e35344d | on-the-importance-of-signer-overlap-for-sign | 2303.10782 | null | https://arxiv.org/abs/2303.10782v1 | https://arxiv.org/pdf/2303.10782v1.pdf | On the Importance of Signer Overlap for Sign Language Detection | Sign language detection, identifying if someone is signing or not, is becoming crucially important for its applications in remote conferencing software and for selecting useful sign data for training sign language recognition or translation tasks. We argue that the current benchmark data sets for sign language detection estimate overly positive results that do not generalize well due to signer overlap between train and test partitions. We quantify this with a detailed analysis of the effect of signer overlap on current sign detection benchmark data sets. Comparing accuracy with and without overlap on the DGS corpus and Signing in the Wild, we observed a relative decrease in accuracy of 4.17% and 6.27%, respectively. Furthermore, we propose new data set partitions that are free of overlap and allow for more realistic performance assessment. We hope this work will contribute to improving the accuracy and generalization of sign language detection systems. | ['Oscar Koller', 'Alessandro Manzotti', 'Cyrine Chaabani', 'Stephan Huber', 'Abhilash Pal'] | 2023-03-19 | null | null | null | null | ['sign-language-recognition'] | ['computer-vision'] | [ 5.16729236e-01 -2.03483105e-01 -9.11188200e-02 -3.74674857e-01
-1.03516126e+00 -7.33325899e-01 6.65464580e-01 -5.02688706e-01
-6.63608313e-01 5.32949388e-01 4.61073160e-01 -4.31565911e-01
-2.00511720e-02 -3.25209409e-01 -2.17207819e-01 -6.51952505e-01
-1.96646407e-01 4.61627185e-01 5.12587786e-01 -7.87083507e-02
2.34443277e-01 5.46385407e-01 -1.85559392e+00 5.97144902e-01
6.55924976e-01 5.75817585e-01 -2.75709748e-01 9.90760744e-01
1.10448673e-01 6.55498326e-01 -9.82487082e-01 -7.78307840e-02
4.10879165e-01 -6.60786390e-01 -5.88710487e-01 -2.19152957e-01
1.15714061e+00 -5.33566177e-01 -2.17333391e-01 5.68283439e-01
1.04122484e+00 -1.05551124e-01 6.84645712e-01 -1.29161644e+00
1.22535489e-01 3.35878909e-01 -2.52768040e-01 1.93994418e-02
6.19706810e-01 4.74644005e-01 8.22814047e-01 -4.90701824e-01
1.06458509e+00 9.95100200e-01 6.49209261e-01 1.03294230e+00
-8.79393578e-01 -9.98292506e-01 -2.45866012e-02 1.93550587e-01
-1.34390450e+00 -6.78435326e-01 4.85768437e-01 -5.28799713e-01
6.68027341e-01 7.54467607e-01 8.12810481e-01 9.98133361e-01
-6.17183149e-01 1.31962609e+00 1.41495502e+00 -6.71505928e-01
-9.06829815e-03 -1.31283075e-01 3.68439496e-01 5.59522867e-01
4.00960565e-01 2.82715887e-01 -6.21284425e-01 -2.58735895e-01
5.44951200e-01 -6.95443332e-01 -4.21588212e-01 -3.48324060e-01
-1.13597286e+00 4.30162698e-01 -6.69957921e-02 4.90390182e-01
-1.21488027e-01 2.42677838e-01 4.57739681e-01 4.39772606e-01
-1.89970568e-01 3.09653133e-01 -3.67287606e-01 -7.61971235e-01
-1.08865988e+00 4.37315494e-01 9.70896244e-01 7.58687019e-01
-4.61222008e-02 -2.01006114e-01 -4.50455815e-01 7.32502103e-01
1.58052281e-01 6.29041016e-01 3.28252226e-01 -8.13784182e-01
3.94197047e-01 6.37623131e-01 1.27289578e-01 -2.40947202e-01
-3.56285185e-01 -1.00319803e-01 9.95176584e-02 8.14392328e-01
1.22400856e+00 -2.86320269e-01 -1.34329987e+00 1.49037027e+00
-1.37284979e-01 1.53526738e-01 9.41465900e-04 9.74097967e-01
1.01167619e+00 -1.97021961e-01 2.63608456e-01 3.81272674e-01
1.40081906e+00 -1.89182565e-01 -3.70191813e-01 -1.58983305e-01
1.11795390e+00 -9.59915340e-01 1.20771706e+00 4.96885210e-01
-5.99436581e-01 6.50094226e-02 -1.02871048e+00 1.23492867e-01
-2.49930680e-01 5.27966380e-01 6.24368906e-01 1.11833787e+00
-9.02073562e-01 9.08931121e-02 -6.71532869e-01 -8.46581340e-01
4.43288535e-01 5.09776592e-01 -3.70759666e-01 -1.46256074e-01
-8.04253459e-01 1.09907222e+00 1.07131891e-01 -3.86494175e-02
-2.24491492e-01 -3.53127211e-01 -5.44269562e-01 -4.75465536e-01
-9.38421488e-03 -1.56608075e-01 1.13070905e+00 -6.29102409e-01
-1.28204751e+00 1.40369570e+00 -8.58833417e-02 -2.56585598e-01
1.18195248e+00 -2.34738231e-01 -5.62162817e-01 -1.46381855e-01
-2.89139390e-01 5.38896382e-01 4.30900365e-01 -1.36582899e+00
-7.98195779e-01 -2.55950242e-01 -2.41658241e-01 1.25410154e-01
1.60775319e-01 4.78465170e-01 -4.06028181e-01 -3.46149683e-01
2.86515713e-01 -1.02635968e+00 1.89408794e-01 9.53541100e-02
-1.66514516e-01 -6.87684715e-02 1.02739036e+00 -1.00408065e+00
1.14164066e+00 -1.92711675e+00 -5.04834354e-01 6.38400733e-01
6.33099601e-02 6.67235255e-01 -2.70191312e-01 2.81031936e-01
2.85312593e-01 -3.66291329e-02 -1.89256400e-01 7.93211386e-02
-5.87319992e-02 2.90044636e-01 -2.29489524e-02 4.94816154e-01
1.94116123e-02 8.23427975e-01 -8.69683862e-01 -6.29938960e-01
3.21933538e-01 5.05603194e-01 -3.79535347e-01 -2.76013613e-01
2.71906823e-01 5.07126808e-01 -3.08159471e-01 9.24748778e-01
4.69700307e-01 2.66764462e-01 2.17262611e-01 1.46817982e-01
-1.99479699e-01 4.57813501e-01 -1.37228954e+00 1.12648475e+00
-3.29896122e-01 1.40346694e+00 9.73243043e-02 -3.75214636e-01
9.01302755e-01 2.71687776e-01 4.54814374e-01 -5.91097891e-01
3.08629066e-01 6.43335223e-01 5.74673355e-01 -6.61257207e-01
4.70516622e-01 -3.34242806e-02 1.41999856e-01 5.22886157e-01
-3.54348183e-01 -1.65055290e-01 4.45094258e-01 -1.72524992e-02
1.42802942e+00 -4.98583205e-02 -1.56640280e-02 -7.55631402e-02
4.16290462e-01 -1.72990151e-02 2.29298010e-01 8.38865280e-01
-5.55226445e-01 8.50566030e-01 3.45014542e-01 -1.00393221e-01
-7.43526638e-01 -1.09818780e+00 -3.64662200e-01 8.82470965e-01
-4.58902027e-03 -2.02936381e-01 -5.30507445e-01 -8.24820578e-01
1.34078652e-01 7.19825506e-01 -3.99327308e-01 9.24161226e-02
-8.74111474e-01 -6.21712625e-01 1.31164372e+00 7.24254489e-01
2.05743760e-01 -1.25460207e+00 -9.89158213e-01 -2.13126466e-01
-1.28723785e-01 -1.09910560e+00 -1.69410691e-01 -9.64868814e-02
-6.63716853e-01 -1.45739889e+00 -1.14646852e+00 -9.47202325e-01
6.93205118e-01 -1.34481162e-01 8.56568098e-01 4.14336890e-01
-4.12188709e-01 5.80124855e-01 -5.22640824e-01 -3.55054259e-01
-8.67217302e-01 -8.63904208e-02 -1.94802076e-01 -3.48592609e-01
7.94526696e-01 -1.04182862e-01 -5.18124640e-01 6.02610052e-01
-6.01313829e-01 -1.30265847e-01 6.32191777e-01 7.46997178e-01
-2.34231368e-01 -7.81819344e-01 1.03598796e-01 -4.48904783e-01
7.83870161e-01 2.59147346e-01 -4.58879232e-01 4.08001542e-01
-3.60586107e-01 2.65822560e-01 -2.62945652e-01 -5.80439806e-01
-8.99377942e-01 2.96211019e-02 -3.01525146e-02 3.37157607e-01
-4.32144701e-01 -1.19779229e-01 7.94861987e-02 -5.15722990e-01
8.69382501e-01 1.22511171e-01 8.44663158e-02 -4.70341444e-01
1.04087502e-01 1.20069671e+00 4.19189662e-01 -4.82961774e-01
5.35596430e-01 5.69916368e-01 -1.63633332e-01 -1.03468633e+00
5.93774803e-02 -9.51526940e-01 -5.97878754e-01 -5.51825345e-01
3.47162724e-01 -5.24557948e-01 -7.83849239e-01 6.43064499e-01
-9.47279155e-01 -4.91031528e-01 -3.21954638e-01 8.13430607e-01
-4.02546436e-01 5.30046463e-01 -2.70167410e-01 -1.14939439e+00
-9.74214748e-02 -1.08779991e+00 1.24456906e+00 -3.61808278e-02
-7.29559541e-01 -4.03885543e-01 1.94549173e-01 6.21729195e-01
3.13656449e-01 3.50259632e-01 4.19838428e-01 -6.29120767e-01
-3.33880603e-01 -8.43584836e-01 -3.70475560e-01 4.19831991e-01
6.76697716e-02 8.44164193e-02 -1.09745097e+00 -1.55178621e-01
-9.13314879e-01 -2.58735448e-01 9.93885696e-01 1.16741069e-01
3.32268804e-01 6.62912950e-02 -4.07734543e-01 1.01389527e-01
8.64043832e-01 1.65905148e-01 7.90686965e-01 2.57379442e-01
4.62958038e-01 7.97858953e-01 5.37024558e-01 2.15268597e-01
3.79162654e-02 9.18257535e-01 -7.02666268e-02 -8.67666900e-02
-7.67041385e-01 -1.56516194e-01 5.45318604e-01 2.70664990e-01
-5.55230558e-01 -1.15629077e-01 -1.27089036e+00 6.21281385e-01
-1.58908617e+00 -1.11338389e+00 -4.65459555e-01 2.38003492e+00
6.25615954e-01 -9.86899212e-02 4.36780930e-01 7.36233771e-01
6.84003592e-01 -8.96932036e-02 -1.25173911e-01 -3.02585751e-01
-3.08398187e-01 3.54497105e-01 9.23868775e-01 6.88575268e-01
-9.89700258e-01 9.88272548e-01 6.63784027e+00 4.73889261e-01
-1.34159482e+00 -2.06780136e-01 -1.91287603e-02 -2.15260059e-01
4.09947010e-03 -1.97905693e-02 -7.42984176e-01 3.35014701e-01
2.81188995e-01 1.80432349e-01 4.57230546e-02 5.52239895e-01
2.34561861e-01 -2.83048838e-01 -1.03611088e+00 1.04702675e+00
2.10077301e-01 -7.74598241e-01 -2.64278859e-01 2.06570148e-01
6.10436201e-01 3.41348648e-01 -3.58742952e-01 2.33202428e-01
4.28676158e-01 -8.18053424e-01 6.89229965e-01 2.70013899e-01
1.02513754e+00 -1.09143935e-01 9.22310293e-01 1.45641476e-01
-9.28504467e-01 1.23113699e-01 4.58831519e-01 -2.16482013e-01
3.68403792e-01 -1.64717063e-01 -1.36584461e+00 -7.26646259e-02
2.93213934e-01 1.46399423e-01 -5.66061914e-01 1.62485206e+00
-4.35396820e-01 7.88251758e-01 -8.10625613e-01 -5.42000294e-01
6.58143386e-02 1.44031629e-01 8.30519140e-01 1.39522779e+00
2.86153316e-01 -2.60381371e-01 -1.14426747e-01 2.30630443e-01
2.78163940e-01 2.43194476e-01 -4.66604173e-01 1.83858469e-01
2.15365708e-01 5.72632670e-01 -7.55939186e-01 -3.81340712e-01
-2.05883637e-01 9.93385196e-01 -3.02786171e-01 5.90402901e-01
-5.53490400e-01 -3.71702701e-01 7.60637224e-01 4.20028776e-01
2.37080138e-02 -2.88335472e-01 -5.01240075e-01 -9.77610469e-01
4.79338199e-01 -8.37379873e-01 4.96723205e-01 -4.49678510e-01
-8.38551879e-01 2.48873919e-01 -7.89840072e-02 -1.62636864e+00
-5.03412783e-01 -1.06456316e+00 -6.16914511e-01 6.39760137e-01
-1.29756451e+00 -1.28913116e+00 -6.60617232e-01 2.10322872e-01
1.33681269e-02 6.43838271e-02 7.48608232e-01 4.96362299e-01
1.07048206e-01 1.04123902e+00 2.71982746e-03 6.47384405e-01
8.22477877e-01 -9.22917128e-01 3.69464844e-01 8.03321719e-01
3.14423710e-01 4.44708616e-01 8.74921203e-01 -4.82554376e-01
-8.71136904e-01 -3.69364530e-01 1.03143847e+00 -8.67276073e-01
4.54304159e-01 -2.77829975e-01 -3.72329593e-01 3.94867331e-01
-4.42210138e-01 -1.35311812e-01 4.99022603e-01 1.88155949e-01
-4.18049306e-01 2.52058268e-01 -1.29347670e+00 8.65490139e-01
1.52104425e+00 -4.21993494e-01 -5.02776682e-01 1.26988113e-01
-4.01001245e-01 -3.91769677e-01 -4.52479810e-01 6.10206544e-01
1.50427067e+00 -9.31779087e-01 7.61679351e-01 -7.02017903e-01
-1.14501625e-01 -2.46803060e-01 -1.48569897e-01 -7.19775081e-01
3.15137208e-01 -5.14998972e-01 1.80000871e-01 9.71784472e-01
5.37991405e-01 -6.61293149e-01 1.44710279e+00 7.68980563e-01
1.35955438e-01 -3.81448716e-01 -1.11908412e+00 -1.14339578e+00
-8.77325907e-02 -8.56947720e-01 2.58690983e-01 8.18283021e-01
2.36967579e-01 -2.23031327e-01 -2.37436354e-01 -2.23311633e-01
4.76761997e-01 -8.80339667e-02 1.20874310e+00 -1.23572731e+00
-1.11163571e-01 -1.02109718e+00 -1.05721283e+00 -1.12667513e+00
-2.78956652e-01 -7.79093206e-01 3.02613914e-01 -1.63074172e+00
-9.45389345e-02 -5.59854209e-01 -1.49721310e-01 6.42455816e-01
-8.79298672e-02 5.72650731e-01 3.81022304e-01 1.16686404e-01
-3.20316166e-01 -3.69388126e-02 1.21464396e+00 -5.69954664e-02
-4.07601416e-01 2.41928950e-01 -1.59875035e-01 6.71914935e-01
8.22437227e-01 -1.97645590e-01 2.30144441e-01 -2.47450426e-01
-2.48140261e-01 -5.41891515e-01 5.76790750e-01 -1.24226999e+00
1.28073702e-02 7.93472975e-02 -3.08278929e-02 -5.22815764e-01
3.15118104e-01 -5.97521186e-01 -4.37228471e-01 8.13807905e-01
-2.84251392e-01 -5.46217680e-01 1.78868294e-01 -8.93801972e-02
-5.28905392e-02 -1.62023772e-02 6.19394124e-01 1.64617345e-01
-1.11708879e+00 -2.20442250e-01 -4.35412496e-01 2.11758241e-01
7.92785823e-01 -9.09501612e-01 -1.67649865e-01 -6.44151986e-01
-5.88947833e-01 1.58796296e-01 4.64261740e-01 6.39336705e-01
3.94427836e-01 -1.10792661e+00 -9.33272243e-01 2.97065139e-01
6.67550445e-01 -6.07478201e-01 -1.76773578e-01 8.50058317e-01
-8.78110349e-01 3.54544669e-01 -2.26087034e-01 -4.67669517e-01
-2.17136168e+00 -6.78279281e-01 2.69449741e-01 -7.68987089e-02
-5.08425355e-01 9.82746124e-01 -4.08027679e-01 -4.98251259e-01
5.21097958e-01 -6.29931748e-01 2.91571338e-02 -8.48912299e-02
4.62150753e-01 6.59563780e-01 -5.26548550e-02 -8.56789052e-01
-7.03483999e-01 7.43969560e-01 1.55931652e-01 -4.73307848e-01
7.46107638e-01 5.75976908e-01 4.59842622e-01 2.37429857e-01
9.08425212e-01 3.90165061e-01 -9.71667826e-01 1.72373727e-01
3.78832579e-01 -5.00086069e-01 -2.48029530e-01 -1.27020252e+00
-6.38322413e-01 7.00701058e-01 1.04071200e+00 -2.50841200e-01
8.20877016e-01 2.66752928e-01 5.45595169e-01 3.85089457e-01
4.60680366e-01 -1.17688346e+00 -5.08737326e-01 4.19207841e-01
9.98621941e-01 -1.35158491e+00 5.53312749e-02 -3.14532638e-01
-5.61870098e-01 8.43761683e-01 4.32618439e-01 -5.10394387e-02
2.23053247e-01 5.52032828e-01 8.00145864e-01 -2.42155626e-01
-5.90717979e-02 -6.05836034e-01 5.38095415e-01 8.10285807e-01
7.89037883e-01 3.87513220e-01 -9.60130870e-01 9.69022810e-02
-4.27708924e-01 3.86038154e-01 1.73157260e-01 1.25012839e+00
-3.48413020e-01 -1.39850760e+00 -5.56901157e-01 7.08479702e-01
-1.08708367e-01 2.04484808e-04 -1.06488431e+00 1.23520899e+00
2.74372160e-01 7.60710120e-01 -1.77259430e-01 -3.83942187e-01
7.22756684e-01 4.29982424e-01 8.92344296e-01 -3.80272657e-01
-5.96230388e-01 -3.68781745e-01 7.09108591e-01 -3.59838158e-01
-2.37481445e-01 -1.03678071e+00 -1.34522295e+00 -5.39755374e-02
-6.08228087e-01 -1.73611149e-01 7.11695492e-01 9.25001323e-01
1.25976861e-01 2.19142020e-01 -2.67000377e-01 -4.76812661e-01
-5.19284844e-01 -1.21143639e+00 -4.35634404e-01 6.83546841e-01
1.25881463e-01 -5.98144770e-01 -3.63507181e-01 -6.64580762e-02] | [9.13172435760498, -6.446225643157959] |
8fd1f449-c902-4785-9f01-a5bdd450125d | sparse-relational-reasoning-with-object | 2207.07512 | null | https://arxiv.org/abs/2207.07512v1 | https://arxiv.org/pdf/2207.07512v1.pdf | Sparse Relational Reasoning with Object-Centric Representations | We investigate the composability of soft-rules learned by relational neural architectures when operating over object-centric (slot-based) representations, under a variety of sparsity-inducing constraints. We find that increasing sparsity, especially on features, improves the performance of some models and leads to simpler relations. Additionally, we observe that object-centric representations can be detrimental when not all objects are fully captured; a failure mode to which CNNs are less prone. These findings demonstrate the trade-offs between interpretability and performance, even for models designed to tackle relational tasks. | ['Murray Shanahan', 'Alessandra Russo', 'Alex F. Spies'] | 2022-07-15 | null | null | null | null | ['relational-reasoning'] | ['natural-language-processing'] | [ 2.34935626e-01 6.02248728e-01 -6.74206197e-01 -4.66479033e-01
-4.46591862e-02 -4.55308646e-01 8.32554221e-01 2.43534446e-01
-3.75462472e-02 4.61664319e-01 5.63639939e-01 -4.12686229e-01
-5.61476886e-01 -8.85002971e-01 -8.42568815e-01 -1.55425921e-01
-1.39254272e-01 5.80825269e-01 -3.00582759e-02 -3.49020779e-01
-2.15114370e-01 5.77311397e-01 -1.57485592e+00 6.01182282e-01
4.98486876e-01 1.00989485e+00 -2.30661437e-01 2.80415446e-01
5.21853045e-02 1.36648071e+00 -4.31444168e-01 -5.39142132e-01
3.70352238e-01 7.93906301e-02 -9.50863898e-01 -7.48309940e-02
6.49101973e-01 -3.10892463e-01 -7.38286972e-01 7.31317699e-01
-3.04655820e-01 2.17306823e-01 7.51227677e-01 -1.08069754e+00
-1.23567653e+00 1.13066101e+00 -8.51008818e-02 4.25688952e-01
2.31418878e-01 2.64915109e-01 1.89307582e+00 -8.17913234e-01
8.81799877e-01 1.49669421e+00 6.61594927e-01 1.78864509e-01
-1.65543759e+00 -3.62425983e-01 4.07048464e-01 -1.29266948e-01
-1.18046999e+00 -7.92894483e-01 4.84690696e-01 -3.11389983e-01
1.54384089e+00 3.06183934e-01 4.04741824e-01 1.03983688e+00
1.63228855e-01 4.91037875e-01 8.68771732e-01 -1.58416688e-01
-8.40168744e-02 -2.54076459e-02 3.54200482e-01 6.27103209e-01
9.67584133e-01 1.22553408e-01 -7.08389938e-01 -1.04712568e-01
8.56386304e-01 2.57168025e-01 2.55994797e-02 -3.70512456e-01
-1.02085078e+00 7.34277904e-01 6.97946429e-01 4.44107771e-01
-3.54512930e-01 3.89647275e-01 3.13659936e-01 8.15776706e-01
3.70553672e-01 1.28294635e+00 -7.19306231e-01 5.06913811e-02
-5.69461048e-01 3.83584917e-01 8.87481749e-01 1.01675212e+00
7.48900473e-01 1.62093773e-01 -2.65912235e-01 7.11930513e-01
1.32710457e-01 2.62021404e-02 2.23323002e-01 -1.08509517e+00
5.47782123e-01 1.02909291e+00 -2.67436534e-01 -1.17763126e+00
-6.48897648e-01 -7.99251318e-01 -5.85870087e-01 -1.79177463e-01
4.59742963e-01 1.21595204e-01 -8.54279995e-01 1.99602115e+00
-1.52357996e-01 -5.13524175e-01 1.26494065e-01 8.01551044e-01
1.02296579e+00 3.50598365e-01 3.66413444e-01 -3.87907913e-03
1.15961564e+00 -8.62667978e-01 -6.14248633e-01 -7.15482235e-01
9.76148009e-01 -3.46905738e-01 1.03678155e+00 1.99743107e-01
-1.12083256e+00 -4.63955373e-01 -1.01612735e+00 -5.12887359e-01
-4.76193339e-01 -6.86249062e-02 1.17651701e+00 3.01585138e-01
-9.18266654e-01 7.58347988e-01 -7.45206475e-01 -4.67744052e-01
6.00457788e-01 4.01517600e-01 -5.52298844e-01 3.52079906e-02
-1.07672620e+00 1.38459563e+00 5.23281157e-01 4.67549786e-02
-8.14626873e-01 -8.34434628e-01 -8.85526836e-01 5.63857436e-01
6.32105529e-01 -6.48834467e-01 1.08303285e+00 -1.05760467e+00
-9.69220519e-01 8.85319829e-01 -5.53739909e-03 -6.41067147e-01
7.31564611e-02 -4.05674964e-01 -3.76736820e-01 -9.79321599e-02
-3.15365419e-02 5.84801912e-01 5.34653306e-01 -1.37085724e+00
-2.06467122e-01 -2.15033531e-01 7.11164117e-01 1.88395634e-01
-4.05245602e-01 -1.50154725e-01 -3.05376900e-03 -5.58785498e-01
2.16685191e-01 -8.38705659e-01 -8.80196691e-02 2.80543398e-02
-6.67321146e-01 -4.83565211e-01 6.02367997e-01 -3.41013074e-01
1.19888067e+00 -2.04391837e+00 4.37133402e-01 2.50658572e-01
4.96904820e-01 5.55612184e-02 -2.61219710e-01 4.51963633e-01
-4.21023458e-01 4.41919208e-01 2.05871746e-01 -3.05034041e-01
2.08205022e-02 7.08658159e-01 -3.53753030e-01 1.39135391e-01
7.16501951e-01 1.15224111e+00 -6.27735555e-01 -3.30031484e-01
-1.52409494e-01 4.41673547e-02 -7.19644666e-01 6.15727454e-02
-5.40231824e-01 -7.32555538e-02 -2.01070413e-01 8.14042687e-01
1.68542489e-01 -6.37491584e-01 6.83762908e-01 -2.17300519e-01
2.35307992e-01 9.90627825e-01 -8.09923232e-01 1.37775731e+00
-5.62333584e-01 5.53095579e-01 -9.55818966e-02 -1.06277931e+00
7.04442322e-01 1.56370029e-01 8.72128233e-02 -7.08167136e-01
-7.35945627e-02 1.87968031e-01 6.38378739e-01 -3.50690633e-01
6.57594621e-01 -2.09584564e-01 1.14684470e-01 4.73259687e-01
3.88704360e-01 -1.92614086e-02 3.02740455e-01 4.42365050e-01
1.30033767e+00 1.05854280e-01 3.48662734e-01 -4.91050690e-01
-1.45907968e-01 5.37312729e-03 5.58072209e-01 9.24951971e-01
3.61120939e-01 3.84857506e-01 1.07929003e+00 -7.23256409e-01
-9.95748281e-01 -8.98055792e-01 -2.77062386e-01 1.42492044e+00
-1.67429537e-01 -7.64313459e-01 4.35346290e-02 -8.07920814e-01
4.45827752e-01 7.07183063e-01 -9.32505727e-01 -5.93002200e-01
-5.41019320e-01 -4.00411159e-01 5.20444036e-01 7.47299790e-01
-1.59220889e-01 -8.84210110e-01 -4.37715322e-01 -8.05105865e-02
1.88052773e-01 -1.18718004e+00 1.76263154e-01 8.04449022e-01
-1.12014842e+00 -1.00892115e+00 2.39864543e-01 -3.00368100e-01
7.24910736e-01 1.47091970e-01 1.61704028e+00 7.23057568e-01
2.15140596e-01 1.54244885e-01 -3.77135038e-01 -2.09991783e-01
-3.32127005e-01 5.92527866e-01 -1.92635413e-02 -3.57618541e-01
2.09656030e-01 -6.96753860e-01 -1.59694418e-01 2.18735069e-01
-9.49821234e-01 -1.50507897e-01 7.24456310e-01 8.87960672e-01
1.64318472e-01 -3.80237736e-02 4.44102526e-01 -1.38637376e+00
6.23090923e-01 -7.09957838e-01 -1.12081677e-01 2.88147300e-01
-6.92835629e-01 5.46373487e-01 7.15442061e-01 -4.96018171e-01
-9.58286524e-01 -2.08135188e-01 3.71165514e-01 -4.73773628e-01
7.61763975e-02 7.52839446e-01 6.17791452e-02 5.79578616e-02
1.05579269e+00 -1.30629450e-01 1.00665569e-01 -4.36588556e-01
4.07689452e-01 2.07461476e-01 2.03972876e-01 -1.07871175e+00
7.69655168e-01 3.84193361e-01 1.28489017e-01 -5.43913305e-01
-1.36022389e+00 1.19712539e-02 -3.84440452e-01 3.39911848e-01
5.38046598e-01 -1.04220593e+00 -5.64413786e-01 -2.56772667e-01
-1.07109046e+00 -4.32446390e-01 -7.06661642e-01 1.10469401e-01
-3.65901351e-01 -1.63866252e-01 -6.24383092e-01 -4.04182762e-01
6.43355101e-02 -1.02847171e+00 8.10986876e-01 -1.26065174e-02
-7.51619518e-01 -1.00315440e+00 -2.66463697e-01 2.08567232e-01
4.60450798e-01 -6.03378601e-02 1.47789097e+00 -1.07105076e+00
-7.40426302e-01 9.85662192e-02 -4.95339274e-01 1.79543823e-01
1.54718354e-01 1.08084649e-01 -8.50603163e-01 -6.02135956e-02
-3.54681730e-01 -7.30078757e-01 1.10161495e+00 1.21422216e-01
1.08359027e+00 -7.06864119e-01 -3.71910185e-01 6.06364429e-01
1.25047064e+00 -7.12912157e-02 5.00382006e-01 2.16853604e-01
8.49228680e-01 6.56158984e-01 2.39994243e-01 2.25736454e-01
5.57652295e-01 6.95385873e-01 6.51169002e-01 -7.73336291e-02
-1.88970998e-01 -4.71862465e-01 1.00696042e-01 4.20064956e-01
-2.98913687e-01 -1.28000066e-01 -1.03153467e+00 6.11172497e-01
-1.91652918e+00 -7.56996155e-01 3.02573368e-02 1.79370403e+00
1.09272885e+00 4.19395328e-01 1.35202199e-01 1.39058719e-03
2.40295857e-01 4.42290068e-01 -4.19431448e-01 -6.40609324e-01
-1.91418156e-01 2.78859973e-01 3.24916899e-01 3.68464559e-01
-9.06313419e-01 1.10282362e+00 7.96570778e+00 3.96233201e-01
-1.03311968e+00 4.80222553e-02 5.96382022e-01 -3.39258194e-01
-8.07576537e-01 2.48418897e-01 -6.82523787e-01 4.08979766e-02
9.60088193e-01 1.94058746e-01 5.07813394e-01 8.78271341e-01
-3.64160419e-01 4.41532284e-02 -1.90983605e+00 4.14742112e-01
-1.20307609e-01 -1.62605369e+00 3.23297232e-01 1.58467859e-01
6.88868225e-01 -1.10006444e-01 1.04470626e-01 4.83261228e-01
8.00889373e-01 -1.52815616e+00 9.43650424e-01 3.91368210e-01
5.06604791e-01 -3.70591909e-01 4.79988962e-01 3.06366701e-02
-7.83058286e-01 -2.73668766e-01 -3.96232158e-01 -6.32526994e-01
-3.01383197e-01 5.45552492e-01 -6.76973760e-01 4.30088371e-01
6.15818024e-01 8.82326603e-01 -8.58966410e-01 3.15157890e-01
-2.75900632e-01 4.47865427e-01 -4.34490412e-01 2.77561843e-01
1.21364214e-01 1.89221859e-01 3.09641570e-01 1.10601425e+00
-1.49543300e-01 5.59966005e-02 1.18618324e-01 1.22228193e+00
-1.49819210e-01 -4.43608969e-01 -1.06709146e+00 -3.06562155e-01
6.65602565e-01 1.10066831e+00 -4.57608312e-01 -3.37045789e-01
-4.89720583e-01 1.63011044e-01 9.39174592e-01 4.36983109e-01
-4.21740562e-01 1.42304689e-01 8.32382619e-01 3.76293182e-01
3.07130307e-01 -4.08573806e-01 -8.46588135e-01 -1.19302666e+00
-5.57400240e-03 -1.12007153e+00 4.83220011e-01 -6.91665888e-01
-1.20818067e+00 2.98751950e-01 1.76784247e-01 -6.00442290e-01
-2.29274869e-01 -6.43460631e-01 -2.20199093e-01 5.57572603e-01
-1.21166682e+00 -1.22251332e+00 1.18391834e-01 3.57960820e-01
8.75600129e-02 3.00346296e-02 8.00183892e-01 8.02125931e-02
-5.82178652e-01 6.66534841e-01 -5.01838744e-01 1.15785927e-01
1.71740174e-01 -1.06225383e+00 2.02609360e-01 6.92210257e-01
4.84748870e-01 1.30990624e+00 7.87466526e-01 -5.86081922e-01
-1.62171769e+00 -9.07406926e-01 9.96454358e-01 -7.79147446e-01
6.87756836e-01 -2.19338939e-01 -9.66161132e-01 1.40421402e+00
8.21495205e-02 3.31351042e-01 4.84166980e-01 1.09511554e+00
-1.02630818e+00 -1.97703511e-01 -8.03039014e-01 6.03102803e-01
1.49507701e+00 -8.23001921e-01 -8.61216009e-01 2.70687848e-01
1.09700406e+00 -4.47809428e-01 -1.25392783e+00 4.04908150e-01
5.10777831e-01 -8.41732800e-01 1.01553452e+00 -1.10562849e+00
1.05330431e+00 -8.27805251e-02 -3.31875205e-01 -1.26453900e+00
-7.19133377e-01 -2.38272101e-01 -4.70595360e-01 9.71869886e-01
9.42771435e-01 -7.40255415e-01 5.51196337e-01 7.37474918e-01
-8.15663785e-02 -8.18157315e-01 -7.38104343e-01 -8.01387429e-01
-3.43368901e-03 -3.23639601e-01 6.79471791e-01 1.18773723e+00
1.56423569e-01 4.97599959e-01 -1.46390185e-01 5.35586989e-03
7.83681050e-02 1.03448801e-01 4.48925227e-01 -1.37752378e+00
-3.21650326e-01 -5.01617610e-01 -2.91759670e-01 -7.06771970e-01
3.48219246e-01 -1.07393098e+00 -3.45961124e-01 -1.46717179e+00
2.17711218e-02 -8.68347824e-01 -4.80016917e-01 1.16495800e+00
8.80683307e-03 4.16376032e-02 4.75779802e-01 4.09386128e-01
-6.10373616e-01 1.98900059e-01 1.22532070e+00 -1.55153260e-01
-7.86024183e-02 -2.71885365e-01 -1.27304995e+00 5.89791775e-01
4.81010050e-01 -5.58969557e-01 -5.50535142e-01 -9.10290003e-01
8.68599713e-01 -1.77252889e-01 3.52952600e-01 -7.53886938e-01
4.53129299e-02 -3.83065253e-01 2.98394471e-01 -1.92327518e-02
4.37848657e-01 -9.24530864e-01 2.66178012e-01 3.26575100e-01
-9.33629274e-01 1.51300371e-01 8.58633071e-02 3.94864827e-01
-2.35471427e-01 1.53645705e-02 5.19972265e-01 -2.87715346e-01
-4.12265062e-01 8.54692608e-02 -9.20402780e-02 1.51555821e-01
7.20256567e-01 -1.21506736e-01 -5.33203363e-01 -2.12026462e-01
-7.66993880e-01 -4.76022512e-02 3.38681161e-01 6.06790483e-01
2.73010671e-01 -1.37295771e+00 -3.77147794e-01 2.33594209e-01
3.09913874e-01 2.10771367e-01 -3.55464011e-01 8.22053134e-01
-3.59873354e-01 6.29429996e-01 -2.13520810e-01 -4.81835663e-01
-6.44119263e-01 5.06601095e-01 3.45685303e-01 -4.60277915e-01
-5.05334675e-01 9.20377672e-01 2.35835105e-01 -5.01957476e-01
1.47572786e-01 -8.21467161e-01 -4.14903387e-02 4.74859364e-02
-5.45210540e-02 9.93561116e-04 2.82457083e-01 -3.07594806e-01
-4.44159418e-01 8.39275792e-02 -4.04500425e-01 3.20507586e-01
1.52388465e+00 3.29975247e-01 -3.59604150e-01 4.28233713e-01
7.11684644e-01 -1.82971776e-01 -1.09492481e+00 -5.55564225e-01
1.40558675e-01 -4.01368976e-01 7.36890137e-02 -7.43245065e-01
-1.24928367e+00 5.17039359e-01 -3.14224511e-01 6.79895282e-01
5.82788825e-01 2.39883035e-01 1.77283883e-01 7.41765380e-01
2.80761898e-01 -8.65700841e-01 -2.82911537e-03 6.35573685e-01
9.77419317e-01 -1.09353042e+00 4.51385587e-01 -5.28201461e-01
-5.53216219e-01 9.48772669e-01 8.75073791e-01 -2.92287320e-01
5.42734683e-01 3.32208902e-01 -5.05743206e-01 -4.26739365e-01
-1.48016489e+00 -3.04830521e-01 1.88859850e-01 5.83644807e-01
6.22514009e-01 2.11187422e-01 -3.58887985e-02 4.90019679e-01
-3.37849230e-01 -2.63229012e-01 3.57750475e-01 7.93880343e-01
-1.49101883e-01 -9.41860557e-01 -1.24232925e-01 9.15447712e-01
-3.71084422e-01 -4.44922715e-01 -6.41579986e-01 9.38117087e-01
1.74537987e-01 7.20234454e-01 4.10788357e-01 -4.24041003e-01
2.91992784e-01 1.56307429e-01 4.29156721e-01 -1.08305728e+00
-8.88568699e-01 -6.16294742e-01 7.05900073e-01 -8.04439425e-01
-2.83214360e-01 -4.46705937e-01 -9.94092107e-01 -4.04879749e-01
-2.82006353e-01 -2.78498590e-01 1.33220151e-01 1.17298424e+00
6.55598938e-01 6.67230546e-01 1.93772480e-01 -3.21211845e-01
-6.42206967e-01 -9.38062847e-01 -4.48901922e-01 6.70557082e-01
4.54931587e-01 -9.41328943e-01 -3.39628667e-01 -1.82201430e-01] | [9.429798126220703, 7.048830986022949] |
27c3c85d-e504-4a1c-a82f-be672f28bea5 | spectral-analysis-network-for-deep | 2009.05235 | null | https://arxiv.org/abs/2009.05235v1 | https://arxiv.org/pdf/2009.05235v1.pdf | Spectral Analysis Network for Deep Representation Learning and Image Clustering | Deep representation learning is a crucial procedure in multimedia analysis and attracts increasing attention. Most of the popular techniques rely on convolutional neural network and require a large amount of labeled data in the training procedure. However, it is time consuming or even impossible to obtain the label information in some tasks due to cost limitation. Thus, it is necessary to develop unsupervised deep representation learning techniques. This paper proposes a new network structure for unsupervised deep representation learning based on spectral analysis, which is a popular technique with solid theory foundations. Compared with the existing spectral analysis methods, the proposed network structure has at least three advantages. Firstly, it can identify the local similarities among images in patch level and thus more robust against occlusion. Secondly, through multiple consecutive spectral analysis procedures, the proposed network can learn more clustering-friendly representations and is capable to reveal the deep correlations among data samples. Thirdly, it can elegantly integrate different spectral analysis procedures, so that each spectral analysis procedure can have their individual strengths in dealing with different data sample distributions. Extensive experimental results show the effectiveness of the proposed methods on various image clustering tasks. | ['Jinghua Wang', 'Jianmin Jiang', 'Adrian Hilton'] | 2020-09-11 | null | null | null | null | ['image-clustering'] | ['computer-vision'] | [ 1.83801144e-01 -6.26254261e-01 -2.06736177e-02 -1.84092566e-01
-4.77305233e-01 -5.33159859e-02 2.01857284e-01 1.97629645e-01
-2.26319999e-01 1.96700349e-01 -1.05675772e-01 1.25683984e-02
-4.45011616e-01 -9.21386600e-01 -3.30643058e-01 -1.16825724e+00
8.73733908e-02 1.03805438e-01 1.00480765e-01 -2.85199601e-02
3.08372155e-02 5.54420233e-01 -1.67266762e+00 1.37960821e-01
9.39926922e-01 1.15643501e+00 6.32232130e-01 3.90866362e-02
-3.71196687e-01 9.76756215e-01 -3.32946807e-01 -8.61723348e-03
-1.23574464e-02 -3.20746601e-01 -5.93489826e-01 5.91574609e-01
9.40873623e-02 8.66999701e-02 -3.79797310e-01 1.31769943e+00
5.61313689e-01 3.80264968e-01 8.31092954e-01 -9.36382294e-01
-8.15549850e-01 4.02272046e-01 -1.00165915e+00 2.27244750e-01
-8.54784325e-02 -2.16973811e-01 8.90025258e-01 -7.40327001e-01
6.52314499e-02 1.05340242e+00 7.32102633e-01 4.89371046e-02
-1.26426649e+00 -6.65316701e-01 -5.72925024e-02 6.36784673e-01
-1.48638940e+00 -2.11262539e-01 1.32720065e+00 -5.69049299e-01
2.81596899e-01 -9.79905128e-02 6.15106046e-01 9.07618344e-01
-2.74402261e-01 9.15109575e-01 1.03960526e+00 -3.56678993e-01
1.09804370e-01 7.89606292e-03 1.16030306e-01 6.24445081e-01
1.25332937e-01 -3.20414990e-01 -1.17570117e-01 2.05686346e-01
7.45749533e-01 3.69940341e-01 -4.69427586e-01 -5.17063618e-01
-1.08620083e+00 8.35209966e-01 8.11482489e-01 5.66367984e-01
-3.88526261e-01 1.08646072e-01 6.00852251e-01 3.09878830e-02
4.04248983e-01 1.04763173e-01 -8.17234442e-02 3.69257182e-01
-9.53358531e-01 -2.88307756e-01 -2.90111210e-02 5.09076655e-01
1.03274512e+00 3.63559544e-01 4.70182002e-01 1.21448147e+00
2.74189949e-01 3.04447681e-01 5.83097458e-01 -8.15356672e-01
2.32662693e-01 6.96750045e-01 -2.39403114e-01 -1.61885750e+00
-6.71051145e-01 -7.33072758e-01 -1.44221175e+00 6.38489649e-02
1.67036101e-01 5.11308834e-02 -6.01093054e-01 1.53224862e+00
1.11286193e-01 2.34139457e-01 -8.86293128e-03 9.92911160e-01
1.08769941e+00 9.93234456e-01 1.12429142e-01 -4.81580496e-01
1.15682197e+00 -6.62079155e-01 -8.48753512e-01 -4.93185967e-02
3.00854295e-01 -8.98556292e-01 9.84359682e-01 4.85370189e-01
-6.37370884e-01 -9.51545000e-01 -1.06450880e+00 8.72499794e-02
-4.31976259e-01 5.14688909e-01 1.04498160e+00 5.01102209e-01
-9.27673757e-01 4.13559228e-01 -6.07657075e-01 -4.38277572e-01
5.53334653e-01 3.40001702e-01 -2.68575490e-01 -2.00087875e-01
-1.14222991e+00 3.46367896e-01 6.61056101e-01 4.57237720e-01
-4.71571892e-01 -3.23019713e-01 -9.38805699e-01 2.13130683e-01
3.04945439e-01 -2.03420728e-01 7.46799886e-01 -1.23043358e+00
-1.23392367e+00 6.91616237e-01 2.70853937e-02 -1.07967064e-01
-8.35589226e-03 6.79983869e-02 -6.14150226e-01 4.81592745e-01
2.06353050e-02 2.82078475e-01 9.25059021e-01 -1.20333600e+00
-3.32849652e-01 -2.97925264e-01 -1.78032443e-01 1.28529295e-01
-7.16375768e-01 -1.23348206e-01 -3.74667645e-01 -8.33563387e-01
4.73960072e-01 -5.74496388e-01 -2.17010006e-01 -5.07166013e-02
-2.03590572e-01 -2.99332708e-01 9.98309672e-01 -5.49437821e-01
9.93097723e-01 -2.42127967e+00 2.33918414e-01 3.12972873e-01
2.12231308e-01 4.18530107e-01 -1.15716577e-01 4.10375923e-01
-3.62464011e-01 -9.09763053e-02 -3.72449040e-01 1.14890113e-01
-3.45302492e-01 2.44940142e-03 -1.82917088e-01 6.75235987e-01
1.83775991e-01 6.08863115e-01 -8.06709766e-01 -5.87530911e-01
6.21411622e-01 6.13276660e-01 -2.38159806e-01 9.11424384e-02
-9.73629504e-02 5.61542869e-01 -4.20172811e-01 5.59794962e-01
9.14541006e-01 -4.63062346e-01 2.38168538e-01 -6.33438468e-01
-6.11768477e-02 -1.73857629e-01 -1.31962049e+00 1.80909657e+00
-4.57160294e-01 7.89463341e-01 5.45946397e-02 -1.83217204e+00
1.03129506e+00 2.27506801e-01 7.75764346e-01 -6.27804577e-01
3.38875771e-01 2.41040513e-01 -5.84566779e-02 -7.45231032e-01
2.05447733e-01 -1.77781209e-01 1.28464907e-01 3.62214774e-01
5.30145429e-02 2.74281621e-01 1.86971024e-01 -2.28149202e-02
4.42404211e-01 -1.39595553e-01 1.14579752e-01 -3.44599545e-01
1.02954876e+00 -3.81679833e-01 6.98073506e-01 9.37948003e-02
-7.44734630e-02 4.16637778e-01 2.33807147e-01 -6.29181981e-01
-7.99785376e-01 -8.81920874e-01 -2.45150417e-01 9.48749125e-01
4.38436478e-01 -2.40634412e-01 -5.08026540e-01 -2.68067926e-01
-3.32100034e-01 9.68024805e-02 -5.65806150e-01 -1.95656762e-01
-4.50826705e-01 -9.46052253e-01 9.37618017e-02 5.55793524e-01
1.01607442e+00 -1.10067773e+00 -2.41703749e-01 6.15355745e-02
-4.38735217e-01 -8.80757093e-01 6.59972355e-02 1.45346999e-01
-8.76535714e-01 -1.22301269e+00 -8.83583367e-01 -1.28366446e+00
6.87894106e-01 9.03768182e-01 6.67829692e-01 1.49168134e-01
-3.56602699e-01 2.33965382e-01 -5.10198414e-01 -2.09139716e-02
-1.42422132e-02 1.50057033e-01 -3.81190851e-02 4.68679905e-01
3.97643179e-01 -7.81679869e-01 -6.15248561e-01 2.13649809e-01
-9.83328640e-01 -3.24998051e-02 5.96443355e-01 8.92142236e-01
7.66602099e-01 9.28608835e-01 8.11815500e-01 -7.70437360e-01
5.18622577e-01 -4.91679996e-01 -4.56555963e-01 1.98579624e-01
-1.37611419e-01 -5.19756153e-02 9.38596189e-01 -3.50932419e-01
-1.15534544e+00 1.21079274e-01 9.99958999e-03 -4.77300704e-01
-1.84773937e-01 9.16543543e-01 -3.74057442e-01 -7.62029514e-02
5.07478416e-01 4.85039204e-01 1.06160745e-01 -5.87809741e-01
2.84188896e-01 6.09899819e-01 4.23888892e-01 -4.77151126e-01
7.81972945e-01 4.51705694e-01 4.73091565e-02 -1.16364539e+00
-6.84877276e-01 -5.76876938e-01 -7.18334973e-01 -3.76129299e-01
1.03599250e+00 -1.00172853e+00 -7.14460969e-01 6.51997209e-01
-8.56239200e-01 2.23353095e-02 1.39173597e-01 5.53990006e-01
-3.35221499e-01 8.45437288e-01 -5.31206369e-01 -8.07105362e-01
-6.50788546e-02 -1.22186494e+00 7.46554554e-01 3.46463144e-01
2.79802501e-01 -1.09555638e+00 -2.70524353e-01 3.29583436e-01
2.35707879e-01 2.28388533e-01 1.04562449e+00 -1.78718448e-01
-4.73232329e-01 -2.65239358e-01 -5.31268299e-01 4.57367152e-01
4.60949600e-01 1.04026951e-01 -1.00203228e+00 -3.45445752e-01
1.00328542e-01 -4.99195158e-01 1.00494027e+00 5.18767416e-01
1.94346762e+00 1.00053988e-01 -9.55179110e-02 7.16160238e-01
1.47317886e+00 2.03991413e-01 6.78296924e-01 3.93860012e-01
8.97832334e-01 8.08283687e-01 3.91305715e-01 2.92335123e-01
6.53311536e-02 4.70307082e-01 4.96442616e-01 -4.37311530e-01
2.30501126e-03 2.78299805e-02 1.11840032e-01 1.22983694e+00
-2.22254753e-01 -3.09001911e-03 -6.97685480e-01 4.35118347e-01
-1.98068714e+00 -1.17909622e+00 -3.55067015e-01 1.97923720e+00
4.37823087e-01 -2.96898723e-01 1.23603918e-01 6.69273734e-01
9.33969438e-01 4.13885862e-01 -4.83978897e-01 -5.62378205e-02
-2.86820561e-01 2.30317131e-01 1.90338165e-01 -8.70311707e-02
-1.36821926e+00 7.68749416e-01 5.39288616e+00 1.08151579e+00
-1.15247738e+00 -8.45829621e-02 6.08614683e-01 3.19679022e-01
-1.60641193e-01 -3.95088166e-01 -6.70596063e-02 6.28663719e-01
4.94577765e-01 4.15586159e-02 3.53626162e-01 7.28137434e-01
1.75924391e-01 -1.68390684e-02 -6.25214040e-01 1.46427202e+00
-3.44634131e-02 -1.29166496e+00 2.00145215e-01 -8.87807235e-02
5.79506934e-01 -3.70136499e-01 3.22334170e-01 1.81020871e-01
-1.39966637e-01 -1.06854856e+00 2.77113497e-01 3.80081862e-01
6.77387536e-01 -1.28476095e+00 9.83487427e-01 3.71773332e-01
-1.54835141e+00 -1.69904172e-01 -7.53821671e-01 -7.35149719e-03
-1.53159857e-01 7.45592117e-01 -1.60765246e-01 7.71427929e-01
8.02029908e-01 1.23665822e+00 -5.22471786e-01 1.03583169e+00
-1.97421923e-01 4.50732648e-01 2.55249105e-02 1.07835583e-01
2.47889563e-01 -6.58620059e-01 3.12886983e-02 1.08150208e+00
3.78513217e-01 -8.46897997e-03 3.87293488e-01 7.73661792e-01
-9.31988731e-02 4.60503608e-01 -5.83599150e-01 -1.06031708e-01
1.99906260e-01 1.46900046e+00 -1.03046620e+00 -1.22070692e-01
-5.35450578e-01 8.76415431e-01 3.02129239e-01 4.88972574e-01
-7.96356499e-01 -3.73469234e-01 3.72239769e-01 -4.25791033e-02
3.43004405e-01 -2.98112154e-01 2.85602594e-03 -1.13567007e+00
-1.37305707e-01 -8.25867951e-01 4.77934808e-01 -8.14048290e-01
-1.37875462e+00 4.13563907e-01 -3.35425258e-01 -1.58390856e+00
1.47956967e-01 -7.67861068e-01 -6.94407225e-01 7.58829236e-01
-1.60696590e+00 -1.01782835e+00 -5.64867914e-01 8.90330970e-01
6.01483762e-01 -3.76661062e-01 7.74948478e-01 6.24141693e-01
-9.02955353e-01 3.39153647e-01 3.73115271e-01 3.10411185e-01
4.80463237e-01 -9.52181995e-01 -3.59738261e-01 8.65814209e-01
2.03253463e-01 7.01342702e-01 2.55574971e-01 -2.06577629e-01
-1.19784892e+00 -1.06947815e+00 1.06106117e-01 3.95971805e-01
6.77736163e-01 -1.06900863e-01 -1.21029747e+00 2.00749651e-01
1.99071512e-01 -2.23271754e-02 1.08063233e+00 1.92223147e-01
-3.57226372e-01 -5.36384642e-01 -8.20692480e-01 2.91223764e-01
5.59125900e-01 -6.78808689e-01 -2.92911232e-01 4.63316202e-01
4.35011238e-01 1.26154229e-01 -7.53554761e-01 4.12302792e-01
1.88723981e-01 -1.08073115e+00 1.09915483e+00 -1.65005520e-01
5.38140774e-01 -5.63355863e-01 -1.89041138e-01 -1.06127906e+00
-7.17820048e-01 -9.14670452e-02 5.00856899e-02 1.41399670e+00
-7.77048944e-03 -4.94373649e-01 6.55957937e-01 -4.86443602e-02
9.11505707e-03 -6.05593979e-01 -6.11970127e-01 -6.50285900e-01
-7.79098496e-02 -4.27888066e-01 3.76920372e-01 1.28260100e+00
-2.71983474e-01 3.25318068e-01 -4.31808829e-01 4.12927508e-01
8.24896157e-01 4.22261059e-01 4.91342753e-01 -1.54198575e+00
-1.19275510e-01 -6.88427567e-01 -6.32497311e-01 -7.62801051e-01
3.15070987e-01 -1.06001759e+00 -2.30320990e-01 -1.74117911e+00
3.03038776e-01 -3.42352867e-01 -7.48285353e-01 2.20645070e-01
-4.79361899e-02 2.36640900e-01 -1.17820911e-01 5.38372815e-01
-5.80554545e-01 7.54501581e-01 1.13221729e+00 -4.39852297e-01
-3.52485180e-02 -1.91596169e-02 -7.66015053e-01 9.02444899e-01
9.29441154e-01 6.28201887e-02 -6.68034315e-01 -4.67176944e-01
2.39590004e-01 -1.54051587e-01 4.08886731e-01 -1.27710354e+00
2.02791542e-01 -2.79191453e-02 7.42983341e-01 -7.77028501e-01
1.62032127e-01 -1.05430841e+00 7.94401318e-02 3.58179867e-01
-1.80157185e-01 -4.12444383e-01 9.31068063e-02 8.17834496e-01
-5.54632306e-01 -2.52662629e-01 1.04673755e+00 -1.99531376e-01
-1.21785152e+00 3.64816338e-01 -4.97353643e-01 -3.59268874e-01
9.72098410e-01 -2.63491094e-01 3.25102992e-02 -2.11990967e-01
-5.69793463e-01 1.59734130e-01 2.42227629e-01 2.12440774e-01
7.40199625e-01 -1.52566063e+00 -3.68614674e-01 1.98187605e-01
1.26009986e-01 1.06722377e-01 7.02061534e-01 7.60875285e-01
-7.11698890e-01 -4.00337344e-03 -2.96896666e-01 -7.65083909e-01
-9.99293804e-01 9.10625041e-01 2.44118705e-01 8.52277502e-02
-7.35525608e-01 7.15077579e-01 4.28312898e-01 -2.87693828e-01
3.37435097e-01 8.34572688e-02 -6.47328794e-01 3.65718305e-01
5.85485637e-01 3.42018157e-01 -1.25130236e-01 -8.29850137e-01
-2.64546365e-01 9.43630993e-01 2.72707269e-02 4.69246060e-01
1.51485777e+00 -7.98937678e-02 -4.24057841e-01 4.36025858e-01
1.39250791e+00 -2.35358670e-01 -9.88092303e-01 -4.63500738e-01
-8.20571929e-02 -3.45327437e-01 3.37458044e-01 -2.61021644e-01
-1.59992123e+00 1.39396465e+00 7.38294899e-01 4.77352411e-01
1.53578079e+00 -1.92218587e-01 8.31163645e-01 4.18975264e-01
2.16230109e-01 -1.20123351e+00 2.40683466e-01 2.00953200e-01
6.59376442e-01 -1.37587273e+00 2.30972633e-01 -5.79252601e-01
-3.48876595e-01 1.39363885e+00 4.86851633e-01 -1.85518876e-01
5.77924192e-01 -2.59611547e-01 1.31959729e-02 -3.56222063e-01
2.85455417e-02 -4.61124718e-01 3.58785957e-01 8.04519892e-01
7.11408794e-01 -5.38439453e-02 -8.26224219e-03 4.80208993e-01
1.54906526e-01 -4.29932654e-01 2.56346345e-01 4.51947153e-01
-6.16337061e-01 -1.00712860e+00 -3.54007453e-01 3.18725854e-01
-2.52679557e-01 6.71655312e-02 -2.04192504e-01 7.16941595e-01
2.50992924e-01 1.02945137e+00 -4.16322649e-02 -5.68646789e-01
-5.95895238e-02 -1.47692785e-01 2.92127907e-01 -4.55906868e-01
-4.41695526e-02 3.25343341e-01 -4.62846160e-01 -2.91267127e-01
-1.07024610e+00 -5.31008899e-01 -1.21696317e+00 -2.11317569e-01
-2.93801606e-01 1.25655785e-01 3.55730772e-01 7.05054760e-01
1.28992647e-01 7.52221286e-01 9.20728147e-01 -9.58389759e-01
4.68425564e-02 -7.40235865e-01 -8.83389294e-01 4.95271593e-01
1.91981465e-01 -9.63606119e-01 -2.15661198e-01 6.54607713e-02] | [9.088578224182129, 3.2151010036468506] |
93870087-7d6f-4880-86d3-ca566fd329d5 | deep-lexical-segmentation-and-syntactic | null | null | https://aclanthology.org/N16-1127 | https://aclanthology.org/N16-1127.pdf | Deep Lexical Segmentation and Syntactic Parsing in the Easy-First Dependency Framework | null | ['Matthieu Constant', 'Nadi Tomeh', 'Joseph Le Roux'] | 2016-06-01 | null | null | null | naacl-2016-6 | ['lexical-analysis'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.445680141448975, 3.6778674125671387] |
95a2f13b-1a47-4e97-99df-34a2447020fc | regression-transformer-concurrent-conditional | 2202.01338 | null | https://arxiv.org/abs/2202.01338v3 | https://arxiv.org/pdf/2202.01338v3.pdf | Regression Transformer: Concurrent sequence regression and generation for molecular language modeling | Despite significant progress of generative models in the natural sciences, their controllability remains challenging. One fundamentally missing aspect of molecular or protein generative models is an inductive bias that can reflect continuous properties of interest. To that end, we propose the Regression Transformer (RT), a novel method that abstracts regression as a conditional sequence modeling problem. This introduces a new paradigm of multitask language models which seamlessly bridge sequence regression and conditional sequence generation. We thoroughly demonstrate that, despite using a nominal-scale training objective, the RT matches or surpasses the performance of conventional regression models in property prediction tasks of small molecules, proteins and chemical reactions. Critically, priming the same model with continuous properties yields a highly competitive conditional generative model that outperforms specialized approaches in a substructure-constrained, property-driven molecule generation benchmark. Our dichotomous approach is facilitated by a novel, alternating training scheme that enables the model to decorate seed sequences by desired properties, e.g., to optimize reaction yield. In sum, the RT is the first report of a multitask model that concurrently excels at predictive and generative tasks in biochemistry. This finds particular application in property-driven, local exploration of the chemical or protein space and could pave the road toward foundation models in material design. The code to reproduce all experiments of the paper is available at: https://github.com/IBM/regression-transformer | ['Matteo Manica', 'Jannis Born'] | 2022-02-01 | null | null | null | null | ['conditional-text-generation'] | ['natural-language-processing'] | [ 7.68482804e-01 5.60943149e-02 -3.26218247e-01 -1.89886808e-01
-1.00189602e+00 -8.54833126e-01 8.35189283e-01 1.04581572e-01
-1.78166866e-01 1.38383210e+00 1.33559238e-02 -7.47994006e-01
-8.72002020e-02 -6.43037140e-01 -1.21120095e+00 -1.14966035e+00
1.52740717e-01 7.20982909e-01 -6.12353198e-02 -4.09996629e-01
3.71659845e-01 5.21907210e-01 -1.21369576e+00 4.80215192e-01
9.42136168e-01 5.26722848e-01 4.25784439e-01 6.61222339e-01
5.06983735e-02 7.16593564e-01 -3.44918996e-01 -1.94485515e-01
-1.57758951e-01 -7.69858956e-01 -7.32360423e-01 -4.58588839e-01
-3.25129479e-02 4.60681349e-01 1.73036963e-01 5.02339065e-01
7.43317604e-01 4.17306274e-02 9.41577256e-01 -9.33618426e-01
-8.38784754e-01 6.48584247e-01 -1.75636128e-01 -6.82737306e-02
4.27251190e-01 3.60764265e-01 1.12604797e+00 -9.42564845e-01
8.67791116e-01 9.58169818e-01 5.15398681e-01 6.84657216e-01
-2.00820684e+00 -5.30641019e-01 1.76482961e-01 -3.35196644e-01
-1.23396385e+00 -4.50484335e-01 4.45209324e-01 -5.85922897e-01
1.34879696e+00 3.13241184e-01 4.66321349e-01 1.49312055e+00
6.74785852e-01 4.66498852e-01 1.28387105e+00 -2.61758834e-01
2.75287896e-01 6.38909563e-02 -3.30724597e-01 5.71168959e-01
6.36838078e-02 3.31301957e-01 -8.01511645e-01 -3.69402468e-01
5.83517969e-01 -1.02335643e-02 -2.02650771e-01 -4.72521514e-01
-1.30976856e+00 8.96708250e-01 1.57503933e-01 2.68458813e-01
-3.97977293e-01 2.79609829e-01 4.39583231e-03 2.18998283e-01
5.41109860e-01 9.15854394e-01 -9.07667816e-01 -6.38100132e-02
-9.37469721e-01 5.00280499e-01 8.22701752e-01 9.54789400e-01
8.76024842e-01 1.01101222e-02 -2.85586208e-01 4.03241217e-01
1.16820887e-01 2.02461615e-01 2.60201782e-01 -3.83960128e-01
-2.40908004e-02 2.74534613e-01 1.60907835e-01 -2.43303016e-01
-4.92836058e-01 -6.65878475e-01 -5.29147983e-01 4.64550257e-02
2.93825209e-01 -1.21741571e-01 -9.83416200e-01 1.94056392e+00
3.30689222e-01 -5.43508865e-02 3.05461511e-02 4.83177245e-01
6.47485077e-01 8.72719705e-01 6.11988008e-01 -4.05770808e-01
1.27765906e+00 -6.00941896e-01 -3.97484630e-01 4.36054654e-02
6.02674186e-01 -7.95026183e-01 9.13191855e-01 5.60456216e-01
-1.09929407e+00 -4.67189550e-01 -9.82381403e-01 -8.69859159e-02
-7.47097373e-01 -2.91483290e-02 1.12667072e+00 6.72680020e-01
-8.76656294e-01 8.53322566e-01 -7.75291145e-01 -9.26414654e-02
3.30876887e-01 5.46862125e-01 -2.46788487e-01 1.92724407e-01
-1.20182514e+00 9.92065191e-01 4.96458381e-01 -1.36447683e-01
-1.16096663e+00 -1.18758559e+00 -5.52896380e-01 -9.05386806e-02
3.74834418e-01 -1.09450865e+00 1.09945893e+00 -8.08033526e-01
-1.71113491e+00 7.44378388e-01 -2.61011750e-01 -3.28999817e-01
4.44521338e-01 -5.60419075e-02 -2.60271318e-02 -4.40503478e-01
1.19497003e-02 6.66431725e-01 6.46953344e-01 -1.13294208e+00
-1.30216300e-01 -1.50200948e-01 -3.43493253e-01 -1.41883254e-01
1.94884300e-01 -4.16075438e-02 1.08847015e-01 -6.78060293e-01
-3.43794942e-01 -9.10180390e-01 -6.59680545e-01 -6.66036189e-01
-6.31317914e-01 -2.26573765e-01 9.51341540e-02 -1.39568955e-01
1.10887992e+00 -1.50426424e+00 7.05528319e-01 2.39179298e-01
2.12046102e-01 6.98668957e-02 -1.57204479e-01 9.89162922e-01
-6.38822734e-01 3.87994707e-01 -3.54118228e-01 7.02940375e-02
-1.45295501e-01 -1.32812172e-01 -2.92103529e-01 3.07264447e-01
7.58423746e-01 1.38747811e+00 -8.12888622e-01 -2.08410695e-02
-9.54756737e-02 6.94755495e-01 -6.75774634e-01 1.36211097e-01
-8.60031247e-01 8.83219719e-01 -6.43018782e-01 6.45304024e-01
3.57724875e-01 -4.99975979e-01 4.83936727e-01 -1.29619747e-01
-4.38652188e-01 4.30566162e-01 -6.49724841e-01 1.80175412e+00
-2.24347562e-01 6.70792162e-02 -5.03339887e-01 -9.98150110e-01
1.06675303e+00 3.45685333e-01 7.01975763e-01 -4.29601967e-01
-7.45498091e-02 3.67054999e-01 1.19698286e-01 -1.16545141e-01
2.46109664e-01 -6.20451689e-01 2.56803865e-03 3.96949112e-01
1.73233524e-01 -3.81405830e-01 2.73178965e-01 1.20403916e-01
9.75572646e-01 1.03934062e+00 5.30102849e-01 -3.67147416e-01
4.36745703e-01 1.49545446e-01 4.68349814e-01 7.23630011e-01
3.82326424e-01 3.79147261e-01 6.37650549e-01 -3.28749001e-01
-1.36920118e+00 -9.79081392e-01 -2.84321249e-01 1.45960557e+00
-4.61801946e-01 -5.36124527e-01 -5.30437529e-01 -2.61969984e-01
6.99665770e-02 6.28016293e-01 -8.37318897e-01 -3.06641936e-01
-5.84349275e-01 -1.28594315e+00 5.05259752e-01 3.06974143e-01
-4.21072930e-01 -1.12320447e+00 -9.67102498e-02 5.36578655e-01
8.62514414e-03 -6.67829394e-01 -3.42540354e-01 9.75643277e-01
-7.25182593e-01 -8.26810122e-01 -6.69101775e-01 -6.61996782e-01
4.33338404e-01 -2.13197097e-01 1.28726912e+00 -2.32190281e-01
-5.07282615e-01 -1.13580935e-01 -1.70300037e-01 -6.00067258e-01
-7.58806109e-01 4.52395529e-01 -1.20269507e-01 -2.79350907e-01
2.88976341e-01 -9.30883765e-01 -5.53551376e-01 1.21776484e-01
-9.08869743e-01 3.11443627e-01 8.15727234e-01 1.02553880e+00
9.44317341e-01 -5.84812820e-01 9.66897309e-01 -1.13836920e+00
6.98133051e-01 -6.01461947e-01 -6.29437625e-01 3.34487110e-01
-9.34447348e-01 6.44114316e-01 6.17557526e-01 -5.54974377e-01
-8.91146898e-01 5.02825379e-01 -2.94972897e-01 1.83983847e-01
2.97499541e-02 6.65676713e-01 -2.01158449e-01 1.22191027e-01
8.15793753e-01 7.05927968e-01 6.81287721e-02 -3.38590026e-01
5.18318772e-01 -2.46276235e-04 1.68923900e-01 -9.89792168e-01
7.00045943e-01 4.93025631e-02 4.13404286e-01 -4.66043472e-01
-5.14371812e-01 -2.16560557e-01 -5.95656157e-01 1.49456218e-01
7.31006622e-01 -7.46813297e-01 -1.10714817e+00 1.15060411e-01
-9.60850477e-01 -5.72447300e-01 -1.93702966e-01 2.42864922e-01
-1.14033067e+00 7.39677623e-02 -4.58063602e-01 -7.10205495e-01
-3.94472122e-01 -1.18820143e+00 1.13883805e+00 -5.30978180e-02
-4.43609327e-01 -9.69389975e-01 3.15233171e-01 5.32048084e-02
3.63936782e-01 4.91911322e-01 1.24087584e+00 -7.88827896e-01
-8.60504150e-01 7.97032714e-02 2.19763592e-01 -9.76388007e-02
3.21391106e-01 1.72806874e-01 -7.68905938e-01 -1.44873768e-01
-4.30806190e-01 -6.03982806e-01 1.13048792e+00 5.41647613e-01
8.92645776e-01 -8.23942274e-02 -5.61951101e-01 6.99194312e-01
1.25949657e+00 3.70213419e-01 7.31061220e-01 2.79140174e-01
5.31626463e-01 4.74399805e-01 4.72286046e-01 3.76636565e-01
-9.48862955e-02 6.16562188e-01 1.03544265e-01 -3.34015101e-01
7.03885257e-02 -6.67671680e-01 3.90090704e-01 1.99580804e-01
-4.91068751e-01 -3.70084673e-01 -7.02239990e-01 -3.50976773e-02
-1.78383100e+00 -1.14016402e+00 -1.39518663e-01 2.15774465e+00
1.50207853e+00 5.39471693e-02 3.38585049e-01 -1.94661289e-01
3.22056264e-01 -1.37464046e-01 -7.77463317e-01 -4.91490155e-01
-3.38546962e-01 6.69702649e-01 5.07052779e-01 6.80705547e-01
-8.67154956e-01 1.20063770e+00 6.96147156e+00 9.44725335e-01
-1.05555928e+00 -5.44363260e-02 8.69693995e-01 -9.22165532e-03
-7.71842659e-01 1.99045599e-01 -1.18524516e+00 3.17295223e-01
1.22483528e+00 -2.65867800e-01 4.10019666e-01 6.06373429e-01
5.15870035e-01 2.01132938e-01 -1.39238620e+00 3.58026236e-01
-2.58951992e-01 -1.82386553e+00 2.49600992e-01 2.96053976e-01
6.61381304e-01 -1.09363206e-01 3.40695441e-01 1.76497251e-01
6.24915779e-01 -1.53904378e+00 5.55335045e-01 6.84283614e-01
9.37273264e-01 -6.99806213e-01 3.79222669e-02 2.99226433e-01
-7.63399899e-01 1.86525822e-01 -2.29983225e-01 2.54999459e-01
7.86181316e-02 4.71688569e-01 -1.15529549e+00 6.02466524e-01
1.01059750e-01 6.28919959e-01 -2.29157582e-01 5.50560713e-01
-7.95649961e-02 5.41691005e-01 1.27139781e-02 -3.18321437e-01
1.77237958e-01 -4.18555290e-01 3.13562483e-01 1.53682065e+00
2.28668377e-01 -6.27600923e-02 1.09618343e-01 1.26044047e+00
6.46877512e-02 2.24103123e-01 -6.47378385e-01 -4.33947057e-01
8.21621045e-02 1.06346643e+00 -5.15167654e-01 -1.32296458e-01
-1.50215933e-02 7.97688067e-01 3.96856755e-01 6.45771861e-01
-1.06828296e+00 1.58033501e-02 6.07003808e-01 9.81887951e-02
5.15351534e-01 -2.17245176e-01 -2.85514802e-01 -8.84220421e-01
-3.98940921e-01 -1.04252613e+00 3.61236185e-02 -5.54397583e-01
-1.11566985e+00 4.44915116e-01 -2.00279891e-01 -7.38629937e-01
-3.39418590e-01 -8.45883965e-01 -1.41475603e-01 1.15730453e+00
-1.47283566e+00 -1.22915840e+00 2.21923247e-01 3.47234428e-01
4.41914618e-01 -5.21338098e-02 9.92081642e-01 6.85032383e-02
-4.91408557e-01 4.16519195e-01 3.90172869e-01 -5.88301957e-01
9.34396923e-01 -1.31607461e+00 6.10616088e-01 3.29154640e-01
-1.25255464e-02 1.16136611e+00 1.09633863e+00 -8.21456969e-01
-1.53444231e+00 -1.00917554e+00 8.59583735e-01 -9.10208166e-01
6.42880619e-01 -6.77147746e-01 -7.59695470e-01 6.82725489e-01
-7.48151988e-02 -4.34221804e-01 1.13561535e+00 2.32836083e-01
-4.84084755e-01 1.71050146e-01 -6.90174639e-01 4.88832504e-01
1.13029397e+00 -2.73058474e-01 1.92778409e-02 6.96012378e-01
8.45338404e-01 -2.77939171e-01 -1.17984462e+00 3.90948445e-01
6.93694234e-01 -7.18757808e-01 1.27531362e+00 -1.32311428e+00
6.69214487e-01 -2.67609894e-01 2.08193976e-02 -1.29659629e+00
-7.32598424e-01 -1.08656549e+00 -2.72156298e-02 9.92817760e-01
1.22645259e+00 -6.73176527e-01 7.63447702e-01 4.13759381e-01
-2.99760461e-01 -9.76319551e-01 -6.70130908e-01 -7.19023705e-01
6.04556859e-01 -6.72865286e-02 5.07583618e-01 6.95355952e-01
8.03154781e-02 7.08083749e-01 -5.66144407e-01 -4.40476537e-01
2.43680209e-01 1.51684299e-01 6.98308647e-01 -1.14515448e+00
-6.12719297e-01 -6.32847250e-01 8.36540759e-02 -1.04959846e+00
3.75680029e-02 -1.06352198e+00 2.95764916e-02 -1.35285759e+00
3.65652204e-01 -4.75585729e-01 -3.61188680e-01 4.77520674e-01
-3.53907794e-01 -7.18452111e-02 -2.11664692e-01 6.85188770e-02
-2.34885573e-01 6.40857458e-01 1.35232043e+00 -6.58634007e-02
-3.82521391e-01 1.67160109e-01 -1.11094940e+00 -1.12416513e-01
8.89163554e-01 -5.33970296e-01 -4.88530040e-01 3.84447634e-01
6.16986871e-01 6.67615011e-02 1.37963578e-01 -4.02669728e-01
2.77421530e-02 -5.84949613e-01 4.86885935e-01 -2.72225678e-01
3.87668520e-01 -2.82750100e-01 6.41210914e-01 5.99478126e-01
-6.10011935e-01 1.17139630e-02 1.74280688e-01 6.83538318e-01
2.65076876e-01 1.56562045e-01 6.02421284e-01 -3.56338531e-01
-4.46925201e-02 5.22819817e-01 -6.50995970e-01 -2.34023973e-01
8.41997027e-01 -3.06513041e-01 -3.23214650e-01 -6.75060079e-02
-7.69323766e-01 -1.76605880e-01 3.54355812e-01 2.72666872e-01
4.39461559e-01 -9.41105604e-01 -9.05367374e-01 1.57911889e-02
1.73482716e-01 -2.88591743e-01 -1.35512188e-01 9.26449060e-01
-1.81507409e-01 9.36553061e-01 -4.62753214e-02 -4.09813732e-01
-1.07276917e+00 9.48678851e-01 4.08948272e-01 -3.87301713e-01
-2.93536652e-02 8.26922774e-01 3.61827105e-01 -3.74665976e-01
-2.60418952e-01 -2.13793784e-01 1.05855919e-01 -5.65465353e-02
3.24031323e-01 -5.33952527e-02 1.53195113e-01 -1.79833978e-01
-2.49045834e-01 2.75842220e-01 -2.72835135e-01 5.64541407e-02
1.66457236e+00 3.51675242e-01 -1.04166128e-01 4.42007780e-01
9.35999036e-01 -4.97780107e-02 -1.25569272e+00 -6.04582578e-02
1.27256326e-02 9.14719626e-02 -4.06057864e-01 -1.15021491e+00
-2.99477011e-01 5.38136721e-01 1.23873122e-01 -1.09309092e-01
7.80860186e-01 1.33502752e-01 2.66765833e-01 4.31800157e-01
2.18217328e-01 -7.47652769e-01 1.63385943e-01 3.70057285e-01
8.05534244e-01 -1.07747567e+00 1.52262345e-01 -3.39035094e-01
-5.68429410e-01 1.07582474e+00 3.18546802e-01 2.08449721e-01
2.60284752e-01 5.91747880e-01 -2.99644619e-01 -2.43577510e-01
-1.14953506e+00 -2.03313291e-01 1.39520094e-01 8.20812225e-01
1.25022137e+00 7.62486830e-02 -5.14452100e-01 3.11342150e-01
-1.18069477e-01 4.17332090e-02 1.03584766e-01 9.23374355e-01
-4.88028556e-01 -1.76600897e+00 -1.26170442e-01 1.24187522e-01
-4.79341120e-01 -5.96225023e-01 -7.20988572e-01 5.06115675e-01
-4.86309193e-02 7.57219791e-01 -5.29890954e-01 -1.84341550e-01
7.45046064e-02 4.39466417e-01 7.56424189e-01 -6.32556677e-01
-8.23110044e-01 2.61808187e-01 1.86286390e-01 -3.05023551e-01
-4.33621764e-01 -7.80300498e-01 -1.11227298e+00 -2.82700509e-01
-2.99072742e-01 4.01829779e-01 6.36922956e-01 6.51461959e-01
5.27939498e-01 6.49794340e-01 4.98763829e-01 -7.51040459e-01
-4.64027256e-01 -8.53192508e-01 -1.46308064e-01 -2.51308698e-02
2.88279891e-01 -6.22819960e-01 9.30224136e-02 3.08393121e-01] | [4.710371017456055, 5.797785758972168] |
a2cf4a70-9163-4ba7-a52c-3effe3b5f5f9 | sparsealign-a-super-resolution-algorithm-for | 2201.08706 | null | https://arxiv.org/abs/2201.08706v1 | https://arxiv.org/pdf/2201.08706v1.pdf | SparseAlign: A Super-Resolution Algorithm for Automatic Marker Localization and Deformation Estimation in Cryo-Electron Tomography | Tilt-series alignment is crucial to obtaining high-resolution reconstructions in cryo-electron tomography. Beam-induced local deformation of the sample is hard to estimate from the low-contrast sample alone, and often requires fiducial gold bead markers. The state-of-the-art approach for deformation estimation uses (semi-)manually labelled marker locations in projection data to fit the parameters of a polynomial deformation model. Manually-labelled marker locations are difficult to obtain when data are noisy or markers overlap in projection data. We propose an alternative mathematical approach for simultaneous marker localization and deformation estimation by extending a grid-free super-resolution algorithm first proposed in the context of single-molecule localization microscopy. Our approach does not require labelled marker locations; instead, we use an image-based loss where we compare the forward projection of markers with the observed data. We equip this marker localization scheme with an additional deformation estimation component and solve for a reduced number of deformation parameters. Using extensive numerical studies on marker-only samples, we show that our approach automatically finds markers and reliably estimates sample deformation without labelled marker data. We further demonstrate the applicability of our approach for a broad range of model mismatch scenarios, including experimental electron tomography data of gold markers on ice. | ['K Joost Batenburg', 'Hermen Jan Hupkes', 'Erik Franken', 'Holger Kohr', 'Felix Lucka', 'Poulami Somanya Ganguly'] | 2022-01-21 | null | null | null | null | ['electron-tomography'] | ['medical'] | [ 6.42188668e-01 -1.83764145e-01 4.49019045e-01 -2.21016496e-01
-1.21925914e+00 -4.67806786e-01 4.12041336e-01 1.57787591e-01
-8.13673139e-01 8.83445740e-01 -4.40212250e-01 1.41339362e-01
-6.69662952e-02 -4.26038325e-01 -8.15653086e-01 -9.78774786e-01
2.05043674e-01 1.24442911e+00 4.56067085e-01 2.23566994e-01
5.05095720e-01 8.75997007e-01 -8.91046703e-01 -2.50210255e-01
5.99969149e-01 5.57510495e-01 5.47286868e-01 7.53653467e-01
3.33984405e-01 1.74041942e-01 -1.90070510e-01 -2.30966713e-02
1.38839468e-01 -4.42334831e-01 -9.24493611e-01 1.92371473e-01
4.00813699e-01 -2.29936585e-01 3.48955721e-01 8.71577322e-01
5.29558957e-01 -9.40194875e-02 6.79429948e-01 -5.22563815e-01
1.51219547e-01 -1.61444262e-01 -6.03230774e-01 8.55683461e-02
2.70150453e-01 -1.86835244e-01 4.43736345e-01 -1.08619738e+00
1.24798131e+00 6.26663864e-01 9.14613903e-01 6.03562951e-01
-1.88339078e+00 -1.19013198e-01 -5.23139119e-01 -2.09263667e-01
-1.24640262e+00 -4.87647444e-01 7.01164246e-01 -7.72308409e-01
5.60816288e-01 2.57736981e-01 4.56920624e-01 4.83048528e-01
3.60145152e-01 -1.94171160e-01 1.54916847e+00 -6.98405385e-01
3.49044740e-01 -7.99604505e-02 4.90077883e-02 6.19786382e-01
1.87734038e-01 -2.77345665e-02 -2.17693478e-01 -5.02743423e-01
1.27509415e+00 6.24122247e-02 -5.23042679e-01 -7.65648484e-01
-1.48214781e+00 4.91744399e-01 1.02239206e-01 3.29731047e-01
-3.05616319e-01 4.58830874e-03 8.22497308e-02 -2.09977344e-01
6.67190254e-01 5.48514962e-01 -2.71916687e-01 -1.57338247e-01
-1.29849064e+00 4.08501893e-01 4.07314003e-01 4.64030594e-01
1.00831664e+00 -4.27945286e-01 4.86387104e-01 5.91460168e-01
1.46773189e-01 3.19393963e-01 2.40318701e-01 -1.15841186e+00
2.35820249e-01 2.91379482e-01 6.10187113e-01 -6.03066623e-01
-5.59977114e-01 9.71304849e-02 -7.07386911e-01 3.75094384e-01
9.42696631e-01 8.34128335e-02 -7.58289933e-01 1.32917392e+00
6.77535951e-01 2.06992254e-01 -2.62588829e-01 9.80760634e-01
8.12147707e-02 2.07298890e-01 -4.95747834e-01 -7.31545627e-01
1.13330114e+00 -3.01108271e-01 -4.23878372e-01 5.78586431e-03
7.25053132e-01 -8.28119397e-01 6.81677639e-01 2.02513054e-01
-1.22035837e+00 -6.31269515e-02 -8.55473220e-01 2.93965824e-02
1.66899070e-01 2.13647261e-01 6.31627962e-02 1.79553092e-01
-7.89041579e-01 1.24359643e+00 -1.22976434e+00 -2.71864504e-01
2.47590423e-01 5.59403181e-01 -5.95114946e-01 2.41172701e-01
-3.37076992e-01 1.14907908e+00 -7.34305680e-02 1.77644461e-01
-2.67029613e-01 -7.19745100e-01 -5.22633612e-01 -4.55442011e-01
1.19021922e-01 -4.88348961e-01 9.76690114e-01 -3.49807680e-01
-1.61895037e+00 1.20326257e+00 -5.29016733e-01 -2.51208276e-01
7.36237168e-01 -3.11087002e-03 2.05233321e-01 4.97246295e-01
-8.68684985e-03 1.48613408e-01 6.05010748e-01 -1.47869098e+00
1.90100715e-01 -5.39359510e-01 -5.30938804e-01 1.51838120e-02
4.55188036e-01 3.12142074e-01 -5.86423650e-02 -1.06712863e-01
6.58210635e-01 -1.06664479e+00 -5.12694478e-01 9.15720090e-02
-3.01125705e-01 5.49018741e-01 8.98162127e-01 -6.39852762e-01
6.25917137e-01 -1.68669987e+00 3.75939429e-01 3.06373775e-01
4.54250395e-01 3.83467339e-02 4.48882699e-01 4.42909300e-01
6.89199148e-03 -1.36712387e-01 -6.44161165e-01 -4.02476400e-01
-3.28603804e-01 7.97190070e-02 -3.05691082e-02 1.02912092e+00
1.15420455e-02 7.45740831e-01 -7.41690576e-01 -7.06564844e-01
3.19312304e-01 7.05052078e-01 -3.54517072e-01 4.94927634e-04
-6.80226535e-02 9.49536443e-01 -2.41741553e-01 4.45342690e-01
8.87132406e-01 -3.80510688e-01 5.41071355e-01 -3.30663443e-01
-4.12090212e-01 1.62315711e-01 -1.11644745e+00 1.53585887e+00
-1.32974938e-01 3.20657790e-01 4.84440714e-01 -8.34366560e-01
8.09867024e-01 2.66252458e-01 5.86558700e-01 -2.13008627e-01
-3.99468876e-02 5.77504456e-01 3.60084474e-02 -1.53930485e-01
4.21765715e-01 -7.56166935e-01 1.14520997e-01 6.62974715e-01
-1.29155472e-01 -5.42380452e-01 -2.42545307e-01 -3.84993777e-02
9.84678030e-01 5.34546256e-01 2.00865015e-01 -5.34752607e-01
3.19792777e-01 -6.86533079e-02 4.05854434e-01 2.49369010e-01
4.15812694e-02 1.06391835e+00 3.07924211e-01 -5.51473200e-01
-1.73267663e+00 -6.40536964e-01 -7.75757313e-01 2.81795681e-01
8.98434296e-02 -2.52551049e-01 -1.15062129e+00 -1.99586928e-01
-1.02080256e-01 -1.55816227e-01 -4.87327397e-01 4.65883553e-01
-9.97274578e-01 -1.27920902e+00 1.12937495e-01 2.76804060e-01
-7.09590688e-02 -7.50948012e-01 -1.00562978e+00 3.36019516e-01
3.01528946e-02 -1.08904290e+00 -1.18854336e-01 4.10338312e-01
-1.12009394e+00 -1.12163687e+00 -8.16663206e-01 -4.05847460e-01
1.28422332e+00 -1.25828743e-01 8.25679183e-01 2.29980469e-01
-2.46426493e-01 4.98533882e-02 7.20774904e-02 2.40027636e-01
-3.92688185e-01 -2.75932878e-01 2.66761780e-01 -2.10903659e-01
-1.46271527e-01 -7.10503340e-01 -5.78408003e-01 6.93617284e-01
-6.02336407e-01 1.08023793e-01 1.44642487e-01 1.06439185e+00
1.13961422e+00 -5.96765876e-01 1.54434323e-01 -9.84137237e-01
2.20153689e-01 1.04455143e-01 -9.15088773e-01 -1.06709965e-01
-3.70467037e-01 1.46884974e-02 5.42479873e-01 -4.98371869e-01
-6.53794408e-01 3.95300925e-01 -1.79925635e-01 -4.35142130e-01
-1.44074664e-01 3.91902328e-01 -2.04283223e-02 -7.58656204e-01
5.32260180e-01 1.40172750e-01 2.99692839e-01 -6.17150247e-01
-1.92851439e-01 2.75147796e-01 6.76970959e-01 -8.78652155e-01
5.09058535e-01 8.73485625e-01 6.14938557e-01 -8.09345126e-01
-2.03113437e-01 -4.15604800e-01 -1.18731761e+00 -1.10209830e-01
5.87987185e-01 -2.91761577e-01 -9.27827895e-01 5.20622849e-01
-1.16226792e+00 -5.88651717e-01 -2.62589723e-01 6.36285365e-01
-8.94324005e-01 8.13563049e-01 -7.08647251e-01 -8.12180221e-01
-1.05607741e-01 -1.40472579e+00 1.45099175e+00 -1.69413447e-01
-2.39274964e-01 -1.04511344e+00 3.99745047e-01 1.87061280e-01
3.02659243e-01 5.65865695e-01 5.94610870e-01 -2.10581601e-01
-6.99075162e-01 -3.55807692e-01 5.66931106e-02 -1.59586102e-01
-9.28126946e-02 7.35950544e-02 -8.62164676e-01 -3.49577338e-01
4.61347401e-01 -1.88295543e-01 6.19737744e-01 4.78372157e-01
9.23065066e-01 2.98213139e-02 -6.12085760e-01 7.70224154e-01
1.67039883e+00 -8.16656202e-02 8.11634541e-01 4.66607839e-01
7.98175216e-01 5.13700128e-01 6.37152612e-01 3.18308741e-01
-9.65989384e-05 1.24515510e+00 1.88286871e-01 8.18005472e-04
2.48768002e-01 1.72860265e-01 4.01480980e-02 7.90962100e-01
-7.04051733e-01 1.71377361e-01 -9.72510695e-01 3.79394978e-01
-1.67141628e+00 -9.14571822e-01 -4.29470420e-01 2.49692607e+00
1.02940583e+00 -9.57120508e-02 1.57925799e-01 1.10410720e-01
7.47993231e-01 -3.82749766e-01 -5.05767465e-01 -2.68568657e-02
1.45981967e-01 4.97936994e-01 6.20053649e-01 9.49725211e-01
-8.58366370e-01 6.34134531e-01 6.51702976e+00 5.09724200e-01
-1.19706678e+00 3.23749036e-01 3.48183662e-01 1.29202873e-01
-1.67672053e-01 3.94628167e-01 -8.45385671e-01 6.97800457e-01
9.57793474e-01 1.34894148e-01 1.61416352e-01 4.61851001e-01
2.90558487e-01 -6.62979305e-01 -1.10882354e+00 6.96333110e-01
-2.66095430e-01 -1.57633388e+00 -4.17490184e-01 4.93241578e-01
4.10772741e-01 4.96609975e-03 -3.88670951e-01 -7.74491191e-01
-3.88879701e-02 -7.78537035e-01 3.84138465e-01 8.59833598e-01
1.03892732e+00 -5.10764360e-01 6.32691562e-01 5.33361316e-01
-9.88226116e-01 6.99242413e-01 -4.79151934e-01 -1.14912307e-02
6.62085354e-01 7.39683747e-01 -9.95427549e-01 4.97254312e-01
2.71789014e-01 2.65984207e-01 -1.55218735e-01 8.34277272e-01
3.20255399e-01 3.36650461e-01 -6.84686482e-01 2.91415542e-01
-2.50276536e-01 -8.05394351e-01 4.42089260e-01 1.04406285e+00
3.71883392e-01 2.22011693e-02 -8.70069042e-02 9.51366425e-01
1.09973475e-01 -1.34781659e-01 -3.03577930e-01 3.37715894e-01
3.80086660e-01 1.39634144e+00 -1.13917124e+00 -1.76534444e-01
-1.53075941e-02 8.37428570e-01 4.29168642e-01 2.34712571e-01
-3.83973002e-01 -3.50094140e-02 2.16174901e-01 8.16600859e-01
2.23377988e-01 -5.11120021e-01 -1.94835544e-01 -1.22491181e+00
1.64257228e-01 -3.59898537e-01 -3.08062762e-01 -8.69531512e-01
-9.50432420e-01 2.94409752e-01 4.17916551e-02 -9.98084188e-01
-4.55468148e-01 -6.90153599e-01 -6.21395648e-01 1.09993005e+00
-1.16940224e+00 -1.00572157e+00 -5.65856285e-02 4.49476615e-02
-2.18309015e-01 5.07701933e-01 1.05835867e+00 5.78113906e-02
-2.83305824e-01 -5.73454832e-04 4.83910024e-01 -2.10034847e-02
7.37861753e-01 -1.32148385e+00 3.47573310e-01 7.64474571e-01
-3.13729286e-01 9.26334500e-01 8.22201014e-01 -8.78481805e-01
-1.29128957e+00 -6.41262054e-01 8.66988957e-01 -7.01072216e-01
6.46036983e-01 -4.83530343e-01 -1.28517616e+00 8.51595640e-01
-3.01521331e-01 6.20094538e-01 4.61452603e-01 -3.95581663e-01
3.85132402e-01 3.61339450e-01 -1.54469526e+00 -1.55820549e-02
5.85547388e-01 -6.52823150e-01 -2.57498771e-01 3.60738248e-01
-4.41361638e-03 -7.78914928e-01 -1.13036501e+00 3.63478929e-01
6.91001296e-01 -8.40437353e-01 9.75746572e-01 -1.30517066e-01
8.54389146e-02 -6.07940316e-01 4.35566269e-02 -9.98962998e-01
-9.83707607e-02 -7.51783788e-01 2.18116879e-01 9.80029047e-01
4.92688678e-02 -6.67196810e-01 1.10254312e+00 7.93100476e-01
-1.47276089e-01 -9.17437792e-01 -1.25283527e+00 -5.55310547e-01
1.09081730e-01 8.76325518e-02 1.26333877e-01 8.42021585e-01
1.09909989e-01 -4.37641293e-02 -1.83038980e-01 5.64086698e-02
9.98106658e-01 2.75444329e-01 6.55939698e-01 -1.40003586e+00
-2.76087016e-01 1.89111501e-01 -5.49432576e-01 -9.29198742e-01
-3.36447991e-02 -5.27653039e-01 3.10041815e-01 -1.08581853e+00
4.70657855e-01 -7.42601335e-01 4.04687524e-01 5.35970815e-02
6.88580871e-02 5.63173234e-01 -1.37748167e-01 6.83882117e-01
-3.52433562e-01 1.67844564e-01 1.32043624e+00 6.26202881e-01
6.79000393e-02 -2.07136422e-01 7.44789317e-02 9.91770089e-01
4.69316870e-01 -6.48849964e-01 2.35653758e-01 -4.62697409e-02
2.53659368e-01 2.81915098e-01 7.68338144e-01 -9.25710261e-01
2.94192970e-01 -7.22689927e-02 4.08172727e-01 -5.88193357e-01
6.39240503e-01 -7.32660711e-01 5.83753169e-01 2.15455636e-01
9.69660729e-02 -2.40390655e-02 -1.55458003e-01 2.74280876e-01
1.81698799e-01 -6.46812260e-01 1.15351546e+00 -4.16622847e-01
1.88466623e-01 5.64026274e-02 -2.63239890e-01 -1.27240196e-01
6.64161801e-01 -2.87779897e-01 -3.31469893e-01 7.89682716e-02
-9.60308611e-01 -3.77335936e-01 1.53318942e+00 -9.04292285e-01
7.55070329e-01 -1.10268307e+00 -3.79786819e-01 1.63265526e-01
-3.22517335e-01 4.93909538e-01 2.37020478e-02 1.40318668e+00
-9.58861053e-01 7.42237121e-02 -9.17700231e-02 -8.98896992e-01
-1.25786698e+00 3.06669116e-01 6.13214433e-01 -4.52875733e-01
-6.57364607e-01 5.59744537e-01 -1.27200097e-01 -5.19421756e-01
-6.33677840e-01 -4.38184232e-01 2.30166391e-01 -2.99120754e-01
4.53005672e-01 4.01125371e-01 3.13185483e-01 -9.60147262e-01
-3.70325834e-01 1.22391212e+00 -8.03337246e-02 -3.43013406e-01
1.62794793e+00 -1.75946549e-01 -3.29888523e-01 4.95873421e-01
1.04819536e+00 1.56193852e-01 -1.45894372e+00 -5.18809445e-02
-7.90390372e-02 -4.56257910e-01 -8.79755318e-02 -3.15745294e-01
-5.45681775e-01 9.12632823e-01 3.19673538e-01 -1.05443090e-01
5.88131309e-01 1.63185760e-01 6.39706790e-01 1.95391342e-01
6.79245114e-01 -8.63313735e-01 -2.35314831e-01 1.17184162e-01
6.39482915e-01 -9.83147442e-01 5.50854862e-01 -5.52427590e-01
5.76791838e-02 1.13656199e+00 2.03050300e-01 -3.89990896e-01
2.57684976e-01 7.12541223e-01 -1.07817836e-01 -4.18075502e-01
-4.12757486e-01 1.82280630e-01 -9.64732915e-02 4.74353373e-01
3.68385881e-01 -2.03018442e-01 -3.77834439e-01 9.58461016e-02
7.06518516e-02 1.52201965e-01 7.74740219e-01 1.27300358e+00
-4.70072120e-01 -1.45412111e+00 -6.91036105e-01 3.56391221e-01
-6.27751708e-01 -5.29003441e-02 -3.45179230e-01 6.05693221e-01
-9.10494253e-02 2.05826327e-01 2.40500480e-01 -4.75559011e-02
2.40361150e-02 2.06175715e-01 1.08823085e+00 -4.72106516e-01
-1.19063362e-01 5.39489388e-01 -1.78592995e-01 -3.58091801e-01
-9.81893361e-01 -1.06726813e+00 -1.48440480e+00 -2.42731467e-01
-7.38968074e-01 2.98703462e-01 8.25385749e-01 1.04352617e+00
2.86331683e-01 1.81871129e-03 3.44359010e-01 -1.70130908e+00
-2.31340066e-01 -9.69401300e-01 -7.17730761e-01 3.24084431e-01
3.42473000e-01 -7.74361491e-01 -5.40429711e-01 3.63291204e-01] | [13.188070297241211, -2.9690043926239014] |
d853d8c6-54b1-4d66-9523-92efe2fc845e | metasci-scalable-and-adaptive-reconstruction | 2103.01786 | null | https://arxiv.org/abs/2103.01786v1 | https://arxiv.org/pdf/2103.01786v1.pdf | MetaSCI: Scalable and Adaptive Reconstruction for Video Compressive Sensing | To capture high-speed videos using a two-dimensional detector, video snapshot compressive imaging (SCI) is a promising system, where the video frames are coded by different masks and then compressed to a snapshot measurement. Following this, efficient algorithms are desired to reconstruct the high-speed frames, where the state-of-the-art results are achieved by deep learning networks. However, these networks are usually trained for specific small-scale masks and often have high demands of training time and GPU memory, which are hence {\bf \em not flexible} to $i$) a new mask with the same size and $ii$) a larger-scale mask. We address these challenges by developing a Meta Modulated Convolutional Network for SCI reconstruction, dubbed MetaSCI. MetaSCI is composed of a shared backbone for different masks, and light-weight meta-modulation parameters to evolve to different modulation parameters for each mask, thus having the properties of {\bf \em fast adaptation} to new masks (or systems) and ready to {\bf \em scale to large data}. Extensive simulation and real data results demonstrate the superior performance of our proposed approach. Our code is available at {\small\url{https://github.com/xyvirtualgroup/MetaSCI-CVPR2021}}. | ['Xin Yuan', 'Bo Chen', 'Ziheng Cheng', 'Hao Zhang', 'Zhengjue Wang'] | 2021-03-02 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Wang_MetaSCI_Scalable_and_Adaptive_Reconstruction_for_Video_Compressive_Sensing_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Wang_MetaSCI_Scalable_and_Adaptive_Reconstruction_for_Video_Compressive_Sensing_CVPR_2021_paper.pdf | cvpr-2021-1 | ['video-compressive-sensing'] | ['computer-vision'] | [ 4.85890716e-01 -2.69287616e-01 -4.88160811e-02 1.23933963e-02
-7.47879803e-01 -2.03250006e-01 3.44620496e-01 -8.01700592e-01
-3.18530053e-01 4.93487537e-01 -1.32687807e-01 -3.04793686e-01
4.39658016e-02 -5.63667238e-01 -9.62104857e-01 -8.54310513e-01
-4.04342175e-01 -3.58881243e-02 2.63366580e-01 1.44817248e-01
4.52534147e-02 2.19706416e-01 -1.45489550e+00 3.30227762e-01
5.92476368e-01 1.29039037e+00 7.12049186e-01 8.68727505e-01
3.23951274e-01 8.61657262e-01 -3.54242444e-01 -2.18385443e-01
5.27568817e-01 -5.83485901e-01 -3.46434414e-01 1.85745567e-01
3.05056989e-01 -7.72100568e-01 -9.94813800e-01 1.12557662e+00
5.66893876e-01 1.22047905e-02 2.44089425e-01 -9.71024334e-01
-6.12515986e-01 4.67452317e-01 -6.61902368e-01 4.57930505e-01
2.59189010e-01 4.66461778e-01 3.96688461e-01 -1.02682173e+00
7.37161100e-01 8.69074404e-01 4.02482569e-01 5.10469019e-01
-7.60077655e-01 -9.94165361e-01 4.50434722e-02 5.92471719e-01
-1.55102158e+00 -7.79262245e-01 7.75467515e-01 -3.57106805e-01
6.20047987e-01 3.26359063e-01 6.72278166e-01 1.05804813e+00
9.92929637e-02 6.90673232e-01 1.05173731e+00 -1.83967903e-01
1.55228257e-01 -2.24115252e-01 -3.50078940e-01 8.88674378e-01
2.10054070e-01 2.62575030e-01 -5.65852582e-01 6.15450405e-02
1.08483636e+00 2.40245908e-01 -9.08105552e-01 -2.42914520e-02
-1.33906722e+00 5.71711957e-01 2.39870623e-01 1.71362460e-01
-4.11710203e-01 4.17916238e-01 1.74370334e-01 4.11606103e-01
9.14930850e-02 6.53950125e-02 -1.45734236e-01 -1.25122815e-01
-9.60185766e-01 -1.75183713e-02 4.74275589e-01 1.17373717e+00
6.92920327e-01 4.47139591e-01 -2.27468088e-01 6.89723253e-01
7.12011829e-02 6.61925793e-01 3.12568724e-01 -1.42151201e+00
5.40676177e-01 2.49284431e-02 1.92513645e-01 -8.96062315e-01
-2.66017675e-01 -7.07983673e-01 -1.35243750e+00 1.25266314e-01
-1.02514505e-01 -3.38974684e-01 -8.24774384e-01 1.74725199e+00
1.52626351e-01 9.36977446e-01 -1.68945998e-01 1.07416332e+00
7.91714549e-01 9.95522082e-01 -4.11768526e-01 -7.36631513e-01
1.14699113e+00 -1.03100407e+00 -5.24024427e-01 -2.09473506e-01
2.82177657e-01 -7.19541609e-01 8.31907034e-01 5.56560516e-01
-1.40810835e+00 -7.36312211e-01 -1.39434409e+00 1.26276866e-01
3.16246867e-01 8.10941085e-02 3.19593817e-01 3.99376065e-01
-1.26340270e+00 4.98470783e-01 -7.35415399e-01 1.58936679e-01
6.45640373e-01 3.74192864e-01 -1.24150096e-02 -5.02161860e-01
-1.11525488e+00 3.48393917e-01 2.38456696e-01 1.96176842e-01
-1.12326908e+00 -5.59494674e-01 -6.65429592e-01 8.42960477e-02
5.87232113e-01 -7.57778227e-01 8.73678565e-01 -7.25464404e-01
-1.51289237e+00 5.77022195e-01 1.92401186e-02 -3.31487983e-01
3.95777106e-01 -4.65836525e-02 -5.85785389e-01 5.74462354e-01
-2.54561100e-02 4.70047385e-01 1.39448094e+00 -1.22423506e+00
-5.04972577e-01 -1.09894291e-01 -2.60782056e-03 7.12780580e-02
-4.92211759e-01 8.34780261e-02 -1.04205847e+00 -6.77804112e-01
4.43637483e-02 -1.04997873e+00 -5.07379808e-02 3.07317264e-03
-3.00436318e-01 3.44904631e-01 9.06111598e-01 -6.88877225e-01
1.43928325e+00 -2.18794799e+00 2.98261523e-01 -1.30395442e-01
5.53691268e-01 6.72545016e-01 -1.62909478e-01 1.66023850e-01
5.21335304e-02 -2.36949414e-01 -3.50514114e-01 -3.35027069e-01
-2.55784959e-01 -4.75578196e-02 -5.93795627e-02 6.22364044e-01
-1.92385122e-01 7.25933254e-01 -7.35342324e-01 -3.71068984e-01
4.14512515e-01 5.54216564e-01 -6.67653978e-01 3.89169037e-01
-9.08533484e-02 6.54937387e-01 -5.13348520e-01 6.38524950e-01
1.05671418e+00 -7.62972474e-01 3.19418192e-01 -5.25539100e-01
-2.34808102e-01 -2.44635060e-01 -1.28506267e+00 1.79083860e+00
-1.68831259e-01 4.88410324e-01 5.23348451e-01 -1.21478200e+00
5.53029239e-01 2.85508364e-01 7.46912062e-01 -8.14804614e-01
4.73739713e-01 2.97543943e-01 -4.00073491e-02 -7.59636641e-01
3.71220320e-01 1.09159060e-01 1.89265147e-01 4.39991683e-01
-1.92285571e-02 -4.85151337e-04 1.80449978e-01 1.22916751e-01
1.21700203e+00 -3.35350842e-03 -1.29196510e-01 -2.23660674e-02
5.97921968e-01 -4.71468180e-01 6.18937492e-01 6.97618723e-01
-1.06254242e-01 9.35329080e-01 1.01818964e-01 -3.02406520e-01
-1.17091560e+00 -8.40137482e-01 -9.21266750e-02 6.98849022e-01
4.05840307e-01 -2.61089951e-01 -6.84292734e-01 -9.35615972e-02
-3.40468943e-01 3.54015082e-01 -2.31662184e-01 -1.80718794e-01
-8.72398138e-01 -6.67749107e-01 3.65449399e-01 1.47112653e-01
8.39487374e-01 -1.06640828e+00 -8.90243232e-01 3.44336092e-01
-2.90256232e-01 -1.50147641e+00 -6.76959991e-01 -5.89684732e-02
-5.70818305e-01 -1.03264618e+00 -1.01789665e+00 -9.01112497e-01
5.21954596e-01 7.32850730e-01 8.81592333e-01 4.12067592e-01
-1.90612972e-01 2.57791042e-01 -4.06831652e-01 1.43153230e-02
-1.19096123e-01 -2.39907786e-01 1.53070554e-01 2.55697101e-01
-2.07043678e-01 -9.40470159e-01 -1.17795372e+00 2.63712168e-01
-1.38338494e+00 3.85432661e-01 7.17101514e-01 8.80531847e-01
6.94095552e-01 -1.78151533e-01 2.88322121e-01 -7.12752521e-01
2.23261654e-01 -5.78326702e-01 -5.89754939e-01 4.90287468e-02
-3.98772150e-01 -4.06556010e-01 9.81828392e-01 -5.30578375e-01
-7.07099378e-01 -1.53323278e-01 -1.11232266e-01 -1.16989768e+00
2.39234135e-01 3.98275018e-01 -1.91399548e-02 -2.45316342e-01
3.55149835e-01 7.06345677e-01 1.03595644e-01 -2.26515070e-01
1.10662907e-01 6.09794021e-01 6.80851340e-01 -4.28950995e-01
7.95386791e-01 6.74828649e-01 1.12741664e-02 -7.78312027e-01
-6.11224234e-01 -1.44635007e-01 -2.71254033e-01 -4.47023958e-01
7.64811397e-01 -1.22388196e+00 -6.78727865e-01 7.65669703e-01
-9.48453665e-01 -4.42680091e-01 -5.81304878e-02 6.41349435e-01
-7.14914918e-01 6.44347668e-01 -7.90471911e-01 -4.23633277e-01
-4.31963712e-01 -1.37672544e+00 9.62109149e-01 2.38264978e-01
4.26568836e-01 -5.16372323e-01 -2.82024890e-01 4.42930549e-01
6.12993300e-01 1.01803042e-01 7.90422678e-01 9.64250639e-02
-1.15844023e+00 -7.53234550e-02 -3.79967839e-01 4.83582228e-01
-7.45366663e-02 -3.79536957e-01 -7.32693315e-01 -9.12596762e-01
2.80516684e-01 -2.79712379e-01 7.94209599e-01 5.27607322e-01
1.70999515e+00 -3.11510116e-01 -3.70634794e-01 1.28680480e+00
1.55532289e+00 5.34320951e-01 7.96858490e-01 2.05828264e-01
6.96024239e-01 -2.98691094e-01 3.02932501e-01 8.06405842e-01
3.91689897e-01 7.60548770e-01 5.87150753e-01 6.34391308e-02
-3.26144129e-01 1.09888867e-01 4.10718471e-01 1.33482027e+00
-1.46395892e-01 -5.01557827e-01 -4.55600888e-01 3.18945199e-01
-1.58714736e+00 -1.05190086e+00 -1.55378096e-02 2.21144557e+00
5.59792280e-01 -1.15099307e-02 -2.44770750e-01 -5.39392456e-02
7.25044966e-01 6.15397573e-01 -7.59322882e-01 1.73902646e-01
-2.63664663e-01 3.69407982e-01 3.76029074e-01 2.71024972e-01
-9.66459274e-01 6.01884425e-01 5.17783117e+00 9.51965570e-01
-1.47087729e+00 4.02861357e-01 6.17236912e-01 -4.73957330e-01
-1.32910609e-01 -3.98071855e-03 -5.02200246e-01 9.54133749e-01
7.32567370e-01 1.89790223e-02 7.83436894e-01 3.27992439e-01
1.33716583e-01 -1.90929510e-02 -8.39022696e-01 1.40030789e+00
2.30166346e-01 -1.81248927e+00 1.34251481e-02 5.89100942e-02
8.65581572e-01 2.25327387e-01 3.01249027e-01 2.33651981e-01
-1.09400235e-01 -6.95222020e-01 7.33263195e-01 5.98036349e-01
1.45317900e+00 -3.97207409e-01 3.68178219e-01 3.30988646e-01
-1.31962395e+00 -2.62290418e-01 -4.27876920e-01 2.34438151e-01
4.41088498e-01 4.76724803e-01 -7.08894804e-02 6.15929961e-01
8.58617961e-01 6.63577139e-01 -1.90688342e-01 1.02489674e+00
2.91815341e-01 5.94681323e-01 -1.88617766e-01 1.92239940e-01
5.80717921e-02 -3.87634903e-01 7.28335381e-01 9.33895230e-01
9.13100421e-01 5.53689778e-01 3.35849047e-01 5.56773067e-01
-2.49368459e-01 -3.52281809e-01 -4.96855617e-01 2.03789875e-01
5.23364186e-01 1.07020223e+00 -3.53681654e-01 -5.20963967e-01
-6.90402389e-01 1.18361318e+00 1.52124956e-01 4.51938540e-01
-1.11592543e+00 -1.22515626e-01 6.14267647e-01 2.64833808e-01
6.92712069e-01 -4.01544034e-01 1.27606943e-01 -1.42873871e+00
2.25196630e-01 -1.02000892e+00 3.09128106e-01 -9.66837704e-01
-9.45469022e-01 7.65830398e-01 -8.02407041e-02 -1.68589497e+00
7.46321827e-02 -4.92273867e-01 -4.41353917e-01 5.73418379e-01
-1.65674126e+00 -7.99151361e-01 -5.71101189e-01 8.77392113e-01
5.98297238e-01 -2.76445746e-01 5.30756116e-01 7.12814450e-01
-6.48116529e-01 6.44362450e-01 3.75845611e-01 -7.83156678e-02
4.75868911e-01 -4.06006664e-01 9.83939096e-02 9.63996053e-01
-2.30264798e-01 2.52668470e-01 4.35088634e-01 -5.23439407e-01
-2.00753999e+00 -1.20569634e+00 1.07068956e-01 1.67066097e-01
5.10926425e-01 -1.02419667e-01 -8.36081862e-01 5.50063193e-01
4.08087879e-01 3.68077725e-01 4.79797304e-01 -7.53238916e-01
-1.52825546e-02 -8.12728405e-02 -9.79971409e-01 5.67203104e-01
1.49130416e+00 -3.94517899e-01 4.92736958e-02 4.78562534e-01
7.64303863e-01 -7.06885099e-01 -7.86054790e-01 5.42253554e-01
3.64948004e-01 -1.32015133e+00 1.06967866e+00 -1.87458824e-02
5.45942605e-01 -4.52004045e-01 -3.18038315e-01 -9.97174323e-01
-5.20133734e-01 -9.07663405e-01 -7.47586131e-01 5.13787210e-01
3.47263962e-02 -5.79076946e-01 8.07797551e-01 6.74772337e-02
-5.14548719e-01 -1.02157342e+00 -1.11527586e+00 -7.34221101e-01
-2.40023300e-01 -4.26056802e-01 5.30495703e-01 9.06351507e-01
-4.40339774e-01 2.73205265e-02 -8.52677226e-01 4.75376427e-01
8.20590377e-01 3.84183347e-01 5.93381882e-01 -7.02246785e-01
-6.96418047e-01 -3.11502188e-01 -2.51188904e-01 -1.64617038e+00
-1.77156284e-01 -6.94437146e-01 -1.91646636e-01 -1.23617351e+00
3.64775270e-01 -5.57201564e-01 -2.85382658e-01 -2.15462726e-02
-2.35947952e-01 4.72705811e-01 4.46418852e-01 5.50229430e-01
-7.44028568e-01 7.28381157e-01 1.52251339e+00 -9.87702757e-02
6.87113702e-02 -1.17519833e-01 -4.93247300e-01 4.81315672e-01
4.92179364e-01 -2.76943266e-01 -4.02572900e-01 -9.62700486e-01
2.38752738e-02 7.26590216e-01 4.17240560e-01 -1.44135165e+00
3.82794648e-01 -9.07663703e-02 4.59590107e-01 -5.17468691e-01
7.68597722e-01 -6.52210057e-01 5.44884443e-01 4.98413414e-01
1.44056544e-01 1.01531081e-01 -7.83270672e-02 5.35828412e-01
-1.22442767e-01 -1.10456884e-01 8.49407077e-01 -2.87035406e-01
-7.32636869e-01 9.38404083e-01 -1.68465957e-01 6.11498356e-02
1.03490484e+00 -3.34550321e-01 -3.55781764e-01 -5.67855716e-01
-5.01710355e-01 2.56417692e-01 4.66984987e-01 2.83059865e-01
8.51066053e-01 -1.35838068e+00 -8.42363954e-01 3.12559217e-01
-2.12707385e-01 1.26050085e-01 7.17184961e-01 9.81622994e-01
-6.52766883e-01 1.47015661e-01 -1.73644423e-01 -7.22142339e-01
-8.78971338e-01 6.86740994e-01 2.79845953e-01 -6.25436381e-03
-8.06876838e-01 1.01789796e+00 1.85886547e-01 -8.49590078e-02
1.00181237e-01 6.17150627e-02 4.35010269e-02 -2.66203403e-01
6.42190933e-01 3.43139529e-01 -1.41205028e-01 -7.11359859e-01
1.66096184e-02 6.67869866e-01 2.88554337e-02 1.50974393e-01
1.50686193e+00 -2.23130792e-01 2.22360040e-03 -1.16069086e-01
1.40241730e+00 -1.76098794e-01 -1.74253869e+00 -5.10149121e-01
-6.92106545e-01 -8.28256011e-01 3.21345702e-02 -3.12129200e-01
-1.56081498e+00 5.57782292e-01 8.40673149e-01 7.60335848e-02
1.46917796e+00 -2.12895855e-01 1.27060997e+00 1.28690049e-01
7.14387894e-01 -8.20313632e-01 2.41575956e-01 4.74444240e-01
7.82241464e-01 -9.53496754e-01 1.41416743e-01 -3.46442282e-01
-4.17892188e-01 9.44216132e-01 5.51923394e-01 -2.04122409e-01
7.41435766e-01 3.55216831e-01 -1.85685143e-01 -2.86461204e-01
-6.66021943e-01 1.02399580e-01 -6.34272248e-02 3.99507731e-01
-5.99203594e-02 -4.08767201e-02 -1.60527855e-01 4.01079476e-01
6.11667968e-02 3.88436526e-01 5.89874923e-01 9.60563362e-01
-2.22351298e-01 -8.54548693e-01 -3.13835949e-01 7.21914887e-01
-3.33072364e-01 -1.33963332e-01 4.43255007e-01 4.61645007e-01
3.54876429e-01 8.24375570e-01 -1.06044069e-01 -6.55893505e-01
3.84522905e-03 -4.51490998e-01 5.10957420e-01 -2.81812072e-01
-1.77848533e-01 6.61932454e-02 -2.04742566e-01 -5.96301734e-01
-4.81695443e-01 -6.66916549e-01 -1.12622070e+00 -5.27688265e-01
-8.67884010e-02 -9.47274789e-02 2.70803720e-01 7.24314630e-01
6.38453305e-01 5.42809367e-01 1.09957564e+00 -1.19839394e+00
-4.15136039e-01 -7.77114749e-01 -4.21321183e-01 2.51075864e-01
3.56784165e-01 -5.63619673e-01 -2.94441611e-01 1.05761096e-01] | [11.071147918701172, -2.0255329608917236] |
38d91348-31b1-4342-b691-3273d8bf556a | correlation-networks-for-extreme-multi-label | null | null | https://dl.acm.org/doi/pdf/10.1145/3394486.3403151 | https://dl.acm.org/doi/pdf/10.1145/3394486.3403151 | Correlation Networks for Extreme Multi-label Text Classification | This paper develops the Correlation Networks (CorNet) architecture for the extreme multi-label text classification (XMTC) task, where the objective is to tag an input text sequence with the most relevant subset of labels from an extremely large label set. XMTC can be found in many real-world applications, such as document tagging and product annotation. Recently, deep learning models have achieved outstanding performances in XMTC tasks. However, these deep XMTC models ignore the useful correlation information among different labels. CorNet addresses this limitation by adding an extra CorNet module at the prediction layer of a deep model, which is able to learn label correlations, enhance raw label predictions with correlation knowledge and output augmented label predictions. We show that CorNet can be easily integrated with deep XMTC models and generalize effectively across different datasets. We further demonstrate that CorNet can bring significant improvements over the existing deep XMTC models in terms of both performance and convergence rate. The models and datasets are available at https://github.com/XunGuangxu/CorNet. | ['Aidong Zhang', 'Jianhui Sun', 'Kishlay Jha', 'Guangxu Xun'] | 2022-08-23 | null | null | null | proceedings-of-the-26th-acm-sigkdd-1 | ['multi-label-text-classification', 'multi-label-text-classification'] | ['methodology', 'natural-language-processing'] | [ 3.27459693e-01 -1.86238661e-01 -2.48616755e-01 -6.30014062e-01
-5.45376122e-01 -4.54641193e-01 3.90378594e-01 1.30494937e-01
-1.66493446e-01 4.16148305e-01 -9.48567390e-02 -2.42525846e-01
-1.71152562e-01 -5.14073491e-01 -2.48928100e-01 -6.09765768e-01
3.34184796e-01 8.20967793e-01 -1.07699580e-01 -3.89188454e-02
-1.02524891e-01 -5.79904467e-02 -1.28981376e+00 6.62247837e-01
7.01414764e-01 1.16677630e+00 1.07335985e-01 5.23195565e-01
-2.09439725e-01 9.51342821e-01 -2.49764472e-01 -5.71070552e-01
9.30527672e-02 -2.21985087e-01 -9.30076480e-01 -2.43985355e-01
3.38380098e-01 5.38955024e-03 2.42804829e-03 9.96752679e-01
5.01737714e-01 -9.09946784e-02 7.33039558e-01 -1.54486537e+00
-7.72459269e-01 1.01615310e+00 -9.13485229e-01 -3.21983159e-01
-1.31554663e-01 -2.96760172e-01 1.48444879e+00 -8.23773146e-01
3.27093482e-01 1.27827013e+00 9.69006300e-01 6.26238585e-01
-1.10254645e+00 -9.64671314e-01 2.28729710e-01 1.61928087e-01
-1.02092826e+00 4.61310819e-02 5.22766292e-01 -3.64457250e-01
8.43238831e-01 9.85056236e-02 8.85585845e-02 1.17816639e+00
6.21304661e-02 1.19267929e+00 9.73047554e-01 -3.64651889e-01
-8.59101713e-02 4.72063459e-02 6.77262127e-01 6.58484757e-01
-1.81345850e-01 -1.08244933e-01 -4.87805307e-01 -2.58218080e-01
2.00029343e-01 2.88499475e-01 1.69945240e-01 -1.02094524e-01
-1.21001518e+00 8.18347633e-01 4.68438119e-01 5.10970294e-01
-5.41639179e-02 4.66526985e-01 6.61711216e-01 3.29584122e-01
9.05840993e-01 4.77688968e-01 -1.03832960e+00 7.92074725e-02
-6.92171574e-01 -1.55141070e-01 5.70822895e-01 1.17210603e+00
6.28549814e-01 -4.16207433e-01 -4.28356528e-01 1.25441372e+00
2.26918519e-01 2.38207787e-01 6.01922214e-01 -8.91252875e-01
4.18789417e-01 7.39774644e-01 -2.62083799e-01 -6.25517130e-01
-8.02445352e-01 -8.83960843e-01 -9.53694999e-01 -1.79973915e-01
8.00408050e-02 -4.59811538e-01 -6.07320309e-01 1.82352078e+00
1.79640114e-01 3.13990146e-01 9.49790180e-02 6.44508362e-01
9.47117388e-01 5.53136945e-01 4.20705020e-01 -5.36173731e-02
1.42658126e+00 -1.40907490e+00 -7.10654676e-01 -3.66254538e-01
1.40127337e+00 -6.90683663e-01 9.69741702e-01 2.28429794e-01
-5.95435381e-01 -5.06122112e-01 -8.16411734e-01 -2.53551662e-01
-5.35154402e-01 5.20425737e-01 9.86310303e-01 2.52997369e-01
-9.37979937e-01 6.05327606e-01 -4.89646345e-01 -3.16245049e-01
6.35771811e-01 5.87508380e-01 -1.72701836e-01 -1.71360388e-01
-1.38795865e+00 7.44998634e-01 5.64943254e-01 1.06634669e-01
-3.64738882e-01 -7.15350688e-01 -5.39923966e-01 3.42229694e-01
4.17839408e-01 -6.58614099e-01 1.65453577e+00 -1.01212454e+00
-1.22658706e+00 6.67284787e-01 -7.65200332e-02 -1.74275890e-01
3.68527025e-01 -2.37533480e-01 -4.19263989e-01 -3.05926532e-01
2.05112487e-01 8.46311688e-01 3.20353776e-01 -1.08468866e+00
-9.34549272e-01 -2.45735630e-01 -2.18999475e-01 1.14390865e-01
-5.94341636e-01 1.53627962e-01 -2.92513400e-01 -4.17182505e-01
-3.93956974e-02 -1.30154514e+00 -2.22641289e-01 -2.59223670e-01
-5.10148048e-01 -6.35425448e-01 9.61569905e-01 -3.41593146e-01
1.09686852e+00 -2.03055358e+00 -4.78714481e-02 -8.33042711e-02
4.38644558e-01 3.98574471e-01 -4.94638562e-01 4.73468572e-01
-1.68800250e-01 1.60800725e-01 -1.09838359e-02 -6.89670563e-01
2.17388526e-01 2.65222155e-02 -1.62477031e-01 1.62165891e-02
6.14119619e-02 1.35932398e+00 -1.09034038e+00 -4.00300235e-01
-8.19236115e-02 2.66351640e-01 -1.87669739e-01 1.83679827e-03
-7.13635921e-01 4.20424640e-01 -4.69203949e-01 4.46269393e-01
5.39425969e-01 -9.86781299e-01 4.60300595e-01 -5.50787710e-02
1.68893799e-01 8.05558041e-02 -5.35703361e-01 1.62340689e+00
-6.98522210e-01 6.15559340e-01 -5.27153432e-01 -8.33706379e-01
9.40381646e-01 5.25216818e-01 8.17573845e-01 -5.72045684e-01
4.91090804e-01 3.60184222e-01 -3.78733017e-02 -2.76136518e-01
4.16598916e-01 -5.21217892e-03 -1.91350058e-01 9.79785621e-01
1.82682037e-01 4.01790977e-01 9.52531323e-02 1.48332208e-01
9.85660434e-01 -3.93205844e-02 1.88318744e-01 -1.27874866e-01
3.62611800e-01 -1.62589744e-01 6.51073873e-01 5.52906752e-01
6.37299642e-02 3.72593284e-01 5.88203490e-01 -5.15121520e-01
-9.10391331e-01 -3.74853194e-01 -5.16697392e-02 1.64710021e+00
-3.55388410e-02 -5.67763448e-01 -3.75213474e-01 -1.15470135e+00
1.43998846e-01 5.99971712e-01 -7.20140398e-01 -2.68023193e-01
-3.53783071e-01 -9.53911483e-01 5.00535727e-01 7.75220931e-01
6.99908495e-01 -8.82382631e-01 -4.81200730e-03 8.88459235e-02
-2.75253534e-01 -1.06242085e+00 -6.35709882e-01 5.55588067e-01
-7.28073597e-01 -8.87844503e-01 -3.35306168e-01 -9.90661383e-01
4.44485635e-01 3.25360328e-01 1.09734893e+00 2.60409623e-01
2.93521974e-02 -1.14960104e-01 -5.36529422e-01 -3.86135697e-01
-5.04227817e-01 6.00799859e-01 -8.30978975e-02 9.44375619e-02
8.82418752e-01 -4.83898103e-01 -2.68272758e-01 5.08072674e-01
-6.71209037e-01 5.51783323e-01 4.90710109e-01 1.20089328e+00
3.43045056e-01 8.24945047e-02 9.57652032e-01 -1.54811478e+00
6.46226943e-01 -6.04697764e-01 -4.36406106e-01 5.28405845e-01
-9.61237013e-01 1.72363818e-01 6.23275220e-01 -5.76647520e-01
-8.81288767e-01 1.38835207e-01 -3.36780131e-01 -2.74685860e-01
1.22591471e-02 6.51624203e-01 1.11187227e-01 1.71814054e-01
4.05269653e-01 -9.13727880e-02 -2.48585492e-01 -6.87859297e-01
3.76290441e-01 1.01460028e+00 1.16104811e-01 -3.70918810e-01
2.66591758e-01 1.44466537e-03 1.45485371e-01 1.44648463e-01
-1.64113331e+00 -6.99330151e-01 -1.02118146e+00 -8.67895931e-02
6.07132792e-01 -8.82354796e-01 -9.22954619e-01 4.15737301e-01
-1.15461171e+00 -5.23991168e-01 1.54411182e-01 3.26677918e-01
-3.76834989e-01 1.36901364e-01 -9.83568251e-01 -5.43479681e-01
-6.02975726e-01 -8.99557769e-01 1.32163680e+00 1.57839969e-01
-1.86777979e-01 -1.36324215e+00 -3.82001922e-02 3.58071715e-01
3.03850412e-01 -2.90945806e-02 1.27852011e+00 -9.96786892e-01
-7.36954287e-02 -2.52443731e-01 -4.21764195e-01 3.34078938e-01
8.32609609e-02 -2.95598030e-01 -1.11180305e+00 -4.23662364e-01
-4.06940460e-01 -5.86502552e-01 1.07209682e+00 1.78825557e-01
1.30746984e+00 6.11423235e-03 -7.84795105e-01 5.28086066e-01
1.33811128e+00 1.98790893e-01 9.11052153e-02 2.65709668e-01
8.74818861e-01 5.39636374e-01 7.08357632e-01 2.34103233e-01
3.95215839e-01 8.24597955e-01 3.42120677e-01 -1.80680305e-01
-8.57730955e-02 -2.12293461e-01 1.71466339e-02 1.05403221e+00
3.64535362e-01 -5.79980314e-01 -1.08671129e+00 2.89266348e-01
-2.38637972e+00 -7.54864156e-01 -3.43520850e-01 1.58233464e+00
7.72786438e-01 -1.48221314e-01 -1.73008457e-01 -3.45125087e-02
9.63012815e-01 -1.05255954e-01 -8.29517365e-01 -2.91596860e-01
-2.53983960e-02 -1.42628253e-01 3.51526856e-01 1.83831573e-01
-1.36422169e+00 1.15159082e+00 6.00835609e+00 8.66045892e-01
-8.77160788e-01 5.39814055e-01 8.37996006e-01 -6.30270168e-02
5.81817422e-03 -1.61664292e-01 -1.00650120e+00 4.32363331e-01
9.85710263e-01 7.50907958e-02 -1.12198017e-04 7.73272216e-01
-1.97006643e-01 3.21997106e-01 -1.29906321e+00 8.73391032e-01
-1.76377118e-01 -1.11621022e+00 -9.21336561e-02 1.49423674e-01
1.00748718e+00 1.74931347e-01 3.56677204e-01 6.81371152e-01
7.48180091e-01 -9.43297267e-01 2.74700671e-01 3.23050171e-01
8.68666708e-01 -7.68491983e-01 1.10818648e+00 4.41593677e-01
-1.08955944e+00 -4.76306707e-01 -4.67873752e-01 3.69937159e-02
3.64076048e-02 8.07849348e-01 -9.89831924e-01 6.03186846e-01
3.53140652e-01 1.24439085e+00 -7.48658776e-01 7.66156077e-01
-2.33155504e-01 6.39313936e-01 -5.65462336e-02 -2.38020256e-01
3.79969776e-01 -8.72590020e-02 -1.17933065e-01 1.26036048e+00
4.04466838e-01 -1.91177294e-01 2.44346276e-01 6.29140198e-01
-3.90732437e-01 2.54804075e-01 -3.00551057e-01 8.40979889e-02
4.92535621e-01 1.53106678e+00 -6.78395987e-01 -4.10533845e-01
-6.18051589e-01 9.29994464e-01 6.70027435e-01 1.93473324e-01
-8.42243612e-01 -1.30628496e-01 5.88684201e-01 -4.68912542e-01
2.05440745e-01 1.01520747e-01 -5.27677298e-01 -1.05552137e+00
-3.01241755e-01 -5.58258653e-01 4.72277522e-01 -8.23051333e-01
-1.58178604e+00 7.22340047e-01 -4.07540381e-01 -1.29450440e+00
-2.23937213e-01 -8.30307007e-01 -2.06216887e-01 8.25905144e-01
-1.39285195e+00 -1.51889205e+00 -2.61394948e-01 2.58289754e-01
4.64633822e-01 -2.99691468e-01 9.57079768e-01 4.18616325e-01
-7.62841702e-01 8.38680923e-01 6.52154863e-01 2.82747984e-01
8.41281772e-01 -1.43197739e+00 5.78265965e-01 2.84756452e-01
1.99766845e-01 3.89360815e-01 2.46159256e-01 -5.61487436e-01
-7.39881873e-01 -1.70148468e+00 1.20788407e+00 -6.14535749e-01
6.48410559e-01 -5.90331614e-01 -6.49777174e-01 8.90116930e-01
1.67678624e-01 5.00752144e-02 1.15361428e+00 6.46925807e-01
-6.69123411e-01 -5.89794666e-02 -6.59201086e-01 5.45502543e-01
1.08644509e+00 -4.70297217e-01 5.60385659e-02 7.81964362e-01
9.18608487e-01 -1.93425000e-01 -1.16801155e+00 3.92293990e-01
7.23594010e-01 -6.03934526e-01 5.64946413e-01 -7.39605963e-01
6.24253452e-01 -9.27308500e-02 9.86371636e-02 -1.48787999e+00
-6.66883409e-01 -2.22288772e-01 -8.90376642e-02 1.24839473e+00
8.46196473e-01 -5.78420103e-01 6.76659882e-01 4.41815257e-01
-2.41910681e-01 -8.89855206e-01 -6.03444278e-01 -6.60260677e-01
2.72012353e-01 -3.85832757e-01 9.34901834e-01 1.28403163e+00
3.14203680e-01 8.81152809e-01 -7.31628478e-01 -1.40776569e-02
3.43890399e-01 3.90957147e-01 3.26009959e-01 -1.77184629e+00
-3.54559124e-01 -3.87036055e-01 -4.75381315e-02 -1.08717215e+00
5.21801233e-01 -1.40598667e+00 3.35240923e-03 -1.56384969e+00
6.41292036e-01 -7.79829860e-01 -6.11069500e-01 1.02730477e+00
-3.08013499e-01 3.02972972e-01 2.63827652e-01 3.52432996e-01
-9.99127328e-01 4.29274023e-01 1.23195744e+00 -2.71086007e-01
7.23528564e-02 3.00700068e-01 -8.68026316e-01 2.91041106e-01
1.12262142e+00 -7.18632579e-01 -4.81074512e-01 -6.97324276e-01
4.93751138e-01 -2.39277948e-02 -2.73150466e-02 -7.80404389e-01
3.48725587e-01 6.15113676e-02 3.13488692e-01 -5.08110106e-01
1.98880687e-01 -9.42679942e-01 3.18137437e-01 3.40961605e-01
-1.01114273e+00 3.79040800e-02 3.00085209e-02 4.13716882e-01
-3.76309603e-02 -3.75485271e-01 6.03123963e-01 4.70629930e-02
-4.94969964e-01 5.20019233e-01 -9.28625762e-02 -4.46068227e-01
9.33312714e-01 3.04217339e-01 -8.25948000e-01 -2.01656416e-01
-6.83340192e-01 6.03815019e-01 1.68722823e-01 6.86023831e-01
1.36462599e-01 -1.64641881e+00 -6.59929633e-01 -2.62657963e-02
3.50888431e-01 -1.49210438e-01 1.26587003e-01 8.39827776e-01
3.85377258e-02 7.57016003e-01 6.15670420e-02 -5.42271376e-01
-1.38890791e+00 7.79826820e-01 3.13949943e-01 -6.54374540e-01
-4.20011669e-01 7.76921332e-01 2.95497835e-01 -6.99950337e-01
1.12563364e-01 -1.20641805e-01 -3.80450577e-01 -1.74251124e-02
4.84855264e-01 6.25399500e-02 3.69062126e-02 -4.88628417e-01
-9.08489600e-02 6.35433078e-01 -5.05596399e-01 2.44373590e-01
1.23851490e+00 -1.64608389e-01 -1.30917579e-01 5.85169554e-01
1.54873931e+00 -6.32787704e-01 -1.11610782e+00 -6.64036810e-01
4.52936947e-01 -3.75939831e-02 6.16680160e-02 -1.21964860e+00
-1.30185091e+00 9.64437068e-01 5.32468021e-01 6.09138794e-02
1.08504856e+00 1.14833392e-01 9.50318038e-01 4.69997197e-01
2.96560407e-01 -1.03496349e+00 2.19531402e-01 8.24499071e-01
3.79477859e-01 -1.43470168e+00 -2.81029671e-01 -4.23836827e-01
-7.59827733e-01 1.09449637e+00 7.68796861e-01 4.60094720e-01
7.12485969e-01 2.59107590e-01 2.52924711e-01 -3.31473023e-01
-1.33714354e+00 -3.48181725e-01 8.04620087e-02 3.92605245e-01
8.63323987e-01 1.19034082e-01 -3.61492544e-01 6.47823036e-01
3.67132053e-02 -1.10388406e-01 1.59587786e-01 5.89735568e-01
-2.63494045e-01 -1.38523245e+00 1.96995735e-01 8.30960929e-01
-3.49603653e-01 -3.50780457e-01 -3.92113090e-01 3.43405843e-01
2.60161966e-01 9.72749770e-01 -1.05938770e-01 -7.73153067e-01
6.19455939e-03 4.13746446e-01 -5.66767454e-02 -7.24595010e-01
-7.93863535e-01 -1.79649740e-02 7.23897815e-02 -6.83314651e-02
-4.12153751e-01 -6.93029404e-01 -1.12707520e+00 -2.59182513e-01
-5.91098368e-01 3.25056344e-01 5.01861691e-01 9.20316339e-01
5.57852089e-01 6.81869924e-01 7.42114723e-01 -3.82301956e-01
-3.21878403e-01 -1.38566983e+00 -6.10186338e-01 3.29279095e-01
8.96559656e-02 -6.18652880e-01 -1.69442147e-02 -8.90432447e-02] | [9.619002342224121, 4.451057434082031] |
5eb25b5a-b285-447a-95f5-a8698046a7b3 | exploiting-neural-query-translation-into | 2010.13659 | null | https://arxiv.org/abs/2010.13659v1 | https://arxiv.org/pdf/2010.13659v1.pdf | Exploiting Neural Query Translation into Cross Lingual Information Retrieval | As a crucial role in cross-language information retrieval (CLIR), query translation has three main challenges: 1) the adequacy of translation; 2) the lack of in-domain parallel training data; and 3) the requisite of low latency. To this end, existing CLIR systems mainly exploit statistical-based machine translation (SMT) rather than the advanced neural machine translation (NMT), limiting the further improvements on both translation and retrieval quality. In this paper, we investigate how to exploit neural query translation model into CLIR system. Specifically, we propose a novel data augmentation method that extracts query translation pairs according to user clickthrough data, thus to alleviate the problem of domain-adaptation in NMT. Then, we introduce an asynchronous strategy which is able to leverage the advantages of the real-time in SMT and the veracity in NMT. Experimental results reveal that the proposed approach yields better retrieval quality than strong baselines and can be well applied into a real-world CLIR system, i.e. Aliexpress e-Commerce search engine. Readers can examine and test their cases on our website: https://aliexpress.com . | ['Boxing Chen', 'Weihua Luo', 'Haibo Zhang', 'Baosong Yang', 'Liang Yao'] | 2020-10-26 | null | null | null | null | ['cross-lingual-information-retrieval'] | ['natural-language-processing'] | [ 5.18062934e-02 -5.46246469e-01 -5.73628843e-01 6.85357228e-02
-1.50557840e+00 -7.12498307e-01 8.04648340e-01 1.96575690e-02
-7.22715497e-01 7.71299481e-01 1.64128438e-01 -7.24012315e-01
-5.66079468e-02 -5.57674170e-01 -7.02502310e-01 -2.91732371e-01
5.01521707e-01 7.25733161e-01 1.33136660e-01 -6.42786264e-01
1.06624730e-01 2.25575253e-01 -8.84941757e-01 3.05321515e-01
1.24194992e+00 8.89614046e-01 1.25707731e-01 2.44531348e-01
-3.33861232e-01 3.57317239e-01 -3.65317464e-01 -5.89882255e-01
3.37814003e-01 -3.02610993e-01 -7.48000383e-01 -4.59597200e-01
1.40191808e-01 -3.57262790e-01 -2.17131227e-01 9.67953146e-01
7.73193121e-01 -4.56790440e-02 4.98362213e-01 -1.07888770e+00
-1.07176650e+00 5.10846436e-01 -6.79565609e-01 2.08198234e-01
2.06712633e-01 3.31938863e-02 1.20209086e+00 -1.21526194e+00
5.90269089e-01 9.39757109e-01 2.42102340e-01 3.95564675e-01
-9.73321140e-01 -7.43725359e-01 -1.39292583e-01 1.56353951e-01
-1.50923920e+00 -5.82509458e-01 6.62664592e-01 -1.37325078e-01
7.80264854e-01 3.18676949e-01 1.73781157e-01 1.22568619e+00
1.02184556e-01 1.05260241e+00 1.00712097e+00 -6.64347708e-01
2.29937304e-02 4.27250803e-01 -1.25220627e-01 2.36656249e-01
2.36606933e-02 6.29502833e-02 -5.23891509e-01 -2.67997175e-01
7.33136833e-01 4.76302393e-03 -2.29468599e-01 -1.18536197e-01
-1.40258431e+00 7.17202544e-01 2.99340755e-01 4.31715935e-01
-5.51421344e-01 -2.76776701e-01 6.04027033e-01 6.49856925e-01
5.75452685e-01 5.96745074e-01 -6.40097737e-01 1.54424697e-01
-8.95462632e-01 7.64337033e-02 5.26243269e-01 1.16388166e+00
6.74049377e-01 -3.01618367e-01 -4.14417714e-01 1.07934368e+00
2.21738368e-01 8.37930918e-01 8.73780429e-01 -3.55187595e-01
7.79718399e-01 5.11125207e-01 2.30047181e-01 -6.67939782e-01
1.92508161e-01 -6.87519908e-01 -7.82582283e-01 -5.66559613e-01
1.08792022e-01 2.16904981e-03 -6.24966085e-01 1.69028234e+00
1.43180802e-01 -1.79021209e-01 3.15528095e-01 1.01453650e+00
5.49815834e-01 6.61249280e-01 8.70463029e-02 -3.45985681e-01
1.45412087e+00 -1.16424608e+00 -8.34697723e-01 -3.58672410e-01
9.04402137e-01 -1.38478804e+00 1.43230355e+00 -2.75960136e-02
-1.20134377e+00 -4.86702412e-01 -9.08320308e-01 -1.41659960e-01
-3.98890764e-01 5.04505932e-01 5.04686713e-01 2.29017451e-01
-1.02491987e+00 8.34945515e-02 -7.12378681e-01 -4.74584967e-01
-1.61067069e-01 2.64607221e-01 -1.64311856e-01 -2.22869337e-01
-1.67520452e+00 8.55481446e-01 3.19941819e-01 1.36884153e-01
-3.37466657e-01 -4.45158392e-01 -4.43320602e-01 1.63949970e-02
4.80818421e-01 -7.76485741e-01 1.49980867e+00 -1.02409244e+00
-1.38020885e+00 6.48753703e-01 -3.10281575e-01 -2.27118239e-01
6.36690497e-01 -4.34347063e-01 -5.39818943e-01 -4.45111580e-02
2.81669438e-01 3.86328012e-01 4.03903693e-01 -9.56950366e-01
-5.76627016e-01 -4.53351945e-01 -1.30638614e-01 4.17104393e-01
-6.70384109e-01 2.62831777e-01 -9.47546363e-01 -6.66869581e-01
5.61208241e-02 -1.02211165e+00 -8.16789493e-02 -3.48958969e-01
-3.11875165e-01 -5.22262454e-01 3.84532273e-01 -7.76504755e-01
1.51567817e+00 -1.99671888e+00 -1.08607858e-01 1.31128162e-01
-1.44849017e-01 5.83943605e-01 -4.39627498e-01 8.70156467e-01
2.88971126e-01 1.94302261e-01 1.57164812e-01 -8.14465657e-02
-2.69782655e-02 -1.25393733e-01 -4.63153273e-01 -1.30107682e-02
1.66503772e-01 1.31210124e+00 -7.48529077e-01 -6.94297552e-01
-2.34422699e-01 2.61983365e-01 -3.25705260e-01 2.39413261e-01
-3.21211308e-01 3.96587640e-01 -9.72654462e-01 6.97074115e-01
5.92138350e-01 -3.77128929e-01 2.08844841e-01 -1.10309742e-01
-1.02692522e-01 5.60911715e-01 -6.69443667e-01 2.00339508e+00
-6.66642129e-01 1.79535702e-01 -1.83734924e-01 -5.97174466e-01
1.04184413e+00 5.80143988e-01 3.61013919e-01 -1.36781085e+00
-5.22235632e-02 7.99422979e-01 -2.05870286e-01 -2.93860048e-01
6.91223502e-01 1.97810158e-01 4.27916907e-02 7.53665388e-01
-2.37441406e-01 3.29557627e-01 1.48571059e-01 2.00936794e-01
7.37872541e-01 2.08911866e-01 1.40909284e-01 -2.23918915e-01
6.36563599e-01 2.37704605e-01 4.14090008e-01 6.01470053e-01
2.29675379e-02 2.63065279e-01 -1.34114772e-01 -7.04366788e-02
-1.23369217e+00 -7.31910467e-01 1.30626470e-01 1.35327363e+00
9.42995772e-02 -1.91591799e-01 -4.61250633e-01 -5.39215744e-01
-2.39180654e-01 6.04986310e-01 -1.56155854e-01 -2.14618489e-01
-8.60223532e-01 -6.92162097e-01 5.51844239e-01 4.32852417e-01
4.87230122e-01 -9.70123470e-01 9.82390195e-02 1.62576467e-01
-7.11225510e-01 -1.28714132e+00 -1.01933360e+00 -2.45131671e-01
-9.56673026e-01 -5.36846340e-01 -8.86161149e-01 -8.91902924e-01
3.30583185e-01 7.46283591e-01 1.19710529e+00 1.45041466e-01
2.60932654e-01 1.45454668e-02 -5.56286633e-01 -2.60967582e-01
-4.74759698e-01 6.57529056e-01 8.45704302e-02 -2.08689079e-01
9.53367352e-01 -4.54693347e-01 -8.56911838e-01 5.30495882e-01
-1.03669739e+00 1.05493397e-01 1.33273494e+00 9.38110709e-01
6.00912273e-01 -3.33413899e-01 8.64380419e-01 -7.48981833e-01
1.13065445e+00 -5.17524123e-01 -6.12374008e-01 7.99800754e-01
-1.07982838e+00 1.99418843e-01 5.51377475e-01 -6.33331001e-01
-1.00500453e+00 -3.26396435e-01 -1.25301078e-01 -3.81518483e-01
3.11844289e-01 8.95519614e-01 2.59636398e-02 2.53300667e-01
7.46339798e-01 6.85070157e-01 -1.10089816e-01 -7.25831985e-01
3.44362587e-01 9.86643195e-01 2.76280522e-01 -7.28550196e-01
8.87470126e-01 1.66086685e-02 -3.91934246e-01 -3.25363189e-01
-5.62255323e-01 -8.27231050e-01 -5.18225849e-01 2.49853432e-01
5.64599872e-01 -1.16536319e+00 -5.09200215e-01 -5.75122274e-02
-1.18960977e+00 1.49216950e-02 2.63048232e-01 7.65489697e-01
-3.18202853e-01 3.93522143e-01 -9.52486157e-01 -7.03357637e-01
-1.00025761e+00 -1.09074199e+00 1.23727369e+00 2.00239792e-02
4.80411202e-02 -8.73816490e-01 1.49315313e-01 5.80722690e-01
6.08714342e-01 -6.26846433e-01 8.84971619e-01 -1.01989770e+00
-7.11382568e-01 -2.27277175e-01 -4.69380260e-01 2.06561223e-01
1.02699853e-01 -2.13608935e-01 -7.59588599e-01 -4.98355508e-01
-1.76959932e-01 -4.38530356e-01 4.08362031e-01 -1.73614979e-01
5.91999292e-01 -5.08002520e-01 -1.96974754e-01 1.47653088e-01
1.43245745e+00 1.59082353e-01 4.29722309e-01 6.65171325e-01
3.41948301e-01 5.75947165e-01 1.01076245e+00 -1.63095794e-03
2.78475076e-01 1.09465265e+00 -1.58958495e-01 -3.49483818e-01
-5.81455715e-02 -7.50152349e-01 3.27231795e-01 1.38515651e+00
7.51563534e-02 -1.83622524e-01 -9.78270948e-01 5.24638057e-01
-2.02180600e+00 -5.42317510e-01 5.28478213e-02 2.40522218e+00
1.15505195e+00 3.82976723e-03 -1.13804437e-01 -3.68380815e-01
6.65367126e-01 -2.67946303e-01 -5.52057862e-01 -2.06533596e-01
-1.65080518e-01 1.73738912e-01 5.13508856e-01 4.28102553e-01
-6.79391503e-01 1.20994604e+00 5.35795259e+00 1.15630567e+00
-1.18828964e+00 3.87958974e-01 4.10122424e-01 1.67527735e-01
-4.54755962e-01 -1.01142988e-01 -8.56788635e-01 3.59135240e-01
9.17211533e-01 -5.20698309e-01 5.18935323e-01 6.26775205e-01
3.18698138e-01 4.75400805e-01 -1.14210832e+00 7.36214042e-01
-7.10205585e-02 -1.06109834e+00 3.64208192e-01 1.81292862e-01
4.62869167e-01 2.08830282e-01 1.26619816e-01 6.62340879e-01
1.59324661e-01 -6.06072724e-01 3.58941197e-01 1.31071806e-01
1.01735044e+00 -5.48222601e-01 8.14670920e-01 7.50874639e-01
-9.05088067e-01 1.94142818e-01 -3.87934417e-01 4.21058655e-01
-8.62442777e-02 2.40928903e-01 -1.01518643e+00 9.39403176e-01
3.51338536e-01 3.07814211e-01 -3.97801757e-01 6.62937462e-01
-2.01597676e-01 4.09203053e-01 -2.57639170e-01 -1.45409599e-01
4.24309671e-01 -3.69579971e-01 2.79617429e-01 1.15842068e+00
4.25235212e-01 -2.31684715e-01 1.87856510e-01 6.12604916e-01
-3.96208256e-01 7.98658490e-01 -5.51197290e-01 -2.56137699e-01
6.94474936e-01 1.17698693e+00 -5.56148440e-02 -3.35004538e-01
-7.28268206e-01 1.15887475e+00 2.70552576e-01 6.69812977e-01
-5.20455420e-01 -1.83518957e-02 4.25829709e-01 6.41915128e-02
-9.79824886e-02 -1.91357359e-01 -7.13659078e-02 -1.37707829e+00
4.86138225e-01 -1.32759941e+00 4.35929567e-01 -6.66171253e-01
-1.39620245e+00 8.95673931e-01 -2.78966010e-01 -1.48133087e+00
-5.91364205e-01 -2.29238063e-01 -9.06871632e-02 1.30366826e+00
-2.01569819e+00 -1.45704293e+00 1.86783403e-01 5.48051596e-01
7.86460996e-01 -2.83264279e-01 8.17391336e-01 8.93322885e-01
-5.24028599e-01 9.78580117e-01 5.32576561e-01 2.38222048e-01
1.22519708e+00 -7.18879342e-01 5.37508309e-01 8.39694977e-01
1.89372614e-01 1.03658175e+00 3.72388810e-01 -5.70746720e-01
-1.61235511e+00 -1.01776671e+00 1.40498424e+00 -4.53045070e-01
8.22037518e-01 -3.01887184e-01 -1.02373040e+00 6.45541966e-01
3.76576990e-01 -4.64957625e-01 7.05097020e-01 2.37761810e-01
-4.37757403e-01 -2.58223891e-01 -6.89866066e-01 8.92449141e-01
6.30313158e-01 -7.15859532e-01 -5.04058063e-01 5.30623317e-01
1.06069362e+00 -1.86541975e-01 -8.09276223e-01 6.26932263e-01
6.53662503e-01 -4.18907791e-01 1.02005017e+00 -8.23000133e-01
4.46491569e-01 -2.20650062e-01 -1.62991136e-01 -1.04987013e+00
-1.62606284e-01 -5.18989205e-01 8.46900269e-02 1.22418630e+00
7.43931770e-01 -7.11084604e-01 5.21576583e-01 6.73587620e-01
2.26931989e-01 -7.35339344e-01 -7.81002402e-01 -7.87382007e-01
3.96232158e-01 1.11093363e-02 7.01218665e-01 9.97998297e-01
-1.07747458e-01 8.55018318e-01 -4.14789528e-01 1.50796352e-02
2.20301419e-01 2.88643152e-01 8.50718439e-01 -8.95744741e-01
-5.04845202e-01 -4.32549596e-01 4.33632344e-01 -1.55270088e+00
-7.27462098e-02 -8.43376219e-01 -1.75286695e-01 -1.27143109e+00
4.98546541e-01 -6.94038749e-01 -5.84999621e-01 3.18401605e-01
-4.25643653e-01 1.85734302e-01 7.67676681e-02 9.09856558e-01
-5.80180824e-01 5.75155377e-01 1.34590936e+00 -4.57050316e-02
-3.04963619e-01 5.96264564e-02 -8.06164920e-01 8.90232325e-02
8.69488716e-01 -4.00029570e-01 -5.45928240e-01 -8.31116259e-01
3.72752368e-01 3.66349131e-01 -1.35086358e-01 -3.68340492e-01
4.21315193e-01 -2.78511554e-01 1.62173249e-02 -3.97789955e-01
1.89900964e-01 -7.58517623e-01 -2.32156649e-01 2.33642742e-01
-5.53248703e-01 6.38785303e-01 1.38021648e-01 3.83396983e-01
-4.52742457e-01 -4.94321110e-03 3.00334990e-01 -2.01341659e-01
-3.28934163e-01 4.30734456e-01 -6.11040182e-02 -6.97944835e-02
4.31623131e-01 3.49136025e-01 -4.80078489e-01 -5.15643895e-01
-2.44912785e-02 4.26662326e-01 3.70673239e-01 7.64881372e-01
3.62986177e-01 -1.43255997e+00 -1.01560509e+00 1.50376335e-01
5.09762585e-01 -3.85852218e-01 -3.47505249e-02 1.03516030e+00
-1.59066156e-01 9.79819715e-01 6.52175844e-02 -5.35782397e-01
-9.96580422e-01 7.26036072e-01 -3.34634222e-02 -7.67979383e-01
-2.13942632e-01 2.87751257e-01 2.06832439e-01 -6.19248271e-01
8.72009918e-02 2.57055819e-01 -9.36824456e-02 -3.03584069e-01
5.25119066e-01 -1.62348658e-01 3.87336671e-01 -4.51513469e-01
-8.16530511e-02 3.42699051e-01 -5.00537753e-01 -4.11141455e-01
9.49025393e-01 -4.85359401e-01 -9.96674523e-02 2.63960779e-01
1.20914817e+00 5.79034165e-02 -3.81874293e-01 -9.30301249e-01
3.61409575e-01 -5.20043194e-01 1.45959947e-02 -1.06958175e+00
-6.74467623e-01 8.75458717e-01 4.75766003e-01 -9.71368030e-02
1.35956538e+00 -2.17328101e-01 1.34518361e+00 7.28246033e-01
4.09998953e-01 -1.02757537e+00 -1.66436005e-02 2.95054018e-01
7.75640249e-01 -1.44976473e+00 -2.57100374e-01 -2.65431285e-01
-6.26377225e-01 8.49016428e-01 4.27863061e-01 3.62324506e-01
2.26106554e-01 -1.65286705e-01 5.91515124e-01 1.45166621e-01
-9.62988794e-01 -1.55980974e-01 5.03164649e-01 8.10864419e-02
8.91678452e-01 -1.35070533e-01 -8.68820310e-01 3.34026754e-01
7.00563490e-02 2.15625137e-01 -1.52184233e-01 8.35177243e-01
-1.32908642e-01 -1.58071744e+00 -1.73430413e-01 1.20733112e-01
-7.50705361e-01 -5.84557474e-01 -4.12005275e-01 5.93611777e-01
-5.56105316e-01 1.01651025e+00 -3.78058076e-01 -2.68942535e-01
3.58016998e-01 2.80416608e-01 1.94660928e-02 -3.92328084e-01
-5.14631629e-01 5.52673995e-01 5.76417260e-02 -3.72246116e-01
-2.53881633e-01 -3.04628283e-01 -7.06368387e-01 -3.38238329e-01
-5.36167920e-01 6.20583117e-01 9.40510035e-01 7.96051681e-01
6.49324775e-01 7.41386116e-02 7.34268606e-01 -1.38772232e-03
-9.65486765e-01 -1.28809726e+00 -1.57847941e-01 4.57329690e-01
1.49449661e-01 -1.94020510e-01 -5.42951338e-02 -1.86702430e-01] | [11.599230766296387, 10.040308952331543] |
c49a6b4c-3916-42ff-983c-49c398114bd7 | implicit-neural-networks-with-fourier-feature | 2305.06822 | null | https://arxiv.org/abs/2305.06822v1 | https://arxiv.org/pdf/2305.06822v1.pdf | Implicit Neural Networks with Fourier-Feature Inputs for Free-breathing Cardiac MRI Reconstruction | In this paper, we propose an approach for cardiac magnetic resonance imaging (MRI), which aims to reconstruct a real-time video of a beating heart from continuous highly under-sampled measurements. This task is challenging since the object to be reconstructed (the heart) is continuously changing during signal acquisition. To address this challenge, we represent the beating heart with an implicit neural network and fit the network so that the representation of the heart is consistent with the measurements. The network in the form of a multi-layer perceptron with Fourier-feature inputs acts as an effective signal prior and enables adjusting the regularization strength in both the spatial and temporal dimensions of the signal. We examine the proposed approach for 2D free-breathing cardiac real-time MRI in different operating regimes, i.e., for different image resolutions, slice thicknesses, and acquisition lengths. Our method achieves reconstruction quality on par with or slightly better than state-of-the-art untrained convolutional neural networks and superior image quality compared to a recent method that fits an implicit representation directly to Fourier-domain measurements. However, this comes at a higher computational cost. Our approach does not require any additional patient data or biosensors including electrocardiography, making it potentially applicable in a wide range of clinical scenarios. | ['Reinhard Heckel', 'Stefan Ruschke', 'Johannes F. Kunz'] | 2023-05-11 | null | null | null | null | ['mri-reconstruction'] | ['computer-vision'] | [ 6.64241970e-01 1.93246573e-01 2.31430292e-01 -1.28395468e-01
-3.93316984e-01 -2.30045572e-01 1.10888956e-02 -8.48695338e-02
-6.81624711e-01 5.89568615e-01 -1.90878406e-01 -1.58288956e-01
-2.62455970e-01 -4.14196134e-01 -6.58132672e-01 -9.08667684e-01
-1.64325729e-01 3.81409585e-01 5.09362258e-02 1.53228164e-01
-1.47242755e-01 6.28416657e-01 -7.94566095e-01 6.97064400e-02
4.58141625e-01 1.04057252e+00 2.46029422e-01 8.69301736e-01
4.81516689e-01 7.87247717e-01 -4.03681308e-01 1.11208729e-01
2.55113661e-01 -6.12329960e-01 -5.66212296e-01 2.72065718e-02
2.43210986e-01 -4.47742105e-01 -4.97452974e-01 8.44959080e-01
9.28920507e-01 -1.24191247e-01 3.32200497e-01 -3.29636306e-01
-1.69266462e-01 4.47097182e-01 -3.74581069e-01 4.26244169e-01
-1.05430901e-01 2.06098586e-01 3.25197637e-01 -6.72751129e-01
6.38275862e-01 3.86561871e-01 1.00170827e+00 6.42737031e-01
-1.71039677e+00 -3.62927616e-01 -2.70531446e-01 -1.09255694e-01
-1.13368547e+00 -5.95535040e-01 9.26988125e-01 -5.95906138e-01
6.20737970e-01 2.23054618e-01 8.01314890e-01 8.82073700e-01
6.56885326e-01 8.72586295e-02 1.09292424e+00 -3.74906451e-01
1.86317652e-01 -1.42809199e-02 -1.15802839e-01 5.12857258e-01
4.86208536e-02 1.48833126e-01 -2.28974193e-01 -2.42121980e-01
1.26041019e+00 2.38913223e-01 -8.54311526e-01 -4.99525428e-01
-1.61149108e+00 4.70277011e-01 2.63166666e-01 6.73061669e-01
-8.10088277e-01 3.60969305e-01 3.52413774e-01 1.35466874e-01
1.14827566e-01 5.59434712e-01 -2.61389405e-01 -1.90491546e-02
-1.13074303e+00 -2.15639278e-01 6.81503236e-01 1.77340791e-01
1.60030320e-01 3.06119353e-01 -1.64550006e-01 5.22325218e-01
2.38447934e-01 2.41179675e-01 6.51715100e-01 -1.17524874e+00
2.90037524e-02 -6.64454326e-02 1.58954397e-01 -7.96352208e-01
-7.56322980e-01 -7.49306500e-01 -1.19264185e+00 2.74940997e-01
4.94209975e-01 -3.50599051e-01 -6.26147807e-01 1.59709120e+00
4.23641384e-01 5.36383510e-01 -2.16323324e-02 1.41137528e+00
8.12988877e-01 4.17939961e-01 -1.92501798e-01 -7.26496100e-01
1.27180171e+00 -4.84322876e-01 -9.28923368e-01 -2.30717182e-01
8.37351009e-02 -3.54853958e-01 7.32352197e-01 5.07216513e-01
-1.25408304e+00 -6.38258278e-01 -1.25212622e+00 3.10627043e-01
3.72833103e-01 1.13254495e-01 1.89192921e-01 4.97185767e-01
-9.35780704e-01 9.20669973e-01 -1.16083860e+00 6.85495734e-02
1.67560250e-01 4.55384821e-01 -2.96840221e-01 9.40428898e-02
-1.10417175e+00 7.80113876e-01 2.28248879e-01 4.35772836e-01
-6.98286057e-01 -1.04780853e+00 -6.01729810e-01 3.17611592e-03
1.72079444e-01 -7.77665913e-01 9.84319687e-01 -8.91117692e-01
-1.95826328e+00 8.30168843e-01 1.27836004e-01 -5.10592818e-01
6.93417728e-01 9.74133238e-03 -3.30684036e-01 4.25295234e-01
-3.77825350e-01 2.75818378e-01 1.02044106e+00 -9.56538141e-01
3.69869947e-01 -3.19616407e-01 -1.33690342e-01 -3.95285338e-02
3.20377350e-02 -1.71912327e-01 -3.18917096e-01 -5.86504638e-01
4.92303699e-01 -9.48045969e-01 -3.98114085e-01 3.81464452e-01
-6.14749342e-02 8.41139138e-01 4.97889429e-01 -7.18850374e-01
9.91364360e-01 -2.14253616e+00 3.54500145e-01 1.24477424e-01
5.48642099e-01 3.27934980e-01 1.54796675e-01 1.18158758e-01
-2.58863688e-01 -2.60871559e-01 -5.15450120e-01 -8.05541575e-02
-6.11347020e-01 1.36608612e-02 1.50332987e-01 9.89195466e-01
2.33465564e-02 7.67728746e-01 -7.37163067e-01 -4.22283024e-01
3.12479496e-01 1.02717674e+00 -3.47145468e-01 4.67599005e-01
2.29488134e-01 1.33258951e+00 -3.71661127e-01 4.90622483e-02
5.38348794e-01 -4.31851864e-01 6.73075020e-01 -5.49773097e-01
-2.01164886e-01 1.04192980e-01 -1.05475080e+00 1.99716365e+00
-6.19381964e-01 5.59470356e-01 5.64517140e-01 -1.48110580e+00
8.50677371e-01 8.45663846e-01 9.91274893e-01 -8.65932107e-01
2.04679474e-01 2.24448964e-01 2.95796514e-01 -9.30266559e-01
-2.36843511e-01 -7.46777833e-01 3.70904326e-01 5.09225011e-01
1.97337661e-03 -1.08875081e-01 -2.93147922e-01 -4.09085274e-01
1.25501263e+00 1.72052994e-01 1.29929230e-01 -3.43062580e-01
5.40516257e-01 -4.91365492e-01 6.86720431e-01 6.91171646e-01
-3.26046824e-01 7.99929023e-01 3.15624952e-01 -8.78136337e-01
-1.01425624e+00 -8.66365552e-01 -4.33482498e-01 3.31006318e-01
-1.07229881e-01 1.20671391e-01 -6.92754447e-01 -2.96910763e-01
-2.56327391e-01 1.40999526e-01 -4.77381915e-01 -5.60084842e-02
-1.05606103e+00 -7.99153984e-01 4.50603276e-01 1.88292906e-01
2.07939178e-01 -1.04613769e+00 -1.42583680e+00 7.93780148e-01
-3.59312117e-01 -1.34684741e+00 -2.34479964e-01 3.81854951e-01
-1.17932773e+00 -8.45146298e-01 -8.79386127e-01 -4.29198384e-01
5.12030602e-01 -3.58762234e-01 1.09638333e+00 -1.20041281e-01
-4.96349335e-01 2.81166553e-01 8.53916407e-02 -1.53102651e-01
-5.26072562e-01 -3.24702740e-01 -7.05026090e-02 2.71310955e-01
-2.78894395e-01 -8.85294259e-01 -1.00193787e+00 1.45638257e-01
-7.97247887e-01 1.60537064e-01 4.99809533e-01 9.25645709e-01
7.75679529e-01 -8.04059356e-02 5.49389899e-01 -7.54882455e-01
3.60818654e-01 -3.32311720e-01 -5.78905284e-01 -8.38747472e-02
-2.88784325e-01 8.60806555e-02 8.07576954e-01 -6.30115688e-01
-6.36008084e-01 2.75753617e-01 -1.92455187e-01 -5.22734284e-01
-4.13231663e-02 5.32896996e-01 3.41797084e-01 -2.86316812e-01
6.84088886e-01 4.15216565e-01 4.84129131e-01 -2.51632869e-01
-1.58175319e-01 4.17730957e-01 7.30648160e-01 -2.71415740e-01
3.10920596e-01 6.43358052e-01 4.34103847e-01 -9.56271350e-01
-3.66782933e-01 -1.68313146e-01 -8.01409364e-01 -4.71269369e-01
8.98556948e-01 -5.56722701e-01 -8.95027459e-01 3.89118433e-01
-1.10628688e+00 -4.71196800e-01 -4.17806119e-01 9.88505840e-01
-7.56215274e-01 5.11756003e-01 -8.90450895e-01 -6.64515853e-01
-7.05226421e-01 -1.21853113e+00 6.78797245e-01 -2.22155556e-01
-2.03324348e-01 -1.01224959e+00 1.85113430e-01 2.93423653e-01
7.57269621e-01 9.23931062e-01 7.43975163e-01 6.90481588e-02
-4.42266852e-01 -1.09946966e-01 1.23479746e-01 3.96356165e-01
1.21367685e-01 -4.20396924e-01 -1.02403438e+00 -5.15209496e-01
7.55767345e-01 -1.41418040e-01 6.01090968e-01 1.07950926e+00
1.21842349e+00 -1.61923170e-01 -4.25995104e-02 7.76729524e-01
1.44390535e+00 1.55912206e-01 6.32842958e-01 -3.59495729e-02
4.68281180e-01 4.52932179e-01 1.01643629e-01 6.19778514e-01
-2.00798884e-02 6.98385358e-01 4.38672841e-01 -3.91147703e-01
-1.37820512e-01 4.05630797e-01 1.16492435e-01 9.30256128e-01
-2.13566914e-01 8.56661946e-02 -8.56689513e-01 2.54438519e-01
-1.63298643e+00 -7.77743518e-01 5.14299273e-02 2.42294312e+00
8.33466411e-01 9.91851091e-02 -1.56664476e-02 4.02363271e-01
5.92130542e-01 2.95749167e-03 -8.24235141e-01 -1.93002790e-01
2.93519408e-01 3.71670544e-01 4.21809196e-01 3.85401785e-01
-1.04167604e+00 -1.69009686e-01 6.37560797e+00 -7.20611662e-02
-1.95609462e+00 4.22372520e-01 6.41872287e-01 -2.19619811e-01
9.89777520e-02 -3.70120347e-01 3.68009545e-02 4.41787869e-01
1.17505133e+00 2.82559514e-01 5.43538630e-01 1.66988805e-01
5.82696676e-01 1.26254454e-01 -1.19173038e+00 1.11355031e+00
-1.85624938e-02 -1.39908230e+00 -6.62068129e-01 -9.19826552e-02
1.99343011e-01 1.32608980e-01 -8.71578455e-02 -3.04729074e-01
-7.51447380e-01 -1.12093174e+00 6.12625539e-01 9.17184293e-01
1.20065987e+00 -3.00034076e-01 6.60911500e-01 6.22227311e-01
-8.35710227e-01 3.46798897e-02 -5.12416661e-02 6.51506037e-02
2.52405137e-01 9.49778497e-01 -5.88445961e-01 3.17513436e-01
4.75205868e-01 5.93018115e-01 1.62361532e-01 9.17792439e-01
3.44755232e-01 7.17824340e-01 -2.88767368e-01 5.49510121e-01
4.64514866e-02 -2.24942982e-01 6.13634527e-01 9.80533481e-01
3.15166473e-01 3.83939534e-01 1.17824994e-01 1.06341457e+00
1.08601362e-01 -1.62113085e-01 -4.10063982e-01 1.19192600e-01
1.54640945e-02 1.18605769e+00 -6.70751929e-01 -1.66952342e-01
-4.05033678e-01 6.66230202e-01 6.20186105e-02 4.20694560e-01
-7.99185693e-01 -1.69426531e-01 1.15404956e-01 7.73366392e-01
2.08404407e-01 -2.40676910e-01 -4.61869054e-02 -1.10467935e+00
2.57495165e-01 -6.77724540e-01 1.03493847e-01 -6.60561204e-01
-8.35138321e-01 7.22408593e-01 -3.05144370e-01 -1.31214643e+00
-2.36440688e-01 -4.09853280e-01 -4.74910706e-01 9.31320786e-01
-1.60035765e+00 -6.04931891e-01 -3.52243334e-01 3.50987107e-01
6.34215847e-02 2.53419876e-01 9.61699486e-01 6.65832400e-01
-2.63727874e-01 5.58028668e-02 -2.04984676e-02 1.23168901e-01
4.92781222e-01 -9.65983331e-01 1.20659843e-02 5.43318093e-01
-2.67787933e-01 4.55236524e-01 7.25975513e-01 -2.97020227e-01
-1.52160192e+00 -8.81785452e-01 3.73991013e-01 9.35611874e-02
3.16810906e-01 -1.80077478e-01 -1.03189576e+00 2.96057016e-01
1.37955368e-01 8.00075531e-01 5.51297486e-01 -5.76009929e-01
2.73505777e-01 -3.53266329e-01 -1.44236481e+00 3.33711654e-02
4.99529600e-01 -4.04666543e-01 -3.44731301e-01 2.07194120e-01
3.58744651e-01 -6.82311535e-01 -1.31270325e+00 3.85882378e-01
8.83713007e-01 -9.27998722e-01 1.01735580e+00 -1.53790504e-01
1.84372574e-01 -2.49035522e-01 2.51152962e-01 -1.19490874e+00
-4.93483782e-01 -8.42412353e-01 -3.81790131e-01 3.31641227e-01
-1.69585217e-02 -7.90562272e-01 7.41849005e-01 4.65650022e-01
-7.87341595e-02 -9.52419817e-01 -1.29281425e+00 -5.16615510e-01
-1.28302630e-02 -2.51750320e-01 2.13862568e-01 9.37710047e-01
-2.29259226e-02 8.46947655e-02 -4.86830652e-01 2.60443300e-01
9.57753420e-01 5.68000786e-02 3.30796242e-02 -1.12063634e+00
-5.97871959e-01 -8.80301073e-02 -4.65274453e-01 -9.23000574e-01
-7.75710866e-02 -7.27856100e-01 9.63481218e-02 -1.12984455e+00
4.82077114e-02 -5.83428562e-01 -5.06607294e-01 5.07418588e-02
8.86903629e-02 3.76622319e-01 -2.68859547e-02 2.50537127e-01
-7.77852163e-02 1.56030715e-01 1.49819982e+00 -8.84229168e-02
-2.24256575e-01 -3.21263750e-03 -1.51358098e-01 6.08323634e-01
6.75514340e-01 -5.85298538e-01 -2.18582571e-01 -4.74802762e-01
-7.65971560e-03 9.24355865e-01 6.03269458e-01 -1.42212820e+00
2.49642223e-01 2.72521317e-01 4.86808896e-01 -1.71533167e-01
4.25148576e-01 -1.14096856e+00 7.90766120e-01 9.68029976e-01
-4.07945275e-01 -9.48880389e-02 5.33760674e-02 3.26598048e-01
-2.14706123e-01 -1.85186163e-01 1.25140893e+00 -4.62274462e-01
1.91665038e-01 2.55168527e-01 -3.91172677e-01 1.10398550e-02
6.57042205e-01 -3.72159481e-01 2.10032463e-01 -2.82732785e-01
-1.17521691e+00 -2.47225806e-01 6.33409545e-02 -2.27273643e-01
7.22904623e-01 -1.14307809e+00 -7.83271074e-01 5.25127113e-01
-3.38478297e-01 -1.61300544e-02 5.40164590e-01 1.43379915e+00
-6.90185666e-01 1.89700797e-01 -2.91157216e-01 -1.00388122e+00
-8.07456613e-01 4.23954606e-01 1.04098463e+00 -4.81033772e-01
-1.11421239e+00 2.33977616e-01 3.99726583e-03 -3.63473177e-01
4.39249873e-02 -5.25129199e-01 -2.15711430e-01 -3.78093153e-01
4.20159340e-01 6.12172410e-02 2.77882606e-01 -5.75296104e-01
-2.75031358e-01 8.73106003e-01 3.34592491e-01 8.36044475e-02
1.48549604e+00 -1.60538983e-02 1.42027130e-02 6.00898623e-01
9.76750672e-01 -2.56666124e-01 -1.49136221e+00 -2.08869845e-01
-3.81160349e-01 -1.77472934e-01 4.45376217e-01 -7.13133872e-01
-1.44016039e+00 1.00365782e+00 9.45073485e-01 2.03431010e-01
1.15749967e+00 -3.50168079e-01 7.79168129e-01 1.73247486e-01
2.62757391e-01 -7.85168588e-01 6.03383183e-02 5.47742434e-02
8.21933210e-01 -1.00709224e+00 1.54120281e-01 -2.35870078e-01
-2.48223886e-01 1.32358861e+00 6.50691688e-02 -1.40222609e-01
8.31927359e-01 4.97179329e-01 3.19634408e-01 -2.69334137e-01
-4.99068022e-01 4.74913001e-01 6.30365461e-02 4.95319158e-01
7.35118091e-01 -3.62651981e-02 -2.13931605e-01 2.87046790e-01
3.80657852e-01 5.69025636e-01 6.07562840e-01 8.97170901e-01
-9.68391374e-02 -5.88472307e-01 -3.90419364e-01 3.47951204e-01
-1.00202990e+00 1.81402475e-01 3.97226423e-01 4.15463924e-01
1.74082488e-01 5.38413227e-01 -1.27961412e-01 2.53745168e-01
4.06523347e-01 8.52798298e-02 7.60933161e-01 -4.29425865e-01
-6.98824465e-01 2.76414514e-01 -1.48488745e-01 -6.55567586e-01
-7.21868038e-01 -5.33680975e-01 -1.29853201e+00 2.50058532e-01
-1.88452095e-01 -8.06352645e-02 7.43625343e-01 7.40424037e-01
3.85522962e-01 8.85317564e-01 5.16309977e-01 -1.07912648e+00
-5.20748377e-01 -8.17222595e-01 -7.04631627e-01 3.20670635e-01
7.93844104e-01 -3.96534234e-01 -1.24632262e-01 1.43899918e-01] | [13.57983112335205, -2.477501392364502] |
d62488aa-f3d3-4e66-9772-9badbb9398dd | secure-multiparty-computation-for-synthetic | 2210.07332 | null | https://arxiv.org/abs/2210.07332v2 | https://arxiv.org/pdf/2210.07332v2.pdf | Secure Multiparty Computation for Synthetic Data Generation from Distributed Data | Legal and ethical restrictions on accessing relevant data inhibit data science research in critical domains such as health, finance, and education. Synthetic data generation algorithms with privacy guarantees are emerging as a paradigm to break this data logjam. Existing approaches, however, assume that the data holders supply their raw data to a trusted curator, who uses it as fuel for synthetic data generation. This severely limits the applicability, as much of the valuable data in the world is locked up in silos, controlled by entities who cannot show their data to each other or a central aggregator without raising privacy concerns. To overcome this roadblock, we propose the first solution in which data holders only share encrypted data for differentially private synthetic data generation. Data holders send shares to servers who perform Secure Multiparty Computation (MPC) computations while the original data stays encrypted. We instantiate this idea in an MPC protocol for the Multiplicative Weights with Exponential Mechanism (MWEM) algorithm to generate synthetic data based on real data originating from many data holders without reliance on a single point of failure. | ['Martine De Cock', 'Rafael T. de Sousa Jr.', 'Anderson Nascimento', 'Sikha Pentyala', 'Mayana Pereira'] | 2022-10-13 | null | null | null | null | ['synthetic-data-generation', 'synthetic-data-generation'] | ['medical', 'miscellaneous'] | [-1.05070714e-02 6.12635016e-01 -1.06419139e-01 -3.59400064e-01
-7.37172544e-01 -1.39919031e+00 6.30092323e-01 9.02992487e-01
-7.73415387e-01 1.08864236e+00 -1.05812913e-02 -3.31953079e-01
1.05168559e-01 -1.36863279e+00 -7.48156428e-01 -9.51507449e-01
2.11799085e-01 4.75596011e-01 1.74103796e-01 -1.46038398e-01
2.76687872e-02 4.53546911e-01 -1.36622036e+00 4.51491117e-01
8.52682292e-01 7.05006301e-01 -5.45390666e-01 3.83773744e-01
-7.52451941e-02 4.34465200e-01 -6.91635191e-01 -9.85654593e-01
9.23098683e-01 8.03473517e-02 -7.34412134e-01 -6.47545099e-01
-5.64373471e-02 -5.39946973e-01 1.73502982e-01 9.85284567e-01
4.69774753e-01 -5.95319331e-01 9.42575857e-02 -2.11733985e+00
-3.30615163e-01 8.18519056e-01 -2.15993017e-01 -5.66582620e-01
1.75743133e-01 2.60103345e-01 6.42920673e-01 -3.79257560e-01
1.07268083e+00 5.24969459e-01 3.95308882e-01 6.31325662e-01
-1.40768874e+00 -7.95625329e-01 -6.50653720e-01 -2.92478949e-01
-1.33431089e+00 -4.77037877e-01 5.04918993e-01 -3.81138921e-01
1.10777795e-01 8.74155641e-01 4.43927139e-01 1.04889297e+00
4.53744940e-02 5.31212509e-01 1.34484196e+00 -2.94809133e-01
6.79509997e-01 9.47971880e-01 2.01578587e-01 -1.03823580e-01
1.21055090e+00 -1.07156567e-01 -7.78299451e-01 -1.09390259e+00
1.50098860e-01 1.58633082e-03 -2.94774383e-01 -7.42602825e-01
-1.05746543e+00 8.40528011e-01 -1.40616000e-01 1.53265104e-01
-5.72422624e-01 4.95788492e-02 3.67613763e-01 8.81096065e-01
5.44394553e-01 2.04008088e-01 -6.85525000e-01 5.92239648e-02
-8.98233831e-01 8.64054263e-01 1.06028342e+00 1.01099837e+00
5.66261947e-01 -6.44189358e-01 2.12703153e-01 -2.37022489e-01
4.20990139e-01 5.68397760e-01 2.03261107e-01 -7.31542110e-01
5.65475762e-01 8.03393960e-01 6.81931198e-01 -9.23694789e-01
2.48052463e-01 3.30998689e-01 -9.06729639e-01 3.09267581e-01
6.88150942e-01 -2.73391217e-01 -2.05246985e-01 1.78486383e+00
1.08085084e+00 -5.28238297e-01 5.02525806e-01 7.69159973e-01
3.53093594e-01 3.44362885e-01 -4.92509492e-02 -1.72439963e-01
1.49962783e+00 -5.83903454e-02 -7.80337453e-01 8.39101791e-01
4.61046606e-01 -5.12438476e-01 5.62029421e-01 5.50353765e-01
-1.19739318e+00 2.18416825e-01 -7.23094344e-01 -2.67992496e-01
-8.98221612e-01 -5.69297850e-01 6.15769267e-01 1.15758240e+00
-1.04832542e+00 3.80522907e-01 -6.81578875e-01 -1.85241513e-02
5.54136157e-01 5.85741282e-01 -8.87646616e-01 3.18813562e-01
-1.53498709e+00 1.62489057e-01 3.04038942e-01 -3.28422993e-01
-5.98684251e-01 -1.07626784e+00 -5.11167467e-01 -6.24346584e-02
-4.40848693e-02 -8.28281701e-01 8.84129941e-01 -4.53636527e-01
-1.10040641e+00 1.02447951e+00 3.89144152e-01 -7.95343280e-01
1.23039579e+00 2.89461434e-01 -4.55324203e-02 8.45569521e-02
3.90099995e-02 1.29335355e-02 6.26483142e-01 -1.45527375e+00
-7.27678597e-01 -7.70537674e-01 -6.15610089e-03 -4.93808538e-01
-3.76596659e-01 2.39721328e-01 5.14688849e-01 -2.63541669e-01
6.89275097e-03 -8.15772593e-01 -3.92895728e-01 6.48094565e-02
-4.31232333e-01 2.23107841e-02 1.01323950e+00 -2.35379234e-01
1.01574767e+00 -1.85658872e+00 -2.78648555e-01 6.22095585e-01
4.57705736e-01 5.35964608e-01 3.41514558e-01 9.64067280e-01
1.13855205e-01 6.59971774e-01 -1.88008025e-01 -4.06906724e-01
3.83445710e-01 -8.95637870e-02 -8.93736184e-01 4.84497845e-01
-1.79697514e-01 6.89685106e-01 -8.76028419e-01 -2.85702825e-01
-1.59196228e-01 4.01587278e-01 -4.82195050e-01 2.73892879e-01
-2.28593141e-01 4.93373603e-01 -6.31779432e-01 3.88725877e-01
1.40525067e+00 -1.84918225e-01 3.84676337e-01 2.88784713e-01
-3.44409823e-01 1.04978181e-01 -1.53158128e+00 1.28900874e+00
-4.24604043e-02 -3.80314998e-02 6.23113990e-01 -3.35769594e-01
5.24518967e-01 9.15494442e-01 7.60509133e-01 -4.94177565e-02
5.23365326e-02 2.97419012e-01 -6.08930886e-01 -3.16295922e-01
6.33522451e-01 -1.84491053e-02 -5.44275343e-01 1.30328500e+00
-3.48649263e-01 2.01126505e-02 -2.51412302e-01 4.68672335e-01
1.18791032e+00 -4.86122146e-02 1.25707477e-01 -1.17169909e-01
4.90682632e-01 8.66179243e-02 5.32148540e-01 4.52871978e-01
-1.24531716e-01 2.79675364e-01 8.64696741e-01 -7.91567445e-01
-1.30595160e+00 -1.07441962e+00 4.21048282e-03 7.48495042e-01
-7.26779476e-02 -6.32043481e-01 -9.76660728e-01 -7.79805541e-01
3.94088954e-01 3.98729533e-01 -4.14561689e-01 1.75944805e-01
-2.68691510e-01 -4.87632155e-01 8.26428652e-01 -2.56989121e-01
3.90725553e-01 -8.54684472e-01 -9.77375567e-01 2.20727578e-01
-2.98362702e-01 -9.81627822e-01 -1.74646214e-01 -2.17019647e-01
-4.27377492e-01 -1.11234963e+00 -3.33286554e-01 1.53973922e-01
1.00614035e+00 8.83538127e-02 6.46815300e-01 -7.57535324e-02
-2.37334713e-01 1.64185315e-01 -1.75335944e-01 -9.72724378e-01
-8.18409204e-01 1.06595308e-01 2.79879242e-01 4.52484339e-01
3.51042777e-01 -7.70513356e-01 -9.77740467e-01 -8.86662677e-02
-1.57746780e+00 -7.21899718e-02 -3.00262799e-03 3.08807880e-01
3.52313161e-01 -6.56229630e-02 6.26508832e-01 -1.42376029e+00
7.98148513e-01 -7.74041295e-01 -1.00741065e+00 2.82339871e-01
-8.97069514e-01 2.53352933e-02 1.02709198e+00 -1.92905888e-01
-6.95922792e-01 1.37505397e-01 4.21023935e-01 -1.10088706e-01
-3.24816145e-02 -4.83900942e-02 -5.07410645e-01 1.01323174e-02
4.48515177e-01 3.39641243e-01 1.53012618e-01 -3.96794736e-01
5.28666437e-01 9.81800318e-01 1.09338626e-01 -7.76473939e-01
1.02433693e+00 8.66305590e-01 2.31289431e-01 -1.28498793e-01
-1.12562582e-01 -3.49416994e-02 -4.09680933e-01 2.59376109e-01
6.03711367e-01 -7.64476061e-01 -1.61016536e+00 7.23715961e-01
-1.30326295e+00 1.80794314e-01 -6.98840201e-01 1.19707324e-01
-2.72038877e-01 2.08868638e-01 -2.51518041e-01 -9.51993883e-01
-6.62362397e-01 -9.02200580e-01 8.50514889e-01 6.66605979e-02
-1.51214376e-01 -5.29312193e-01 1.12549245e-01 6.91870272e-01
6.92541599e-01 8.56674254e-01 5.46492159e-01 -1.02648842e+00
-1.01787114e+00 -7.01325178e-01 3.70468050e-02 1.54026404e-01
4.99117486e-02 1.15938805e-01 -1.13583577e+00 -4.00666714e-01
2.17666477e-01 -2.94370770e-01 -6.12161122e-02 -5.23958564e-01
1.03031790e+00 -1.04647136e+00 -8.44682381e-02 3.70464057e-01
1.48662817e+00 -3.69350225e-01 4.26820844e-01 -1.40721630e-02
2.09954619e-01 9.67258871e-01 2.19018951e-01 8.30625653e-01
8.54658127e-01 2.25131229e-01 2.86674827e-01 1.47696301e-01
6.99902892e-01 -4.92419302e-01 1.27468213e-01 3.73199791e-01
1.56550482e-02 -1.78744435e-01 -5.96994579e-01 5.95739186e-01
-1.86073089e+00 -9.24922347e-01 -4.61970329e-01 2.53593516e+00
1.33910370e+00 -3.06249350e-01 1.23699278e-01 1.27286971e-01
2.63045698e-01 -1.15637094e-01 -2.70748347e-01 -3.67838860e-01
7.50989690e-02 4.46141809e-01 1.08049226e+00 2.73174316e-01
-6.80262923e-01 5.69758117e-01 4.76490068e+00 3.08741450e-01
-1.04861212e+00 3.81309927e-01 5.80695212e-01 -3.87425385e-02
-9.48460937e-01 4.27518964e-01 -5.63674033e-01 9.18390155e-01
1.28354716e+00 -1.02629077e+00 3.99010301e-01 7.28994846e-01
3.05517733e-01 -9.64819118e-02 -1.18989444e+00 4.79010493e-01
-6.11486614e-01 -1.48597729e+00 -1.13078319e-01 6.04265213e-01
6.99550629e-01 -2.30498508e-01 1.25168294e-01 -3.50427091e-01
8.54727447e-01 -7.06044972e-01 5.63178360e-01 7.97399342e-01
5.85643232e-01 -9.03899848e-01 5.82305729e-01 7.09404647e-01
-5.85655451e-01 1.43676713e-01 -2.42901862e-01 4.32278454e-01
-9.33978558e-02 8.18553984e-01 -9.06682074e-01 9.22032416e-01
4.24479634e-01 -5.75080395e-01 -1.82112619e-01 3.77046496e-01
-5.01120538e-02 4.13179606e-01 -5.46604693e-01 1.96417570e-01
-6.32989854e-02 -4.03709441e-01 2.64264613e-01 6.24620736e-01
2.28769034e-01 2.81012893e-01 -3.20936203e-01 1.02075648e+00
-5.96567929e-01 1.71450093e-01 -9.30992961e-01 -1.90493874e-02
8.31095755e-01 1.24666917e+00 -3.95650595e-01 -2.98614562e-01
-5.07182330e-02 8.23264182e-01 1.83591414e-02 1.63118586e-01
-3.52090031e-01 -3.63111615e-01 8.31387877e-01 5.42339265e-01
-6.40529916e-02 -1.20476902e-01 -3.25387329e-01 -1.16166639e+00
4.02753383e-01 -1.01517749e+00 5.57293355e-01 -2.64555633e-01
-1.27170968e+00 1.53407931e-01 -1.26753494e-01 -1.14409578e+00
-1.25644103e-01 1.38436109e-01 -3.95023048e-01 9.76678610e-01
-1.44149280e+00 -1.43760896e+00 7.58354040e-03 9.13689554e-01
-6.77689672e-01 9.14410353e-02 1.19378138e+00 1.62183926e-01
1.53292120e-02 8.63976181e-01 2.13886559e-01 6.85936138e-02
7.84177542e-01 -1.18772018e+00 4.35843825e-01 8.18299532e-01
-1.79101676e-01 1.20594954e+00 7.19851911e-01 -6.67144477e-01
-1.80000949e+00 -1.08686244e+00 1.61037767e+00 -8.50580275e-01
4.93369609e-01 -1.05424988e+00 -8.77539635e-01 5.88109553e-01
1.52337044e-01 4.66133982e-01 1.09062362e+00 -6.23959541e-01
-5.95583737e-01 -5.98410904e-01 -2.04354262e+00 3.00710559e-01
3.48534882e-01 -6.63414598e-01 -4.00964394e-02 4.36264962e-01
9.20817792e-01 -7.65501335e-02 -1.08658254e+00 1.19292461e-04
7.54075170e-01 -9.27016973e-01 4.84263718e-01 -9.76880372e-01
9.36254635e-02 -5.73054254e-01 -1.11287527e-01 -7.66619742e-01
8.83398592e-01 -1.57214785e+00 -1.28659680e-01 1.48756528e+00
2.30515629e-01 -1.24372244e+00 1.01815498e+00 1.72595727e+00
9.09449160e-01 -2.10837945e-01 -1.21959770e+00 -3.98911208e-01
3.15848500e-01 -1.68004751e-01 1.60288215e+00 1.35548735e+00
5.04892468e-02 -5.23012578e-01 -2.19544649e-01 4.28447574e-01
1.16433287e+00 2.91699737e-01 1.45258713e+00 -1.21638572e+00
2.42244750e-01 3.52365285e-01 -2.17323795e-01 4.44983989e-02
-1.58434436e-01 -8.57995510e-01 -4.96301740e-01 -1.15079331e+00
6.96739601e-03 -8.51257324e-01 -1.68116808e-01 5.88520586e-01
1.61728740e-01 -6.91240951e-02 2.30474621e-01 1.17883354e-01
-2.91830987e-01 2.94481814e-01 8.54451895e-01 2.73512989e-01
-1.87388752e-02 2.12637112e-01 -1.00202465e+00 2.73678482e-01
6.27546847e-01 -9.83888865e-01 -5.63163340e-01 1.98480248e-01
7.70566940e-01 2.38621414e-01 5.07578850e-01 -7.00968385e-01
7.58550823e-01 -5.05866885e-01 -9.20947045e-02 -4.33162570e-01
-3.84625763e-01 -1.47402716e+00 8.77470195e-01 5.29112875e-01
-4.86547083e-01 2.93936227e-02 -4.77604896e-01 3.01189125e-01
-3.02243270e-02 5.76994382e-02 4.91571128e-01 -1.46101445e-01
3.56950760e-01 5.33780575e-01 -1.75922558e-01 -1.11873128e-01
1.71475744e+00 4.80021574e-02 -6.13145471e-01 -3.32288176e-01
-6.29482031e-01 3.54917318e-01 9.69028771e-01 9.22618806e-02
2.98906207e-01 -1.05923259e+00 -9.53861773e-01 7.12444127e-01
8.26474801e-02 3.77155453e-01 1.55742109e-01 5.35779536e-01
-3.70529324e-01 1.92097813e-01 -1.39947399e-01 2.94061787e-02
-1.42953849e+00 6.84131920e-01 -4.56845574e-02 -2.43700668e-01
-1.85910419e-01 3.37038338e-01 -4.18234795e-01 -6.58132076e-01
-2.40259208e-02 -3.58036131e-01 6.24050379e-01 4.18360680e-01
6.64352894e-01 2.70447761e-01 1.86467052e-01 -3.06682527e-01
-1.45762578e-01 -2.98606336e-01 -7.74442218e-03 -3.59359801e-01
1.46663678e+00 -1.52080238e-01 -7.43082643e-01 8.56116861e-02
1.04846883e+00 6.61785901e-01 -6.03186727e-01 -1.01552725e-01
-3.06474306e-02 -7.71653950e-01 -6.11240268e-01 -4.51265067e-01
-8.72031868e-01 4.69343454e-01 3.14140648e-01 6.59623206e-01
9.48580265e-01 -2.72175729e-01 9.58503306e-01 6.31341785e-02
9.79740858e-01 -9.91766274e-01 -6.95749521e-01 -3.77664089e-01
3.94681841e-01 -1.08889365e+00 1.47960857e-01 -1.66483194e-01
-4.87222672e-01 8.67642939e-01 1.23968579e-01 2.40630463e-01
7.13927627e-01 4.86705989e-01 7.82750323e-02 -1.41384333e-01
-1.09337711e+00 5.06157160e-01 -6.40759110e-01 5.62529743e-01
-1.50021791e-01 2.81858414e-01 -8.14469218e-01 9.23912883e-01
-2.99842715e-01 3.55376631e-01 9.78511095e-01 1.27585638e+00
1.58043936e-01 -2.05359745e+00 -2.87450254e-01 -3.31112221e-02
-9.77439284e-01 4.78688162e-03 -6.06258094e-01 6.52143419e-01
4.44778740e-01 7.37352014e-01 -2.32643247e-01 9.89283696e-02
2.51010895e-01 2.23725826e-01 -2.60517925e-01 -2.44842634e-01
-1.25070488e+00 -7.20528245e-01 -2.20746502e-01 -4.67425764e-01
-2.73048222e-01 -5.89474678e-01 -1.27255976e+00 -9.32901919e-01
-6.24943152e-02 7.15253174e-01 1.13085401e+00 1.42424420e-01
8.37822378e-01 -4.31155294e-01 1.44857717e+00 3.27420622e-01
-6.79716527e-01 -2.81052381e-01 -5.77284515e-01 6.19690597e-01
4.70581561e-01 4.40655023e-01 -2.45223895e-01 4.24810141e-01] | [5.8922438621521, 6.665071487426758] |
6f36344b-d70e-4c6a-9855-170cf4aa7cb7 | keyphrase-extraction-using-neighborhood | 2111.07198 | null | https://arxiv.org/abs/2111.07198v1 | https://arxiv.org/pdf/2111.07198v1.pdf | Keyphrase Extraction Using Neighborhood Knowledge Based on Word Embeddings | Keyphrase extraction is the task of finding several interesting phrases in a text document, which provide a list of the main topics within the document. Most existing graph-based models use co-occurrence links as cohesion indicators to model the relationship of syntactic elements. However, a word may have different forms of expression within the document, and may have several synonyms as well. Simply using co-occurrence information cannot capture this information. In this paper, we enhance the graph-based ranking model by leveraging word embeddings as background knowledge to add semantic information to the inter-word graph. Our approach is evaluated on established benchmark datasets and empirical results show that the word embedding neighborhood information improves the model performance. | ['Mohammed J. Zaki', 'Yuchen Liang'] | 2021-11-13 | null | null | null | null | ['keyphrase-extraction'] | ['natural-language-processing'] | [-3.45492095e-01 -1.53044194e-01 -8.11228454e-01 -2.81139910e-02
-4.05694544e-02 -5.77815950e-01 8.03353727e-01 1.07401359e+00
-4.83140439e-01 4.73225445e-01 1.00848448e+00 -1.61525488e-01
-4.05551910e-01 -1.12836528e+00 -1.75918877e-01 -3.46661359e-01
-1.51733503e-01 4.43155169e-02 5.63650846e-01 -5.14262021e-01
5.84919572e-01 2.12189898e-01 -1.01940060e+00 2.51869142e-01
6.60676420e-01 6.09207690e-01 1.79780617e-01 1.88800618e-01
-9.82278109e-01 7.21123695e-01 -7.33651698e-01 -3.41010779e-01
-2.16706753e-01 -1.03906199e-01 -6.10999346e-01 -2.03787923e-01
1.24805346e-01 1.37476847e-04 -5.97687960e-01 1.32293487e+00
4.16347124e-02 1.74051046e-01 6.27482533e-01 -1.21479821e+00
-8.72480810e-01 9.09982264e-01 -7.48867154e-01 4.77713257e-01
5.05825520e-01 -6.28330886e-01 1.96782458e+00 -1.03693116e+00
5.81727982e-01 1.27470100e+00 4.05849874e-01 -7.50252530e-02
-7.59010315e-01 -5.22652805e-01 4.73811358e-01 4.56918091e-01
-1.46711981e+00 4.50344950e-01 1.04568315e+00 -1.47643909e-01
1.17484665e+00 1.41169295e-01 8.87921751e-01 7.07927704e-01
4.13471162e-01 4.94224876e-01 7.45975196e-01 -4.95948315e-01
-1.78030238e-01 8.79229419e-03 8.57909799e-01 6.56661332e-01
9.01209652e-01 -5.18082142e-01 -5.96646667e-01 -4.72300291e-01
3.38478386e-01 4.03954417e-01 -3.81075501e-01 -7.28787184e-02
-1.09830606e+00 1.03732204e+00 5.62610030e-01 7.84025490e-01
-2.62022167e-01 3.03830840e-02 4.62801069e-01 5.58344461e-02
5.84029198e-01 7.99908638e-01 -2.43724123e-01 -8.95935763e-03
-5.76856017e-01 2.45537609e-01 7.25846887e-01 7.76270568e-01
9.55349803e-01 -5.30158162e-01 -1.91685021e-01 8.18669856e-01
7.05871880e-01 6.12434149e-02 8.10333252e-01 3.37496810e-02
5.27828991e-01 1.25300014e+00 -2.23860621e-01 -1.72523630e+00
-3.65145355e-01 -5.74597597e-01 -4.46292222e-01 -5.06801367e-01
-3.36236060e-01 -5.75919822e-02 -7.98606634e-01 1.24846125e+00
2.45621368e-01 3.93529862e-01 -3.14051993e-02 6.13679051e-01
1.28297722e+00 1.01103377e+00 3.67403142e-02 -2.23750174e-01
1.50653160e+00 -9.62642372e-01 -1.14783061e+00 -2.08094150e-01
7.07365036e-01 -9.75294650e-01 9.88533974e-01 -9.86651983e-03
-3.20332378e-01 -2.08164424e-01 -1.14723337e+00 -4.83817533e-02
-9.07344759e-01 -2.67034113e-01 5.94790161e-01 2.75053114e-01
-6.28767312e-01 3.78351718e-01 -3.40531617e-01 -4.70536679e-01
9.34950262e-02 9.56119969e-03 -3.16291392e-01 -2.07506239e-01
-1.64877486e+00 9.07257855e-01 9.16350484e-01 -3.23826373e-01
8.55031610e-02 -6.44207239e-01 -8.95577431e-01 2.36293852e-01
5.60119390e-01 -4.64774758e-01 4.50059175e-01 -2.46773332e-01
-7.09140539e-01 4.54591721e-01 -3.66327226e-01 -1.46560535e-01
-5.63819945e-01 -2.89038509e-01 -9.06718314e-01 1.61729425e-01
1.15182288e-01 1.97413087e-01 5.43908000e-01 -1.11937547e+00
-7.96886623e-01 -9.56189856e-02 4.06764835e-01 3.47253919e-01
-1.04535794e+00 1.54837966e-01 -8.22156131e-01 -9.87759888e-01
1.15447298e-01 -5.54180443e-01 -6.72628880e-02 -4.07407343e-01
-5.25316894e-01 -6.49998665e-01 9.63242233e-01 -3.96695048e-01
2.21179652e+00 -1.99517477e+00 5.14681451e-02 5.79122186e-01
5.90174139e-01 7.89065361e-02 -1.09647661e-01 1.04101408e+00
6.31583035e-02 5.68889260e-01 4.74604368e-02 2.64684767e-01
-8.05963874e-02 3.70205253e-01 -4.14110601e-01 -1.23383574e-01
-2.10630521e-02 9.93985891e-01 -1.11502635e+00 -7.01033473e-01
-1.44156620e-01 2.99154520e-01 -2.73467422e-01 3.84381860e-02
-1.84301019e-01 -4.59770411e-01 -8.81458223e-01 2.85767883e-01
3.70611072e-01 -3.09633702e-01 2.61202872e-01 -5.57629585e-01
2.50941813e-01 5.18891215e-01 -1.13245010e+00 1.44855142e+00
-3.85969609e-01 5.54480851e-01 -6.37237906e-01 -9.26798344e-01
8.32030416e-01 -8.52778275e-03 6.20035291e-01 -4.15221751e-01
2.02724803e-02 -2.70049158e-03 1.73976526e-01 -4.51967925e-01
8.18370521e-01 6.77958205e-02 -1.67268962e-01 4.04231608e-01
8.99124965e-02 1.17709659e-01 5.54652870e-01 7.74580419e-01
1.24745035e+00 -5.04384041e-01 6.74629867e-01 -5.22893965e-01
5.18905163e-01 1.69106591e-02 4.95559305e-01 3.89474511e-01
3.18986624e-01 2.26236314e-01 7.45161891e-01 -1.37511209e-01
-6.14792466e-01 -9.15732801e-01 -7.43180932e-03 9.32403326e-01
4.08037037e-01 -1.63387132e+00 -1.61072075e-01 -1.01088023e+00
2.07556441e-01 5.67443967e-01 -7.95840740e-01 -1.68992117e-01
-4.39964265e-01 -7.23964989e-01 1.58285379e-01 4.56865877e-01
1.11851156e-01 -5.30041575e-01 9.00581330e-02 3.10093433e-01
-1.30421296e-01 -1.00258863e+00 -6.74695611e-01 -2.33180225e-02
-7.33355522e-01 -1.15498459e+00 -4.88391727e-01 -9.34000552e-01
7.77174652e-01 6.31380916e-01 1.14709878e+00 4.85839188e-01
-1.84310988e-01 3.70910496e-01 -8.60795677e-01 -3.04452807e-01
7.72574395e-02 1.59242570e-01 -6.86771646e-02 -1.75846800e-01
7.56294012e-01 -3.69192272e-01 -4.67638999e-01 6.93644062e-02
-1.11019146e+00 -1.39577255e-01 2.88080037e-01 8.84274721e-01
5.66911936e-01 5.96982300e-01 3.35340708e-01 -9.45833981e-01
1.25546932e+00 -7.00729907e-01 -2.00526997e-01 4.75187719e-01
-1.12602532e+00 3.21546197e-01 4.64455992e-01 -3.83918166e-01
-7.12931752e-01 -6.24995291e-01 1.76212713e-01 1.61401536e-02
4.36855346e-01 1.33386016e+00 -1.53438106e-01 2.32418895e-01
1.64907515e-01 2.01588288e-01 -5.76313317e-01 -4.26802903e-01
7.01888561e-01 5.05745590e-01 -1.29249111e-01 -5.21428943e-01
8.96947384e-01 2.48027131e-01 2.07395598e-01 -8.47730398e-01
-1.07551944e+00 -1.04298151e+00 -4.96582299e-01 -2.39555258e-04
7.21778870e-01 -8.49290013e-01 -2.26799212e-02 -2.46779487e-01
-1.18959188e+00 5.95270336e-01 -1.98053885e-02 5.88648498e-01
4.09715623e-01 5.69275737e-01 -4.07996982e-01 -1.90490574e-01
-3.20853174e-01 -7.11574614e-01 7.60351479e-01 3.27451587e-01
-4.05781657e-01 -1.44270337e+00 3.09684157e-01 -1.48540154e-01
1.65305689e-01 4.63066287e-02 1.33440173e+00 -1.05279171e+00
-2.41552651e-01 -5.09657502e-01 -2.69426674e-01 1.52764663e-01
8.69079709e-01 1.77219585e-01 -2.40435421e-01 -1.28421038e-01
-5.73201656e-01 2.48324543e-01 1.14264190e+00 -8.00248981e-02
9.82536554e-01 -6.19553030e-01 -7.42952585e-01 8.31079334e-02
1.58015192e+00 5.52254990e-02 3.87458682e-01 4.94508654e-01
1.14404166e+00 5.94723940e-01 6.61516547e-01 3.57023269e-01
4.25941855e-01 5.66345215e-01 1.85119554e-01 2.14589193e-01
-8.31193551e-02 -6.82282269e-01 2.29797468e-01 1.50599432e+00
1.77160233e-01 -5.46230078e-01 -9.83236909e-01 5.68415880e-01
-1.79245019e+00 -8.67571831e-01 -4.69493598e-01 1.77115452e+00
8.59702706e-01 4.15154189e-01 -6.68075979e-02 5.33618405e-02
7.93098569e-01 7.18728065e-01 1.05034709e-01 -3.26666176e-01
-9.53224376e-02 2.27423653e-01 5.13155043e-01 5.43705225e-01
-7.75721014e-01 1.13176227e+00 5.97039032e+00 1.00747097e+00
-7.69644201e-01 -8.40703994e-02 -7.05569014e-02 2.27481842e-01
-8.31631720e-01 1.93045169e-01 -1.03784907e+00 4.67327386e-01
5.33892870e-01 -9.22689319e-01 -2.15053558e-01 6.35259390e-01
-7.36903176e-02 4.62299176e-02 -7.86016285e-01 8.16342413e-01
5.33453226e-01 -1.41853786e+00 6.34512484e-01 -4.32939790e-02
6.40468776e-01 -1.59600466e-01 -2.20885590e-01 7.73474202e-02
3.47823113e-01 -7.74704933e-01 -2.41628774e-02 3.76100957e-01
1.18723452e-01 -9.28502798e-01 8.24158967e-01 2.51263138e-02
-1.70673835e+00 8.93575326e-02 -5.08214593e-01 -4.88009397e-03
1.01161122e-01 8.62763941e-01 -7.61532485e-01 8.78042281e-01
5.29677033e-01 1.15625644e+00 -8.23504388e-01 9.22576249e-01
-7.85954416e-01 7.57198274e-01 -1.25882968e-01 -8.38226199e-01
5.23486793e-01 -2.33224630e-01 6.08940184e-01 1.49588180e+00
2.43667230e-01 3.54398191e-02 2.42613941e-01 4.36859012e-01
-3.47244531e-01 7.56408513e-01 -6.46814287e-01 -4.16788518e-01
6.09761417e-01 1.40390813e+00 -9.36618447e-01 -4.02598739e-01
-8.00940156e-01 7.62375951e-01 2.49328449e-01 1.84032932e-01
-5.82708120e-01 -8.14184487e-01 8.50651622e-01 1.47947863e-01
2.08104879e-01 -5.25547683e-01 2.06432357e-01 -1.13147461e+00
1.40164047e-01 -4.31639552e-01 6.90636516e-01 -6.80304050e-01
-1.54852271e+00 3.87904733e-01 2.37491608e-01 -1.17178047e+00
8.15139040e-02 -6.43581331e-01 -8.71723056e-01 6.87237620e-01
-1.64164066e+00 -9.68429685e-01 -1.75583601e-01 3.29708904e-01
4.81985599e-01 -2.47369871e-01 8.25399280e-01 1.24653652e-01
-4.21357542e-01 3.25930059e-01 1.53898910e-01 3.92524242e-01
5.69696963e-01 -1.40481627e+00 3.57616752e-01 8.48419726e-01
8.88383269e-01 1.10941815e+00 6.43756747e-01 -8.84632945e-01
-1.29460096e+00 -9.73201931e-01 1.31497228e+00 -3.05422366e-01
1.31928420e+00 -2.74846196e-01 -1.12293684e+00 4.99936074e-01
4.92872059e-01 9.20426250e-02 1.12965930e+00 4.53258932e-01
-5.90964615e-01 2.72856168e-02 -4.27212775e-01 7.24742174e-01
9.97566342e-01 -5.48042595e-01 -1.36974275e+00 3.38595748e-01
1.24877357e+00 1.04601011e-01 -8.05756688e-01 2.44778216e-01
3.22148412e-01 -2.32582837e-01 9.64775920e-01 -8.03483129e-01
5.91414630e-01 -5.03480613e-01 -1.79567903e-01 -1.66476238e+00
-5.47826529e-01 -2.93678701e-01 -5.29395700e-01 1.55637825e+00
5.77059865e-01 -5.65106988e-01 5.88522553e-01 5.94693162e-02
2.43395954e-01 -8.20981324e-01 -6.48468792e-01 -9.12260234e-01
2.63550524e-02 -2.62904972e-01 7.07482100e-01 1.17975557e+00
6.30952418e-01 7.45743454e-01 -1.22951122e-03 1.24092825e-01
2.31481254e-01 1.43490642e-01 3.46772283e-01 -1.51888525e+00
-1.61229968e-02 -6.52165473e-01 -7.20847547e-01 -1.01978445e+00
3.92452776e-01 -1.26606476e+00 -4.65741605e-01 -2.12306762e+00
3.08420330e-01 -2.01716557e-01 -8.93334031e-01 2.56547481e-01
-6.17701232e-01 -3.09139431e-01 -5.64892739e-02 2.45188400e-01
-6.11337423e-01 5.59762955e-01 1.06802666e+00 -4.67732072e-01
4.53846902e-02 -4.32492733e-01 -8.13250780e-01 7.73394406e-01
6.83268249e-01 -7.75939107e-01 -6.76049650e-01 -4.70146984e-01
7.66002953e-01 -6.19444072e-01 -5.30200824e-02 -5.32900155e-01
4.94469285e-01 -3.53227198e-01 3.96224409e-02 -4.23731863e-01
8.44825283e-02 -1.11037719e+00 -4.07883853e-01 3.68992388e-01
-4.52679813e-01 5.19236326e-01 5.35922647e-02 8.30136061e-01
-5.38371444e-01 -4.36029404e-01 -6.51469547e-03 -2.80485433e-02
-1.10313177e+00 3.93616319e-01 -1.59145549e-01 1.77859351e-01
7.30922937e-01 7.46799260e-02 -4.24873084e-01 -1.99640885e-01
-2.45787710e-01 3.02619338e-01 1.93185806e-01 9.67780769e-01
1.00345814e+00 -1.73318541e+00 -5.69620252e-01 -3.53228807e-01
7.81791151e-01 -4.13936824e-01 -3.08200061e-01 2.78128296e-01
-3.38233352e-01 3.30228955e-01 1.93214491e-01 -4.05873097e-02
-1.68520474e+00 5.96120894e-01 -1.44790903e-01 -5.22645950e-01
-6.66337013e-01 6.99393272e-01 -3.53319757e-02 4.61997315e-02
2.02660654e-02 -4.17202294e-01 -9.74560142e-01 5.48588097e-01
6.44083798e-01 1.27467528e-01 -1.23422891e-01 -7.00539947e-01
-6.31816089e-01 8.36087108e-01 -2.77609736e-01 -1.12407871e-01
1.17496729e+00 -5.63774109e-02 -7.51509428e-01 5.80997765e-01
1.55636013e+00 3.67947608e-01 -1.16581514e-01 -7.07998276e-01
5.50249994e-01 -6.89524353e-01 4.28420603e-01 -3.23486060e-01
-9.56348062e-01 6.68885469e-01 -4.89640124e-02 4.13126498e-01
8.21897030e-01 2.24828839e-01 8.75222027e-01 5.98603129e-01
1.95759371e-01 -1.20957446e+00 2.98982978e-01 6.73350275e-01
8.30528617e-01 -8.86150062e-01 5.50914824e-01 -7.74985790e-01
-4.42856342e-01 1.39948797e+00 5.89792967e-01 -1.81128141e-02
1.15557325e+00 3.33632976e-02 -1.25054777e-01 -5.92295229e-01
-6.72905803e-01 -5.62875390e-01 8.79174650e-01 2.16568068e-01
6.50958955e-01 -8.30301344e-02 -1.01010144e+00 6.38049603e-01
-1.17861331e-01 -6.14071488e-01 5.00740111e-01 9.96636212e-01
-7.82898247e-01 -1.44098175e+00 7.92770833e-02 6.59230888e-01
-5.50953269e-01 -5.02599180e-01 -6.84712589e-01 5.81441522e-01
1.08257189e-01 1.07720160e+00 9.64744389e-02 -5.72943985e-01
1.86704755e-01 1.82227716e-02 -1.12810181e-02 -1.05806565e+00
-4.21069443e-01 -4.44654934e-02 8.77920538e-02 -1.75156847e-01
-4.69514132e-01 -1.52242854e-01 -1.59471798e+00 -1.34081066e-01
-5.20083010e-01 5.53718328e-01 3.98278415e-01 9.52892661e-01
3.11756432e-01 7.85075665e-01 7.85296500e-01 2.62915164e-01
1.22666312e-02 -9.12848651e-01 -6.20118916e-01 5.31169355e-01
8.28817636e-02 -6.52918160e-01 -3.10758919e-01 -2.37298265e-01] | [10.395464897155762, 8.429976463317871] |
9843b515-3a52-4459-87a2-6025192596e9 | wipin-operation-free-person-identification | 1810.04106 | null | https://arxiv.org/abs/1810.04106v2 | https://arxiv.org/pdf/1810.04106v2.pdf | WiPIN: Operation-free Passive Person Identification Using Wi-Fi Signals | Wi-Fi signals-based person identification attracts increasing attention in the booming Internet-of-Things era mainly due to its pervasiveness and passiveness. Most previous work applies gaits extracted from WiFi distortions caused by the person walking to achieve the identification. However, to extract useful gait, a person must walk along a pre-defined path for several meters, which requires user high collaboration and increases identification time overhead, thus limiting use scenarios. Moreover, gait based work has severe shortcoming in identification performance, especially when the user volume is large. In order to eliminate the above limitations, in this paper, we present an operation-free person identification system, namely WiPIN, that requires least user collaboration and achieves good performance. WiPIN is based on an entirely new insight that Wi-Fi signals would carry person body information when propagating through the body, which is potentially discriminated for person identification. Then we demonstrate the feasibility on commodity off-the-shelf Wi-Fi devices by well-designed signal pre-processing, feature extraction, and identity matching algorithms. Results show that WiPIN achieves 92% identification accuracy over 30 users, high robustness to various experimental settings, and low identifying time overhead, i.e., less than 300ms. | ['Feng Lin', 'Fei Wang', 'Kui Ren', 'Jinsong Han'] | 2018-10-06 | null | null | null | null | ['person-identification'] | ['computer-vision'] | [ 2.22358599e-01 -4.20851141e-01 -9.10027996e-02 -1.05119124e-01
-2.69298434e-01 -4.20935452e-01 -6.41495064e-02 -2.71266103e-01
-3.74553084e-01 7.50356257e-01 -1.68080069e-02 2.60940474e-02
-4.52465892e-01 -9.99205410e-01 -9.84286815e-02 -5.87535024e-01
-8.16118792e-02 1.10885061e-01 1.88909590e-01 4.53129113e-02
-1.91757873e-01 1.74757913e-01 -1.34179223e+00 -4.06175584e-01
8.93424392e-01 1.17437184e+00 -9.31974575e-02 4.23135191e-01
2.48562008e-01 8.01957175e-02 -7.94843376e-01 -5.26028335e-01
2.14005813e-01 -3.20020139e-01 -2.40601107e-01 -3.85709614e-01
-6.08202480e-02 -5.33817053e-01 -6.07498765e-01 7.29748666e-01
9.77981687e-01 -1.03473358e-01 3.84515613e-01 -1.54051113e+00
-2.66374022e-01 4.71150696e-01 -5.62867761e-01 -4.67501730e-02
9.48426187e-01 9.71100181e-02 4.79072064e-01 -2.55231202e-01
9.93166193e-02 8.98785472e-01 1.29979670e+00 4.08820629e-01
-9.34067249e-01 -1.06442857e+00 -3.11113536e-01 3.09453160e-01
-1.94409621e+00 -5.84186733e-01 6.99120343e-01 -3.01382810e-01
5.96233428e-01 5.16211808e-01 7.82227159e-01 1.15115035e+00
-1.73395216e-01 3.31436455e-01 4.89006668e-01 -2.80010462e-01
4.35321666e-02 3.84769663e-02 2.23173842e-01 4.07527208e-01
8.73352408e-01 -3.10673676e-02 -6.15271389e-01 -3.00642282e-01
7.37203240e-01 1.62942618e-01 -5.31319141e-01 4.00741436e-02
-1.19827628e+00 8.79455879e-02 1.51812986e-01 5.36097109e-01
-3.57854337e-01 1.07368931e-01 3.16687882e-01 1.07281141e-01
-2.31658891e-01 9.11723077e-02 6.18460886e-02 -6.80564642e-01
-1.00161886e+00 4.16372977e-02 5.90410292e-01 1.07846308e+00
4.26502317e-01 4.19226140e-02 -2.92274281e-02 6.95722044e-01
4.34127092e-01 1.13358033e+00 4.47150975e-01 -5.60833871e-01
6.60648286e-01 3.18174362e-01 5.11529207e-01 -1.37420595e+00
-6.14791811e-01 -7.09484935e-01 -1.18213928e+00 -4.97408003e-01
6.29761934e-01 -5.79655230e-01 -1.79371580e-01 1.67191684e+00
9.65414047e-02 4.42808121e-01 -3.42810869e-01 8.93681109e-01
5.67372143e-01 1.64441228e-01 -2.15575993e-02 -2.40980014e-01
1.53385818e+00 -2.58659184e-01 -9.00385678e-01 -1.67929485e-01
3.04844528e-01 -5.66449106e-01 8.44645262e-01 4.87660497e-01
-7.91563392e-01 -8.23368251e-01 -1.24511135e+00 7.38000572e-01
4.68976721e-02 9.36855450e-02 5.27444959e-01 1.64193225e+00
-5.77792704e-01 3.59097749e-01 -5.76229036e-01 -7.18172729e-01
2.63824105e-01 7.73357213e-01 -8.59222040e-02 3.04558594e-02
-1.51778615e+00 3.02195817e-01 -1.04909642e-02 1.75933734e-01
2.08418041e-01 -5.83656192e-01 -6.30748093e-01 4.20082323e-02
-1.16321715e-02 -8.88246059e-01 9.25707519e-01 -4.82566804e-01
-1.53703499e+00 3.59678358e-01 -4.43396181e-01 -3.46412838e-01
7.09934056e-01 -2.41820112e-01 -1.25538278e+00 1.62990183e-01
2.06869051e-01 -1.27742380e-01 7.60826349e-01 -8.62411320e-01
-6.49412453e-01 -5.99192083e-01 -8.65818337e-02 -1.80829003e-01
-7.85123408e-01 -2.48825885e-02 -5.67007899e-01 -6.22241318e-01
3.92084688e-01 -9.68493462e-01 1.64993405e-01 -4.70953792e-01
-4.64549422e-01 2.51794368e-01 6.89102948e-01 -7.11260200e-01
1.70402956e+00 -2.04917312e+00 -6.23411536e-01 6.70791507e-01
1.17022894e-01 2.22681493e-01 3.31223220e-01 3.09678912e-01
4.01984334e-01 7.52296224e-02 1.83521975e-02 -1.27205566e-01
1.72264516e-01 -2.50622898e-01 1.00122698e-01 6.72273278e-01
-4.53822941e-01 6.66492522e-01 -8.33394229e-01 -3.34541470e-01
4.38780606e-01 4.05051261e-01 -2.45901167e-01 -1.11847878e-01
1.02535832e+00 4.74649310e-01 -6.68850839e-01 7.49334633e-01
1.09697378e+00 1.53742909e-01 2.28001133e-01 -5.18692017e-01
-1.57069638e-01 2.19803918e-02 -1.55678737e+00 1.31828547e+00
-4.76003706e-01 2.47705162e-01 1.58016816e-01 -9.24827814e-01
9.53288853e-01 5.59304237e-01 8.62415969e-01 -8.02640200e-01
3.18775743e-01 2.79556811e-01 -8.60053673e-02 -6.50998890e-01
1.54781684e-01 1.97946489e-01 -2.80356467e-01 5.13738334e-01
-3.14727962e-01 7.26358056e-01 1.36627525e-01 -3.48179601e-02
1.19651544e+00 -1.07585907e-01 4.00271535e-01 -2.25207746e-01
8.19438159e-01 -5.28875709e-01 6.32281005e-01 8.36976767e-01
-5.13160586e-01 2.50649780e-01 -4.68671232e-01 -8.66741408e-03
-2.57381648e-01 -1.11601973e+00 -1.39856920e-01 5.42969882e-01
6.50101662e-01 -4.80814606e-01 -7.09354043e-01 -3.48489940e-01
3.04735571e-01 4.85460423e-02 -5.90758063e-02 -1.68758661e-01
-8.10423374e-01 -7.79205203e-01 1.27072144e+00 5.31890273e-01
1.21120083e+00 -3.69310051e-01 -5.69618464e-01 5.89792192e-01
-7.50759423e-01 -1.24020028e+00 -5.69719374e-01 -6.07530892e-01
-3.30259055e-01 -9.10071015e-01 -9.79507506e-01 -6.01173878e-01
4.12729412e-01 6.72787488e-01 5.03398478e-01 7.25164115e-02
-2.11574525e-01 5.58623672e-01 -1.42484143e-01 -3.63041371e-01
4.87819612e-01 1.31478086e-01 7.30138361e-01 5.31081498e-01
8.30395281e-01 -7.63460934e-01 -8.54708791e-01 8.60445797e-01
1.20382272e-02 -4.17647749e-01 2.15673491e-01 4.08033788e-01
-1.81993306e-01 5.67376256e-01 8.52011681e-01 -5.02284132e-02
4.67577845e-01 -3.60514164e-01 -1.44089982e-01 5.38460724e-02
-3.65333200e-01 -4.88766760e-01 6.30783617e-01 -2.68085778e-01
-9.38698590e-01 -1.01174094e-01 -3.79045606e-01 4.12287652e-01
-1.57518566e-01 1.70501620e-01 -7.39505827e-01 -4.25370246e-01
5.79971790e-01 2.63883501e-01 -2.05457300e-01 -3.33879411e-01
-1.26919150e-01 1.34908760e+00 7.54124582e-01 -1.93696693e-01
1.15126276e+00 6.22060478e-01 -1.78088188e-01 -1.24454546e+00
-3.94772664e-02 -9.15548027e-01 -5.59402943e-01 -4.97923046e-01
4.28593367e-01 -9.57513690e-01 -1.51261103e+00 8.14238548e-01
-8.98149073e-01 1.90058082e-01 3.44731897e-01 7.47948647e-01
-6.02516197e-02 6.69779956e-01 -4.66538966e-01 -1.22531676e+00
-6.30478859e-01 -5.99419355e-01 5.29674053e-01 4.86584991e-01
-7.88098693e-01 -5.98862529e-01 -3.97897393e-01 5.89163542e-01
7.04487622e-01 1.51231006e-01 1.50728419e-01 9.99165103e-02
-2.18091384e-01 -6.09201550e-01 -2.74665803e-01 -5.24549842e-01
4.90215421e-01 -5.59593916e-01 -9.65286016e-01 -3.76826763e-01
-1.39569998e-01 3.07764232e-01 2.65683979e-01 5.91469586e-01
9.50711429e-01 -1.54832616e-01 -9.05763388e-01 8.36207986e-01
1.11612153e+00 2.97334939e-01 8.14487875e-01 4.05374229e-01
6.36886597e-01 2.88457543e-01 5.63429117e-01 6.79079652e-01
3.58358413e-01 1.02376699e+00 -3.66713107e-02 -2.45716777e-02
1.46530941e-01 -2.19842553e-01 1.66678846e-01 5.53669930e-01
-4.76075739e-01 -5.47933996e-01 -7.44347990e-01 2.80795753e-01
-1.82771659e+00 -1.16449547e+00 -3.93507272e-01 2.60242915e+00
2.31524572e-01 1.68934420e-01 6.63015723e-01 8.32627296e-01
9.87495065e-01 -2.03647166e-01 -4.29016352e-01 2.37987444e-01
1.14721514e-01 9.43076015e-02 9.29621160e-01 3.34778965e-01
-1.17086327e+00 1.48423329e-01 5.26435375e+00 6.20222270e-01
-1.07280016e+00 1.12644091e-01 -2.71319970e-02 2.74192113e-02
1.36110097e-01 -6.06890500e-01 -8.22974801e-01 8.11043382e-01
7.96560526e-01 -2.27934584e-01 5.84239513e-02 5.55924416e-01
5.90860665e-01 4.14772928e-02 -7.45128512e-01 1.76837146e+00
-4.63772044e-02 -7.25464106e-01 -3.39696079e-01 2.59327710e-01
3.38979289e-02 -5.80973685e-01 -1.99681640e-01 6.72816560e-02
-3.58153224e-01 -6.43594980e-01 5.59144914e-01 2.92279899e-01
1.13373077e+00 -8.91294301e-01 9.73226845e-01 1.85721144e-01
-2.00054264e+00 -1.84680849e-01 -1.68018043e-01 -4.33366835e-01
6.11007690e-01 9.55824852e-01 -4.93457228e-01 8.49728525e-01
6.24333739e-01 4.98230577e-01 -1.25958875e-01 1.31399584e+00
3.52466479e-03 8.43133509e-01 -7.23489642e-01 -1.52529418e-01
-3.82028610e-01 -1.65051490e-01 3.71984214e-01 1.39665210e+00
8.89437616e-01 2.19221085e-01 1.21412985e-01 4.45410579e-01
1.53457657e-01 -6.44752085e-02 -5.02325416e-01 5.17504036e-01
9.73960876e-01 9.99694467e-01 -5.53918719e-01 -7.60283292e-05
-3.00664008e-01 1.13303542e+00 -6.00089192e-01 4.14979130e-01
-1.03155422e+00 -9.44058597e-01 7.08238661e-01 6.05786562e-01
-1.03846394e-01 -4.52920943e-01 -4.19063628e-01 -1.08822381e+00
2.85278916e-01 -3.70763987e-01 1.23611107e-01 -1.77034475e-02
-1.06663859e+00 1.31163761e-01 -4.60739791e-01 -1.65898108e+00
-2.06353232e-01 -1.02457516e-01 -5.34354389e-01 9.71859157e-01
-9.69880939e-01 -1.00731170e+00 -8.92358959e-01 8.44040513e-01
-9.80617385e-03 2.23766100e-02 8.40233088e-01 1.01783526e+00
-6.96132004e-01 1.38558650e+00 7.48769045e-02 3.64697158e-01
6.14345491e-01 -7.11348295e-01 5.74909389e-01 8.84582579e-01
-2.81978786e-01 1.07354128e+00 5.51677465e-01 -6.54384851e-01
-1.60649633e+00 -9.30146217e-01 1.07339597e+00 -1.28530279e-01
1.30707830e-01 -4.82906818e-01 -3.58697593e-01 2.02610672e-01
-3.77612352e-01 -2.65122741e-01 9.47265625e-01 1.03271559e-01
6.56976476e-02 -6.43642008e-01 -1.45737779e+00 7.20830500e-01
1.65110242e+00 -3.85244876e-01 -2.22868949e-01 -1.55095533e-01
-1.22092530e-01 8.65021348e-02 -8.76873016e-01 3.05450588e-01
1.37848115e+00 -8.94168794e-01 1.42836750e+00 4.35428649e-01
-6.97374582e-01 -4.66482967e-01 3.02655920e-02 -8.82635057e-01
-7.62774289e-01 -9.82594669e-01 -1.62189126e-01 1.63485301e+00
8.29060748e-02 -1.08492458e+00 8.53842139e-01 6.98462009e-01
5.56924224e-01 -1.51012644e-01 -8.96207035e-01 -1.12729836e+00
-9.07352269e-01 -5.50391376e-01 1.01207137e+00 8.24275672e-01
4.43524927e-01 3.09038728e-01 -7.76323915e-01 3.48540694e-01
1.07508743e+00 -3.53253841e-01 9.39574003e-01 -1.52483606e+00
-3.72657001e-01 -1.72284067e-01 -7.77052462e-01 -1.44973755e+00
-5.12571871e-01 -3.26655060e-01 -2.01921031e-01 -1.39749002e+00
-2.26621866e-01 -6.71797156e-01 -1.60870925e-01 2.09539086e-01
4.24180180e-02 6.05900347e-01 -6.72687665e-02 1.17716745e-01
-3.62297237e-01 2.88075328e-01 5.18353343e-01 -2.74395287e-01
-4.22538370e-01 5.21735907e-01 -5.52153349e-01 8.00672770e-01
9.32085574e-01 3.62887830e-02 -4.53977138e-01 -2.31321096e-01
-1.33527756e-01 -7.97968358e-02 2.76085794e-01 -1.83055162e+00
1.94953591e-01 1.22810215e-01 6.24587655e-01 -2.01276228e-01
4.93663847e-01 -1.05425417e+00 5.03991604e-01 7.79594421e-01
4.88387227e-01 -3.46142381e-01 -3.02221645e-02 2.77544796e-01
6.85537383e-02 7.83879086e-02 4.82542902e-01 2.03846216e-01
-4.52107996e-01 2.96844751e-01 -5.39297819e-01 -3.65653962e-01
8.29836845e-01 -8.13498437e-01 -1.43869027e-01 -6.58207178e-01
-3.13308537e-01 2.14740988e-02 3.23788464e-01 3.28527212e-01
4.16817188e-01 -1.47602987e+00 -3.07911277e-01 4.78372842e-01
4.75638211e-02 -8.09029996e-01 5.00518084e-01 8.68051171e-01
-2.72185922e-01 6.52443647e-01 -2.24046797e-01 -5.46643078e-01
-1.44732523e+00 1.01051725e-01 1.50025606e-01 1.10711932e-01
-6.34051859e-01 5.10513604e-01 -3.56895983e-01 -8.42863321e-02
2.61914194e-01 -1.72509830e-02 -2.23250985e-01 1.84241217e-02
9.72586691e-01 8.65095377e-01 1.38933167e-01 -7.35456228e-01
-9.14093673e-01 1.17426586e+00 4.90575463e-01 -1.74766898e-01
5.61693609e-01 -7.00329602e-01 5.52702665e-01 -1.16674632e-01
8.47189724e-01 4.04243708e-01 -7.32468009e-01 -4.75516245e-02
-1.15642458e-01 -6.50157511e-01 -2.52899259e-01 -3.92020911e-01
-8.28006446e-01 4.87603754e-01 9.04954076e-01 3.90871078e-01
1.03356075e+00 -6.31660819e-01 1.46815467e+00 2.44418278e-01
1.32732224e+00 -9.14411604e-01 -4.66049314e-01 9.65057835e-02
2.45394632e-01 -8.75117838e-01 -3.77252921e-02 -7.83123195e-01
-3.17985713e-02 8.83690834e-01 3.81025404e-01 2.93355465e-01
7.22547472e-01 2.79550076e-01 -4.83467467e-02 2.94354022e-01
3.78311187e-01 -3.33772570e-01 5.98485675e-03 1.29349267e+00
3.18821430e-01 2.53323525e-01 -4.43129212e-01 1.35486341e+00
-5.69744289e-01 2.42252335e-01 5.14525138e-02 9.49797750e-01
-4.33157772e-01 -1.02069592e+00 -7.68754244e-01 6.94452882e-01
-5.85035563e-01 3.13174427e-01 -1.28164953e-02 4.46977109e-01
5.09297073e-01 1.54099584e+00 -3.71050239e-01 -8.33068728e-01
6.08901143e-01 -1.56151906e-01 4.53107208e-01 7.92912394e-02
-4.02340919e-01 -2.15671539e-01 3.35613251e-01 -5.89867353e-01
-4.40971047e-01 -6.28751099e-01 -1.18805230e+00 -7.39152789e-01
-3.02220374e-01 1.74619943e-01 2.80346274e-01 1.03154182e+00
4.27748173e-01 4.27748293e-01 5.78589857e-01 -8.21640730e-01
-6.73202351e-02 -8.54837537e-01 -7.75567234e-01 4.13244724e-01
5.96095249e-02 -8.40237200e-01 -1.16235025e-01 -1.33402452e-01] | [6.726963043212891, 0.6799643039703369] |
efbe5472-18cd-47a3-b9fb-83771352565c | learning-the-distribution-of-errors-in-stereo | 2304.00152 | null | https://arxiv.org/abs/2304.00152v1 | https://arxiv.org/pdf/2304.00152v1.pdf | Learning the Distribution of Errors in Stereo Matching for Joint Disparity and Uncertainty Estimation | We present a new loss function for joint disparity and uncertainty estimation in deep stereo matching. Our work is motivated by the need for precise uncertainty estimates and the observation that multi-task learning often leads to improved performance in all tasks. We show that this can be achieved by requiring the distribution of uncertainty to match the distribution of disparity errors via a KL divergence term in the network's loss function. A differentiable soft-histogramming technique is used to approximate the distributions so that they can be used in the loss. We experimentally assess the effectiveness of our approach and observe significant improvements in both disparity and uncertainty prediction on large datasets. | ['Philippos Mordohai', 'Weihan Wang', 'Liyan Chen'] | 2023-03-31 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Chen_Learning_the_Distribution_of_Errors_in_Stereo_Matching_for_Joint_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Chen_Learning_the_Distribution_of_Errors_in_Stereo_Matching_for_Joint_CVPR_2023_paper.pdf | cvpr-2023-1 | ['stereo-matching-1'] | ['computer-vision'] | [-1.33805975e-01 1.06246984e-02 -7.65863210e-02 -8.40902328e-01
-1.38699114e+00 -3.28609109e-01 4.24848020e-01 2.02542111e-01
-6.45462632e-01 1.08919513e+00 4.48023558e-01 -8.82522203e-03
-7.28661418e-02 -4.73503351e-01 -1.13897395e+00 -4.05075848e-01
2.19369322e-01 4.39612269e-01 3.34150165e-01 2.62948602e-01
4.58927989e-01 2.49471769e-01 -1.67586839e+00 -1.72123909e-02
1.05138099e+00 1.20028138e+00 4.85031232e-02 5.88389575e-01
4.54626232e-02 6.48618519e-01 -5.08691430e-01 -5.01352489e-01
6.80589259e-01 -1.42291903e-01 -6.48523927e-01 -2.79026300e-01
1.08608317e+00 -7.09960163e-01 -3.29956830e-01 1.12379849e+00
6.82529867e-01 4.12572026e-01 8.53907466e-01 -1.19077945e+00
1.51337907e-02 2.27871090e-01 -7.14319527e-01 2.27821991e-01
1.57744810e-01 -1.38606727e-01 1.10822535e+00 -6.86376154e-01
4.57886606e-01 1.40583766e+00 8.14473093e-01 3.93370181e-01
-1.35638309e+00 -6.76693380e-01 -7.13650212e-02 7.79216290e-02
-1.37648165e+00 -6.28918231e-01 4.49955016e-01 -5.72799325e-01
7.42464900e-01 -1.76053509e-01 2.61288017e-01 6.96580708e-01
5.42169869e-01 7.46955395e-01 9.80672538e-01 -2.69028783e-01
2.72509813e-01 9.10428390e-02 -1.00183763e-01 6.25505388e-01
2.90778518e-01 4.78190362e-01 -6.71546757e-01 -2.07603201e-01
9.42038953e-01 -2.66248852e-01 -2.74971128e-01 -6.53113425e-01
-6.64689779e-01 9.16128755e-01 5.01619816e-01 -2.66222537e-01
-7.65330493e-02 7.27841258e-01 4.09296453e-01 2.28340045e-01
6.52269781e-01 2.05603167e-01 -1.65443391e-01 -2.96255797e-01
-9.41949368e-01 5.21523535e-01 6.44374669e-01 7.80483007e-01
9.84209299e-01 -2.98036814e-01 -2.68736809e-01 9.09548700e-01
3.07920396e-01 2.86367267e-01 1.11987339e-02 -1.60054588e+00
5.83117604e-01 3.27362828e-02 2.69434810e-01 -6.69557512e-01
-6.71625361e-02 -1.93446279e-01 -4.34712946e-01 7.15920508e-01
7.72868872e-01 -3.57738107e-01 -9.58164036e-01 2.01670003e+00
1.96909830e-01 3.98433030e-01 -1.50450781e-01 7.48955488e-01
2.09662870e-01 2.20029011e-01 -6.90519437e-02 1.79991871e-01
6.06995225e-01 -6.55411839e-01 -3.45965117e-01 -3.53775471e-01
1.90586120e-01 -6.55027568e-01 7.42188096e-01 7.76100680e-02
-1.15126562e+00 -4.60029900e-01 -1.11278117e+00 -2.92831481e-01
1.04898624e-01 -3.97549719e-01 4.21658814e-01 4.35198367e-01
-9.64689195e-01 9.59974468e-01 -7.19861507e-01 7.41394013e-02
5.98542273e-01 5.31377196e-01 -1.36009708e-01 -1.67598486e-01
-9.42221940e-01 1.09510398e+00 3.83449167e-01 -1.86555728e-01
-7.02241719e-01 -8.63644719e-01 -1.09771872e+00 -5.67535162e-02
3.83097082e-02 -9.80166018e-01 1.27415562e+00 -6.16849959e-01
-1.31060159e+00 7.57047594e-01 -2.49527216e-01 -6.62559748e-01
8.75069082e-01 -4.59250927e-01 4.35392648e-01 -4.81199920e-02
1.77149668e-01 1.02867103e+00 8.51533830e-01 -1.13893712e+00
-9.16925669e-01 -3.98591548e-01 1.08932935e-01 4.39660043e-01
1.17896885e-01 -4.31547046e-01 -6.00604951e-01 -4.85212773e-01
1.05571067e-02 -7.57366598e-01 -2.96633512e-01 4.91619587e-01
-1.67216823e-01 -1.50304493e-02 3.05650592e-01 -5.80209017e-01
8.79879117e-01 -2.14197397e+00 2.24284772e-02 4.19454008e-01
2.73452282e-01 -3.69713545e-01 4.44422849e-02 -1.08756281e-01
4.33373332e-01 -2.36369744e-01 -3.59123766e-01 -7.14318991e-01
1.69776812e-01 3.20481509e-01 -3.14110607e-01 5.81835628e-01
2.97460519e-02 7.46064961e-01 -8.32078099e-01 -5.30452788e-01
3.28291386e-01 4.72385198e-01 -7.63696790e-01 1.71846136e-01
-1.99126199e-01 4.52028483e-01 -4.35512476e-02 2.81069726e-01
7.49806762e-01 -1.34682387e-01 -3.06815177e-01 -1.02898471e-01
1.92931488e-01 3.60813409e-01 -1.27438843e+00 1.79031789e+00
-6.63907111e-01 8.83648932e-01 1.20602855e-02 -4.42048192e-01
6.82464361e-01 -1.01153292e-01 3.86982173e-01 -4.65999007e-01
3.87250073e-02 5.16990304e-01 -2.20500395e-01 6.79786801e-02
4.22923774e-01 -3.75025988e-01 6.71434402e-03 1.48469508e-01
-1.18478909e-01 -6.19686961e-01 9.53448843e-03 -2.26089120e-01
6.63978517e-01 2.49146804e-01 8.40640292e-02 -5.10378897e-01
1.64018810e-01 -2.02047735e-01 6.81576669e-01 8.03555906e-01
-3.46174002e-01 7.39742756e-01 3.43994945e-01 -1.91464588e-01
-1.26198697e+00 -1.41488051e+00 -5.96413255e-01 7.51067936e-01
2.51013696e-01 2.43496940e-01 -5.49364865e-01 -5.79133272e-01
6.73932731e-01 5.45451880e-01 -5.28719425e-01 -2.21023373e-02
-5.04931152e-01 -2.87773579e-01 4.26110297e-01 8.96404922e-01
5.15292048e-01 -2.95058757e-01 -4.32868600e-01 1.28970429e-01
-5.13222665e-02 -1.02012265e+00 -7.25787222e-01 2.53852695e-01
-1.00093794e+00 -9.29854870e-01 -6.94516480e-01 -5.38329363e-01
4.84317124e-01 -1.28673047e-01 1.15079594e+00 -2.39382178e-01
-1.99804440e-01 3.58374804e-01 4.82031792e-01 -4.10034001e-01
-2.02696636e-01 -3.32432508e-01 2.75309645e-02 -5.06685793e-01
2.96071529e-01 -7.20768213e-01 -6.38564050e-01 2.22998470e-01
-7.10973859e-01 -3.07921290e-01 1.53097436e-01 9.26070750e-01
7.12477028e-01 -1.17513835e-01 3.36545557e-01 -4.81070995e-01
5.87340474e-01 -1.75689921e-01 -1.12687087e+00 -2.70675216e-02
-7.60785818e-01 6.88239694e-01 4.08093333e-02 -1.33461580e-01
-9.77304757e-01 6.13302700e-02 -1.96795642e-01 -7.11020947e-01
2.00726569e-01 1.71528265e-01 2.85706431e-01 -4.20055777e-01
6.26683831e-01 -2.57511914e-01 9.86974239e-02 -2.77226806e-01
3.02446395e-01 5.78901291e-01 7.97387838e-01 -8.45618963e-01
5.48080206e-01 6.01740956e-01 3.38126719e-01 -4.37496096e-01
-1.07411242e+00 -5.08583605e-01 -3.09097230e-01 -1.00552790e-01
6.81204319e-01 -1.19343460e+00 -6.96321845e-01 4.73456740e-01
-1.13590181e+00 -2.68829048e-01 -3.37808758e-01 6.25831544e-01
-8.54136348e-01 4.61443424e-01 -5.32374263e-01 -8.96964252e-01
-1.87338069e-01 -1.18822634e+00 1.36072755e+00 2.68852293e-01
-1.05254292e-01 -1.32597125e+00 3.42375070e-01 3.72538939e-02
3.85174364e-01 2.31330723e-01 8.39530230e-01 -2.70100951e-01
-7.50097990e-01 9.06094443e-03 -5.88421226e-01 5.20327389e-01
-7.39837587e-02 -2.43052825e-01 -1.22895288e+00 -2.83258349e-01
-7.70227388e-02 -6.40449822e-01 1.34142435e+00 9.67236400e-01
1.22660768e+00 1.79947719e-01 -2.54380524e-01 8.68976176e-01
1.83963382e+00 -9.73816290e-02 7.64243066e-01 3.63674164e-01
5.07546008e-01 5.12291014e-01 5.37515283e-01 5.42202950e-01
4.77499336e-01 7.61022031e-01 4.13092524e-01 3.65211189e-01
-5.51067553e-02 -3.52028638e-01 -1.41562987e-02 1.60105169e-01
2.91357934e-01 -9.37298778e-03 -9.23051894e-01 5.56132495e-01
-1.96855426e+00 -7.36163437e-01 2.25979909e-01 2.57875037e+00
9.59329903e-01 4.30275530e-01 -4.08881195e-02 -3.23209852e-01
6.89565063e-01 2.92309690e-02 -8.41238439e-01 -3.49070728e-01
3.18013318e-02 1.40381128e-01 8.37970316e-01 1.15521145e+00
-1.10029960e+00 5.85687697e-01 7.85530472e+00 8.78114283e-01
-7.71872044e-01 -1.42903090e-01 7.35925078e-01 -2.34454170e-01
-6.01416230e-01 -1.74438775e-01 -9.40093696e-01 5.46151638e-01
6.45111918e-01 -3.55820715e-01 2.55088985e-01 6.67876601e-01
3.63175124e-02 -6.04980826e-01 -1.32736707e+00 1.22832274e+00
-2.21991211e-01 -1.31411481e+00 -3.08104277e-01 1.41260579e-01
9.56343532e-01 2.83824891e-01 1.73683330e-01 9.20749456e-02
5.53855717e-01 -9.48639035e-01 7.05419004e-01 5.11409342e-01
7.57466018e-01 -9.43203032e-01 6.79971874e-01 1.09559134e-01
-1.00976598e+00 -2.47607138e-02 -6.51331902e-01 -2.33856887e-02
2.60336071e-01 8.93359125e-01 -5.79456806e-01 1.36251032e-01
4.52801079e-01 5.67045748e-01 -8.31157714e-02 1.56118941e+00
-4.22591753e-02 -1.48824096e-01 -6.95832491e-01 1.35949641e-01
1.64325058e-01 -2.76744425e-01 5.09865642e-01 1.04841411e+00
2.27164239e-01 -4.83936340e-01 3.36341150e-02 9.72379804e-01
-1.92486241e-01 -1.74165979e-01 -6.12345457e-01 3.86543542e-01
5.83717048e-01 5.19614458e-01 -5.19625880e-02 -1.47884488e-01
-3.19368482e-01 9.93079245e-01 6.78911746e-01 1.98478356e-01
-6.90003276e-01 -4.64019001e-01 1.11054063e+00 -1.34624526e-01
3.66913646e-01 -1.50380120e-01 -6.30876958e-01 -9.94240880e-01
2.56520122e-01 -3.35868090e-01 3.54942977e-01 -6.34801865e-01
-1.47622895e+00 3.22014600e-01 1.55824199e-01 -1.05857873e+00
-7.05252051e-01 -6.14827514e-01 -5.45455277e-01 1.17193794e+00
-1.84081376e+00 -6.16766512e-01 -1.46825299e-01 4.03654039e-01
2.34109968e-01 8.50762948e-02 4.06990081e-01 4.70897019e-01
-9.06059891e-02 7.06021726e-01 3.21910948e-01 -2.61581212e-01
8.44020128e-01 -1.56412292e+00 4.31287855e-01 6.67983711e-01
-1.57965034e-01 1.99291602e-01 9.55542862e-01 -4.96620476e-01
-7.10748971e-01 -8.31293881e-01 6.99945092e-01 -4.80876029e-01
6.79431140e-01 -1.46219417e-01 -7.11797297e-01 6.31003022e-01
-2.82624383e-02 1.03889711e-01 3.65033984e-01 1.32200494e-01
-5.00065207e-01 -1.16678946e-01 -1.40329397e+00 4.22249615e-01
8.82234812e-01 -8.35105240e-01 -4.99268621e-01 9.87280086e-02
6.13589704e-01 -8.99943233e-01 -7.85844743e-01 5.52282631e-01
8.06864262e-01 -1.20337820e+00 8.72631907e-01 -3.35900038e-01
5.75269818e-01 -5.78261241e-02 -5.36265433e-01 -1.29849792e+00
5.58711663e-02 -3.38592708e-01 -1.80737972e-01 8.06607068e-01
3.97739947e-01 -6.01262748e-01 1.01667058e+00 9.99079347e-01
-1.19127110e-01 -7.90764868e-01 -1.35506594e+00 -9.00041461e-01
3.41229647e-01 -5.52201152e-01 3.01393390e-01 1.71284914e-01
-1.72708377e-01 -9.69677716e-02 -3.68793845e-01 6.86180368e-02
1.18809271e+00 7.19344616e-02 5.48458874e-01 -1.16032827e+00
-5.69219053e-01 -5.63656211e-01 -7.24464715e-01 -1.53735912e+00
3.37799609e-01 -4.77206856e-01 4.65290844e-01 -1.45787847e+00
3.12348872e-01 -4.15853024e-01 -1.62058324e-01 -6.56486452e-02
-2.81356812e-01 7.78186768e-02 1.06033936e-01 8.14729091e-03
-3.51898015e-01 6.50813162e-01 1.11176705e+00 -7.88361300e-03
-1.41883612e-01 2.02736527e-01 -4.68052357e-01 6.78782105e-01
5.34246624e-01 -3.68239582e-01 -3.84974718e-01 -5.89719117e-01
1.99041203e-01 -9.71191563e-03 3.47573608e-01 -1.18253243e+00
2.74022996e-01 -5.82987294e-02 2.88807243e-01 -6.29049301e-01
6.28230989e-01 -8.31799150e-01 -9.26958472e-02 3.03890198e-01
-3.52818310e-01 -1.42531931e-01 5.03432274e-01 6.85427248e-01
-2.47352943e-01 -2.89501488e-01 1.16249454e+00 1.70393065e-01
-7.28561044e-01 3.39745492e-01 3.03209990e-01 4.20559108e-01
7.71011472e-01 -1.17671207e-01 -2.31315851e-01 -7.24655747e-01
-4.31386799e-01 5.21040320e-01 5.85054934e-01 1.08537666e-01
6.66689754e-01 -1.57190013e+00 -6.50631368e-01 3.57609615e-02
1.14068881e-01 1.92608789e-01 -1.34251177e-01 5.98885238e-01
-5.04188836e-01 2.61174202e-01 -2.00825483e-01 -8.21087301e-01
-9.62090194e-01 5.35310572e-03 8.32472861e-01 -2.11274534e-01
-3.05712789e-01 1.13705409e+00 8.15058127e-02 -2.30788767e-01
7.22674191e-01 -3.12109619e-01 3.21683079e-01 -2.65043199e-01
4.09286559e-01 5.94622910e-01 -1.93016604e-01 -1.32570386e-01
-1.87171191e-01 6.56578124e-01 2.63773464e-02 -4.79542971e-01
1.13810647e+00 -4.57867056e-01 1.81565344e-01 5.63424289e-01
1.52884793e+00 1.15172304e-02 -1.95257199e+00 -3.22241515e-01
-2.06170119e-02 -8.94971192e-01 2.81685561e-01 -5.33259392e-01
-8.36724877e-01 8.08557689e-01 8.16602051e-01 -4.20031130e-01
7.16054261e-01 -4.22762670e-02 9.46365535e-01 2.72386163e-01
5.74736953e-01 -1.16682482e+00 -1.22300178e-01 5.34593165e-01
6.20041370e-01 -1.61684716e+00 4.50937264e-02 -2.48975828e-01
-4.28514004e-01 8.81198227e-01 6.90445840e-01 -3.35248172e-01
8.66027117e-01 5.48674822e-01 2.26841960e-02 9.77830067e-02
-5.10161877e-01 -2.35823199e-01 6.17776453e-01 5.73049188e-01
2.81485319e-01 -2.33572185e-01 -7.01566190e-02 -1.89200193e-01
-7.99422637e-02 8.30938071e-02 2.38906562e-01 7.96225011e-01
-7.78257310e-01 -1.04846263e+00 -1.50098681e-01 6.53854966e-01
-4.02612031e-01 -9.26331952e-02 1.43906549e-01 3.68096441e-01
-1.69118658e-01 6.54314935e-01 4.11406964e-01 -1.88927889e-01
1.68126583e-01 -4.03453819e-02 8.20497096e-01 -4.48807359e-01
8.89829993e-02 7.18970131e-03 1.14946380e-01 -7.57684112e-01
-2.38304257e-01 -5.55660784e-01 -1.18482494e+00 -4.39867824e-01
-2.18426615e-01 -7.23569319e-02 6.95291400e-01 1.01895416e+00
1.74421906e-01 1.25516400e-01 5.29143691e-01 -6.97139502e-01
-9.59708452e-01 -5.45376837e-01 -7.32832134e-01 3.79355878e-01
6.40014768e-01 -8.18578959e-01 -6.90286279e-01 -3.84956628e-01] | [8.30721664428711, -2.2283220291137695] |
7182d828-5ea2-4928-9ec1-1f4dc0e447d7 | electronics-and-sensor-subsystem-design-for | 2211.02870 | null | https://arxiv.org/abs/2211.02870v1 | https://arxiv.org/pdf/2211.02870v1.pdf | Electronics and Sensor Subsystem Design for Daedalus 2 on REXUS 29: An Autorotation Probe for Sub-Orbital Re-Entry | The Daedalus 2 mission aboard REXUS 29 is a technology demonstrator for an alternative descent mechanism for very high altitude drops based on auto-rotation. It consists of two probes that are ejected from a sounding rocket at an altitude of about 80 km and decelerate to a soft landing using only a passive rotor with pitch control. This type of autonomous, scientific experiment poses great challenges upon the electronics subsystem, which include mechanical stress, power system reliability, sensor redundancy, subsystem communication, and development procedures. Based on the data gathered in Daedalus 1 multiple new approaches were developed to fulfill these requirements, such as redundant communication links, mechanical decoupling of PCBs and fault-tolerant power source selection. | ['Frederik Dunschen', 'Clemens Riegler', 'Philip Bergmann', 'Lennart Werner', 'Jan M. Wolf'] | 2022-11-05 | null | null | null | null | ['pitch-control'] | ['audio'] | [-4.43537802e-01 1.42324820e-01 1.64667889e-01 1.08770147e-01
4.92368370e-01 -1.04473543e+00 -4.55616899e-02 -2.73412287e-01
-6.10696562e-02 8.27391088e-01 -5.57842135e-01 -3.28379095e-01
-7.12563038e-01 -4.86523777e-01 -4.16624606e-01 -3.50888729e-01
-3.92286748e-01 3.64300638e-01 2.51627803e-01 -6.19576573e-01
1.33244231e-01 1.19024682e+00 -1.68100989e+00 -6.61922097e-01
5.43205500e-01 7.61136949e-01 1.02936521e-01 3.45109522e-01
7.74117112e-01 2.03420490e-01 -8.64469647e-01 3.46973002e-01
2.70204451e-02 -4.69279647e-01 -6.28377676e-01 -2.42966428e-01
-3.11446130e-01 -1.95179537e-01 7.14123398e-02 6.65517569e-01
3.44037890e-01 -1.61776375e-02 4.86147314e-01 -1.29418314e+00
8.53494167e-01 6.66107714e-01 -2.01183662e-01 1.25940338e-01
4.64215696e-01 -6.48243427e-02 3.70269090e-01 -3.97084177e-01
6.09155774e-01 6.89558625e-01 8.36895406e-01 3.35973710e-01
-1.18019390e+00 -5.68500102e-01 -8.72256756e-01 -1.02391869e-01
-1.33528042e+00 -4.76657778e-01 5.15181065e-01 -4.63291913e-01
1.00116718e+00 5.73447764e-01 9.54312384e-01 7.39919364e-01
8.60659003e-01 -3.56945604e-01 8.75232935e-01 -4.63985622e-01
5.26291728e-01 4.41558003e-01 -6.86986372e-02 5.73235571e-01
8.47768843e-01 1.05514929e-01 -5.37128985e-01 -1.16910532e-01
5.60980082e-01 -7.97153056e-01 -5.62468767e-01 -1.67140916e-01
-7.71309257e-01 1.34422660e-01 -2.62740385e-02 3.98092419e-01
-1.04429070e-02 2.86027044e-02 4.01763052e-01 5.16430318e-01
-2.46543854e-01 1.18768001e+00 -8.05491984e-01 -5.09501219e-01
-5.75382829e-01 4.77094232e-04 1.04121792e+00 7.59130955e-01
2.54841715e-01 6.72390699e-01 6.46635056e-01 1.96943164e-01
3.78142476e-01 6.55685484e-01 2.59285629e-01 -3.90595704e-01
1.92521550e-02 4.89265800e-01 2.62316763e-01 -6.49996340e-01
-1.30642617e+00 -5.23761809e-01 -1.17628567e-01 3.89654070e-01
-4.10051972e-01 -1.16701341e+00 -1.35801867e-01 1.15707374e+00
4.65730309e-01 -6.23225331e-01 1.76443681e-01 1.23212302e+00
4.10072625e-01 5.76016068e-01 -5.18039286e-01 -1.84308663e-01
1.28995192e+00 -1.09012678e-01 -5.52239060e-01 2.63970140e-02
5.98116636e-01 -6.68451250e-01 5.26310503e-01 6.65838242e-01
-1.00474727e+00 -5.46709776e-01 -2.06076574e+00 5.20088255e-01
-6.85178638e-02 5.18082500e-01 2.18463555e-01 9.94180918e-01
-8.35237563e-01 7.83487558e-01 -8.21319163e-01 -5.89676201e-01
-8.12479854e-01 3.16116720e-01 -2.39564344e-01 8.81234705e-01
-1.18687820e+00 1.12670636e+00 1.27823114e-01 1.73467487e-01
-7.58317947e-01 -6.07816219e-01 -4.96734202e-01 -1.42014638e-01
-1.87419713e-01 -8.21414649e-01 8.42689037e-01 -6.74890950e-02
-2.50984502e+00 1.82698920e-01 9.95958507e-01 -5.45772970e-01
1.66812867e-01 -6.91153824e-01 -8.02585900e-01 6.55244350e-01
2.16444228e-02 -2.48688802e-01 6.58900499e-01 -4.95671183e-01
-6.72262192e-01 9.29230265e-03 -3.89001489e-01 1.69711828e-01
-3.40032578e-01 7.73438811e-02 6.14403844e-01 -2.18793914e-01
1.84559435e-01 -1.09334040e+00 6.33595511e-02 -4.39834863e-01
-6.98345959e-01 1.08563296e-01 1.09972978e+00 -1.30548373e-01
4.25008774e-01 -1.91064310e+00 6.55630589e-01 4.38973099e-01
-5.43542624e-01 -2.11125195e-01 7.72331595e-01 1.07289803e+00
-2.45955330e-03 -1.18240952e-01 1.86600685e-01 4.08725500e-01
-2.26923183e-01 -1.77573800e-01 -3.76442671e-01 9.82044876e-01
-6.68797418e-02 -3.50942492e-01 -3.76149178e-01 2.58369967e-02
1.00772791e-01 2.85162479e-02 -2.53981322e-01 2.58723676e-01
3.69115204e-01 4.98171657e-01 -6.35830641e-01 7.82426417e-01
5.67079544e-01 7.87119567e-01 3.92931998e-01 -4.21968371e-01
-9.83514488e-01 2.62232572e-01 -1.11924243e+00 1.53387392e+00
-5.17950356e-01 4.50528413e-01 9.97406185e-01 -6.77691102e-01
1.38905454e+00 3.36574435e-01 7.05542088e-01 -2.93319911e-01
6.68704987e-01 3.59400839e-01 -2.67497331e-01 -8.71895432e-01
1.07273376e+00 -3.66338134e-01 -5.37490964e-01 -1.77587457e-02
2.79960066e-01 -8.56214345e-01 -7.90605173e-02 1.87389895e-01
1.18878210e+00 5.66013873e-01 -4.40158337e-01 -1.09666550e+00
3.96066755e-01 5.59039950e-01 8.28163505e-01 -2.03409180e-01
3.03978056e-01 7.57476687e-02 5.07093966e-01 2.53648669e-01
-7.56550193e-01 -9.94467199e-01 -4.02274340e-01 1.82743922e-01
7.66197801e-01 -6.65280282e-01 -4.02470738e-01 -1.95700392e-01
2.13253647e-01 6.63677573e-01 2.46782497e-01 -6.59017801e-01
-5.15676558e-01 -5.06747901e-01 7.51397491e-01 1.24446802e-01
2.09391326e-01 3.69464420e-02 -1.50147998e+00 2.26780787e-01
3.66399229e-01 -1.09129930e+00 5.44524431e-01 5.99047482e-01
-1.02051055e+00 -1.13093030e+00 3.62222284e-01 -3.30194890e-01
4.46292788e-01 -1.12461157e-01 9.76212084e-01 1.84918121e-01
-5.93415201e-01 5.26239157e-01 -5.69235682e-01 -3.69045198e-01
-7.04329386e-02 3.99031192e-02 8.76894593e-01 -4.96749252e-01
-7.48482108e-01 -4.92610484e-01 -4.65488255e-01 9.41718459e-01
-1.54521629e-01 -3.16809595e-01 5.28000891e-01 3.85060936e-01
3.72538194e-02 6.77984238e-01 7.63703406e-01 -2.47894883e-01
2.76246697e-01 -8.07274461e-01 -9.87377763e-01 -3.78774256e-01
-3.87553185e-01 -2.05079183e-01 1.16405916e+00 1.70966282e-01
-9.99238074e-01 7.79983774e-02 -1.48972824e-01 2.91833039e-02
2.27445513e-01 3.38858843e-01 -1.35944709e-01 -4.84778732e-01
5.77835083e-01 -4.21544820e-01 4.70445275e-01 -4.16999102e-01
-2.07739890e-01 5.08605242e-01 8.46923530e-01 -5.92053354e-01
1.07170069e+00 3.29249591e-01 7.83160925e-01 -1.45403850e+00
5.84459752e-02 1.50719926e-01 -2.29432523e-01 -5.25487244e-01
4.90157038e-01 -8.95581067e-01 -9.13486600e-01 1.45571962e-01
-6.21707022e-01 -6.03524670e-02 -3.46859157e-01 1.09673679e+00
-5.84771149e-02 8.93281028e-02 -3.82356048e-01 -7.18067586e-01
-5.72448313e-01 -8.80134523e-01 5.42754173e-01 6.13902748e-01
-5.92593670e-01 -5.64784765e-01 1.53573439e-01 -1.06025428e-01
4.52818155e-01 7.44465947e-01 5.40261626e-01 1.53705448e-01
1.25510497e-02 -4.76345360e-01 8.07794392e-01 1.61737919e-01
-1.29996479e-01 5.26203454e-01 -2.93418199e-01 -9.08390403e-01
5.63006103e-01 -3.63819361e-01 6.96215034e-03 -2.99272746e-01
3.79561692e-01 2.35372454e-01 -9.41635430e-01 7.49244213e-01
1.51112235e+00 9.45431218e-02 2.03245461e-01 3.49888921e-01
4.90770079e-02 6.88404500e-01 1.24101520e+00 5.88124514e-01
-3.59318107e-01 6.58327162e-01 9.22296822e-01 1.28773093e-01
2.76815712e-01 2.65363693e-01 7.32410192e-01 8.86734843e-01
3.03156320e-02 -2.00602636e-02 -4.74861443e-01 1.77634835e-01
-7.73906410e-01 -1.08432069e-01 -7.90425003e-01 2.30950522e+00
4.64460820e-01 3.20292115e-01 -7.08306655e-02 4.15751755e-01
5.35504222e-01 -5.43079317e-01 -2.15194300e-01 -8.20569038e-01
1.49617329e-01 1.44417137e-01 1.10959029e+00 2.32059136e-01
-4.43220019e-01 1.38668388e-01 6.75739574e+00 1.11558199e-01
-1.42682528e+00 -2.43565202e-01 -7.14990854e-01 -3.22492421e-01
-9.83584523e-02 5.49248338e-01 -1.03346109e+00 4.38627541e-01
1.27956486e+00 -3.23717177e-01 -2.16000065e-01 7.06950307e-01
-1.18859582e-01 -6.20084703e-01 -7.56522954e-01 2.00332612e-01
4.38628048e-02 -7.81626463e-01 -5.58059990e-01 -2.39157274e-01
6.54068068e-02 -1.78470865e-01 -4.98220325e-01 4.69776914e-02
-3.61256868e-01 -1.54917613e-01 1.11056840e+00 5.14088333e-01
8.97464275e-01 -9.54496205e-01 5.36109746e-01 3.75103384e-01
-9.42998171e-01 -1.42624572e-01 -3.53838086e-01 -4.44442332e-01
5.18816471e-01 7.85176396e-01 -9.04644310e-01 1.22367108e+00
9.60973442e-01 6.13945015e-02 -3.55027974e-01 7.97888756e-01
-3.16975236e-01 7.38038242e-01 -8.22769463e-01 -4.05369848e-01
-2.30699450e-01 -4.97525275e-01 1.28071010e+00 2.48207092e-01
7.67984807e-01 -6.31839111e-02 -3.05401593e-01 5.36266565e-01
4.72779006e-01 -2.29552597e-01 -6.41134739e-01 4.09185797e-01
6.42163336e-01 1.90268922e+00 -8.08218718e-01 1.44921958e-01
2.58204311e-01 3.44689012e-01 -6.07615650e-01 -2.43485540e-01
-1.15519500e+00 -1.50585496e+00 6.23539925e-01 6.94521010e-01
-3.20201755e-01 -8.25489819e-01 1.01554804e-02 -5.67652881e-01
-7.62285218e-02 1.46238163e-01 -3.53047065e-02 -7.93714881e-01
-5.44663310e-01 5.47416866e-01 2.68064767e-01 -1.78829122e+00
-2.95553416e-01 -6.96688831e-01 -8.27039123e-01 6.07369304e-01
-6.71423793e-01 -6.00256085e-01 -3.57394487e-01 8.67808238e-02
-1.07945554e-01 -4.12382841e-01 7.05807149e-01 2.42587268e-01
-7.85038829e-01 5.97409308e-02 3.61964524e-01 -8.55419576e-01
7.04446793e-01 -1.03460252e+00 -2.93116897e-01 9.39362824e-01
-1.02722585e+00 3.09572607e-01 1.33215356e+00 -9.75426078e-01
-2.84364915e+00 -6.32054925e-01 8.36040825e-02 1.37651503e-01
7.92329729e-01 -5.13669848e-01 -2.42934376e-01 3.99082780e-01
4.98848945e-01 -3.98012191e-01 4.31408823e-01 -1.27684157e-02
8.91897619e-01 -5.77293456e-01 -1.32758844e+00 4.39701945e-01
6.29010975e-01 7.44452253e-02 -5.12098968e-01 4.27399367e-01
4.14456606e-01 -6.13197923e-01 -1.15368533e+00 5.02069175e-01
5.48782408e-01 -6.29107714e-01 3.51554573e-01 2.86245774e-02
-2.25065529e-01 -5.85498214e-01 4.39767540e-01 -1.48271143e+00
-1.40050158e-01 -1.40503705e+00 5.79878807e-01 1.49771321e+00
2.46242702e-01 -8.81941199e-01 3.13842803e-01 1.92003436e-02
-8.56121182e-01 -2.86143214e-01 -1.31326258e+00 -1.04533219e+00
-4.98120636e-01 4.11067039e-01 2.34314039e-01 6.78597212e-01
8.76344025e-01 7.09791481e-01 -1.78508148e-01 7.73225665e-01
4.41449106e-01 1.13465436e-01 9.16619778e-01 -1.52238107e+00
-3.57259333e-01 1.74621388e-01 -5.46501338e-01 -3.75245839e-01
-3.60550918e-02 -5.69992900e-01 8.43723640e-02 -1.27106464e+00
-1.19268250e+00 -2.88876474e-01 4.82859403e-01 6.25063032e-02
7.81316757e-01 -1.30540550e-01 -3.86951983e-01 9.63473916e-02
1.26886949e-01 6.25145972e-01 5.33805966e-01 3.58737826e-01
-1.78177774e-01 2.67524332e-01 8.05178657e-02 7.63935804e-01
8.32176745e-01 -5.94013035e-01 -3.55893731e-01 -1.47309244e-01
8.40798020e-01 5.96191764e-01 9.86977965e-02 -1.75260127e+00
3.93387794e-01 2.56096840e-01 2.42617831e-01 -5.57039678e-01
3.28529835e-01 -1.12847507e+00 8.30191553e-01 1.20279372e+00
6.69395268e-01 3.47862780e-01 5.33040285e-01 1.12463623e-01
-9.58149284e-02 -2.74371564e-01 7.32092917e-01 4.51181203e-01
-1.66029334e-01 -7.62616575e-01 -1.13157046e+00 -7.30763078e-01
1.47221088e+00 -1.26592159e-01 -6.82036102e-01 4.33114916e-01
-3.45603049e-01 5.09198666e-01 6.54263735e-01 1.67454720e-01
2.19758034e-01 -8.85122895e-01 -3.57871950e-01 4.16701227e-01
-4.90241021e-01 -8.73313472e-02 3.54476482e-01 9.83405888e-01
-1.24174428e+00 1.59179285e-01 -1.08791351e+00 -5.37847221e-01
-1.10467362e+00 -1.87245253e-02 5.96254826e-01 5.78225136e-01
-8.07334781e-01 5.39839327e-01 -7.39759624e-01 1.04639400e-02
-6.39350712e-01 -1.44475281e-01 -2.71878183e-01 1.17426813e-01
-1.70159638e-01 8.24947536e-01 5.29052794e-01 -1.97816312e-01
-6.57499015e-01 8.28036487e-01 7.43380845e-01 -3.91175002e-01
1.13157034e+00 -2.69992590e-01 -3.06338608e-01 4.27457422e-01
4.74346906e-01 7.62115359e-01 -6.29014075e-01 1.02982652e+00
-2.55009085e-01 -3.92201960e-01 5.18349707e-02 -6.48252964e-01
-7.76046932e-01 1.39009982e-01 2.18066216e-01 6.33275986e-01
1.10868633e+00 -5.09268701e-01 2.33520508e-01 3.57555151e-01
8.19290996e-01 -1.40098357e+00 -8.44811648e-02 2.78538495e-01
1.06508946e+00 1.32176727e-02 6.88056171e-01 -2.99909413e-01
-1.94179803e-01 1.48629940e+00 8.98415387e-01 -4.90946770e-01
4.83329982e-01 8.16241860e-01 -3.61119807e-01 -2.50828564e-01
-6.42958999e-01 4.23963755e-01 -4.06164825e-01 3.33737314e-01
2.75100708e-01 -1.49935409e-01 -9.09470201e-01 9.10746872e-01
-4.66930151e-01 -4.11176711e-01 1.03061032e+00 1.45745397e+00
-5.82958817e-01 -8.22561204e-01 -8.46256256e-01 7.84749240e-02
-4.38939154e-01 8.12854886e-01 -3.19913417e-01 1.08109868e+00
3.61508168e-02 8.85534286e-01 1.74265698e-01 -8.23657215e-01
1.03554595e+00 -3.90831977e-01 3.32667053e-01 -3.49697351e-01
-1.15561783e+00 -9.05474275e-02 8.67098629e-01 -4.90703195e-01
1.16587788e-01 -5.28130293e-01 -1.66897500e+00 -1.11966580e-01
-6.23863578e-01 9.77627218e-01 1.53607714e+00 3.67244214e-01
7.22250581e-01 9.39363122e-01 1.25285828e+00 -7.83403337e-01
-6.19992018e-01 -1.05772197e+00 -1.35544991e+00 -6.40998006e-01
-1.31210566e-01 -1.09550738e+00 -6.83530211e-01 -5.57511508e-01] | [5.427677154541016, 2.3380684852600098] |
b654a539-574a-4e4c-9585-6d51f31a1589 | implicit-bilevel-optimization-differentiating | 2302.14473 | null | https://arxiv.org/abs/2302.14473v1 | https://arxiv.org/pdf/2302.14473v1.pdf | Implicit Bilevel Optimization: Differentiating through Bilevel Optimization Programming | Bilevel Optimization Programming is used to model complex and conflicting interactions between agents, for example in Robust AI or Privacy-preserving AI. Integrating bilevel mathematical programming within deep learning is thus an essential objective for the Machine Learning community. Previously proposed approaches only consider single-level programming. In this paper, we extend existing single-level optimization programming approaches and thus propose Differentiating through Bilevel Optimization Programming (BiGrad) for end-to-end learning of models that use Bilevel Programming as a layer. BiGrad has wide applicability and can be used in modern machine learning frameworks. BiGrad is applicable to both continuous and combinatorial Bilevel optimization problems. We describe a class of gradient estimators for the combinatorial case which reduces the requirements in terms of computation complexity; for the case of the continuous variable, the gradient computation takes advantage of the push-back approach (i.e. vector-jacobian product) for an efficient implementation. Experiments show that the BiGrad successfully extends existing single-level approaches to Bilevel Programming. | ['Francesco Alesiani'] | 2023-02-28 | null | null | null | null | ['bilevel-optimization'] | ['methodology'] | [-3.70509475e-01 -1.11663684e-01 -3.73336494e-01 -2.49591887e-01
-7.91417897e-01 -5.57434738e-01 7.43155956e-01 4.22589481e-01
-7.82425225e-01 7.23101616e-01 -3.85529101e-01 -4.11374480e-01
-5.88834047e-01 -7.36379087e-01 -1.07059538e+00 -8.07483852e-01
-3.19463491e-01 9.07932818e-01 -2.79259264e-01 -1.09982930e-01
-4.16980609e-02 5.32697082e-01 -8.66160870e-01 -8.10569823e-02
7.28412330e-01 1.06143868e+00 -3.44365776e-01 7.32341647e-01
-1.45535797e-01 6.11703098e-01 -3.06374371e-01 -6.72373176e-01
8.01311374e-01 4.43715714e-02 -1.99601009e-01 -3.33196551e-01
6.54068649e-01 -1.07801050e-01 -6.58071339e-02 9.36304092e-01
4.69988674e-01 6.58518970e-02 4.07474995e-01 -1.65048635e+00
-2.55141765e-01 5.00815272e-01 -6.20796263e-01 -3.18967849e-01
-1.11340612e-01 2.22836837e-01 1.16086638e+00 -9.86979246e-01
4.35026109e-01 1.45507467e+00 8.13013613e-01 2.39066139e-01
-1.63211620e+00 -4.80301499e-01 3.29829633e-01 9.93299037e-02
-1.03937078e+00 -1.63662553e-01 6.55635893e-01 -6.62589848e-01
5.66560268e-01 5.31931341e-01 8.67760122e-01 5.64791858e-01
1.99229807e-01 1.05735302e+00 1.37210071e+00 -1.31047860e-01
3.72289091e-01 2.69380718e-01 6.35499477e-01 1.06151569e+00
4.47758198e-01 5.81035733e-01 -3.90910029e-01 -5.09594262e-01
4.12099689e-01 5.85678928e-02 -4.62033600e-02 -1.01409388e+00
-1.20690107e+00 1.21279061e+00 6.11206353e-01 -1.89249784e-01
-4.59512293e-01 8.08404326e-01 4.84619677e-01 6.58334494e-01
4.26954836e-01 6.50735557e-01 -3.87979239e-01 -6.74101263e-02
-9.55527186e-01 6.58790648e-01 1.40273404e+00 5.99239051e-01
6.63406312e-01 1.84146106e-01 -1.74694210e-01 7.59298146e-01
7.89887249e-01 4.37495232e-01 -2.27633134e-01 -7.96416283e-01
5.05628228e-01 5.12419999e-01 1.22801587e-01 -1.17138028e+00
-5.44232011e-01 -6.11175478e-01 -8.01356256e-01 6.59603715e-01
7.02825367e-01 -4.15278584e-01 -7.59730816e-01 1.65835321e+00
7.03242958e-01 1.22027911e-01 7.89979622e-02 1.00444806e+00
6.04274631e-01 8.18843722e-01 -1.39011979e-01 -3.54421556e-01
1.17414403e+00 -1.31529498e+00 -5.42047679e-01 -2.09400594e-01
3.05118531e-01 -3.17679942e-01 8.81770372e-01 5.50802946e-01
-1.19414151e+00 2.31063351e-01 -9.89669144e-01 -2.10671932e-01
-7.85253286e-01 1.39899841e-02 9.55316246e-01 8.56106639e-01
-1.06850004e+00 2.25654110e-01 -7.33563364e-01 2.71137416e-01
6.29098535e-01 8.10613155e-01 -2.76992410e-01 3.17229182e-02
-1.08592725e+00 9.07624364e-01 2.12220058e-01 4.31721777e-01
-1.28490126e+00 -1.00012362e+00 -9.45172668e-01 7.19975233e-02
7.35157609e-01 -8.84034753e-01 8.63432646e-01 -7.88811922e-01
-1.80457747e+00 6.50709510e-01 2.26079240e-01 -7.42768407e-01
1.28165174e+00 -1.47270516e-01 5.88482618e-01 -5.12099326e-01
-3.69301796e-01 2.32996047e-01 9.03919339e-01 -1.20015287e+00
-3.38120937e-01 -5.10574579e-01 3.73884797e-01 2.23110363e-01
-1.84128314e-01 -1.11043647e-01 -3.95664096e-01 -5.11725664e-01
-4.53530520e-01 -1.09015751e+00 -6.79267645e-01 3.71143281e-01
-2.44602859e-01 -1.78215802e-01 4.68974799e-01 -5.19311011e-01
9.85193729e-01 -1.68290579e+00 8.27008188e-01 6.47642136e-01
4.44438905e-01 2.12857544e-01 -2.17394561e-01 2.92018890e-01
2.43971542e-01 -9.33791138e-03 -4.24612939e-01 -7.23813653e-01
5.89211047e-01 2.40733027e-01 -3.22265588e-02 7.55482852e-01
-1.65628493e-01 1.10771394e+00 -6.54837906e-01 -2.79962361e-01
1.77619189e-01 4.52267945e-01 -7.71151841e-01 1.32162541e-01
-5.19995689e-01 5.52337524e-03 -3.74754697e-01 4.46780503e-01
7.12576926e-01 1.44143179e-01 -7.98806250e-02 -3.83998733e-03
-1.64783165e-01 -1.28729492e-01 -1.30720711e+00 1.39018607e+00
-7.74276376e-01 6.47607505e-01 7.66004145e-01 -1.07275391e+00
7.90872931e-01 -1.70691922e-01 6.11786783e-01 -3.08250904e-01
9.27470326e-02 4.18485969e-01 -9.95459408e-02 -1.50789842e-01
2.12547645e-01 1.44220458e-03 -2.81809300e-01 2.55361646e-01
-3.37897301e-01 -2.74898916e-01 3.87905627e-01 -8.57504457e-02
6.84475124e-01 -1.80466250e-01 2.86186963e-01 -5.11424601e-01
7.14966297e-01 2.68388242e-01 4.46633935e-01 9.35370743e-01
1.64561570e-01 6.79224208e-02 9.51713204e-01 -8.64926398e-01
-6.43106222e-01 -8.26193750e-01 3.20866294e-02 1.17074275e+00
-9.70737115e-02 -3.31813365e-01 -6.13934755e-01 -5.78236639e-01
5.81068456e-01 3.00937206e-01 -6.46933258e-01 2.28282571e-01
-7.14176476e-01 -1.35914123e+00 3.13966542e-01 -8.11690167e-02
1.51832804e-01 -6.77120388e-01 -6.53811991e-01 2.76187330e-01
4.90402937e-01 -5.82994640e-01 -5.81451714e-01 4.03914511e-01
-5.85660636e-01 -1.03397465e+00 -7.64422655e-01 -4.30287451e-01
5.49580514e-01 -2.37264305e-01 9.47417736e-01 -1.05426766e-01
-3.85488600e-01 5.26912510e-01 2.67933011e-01 -5.39656639e-01
-2.99810201e-01 6.35747612e-02 -4.38174233e-02 2.67270893e-01
2.20003974e-04 -2.64256120e-01 -1.65253997e-01 7.97408447e-02
-6.33608162e-01 -3.10119212e-01 4.85268325e-01 1.26009703e+00
7.30146050e-01 -4.76399302e-01 3.24881732e-01 -8.69352281e-01
1.13828039e+00 -5.39727747e-01 -1.56541824e+00 3.72668147e-01
-7.68451393e-01 2.51867861e-01 6.60675943e-01 -5.54132700e-01
-3.84529382e-01 1.24246031e-01 1.40343279e-01 -5.62624991e-01
6.04964197e-01 9.57105637e-01 4.02005240e-02 -5.93803465e-01
2.47477174e-01 8.16782713e-02 3.34434181e-01 -1.74660876e-01
5.53704858e-01 2.14266136e-01 1.34555697e-02 -7.78143823e-01
6.68594539e-01 2.46403486e-01 5.15498817e-01 -7.06224620e-01
-3.25651437e-01 -9.87171307e-02 -1.49289086e-01 -4.37073633e-02
6.51882827e-01 -7.31820941e-01 -1.41313672e+00 3.25353175e-01
-1.16489446e+00 -5.96446335e-01 -4.58796948e-01 4.78085518e-01
-6.24240041e-01 5.56936450e-02 -2.91381329e-01 -8.39399219e-01
-3.73043865e-01 -1.61162794e+00 7.70863175e-01 1.42568603e-01
3.33949864e-01 -1.36073101e+00 3.17325383e-01 4.10278022e-01
4.19182241e-01 4.56343353e-01 7.71415770e-01 -6.94287181e-01
-8.90397072e-01 -2.84132779e-01 -1.34005859e-01 -2.80068703e-02
-4.70619887e-01 -2.43624687e-01 -5.96935332e-01 -4.29644495e-01
-6.04109988e-02 -3.93418074e-01 9.45679069e-01 6.27715409e-01
9.23733175e-01 -7.76963532e-01 -7.15379342e-02 9.87214506e-01
1.52503765e+00 3.20448205e-02 1.31665066e-01 4.71862465e-01
9.28782523e-01 6.77155972e-01 4.86583501e-01 4.04262602e-01
6.73415482e-01 9.31064606e-01 7.34867573e-01 -2.48706505e-01
3.94121557e-01 5.56013249e-02 7.17128932e-01 3.62023503e-01
3.03230524e-01 -6.80896118e-02 -8.99405181e-01 5.24048209e-01
-2.31681371e+00 -6.71001256e-01 -2.30812922e-01 2.20395660e+00
9.16993558e-01 -2.78050423e-01 2.57756054e-01 -4.76449460e-01
2.33254150e-01 2.34268859e-01 -7.97040522e-01 -9.96695578e-01
1.51006073e-01 -1.50173411e-01 8.87269020e-01 1.07405162e+00
-1.14040697e+00 7.76851237e-01 5.77916479e+00 8.92383158e-01
-1.07806969e+00 2.90430278e-01 6.65217817e-01 -3.50497127e-01
-3.34830374e-01 -1.71608388e-01 -7.56772935e-01 2.69216329e-01
5.02459943e-01 -1.95426002e-01 9.30312574e-01 9.96927023e-01
2.79150456e-01 1.74367681e-01 -1.40540481e+00 1.23182344e+00
1.30923754e-02 -1.39247334e+00 -2.82548100e-01 4.48628694e-01
6.39908075e-01 2.12559029e-01 3.67426574e-01 2.70543873e-01
6.97426498e-01 -1.25598586e+00 5.18940449e-01 3.94317627e-01
2.76282966e-01 -7.61542737e-01 5.41404247e-01 3.57638717e-01
-8.29662025e-01 -2.23543137e-01 -1.79361358e-01 3.03587019e-01
1.25207797e-01 6.82707787e-01 -6.12602949e-01 3.22203308e-01
3.71881187e-01 4.60601896e-01 -1.94610745e-01 1.15027952e+00
5.19668795e-02 1.29641235e-01 -8.93965781e-01 -3.94788474e-01
7.29064703e-01 -7.80718446e-01 1.06197071e+00 1.33177376e+00
-6.29814565e-02 -5.03448844e-01 5.55924296e-01 1.00900793e+00
-2.08417013e-01 3.28849614e-01 -5.58598757e-01 -1.77704878e-02
1.51479070e-03 1.27189207e+00 -1.28500193e-01 -2.99252663e-03
-2.30225354e-01 4.36995476e-01 4.47981030e-01 1.29021466e-01
-7.59004176e-01 -1.71558142e-01 8.22670937e-01 -3.59787971e-01
1.06996097e-01 -4.38697219e-01 -4.01459455e-01 -1.18417919e+00
1.70782760e-01 -1.18430531e+00 5.53607523e-01 1.19384810e-01
-1.18212676e+00 3.10464531e-01 1.15338162e-01 -5.96937239e-01
-3.41584980e-01 -7.84151256e-01 -4.09610271e-01 6.50506556e-01
-1.83956981e+00 -1.40053380e+00 -5.77432401e-02 5.53909540e-01
4.37143475e-01 -3.69750440e-01 4.66411054e-01 3.05878282e-01
-6.09826684e-01 7.39447057e-01 4.39720631e-01 -1.45618156e-01
3.11846942e-01 -1.54974902e+00 -1.35246456e-01 7.37677813e-01
2.09994629e-01 6.17005765e-01 6.92648590e-01 -3.37832361e-01
-2.23597622e+00 -1.00729501e+00 5.13102770e-01 -2.70697445e-01
7.66799569e-01 -6.40352666e-01 -3.72407466e-01 5.90631366e-01
8.38846266e-02 2.31550470e-01 3.58966142e-01 9.47069377e-02
-2.69801348e-01 -5.84755301e-01 -1.32992446e+00 7.51854062e-01
3.75608295e-01 -9.08369124e-02 2.69912705e-02 3.69653195e-01
4.98713076e-01 -4.74300653e-01 -1.05665290e+00 2.43554227e-02
7.41986632e-01 -4.16135997e-01 1.21164620e+00 -6.16023362e-01
-8.92406702e-02 -4.60182071e-01 -1.52672440e-01 -1.48151803e+00
2.20225424e-01 -1.06868136e+00 -5.01119554e-01 7.68861413e-01
5.68309963e-01 -9.68864679e-01 7.73894966e-01 9.04277682e-01
2.70896584e-01 -9.39392865e-01 -1.15562010e+00 -6.55910015e-01
3.24603260e-01 -1.90352082e-01 4.04225081e-01 8.67182553e-01
-1.28020465e-01 2.08846852e-02 -8.68209541e-01 1.30449146e-01
1.05323052e+00 2.27404967e-01 1.22295666e+00 -1.06867683e+00
-5.89879930e-01 -7.70482302e-01 -3.02809983e-01 -1.02527368e+00
2.50036687e-01 -1.04962552e+00 1.91878434e-02 -1.20636475e+00
-7.66204223e-02 -9.28462446e-01 -5.42465262e-02 4.02645856e-01
2.77519841e-02 -6.02425970e-02 7.22043216e-01 -1.62726566e-01
-3.88910294e-01 5.90545475e-01 9.58462715e-01 -5.27717173e-01
-4.96773660e-01 1.52979851e-01 -5.95946252e-01 6.10103607e-01
6.26706123e-01 -5.75020194e-01 -2.45064080e-01 -6.63905859e-01
5.71915448e-01 4.40808311e-02 5.61955273e-01 -5.15353978e-01
6.64330065e-01 -5.45910120e-01 -2.46283576e-01 -2.31994495e-01
6.02046132e-01 -1.30708838e+00 3.18533719e-01 8.34054291e-01
-6.62801206e-01 5.96809275e-02 1.96998820e-01 5.62886894e-01
-1.54596269e-01 -2.79015332e-01 8.95011127e-01 -8.50925893e-02
-2.13662252e-01 6.12687826e-01 2.36422513e-02 -1.09043963e-01
1.36794317e+00 1.67465776e-01 -2.77264565e-01 -3.44526291e-01
-6.96915746e-01 9.73416209e-01 1.57978535e-01 2.28254765e-01
7.07765043e-01 -1.07417679e+00 -9.99353647e-01 4.32086810e-02
-2.14822710e-01 8.40942487e-02 -3.86196882e-01 1.17770290e+00
-6.69846773e-01 3.29827964e-01 4.34633084e-02 -4.83842969e-01
-1.40334487e+00 5.21579027e-01 7.76435435e-01 -8.06617200e-01
-7.81432465e-02 8.18500102e-01 2.12919608e-01 -8.26945543e-01
5.65389335e-01 -2.46345520e-01 5.21328393e-03 3.35168034e-01
2.93920815e-01 6.00192368e-01 -3.26727659e-01 -3.18817854e-01
-4.28674996e-01 7.59001017e-01 -1.56628460e-01 -2.41883233e-01
1.82444942e+00 1.53874904e-01 -5.55702269e-01 3.32348078e-01
1.31220877e+00 -8.61847997e-02 -1.18613183e+00 -4.88887094e-02
3.53908986e-02 -3.41402709e-01 3.42189044e-01 -7.85114288e-01
-1.28827810e+00 1.05199885e+00 6.20413721e-01 7.30965734e-02
6.34006500e-01 -3.60565096e-01 2.72641152e-01 7.32865632e-01
2.99399257e-01 -8.84628832e-01 -2.01194838e-01 4.50007856e-01
1.17953646e+00 -1.31117523e+00 3.26338857e-01 -1.19396433e-01
-4.92328435e-01 1.18432105e+00 3.11882794e-01 -3.39134842e-01
7.59728193e-01 4.17333812e-01 -1.88761637e-01 -1.92979574e-01
-8.04466784e-01 5.74937724e-02 2.60180384e-01 3.75679970e-01
-2.19423309e-01 1.44870743e-01 -5.89536011e-01 2.24262252e-01
1.11057773e-01 1.51236318e-02 4.19992983e-01 5.72111309e-01
1.29347015e-02 -1.33273780e+00 -5.03917992e-01 3.51060778e-01
-5.79556465e-01 -1.16538383e-01 -4.16004270e-01 6.12711668e-01
-1.03455871e-01 6.18810177e-01 -2.41514757e-01 1.31130263e-01
2.35532492e-01 -3.78473490e-01 2.74566203e-01 -1.41272157e-01
-1.13469517e+00 -7.07517564e-02 5.63432015e-02 -6.73593581e-01
-1.83175817e-01 -6.29127979e-01 -6.65369093e-01 -2.29380354e-01
-2.08237499e-01 1.36386707e-01 1.18009412e+00 5.67165494e-01
2.54045397e-01 9.72956643e-02 5.66556454e-01 -7.70329535e-01
-9.81860161e-01 -1.53726667e-01 -1.91766992e-01 -2.09269106e-01
9.68723714e-01 -4.64688152e-01 -1.41550511e-01 -3.99732769e-01] | [6.6822075843811035, 4.3949432373046875] |
a1188305-a616-4b7c-a015-15f4bee8bc6d | multi-task-regression-based-learning-for | 1907.08320 | null | https://arxiv.org/abs/1907.08320v1 | https://arxiv.org/pdf/1907.08320v1.pdf | Multi-Task Regression-based Learning for Autonomous Unmanned Aerial Vehicle Flight Control within Unstructured Outdoor Environments | Increased growth in the global Unmanned Aerial Vehicles (UAV) (drone) industry has expanded possibilities for fully autonomous UAV applications. A particular application which has in part motivated this research is the use of UAV in wide area search and surveillance operations in unstructured outdoor environments. The critical issue with such environments is the lack of structured features that could aid in autonomous flight, such as road lines or paths. In this paper, we propose an End-to-End Multi-Task Regression-based Learning approach capable of defining flight commands for navigation and exploration under the forest canopy, regardless of the presence of trails or additional sensors (i.e. GPS). Training and testing are performed using a software in the loop pipeline which allows for a detailed evaluation against state-of-the-art pose estimation techniques. Our extensive experiments demonstrate that our approach excels in performing dense exploration within the required search perimeter, is capable of covering wider search regions, generalises to previously unseen and unexplored environments and outperforms contemporary state-of-the-art techniques. | ['Toby P. Breckon', 'Amir Atapour-Abarghouei', 'Bruna G. Maciel-Pearson', 'Samet Akcay', 'Christopher Holder'] | 2019-07-18 | null | null | null | null | ['autonomous-flight-dense-forest'] | ['computer-vision'] | [ 3.99016827e-01 -2.69486785e-01 1.81066468e-01 -2.01343015e-01
-2.95616090e-01 -1.07295787e+00 5.31612396e-01 1.33206114e-01
-5.82105637e-01 9.00377452e-01 -5.02229214e-01 -4.54187810e-01
-6.79807365e-01 -8.67329001e-01 -5.29034019e-01 -2.60028034e-01
-7.85177290e-01 7.52094686e-01 7.82943487e-01 -8.35936725e-01
1.90899093e-02 1.01725674e+00 -1.86292398e+00 -2.34583035e-01
5.65274715e-01 1.09016788e+00 4.20526892e-01 1.01175296e+00
5.75129807e-01 1.03201814e-01 -5.17670453e-01 2.95292169e-01
7.03407824e-01 6.95461482e-02 -6.76149309e-01 1.01649292e-01
6.42263830e-01 -3.68366838e-01 1.45845845e-01 6.19334400e-01
2.52071261e-01 4.40633208e-01 3.37769091e-01 -1.09344983e+00
5.74020326e-01 -1.83379412e-01 -2.43275821e-01 2.05477685e-01
4.30421412e-01 2.74345726e-01 7.31468499e-01 -3.98873776e-01
5.90632141e-01 7.39128053e-01 7.65385926e-01 -1.49183705e-01
-9.98277068e-01 -3.60290468e-01 3.25613827e-01 -1.44051105e-01
-1.42425942e+00 -9.31802019e-02 1.30087733e-01 -4.56890851e-01
1.14838028e+00 3.01707536e-01 9.52870250e-01 7.81619847e-01
5.16961992e-01 2.04113364e-01 9.22426164e-01 -2.86588132e-01
1.99793637e-01 -1.27417937e-01 -6.45214975e-01 1.06360304e+00
5.40532589e-01 7.05199897e-01 -2.88065553e-01 -5.22239618e-02
7.69525230e-01 -1.27438888e-01 -4.03270274e-01 -1.01395905e+00
-1.40101361e+00 8.12879086e-01 8.59288454e-01 -8.27594846e-02
-6.03687167e-01 -4.90644760e-02 2.49187216e-01 3.86292994e-01
1.70668229e-01 9.17253971e-01 -8.94033492e-01 -1.18370898e-01
-1.22040093e+00 6.53196096e-01 9.36146498e-01 1.10747921e+00
9.17365253e-01 1.31471112e-01 4.21393096e-01 3.42683643e-01
3.15100402e-01 6.23297811e-01 7.88649172e-03 -7.17463970e-01
2.39900723e-01 7.13235915e-01 5.00538051e-01 -7.82053292e-01
-6.75955892e-01 -6.76611543e-01 -3.84609967e-01 7.85650074e-01
4.37344201e-02 -7.19074786e-01 -9.18959737e-01 1.01995397e+00
5.85187435e-01 -2.04583287e-01 1.73328593e-01 9.53264236e-01
3.52857500e-01 3.14258248e-01 -4.06141043e-01 -2.45888550e-02
1.17345214e+00 -8.97436321e-01 -8.22471380e-02 -8.68759990e-01
4.96678889e-01 -8.18341553e-01 5.16425312e-01 6.32133782e-01
-3.16406727e-01 -6.11931443e-01 -1.24971831e+00 4.48354870e-01
-7.67534912e-01 4.70775574e-01 8.02634060e-01 5.37196457e-01
-1.10588586e+00 5.42920172e-01 -8.59465182e-01 -8.41399848e-01
2.93087095e-01 7.26125956e-01 -4.53132927e-01 -2.06683408e-02
-9.03087437e-01 1.17221522e+00 5.89040875e-01 2.28257582e-01
-1.27295697e+00 -2.96598494e-01 -1.01771975e+00 -4.48000133e-01
8.83293211e-01 -5.97659171e-01 1.14708877e+00 -5.18448234e-01
-1.43565524e+00 5.69679141e-01 2.94862032e-01 -8.36361945e-01
5.37657380e-01 -8.32644582e-01 -2.67852604e-01 4.83787656e-02
1.53326616e-01 9.84312415e-01 9.15148377e-01 -1.08752787e+00
-1.21270442e+00 -5.06869614e-01 2.62377650e-01 4.10882175e-01
2.75824070e-01 -3.85725141e-01 3.14368010e-02 -3.47395837e-01
5.19258641e-02 -1.48263574e+00 -6.13592029e-01 1.81014150e-01
-7.80528188e-02 4.25884455e-01 1.32476997e+00 -2.75269866e-01
9.29688632e-01 -1.48733711e+00 3.50968450e-01 3.33317280e-01
-2.93724447e-01 3.33068967e-01 1.69365287e-01 8.22305620e-01
4.23751324e-01 -2.21685454e-01 -3.84472638e-01 2.93679684e-01
-4.40668732e-01 2.68295825e-01 -4.07117814e-01 5.28154731e-01
6.85486495e-02 4.67021197e-01 -8.57549667e-01 -1.16498403e-01
6.17236972e-01 3.54372472e-01 -2.24616721e-01 1.62729189e-01
-5.87388515e-01 5.06716967e-01 -6.69877172e-01 1.05843985e+00
4.42021787e-01 4.47440594e-01 -2.35519670e-02 1.89919025e-01
-8.07255864e-01 -1.29487216e-01 -1.26918113e+00 1.91820312e+00
-4.84121621e-01 8.24134231e-01 5.52321255e-01 -3.88561338e-01
1.18006921e+00 -1.78979725e-01 2.32484087e-01 9.92075279e-02
1.82040006e-01 3.56532484e-01 -7.37509578e-02 -8.76474902e-02
9.19064224e-01 3.36892247e-01 -1.14042491e-01 -2.79756814e-01
8.53850693e-02 -8.73125255e-01 1.41304716e-01 -4.37240571e-01
1.11679339e+00 6.92207575e-01 8.43121409e-01 -4.46356654e-01
6.02347791e-01 8.08394909e-01 4.54865582e-02 5.28267682e-01
-8.66047144e-02 7.76587129e-02 -1.77499965e-01 -7.47068942e-01
-4.33233112e-01 -7.22277939e-01 -1.77675128e-01 9.36290622e-01
3.93040031e-01 -4.87308592e-01 -2.58567929e-01 -5.67150235e-01
1.92069009e-01 4.45526987e-01 -4.59271789e-01 2.49323770e-01
-2.94079661e-01 -4.63590086e-01 3.89536262e-01 1.71297595e-01
5.80208242e-01 -8.42755377e-01 -1.97282350e+00 4.09267873e-01
1.97002530e-01 -1.17620373e+00 4.83515143e-01 5.48264027e-01
-9.72839773e-01 -1.21608078e+00 -2.38421530e-01 -3.30749661e-01
3.14591140e-01 5.89392662e-01 8.87276709e-01 -2.93045819e-01
-6.37443662e-01 7.23153889e-01 -7.22328842e-01 -7.53141463e-01
1.28023475e-01 3.37370276e-01 1.18464634e-01 -5.96551299e-01
4.66600545e-02 -4.07208592e-01 -2.89798439e-01 6.74529433e-01
-5.25156319e-01 -3.77717346e-01 9.24557269e-01 6.04585826e-01
7.76189268e-01 3.73869479e-01 -1.21602267e-01 -4.30793822e-01
4.85695481e-01 -3.03715855e-01 -1.29756713e+00 1.12485483e-01
-3.87046844e-01 -1.20052502e-01 5.33924401e-01 -8.24158415e-02
-2.47325972e-01 6.84634089e-01 2.89974868e-01 -3.36017668e-01
-6.49152875e-01 5.21063030e-01 7.37834647e-02 -9.36949134e-01
7.20127285e-01 -7.37719908e-02 2.46100575e-02 -7.01318681e-02
1.89841077e-01 3.90644759e-01 3.48140597e-01 -1.98371261e-01
1.10625875e+00 4.03660864e-01 7.44675517e-01 -1.43204403e+00
-5.26812851e-01 -6.05284631e-01 -1.23525441e+00 -3.98433059e-01
6.01481497e-01 -9.72347796e-01 -2.90831476e-01 -2.03671884e-02
-7.64461160e-01 -4.16463345e-01 -1.49248242e-01 5.04804373e-01
-6.45206213e-01 -2.86707073e-03 6.02157891e-01 -9.65731025e-01
-2.09270597e-01 -1.25096798e+00 1.35777867e+00 2.96233535e-01
-2.47641802e-01 -8.77086937e-01 2.67612696e-01 -1.03801765e-01
4.46746767e-01 9.11678970e-01 1.69724762e-01 -3.85330856e-01
-1.00539017e+00 -6.46676719e-01 3.09639335e-01 -1.23515829e-01
5.34128211e-02 9.01014637e-03 -7.18031049e-01 -6.02400541e-01
-5.02593696e-01 -4.97405708e-01 8.23241115e-01 4.04836923e-01
5.01041889e-01 8.15323740e-03 -8.09930801e-01 8.05088103e-01
1.64366269e+00 2.20419660e-01 1.47588536e-01 7.22872972e-01
1.47075012e-01 7.29281127e-01 1.56242776e+00 5.61002731e-01
2.49334909e-02 8.10411751e-01 1.24276805e+00 -2.63316203e-02
6.67851329e-01 -1.37455508e-01 2.68923551e-01 -6.49396896e-01
-3.30835491e-01 -5.46374440e-01 -1.02332103e+00 5.24122417e-01
-1.71886837e+00 -5.02017379e-01 3.76386419e-02 2.40039229e+00
-1.74250200e-01 2.16485545e-01 -6.59777448e-02 -1.30487427e-01
1.33818880e-01 4.33728486e-01 -4.51269478e-01 -3.57028723e-01
2.38516614e-01 1.64636761e-01 1.31337452e+00 5.85822284e-01
-1.75212038e+00 1.36870885e+00 5.94691610e+00 2.97699392e-01
-9.28173661e-01 -5.00743806e-01 -5.06275296e-01 6.93302155e-02
5.01322687e-01 2.12907150e-01 -1.02363181e+00 -3.40573311e-01
8.38013411e-01 4.84265506e-01 5.81491828e-01 1.12650740e+00
-8.33186731e-02 -6.88675284e-01 -6.83533013e-01 4.28257227e-01
-1.04446150e-02 -1.03501403e+00 -1.16934709e-01 3.60152543e-01
4.06518281e-01 4.48439837e-01 -1.35971889e-01 1.39905140e-01
3.83738071e-01 -8.43385160e-01 5.59164405e-01 2.38352701e-01
8.64481688e-01 -8.49100590e-01 8.43835294e-01 5.78678191e-01
-1.53388155e+00 -2.69849539e-01 -4.24500525e-01 -2.82400101e-01
1.75946981e-01 -4.62064520e-02 -1.34969938e+00 9.31315541e-01
6.80699587e-01 5.17889142e-01 -8.04028869e-01 1.38423264e+00
-3.39333683e-01 1.92996800e-01 -7.68467069e-01 -1.99480221e-01
9.91351366e-01 -3.11723262e-01 1.06554651e+00 9.86575723e-01
5.10264218e-01 -1.20121144e-01 8.85427713e-01 2.98119783e-01
5.51490545e-01 -6.34486228e-02 -1.21400213e+00 -7.91528746e-02
2.72275031e-01 1.34627318e+00 -9.00885046e-01 3.49920005e-01
-5.89004271e-02 8.59955430e-01 -1.10036641e-01 -7.27655441e-02
-5.67432940e-01 -5.86060166e-01 7.64129996e-01 3.20802718e-01
5.99953294e-01 -8.79309118e-01 3.82468075e-01 -4.44786906e-01
-1.43101186e-01 -6.70189917e-01 3.39143351e-02 -6.39986753e-01
-3.67590457e-01 1.09508908e+00 3.07342738e-01 -1.62037754e+00
-5.61298013e-01 -8.64175141e-01 -4.94548321e-01 7.30616987e-01
-1.66241515e+00 -1.64492238e+00 -8.45399797e-01 1.60586685e-01
7.60547101e-01 -7.05685973e-01 1.17517960e+00 -5.09980619e-01
7.95029849e-02 -3.32400829e-01 -2.04025239e-01 -4.93150741e-01
5.03184795e-01 -1.09015822e+00 3.15786093e-01 1.11243677e+00
2.87995130e-01 1.91802904e-01 8.92481983e-01 -1.12873280e+00
-1.50624990e+00 -1.17988718e+00 1.43620804e-01 -4.99811441e-01
5.95409870e-01 -2.11971164e-01 -1.84019387e-01 7.50199497e-01
3.39059830e-02 -6.54706061e-02 3.95829886e-01 1.95625395e-01
3.97508353e-01 -1.30805999e-01 -1.06637490e+00 2.48679385e-01
8.44521701e-01 1.06589481e-01 -2.60183513e-01 5.02833307e-01
4.87705350e-01 -7.49009371e-01 -5.59368432e-01 8.91215801e-01
8.21142733e-01 -1.14002621e+00 9.34900582e-01 -4.36020911e-01
-1.28461167e-01 -5.94846189e-01 -3.88728499e-01 -1.46194696e+00
-1.83833659e-01 -9.21988189e-01 1.23914950e-01 4.82875377e-01
4.48888093e-01 -3.90072137e-01 7.30589807e-01 -1.56293824e-01
-3.24670285e-01 -7.74338126e-01 -8.54821980e-01 -8.23618293e-01
-7.03774869e-01 -9.58939493e-02 2.52196163e-01 1.67583719e-01
-6.65825546e-01 2.56279022e-01 -4.07534719e-01 9.49259341e-01
3.20663601e-01 1.98732182e-01 1.16816390e+00 -1.75830019e+00
4.69201580e-02 1.70493685e-03 -8.18098664e-01 -8.26064467e-01
7.74259046e-02 -3.07838470e-01 2.59793133e-01 -1.65196490e+00
-1.18070459e+00 -3.88691515e-01 4.60447371e-01 4.67236429e-01
4.00975674e-01 -8.41588676e-02 -9.43789706e-02 6.78148121e-02
-4.44298118e-01 4.94286835e-01 9.33558702e-01 2.60081738e-01
-4.73511428e-01 6.20532513e-01 -1.86962575e-01 7.33719945e-01
7.34517395e-01 -2.35909313e-01 -5.98840356e-01 -2.63085127e-01
7.39163905e-02 1.79974234e-03 4.95225370e-01 -1.81519055e+00
4.12081093e-01 -1.15326077e-01 4.26722705e-01 -9.40769613e-01
7.33642459e-01 -1.42529917e+00 -8.31163861e-03 6.17111444e-01
2.89774925e-01 4.84188110e-01 6.31601691e-01 7.48120368e-01
-2.45979145e-01 -2.86426216e-01 6.70888722e-01 -4.40232962e-01
-1.33089733e+00 2.27689579e-01 -5.66989779e-01 -2.20124766e-01
1.54359746e+00 -6.41276896e-01 8.78729671e-02 -1.29951149e-01
-4.91659522e-01 5.92823267e-01 6.79422855e-01 4.70122993e-01
7.58414030e-01 -3.42551887e-01 -4.95350927e-01 7.44098961e-01
2.08715484e-01 3.60862404e-01 -1.64680988e-01 4.91060615e-01
-1.16000307e+00 8.39026928e-01 -6.65115833e-01 -6.32321656e-01
-1.46438611e+00 1.87599003e-01 3.70597154e-01 -9.04656351e-02
-4.08491820e-01 8.89986873e-01 -4.77755487e-01 -4.81422812e-01
-3.73614021e-03 -5.51623404e-01 -2.50923038e-01 1.28649637e-01
2.45121449e-01 4.24856454e-01 1.47327721e-01 -7.36340880e-01
-5.48915863e-01 8.41275632e-01 1.97533831e-01 -1.62452638e-01
1.37369704e+00 1.27966674e-02 2.48316973e-01 1.70917243e-01
3.18450987e-01 9.61465761e-02 -1.48050833e+00 3.35549712e-01
-2.00989507e-02 -6.74597919e-01 5.76155424e-01 -1.05269039e+00
-3.32599670e-01 6.81403458e-01 7.89489508e-01 1.61023080e-01
1.01753509e+00 -4.83275145e-01 4.39790100e-01 9.89248514e-01
1.14573169e+00 -7.86097646e-01 -4.26368177e-01 7.20043421e-01
1.00546396e+00 -1.37210572e+00 4.84200120e-01 -4.70202476e-01
-8.67858410e-01 1.47736323e+00 5.87944090e-01 -3.27641726e-01
4.26758885e-01 2.72722542e-01 -5.11105321e-02 -2.89004892e-01
-5.94927669e-01 -7.18107402e-01 3.04267913e-01 1.05595863e+00
-1.34361774e-01 9.26197544e-02 3.15583915e-01 -1.12714224e-01
-4.83548075e-01 -2.29752332e-01 4.31939840e-01 1.42601597e+00
-1.06041157e+00 -9.13501441e-01 -3.90276909e-01 3.80039334e-01
1.64496556e-01 9.25332755e-02 -7.88505435e-01 1.40361011e+00
4.92664307e-01 7.74514496e-01 -2.51694411e-01 -4.91704315e-01
6.93290591e-01 -2.11545646e-01 6.12426460e-01 -8.07852209e-01
-7.34155416e-01 -5.76719269e-02 3.57695520e-01 -6.95926368e-01
-2.48333693e-01 -8.25401962e-01 -6.94769323e-01 4.61393952e-01
-5.79759836e-01 1.30896419e-01 9.91701841e-01 6.30981028e-01
3.66524339e-01 4.92468894e-01 2.37120256e-01 -1.27193737e+00
-5.22953570e-01 -7.26320386e-01 -4.48789358e-01 -9.19475138e-01
6.21316135e-01 -1.13411224e+00 -5.94852604e-02 -3.18634450e-01] | [7.329524993896484, -1.9234261512756348] |
2af2903a-9cd0-43a7-97dc-3e4b8c6ecb20 | zero-shot-learning-for-code-education-rubric | 1809.01357 | null | http://arxiv.org/abs/1809.01357v2 | http://arxiv.org/pdf/1809.01357v2.pdf | Zero Shot Learning for Code Education: Rubric Sampling with Deep Learning Inference | In modern computer science education, massive open online courses (MOOCs) log
thousands of hours of data about how students solve coding challenges. Being so
rich in data, these platforms have garnered the interest of the machine
learning community, with many new algorithms attempting to autonomously provide
feedback to help future students learn. But what about those first hundred
thousand students? In most educational contexts (i.e. classrooms), assignments
do not have enough historical data for supervised learning. In this paper, we
introduce a human-in-the-loop "rubric sampling" approach to tackle the "zero
shot" feedback challenge. We are able to provide autonomous feedback for the
first students working on an introductory programming assignment with accuracy
that substantially outperforms data-hungry algorithms and approaches human
level fidelity. Rubric sampling requires minimal teacher effort, can associate
feedback with specific parts of a student's solution and can articulate a
student's misconceptions in the language of the instructor. Deep learning
inference enables rubric sampling to further improve as more assignment
specific student data is acquired. We demonstrate our results on a novel
dataset from Code.org, the world's largest programming education platform. | ['Noah Goodman', 'Milan Mosse', 'Mike Wu', 'Chris Piech'] | 2018-09-05 | null | null | null | null | ['misconceptions'] | ['miscellaneous'] | [-5.57524078e-02 1.54623806e-01 -2.32752010e-01 -4.38148826e-01
-7.11541295e-01 -9.32741582e-01 2.77940810e-01 7.96488643e-01
-1.23833492e-01 4.06038523e-01 -2.88869925e-02 -8.02090228e-01
-1.11062095e-01 -8.64879966e-01 -7.49987185e-01 -5.52785993e-02
3.34840566e-01 4.20308739e-01 2.35074759e-01 -7.34096467e-01
6.48964882e-01 2.91775852e-01 -1.85325229e+00 5.16110539e-01
1.13067603e+00 2.96542734e-01 4.18521047e-01 8.53286922e-01
-3.08219939e-01 1.56330359e+00 -7.13247359e-01 -3.70784134e-01
2.29825377e-02 -3.09463471e-01 -1.19400251e+00 -2.61103094e-01
1.07571828e+00 -4.58909303e-01 -2.44124338e-01 1.16400552e+00
9.74432230e-02 3.81487429e-01 1.61099747e-01 -1.02859592e+00
-6.92323029e-01 7.99718618e-01 -3.26824158e-01 2.44306341e-01
8.35371077e-01 2.66111523e-01 9.98653173e-01 -7.69844353e-01
7.65706420e-01 7.51169503e-01 4.96862680e-01 6.38941228e-01
-1.21511853e+00 -5.09882629e-01 -4.72901054e-02 2.13559031e-01
-8.91512394e-01 -4.25404124e-02 5.63762128e-01 -1.07531738e+00
4.16864961e-01 1.74193516e-01 1.04561460e+00 6.39508009e-01
7.27210939e-02 9.19319034e-01 9.10089672e-01 -5.25413156e-01
9.53367949e-02 6.09621227e-01 4.96030897e-01 1.10097122e+00
1.48399770e-01 -1.61066160e-01 -6.74351871e-01 6.23169988e-02
4.46535617e-01 3.98102820e-01 -2.39487439e-01 -5.45081556e-01
-9.83516693e-01 1.01886225e+00 2.77771622e-01 4.58281189e-02
1.84486166e-01 1.56418800e-01 1.94806084e-01 7.68784165e-01
-1.68165430e-01 7.65233994e-01 -4.48545754e-01 -8.01308393e-01
-1.07326448e+00 6.11072183e-01 1.15357220e+00 1.11878812e+00
8.54530573e-01 2.91837621e-02 2.53998280e-01 7.92134166e-01
1.94280192e-01 5.03819138e-02 7.14693427e-01 -1.32659328e+00
2.65703559e-01 1.00756884e+00 -1.98591799e-01 -7.41129100e-01
3.41120601e-01 -5.06683826e-01 -1.95140347e-01 5.88317513e-01
6.76832259e-01 -2.14252457e-01 -6.42904937e-01 1.40891361e+00
3.60436201e-01 2.14112818e-01 -2.17450857e-01 7.90628076e-01
1.03502822e+00 4.66789901e-01 -1.52857646e-01 2.00592637e-01
1.36620605e+00 -8.68345797e-01 -3.81115019e-01 -1.02627732e-01
9.34345603e-01 -9.84968603e-01 1.30325711e+00 1.10749996e+00
-1.29882002e+00 -6.64524674e-01 -1.03555036e+00 -3.91212374e-01
-2.21041605e-01 -4.06384438e-01 6.79937124e-01 7.29220688e-01
-9.91245985e-01 7.73322165e-01 -5.32122493e-01 9.24207121e-02
5.80147743e-01 1.78524613e-01 -7.43298754e-02 -4.48188365e-01
-7.18355596e-01 7.04870701e-01 -1.26123324e-01 -8.54436338e-01
-1.28713524e+00 -1.51025975e+00 -7.50869691e-01 5.05882382e-01
4.10792947e-01 -2.47892201e-01 1.87865424e+00 -8.14054608e-01
-1.59317899e+00 1.01251185e+00 2.22123012e-01 -7.21603781e-02
5.66503644e-01 -2.63252556e-01 4.27353472e-01 -9.66808200e-02
-3.68912257e-02 5.97960114e-01 2.52695113e-01 -7.89469600e-01
-4.68500108e-01 -1.68882713e-01 4.91563201e-01 4.35837246e-02
-4.47100163e-01 -8.28115717e-02 1.74718216e-01 -2.93228716e-01
2.44299039e-01 -7.73008883e-01 -1.89291701e-01 1.14827111e-01
4.82072324e-01 -7.33709633e-01 5.60230911e-01 -2.27649420e-01
1.30583024e+00 -1.92036021e+00 2.19002403e-02 -5.30973971e-02
5.85026026e-01 2.84145743e-01 1.99564576e-01 4.51508194e-01
-2.22318113e-01 -3.88319255e-03 1.39852822e-01 1.01175047e-01
1.63274810e-01 -5.47027439e-02 -4.46191996e-01 1.06421858e-01
-2.79140651e-01 4.74905610e-01 -1.60644352e+00 -3.79902989e-01
2.18422979e-01 -6.56847656e-02 -1.05581951e+00 6.24838591e-01
-2.90743440e-01 5.25311939e-02 -2.32303038e-01 4.28272992e-01
1.07719935e-01 -2.88364202e-01 -3.58506888e-02 9.44139719e-01
-1.80290297e-01 7.33456254e-01 -1.19453406e+00 2.05012465e+00
-8.54628682e-01 1.01909304e+00 1.84426486e-01 -8.66401792e-01
1.08575976e+00 4.01839197e-01 3.30981314e-01 -1.87029809e-01
-1.02939628e-01 1.33651882e-01 2.76549846e-01 -5.22905529e-01
6.99779391e-01 1.49888815e-02 1.15544915e-01 8.32812667e-01
3.71224105e-01 -7.21890926e-01 4.53628898e-01 7.21731365e-01
1.43423748e+00 1.35781005e-01 1.83287635e-01 -3.82819772e-01
2.62234330e-01 2.16008514e-01 2.71960407e-01 7.56091893e-01
-3.26592296e-01 3.88059616e-01 4.78743315e-01 -9.02444661e-01
-7.47418642e-01 -9.46917117e-01 1.17527358e-01 1.88397777e+00
-3.37543160e-01 -6.60204887e-01 -8.51044357e-01 -4.18870836e-01
2.95119621e-02 5.80808640e-01 -2.47158602e-01 4.32607543e-04
-3.63452852e-01 2.88543820e-01 2.30675310e-01 3.65413308e-01
6.96470514e-02 -9.73302484e-01 -8.84016871e-01 3.73065799e-01
1.90818608e-02 -8.36284101e-01 -3.76611292e-01 2.57483721e-01
-8.94842327e-01 -8.99047136e-01 -1.55277401e-01 -8.77549469e-01
8.80549133e-01 4.33030576e-01 1.61559725e+00 7.44493544e-01
-6.22973919e-01 6.36779726e-01 -1.47652045e-01 -6.75314188e-01
-6.22831881e-01 -5.29811494e-02 -1.72377732e-02 -8.22487056e-01
8.66818011e-01 -9.20990229e-01 -3.37904245e-01 -1.61324099e-01
-6.25936806e-01 1.97274789e-01 2.71095455e-01 6.22911036e-01
-1.04799882e-01 -2.70600200e-01 3.10600996e-01 -1.40053880e+00
6.97177410e-01 -6.57106459e-01 -9.04531896e-01 -1.76327918e-02
-7.11151481e-01 1.05432853e-01 8.36874306e-01 -4.79810596e-01
-6.44656360e-01 3.57001983e-02 -3.99433896e-02 -4.56305534e-01
-3.70820105e-01 6.31910801e-01 4.82028574e-01 -3.10686976e-01
1.13239360e+00 1.30404040e-01 -1.71057612e-01 -3.49308640e-01
4.93646115e-02 4.83674109e-01 6.39750779e-01 -1.29985595e+00
7.39057362e-01 -1.60397351e-01 -3.08899820e-01 -6.21253908e-01
-1.17553973e+00 -7.23473012e-01 -5.93552887e-01 -5.50250292e-01
3.64017844e-01 -1.05408669e+00 -1.43454921e+00 9.55995098e-02
-1.20956445e+00 -6.58443153e-01 -3.83744180e-01 2.80774742e-01
-4.13385659e-01 -1.53634176e-01 -7.76395857e-01 -5.00085890e-01
2.16205314e-01 -1.52373362e+00 1.85615644e-01 6.64791405e-01
-5.19839585e-01 -1.00593066e+00 4.39763099e-01 9.49805677e-01
3.61871690e-01 -8.57228264e-02 8.23021114e-01 -6.06718302e-01
-9.06765103e-01 -9.20211747e-02 2.35276252e-01 2.24370331e-01
-3.52462769e-01 1.10370502e-01 -1.04904497e+00 -1.42118618e-01
-2.75094025e-02 -8.59128833e-01 4.27154243e-01 -3.10770065e-01
1.48433959e+00 -4.29476440e-01 1.85870305e-01 5.12743831e-01
1.38199043e+00 -2.40528256e-01 2.20639020e-01 2.97757030e-01
5.99435627e-01 7.41900325e-01 2.36603707e-01 4.44557279e-01
3.90029609e-01 2.13444784e-01 4.06356275e-01 5.96362591e-01
-8.79438668e-02 -4.67111379e-01 5.19883037e-01 1.36576986e+00
1.51302248e-01 2.60871053e-01 -1.30937898e+00 9.69575942e-01
-1.53938842e+00 -9.33744848e-01 -4.05673772e-01 1.99828017e+00
1.35276270e+00 2.08258986e-01 -8.79805759e-02 3.43341343e-02
2.07894862e-01 -1.01669520e-01 -2.74992824e-01 -1.07332087e+00
9.75696027e-01 7.42361248e-01 3.82586606e-02 8.01469326e-01
-3.97501975e-01 6.66722059e-01 5.29208469e+00 5.07843196e-01
-7.60176301e-01 -7.12769851e-02 2.37511411e-01 6.41785115e-02
-5.57692170e-01 9.03850943e-02 -9.12099957e-01 2.77822435e-01
1.15695000e+00 -2.91829139e-01 6.65655196e-01 1.37059295e+00
-1.45350501e-01 -1.05249494e-01 -1.57564890e+00 7.50454426e-01
-1.48604870e-01 -1.67683542e+00 -1.72305152e-01 5.93812317e-02
1.16703987e+00 -7.71746635e-02 3.79463062e-02 6.86292231e-01
1.24990201e+00 -1.28544772e+00 4.48451638e-01 2.14328468e-01
3.42077553e-01 -6.33150101e-01 3.27413648e-01 8.30464542e-01
-6.89825356e-01 -2.42234483e-01 -3.43318611e-01 -7.99216449e-01
-7.84914792e-01 3.03607672e-01 -1.40455389e+00 -1.51463136e-01
5.68708003e-01 4.95559514e-01 -7.18349159e-01 9.91686821e-01
-4.74685282e-01 8.85161102e-01 -1.11173047e-02 -6.33549571e-01
2.66189337e-01 -1.41374961e-01 2.02348500e-01 1.03741264e+00
2.41177514e-01 5.70123374e-01 5.82497120e-01 1.16637993e+00
-3.79755288e-01 -8.81977677e-02 -7.09819257e-01 -2.50459593e-02
7.70806015e-01 1.41409492e+00 -2.93370217e-01 -4.17141557e-01
-4.36241031e-01 5.04723191e-01 5.39411366e-01 -8.57382938e-02
-1.31172895e-01 -6.69469118e-01 7.76334643e-01 4.58334982e-01
-1.13582365e-01 -2.50012219e-01 -3.65893424e-01 -1.00717831e+00
-2.75847435e-01 -1.34249306e+00 1.36448339e-01 -9.42152858e-01
-8.99152577e-01 -2.25955889e-01 -4.63083684e-01 -9.91664171e-01
-3.85107517e-01 -7.56241262e-01 -1.07353950e+00 7.68739760e-01
-1.06825912e+00 -1.63966611e-01 -7.69817114e-01 3.60084951e-01
7.89900362e-01 -3.99075508e-01 9.72503304e-01 9.20774713e-02
-7.84617011e-03 5.63708663e-01 -2.67248452e-02 4.46188152e-01
8.23657095e-01 -1.83595383e+00 -6.28826534e-03 6.04717910e-01
4.79905009e-01 9.77855623e-01 9.44062293e-01 -1.92824781e-01
-1.77668083e+00 -4.42504048e-01 7.42725909e-01 -8.75595868e-01
8.74138176e-01 -2.66373575e-01 -1.11552763e+00 7.33721733e-01
3.64982188e-01 1.60078019e-01 9.88579929e-01 4.11535561e-01
-5.63887477e-01 -1.10363655e-01 -8.81290793e-01 4.30431247e-01
5.62103212e-01 -6.63230360e-01 -1.03645372e+00 5.70842683e-01
7.16827393e-01 -9.27969396e-01 -1.10449576e+00 -3.12679201e-01
4.19190437e-01 -8.66764784e-01 7.26990998e-01 -8.94189715e-01
1.28902268e+00 2.72732619e-02 1.33657902e-01 -1.41029859e+00
4.92120311e-02 -7.64382541e-01 -1.04584701e-01 1.01488256e+00
-4.18120325e-02 7.82849384e-04 1.20533895e+00 7.82673419e-01
-5.61163068e-01 -8.44193578e-01 -1.23268284e-01 -2.49021515e-01
3.78835231e-01 -3.32680970e-01 3.79684120e-01 1.47517490e+00
6.29265428e-01 3.06385010e-01 4.71592277e-01 -1.86031744e-01
7.12254405e-01 3.70967746e-01 1.12778246e+00 -1.61509860e+00
-3.85107964e-01 -7.10760117e-01 -3.90818089e-01 -1.06885219e+00
-1.11978194e-02 -1.23080075e+00 6.96424320e-02 -9.96845067e-01
2.23178849e-01 -5.93786657e-01 -7.52097741e-02 5.06137669e-01
-1.93203300e-01 -1.64554581e-01 1.58465415e-01 -3.07234824e-01
-8.30611587e-01 -1.89375818e-01 1.29945362e+00 1.03506178e-01
-8.36156681e-03 -6.98224977e-02 -7.23383605e-01 9.84127045e-01
4.22727525e-01 -4.81669635e-01 -5.18533707e-01 -4.77545083e-01
6.91494942e-01 3.92863303e-01 2.25735113e-01 -1.13406289e+00
8.40154767e-01 -6.81745231e-01 3.03710699e-01 -2.68536925e-01
-3.33317995e-01 -6.07639670e-01 -5.89061618e-01 4.77868229e-01
-8.16696882e-01 1.16712358e-02 2.45251060e-01 1.56490117e-01
-2.70648628e-01 -9.21205819e-01 8.01955163e-01 -5.83234549e-01
-5.06847620e-01 8.68368745e-02 -5.76758564e-01 5.04285216e-01
8.59746575e-01 -3.71500589e-02 -5.45553267e-01 -3.71050745e-01
-5.57949364e-01 5.02170265e-01 3.38351130e-01 3.55785906e-01
6.62188828e-01 -1.09765220e+00 -7.18991101e-01 2.37730294e-01
5.20315813e-03 4.50726092e-01 -4.88891378e-02 4.00137991e-01
-9.04658079e-01 3.79632145e-01 -4.41330284e-01 -7.46181309e-01
-1.42224801e+00 3.83024812e-01 -1.39839668e-02 -2.84929127e-01
-3.94293845e-01 1.22676611e+00 -1.11801766e-01 -1.05671966e+00
4.67150778e-01 -3.46407950e-01 7.32855648e-02 2.25803175e-04
9.07810628e-01 3.72674137e-01 2.45897472e-01 5.64851880e-01
3.32111210e-01 -1.75805226e-01 -3.74211043e-01 1.86976463e-01
1.67410135e+00 4.17570412e-01 6.68858588e-02 7.01641023e-01
9.13126349e-01 2.53671050e-01 -1.08170128e+00 -2.64038324e-01
2.05949008e-01 -5.47893703e-01 2.52756625e-02 -9.94465292e-01
-6.74719989e-01 1.51626420e+00 2.14617819e-01 2.80985057e-01
4.93906021e-01 -3.02879244e-01 5.31931818e-01 7.01006889e-01
5.28543293e-01 -9.75496411e-01 5.33940792e-01 8.66994500e-01
3.53266090e-01 -1.40561676e+00 1.87834203e-01 -4.61672470e-02
-7.15185106e-02 1.59371173e+00 1.18039513e+00 -3.39204073e-01
2.41555765e-01 1.92374721e-01 -5.17081358e-02 -2.77177244e-01
-1.38955498e+00 5.34198046e-01 1.12109542e-01 1.31852463e-01
1.23081625e+00 1.49234220e-01 2.46952683e-01 6.06726170e-01
-7.22618759e-01 1.07069775e-01 1.43631065e+00 1.15491676e+00
-1.21011639e+00 -1.01483905e+00 -5.67131698e-01 3.84511352e-01
-3.69256169e-01 -2.49244064e-01 -3.42677593e-01 5.28620303e-01
4.77846377e-02 5.76264858e-01 -9.15805697e-02 -1.29386380e-01
1.76025312e-02 5.14509976e-01 8.76678944e-01 -1.32709098e+00
-1.13314128e+00 -8.16287041e-01 -5.29127717e-01 -4.64476705e-01
3.04331452e-01 -5.96689641e-01 -1.40393531e+00 -9.28876102e-01
-1.51848868e-01 4.14911300e-01 7.06560612e-01 8.64992261e-01
-4.79495880e-04 7.19152033e-01 4.08086836e-01 -3.16704959e-01
-9.92548943e-01 -7.72542775e-01 -4.25119959e-02 2.10716441e-01
4.69563812e-01 -3.85968477e-01 -3.29320490e-01 3.44170332e-01] | [9.776865005493164, 7.364322185516357] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.