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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
369ca808-3fea-44c0-9f7e-a4c2f47e596f | the-cone-of-silence-speech-separation-by | 2010.06007 | null | https://arxiv.org/abs/2010.06007v1 | https://arxiv.org/pdf/2010.06007v1.pdf | The Cone of Silence: Speech Separation by Localization | Given a multi-microphone recording of an unknown number of speakers talking concurrently, we simultaneously localize the sources and separate the individual speakers. At the core of our method is a deep network, in the waveform domain, which isolates sources within an angular region $\theta \pm w/2$, given an angle of interest $\theta$ and angular window size $w$. By exponentially decreasing $w$, we can perform a binary search to localize and separate all sources in logarithmic time. Our algorithm allows for an arbitrary number of potentially moving speakers at test time, including more speakers than seen during training. Experiments demonstrate state-of-the-art performance for both source separation and source localization, particularly in high levels of background noise. | ['Ira Kemelmacher-Shlizerman', 'Steve Seitz', 'Vivek Jayaram', 'Teerapat Jenrungrot'] | 2020-10-12 | null | http://proceedings.neurips.cc/paper/2020/hash/f056bfa71038e04a2400266027c169f9-Abstract.html | http://proceedings.neurips.cc/paper/2020/file/f056bfa71038e04a2400266027c169f9-Paper.pdf | neurips-2020-12 | ['audio-source-separation'] | ['audio'] | [ 8.69290158e-02 -2.27824867e-01 2.57487297e-01 5.19568361e-02
-1.55137861e+00 -8.19460392e-01 -7.34260157e-02 3.80657874e-02
-4.93225187e-01 5.47946095e-01 -1.02464072e-01 -2.47541398e-01
-4.53007907e-01 -5.60605168e-01 -5.80753386e-01 -6.53242290e-01
-8.52227151e-01 3.16329867e-01 2.91881830e-01 1.84964374e-01
1.11772884e-02 5.45926511e-01 -1.34798837e+00 -7.27428794e-02
3.85416389e-01 1.10176516e+00 4.83637825e-02 1.30114913e+00
1.34366095e-01 2.94107556e-01 -1.20775044e+00 -1.53419413e-02
2.79017061e-01 -5.86277604e-01 -3.53768975e-01 -1.62368298e-01
3.22489709e-01 1.23671349e-02 -2.14982942e-01 1.06099117e+00
9.04365718e-01 4.63832289e-01 2.62835532e-01 -1.01157475e+00
-4.24769893e-02 7.34952033e-01 -5.92235863e-01 8.52633238e-01
4.82160240e-01 -2.89257523e-02 8.95001113e-01 -1.01968610e+00
-1.19247541e-01 9.22032297e-01 8.28193486e-01 1.77611597e-02
-1.25587046e+00 -1.26586044e+00 4.71105836e-02 7.33451322e-02
-1.83245158e+00 -8.69695961e-01 9.77052808e-01 -1.89038381e-01
1.00447392e+00 3.39725912e-01 4.08597678e-01 6.74707532e-01
-3.30132335e-01 1.54324397e-01 4.19417202e-01 -4.32964385e-01
3.80989850e-01 -1.55058384e-01 -9.18662399e-02 3.31377715e-01
-5.62152788e-02 4.42886278e-02 -1.14499652e+00 -4.58839923e-01
5.93914568e-01 -3.86071026e-01 -5.45329094e-01 1.83249891e-01
-8.68733644e-01 7.14109421e-01 -5.00461720e-02 2.41366729e-01
2.92320251e-02 3.79328370e-01 -5.96580503e-04 2.56775051e-01
4.97582614e-01 4.03534412e-01 -5.17045319e-01 -4.55127656e-01
-1.18446493e+00 5.45262545e-02 8.75180066e-01 6.73511147e-01
5.69250762e-01 3.51745456e-01 3.69485766e-01 9.62859511e-01
2.41896138e-01 8.83303404e-01 1.84692517e-01 -1.10160613e+00
5.63060582e-01 -1.98934764e-01 2.17948452e-01 -7.45045662e-01
-2.54518420e-01 -7.31561601e-01 -7.19572604e-01 2.04380274e-01
5.30371726e-01 -6.57930195e-01 -6.01640522e-01 2.05770922e+00
3.34842056e-01 7.74362564e-01 -8.21925625e-02 6.26930058e-01
1.78349838e-01 7.30499148e-01 -5.96546292e-01 -6.13042712e-01
1.21320724e+00 -6.73372447e-01 -5.43410122e-01 -6.76389337e-01
-2.71982640e-01 -1.00027418e+00 6.89776003e-01 6.97800994e-01
-1.59540272e+00 -6.04473531e-01 -1.00923920e+00 4.41460013e-01
-5.92607893e-02 -2.72665352e-01 2.00199217e-01 1.05506134e+00
-1.06121111e+00 2.45034337e-01 -9.13735628e-01 4.36843067e-01
2.20808715e-01 5.13185143e-01 4.93061505e-02 2.07301050e-01
-1.06779504e+00 2.01448709e-01 -4.88831580e-01 2.68214464e-01
-1.07641196e+00 -8.53368223e-01 -7.92232394e-01 3.00124705e-01
1.28850579e-01 -3.10231782e-02 1.34124148e+00 -6.02957904e-01
-1.40723014e+00 3.91076535e-01 -8.47089648e-01 -4.46699023e-01
2.07555816e-01 -2.70631880e-01 -1.03078413e+00 3.36770773e-01
1.92678601e-01 -1.67495403e-02 1.05629110e+00 -1.04402161e+00
-7.42236555e-01 -2.64515340e-01 -3.07270527e-01 1.74832091e-01
-2.20916286e-01 2.72310853e-01 -5.56428552e-01 -5.29097497e-01
3.11945051e-01 -6.37121379e-01 -2.15407629e-02 -1.62199065e-01
-3.64760727e-01 9.33438633e-03 4.16153848e-01 -6.60290241e-01
1.04585898e+00 -2.50077629e+00 -1.06339261e-01 4.97577578e-01
2.33277634e-01 -3.90624329e-02 -1.40156910e-01 2.35073656e-01
-1.85583547e-01 -4.55517583e-02 -2.56055295e-01 -8.12334538e-01
-1.64147943e-01 -3.47225249e-01 -3.53384852e-01 6.62896216e-01
-1.85038581e-01 1.00081407e-01 -5.10879040e-01 -1.29745111e-01
-1.27520561e-01 5.92162907e-01 -6.21563494e-01 3.53090584e-01
3.09985191e-01 2.08873644e-01 1.07873790e-01 4.16725785e-01
8.51299942e-01 -2.04813316e-01 5.91310039e-02 3.27864647e-01
-1.79829493e-01 7.22331941e-01 -2.03145385e+00 1.55592585e+00
-6.12195432e-01 1.07449400e+00 9.51714277e-01 -9.59534526e-01
7.45762646e-01 6.60906971e-01 5.00497580e-01 -3.09040874e-01
-2.92502567e-02 2.99641162e-01 1.20217130e-01 -2.28502661e-01
1.22203864e-01 -2.62189597e-01 -7.30255768e-02 7.01798320e-01
1.58605069e-01 1.47065483e-02 8.08269009e-02 -1.34757131e-01
1.31299210e+00 -8.97211492e-01 4.31747139e-02 -1.63366109e-01
2.35612690e-01 -6.67386651e-01 5.23743272e-01 9.15194511e-01
-2.81762272e-01 6.05357409e-01 2.30790555e-01 6.04559816e-02
-3.79089952e-01 -1.67657650e+00 -1.02876596e-01 1.31069040e+00
1.69983909e-01 -9.21792835e-02 -4.99891251e-01 2.71765105e-02
-1.71742305e-01 4.52413797e-01 -3.67806762e-01 5.81554100e-02
-1.09018183e+00 -7.63789952e-01 8.78864229e-01 4.83316451e-01
3.12041163e-01 -9.10976946e-01 -7.72178113e-01 3.26372117e-01
-3.00881475e-01 -9.51219559e-01 -8.99224460e-01 7.69354045e-01
-4.72947806e-01 -8.01880538e-01 -4.98805583e-01 -9.41747725e-01
4.85056698e-01 3.81467760e-01 1.05637896e+00 -1.89647198e-01
-3.39588612e-01 1.19998217e-01 1.15176953e-01 -4.54215407e-01
-1.82175413e-02 -1.42826006e-01 2.13439286e-01 -5.71789108e-02
1.99354261e-01 -1.06212509e+00 -6.83030903e-01 1.54405639e-01
-3.54301721e-01 -7.66458869e-01 8.44801515e-02 4.27514941e-01
2.14727685e-01 6.82131827e-01 5.03512800e-01 -8.23399425e-02
6.66135669e-01 -5.12001932e-01 -5.71764052e-01 -2.14756116e-01
5.49754091e-02 -3.65855187e-01 6.64828718e-01 -7.86158085e-01
-5.57240307e-01 -3.08159322e-01 -3.38089436e-01 -2.84402698e-01
-2.21927166e-01 1.08998276e-01 -1.61808327e-01 6.80538192e-02
6.94209158e-01 2.97740072e-01 -4.25802320e-01 -5.37026346e-01
1.15387142e-01 4.33740735e-01 7.12383449e-01 -3.39691788e-01
8.00633848e-01 3.54434252e-01 -2.74469405e-01 -9.51605022e-01
-4.45195019e-01 -4.84299928e-01 -1.00288466e-01 9.20223910e-03
4.89449471e-01 -9.62981343e-01 -1.04596543e+00 4.39541727e-01
-1.18403733e+00 -3.80029559e-01 -4.46062624e-01 6.89928591e-01
6.98290914e-02 -1.77036528e-03 -3.58304799e-01 -1.35668695e+00
-1.23724274e-01 -8.55298340e-01 9.13447559e-01 3.16994131e-01
-2.32163057e-01 -8.34960639e-01 2.34457314e-01 -2.97350883e-02
3.76307577e-01 -2.64522046e-01 4.95238334e-01 -7.01460600e-01
-4.80493069e-01 -2.07616881e-01 2.84678251e-01 2.25644842e-01
2.99064487e-01 -2.95028150e-01 -1.16135252e+00 -3.40975910e-01
3.84930164e-01 6.08069934e-02 5.25862157e-01 6.95869446e-01
1.12309325e+00 -3.31788510e-01 -4.94588226e-01 7.65652657e-01
1.12108958e+00 7.88338542e-01 2.51845211e-01 -2.11318150e-01
1.51464716e-01 2.34227702e-01 -1.66087329e-01 5.64346194e-01
2.73451488e-02 3.38097125e-01 2.56182790e-01 -1.46007836e-01
9.34466347e-03 9.92099717e-02 1.45686105e-01 7.59975076e-01
3.45835984e-01 -6.92251563e-01 -7.88199723e-01 8.39114845e-01
-9.45527852e-01 -1.05201340e+00 3.89856368e-01 2.51185131e+00
9.58356500e-01 5.07665992e-01 2.59943783e-01 7.00334787e-01
9.35279727e-01 2.87159413e-01 -5.69117486e-01 -2.09181368e-01
-3.73354033e-02 1.07534051e+00 5.36425889e-01 1.07690871e+00
-7.94256270e-01 1.56366810e-01 7.51297760e+00 6.54434025e-01
-1.19950867e+00 1.87006041e-01 3.92440200e-01 -8.45488966e-01
-1.33822426e-01 -4.67028499e-01 -7.98134148e-01 6.08382821e-01
1.18052506e+00 -1.55502930e-01 8.64003897e-01 5.12013435e-01
1.58155844e-01 -2.59600997e-01 -1.06246579e+00 1.09988809e+00
2.44158357e-01 -1.12530386e+00 -8.44994128e-01 1.34226270e-02
3.10915858e-01 7.91081488e-02 2.84181625e-01 2.89424062e-02
5.08681595e-01 -1.16193497e+00 8.42486262e-01 -8.30696523e-02
6.93489492e-01 -8.81724238e-01 1.41969964e-01 4.75009054e-01
-1.56461024e+00 -2.75800884e-01 2.18246169e-02 -1.47879466e-01
4.90774900e-01 8.71968806e-01 -7.49342620e-01 4.68045920e-02
1.10780644e+00 -2.02860951e-01 1.01567961e-01 1.03219283e+00
-2.61025012e-01 1.01087058e+00 -1.01178753e+00 9.62199122e-02
-1.67735189e-01 2.11184502e-01 6.49870515e-01 1.30833673e+00
8.38421047e-01 2.04306409e-01 1.27450749e-01 6.22026443e-01
-2.68263817e-01 -2.98120528e-01 -2.31304705e-01 4.81422991e-01
1.36069047e+00 8.26647520e-01 -6.71610713e-01 -1.65070578e-01
-1.57596290e-01 7.35863864e-01 7.16068000e-02 5.76450706e-01
-9.12055433e-01 -9.19750571e-01 9.60338354e-01 5.27620651e-02
6.63577199e-01 -6.44979894e-01 -3.51459265e-01 -5.88496208e-01
1.93802312e-01 -7.40735233e-01 3.24329168e-01 -4.86266285e-01
-9.34947073e-01 6.53311789e-01 -1.63317904e-01 -1.05989647e+00
-1.63440228e-01 -2.78635740e-01 -1.05608392e+00 1.21354365e+00
-1.27346289e+00 -2.15788424e-01 7.74483606e-02 7.78248906e-01
6.58769608e-01 2.15066094e-02 9.53548610e-01 4.13866013e-01
-4.64694679e-01 7.71013379e-01 1.09608889e-01 2.63346046e-01
5.83335578e-01 -1.11960530e+00 4.00944412e-01 1.12504661e+00
4.61773187e-01 7.95378745e-01 9.28688347e-01 -1.35712132e-01
-9.93565857e-01 -6.79732084e-01 8.84051085e-01 -2.73419291e-01
5.45864999e-01 -8.90207708e-01 -7.63692856e-01 4.09997582e-01
3.54153812e-01 2.41834715e-01 1.17536092e+00 3.28217626e-01
-2.96671808e-01 -3.42464536e-01 -1.01967812e+00 2.81515419e-01
9.74167049e-01 -7.11068809e-01 -6.12542510e-01 2.46161908e-01
6.02092147e-01 -6.09639823e-01 -5.06601751e-01 -6.72116950e-02
4.38671917e-01 -9.99454975e-01 1.30476868e+00 9.69759598e-02
-2.29917452e-01 -2.05148235e-01 -2.31426746e-01 -1.35312438e+00
-1.75871834e-01 -1.21615028e+00 -1.79365262e-01 1.05146480e+00
7.12834179e-01 -6.81468725e-01 7.68967867e-01 2.01621756e-01
-1.99668445e-02 -2.86552072e-01 -1.64577210e+00 -8.37521255e-01
-3.49911824e-02 -1.02452815e+00 4.77038175e-01 4.43464905e-01
1.27936676e-01 2.38376975e-01 -2.13217497e-01 9.28344727e-01
6.93444550e-01 6.65890127e-02 3.96988988e-01 -1.07780433e+00
-8.44743848e-01 -4.41779166e-01 1.67280048e-01 -1.37104499e+00
-1.94308385e-02 -4.16745454e-01 4.06271249e-01 -9.61956680e-01
-3.82092088e-01 -4.53602433e-01 -5.32570779e-01 1.20439783e-01
-1.78957194e-01 4.35942322e-01 -6.55726269e-02 -9.65024978e-02
-3.40300560e-01 1.35675654e-01 5.47369003e-01 -1.03175379e-01
-4.15489137e-01 3.02087784e-01 -7.27054417e-01 8.63250554e-01
6.18293405e-01 -7.95937777e-01 -3.90901953e-01 -7.97780275e-01
9.34910551e-02 4.24870223e-01 1.74470708e-01 -1.21263301e+00
4.62433279e-01 -9.91536770e-03 4.36188191e-01 -5.39434910e-01
8.67141306e-01 -7.18769550e-01 7.88632482e-02 1.62871778e-01
-4.49029744e-01 1.35967091e-01 4.76457119e-01 3.80280614e-01
-2.55165607e-01 -7.51259848e-02 8.14244807e-01 9.40094665e-02
2.55012233e-02 9.31089371e-02 -4.75943059e-01 4.17122304e-01
6.14316821e-01 -5.07549681e-02 -8.29359517e-02 -5.59119821e-01
-8.62706125e-01 -1.49762155e-02 -3.39765847e-01 1.18217520e-01
3.62655729e-01 -1.17164063e+00 -4.78464037e-01 6.18354619e-01
-4.46104944e-01 1.96519390e-01 3.68638664e-01 6.05228901e-01
-1.02110289e-01 4.11855504e-02 4.93081361e-01 -5.82490087e-01
-1.23933172e+00 9.96344388e-02 6.65872991e-01 -8.91330019e-02
-2.05578923e-01 1.83582723e+00 -2.48125289e-02 8.32743719e-02
5.62653184e-01 -3.35274458e-01 1.57353088e-01 2.39756685e-02
8.95868123e-01 6.43063784e-01 5.16126305e-02 -2.25797936e-01
-7.14095175e-01 6.30480826e-01 5.57189763e-01 -1.01679134e+00
1.10371983e+00 -2.05994099e-01 2.56730560e-02 5.44571519e-01
1.32442629e+00 1.01504922e+00 -1.47536182e+00 -1.34060562e-01
-4.31345075e-01 -4.87801671e-01 -2.11537749e-01 -6.92262113e-01
-1.01164007e+00 1.03910077e+00 7.19529569e-01 7.78532326e-01
1.30452681e+00 1.38983235e-01 6.36793137e-01 1.43801242e-01
9.27946717e-02 -8.70158255e-01 1.91760272e-01 4.21327472e-01
4.91808712e-01 -7.37226188e-01 -4.44859684e-01 -2.24229738e-01
6.40734211e-02 9.02609348e-01 2.70252854e-01 -2.31336206e-01
1.05842066e+00 1.17561018e+00 2.89909750e-01 9.37335938e-02
-5.31766534e-01 1.16676278e-01 -1.62721686e-02 6.95658743e-01
3.54247987e-01 -5.47563359e-02 4.52861220e-01 8.48723650e-01
-6.00486755e-01 -4.27225083e-01 1.19264819e-01 8.77903938e-01
-8.50039065e-01 -8.63049746e-01 -9.25669551e-01 1.44514516e-01
-8.52187753e-01 -2.93242246e-01 8.01101923e-02 3.89792800e-01
3.79562676e-01 1.63912559e+00 4.86367047e-01 -1.07442498e-01
4.44009602e-01 1.97063774e-01 2.63173759e-01 -4.26891595e-01
-4.82957959e-01 6.74639702e-01 -5.60481250e-02 -2.32625321e-01
-8.50341171e-02 -7.90732741e-01 -1.27443647e+00 -2.88451165e-01
-2.22881898e-01 6.81864738e-01 6.50347948e-01 8.57120335e-01
1.37678280e-01 7.78669357e-01 9.03610826e-01 -8.05837154e-01
-2.71017432e-01 -1.13810551e+00 -7.28384495e-01 -1.76385015e-01
1.00174403e+00 -2.87812561e-01 -1.00747168e+00 1.82940379e-01] | [15.147488594055176, 5.65294885635376] |
e217ad1d-0a96-4751-a462-e32f93e4b62f | one-step-distributional-reinforcement | 2304.14421 | null | https://arxiv.org/abs/2304.14421v1 | https://arxiv.org/pdf/2304.14421v1.pdf | One-Step Distributional Reinforcement Learning | Reinforcement learning (RL) allows an agent interacting sequentially with an environment to maximize its long-term expected return. In the distributional RL (DistrRL) paradigm, the agent goes beyond the limit of the expected value, to capture the underlying probability distribution of the return across all time steps. The set of DistrRL algorithms has led to improved empirical performance. Nevertheless, the theory of DistrRL is still not fully understood, especially in the control case. In this paper, we present the simpler one-step distributional reinforcement learning (OS-DistrRL) framework encompassing only the randomness induced by the one-step dynamics of the environment. Contrary to DistrRL, we show that our approach comes with a unified theory for both policy evaluation and control. Indeed, we propose two OS-DistrRL algorithms for which we provide an almost sure convergence analysis. The proposed approach compares favorably with categorical DistrRL on various environments. | ['Eric Moulines', 'Kirill Fedyanin', 'Yasser Abdelaziz Dahou Djilali', 'REDA ALAMI', 'Mastane Achab'] | 2023-04-27 | null | null | null | null | ['distributional-reinforcement-learning'] | ['methodology'] | [-2.66162068e-01 8.96814242e-02 -2.38230914e-01 1.69265315e-01
-5.27264357e-01 -5.57307422e-01 9.10917163e-01 2.91500807e-01
-1.01818299e+00 1.21237409e+00 1.00345723e-02 -4.38434631e-01
-5.02948344e-01 -8.73789608e-01 -6.31004155e-01 -1.05494034e+00
-1.90407574e-01 4.44205105e-01 9.14487150e-03 -2.33910978e-01
1.76949814e-01 3.91095340e-01 -1.48314500e+00 -5.46078980e-01
7.29255736e-01 6.72321022e-01 4.77420092e-01 7.93947995e-01
-3.39961238e-02 9.93838608e-01 -4.71761614e-01 -2.51244213e-02
1.58587590e-01 -5.33692777e-01 -5.27616620e-01 -2.83248425e-01
-4.37015712e-01 -4.12566125e-01 -2.44942337e-01 1.15515554e+00
5.94524920e-01 3.40057582e-01 7.59678245e-01 -1.15995336e+00
-3.66012603e-01 6.85902536e-01 -5.28878033e-01 -5.16464971e-02
2.64196396e-01 3.94503325e-01 1.09923613e+00 -2.82540023e-01
6.25118375e-01 1.27168405e+00 2.12857455e-01 5.08678973e-01
-1.31784964e+00 -1.09057434e-01 3.40253800e-01 -1.83922797e-02
-8.68779421e-01 3.26732770e-02 4.72234547e-01 -2.89894551e-01
6.85256898e-01 -1.71918333e-01 7.47920454e-01 1.06474245e+00
3.67061764e-01 1.02039742e+00 1.56328940e+00 -6.54297113e-01
8.93676698e-01 5.11754975e-02 -3.76163796e-02 2.05173075e-01
3.96146774e-01 8.51720572e-01 -6.05190881e-02 -1.82389289e-01
7.23646164e-01 -1.06578380e-01 1.75511092e-02 -7.82174170e-01
-7.30525851e-01 8.58151972e-01 6.28641695e-02 4.44639355e-01
-8.56344163e-01 3.63011122e-01 3.17903996e-01 5.86879432e-01
2.30133727e-01 2.59087563e-01 -2.73038298e-01 -4.98989671e-01
-5.27538836e-01 7.59062707e-01 8.39377165e-01 5.48115611e-01
7.00617552e-01 1.87733576e-01 -3.05501044e-01 3.46938848e-01
2.73195326e-01 6.97399020e-01 4.06829596e-01 -1.12070608e+00
1.70464218e-01 -5.89751974e-02 6.88156486e-01 -5.33559263e-01
-3.29338312e-01 -6.06709898e-01 -5.53263783e-01 6.00153387e-01
6.13528371e-01 -5.72994769e-01 -8.34388062e-02 2.20763874e+00
2.06997439e-01 2.26978753e-02 3.67850840e-01 5.16012788e-01
-3.17076921e-01 6.00768149e-01 1.19574502e-01 -7.63932765e-01
5.15186608e-01 -3.38080704e-01 -6.65436268e-01 1.56579271e-01
5.53245842e-01 -2.13678643e-01 1.28122211e+00 5.85385859e-01
-1.02517819e+00 -2.18943104e-01 -7.35867679e-01 6.98245645e-01
-8.04217681e-02 -3.77465516e-01 2.20971569e-01 5.30052543e-01
-1.22637486e+00 6.48066342e-01 -6.89475119e-01 -2.43884042e-01
6.89524859e-02 1.48044184e-01 1.74885571e-01 1.05151169e-01
-1.05468547e+00 8.91516745e-01 3.45428765e-01 -2.58356661e-01
-1.17496598e+00 -2.11298689e-01 -5.43114424e-01 -1.19090505e-01
7.78956115e-01 -5.14674544e-01 1.69285274e+00 -8.57980847e-01
-1.98854673e+00 1.94288343e-01 -5.87265603e-02 -7.91985154e-01
9.19816315e-01 -2.46167675e-01 2.13705257e-01 -4.76137474e-02
-2.50499360e-02 1.73775330e-01 7.95157909e-01 -1.24067259e+00
-5.52299857e-01 -2.49283716e-01 2.41945982e-01 3.17549199e-01
-1.26072047e-02 -3.42107356e-01 4.34243947e-01 -4.45358783e-01
-7.44498193e-01 -8.98647010e-01 -4.27219003e-01 -5.56380749e-01
1.01846687e-01 -5.45391440e-01 3.27308655e-01 1.68764785e-01
1.22558308e+00 -2.13581800e+00 1.82144225e-01 1.32062053e-02
-3.40366773e-02 1.42915189e-01 -3.56195092e-01 8.83969426e-01
9.05032754e-02 -1.50492966e-01 -1.78102404e-01 -2.34544724e-01
5.23157716e-01 4.95863944e-01 -5.15552521e-01 6.63756728e-01
-1.19283125e-01 7.69238055e-01 -1.19782865e+00 -2.07223743e-01
3.15076590e-01 7.51766236e-03 -7.01062620e-01 4.17970240e-01
-5.28931022e-01 1.97160393e-01 -6.92196727e-01 9.07386839e-02
4.62087423e-01 1.25550777e-01 3.80016148e-01 6.67851031e-01
-5.45899272e-01 5.74613996e-02 -1.21827853e+00 1.10452425e+00
-4.66383606e-01 2.33225048e-01 1.93621591e-02 -1.08280933e+00
7.01144934e-01 1.86975867e-01 7.78557420e-01 -7.91493893e-01
1.80112883e-01 3.08560759e-01 5.98254651e-02 -4.96924847e-01
4.39529181e-01 -4.91425306e-01 -1.67073578e-01 7.57845461e-01
-5.96213937e-02 -3.31786904e-03 3.49958152e-01 4.57946286e-02
1.08551311e+00 5.45196056e-01 7.64084756e-01 -4.06131268e-01
3.36712629e-01 -1.56769380e-01 3.47759843e-01 1.19723547e+00
-4.44360465e-01 -2.24769637e-01 7.47859776e-01 -1.88403934e-01
-8.56806457e-01 -1.38589847e+00 7.17537850e-02 1.09374261e+00
1.23221517e-01 -7.83018842e-02 -7.74197102e-01 -4.45510417e-01
2.69695878e-01 1.09395158e+00 -6.29960358e-01 -5.43246232e-02
-4.36847866e-01 -7.18655586e-01 3.00781012e-01 -4.45134751e-03
2.92524189e-01 -1.47628784e+00 -1.06941903e+00 5.39763212e-01
2.05818594e-01 -5.33897221e-01 -2.40856484e-01 3.41599345e-01
-6.00515962e-01 -9.95941579e-01 -5.87746143e-01 -1.16426617e-01
1.33396119e-01 -1.26902550e-01 9.14589465e-01 -4.86787856e-01
-6.69530109e-02 1.03660345e+00 -3.93669039e-01 -5.30519307e-01
-5.77784121e-01 -8.23017210e-02 3.91658962e-01 3.16350609e-02
2.30052382e-01 -4.61674184e-01 -4.00775045e-01 -2.14105561e-01
-9.83036816e-01 -5.58402419e-01 6.08498096e-01 7.40739167e-01
5.32711029e-01 1.34586960e-01 9.46828902e-01 -5.30458748e-01
1.22368205e+00 -5.80260038e-01 -9.44513679e-01 2.03104272e-01
-8.41454506e-01 5.74829102e-01 7.47427762e-01 -5.33332765e-01
-1.04931486e+00 -3.01353008e-01 1.07636163e-02 -6.58213496e-02
-3.23183119e-01 3.41237783e-01 8.74710307e-02 2.96852499e-01
3.34130794e-01 6.23853862e-01 3.28644931e-01 -3.17426175e-01
4.33308929e-01 4.25697923e-01 1.77544698e-01 -1.04048538e+00
5.58577061e-01 1.13095976e-01 1.36423588e-01 -9.02255058e-01
-5.37006378e-01 1.76853295e-02 -2.71162182e-01 -4.71559614e-01
6.11873746e-01 -5.23558915e-01 -1.23866558e+00 3.76732409e-01
-8.72576416e-01 -8.80124807e-01 -9.55137670e-01 6.69821262e-01
-1.39691925e+00 2.06401676e-01 -3.68720502e-01 -1.65648341e+00
1.17905557e-01 -1.08615875e+00 5.31219840e-01 1.61170244e-01
9.44788754e-02 -1.22088051e+00 6.86778247e-01 -6.48318291e-01
4.19889867e-01 3.38957310e-01 8.64482939e-01 -5.33066809e-01
-3.67621869e-01 1.32996812e-01 2.90922433e-01 2.94015050e-01
-1.60290301e-02 -1.86649680e-01 -5.95677614e-01 -4.94089276e-01
2.07609326e-01 -5.57570994e-01 6.55951679e-01 7.70255148e-01
8.05994153e-01 -3.35242152e-01 1.86899841e-01 3.59056257e-02
1.71955478e+00 4.26691324e-01 3.72128099e-01 5.98021626e-01
-8.73801559e-02 6.32947981e-01 9.20390964e-01 1.00028622e+00
3.81104261e-01 3.52416784e-01 7.09199667e-01 3.43692601e-01
4.05029118e-01 -4.25477922e-01 8.32249522e-01 3.22698712e-01
7.20332889e-03 -3.59785438e-01 -5.42703807e-01 3.54640186e-01
-2.04668713e+00 -1.44852304e+00 2.79277891e-01 2.57485676e+00
6.63925469e-01 9.79517549e-02 5.99363029e-01 -8.95091817e-02
5.44795990e-01 1.67127818e-01 -8.53758454e-01 -5.63740134e-01
-8.51920694e-02 1.13499373e-01 5.17694592e-01 8.00785601e-01
-6.50943637e-01 6.21165991e-01 6.96586704e+00 7.65290380e-01
-9.45287645e-01 -3.67873162e-02 3.70034903e-01 -1.72392860e-01
-3.17069620e-01 -1.11694649e-01 -6.43391013e-01 4.39430773e-01
9.77743626e-01 -6.33843780e-01 7.82478273e-01 8.54870081e-01
6.32784307e-01 -3.86091799e-01 -9.59036469e-01 7.15388060e-01
-4.46954191e-01 -7.70458341e-01 -2.55813837e-01 6.07641518e-01
5.70879221e-01 1.13229200e-01 2.46396929e-01 6.84779167e-01
9.16256011e-01 -8.34213316e-01 8.60613644e-01 7.41028011e-01
4.25590813e-01 -1.11836517e+00 6.99734867e-01 9.08002019e-01
-8.91926944e-01 -4.45318878e-01 -2.95248359e-01 -3.77292931e-01
-7.91388825e-02 3.44312191e-01 -5.87781549e-01 6.07848108e-01
2.19899327e-01 5.40605962e-01 -1.00956507e-01 7.90996790e-01
-8.35854337e-02 5.70043087e-01 -2.35089704e-01 -6.37023687e-01
6.25711918e-01 -6.22026205e-01 7.29972720e-01 8.60618711e-01
4.07118678e-01 -1.50070757e-01 3.39534491e-01 6.75663471e-01
3.36645126e-01 2.61842251e-01 -9.76445913e-01 -7.28638470e-02
3.59150708e-01 7.42630601e-01 -3.22913736e-01 -2.19634920e-01
-1.46422625e-01 4.63139147e-01 5.76988637e-01 3.89986515e-01
-6.14835799e-01 -1.78308353e-01 7.06943512e-01 5.36158076e-03
2.92507619e-01 -3.13744545e-01 2.25712880e-01 -8.53742063e-01
-1.37947395e-01 -8.04678500e-01 2.66175181e-01 -1.92752942e-01
-1.29839003e+00 3.20095241e-01 3.40977103e-01 -1.06470311e+00
-8.13205063e-01 -4.97275680e-01 -4.02079850e-01 6.06467187e-01
-1.64528799e+00 -3.52593929e-01 4.85724866e-01 6.25153363e-01
4.30179447e-01 -1.69449881e-01 6.16306424e-01 -2.28841901e-01
-5.32242298e-01 6.52787313e-02 7.38291085e-01 -3.89477998e-01
5.18919706e-01 -1.72417450e+00 -4.00600135e-01 4.79994267e-01
-1.83907852e-01 4.83461648e-01 1.09123623e+00 -4.08177942e-01
-1.41665244e+00 -8.37316215e-01 3.82441700e-01 -1.87502876e-01
9.60976005e-01 6.06769249e-02 -5.71522772e-01 3.80529344e-01
4.25721496e-01 -2.93477803e-01 4.31528091e-01 -8.94799978e-02
1.03745557e-01 -9.95397195e-02 -1.02926421e+00 8.08427930e-01
6.97062433e-01 -1.90087572e-01 -5.95988274e-01 -1.40912190e-01
5.47415853e-01 1.69744506e-01 -5.20558417e-01 8.66254643e-02
4.45577413e-01 -1.08629048e+00 5.44023693e-01 -4.73637402e-01
2.27548853e-01 -1.95011973e-01 -3.42911899e-01 -1.62493658e+00
-1.66102380e-01 -9.05356228e-01 -1.11180685e-01 8.76295149e-01
-2.80766021e-02 -9.99283433e-01 2.89040416e-01 2.57682540e-02
2.65788257e-01 -6.45421267e-01 -1.03330457e+00 -1.08970094e+00
6.43152773e-01 -6.15147889e-01 4.41684484e-01 3.13545525e-01
1.01267181e-01 6.55400706e-03 -3.41913432e-01 -1.27150744e-01
1.06021643e+00 2.05217808e-01 6.21724188e-01 -9.92768347e-01
-6.62545323e-01 -7.23797917e-01 6.85573444e-02 -1.00998104e+00
4.93824512e-01 -5.49389124e-01 2.93025732e-01 -1.26959753e+00
1.54312387e-01 -3.44044536e-01 -5.42112231e-01 7.81044364e-02
-6.41488135e-02 -5.18370271e-01 2.98071593e-01 -2.62205955e-02
-8.54337931e-01 9.33381975e-01 1.16787791e+00 2.50158310e-01
-3.73641968e-01 2.21958324e-01 -5.55042803e-01 6.32755101e-01
1.07037580e+00 -5.09450555e-01 -7.00889289e-01 -2.27245707e-02
2.84828305e-01 3.47436011e-01 3.22120696e-01 -7.07233965e-01
1.37682557e-02 -6.95617318e-01 -1.31400898e-01 -4.02612656e-01
-1.40233189e-01 -5.97737908e-01 -9.91246551e-02 8.17265511e-01
-6.88880324e-01 2.78906107e-01 -1.14601217e-01 8.83569241e-01
1.37056904e-02 -4.61311579e-01 8.09653461e-01 -2.45584771e-01
-2.48319760e-01 1.64061144e-01 -7.89653420e-01 2.71813065e-01
1.21762526e+00 2.59820610e-01 -5.96653521e-02 -6.18271708e-01
-8.80105019e-01 2.88178980e-01 4.79452491e-01 -3.57240662e-02
4.96531725e-01 -1.10611081e+00 -6.02050662e-01 -1.57736748e-01
-2.24501505e-01 -3.62286001e-01 5.58635928e-02 6.96933687e-01
-9.42384973e-02 4.07764822e-01 -9.41629037e-02 -2.96750188e-01
-7.03925312e-01 7.79567719e-01 4.38700706e-01 -7.38750994e-01
-4.73185897e-01 8.49040374e-02 1.40396208e-01 -4.05052602e-01
2.74826914e-01 -1.68432355e-01 -1.02998607e-01 1.37715936e-01
5.76228499e-01 4.23866600e-01 -4.88327265e-01 -2.14857742e-01
3.17857452e-02 3.60343963e-01 1.60627261e-01 -7.29975462e-01
1.44237244e+00 -3.36586922e-01 1.25400629e-02 8.98269713e-01
7.44209409e-01 -2.66487733e-03 -1.61744714e+00 -1.69183895e-01
2.60103077e-01 -1.79144517e-01 -8.26755464e-02 -5.79510152e-01
-4.13059801e-01 6.77500546e-01 5.48954606e-01 6.04006946e-01
9.05434191e-01 -2.57262945e-01 6.78579956e-02 5.26845157e-01
8.57513249e-01 -1.24319911e+00 2.77451843e-01 6.75410509e-01
7.17254281e-01 -9.39869046e-01 -2.70475388e-01 6.08757794e-01
-7.65682876e-01 9.49209690e-01 3.06559533e-01 -4.83834743e-01
5.37962437e-01 4.26551610e-01 -3.43191952e-01 3.20931643e-01
-1.08966434e+00 -7.16477036e-01 -5.88882387e-01 7.97672272e-01
1.77832931e-01 3.54307026e-01 -4.77521122e-01 1.44299805e-01
-4.99322079e-02 1.80993795e-01 6.97605014e-01 9.41093564e-01
-7.82028377e-01 -1.41714299e+00 -3.25105608e-01 1.21230759e-01
-3.76994699e-01 2.36972034e-01 -1.89073738e-02 9.39545155e-01
-2.43400052e-01 9.72416461e-01 1.18567739e-02 1.80264682e-01
2.73300916e-01 5.15163466e-02 6.33110225e-01 -4.10558641e-01
-2.68733650e-01 2.20879346e-01 -3.38623613e-01 -5.17429233e-01
-4.20703411e-01 -9.64953661e-01 -1.16932201e+00 -2.86912471e-01
2.22314715e-01 4.20662791e-01 4.45053101e-01 9.49442387e-01
3.57469358e-02 4.56512630e-01 1.00427306e+00 -6.06494308e-01
-1.44188654e+00 -7.53924251e-01 -9.78188753e-01 3.08803283e-02
7.13587463e-01 -7.69765973e-01 -4.49603766e-01 -4.71277118e-01] | [4.166033744812012, 2.4865119457244873] |
7273bda7-1a81-40e7-abd7-446dda1a1933 | the-acl-ocl-corpus-advancing-open-science-in | 2305.14996 | null | https://arxiv.org/abs/2305.14996v1 | https://arxiv.org/pdf/2305.14996v1.pdf | The ACL OCL Corpus: advancing Open science in Computational Linguistics | We present a scholarly corpus from the ACL Anthology to assist Open scientific research in the Computational Linguistics domain, named as ACL OCL. Compared with previous ARC and AAN versions, ACL OCL includes structured full-texts with logical sections, references to figures, and links to a large knowledge resource (semantic scholar). ACL OCL contains 74k scientific papers, together with 210k figures extracted up to September 2022. To observe the development in the computational linguistics domain, we detect the topics of all OCL papers with a supervised neural model. We observe ''Syntax: Tagging, Chunking and Parsing'' topic is significantly shrinking and ''Natural Language Generation'' is resurging. Our dataset is open and available to download from HuggingFace in https://huggingface.co/datasets/ACL-OCL/ACL-OCL-Corpus. | ['Min-Yen Kan', 'Niranjana Unnithan', 'Benjamin Aw', 'Yanxia Qin', 'Shaurya Rohatgi'] | 2023-05-24 | null | null | null | null | ['chunking'] | ['natural-language-processing'] | [-5.36316395e-01 6.89313710e-01 -6.08252287e-01 -4.51534316e-02
-1.11575389e+00 -7.87194073e-01 6.64675295e-01 5.46793163e-01
-3.85063887e-01 1.07810581e+00 6.94789469e-01 -5.20195544e-01
-1.50148615e-01 -4.57213014e-01 -9.91944313e-01 8.34565982e-02
1.56338230e-01 6.84342027e-01 -9.79550853e-02 -7.68032670e-02
9.18207645e-01 6.60603642e-02 -1.20850396e+00 2.54335821e-01
1.14742303e+00 2.37900540e-01 4.43628728e-01 6.21814728e-01
-7.37039983e-01 6.54912114e-01 -8.52290690e-01 -8.10499847e-01
-2.51665473e-01 -2.85615027e-01 -1.19145751e+00 -4.96885180e-01
7.67711639e-01 3.44185263e-01 -6.03734553e-01 9.56211805e-01
2.66302168e-01 -1.77266285e-01 3.19508433e-01 -1.24789631e+00
-7.87240863e-01 1.73884106e+00 -4.21979547e-01 4.45120513e-01
5.43219626e-01 -3.38094085e-01 1.25703645e+00 -9.01054323e-01
1.42150235e+00 1.45383251e+00 6.26179516e-01 4.53144729e-01
-6.07774913e-01 -7.43089736e-01 -1.82008728e-01 3.74067724e-01
-1.06770194e+00 -5.86005270e-01 8.58408570e-01 -5.38503408e-01
8.24304640e-01 2.04071845e-03 7.18025684e-01 1.28453577e+00
8.60102400e-02 9.96140122e-01 1.05215931e+00 -8.40141356e-01
9.11522210e-02 2.06816137e-01 5.48381507e-01 6.16190016e-01
6.48015976e-01 -4.88837987e-01 -1.02143681e+00 -9.76244435e-02
4.74781275e-01 -6.32682860e-01 -8.84226933e-02 4.84288663e-01
-1.19062662e+00 5.73801458e-01 -3.90276998e-01 6.54791892e-01
-8.68901983e-02 1.41436867e-02 6.71268761e-01 1.39411002e-01
6.73863053e-01 6.48834169e-01 -4.42431957e-01 -6.53390348e-01
-9.97892082e-01 6.92988217e-01 1.31788719e+00 1.43460000e+00
5.38394332e-01 -2.73564637e-01 2.60273516e-01 1.12150645e+00
2.84908533e-01 5.15014291e-01 4.91631866e-01 -1.51949680e+00
8.41329992e-01 6.31481528e-01 -2.20388129e-01 -9.40229058e-01
-3.15006495e-01 -3.63096297e-01 -3.10636699e-01 -6.01967752e-01
7.02307820e-01 -1.74154013e-01 -3.74780774e-01 1.49073648e+00
-1.40902482e-03 -4.10567641e-01 2.77214289e-01 3.75143468e-01
1.45809734e+00 9.16215599e-01 3.24657917e-01 -3.24104488e-01
1.68034136e+00 -7.27816701e-01 -1.16439331e+00 -1.21776395e-01
1.06521356e+00 -1.01149380e+00 9.49415088e-01 2.85078764e-01
-1.57898879e+00 -1.96318656e-01 -9.36240017e-01 -6.28554940e-01
-8.61083388e-01 2.96615601e-01 5.50075293e-01 3.19527835e-01
-9.62617815e-01 4.94162798e-01 -4.83042508e-01 -6.71626210e-01
6.33758366e-01 -4.14038718e-01 -1.88917652e-01 -2.17359271e-02
-1.37431371e+00 7.37184107e-01 5.93647838e-01 -3.15966457e-01
-2.41544902e-01 -1.18883467e+00 -8.28319907e-01 -2.14564919e-01
4.94106352e-01 1.94610968e-01 1.18295574e+00 -1.28485158e-01
-1.13575101e+00 1.30496955e+00 -2.35565737e-01 -4.89564836e-01
5.08808017e-01 -5.00278234e-01 -4.71530288e-01 2.90840030e-01
6.72396839e-01 6.08657062e-01 -3.92552055e-02 -1.04763460e+00
-7.47870922e-01 -2.23402813e-01 -2.26876408e-01 -1.19160309e-01
-4.82653797e-01 6.56399071e-01 -6.37551248e-01 -9.59017098e-01
-1.82947353e-01 -6.64701700e-01 1.94648370e-01 -4.30174440e-01
-5.76828599e-01 -1.00359321e+00 7.38922954e-01 -1.18665016e+00
1.52999222e+00 -1.75932944e+00 -2.32592635e-02 -2.75176793e-01
4.93353039e-01 -7.60481060e-02 2.36678317e-01 7.45525241e-01
2.31641933e-01 6.38098240e-01 -6.30804747e-02 -2.57418752e-01
4.43250716e-01 1.83769569e-01 -2.83002317e-01 4.77694362e-01
-2.61702091e-01 9.52063441e-01 -9.54502344e-01 -1.01469958e+00
-3.13348144e-01 1.14913739e-01 -2.28825063e-01 -2.18525589e-01
-4.61121798e-01 -6.74202517e-02 -4.74173784e-01 8.36200535e-01
3.89691174e-01 -2.30108887e-01 2.65807420e-01 1.77649602e-01
-7.43951142e-01 8.07078660e-01 -7.69029498e-01 2.00422025e+00
-3.15845519e-01 1.11482298e+00 1.89778432e-01 -9.59288120e-01
1.06172037e+00 4.05816436e-01 2.33078673e-01 -5.82473576e-01
2.04967096e-01 5.65229297e-01 -1.39545202e-01 -4.29420084e-01
7.32133389e-01 5.09557486e-01 -5.62905610e-01 7.23405421e-01
3.45107019e-01 -1.55471399e-01 8.54877710e-01 8.99124086e-01
9.63503420e-01 3.91419202e-01 1.64457604e-01 -7.46342301e-01
4.10537124e-01 5.62830269e-01 7.61270583e-01 6.94669127e-01
-1.75800771e-01 -4.32488173e-02 8.23525846e-01 -3.97801101e-01
-1.51117277e+00 -6.14912450e-01 -7.16943204e-01 9.53642190e-01
-3.18748146e-01 -9.22251821e-01 -1.03193355e+00 -4.32001054e-01
3.99768576e-02 9.29116428e-01 -5.56172073e-01 5.16512215e-01
-1.00409901e+00 -3.35914671e-01 1.06848156e+00 2.58511245e-01
4.65592057e-01 -1.49112153e+00 -1.97446376e-01 1.66766852e-01
-3.67987961e-01 -1.46806479e+00 -1.23509824e-01 9.44569185e-02
-5.13724148e-01 -9.81610775e-01 -6.94593132e-01 -9.32970166e-01
2.19050497e-01 -5.71478724e-01 1.12278748e+00 -5.34897596e-02
-5.56401730e-01 1.04850516e-01 -4.98154581e-01 -9.46220100e-01
-8.86008501e-01 5.05243838e-01 -7.49243945e-02 -1.07457590e+00
5.50972879e-01 -2.14472502e-01 1.45855442e-01 -4.25192058e-01
-3.33752453e-01 -4.49927263e-02 2.55893201e-01 3.66723657e-01
1.88668519e-01 -2.48663008e-01 8.37024152e-01 -1.13015878e+00
5.63363612e-01 -6.15545392e-01 -3.89492661e-01 2.61055022e-01
-7.13874161e-01 -2.26935491e-01 3.97595644e-01 6.33800328e-02
-1.11360061e+00 -7.30712414e-01 5.70725091e-02 1.31424563e-02
-4.54173565e-01 7.51155138e-01 -1.73546270e-01 5.84154308e-01
3.94702226e-01 -1.40081644e-01 -9.11357924e-02 -9.17912602e-01
3.04278791e-01 8.91433775e-01 8.24445605e-01 -1.06236899e+00
6.97341681e-01 2.33559273e-02 -1.54598862e-01 -1.18286312e+00
-9.54662681e-01 -2.40796715e-01 -7.76553810e-01 -5.10940313e-01
5.61414123e-01 -1.00116622e+00 -6.87769234e-01 2.16273859e-01
-1.34024024e+00 -2.97441095e-01 -4.29243982e-01 4.72024500e-01
-3.24395984e-01 4.54801917e-01 -1.22772980e+00 -4.87707287e-01
-3.96562576e-01 -6.33021474e-01 7.49162555e-01 4.44699377e-01
-5.71742773e-01 -1.18703365e+00 2.34047785e-01 6.75427437e-01
-1.80952162e-01 1.33540347e-01 1.04029250e+00 -9.40550089e-01
-8.53475183e-02 3.21837165e-03 -1.97150216e-01 -2.32410446e-01
-5.19309044e-01 2.99779326e-01 -6.03194714e-01 1.27839193e-01
-5.96738875e-01 -4.38151836e-01 6.16757393e-01 2.36041859e-01
8.54268193e-01 -6.87949538e-01 -4.66405839e-01 2.01371923e-01
1.23703897e+00 2.09461868e-01 7.16893852e-01 9.78013873e-01
6.12679720e-01 8.33071709e-01 3.62028688e-01 4.41658258e-01
7.17073619e-01 2.11899400e-01 -4.38255489e-01 3.47437561e-01
-3.62863183e-01 -4.81924117e-01 2.18844086e-01 1.57781374e+00
-1.94806814e-01 -2.81468570e-01 -1.51074100e+00 8.74126792e-01
-1.82275796e+00 -8.43902588e-01 -4.77135181e-01 1.45306897e+00
1.24999261e+00 2.40754664e-01 -1.38679589e-03 -5.99543229e-02
8.58728886e-01 1.38722092e-01 -1.80249021e-03 -8.12519908e-01
-4.44401234e-01 4.35917169e-01 4.53264982e-01 5.42026579e-01
-7.70506322e-01 1.42739010e+00 6.21854115e+00 1.21324801e+00
-4.44545090e-01 3.38459432e-01 2.90232450e-01 9.26330388e-02
-3.95908326e-01 3.43299687e-01 -1.31035841e+00 6.85693920e-01
1.53934538e+00 -6.54220283e-01 -1.53896570e-01 6.93480194e-01
2.76506066e-01 -4.81436931e-04 -5.66872597e-01 3.59471023e-01
7.20606446e-02 -1.99483597e+00 -5.17779440e-02 3.02415788e-01
4.02143061e-01 2.53998265e-02 -3.73454988e-01 4.82658505e-01
5.79554915e-01 -8.14537823e-01 9.50513959e-01 4.25294399e-01
9.27854419e-01 -5.78305304e-01 5.62946141e-01 1.65398970e-01
-7.21023381e-01 3.65583390e-01 -4.58854884e-01 -1.37976125e-01
6.56286478e-02 6.35941923e-01 -4.85446483e-01 5.22394776e-01
9.18702543e-01 1.14072847e+00 -6.32604003e-01 7.55280554e-01
-4.51651543e-01 1.20737529e+00 -1.61443681e-01 -4.65379566e-01
2.22289279e-01 -1.68508589e-01 9.28460419e-01 1.72632372e+00
7.64220729e-02 1.15158036e-01 1.77462459e-01 1.06635630e+00
-5.87195575e-01 5.70796788e-01 -3.71598840e-01 -7.27349102e-01
1.02806592e+00 1.19246268e+00 -1.07232738e+00 -4.15032655e-01
-2.52226353e-01 4.12479728e-01 5.17130196e-01 1.81430459e-01
-5.71355820e-01 -9.49459374e-01 5.29606752e-02 2.37300813e-01
-1.21234410e-01 -2.23921672e-01 -5.17413080e-01 -9.49066818e-01
-1.59917071e-01 -5.73457062e-01 3.30130041e-01 -6.59631908e-01
-1.16606188e+00 2.37368330e-01 3.88158560e-02 -5.72005868e-01
1.89713798e-02 -4.66065258e-01 -4.59570646e-01 7.09990203e-01
-1.02933490e+00 -1.19663751e+00 1.57621533e-01 -1.07921496e-01
7.51311600e-01 -4.68818188e-01 5.39376616e-01 4.05663937e-01
-7.98204124e-01 6.61428511e-01 1.42149955e-01 5.71113706e-01
9.53811407e-01 -1.37589002e+00 3.16335827e-01 6.70721233e-01
-4.02570628e-02 4.91627425e-01 7.43059695e-01 -1.11121333e+00
-1.41232312e+00 -7.19556689e-01 1.74473023e+00 -6.15073800e-01
1.35650313e+00 -3.82902116e-01 -9.45077896e-01 7.76092529e-01
8.12429965e-01 -6.10726774e-01 6.58377528e-01 2.10823625e-01
-2.79690862e-01 4.13074523e-01 -7.04390407e-01 5.68698883e-01
9.82441366e-01 -4.28329229e-01 -1.13143837e+00 6.43248916e-01
9.56139147e-01 -5.49636185e-01 -1.39357555e+00 -7.17118308e-02
6.36490941e-01 -1.51342109e-01 5.22655368e-01 -4.96017367e-01
1.19464552e+00 2.81963229e-01 1.48531094e-01 -8.13577056e-01
-1.06286310e-01 -8.81245255e-01 -2.64394619e-02 1.76845872e+00
6.65988445e-01 -2.94677913e-01 8.26033950e-01 2.72603720e-01
-7.12681711e-01 -4.84655857e-01 -9.55680072e-01 -5.83221674e-01
8.90621543e-01 -6.01175904e-01 1.76077843e-01 1.43367183e+00
7.89342821e-01 3.37729573e-01 1.51876822e-01 -4.00194019e-01
1.00639415e+00 -9.93970558e-02 3.53995383e-01 -1.54382026e+00
1.92628786e-01 -8.01353157e-01 6.20290861e-02 -4.78952736e-01
7.55427837e-01 -1.24379611e+00 -1.81371063e-01 -1.72856319e+00
2.13050172e-01 -6.11926019e-01 3.84526134e-01 7.23666012e-01
1.62730038e-01 1.43637015e-02 9.61564928e-02 6.11380458e-01
-5.70284247e-01 -4.11445908e-02 1.09484911e+00 6.42197877e-02
-1.20563500e-01 -8.37884367e-01 -6.85163200e-01 9.48481441e-01
8.78046095e-01 -5.02925932e-01 5.71593761e-01 -7.84997791e-02
4.41287935e-01 1.25056475e-01 4.14573476e-02 -8.17441583e-01
3.23954254e-01 -1.96994275e-01 2.98105896e-01 -7.93842018e-01
-2.26760313e-01 8.28943178e-02 -3.86550277e-01 2.57858068e-01
-7.09738195e-01 2.60679964e-02 4.56950963e-01 -2.69536627e-03
-7.58287385e-02 -5.94217956e-01 3.80065620e-01 -4.20335114e-01
-5.51634669e-01 -1.16506144e-01 -7.81332612e-01 9.52139676e-01
6.35637581e-01 -7.58113936e-02 -1.02121651e+00 -8.26660618e-02
-6.66100562e-01 4.09345418e-01 3.37755978e-01 4.01783824e-01
1.82731539e-01 -8.19336414e-01 -1.04325771e+00 -4.44547355e-01
-8.58189389e-02 -1.84203863e-01 1.77410036e-01 5.91102183e-01
-8.15572143e-01 9.63245213e-01 1.21743642e-02 9.67513844e-02
-7.35940754e-01 2.43833765e-01 -5.26328206e-01 -2.71619469e-01
-1.04078400e+00 6.70336843e-01 -5.06612539e-01 -4.03202504e-01
5.83092153e-01 1.46024942e-01 -4.72062498e-01 2.25462139e-01
4.40941960e-01 8.31264913e-01 -2.80338556e-01 -6.21669173e-01
-2.43703678e-01 1.37919024e-01 -1.07704014e-01 -3.07479769e-01
1.55110312e+00 -3.32789510e-01 -4.74373996e-01 8.96075368e-01
1.12352777e+00 5.32686114e-01 -1.99422136e-01 -1.52510986e-01
4.68664587e-01 2.73411423e-01 7.33267218e-02 -1.01056445e+00
-5.92992127e-01 7.34549582e-01 -2.11182296e-01 1.52354568e-01
1.33899167e-01 4.80940074e-01 9.04006302e-01 3.37947965e-01
-9.97919738e-02 -1.68511724e+00 -2.75510967e-01 8.16434324e-01
1.01422274e+00 -7.98543334e-01 3.35617006e-01 -4.73851740e-01
-4.77119982e-01 1.30840588e+00 5.94417274e-01 1.24293426e-02
7.86865056e-01 7.14752793e-01 7.44626448e-02 -5.22306263e-01
-7.30342448e-01 2.85904884e-01 -4.80467118e-02 3.05828542e-01
7.44936764e-01 -2.76223011e-02 -1.05834806e+00 8.41699898e-01
-1.01350820e+00 -1.09151401e-01 1.11594903e+00 9.40404713e-01
-4.36968356e-01 -1.15396070e+00 -4.13269669e-01 1.60708040e-01
-9.01480317e-01 -3.39421511e-01 -4.87403065e-01 1.04877305e+00
-3.49733718e-02 5.52051902e-01 5.97001433e-01 1.72920182e-01
-2.52928644e-01 4.95418787e-01 3.39280427e-01 -6.30926967e-01
-5.49294829e-01 3.90440762e-01 5.73655546e-01 -1.43770531e-01
-6.20041966e-01 -1.12763536e+00 -1.53901196e+00 -6.98831618e-01
6.88272268e-02 8.19166839e-01 7.51223803e-01 9.79595780e-01
3.06011528e-01 5.24549842e-01 -2.11882621e-01 -5.24801910e-01
2.57129192e-01 -1.22268736e+00 -7.55361319e-01 -2.14306191e-01
-3.76940906e-01 -4.95851219e-01 -6.41705990e-01 3.17117184e-01] | [9.779560089111328, 8.540742874145508] |
c6e2c46c-f3f2-4903-adcd-7284a43d8254 | quad-networks-unsupervised-learning-to-rank | 1611.07571 | null | http://arxiv.org/abs/1611.07571v2 | http://arxiv.org/pdf/1611.07571v2.pdf | Quad-networks: unsupervised learning to rank for interest point detection | Several machine learning tasks require to represent the data using only a
sparse set of interest points. An ideal detector is able to find the
corresponding interest points even if the data undergo a transformation typical
for a given domain. Since the task is of high practical interest in computer
vision, many hand-crafted solutions were proposed. In this paper, we ask a
fundamental question: can we learn such detectors from scratch? Since it is
often unclear what points are "interesting", human labelling cannot be used to
find a truly unbiased solution. Therefore, the task requires an unsupervised
formulation. We are the first to propose such a formulation: training a neural
network to rank points in a transformation-invariant manner. Interest points
are then extracted from the top/bottom quantiles of this ranking. We validate
our approach on two tasks: standard RGB image interest point detection and
challenging cross-modal interest point detection between RGB and depth images.
We quantitatively show that our unsupervised method performs better or on-par
with baselines. | ['Marc Pollefeys', 'Torsten Sattler', 'Lubor Ladicky', 'Akihito Seki', 'Nikolay Savinov'] | 2016-11-22 | quad-networks-unsupervised-learning-to-rank-1 | http://openaccess.thecvf.com/content_cvpr_2017/html/Savinov_Quad-Networks_Unsupervised_Learning_CVPR_2017_paper.html | http://openaccess.thecvf.com/content_cvpr_2017/papers/Savinov_Quad-Networks_Unsupervised_Learning_CVPR_2017_paper.pdf | cvpr-2017-7 | ['interest-point-detection'] | ['computer-vision'] | [ 3.31813842e-01 1.23066135e-01 -8.68841708e-02 -4.04279530e-01
-1.25511980e+00 -5.47040641e-01 7.68435657e-01 3.88451874e-01
-5.87946951e-01 3.23948234e-01 -2.05677822e-01 8.67410898e-02
-1.17438316e-01 -5.74716985e-01 -7.71314740e-01 -7.51325071e-01
9.39594209e-03 7.36963809e-01 5.91820598e-01 -5.04268780e-02
5.44686198e-01 7.77350605e-01 -1.75994575e+00 5.33896163e-02
4.83370572e-01 1.15071154e+00 2.65664667e-01 3.41413528e-01
-2.21825968e-02 3.32000166e-01 -2.24577010e-01 4.15108502e-02
5.35912454e-01 -3.52562070e-01 -9.31793869e-01 2.30744958e-01
6.71279252e-01 -4.95576113e-02 2.05572426e-01 1.17369294e+00
3.48677307e-01 1.83972165e-01 9.74952042e-01 -1.11720002e+00
2.56328434e-02 8.90802741e-02 -7.61589587e-01 -1.81080913e-03
3.42285484e-01 -9.41393822e-02 1.32221818e+00 -1.05425143e+00
7.34092891e-01 8.87518644e-01 5.62583387e-01 4.74685758e-01
-1.33450246e+00 -1.50768861e-01 5.01212524e-03 -1.89660955e-02
-1.33865452e+00 -4.67647284e-01 1.04651213e+00 -5.21861911e-01
4.26460505e-01 1.90031737e-01 3.99425030e-01 7.71974325e-01
-2.12496206e-01 9.54314947e-01 1.13421512e+00 -6.29348695e-01
5.07020414e-01 2.60962963e-01 3.83942686e-02 6.16151214e-01
2.02881724e-01 -3.92559431e-02 -5.75553894e-01 -1.17907830e-01
7.40361869e-01 4.17649671e-02 -3.17730635e-01 -1.12427139e+00
-1.37580073e+00 8.70659232e-01 8.32784235e-01 3.13863218e-01
-3.88414502e-01 7.40932450e-02 7.46276751e-02 2.20410898e-02
3.26220453e-01 6.28029406e-01 -3.56780916e-01 2.14755937e-01
-1.22486913e+00 1.90805942e-01 6.41537786e-01 7.54224181e-01
1.06621313e+00 -4.99709666e-01 2.28999153e-01 6.86286569e-01
3.92759115e-01 3.73340160e-01 3.52081597e-01 -9.57817316e-01
1.79595742e-02 6.94247425e-01 2.17650309e-01 -9.21921611e-01
-4.91417736e-01 -2.93977976e-01 -6.08749747e-01 5.76382518e-01
8.93115878e-01 2.32896239e-01 -7.91909873e-01 1.50506091e+00
5.29449403e-01 -1.68505147e-01 -2.15700060e-01 1.07699490e+00
5.37992895e-01 4.54421312e-01 -2.58686572e-01 -4.49867062e-02
1.21872330e+00 -4.22149807e-01 -9.43069085e-02 -2.95892119e-01
5.61689675e-01 -6.94465697e-01 1.07984519e+00 5.40097654e-01
-8.57713163e-01 -3.34037781e-01 -9.00793850e-01 -1.25892758e-01
-2.05765963e-01 2.54074991e-01 5.59809029e-01 3.33937556e-01
-7.39938736e-01 6.04879916e-01 -8.86952043e-01 -6.63818538e-01
3.50480169e-01 4.97230262e-01 -5.53900421e-01 1.16978250e-01
-6.10485733e-01 8.34397554e-01 2.73605943e-01 2.56366320e-02
-6.76010966e-01 -3.53789300e-01 -7.15772867e-01 -2.30640113e-01
4.06914562e-01 -5.92989087e-01 1.15225530e+00 -1.05393946e+00
-1.21838248e+00 1.30895877e+00 -3.05001050e-01 -4.32839960e-01
6.25722051e-01 -8.29357654e-02 2.15095177e-01 2.22259939e-01
2.18868986e-01 6.68666124e-01 9.41175580e-01 -1.58396983e+00
-9.45192218e-01 -5.60310066e-01 3.53210568e-02 1.90134626e-02
-7.66455978e-02 -3.56535949e-02 -3.40928704e-01 -2.51766890e-01
7.37413585e-01 -9.72837687e-01 -2.11204857e-01 3.04263115e-01
-4.83460009e-01 -3.93671393e-01 7.04937935e-01 -8.60289037e-02
5.33113778e-01 -2.04103112e+00 1.03445761e-01 3.90814602e-01
2.15201318e-01 -1.02030896e-01 2.85530418e-01 3.02040279e-01
-6.27838715e-04 -1.41377300e-01 -4.15244460e-01 -2.21026734e-01
8.23146552e-02 1.74762961e-02 -3.16046059e-01 9.62833762e-01
2.01446578e-01 6.51731610e-01 -1.00255132e+00 -6.51954114e-01
4.17603940e-01 3.76156330e-01 -4.93267089e-01 6.80275634e-02
-2.25616813e-01 3.92146796e-01 -5.67580819e-01 6.79793179e-01
4.51603919e-01 -2.74922639e-01 -2.57013232e-01 -3.97911608e-01
-7.89049789e-02 1.34300068e-01 -1.40878057e+00 1.70464814e+00
-2.64271885e-01 7.57232368e-01 -1.25513285e-01 -1.23692060e+00
1.06214643e+00 -3.97262797e-02 7.78729498e-01 -4.63742554e-01
1.80616394e-01 4.21124905e-01 -1.40012935e-01 -2.61541754e-01
5.25976241e-01 -4.02342170e-01 -1.01926692e-01 2.93983847e-01
5.37705049e-03 -3.36646885e-01 -5.15005663e-02 2.62078233e-02
1.10974979e+00 3.23412158e-02 4.38764274e-01 -3.27248871e-01
5.08076906e-01 1.30645007e-01 3.20597291e-01 7.44033337e-01
-1.38493598e-01 1.10531259e+00 5.16116917e-01 -5.64277768e-01
-8.64763141e-01 -1.11765313e+00 -3.21678698e-01 8.39169800e-01
3.83593410e-01 -1.07413433e-01 -6.95057333e-01 -8.07959318e-01
-7.46968687e-02 3.09125483e-01 -6.26532197e-01 7.51121640e-02
-4.48900253e-01 -4.71036166e-01 1.18866581e-02 1.61508396e-01
2.20439121e-01 -1.06846547e+00 -9.90666628e-01 -3.84270921e-02
-3.35272551e-02 -1.01690733e+00 -1.21555611e-01 4.32834208e-01
-8.98139298e-01 -1.20631421e+00 -8.35521817e-01 -7.52124846e-01
1.01920736e+00 3.32469285e-01 1.12234449e+00 -6.19463511e-02
-2.15168566e-01 3.73308480e-01 -3.76050889e-01 -4.05774415e-01
-1.00164905e-01 1.90240726e-01 3.14659514e-02 1.87447280e-01
4.88371849e-01 -3.32136780e-01 -6.74426854e-01 3.35099846e-01
-8.65980685e-01 -4.28072006e-01 7.37586737e-01 7.08460808e-01
8.15075576e-01 -1.89076345e-02 2.19891235e-01 -7.07876384e-01
2.23499075e-01 -2.26652976e-02 -7.76003599e-01 8.17092955e-02
-1.38579994e-01 3.75207812e-01 4.87822443e-01 -1.15606628e-01
-4.62224007e-01 9.76229668e-01 -1.62091285e-01 -2.97160417e-01
-4.59683210e-01 2.07486749e-01 -1.84655070e-01 -2.84420133e-01
8.54539454e-01 1.37361333e-01 -1.35715470e-01 -4.42034662e-01
4.28733528e-01 5.25347590e-01 6.45147622e-01 -6.43471360e-01
9.55802083e-01 8.35537612e-01 2.70956993e-01 -1.03579021e+00
-7.86670923e-01 -9.92640853e-01 -9.93650615e-01 -2.62593269e-01
6.88602567e-01 -6.29405558e-01 -5.39543390e-01 6.24754690e-02
-1.21343565e+00 -7.57464543e-02 -5.96053720e-01 3.22302490e-01
-9.14336562e-01 4.26197141e-01 4.59370762e-02 -8.65347505e-01
8.12168606e-03 -1.15913033e+00 1.42849314e+00 1.00545764e-01
-3.01778615e-01 -7.84008384e-01 2.43174672e-01 1.24495383e-02
-9.22486410e-02 4.57124799e-01 5.19096971e-01 -6.57137871e-01
-6.14168704e-01 -5.32407880e-01 -1.86320573e-01 1.55798957e-01
1.56096891e-01 8.40709433e-02 -1.13209391e+00 -1.82937458e-01
6.04074672e-02 -3.27054858e-01 1.00590050e+00 5.74502349e-01
1.09095466e+00 1.61021709e-01 -3.91459554e-01 5.28938293e-01
1.47396636e+00 -2.44902581e-01 4.25682455e-01 5.37262857e-01
4.94440883e-01 8.81166041e-01 6.51382148e-01 3.23849618e-01
1.92589775e-01 8.30633104e-01 6.94407403e-01 -2.02626899e-01
2.49792829e-01 -2.58546382e-01 3.50352973e-02 2.06590727e-01
-1.28164589e-01 2.99387984e-02 -1.06616843e+00 7.58906424e-01
-1.83317304e+00 -6.95504367e-01 -2.89254904e-01 2.49875998e+00
5.01647890e-01 2.57922560e-01 3.37148607e-01 4.10623908e-01
4.54965681e-01 -1.44946769e-01 -4.88859564e-01 1.50066718e-01
3.66676487e-02 1.89547211e-01 6.27789915e-01 2.99942195e-01
-1.33941090e+00 7.62866795e-01 5.81167793e+00 5.48046887e-01
-1.19559157e+00 -1.78806856e-01 5.23802698e-01 1.71658784e-01
-3.49855646e-02 2.26857841e-01 -7.27615058e-01 2.96923041e-01
4.47833210e-01 2.06102461e-01 -2.00172573e-01 9.92784679e-01
8.55903849e-02 -4.02416617e-01 -1.45419407e+00 1.19424689e+00
-1.20087124e-01 -1.01168036e+00 -2.07389459e-01 1.91256568e-01
6.09809935e-01 1.19335145e-01 1.17175177e-01 -1.89419121e-01
1.59878626e-01 -8.38433623e-01 5.99455416e-01 6.06460094e-01
3.34655285e-01 -5.74042439e-01 5.22210062e-01 5.79365253e-01
-9.76730645e-01 2.60817349e-01 -5.06668031e-01 1.97126105e-01
1.18033491e-01 6.87799871e-01 -8.06705415e-01 3.44729066e-01
4.91539031e-01 6.53607488e-01 -6.37547314e-01 1.51077282e+00
-1.76403373e-01 3.41484070e-01 -7.52494872e-01 -3.50847058e-02
3.20454091e-01 8.65329057e-03 5.22832096e-01 1.01489139e+00
4.71047461e-01 -1.26400039e-01 5.87887987e-02 7.43894517e-01
-5.43675236e-02 1.16281815e-01 -7.26001799e-01 2.95168072e-01
1.85377449e-01 1.52703118e+00 -1.34178102e+00 4.08155508e-02
-2.61076421e-01 9.29412305e-01 3.11470777e-01 1.11178532e-01
-3.57876271e-01 -2.80365348e-01 3.66129518e-01 5.08986354e-01
2.32609063e-01 -3.25974911e-01 -2.91945755e-01 -1.09337270e+00
6.97579160e-02 -5.57764769e-01 2.20502719e-01 -6.94131255e-01
-1.28928530e+00 4.22472388e-01 -9.05362368e-02 -1.57137609e+00
-5.11897027e-01 -8.54495883e-01 -4.34300780e-01 6.46306872e-01
-1.47192025e+00 -9.49414551e-01 -3.47131789e-01 5.30763507e-01
3.61705124e-01 1.12143472e-01 5.98343551e-01 7.23643154e-02
-6.97802305e-02 2.23574877e-01 2.37735882e-01 1.38975412e-01
7.14387476e-01 -1.53356147e+00 1.39324561e-01 8.65429342e-01
6.64869130e-01 6.16696000e-01 9.96515691e-01 -1.85201854e-01
-1.30852330e+00 -7.56493568e-01 7.63628364e-01 -8.20065320e-01
4.60606039e-01 -4.63814944e-01 -9.06286418e-01 4.56787676e-01
-2.46933743e-01 5.24103761e-01 2.32276052e-01 -3.56950574e-02
-1.45869523e-01 -1.74765259e-01 -1.06704450e+00 3.07686239e-01
6.87133610e-01 -5.33339500e-01 -6.06622159e-01 5.00784457e-01
1.19653024e-01 -2.40057886e-01 -4.81210530e-01 3.63816291e-01
3.99577975e-01 -1.11537039e+00 1.22863829e+00 -2.55060494e-01
3.35442573e-01 -5.20855367e-01 -2.09321350e-01 -1.18764699e+00
9.41145271e-02 -3.97335529e-01 1.13521598e-01 1.00654137e+00
2.50796974e-01 -3.62201959e-01 1.16313243e+00 5.57904661e-01
1.70836151e-01 -6.40893102e-01 -1.06762052e+00 -7.03164518e-01
-2.91582737e-02 -4.92683858e-01 1.89001933e-01 7.67757893e-01
-1.51738554e-01 3.80190790e-01 -1.61610365e-01 2.84273028e-01
9.24379647e-01 2.97870547e-01 1.00065732e+00 -1.61124468e+00
-3.92017886e-02 -5.41695476e-01 -8.39807630e-01 -1.22203410e+00
7.29616452e-03 -7.72008598e-01 3.67957830e-01 -1.57178974e+00
-1.09456116e-02 -6.81193352e-01 -2.66441464e-01 2.94181556e-01
8.70160758e-02 2.62425989e-01 2.54906137e-02 5.15694201e-01
-7.27537155e-01 3.55342954e-01 7.07435966e-01 -6.64313212e-02
-1.86272532e-01 2.60893732e-01 -5.51288247e-01 9.75024462e-01
6.47429049e-01 -5.72756708e-01 -1.36237279e-01 -1.62752435e-01
2.45544955e-01 -3.36593449e-01 6.13619804e-01 -1.21678770e+00
5.21435201e-01 -1.44565906e-02 3.10275853e-01 -7.91231990e-01
3.71347129e-01 -1.14148045e+00 -1.57327816e-01 2.30728865e-01
-3.28205436e-01 -2.77181298e-01 -2.16368482e-01 5.47604024e-01
-2.57913589e-01 -6.29955530e-01 1.01816094e+00 -2.76266694e-01
-8.77432466e-01 2.27122083e-01 -8.69742632e-02 -3.17192562e-02
9.59622681e-01 -3.02215338e-01 2.13871762e-01 -3.74120235e-01
-6.21263146e-01 -3.17760333e-02 7.70160437e-01 1.21795684e-01
7.37362087e-01 -1.14374340e+00 -4.34590697e-01 1.65732160e-01
5.02064824e-01 3.54446650e-01 -2.34485850e-01 6.93216443e-01
-4.96284634e-01 3.04365575e-01 -4.33039889e-02 -1.04535758e+00
-1.16993761e+00 4.81559604e-01 4.04956639e-01 3.74277309e-02
-5.82007647e-01 8.81414294e-01 4.02010173e-01 -4.13100749e-01
2.07943410e-01 -5.54604530e-01 -7.15778247e-02 1.68798506e-01
2.91088521e-01 1.18807301e-01 1.70435995e-01 -1.05232763e+00
-4.46639180e-01 9.29259777e-01 1.16024964e-01 -1.51901454e-01
1.36837375e+00 -4.51333299e-02 8.92842188e-02 5.56846738e-01
1.35046721e+00 -9.46122333e-02 -1.28707159e+00 -4.78238165e-01
4.19328630e-01 -5.49954772e-01 7.92458132e-02 -3.08852285e-01
-8.49417865e-01 1.07468033e+00 6.85450912e-01 5.03107846e-01
1.07890451e+00 4.01057392e-01 3.75051528e-01 5.56267321e-01
6.01914287e-01 -9.87660170e-01 -1.51470490e-02 2.37305298e-01
7.45146513e-01 -1.67358267e+00 1.45134911e-01 -2.38216415e-01
-3.57289016e-01 1.25490713e+00 2.11779967e-01 -4.33420837e-01
5.26203215e-01 -8.13968778e-02 4.33707759e-02 -4.17314559e-01
-2.82413453e-01 -6.19411826e-01 5.20984292e-01 6.21889949e-01
2.93341339e-01 -1.49055377e-01 -3.66009623e-02 6.66970015e-02
-2.00494260e-01 -1.63990468e-01 3.50947887e-01 8.93257201e-01
-8.44333470e-01 -1.11136377e+00 -5.46318293e-01 4.47130799e-01
-3.47571880e-01 2.89084166e-01 -5.87327719e-01 9.12282944e-01
8.44416842e-02 6.22905612e-01 -5.18024107e-03 -2.22544789e-01
4.68961209e-01 -5.05571254e-02 4.84017313e-01 -6.31394863e-01
-1.85143039e-01 1.97461098e-01 -4.56410468e-01 -5.60372829e-01
-7.40877986e-01 -1.01476288e+00 -1.29471624e+00 2.78108418e-01
-3.34529757e-01 7.52208978e-02 9.06713784e-01 9.73649383e-01
-1.32943183e-01 1.16520219e-01 7.01116383e-01 -1.05346775e+00
-4.19161230e-01 -5.84278822e-01 -6.75232232e-01 4.36222970e-01
5.42954266e-01 -7.79188275e-01 -5.17101049e-01 6.12683184e-02] | [8.065537452697754, -1.941015601158142] |
18a664c5-c5f4-4850-986d-f3c1305c74b9 | neuralfield-ldm-scene-generation-with | 2304.09787 | null | https://arxiv.org/abs/2304.09787v1 | https://arxiv.org/pdf/2304.09787v1.pdf | NeuralField-LDM: Scene Generation with Hierarchical Latent Diffusion Models | Automatically generating high-quality real world 3D scenes is of enormous interest for applications such as virtual reality and robotics simulation. Towards this goal, we introduce NeuralField-LDM, a generative model capable of synthesizing complex 3D environments. We leverage Latent Diffusion Models that have been successfully utilized for efficient high-quality 2D content creation. We first train a scene auto-encoder to express a set of image and pose pairs as a neural field, represented as density and feature voxel grids that can be projected to produce novel views of the scene. To further compress this representation, we train a latent-autoencoder that maps the voxel grids to a set of latent representations. A hierarchical diffusion model is then fit to the latents to complete the scene generation pipeline. We achieve a substantial improvement over existing state-of-the-art scene generation models. Additionally, we show how NeuralField-LDM can be used for a variety of 3D content creation applications, including conditional scene generation, scene inpainting and scene style manipulation. | ['Sanja Fidler', 'Antonio Torralba', 'Robin Rombach', 'Daiqing Li', 'Katja Schwarz', 'Karsten Kreis', 'Kangxue Yin', 'Bradley Brown', 'Seung Wook Kim'] | 2023-04-19 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Kim_NeuralField-LDM_Scene_Generation_With_Hierarchical_Latent_Diffusion_Models_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Kim_NeuralField-LDM_Scene_Generation_With_Hierarchical_Latent_Diffusion_Models_CVPR_2023_paper.pdf | cvpr-2023-1 | ['scene-generation'] | ['computer-vision'] | [ 4.94124353e-01 3.21436018e-01 4.57369417e-01 -3.50529015e-01
-4.93407100e-01 -4.81123418e-01 1.05820227e+00 -1.89265117e-01
1.74434111e-01 5.09317756e-01 5.33834100e-01 4.79354300e-02
1.94451481e-01 -1.21411920e+00 -1.08092451e+00 -4.55195874e-01
4.57089534e-03 8.71263325e-01 5.79336397e-02 -1.58816308e-01
1.38945743e-01 8.17494035e-01 -1.84326172e+00 3.88036817e-01
6.32000744e-01 6.15272760e-01 6.44019842e-01 8.49084318e-01
-2.36127988e-01 8.76195550e-01 -5.82437754e-01 -1.09787539e-01
2.68382996e-01 -4.51789051e-01 -6.43167555e-01 4.65770096e-01
4.86101359e-01 -7.87315130e-01 -4.76079017e-01 6.89748108e-01
4.01969701e-01 2.97152907e-01 9.04075980e-01 -9.47677851e-01
-6.92342281e-01 2.53041267e-01 -3.32572073e-01 -3.23076665e-01
7.19581068e-01 2.53431618e-01 5.67060828e-01 -9.89962339e-01
1.26495838e+00 1.53891766e+00 3.09217542e-01 5.76434553e-01
-1.58973587e+00 -2.18599543e-01 3.83090936e-02 -3.09787363e-01
-1.13499868e+00 -3.20931852e-01 1.06164396e+00 -5.80580294e-01
1.28328586e+00 4.29199822e-02 1.04259086e+00 1.26052749e+00
4.34306473e-01 7.73313820e-01 1.05938542e+00 -3.26436698e-01
4.76129919e-01 5.04391380e-02 -7.86281347e-01 7.14808583e-01
-2.85850316e-01 1.32578194e-01 -6.88430071e-01 -3.61407697e-02
1.53097856e+00 -3.84717956e-02 -1.29295420e-03 -8.18107069e-01
-1.34151745e+00 9.31205094e-01 6.45499110e-01 -1.37497351e-01
-7.46307433e-01 5.14859319e-01 -3.52534465e-02 -1.29980534e-01
7.81636000e-01 5.88595688e-01 -8.78721327e-02 3.24448049e-02
-8.34750175e-01 7.51007438e-01 6.35512769e-01 1.14960587e+00
7.07183063e-01 4.16727781e-01 -2.09362552e-01 7.28219271e-01
4.27158535e-01 4.23266858e-01 1.93899155e-01 -1.43269193e+00
6.72365725e-02 3.26922089e-01 2.44421139e-02 -9.11971450e-01
-1.65375452e-02 -2.07170919e-01 -7.91803241e-01 4.81104761e-01
-4.09217238e-01 1.72209010e-01 -1.24455273e+00 1.48551464e+00
5.20193100e-01 4.36688572e-01 -2.62958277e-02 7.30759859e-01
7.21579731e-01 1.05527341e+00 -1.15459844e-01 9.87479016e-02
9.03039098e-01 -8.37938130e-01 -5.11102855e-01 -2.14870155e-01
2.76592314e-01 -7.97723293e-01 8.97457600e-01 2.37380162e-01
-1.47617686e+00 -5.48181355e-01 -8.64428937e-01 -4.61274058e-01
-3.01694751e-01 -4.06502873e-01 7.51050234e-01 1.23469099e-01
-1.26630688e+00 5.16809046e-01 -9.40882504e-01 -2.70398736e-01
6.02458477e-01 -1.24468410e-03 -3.81381720e-01 -1.99679509e-01
-6.52298629e-01 8.14087033e-01 3.39211881e-01 -2.76118904e-01
-1.53255272e+00 -8.33966076e-01 -1.31859422e+00 -1.11338131e-01
-7.51971304e-02 -1.49983394e+00 1.22811580e+00 -2.99764037e-01
-1.74981105e+00 9.99777198e-01 6.34120107e-02 -3.94781053e-01
3.79608631e-01 -1.50448188e-01 1.05103940e-01 3.49883214e-02
3.25023383e-01 1.39178443e+00 1.04836786e+00 -1.76446021e+00
-2.68581539e-01 -1.49608934e-02 9.60590839e-02 5.76901197e-01
1.99417040e-01 -3.35239530e-01 -4.73116457e-01 -8.02124560e-01
2.11162224e-01 -6.94324732e-01 -6.73151970e-01 1.76777437e-01
-5.80661476e-01 2.21430063e-01 1.00962198e+00 -4.70273107e-01
4.32891160e-01 -1.90642846e+00 7.89304733e-01 3.38625386e-02
2.84099996e-01 -1.96612880e-01 -2.05654800e-01 3.94834578e-01
-7.73309963e-03 -1.54574767e-01 -4.38985348e-01 -9.63414967e-01
1.13563329e-01 3.55174184e-01 -5.08157074e-01 9.06391963e-02
3.90792698e-01 1.07656944e+00 -1.01485002e+00 -2.58450240e-01
8.61016214e-01 8.53589475e-01 -1.07509863e+00 5.61016023e-01
-8.88811588e-01 7.87795722e-01 -2.25479931e-01 1.90713897e-01
6.19727314e-01 -2.49005824e-01 -2.25676522e-01 -1.11434050e-01
-6.44442588e-02 4.03817058e-01 -9.56266284e-01 2.59631753e+00
-7.26112306e-01 5.18753946e-01 -9.51705948e-02 -5.35432518e-01
9.94819999e-01 6.42046239e-03 4.84920144e-01 -4.95514393e-01
1.52496025e-01 -1.36092052e-01 -7.15937078e-01 -2.56386667e-01
8.77893567e-01 -3.52552950e-01 -1.58626854e-01 5.99376976e-01
3.81186128e-01 -1.27005661e+00 -7.63807371e-02 5.53392708e-01
8.64419997e-01 7.06094623e-01 -1.10911854e-01 1.03527121e-01
-9.88827180e-03 -2.69059334e-02 -3.16919722e-02 5.41226685e-01
5.68522274e-01 9.98008609e-01 3.24937016e-01 -3.45128208e-01
-1.51104999e+00 -1.53495717e+00 8.75813961e-02 4.53171045e-01
3.33666578e-02 -4.07337010e-01 -6.60592735e-01 -2.86315382e-01
-3.97409918e-03 1.16616225e+00 -6.15624011e-01 -2.27698714e-01
-3.16540450e-01 -2.92841583e-01 6.34022877e-02 4.12743211e-01
2.56624073e-01 -1.28177226e+00 -8.12373936e-01 3.85607481e-01
3.37959155e-02 -1.16737258e+00 -5.89437038e-02 1.95350394e-01
-8.28812957e-01 -3.97746652e-01 -7.82349169e-01 -7.75314033e-01
7.80774117e-01 2.56854594e-01 1.47997630e+00 -3.83903205e-01
-5.56091428e-01 5.73915601e-01 -1.22642234e-01 -3.94436002e-01
-9.17096436e-01 -1.25297979e-01 3.58017646e-02 -8.87024179e-02
-2.96585530e-01 -1.11012208e+00 -5.50226450e-01 -2.86482513e-01
-1.26828349e+00 8.36060822e-01 3.36378306e-01 6.64247811e-01
9.25437272e-01 -9.91529748e-02 9.71276760e-02 -9.31632876e-01
5.60554624e-01 -4.54090834e-01 -7.64470935e-01 -2.64288098e-01
1.13732368e-01 2.71776825e-01 4.03181553e-01 -3.07060301e-01
-1.35745990e+00 3.35220158e-01 -4.86077577e-01 -8.32429588e-01
-4.76270288e-01 2.84917474e-01 -9.24233124e-02 1.83760270e-01
7.45681763e-01 2.96340495e-01 -1.46771297e-01 -4.41516876e-01
1.13232172e+00 2.14163423e-01 6.26661003e-01 -5.06273627e-01
9.61079359e-01 6.03149831e-01 9.75693688e-02 -7.51819313e-01
-7.03314722e-01 -5.78333177e-02 -8.93144906e-01 -1.83000833e-01
1.07891631e+00 -9.50052142e-01 -1.83954149e-01 4.87317830e-01
-1.46050215e+00 -8.26601982e-01 -8.68864059e-01 1.56969950e-01
-1.18379700e+00 3.36101055e-02 -5.45976043e-01 -5.60935676e-01
-1.76353753e-02 -1.12813044e+00 1.74232066e+00 -1.17664197e-02
-3.05852920e-01 -1.06095672e+00 3.11405867e-01 2.65271459e-02
2.21922308e-01 7.22034574e-01 9.37202454e-01 4.10825282e-01
-1.07755888e+00 1.06202804e-01 -6.17481060e-02 7.61338770e-02
1.38784468e-01 -1.70291901e-01 -1.03227806e+00 -1.15023572e-02
-8.97402465e-02 -4.67384040e-01 7.33040392e-01 6.36806726e-01
1.19781148e+00 -1.08612895e-01 -4.27388281e-01 1.08076429e+00
1.38996804e+00 4.03373353e-02 8.17382872e-01 -1.35008609e-02
9.36111152e-01 5.01757205e-01 1.03819817e-01 6.41015887e-01
4.17111874e-01 7.07378983e-01 5.09133577e-01 -3.13419610e-01
-5.64973950e-01 -9.24636066e-01 7.54322633e-02 8.32495987e-01
3.19210887e-01 -2.06890181e-01 -7.13683784e-01 6.14906311e-01
-1.49390292e+00 -9.39679921e-01 1.50995180e-01 1.77999699e+00
6.46126747e-01 1.49520606e-01 -1.80663735e-01 -1.30675927e-01
1.59694597e-01 3.08857828e-01 -7.12444365e-01 -5.03012538e-01
-9.48305577e-02 3.86920303e-01 5.44285625e-02 7.14223325e-01
-7.72005558e-01 1.35974908e+00 6.83601284e+00 5.78781903e-01
-1.02968121e+00 4.54555973e-02 6.67364299e-01 -9.37259421e-02
-8.93553793e-01 -1.59410108e-02 -4.58598554e-01 6.45406358e-03
5.95576286e-01 -2.25624144e-02 5.49790442e-01 1.01820040e+00
1.46346271e-01 -1.21562704e-01 -1.08659875e+00 1.14394021e+00
2.87384510e-01 -1.88352692e+00 5.43125629e-01 1.91598535e-01
1.20283043e+00 -4.82880287e-02 1.92814857e-01 1.18950874e-01
8.99990976e-01 -1.06038511e+00 9.97628331e-01 6.73709035e-01
1.14488602e+00 -6.94862664e-01 4.81207818e-02 4.44889337e-01
-8.48608136e-01 3.38777989e-01 -4.67872113e-01 -8.38639587e-03
8.14462066e-01 6.74006462e-01 -9.73929286e-01 3.43030035e-01
5.08220315e-01 8.20982933e-01 -2.94935703e-01 8.45615327e-01
-3.04864913e-01 3.72380242e-02 -4.93630439e-01 2.99824536e-01
4.05871160e-02 -1.71042398e-01 5.84349036e-01 9.10069644e-01
4.91871417e-01 -1.23261828e-02 1.28047327e-02 1.64256215e+00
-9.96877253e-02 -2.60228872e-01 -1.29341447e+00 9.86019969e-02
2.41138786e-01 1.14074147e+00 -8.05635154e-01 -5.45307875e-01
-8.58406276e-02 1.63751888e+00 4.21960264e-01 3.46994996e-01
-8.09530973e-01 -5.26437834e-02 4.60337609e-01 1.49758115e-01
3.24981749e-01 -6.82903588e-01 -3.38872135e-01 -1.32773674e+00
-4.14479196e-01 -5.06193638e-01 -3.56671482e-01 -1.50589263e+00
-1.12266994e+00 7.10687399e-01 3.04881275e-01 -1.08028138e+00
-6.02516592e-01 -3.38590056e-01 -3.46871138e-01 8.86118591e-01
-1.19175613e+00 -1.43053281e+00 -5.38441598e-01 5.35229862e-01
8.25705409e-01 -5.14094532e-02 1.07928729e+00 -6.76515996e-02
-1.84328966e-02 -8.33751857e-02 -2.16743454e-01 -3.04614753e-01
2.50341028e-01 -1.29325593e+00 1.25745702e+00 7.85599947e-01
4.59155500e-01 4.33233738e-01 6.99524224e-01 -8.74630749e-01
-1.28704309e+00 -1.39773118e+00 4.22579497e-01 -6.79740548e-01
9.05091390e-02 -7.84244061e-01 -5.05984187e-01 7.88596809e-01
2.10336387e-01 -1.61902875e-01 3.76452684e-01 -2.76552945e-01
-9.26010981e-02 4.70468223e-01 -1.03652692e+00 8.46944690e-01
1.38953483e+00 -6.17443562e-01 -3.81020039e-01 3.78403902e-01
1.01910985e+00 -8.63211930e-01 -8.03461194e-01 1.47195905e-01
3.20471197e-01 -1.05038929e+00 1.39454997e+00 -2.93592662e-01
9.02245939e-01 -2.60234654e-01 -1.79077700e-01 -1.63845909e+00
-5.31774342e-01 -5.69960654e-01 -4.18852925e-01 8.44410062e-01
3.72500606e-02 -1.06513485e-01 8.96349072e-01 5.10642886e-01
-4.22600240e-01 -5.38243055e-01 -5.88047922e-01 -2.69325703e-01
2.90901568e-02 -5.79592288e-01 8.20847452e-01 7.61752248e-01
-5.89576960e-01 5.23315132e-01 -3.98521364e-01 -5.48793785e-02
7.53184974e-01 1.69897392e-01 1.23181272e+00 -9.38142896e-01
-5.45878172e-01 -3.34027976e-01 -3.83326054e-01 -1.66152179e+00
4.69253808e-02 -8.87159169e-01 2.50171304e-01 -2.11134863e+00
7.28960661e-03 -5.07331908e-01 3.57691437e-01 1.70262575e-01
5.99284098e-02 3.41483355e-01 2.82510400e-01 3.34662385e-02
-2.69346982e-01 1.05291629e+00 1.65933609e+00 -4.86106426e-02
-4.44832295e-01 -4.21922415e-01 -4.64635104e-01 6.73341572e-01
4.37064618e-01 -3.22592854e-01 -6.87633455e-01 -7.22025394e-01
2.46683106e-01 2.90174246e-01 5.00665843e-01 -1.16177583e+00
1.48235653e-02 -3.07649881e-01 7.31727362e-01 -8.99384797e-01
9.11919355e-01 -5.64107358e-01 4.55772191e-01 6.09360561e-02
-2.86724299e-01 -1.03754260e-01 2.64103085e-01 4.69519228e-01
1.75063685e-02 1.56260610e-01 5.33806324e-01 -4.59633201e-01
-6.03405058e-01 5.86004734e-01 -4.39668059e-01 -3.69964093e-01
9.64983523e-01 -2.77400553e-01 1.16696194e-01 -7.18118846e-01
-7.48854578e-01 -1.80081636e-01 8.11963737e-01 5.61290562e-01
1.11225891e+00 -1.66812313e+00 -6.65920317e-01 6.87321544e-01
4.88874950e-02 8.26027453e-01 4.78373945e-01 7.83489197e-02
-1.00188422e+00 2.13288546e-01 -4.01871949e-01 -8.20322394e-01
-7.10800529e-01 5.74975073e-01 1.96011141e-01 -1.62673607e-01
-9.45377231e-01 1.03013396e+00 6.42501175e-01 -6.39404297e-01
-1.06835514e-01 -4.77169156e-01 2.20293358e-01 -4.39569920e-01
2.73985386e-01 -1.35558963e-01 -3.74192804e-01 -7.94187665e-01
1.45504504e-01 5.50065339e-01 2.27935970e-01 -6.31956697e-01
1.63560534e+00 -4.90631275e-02 9.65783447e-02 4.26434457e-01
1.26822186e+00 -1.52499333e-01 -1.76080179e+00 3.02760359e-02
-6.95943356e-01 -6.02857471e-01 2.15210468e-01 -4.48659480e-01
-7.33439147e-01 1.06671846e+00 3.15601677e-01 -8.75345469e-02
9.14257824e-01 1.22583874e-01 8.79140377e-01 -3.62293273e-02
6.65850341e-01 -6.96105361e-01 4.81602699e-01 5.35762846e-01
1.25041974e+00 -7.56094277e-01 4.43995930e-02 -4.27380800e-01
-5.78095138e-01 7.89048135e-01 4.45510924e-01 -4.83574688e-01
6.02040052e-01 4.57199246e-01 -1.71010807e-01 -4.78737772e-01
-7.99647152e-01 -4.49577384e-02 3.35861355e-01 1.00553429e+00
1.63947627e-01 4.00501341e-02 5.27921140e-01 -5.05463108e-02
-6.37507558e-01 2.61745434e-02 5.80692828e-01 8.60598147e-01
-1.23455562e-01 -1.05470955e+00 -2.31069878e-01 3.53390574e-01
1.91767633e-01 -2.30100448e-03 -3.04153711e-01 2.79172510e-01
2.04194844e-01 3.87284487e-01 3.28392416e-01 -3.31027597e-01
2.87104100e-01 -1.21511333e-01 8.60095084e-01 -1.10598147e+00
3.44055891e-02 3.33942354e-01 -1.51978120e-01 -7.41841793e-01
-3.25887620e-01 -5.54901719e-01 -1.07326913e+00 -1.78141370e-01
5.89735620e-02 -3.76347274e-01 8.36759806e-01 5.81226170e-01
5.42589843e-01 8.82964969e-01 3.86142075e-01 -1.64046395e+00
2.06507221e-01 -7.20864892e-01 -4.08618152e-01 6.44306302e-01
2.14713141e-01 -7.76162863e-01 7.67581686e-02 5.56527197e-01] | [9.153534889221191, -3.157683849334717] |
d46c71a7-3c2d-45ed-a6d7-c7d3b415d059 | focus-manipulation-detection-via-photometric | null | null | http://openaccess.thecvf.com/content_cvpr_2018/html/Chen_Focus_Manipulation_Detection_CVPR_2018_paper.html | http://openaccess.thecvf.com/content_cvpr_2018/papers/Chen_Focus_Manipulation_Detection_CVPR_2018_paper.pdf | Focus Manipulation Detection via Photometric Histogram Analysis | With the rise of misinformation spread via social media channels, enabled by the increasing automation and realism of image manipulation tools, image forensics is an increasingly relevant problem. Classic image forensic methods leverage low-level cues such as metadata, sensor noise fingerprints, and others that are easily fooled when the image is re-encoded upon upload to facebook, etc. This necessitates the use of higher-level physical and semantic cues that, once hard to estimate reliably in the wild, have become more effective due to the increasing power of computer vision. In particular, we detect manipulations introduced by artificial blurring of the image, which creates inconsistent photometric relationships between image intensity and various cues. We achieve 98% accuracy on the most challenging cases in a new dataset of blur manipulations, where the blur is geometrically correct and consistent with the scene's physical arrangement. Such manipulations are now easily generated, for instance, by smartphone cameras having hardware to measure depth, e.g. `Portrait Mode' of the iPhone7Plus. We also demonstrate good performance on a challenge dataset evaluating a wider range of manipulations in imagery representing `in the wild' conditions. | ['Can Chen', 'Scott McCloskey', 'Jingyi Yu'] | 2018-06-01 | null | null | null | cvpr-2018-6 | ['image-forensics'] | ['computer-vision'] | [ 5.91689229e-01 -2.13115156e-01 4.07932162e-01 3.57692018e-02
-7.68645823e-01 -8.92162800e-01 6.36252701e-01 3.91314505e-03
-5.33319354e-01 5.89930236e-01 -9.09066014e-03 6.48749387e-03
-2.17847452e-02 -4.13000643e-01 -8.52047265e-01 -5.19423306e-01
-3.70226800e-02 -7.70992190e-02 2.50074923e-01 -8.24793875e-02
5.59861183e-01 3.89115989e-01 -1.71945620e+00 1.16480030e-01
4.79917884e-01 1.22816575e+00 2.66246319e-01 8.14624131e-01
1.87251478e-01 6.37799859e-01 -9.01945770e-01 -8.30498755e-01
5.13260186e-01 8.00715089e-02 -4.02452052e-01 3.66984308e-01
9.36189294e-01 -7.94977248e-01 -5.21710634e-01 1.45633805e+00
2.01990455e-01 -2.75785476e-01 -4.46886197e-02 -1.37348950e+00
-5.66036820e-01 -1.52267605e-01 -9.76704895e-01 3.37634593e-01
7.66385317e-01 5.57527959e-01 2.86336303e-01 -5.58287919e-01
9.49621081e-01 1.21982038e+00 8.85678709e-01 -1.39994428e-01
-1.39167893e+00 -7.07568884e-01 -5.52338839e-01 3.82346392e-01
-1.29517329e+00 -7.82389164e-01 7.10780501e-01 -4.17795897e-01
3.30204278e-01 3.29113394e-01 1.92368418e-01 1.33442855e+00
1.60442695e-01 2.12884262e-01 1.40339196e+00 -2.91780680e-01
8.66395533e-02 3.49321157e-01 -2.95937538e-01 4.01355714e-01
6.68838978e-01 1.63407579e-01 -7.44788349e-01 -2.80188590e-01
8.13694656e-01 1.57305285e-01 -6.81679249e-01 -2.62103558e-01
-1.16338110e+00 1.51177555e-01 1.07203804e-01 6.78949174e-04
-3.59450459e-01 2.28441074e-01 9.06288400e-02 2.00917765e-01
2.90868372e-01 6.45474553e-01 -5.77133782e-02 -5.53895712e-01
-1.03615880e+00 8.11637789e-02 4.56500173e-01 7.55501866e-01
7.74779260e-01 -3.04669648e-01 3.35503250e-01 4.66456532e-01
-9.80315059e-02 7.30578065e-01 9.05690938e-02 -1.40711725e+00
6.55032694e-01 1.68415397e-01 5.66912115e-01 -1.54501247e+00
1.74531475e-01 -6.08608201e-02 -7.91113257e-01 4.41256523e-01
9.07739639e-01 6.53472990e-02 -5.29117465e-01 1.33818483e+00
3.93033147e-01 3.94106269e-01 -3.45717460e-01 9.37662542e-01
2.02881336e-01 9.68086049e-02 -3.58002484e-01 -2.38176003e-01
1.47458446e+00 -9.20426548e-02 -7.50606418e-01 -1.99306890e-01
1.21702865e-01 -1.09074235e+00 9.70774412e-01 9.63898659e-01
-7.82775640e-01 -3.66872966e-01 -8.92760754e-01 -7.35178366e-02
-4.34701592e-01 -3.51088434e-01 4.70991135e-01 7.82953799e-01
-9.12437558e-01 6.89640284e-01 -4.95380223e-01 -1.60634279e-01
7.44677782e-01 -4.05478803e-03 -8.12381744e-01 -5.27751923e-01
-8.69356215e-01 7.04844117e-01 -1.26309320e-01 1.45922184e-01
-5.55512667e-01 -8.92331362e-01 -7.04229712e-01 -1.17486738e-01
6.18703008e-01 -7.73500949e-02 8.51430118e-01 -7.68516481e-01
-1.11036551e+00 1.09301531e+00 7.32193962e-02 -3.52420509e-01
8.66286516e-01 -2.99331397e-01 -4.58057940e-01 5.62381744e-01
2.75101334e-01 4.51766878e-01 1.48851335e+00 -1.41759479e+00
-1.89535677e-01 -5.34811318e-01 5.76531105e-02 -3.12438577e-01
-3.20191860e-01 1.37254298e-01 -2.87961602e-01 -3.73731136e-01
-1.15419529e-01 -9.11062956e-01 2.56805032e-01 4.07265007e-01
-5.70505083e-01 6.07241631e-01 1.25161183e+00 -9.47089493e-01
9.44669306e-01 -2.53198910e+00 -3.45135421e-01 6.46336228e-02
3.10062349e-01 3.92229825e-01 2.13247776e-01 3.44737560e-01
-6.70655519e-02 2.50847459e-01 -1.87299401e-01 -4.44738030e-01
-1.26115516e-01 1.85415447e-02 -4.52157408e-01 7.90797114e-01
-7.08573386e-02 6.72035873e-01 -9.99574721e-01 -4.64085281e-01
4.29570317e-01 4.61193174e-01 -2.03563094e-01 1.13789670e-01
1.69810709e-02 5.17768085e-01 3.37879136e-02 8.00957263e-01
1.12381411e+00 -2.87547857e-01 -1.93949029e-01 -3.81945550e-01
-9.96822342e-02 -1.13458328e-01 -1.15998662e+00 1.78618622e+00
-3.70401442e-01 1.05621731e+00 4.01722223e-01 -3.61811161e-01
4.63954508e-01 -2.21142713e-02 3.04148972e-01 -5.98471463e-01
1.38050959e-01 4.74513918e-02 -3.36723000e-01 -9.26984310e-01
7.30445385e-01 2.49667048e-01 -1.31025417e-02 4.32457447e-01
-3.89873803e-01 -4.05842364e-01 -8.08360502e-02 3.03417146e-01
1.53884542e+00 1.50938242e-04 1.06221154e-01 1.03924125e-01
-8.01116321e-03 -1.06074318e-01 1.44988641e-01 8.53006065e-01
-4.29700732e-01 1.05504799e+00 3.24625075e-01 -3.21816832e-01
-1.08084524e+00 -8.98102999e-01 -2.17660621e-01 3.01850021e-01
3.45377862e-01 -4.16712850e-01 -8.16296756e-01 -3.24458838e-01
1.74979359e-01 4.59047168e-01 -4.53604907e-01 3.12162545e-02
-1.92131713e-01 -5.08806050e-01 6.53159440e-01 -3.44537385e-02
9.43372309e-01 -4.18997914e-01 -9.57472265e-01 -4.28026579e-02
-2.83791274e-01 -1.74467051e+00 -3.10989976e-01 -5.32960117e-01
-4.75466460e-01 -1.46545696e+00 -4.78877246e-01 7.40399882e-02
5.43012738e-01 7.71003306e-01 7.88593829e-01 2.85270065e-01
-7.26635039e-01 6.68958724e-01 -2.76521206e-01 -1.41774088e-01
-1.62476495e-01 -5.99456370e-01 -6.45741373e-02 4.54108179e-01
-3.99430543e-02 -8.24945271e-01 -7.83121943e-01 2.22318903e-01
-1.20383012e+00 -1.13330223e-01 3.49489212e-01 4.47798103e-01
-6.12887926e-02 3.48962396e-01 -1.92703351e-01 -6.32638633e-01
4.66808707e-01 -6.30914330e-01 -4.87396538e-01 -8.55787396e-02
-2.54538447e-01 -2.40726784e-01 3.78504008e-01 -4.99579906e-01
-1.02566147e+00 -2.17412502e-01 5.08970261e-01 -8.27279150e-01
-4.53329682e-01 -1.82787031e-02 -5.48610091e-02 -2.71607161e-01
8.84425104e-01 -2.36226227e-02 1.10346928e-01 -4.46330041e-01
1.60926104e-01 9.02731776e-01 1.11833346e+00 -3.33355933e-01
1.07673335e+00 1.08428371e+00 1.82652667e-01 -1.23619616e+00
-6.33123934e-01 -3.71381342e-01 -3.86273831e-01 -4.25551593e-01
5.98058820e-01 -7.38833010e-01 -9.51403439e-01 9.19593573e-01
-1.35358536e+00 -1.25365257e-01 1.00856845e-03 2.87058741e-01
-1.94626361e-01 9.56569374e-01 -5.11311293e-01 -9.44594085e-01
2.39029631e-01 -1.15928352e+00 1.37098539e+00 9.38812271e-02
-1.48195699e-01 -5.67703009e-01 -4.61277217e-01 6.56470299e-01
5.30921221e-01 7.36598909e-01 4.23252016e-01 2.12285504e-01
-9.72622335e-01 -4.33961570e-01 -6.45383775e-01 3.57762158e-01
3.81333649e-01 -1.24100648e-01 -1.36585176e+00 -1.44424990e-01
2.98648953e-01 -1.06098227e-01 3.21517557e-01 -7.38125890e-02
1.29545176e+00 -2.91636497e-01 -1.75427794e-01 6.45260751e-01
1.44269001e+00 -1.96407095e-01 1.13413227e+00 4.65246528e-01
8.40772033e-01 6.82317495e-01 5.43353915e-01 5.81933975e-01
1.17851272e-01 9.03191566e-01 7.09456861e-01 2.33012944e-01
-2.51708418e-01 -9.10795853e-02 1.74205765e-01 7.46748969e-02
-4.60956655e-02 -1.56105995e-01 -8.20979595e-01 2.16908947e-01
-1.42935169e+00 -1.15776002e+00 -3.01985741e-01 2.57376528e+00
7.34922230e-01 2.15833753e-01 -2.80863583e-01 3.27031106e-01
9.03368235e-01 1.78727806e-01 -4.11742955e-01 8.68170261e-02
-1.95296362e-01 9.08594299e-03 8.89611304e-01 2.88235128e-01
-8.47313941e-01 4.70367163e-01 5.63054228e+00 5.81032455e-01
-1.15160096e+00 8.91171694e-02 7.33985960e-01 -1.34462461e-01
-1.31798714e-01 7.70303905e-02 -2.95385152e-01 1.19477129e+00
8.56201947e-01 3.07591379e-01 7.83393264e-01 4.47681874e-01
4.00430948e-01 -8.33339393e-01 -8.78684759e-01 1.72766435e+00
4.01082993e-01 -1.21710122e+00 -6.06090784e-01 4.38483119e-01
2.90506721e-01 -3.18184406e-01 3.31791520e-01 -4.56978261e-01
-8.32440555e-02 -7.75186241e-01 6.52525246e-01 6.14657164e-01
1.18556845e+00 -3.45697194e-01 4.44582611e-01 2.67477393e-01
-5.44467986e-01 1.92467384e-02 -1.92018911e-01 -1.57917425e-01
3.08285534e-01 9.36069846e-01 -9.26755369e-01 1.44053102e-01
9.02132213e-01 5.50804079e-01 -8.94903302e-01 1.03880548e+00
-1.36540592e-01 2.64590561e-01 -5.99203169e-01 6.24555886e-01
-2.19958663e-01 -3.16210419e-01 7.09640205e-01 8.83087814e-01
4.33109760e-01 -1.01916000e-01 -4.63479280e-01 7.26904750e-01
-1.79532439e-01 -5.95605612e-01 -9.79620695e-01 1.60030156e-01
5.21997154e-01 1.28949201e+00 -6.71447515e-01 2.33833455e-02
-1.96449324e-01 1.30981827e+00 -1.40795996e-02 2.92563885e-01
-8.69388163e-01 -1.43230557e-01 8.17170322e-01 5.64162731e-01
1.07111633e-01 -3.79207939e-01 -9.63811353e-02 -1.22304511e+00
4.18115675e-01 -9.83258784e-01 -2.43308365e-01 -1.42961454e+00
-1.22170484e+00 1.26875177e-01 8.14947262e-02 -1.17910659e+00
-1.14938520e-01 -5.30649304e-01 -2.96928436e-01 4.62241113e-01
-1.31371891e+00 -6.97412133e-01 -6.96475387e-01 5.41253269e-01
1.77165121e-01 3.36042970e-01 5.59424341e-01 5.65385580e-01
-2.21144259e-01 2.86286056e-01 1.09384477e-01 4.59371693e-02
1.07964551e+00 -9.69105721e-01 2.90205479e-01 1.04062700e+00
9.60178450e-02 4.77582514e-01 8.87563646e-01 -6.84773028e-01
-1.84319615e+00 -6.27865553e-01 3.68554622e-01 -8.00770879e-01
7.32820809e-01 -5.51596284e-01 -8.69612515e-01 3.87560606e-01
4.78512570e-02 2.11278513e-01 2.54490346e-01 -4.55193132e-01
-7.33212888e-01 -1.94484949e-01 -1.39957285e+00 4.63759452e-01
1.03885972e+00 -9.42902446e-01 -3.34195733e-01 4.56778526e-01
4.80056882e-01 -5.49139917e-01 -6.94776356e-01 1.34087682e-01
5.93596399e-01 -1.37573993e+00 1.19157088e+00 -1.21668288e-02
4.49552894e-01 -3.63493264e-01 -2.21526191e-01 -9.77915943e-01
4.10199910e-01 -1.15783536e+00 -6.20930232e-02 1.35246599e+00
-2.17550471e-01 -5.60162306e-01 6.17250085e-01 9.14424419e-01
3.40676576e-01 -4.34586704e-02 -9.54291284e-01 -6.97770119e-01
-8.30303490e-01 -6.65436327e-01 5.90385616e-01 1.13329113e+00
-1.55738369e-01 -2.53240883e-01 -5.78784645e-01 2.29865506e-01
1.01208913e+00 -1.90559745e-01 1.01817775e+00 -1.04847002e+00
-4.71384376e-01 -3.61413918e-02 -8.85016799e-01 -8.43060315e-01
-2.17588022e-01 -2.08470970e-01 -1.54825255e-01 -5.58556497e-01
9.42559913e-02 -3.51181477e-01 3.49628597e-01 -3.67860496e-02
4.34815958e-02 6.92887127e-01 3.85769397e-01 4.92176503e-01
-5.83259881e-01 -9.24209133e-02 9.10631001e-01 8.18533897e-02
2.90912002e-01 -3.82254809e-01 -5.82850933e-01 6.79973602e-01
5.94706237e-01 -5.00042617e-01 -9.77256298e-02 -5.06000042e-01
5.75608253e-01 5.77725358e-02 9.36702371e-01 -1.21528471e+00
3.12397271e-01 9.87378210e-02 4.33522075e-01 -2.44208500e-01
6.77586675e-01 -9.80055153e-01 4.42907184e-01 7.46592805e-02
9.80646312e-02 1.48063704e-01 1.38182938e-01 8.66556585e-01
-6.43930286e-02 -1.48083195e-01 6.14282608e-01 -1.83555707e-01
-5.30243158e-01 -9.06512961e-02 2.74639390e-02 9.22264457e-02
8.41788828e-01 -4.77129996e-01 -7.51695991e-01 -6.29352987e-01
-1.58653975e-01 -4.94903952e-01 9.83625889e-01 2.55832106e-01
6.20304346e-01 -9.26845312e-01 -1.98492065e-01 1.49502113e-01
-2.18938924e-02 -1.40710007e-02 2.32076183e-01 1.02883804e+00
-7.85682321e-01 -1.87883481e-01 -1.60688967e-01 -8.34186435e-01
-1.20483947e+00 6.44316435e-01 3.89782526e-02 2.66483277e-01
-5.44665813e-01 5.02781689e-01 -6.03228249e-02 4.63761985e-01
1.36124771e-02 -1.32487401e-01 5.13532698e-01 -4.30737846e-02
8.81378055e-01 6.27503872e-01 1.77460596e-01 -5.96737921e-01
-1.94956586e-01 3.90395731e-01 2.37101302e-01 -1.15928777e-01
1.20683157e+00 -5.64125478e-01 -6.16107769e-02 2.27461323e-01
1.22704518e+00 2.68124908e-01 -1.56222701e+00 -1.32224977e-01
-4.39684838e-02 -1.35577548e+00 1.78752467e-01 -4.91721869e-01
-9.91496503e-01 7.81840622e-01 7.01706111e-01 4.60127026e-01
8.97544742e-01 -2.77711183e-01 9.46227610e-01 8.42554942e-02
7.60536849e-01 -9.12874043e-01 1.69062480e-01 -1.11724347e-01
6.47505343e-01 -1.41426146e+00 2.83141255e-01 -5.02989829e-01
-2.49155864e-01 1.05477262e+00 3.29494551e-02 5.77826286e-03
5.03905594e-01 4.11036253e-01 -1.17672615e-01 -2.09143981e-01
-1.79174185e-01 7.15534166e-02 -2.92633384e-01 6.79105282e-01
-2.55769551e-01 -1.72715992e-01 5.07984579e-01 -7.60616288e-02
-3.13134909e-01 -1.07683271e-01 9.53307450e-01 8.29906225e-01
-9.39302966e-02 -7.26377845e-01 -1.05124795e+00 3.99415225e-01
-6.03448510e-01 8.53117928e-03 -2.45791242e-01 8.52177382e-01
3.26645970e-01 1.19833910e+00 1.33174419e-01 -1.45591080e-01
5.34918047e-02 -2.24451050e-01 5.81142902e-01 -3.91075343e-01
-1.57542214e-01 -3.56760174e-01 1.26982890e-02 -1.14579248e+00
-5.22286534e-01 -1.01983535e+00 -5.14411509e-01 -5.78213453e-01
-2.39177763e-01 -4.45953608e-01 1.06913209e+00 9.08881783e-01
6.53852642e-01 -1.47209629e-01 4.51866925e-01 -1.21075428e+00
-4.20088738e-01 -7.26748586e-01 -5.87743104e-01 1.11141992e+00
5.72389245e-01 -7.89099276e-01 -7.60664165e-01 3.92317116e-01] | [12.40837287902832, 1.0267765522003174] |
47fdb896-9ee5-4f17-82a7-9e081f59f14c | towards-robust-and-transferable-iiot-sensor | 2110.03440 | null | https://arxiv.org/abs/2110.03440v1 | https://arxiv.org/pdf/2110.03440v1.pdf | Towards Robust and Transferable IIoT Sensor based Anomaly Classification using Artificial Intelligence | The increasing deployment of low-cost industrial IoT (IIoT) sensor platforms on industrial assets enables great opportunities for anomaly classification in industrial plants. The performance of such a classification model depends highly on the available training data. Models perform well when the training data comes from the same machine. However, as soon as the machine is changed, repaired, or put into operation in a different environment, the prediction often fails. For this reason, we investigate whether it is feasible to have a robust and transferable method for AI based anomaly classification using different models and pre-processing steps on centrifugal pumps which are dismantled and put back into operation in the same as well as in different environments. Further, we investigate the model performance on different pumps from the same type compared to those from the training data. | ['Daniel Schall', 'Stefan von Dosky', 'Herbert Grieb', 'Thomas Bierweiler', 'Jana Kemnitz'] | 2021-10-07 | null | null | null | null | ['anomaly-classification'] | ['computer-vision'] | [ 9.64244008e-02 -3.46880406e-01 2.77565897e-01 -1.51031524e-01
4.79058921e-01 -4.97068703e-01 4.07027721e-01 6.90804601e-01
1.30146891e-01 4.13006723e-01 -8.45538735e-01 -5.27234495e-01
-4.76872712e-01 -1.04310811e+00 -4.92107779e-01 -7.28301048e-01
-1.26427501e-01 6.01786852e-01 1.73411682e-01 -1.85269609e-01
2.47734487e-01 1.03182852e+00 -1.78791678e+00 1.39084220e-01
6.27750814e-01 1.33498180e+00 1.67088151e-01 5.88545144e-01
-2.89607197e-02 5.08080065e-01 -9.40758049e-01 -8.80834684e-02
7.00139463e-01 -4.10996050e-01 -6.42060041e-01 2.47731850e-01
9.93843451e-02 -1.00096114e-01 -1.49250880e-01 1.01431072e+00
8.76025185e-02 -1.04159955e-02 4.76030976e-01 -1.44074690e+00
-2.01691717e-01 3.15726250e-01 1.58211961e-01 3.18029851e-01
2.26057991e-01 2.16804221e-01 5.14135480e-01 -4.03780997e-01
2.97544271e-01 7.11564124e-01 5.56301296e-01 8.27860385e-02
-1.17934144e+00 -3.16990554e-01 1.99810550e-01 3.33802730e-01
-8.60784948e-01 -6.78276569e-02 7.98680782e-01 -4.09156889e-01
8.25604260e-01 2.84640521e-01 9.22760010e-01 1.03237295e+00
8.57897818e-01 1.98060021e-01 7.01144934e-01 -2.13249907e-01
8.04225683e-01 5.46310805e-02 4.93897162e-02 2.06067339e-01
7.26664662e-01 6.58969209e-02 -1.90507755e-01 7.41322618e-03
4.03405190e-01 4.78304833e-01 9.72743928e-02 -4.11064148e-01
-7.52372682e-01 4.18280154e-01 4.88586500e-02 8.50973248e-01
-7.93674827e-01 -2.98185974e-01 5.55295944e-01 1.01722884e+00
1.02286913e-01 1.11502230e+00 -1.08503735e+00 -2.25307703e-01
-5.37295401e-01 -6.05192184e-02 1.11652219e+00 9.36251402e-01
5.42413950e-01 2.71358997e-01 4.11237031e-01 3.32976460e-01
-2.53880229e-02 1.70722738e-01 6.41397297e-01 -5.47378004e-01
1.59285933e-01 9.50944901e-01 1.01597279e-01 -9.72314060e-01
-5.33872366e-01 -3.17982554e-01 -7.74594605e-01 6.08675897e-01
5.16338646e-01 -1.00434646e-01 -8.28483284e-01 8.73145580e-01
1.46895394e-01 1.19260542e-01 4.08652425e-01 4.86000448e-01
1.60948902e-01 5.56323647e-01 4.09417078e-02 -2.48277873e-01
9.35849786e-01 -5.30621707e-01 -7.22741663e-01 -5.27767912e-02
7.46447206e-01 -6.05796158e-01 8.23570132e-01 1.04079878e+00
-6.62248850e-01 -7.50261486e-01 -1.24810112e+00 6.69466197e-01
-9.15067136e-01 -2.23112643e-01 5.07204533e-01 5.92230320e-01
-3.22172850e-01 1.32817614e+00 -1.00155187e+00 -6.10427022e-01
1.89065963e-01 4.42653418e-01 -4.06531215e-01 -6.50311038e-02
-8.42202246e-01 1.09981382e+00 6.63340807e-01 8.50160122e-02
-7.99970448e-01 -5.68710387e-01 -5.92735410e-01 -3.23699191e-02
1.54203489e-01 -1.39093146e-01 9.24408495e-01 -7.50185549e-01
-1.70540476e+00 1.86930254e-01 4.63035315e-01 -5.86886108e-01
4.79916126e-01 -2.15800837e-01 -1.00091815e+00 -1.41185656e-01
-3.06810558e-01 -3.81886810e-01 8.64352167e-01 -9.43818748e-01
-7.62773871e-01 -6.88278198e-01 -4.30865958e-02 -5.49360514e-01
-3.31390202e-01 -1.36703089e-01 5.37074804e-01 -2.58820295e-01
3.87879461e-01 -1.02271378e+00 2.19381601e-02 -3.75892073e-01
-2.93152809e-01 -9.73431244e-02 1.37285995e+00 -6.23694062e-01
8.43388140e-01 -2.22413015e+00 -8.14702064e-02 5.11395454e-01
-4.46905911e-01 4.11984056e-01 1.33835897e-01 6.73750997e-01
-5.35637558e-01 1.38495443e-03 1.62060757e-03 2.75992483e-01
-1.08526357e-01 5.85634232e-01 -1.02685623e-01 1.60316467e-01
3.64821017e-01 2.95992166e-01 -8.83352697e-01 2.10453719e-01
7.14401364e-01 -2.89727658e-01 -1.15014963e-01 2.98535615e-01
-1.01668537e-01 8.33718181e-01 -5.68185449e-01 9.71256971e-01
5.08696973e-01 2.46575490e-01 1.87132776e-01 -1.70558572e-01
-2.28146642e-01 -5.74786216e-02 -1.11845875e+00 1.47698057e+00
-6.90010428e-01 3.29361349e-01 -1.37250692e-01 -1.48577201e+00
1.18562341e+00 6.39835715e-01 9.02279675e-01 -5.83096743e-01
1.50998458e-01 3.44950289e-01 3.12255621e-01 -7.40066171e-01
2.37880051e-01 -1.19297005e-01 2.50083953e-02 -2.65993252e-02
9.56889391e-02 -4.69846547e-01 4.64248002e-01 -5.41140497e-01
1.57319832e+00 1.39134184e-01 1.48802251e-01 -3.86031866e-01
5.52557647e-01 6.75039366e-02 6.12364531e-01 4.67929602e-01
-1.57848760e-01 4.94499393e-02 4.12270874e-01 -9.31332886e-01
-1.01253188e+00 -9.35917020e-01 -4.14193094e-01 5.56597888e-01
1.76731378e-01 -2.77889818e-01 -1.95815399e-01 -7.91533470e-01
4.55458939e-01 9.97514069e-01 -1.28828496e-01 -5.19260466e-01
-2.32620433e-01 -8.60787690e-01 1.61922619e-01 3.80515039e-01
3.91060919e-01 -9.90064859e-01 -7.23990798e-01 5.86060643e-01
4.51863140e-01 -9.64925885e-01 7.32944369e-01 5.68639696e-01
-1.22844458e+00 -1.37697554e+00 2.04443678e-01 -2.14042708e-01
6.22513235e-01 -2.54427910e-01 1.09677148e+00 1.94856264e-02
-3.25567961e-01 2.03153253e-01 -4.38386798e-01 -1.02996147e+00
-8.79416287e-01 -3.89443561e-02 4.93580759e-01 -1.76630497e-01
3.86360824e-01 -6.94795787e-01 -1.95160598e-01 4.47511464e-01
-8.18587899e-01 -7.35879898e-01 5.08336365e-01 4.56287622e-01
2.31751621e-01 1.12797034e+00 8.01188707e-01 -5.70826948e-01
3.51839125e-01 -6.66396439e-01 -8.17228258e-01 2.47017846e-01
-1.01122570e+00 -6.75731748e-02 1.05604446e+00 -3.26527387e-01
-7.61092603e-01 1.06268421e-01 -5.93297109e-02 -4.97904360e-01
-8.10052276e-01 2.10236013e-01 -3.24369192e-01 2.15599582e-01
3.57023299e-01 -2.89140999e-01 -2.76200827e-02 -8.42355967e-01
-2.21396267e-01 7.18025923e-01 2.76492059e-01 -4.02788758e-01
1.03898370e+00 1.13981508e-01 2.90444732e-01 -1.03153026e+00
-3.84986609e-01 -1.87653795e-01 -8.87318671e-01 -2.83397764e-01
7.34739959e-01 -3.37048888e-01 -7.06017733e-01 6.77768648e-01
-8.06227803e-01 -2.16941759e-01 -8.14578474e-01 7.19684243e-01
-2.38169536e-01 2.00334013e-01 -3.98968548e-01 -8.02588463e-01
-1.61055416e-01 -9.36769664e-01 5.39675653e-01 9.71566588e-02
-2.07390815e-01 -9.31709170e-01 -2.67041206e-01 1.26842022e-01
3.38623226e-01 4.28112924e-01 1.08322978e+00 -1.14439988e+00
-2.57035881e-01 -7.62078464e-01 4.95046735e-01 9.31051850e-01
7.67573237e-01 3.47921550e-01 -6.74721897e-01 -4.26228315e-01
3.67869467e-01 1.27990469e-01 1.30161479e-01 -1.04477413e-01
1.24152422e+00 -9.17758867e-02 -2.51317352e-01 6.17258698e-02
1.18428540e+00 6.94472253e-01 6.64152443e-01 5.69680870e-01
3.26569885e-01 5.12090862e-01 9.13526297e-01 2.10335866e-01
-4.90699023e-01 3.47201973e-01 8.28266025e-01 2.45846212e-01
6.07288420e-01 1.42411038e-01 4.26723242e-01 9.71123517e-01
-1.75882950e-01 -2.43309394e-01 -6.04818106e-01 2.56166846e-01
-1.61245346e+00 -9.06013370e-01 -1.88381225e-01 2.33753300e+00
3.04268181e-01 4.04539555e-01 -9.64700058e-02 7.25545228e-01
5.26653767e-01 -4.09602046e-01 -6.82448506e-01 -7.35614002e-01
2.12437332e-01 2.69136220e-01 3.53367954e-01 -1.84983566e-01
-1.00541294e+00 1.33182853e-01 6.61051559e+00 1.66691825e-01
-1.09581614e+00 -2.17857882e-01 4.68790561e-01 1.06637403e-01
4.80972201e-01 1.14880297e-02 -2.71243483e-01 8.37334454e-01
1.35500801e+00 6.16061129e-02 2.94621795e-01 1.27171934e+00
9.70481709e-02 -1.65885493e-01 -1.53258908e+00 5.95478415e-01
-3.19905400e-01 -7.70887852e-01 -2.52747983e-01 2.38110438e-01
4.36808228e-01 -9.56631154e-02 -5.45100808e-01 2.02328533e-01
9.23815183e-03 -5.91457129e-01 5.16091943e-01 4.99122888e-01
-3.97311449e-02 -6.01189196e-01 1.26966465e+00 2.51998425e-01
-1.02687001e+00 -6.56286657e-01 -3.53402525e-01 -3.95190865e-01
-7.82015398e-02 8.56541097e-01 -7.93426991e-01 1.11120653e+00
8.72476637e-01 5.67397118e-01 -5.83045959e-01 7.59221196e-01
5.78103475e-02 4.48803693e-01 -3.08948845e-01 2.15716898e-01
-2.78154016e-01 -5.67990780e-01 5.35829782e-01 3.63144159e-01
8.85977209e-01 -4.25535321e-01 1.93535313e-01 5.96958578e-01
4.79172796e-01 -1.18017621e-01 -9.98071909e-01 -2.85745680e-01
2.09582955e-01 1.25304377e+00 -7.10790157e-01 -1.65813372e-01
-5.10810316e-01 8.07307839e-01 -2.58739740e-01 1.51630901e-02
-5.85590303e-01 -3.89665216e-01 8.28677714e-01 8.85309875e-02
2.34339401e-01 -2.72722840e-01 -1.20551601e-01 -8.12552094e-01
9.75769684e-02 -5.81885934e-01 4.15480047e-01 -7.87327170e-01
-1.77947414e+00 3.44935507e-01 1.78892724e-02 -1.46561325e+00
-2.94490993e-01 -1.20046794e+00 -7.97738135e-01 3.34243685e-01
-1.12124300e+00 -5.08365929e-01 -4.16524202e-01 3.53888661e-01
4.69192564e-01 -5.38940787e-01 1.05879903e+00 2.57740110e-01
-9.10376370e-01 -9.88267064e-02 3.18315059e-01 -1.52900308e-01
3.16071928e-01 -1.35302413e+00 3.64263430e-02 1.00456572e+00
-1.36523500e-01 2.11365506e-01 8.61422241e-01 -6.00472271e-01
-1.52975786e+00 -1.13535798e+00 2.68566817e-01 -4.92029160e-01
7.32336223e-01 8.68704170e-02 -1.06495690e+00 7.72360444e-01
1.23922206e-01 1.43630624e-01 6.38552606e-01 -3.53260641e-03
4.03036237e-01 -5.39061725e-01 -1.50043643e+00 1.99627057e-01
6.83288217e-01 -1.06205456e-01 -8.31502676e-01 4.54191893e-01
2.07093328e-01 -1.41516909e-01 -1.53967893e+00 5.46521664e-01
2.89115727e-01 -8.92797709e-01 6.44651532e-01 -8.45291793e-01
2.19588101e-01 -2.90995002e-01 -6.75546005e-02 -1.53027093e+00
-1.25282854e-01 -5.13315737e-01 -1.77723408e-01 1.23821211e+00
7.33830407e-02 -1.20824587e+00 5.54617167e-01 8.53441119e-01
-4.33432817e-01 -5.64335585e-01 -8.92307818e-01 -1.33386338e+00
-1.20788574e-01 -3.56634021e-01 1.10708821e+00 1.14067507e+00
2.02132285e-01 4.86691482e-02 2.19392657e-01 4.48550135e-01
1.39993310e-01 1.86456725e-01 8.05367172e-01 -2.12411714e+00
1.57865301e-01 4.81131487e-02 -1.03738725e+00 -8.24446529e-02
-7.42826471e-03 -6.29289508e-01 -1.52920052e-01 -1.26914167e+00
-7.10498869e-01 -7.18289912e-01 -7.15782821e-01 5.28935373e-01
4.39869165e-01 -2.26430207e-01 1.47765145e-01 1.05164707e-01
7.24049583e-02 2.60921568e-01 8.21079314e-01 -7.41026998e-02
-3.24850678e-01 4.28615749e-01 1.29529029e-01 8.27189147e-01
1.23535764e+00 -5.20416975e-01 -2.36917093e-01 1.54483557e-01
1.46794677e-01 -1.38402447e-01 3.12290043e-01 -1.54777825e+00
-2.48059496e-01 -1.98240027e-01 6.34824097e-01 -3.92569125e-01
4.70625162e-02 -1.73230505e+00 6.16510928e-01 9.11058903e-01
2.46760100e-01 4.56176996e-01 2.07798168e-01 5.89001000e-01
-1.33178085e-02 -7.38103986e-01 4.91400272e-01 -1.81070462e-01
-6.32585227e-01 6.89438358e-02 -6.05932474e-01 -6.43443823e-01
1.45382917e+00 -3.63065183e-01 -2.16769040e-01 2.66813219e-01
-9.57321942e-01 2.34206244e-02 7.00013876e-01 7.23060250e-01
9.54971388e-02 -1.18783796e+00 -2.78485745e-01 6.95235729e-01
1.87975585e-01 3.11493110e-02 -2.10035607e-01 6.82477057e-01
-5.08976698e-01 2.24336505e-01 -6.48735166e-01 -6.37843370e-01
-9.92336333e-01 8.11395407e-01 3.82891893e-01 -1.31233171e-01
-7.81610489e-01 -7.68431127e-02 -5.04422605e-01 -3.88720959e-01
-2.96696365e-01 -8.71125698e-01 5.29990830e-02 -3.89390215e-02
5.86031228e-02 6.57002866e-01 8.04217517e-01 -1.19807795e-01
-2.91359365e-01 7.10122287e-02 2.23079145e-01 4.55962867e-01
1.42381585e+00 2.55303562e-01 -2.16956049e-01 8.57753217e-01
6.39961362e-01 -1.77156702e-01 -8.63188684e-01 2.86321253e-01
2.61856854e-01 -6.66476846e-01 -9.51149389e-02 -5.27879715e-01
-1.23790562e+00 4.52599615e-01 9.39218402e-01 1.16377008e+00
1.03318405e+00 -3.41491222e-01 5.22092879e-01 8.43519211e-01
5.54623485e-01 -1.49415648e+00 1.51056647e-01 3.63930404e-01
7.09655225e-01 -9.57458794e-01 2.86370754e-01 -3.09126437e-01
-2.44858518e-01 1.39210737e+00 4.10721809e-01 -7.59510696e-02
9.90300655e-01 2.33953029e-01 -1.01976760e-01 -4.09940660e-01
-4.09240901e-01 1.68813109e-01 -2.05101013e-01 7.93523312e-01
1.53526012e-02 6.29824772e-02 1.73582554e-01 3.76950473e-01
-1.79199189e-01 7.29228705e-02 5.49961507e-01 1.37157774e+00
-2.55011767e-01 -1.19585121e+00 -3.27430964e-01 8.04988921e-01
-4.50385720e-01 5.71214378e-01 -1.69697985e-01 9.53515351e-01
3.79505396e-01 1.24471700e+00 3.55091363e-01 -4.49458897e-01
8.85414779e-01 6.24114871e-01 2.86862642e-01 -5.47047973e-01
-5.07108510e-01 -6.55077994e-01 1.26435310e-01 -6.39800370e-01
-3.03103507e-01 -8.44535530e-01 -1.23264110e+00 -1.54156059e-01
-5.53751886e-01 1.72691554e-01 9.68194842e-01 1.11191225e+00
4.63250190e-01 1.01600957e+00 9.63531494e-01 -8.24187279e-01
-6.89292729e-01 -1.06977761e+00 -1.04074574e+00 6.36501849e-01
-1.42345224e-02 -8.85656178e-01 -6.81880057e-01 3.16385813e-02] | [6.8937788009643555, 2.3538951873779297] |
51b9905f-e6aa-4783-b6c8-6ad27bbdab88 | a-vision-transformer-approach-for-efficient | 2305.02074 | null | https://arxiv.org/abs/2305.02074v2 | https://arxiv.org/pdf/2305.02074v2.pdf | A Vision Transformer Approach for Efficient Near-Field Irregular SAR Super-Resolution | In this paper, we develop a novel super-resolution algorithm for near-field synthetic-aperture radar (SAR) under irregular scanning geometries. As fifth-generation (5G) millimeter-wave (mmWave) devices are becoming increasingly affordable and available, high-resolution SAR imaging is feasible for end-user applications and non-laboratory environments. Emerging applications such freehand imaging, wherein a handheld radar is scanned throughout space by a user, unmanned aerial vehicle (UAV) imaging, and automotive SAR face several unique challenges for high-resolution imaging. First, recovering a SAR image requires knowledge of the array positions throughout the scan. While recent work has introduced camera-based positioning systems capable of adequately estimating the position, recovering the algorithm efficiently is a requirement to enable edge and Internet of Things (IoT) technologies. Efficient algorithms for non-cooperative near-field SAR sampling have been explored in recent work, but suffer image defocusing under position estimation error and can only produce medium-fidelity images. In this paper, we introduce a mobile-friend vision transformer (ViT) architecture to address position estimation error and perform SAR image super-resolution (SR) under irregular sampling geometries. The proposed algorithm, Mobile-SRViT, is the first to employ a ViT approach for SAR image enhancement and is validated in simulation and via empirical studies. | ['Murat Torlak', 'Geetika Vedula', 'Yusef Alimam', 'Josiah Smith'] | 2023-05-03 | null | null | null | null | ['image-super-resolution', 'image-enhancement'] | ['computer-vision', 'computer-vision'] | [ 9.19214249e-01 -2.48204827e-01 1.85391322e-01 -3.70264977e-01
-1.03894448e+00 -5.34039140e-01 3.68584305e-01 -8.47230494e-01
-5.88130802e-02 9.77218270e-01 3.23806480e-02 -1.89056873e-01
-6.94194436e-01 -7.76782811e-01 -2.30675116e-01 -8.00776482e-01
-1.57083362e-01 4.61353660e-01 -4.03859876e-02 -2.26987854e-01
1.00629486e-01 7.91965067e-01 -1.22384918e+00 -2.01585621e-01
1.15070403e+00 1.08594263e+00 8.29426706e-01 6.67751551e-01
8.16223919e-01 6.93471611e-01 -3.41130227e-01 1.81854710e-01
6.03714526e-01 -3.76695812e-01 -1.24277443e-01 2.00026512e-01
6.46686614e-01 -6.76766694e-01 -2.96943158e-01 1.24789929e+00
5.28652072e-01 -2.07759380e-01 3.33921969e-01 -4.05980170e-01
-6.54890537e-01 2.14870289e-01 -9.13706899e-01 3.71069103e-01
3.04120004e-01 9.00844708e-02 2.89088517e-01 -1.09471250e+00
4.54427183e-01 5.95885873e-01 6.62287652e-01 6.83037937e-02
-6.04908228e-01 -5.62239945e-01 -6.37967944e-01 1.64504722e-01
-1.53441989e+00 -6.38112724e-01 7.06034362e-01 -1.84364498e-01
3.26305777e-01 3.07928771e-01 5.63453317e-01 4.83927518e-01
4.41397250e-01 -3.01595405e-02 1.25955760e+00 -4.91508484e-01
1.68071330e-01 -2.23057672e-01 -6.43509924e-02 4.75361407e-01
9.99758005e-01 4.14104670e-01 -4.31574881e-01 -8.09162855e-02
1.12312222e+00 1.83729321e-01 -9.54559684e-01 -3.04433733e-01
-1.47671103e+00 6.50710881e-01 5.13777792e-01 6.20143354e-01
-9.01371717e-01 -5.55471070e-02 -6.83663487e-01 2.63069957e-01
4.04549450e-01 7.61026382e-01 -3.47272530e-02 2.41461635e-01
-1.41687262e+00 -1.15149997e-01 2.44549394e-01 8.60883772e-01
6.60196960e-01 6.54055417e-01 1.78080034e-02 4.04462695e-01
3.34978461e-01 1.27904832e+00 2.35046029e-01 -7.79926836e-01
3.11383218e-01 -3.29814628e-02 5.09125769e-01 -7.75454223e-01
-2.97389895e-01 -1.37965226e+00 -1.27510595e+00 1.78098857e-01
-1.04308352e-02 -3.46608758e-01 -9.63274240e-01 1.15890622e+00
2.17469454e-01 3.81781131e-01 6.77383482e-01 1.38267660e+00
6.43477976e-01 5.54467618e-01 -6.12144411e-01 -8.65261376e-01
1.40931964e+00 -4.59246218e-01 -8.26436460e-01 -9.08888876e-01
1.48506658e-02 -8.54061544e-01 2.22324908e-01 5.07708192e-01
-1.07976711e+00 -4.51411843e-01 -1.48695230e+00 4.33676988e-01
5.29800951e-01 3.42999071e-01 4.30392861e-01 5.88448822e-01
-1.01860392e+00 2.20103294e-01 -5.99765480e-01 -1.97868496e-01
5.26385963e-01 -6.03052080e-02 5.39393798e-02 -7.71881938e-01
-1.00185156e+00 8.69281888e-01 -2.98750669e-01 1.15715273e-01
-5.83861530e-01 -1.01702201e+00 -6.17009342e-01 -2.04717770e-01
1.04334876e-01 -8.31629336e-01 7.74851501e-01 -6.31788790e-01
-1.24444628e+00 4.29150701e-01 -2.63656490e-02 -7.63037741e-01
-1.22661358e-02 -2.04508513e-01 -7.96665549e-01 6.62146807e-01
9.52088386e-02 2.16858797e-02 1.26954293e+00 -1.19581926e+00
-3.34012121e-01 -7.95793653e-01 -3.26307476e-01 3.19355488e-01
5.23881137e-01 -2.29527280e-01 5.28161526e-01 -4.65244919e-01
7.99348652e-01 -6.58495665e-01 -4.34944421e-01 -2.52436936e-01
1.73277870e-01 8.12730134e-01 1.23727095e+00 -6.58918798e-01
8.24711621e-01 -2.02221465e+00 -2.71274298e-01 -1.24947779e-01
1.08082160e-01 4.35158759e-01 5.50738238e-02 2.08162069e-01
1.30279005e-01 -6.50824785e-01 -5.63345253e-01 2.30286822e-01
-7.75294900e-01 -2.39338621e-01 -5.23293734e-01 8.45921397e-01
-2.96889693e-01 9.35286403e-01 -5.34350038e-01 -4.66110557e-02
3.13414127e-01 7.20609069e-01 -9.90633965e-02 -9.30406675e-02
2.18839258e-01 9.55264390e-01 -7.74452806e-01 1.07691312e+00
1.38728547e+00 -4.82847422e-01 -4.03187796e-02 -7.70158470e-01
-5.39811730e-01 -6.80139482e-01 -1.02318776e+00 1.41140413e+00
-6.79303825e-01 5.80383837e-01 5.09740651e-01 -7.76072562e-01
1.13567257e+00 1.34127378e-01 2.66333520e-01 -9.74946439e-01
-5.43175712e-02 3.09776425e-01 -3.06570411e-01 -3.61332297e-01
5.97095132e-01 -3.73054177e-01 8.69419202e-02 3.71650368e-01
-4.62619543e-01 -5.33208370e-01 -4.39012527e-01 -3.78382578e-02
1.16703701e+00 -2.64170378e-01 7.28281796e-01 -1.69692844e-01
5.78162611e-01 3.60173494e-01 2.92632669e-01 6.70451045e-01
1.13767982e-01 6.16183162e-01 -9.88095880e-01 -3.78851473e-01
-8.05315137e-01 -1.04411077e+00 -2.55396038e-01 -1.67794768e-02
4.41867203e-01 4.22137529e-01 -2.70748198e-01 6.79449439e-02
-5.32132983e-01 8.05661917e-01 2.34540224e-01 3.15571696e-01
-6.33257270e-01 -8.03702354e-01 -9.30748954e-02 -4.81054783e-02
1.19287872e+00 -3.49411070e-01 -1.40276289e+00 1.96173698e-01
-3.08835685e-01 -1.59233689e+00 -7.70438164e-02 -1.71357855e-01
-9.17303383e-01 -1.07204795e+00 -7.54290462e-01 -6.39495611e-01
5.96413434e-01 1.19266510e+00 8.39807034e-01 -3.97815049e-01
-5.08740902e-01 5.86198866e-01 -4.33050990e-01 -2.00155288e-01
1.75175875e-01 -6.70714617e-01 1.14492089e-01 2.14673951e-01
-9.42197070e-02 -8.36183429e-01 -9.82531190e-01 3.46534789e-01
-7.25360215e-01 -5.50782774e-03 1.32942915e+00 7.71354496e-01
7.40273118e-01 1.68739095e-01 5.63057661e-01 -5.77277958e-01
1.01387702e-01 -2.18430161e-01 -1.03722203e+00 1.43118724e-01
-4.78317708e-01 -2.23034158e-01 6.09866738e-01 -1.11711640e-02
-1.33781803e+00 1.55228451e-01 2.30015576e-01 -3.35171878e-01
-2.91355029e-02 6.56576633e-01 -1.29286215e-01 -8.13928843e-01
7.05882192e-01 7.26264358e-01 -4.78225015e-02 -1.41727671e-01
2.25345567e-02 8.70607913e-01 8.75954509e-01 4.13719922e-01
1.55516267e+00 8.77970099e-01 4.59225237e-01 -1.53177643e+00
-1.10806918e+00 -3.01666886e-01 -2.56692350e-01 -2.17190519e-01
8.46262634e-01 -1.43853319e+00 -2.07303137e-01 1.72800526e-01
-7.83807516e-01 -1.13183223e-01 -9.03486162e-02 1.05590189e+00
-3.71419072e-01 6.50482714e-01 1.19029604e-01 -8.70792687e-01
-6.91809058e-01 -7.52871692e-01 1.08272886e+00 3.81155998e-01
2.62277842e-01 -5.21648169e-01 -1.48015678e-01 9.32735145e-01
1.06705606e+00 2.31823653e-01 2.11943746e-01 3.19051206e-01
-1.40828681e+00 -8.82204995e-02 -3.88874620e-01 -1.24798968e-01
1.53780878e-01 -1.20439351e+00 -5.93817472e-01 -7.24164009e-01
6.87377095e-01 5.91982380e-02 6.12423599e-01 1.13446844e+00
6.17563725e-01 -2.69343048e-01 -4.65410650e-01 1.22918129e+00
1.92559540e+00 2.34606981e-01 8.33264470e-01 2.87425201e-02
3.18137318e-01 3.30784842e-02 1.42012835e+00 5.47509551e-01
-7.10713118e-03 7.29643285e-01 5.88559389e-01 2.03830265e-02
-2.05081761e-01 2.88038284e-01 9.07569155e-02 2.01402143e-01
-2.41160870e-01 -2.49347344e-01 -2.78000444e-01 3.59921038e-01
-1.26685929e+00 -1.27105069e+00 -2.83715367e-01 2.19634199e+00
1.69759676e-01 -6.36144578e-01 -1.03076100e+00 -7.37237334e-02
4.65672463e-01 4.58614498e-01 -5.66814303e-01 4.32373375e-01
-3.05551648e-01 4.74750221e-01 1.10882366e+00 7.94786334e-01
-6.88305616e-01 4.89036947e-01 4.96405363e+00 5.91258049e-01
-1.22839487e+00 3.52010816e-01 2.31839359e-01 3.36663216e-01
-4.46975470e-01 -3.60173136e-02 -7.10454941e-01 4.20896970e-02
5.36610901e-01 -1.22130245e-01 4.26249295e-01 4.37632084e-01
2.87363261e-01 -6.13255978e-01 -3.64564151e-01 1.52315652e+00
2.91082025e-01 -1.43695426e+00 -1.45759925e-01 1.53985396e-01
9.11229193e-01 -3.12083568e-02 2.27918968e-01 -2.92812198e-01
-1.59507543e-02 -9.30385590e-01 1.60788476e-01 6.66095674e-01
1.27047145e+00 -8.48863304e-01 6.47543132e-01 5.41288257e-01
-1.14920795e+00 -1.46794185e-01 -2.53186345e-01 -4.11626622e-02
5.67703247e-01 1.28196454e+00 -8.22123587e-01 1.09766805e+00
5.97948194e-01 3.94964844e-01 -1.23834737e-01 8.36663723e-01
-9.28569660e-02 3.21991205e-01 5.45456558e-02 2.89987385e-01
1.39814451e-01 -4.40155387e-01 1.10217786e+00 4.13288891e-01
1.37991858e+00 9.25257266e-01 4.01547253e-02 6.05261505e-01
5.10173678e-01 -6.39427066e-01 -8.77117574e-01 2.34877288e-01
5.20663142e-01 1.47813034e+00 -3.10191751e-01 6.79911673e-02
-2.71441758e-01 6.57876015e-01 -7.35038340e-01 5.21598637e-01
-6.83066607e-01 -3.36316049e-01 4.96026486e-01 6.10094607e-01
5.82252145e-01 -7.48109818e-01 -1.01399288e-01 -9.03382480e-01
-4.16394845e-02 -6.92600846e-01 2.75648553e-02 -1.35739088e+00
-6.81578755e-01 8.99251282e-01 1.44051146e-02 -1.75848007e+00
-3.51697236e-01 6.00101519e-03 -1.92306772e-01 8.96667123e-01
-2.10132766e+00 -1.44650662e+00 -9.22245920e-01 5.21286666e-01
6.00470066e-01 -4.51970488e-01 5.05111217e-01 6.20354749e-02
3.95692945e-01 -2.21079826e-01 2.50631571e-02 -3.98592412e-01
4.71693933e-01 -3.86397392e-01 -1.25978082e-01 1.42910278e+00
-2.26843223e-01 2.37306491e-01 1.08576834e+00 -9.36165988e-01
-2.13806081e+00 -1.48883820e+00 4.49234456e-01 1.63228586e-01
1.80002913e-01 1.20439433e-01 -2.81054795e-01 6.95844293e-01
2.03337118e-01 4.49547887e-01 2.13346645e-01 -8.55736971e-01
2.39230707e-01 -3.08716208e-01 -1.60122669e+00 1.56628758e-01
1.11512041e+00 5.17751798e-02 -4.03414905e-01 3.75799060e-01
4.27580953e-01 -4.46714789e-01 -4.52777714e-01 9.50936913e-01
1.77286446e-01 -1.03090227e+00 1.33257997e+00 6.98550582e-01
-4.36128005e-02 -8.05711448e-01 -7.02533543e-01 -1.47926986e+00
-5.15007079e-01 -7.46836722e-01 -8.08966458e-02 5.18409431e-01
-5.24740592e-02 -9.74876940e-01 8.84275854e-01 -4.79783118e-01
-2.78436333e-01 -1.68129653e-01 -9.33273673e-01 -7.90526032e-01
-8.07281852e-01 -2.47874781e-02 3.95537257e-01 8.41118276e-01
-6.62958920e-01 3.69576305e-01 -6.90774858e-01 1.21501648e+00
1.57013011e+00 7.19608366e-01 4.83283430e-01 -1.48396122e+00
-3.94049466e-01 4.90147442e-01 -3.51244897e-01 -1.33019614e+00
-3.49272579e-01 -6.05436504e-01 -7.22582936e-02 -1.77051258e+00
-1.80032905e-02 -2.74390370e-01 6.05865896e-01 -3.83888602e-01
4.74765360e-01 7.79457152e-01 -2.71940112e-01 3.39263558e-01
-2.67590165e-01 4.52748120e-01 1.45878148e+00 -7.68260285e-02
6.93391450e-03 4.52440679e-01 -6.64042115e-01 4.89385873e-01
6.25307202e-01 -1.01296075e-01 -5.46909153e-01 -6.09384596e-01
2.40247741e-01 9.03414607e-01 7.75839031e-01 -1.53361177e+00
7.99008489e-01 -5.27809635e-02 7.11443841e-01 -1.01749718e+00
7.59383082e-01 -1.11174178e+00 7.37220943e-01 6.04717016e-01
5.57093441e-01 -2.68653482e-01 -3.81132632e-01 6.82682633e-01
-2.82264411e-01 -1.23323932e-01 1.25594354e+00 -2.46449560e-01
-7.98021555e-01 4.00415212e-01 -2.75910378e-01 -1.69260889e-01
1.14666164e+00 -5.96165180e-01 -5.32158434e-01 -7.47904837e-01
-2.74487019e-01 -1.41750380e-01 3.27125728e-01 -3.66830885e-01
1.16738546e+00 -1.01041174e+00 -1.02869165e+00 4.81719911e-01
2.47282279e-03 -1.93399131e-01 7.44314134e-01 1.05814648e+00
-6.06289625e-01 6.14010513e-01 -1.51777446e-01 -5.51396430e-01
-1.42117739e+00 5.19819185e-02 3.07875216e-01 -2.21775740e-01
-5.70855141e-01 4.91141140e-01 -7.83442408e-02 -2.61093844e-02
-5.68858683e-01 6.18688501e-02 -2.04540282e-01 -3.47325355e-01
1.04482174e+00 2.02489868e-01 4.49739881e-02 -5.05230010e-01
-3.40686381e-01 1.28248703e+00 2.51545846e-01 -1.11596487e-01
1.66456366e+00 -5.31385362e-01 7.35731199e-02 -3.23795199e-01
4.53504890e-01 4.85015064e-01 -1.01497829e+00 -3.53251636e-01
-4.39826101e-01 -7.34875202e-01 7.25537777e-01 -6.89030170e-01
-9.71252322e-01 4.04694229e-01 8.02143514e-01 -5.80765419e-02
1.49115598e+00 3.63769494e-02 7.23514259e-01 5.24977744e-01
1.06283617e+00 -4.85964388e-01 -5.67088388e-02 4.57631946e-01
7.63837874e-01 -1.11388278e+00 6.41359031e-01 -5.46881199e-01
-5.23586810e-01 1.06302583e+00 7.89626129e-03 -4.18085493e-02
5.67669690e-01 5.39190888e-01 3.32583152e-02 -5.12490809e-01
-2.86507607e-01 -2.00071380e-01 -9.44694038e-03 9.79092300e-01
-2.96794828e-02 -8.08259323e-02 -2.24229217e-01 2.74735361e-01
-1.45218402e-01 9.50159431e-02 8.44335020e-01 8.81947279e-01
-1.00628281e+00 -4.92785066e-01 -8.33115935e-01 3.83834183e-01
-2.05475315e-01 -1.66011125e-01 3.60219687e-01 5.90112567e-01
-2.28739738e-01 1.15400529e+00 -8.17929879e-02 -5.23118256e-03
1.20938107e-01 -7.36973107e-01 7.88985848e-01 -3.04558367e-01
1.74027026e-01 -1.26620024e-01 7.13657215e-02 -4.49788183e-01
-6.12331212e-01 -6.15936399e-01 -7.70592093e-01 1.92392424e-01
-1.77724883e-01 3.37616116e-01 7.60925412e-01 7.52355874e-01
4.31293219e-01 2.42439926e-01 9.61208999e-01 -8.56095493e-01
-5.62768757e-01 -8.36542368e-01 -9.79257405e-01 -7.74992883e-01
5.58189392e-01 -4.21269983e-01 -6.20170712e-01 -2.09155768e-01] | [6.779852867126465, 0.932944118976593] |
0218097b-43c5-411f-8459-08771c93e73c | development-of-a-fire-detection-system-on | 2212.03709 | null | https://arxiv.org/abs/2212.03709v1 | https://arxiv.org/pdf/2212.03709v1.pdf | Development Of A Fire Detection System On Satellite Images | This paper discusses the development of a convolutional architecture of a deep neural network for the recognition of wildfires on satellite images. Based on the results of image classification, a fuzzy cognitive map of the analysis of the macroeconomic situation was built. The paper also considers the prospect of using hybrid cognitive models for forecasting macroeconomic indicators based on fuzzy cognitive maps using data on recognized wildfires on satellite images. | ['Alexey Averkin', 'Sergey Yarushev'] | 2022-12-07 | null | null | null | null | ['fire-detection'] | ['time-series'] | [-2.55648464e-01 -2.81053066e-01 6.34740472e-01 -3.50616753e-01
2.40881130e-01 -3.60727280e-01 6.49503887e-01 1.54106155e-01
-5.25077701e-01 3.03464472e-01 2.91236669e-01 -6.86297894e-01
-5.26846230e-01 -1.36512208e+00 -4.89906937e-01 -4.28312391e-01
-3.98598760e-01 2.09901974e-01 -2.27824256e-01 -9.73328769e-01
5.95725775e-01 6.23421371e-01 -1.72475052e+00 7.54451811e-01
4.18673187e-01 1.34178662e+00 6.29123092e-01 7.04322040e-01
-1.70786113e-01 1.08520329e+00 -7.47864366e-01 1.18892655e-01
2.18489885e-01 3.12084079e-01 -4.44184154e-01 -4.30784047e-01
1.36161432e-01 -2.16799691e-01 -3.55860859e-01 1.40372550e+00
1.74792215e-01 4.49342549e-01 7.68544972e-01 -5.75577617e-01
-1.17548847e+00 5.38689494e-01 2.47156501e-01 1.28379273e+00
1.03218593e-01 -1.07703827e-01 1.61969617e-01 -1.17218852e+00
1.39656827e-01 1.40155220e+00 1.01146245e+00 -5.75150788e-01
-4.58657116e-01 -6.05581462e-01 -2.60363013e-01 6.38147354e-01
-1.39987528e+00 -1.11967877e-01 3.45299870e-01 -9.23426926e-01
1.20690250e+00 -9.36332718e-02 8.38919044e-01 1.74756736e-01
9.18304324e-01 -3.46292794e-01 1.72946203e+00 -6.14935040e-01
3.77968997e-01 -3.78242165e-01 6.69253588e-01 7.72665858e-01
2.32815936e-01 9.81970906e-01 -2.13945672e-01 2.71604717e-01
6.03783429e-01 3.29599567e-02 2.40642145e-01 8.46961200e-01
-4.53424275e-01 9.62129593e-01 1.14884877e+00 8.43626022e-01
-9.05299187e-01 -2.82474458e-01 2.25784153e-01 4.19536680e-01
7.12006092e-01 2.41211727e-01 1.55500814e-01 6.09329522e-01
-1.16978991e+00 2.48435929e-01 1.56800553e-01 -2.27128193e-01
8.40098858e-01 4.44102764e-01 2.35390380e-01 1.14265189e-01
2.38941342e-01 1.07765400e+00 2.77395099e-01 -8.98584485e-01
4.87853065e-02 6.54079080e-01 1.50525182e-01 -1.72244716e+00
-7.65807033e-01 -3.33393693e-01 -9.11471665e-01 9.58268762e-01
-1.23706378e-01 -6.61185361e-04 -1.31010520e+00 7.59745479e-01
-7.43054688e-01 1.14832424e-01 3.25900018e-01 9.36850250e-01
1.04321003e+00 8.42762351e-01 1.90394387e-01 1.80966526e-01
1.21924758e+00 -1.14471361e-01 -9.55706477e-01 -4.33036909e-02
-2.91364133e-01 2.45400760e-02 4.96813387e-01 2.59512275e-01
-5.73006988e-01 -8.75721753e-01 -1.25058055e+00 3.74397278e-01
-1.41731715e+00 -1.01906493e-01 5.76740444e-01 5.22334814e-01
-1.55517578e+00 4.27692652e-01 -4.79239136e-01 -1.50997013e-01
2.93224126e-01 2.95513093e-01 2.12915130e-02 2.44693965e-01
-1.64616203e+00 1.58571839e+00 9.44990993e-01 7.40679145e-01
-5.74589133e-01 3.01457942e-04 -6.60844982e-01 5.69820285e-01
-5.02821982e-01 -2.12615639e-01 4.08485651e-01 -1.34105182e+00
-6.93230331e-01 8.86700809e-01 3.50638956e-01 -8.23754489e-01
-1.16936587e-01 2.55203635e-01 -7.14279473e-01 1.39858037e-01
2.29734331e-01 3.28145921e-01 2.41793513e-01 -8.58262718e-01
-8.77534211e-01 -5.51915944e-01 3.01903009e-01 7.91341066e-04
-5.30756488e-02 1.60169393e-01 6.11981392e-01 -7.35302448e-01
2.18233131e-02 -3.53075266e-01 -1.77362382e-01 -7.96821654e-01
7.58158982e-01 -2.13416144e-01 5.59020460e-01 -1.16221869e+00
1.02820706e+00 -2.01593280e+00 -3.92385393e-01 6.19166255e-01
-2.20413491e-01 7.76136339e-01 1.45453095e-01 -1.22175649e-01
-1.75310180e-01 3.85962009e-01 -2.76394963e-01 7.58378625e-01
-8.31119195e-02 4.80045706e-01 -5.57349086e-01 -5.09470329e-02
-1.37051567e-02 7.18014359e-01 -7.65918076e-01 -9.26348716e-02
6.65357828e-01 3.42411369e-01 -1.11684501e-02 -3.03162094e-02
1.15864672e-01 -8.16470236e-02 -6.30392432e-02 6.74890578e-01
9.98910427e-01 4.23466384e-01 1.44718379e-01 4.30562953e-03
-8.74364555e-01 -8.43411833e-02 -8.53330255e-01 9.15159404e-01
-1.77712694e-01 9.15850520e-01 -7.85247386e-02 -1.28101945e+00
1.27039659e+00 4.71133649e-01 -5.58234975e-02 -1.23849404e+00
4.21974599e-01 1.15662634e-01 -3.25257421e-01 -7.84208179e-01
2.93845475e-01 -5.68745971e-01 -3.12829055e-02 3.65808070e-01
2.00493410e-01 3.16675901e-01 2.73470104e-01 -5.90798199e-01
6.92733407e-01 -6.40679151e-02 7.83481151e-02 -9.68318760e-01
5.68544269e-01 3.94304454e-01 2.57512331e-01 6.94943428e-01
-3.87591541e-01 1.30961508e-01 -3.15671042e-02 -1.74294639e+00
-4.65478420e-01 -1.10329020e+00 -2.24232674e-01 1.27567351e+00
-6.26009628e-02 3.35367143e-01 -7.50343442e-01 -1.59379870e-01
-4.30307388e-02 7.23728180e-01 -1.08946466e+00 1.87613636e-01
-3.92841488e-01 -1.36663175e+00 7.63740361e-01 5.72955966e-01
1.08045995e+00 -1.57289410e+00 -1.00821400e+00 2.26359457e-01
-2.65025079e-01 -4.24780428e-01 6.28146827e-01 5.35516262e-01
-7.78951526e-01 -1.17795622e+00 -9.52363163e-02 -9.91855979e-01
3.92859638e-01 3.86694610e-01 1.05375898e+00 2.35781401e-01
2.84746826e-01 -3.47401090e-02 -3.02711278e-01 -7.30426013e-01
-3.89892757e-01 -3.01662594e-01 -9.24683809e-02 -2.70310760e-01
6.66400313e-01 -5.62564254e-01 -2.36923367e-01 -1.93359762e-01
-1.03336811e+00 -4.95164722e-01 2.50868350e-01 4.29792821e-01
1.47586077e-01 1.03643680e+00 5.26996791e-01 6.31701201e-02
8.95309448e-01 -6.80977821e-01 -6.94993854e-01 3.59100312e-01
-7.37244129e-01 1.35596782e-01 1.25098139e-01 3.21551770e-01
-1.38101554e+00 -2.10022777e-01 5.30410409e-02 -4.27437782e-01
-3.26421559e-01 1.39605582e+00 5.22280574e-01 -1.06868908e-01
8.70072067e-01 9.00694579e-02 -6.58384115e-02 -4.33445275e-01
-1.36029884e-01 8.57268810e-01 1.19289088e+00 2.31673308e-02
3.13547701e-01 3.46705377e-01 -3.34185570e-01 -6.85306430e-01
-5.21102548e-01 -1.50247768e-01 -8.01812530e-01 -1.11829627e+00
1.08544612e+00 -9.10371304e-01 -5.04709065e-01 7.06121981e-01
-1.00672376e+00 -9.76359285e-03 1.78947017e-01 3.22547615e-01
-5.17825596e-02 -4.03668970e-01 -4.21074688e-01 -9.19811368e-01
-4.78294939e-01 -7.76069462e-01 1.19690798e-01 3.63871276e-01
2.77229398e-01 -1.16484475e+00 5.25352120e-01 4.19795901e-01
8.71075988e-01 4.51338708e-01 9.26998913e-01 1.67825427e-02
-6.39411807e-02 -8.15346688e-02 -2.05685094e-01 3.96406651e-01
-4.76117656e-02 5.86542971e-02 -8.92482162e-01 5.57095408e-01
3.65711480e-01 5.91186099e-02 1.36806107e+00 5.30635536e-01
4.87899870e-01 -2.50091732e-01 2.05176353e-01 4.85469997e-01
1.65774095e+00 6.83616757e-01 9.93855357e-01 9.63585496e-01
-1.33596122e-01 4.70155180e-01 4.89386469e-01 9.26295370e-02
5.22167802e-01 1.16046466e-01 8.17269444e-01 -3.58412266e-02
9.07848850e-02 4.60864633e-01 4.57402885e-01 1.66094452e-01
-9.50697899e-01 1.02342024e-01 -1.54958975e+00 4.20883119e-01
-1.66281474e+00 -1.64106119e+00 -6.63196072e-02 1.35751760e+00
-1.95314854e-01 9.37972441e-02 -1.74998134e-01 3.31041068e-01
1.31450772e+00 2.02801302e-01 1.24327444e-01 -8.37195098e-01
-7.50094533e-01 8.67153823e-01 5.60138106e-01 6.28232479e-01
-1.80413795e+00 1.00964546e+00 7.19624329e+00 2.33061224e-01
-1.37588179e+00 2.49931738e-01 5.73052227e-01 2.27387473e-01
1.93578467e-01 -3.85161877e-01 3.78222615e-02 5.27486622e-01
1.16145110e+00 4.59194295e-02 6.52538240e-01 3.07294071e-01
5.77030778e-01 -4.70284760e-01 5.35618126e-01 3.80574703e-01
3.26540284e-02 -1.75344431e+00 3.41840893e-01 -1.29387811e-01
5.19679070e-01 5.38542151e-01 7.36205205e-02 3.01438570e-01
6.83350563e-01 -1.13962090e+00 1.09097195e+00 1.34192109e+00
2.46508554e-01 -1.09908247e+00 1.41674113e+00 2.34764785e-01
-1.11118579e+00 -6.73969626e-01 -5.03295302e-01 -9.70884144e-01
-2.88257509e-01 2.52797514e-01 -3.32116604e-01 3.40940326e-01
1.27854753e+00 3.98662150e-01 -8.17548394e-01 5.19334555e-01
-1.45840701e-02 5.43809295e-01 -3.34186107e-01 2.94383645e-01
5.44690311e-01 -2.86307871e-01 1.20900527e-01 1.39665091e+00
6.35322332e-01 6.08489037e-01 3.92894335e-02 6.92457616e-01
5.14334202e-01 -2.84261674e-01 -7.79191196e-01 1.93636000e-01
1.72428340e-01 7.99956918e-01 -9.67543006e-01 -4.93747324e-01
1.51209638e-01 6.49917126e-02 1.43025011e-01 1.30341962e-01
-4.17357832e-01 -1.85097471e-01 3.29248279e-01 1.29742175e-01
3.76423925e-01 -2.25632861e-01 -9.72318172e-01 -7.69404531e-01
-2.79762924e-01 -2.22152039e-01 5.76206744e-01 -1.35643554e+00
-8.19223881e-01 8.91288400e-01 6.70952350e-02 -7.54311442e-01
-7.46184215e-02 -6.84736609e-01 -1.04209960e+00 1.07562339e+00
-1.41400170e+00 -9.01548028e-01 -3.86058033e-01 6.06832683e-01
-5.23683429e-02 -5.98843336e-01 1.22438073e+00 6.28032070e-03
-1.18494645e-01 -4.66327280e-01 1.88561648e-01 3.34963799e-01
-4.56916131e-02 -9.67250824e-01 3.45247805e-01 1.00857401e+00
-2.44085714e-01 9.88395810e-02 3.22198629e-01 -8.82842243e-01
-4.03321177e-01 -1.04802084e+00 1.09128821e+00 3.83770838e-03
4.83378321e-01 3.86524379e-01 -8.24393570e-01 5.73041797e-01
5.83495140e-01 -4.12943274e-01 4.76897568e-01 -2.40797952e-01
-1.31032497e-01 -3.05597395e-01 -1.53692889e+00 -1.00653008e-01
4.60319102e-01 -4.96266544e-01 -1.26390171e+00 7.48751238e-02
-3.13861698e-01 2.10106224e-01 -6.27183795e-01 5.16454577e-01
6.01733923e-01 -1.24508262e+00 1.02723253e+00 -4.52855676e-01
2.90867001e-01 -5.31634212e-01 -4.25332725e-01 -1.31266403e+00
-1.12244761e+00 4.29551423e-01 4.46960777e-01 3.30380499e-01
3.09265368e-02 -4.57874119e-01 1.51572317e-01 9.98941883e-02
-3.04879963e-01 2.87896961e-01 -1.17892230e+00 -4.84187216e-01
2.41388157e-01 -3.50701749e-01 7.09650636e-01 8.83328795e-01
-1.05841774e-02 -2.20381424e-01 3.17009538e-01 4.87584919e-01
3.91629875e-01 -1.04069710e-01 -3.63910705e-01 -1.77263105e+00
7.88656116e-01 -7.29710460e-01 -9.45286572e-01 4.65146393e-01
2.91192591e-01 -9.63173568e-01 2.08116509e-02 -1.58286333e+00
-1.10931419e-01 -1.60552830e-01 -1.00282836e+00 7.18771338e-01
-3.27096134e-02 7.77440667e-01 4.42411721e-01 3.19019169e-01
-3.38683128e-02 -2.38354057e-02 2.75916338e-01 -3.95913363e-01
8.59560668e-02 -1.00375295e-01 -4.90661800e-01 8.47146392e-01
1.18654227e+00 -2.49010444e-01 1.97623640e-01 -6.35906994e-01
4.98105824e-01 3.59754950e-01 6.27864003e-01 -1.73034513e+00
3.88855748e-02 -1.69371501e-01 4.81243670e-01 -7.07815289e-01
-2.28727058e-01 -7.24267185e-01 6.21734262e-01 8.23824108e-01
-9.33979154e-02 4.93048370e-01 3.12014759e-01 1.77559838e-01
-5.04768193e-01 -1.03971682e-01 6.77325368e-01 -5.64073265e-01
-1.18823338e+00 -4.88252342e-01 -1.43559837e+00 -6.68331981e-01
6.68682694e-01 -2.68545389e-01 -4.09651190e-01 5.41615952e-03
-1.24930084e+00 -1.56908378e-01 1.84666961e-01 4.72098500e-01
6.03862464e-01 -1.19730914e+00 -6.82850957e-01 4.46727097e-01
-5.49013674e-01 -7.17483819e-01 3.45702112e-01 2.64063686e-01
-1.36827970e+00 7.07395554e-01 -1.29301417e+00 -8.88895094e-02
-9.06756878e-01 4.89469707e-01 8.35773647e-01 -1.72362447e-01
-2.45826796e-01 2.01472044e-01 -1.60856739e-01 -2.27878779e-01
-1.73311591e-01 -2.81336427e-01 -1.17234874e+00 5.50550938e-01
8.01493466e-01 5.95358133e-01 6.10462070e-01 -1.06536543e+00
-4.53998715e-01 2.78224856e-01 6.91131771e-01 -3.70965779e-01
1.81187320e+00 -1.08494595e-01 -4.52407748e-01 2.29386047e-01
4.28083032e-01 -9.12058175e-01 -8.08443069e-01 4.73774932e-02
3.99892807e-01 -5.67288809e-02 8.99712443e-01 -1.44396603e+00
-1.50549865e+00 9.74149525e-01 1.45896888e+00 2.78784811e-01
1.62256837e+00 -7.27066994e-01 9.31705907e-02 7.71490335e-01
2.93959677e-01 -1.35440934e+00 -5.41391313e-01 1.19481945e+00
1.11537838e+00 -1.03555417e+00 -2.43757293e-01 4.54028130e-01
-4.63909060e-01 1.79015660e+00 -1.64364517e-01 -6.73694909e-01
1.41315806e+00 4.61850941e-01 4.82437462e-01 -7.64041185e-01
-1.97660834e-01 -5.64630866e-01 3.43916081e-02 8.26814950e-01
-1.07610717e-01 9.52286363e-01 -1.32332295e-01 8.15871835e-01
-1.03626154e-01 6.33444130e-01 5.99781632e-01 1.03898644e+00
-1.32312751e+00 -2.98712164e-01 -1.24073863e+00 3.65471184e-01
-5.02179205e-01 -4.17813629e-01 -5.46216846e-01 3.42597514e-01
8.36898983e-01 1.17013133e+00 5.50064802e-01 -8.31750214e-01
7.43900463e-02 3.04354846e-01 -1.33262783e-01 -9.68800932e-02
-1.41736722e+00 -3.07972461e-01 1.54955890e-02 4.92937975e-02
-8.89182448e-01 -5.00153482e-01 -9.67323303e-01 -6.78816497e-01
1.06686130e-01 2.22475320e-01 8.74608696e-01 1.24743998e+00
1.61820218e-01 4.54110324e-01 4.28931296e-01 -1.20097661e+00
2.22052962e-01 -1.20310450e+00 -8.03602099e-01 -1.32197170e-02
1.21152058e-01 -1.02062893e+00 -3.63390952e-01 2.73892611e-01] | [9.558145523071289, -1.4554810523986816] |
784dd9cb-0c7e-4cbf-b826-a635f60e83b0 | sketch-based-video-object-localization | 2304.00450 | null | https://arxiv.org/abs/2304.00450v1 | https://arxiv.org/pdf/2304.00450v1.pdf | Sketch-based Video Object Localization | We introduce Sketch-based Video Object Localization (SVOL), a new task aimed at localizing spatio-temporal object boxes in video queried by the input sketch. We first outline the challenges in the SVOL task and build the Sketch-Video Attention Network (SVANet) with the following design principles: (i) to consider temporal information of video and bridge the domain gap between sketch and video; (ii) to accurately identify and localize multiple objects simultaneously; (iii) to handle various styles of sketches; (iv) to be classification-free. In particular, SVANet is equipped with a Cross-modal Transformer that models the interaction between learnable object tokens, query sketch, and video through attention operations, and learns upon a per-frame set macthing strategy that enables frame-wise prediction while utilizing global video context. We evaluate SVANet on a newly curated SVOL dataset. By design, SVANet successfully learns the mapping between the query sketch and video objects, achieving state-of-the-art results on the SVOL benchmark. We further confirm the effectiveness of SVANet via extensive ablation studies and visualizations. Lastly, we demonstrate its zero-shot capability on unseen datasets and novel categories, suggesting its high scalability in real-world applications. | ['Changick Kim', 'Sumin Lee', 'Minji Son', 'Jinyoung Park', 'So-Yeong Jeon', 'Sangmin Woo'] | 2023-04-02 | null | null | null | null | ['object-localization'] | ['computer-vision'] | [-8.46643746e-02 -5.62951386e-01 -3.65645677e-01 -3.45396325e-02
-6.76341712e-01 -6.76269412e-01 6.35247111e-01 -2.93618411e-01
-2.98486650e-01 2.38995224e-01 1.82287946e-01 -2.01750305e-02
-7.56143034e-02 -4.87043709e-01 -8.56362820e-01 -2.83768624e-01
-3.96067142e-01 2.94754475e-01 2.67838746e-01 1.64642021e-01
2.76591182e-01 5.49249351e-01 -1.55354369e+00 5.89786112e-01
3.36227059e-01 1.32657468e+00 4.52999175e-01 7.48357296e-01
-1.12734191e-01 1.01680064e+00 -4.28715348e-01 -2.22433031e-01
2.25019976e-01 -1.33749604e-01 -7.39978731e-01 3.89883593e-02
1.06018853e+00 -6.88990831e-01 -8.94549549e-01 6.51548266e-01
2.65204579e-01 2.66655087e-01 5.48917055e-01 -1.67729068e+00
-1.06332707e+00 3.78077656e-01 -3.80919129e-01 3.98564339e-01
4.71155077e-01 4.73201334e-01 1.19522858e+00 -1.47949100e+00
8.34782243e-01 1.36685503e+00 4.97232527e-01 6.11546874e-01
-1.15805340e+00 -7.74265647e-01 6.60483301e-01 3.50066572e-01
-1.65275776e+00 -5.06174862e-01 8.91982794e-01 -5.45028865e-01
8.99534523e-01 2.69844711e-01 8.22690129e-01 1.33436143e+00
-1.69251025e-01 1.24121702e+00 4.50529724e-01 6.78674364e-03
9.63565558e-02 -1.50436580e-01 -1.76857069e-01 9.65307474e-01
-3.00546885e-01 -7.39794672e-02 -1.01262987e+00 -3.39264236e-02
1.14259899e+00 4.69164073e-01 -3.75170469e-01 -4.58336383e-01
-1.65244496e+00 4.19865221e-01 4.73683119e-01 2.54649937e-01
-3.21352243e-01 7.32686520e-01 5.33753037e-01 1.84054449e-01
3.77765030e-01 3.81331712e-01 -2.43121013e-01 -9.60100517e-02
-8.93662453e-01 2.28550494e-01 4.38238412e-01 1.40248513e+00
4.32165742e-01 9.02026892e-02 -6.37297451e-01 6.14015877e-01
3.35427783e-02 5.16229689e-01 1.88243151e-01 -1.11137307e+00
5.78708529e-01 4.59941179e-01 1.03432655e-01 -1.08918774e+00
1.54557258e-01 2.77808495e-02 -5.79787493e-01 -1.51722254e-02
2.78904855e-01 2.88258195e-01 -7.76090086e-01 1.87680387e+00
-1.18914798e-01 6.45045519e-01 -3.40341657e-01 1.10985470e+00
1.11000788e+00 6.97584391e-01 4.17343974e-01 1.90835923e-01
1.22316778e+00 -1.14211380e+00 -5.84126174e-01 -6.71330839e-02
5.79230227e-02 -2.10131913e-01 1.51655400e+00 2.48557121e-01
-1.08527720e+00 -8.09651732e-01 -7.22916901e-01 -2.63163298e-01
-4.31983352e-01 4.51916397e-01 6.69595838e-01 7.20134377e-02
-1.08120179e+00 4.52506751e-01 -7.65987813e-01 -4.27434027e-01
8.29996765e-01 2.40932167e-01 -5.95964968e-01 -1.81324542e-01
-8.60867560e-01 2.62001991e-01 1.77146360e-01 -2.14269180e-02
-1.37299907e+00 -9.85056698e-01 -8.83242786e-01 5.14485478e-01
5.77619851e-01 -6.12962365e-01 9.27519619e-01 -1.09033024e+00
-1.21312916e+00 6.87961876e-01 -2.84340680e-01 -9.27503929e-02
5.11359870e-01 -3.57647508e-01 -2.53051400e-01 6.36675060e-01
2.29582325e-01 1.08404005e+00 1.10335040e+00 -1.18087411e+00
-6.77949369e-01 4.55772802e-02 3.09758633e-01 1.41877860e-01
-3.99292588e-01 2.18160800e-03 -1.07054877e+00 -1.13264906e+00
-1.47655278e-01 -6.69221222e-01 4.12642092e-01 7.70587921e-01
-8.99879411e-02 -5.32463253e-01 1.32282722e+00 -4.47428554e-01
1.23342025e+00 -2.24871325e+00 3.28217149e-01 -1.31773368e-01
5.24903715e-01 3.05648029e-01 -4.98661160e-01 5.32604814e-01
-9.47408006e-02 2.35540450e-01 2.19578668e-02 -5.67793012e-01
1.98884625e-02 8.75080004e-02 -8.24924946e-01 2.90003777e-01
3.23581666e-01 1.48732841e+00 -1.07646096e+00 -6.49795234e-01
3.85889560e-01 4.78233933e-01 -6.63967669e-01 4.04248029e-01
-4.49029326e-01 1.73276603e-01 -3.16485822e-01 1.10723948e+00
2.27269039e-01 -4.76501882e-01 -3.73180099e-02 -4.15564358e-01
6.28638566e-02 -2.34636173e-01 -9.02037084e-01 2.04281425e+00
-3.67541283e-01 1.04944968e+00 4.60714987e-03 -7.04184592e-01
6.14457548e-01 3.27299505e-01 5.67629099e-01 -6.11214697e-01
-2.49133646e-01 -2.31372163e-01 -6.49821639e-01 -7.41469502e-01
2.38708466e-01 2.84689575e-01 4.24540229e-02 5.20730674e-01
5.39154530e-01 2.97848195e-01 -3.43219265e-02 5.70004046e-01
9.83437896e-01 2.88556725e-01 -2.13574711e-02 -1.36613712e-01
3.78739387e-01 -5.56569278e-01 1.96984038e-01 9.32047307e-01
-2.51757860e-01 6.23818457e-01 5.38244724e-01 -7.97937512e-01
-7.86028862e-01 -1.14158773e+00 2.71179378e-01 1.54202557e+00
4.22644705e-01 -6.18558586e-01 -4.36708540e-01 -9.63686883e-01
2.25465119e-01 3.13066512e-01 -8.99151325e-01 1.23513237e-01
-5.25600195e-01 2.63928562e-01 4.35656667e-01 8.20174396e-01
5.30311048e-01 -1.32624948e+00 -7.11082458e-01 -5.04879318e-02
-1.44808516e-01 -1.27087033e+00 -1.20966315e+00 -3.81931067e-01
-4.16387826e-01 -1.17010200e+00 -7.68079519e-01 -8.79004359e-01
5.95086575e-01 6.46390498e-01 1.13246191e+00 3.80155981e-01
-3.99062574e-01 1.01894355e+00 -2.89656490e-01 1.10316508e-01
3.22780639e-01 -8.41447785e-02 7.79937878e-02 3.24751168e-01
2.23822854e-02 -4.87231463e-01 -7.19465137e-01 4.68226552e-01
-8.43584955e-01 7.59272203e-02 3.96905810e-01 8.43555868e-01
4.74126071e-01 -2.44559973e-01 5.11593878e-01 -3.75113994e-01
3.61388832e-01 -4.12481397e-01 -5.26323497e-01 6.32160306e-01
-1.27175348e-02 -1.02406509e-01 5.02491415e-01 -7.96105504e-01
-6.16144538e-01 1.31424010e-01 1.76020876e-01 -1.40052378e+00
-5.90511151e-02 1.42410889e-01 -2.36495048e-01 -2.01371863e-01
2.30597287e-01 4.25278962e-01 -2.17814192e-01 -3.06122631e-01
5.27828455e-01 3.26874793e-01 8.48349035e-01 -8.70883405e-01
6.03566706e-01 6.31796002e-01 -2.28765234e-01 -8.98029268e-01
-5.47342539e-01 -2.73646206e-01 -6.57944798e-01 -6.13897622e-01
9.15629208e-01 -9.16077912e-01 -1.23518622e+00 1.46943972e-01
-1.20983171e+00 -5.96416831e-01 -2.44223997e-01 1.63000971e-01
-6.68128252e-01 3.22125942e-01 -5.03076375e-01 -8.22006106e-01
-3.47323149e-01 -1.19181836e+00 1.44714200e+00 -7.67007098e-02
-1.25825375e-01 -9.26353872e-01 -3.08430374e-01 4.03824123e-03
2.50259817e-01 8.39850157e-02 8.98815930e-01 -3.54939908e-01
-1.22985721e+00 -3.75513174e-02 -6.61183536e-01 5.27309068e-02
-8.02374110e-02 2.85984408e-02 -9.26739335e-01 -4.89673346e-01
-5.07174790e-01 -5.42922735e-01 9.95236099e-01 1.83557346e-01
1.72703731e+00 -5.11321187e-01 -3.87876093e-01 9.27339971e-01
1.26852405e+00 9.13946778e-02 3.88486564e-01 -1.45462930e-01
8.23562086e-01 2.78378040e-01 5.30922174e-01 2.36773968e-01
3.25657994e-01 8.77805471e-01 5.51452577e-01 1.22925444e-02
-4.01589155e-01 -8.02457869e-01 3.14156651e-01 4.72067714e-01
4.10632305e-02 -4.14798915e-01 -7.25862086e-01 7.42557168e-01
-2.00391102e+00 -1.31659889e+00 5.52109420e-01 1.91685712e+00
2.22201094e-01 -9.04833823e-02 3.09711128e-01 -3.68770242e-01
5.90424240e-01 4.86212969e-01 -6.49196565e-01 2.84996539e-01
7.95725584e-02 -5.71465865e-02 8.30710232e-02 2.29201868e-01
-1.31420135e+00 1.10457349e+00 6.45891094e+00 8.48472774e-01
-1.24960411e+00 1.45589650e-01 4.56896424e-01 -2.92352885e-01
-1.81299269e-01 -1.66811168e-01 -3.56666267e-01 6.60633683e-01
8.44951123e-02 5.71417287e-02 8.42732012e-01 8.66659105e-01
-2.07904160e-01 1.09163120e-01 -1.50495684e+00 1.40641201e+00
3.12294275e-01 -1.80020738e+00 3.50178987e-01 -3.05961907e-01
4.35055435e-01 -1.15374632e-01 2.33538046e-01 4.91403550e-01
-1.64753705e-01 -1.11112046e+00 1.04085600e+00 6.72416866e-01
1.43111312e+00 -4.46754009e-01 1.37495324e-01 3.95809524e-02
-1.67991424e+00 -3.91047239e-01 -2.19181910e-01 1.62426040e-01
3.13271768e-02 -2.34721497e-01 -4.78698283e-01 2.35688224e-01
8.68534625e-01 1.20287633e+00 -5.09768307e-01 9.22578990e-01
-9.50720310e-02 3.44556719e-01 -1.59834832e-01 4.10272367e-02
2.93377787e-01 1.19709320e-01 6.06564999e-01 1.31344581e+00
2.02365503e-01 2.49611333e-01 2.95308113e-01 1.28980386e+00
-3.28813285e-01 -2.50966758e-01 -7.24049211e-01 -2.65725076e-01
7.94051647e-01 9.23075974e-01 -7.24884391e-01 -3.76178354e-01
-5.07741392e-01 1.11554456e+00 5.19779027e-01 8.21915507e-01
-8.78310800e-01 -3.30951035e-01 1.06187284e+00 1.00161560e-01
6.52303040e-01 -3.07782471e-01 1.53644979e-01 -1.30910611e+00
2.64253467e-01 -5.95157266e-01 3.07559937e-01 -1.07197070e+00
-1.27160490e+00 6.58606350e-01 -2.82151438e-02 -1.26559043e+00
1.68302640e-01 -6.07737720e-01 -7.08739400e-01 5.87524891e-01
-1.19207513e+00 -1.47511733e+00 -5.90351284e-01 7.89905608e-01
1.03273582e+00 -3.79913390e-01 6.16437018e-01 4.70779508e-01
-5.11586905e-01 7.54545987e-01 -3.25619251e-01 4.90440577e-01
6.64466679e-01 -9.61703479e-01 6.24822915e-01 6.13864362e-01
4.78264451e-01 6.19725585e-01 9.40070674e-02 -6.78562164e-01
-1.74868715e+00 -1.29999447e+00 4.26535428e-01 -8.95212710e-01
5.60214281e-01 -8.77925277e-01 -8.58776629e-01 8.59358847e-01
-4.66953292e-02 6.01471603e-01 1.71512187e-01 -1.39659777e-01
-7.34244287e-01 -1.02624439e-01 -7.63013601e-01 5.85831225e-01
1.44686890e+00 -1.11489630e+00 -3.90307277e-01 2.42485777e-01
8.91073346e-01 -4.27436799e-01 -6.84307873e-01 1.18738227e-01
1.08984566e+00 -8.04442704e-01 1.42491376e+00 -9.20782268e-01
4.43572432e-01 -3.38771731e-01 -3.01467359e-01 -8.12490165e-01
-3.80977511e-01 -6.36555433e-01 -5.28101742e-01 1.12839234e+00
-1.28433079e-01 -1.28391549e-01 7.34803438e-01 6.56671286e-01
5.16300499e-02 -9.52618957e-01 -9.38032985e-01 -7.44051516e-01
-4.11426693e-01 -5.62136829e-01 5.96333086e-01 8.66869390e-01
-1.08928591e-01 7.99594671e-02 -7.56488860e-01 9.14572477e-02
4.19371873e-01 3.14987719e-01 7.94377863e-01 -8.68553340e-01
-1.65563822e-01 -5.64667642e-01 -4.55387235e-01 -1.39835048e+00
4.90629077e-01 -7.80579388e-01 -1.91963539e-01 -1.28820062e+00
2.90169001e-01 -4.80290473e-01 -2.73915321e-01 6.75850511e-01
-1.95378542e-01 4.21904534e-01 6.84140027e-01 3.22005689e-01
-1.22225153e+00 6.72499537e-01 1.04763687e+00 -4.30862844e-01
-9.80897099e-02 -2.54720062e-01 -3.14805180e-01 3.78760099e-01
1.19239882e-01 -2.62788266e-01 -5.73535204e-01 -6.99334979e-01
-7.60267824e-02 4.09750581e-01 8.95237207e-01 -8.82791817e-01
3.41200203e-01 -2.42667437e-01 4.35274690e-01 -6.29616559e-01
5.88251173e-01 -9.41133440e-01 1.04159047e-03 8.37613791e-02
-6.56172156e-01 1.46002993e-01 2.06255287e-01 8.51017356e-01
-9.44106951e-02 3.01431924e-01 5.36585212e-01 2.46224310e-02
-1.16103148e+00 8.66976321e-01 -1.06859304e-01 7.63722733e-02
1.06528020e+00 -1.01402536e-01 -4.00125712e-01 -5.31133235e-01
-5.51712155e-01 3.85845035e-01 5.27498901e-01 7.88985312e-01
9.48140323e-01 -1.59950781e+00 -3.92448187e-01 4.53367174e-01
3.90852302e-01 -2.38189876e-01 4.60283816e-01 5.07518649e-01
-3.65268707e-01 5.98294199e-01 -2.21571192e-01 -7.92454839e-01
-1.14417112e+00 1.19249535e+00 2.25820109e-01 3.40635121e-01
-9.54516828e-01 9.85809445e-01 6.96506917e-01 7.33054057e-02
8.13408017e-01 -3.95980299e-01 1.66305348e-01 -1.23883322e-01
6.93840683e-01 1.64885163e-01 -3.60044479e-01 -5.19273877e-01
-4.07201678e-01 6.56962872e-01 2.18517497e-01 1.20017538e-02
1.11575747e+00 -5.72049692e-02 1.38754293e-01 3.27995688e-01
1.29214096e+00 -3.31967264e-01 -1.85318875e+00 -3.57282192e-01
-2.63825417e-01 -9.12549078e-01 -3.14376019e-02 -7.33151376e-01
-1.25737107e+00 1.02129149e+00 4.49911088e-01 -1.56140581e-01
1.02597356e+00 2.12359875e-01 6.42916620e-01 5.00452101e-01
4.04438823e-01 -6.78512871e-01 5.36637545e-01 2.02784032e-01
1.29889309e+00 -1.18325591e+00 -1.18553296e-01 -2.13600080e-02
-7.76205659e-01 1.12424219e+00 7.19510019e-01 -1.65037528e-01
5.69765866e-01 5.67441732e-02 -2.87623584e-01 -2.84984440e-01
-1.04577470e+00 -9.76562500e-02 5.62510073e-01 4.93610919e-01
-4.47900109e-02 -2.25234777e-01 5.44709086e-01 6.40201271e-01
5.06731868e-01 1.19712085e-01 -9.97611210e-02 6.91439509e-01
-1.72444489e-02 -6.42758191e-01 -1.03608862e-01 4.06910390e-01
4.24565785e-02 -3.09407450e-02 -4.11176652e-01 9.71163809e-01
-3.61683108e-02 4.66752082e-01 3.06690395e-01 -4.16058689e-01
1.64606646e-01 2.90879291e-02 3.89626205e-01 -3.61781657e-01
-3.39774013e-01 -4.90302034e-02 -4.13931459e-01 -9.68526542e-01
-2.84726918e-01 -5.00581503e-01 -7.52451479e-01 -1.56213760e-01
7.00768605e-02 -7.66606182e-02 3.05433005e-01 7.72547364e-01
6.83523774e-01 4.61727023e-01 3.85630459e-01 -1.38923562e+00
-3.51624824e-02 -5.76633096e-01 -3.17961097e-01 4.30521816e-01
7.57629633e-01 -9.25165057e-01 -2.67037358e-02 8.54196176e-02] | [9.74278450012207, 0.7647556662559509] |
51598989-89d8-40fb-85db-aecb6cd4ec05 | not-only-generative-art-stable-diffusion-for | 2304.10278 | null | https://arxiv.org/abs/2304.10278v1 | https://arxiv.org/pdf/2304.10278v1.pdf | Not Only Generative Art: Stable Diffusion for Content-Style Disentanglement in Art Analysis | The duality of content and style is inherent to the nature of art. For humans, these two elements are clearly different: content refers to the objects and concepts in the piece of art, and style to the way it is expressed. This duality poses an important challenge for computer vision. The visual appearance of objects and concepts is modulated by the style that may reflect the author's emotions, social trends, artistic movement, etc., and their deep comprehension undoubtfully requires to handle both. A promising step towards a general paradigm for art analysis is to disentangle content and style, whereas relying on human annotations to cull a single aspect of artworks has limitations in learning semantic concepts and the visual appearance of paintings. We thus present GOYA, a method that distills the artistic knowledge captured in a recent generative model to disentangle content and style. Experiments show that synthetically generated images sufficiently serve as a proxy of the real distribution of artworks, allowing GOYA to separately represent the two elements of art while keeping more information than existing methods. | ['Noa Garcia', 'Yuta Nakashima', 'Yankun Wu'] | 2023-04-20 | null | null | null | null | ['art-analysis'] | ['computer-vision'] | [ 1.52261004e-01 -1.70134619e-01 3.48799825e-02 -2.80544162e-01
1.28246203e-01 -1.00147617e+00 1.12653530e+00 4.47634608e-02
7.53658190e-02 4.38200623e-01 4.41401601e-01 3.09961110e-01
-1.44568747e-02 -9.55815613e-01 -4.55175728e-01 -6.21811688e-01
4.42230225e-01 6.73420846e-01 3.03495415e-02 -4.32886273e-01
3.73307616e-01 6.45991147e-01 -1.71803105e+00 5.42556524e-01
4.34698462e-01 8.66161764e-01 2.06124425e-01 7.43791699e-01
-7.40490198e-01 9.32054937e-01 -8.53547394e-01 -5.37538767e-01
9.57241878e-02 -6.60533011e-01 -6.63953781e-01 4.20817047e-01
7.55616307e-01 2.04236314e-01 -1.04546890e-01 1.13641036e+00
7.03135692e-03 -2.36612961e-01 1.01335001e+00 -1.14158297e+00
-8.51308465e-01 4.28700835e-01 -4.25350010e-01 -2.07577348e-02
3.20823878e-01 1.07474171e-01 1.09610689e+00 -5.38646281e-01
7.87884176e-01 1.57138264e+00 4.20671880e-01 4.32754159e-01
-1.68476927e+00 -4.22812283e-01 2.30917260e-01 -6.40356839e-02
-1.16427445e+00 -3.31157804e-01 1.19690192e+00 -9.35740590e-01
8.86906311e-03 4.16878670e-01 1.17105663e+00 1.42894375e+00
-4.07453254e-03 7.59320557e-01 1.45826578e+00 -4.82157469e-01
4.35002148e-01 4.90403205e-01 -2.02710792e-01 4.25658256e-01
1.28045097e-01 -1.82879448e-01 -5.55125177e-01 -4.76363041e-02
1.06297982e+00 1.37228221e-01 -2.56977528e-01 -6.40004039e-01
-1.19774520e+00 6.88579559e-01 4.39188898e-01 4.88409400e-01
-2.57203281e-01 2.16473505e-01 1.22522816e-01 2.65001684e-01
3.82390052e-01 7.37530470e-01 -1.85700029e-01 -1.28000677e-01
-1.01984131e+00 3.22722405e-01 7.92832255e-01 7.75933743e-01
8.38696778e-01 -5.40264882e-02 -1.50789618e-01 7.46981680e-01
2.55626440e-01 5.58067858e-01 2.59809703e-01 -9.90978122e-01
-1.62251547e-01 9.20732975e-01 -8.78702477e-03 -1.24190366e+00
1.38088539e-01 -4.45391208e-01 -4.90656793e-01 6.72918975e-01
6.73567474e-01 3.99283797e-01 -9.66853976e-01 1.75766528e+00
6.32567927e-02 -2.58817464e-01 -3.28991085e-01 7.66548395e-01
6.11882091e-01 3.11035305e-01 2.16758132e-01 1.88232958e-01
1.69846177e+00 -6.94640100e-01 -6.84040785e-01 -5.13311923e-01
-9.60309803e-02 -9.90689576e-01 1.36096513e+00 3.90570670e-01
-1.28913403e+00 -5.47811389e-01 -9.75147128e-01 -4.85442914e-02
-6.12682283e-01 6.47457987e-02 5.72627902e-01 5.69627345e-01
-9.10388947e-01 5.50304353e-01 -4.88091826e-01 -4.51427251e-01
6.81974053e-01 -1.01780608e-01 -1.05146252e-01 1.44159392e-01
-5.87576330e-01 8.55103016e-01 2.81536609e-01 -1.27389073e-01
-3.82888466e-01 -6.45768225e-01 -6.45576477e-01 -8.55869129e-02
2.34738901e-01 -8.34065259e-01 9.77690399e-01 -1.58762562e+00
-1.27538955e+00 1.27258098e+00 7.16376060e-04 1.67869508e-01
7.78434277e-01 -8.20465013e-02 -3.16975474e-01 6.12660199e-02
-5.34731895e-02 6.96955383e-01 1.20483625e+00 -1.84532464e+00
-3.32098633e-01 -6.00401044e-01 2.34003335e-01 1.31129497e-03
-2.94939876e-01 -2.58560240e-01 -3.89062643e-01 -9.30744588e-01
1.26206875e-01 -8.82632375e-01 1.86846420e-01 3.95055711e-01
-2.50359327e-01 7.00784251e-02 9.08831537e-01 -2.59959906e-01
9.38832581e-01 -2.16509056e+00 4.33767498e-01 1.01458594e-01
4.56937402e-01 2.91988365e-02 -7.90092498e-02 6.25032067e-01
1.31296575e-01 1.12949580e-01 -1.11352310e-01 -3.25642616e-01
2.55433887e-01 3.75181466e-01 -4.78179961e-01 2.61763722e-01
2.44242698e-01 1.01873112e+00 -1.03564239e+00 -4.42951798e-01
2.60174334e-01 7.72443116e-01 -2.62994319e-01 8.40424076e-02
-6.64114475e-01 5.95999002e-01 -5.17362714e-01 7.08647430e-01
4.22703058e-01 -2.59554923e-01 2.62588471e-01 -3.76682609e-01
1.02795541e-01 7.85883814e-02 -1.05220163e+00 1.79979944e+00
-4.68942612e-01 1.05225337e+00 9.18513611e-02 -4.85326052e-01
1.00175488e+00 7.10218474e-02 6.34227470e-02 -6.26804233e-01
3.81973445e-01 -1.16394609e-01 -2.11329073e-01 -5.42281330e-01
4.28660333e-01 -5.08564055e-01 7.36834630e-02 5.50351083e-01
-1.55959100e-01 -5.91917396e-01 1.35051161e-01 1.39051244e-01
8.62826049e-01 3.48435342e-01 1.17775008e-01 -3.92282933e-01
1.46079481e-01 -5.97098768e-02 1.76553160e-01 5.69532096e-01
4.78154384e-02 7.28229523e-01 6.03379548e-01 -6.03006721e-01
-1.05338275e+00 -1.37971044e+00 -4.32269163e-02 1.09549701e+00
5.29530197e-02 -1.70841694e-01 -6.72985613e-01 -4.03746545e-01
5.86198531e-02 6.95583344e-01 -1.06307685e+00 2.97694933e-02
-3.61671329e-01 -4.03167844e-01 2.80659676e-01 3.97368819e-01
1.25823572e-01 -1.16664994e+00 -8.26177061e-01 -9.40750092e-02
-5.34077398e-02 -1.07341743e+00 -2.33199984e-01 -2.50593305e-01
-7.83010423e-01 -1.12343001e+00 -6.03983343e-01 -5.39549410e-01
7.21779883e-01 -4.95682508e-02 1.79359460e+00 8.10152143e-02
-4.68087524e-01 6.55832529e-01 -4.12765652e-01 -6.11439228e-01
-6.02862060e-01 -2.28392243e-01 -4.74078298e-01 2.78404683e-01
3.00721616e-01 -1.02842271e+00 -6.44963324e-01 4.02811505e-02
-1.11792278e+00 2.31374353e-01 6.39626384e-01 4.10806119e-01
1.78793624e-01 5.82129434e-02 -1.10097672e-03 -1.21943128e+00
4.38571364e-01 -2.59347528e-01 -1.31698072e-01 3.13700318e-01
-4.02051777e-01 1.88408211e-01 2.92086452e-01 -6.48492396e-01
-1.00159931e+00 -7.58790523e-02 3.76095265e-01 -5.65570354e-01
-5.48516989e-01 -2.82285456e-02 -3.57555538e-01 2.12577283e-01
5.02999723e-01 1.54728025e-01 8.36313069e-02 -5.23954272e-01
6.63777590e-01 2.72450000e-01 6.71396613e-01 -8.31389546e-01
8.14062178e-01 9.91141200e-01 1.63566638e-02 -9.22241986e-01
-9.58772779e-01 -2.12290034e-01 -8.63963783e-01 -3.27140301e-01
8.08139443e-01 -6.86102271e-01 -6.72094822e-01 1.42049477e-01
-1.18142772e+00 -1.85379699e-01 -6.69615448e-01 8.23431537e-02
-6.23658478e-01 1.68940127e-01 -3.08667213e-01 -8.08703482e-01
1.49652660e-01 -9.17612612e-01 1.30716050e+00 3.53997014e-02
-7.00114012e-01 -1.29376817e+00 1.80729166e-01 3.00295293e-01
3.83954287e-01 7.22642243e-01 1.18207812e+00 -1.72114566e-01
-6.07161760e-01 -1.40576541e-01 -2.56085753e-01 1.05714388e-01
3.71534497e-01 2.44156092e-01 -1.20600057e+00 5.09707257e-02
1.58789292e-01 -2.02455103e-01 7.49039888e-01 9.84634683e-02
8.82136822e-01 -3.12739313e-01 -2.15206277e-02 3.61278683e-01
1.64613998e+00 6.90940320e-02 7.73015678e-01 2.28846014e-01
9.27571535e-01 9.46656108e-01 3.77799422e-02 3.52809757e-01
1.48509875e-01 6.77841723e-01 3.28676373e-01 -1.78041771e-01
-3.73736918e-01 -3.64753664e-01 1.97465822e-01 5.07315397e-01
-3.79458964e-01 -1.88688084e-01 -7.32341230e-01 2.20377326e-01
-1.62871325e+00 -1.14796877e+00 -2.18276560e-01 1.90543830e+00
8.44562471e-01 1.83739200e-01 2.00864017e-01 1.30812198e-01
5.35955369e-01 2.75969297e-01 -3.93190771e-01 -3.19578111e-01
-2.08901659e-01 1.43004552e-01 4.73835431e-02 1.32097915e-01
-6.49081349e-01 7.98076212e-01 7.10750818e+00 8.19528222e-01
-1.15186226e+00 -1.68476418e-01 4.56595540e-01 -4.22053784e-02
-6.85628355e-01 -3.00791282e-02 -2.76134908e-01 5.44641733e-01
3.01102519e-01 1.79667339e-01 5.61743319e-01 6.01771295e-01
3.21586654e-02 -6.32977635e-02 -1.36761248e+00 1.06618631e+00
3.32684934e-01 -1.21150005e+00 4.21276063e-01 1.34685352e-01
6.55481994e-01 -5.26762366e-01 2.22121149e-01 -1.56117260e-01
3.23038310e-01 -1.10071468e+00 1.22592783e+00 9.41352129e-01
5.97266912e-01 -2.80251443e-01 2.37106889e-01 8.01564530e-02
-1.03432870e+00 9.64388698e-02 -1.90287009e-02 -2.85950422e-01
2.13146489e-03 5.08062303e-01 -5.03809810e-01 9.22054797e-02
4.43996578e-01 6.81929529e-01 -8.66523445e-01 4.77121949e-01
-2.36389443e-01 1.84005246e-01 1.69091433e-01 -1.09804869e-02
-1.97647642e-02 -4.44917530e-01 6.60460889e-01 1.16009498e+00
1.17170632e-01 -1.93532541e-01 7.19139054e-02 1.48884642e+00
1.17336139e-01 -1.78703234e-01 -5.37703395e-01 -5.34786284e-01
3.07426900e-01 1.38306034e+00 -1.02282345e+00 -3.00153375e-01
-1.65288851e-01 1.07341003e+00 2.40886942e-01 3.76690477e-01
-4.45879042e-01 1.81920618e-01 7.31468141e-01 4.56463218e-01
1.83536425e-01 -3.36811066e-01 -5.49778879e-01 -1.04156959e+00
-1.08548373e-01 -7.30378568e-01 -1.09522134e-01 -1.30020034e+00
-1.46224725e+00 6.94895446e-01 -1.67124882e-01 -1.19317424e+00
5.65288998e-02 -8.44332457e-01 -7.14503467e-01 7.91913748e-01
-1.13429570e+00 -1.67216408e+00 -6.23920202e-01 2.90873587e-01
7.01377451e-01 6.03998490e-02 8.24340701e-01 -1.09006122e-01
-7.18564466e-02 1.95985325e-02 -1.18796840e-01 6.56569302e-02
6.12034619e-01 -1.60278916e+00 1.11062735e-01 3.55455548e-01
6.99093699e-01 5.09828627e-01 1.07400072e+00 -4.09449816e-01
-1.33808160e+00 -6.52310967e-01 5.88500857e-01 -1.01639891e+00
7.64417529e-01 -4.82619047e-01 -7.59996116e-01 4.33157116e-01
1.97491914e-01 -4.42089081e-01 6.93646371e-01 2.90378660e-01
-8.21728587e-01 1.06297977e-01 -8.56465638e-01 8.06099594e-01
1.07816124e+00 -7.77673244e-01 -7.78487146e-01 2.57971287e-02
3.19316030e-01 -7.14993402e-02 -7.25913346e-01 1.09162331e-01
7.80244052e-01 -1.38260925e+00 1.01422966e+00 -4.22055364e-01
7.18253076e-01 -3.10442328e-01 -1.95506960e-01 -1.27605844e+00
-4.45846915e-01 -1.98061168e-01 1.78906024e-02 1.43864584e+00
2.62699872e-02 -1.57289118e-01 7.61913180e-01 5.45285285e-01
2.60842741e-01 -4.93436307e-01 -1.91444337e-01 -5.67084610e-01
1.90579761e-02 -4.20162201e-01 7.51450062e-01 1.10702658e+00
-3.68390232e-01 5.10586262e-01 -1.93658564e-02 -2.48160005e-01
4.62675661e-01 3.65511626e-01 7.69302905e-01 -1.84012687e+00
-3.31460595e-01 -1.00154901e+00 -6.41013384e-01 -4.92032021e-01
1.63516961e-02 -7.24476516e-01 -2.85660952e-01 -1.47582138e+00
3.00433964e-01 -5.38640082e-01 -7.17118084e-02 6.19781800e-02
1.16499029e-01 6.54365063e-01 2.89194375e-01 4.21179205e-01
-2.92601496e-01 4.32340622e-01 1.54323351e+00 -3.39751482e-01
-1.04264066e-01 -3.03937763e-01 -8.54044318e-01 1.02458441e+00
6.24544799e-01 -1.49399787e-01 -4.99086767e-01 -5.10304451e-01
5.04411221e-01 -3.53767335e-01 7.59959638e-01 -8.98047268e-01
-7.86378011e-02 -3.74307632e-01 6.45544946e-01 -2.73948640e-01
6.67893171e-01 -9.71307397e-01 5.02559125e-01 2.57296532e-01
-3.83231550e-01 -1.78106986e-02 -3.31221446e-02 8.02351356e-01
-1.92382559e-01 -9.76203457e-02 6.88793063e-01 -5.81888199e-01
-5.74003279e-01 2.58767996e-02 -1.46763071e-01 2.05105603e-01
6.72890246e-01 -4.54789609e-01 -2.52802432e-01 -2.85158694e-01
-7.92632937e-01 -3.48629981e-01 1.10275733e+00 6.53440773e-01
3.05275887e-01 -1.29592955e+00 -4.89850938e-01 1.72811061e-01
2.40473926e-01 -2.79953182e-01 2.07909048e-01 3.39481205e-01
-4.69633609e-01 -1.80328265e-01 -4.64525491e-01 -6.12759054e-01
-1.18754649e+00 5.60865521e-01 1.44116238e-01 1.52835593e-01
-5.86511552e-01 7.05726087e-01 6.10720456e-01 5.04008681e-02
4.10951786e-02 -5.84395379e-02 -2.43076593e-01 4.25303698e-01
4.85365540e-01 2.98432112e-01 -2.92872995e-01 -7.88662851e-01
-1.16529325e-02 8.42475832e-01 1.70093536e-01 -2.73167372e-01
1.19514680e+00 -1.42962202e-01 -3.51098329e-01 1.15458202e+00
8.75526726e-01 3.38632643e-01 -1.18362820e+00 -2.04901487e-01
-6.83794543e-02 -8.40895057e-01 -2.45247688e-02 -1.04448867e+00
-9.27303851e-01 1.05231810e+00 4.48786855e-01 6.39042914e-01
1.03877032e+00 3.09490830e-01 4.05186206e-01 -6.06369004e-02
4.03730065e-01 -1.05547154e+00 4.81883436e-01 4.99147438e-02
1.08099139e+00 -9.91722465e-01 5.76791354e-02 -4.70567018e-01
-6.20427310e-01 1.20343685e+00 2.57926077e-01 -1.17381841e-01
5.25860727e-01 1.11700505e-01 2.27169394e-01 -5.71006477e-01
-5.00934660e-01 -3.07378769e-01 6.92179620e-01 5.09986103e-01
5.60718060e-01 2.16886804e-01 6.51876330e-02 2.29957312e-01
-4.67404723e-01 -2.15014324e-01 3.70323330e-01 8.73782337e-01
-3.42524737e-01 -1.18284750e+00 -4.70044255e-01 9.43090320e-02
-2.82871872e-01 2.52114922e-01 -9.44725752e-01 6.80126488e-01
5.77577710e-01 6.36206090e-01 4.18409616e-01 -1.97601736e-01
3.10381204e-01 1.27228484e-01 9.31355536e-01 -5.26750386e-01
-3.55272651e-01 8.95072445e-02 -2.81777710e-01 -3.30038846e-01
-7.30571032e-01 -4.41490531e-01 -6.90440297e-01 -3.08531433e-01
5.58765978e-02 -1.05061248e-01 8.53395164e-01 9.17350769e-01
8.59066993e-02 6.48205817e-01 3.81905168e-01 -1.01809597e+00
1.21225707e-01 -7.37849057e-01 -7.95645475e-01 8.89466524e-01
2.22841233e-01 -8.16091418e-01 -4.19796795e-01 5.10804892e-01] | [11.369782447814941, 0.13686326146125793] |
5e79a28b-253c-401a-8417-10a83d83754d | face-phylogeny-tree-using-basis-functions | 2002.09068 | null | https://arxiv.org/abs/2002.09068v2 | https://arxiv.org/pdf/2002.09068v2.pdf | Face Phylogeny Tree Using Basis Functions | Photometric transformations, such as brightness and contrast adjustment, can be applied to a face image repeatedly creating a set of near-duplicate images. Identifying the original image from a set of such near-duplicates and deducing the relationship between them are important in the context of digital image forensics. This is commonly done by generating an image phylogeny tree \textemdash \hspace{0.08cm} a hierarchical structure depicting the relationship between a set of near-duplicate images. In this work, we utilize three different families of basis functions to model pairwise relationships between near-duplicate images. The basis functions used in this work are orthogonal polynomials, wavelet basis functions and radial basis functions. We perform extensive experiments to assess the performance of the proposed method across three different modalities, namely, face, fingerprint and iris images; across different image phylogeny tree configurations; and across different types of photometric transformations. We also utilize the same basis functions to model geometric transformations and deep-learning based transformations. We also perform extensive analysis of each basis function with respect to its ability to model arbitrary transformations and to distinguish between the original and the transformed images. Finally, we utilize the concept of approximate von Neumann graph entropy to explain the success and failure cases of the proposed IPT generation algorithm. Experiments indicate that the proposed algorithm generalizes well across different scenarios thereby suggesting the merits of using basis functions to model the relationship between photometrically and geometrically modified images. | ['Sudipta Banerjee', 'Arun Ross'] | 2020-02-21 | null | null | null | null | ['image-forensics'] | ['computer-vision'] | [ 4.43114549e-01 -1.28441155e-01 4.01151717e-01 -4.24099475e-01
-3.16125870e-01 -7.52266824e-01 6.93085670e-01 -1.92190737e-01
1.62514150e-02 4.28682655e-01 -2.63764173e-01 -2.62290567e-01
-4.10005629e-01 -7.28602886e-01 -7.54342556e-01 -8.20062935e-01
2.18620244e-02 3.42829973e-01 -5.79694547e-02 -6.24475963e-02
4.15238976e-01 9.58440423e-01 -1.68873000e+00 1.19419485e-01
6.44878924e-01 8.45472813e-01 -1.95337474e-01 6.08023345e-01
1.80417243e-02 2.64004767e-01 -6.66902959e-01 -9.58016336e-01
6.88645780e-01 -4.46587771e-01 -8.92233551e-01 3.92901033e-01
8.27033579e-01 -2.40002245e-01 -2.96382606e-01 1.21987987e+00
4.20428246e-01 1.54511064e-01 7.30291247e-01 -1.35000932e+00
-8.82778287e-01 9.72094983e-02 -7.84173369e-01 1.83993757e-01
4.19731259e-01 2.35710412e-01 6.21226311e-01 -5.24930537e-01
6.77478194e-01 1.45633042e+00 6.82504058e-01 3.14179361e-01
-1.47311711e+00 -6.09630644e-01 -5.38363039e-01 4.54249024e-01
-1.52529764e+00 -5.30781209e-01 9.36147332e-01 -3.99028212e-01
4.15198296e-01 3.70539635e-01 2.55177557e-01 9.13275123e-01
3.43370765e-01 -8.34522396e-02 1.55625522e+00 -6.12635553e-01
2.69965231e-02 6.89477399e-02 7.93526247e-02 7.78051198e-01
3.49572510e-01 2.35917225e-01 -4.52372879e-01 -5.17201781e-01
5.79565942e-01 -3.50402504e-01 -3.27469736e-01 -2.55122066e-01
-7.10870683e-01 6.40087962e-01 2.76625514e-01 1.52017653e-01
-2.91750133e-01 7.75195584e-02 -8.49350765e-02 1.51328459e-01
1.05438575e-01 2.86154747e-01 2.39180401e-01 3.39520633e-01
-7.09971428e-01 8.59702844e-03 7.12349951e-01 7.61860609e-01
9.59706903e-01 -2.59002931e-02 -1.00792088e-01 7.97343016e-01
7.50963986e-02 3.42720032e-01 2.86564142e-01 -1.22505736e+00
1.06616668e-01 3.43257070e-01 -3.44937108e-02 -1.31244719e+00
-2.16529921e-01 6.49135420e-03 -7.54845440e-01 3.21651071e-01
6.00307286e-01 2.51019627e-01 -9.72621262e-01 1.66887319e+00
5.04799902e-01 4.37262058e-01 -1.18927561e-01 5.98738134e-01
6.85301363e-01 3.19580108e-01 -3.09170902e-01 -1.26955435e-01
1.42875195e+00 -2.38329113e-01 -3.30340952e-01 2.35129535e-01
1.18282013e-01 -9.87577319e-01 8.10734034e-01 3.42384547e-01
-9.81961548e-01 -6.09160304e-01 -9.39724684e-01 -7.83877745e-02
-2.72096753e-01 1.15829892e-01 2.18567058e-01 9.26688433e-01
-1.20930803e+00 7.08931386e-01 -4.52992320e-01 -5.15771925e-01
2.16190293e-01 5.34316778e-01 -5.52762866e-01 -1.42707109e-01
-8.40624750e-01 6.42154217e-01 3.00032884e-01 1.84357092e-01
-4.64299351e-01 -4.48456764e-01 -6.15782022e-01 1.48790210e-01
-5.07426225e-02 -6.95989013e-01 5.70344806e-01 -9.67101216e-01
-1.30265796e+00 1.19366884e+00 -2.67493039e-01 -3.09330940e-01
4.11055475e-01 4.54434752e-01 -5.15209675e-01 3.51862788e-01
-1.57789677e-01 4.79864478e-01 1.20813096e+00 -1.44008672e+00
-1.77683800e-01 -6.31739259e-01 -9.91481394e-02 -1.64027978e-02
-2.85223752e-01 -2.78644953e-02 -3.79026026e-01 -3.81339908e-01
2.02844352e-01 -1.27377510e+00 3.27514738e-01 3.99077795e-02
-6.55131280e-01 8.16656128e-02 7.95894325e-01 -8.17167699e-01
6.35781586e-01 -2.21731663e+00 -4.27807644e-02 6.46622241e-01
7.80914649e-02 9.72084552e-02 -3.65123570e-01 4.17161942e-01
-2.26932287e-01 1.43200636e-01 -3.38401139e-01 -1.75889641e-01
-1.70475304e-01 2.26477623e-01 -2.39926532e-01 8.00012648e-01
1.76150620e-01 4.35418516e-01 -4.83339638e-01 -4.40463126e-01
2.57752478e-01 7.13423312e-01 -2.45224267e-01 2.96891611e-02
2.31056526e-01 6.04595065e-01 -1.33926436e-01 4.88001525e-01
1.12616813e+00 9.57678929e-02 2.88522631e-01 -4.42306936e-01
2.38014802e-01 -2.09234759e-01 -9.93237257e-01 1.15804398e+00
-2.49046832e-01 6.47358179e-01 -4.43255752e-02 -9.31289554e-01
1.25883412e+00 -5.71184210e-04 4.18424398e-01 -6.30407691e-01
1.77578554e-01 -3.51467878e-02 2.19440058e-01 -4.84747082e-01
4.16985154e-01 -1.39685512e-01 3.95130813e-01 5.12922645e-01
5.31866923e-02 -3.75498347e-02 1.59209460e-01 -1.64034188e-01
8.71440411e-01 -2.80595589e-02 -1.03061004e-02 -2.18798786e-01
7.52919793e-01 -4.12875026e-01 3.93100619e-01 6.25283539e-01
-1.61171228e-01 7.88103282e-01 5.34147799e-01 -4.68680650e-01
-1.17349815e+00 -1.15801549e+00 -4.74151075e-01 5.49883127e-01
4.71639246e-01 -7.38316216e-03 -7.81880498e-01 -2.70937711e-01
1.61288917e-01 7.74139166e-01 -5.75509608e-01 -3.66804600e-01
-4.29475754e-01 -8.60777557e-01 9.27738011e-01 -9.40692052e-02
6.86119974e-01 -7.71728814e-01 -4.33316648e-01 -3.86929750e-01
-1.41198114e-01 -1.35699522e+00 -1.71164885e-01 -4.20055658e-01
-6.06597602e-01 -1.40865576e+00 -2.80796677e-01 -5.80655098e-01
7.79112875e-01 2.20743403e-01 8.70026052e-01 1.15316562e-01
-5.36988497e-01 7.68549025e-01 -7.70307630e-02 2.47488655e-02
-8.33829761e-01 -5.08910179e-01 1.69237792e-01 6.81438446e-01
1.98730290e-01 -7.39904523e-01 -5.82790375e-01 4.42649901e-01
-1.14062142e+00 -3.72806042e-01 3.25964332e-01 6.83883011e-01
5.02497792e-01 3.41945142e-01 -1.10212460e-01 -6.40832126e-01
6.94882989e-01 -3.28558475e-01 -7.29364276e-01 6.51940644e-01
-5.22093117e-01 1.87344775e-01 6.00417614e-01 -3.12880009e-01
-1.26631534e+00 -6.99122176e-02 3.07726920e-01 -5.99847615e-01
-5.06979115e-02 1.35108605e-01 1.34696541e-02 -5.39666712e-01
6.80629671e-01 3.03736031e-01 3.27646255e-01 -3.97149891e-01
4.83802855e-01 5.64407289e-01 8.82367730e-01 -7.68028557e-01
1.04425645e+00 9.13871408e-01 5.58013380e-01 -1.03987539e+00
-4.07512560e-02 -7.02318028e-02 -7.55546331e-01 -2.55240500e-01
6.47163272e-01 -3.18247169e-01 -1.04027915e+00 5.69397807e-01
-1.16234970e+00 2.70521909e-01 1.13400631e-02 1.58482805e-01
-4.23328161e-01 9.89332914e-01 -4.19162184e-01 -7.85328448e-01
-2.54680812e-01 -1.17209375e+00 8.85274887e-01 4.03764278e-01
8.06444362e-02 -9.85573053e-01 -1.73903424e-02 5.31316996e-01
-5.38200606e-03 5.08868992e-01 1.26688409e+00 -4.23193425e-01
-4.96176124e-01 -2.29867890e-01 -2.06712455e-01 4.37123865e-01
4.71441090e-01 5.33741534e-01 -1.12307465e+00 -4.08446759e-01
1.44278973e-01 9.68601927e-02 6.61811113e-01 3.04128051e-01
1.25255203e+00 -1.85364589e-01 -1.19195394e-01 9.60663497e-01
1.20350218e+00 3.85393471e-01 8.63573194e-01 1.46927580e-01
6.74210727e-01 9.77663815e-01 1.07072823e-01 3.35507840e-01
4.24776562e-02 7.65313208e-01 4.81159121e-01 1.29684269e-01
-2.58392930e-01 -7.21009597e-02 2.58019626e-01 3.28635812e-01
-9.82766971e-02 -3.01030248e-01 -8.37895751e-01 1.96201861e-01
-1.34739292e+00 -1.13190520e+00 -8.76324922e-02 2.60827565e+00
3.94709915e-01 -2.63608098e-01 -1.77300759e-02 -4.12894562e-02
1.33133829e+00 5.98566793e-03 -5.38638890e-01 -6.20681345e-01
-2.94086039e-01 4.83611941e-01 4.57433164e-01 2.88944632e-01
-7.97814190e-01 6.55753672e-01 6.23783255e+00 6.82715058e-01
-1.35476625e+00 -2.46187195e-01 7.78512359e-01 1.75609648e-01
-3.01697224e-01 3.82719666e-01 -4.39973593e-01 5.31536639e-01
7.50204325e-01 -1.94692418e-01 7.66985297e-01 3.95385951e-01
2.96485815e-02 -1.76659301e-01 -9.90749538e-01 1.14747500e+00
2.61668026e-01 -1.18236518e+00 1.88683495e-01 2.95436829e-01
6.02011442e-01 -3.55845988e-01 4.01719242e-01 -4.88109767e-01
1.82623401e-01 -1.06719232e+00 3.74421179e-01 5.13653934e-01
8.58822703e-01 -5.55549860e-01 3.74255389e-01 -2.13667601e-02
-9.23603058e-01 -9.63124782e-02 -5.69648683e-01 3.03874671e-01
-1.75392091e-01 3.75080138e-01 -1.10073721e+00 7.09773242e-01
7.26435006e-01 2.46309608e-01 -8.58175337e-01 9.15826142e-01
5.79920672e-02 3.01132083e-01 -4.92689639e-01 4.77979869e-01
-1.96458235e-01 -7.44167387e-01 5.63954055e-01 7.58685112e-01
7.26366580e-01 -1.03292745e-02 -4.74265784e-01 1.21776199e+00
-1.69159889e-01 -9.81460288e-02 -6.69507146e-01 4.30619754e-02
7.79326081e-01 1.20634615e+00 -8.08090091e-01 3.21352817e-02
-1.11501582e-01 1.01950753e+00 8.71351808e-02 5.14860809e-01
-7.40084887e-01 -1.65989861e-01 6.70389533e-01 9.36799943e-02
2.97831029e-01 -1.23329751e-01 -6.06605895e-02 -9.56604004e-01
1.41588390e-01 -7.09389627e-01 3.21747005e-01 -1.04372609e+00
-1.33805752e+00 5.69405913e-01 2.69523770e-01 -1.15182865e+00
-1.49171889e-01 -5.50740123e-01 -9.22839701e-01 1.02137148e+00
-1.31025660e+00 -1.17790127e+00 -4.36195642e-01 7.69756854e-01
-1.45803034e-01 -1.88293740e-01 5.50932467e-01 4.68982421e-02
-5.20502627e-01 7.40417838e-01 3.99999708e-01 -1.21939778e-02
6.51590705e-01 -9.20166671e-01 3.44290942e-01 9.97899294e-01
3.03954959e-01 8.41992676e-01 6.01258576e-01 -4.95816320e-01
-1.40301764e+00 -7.61827052e-01 3.76692146e-01 -2.23215505e-01
3.18454117e-01 -1.12538077e-01 -9.78527665e-01 4.73729163e-01
1.31831303e-01 1.34342551e-01 5.51050067e-01 -2.51256406e-01
-6.41911209e-01 -3.66627097e-01 -1.49813497e+00 4.56891835e-01
9.06981051e-01 -7.37846613e-01 -1.65931553e-01 3.09823096e-01
9.60285589e-02 -1.47127837e-01 -9.72424567e-01 3.05275798e-01
6.32572353e-01 -1.44235015e+00 1.14763463e+00 -3.94576699e-01
2.44484156e-01 -3.07893544e-01 -2.82559786e-02 -1.03317094e+00
-2.49513209e-01 -8.22150946e-01 2.12754428e-01 1.50665879e+00
-1.92252070e-01 -8.99420917e-01 6.70063853e-01 6.79608822e-01
2.39728838e-01 -2.99614668e-01 -1.22977901e+00 -8.98382843e-01
5.21990173e-02 -3.18593383e-02 8.39263797e-01 9.08099174e-01
-5.64816236e-01 -1.32086715e-02 -3.36456925e-01 4.46560800e-01
8.85292411e-01 2.07247540e-01 9.18458760e-01 -1.26328528e+00
-3.01250845e-01 -4.66943383e-01 -8.45043898e-01 -2.66697168e-01
3.31709743e-01 -8.55105996e-01 -4.59340543e-01 -8.42684329e-01
1.64590344e-01 -4.41581577e-01 -1.12650104e-01 2.30510607e-01
6.43653125e-02 3.75777304e-01 3.05635750e-01 4.75672424e-01
3.71837854e-01 4.18636560e-01 1.04032171e+00 -1.28450856e-01
8.62331539e-02 -4.76974472e-02 -4.67795521e-01 5.39713740e-01
7.17456162e-01 -4.04306144e-01 -4.70294446e-01 -1.80118486e-01
4.11666594e-02 -2.70620957e-02 7.15750515e-01 -8.66378367e-01
2.77668573e-02 -4.91649918e-02 3.83233666e-01 -7.62774721e-02
4.87709194e-01 -7.81682491e-01 6.26686871e-01 3.63979638e-01
-5.34786135e-02 1.25392586e-01 1.13964617e-01 5.58566809e-01
5.08675235e-04 -4.55871642e-01 9.17146027e-01 4.31317538e-02
-5.66934526e-01 1.26089349e-01 1.84940785e-01 -2.69290805e-01
1.06433022e+00 -6.59479201e-01 -4.27156091e-01 -4.96945888e-01
-6.12090647e-01 -3.66369128e-01 7.33557045e-01 3.59035105e-01
6.48520827e-01 -1.22982836e+00 -6.25561595e-01 4.53769296e-01
3.25088650e-02 -5.54812908e-01 2.36552253e-01 6.77577853e-01
-8.30230534e-01 1.68294425e-03 -5.78038156e-01 -7.27023542e-01
-1.59936678e+00 5.83163679e-01 5.30757606e-01 1.71782240e-01
-2.81843424e-01 5.95474482e-01 3.24482113e-01 -3.27916622e-01
-2.02450678e-01 2.63144588e-03 3.10531501e-02 -1.73160017e-01
9.17193145e-02 5.45832574e-01 4.11027782e-02 -1.08279979e+00
-3.05771440e-01 8.69421482e-01 1.42962188e-01 -3.79524902e-02
9.97875929e-01 -1.86105788e-01 -4.40100402e-01 -1.46895558e-01
1.43794107e+00 -7.68521801e-03 -9.91551518e-01 -1.64323345e-01
-1.49256080e-01 -8.23435903e-01 -1.97980061e-01 -3.96054059e-01
-1.12671590e+00 7.96221972e-01 9.55432057e-01 2.37493500e-01
1.19151568e+00 -3.00425738e-01 4.77950037e-01 2.18799159e-01
2.02181101e-01 -7.67624319e-01 -1.11511517e-02 2.27687936e-02
7.49857485e-01 -1.08095193e+00 2.19258711e-01 -5.20040214e-01
-3.31335545e-01 1.41318619e+00 2.25210175e-01 -3.31227444e-02
5.08479297e-01 -2.47694716e-01 -7.04709068e-02 -2.54515380e-01
-7.01718554e-02 -1.40346512e-02 4.44981009e-01 8.43474925e-01
2.32983202e-01 -1.19881131e-01 -2.31261641e-01 -3.67257327e-01
-4.36096489e-01 -2.59457320e-01 6.89973414e-01 6.07365727e-01
1.41858473e-01 -1.26399195e+00 -8.98045480e-01 2.71097779e-01
-2.89487273e-01 1.00452967e-01 -6.68046713e-01 8.16221356e-01
1.42720073e-01 9.92744386e-01 2.06065312e-01 -4.66492444e-01
-4.75952551e-02 1.05524264e-01 8.19004595e-01 -1.27975017e-01
-4.16961789e-01 -1.96266696e-01 -2.06861928e-01 -2.86136061e-01
-5.35639644e-01 -6.88014030e-01 -9.18769240e-01 -6.94229007e-01
-2.16286331e-01 -1.85794130e-01 6.89644158e-01 7.11486876e-01
5.51464617e-01 -3.86630856e-02 8.90106499e-01 -6.86496854e-01
-6.63501859e-01 -6.13670170e-01 -6.91343963e-01 9.39369678e-01
7.15443939e-02 -6.82591915e-01 -3.62105101e-01 2.49746919e-01] | [12.7443265914917, 0.7849080562591553] |
5dbdffc5-a55f-4a86-8f5d-b98c3527d194 | medical-diffusion-denoising-diffusion | 2211.03364 | null | https://arxiv.org/abs/2211.03364v7 | https://arxiv.org/pdf/2211.03364v7.pdf | Medical Diffusion: Denoising Diffusion Probabilistic Models for 3D Medical Image Generation | Recent advances in computer vision have shown promising results in image generation. Diffusion probabilistic models in particular have generated realistic images from textual input, as demonstrated by DALL-E 2, Imagen and Stable Diffusion. However, their use in medicine, where image data typically comprises three-dimensional volumes, has not been systematically evaluated. Synthetic images may play a crucial role in privacy preserving artificial intelligence and can also be used to augment small datasets. Here we show that diffusion probabilistic models can synthesize high quality medical imaging data, which we show for Magnetic Resonance Images (MRI) and Computed Tomography (CT) images. We provide quantitative measurements of their performance through a reader study with two medical experts who rated the quality of the synthesized images in three categories: Realistic image appearance, anatomical correctness and consistency between slices. Furthermore, we demonstrate that synthetic images can be used in a self-supervised pre-training and improve the performance of breast segmentation models when data is scarce (dice score 0.91 vs. 0.95 without vs. with synthetic data). The code is publicly available on GitHub: https://github.com/FirasGit/medicaldiffusion. | ['Daniel Truhn', 'Jakob Nikolas Kather', 'Sven Nebelung', 'Christiane Kuhl', 'Johannes Stegmaier', 'Sebastian Foersch', 'Bettina Baessler', 'Sandy Engelhardt', 'Philipp Schad', 'Maximilian Schulze-Hagen', 'Christoph Haarburger', 'Tianyu Han', 'Soroosh Tayebi Arasteh', 'Gustav Mueller-Franzes', 'Firas Khader'] | 2022-11-07 | null | null | null | null | ['medical-image-generation'] | ['medical'] | [ 3.64544898e-01 6.72613263e-01 3.62686776e-02 -5.12523413e-01
-1.01959574e+00 -5.89170694e-01 5.77276170e-01 2.42029831e-01
-6.22660339e-01 8.54012311e-01 3.66459996e-01 -4.56400961e-01
-1.34514064e-01 -5.29457748e-01 -6.57667577e-01 -7.39272952e-01
-1.46946888e-02 7.83060908e-01 4.00927365e-02 2.95064181e-01
-3.49728833e-03 4.02306646e-01 -8.07598591e-01 4.61251825e-01
9.47312593e-01 7.56570220e-01 2.73811352e-02 7.68600285e-01
2.07456216e-01 6.00179791e-01 -5.70938230e-01 -7.10735261e-01
3.64284515e-01 -4.09423560e-01 -8.04720104e-01 2.81225085e-01
1.64929956e-01 -5.04765213e-01 -3.49823862e-01 1.19029725e+00
7.23219275e-01 -4.21950251e-01 8.86702240e-01 -1.23955381e+00
-9.57752705e-01 7.62566268e-01 -5.16240418e-01 1.36896163e-01
1.12590484e-01 4.47349817e-01 4.78954822e-01 -5.33381402e-01
1.12867558e+00 1.00405586e+00 5.90482652e-01 8.00752699e-01
-1.49305940e+00 -4.96784478e-01 -5.11595368e-01 3.01028937e-02
-1.22184348e+00 -4.32614386e-01 7.09123790e-01 -6.07718647e-01
2.95109838e-01 3.82518232e-01 6.09182298e-01 1.26783645e+00
3.23192865e-01 8.63619328e-01 1.56718087e+00 -2.63756990e-01
3.40833426e-01 5.14311790e-01 -2.31132761e-01 7.80988514e-01
3.13539654e-01 1.09122328e-01 -1.67522460e-01 -4.58886862e-01
8.41201067e-01 -3.03736866e-01 -4.97199923e-01 -5.03168523e-01
-1.60272467e+00 1.10467041e+00 7.11714804e-01 3.52396369e-01
-5.83889067e-01 7.54936785e-02 1.76610813e-01 -6.07682653e-02
5.50326884e-01 3.95087689e-01 -2.14664708e-03 3.39945108e-01
-1.23611736e+00 1.67947635e-01 6.87754750e-01 7.23012626e-01
-2.24590302e-02 -1.41899064e-01 -2.64804095e-01 7.91666090e-01
2.70906746e-01 6.93393707e-01 7.05079556e-01 -1.30129302e+00
7.76045099e-02 3.35155167e-02 6.87886178e-02 -9.92814064e-01
-3.43546003e-01 -9.64109674e-02 -1.17942953e+00 2.51581967e-01
6.26822770e-01 -1.75224602e-01 -1.05040836e+00 1.45460868e+00
3.49494725e-01 -1.17573507e-01 9.25972685e-02 9.50257897e-01
8.47346067e-01 2.95437664e-01 1.39848024e-01 -1.72904924e-01
1.24959707e+00 -7.29157925e-01 -7.17955530e-01 7.84173049e-03
5.80866396e-01 -7.45256305e-01 7.67297029e-01 3.78919929e-01
-1.46436524e+00 -1.21519528e-01 -9.22949553e-01 1.41540319e-01
-1.28331870e-01 -4.21175137e-02 6.07976317e-01 1.02468657e+00
-1.38576710e+00 6.07956946e-01 -9.93108869e-01 -1.93309233e-01
1.14654624e+00 2.63549656e-01 -5.34888446e-01 -3.59880537e-01
-1.07664323e+00 8.44568908e-01 2.00415149e-01 -1.92979425e-01
-9.28476214e-01 -8.60039413e-01 -7.30040848e-01 -4.90276456e-01
2.72139180e-02 -8.50096345e-01 1.01044261e+00 -1.03609955e+00
-1.07515860e+00 1.28682160e+00 1.51716679e-01 -8.21899116e-01
1.21734536e+00 4.37608033e-01 -1.35161445e-01 6.23027861e-01
2.27404058e-01 1.38389432e+00 7.39049613e-01 -1.43932819e+00
-2.11293295e-01 -3.55039239e-01 -2.99788624e-01 -9.08224785e-04
1.12111373e-02 -3.71564366e-02 -2.31966227e-01 -8.48403394e-01
1.21428393e-01 -1.13997996e+00 -6.76380336e-01 7.10716367e-01
-6.05770588e-01 4.05710191e-01 3.56088519e-01 -1.02055609e+00
5.24135649e-01 -1.94068360e+00 -2.71494359e-01 3.98791403e-01
5.25415540e-01 8.56377929e-02 3.84987183e-02 -2.91901886e-01
-9.17440504e-02 5.12718141e-01 -9.60816860e-01 -1.82361349e-01
-1.69920549e-01 3.39849181e-02 2.51664609e-01 7.17929542e-01
1.04323298e-01 1.18229938e+00 -8.91138494e-01 -8.14409971e-01
1.72591552e-01 7.50154018e-01 -4.19767737e-01 -1.46424279e-01
9.19479728e-02 8.97991419e-01 -2.89103419e-01 5.11999249e-01
8.12746584e-01 -5.23384571e-01 2.41414607e-01 -7.52025843e-02
5.55736363e-01 -3.88693601e-01 -8.31520736e-01 1.57536471e+00
-9.76161957e-02 6.10112309e-01 2.25468297e-02 -6.37599051e-01
5.85541964e-01 4.68770653e-01 7.87035167e-01 -6.69957817e-01
1.21597804e-01 2.05212548e-01 2.27037370e-01 -4.48379129e-01
1.64582357e-01 -3.58738095e-01 2.89584368e-01 7.84941196e-01
-3.35061014e-01 -6.83612823e-01 6.96015581e-02 4.59879220e-01
9.50415671e-01 -2.16215357e-01 5.56612387e-02 -3.08830708e-01
1.80770621e-01 3.41981739e-01 1.00382008e-01 7.69138992e-01
-5.73153973e-01 1.21383631e+00 4.66082364e-01 -1.31822184e-01
-1.36495018e+00 -1.22929358e+00 -3.72344434e-01 1.22670449e-01
-1.39080629e-01 1.62148416e-01 -1.15673673e+00 -8.11489940e-01
-1.41610548e-01 8.47284377e-01 -8.88612449e-01 2.33654324e-02
-2.22029150e-01 -1.08087742e+00 6.62423849e-01 3.40662092e-01
2.30297938e-01 -1.15087712e+00 -7.54426479e-01 2.79191341e-02
-3.47949326e-01 -1.12313068e+00 -5.78430235e-01 -2.86336243e-01
-9.77416217e-01 -1.02567744e+00 -1.57043302e+00 -5.12167037e-01
1.04952812e+00 -7.17422664e-02 1.13041270e+00 1.39480144e-01
-6.80252314e-01 5.80977321e-01 -1.08323641e-01 -4.73455399e-01
-9.94396389e-01 -2.95371622e-01 -7.92644694e-02 -1.03261203e-01
-1.75266847e-01 -4.04416651e-01 -1.16575289e+00 3.63561660e-01
-1.31618297e+00 2.96705842e-01 5.93566239e-01 8.61523628e-01
8.40453446e-01 -3.18473160e-01 2.16236144e-01 -1.31973422e+00
8.65214229e-01 -5.37697732e-01 -4.11278814e-01 1.35051697e-01
-8.37788701e-01 -7.91754425e-02 5.36982417e-02 -4.41987127e-01
-9.14418042e-01 1.63326368e-01 -1.55168369e-01 -3.96162003e-01
-2.46014997e-01 3.33580375e-01 5.11619091e-01 -1.77770600e-01
1.01903546e+00 2.25903660e-01 3.62400055e-01 -3.87089364e-02
4.44231838e-01 6.37124121e-01 3.65458667e-01 -2.79611915e-01
4.99300301e-01 9.54726338e-01 -4.47832234e-02 -5.58631659e-01
-1.94005385e-01 9.12515521e-02 -7.54113674e-01 -1.14712536e-01
9.73069072e-01 -5.80315650e-01 -3.00743163e-01 3.01930547e-01
-9.86393392e-01 -4.23748761e-01 -4.37492758e-01 5.31999111e-01
-6.48382664e-01 4.35849696e-01 -5.79010546e-01 -5.02986073e-01
-4.80873376e-01 -1.60213566e+00 7.57353723e-01 6.32286817e-02
-4.10178214e-01 -1.20740163e+00 -2.01430514e-01 6.49240136e-01
5.51526427e-01 5.91111481e-01 8.77152920e-01 -5.88228166e-01
-3.75005871e-01 -1.83190554e-01 -2.96787351e-01 4.68206495e-01
2.58738585e-02 -1.27617672e-01 -9.45106626e-01 -1.39581323e-01
6.26080185e-02 -3.79581809e-01 5.93851149e-01 9.58619058e-01
1.25069904e+00 -3.03380072e-01 -4.61695969e-01 5.41122615e-01
1.14944577e+00 2.05059275e-01 7.85405278e-01 5.15537262e-02
4.83476102e-01 9.16027248e-01 3.29955935e-01 1.90379024e-01
2.50148416e-01 2.69709319e-01 1.49681702e-01 -4.09917235e-01
-4.12652761e-01 -1.00590824e-03 -1.49476111e-01 4.43977624e-01
1.78302333e-01 -7.75865912e-02 -1.21072578e+00 7.00294375e-01
-1.42041874e+00 -5.78795910e-01 -3.42221409e-01 2.01296496e+00
1.02550697e+00 5.19493632e-02 -1.32745758e-01 -4.89419103e-02
7.25200713e-01 -5.20270839e-02 -7.12685347e-01 -1.56025931e-01
-2.71782011e-01 6.71003684e-02 7.43701041e-01 3.94178808e-01
-1.02812254e+00 5.82711279e-01 6.66112041e+00 6.98394716e-01
-1.06022346e+00 4.54310715e-01 1.52353156e+00 -5.45570031e-02
-6.02042615e-01 -3.09658319e-01 9.18270871e-02 5.00131845e-01
9.11166847e-01 -3.02840292e-01 9.00988840e-03 6.82500541e-01
2.67005593e-01 -3.22403163e-01 -9.45197821e-01 1.06462812e+00
4.16466147e-02 -1.71714902e+00 8.16869959e-02 2.83645689e-01
1.06093979e+00 1.12024903e-01 5.70487380e-01 -3.93420011e-01
4.20544058e-01 -1.31780922e+00 6.77391768e-01 6.13272309e-01
9.66021538e-01 -4.59663898e-01 7.99994767e-01 1.55330166e-01
-2.65147924e-01 5.77246368e-01 -2.33623609e-01 8.12169969e-01
2.36260921e-01 7.27921247e-01 -1.27473891e+00 4.02883232e-01
6.59507215e-01 2.52273321e-01 -6.81693912e-01 1.12715507e+00
-5.72694167e-02 5.55241466e-01 -3.17408472e-01 2.48521388e-01
1.76161617e-01 -1.14443570e-01 4.86635476e-01 9.97457206e-01
2.12346017e-01 1.97936058e-01 -2.38665640e-01 1.07162023e+00
-1.34074211e-01 2.64857143e-01 -5.52730620e-01 -3.18736546e-02
4.64526527e-02 1.17530799e+00 -1.14139152e+00 -3.88930500e-01
-2.80105006e-02 1.07216632e+00 -2.49104336e-01 2.20153362e-01
-8.41128647e-01 1.22300372e-01 1.27319968e-03 3.07183146e-01
2.20432840e-02 5.87682128e-02 -5.92376053e-01 -7.78961360e-01
-1.29628628e-01 -9.65902209e-01 3.47269297e-01 -1.12185907e+00
-1.35105658e+00 9.40341473e-01 5.77458702e-02 -9.68084872e-01
-2.15044767e-01 -4.12579805e-01 -9.09481868e-02 9.09311652e-01
-1.17100525e+00 -1.00977850e+00 -1.81075066e-01 5.64076126e-01
2.97351718e-01 -9.40948538e-03 8.07583094e-01 8.20165947e-02
-1.20933048e-01 6.81526899e-01 3.52088720e-01 1.89866990e-01
6.67474926e-01 -1.23290980e+00 3.23929846e-01 5.29537976e-01
1.52310416e-01 3.48803610e-01 6.20189548e-01 -7.91044474e-01
-8.56612325e-01 -1.00380528e+00 4.73038018e-01 -5.17447352e-01
4.39061493e-01 1.02506414e-01 -8.12735796e-01 4.69465852e-01
3.21619302e-01 2.55618423e-01 8.16067159e-01 -8.34949911e-01
-6.30192608e-02 2.94477493e-01 -1.93030596e+00 6.87770784e-01
6.01644039e-01 -1.45237505e-01 -1.67895943e-01 7.60335743e-01
4.04197901e-01 -4.99981523e-01 -1.14683926e+00 1.92450106e-01
4.19756949e-01 -1.01216757e+00 9.08764660e-01 -3.01038474e-01
7.43794918e-01 1.22527041e-01 2.19986755e-02 -1.23753417e+00
7.75102898e-02 -6.19966924e-01 3.51980746e-01 7.64043927e-01
7.74891615e-01 -5.85548520e-01 9.63817298e-01 1.11588430e+00
4.89361405e-01 -6.62313163e-01 -8.67171586e-01 -3.74833822e-01
3.58203739e-01 -4.54033315e-01 3.39717746e-01 1.03999341e+00
-3.73685777e-01 -2.79239774e-01 -1.18521541e-01 -7.93875009e-02
9.63338733e-01 -3.43293995e-01 3.61004263e-01 -9.16412234e-01
-1.39579087e-01 -5.66742063e-01 -3.67104888e-01 -3.89952570e-01
-1.32820100e-01 -1.37498426e+00 -1.07689694e-01 -1.69384050e+00
4.32179242e-01 -6.97779238e-01 -1.47517934e-01 2.78359622e-01
-1.22680530e-01 6.23905182e-01 1.70380190e-01 5.10675073e-01
-5.37428521e-02 2.87423879e-01 1.61153245e+00 -4.27249402e-01
1.74719647e-01 -1.35975694e-02 -8.39467347e-01 6.23768270e-01
9.76076961e-01 -8.77451122e-01 -5.01021266e-01 -3.88104945e-01
-3.38105708e-01 2.18181595e-01 4.13022518e-01 -8.35450113e-01
1.11994810e-01 1.08699828e-01 6.78216636e-01 -2.24664420e-01
2.80120730e-01 -8.13331604e-01 4.06969160e-01 9.30804610e-01
-4.80459899e-01 1.54254213e-02 1.42452076e-01 4.90964472e-01
2.19649971e-02 -3.11376601e-01 1.00817180e+00 -4.82031316e-01
-1.37431204e-01 4.65207815e-01 -2.37659588e-01 1.31430075e-01
1.12713885e+00 -1.69723198e-01 -7.44261965e-03 -7.55356491e-01
-1.02992547e+00 -3.01406868e-02 5.24699092e-01 6.44522682e-02
9.57702398e-01 -1.14944983e+00 -1.19420934e+00 -4.27245833e-02
-8.35875198e-02 -1.01624556e-01 4.13219988e-01 1.15710342e+00
-9.79366899e-01 1.62288889e-01 -3.36817414e-01 -8.89116347e-01
-1.19644642e+00 5.08353591e-01 3.14292431e-01 -2.99521267e-01
-6.46004796e-01 9.65097010e-01 2.97609717e-01 -3.28585684e-01
1.07669219e-01 -1.47140041e-01 1.24859259e-01 -8.20035636e-02
4.84018564e-01 2.68067350e-04 -1.51758626e-01 -7.44885087e-01
-3.08223397e-01 1.79139078e-01 -2.41721407e-01 -6.84633374e-01
1.33126330e+00 -1.02199301e-01 -8.12544897e-02 8.76126513e-02
1.09254897e+00 -1.90031275e-01 -1.15185058e+00 -7.17606321e-02
-8.42810646e-02 -5.45537829e-01 1.75532997e-01 -1.00295615e+00
-1.42948461e+00 8.20377946e-01 1.04318762e+00 7.25529566e-02
8.92461181e-01 9.79524255e-02 6.56145692e-01 -1.12250537e-01
3.07988822e-01 -6.86875343e-01 -7.43029639e-02 -2.92806923e-01
1.10829818e+00 -1.50880075e+00 2.18789831e-01 -3.96140426e-01
-1.25498378e+00 7.73068309e-01 -2.91819777e-02 7.47502670e-02
8.10575724e-01 3.19416463e-01 3.66199195e-01 -1.27991036e-01
-2.84873068e-01 2.17697263e-01 3.54767561e-01 1.09097421e+00
3.55687350e-01 3.43458265e-01 -3.44349116e-01 3.13336909e-01
-2.34527588e-01 1.18867502e-01 6.61842585e-01 8.30199480e-01
1.65440366e-01 -1.00823820e+00 -5.46925545e-01 5.97950161e-01
-7.05787539e-01 -2.18599126e-01 -3.76736283e-01 6.58015311e-01
-1.76384911e-01 6.82372808e-01 -1.78368554e-01 3.74719091e-02
4.25984077e-02 -9.73190144e-02 5.67609966e-01 -3.40201795e-01
-5.59022844e-01 -1.80071034e-02 5.07137319e-03 -4.25383449e-01
-4.24874902e-01 -7.48988569e-01 -1.08708847e+00 -3.02235305e-01
3.70867886e-02 1.13149643e-01 1.01833773e+00 5.06341755e-01
2.60146230e-01 4.14064884e-01 3.30231905e-01 -6.80690587e-01
-3.56802791e-01 -8.82930577e-01 -4.91854340e-01 5.70598006e-01
1.35571882e-01 -2.39941597e-01 -1.14785880e-01 4.95047152e-01] | [14.291831970214844, -1.9556490182876587] |
71c09a32-091f-4884-b98d-ac81bbd00d94 | sess-saliency-enhancing-with-scaling-and | 2207.01769 | null | https://arxiv.org/abs/2207.01769v1 | https://arxiv.org/pdf/2207.01769v1.pdf | SESS: Saliency Enhancing with Scaling and Sliding | High-quality saliency maps are essential in several machine learning application areas including explainable AI and weakly supervised object detection and segmentation. Many techniques have been developed to generate better saliency using neural networks. However, they are often limited to specific saliency visualisation methods or saliency issues. We propose a novel saliency enhancing approach called SESS (Saliency Enhancing with Scaling and Sliding). It is a method and model agnostic extension to existing saliency map generation methods. With SESS, existing saliency approaches become robust to scale variance, multiple occurrences of target objects, presence of distractors and generate less noisy and more discriminative saliency maps. SESS improves saliency by fusing saliency maps extracted from multiple patches at different scales from different areas, and combines these individual maps using a novel fusion scheme that incorporates channel-wise weights and spatial weighted average. To improve efficiency, we introduce a pre-filtering step that can exclude uninformative saliency maps to improve efficiency while still enhancing overall results. We evaluate SESS on object recognition and detection benchmarks where it achieves significant improvement. The code is released publicly to enable researchers to verify performance and further development. Code is available at: https://github.com/neouyghur/SESS | ['Clinton Fookes', 'Sridha Sridharan', 'Simon Denman', 'Osman Tursun'] | 2022-07-05 | null | null | null | null | ['weakly-supervised-object-detection'] | ['computer-vision'] | [ 4.81802434e-01 1.38816284e-02 -1.75517216e-01 -3.67571563e-01
-5.88632584e-01 -2.97905564e-01 4.45809335e-01 2.06375584e-01
-2.77340084e-01 6.13412917e-01 2.57567257e-01 1.64128821e-02
7.06090778e-02 -6.10262632e-01 -6.63895249e-01 -4.76233929e-01
9.62471366e-02 -1.47746220e-01 1.21499336e+00 -3.51610720e-01
6.13983631e-01 1.92978367e-01 -1.85732353e+00 3.25153530e-01
1.29959071e+00 8.69074881e-01 7.49207616e-01 4.60677594e-01
3.53598148e-02 5.52411675e-01 -5.48351169e-01 -3.09577417e-02
2.98452705e-01 -3.62113684e-01 -7.27930427e-01 -1.40888050e-01
3.15001279e-01 2.70628054e-02 1.56978816e-01 1.25060534e+00
5.12311459e-01 8.24283063e-02 4.12715316e-01 -1.53172565e+00
-8.28954577e-01 5.39799869e-01 -8.38385940e-01 6.30169690e-01
1.20152228e-01 6.42501637e-02 8.91692102e-01 -1.05661809e+00
4.91349459e-01 1.14582014e+00 5.50803959e-01 3.46834600e-01
-1.15741909e+00 -6.39310181e-01 1.03140026e-01 3.99620354e-01
-1.31044543e+00 -2.98254818e-01 9.37494397e-01 -5.23591861e-02
5.87971210e-01 5.44843674e-01 5.77519655e-01 7.28164136e-01
6.86945990e-02 1.30153406e+00 1.26025093e+00 -4.41802412e-01
2.78565317e-01 2.80704945e-01 -3.41750234e-02 3.97326976e-01
2.73665130e-01 1.48914764e-02 -6.46748185e-01 1.08273871e-01
8.11328650e-01 7.43122622e-02 -1.55448228e-01 -5.33100486e-01
-1.48699772e+00 8.29546988e-01 1.04006231e+00 3.26474816e-01
-4.64417130e-01 -1.25655234e-01 8.82304385e-02 -1.61908701e-01
4.73465055e-01 7.77806103e-01 -3.51689309e-01 1.23398498e-01
-1.27130163e+00 4.36990708e-01 2.39708483e-01 9.42617476e-01
8.93185496e-01 1.65383577e-01 -4.22145098e-01 9.05229211e-01
1.39063731e-01 5.46270013e-01 6.79052711e-01 -8.32029641e-01
-1.07833939e-02 8.37924838e-01 7.69013166e-02 -1.26853585e+00
-4.34043050e-01 -5.34332395e-01 -4.00241971e-01 3.68722707e-01
-1.25082638e-02 7.06178769e-02 -1.25089550e+00 1.46923709e+00
3.99572700e-01 4.25066710e-01 -2.05268174e-01 1.33433080e+00
1.20261824e+00 4.33240920e-01 2.36826196e-01 2.76953697e-01
1.31937575e+00 -1.27776456e+00 -5.27471125e-01 -5.47524452e-01
1.26721650e-01 -9.34269249e-01 1.30442894e+00 -3.50436941e-02
-1.15007389e+00 -3.99158925e-01 -1.19014287e+00 -8.63791183e-02
-7.15227067e-01 1.94242954e-01 6.58429205e-01 3.66986036e-01
-1.16474712e+00 4.43074256e-01 -6.94783628e-01 -2.76288927e-01
7.91399479e-01 3.94980043e-01 1.34999538e-02 3.00082207e-01
-1.26537025e+00 1.15062726e+00 5.59228539e-01 -2.59124368e-01
-6.66308582e-01 -7.01254189e-01 -1.03562176e+00 3.67545374e-02
5.16951621e-01 -3.15703571e-01 1.24312806e+00 -1.34168947e+00
-1.08640444e+00 5.89687049e-01 -4.30829108e-01 -5.78254223e-01
2.28083059e-01 -2.20454410e-01 -2.78940260e-01 1.35206640e-01
5.03840148e-01 1.28704464e+00 1.02148592e+00 -1.16851187e+00
-8.55583847e-01 -3.76630574e-02 -5.13511635e-02 4.14020479e-01
-9.10945162e-02 2.96862960e-01 -3.08149010e-01 -8.95273149e-01
2.66275883e-01 -7.94049740e-01 -2.73707986e-01 -1.75005823e-01
-5.18607736e-01 -9.70953181e-02 1.08926010e+00 -4.58066761e-01
1.00826550e+00 -2.01795650e+00 5.35294525e-02 3.11947912e-02
2.92645872e-01 5.84475517e-01 -2.85690248e-01 -7.63575882e-02
-2.38864552e-02 6.00194409e-02 -4.57955182e-01 -2.19002496e-02
-2.22922906e-01 -3.35672230e-01 -1.06195830e-01 1.21763006e-01
7.38843977e-01 1.20631611e+00 -1.06373215e+00 -6.32822871e-01
4.68583137e-01 4.31284040e-01 -3.20559323e-01 -1.60664886e-01
-1.72560304e-01 1.58866331e-01 -4.55417126e-01 1.07360029e+00
6.81639850e-01 -3.42257172e-01 -5.37454486e-01 -2.05927268e-01
-1.99585259e-01 2.86937654e-01 -1.41921616e+00 1.52499008e+00
-1.36742322e-02 6.92577243e-01 -2.84110397e-01 -7.22474158e-01
9.65719402e-01 -1.37311012e-01 1.36366025e-01 -8.84793639e-01
2.29021549e-01 3.40544581e-01 8.13553110e-02 -9.91743505e-02
8.99335206e-01 7.33958483e-02 1.17011264e-01 2.83545285e-01
6.12698570e-02 -3.64334702e-01 2.32952505e-01 4.27999258e-01
7.64163733e-01 2.18025580e-01 3.98907870e-01 -5.42018294e-01
2.59280413e-01 2.58521199e-01 5.73519826e-01 6.82090640e-01
-4.68364120e-01 1.03673995e+00 8.41186121e-02 -1.40134022e-01
-9.33586419e-01 -1.06986666e+00 -1.00719310e-01 1.31547165e+00
8.04433644e-01 -1.97703928e-01 -8.06946695e-01 -5.94730198e-01
-2.07315683e-02 6.48215830e-01 -7.46147633e-01 -2.58781672e-01
-2.88695186e-01 -7.31540501e-01 1.02688491e-01 6.05935991e-01
7.02933192e-01 -1.61998153e+00 -1.10746849e+00 1.08524032e-01
-1.29253343e-01 -6.93732023e-01 -5.89097679e-01 1.35367200e-01
-7.29825497e-01 -9.45600152e-01 -1.16791487e+00 -9.67750788e-01
7.48496532e-01 7.62527645e-01 1.00047696e+00 2.07444310e-01
-4.02214766e-01 -7.29577988e-02 -4.92203295e-01 -9.51224089e-01
4.61800722e-04 3.17372799e-01 -1.63763925e-01 -3.08003068e-01
3.50611776e-01 -2.32726932e-01 -8.30504417e-01 4.97072488e-01
-9.29892957e-01 5.23121357e-01 7.56039739e-01 7.73877800e-01
6.02391064e-01 -2.46701866e-01 8.41119647e-01 -6.23680651e-01
6.47855878e-01 -4.75973755e-01 -4.82569337e-01 5.30729331e-02
-4.56062436e-01 -2.04464663e-02 1.02641270e-01 -4.30279881e-01
-9.15664971e-01 4.82844338e-02 1.83126703e-01 -1.99289501e-01
-2.38600045e-01 2.88881391e-01 1.07599728e-01 -3.27131510e-01
7.69744337e-01 2.83414245e-01 -3.62656340e-02 -2.76930243e-01
2.87421197e-01 4.70474422e-01 5.12563705e-01 -6.29141880e-03
7.49453187e-01 3.38329643e-01 -3.42340142e-01 -5.15579522e-01
-8.78176451e-01 -4.16727513e-01 -5.01889527e-01 -2.64224023e-01
4.84187961e-01 -7.12486744e-01 -1.59352392e-01 5.22316754e-01
-7.72118628e-01 -2.92805821e-01 -2.36251071e-01 2.68891990e-01
-2.77246565e-01 -4.15780842e-02 -2.55264461e-01 -6.05711043e-01
-4.63511318e-01 -1.23118246e+00 1.19713259e+00 9.85277236e-01
-2.89194107e-01 -7.30079591e-01 -2.81698942e-01 4.89871651e-02
8.97837877e-01 3.19819421e-01 3.21457773e-01 -5.44740796e-01
-5.61596811e-01 2.44184453e-02 -5.56790590e-01 1.75568193e-01
3.97722036e-01 7.42382044e-03 -8.36834192e-01 -1.11252949e-01
-3.32376420e-01 -1.36033505e-01 1.05994773e+00 6.75135911e-01
9.37363863e-01 -2.40838185e-01 -4.53230560e-01 2.34957308e-01
1.24844813e+00 2.04414904e-01 5.21500409e-01 7.53150284e-01
6.99101031e-01 5.95812619e-01 8.91178071e-01 1.11642741e-01
4.11884785e-01 6.71877921e-01 5.07787704e-01 -6.33800745e-01
-3.33003700e-01 -3.04313861e-02 1.32347450e-01 4.24136549e-01
8.38096291e-02 2.74850845e-01 -9.45817649e-01 1.12194574e+00
-1.83175302e+00 -8.17534983e-01 -1.83824778e-01 2.01049399e+00
9.05094922e-01 3.24182004e-01 4.16492701e-01 1.76774293e-01
9.73399639e-01 6.20784387e-02 -8.00506234e-01 -2.29268655e-01
-4.52013910e-01 3.26319546e-01 5.54425299e-01 3.75901133e-01
-1.28232563e+00 1.32193196e+00 6.19011688e+00 8.91454697e-01
-1.41722775e+00 1.68333501e-01 6.28407478e-01 -2.31356129e-01
-3.89095604e-01 1.74649302e-02 -5.86119592e-01 6.54069424e-01
3.15766037e-01 -4.21422958e-01 1.12102583e-01 9.70167220e-01
1.93994150e-01 -6.31441951e-01 -4.03169394e-01 5.43712199e-01
1.39952987e-01 -1.47024131e+00 -2.19807208e-01 -4.32767063e-01
1.02968550e+00 2.44135618e-01 3.58637691e-01 2.34425999e-02
3.13165456e-01 -9.76585746e-01 9.86463904e-01 2.81262368e-01
2.41057605e-01 -6.75946772e-01 9.23670173e-01 4.97839637e-02
-1.09742105e+00 4.30266075e-02 -2.22101539e-01 -4.38735895e-02
7.10635781e-02 5.11057615e-01 -9.80631232e-01 2.76692420e-01
9.88863885e-01 7.61686087e-01 -1.22203243e+00 1.58511651e+00
-2.24614292e-01 5.13080835e-01 -4.21822309e-01 -2.89875388e-01
3.68435264e-01 1.44350946e-01 8.20699096e-01 1.16709387e+00
2.29217619e-01 -2.01874420e-01 -1.00356646e-01 1.13506806e+00
3.41182739e-01 1.52394339e-01 -4.08086300e-01 3.02253783e-01
6.21391833e-01 1.38385189e+00 -1.27317417e+00 -4.01943237e-01
-2.04858974e-01 8.30934346e-01 -1.44529762e-02 2.67550230e-01
-8.90474260e-01 -6.45877659e-01 3.00681353e-01 1.79830447e-01
4.68021005e-01 1.10672034e-01 -7.58642077e-01 -8.26107442e-01
-1.56114073e-02 -7.32819080e-01 1.63994536e-01 -1.01421440e+00
-7.39340246e-01 6.01886630e-01 1.77696154e-01 -1.22020411e+00
3.33597064e-02 -1.61657616e-01 -7.58869350e-01 7.79723287e-01
-1.64212501e+00 -1.19286120e+00 -4.28785801e-01 3.86123955e-01
6.79727256e-01 -4.59207557e-02 3.21383417e-01 9.76111963e-02
-3.73817772e-01 3.47844094e-01 -2.70039797e-01 -2.23402128e-01
6.29112780e-01 -1.35675013e+00 5.37048995e-01 1.09560215e+00
1.05481908e-01 5.37687898e-01 9.90589976e-01 -7.03980505e-01
-6.81167781e-01 -1.13879192e+00 7.15267599e-01 -3.67141008e-01
4.16847736e-01 -2.78843015e-01 -1.17226315e+00 3.42329651e-01
2.46215105e-01 -1.47779034e-02 1.63198143e-01 -1.67075872e-01
9.06508416e-02 3.52272503e-02 -1.19467652e+00 8.50493610e-01
7.65600264e-01 -1.54120103e-01 -7.31082439e-01 -5.78995869e-02
7.02161133e-01 -4.92383301e-01 -2.43525311e-01 6.82835937e-01
2.63698339e-01 -1.07712340e+00 8.97756279e-01 -8.33664387e-02
4.96559530e-01 -8.88057709e-01 2.65152812e-01 -1.36044610e+00
-4.32979524e-01 -3.44008178e-01 1.38577431e-01 1.10146332e+00
5.15452743e-01 -5.55511355e-01 7.76927054e-01 3.85490447e-01
-2.37266779e-01 -7.44730830e-01 -7.37487435e-01 -6.51886284e-01
-3.45829546e-01 -1.28535524e-01 7.52863586e-01 7.76109219e-01
-5.82937570e-03 1.56997398e-01 -3.56094956e-01 6.48360401e-02
4.75414127e-01 2.82445639e-01 4.97554272e-01 -1.02500784e+00
2.47519478e-01 -7.53500402e-01 -5.63583314e-01 -6.07865810e-01
-2.55917072e-01 -7.82950938e-01 3.55738431e-01 -1.60518897e+00
3.86946440e-01 -2.65866518e-01 -5.45367718e-01 7.80504882e-01
-6.55803442e-01 8.17376077e-01 3.60738963e-01 1.80393711e-01
-9.18819547e-01 5.61128318e-01 1.44021773e+00 -8.03058445e-02
-3.28624994e-01 -8.07033181e-02 -1.05906379e+00 7.27409422e-01
1.12976956e+00 -5.07803559e-01 -3.45534772e-01 -2.12334991e-01
-3.00793231e-01 -5.83310485e-01 7.36849427e-01 -1.25148463e+00
1.58742726e-01 -3.24275434e-01 4.27347898e-01 -6.76078856e-01
1.15356803e-01 -3.52348447e-01 -5.20562939e-02 4.33330923e-01
-3.00964326e-01 -1.37052655e-01 5.23180902e-01 1.53000936e-01
-3.47389072e-01 -1.74200460e-01 9.71282601e-01 -6.54699877e-02
-1.19348037e+00 3.61540504e-02 -1.48478419e-01 -1.00462496e-01
1.18792701e+00 -4.51429456e-01 -3.08929294e-01 -3.57478648e-01
-3.71364176e-01 3.14991057e-01 6.74558938e-01 7.96381772e-01
7.76942432e-01 -1.43815649e+00 -6.60492301e-01 1.33032829e-01
2.16726348e-01 3.78036126e-02 1.26424864e-01 9.02822971e-01
-2.67878473e-01 3.57364714e-01 -5.38526118e-01 -6.29815817e-01
-1.43532240e+00 4.05878872e-01 1.10389598e-01 3.05342257e-01
-4.09639686e-01 1.02848315e+00 2.86360264e-01 -1.62105784e-01
3.12095508e-02 -4.57957923e-01 -4.57885563e-01 -1.33346468e-01
6.60787880e-01 3.06571990e-01 -7.44332150e-02 -8.88163090e-01
-5.51231265e-01 3.27613592e-01 -1.98369741e-01 3.23919281e-02
1.23568368e+00 -1.18989952e-01 1.78617537e-01 4.19251442e-01
6.98802054e-01 -1.65263489e-01 -1.49163318e+00 -3.47552210e-01
2.15765327e-01 -5.58207750e-01 1.37983903e-01 -1.00297403e+00
-1.16629195e+00 6.28672063e-01 8.44379425e-01 3.80229831e-01
1.30168879e+00 2.26804703e-01 7.21573710e-01 -3.42508912e-01
1.74585864e-01 -1.13059926e+00 1.97786510e-01 3.73452842e-01
1.07963920e+00 -1.57959294e+00 9.22123790e-02 -4.76301759e-01
-1.03175354e+00 5.26321828e-01 8.12292933e-01 -3.86250407e-01
4.58965838e-01 2.79218644e-01 1.39618888e-01 -1.43486857e-01
-4.24024612e-01 -5.99238753e-01 7.75858462e-01 6.73211813e-01
2.85282940e-01 1.01266526e-01 -4.27411646e-01 5.10240674e-01
-1.88959241e-01 7.55995139e-02 3.04509848e-01 1.07607377e+00
-6.77341580e-01 -8.27959776e-01 -4.46904302e-01 6.87015474e-01
-3.71257871e-01 -3.88013989e-01 -4.27202612e-01 5.65781474e-01
1.99961290e-01 8.16227376e-01 1.26313581e-03 -1.98621869e-01
1.84748024e-01 -3.19978088e-01 5.60070165e-02 -6.75433815e-01
-4.55105901e-01 1.03755981e-01 -3.35434645e-01 -5.44552803e-01
-6.82165623e-01 -7.20537841e-01 -1.33367074e+00 3.16592678e-02
-7.12630093e-01 2.85880957e-02 6.13526821e-01 8.09508979e-01
6.51917100e-01 6.48168027e-01 3.61606717e-01 -1.24990046e+00
4.07850929e-02 -1.11197627e+00 -4.07470763e-01 4.53695685e-01
3.75590175e-01 -1.14239657e+00 -2.05555812e-01 9.87198874e-02] | [9.807555198669434, -0.35700753331184387] |
07b3bd00-cc1d-4386-98ac-a9e93fe9ee88 | an-alternative-to-wsss-an-empirical-study-of | 2305.01586 | null | https://arxiv.org/abs/2305.01586v2 | https://arxiv.org/pdf/2305.01586v2.pdf | An Alternative to WSSS? An Empirical Study of the Segment Anything Model (SAM) on Weakly-Supervised Semantic Segmentation Problems | The Segment Anything Model (SAM) has demonstrated exceptional performance and versatility, making it a promising tool for various related tasks. In this report, we explore the application of SAM in Weakly-Supervised Semantic Segmentation (WSSS). Particularly, we adapt SAM as the pseudo-label generation pipeline given only the image-level class labels. While we observed impressive results in most cases, we also identify certain limitations. Our study includes performance evaluations on PASCAL VOC and MS-COCO, where we achieved remarkable improvements over the latest state-of-the-art methods on both datasets. We anticipate that this report encourages further explorations of adopting SAM in WSSS, as well as wider real-world applications. | ['Nick Barnes', 'Yiran Zhong', 'Yanhao Zhang', 'Zheyuan Liu', 'Weixuan Sun'] | 2023-05-02 | null | null | null | null | ['pseudo-label'] | ['miscellaneous'] | [ 5.42477369e-01 2.25776464e-01 -3.47896725e-01 -7.49768853e-01
-1.01142550e+00 -7.06799388e-01 6.08923912e-01 -4.53694537e-02
-6.93822563e-01 5.84073365e-01 -1.46159008e-01 -1.54605865e-01
3.79215330e-01 -2.68848300e-01 -6.46999657e-01 -5.29171646e-01
1.31752044e-01 5.75138748e-01 8.45214009e-01 1.08423904e-01
2.35930577e-01 3.18133175e-01 -1.52400041e+00 4.44593489e-01
7.38665521e-01 8.63201618e-01 2.85360485e-01 4.64847833e-01
-2.83471495e-01 4.73821849e-01 -7.17227817e-01 -5.10456324e-01
1.63482413e-01 -2.97051668e-01 -1.32089937e+00 3.16968083e-01
6.78361595e-01 1.82889879e-01 2.54963547e-01 9.58165050e-01
3.30739141e-01 1.70464218e-01 5.20308793e-01 -1.17772079e+00
-2.96229362e-01 4.99503344e-01 -5.85785866e-01 -1.00527192e-02
2.71650463e-01 3.49320248e-02 1.05331945e+00 -6.79578066e-01
1.03459358e+00 1.06058311e+00 5.98438263e-01 7.91648567e-01
-1.12363815e+00 -4.76307660e-01 4.20254290e-01 1.21050015e-01
-1.20446980e+00 -3.61121148e-01 5.48496485e-01 -3.57720941e-01
8.36680651e-01 3.89551252e-01 5.31043589e-01 1.06079078e+00
-4.54163611e-01 1.72254348e+00 1.60321069e+00 -5.67148626e-01
3.17396045e-01 1.81366101e-01 2.43482009e-01 7.16306508e-01
-1.06506892e-01 -2.62749076e-01 -4.24760520e-01 -3.16371545e-02
3.34609151e-01 -5.89505613e-01 1.65876858e-02 -2.83233345e-01
-1.48357451e+00 7.33131170e-01 4.85355049e-01 3.53181720e-01
5.49243242e-02 8.79355669e-02 3.73698801e-01 -2.24180296e-01
7.03387499e-01 6.77876890e-01 -5.71459472e-01 -4.68589216e-02
-1.42741919e+00 3.19694847e-01 6.66312754e-01 1.32997143e+00
4.26366657e-01 -1.74458802e-01 -2.36253157e-01 1.05890954e+00
1.49553537e-01 7.51044080e-02 2.21590236e-01 -1.24751985e+00
2.43139401e-01 4.03715611e-01 4.93207760e-03 -2.99210548e-01
-5.24083018e-01 -6.27153695e-01 -1.81110770e-01 -3.47088128e-02
3.42860699e-01 1.42183259e-01 -1.38959646e+00 1.55448210e+00
3.98044407e-01 2.37066641e-01 -5.73680662e-02 9.46753740e-01
9.95998144e-01 3.90400529e-01 5.60211658e-01 -1.00579401e-02
1.24667621e+00 -1.58229601e+00 -6.37809992e-01 -4.18775439e-01
5.45390069e-01 -9.01000679e-01 1.15711749e+00 4.90799278e-01
-1.03858614e+00 -5.84483147e-01 -9.59218562e-01 -1.65412035e-02
-5.31072438e-01 3.90031219e-01 9.67826426e-01 4.79778230e-01
-1.10884988e+00 6.15527391e-01 -9.95986640e-01 -7.74269402e-01
6.85517669e-01 2.22325146e-01 -6.41799420e-02 5.33963591e-02
-9.26730812e-01 6.83973014e-01 5.22859752e-01 6.56042993e-02
-7.24426031e-01 -3.98112655e-01 -7.49211311e-01 -4.91019130e-01
7.11539686e-01 -4.77283001e-01 1.58609009e+00 -8.81345093e-01
-1.51167464e+00 1.50358438e+00 -2.39523143e-01 -4.74226356e-01
6.73196018e-01 -2.45460719e-01 -5.01125515e-01 2.40564615e-01
3.74735504e-01 1.34812593e+00 4.86237317e-01 -1.47275269e+00
-8.47984493e-01 -2.14342266e-01 -2.30591297e-02 2.83162832e-01
-3.01863421e-02 3.00252914e-01 -9.39217865e-01 -7.03894138e-01
2.57261574e-01 -1.22272444e+00 -3.85547519e-01 -7.97122493e-02
-6.66987956e-01 -4.82666969e-01 7.07638025e-01 -4.00298744e-01
9.73372161e-01 -2.10526323e+00 6.92728460e-02 -1.75534725e-01
-2.17263699e-01 5.97335398e-01 -1.64383441e-01 1.15357041e-01
1.58292338e-01 2.12022200e-01 -7.48955071e-01 -6.65374696e-01
-6.21810555e-02 1.67862177e-01 -1.90854874e-02 2.85138935e-01
3.79101276e-01 1.13655436e+00 -1.03545761e+00 -7.66369879e-01
3.04512054e-01 5.92109896e-02 -1.50879964e-01 2.41039887e-01
-5.95018625e-01 6.70881152e-01 -4.83857274e-01 9.36358213e-01
5.73177457e-01 -3.20307136e-01 -2.47327805e-01 -1.49774417e-01
-1.27344549e-01 2.15469405e-01 -9.62365866e-01 2.16061020e+00
-1.39303744e-01 5.39672613e-01 -6.86748326e-02 -8.44871163e-01
6.38765097e-01 3.69329527e-02 2.92624295e-01 -5.85704625e-01
1.29581355e-02 3.39196265e-01 -2.85711974e-01 -4.95058894e-01
6.14794493e-01 -8.50314796e-02 -5.11968508e-02 2.13566706e-01
2.90208250e-01 -5.55679202e-01 5.84027529e-01 3.91080588e-01
7.31722832e-01 7.25312531e-01 2.34237537e-01 -4.47438598e-01
4.33164001e-01 4.76048261e-01 2.99572110e-01 7.40478516e-01
-5.75433671e-01 1.07420528e+00 2.96665460e-01 -1.02878511e-01
-6.86723471e-01 -1.19834733e+00 -3.81164283e-01 1.29651213e+00
3.35818022e-01 -3.58669817e-01 -1.18431365e+00 -1.26156902e+00
-2.96579123e-01 8.87959957e-01 -5.51608384e-01 4.09927845e-01
-4.03628081e-01 -8.55129778e-01 7.12616026e-01 6.46801949e-01
6.69614673e-01 -1.26489687e+00 -3.75256270e-01 1.60532132e-01
-1.44320235e-01 -1.64635217e+00 -3.45620990e-01 1.06013313e-01
-7.71002889e-01 -1.12331748e+00 -9.15264845e-01 -9.91856277e-01
5.74815094e-01 1.80742741e-01 1.06608331e+00 -1.03746705e-01
-3.04594308e-01 2.85283983e-01 -6.35207593e-01 -4.16935563e-01
-4.02680278e-01 4.29980218e-01 -4.77835983e-01 -2.15089083e-01
2.79260874e-01 1.19091377e-01 -6.31725132e-01 4.58408982e-01
-9.08758819e-01 4.72657472e-01 2.37750560e-01 4.19363856e-01
8.94691885e-01 -3.69048536e-01 6.69282615e-01 -1.57825434e+00
2.86643922e-01 -2.09200382e-01 -4.01136249e-01 2.75179178e-01
-6.36464119e-01 -2.35887840e-01 2.76709497e-01 -4.41507250e-02
-1.39840937e+00 3.67432028e-01 -5.58014035e-01 2.63429850e-01
-6.60674930e-01 2.76834697e-01 2.31349543e-02 -3.38544399e-02
5.66558301e-01 -4.08478975e-02 -4.17804450e-01 -5.72988093e-01
4.22935039e-01 9.30057287e-01 6.97042763e-01 -7.80875087e-01
3.86534959e-01 6.31291747e-01 -2.02605039e-01 -8.03220391e-01
-1.43554282e+00 -9.26823735e-01 -7.33114362e-01 -1.36759773e-01
1.19906318e+00 -8.78073037e-01 5.50445095e-02 6.55515552e-01
-1.10976660e+00 -5.32466650e-01 -1.53724253e-01 1.26186192e-01
-5.75386763e-01 2.54316628e-01 -8.26270163e-01 -4.97106880e-01
-1.32960737e-01 -1.35519838e+00 1.38564408e+00 4.61649299e-01
-3.77196848e-01 -1.10484648e+00 -6.10955209e-02 9.24753487e-01
1.69864148e-01 3.66663337e-01 3.44160378e-01 -8.65148664e-01
-4.65490222e-01 -1.04970068e-01 -4.04103309e-01 5.04306138e-01
-8.64565000e-02 2.90452167e-02 -1.36020422e+00 -1.83358163e-01
-4.90667969e-01 -5.80426872e-01 1.09144163e+00 2.73096174e-01
1.49114847e+00 4.76899624e-01 -4.30192471e-01 5.11404216e-01
1.16845703e+00 7.00694770e-02 4.10637230e-01 6.14646494e-01
7.70753980e-01 7.66672611e-01 1.08840322e+00 -2.01413497e-01
5.31388223e-01 7.33939350e-01 1.82695299e-01 -4.94575024e-01
-6.74872816e-01 -2.01620996e-01 -8.97442773e-02 7.80430496e-01
-4.70131487e-02 -3.15876901e-01 -8.54653597e-01 6.44950926e-01
-1.90179276e+00 -3.50524962e-01 -5.00972033e-01 1.89586973e+00
8.94438207e-01 3.27631086e-01 1.46221563e-01 -1.31618336e-01
7.06316650e-01 3.57643038e-01 -5.12538910e-01 -3.02737206e-01
-3.18303436e-01 3.80135387e-01 4.30379629e-01 3.14862311e-01
-1.53493106e+00 1.55771554e+00 7.01779175e+00 1.01272535e+00
-7.82747447e-01 3.77220273e-01 9.06749845e-01 1.14274397e-01
-1.86889246e-02 2.24745143e-02 -8.84810090e-01 3.25536370e-01
6.67364717e-01 4.23121959e-01 1.24669820e-01 9.13824499e-01
-6.82887733e-02 -4.41804111e-01 -1.00746953e+00 5.47864795e-01
1.48777977e-01 -1.18219578e+00 -1.90553904e-01 -3.73512328e-01
1.19188738e+00 3.05371523e-01 -2.21871063e-01 1.91574633e-01
2.32641801e-01 -8.36525083e-01 9.00366127e-01 -1.55499415e-03
8.50231409e-01 -3.76229405e-01 8.71703327e-01 1.26974985e-01
-9.98642623e-01 3.89606535e-01 -6.23357706e-02 2.60492802e-01
4.03170228e-01 5.63629806e-01 -8.22742403e-01 5.75529695e-01
7.20269740e-01 8.29873860e-01 -9.00106847e-01 1.19221151e+00
-6.28315330e-01 8.68233085e-01 -2.34214619e-01 1.85985625e-01
5.24010897e-01 -9.11012292e-02 2.31255949e-01 1.86530542e+00
-1.27699256e-01 -6.85784072e-02 3.52637380e-01 7.63739407e-01
-5.44396080e-02 8.68126228e-02 -1.71978101e-01 -4.89572883e-02
2.29515970e-01 1.49557316e+00 -1.53306103e+00 -5.63354373e-01
-4.69320029e-01 1.19179702e+00 2.91003853e-01 5.06757021e-01
-7.15938330e-01 -1.27240568e-01 2.39081487e-01 3.10605299e-02
8.06245059e-02 -2.11860165e-01 -6.85660779e-01 -1.02409399e+00
-9.09191594e-02 -5.26439786e-01 4.92938995e-01 -8.20952177e-01
-1.12307286e+00 6.94038093e-01 2.74646342e-01 -9.34880793e-01
8.90287682e-02 -5.57841778e-01 -2.84390986e-01 3.60152066e-01
-1.78216386e+00 -1.45741975e+00 -1.82499677e-01 1.52024433e-01
1.09677041e+00 9.00228843e-02 6.91265047e-01 9.07729417e-02
-5.91275513e-01 5.03855467e-01 -1.19047493e-01 -3.03764865e-02
8.31011534e-01 -1.40668428e+00 7.57394493e-01 7.53537178e-01
5.26963651e-01 2.74858952e-01 6.69637680e-01 -6.78182840e-01
-6.44555509e-01 -1.22286630e+00 6.79547906e-01 -4.67935294e-01
5.81376910e-01 -3.56402665e-01 -7.36101747e-01 7.17237771e-01
2.53269970e-01 4.51739877e-02 5.67912221e-01 -3.88653576e-02
-1.46945417e-01 3.86070788e-01 -1.28032637e+00 5.84146500e-01
1.37599027e+00 -3.49696547e-01 -4.97509748e-01 4.85187322e-01
8.53733480e-01 -8.33369851e-01 -6.89131618e-01 6.87167645e-01
3.74510169e-01 -9.94191051e-01 7.69142807e-01 -3.20334256e-01
3.80683780e-01 -4.11683857e-01 -9.92496684e-03 -1.25780499e+00
-1.15978926e-01 -2.91942179e-01 1.50381491e-01 1.50729489e+00
6.34175479e-01 -4.41369742e-01 1.05124474e+00 5.05334437e-01
-5.07453978e-01 -8.02446127e-01 -8.16408277e-01 -7.24869847e-01
-4.65046708e-03 -7.12912679e-01 2.53284544e-01 9.58724320e-01
-2.47915909e-01 1.40612602e-01 -7.80834910e-03 -5.32814227e-02
7.31947064e-01 1.44888401e-01 4.90767688e-01 -9.45602000e-01
-1.85717747e-01 -3.79982144e-01 -1.98689461e-01 -1.04917288e+00
4.51023638e-01 -9.95426178e-01 3.59635293e-01 -1.66066515e+00
4.33444619e-01 -5.80274820e-01 -3.52231860e-01 5.27017117e-01
-3.47068995e-01 8.55926216e-01 3.77751887e-01 1.16953902e-01
-1.25308371e+00 2.40830898e-01 1.53265047e+00 -1.49427518e-01
-2.53313105e-04 2.65020639e-01 -5.82788944e-01 9.05819356e-01
9.05063331e-01 -4.71850187e-01 -2.71677822e-01 -4.33890492e-01
-2.12694585e-01 -4.31509495e-01 1.50062859e-01 -8.40069771e-01
2.62800511e-02 -2.06934795e-01 1.55333072e-01 -5.40818572e-01
1.42031401e-01 -4.39347625e-01 -1.61592141e-01 1.79443702e-01
-5.77026606e-01 -4.60914344e-01 2.15876028e-01 2.79143333e-01
-3.33863974e-01 -4.86825824e-01 9.10677195e-01 -3.69719654e-01
-1.25859237e+00 1.70344487e-01 -1.89655319e-01 3.02592814e-01
1.19178438e+00 -3.68782729e-01 -2.65767872e-01 6.92485422e-02
-7.72263646e-01 3.84003460e-01 6.85930908e-01 5.61056674e-01
3.81747901e-01 -9.24911916e-01 -5.10545313e-01 5.65829612e-02
5.00956357e-01 4.15974021e-01 -7.34494850e-02 6.70668662e-01
-5.80547690e-01 4.59304273e-01 1.52216613e-01 -8.38947415e-01
-1.22756827e+00 2.26947889e-01 -9.95979365e-03 -1.25863314e-01
-4.59817469e-01 1.07088411e+00 9.20973942e-02 -6.10853851e-01
3.33147168e-01 -4.49035093e-02 -1.99306577e-01 -7.83763304e-02
2.98039973e-01 3.32661957e-01 1.47585988e-01 -5.70528269e-01
-4.62248117e-01 5.05996346e-01 -2.03005135e-01 -1.83518276e-01
1.20105600e+00 -1.43636629e-01 -1.86702935e-03 5.78864455e-01
1.04058206e+00 -2.74578452e-01 -1.49249339e+00 -9.68409032e-02
4.22421128e-01 -2.87987471e-01 8.40121508e-03 -1.19415128e+00
-1.01763010e+00 6.83765948e-01 6.32268906e-01 1.61551431e-01
1.04471946e+00 5.53402781e-01 8.68989408e-01 4.41487432e-02
7.13352740e-01 -1.43311203e+00 -1.18180782e-01 3.07649165e-01
5.33988655e-01 -1.51632607e+00 9.49962214e-02 -8.60211492e-01
-9.51325119e-01 7.79357016e-01 6.91654444e-01 3.06388084e-03
3.07767510e-01 2.00174317e-01 3.00730586e-01 -1.84782848e-01
-3.04374665e-01 -5.62465906e-01 4.69503403e-01 7.34155059e-01
7.08769441e-01 2.37526730e-01 -6.05700612e-01 2.70662487e-01
-1.53224051e-01 9.37330872e-02 4.19684738e-01 1.09673870e+00
-3.47628921e-01 -1.34012926e+00 -1.03909455e-01 4.57105160e-01
-5.77522993e-01 -1.20169111e-02 -5.95003128e-01 7.85673559e-01
3.36558223e-01 1.01310849e+00 -1.73074067e-01 -1.27773315e-01
3.39491516e-01 9.06680003e-02 6.16742313e-01 -1.00780916e+00
-5.12171388e-01 1.58512965e-01 4.38691825e-01 -7.63929367e-01
-9.52321053e-01 -9.33098078e-01 -1.54260695e+00 3.08566242e-01
-3.90418738e-01 2.96909939e-02 9.85430479e-01 1.12691402e+00
1.20627455e-01 5.46306372e-01 1.40466243e-01 -8.57091665e-01
-6.95364103e-02 -9.37188864e-01 -5.49427271e-01 6.01357281e-01
-7.83938766e-02 -6.61755800e-01 -2.04485834e-01 2.78912812e-01] | [9.543098449707031, 0.5813727378845215] |
c075f15b-3cbf-4b1c-9837-e8643e31ffd5 | a-comprehensive-analysis-of-acknowledgement | 2210.09716 | null | https://arxiv.org/abs/2210.09716v1 | https://arxiv.org/pdf/2210.09716v1.pdf | A Comprehensive Analysis of Acknowledgement Texts in Web of Science: a case study on four scientific domains | Analysis of acknowledgments is particularly interesting as acknowledgments may give information not only about funding, but they are also able to reveal hidden contributions to authorship and the researcher's collaboration patterns, context in which research was conducted, and specific aspects of the academic work. The focus of the present research is the analysis of a large sample of acknowledgement texts indexed in the Web of Science (WoS) Core Collection. Record types 'article' and 'review' from four different scientific domains, namely social sciences, economics, oceanography and computer science, published from 2014 to 2019 in a scientific journal in English were considered. Six types of acknowledged entities, i.e., funding agency, grant number, individuals, university, corporation and miscellaneous, were extracted from the acknowledgement texts using a Named Entity Recognition (NER) tagger and subsequently examined. A general analysis of the acknowledgement texts showed that indexing of funding information in WoS is incomplete. The analysis of the automatically extracted entities revealed differences and distinct patterns in the distribution of acknowledged entities of different types between different scientific domains. A strong association was found between acknowledged entity and scientific domain and acknowledged entity and entity type. Only negligible correlation was found between the number of citations and the number of acknowledged entities. Generally, the number of words in the acknowledgement texts positively correlates with the number of acknowledged funding organizations, universities, individuals and miscellaneous entities. At the same time, acknowledgement texts with the larger number of sentences have more acknowledged individuals and miscellaneous categories. | ['Philipp Mayr', 'Nina Smirnova'] | 2022-10-18 | null | null | null | null | ['miscellaneous'] | ['miscellaneous'] | [-4.53914642e-01 1.81319043e-01 -3.12603205e-01 2.42107660e-01
-3.00651163e-01 -8.13616753e-01 8.25018585e-01 5.43024242e-01
-8.68395507e-01 8.15497756e-01 7.47637510e-01 -6.50784671e-01
-2.84131706e-01 -8.39252114e-01 -3.66856992e-01 -2.79580772e-01
3.63978446e-01 2.22173318e-01 -3.58449831e-03 5.33217564e-02
1.15357804e+00 6.61836505e-01 -1.21553862e+00 -9.39981341e-02
8.57938766e-01 4.69031870e-01 4.99456376e-01 4.47273284e-01
-1.03932905e+00 7.72707880e-01 -1.09596455e+00 -3.05377185e-01
-2.67381907e-01 -3.77011716e-01 -8.59703362e-01 -2.78168947e-01
2.28268579e-01 3.85946304e-01 -2.17224941e-01 8.87536824e-01
3.79200608e-01 -3.28561813e-01 7.04156518e-01 -8.52332115e-01
-7.32340515e-01 8.06644320e-01 -4.98581827e-01 6.66975379e-01
5.37415624e-01 -2.00698450e-01 8.94889832e-01 -9.85301018e-01
1.29128921e+00 1.09839582e+00 5.31663775e-01 6.90668300e-02
-6.82903647e-01 -1.09177291e+00 -3.74866366e-01 -8.48940611e-02
-1.24743569e+00 -4.19411331e-01 4.07515407e-01 -1.06230831e+00
5.78924537e-01 1.93189662e-02 4.10130382e-01 8.38000178e-01
2.42254391e-01 -1.04760900e-01 1.10892129e+00 -7.31605172e-01
1.64230019e-01 5.52192152e-01 3.32905293e-01 3.20703626e-01
9.34126675e-01 -4.76292938e-01 -7.13580668e-01 -4.89491940e-01
5.60071409e-01 6.50277212e-02 -2.34550744e-01 3.58941287e-01
-1.67755377e+00 6.59917235e-01 -2.46475488e-01 1.10010171e+00
-7.21234858e-01 -2.35273436e-01 4.58986491e-01 2.36514851e-01
4.09132957e-01 6.93719208e-01 -5.69568813e-01 -4.39393848e-01
-9.95021522e-01 1.28644742e-02 1.39540267e+00 9.63054836e-01
6.55141890e-01 -1.14681058e-01 8.17895830e-02 7.40646183e-01
2.95120984e-01 5.21030486e-01 5.42263150e-01 -1.05139315e+00
6.30363703e-01 1.15695262e+00 2.53170282e-01 -1.30936909e+00
-2.63898104e-01 -4.66119528e-01 -5.67423165e-01 -3.51178020e-01
5.29880524e-01 -2.74694473e-01 -3.26847553e-01 1.37995720e+00
1.36280417e-01 -3.05554539e-01 3.55100751e-01 4.79352742e-01
1.55815876e+00 6.04664326e-01 2.61386603e-01 -2.45623454e-01
1.82021224e+00 -3.10665697e-01 -1.03428936e+00 1.03550322e-01
5.25568724e-01 -1.30870843e+00 1.82517380e-01 -2.67650455e-01
-8.91054749e-01 -3.27051878e-01 -3.45825851e-01 1.63815603e-01
-8.97543669e-01 4.60469127e-01 3.03923696e-01 3.48370910e-01
-8.32347453e-01 4.77853656e-01 -2.33586952e-01 -5.06846189e-01
3.21460217e-01 -3.51507962e-02 -5.36853731e-01 2.88937539e-01
-1.11678386e+00 1.06496596e+00 3.10229808e-02 -2.95259148e-01
7.81029761e-02 -7.72643507e-01 -6.79400742e-01 -1.43337160e-01
2.76414454e-01 -3.54942888e-01 6.40450358e-01 -6.78010643e-01
-6.71624005e-01 1.34760606e+00 -4.42293048e-01 -1.47257045e-01
1.49141759e-01 9.95833892e-03 -8.96140277e-01 2.59813696e-01
7.22957313e-01 5.97782619e-02 1.92500979e-01 -6.54165924e-01
-9.90917385e-01 -3.93234700e-01 -1.99799582e-01 -3.71967912e-01
-4.21285748e-01 7.10517943e-01 3.87095772e-02 -8.42267275e-01
-9.89387110e-02 -5.43683648e-01 1.23859107e-01 -3.95656139e-01
-1.46309868e-01 -8.98146927e-01 5.20412982e-01 -9.72097099e-01
1.47817278e+00 -2.08103013e+00 1.16390288e-01 6.51800707e-02
6.20560467e-01 -8.99820700e-02 4.20699358e-01 9.05357659e-01
1.06869146e-01 8.79678607e-01 9.74688008e-02 4.22079563e-01
-1.32111490e-01 7.96941891e-02 -1.93041161e-01 6.60936296e-01
-1.89451024e-01 5.53026021e-01 -8.79800558e-01 -6.50595367e-01
-4.50432837e-01 1.30115122e-01 1.10902883e-01 5.37729748e-02
6.36397600e-01 2.56033540e-01 -7.83938825e-01 5.86802304e-01
4.65710312e-01 -3.32569718e-01 1.19257085e-02 1.73348606e-01
-1.18391979e+00 5.29670119e-01 -1.09492040e+00 8.80541801e-01
-4.04766351e-01 1.24981594e+00 3.41708630e-01 -6.67273879e-01
1.05480611e+00 5.18753886e-01 3.97407293e-01 -3.03016692e-01
1.00427300e-01 7.14060485e-01 1.82045281e-01 -5.88254750e-01
8.43926728e-01 1.17775664e-01 -1.75719172e-01 5.39979458e-01
-6.93647340e-02 3.96659374e-01 4.07435775e-01 6.20794535e-01
9.96806562e-01 -1.39313370e-01 7.11884558e-01 -6.65609300e-01
6.96776032e-01 1.50686249e-01 7.30121970e-01 5.69453001e-01
-5.65059222e-02 5.09225018e-02 8.14583898e-01 -2.32840270e-01
-1.20506489e+00 -3.11212212e-01 -5.11321962e-01 7.16534674e-01
-1.24298774e-01 -2.54338771e-01 -4.39935714e-01 -1.53263465e-01
2.26449192e-01 5.40403187e-01 -6.23624146e-01 3.15333366e-01
-5.09508193e-01 -1.62086442e-01 5.95880687e-01 6.43108040e-02
3.07277054e-01 -1.38363290e+00 -9.37495589e-01 1.94383189e-01
-2.73505628e-01 -1.24477136e+00 -3.17619950e-01 -9.50247347e-02
-7.81040788e-01 -1.22643387e+00 -1.24467587e+00 -8.43430042e-01
5.99533975e-01 -6.64406791e-02 1.15707910e+00 1.77031979e-01
-9.24046785e-02 4.33141679e-01 -3.60765308e-01 -6.67942524e-01
-6.37036085e-01 2.54354239e-01 -7.12299123e-02 -5.38122833e-01
6.04242682e-01 -2.00479746e-01 -1.99829340e-01 8.08633342e-02
-6.30969048e-01 -5.40852249e-01 5.84477901e-01 7.09079146e-01
8.46266672e-02 -2.43993297e-01 8.85913908e-01 -1.06817675e+00
5.05477309e-01 -9.56134558e-01 -2.72936672e-01 1.40658334e-01
-8.79509211e-01 -2.59410702e-02 4.08140302e-01 -3.07419896e-01
-8.75664294e-01 -8.17630351e-01 2.75857389e-01 1.67734027e-01
-5.74849188e-01 8.18329096e-01 2.01910809e-02 -2.85961665e-03
2.82983989e-01 -8.97535495e-03 -2.20035780e-02 -7.50006318e-01
-3.47555637e-01 1.14799905e+00 3.12891245e-01 -3.86161774e-01
5.44924915e-01 2.48109117e-01 -6.39247522e-03 -1.15969920e+00
-5.87041676e-01 -9.30532277e-01 -5.66908956e-01 -2.97473401e-01
7.18154192e-01 -8.69262874e-01 -7.55161464e-01 4.21234220e-01
-1.39886630e+00 4.66820687e-01 -1.63384765e-01 8.34073603e-01
3.91657740e-01 5.42458713e-01 -4.24962014e-01 -7.71283150e-01
-3.29209864e-01 -8.05720806e-01 5.86393774e-01 5.29810786e-01
-5.47955573e-01 -1.12151492e+00 7.84179568e-02 4.58212525e-01
4.32946265e-01 4.11814094e-01 8.10496926e-01 -9.24371421e-01
-6.81022042e-03 -8.02771747e-02 -2.48964101e-01 -1.44884229e-01
1.01494528e-01 2.78599262e-01 -3.03709447e-01 2.34095842e-01
-9.72542018e-02 4.30998445e-01 6.52642846e-01 1.45592168e-01
5.49293041e-01 -8.14229727e-01 -4.37363982e-01 5.83023243e-02
1.37052548e+00 2.71251529e-01 5.33914864e-01 9.75478053e-01
6.20644331e-01 1.06956148e+00 3.26515108e-01 6.50290191e-01
3.91472191e-01 4.03361827e-01 -1.07383303e-01 3.60451221e-01
-9.27568506e-03 1.83741614e-01 4.55050915e-01 1.03757370e+00
-5.94148099e-01 -1.35922447e-01 -1.27964044e+00 9.69621122e-01
-1.10998178e+00 -1.18664753e+00 -1.07923567e+00 2.01244998e+00
7.60214210e-01 -1.03706114e-01 3.70564274e-02 -1.70723334e-01
1.01566994e+00 -1.24827679e-02 -1.08147740e-01 -4.13299859e-01
-3.77327025e-01 1.85998738e-01 9.87139404e-01 4.71076593e-02
-3.51520598e-01 6.75598800e-01 5.85287189e+00 6.26042128e-01
-9.33588028e-01 -1.97821725e-02 1.50831267e-01 2.42383361e-01
-4.00134206e-01 3.93483639e-01 -1.11560166e+00 7.63653815e-01
1.34480166e+00 -7.24227548e-01 -1.50745139e-01 6.99723423e-01
3.67255598e-01 -4.68969166e-01 -3.58244985e-01 5.19063890e-01
-1.30401328e-01 -1.65482652e+00 -2.73607284e-01 7.34440625e-01
4.68321949e-01 5.24274111e-02 -4.18502897e-01 -1.43944860e-01
7.56302178e-02 -5.40993333e-01 1.05403697e+00 7.27514863e-01
7.17984140e-01 -5.55935144e-01 1.05919147e+00 2.77821064e-01
-7.98100173e-01 1.45323589e-01 -4.71664578e-01 -3.23124290e-01
4.17922735e-02 8.84808540e-01 -4.22104329e-01 6.74950600e-01
8.87934446e-01 7.48659611e-01 -4.71456766e-01 7.73387671e-01
-2.48280078e-01 1.06717253e+00 3.46708782e-02 -3.40060145e-01
1.93955377e-01 -4.54369009e-01 8.13951254e-01 1.43544352e+00
5.57352066e-01 3.12408090e-01 -4.82939124e-01 8.10123861e-01
-4.53109205e-01 3.16358030e-01 -6.66674674e-01 -6.67573333e-01
1.01647234e+00 1.34328651e+00 -8.91309798e-01 -4.85898674e-01
-5.10472536e-01 3.22531492e-01 1.13081262e-01 2.70022243e-01
-2.03917399e-01 -9.01521266e-01 2.62998849e-01 4.91801977e-01
4.01267141e-01 -2.04442158e-01 -3.89054239e-01 -9.44671750e-01
-1.61242932e-01 -5.41007340e-01 2.91146338e-01 -4.25758481e-01
-1.10118771e+00 3.96598011e-01 -1.79181531e-01 -9.31343257e-01
1.26729091e-03 -4.29345399e-01 -5.03776491e-01 1.28792381e+00
-1.19946647e+00 -4.98880267e-01 -1.16083041e-01 2.33110394e-02
2.13905461e-02 -6.90432310e-01 5.94517469e-01 2.99105376e-01
-6.33276284e-01 1.41843259e-01 5.24073899e-01 4.17400777e-01
7.88162708e-01 -1.11443269e+00 -5.39252982e-02 3.94526094e-01
-2.16066658e-01 1.03454244e+00 6.97301447e-01 -9.25504565e-01
-1.19731402e+00 -6.62284791e-01 1.98931301e+00 -4.28321451e-01
1.20184100e+00 1.67815134e-01 -9.30707097e-01 2.02008963e-01
6.32655323e-01 -4.97621030e-01 1.04113710e+00 1.10655561e-01
-6.17400743e-02 3.41859668e-01 -1.00048614e+00 3.32229376e-01
8.51178050e-01 -3.93837363e-01 -1.03363907e+00 2.54897922e-01
3.96362931e-01 -8.16105753e-02 -1.52387333e+00 -2.91496307e-01
5.60333729e-01 -2.98626482e-01 7.32177615e-01 -4.97268766e-01
9.12469804e-01 4.23219949e-02 3.23428690e-01 -7.37543404e-01
-5.46594143e-01 -3.84817541e-01 1.72537386e-01 1.92186093e+00
4.55145419e-01 -1.00646317e+00 4.24001485e-01 6.23575568e-01
-1.78144768e-01 -4.38538343e-01 -1.13264620e+00 -6.11979425e-01
2.69121468e-01 2.01738536e-01 2.57055223e-01 1.36554098e+00
-1.96065363e-02 2.03695014e-01 1.45301849e-01 -3.23637158e-01
5.87630749e-01 2.95915484e-01 5.02213299e-01 -1.73799992e+00
3.50792795e-01 -6.62880719e-01 -4.08252597e-01 -3.30063879e-01
3.72029006e-01 -7.56900847e-01 -5.31928062e-01 -1.95308292e+00
8.81015956e-02 -4.59433347e-01 1.70065865e-01 4.29504007e-01
8.00084136e-03 -3.63512546e-01 -2.99809910e-02 1.05316734e+00
3.74896266e-02 -1.13241356e-02 9.96655464e-01 2.11822659e-01
-1.50248423e-01 -1.58829108e-01 -8.97575617e-01 8.13763916e-01
5.83661973e-01 -8.99263799e-01 7.30071902e-01 -3.88797112e-02
4.71716851e-01 -3.39492932e-02 2.78420210e-01 -5.79646051e-01
4.79266822e-01 -4.50455576e-01 6.37947842e-02 -5.38925469e-01
-6.43804908e-01 -6.75150573e-01 2.87520647e-01 5.30650735e-01
-3.11173350e-01 2.01098636e-01 2.15150956e-02 4.02095050e-01
-3.78631502e-01 -7.73221254e-01 1.96131811e-01 -2.61338890e-01
-4.90436196e-01 -3.67668331e-01 -8.61086488e-01 3.56011927e-01
7.83233762e-01 -4.76039112e-01 -5.05614460e-01 -6.68355916e-03
-1.70895815e-01 -1.03502154e-01 7.14555383e-01 2.48627260e-01
2.30252624e-01 -8.61921191e-01 -9.55108762e-01 -7.31232703e-01
5.37996441e-02 -3.54004055e-01 5.30430414e-02 1.05326843e+00
-5.14098227e-01 5.39217472e-01 -2.09565505e-01 2.62968868e-01
-1.18738842e+00 2.40900159e-01 -3.34974080e-01 -1.85304508e-01
-5.74005961e-01 3.03799152e-01 -1.48592338e-01 -1.96050063e-01
1.01279937e-01 9.01883189e-03 -9.33387220e-01 7.90791869e-01
2.64520764e-01 9.67643857e-01 -2.95967013e-01 -1.27492440e+00
-5.81015527e-01 5.24432361e-01 8.26627240e-02 5.43281762e-03
1.51846159e+00 -2.28699401e-01 -6.85682058e-01 8.91059875e-01
1.08931816e+00 7.04159677e-01 -5.33976629e-02 -1.51594967e-01
6.33245170e-01 -2.15474561e-01 9.53107234e-03 -6.03622377e-01
-1.10312486e+00 5.86373687e-01 5.10406457e-02 4.03684944e-01
4.11074728e-01 3.17901522e-01 5.04389644e-01 -1.22605123e-01
1.67785689e-01 -1.18257582e+00 -5.60000181e-01 6.88961923e-01
7.77520835e-01 -8.38869572e-01 2.86516130e-01 -4.02613312e-01
-4.02686089e-01 1.44421768e+00 1.15986496e-01 3.53369892e-01
6.49513185e-01 1.99073136e-01 4.74114204e-03 -6.55483365e-01
-2.04744488e-01 1.66971356e-01 3.88756961e-01 1.39849007e-01
9.88379419e-01 -1.23524010e-01 -1.35882413e+00 8.25819433e-01
-2.73385763e-01 4.06116880e-02 1.20066333e+00 9.03558671e-01
-4.85895813e-01 -9.91884708e-01 -7.09172368e-01 6.84364498e-01
-1.45869017e+00 -1.71148777e-01 -7.94182360e-01 7.75058687e-01
2.54432499e-01 8.13780904e-01 2.22772911e-01 1.52835175e-01
2.12444589e-01 -1.28123268e-01 -1.32981867e-01 -6.00850761e-01
-1.11643267e+00 -2.30469257e-01 2.46077493e-01 2.23368615e-01
-9.74088848e-01 -9.93537486e-01 -1.42491639e+00 -5.52580595e-01
-4.03831720e-01 9.52592492e-01 9.49170232e-01 9.89951789e-01
6.94658637e-01 5.15856206e-01 2.76080996e-01 -1.81139588e-01
3.78448740e-02 -1.14859986e+00 -7.65439212e-01 8.97463486e-02
-1.19766533e-01 -6.91840649e-01 -9.06012952e-01 2.28088453e-01] | [9.590310096740723, 8.296892166137695] |
009a51d0-b4eb-48e2-8ae1-d8f7b95794ea | towards-optimizing-reiters-hs-tree-for | 1907.12130 | null | https://arxiv.org/abs/1907.12130v1 | https://arxiv.org/pdf/1907.12130v1.pdf | Towards Optimizing Reiter's HS-Tree for Sequential Diagnosis | Reiter's HS-Tree is one of the most popular diagnostic search algorithms due to its desirable properties and general applicability. In sequential diagnosis, where the addressed diagnosis problem is subject to successive change through the acquisition of additional knowledge about the diagnosed system, HS-Tree is used in a stateless fashion. That is, the existing search tree is discarded when new knowledge is obtained, albeit often large parts of the tree are still relevant and have to be rebuilt in the next iteration, involving redundant operations and costly reasoner calls. As a remedy to this, we propose DynamicHS, a variant of HS-Tree that avoids these redundancy issues by maintaining state throughout sequential diagnosis while preserving all desirable properties of HS-Tree. Preliminary results of ongoing evaluations in a problem domain where HS-Tree is the state-of-the-art diagnostic method suggest significant time savings achieved by DynamicHS by reducing expensive reasoner calls. | ['Patrick Rodler'] | 2019-07-28 | null | null | null | null | ['sequential-diagnosis'] | ['medical'] | [ 3.73173058e-01 5.18621802e-01 -3.00843388e-01 -2.11791620e-01
-5.71124911e-01 -4.89739239e-01 3.28049630e-01 4.03206497e-01
3.04249451e-02 8.62804592e-01 -2.19127864e-01 -7.18511283e-01
-6.11780524e-01 -5.44840574e-01 -7.55060185e-03 -5.27214646e-01
-2.42888164e-02 1.03856552e+00 8.79867315e-01 1.20607583e-04
2.60741472e-01 7.44551182e-01 -1.62856579e+00 -2.56370730e-03
1.04614425e+00 7.55877197e-01 -5.55208661e-02 5.25728941e-01
-6.65920600e-02 1.13377154e+00 -5.48012972e-01 -1.21264122e-01
1.83605745e-01 -8.55424225e-01 -1.61148119e+00 4.30857480e-01
-1.90085679e-01 -1.68771520e-01 -1.75194815e-01 1.14710796e+00
7.69929290e-02 3.32575222e-03 4.18906920e-02 -1.33212292e+00
2.58253247e-01 7.21632779e-01 -4.30369437e-01 2.18881860e-01
7.00758755e-01 -1.82640925e-02 7.76567221e-01 -3.86156648e-01
7.61139810e-01 1.14065659e+00 6.51588798e-01 5.12225688e-01
-1.38865054e+00 -4.10451025e-01 1.44071966e-01 4.11182642e-01
-1.67929792e+00 -3.70805651e-01 4.45882529e-01 -1.75110996e-01
1.44423437e+00 6.85046017e-01 7.67750680e-01 2.07859188e-01
4.30335373e-01 5.80714762e-01 9.77578998e-01 -6.22703493e-01
6.16703510e-01 1.93397626e-01 1.45487845e-01 1.12968588e+00
3.76919925e-01 1.41322404e-01 -2.63255805e-01 -6.45542800e-01
5.30204952e-01 -1.87276214e-01 -1.07771009e-01 -6.35945976e-01
-1.02708399e+00 5.15938818e-01 4.94540557e-02 5.44427931e-01
-4.11198288e-01 -4.66268882e-02 5.27042210e-01 6.28021359e-01
2.56981343e-01 6.91552043e-01 -5.10492623e-01 -2.61965722e-01
-1.22807324e+00 3.94991308e-01 9.87621665e-01 6.00331485e-01
6.29404783e-01 -4.70674969e-02 1.04950011e-01 4.40541118e-01
-3.92549112e-03 1.37003869e-01 5.06151915e-01 -1.05378413e+00
-8.90369155e-03 1.12240589e+00 2.59068627e-02 -6.11406922e-01
-4.54861850e-01 -7.18028784e-01 -4.84095484e-01 3.39704722e-01
8.38649459e-03 -8.47625211e-02 -1.04030156e+00 1.41036212e+00
5.43292165e-01 1.66509211e-01 1.25997722e-01 6.02357149e-01
2.02750877e-01 3.83099437e-01 -3.57604980e-01 -6.98822498e-01
9.77694631e-01 -8.52637351e-01 -4.15448248e-01 -2.15685248e-01
8.68404627e-01 -4.53585237e-01 2.51952380e-01 7.95445025e-01
-1.36953616e+00 -9.93548632e-02 -9.43749070e-01 5.09971499e-01
-2.95107663e-02 -5.01222253e-01 8.11855972e-01 5.40628374e-01
-1.43977380e+00 5.72643697e-01 -1.02161026e+00 -6.30539536e-01
-1.34003878e-01 6.01002932e-01 -2.86744863e-01 -2.55607516e-01
-9.18443322e-01 1.10811484e+00 5.03176570e-01 -7.45136738e-02
-7.18334973e-01 -4.80733931e-01 -6.32516026e-01 1.44802228e-01
9.45876181e-01 -6.45315111e-01 1.82915664e+00 -5.56970417e-01
-1.11097360e+00 6.65801525e-01 -3.46666813e-01 -4.09963667e-01
6.10422671e-01 2.60045141e-01 -7.28121400e-01 3.39707583e-01
2.93377012e-01 7.38675222e-02 6.78565800e-01 -9.99323308e-01
-1.07027280e+00 -2.41072461e-01 2.47076213e-01 7.85031468e-02
-2.46465788e-03 -9.62805450e-02 -5.24780214e-01 -1.11539170e-01
8.09063613e-01 -1.08944488e+00 -6.02232873e-01 -2.97771573e-01
-3.71920288e-01 -1.89468771e-01 8.64023507e-01 -4.72379297e-01
1.73222840e+00 -1.88686490e+00 4.08221185e-01 7.06132948e-01
4.34271157e-01 2.80685484e-01 2.65151113e-01 6.86830044e-01
-3.87205988e-01 -1.39483929e-01 -3.94915015e-01 1.26666307e-01
-1.88137323e-01 4.36729729e-01 -8.33957493e-02 4.07654464e-01
-8.57871622e-02 5.58624506e-01 -1.14168668e+00 -8.07611883e-01
2.08688810e-01 -3.29405367e-01 -4.86604840e-01 -1.83242083e-01
-2.54630327e-01 2.38604918e-01 -6.73772097e-01 8.49750340e-01
2.90472567e-01 -5.27421594e-01 4.90500420e-01 2.04328343e-01
-9.57748294e-02 3.00062239e-01 -1.41360617e+00 1.44595945e+00
-3.15865606e-01 1.50000125e-01 1.24578871e-01 -1.25216269e+00
5.59823990e-01 5.83199799e-01 6.94853723e-01 -5.02562761e-01
4.48281690e-02 3.80583972e-01 1.29384711e-01 -4.56700683e-01
2.24370167e-01 -2.11149886e-01 -8.47431421e-02 6.38954699e-01
-4.76261646e-01 -4.49489772e-01 4.62588876e-01 2.66798884e-01
1.71283257e+00 -4.98505622e-01 8.72363031e-01 -3.92251551e-01
7.70738363e-01 5.33229709e-01 7.18997836e-01 7.08242774e-01
-2.82971542e-02 1.15050979e-01 3.75747144e-01 -3.16606343e-01
-4.98200119e-01 -8.59532714e-01 -2.03499973e-01 4.19041663e-01
1.93167552e-01 -5.17857552e-01 -5.90054691e-01 -9.46442664e-01
-1.64576676e-02 8.82916570e-01 -4.51863974e-01 -3.63660604e-01
-4.70627338e-01 -1.76820636e-01 3.42808694e-01 5.09691179e-01
4.04612929e-01 -8.69749188e-01 -1.14056206e+00 6.13601863e-01
1.67494401e-01 -5.74432790e-01 -9.37526301e-03 4.30832237e-01
-1.31614542e+00 -1.39946389e+00 -3.48192602e-01 -6.82724833e-01
9.92040217e-01 1.39187619e-01 8.95325184e-01 5.05450368e-01
-5.10453165e-01 2.76099443e-01 -2.46366382e-01 2.25187764e-01
-7.68559813e-01 1.00481294e-01 -2.07074657e-01 -5.46557367e-01
-8.82092938e-02 -5.00923216e-01 -1.23612314e-01 2.74684966e-01
-8.90175641e-01 -9.56755877e-03 4.77272779e-01 9.61195171e-01
5.03919959e-01 1.11859739e+00 2.42725119e-01 -1.16954994e+00
5.75515032e-01 -2.97936201e-01 -8.61499190e-01 4.24741060e-01
-1.22449517e+00 3.16646099e-01 4.79971647e-01 -1.80744275e-01
-1.23736000e+00 6.55803680e-02 1.20063700e-01 -2.64891088e-01
6.23360239e-02 9.16290462e-01 1.21042706e-01 -8.18226486e-02
5.54419875e-01 1.99471712e-01 -7.49926567e-02 -3.76717240e-01
-5.15611023e-02 3.67314160e-01 6.16237998e-01 -3.88148129e-01
6.20079517e-01 3.07562798e-01 3.46142083e-01 -2.65791684e-01
-5.24378121e-01 -5.72113812e-01 -3.01863462e-01 1.06815761e-02
1.51574671e-01 -3.50543797e-01 -5.97321570e-01 3.41218114e-01
-8.54018986e-01 3.34461918e-03 -6.14527881e-01 1.48382366e-01
-3.03204507e-01 4.34428126e-01 -5.80240130e-01 -8.03120077e-01
-2.34325320e-01 -1.24303901e+00 7.33105779e-01 -2.48512579e-03
-8.89741659e-01 -9.49587584e-01 1.23863548e-01 9.69521403e-02
1.40256301e-01 5.80973327e-02 1.39923394e+00 -7.30775714e-01
-3.36609781e-01 -5.72544217e-01 1.70627162e-02 -1.05857261e-01
5.21650910e-01 -1.60260007e-01 -4.07162309e-01 -4.78566527e-01
2.63069004e-01 1.26398012e-01 2.78704345e-01 1.21681996e-01
9.31298077e-01 -2.54954457e-01 -9.75083470e-01 1.30340923e-02
1.31587934e+00 8.43185663e-01 3.30727786e-01 2.75260091e-01
2.23530844e-01 5.22368252e-01 8.71928632e-01 5.48592508e-01
1.94603711e-01 4.76468503e-01 2.04701588e-01 -1.30320594e-01
2.16228794e-02 9.83641744e-02 -9.18635726e-02 6.33322537e-01
1.80128917e-01 -5.81411421e-02 -1.36163449e+00 9.00045037e-01
-1.98326528e+00 -7.48052955e-01 -1.36458445e-02 2.23880243e+00
8.74698400e-01 5.77370167e-01 4.26486693e-02 8.25203896e-01
6.33025467e-01 -3.02210122e-01 -8.90646935e-01 -5.83166301e-01
4.78318661e-01 2.94331193e-01 2.68609464e-01 6.70863867e-01
-4.65515792e-01 7.10557461e-01 7.33358192e+00 6.10492826e-01
-1.04222894e+00 -1.17045932e-01 2.16282941e-02 -4.78542363e-03
-2.87868679e-01 3.83883297e-01 -3.87430727e-01 7.56503865e-02
8.82451952e-01 -5.31340063e-01 3.41779262e-01 9.60316241e-01
-1.34059697e-01 -6.35592103e-01 -1.40512657e+00 7.24485636e-01
-1.12843737e-01 -1.16224313e+00 -3.15539151e-01 -2.07785945e-02
6.81784093e-01 -4.19641286e-01 -3.10001880e-01 1.05670825e-01
5.28457940e-01 -5.37344217e-01 5.79602063e-01 5.10659404e-02
6.91182256e-01 -8.35258842e-01 8.32849622e-01 3.78596991e-01
-1.30131161e+00 -3.48335594e-01 3.11809301e-01 -9.20152590e-02
2.59720743e-01 7.63213754e-01 -1.18553817e+00 7.17911184e-01
5.17905772e-01 3.30192655e-01 -4.26799864e-01 1.05010247e+00
-2.89624661e-01 4.88755137e-01 -2.11376011e-01 3.63112152e-01
2.70855427e-01 2.92556852e-01 7.29649067e-01 8.83714318e-01
4.09347504e-01 1.31386563e-01 5.12415528e-01 5.61294734e-01
6.65440023e-01 -4.70740497e-01 -3.79196197e-01 -5.85698709e-03
5.92964649e-01 7.35192895e-01 -1.07070637e+00 -7.80490875e-01
-2.09901080e-01 9.67576802e-01 -1.32729232e-01 3.40940766e-02
-5.90653598e-01 -3.63018394e-01 2.67923385e-01 2.27022156e-01
2.16275886e-01 2.26104245e-01 -2.08305046e-01 -7.05782890e-01
-4.72820699e-02 -1.12546623e+00 7.11079061e-01 -6.08993769e-01
-7.94175625e-01 6.46297514e-01 3.30745935e-01 -1.09587801e+00
-1.03355372e+00 8.27115178e-02 -1.11327060e-01 6.40429914e-01
-1.09322190e+00 -6.52246892e-01 8.17494094e-02 8.31337452e-01
6.54911101e-01 8.29767361e-02 9.05452669e-01 1.33219555e-01
-4.09115106e-01 3.17100614e-01 -1.14859544e-01 -6.64650261e-01
2.64875859e-01 -1.33923483e+00 2.94279307e-01 9.37718630e-01
-3.31613272e-01 6.97310507e-01 1.08576560e+00 -1.06103158e+00
-1.39107001e+00 -6.57553852e-01 1.23665798e+00 1.29586652e-01
8.00910413e-01 1.69902563e-01 -9.52949643e-01 6.66131437e-01
-1.76401094e-01 -2.75758356e-01 2.43528008e-01 1.33936822e-01
-8.49301964e-02 -2.55763292e-01 -1.48704946e+00 5.00787079e-01
1.11076272e+00 -5.82566082e-01 -7.57624805e-01 2.56664425e-01
4.60467696e-01 -4.53717738e-01 -8.25458765e-01 4.50735033e-01
1.96284920e-01 -9.33928311e-01 5.06651402e-01 -1.71540901e-01
-7.36547410e-02 -5.37683904e-01 1.99405238e-01 -1.21732330e+00
-5.62662423e-01 -8.04874718e-01 -3.50698948e-01 7.39587903e-01
4.12999898e-01 -9.88645077e-01 7.76250064e-01 9.44639802e-01
-2.39049464e-01 -7.90661931e-01 -1.08854365e+00 -9.83754277e-01
-5.18039107e-01 -2.34132722e-01 7.92692602e-01 1.01003850e+00
7.94456005e-01 1.80307269e-01 3.89376283e-02 3.12961608e-01
5.01599669e-01 4.36609060e-01 2.98167914e-01 -1.45473135e+00
-6.22270882e-01 -2.25525960e-01 -3.82072806e-01 -6.00433469e-01
4.56510372e-02 -6.41414404e-01 2.42378071e-01 -1.65188003e+00
2.52878785e-01 -4.60395157e-01 -3.38915497e-01 9.49324429e-01
1.62483215e-01 -3.05264980e-01 -1.15162350e-01 2.35818878e-01
-6.41003072e-01 -2.90482491e-01 1.11199594e+00 -1.23486206e-01
-4.16037470e-01 1.00866906e-01 -4.69919115e-01 7.28721797e-01
4.50836957e-01 -5.74592233e-01 -9.71454382e-01 1.44097805e-02
1.12721756e-01 8.76738787e-01 2.62512211e-02 -1.07899904e+00
7.00421810e-01 -3.39551926e-01 -1.82706594e-01 -6.87408805e-01
1.49600580e-01 -1.22857618e+00 8.15296412e-01 1.24538469e+00
-2.59220958e-01 3.37689877e-01 2.82572240e-01 1.60971671e-01
-2.99865842e-01 -5.58090389e-01 6.16481602e-01 -1.42998800e-01
-9.47082341e-01 -6.60061557e-03 -8.05198073e-01 -3.38971198e-01
1.29608595e+00 -6.64915979e-01 -2.73977201e-02 -9.03659016e-02
-9.25679803e-01 2.89412767e-01 5.61002076e-01 2.09091026e-02
4.92356718e-01 -6.72620714e-01 -1.11126192e-01 1.15433872e-01
1.14291579e-01 3.36160623e-02 1.37133390e-01 8.80290747e-01
-3.39281231e-01 6.35284781e-01 1.75166026e-01 -5.28406680e-01
-1.61755729e+00 7.64262617e-01 1.95451692e-01 -6.65390491e-01
-9.71351504e-01 6.83046222e-01 -2.13711970e-02 -7.97163248e-02
1.22847050e-01 -4.46466058e-01 3.14468265e-01 -2.06359699e-01
4.28342819e-01 6.87938631e-01 4.62490380e-01 1.50670037e-02
-8.11946571e-01 2.69721240e-01 -4.10351306e-01 -2.48029992e-01
1.18938828e+00 -3.44729811e-01 -4.50261861e-01 2.62971908e-01
5.38716435e-01 -2.03384772e-01 -6.57192171e-01 -2.84630537e-01
2.71922112e-01 -3.91589612e-01 3.70247722e-01 -1.19898152e+00
-1.17244053e+00 1.45906329e-01 3.19589972e-01 4.65204984e-01
1.52792799e+00 9.16252658e-02 6.62526190e-01 5.00144005e-01
8.90054047e-01 -9.05385911e-01 -5.00311196e-01 1.78201795e-01
5.95538378e-01 -5.48825502e-01 3.17656696e-01 -6.23133838e-01
-3.36064309e-01 8.23025286e-01 5.07014811e-01 3.24368358e-01
4.50278878e-01 4.71865356e-01 -2.40804181e-01 -5.39683938e-01
-9.19573724e-01 3.66498064e-03 -1.13814071e-01 3.07835490e-01
-2.07441136e-01 -1.90477401e-01 -3.24468553e-01 9.96011868e-02
-7.68817365e-02 3.18056852e-01 3.70987236e-01 1.68972826e+00
-3.68027985e-01 -1.29960859e+00 -4.49501067e-01 4.87248600e-01
-1.01758562e-01 1.30528659e-01 -4.67357248e-01 9.48255956e-01
6.43970892e-02 1.10582626e+00 -3.07765871e-01 -2.12591082e-01
4.60552722e-01 6.48569241e-02 7.72671938e-01 -8.70182931e-01
-5.59994280e-01 1.64013863e-01 4.01818842e-01 -8.39452446e-01
-1.56208664e-01 -8.65038097e-01 -1.75769162e+00 -3.89963150e-01
-5.02926409e-01 6.02894723e-01 2.81883746e-01 1.28319597e+00
3.13062400e-01 7.18889117e-01 4.69054431e-01 -2.49259859e-01
-6.25121474e-01 -3.68665218e-01 -7.79196918e-01 -3.16454582e-02
4.95172411e-01 -8.56674671e-01 -2.56738007e-01 -6.67120442e-02] | [5.419331073760986, 2.7204294204711914] |
c21da666-90b7-4a3b-8fd7-404a258aee07 | instructedit-improving-automatic-masks-for | 2305.18047 | null | https://arxiv.org/abs/2305.18047v1 | https://arxiv.org/pdf/2305.18047v1.pdf | InstructEdit: Improving Automatic Masks for Diffusion-based Image Editing With User Instructions | Recent works have explored text-guided image editing using diffusion models and generated edited images based on text prompts. However, the models struggle to accurately locate the regions to be edited and faithfully perform precise edits. In this work, we propose a framework termed InstructEdit that can do fine-grained editing based on user instructions. Our proposed framework has three components: language processor, segmenter, and image editor. The first component, the language processor, processes the user instruction using a large language model. The goal of this processing is to parse the user instruction and output prompts for the segmenter and captions for the image editor. We adopt ChatGPT and optionally BLIP2 for this step. The second component, the segmenter, uses the segmentation prompt provided by the language processor. We employ a state-of-the-art segmentation framework Grounded Segment Anything to automatically generate a high-quality mask based on the segmentation prompt. The third component, the image editor, uses the captions from the language processor and the masks from the segmenter to compute the edited image. We adopt Stable Diffusion and the mask-guided generation from DiffEdit for this purpose. Experiments show that our method outperforms previous editing methods in fine-grained editing applications where the input image contains a complex object or multiple objects. We improve the mask quality over DiffEdit and thus improve the quality of edited images. We also show that our framework can accept multiple forms of user instructions as input. We provide the code at https://github.com/QianWangX/InstructEdit. | ['Peter Wonka', 'Michael Birsak', 'Biao Zhang', 'Qian Wang'] | 2023-05-29 | null | null | null | null | ['text-guided-image-editing'] | ['computer-vision'] | [ 4.36279476e-01 4.07366678e-02 2.55568653e-01 -4.17801708e-01
-6.56617343e-01 -6.51490033e-01 5.65088332e-01 -1.20267812e-02
-5.50452948e-01 3.73045564e-01 -1.78272307e-01 -4.09374863e-01
2.53218919e-01 -7.22768247e-01 -6.46962523e-01 -3.67090911e-01
5.65976679e-01 5.69163859e-01 7.56958187e-01 -4.18834761e-02
7.62780547e-01 4.63434488e-01 -1.22182822e+00 5.22889137e-01
1.39522982e+00 6.58066154e-01 8.75775039e-01 1.04600787e+00
-6.50835454e-01 6.09308064e-01 -7.45768487e-01 -1.09076962e-01
3.24323177e-01 -6.32507145e-01 -7.27233708e-01 1.16745204e-01
4.20911103e-01 -6.00567222e-01 1.60614941e-02 1.21981180e+00
5.64724207e-01 1.81810632e-01 6.34471416e-01 -1.11954153e+00
-8.15525770e-01 5.78972816e-01 -7.62972951e-01 2.56180614e-01
2.54539490e-01 4.32193130e-01 1.85848057e-01 -9.83810484e-01
7.93058395e-01 1.23367822e+00 1.56166896e-01 6.58026159e-01
-8.99596989e-01 -7.24501371e-01 2.82064617e-01 -1.51561543e-01
-1.24200773e+00 -3.88963431e-01 5.79376101e-01 -6.91385031e-01
8.42882991e-01 3.43781918e-01 4.50564444e-01 4.76310611e-01
5.53440392e-01 5.53145468e-01 1.27972698e+00 -7.26039171e-01
4.35191132e-02 1.92642167e-01 3.68835390e-01 9.96733427e-01
-4.44828451e-01 8.76639932e-02 -1.78641006e-01 1.40796915e-01
1.31262529e+00 -1.77998766e-01 -3.41802418e-01 7.66322091e-02
-1.33274925e+00 6.65663004e-01 1.23532325e-01 3.26828659e-01
-4.93808031e-01 1.50205772e-02 2.57740587e-01 3.02026987e-01
6.34763658e-01 2.67295003e-01 -1.16039984e-01 -1.88042864e-01
-1.41459608e+00 8.28337744e-02 6.22575223e-01 1.13685179e+00
7.83155680e-01 -9.00863409e-02 -7.28622854e-01 8.82598698e-01
1.58304498e-01 4.00687337e-01 4.06607300e-01 -1.03947365e+00
5.13419271e-01 4.92154360e-01 6.71077371e-02 -9.04579878e-01
6.51215687e-02 -3.41624208e-02 -4.47818130e-01 6.97667539e-01
3.15953732e-01 -2.40227923e-01 -1.36350524e+00 1.40469718e+00
4.00563508e-01 1.00418665e-02 -4.00377303e-01 1.00863397e+00
7.86695540e-01 7.90593088e-01 1.44471243e-01 -4.15019803e-02
1.37126160e+00 -1.43572092e+00 -8.66144478e-01 -2.13798597e-01
4.07502681e-01 -1.28382850e+00 1.36632395e+00 3.41752917e-01
-1.33597147e+00 -7.21365571e-01 -8.77360284e-01 -3.33943397e-01
-4.27455306e-01 4.61826921e-01 6.80316016e-02 3.07334304e-01
-1.35808027e+00 5.81380963e-01 -8.33070934e-01 -4.11904573e-01
2.73338705e-01 3.53219688e-01 -6.45229146e-02 2.22811587e-02
-7.40531921e-01 8.69805634e-01 3.09306443e-01 -1.31283462e-01
-7.39645362e-01 -6.82882667e-01 -7.20309198e-01 -1.07680209e-01
4.48708594e-01 -7.02407360e-01 1.49028504e+00 -1.31863034e+00
-1.72201252e+00 9.34449732e-01 -2.49650955e-01 -1.56188339e-01
8.66744757e-01 -7.85613582e-02 -6.84847683e-02 3.21165919e-01
2.15473920e-01 1.06721807e+00 1.12863755e+00 -1.27046931e+00
-6.42903447e-01 -1.25904262e-01 4.64237928e-02 2.56327420e-01
1.51523232e-01 3.09689552e-01 -1.21535873e+00 -9.57325339e-01
-3.64283055e-01 -7.93486893e-01 -2.48054788e-01 1.86942980e-01
-6.89554870e-01 -2.84014158e-02 1.13361728e+00 -9.18993771e-01
1.32199347e+00 -2.27476883e+00 1.02280758e-01 2.85523623e-01
3.56002569e-01 3.77463609e-01 -4.04826283e-01 2.07596526e-01
-1.22128680e-01 1.83878273e-01 -4.71961856e-01 -4.94718999e-01
-2.18371786e-02 1.37313440e-01 -1.31224677e-01 7.40486160e-02
-2.07386725e-02 9.40732896e-01 -6.34050906e-01 -8.24280322e-01
4.11041647e-01 3.95697713e-01 -4.72150892e-01 5.16349077e-01
-3.78563285e-01 5.08411050e-01 -3.48240107e-01 1.70272052e-01
7.43893921e-01 4.60452624e-02 -3.03442895e-01 -2.34444976e-01
-4.55306262e-01 5.65361581e-04 -1.09618115e+00 1.74590147e+00
-4.25119817e-01 5.25873244e-01 4.68652099e-01 -5.66539049e-01
9.26367164e-01 1.16410524e-01 2.26831585e-02 -6.81771874e-01
6.14867806e-02 8.96071494e-02 -1.37088805e-01 -5.31796038e-01
6.64987624e-01 2.32327342e-01 1.27651602e-01 1.03194988e+00
-1.06888354e-01 -5.59692323e-01 4.46721375e-01 6.08288467e-01
8.57981026e-01 3.64784807e-01 1.81338295e-01 -2.22181439e-01
5.54913640e-01 1.70421034e-01 2.52168983e-01 8.54923368e-01
-4.01392207e-02 7.43943572e-01 4.14825350e-01 -3.73728424e-02
-1.05657279e+00 -7.92326748e-01 3.07901531e-01 1.24609578e+00
1.15690686e-01 -3.18950713e-01 -1.50436366e+00 -7.49598980e-01
-3.82773429e-01 9.36767876e-01 -5.28591096e-01 1.59751162e-01
-5.45975745e-01 -1.07119471e-01 2.63920397e-01 3.81150097e-01
6.00080967e-01 -1.50974274e+00 -6.97266817e-01 1.93861172e-01
-8.72046966e-03 -8.99165988e-01 -1.49167275e+00 -1.12933822e-01
-6.66163445e-01 -9.28865790e-01 -8.99386108e-01 -9.23552811e-01
1.28921854e+00 1.31036982e-01 8.83971632e-01 4.80025113e-01
-2.15044603e-01 5.82085907e-01 -2.86648601e-01 -3.20076048e-01
-8.70195210e-01 7.23677427e-02 -3.62362385e-01 -2.70482805e-03
-9.65913851e-03 -2.26827681e-01 -6.40486419e-01 4.57839072e-01
-1.39566803e+00 6.94118440e-01 4.88690615e-01 4.06492531e-01
8.47172558e-01 -1.73267588e-01 2.85243541e-01 -1.34155786e+00
1.17423677e+00 5.57817854e-02 -8.42168748e-01 3.01225036e-01
-3.20766777e-01 2.74865389e-01 6.86500728e-01 -4.90670264e-01
-1.29425406e+00 1.78811580e-01 -2.97516465e-01 -3.76978189e-01
-3.05017173e-01 3.18697482e-01 7.95310736e-03 -1.62245438e-01
5.38297176e-01 2.43619293e-01 4.92410325e-02 -4.56637532e-01
6.25362933e-01 8.60210240e-01 6.98029220e-01 -5.96146345e-01
7.33586669e-01 9.53892618e-02 -7.93207228e-01 -5.45768559e-01
-3.35975379e-01 -7.83043057e-02 -9.21217680e-01 -3.27669710e-01
1.06149113e+00 -3.69730562e-01 -1.12983167e-01 7.31715441e-01
-1.51895928e+00 -9.02905762e-01 -2.97138423e-01 1.42853528e-01
-5.25583446e-01 4.15290534e-01 -7.80875385e-01 -3.87645274e-01
-6.16547585e-01 -1.64516354e+00 9.85581219e-01 4.76823270e-01
-2.76436985e-01 -9.07567203e-01 -1.54430896e-01 1.76602051e-01
6.15192235e-01 1.31170586e-01 9.20234859e-01 -4.73541737e-01
-8.00831974e-01 1.44352973e-01 -4.04455245e-01 2.73604780e-01
2.62065679e-01 3.64873767e-01 -6.67262614e-01 5.58523089e-02
-1.64056066e-02 1.06652409e-01 6.30510509e-01 4.38609809e-01
1.26567650e+00 -1.29975051e-01 -3.49869609e-01 5.62709570e-01
1.09818041e+00 4.68055993e-01 6.59874380e-01 1.33416191e-01
7.21941769e-01 6.04307711e-01 7.77628243e-01 3.21663939e-03
2.39963666e-01 6.74493194e-01 -5.83813936e-02 -3.72737944e-01
-5.04405022e-01 -1.81550205e-01 4.25153583e-01 1.03849721e+00
1.59010842e-01 -6.99599013e-02 -8.03215921e-01 4.79865909e-01
-1.79646969e+00 -7.19017327e-01 -2.44546786e-01 1.90643406e+00
1.12307143e+00 1.08262740e-01 -2.09284976e-01 -3.22391182e-01
8.59439313e-01 5.17476816e-03 -7.43750215e-01 -8.58902037e-01
3.90102625e-01 3.42375249e-01 2.84791976e-01 1.06122577e+00
-7.77155936e-01 1.43387020e+00 5.27479553e+00 9.67074931e-01
-1.41327083e+00 3.10651302e-01 7.38913119e-01 9.54974517e-02
-2.57151484e-01 -1.92073673e-01 -7.08159089e-01 5.15478075e-01
6.60857201e-01 -1.76536292e-01 6.30361795e-01 4.81287926e-01
7.90554941e-01 -4.90330517e-01 -1.02651536e+00 7.92161763e-01
1.38730943e-01 -1.29975152e+00 2.01763406e-01 -8.48633647e-02
6.56334996e-01 -1.64574921e-01 2.48100422e-02 7.99536407e-02
3.22221130e-01 -8.85334671e-01 9.31700647e-01 7.35034585e-01
1.19446814e+00 -5.57243943e-01 2.36564428e-01 5.23659170e-01
-1.06346381e+00 4.24371153e-01 -1.56084284e-01 2.36847043e-01
2.39976570e-01 6.57584906e-01 -8.15940320e-01 1.47223949e-01
3.73515069e-01 3.41018081e-01 -6.28472984e-01 9.90364730e-01
-4.20534283e-01 3.72585505e-01 -2.47848883e-01 3.75518531e-01
1.76556438e-01 -3.82738352e-01 3.47381920e-01 1.47120416e+00
2.73162872e-01 3.04944366e-01 3.86645555e-01 1.14560270e+00
-5.70574589e-02 2.52736032e-01 -4.16978747e-01 -8.65468904e-02
2.58375257e-01 1.36258066e+00 -1.06862640e+00 -6.94872081e-01
-2.23440066e-01 1.68323922e+00 2.68558174e-01 5.95492184e-01
-9.59128022e-01 -7.97497392e-01 1.68792143e-01 1.79771706e-01
3.60651374e-01 -2.96231419e-01 -2.89128244e-01 -8.62620771e-01
-4.06960174e-02 -1.14516187e+00 3.11801676e-02 -1.11606967e+00
-8.36122692e-01 9.66750741e-01 1.43355578e-01 -6.65772021e-01
-1.49075583e-01 -4.48910952e-01 -9.29041445e-01 1.27590394e+00
-1.28841484e+00 -9.18446302e-01 -4.54724789e-01 7.66855299e-01
1.06015515e+00 2.57799774e-01 4.93096650e-01 2.92826533e-01
-6.42452300e-01 3.78044426e-01 -2.71044314e-01 2.22350657e-01
9.47002351e-01 -1.36007226e+00 5.37361622e-01 9.87817824e-01
-6.31231302e-03 6.20567560e-01 5.05851507e-01 -1.04484332e+00
-7.82765031e-01 -1.20987928e+00 7.05644846e-01 -3.51132184e-01
2.39285588e-01 -4.01000708e-01 -1.00213408e+00 7.32629657e-01
7.39008248e-01 -1.84233278e-01 2.41751716e-01 -8.64291549e-01
1.62091240e-01 2.46887133e-01 -1.21203673e+00 6.85610056e-01
7.64387906e-01 -4.73086745e-01 -5.26748896e-01 2.55953163e-01
7.09958494e-01 -9.56471443e-01 -5.36000252e-01 -3.18033457e-01
1.86328083e-01 -7.63350904e-01 5.56605816e-01 1.43468678e-02
3.62758040e-01 -5.52164018e-01 5.02829373e-01 -1.53648973e+00
-1.09634548e-01 -7.58288860e-01 2.97628045e-01 1.30193853e+00
5.30354261e-01 -5.04395366e-01 3.00790161e-01 8.73573661e-01
-2.44724467e-01 -6.28516257e-01 -4.38932210e-01 -2.54755735e-01
-1.01888394e-02 -4.03840721e-01 4.88362104e-01 6.83477998e-01
-3.18124056e-01 2.02035502e-01 -2.10045427e-02 7.01858550e-02
4.64634061e-01 1.16603203e-01 7.79820800e-01 -7.42917895e-01
-2.15067744e-01 -3.68923813e-01 3.97580564e-01 -1.40016615e+00
-1.55042103e-02 -8.46527219e-01 2.88486689e-01 -1.75791788e+00
9.95258242e-02 -4.44340438e-01 3.09664994e-01 6.05954289e-01
-3.77446532e-01 1.98225796e-01 5.32405972e-01 2.60830998e-01
-4.70046818e-01 1.64705083e-01 1.65563345e+00 -1.54330954e-01
-5.68142474e-01 -2.15728521e-01 -6.24013364e-01 7.10449219e-01
1.00331903e+00 -5.23582101e-01 -4.63942319e-01 -7.47683048e-01
-4.07260269e-01 -1.21912301e-01 2.58146256e-01 -8.19427788e-01
4.87117648e-01 -2.11589783e-01 3.54057699e-01 -5.32391608e-01
9.00568999e-03 -7.00612247e-01 6.69243932e-02 2.91668296e-01
-4.32069093e-01 3.89899671e-01 4.14362222e-01 4.74581346e-02
-4.38271016e-02 -4.69564199e-01 8.69867623e-01 -3.84261996e-01
-6.65481448e-01 3.11482906e-01 -5.98288357e-01 -1.68992560e-02
1.22502887e+00 -3.10893565e-01 -3.33683103e-01 -4.25885439e-01
-8.69671702e-01 2.75105774e-01 7.17937827e-01 3.40248942e-01
8.09246361e-01 -9.82782543e-01 -6.06752157e-01 4.73336637e-01
-3.18024158e-01 7.93801919e-02 1.27281040e-01 9.46631908e-01
-9.66754556e-01 1.70712844e-01 -1.96103245e-01 -4.77869362e-01
-1.42938256e+00 3.85420144e-01 5.06319821e-01 -3.23763460e-01
-6.66872203e-01 6.62657022e-01 5.25539994e-01 -5.11779010e-01
4.01622839e-02 -8.12335014e-01 7.46318512e-03 -2.47663289e-01
7.81847596e-01 -6.95601664e-03 -6.10457212e-02 -5.90216041e-01
-4.52540666e-02 8.13311458e-01 -4.39416796e-01 -4.87636805e-01
1.01935947e+00 -2.83128709e-01 -2.36420080e-01 1.10157743e-01
8.67296219e-01 2.50191838e-01 -1.47221684e+00 -4.15605791e-02
-2.80890793e-01 -4.41735744e-01 1.88887537e-01 -1.13163388e+00
-1.16555715e+00 1.00411940e+00 6.43033266e-01 4.77152430e-02
1.21920359e+00 -2.03597724e-01 9.26696956e-01 -2.53709048e-01
2.31373668e-01 -1.12948942e+00 -6.82482645e-02 5.24882793e-01
1.11037922e+00 -8.97507906e-01 -3.44721258e-01 -5.51933050e-01
-8.75861049e-01 1.03162038e+00 8.99017155e-01 -6.04656152e-02
4.81756300e-01 6.00883186e-01 5.96425235e-01 -2.52602905e-01
-4.90144312e-01 1.10298835e-01 3.53514344e-01 7.02177823e-01
4.21670586e-01 -9.67523083e-02 -4.61351216e-01 4.44403827e-01
-1.31269768e-01 3.31819564e-01 5.05999684e-01 8.56603444e-01
-5.75642884e-01 -1.32692933e+00 -7.36872911e-01 5.24337351e-01
-3.57044667e-01 -4.19299036e-01 -5.71229100e-01 4.33095038e-01
1.67476684e-01 8.00619543e-01 1.40138790e-01 -1.58263948e-02
4.11939085e-01 7.64100328e-02 3.51926893e-01 -9.44579720e-01
-1.05424941e+00 1.68988451e-01 -1.98518962e-01 -6.71029389e-01
-1.46884322e-01 -1.70127243e-01 -1.63446248e+00 -2.46469274e-01
-1.00538753e-01 2.27938473e-01 6.19126439e-01 8.53001773e-01
6.20586336e-01 6.70249939e-01 1.41222224e-01 -1.02433074e+00
-6.44434318e-02 -9.73360717e-01 -2.44877368e-01 3.04737419e-01
1.32397786e-01 -2.70242780e-01 5.13647124e-02 4.75649685e-01] | [11.265953063964844, -0.52339106798172] |
f2d03543-9b5f-4b44-ad0d-cd91efcc13d3 | one-shot-recognition-of-any-material-anywhere | 2212.00648 | null | https://arxiv.org/abs/2212.00648v4 | https://arxiv.org/pdf/2212.00648v4.pdf | One-shot recognition of any material anywhere using contrastive learning with physics-based rendering | Visual recognition of materials and their states is essential for understanding most aspects of the world, from determining whether food is cooked, metal is rusted, or a chemical reaction has occurred. However, current image recognition methods are limited to specific classes and properties and can't handle the vast number of material states in the world. To address this, we present MatSim: the first dataset and benchmark for computer vision-based recognition of similarities and transitions between materials and textures, focusing on identifying any material under any conditions using one or a few examples. The dataset contains synthetic and natural images. The synthetic images were rendered using giant collections of textures, objects, and environments generated by computer graphics artists. We use mixtures and gradual transitions between materials to allow the system to learn cases with smooth transitions between states (like gradually cooked food). We also render images with materials inside transparent containers to support beverage and chemistry lab use cases. We use this dataset to train a siamese net that identifies the same material in different objects, mixtures, and environments. The descriptor generated by this net can be used to identify the states of materials and their subclasses using a single image. We also present the first few-shot material recognition benchmark with images from a wide range of fields, including the state of foods and drinks, types of grounds, and many other use cases. We show that a net trained on the MatSim synthetic dataset outperforms state-of-the-art models like Clip on the benchmark and also achieves good results on other unsupervised material classification tasks. | ['Alan Aspuru-Guzik', 'Han Hao', 'Jolina Li', 'Sagi Eppel', 'Manuel S. Drehwald'] | 2022-12-01 | null | null | null | null | ['material-classification', 'material-recognition', 'one-shot-learning'] | ['computer-vision', 'computer-vision', 'methodology'] | [ 5.05733311e-01 -6.02465332e-01 -2.80489862e-01 -1.76313639e-01
-1.74448162e-01 -7.95434833e-01 7.99818695e-01 3.16215456e-01
7.45519772e-02 5.70838079e-02 -1.72587052e-01 9.66841653e-02
1.42411903e-01 -8.94009888e-01 -9.46026921e-01 -9.43313718e-01
-2.54094470e-02 6.17475569e-01 3.30970913e-01 -4.47331995e-01
3.40201259e-01 7.40851164e-01 -1.99842525e+00 8.82371187e-01
4.31583703e-01 1.15469480e+00 2.22721741e-01 7.58696437e-01
-4.00861055e-01 5.66940367e-01 -3.07385087e-01 -1.46627668e-02
5.28211534e-01 -2.10210308e-01 -5.74132264e-01 5.25563538e-01
1.08725619e+00 -1.33992836e-01 -1.82026908e-01 1.25999427e+00
1.70614213e-01 1.64180532e-01 1.06817126e+00 -1.14105618e+00
-9.31122422e-01 4.21997309e-01 -3.84572864e-01 1.30510284e-02
8.85852933e-01 3.58089209e-01 3.73497099e-01 -6.95024490e-01
8.18654120e-01 1.35888863e+00 5.45204043e-01 5.98144233e-01
-1.42932951e+00 -4.48859870e-01 1.30784720e-01 3.87039185e-01
-9.81121540e-01 -3.21576327e-01 8.17369580e-01 -6.50578737e-01
1.01124060e+00 4.82808143e-01 8.07673633e-01 1.34563291e+00
2.82190114e-01 1.17436194e+00 1.26166070e+00 -3.33552867e-01
4.48247075e-01 8.94705802e-02 3.94584030e-01 7.86106706e-01
2.39600554e-01 2.68303193e-02 -4.03518200e-01 -7.28329644e-03
5.98458588e-01 1.78969011e-01 3.08110956e-02 -7.25642562e-01
-1.41287994e+00 3.76978844e-01 2.22744301e-01 1.41232550e-01
-1.99047670e-01 -2.57796437e-01 4.62872982e-01 3.00156981e-01
6.56327829e-02 5.99602282e-01 -2.60976046e-01 7.23976567e-02
-6.65494323e-01 1.35357276e-01 1.26515138e+00 1.03085792e+00
6.96503460e-01 1.09698907e-01 1.47716813e-02 1.01181602e+00
-1.52888104e-01 7.34284222e-01 3.40348721e-01 -7.50195801e-01
2.67476529e-01 7.52476037e-01 2.51551837e-01 -1.06047416e+00
-4.41850811e-01 3.53144765e-01 -8.39930832e-01 4.77539271e-01
6.39056981e-01 5.65199852e-01 -1.43578196e+00 1.09109092e+00
3.49121302e-01 -2.24498864e-02 1.20383151e-01 7.48621762e-01
1.19166481e+00 9.34519768e-01 4.56434451e-02 2.58367091e-01
1.51178777e+00 -8.98629129e-01 -5.31434894e-01 -1.26719669e-01
1.96137801e-01 -1.10726249e+00 1.20786941e+00 7.92888522e-01
-1.05739355e+00 -7.26368546e-01 -1.29148948e+00 1.56184375e-01
-1.14545310e+00 -2.62041707e-02 6.93802357e-01 6.68749273e-01
-6.03464901e-01 9.22469318e-01 -1.03060949e+00 -7.31946588e-01
3.06054145e-01 1.75743029e-01 -3.67779523e-01 -4.06773001e-01
-6.27325237e-01 8.21500123e-01 3.72707218e-01 1.90530777e-01
-1.12688243e+00 -7.22725332e-01 -1.25503087e+00 -2.81075835e-01
3.80001701e-02 -2.44708031e-01 8.25573862e-01 -1.12958193e+00
-1.49930358e+00 1.11856794e+00 2.01110944e-01 -5.77542968e-02
5.70185542e-01 6.71354607e-02 -8.21161985e-01 2.07742155e-01
4.76826206e-02 6.76902652e-01 8.30473781e-01 -1.38617241e+00
-2.70564318e-01 -8.54895562e-02 -1.51278555e-01 -2.83704661e-02
3.89553024e-03 8.22215900e-02 -3.24264139e-01 -3.66191953e-01
1.26698598e-01 -1.08073783e+00 7.48670846e-02 2.46000171e-01
-7.37899244e-01 -1.15870661e-03 8.72437716e-01 -4.93707210e-01
3.81186813e-01 -2.26071835e+00 -2.31046796e-01 1.15012057e-01
-4.67363149e-01 1.91314057e-01 -3.86240304e-01 2.70351082e-01
-1.98250100e-01 -8.08603242e-02 -2.18128171e-02 1.69651374e-01
2.69867688e-01 3.38541090e-01 2.99060829e-02 6.08571947e-01
3.59257728e-01 5.39528310e-01 -1.09200799e+00 -3.12841237e-01
5.51058769e-01 4.09875333e-01 -1.58642396e-01 6.23284243e-02
-4.98115003e-01 1.76776513e-01 -2.10367352e-01 1.08311951e+00
9.50603187e-01 -2.08268464e-01 1.41650870e-01 -7.98959076e-01
-1.89474985e-01 -9.36046336e-03 -1.61523962e+00 1.61892688e+00
-1.45200565e-01 3.57376009e-01 -4.24951836e-02 -8.46392334e-01
7.65674055e-01 5.51164895e-03 6.11932635e-01 -8.24414849e-01
1.84546225e-02 7.12902695e-02 -4.06591967e-02 -7.69711196e-01
5.87847173e-01 2.32099876e-01 -8.35856795e-02 4.69751209e-01
-1.31377116e-01 -5.38006365e-01 7.64346123e-01 -8.47796574e-02
9.61515784e-01 5.04329562e-01 -5.63029274e-02 -6.22484863e-01
5.40278899e-03 3.54987174e-01 2.35317498e-01 7.79388309e-01
-1.71146661e-01 7.63371587e-01 -1.91952172e-03 -1.00624025e+00
-1.28451657e+00 -1.56213748e+00 -2.21105158e-01 1.26923037e+00
6.58270121e-01 6.88817799e-02 -2.94748396e-01 -3.10554415e-01
1.13752045e-01 5.59754908e-01 -8.89718473e-01 -3.47629964e-01
-5.87201416e-01 -8.74605417e-01 4.16815393e-02 5.53677261e-01
6.41816735e-01 -1.40579796e+00 -5.85982382e-01 1.06053621e-01
1.01331733e-01 -1.08653009e+00 -5.09611785e-01 3.82172734e-01
-5.09447515e-01 -1.43061984e+00 -5.14502347e-01 -1.18125212e+00
7.07670867e-01 2.55568147e-01 1.47162485e+00 -3.30594271e-01
-7.32645273e-01 4.93908018e-01 -1.73831299e-01 -3.54935706e-01
-9.71741140e-01 -3.07177842e-01 6.54259846e-02 1.16061747e-01
2.74454415e-01 -1.85193226e-01 -7.99808621e-01 6.53008997e-01
-1.14572656e+00 2.10563630e-01 2.25030288e-01 5.26506484e-01
6.01157486e-01 9.35980976e-02 -1.27425939e-02 -7.01427042e-01
2.49909773e-01 -3.97402942e-01 -3.46474916e-01 7.16697097e-01
-4.05105390e-02 2.45815203e-01 6.60645187e-01 -1.16142452e+00
-1.04398739e+00 4.11861122e-01 2.89879858e-01 -3.19691509e-01
-6.22147262e-01 -1.79361388e-01 -2.33080551e-01 -1.83040649e-02
8.78034949e-01 2.89392978e-01 -1.99244812e-01 -4.56952512e-01
3.78216445e-01 4.12749171e-01 6.89631283e-01 -1.04941392e+00
3.90225589e-01 6.71011388e-01 -3.38485166e-02 -9.78280902e-01
-6.85658097e-01 -5.37023127e-01 -5.17635226e-01 -2.57091612e-01
8.93378735e-01 -5.92846274e-01 -7.03703642e-01 7.95111954e-01
-7.49334216e-01 -7.56509066e-01 -3.65381777e-01 2.09217906e-01
-5.43061554e-01 2.85800874e-01 -1.04739285e+00 -4.57804650e-01
-1.29836500e-01 -1.35144889e+00 1.07707989e+00 3.04029197e-01
-1.68817326e-01 -8.96063685e-01 -6.41939491e-02 -4.63559814e-02
3.95956665e-01 6.66688204e-01 1.01455104e+00 -2.84657031e-01
-3.70876640e-01 3.40745077e-02 -1.13527454e-01 2.22910076e-01
7.58502126e-01 5.21497369e-01 -7.69121706e-01 -1.56039923e-01
-3.12562406e-01 -4.35707390e-01 8.79230678e-01 1.91547930e-01
1.08861697e+00 -1.75929502e-01 -2.39289999e-01 5.26471794e-01
1.38169134e+00 7.03171432e-01 8.41298819e-01 3.80913168e-01
7.44877338e-01 5.60835302e-01 3.37535292e-01 2.37369165e-01
-2.04375223e-03 5.38743556e-01 4.51442033e-01 -2.90803999e-01
-3.05137962e-01 3.07741016e-02 6.17513597e-01 6.93025112e-01
-3.27136926e-02 5.32360561e-02 -6.53202415e-01 4.27317113e-01
-1.41328239e+00 -1.10215569e+00 -2.25112140e-02 2.15956926e+00
9.02258635e-01 9.29376781e-02 1.75698727e-01 -5.46322167e-02
8.75504911e-01 -8.22560713e-02 -8.83829176e-01 -6.94217265e-01
-2.52011657e-01 3.30361784e-01 4.67488557e-01 -1.11809671e-01
-1.46025455e+00 7.08264291e-01 7.29652119e+00 7.19097733e-01
-1.32362068e+00 -4.04941648e-01 7.12578893e-01 1.15394264e-01
6.10037334e-02 -6.01110041e-01 -6.87704802e-01 5.03294587e-01
5.73216617e-01 5.17441869e-01 7.02801347e-01 9.49697614e-01
-1.78150088e-01 -1.44291058e-01 -1.56829286e+00 1.06984234e+00
3.90290201e-01 -1.32087028e+00 1.15705639e-01 -4.31043386e-01
1.06082904e+00 -1.04973234e-01 1.35169372e-01 9.94204804e-02
7.41103172e-01 -7.79929757e-01 1.12943041e+00 5.06946146e-01
6.86029673e-01 -7.38057345e-02 1.95103481e-01 3.50047299e-03
-1.39605582e+00 1.37536153e-01 -4.32736903e-01 2.96039075e-01
-2.23130733e-01 1.95331424e-01 -6.56770468e-01 2.51177281e-01
9.33954179e-01 7.33301818e-01 -6.39504433e-01 1.02349758e+00
2.77905256e-01 2.53282219e-01 -6.76515698e-01 -2.64824152e-01
8.63170996e-02 -4.53189552e-01 1.94849670e-01 1.51201940e+00
2.03718349e-01 -3.22898805e-01 5.30043840e-01 9.86843765e-01
2.51611441e-01 -2.30264664e-02 -7.31052756e-01 -4.33830731e-02
-5.05859293e-02 1.34054387e+00 -1.26698732e+00 -6.40521526e-01
-2.13284597e-01 9.98384416e-01 -1.47430867e-01 2.82365769e-01
-6.51979029e-01 -7.97808319e-02 4.95670646e-01 8.35943148e-02
3.16721320e-01 -1.57944530e-01 1.90254056e-03 -1.22351217e+00
-1.73428431e-01 -1.01053715e+00 3.46683830e-01 -1.00138927e+00
-1.74779177e+00 3.51157159e-01 1.55774370e-01 -1.48233056e+00
1.15872577e-01 -1.20848989e+00 -5.43077111e-01 3.10680300e-01
-1.16537392e+00 -1.04552579e+00 -6.36389196e-01 5.89068711e-01
6.68140888e-01 1.91084910e-02 8.69084954e-01 2.10758448e-01
-5.90330064e-01 8.01671520e-02 3.08808059e-01 2.41369754e-01
6.28070533e-01 -1.36217213e+00 4.77649301e-01 3.70958388e-01
1.20439015e-01 4.65069294e-01 9.16513860e-01 -4.73959804e-01
-1.91192162e+00 -1.01698661e+00 -2.04024184e-03 -4.07831430e-01
5.94908774e-01 -6.82392776e-01 -8.78562748e-01 3.59135985e-01
2.11894035e-01 8.59115422e-02 5.28956532e-01 -1.69391081e-01
-7.98799098e-01 -1.71996564e-01 -1.36051691e+00 7.60779023e-01
9.56647992e-01 -3.55721295e-01 -4.15691853e-01 7.93237805e-01
1.41556710e-01 -5.89513719e-01 -9.69138145e-01 3.11509073e-01
8.57802570e-01 -8.37470114e-01 1.23021913e+00 -7.78274715e-01
5.23975372e-01 -2.73623019e-01 -3.94496143e-01 -1.34297025e+00
-3.27303469e-01 -3.17091435e-01 -1.03122696e-01 9.34948444e-01
2.39322126e-01 -4.44970191e-01 8.72003436e-01 7.72735894e-01
-2.67947644e-01 -4.04743582e-01 -3.06154698e-01 -1.05891538e+00
-5.32899164e-02 -2.15621158e-01 7.45772421e-01 1.07932329e+00
-2.20775068e-01 -8.60846192e-02 -6.25920892e-02 -4.79366370e-02
6.08130217e-01 7.00849533e-01 6.70232356e-01 -9.44382668e-01
-1.98325008e-01 -6.19677305e-01 -4.39961851e-01 -7.25914955e-01
-1.34796813e-01 -9.58485544e-01 2.77817756e-01 -1.59842718e+00
3.58028263e-01 -6.61436379e-01 -3.97195220e-01 5.19325435e-01
2.29727432e-01 3.42127413e-01 2.16170654e-01 2.68606283e-02
-5.65442383e-01 -1.93905495e-02 1.46490097e+00 -1.06862140e+00
-2.55547732e-01 -2.07472533e-01 -3.38056594e-01 6.68613434e-01
7.92731225e-01 -1.66644335e-01 -4.68259454e-02 -2.24343270e-01
3.20169657e-01 -3.64560843e-01 4.18877870e-01 -1.19200528e+00
-1.59697801e-01 -5.90482950e-01 6.28224790e-01 -6.07029140e-01
4.71106797e-01 -8.33162248e-01 6.29342854e-01 5.56333005e-01
-2.96087444e-01 1.04509825e-02 2.26409718e-01 2.13267207e-01
2.23741367e-01 -1.89810097e-01 9.25437927e-01 -5.05994380e-01
-9.63140428e-01 2.70682305e-01 -2.85164088e-01 -1.68266550e-01
9.39141512e-01 -5.12854457e-01 -7.33876526e-01 1.69413120e-01
-9.56658304e-01 -2.38857031e-01 9.11332071e-01 6.00341797e-01
4.84323889e-01 -1.57106292e+00 -3.95542771e-01 5.72151959e-01
3.28397423e-01 -1.14867285e-01 3.29876512e-01 4.01784718e-01
-1.09015131e+00 -1.95064053e-01 -6.86238706e-01 -9.32188570e-01
-1.02909160e+00 9.92562115e-01 3.69597226e-01 -7.19605461e-02
-6.43600881e-01 2.15245619e-01 3.85765970e-01 -3.69757742e-01
5.73946312e-02 -9.58757281e-01 -4.66958880e-02 -5.27525209e-02
6.19764447e-01 2.78828502e-01 2.96466172e-01 -5.33857465e-01
-2.20276803e-01 7.94721067e-01 -8.97407904e-02 5.36145210e-01
1.41684258e+00 5.00097394e-01 1.56534746e-01 7.88126349e-01
1.30789244e+00 -3.13190669e-01 -1.48277307e+00 -9.61145945e-03
-2.53366947e-01 -1.80482373e-01 -6.25358880e-01 -8.35440636e-01
-1.25188673e+00 6.78380311e-01 1.04632413e+00 5.41095197e-01
8.84527087e-01 6.71536475e-03 7.61597574e-01 5.71257710e-01
3.95652145e-01 -1.29772985e+00 4.65691596e-01 3.04236531e-01
8.48432302e-01 -1.29455805e+00 7.65150487e-02 -5.77353239e-01
-5.28449535e-01 1.30795586e+00 5.17269909e-01 -4.03850496e-01
7.32367158e-01 5.80273569e-01 1.55018225e-01 -4.01525229e-01
-3.63293976e-01 -1.86840165e-02 3.20662320e-01 7.82201111e-01
1.29786387e-01 4.44872141e-01 5.32638252e-01 -1.64504364e-01
-6.55380264e-02 -3.41001213e-01 4.09813404e-01 1.32731938e+00
-3.24133754e-01 -9.25411880e-01 -6.18345678e-01 6.45494044e-01
-2.74577379e-01 7.47491699e-03 -1.78075284e-01 6.79765582e-01
2.47927964e-01 7.51503646e-01 3.99723113e-01 -2.04622716e-01
4.71097708e-01 -4.43861894e-02 1.01043618e+00 -5.13334215e-01
-6.60841525e-01 -8.49052146e-02 1.44009605e-01 -6.13610268e-01
-6.89003646e-01 -8.56078684e-01 -9.01444256e-01 -1.41306192e-01
-4.98722382e-02 -3.77789617e-01 7.95210600e-01 5.82570493e-01
1.06107518e-01 3.43606949e-01 4.02473539e-01 -1.44194818e+00
-3.17200869e-01 -7.72827923e-01 -7.76128531e-01 1.35217059e+00
9.64612439e-02 -9.47574914e-01 -2.51326531e-01 6.55473888e-01] | [10.176905632019043, -0.08873289823532104] |
f4b98134-212b-4d0e-8c35-4e9a7d8f101f | back-to-the-feature-learning-robust-camera | 2103.09213 | null | https://arxiv.org/abs/2103.09213v2 | https://arxiv.org/pdf/2103.09213v2.pdf | Back to the Feature: Learning Robust Camera Localization from Pixels to Pose | Camera pose estimation in known scenes is a 3D geometry task recently tackled by multiple learning algorithms. Many regress precise geometric quantities, like poses or 3D points, from an input image. This either fails to generalize to new viewpoints or ties the model parameters to a specific scene. In this paper, we go Back to the Feature: we argue that deep networks should focus on learning robust and invariant visual features, while the geometric estimation should be left to principled algorithms. We introduce PixLoc, a scene-agnostic neural network that estimates an accurate 6-DoF pose from an image and a 3D model. Our approach is based on the direct alignment of multiscale deep features, casting camera localization as metric learning. PixLoc learns strong data priors by end-to-end training from pixels to pose and exhibits exceptional generalization to new scenes by separating model parameters and scene geometry. The system can localize in large environments given coarse pose priors but also improve the accuracy of sparse feature matching by jointly refining keypoints and poses with little overhead. The code will be publicly available at https://github.com/cvg/pixloc. | ['Torsten Sattler', 'Fredrik Kahl', 'Lars Hammarstrand', 'Vincent Lepetit', 'Marc Pollefeys', 'Viktor Larsson', 'Carl Toft', 'Hugo Germain', 'Måns Larsson', 'Ajaykumar Unagar', 'Paul-Edouard Sarlin'] | 2021-03-16 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Sarlin_Back_to_the_Feature_Learning_Robust_Camera_Localization_From_Pixels_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Sarlin_Back_to_the_Feature_Learning_Robust_Camera_Localization_From_Pixels_CVPR_2021_paper.pdf | cvpr-2021-1 | ['camera-localization'] | ['computer-vision'] | [-1.59455732e-01 2.56780796e-02 -1.24706984e-01 -5.22245526e-01
-1.13274455e+00 -8.42344999e-01 5.60641587e-01 -1.94453999e-01
-3.32537025e-01 1.43234387e-01 9.32134241e-02 1.18929952e-01
-1.51898891e-01 -4.43258464e-01 -1.14653659e+00 -4.06695753e-01
2.26347715e-01 1.00556755e+00 6.56645745e-02 1.12577572e-01
3.41641933e-01 9.03594971e-01 -1.23601437e+00 -2.40143910e-01
2.40510434e-01 1.05453992e+00 3.72193098e-01 7.25259364e-01
2.58316308e-01 5.51961124e-01 -7.69994482e-02 -2.03670979e-01
7.58739233e-01 7.11042732e-02 -6.61905527e-01 3.76184821e-01
1.09068751e+00 -5.15825331e-01 -3.16561759e-01 1.07057345e+00
2.96740383e-01 -9.15509984e-02 4.49756891e-01 -1.20586622e+00
-3.17666978e-01 1.27812073e-01 -6.20641530e-01 -2.15253651e-01
5.46822548e-01 2.82262694e-02 1.12189579e+00 -1.10386431e+00
6.99135065e-01 1.29594862e+00 9.88531709e-01 9.53479633e-02
-1.18521845e+00 -2.03113437e-01 2.10502028e-01 7.84188136e-02
-1.67903566e+00 -4.06909913e-01 1.14057302e+00 -7.16867089e-01
8.58427942e-01 3.76424007e-02 5.53743482e-01 9.82506633e-01
2.48438492e-02 5.84868789e-01 7.05035985e-01 -2.96420068e-01
1.87787339e-01 -1.46245793e-01 -3.18983734e-01 8.41898561e-01
1.52021438e-01 8.93318951e-02 -4.45345163e-01 1.97360180e-02
1.17153966e+00 2.50125945e-01 -1.63259327e-01 -1.29545701e+00
-1.37639177e+00 8.20448160e-01 6.72214150e-01 -1.83929563e-01
-2.75372744e-01 5.46456456e-01 -6.83755279e-02 -1.38679389e-02
3.00281048e-01 6.74698710e-01 -8.08800220e-01 5.63772721e-03
-7.39909172e-01 2.63881296e-01 6.56898677e-01 1.22156131e+00
1.20654511e+00 -2.67164409e-01 4.75425363e-01 4.49573696e-01
4.91062939e-01 8.52047503e-01 8.99692774e-02 -1.56676602e+00
2.21588328e-01 4.56053048e-01 2.23355010e-01 -1.18463111e+00
-7.23081470e-01 -4.96133983e-01 -4.80808675e-01 1.49413750e-01
3.16978544e-01 3.79448850e-03 -6.91162527e-01 1.64011824e+00
5.30363858e-01 9.99114662e-02 -3.67605001e-01 1.08543789e+00
5.69505692e-01 3.25885981e-01 -4.69106376e-01 3.43978494e-01
1.05948770e+00 -9.25957739e-01 5.87429591e-02 -6.48196936e-01
4.58065897e-01 -9.74852979e-01 7.74464726e-01 3.60696435e-01
-1.02518868e+00 -5.78375220e-01 -9.35906231e-01 -4.79378581e-01
-1.83266446e-01 1.99391887e-01 4.49997932e-01 2.54595071e-01
-1.25330269e+00 3.13220650e-01 -9.74821270e-01 -5.78463852e-01
3.04394215e-01 5.00616014e-01 -5.99613488e-01 -2.22224519e-01
-5.51503241e-01 1.03137994e+00 2.45902359e-01 9.97716710e-02
-8.91317546e-01 -6.94490910e-01 -1.25540602e+00 -1.29468054e-01
3.96073043e-01 -1.15092754e+00 1.37465596e+00 -7.41517186e-01
-1.39296663e+00 1.01204634e+00 -1.14355341e-01 -2.89301038e-01
6.06216371e-01 -6.61179185e-01 3.23899716e-01 2.40969077e-01
3.25231493e-01 9.72759247e-01 9.18206155e-01 -1.39347649e+00
-3.56164843e-01 -4.88847852e-01 3.62971365e-01 3.09150636e-01
2.02230319e-01 -2.64968067e-01 -5.94192684e-01 -2.55714327e-01
7.70226955e-01 -1.14033282e+00 -3.78446460e-01 4.67670351e-01
-2.72175968e-01 1.23293146e-01 7.59434938e-01 -3.66944015e-01
1.21047385e-01 -1.99261582e+00 4.64614123e-01 5.98166697e-02
2.37521574e-01 -3.07298005e-01 -2.26623222e-01 3.32855821e-01
-9.26852077e-02 -2.93756127e-01 -1.23200230e-01 -6.18541241e-01
1.09406747e-01 7.74428248e-02 -2.06563175e-01 1.10385025e+00
1.77164137e-01 9.40497041e-01 -8.66785109e-01 -2.44705930e-01
6.51165187e-01 6.22923076e-01 -8.01477075e-01 2.46013924e-01
-2.74251014e-01 4.02466059e-01 -2.71147341e-01 7.44247556e-01
9.50079739e-01 -4.59810197e-01 -1.43496946e-01 -5.47421098e-01
-7.80902579e-02 3.39679234e-02 -1.43186450e+00 2.44766736e+00
-6.53280556e-01 5.46768427e-01 2.30702654e-01 -8.98852468e-01
9.83050764e-01 -5.62879816e-02 6.86881125e-01 -2.25238711e-01
3.54441464e-01 8.88793096e-02 -5.07676005e-01 -3.01841468e-01
3.25098932e-01 1.08998537e-01 -2.88988382e-01 1.50762349e-01
3.39701593e-01 -7.98118412e-01 -2.44517371e-01 -2.41230242e-02
8.19178879e-01 7.01027036e-01 5.02180696e-01 -1.59459174e-01
4.06016558e-01 1.96423698e-02 5.48586845e-01 5.18810213e-01
4.19103466e-02 9.46557641e-01 1.42597318e-01 -7.01374531e-01
-1.23924661e+00 -1.27745020e+00 -1.54865146e-01 7.56478250e-01
4.08867985e-01 -3.93613279e-01 -5.05002141e-01 -4.43837076e-01
1.78973362e-01 2.08580330e-01 -4.01173711e-01 4.67196219e-02
-5.96865416e-01 -1.80242330e-01 -7.60945901e-02 4.63140070e-01
3.74747306e-01 -4.60708290e-01 -6.55118525e-01 2.83418298e-02
-2.97592022e-02 -1.42100227e+00 -5.79860806e-01 3.49598199e-01
-7.83901572e-01 -1.24310374e+00 -5.41800916e-01 -7.24036515e-01
7.82981813e-01 5.24786115e-01 1.09696972e+00 -2.97436476e-01
-1.89872533e-01 8.43184888e-01 -1.92033514e-01 -1.71783939e-01
4.79251631e-02 2.48012558e-01 2.40948170e-01 2.16087755e-02
2.69168347e-01 -8.46936643e-01 -6.66400194e-01 4.65024531e-01
-4.48645443e-01 1.03927851e-01 6.53992534e-01 5.49633563e-01
8.89096677e-01 -5.06176531e-01 -1.84295788e-01 -5.56949317e-01
-1.87501848e-01 -2.38996461e-01 -1.07380092e+00 -5.61587177e-02
-1.10661156e-01 9.45103317e-02 4.78447914e-01 -6.51046410e-02
-5.22770166e-01 9.55513120e-01 -7.78472573e-02 -8.73065591e-01
-3.91534597e-01 2.33079672e-01 -3.79080623e-01 -4.57745969e-01
7.09240556e-01 -3.60624939e-02 -6.63871616e-02 -5.95607638e-01
7.23237157e-01 1.13118157e-01 6.85211062e-01 -7.54597306e-01
1.15474355e+00 7.54317105e-01 2.45741054e-01 -6.33274555e-01
-1.14012849e+00 -7.32524633e-01 -1.42924714e+00 -4.49835695e-02
8.85362029e-01 -1.38637841e+00 -6.00338340e-01 2.98094720e-01
-1.39484382e+00 -2.14294076e-01 -2.24179193e-01 6.42254412e-01
-1.01501656e+00 4.60411608e-01 -2.73559660e-01 -2.64282882e-01
-6.11274950e-02 -1.29150879e+00 1.74214220e+00 -8.08255747e-02
-2.86402047e-01 -9.54860151e-01 2.64124423e-01 3.96009654e-01
-3.80358063e-02 4.56497520e-01 2.50780433e-01 -1.37799710e-01
-1.21417868e+00 -4.14151609e-01 -2.02318370e-01 1.59870759e-01
-2.02453099e-02 -3.24515998e-02 -1.15429175e+00 -4.76526886e-01
2.22186074e-01 -4.17666584e-01 7.38545716e-01 6.52393043e-01
1.02664685e+00 -1.68878794e-01 -2.34505430e-01 1.38824868e+00
1.68859816e+00 -3.34503263e-01 8.14176649e-02 4.22302961e-01
1.01094651e+00 4.06346172e-01 3.72446567e-01 3.85748386e-01
5.49047351e-01 8.09478283e-01 9.45878804e-01 -3.58399935e-02
-2.25091781e-02 -4.52233464e-01 1.90914556e-01 7.19326794e-01
2.01705456e-01 1.35550827e-01 -1.04805410e+00 4.07373011e-01
-1.77651465e+00 -6.51324928e-01 1.36500075e-01 2.02119446e+00
3.60332251e-01 2.88919564e-02 -4.13015299e-02 -3.02566260e-01
5.58309436e-01 1.80344909e-01 -7.65607119e-01 1.63883027e-02
-1.14801854e-01 -1.32615760e-01 7.29394138e-01 9.29352462e-01
-1.27493620e+00 1.07097220e+00 5.54105902e+00 2.90511101e-01
-1.09686971e+00 6.08224645e-02 3.75855356e-01 -1.25698820e-01
-2.64735639e-01 3.11021686e-01 -9.38996851e-01 -3.83960344e-02
3.97419453e-01 2.40153223e-01 3.87990206e-01 1.22063994e+00
-1.73757020e-02 4.15151678e-02 -1.48307550e+00 1.45292771e+00
3.64697069e-01 -1.45383823e+00 -2.68047135e-02 9.01821777e-02
8.35064352e-01 5.97523868e-01 -1.04571111e-03 -1.23443767e-01
4.44901198e-01 -8.43312025e-01 1.10426855e+00 6.16763115e-01
6.63043261e-01 -5.86598933e-01 4.51013267e-01 4.49977398e-01
-1.12144542e+00 -3.10622137e-02 -5.41089594e-01 -1.71985075e-01
2.50597626e-01 4.20413345e-01 -8.97237360e-01 3.97327304e-01
6.65080965e-01 1.10188568e+00 -7.62721360e-01 1.34245992e+00
-2.64606297e-01 -1.58551961e-01 -5.84727645e-01 3.23858321e-01
2.63181806e-01 -1.86717078e-01 5.31697631e-01 8.75842750e-01
4.90080386e-01 -2.55663663e-01 4.02829379e-01 9.34008896e-01
-1.62826404e-02 -2.64168203e-01 -7.53488183e-01 5.75405896e-01
4.21137363e-01 1.44361603e+00 -6.34371281e-01 9.93760377e-02
-3.87084424e-01 9.69284534e-01 4.97790128e-01 2.05336586e-01
-7.35618770e-01 -1.42416451e-02 7.68962622e-01 1.68986380e-01
4.39752877e-01 -7.09647298e-01 -3.25299293e-01 -1.33027387e+00
1.51818842e-01 -5.23159564e-01 2.07972825e-02 -1.16023338e+00
-1.20068061e+00 3.76692683e-01 5.77476211e-02 -1.49822152e+00
-2.80050069e-01 -8.66707146e-01 -3.25231731e-01 6.48228884e-01
-1.43607318e+00 -1.34239447e+00 -5.44727027e-01 5.98440051e-01
5.24889708e-01 1.49153948e-01 6.07283533e-01 3.06502078e-02
-7.00577721e-02 3.81795436e-01 1.80849154e-02 2.55733818e-01
7.33029723e-01 -1.26211071e+00 5.24376571e-01 7.65655696e-01
4.35503274e-01 4.38025415e-01 6.36690497e-01 -9.72486809e-02
-1.74179935e+00 -1.04461575e+00 7.75195062e-01 -1.08333850e+00
5.95317483e-01 -6.86133444e-01 -4.73923266e-01 9.10598755e-01
5.85024543e-02 3.16957921e-01 2.46649221e-01 6.33563697e-02
-5.80308080e-01 -2.81353980e-01 -9.21565354e-01 3.87222409e-01
1.17651927e+00 -7.85068095e-01 -3.74072403e-01 6.66699827e-01
8.54575098e-01 -8.32510471e-01 -8.46919179e-01 2.47116327e-01
3.77590269e-01 -9.20564473e-01 1.29201508e+00 -2.71419585e-01
9.66738984e-02 -6.52239323e-01 -6.31206155e-01 -1.08499038e+00
-4.75109816e-01 -5.87027848e-01 6.68172166e-02 7.48456836e-01
2.12442666e-01 -2.90199876e-01 9.36140060e-01 4.89977032e-01
-2.47899279e-01 -4.44197416e-01 -9.87987280e-01 -7.16753602e-01
1.55552328e-02 -6.21192873e-01 6.48684442e-01 8.57770443e-01
-4.51028764e-01 4.06583011e-01 -3.93620938e-01 7.14564860e-01
9.37949002e-01 4.08817708e-01 1.18677735e+00 -1.34688544e+00
-3.49329442e-01 -4.27918434e-01 -6.96993887e-01 -1.61389506e+00
2.17598319e-01 -8.58845770e-01 5.37482724e-02 -1.36899269e+00
1.62009910e-01 -1.96910605e-01 1.86437190e-01 3.35156053e-01
3.35804343e-01 2.74597496e-01 1.99059471e-01 2.49936715e-01
-8.48335505e-01 5.86768627e-01 1.16808128e+00 -1.42100662e-01
-3.92978871e-03 -3.05211917e-03 -5.22541106e-01 1.02653241e+00
7.51104832e-01 -3.95394504e-01 -2.08508089e-01 -9.16735888e-01
6.15974605e-01 8.12682435e-02 8.03099632e-01 -1.08666861e+00
4.39157933e-01 -1.45324796e-01 6.57232225e-01 -7.76854932e-01
6.59987688e-01 -1.14916539e+00 1.99669152e-01 2.26690322e-01
-1.57438755e-01 1.77816316e-01 3.97179052e-02 5.41696966e-01
-4.97677214e-02 -1.28453895e-01 7.01539516e-01 -4.83522326e-01
-9.27599549e-01 6.95251584e-01 2.30538875e-01 -2.04614215e-02
7.77307570e-01 -3.98315489e-01 8.13927781e-03 -4.61741745e-01
-8.10920537e-01 6.50998205e-02 9.48821068e-01 4.04484183e-01
7.10684478e-01 -1.38160515e+00 -5.62243223e-01 2.74397522e-01
2.53988743e-01 6.47943318e-01 5.34329079e-02 7.14039981e-01
-9.69591439e-01 5.25860488e-01 -4.14824970e-02 -1.27701759e+00
-9.80123818e-01 6.64433360e-01 7.12172985e-01 3.91868323e-01
-4.80305761e-01 1.00949705e+00 3.92851561e-01 -9.27342296e-01
2.70996660e-01 -4.99782354e-01 3.14918607e-01 -2.24776268e-01
1.14428401e-01 4.37253304e-02 9.53468122e-03 -9.91262078e-01
-4.44686681e-01 1.32043219e+00 1.26268327e-01 9.63792354e-02
1.41913736e+00 -4.18812782e-01 -1.13471985e-01 2.37424970e-01
1.65632379e+00 2.67155748e-03 -1.95004892e+00 -4.29418504e-01
-8.33508521e-02 -6.82338357e-01 1.08500011e-01 -3.85859132e-01
-1.01531780e+00 8.93426299e-01 4.99726027e-01 -4.21016067e-01
7.13611186e-01 3.91203254e-01 4.98402894e-01 7.06757784e-01
4.82925296e-01 -7.32476830e-01 2.20523745e-01 7.47647524e-01
1.01438820e+00 -1.48012257e+00 2.00766891e-01 -1.78183630e-01
-2.08834589e-01 1.29103136e+00 6.08188927e-01 -5.04153967e-01
8.84447813e-01 1.86084896e-01 1.66555658e-01 -2.47520313e-01
-2.80180484e-01 -9.40489247e-02 4.26375300e-01 7.01034367e-01
6.87597841e-02 -1.06165081e-01 5.93775749e-01 4.72799428e-02
-4.90262717e-01 -3.19665313e-01 3.47451121e-01 6.04200602e-01
-4.58961666e-01 -9.42889750e-01 -4.60835576e-01 -1.30078927e-01
6.74344599e-02 4.88525406e-02 -4.42585975e-01 8.40754330e-01
2.77361393e-01 4.89395350e-01 1.75610855e-01 -2.67832398e-01
2.48762622e-01 -2.75542319e-01 7.29654014e-01 -6.79760933e-01
7.05785900e-02 2.51871616e-01 -3.24520230e-01 -8.86148214e-01
-5.45710683e-01 -1.00373995e+00 -8.27718914e-01 -9.61353183e-02
-9.53432396e-02 -2.38746285e-01 7.19198525e-01 8.01932752e-01
3.63562882e-01 -2.54273936e-02 6.86666667e-01 -1.49619067e+00
-4.59708929e-01 -5.52753925e-01 -3.94761652e-01 2.45780706e-01
6.93787813e-01 -6.53407335e-01 -4.03397083e-01 1.25863299e-01] | [7.765594005584717, -2.484361410140991] |
6e33fe4f-fb8c-4876-a4d8-3df6c271cb4c | audiocaps-generating-captions-for-audios-in | null | null | https://aclanthology.org/N19-1011 | https://aclanthology.org/N19-1011.pdf | AudioCaps: Generating Captions for Audios in The Wild | We explore the problem of Audio Captioning: generating natural language description for any kind of audio in the wild, which has been surprisingly unexplored in previous research. We contribute a large-scale dataset of 46K audio clips with human-written text pairs collected via crowdsourcing on the AudioSet dataset. Our thorough empirical studies not only show that our collected captions are indeed faithful to audio inputs but also discover what forms of audio representation and captioning models are effective for the audio captioning. From extensive experiments, we also propose two novel components that help improve audio captioning performance: the top-down multi-scale encoder and aligned semantic attention. | ['Chris Dongjoo Kim', 'Byeongchang Kim', 'Hyunmin Lee', 'Gunhee Kim'] | 2019-06-01 | null | null | null | naacl-2019-6 | ['audio-captioning'] | ['audio'] | [ 5.62855601e-01 3.53828043e-01 1.07454151e-01 -4.40483272e-01
-1.91426408e+00 -7.81369686e-01 3.57445538e-01 -1.83195651e-01
1.06864177e-01 8.23520839e-01 1.05386484e+00 3.00203592e-01
2.91617453e-01 -3.39495420e-01 -1.11409116e+00 -1.85392827e-01
-2.01902002e-01 5.07033169e-01 1.84147462e-01 -3.56530339e-01
-1.08074971e-01 -3.02521497e-01 -1.64748275e+00 1.09908915e+00
2.07247689e-01 1.30472493e+00 -8.54109451e-02 1.19981217e+00
1.50571028e-02 1.06513476e+00 -7.63292968e-01 -5.45023143e-01
1.38925627e-01 -6.81899786e-01 -1.14588261e+00 2.16477122e-02
6.55423224e-01 -3.72301549e-01 -4.82561439e-01 7.54185379e-01
1.00361109e+00 -9.65935364e-02 1.87503144e-01 -1.55452871e+00
-9.43014085e-01 1.20308185e+00 -2.43062317e-01 2.39129230e-01
1.13318360e+00 7.24526495e-03 1.48897290e+00 -8.93721640e-01
3.23705018e-01 1.27596068e+00 7.41703033e-01 7.55036414e-01
-8.85854244e-01 -9.77624357e-01 -1.16371401e-01 3.62697095e-01
-1.59176016e+00 -8.95183623e-01 8.62501323e-01 -3.08732241e-01
4.59187657e-01 4.39319372e-01 6.50410473e-01 1.68722892e+00
-5.53431392e-01 9.13570404e-01 5.45436382e-01 -3.50330979e-01
1.50995478e-01 6.03966676e-02 -2.91409463e-01 -1.44515876e-02
-4.85865623e-01 -1.77358702e-01 -1.29642999e+00 -4.48251665e-01
6.83613777e-01 -6.36486888e-01 -3.87221336e-01 2.55830973e-01
-1.48852336e+00 6.72939837e-01 1.70698181e-01 6.60427511e-02
-2.38093078e-01 5.31257808e-01 7.94801056e-01 3.34539920e-01
3.09380561e-01 7.03560948e-01 -3.10436457e-01 -4.36010361e-01
-7.16351986e-01 4.09991920e-01 5.41410029e-01 1.41780293e+00
3.41848820e-01 1.27858788e-01 -6.54141843e-01 9.62540686e-01
7.36849979e-02 7.27579653e-01 6.51142478e-01 -1.07678270e+00
8.72358203e-01 -1.38601929e-01 3.30525488e-01 -7.56684482e-01
9.99332666e-02 -1.15358420e-01 -4.83943194e-01 -9.89139199e-01
3.48350815e-02 -3.76410097e-01 -3.48239988e-01 1.84803629e+00
1.15268260e-01 8.10325801e-01 2.14766979e-01 1.21303475e+00
1.14381909e+00 9.15125310e-01 1.03576273e-01 7.22029060e-02
1.61529315e+00 -1.22706747e+00 -9.65092480e-01 -3.15338343e-01
1.40177369e-01 -9.44119632e-01 1.32695973e+00 2.27554977e-01
-1.17693341e+00 -6.91878259e-01 -7.17507958e-01 -1.03802495e-01
2.10302994e-01 8.10666978e-02 5.68284094e-01 1.71767026e-01
-9.30881500e-01 9.75861847e-02 -2.41537333e-01 -1.56360313e-01
1.78085685e-01 -6.36071786e-02 -3.51919234e-01 1.04626536e-01
-1.80263865e+00 1.17782494e-02 2.40107715e-01 1.25545606e-01
-1.39767623e+00 -7.29535937e-01 -8.63663137e-01 -2.01185122e-02
4.08530354e-01 -5.24919748e-01 1.92273176e+00 -1.28742349e+00
-1.41483617e+00 7.03047872e-01 -2.65792012e-01 -7.99805105e-01
1.33410573e-01 -4.96459067e-01 -6.49951100e-01 6.71392322e-01
2.52720237e-01 1.11812329e+00 9.43812132e-01 -1.20611548e+00
-5.11198521e-01 1.79331645e-01 1.32268786e-01 3.64204228e-01
-5.83856046e-01 3.46079469e-01 -3.97655636e-01 -1.01990020e+00
-3.81100535e-01 -8.73650312e-01 1.51880607e-01 -5.89976609e-02
-6.21223569e-01 -4.07105498e-02 2.87863761e-01 -7.99129486e-01
1.22575259e+00 -2.28783107e+00 -6.30223751e-02 -1.97693869e-01
-1.79618731e-01 -1.38194978e-01 -5.36435366e-01 7.41229951e-01
-7.08445534e-02 8.92939493e-02 -1.56006172e-01 -2.94147223e-01
1.04796134e-01 -8.41014609e-02 -1.24487114e+00 -2.61628658e-01
4.89155561e-01 9.52706873e-01 -1.15046382e+00 -6.42694712e-01
-3.87910008e-01 5.66620648e-01 -7.07220912e-01 7.80965745e-01
-2.49796435e-01 5.18405914e-01 -4.57105041e-01 5.68663716e-01
1.67789757e-01 -8.53695944e-02 -3.79037440e-01 -1.13271952e-01
3.97497177e-01 5.72571933e-01 -1.13078392e+00 2.13718033e+00
-4.48783815e-01 8.61983180e-01 8.67023915e-02 -6.66895270e-01
9.24422264e-01 1.12229776e+00 2.99858153e-01 -3.84101689e-01
-8.54432881e-02 2.43794605e-01 -4.71160680e-01 -9.21640217e-01
6.19053662e-01 -2.46454284e-01 -5.71660757e-01 4.17250484e-01
3.42628777e-01 -2.89375156e-01 -2.40679100e-01 2.33159408e-01
1.12280750e+00 -1.49145275e-01 -1.81427281e-02 1.71715617e-01
3.54359299e-01 -4.78220955e-02 2.76879877e-01 7.99827874e-01
-1.64985970e-01 1.45411730e+00 2.01441884e-01 -2.90427543e-02
-1.19462311e+00 -9.71824467e-01 6.69998676e-02 1.64875364e+00
-6.73721731e-02 -5.40187180e-01 -8.60348701e-01 -2.29950398e-01
-3.14559698e-01 1.66283622e-01 -6.30174100e-01 -4.29029651e-02
-3.26457798e-01 -8.66650417e-03 1.24874961e+00 7.90719986e-01
3.20621759e-01 -1.23686993e+00 -4.72888164e-02 3.90680343e-01
-8.34270537e-01 -1.63650465e+00 -9.92819667e-01 -2.12027684e-01
-3.63554418e-01 -8.33949804e-01 -1.10037374e+00 -9.71834421e-01
6.82382360e-02 2.48599172e-01 1.24570417e+00 -3.62086982e-01
-9.95143503e-02 6.48390293e-01 -8.52142334e-01 -7.92919576e-01
-4.11787361e-01 3.11995238e-01 1.25397108e-02 3.54671448e-01
3.79707128e-01 -7.29500413e-01 -3.98755729e-01 3.41673076e-01
-7.33784139e-01 4.88079861e-02 3.75690460e-01 5.48838735e-01
6.00521922e-01 -5.58341086e-01 1.28375196e+00 -3.95596206e-01
8.22079420e-01 -7.26789832e-01 1.27868488e-01 1.71986148e-02
4.10117030e-01 -2.95336694e-01 7.72750437e-01 -6.57023430e-01
-6.22268438e-01 2.30716258e-01 -3.67780417e-01 -6.04321301e-01
-3.58316302e-01 1.40495673e-01 -1.05385289e-01 4.14294660e-01
6.82318866e-01 2.67856389e-01 -2.63736695e-01 -4.71229613e-01
3.73044461e-01 1.39688468e+00 1.13920975e+00 -7.57390022e-01
8.35305631e-01 5.37543952e-01 -6.89265907e-01 -7.67812490e-01
-1.20342588e+00 -5.67285836e-01 -2.45720074e-01 -1.79580301e-01
9.32684481e-01 -1.69962931e+00 -4.88549769e-01 2.90812291e-02
-1.45803463e+00 -3.68041620e-02 -5.13988078e-01 4.64523196e-01
-9.35759485e-01 -1.71551239e-02 -5.62325358e-01 -9.94448006e-01
-5.79085946e-01 -8.81142139e-01 1.83780885e+00 -1.91933453e-01
-6.01017296e-01 -3.02169472e-01 3.72371584e-01 6.62398636e-01
1.80203795e-01 3.74005511e-02 1.84296608e-01 -9.37947989e-01
-4.81886655e-01 -2.25687653e-01 -8.14126357e-02 1.30062550e-01
-2.08428368e-01 -2.92718709e-01 -1.63906145e+00 -4.55167666e-02
-3.56588483e-01 -9.11609828e-01 4.44696963e-01 -6.24065250e-02
1.30056143e+00 -5.66305816e-01 2.21253693e-01 1.69076681e-01
8.90619099e-01 1.09232850e-01 6.12157524e-01 5.37687028e-03
4.23591554e-01 6.56274319e-01 7.90172160e-01 6.71420693e-01
3.17605674e-01 9.19557869e-01 2.17338234e-01 3.55987288e-02
-2.69081384e-01 -9.97841477e-01 4.44999576e-01 1.24840593e+00
3.68842483e-01 -2.17280477e-01 -7.22418487e-01 1.12086499e+00
-1.78328609e+00 -1.07260585e+00 1.67460755e-01 1.67970490e+00
1.29451466e+00 -1.68864101e-01 4.07710314e-01 4.80381697e-01
1.08619928e+00 -6.07048646e-02 -2.73886561e-01 -1.73809096e-01
-9.09476429e-02 1.80434987e-01 -1.07046440e-01 4.12709594e-01
-1.05855846e+00 7.04460323e-01 7.09295893e+00 8.95731211e-01
-7.57442832e-01 2.89346009e-01 1.96010277e-01 -2.84960955e-01
-5.94426692e-01 -2.50573337e-01 -5.13773561e-01 5.91713846e-01
1.25946629e+00 -4.04656351e-01 6.01863027e-01 6.92846060e-01
2.64751673e-01 4.64584798e-01 -1.50740194e+00 1.17890584e+00
1.46257058e-01 -1.32787263e+00 2.49091685e-01 -5.76330721e-01
6.66236997e-01 -1.36467591e-01 -1.39238372e-01 3.09624016e-01
-1.32682100e-01 -1.04957640e+00 1.32118392e+00 2.67136663e-01
9.51071858e-01 -4.50458854e-01 7.65946448e-01 -5.39262369e-02
-1.52327025e+00 -1.54344320e-01 -1.53338999e-01 -2.53156513e-01
5.27065754e-01 2.92763978e-01 -1.07618189e+00 3.41226667e-01
8.91849816e-01 5.61701477e-01 -3.77795815e-01 1.11170566e+00
-3.43922168e-01 9.76255894e-01 -2.18158528e-01 -2.08504684e-02
9.97239277e-02 7.40851164e-01 6.11211896e-01 1.34429646e+00
3.73429418e-01 7.98648894e-02 -1.02795213e-02 6.73915803e-01
-4.02112067e-01 2.23440066e-01 -6.81681573e-01 -3.74764591e-01
1.05884027e+00 8.85098338e-01 -1.41588643e-01 -2.18972892e-01
-1.52564868e-01 8.14030170e-01 -7.33996779e-02 3.14082623e-01
-1.03804314e+00 -6.60921276e-01 6.78615570e-01 2.13657409e-01
1.08082287e-01 2.00714141e-01 1.07901953e-01 -1.09514332e+00
4.06889409e-01 -9.22461987e-01 3.27164769e-01 -1.47974920e+00
-1.45702922e+00 9.04672682e-01 -2.51732543e-02 -1.52804172e+00
-4.20071125e-01 1.34257125e-02 -5.04865408e-01 4.02568966e-01
-1.41862059e+00 -1.01810491e+00 -3.17901343e-01 6.79117680e-01
9.44566965e-01 -2.06479341e-01 9.76968110e-01 7.03474164e-01
-9.65434313e-03 7.71988213e-01 -3.49849224e-01 4.72821206e-01
1.03339624e+00 -8.73448074e-01 5.50641775e-01 4.48224127e-01
6.62837446e-01 2.57980555e-01 1.05147004e+00 -1.30546883e-01
-1.11041188e+00 -1.30562890e+00 8.80862355e-01 -6.60026073e-01
9.63502407e-01 -9.07385111e-01 -8.12027872e-01 7.73035765e-01
3.87631446e-01 -1.04567267e-01 9.64050531e-01 -2.37566873e-01
-5.63554108e-01 -1.04012236e-01 -7.21263289e-01 2.16935292e-01
1.16250002e+00 -1.31236744e+00 -9.25188005e-01 4.25185829e-01
1.60715997e+00 -4.52259809e-01 -7.79977500e-01 1.06738597e-01
5.63895762e-01 -4.60121334e-01 1.02897263e+00 -8.14296067e-01
7.97868133e-01 -2.39018574e-01 -4.49921966e-01 -1.01772094e+00
1.47355705e-01 -1.24396574e+00 4.24572416e-02 1.82577717e+00
5.41923821e-01 5.90010881e-02 5.52824676e-01 1.17132954e-01
-4.13056910e-01 -2.81357318e-01 -1.09905899e+00 -7.94623613e-01
-2.71929324e-01 -8.64113331e-01 1.01808918e+00 8.41385484e-01
4.23970640e-01 7.48498201e-01 -8.76308441e-01 2.21982211e-01
2.09895313e-01 3.25950235e-02 8.29974473e-01 -1.06953800e+00
-5.20038664e-01 2.41138011e-01 -4.68420208e-01 -9.79911208e-01
2.99587101e-01 -7.23320186e-01 4.80674475e-01 -1.23056102e+00
2.42847636e-01 -1.83462314e-02 2.76105311e-02 5.59800744e-01
-2.41773278e-02 7.89669037e-01 1.98873624e-01 2.47391298e-01
-9.89433467e-01 7.71580279e-01 1.00207734e+00 -2.76906997e-01
-4.02076170e-02 -2.22006008e-01 -1.10455239e+00 3.56367350e-01
3.88712972e-01 -6.97475553e-01 -5.63708007e-01 -8.00849855e-01
4.19379175e-01 4.63482499e-01 4.84522700e-01 -1.18279529e+00
7.53432363e-02 -1.27120009e-02 -2.64349222e-01 -1.67568818e-01
6.74061835e-01 -7.17236578e-01 1.40018299e-01 -3.00677925e-01
-1.22156322e+00 -2.84452327e-02 3.13691676e-01 7.96535492e-01
-8.27084064e-01 -2.05435410e-01 3.27569038e-01 1.11793354e-01
-3.72169346e-01 2.18935966e-01 -2.12917209e-01 7.05148518e-01
7.25556672e-01 1.30734235e-01 -1.42176554e-01 -1.12804461e+00
-8.33982944e-01 2.06836671e-01 -2.23685220e-01 8.05601597e-01
4.34239119e-01 -2.05614662e+00 -1.04965711e+00 -1.35156125e-01
5.29345393e-01 5.49687259e-02 2.17751399e-01 2.59489745e-01
4.29317169e-02 5.18231988e-01 -5.77887036e-02 -4.81071085e-01
-1.03687036e+00 3.98568094e-01 -1.73747782e-02 2.85462588e-01
-4.45650339e-01 1.15603125e+00 3.72662917e-02 -3.48683186e-02
6.91182733e-01 -1.18913829e-01 -7.23979250e-02 1.14577554e-01
9.09078360e-01 5.04626408e-02 -7.49716014e-02 -8.75143111e-01
-2.25715473e-01 4.86669511e-01 3.19111556e-01 -5.82359552e-01
1.16774225e+00 -3.46399158e-01 3.60034198e-01 8.10449779e-01
1.32565069e+00 7.09214807e-03 -9.80635941e-01 -1.47540212e-01
-2.44439319e-01 -4.57016259e-01 -3.24379027e-01 -5.21049321e-01
-5.53272903e-01 1.13364780e+00 2.22781986e-01 4.80992883e-01
1.12557900e+00 3.32287580e-01 1.28822339e+00 4.58697528e-01
4.44661319e-01 -1.10037827e+00 6.41472638e-01 4.30352658e-01
1.33607352e+00 -9.95155931e-01 -7.34849930e-01 -4.59663838e-01
-1.08374715e+00 8.52577627e-01 5.33767045e-01 -6.75108731e-02
1.08350158e-01 2.64456749e-01 1.26123384e-01 1.07926883e-01
-9.86559331e-01 -2.32343733e-01 2.63017535e-01 8.51816356e-01
4.60783005e-01 -2.14756548e-01 3.40566009e-01 1.41838872e+00
-7.67693162e-01 2.02965617e-01 6.39566660e-01 5.32086253e-01
-3.89158666e-01 -7.18823850e-01 -5.48994601e-01 -1.23960145e-01
-5.14112353e-01 -9.24434364e-02 -8.67154777e-01 6.67201653e-02
-1.92943260e-01 1.11345923e+00 1.25602439e-01 -4.71410722e-01
4.42451417e-01 3.51018310e-01 8.84430557e-02 -7.87346900e-01
-6.38066590e-01 3.90173867e-02 2.87247390e-01 -4.68922019e-01
-4.20761287e-01 -3.43524754e-01 -1.21967685e+00 4.83451009e-01
-2.30297714e-01 7.74239421e-01 3.96195740e-01 6.86853886e-01
6.29706323e-01 5.75344682e-01 8.86576951e-01 -9.03072655e-01
-4.84831154e-01 -1.14824212e+00 -4.51485187e-01 5.39332509e-01
6.59740627e-01 -2.77250230e-01 -4.73013550e-01 5.51946402e-01] | [15.294219017028809, 4.93602180480957] |
76e49071-3c13-47f6-9f67-6a4ac8eeb550 | synthesizing-annotated-image-and-video-data | 2209.14448 | null | https://arxiv.org/abs/2209.14448v1 | https://arxiv.org/pdf/2209.14448v1.pdf | Synthesizing Annotated Image and Video Data Using a Rendering-Based Pipeline for Improved License Plate Recognition | An insufficient number of training samples is a common problem in neural network applications. While data augmentation methods require at least a minimum number of samples, we propose a novel, rendering-based pipeline for synthesizing annotated data sets. Our method does not modify existing samples but synthesizes entirely new samples. The proposed rendering-based pipeline is capable of generating and annotating synthetic and partly-real image and video data in a fully automatic procedure. Moreover, the pipeline can aid the acquisition of real data. The proposed pipeline is based on a rendering process. This process generates synthetic data. Partly-real data bring the synthetic sequences closer to reality by incorporating real cameras during the acquisition process. The benefits of the proposed data generation pipeline, especially for machine learning scenarios with limited available training data, are demonstrated by an extensive experimental validation in the context of automatic license plate recognition. The experiments demonstrate a significant reduction of the character error rate and miss rate from 73.74% and 100% to 14.11% and 41.27% respectively, compared to an OCR algorithm trained on a real data set solely. These improvements are achieved by training the algorithm on synthesized data solely. When additionally incorporating real data, the error rates can be decreased further. Thereby, the character error rate and miss rate can be reduced to 11.90% and 39.88% respectively. All data used during the experiments as well as the proposed rendering-based pipeline for the automated data generation is made publicly available under (URL will be revealed upon publication). | ['André Kaup', 'Christian Riess', 'Jürgen Seiler', 'Denise Moussa', 'Anatol Maier', 'Maximilane Gruber', 'Andreas Spruck'] | 2022-09-28 | null | null | null | null | ['license-plate-recognition'] | ['computer-vision'] | [ 8.59464049e-01 1.24628522e-01 3.31858754e-01 -3.74788254e-01
-8.13625634e-01 -5.62545717e-01 6.65193081e-01 5.04771285e-02
-7.86232173e-01 7.74440348e-01 -3.76443356e-01 -7.35784993e-02
4.18308884e-01 -7.92039812e-01 -8.75691414e-01 -6.21550083e-01
4.90571469e-01 5.11872113e-01 3.01517189e-01 6.96239769e-02
4.92852986e-01 7.46347129e-01 -1.96516132e+00 3.75802010e-01
9.97272551e-01 7.73659408e-01 4.43950027e-01 7.03854024e-01
-1.35914207e-01 4.11155522e-01 -8.85656238e-01 -3.54284614e-01
5.71763933e-01 -2.30420470e-01 -2.01447755e-01 6.85236514e-01
6.51831329e-01 -5.13857961e-01 -7.48837069e-02 1.08073628e+00
2.85563529e-01 3.19114119e-01 4.18087691e-01 -1.12334001e+00
-4.15424973e-01 1.93908423e-01 -4.83336866e-01 -2.04064146e-01
-2.22006310e-02 3.32600981e-01 3.64198029e-01 -8.60351562e-01
6.23194456e-01 6.16251588e-01 4.66440022e-01 6.72340274e-01
-8.92104745e-01 -7.25106835e-01 -2.29188666e-01 1.67960878e-02
-1.26835799e+00 -5.41171730e-01 7.86469102e-01 -4.74213958e-01
7.08962739e-01 4.57144529e-01 5.53004324e-01 7.39090085e-01
-4.04307902e-01 6.03535056e-01 1.21488619e+00 -7.78054535e-01
2.04436511e-01 4.97997493e-01 2.70030588e-01 6.02222919e-01
4.71739769e-01 -1.55181482e-01 -2.36556649e-01 3.16783071e-01
8.63164306e-01 -8.02583918e-02 -3.02547842e-01 2.72924323e-02
-1.03356087e+00 4.67285752e-01 -1.44059837e-01 6.10357802e-03
-2.76324242e-01 -7.05406740e-02 5.02085507e-01 -8.71040076e-02
3.23617101e-01 4.54415441e-01 -1.95430696e-01 -1.58618718e-01
-1.27980483e+00 1.16213799e-01 3.61772239e-01 1.20194876e+00
8.40419054e-01 4.97519463e-01 3.52176465e-02 9.72429931e-01
1.72231451e-01 4.93641585e-01 7.13815033e-01 -6.87291801e-01
6.77692235e-01 8.40579569e-01 2.29914099e-01 -8.50345075e-01
-6.89439550e-02 -4.93745893e-01 -6.44692659e-01 4.34573203e-01
3.45674753e-01 -1.14699431e-01 -1.22766006e+00 1.26541185e+00
2.50477940e-01 2.52502799e-01 2.71982133e-01 7.52571821e-01
7.24283338e-01 6.98155999e-01 -2.09591165e-01 -1.45090371e-01
1.27979994e+00 -1.00189960e+00 -7.82245278e-01 -1.59197867e-01
5.21930635e-01 -1.01145041e+00 1.16936135e+00 5.81777096e-01
-8.45680118e-01 -7.65759408e-01 -1.36620069e+00 5.70764057e-02
-2.69171298e-01 9.11697328e-01 2.95918435e-01 9.46890831e-01
-8.15079033e-01 5.13011336e-01 -6.60729170e-01 -2.69093484e-01
4.73931342e-01 5.12024820e-01 -5.50388396e-01 -1.12458952e-01
-8.04820240e-01 5.64872444e-01 6.52186334e-01 2.40523040e-01
-7.26256132e-01 -5.03116429e-01 -8.26628745e-01 -1.69905424e-01
2.44873852e-01 -4.77799550e-02 9.28157806e-01 -1.06237948e+00
-1.51253498e+00 7.06064284e-01 9.44313183e-02 -4.86438394e-01
7.76883006e-01 -3.30897003e-01 -4.75609511e-01 2.44492471e-01
-2.05923647e-01 6.83418870e-01 7.19153225e-01 -1.31467295e+00
-7.51182556e-01 -1.83717638e-01 -2.25982815e-01 2.26180136e-01
-5.18844545e-01 -1.11676812e-01 -5.96139431e-01 -7.09907115e-01
-1.65938631e-01 -1.15294349e+00 1.05770960e-01 2.47542560e-02
-3.81330788e-01 3.37706268e-01 9.79282856e-01 -8.79942060e-01
5.71980476e-01 -2.04462504e+00 -5.80160499e-01 2.74174903e-02
-1.61401972e-01 9.55582440e-01 -8.74155909e-02 1.66505978e-01
-1.87044814e-01 1.47520170e-01 -5.67527711e-01 -6.35736644e-01
-4.28663373e-01 -9.92366821e-02 -7.08992630e-02 3.63952577e-01
5.97595215e-01 2.48875305e-01 -4.89809185e-01 -1.47707298e-01
5.21458805e-01 5.72751224e-01 -3.28358203e-01 3.06036830e-01
-1.44655854e-01 3.27825785e-01 6.74682334e-02 3.91900718e-01
9.98310804e-01 3.31256092e-01 -8.99889097e-02 -2.98339695e-01
-2.28900596e-01 -1.43991634e-01 -1.49526572e+00 1.38707590e+00
-3.87591958e-01 1.03750455e+00 -3.51540595e-01 -7.18085766e-01
1.44766951e+00 3.87435019e-01 1.38143957e-01 -5.20269692e-01
2.02750802e-01 1.24740966e-01 -1.34247258e-01 -4.83574152e-01
9.81437683e-01 -3.37445103e-02 2.16370955e-01 3.37376952e-01
-7.50394985e-02 4.30724844e-02 4.56055641e-01 -8.79226401e-02
7.37919569e-01 5.57797015e-01 -1.19894650e-02 1.24702856e-01
7.81954288e-01 3.14454317e-01 5.54436028e-01 3.67832571e-01
8.49255919e-02 7.10325241e-01 6.70369044e-02 -2.54454821e-01
-1.76542926e+00 -5.65427542e-01 -5.23777641e-02 3.06926280e-01
-3.94895822e-02 -3.84314358e-02 -9.11470592e-01 -3.77064735e-01
-4.65863228e-01 8.67538810e-01 -4.94262487e-01 3.38294245e-02
-6.76667452e-01 -7.46208429e-01 6.43754244e-01 3.66331995e-01
8.18191290e-01 -9.86157417e-01 -5.73994160e-01 -1.18752476e-02
6.38066009e-02 -1.55871367e+00 -2.07420528e-01 -2.18374237e-01
-1.08154619e+00 -9.01207924e-01 -5.69599867e-01 -7.16016293e-01
1.05102265e+00 1.23669721e-01 5.57351351e-01 2.31150299e-01
-2.76234120e-01 -9.50926393e-02 -4.23279822e-01 -3.96795064e-01
-9.32263970e-01 -1.81695923e-01 -1.10265799e-01 3.33809048e-01
2.35167399e-01 -2.31025219e-01 -4.27377075e-01 2.86980152e-01
-1.23928654e+00 5.94199896e-01 7.26819873e-01 6.92005873e-01
5.41589916e-01 6.95381761e-02 4.95901853e-01 -1.15847647e+00
4.61592674e-01 -7.96587672e-03 -9.91789341e-01 4.60238829e-02
-5.16713202e-01 6.15434498e-02 9.64851975e-01 -4.19275641e-01
-1.44573939e+00 4.33902115e-01 -1.52873546e-01 -4.42966610e-01
-7.14988112e-01 1.38138294e-01 -9.03533846e-02 5.85715249e-02
5.49542665e-01 4.89096820e-01 6.75215647e-02 -5.90160072e-01
2.21353350e-03 9.82832134e-01 6.11800671e-01 -1.66831806e-01
1.00517857e+00 3.15939158e-01 -8.96809176e-02 -1.07442892e+00
-3.76982428e-02 -3.78604889e-01 -6.86401010e-01 -4.50858831e-01
6.94875360e-01 -8.30940843e-01 -4.46831822e-01 8.83325577e-01
-9.76681292e-01 -1.24585927e-01 -1.75869659e-01 5.55828750e-01
-2.70229757e-01 6.02556169e-01 -3.73255908e-01 -1.05305529e+00
-4.52683657e-01 -1.34102821e+00 8.74870121e-01 4.05437231e-01
1.07862502e-01 -5.36003947e-01 -3.45635980e-01 7.22027659e-01
1.78862527e-01 3.91812712e-01 5.69154263e-01 -8.92701924e-01
-8.00957859e-01 -6.27981424e-01 -2.68629462e-01 6.44314885e-01
2.45581374e-01 4.72669363e-01 -1.09947014e+00 -6.75837398e-02
-1.34915993e-01 -2.48899862e-01 5.45603037e-01 -1.21640593e-01
8.04206431e-01 -1.51437625e-01 1.39702037e-01 3.06259245e-01
1.73689866e+00 5.41868687e-01 7.74962366e-01 5.17797172e-01
7.63666868e-01 5.70967853e-01 7.24264860e-01 5.22517741e-01
-3.29461880e-02 6.72816873e-01 4.65265989e-01 1.23273917e-02
-4.48739678e-01 -1.37944296e-01 2.22315252e-01 8.12414646e-01
-8.41747895e-02 -3.22418183e-01 -9.82885122e-01 5.42607069e-01
-1.37694073e+00 -8.66756678e-01 -7.34962046e-01 2.48577309e+00
9.25593138e-01 2.93759257e-01 1.84038933e-02 6.36166275e-01
1.01702988e+00 -2.42346898e-01 -3.47839206e-01 -4.94633704e-01
7.14499205e-02 6.75265044e-02 6.90546572e-01 5.29290915e-01
-9.93938327e-01 9.33322608e-01 4.65652514e+00 5.39837182e-01
-1.39442194e+00 -2.10799769e-01 4.76736188e-01 9.87612829e-02
-8.53000060e-02 -1.44022807e-01 -1.02424037e+00 4.60633129e-01
1.06874919e+00 2.93839853e-02 1.90665424e-01 7.45826602e-01
4.49840069e-01 -2.56539822e-01 -8.50759864e-01 1.01609468e+00
3.55514497e-01 -1.51054442e+00 2.62045652e-01 -1.13791414e-02
7.27240264e-01 -2.83001959e-01 -4.28963453e-02 -2.62992503e-03
-8.72358587e-03 -6.85105026e-01 6.75522685e-01 5.64666510e-01
9.08573568e-01 -9.30885792e-01 9.80367303e-01 4.76802349e-01
-8.67370725e-01 2.16755882e-01 -3.06888193e-01 6.21236339e-02
4.34444807e-02 5.66448450e-01 -1.33741617e+00 5.00386953e-01
1.59999698e-01 3.79189551e-01 -8.82164121e-01 1.24519718e+00
-6.19774796e-02 6.24727428e-01 -2.72947460e-01 -4.76946048e-02
1.61838438e-02 -2.44592562e-01 3.93247217e-01 1.36342990e+00
3.00578445e-01 -1.93299055e-01 -2.45026186e-01 7.85500526e-01
-1.12813719e-01 2.28425801e-01 -4.75653529e-01 -7.81994835e-02
4.25151199e-01 1.33150303e+00 -7.79960632e-01 -4.58554238e-01
-2.38562390e-01 1.03157496e+00 -4.77190316e-02 1.80097178e-01
-1.01893854e+00 -7.28692234e-01 1.20942473e-01 1.60392091e-01
1.24359258e-01 -1.96905270e-01 -3.17369550e-01 -1.02365828e+00
2.21797660e-01 -9.59352374e-01 -1.84494019e-01 -9.32466805e-01
-5.64310014e-01 9.36369956e-01 -5.66265173e-02 -1.55295193e+00
-2.00586870e-01 -7.13591754e-01 -3.15185905e-01 9.45682943e-01
-1.24557865e+00 -1.07902348e+00 -8.15165460e-01 2.29654431e-01
9.99038041e-01 -5.28933942e-01 7.69775569e-01 4.40732718e-01
-9.40314531e-01 7.02830970e-01 2.77547836e-01 3.98926497e-01
6.17759824e-01 -9.63361382e-01 4.49198127e-01 1.27315915e+00
6.49358938e-03 4.32833523e-01 7.18558967e-01 -6.12675607e-01
-1.08984578e+00 -1.28926468e+00 3.36609185e-01 7.58936480e-02
1.43871427e-01 -5.17982483e-01 -9.21772361e-01 4.00451958e-01
8.08143839e-02 -7.58602843e-02 7.84181714e-01 -7.38323390e-01
1.43392971e-02 -1.35964334e-01 -1.40158486e+00 6.21493042e-01
3.92843068e-01 -1.18815675e-01 -3.80723357e-01 2.58961618e-01
3.97029579e-01 -4.01455998e-01 -8.16157579e-01 1.57397538e-01
5.58353603e-01 -9.33830857e-01 5.10151207e-01 -1.32855669e-01
6.01129591e-01 -6.98458791e-01 -9.93409455e-02 -9.62788939e-01
2.13781893e-01 -3.56859535e-01 2.81564444e-01 1.57402670e+00
5.46569645e-01 -4.93359089e-01 1.07312810e+00 8.94829929e-01
-2.75003642e-01 -4.04389113e-01 -5.76869667e-01 -6.31525636e-01
-3.25446159e-01 -5.74092388e-01 3.45471948e-01 8.38232517e-01
-5.95088303e-01 4.61091474e-02 -4.02308017e-01 3.49896669e-01
7.23101020e-01 -2.44410187e-01 1.11066580e+00 -9.94427264e-01
-3.27016264e-01 7.99468234e-02 -6.96234047e-01 -6.13641441e-01
-2.30707631e-01 -6.12008214e-01 4.90958616e-02 -1.49020338e+00
5.88889718e-02 -4.59970057e-01 1.96809456e-01 2.92805552e-01
-7.84873962e-02 7.09356666e-01 3.89942795e-01 4.41958278e-01
3.51960137e-02 3.39673340e-01 1.17412746e+00 1.07438363e-01
-1.92915320e-01 -1.73106119e-02 -1.77704841e-01 7.48114944e-01
1.02866530e+00 -4.40257490e-01 -5.44790924e-01 -3.66360694e-01
-1.97154343e-01 -2.18057901e-01 1.16024792e-01 -1.30036986e+00
-4.59716050e-03 1.05222249e-02 5.83186388e-01 -7.14790106e-01
5.74229538e-01 -9.02745783e-01 3.60623062e-01 4.01523083e-01
-2.18373433e-01 1.12367338e-02 6.06210232e-01 5.00355422e-01
-1.85856700e-01 -7.60851562e-01 9.10273492e-01 -4.37908918e-02
-7.73859501e-01 -1.92085519e-01 -4.00369227e-01 -3.40075016e-01
1.33727908e+00 -8.32763374e-01 -3.59059930e-01 -3.00053090e-01
-5.80853403e-01 -4.43455219e-01 5.40754855e-01 3.33146960e-01
6.78186834e-01 -1.17126143e+00 -6.95780396e-01 4.21817482e-01
1.33242446e-03 1.06630296e-01 1.99140772e-01 4.70515281e-01
-1.18313634e+00 3.50662619e-01 -5.79149187e-01 -4.47230667e-01
-1.62971628e+00 3.54150802e-01 1.44629568e-01 -8.64061806e-03
-4.45241302e-01 4.88992453e-01 -2.42328584e-01 -3.28236997e-01
9.66054853e-03 -3.71705115e-01 -3.80451143e-01 -1.17794253e-01
4.68492866e-01 4.67087686e-01 3.42813849e-01 -7.79956698e-01
-7.67426565e-02 3.94626230e-01 -1.96870148e-01 -3.07422608e-01
1.52805865e+00 2.87603676e-01 2.09971666e-01 3.25273752e-01
1.07223940e+00 1.99532777e-01 -1.33356428e+00 7.07566291e-02
-1.55661806e-01 -7.06274688e-01 -6.67177141e-02 -7.76453316e-01
-1.07263052e+00 8.25909734e-01 7.83155560e-01 -1.34464219e-01
1.29424417e+00 -7.24271894e-01 5.76283336e-01 3.09696525e-01
5.11017665e-02 -1.02731049e+00 -1.00032359e-01 2.10569233e-01
8.89060676e-01 -1.13326812e+00 1.23613454e-01 -5.72781622e-01
-7.37356007e-01 1.35551381e+00 7.99037933e-01 -2.31540263e-01
3.02516688e-02 3.96064699e-01 8.72452781e-02 3.71958524e-01
-5.03641248e-01 1.24761149e-01 5.30920997e-02 6.18641257e-01
5.45428276e-01 -2.32973292e-01 -2.84908921e-01 2.60677487e-01
-1.65582061e-01 1.61135137e-01 1.21206176e+00 9.54558551e-01
-3.04516047e-01 -1.31710172e+00 -6.66458666e-01 4.47694480e-01
-3.40858132e-01 -1.23052590e-01 -2.95485795e-01 9.60868239e-01
2.65415728e-01 8.46111178e-01 2.51779199e-01 -3.44392031e-01
3.02706420e-01 1.78041771e-01 1.55803919e-01 -7.33189225e-01
-4.92848337e-01 4.06485535e-02 3.40793639e-01 6.79860711e-02
-4.36039418e-01 -6.96843803e-01 -1.30382442e+00 -1.14946887e-01
-5.90055585e-01 -2.30903588e-02 1.18803215e+00 8.19826841e-01
2.83642441e-01 6.61234677e-01 4.39353108e-01 -6.71806157e-01
-2.49783576e-01 -1.12094903e+00 -1.49927363e-01 4.91922736e-01
1.22615829e-01 -3.26301962e-01 -1.65528744e-01 7.26641774e-01] | [9.949886322021484, -4.577886581420898] |
7c407fe7-cad2-43b9-93fe-5cc11413d099 | sketch-guided-text-to-image-diffusion-models | 2211.13752 | null | https://arxiv.org/abs/2211.13752v1 | https://arxiv.org/pdf/2211.13752v1.pdf | Sketch-Guided Text-to-Image Diffusion Models | Text-to-Image models have introduced a remarkable leap in the evolution of machine learning, demonstrating high-quality synthesis of images from a given text-prompt. However, these powerful pretrained models still lack control handles that can guide spatial properties of the synthesized images. In this work, we introduce a universal approach to guide a pretrained text-to-image diffusion model, with a spatial map from another domain (e.g., sketch) during inference time. Unlike previous works, our method does not require to train a dedicated model or a specialized encoder for the task. Our key idea is to train a Latent Guidance Predictor (LGP) - a small, per-pixel, Multi-Layer Perceptron (MLP) that maps latent features of noisy images to spatial maps, where the deep features are extracted from the core Denoising Diffusion Probabilistic Model (DDPM) network. The LGP is trained only on a few thousand images and constitutes a differential guiding map predictor, over which the loss is computed and propagated back to push the intermediate images to agree with the spatial map. The per-pixel training offers flexibility and locality which allows the technique to perform well on out-of-domain sketches, including free-hand style drawings. We take a particular focus on the sketch-to-image translation task, revealing a robust and expressive way to generate images that follow the guidance of a sketch of arbitrary style or domain. Project page: sketch-guided-diffusion.github.io | ['Daniel Cohen-Or', 'Kfir Aberman', 'Andrey Voynov'] | 2022-11-24 | null | null | null | null | ['sketch-to-image-translation'] | ['computer-vision'] | [ 6.25664949e-01 1.94437444e-01 -1.42340243e-01 -3.41165870e-01
-8.92442703e-01 -6.11263990e-01 1.05093050e+00 -5.62190354e-01
-4.67122048e-02 6.14273906e-01 9.87733081e-02 -9.03809816e-02
4.35544029e-02 -9.94203389e-01 -1.10253513e+00 -9.15061295e-01
4.30027694e-01 5.74374080e-01 1.78518385e-01 -3.93781215e-02
3.85852277e-01 8.17206919e-01 -1.03775716e+00 6.18912458e-01
4.81064588e-01 9.24912810e-01 5.94857335e-01 9.22906220e-01
-2.57066488e-01 6.96763575e-01 -4.28881019e-01 -4.01778221e-01
3.00811708e-01 -6.21390641e-01 -5.64120471e-01 1.42509833e-01
5.54165900e-01 -6.00680292e-01 -5.45319736e-01 9.39450324e-01
4.60770249e-01 -2.47446299e-01 9.87684011e-01 -1.00396359e+00
-1.12290978e+00 4.48972374e-01 -4.00260568e-01 -3.15062940e-01
1.91902816e-01 3.81638229e-01 8.20627272e-01 -1.05330336e+00
1.30483639e+00 1.32545996e+00 5.49784243e-01 7.38398314e-01
-1.62876856e+00 -4.85272944e-01 -1.68346483e-02 -2.52563924e-01
-1.20323265e+00 -3.27791214e-01 1.11113942e+00 -5.89934826e-01
6.57524288e-01 -4.06983346e-02 3.51059914e-01 1.59988403e+00
2.32617766e-01 7.04736352e-01 1.20351052e+00 -4.48098302e-01
2.07776070e-01 3.24041456e-01 -7.61485040e-01 8.50765109e-01
-3.81191820e-01 2.16704011e-01 -6.96537375e-01 3.34452838e-02
1.33744717e+00 -1.45441324e-01 -1.03650540e-01 -4.83215630e-01
-1.29946172e+00 8.05101633e-01 4.77068305e-01 2.19554871e-01
-4.92252529e-01 4.77995604e-01 1.21651225e-01 3.92011404e-01
4.94452924e-01 3.36099386e-01 -2.78246999e-01 -3.75900907e-03
-1.37421227e+00 2.80805081e-01 6.74816668e-01 9.18588221e-01
8.98593307e-01 -7.72186089e-03 -2.30957538e-01 7.80103266e-01
1.62628829e-01 6.89158738e-01 1.09505735e-01 -1.32779980e+00
3.93113732e-01 1.91487402e-01 4.29775566e-02 -1.10643566e+00
2.08789960e-01 -2.05239713e-01 -1.08162463e+00 6.76157892e-01
4.72053617e-01 1.11295536e-01 -8.62235487e-01 1.70770764e+00
-3.28281079e-03 -1.75196260e-01 -2.26267532e-01 8.05096447e-01
3.12567860e-01 1.04238641e+00 -1.20437257e-01 2.56460369e-01
7.79870749e-01 -9.67405081e-01 -2.46044368e-01 -1.22529969e-01
3.53095293e-01 -9.43362713e-01 1.30800641e+00 6.04031503e-01
-1.40330231e+00 -6.92561150e-01 -1.04301322e+00 -2.99944133e-01
-4.71303880e-01 3.07272375e-01 1.07455008e-01 2.31379420e-01
-1.27082133e+00 1.09704161e+00 -4.93043661e-01 -3.17496657e-01
6.48934782e-01 1.16039678e-01 -4.30763751e-01 -2.35289469e-01
-8.37301672e-01 8.96333396e-01 -2.25516073e-02 -1.58808693e-01
-1.11819196e+00 -8.78264368e-01 -4.58864331e-01 -2.17509970e-01
-8.25425610e-02 -9.07260299e-01 8.06226492e-01 -1.31888485e+00
-2.05490208e+00 1.06423080e+00 -5.97088113e-02 -2.51700729e-01
9.59688485e-01 1.58403680e-01 1.18596591e-01 2.70442724e-01
2.14350969e-01 1.36392212e+00 1.53508127e+00 -1.41981339e+00
-2.94214755e-01 1.63166508e-01 -1.68540195e-01 -1.07207902e-01
-3.76627855e-02 -6.62904829e-02 -5.65397322e-01 -1.04993582e+00
-1.25375405e-01 -7.90502906e-01 -3.83395068e-02 7.44130969e-01
-5.16857922e-01 4.72642574e-03 9.66471910e-01 -6.83089256e-01
7.61319220e-01 -2.14770365e+00 6.96048319e-01 2.10885957e-01
-5.99009097e-02 -3.73281054e-02 -4.19685721e-01 5.99173427e-01
-3.25913454e-04 1.16930358e-01 -6.32937729e-01 -6.32869065e-01
2.36675054e-01 3.07470322e-01 -7.59563684e-01 3.47626120e-01
5.68351924e-01 1.03896201e+00 -7.67154932e-01 -4.63713527e-01
3.80130857e-01 7.79006064e-01 -6.00408435e-01 3.30367148e-01
-6.59115791e-01 7.36441076e-01 -2.89547086e-01 4.15863872e-01
7.47195840e-01 -2.28411436e-01 -1.57579586e-01 -2.57294089e-01
-6.97304457e-02 1.10011145e-01 -9.56792057e-01 2.22315168e+00
-7.03272879e-01 7.86750197e-01 1.20910883e-01 -7.55578816e-01
1.07529318e+00 2.93023568e-02 3.05089355e-01 -7.17642248e-01
-1.66172624e-01 4.48371381e-01 -5.55559754e-01 -1.90807119e-01
2.32652023e-01 -2.51337260e-01 8.63178074e-03 7.49221504e-01
1.86035737e-01 -7.53220081e-01 9.22474172e-03 2.55314916e-01
9.74589586e-01 6.63681686e-01 -4.14217114e-01 -2.21372232e-01
4.58160996e-01 -1.46370918e-01 -8.91659129e-03 6.23685002e-01
3.31826091e-01 1.12018609e+00 6.38533413e-01 -4.63264674e-01
-1.84778678e+00 -1.34412873e+00 5.79150207e-03 8.43824029e-01
-1.98785484e-01 -1.75651357e-01 -8.70875061e-01 -7.14057803e-01
-1.18384860e-01 7.53427625e-01 -6.82851732e-01 9.20187086e-02
-6.30221665e-01 -3.24242622e-01 6.09685600e-01 3.12647820e-01
4.96338844e-01 -1.26225317e+00 -1.48327142e-01 2.16056600e-01
1.08569123e-01 -1.02072310e+00 -6.29070401e-01 1.16684977e-02
-7.57980168e-01 -5.48655152e-01 -1.31784320e+00 -7.93217719e-01
9.48186994e-01 -2.96704113e-01 1.04721701e+00 -1.05410770e-01
-1.87154129e-01 3.30883116e-01 1.82902098e-01 9.26896185e-02
-8.11767638e-01 1.52576417e-01 -3.15801859e-01 1.80665284e-01
-2.93903112e-01 -8.59881043e-01 -7.18103826e-01 3.51111263e-01
-1.08686101e+00 4.19414192e-01 8.58316779e-01 9.69451308e-01
7.09612310e-01 -2.33135343e-01 2.70881742e-01 -6.44164562e-01
5.52208483e-01 -2.30273515e-01 -5.71718097e-01 1.32408693e-01
-4.63109910e-01 3.39871407e-01 8.98562074e-01 -5.74018121e-01
-9.94714797e-01 3.66608500e-01 -2.30334505e-01 -5.58271468e-01
-7.73153901e-02 -1.24054067e-02 -1.27642557e-01 -1.82895511e-01
6.75365448e-01 5.21715403e-01 1.33185368e-02 -4.19145405e-01
9.12745774e-01 3.57220024e-01 5.23073018e-01 -8.18749368e-01
9.69778180e-01 7.35935211e-01 8.52469802e-02 -7.54315436e-01
-3.70138675e-01 2.48077035e-01 -9.83421683e-01 -1.25743538e-01
8.60205114e-01 -6.34884834e-01 -2.78367102e-01 5.39651692e-01
-1.49483705e+00 -1.01740730e+00 -4.80719149e-01 -1.03359118e-01
-9.81930137e-01 1.83257222e-01 -8.18876684e-01 -4.92253423e-01
-1.97344333e-01 -1.16753447e+00 1.42078304e+00 -1.06847905e-01
-3.04377407e-01 -1.12685978e+00 4.05349322e-02 -8.55166614e-02
6.26620650e-01 1.77545875e-01 1.01751459e+00 2.22129568e-01
-1.11684787e+00 -5.11970744e-03 -4.72699612e-01 7.42838621e-01
-8.00653398e-02 3.36211741e-01 -9.86241639e-01 -2.04723747e-03
-1.95634156e-01 -4.36986923e-01 9.37107503e-01 3.73106241e-01
1.27185082e+00 -3.33745539e-01 -2.88323969e-01 8.97155046e-01
1.37866116e+00 -2.85391599e-01 7.88424492e-01 1.22903332e-01
6.35779083e-01 5.87278247e-01 8.14696401e-02 6.53566495e-02
7.44648352e-02 6.46548569e-01 9.83263105e-02 -2.88292050e-01
-7.29524434e-01 -7.32817471e-01 5.12719870e-01 6.79102123e-01
8.36381987e-02 -1.14279225e-01 -4.84235257e-01 3.45981061e-01
-1.47424972e+00 -8.45711112e-01 1.52108833e-01 1.88770378e+00
1.15126419e+00 1.29310250e-01 -2.01078758e-01 -8.39622170e-02
4.25038457e-01 3.87820721e-01 -4.01324391e-01 -4.92463976e-01
-4.22689617e-01 2.83091873e-01 4.18925643e-01 8.26179028e-01
-8.08072686e-01 1.11681652e+00 6.11250687e+00 1.22347987e+00
-1.49659812e+00 1.18717067e-01 8.97931039e-01 8.99393298e-03
-6.89720392e-01 -9.41252559e-02 -4.41711068e-01 5.65361977e-01
6.40094578e-01 2.66084015e-01 7.97742307e-01 7.17773438e-01
2.43051291e-01 2.11562347e-02 -1.23691165e+00 9.47388291e-01
-1.47726238e-02 -1.86773825e+00 3.45961273e-01 7.55458549e-02
1.00353909e+00 -5.03615141e-02 4.85175997e-01 -1.53411210e-01
3.15675944e-01 -1.10797632e+00 1.01621139e+00 1.01200688e+00
1.26327109e+00 -4.07267332e-01 3.07791121e-02 3.70660961e-01
-6.64093435e-01 2.68381387e-01 -5.01108825e-01 3.51074725e-01
3.09315860e-01 6.64415836e-01 -5.65288723e-01 1.81597024e-01
4.31764603e-01 6.90668941e-01 -2.67270207e-01 3.41527849e-01
-4.66606289e-01 3.46973538e-01 -3.47957641e-01 1.81771368e-01
4.73098665e-01 -3.92212659e-01 3.97705793e-01 1.39029074e+00
5.81320167e-01 -2.26665571e-01 -1.77390411e-01 1.62838745e+00
-2.55352944e-01 -1.67664781e-01 -9.03941989e-01 -7.41087198e-02
7.27653503e-02 1.29207242e+00 -6.46000445e-01 -3.99845898e-01
-1.45604774e-01 1.66282916e+00 2.56962717e-01 5.83930314e-01
-7.95139134e-01 -2.08776936e-01 3.40325862e-01 3.68900388e-01
5.20970047e-01 -4.57208902e-01 -5.77486455e-01 -7.88901806e-01
9.16330442e-02 -7.34005511e-01 -3.67486298e-01 -1.27526557e+00
-1.46101046e+00 5.17217577e-01 -3.02256405e-01 -9.10811484e-01
-1.83738396e-01 -6.80817068e-01 -6.51920438e-01 1.31051970e+00
-1.52550113e+00 -1.51668346e+00 -1.03082277e-01 4.75602806e-01
5.01030803e-01 -1.21404991e-01 7.94082046e-01 3.00578386e-01
-1.22363754e-01 4.03898329e-01 1.23091467e-01 7.28257298e-02
8.92575920e-01 -1.15341568e+00 5.95932424e-01 5.48282743e-01
1.61197558e-01 4.48360771e-01 5.89366913e-01 -6.86497271e-01
-1.37872124e+00 -1.03418148e+00 9.03761327e-01 -6.84775531e-01
7.70697236e-01 -4.62686002e-01 -7.38779962e-01 4.66982514e-01
3.36387664e-01 -2.86857821e-02 4.18293150e-03 -6.02163553e-01
-4.12698120e-01 -1.08147852e-01 -1.03806257e+00 7.40182757e-01
1.03575563e+00 -1.01749551e+00 -2.02419609e-01 3.21102381e-01
5.33668995e-01 -3.17089468e-01 -7.56482661e-01 -2.26325989e-01
6.21046782e-01 -9.18784857e-01 1.18253350e+00 -3.01994741e-01
1.09073210e+00 -2.45460913e-01 -1.80199578e-01 -1.32961154e+00
-3.11931491e-01 -9.45783019e-01 7.80877769e-02 1.24634814e+00
6.57336414e-01 -3.15880686e-01 8.79394352e-01 4.23844784e-01
2.91664358e-02 -6.77826107e-01 -7.71651208e-01 -5.92051804e-01
5.09957671e-01 -4.00197983e-01 4.22111601e-01 7.76264429e-01
-4.75027919e-01 1.30215645e-01 -5.30188024e-01 -1.16701566e-01
7.24628806e-01 -6.57825992e-02 7.48584449e-01 -7.36943245e-01
-3.98736238e-01 -6.95698917e-01 8.09443556e-03 -1.45967925e+00
4.12902497e-02 -9.89907563e-01 2.27381717e-02 -1.47788382e+00
-1.74417734e-01 -6.37696743e-01 1.28783450e-01 2.47745991e-01
3.05107027e-01 4.03478920e-01 3.11901063e-01 3.13013285e-01
-1.86556205e-01 6.04762554e-01 1.65771854e+00 -4.78625625e-01
-3.58563326e-02 -1.58873960e-01 -2.87636191e-01 5.73094130e-01
6.03116751e-01 -6.86167538e-01 -2.40816861e-01 -6.61407769e-01
3.89074117e-01 -1.11817755e-01 7.21263945e-01 -8.05502355e-01
3.61792415e-01 -6.81592301e-02 5.86529434e-01 -3.45707119e-01
5.01302361e-01 -7.06678391e-01 2.00536355e-01 2.46790469e-01
-5.55822372e-01 -1.83425352e-01 -1.12965867e-01 3.68531018e-01
2.37848032e-02 -2.33895361e-01 9.17761087e-01 -2.30853558e-01
-3.59803706e-01 4.00019526e-01 -8.28063041e-02 -2.48643920e-01
6.60415173e-01 -2.31304199e-01 -1.50529295e-01 -5.10251701e-01
-7.65896797e-01 -3.91765058e-01 8.24690998e-01 2.83184737e-01
6.49537504e-01 -1.52477241e+00 -7.92061567e-01 4.12790120e-01
-1.74392238e-01 1.79746607e-03 2.36016691e-01 6.16927624e-01
-6.94989145e-01 3.52424651e-01 -3.07536751e-01 -7.27739632e-01
-5.48090875e-01 5.52969933e-01 3.76430035e-01 -1.87103540e-01
-9.40378070e-01 9.44590092e-01 3.09347749e-01 -5.46195328e-01
1.47492960e-01 -4.69133377e-01 6.81564033e-01 -1.79159328e-01
4.38138038e-01 -1.34953391e-02 -2.37945184e-01 -3.86921346e-01
1.24070518e-01 1.00583005e+00 2.92560101e-01 -6.32899284e-01
1.50856793e+00 1.05129834e-02 -1.44491315e-01 3.07095319e-01
1.41573727e+00 -7.26611307e-03 -2.08072281e+00 -2.12282166e-01
-2.31376827e-01 -4.53153789e-01 3.95063199e-02 -9.93904948e-01
-1.02355206e+00 1.28098321e+00 3.17911923e-01 -6.38975063e-03
8.61096323e-01 4.99517992e-02 8.29643905e-01 5.90041094e-02
2.72845984e-01 -1.20085180e+00 4.74283427e-01 2.72656798e-01
1.48928678e+00 -9.99176621e-01 -1.44208208e-01 -4.86579835e-02
-6.54627085e-01 1.28302097e+00 -2.64303330e-02 -4.98884737e-01
6.46920562e-01 4.25490886e-01 -2.27457192e-02 -3.61639932e-02
-6.79683030e-01 2.63601899e-01 3.20337504e-01 8.20268571e-01
9.87721607e-02 -7.94119015e-02 2.97793478e-01 1.17316090e-01
-7.34072775e-02 2.77358443e-01 2.23381221e-01 5.50079823e-01
-2.29022801e-01 -1.32259858e+00 -2.95965731e-01 4.96739633e-02
3.01461574e-03 -1.88274100e-01 -4.85878944e-01 5.37635624e-01
2.86990881e-01 3.45979035e-01 6.42316341e-02 -1.47786379e-01
1.76119551e-01 1.11823440e-01 7.69241512e-01 -3.29186887e-01
-3.51418138e-01 1.02487311e-01 -1.91870049e-01 -6.53670371e-01
-1.17906496e-01 -4.76176441e-01 -9.19817209e-01 -5.57938337e-01
2.80034572e-01 -4.17906463e-01 9.21050191e-01 6.81381047e-01
3.56732965e-01 4.47745413e-01 6.08797491e-01 -1.36031127e+00
-2.87954837e-01 -8.08443785e-01 -5.06064355e-01 3.19764823e-01
2.57726341e-01 -3.42804015e-01 -2.21743733e-01 4.02773261e-01] | [11.51987361907959, -0.5101370811462402] |
aadecf70-6a71-45ea-b0e9-cdd5ab157847 | 3d-colored-shape-reconstruction-from-a-single | 2302.05573 | null | https://arxiv.org/abs/2302.05573v1 | https://arxiv.org/pdf/2302.05573v1.pdf | 3D Colored Shape Reconstruction from a Single RGB Image through Diffusion | We propose a novel 3d colored shape reconstruction method from a single RGB image through diffusion model. Diffusion models have shown great development potentials for high-quality 3D shape generation. However, most existing work based on diffusion models only focus on geometric shape generation, they cannot either accomplish 3D reconstruction from a single image, or produce 3D geometric shape with color information. In this work, we propose to reconstruct a 3D colored shape from a single RGB image through a novel conditional diffusion model. The reverse process of the proposed diffusion model is consisted of three modules, shape prediction module, color prediction module and NeRF-like rendering module. In shape prediction module, the reference RGB image is first encoded into a high-level shape feature and then the shape feature is utilized as a condition to predict the reverse geometric noise in diffusion model. Then the color of each 3D point updated in shape prediction module is predicted by color prediction module. Finally, a NeRF-like rendering module is designed to render the colored point cloud predicted by the former two modules to 2D image space to guide the training conditioned only on a reference image. As far as the authors know, the proposed method is the first diffusion model for 3D colored shape reconstruction from a single RGB image. Experimental results demonstrate that the proposed method achieves competitive performance on colored 3D shape reconstruction, and the ablation study validates the positive role of the color prediction module in improving the reconstruction quality of 3D geometric point cloud. | ['Bin Liu', 'Fengwei Chen', 'Xiaolin Wei', 'Bo Li'] | 2023-02-11 | null | null | null | null | ['3d-shape-generation', '3d-shape-reconstruction'] | ['computer-vision', 'computer-vision'] | [ 1.18747979e-01 -2.05607489e-01 4.69618469e-01 2.49039866e-02
-3.81838471e-01 -3.65688443e-01 4.30196911e-01 -4.34861004e-01
-1.32590503e-01 1.57283917e-01 -2.24100873e-01 -3.58700871e-01
6.92290217e-02 -1.31900811e+00 -4.90875393e-01 -9.31950331e-01
5.30605853e-01 5.22963464e-01 4.35029000e-01 -1.22572429e-01
3.26734692e-01 1.03681159e+00 -1.32136977e+00 1.48676381e-01
9.04189408e-01 1.16408491e+00 5.00201225e-01 6.16312027e-01
-8.60012829e-01 3.32479358e-01 -4.01289105e-01 -2.76634216e-01
5.27040482e-01 -5.08713961e-01 -2.43043840e-01 3.50649178e-01
-1.88057289e-01 -2.71851480e-01 -1.37248412e-01 1.12884641e+00
7.75104105e-01 -1.37162209e-01 8.60889554e-01 -1.06262290e+00
-1.13973844e+00 -1.99984834e-01 -6.57037199e-01 -4.69337851e-01
3.21372539e-01 1.27388209e-01 2.41148725e-01 -1.24298799e+00
7.85030365e-01 1.28228724e+00 4.75531787e-01 7.60494709e-01
-9.99413669e-01 -7.02790618e-01 1.11401558e-01 -2.17497081e-01
-1.42837167e+00 1.04199789e-01 1.52170670e+00 -1.88693330e-01
4.95325685e-01 2.46967450e-01 1.14689291e+00 5.21344304e-01
7.54395314e-03 7.74460196e-01 1.46590054e+00 -3.09235841e-01
3.03928554e-01 -7.15738442e-03 -5.37512124e-01 9.20935631e-01
-9.10730734e-02 3.00428331e-01 -1.74903423e-01 -1.27855211e-01
1.23001850e+00 2.06785977e-01 -3.19862753e-01 -5.10096192e-01
-9.93929982e-01 4.70776647e-01 7.03424454e-01 2.53040761e-01
-3.61749917e-01 -9.71770566e-03 -4.62432861e-01 1.38981327e-01
7.67076015e-01 -2.74669826e-01 -3.07039529e-01 -7.16696158e-02
-5.85936129e-01 -1.65063322e-01 5.92915535e-01 1.03701055e+00
9.54542875e-01 2.91657150e-01 6.84576333e-02 8.01667392e-01
8.64036918e-01 1.32542884e+00 2.69509882e-01 -7.32546628e-01
2.10400403e-01 1.08823216e+00 1.24984726e-01 -1.15902042e+00
-3.60337526e-01 -9.86663774e-02 -1.03703403e+00 7.52909958e-01
3.12160794e-02 -6.26117736e-02 -1.05526328e+00 9.94847178e-01
6.97162211e-01 1.66230977e-01 1.42942503e-01 1.03735924e+00
1.13830042e+00 9.78294909e-01 -3.74739677e-01 -1.98956013e-01
7.48161972e-01 -6.48675978e-01 -3.76502723e-01 3.05021495e-01
2.51194358e-01 -1.09900010e+00 1.01926196e+00 6.06512368e-01
-1.09110951e+00 -6.38146400e-01 -8.57545078e-01 -1.64098516e-02
-3.25014770e-01 3.17467004e-01 3.99540186e-01 8.76565337e-01
-1.04740310e+00 3.49462688e-01 -5.45464396e-01 -8.63982961e-02
2.71699488e-01 1.75018497e-02 -1.46699660e-02 -4.95420724e-01
-5.95831394e-01 6.27418458e-01 -3.17919962e-02 2.79769242e-01
-4.94397879e-01 -5.58376968e-01 -4.16251183e-01 -4.37056094e-01
-4.33580503e-02 -7.32441902e-01 5.41585088e-01 -6.85248375e-01
-1.85135293e+00 7.96394706e-01 -1.79344386e-01 4.21975136e-01
6.38347864e-01 3.22734177e-01 -4.64780360e-01 3.96310575e-02
-1.67040020e-01 4.72529054e-01 1.01235032e+00 -2.12083697e+00
-4.85617697e-01 -4.07318175e-01 -3.14047188e-01 3.40185285e-01
7.04540685e-02 -4.87651676e-01 -9.11716640e-01 -6.00068986e-01
8.60481322e-01 -8.05848300e-01 -3.70101929e-01 4.73742127e-01
-2.04406604e-01 9.47355181e-02 1.20248508e+00 -3.90101850e-01
7.41947651e-01 -2.24832106e+00 -6.23835698e-02 7.00261593e-01
-4.77405004e-02 1.23516753e-01 -3.24709088e-01 2.52099097e-01
1.22545585e-01 2.20341742e-01 -4.54719931e-01 -4.59661990e-01
-7.61887357e-02 1.77834183e-01 -1.84488878e-01 3.41784656e-01
2.49572545e-01 9.50848401e-01 -9.11338449e-01 -3.71781051e-01
4.13899273e-01 9.56752479e-01 -3.52286518e-01 2.27716565e-01
-1.82846397e-01 7.88828015e-01 -5.42905211e-01 7.81161726e-01
1.41329324e+00 7.25852549e-02 -2.00045153e-01 -4.21138555e-01
-2.50383228e-01 -5.71711361e-01 -1.44550824e+00 1.86418366e+00
-4.70833719e-01 -1.96040884e-01 1.44499252e-02 -3.10208410e-01
1.81603098e+00 8.73149838e-03 6.88022316e-01 -1.03120708e+00
-1.68828312e-02 3.27014595e-01 -3.34047377e-01 -2.92474002e-01
4.86275047e-01 -3.43514532e-01 2.38747135e-01 6.12156332e-01
-5.16519368e-01 -1.05282187e+00 -3.66476208e-01 1.99235737e-01
6.11703396e-01 5.26648819e-01 -5.56729376e-01 2.02498302e-01
6.12022340e-01 1.71642676e-01 4.47142899e-01 3.74559283e-01
2.35684425e-01 1.04045451e+00 1.26257492e-05 -2.69264936e-01
-1.21183419e+00 -1.30849063e+00 1.42538562e-01 1.31481692e-01
7.64987230e-01 -9.11582187e-02 -4.12868261e-01 -7.13266134e-01
-1.09621815e-01 5.96205711e-01 -4.51789826e-01 -2.18534991e-01
-4.41251695e-01 -7.39700377e-01 1.14480808e-01 2.04524919e-01
8.23028743e-01 -1.05387044e+00 -1.74654484e-01 2.80999020e-02
1.95869267e-01 -6.68737233e-01 -3.83393019e-01 -2.90061533e-01
-1.22687817e+00 -1.14655840e+00 -1.05409992e+00 -7.87159026e-01
1.22036040e+00 6.56801999e-01 7.69017339e-01 5.00744939e-01
-2.38146663e-01 7.51694262e-01 -5.78894913e-01 -3.16543430e-01
-4.09952343e-01 -6.54074192e-01 -3.06563675e-01 4.42609221e-01
-6.78056926e-02 -5.72823524e-01 -9.11853492e-01 3.61850172e-01
-1.09428942e+00 3.64309937e-01 7.34068751e-01 4.85454619e-01
1.13576543e+00 1.87219068e-01 6.10146709e-02 -7.24014223e-01
6.38393223e-01 -1.40306607e-01 -6.64920449e-01 1.74542695e-01
-7.14829266e-01 -6.05873801e-02 5.66351354e-01 -3.99703503e-01
-1.45855701e+00 4.49508607e-01 -3.65519971e-01 -5.62478483e-01
-8.61027613e-02 1.79822132e-01 -2.50763685e-01 -3.30738783e-01
2.82921493e-01 9.17851806e-01 8.84889066e-02 -7.03472197e-01
5.53395808e-01 4.11483914e-01 2.32884780e-01 -5.90539694e-01
1.33071780e+00 8.18400502e-01 2.28948772e-01 -7.03809798e-01
-1.09109588e-01 -1.96127906e-01 -8.46525550e-01 -6.16379738e-01
8.90332758e-01 -5.99358439e-01 -7.69636154e-01 7.32512414e-01
-1.17122328e+00 -4.09793317e-01 -3.08506042e-01 2.44993880e-01
-5.35585880e-01 6.65398300e-01 -4.16315258e-01 -1.02518940e+00
-4.31060791e-01 -1.01125646e+00 1.25376701e+00 3.26773345e-01
5.83591878e-01 -9.12277281e-01 -3.89645435e-02 9.74416211e-02
4.07262057e-01 1.68281719e-01 1.06971657e+00 4.43428010e-01
-9.55324650e-01 -1.71306193e-01 -4.59195018e-01 2.93219954e-01
2.48954743e-01 2.12371945e-01 -7.10777640e-01 3.29203345e-02
8.92545506e-02 4.82373089e-02 5.47518671e-01 1.06926635e-01
1.04580700e+00 2.88442135e-01 -1.77987441e-01 8.18367183e-01
1.79307318e+00 5.75433552e-01 7.99659610e-01 1.13278940e-01
8.84296656e-01 3.61413993e-02 7.05876112e-01 4.19528306e-01
4.32372123e-01 3.23627114e-01 6.27794683e-01 -4.57746625e-01
-6.20249271e-01 -6.17715776e-01 1.56272605e-01 1.24332345e+00
-4.00770038e-01 1.48505449e-01 -7.56828010e-01 8.55204687e-02
-1.50552130e+00 -5.58874547e-01 -6.29435539e-01 2.21266174e+00
4.52877790e-01 -1.05771534e-01 -2.53010809e-01 2.30598584e-01
5.31817615e-01 -1.27471462e-01 -6.03064001e-01 -1.34138927e-01
-4.28183198e-01 1.07374735e-01 1.19899966e-01 4.67906356e-01
-3.44624698e-01 1.00517583e+00 5.38024759e+00 8.16250980e-01
-1.31378138e+00 -1.11378841e-02 4.88157541e-01 3.51135731e-01
-9.04598534e-01 7.60002136e-02 -5.21214247e-01 3.03885877e-01
-5.53742163e-02 2.43799258e-02 3.89501959e-01 6.69739068e-01
3.39161634e-01 -2.76622653e-01 -2.87945062e-01 1.44843459e+00
1.40052751e-01 -1.08229029e+00 4.54169393e-01 6.60230741e-02
1.00448310e+00 -3.06666225e-01 3.68024737e-01 -2.57669270e-01
3.63081247e-01 -6.26524866e-01 7.38929451e-01 1.22872055e+00
9.13907945e-01 -7.90361047e-01 3.27365458e-01 6.13885820e-01
-1.26494217e+00 2.12867126e-01 -6.82160735e-01 3.91802907e-01
1.92889974e-01 8.34620953e-01 -6.24824524e-01 8.28617871e-01
5.78333914e-01 6.07599795e-01 -5.36607504e-01 1.32231474e+00
-4.17547256e-01 3.22987080e-01 -2.48591959e-01 -1.61973462e-01
1.52965426e-01 -7.86242247e-01 3.44033897e-01 7.73087204e-01
9.01966035e-01 5.59249938e-01 9.14100856e-02 1.17728209e+00
2.05202118e-01 3.74724835e-01 -6.31693780e-01 2.93451279e-01
1.99902758e-01 1.27250290e+00 -1.00895154e+00 -3.11410844e-01
-3.36377114e-01 1.32383525e+00 -1.79396588e-02 5.62515736e-01
-6.84697628e-01 -1.57727778e-01 -5.56667149e-02 -1.14292540e-01
3.18187326e-01 -6.32009327e-01 -6.11563683e-01 -9.60702717e-01
-5.30907810e-02 -2.83529222e-01 -1.55474499e-01 -1.51603472e+00
-1.31621313e+00 6.35360837e-01 -5.72238624e-01 -1.61335659e+00
4.79755849e-01 -5.54795504e-01 -6.02325916e-01 1.29930127e+00
-1.71858394e+00 -1.23418891e+00 -5.90914667e-01 1.00263560e+00
2.50082016e-01 5.99084003e-03 8.88226032e-01 9.55908671e-02
-2.26436213e-01 -5.20033855e-03 2.32795119e-01 -5.10352887e-02
4.16185737e-01 -1.09795570e+00 1.51667058e-01 6.22404695e-01
-1.00790132e-02 1.04300104e-01 -5.79717793e-02 -1.09615302e+00
-1.87309432e+00 -1.03398108e+00 4.18381155e-01 -2.23106250e-01
4.29331623e-02 -1.09135330e-01 -7.35858440e-01 -7.93845281e-02
-3.17157835e-01 2.04262212e-01 3.10917705e-01 -5.81146181e-01
-1.34177938e-01 -1.97168529e-01 -1.35692382e+00 5.06631494e-01
1.29089355e+00 -4.49160695e-01 -1.36695042e-01 3.31622437e-02
4.81211305e-01 -5.73466063e-01 -9.82808948e-01 3.35112512e-01
3.32244664e-01 -9.48881924e-01 9.77526546e-01 3.84448990e-02
3.66711169e-01 -7.68663406e-01 -1.18765414e-01 -1.30728757e+00
-3.50951493e-01 -2.54247665e-01 1.07282825e-01 1.15474403e+00
3.61038774e-01 -4.22432989e-01 1.08127761e+00 8.11504960e-01
-1.88786194e-01 -6.72634184e-01 -7.39733756e-01 -6.03242636e-01
8.90143365e-02 -7.22077727e-01 9.39314604e-01 7.77158380e-01
-7.20722079e-01 -4.70511951e-02 -1.96765110e-01 3.41073126e-01
6.79394722e-01 7.55012870e-01 9.16947603e-01 -1.10413671e+00
9.19787511e-02 -4.46582526e-01 -8.52472112e-02 -1.48451948e+00
-3.96877199e-01 -1.03643918e+00 -1.10521540e-01 -2.10971737e+00
-3.85260135e-02 -1.11180186e+00 1.08354308e-01 7.96538740e-02
-1.04267761e-01 6.00917101e-01 5.01969337e-01 2.77673811e-01
-1.84889436e-01 8.74562263e-01 2.06987667e+00 -4.25886095e-01
-4.25090998e-01 7.33280405e-02 -4.40154940e-01 5.09302557e-01
5.64312637e-01 -3.51763934e-01 -5.19697607e-01 -5.41899323e-01
3.19580644e-01 9.73912925e-02 2.45355144e-01 -8.90040159e-01
3.40626001e-01 -2.73986906e-01 8.62809539e-01 -1.05607224e+00
5.53527653e-01 -1.38553214e+00 3.46313149e-01 5.06406665e-01
4.26743954e-01 -8.02962258e-02 7.11333826e-02 6.50360644e-01
-4.37553460e-03 -7.54899532e-02 6.78938866e-01 -2.56201595e-01
-8.74948204e-01 7.90485084e-01 2.51889378e-02 -3.05231452e-01
9.18676615e-01 -7.75383413e-01 -3.13127190e-02 -3.33016247e-01
-9.55158293e-01 -2.84904152e-01 7.03439116e-01 3.01087588e-01
1.36289454e+00 -1.89306021e+00 -5.15293241e-01 6.02743030e-01
-1.05027080e-01 2.68880457e-01 5.92951000e-01 5.21445036e-01
-6.75076842e-01 -1.56297132e-01 -4.52620722e-03 -6.10338509e-01
-1.01221251e+00 6.59180939e-01 3.11321169e-01 1.75355345e-01
-7.74270713e-01 7.58612454e-01 5.92866400e-03 -7.58541942e-01
-1.01102710e-01 -3.06666136e-01 8.25018212e-02 -3.77616644e-01
3.14927012e-01 1.34019524e-01 6.21218197e-02 -8.62196803e-01
-2.71989740e-02 1.38008416e+00 6.12444520e-01 -3.43336523e-01
1.37400997e+00 -2.04715043e-01 -2.26722479e-01 3.56732696e-01
1.11033165e+00 4.27167028e-01 -1.49308467e+00 -2.13497773e-01
-6.09294951e-01 -8.40375483e-01 -8.25437251e-04 -1.00436306e+00
-1.61650050e+00 1.03030944e+00 7.60398030e-01 8.08809772e-02
1.35784519e+00 -1.37198165e-01 7.70033419e-01 -8.92947987e-02
5.55865645e-01 -8.47493470e-01 3.12417120e-01 4.75783944e-01
1.09666443e+00 -9.24088478e-01 -4.38829884e-02 -6.08851194e-01
-6.17650270e-01 1.38239121e+00 6.26264274e-01 -1.14961103e-01
9.98628855e-01 2.13118196e-01 3.07614475e-01 -2.85789579e-01
-2.40934432e-01 -3.13014835e-01 4.25085396e-01 1.01477778e+00
1.76128745e-02 -7.04219565e-02 -1.75376147e-01 4.73045021e-01
-3.53345275e-02 3.13621014e-02 3.67428660e-01 6.48797452e-01
-3.87793481e-01 -1.35989606e+00 -7.14989364e-01 7.67182037e-02
2.66487122e-01 -2.00177561e-02 -5.30904293e-01 4.52562541e-01
2.90195912e-01 8.17586243e-01 3.20746936e-02 -6.61108553e-01
5.84501982e-01 -5.80318607e-02 5.19611716e-01 -3.12754571e-01
-1.48960128e-01 1.98228806e-01 -3.82183552e-01 -4.64460373e-01
-4.73653376e-01 -3.30074012e-01 -1.60627389e+00 -1.60160393e-01
-6.05289564e-02 5.67204133e-03 1.05492830e+00 4.05739963e-01
3.64181459e-01 6.91952333e-02 9.82422590e-01 -1.00516760e+00
1.00041106e-01 -3.95664185e-01 -8.60473990e-01 5.56549072e-01
-8.80846754e-02 -4.14584160e-01 -2.06345454e-01 -6.29645437e-02] | [8.7689847946167, -3.3939883708953857] |
ec2d31b2-3317-4ff1-8350-fed701549108 | aggressionnet-generalised-multi-modal-deep | 2001.05493 | null | https://arxiv.org/abs/2001.05493v2 | https://arxiv.org/pdf/2001.05493v2.pdf | A Unified System for Aggression Identification in English Code-Mixed and Uni-Lingual Texts | Wide usage of social media platforms has increased the risk of aggression, which results in mental stress and affects the lives of people negatively like psychological agony, fighting behavior, and disrespect to others. Majority of such conversations contains code-mixed languages[28]. Additionally, the way used to express thought or communication style also changes from one social media plat-form to another platform (e.g., communication styles are different in twitter and Facebook). These all have increased the complexity of the problem. To solve these problems, we have introduced a unified and robust multi-modal deep learning architecture which works for English code-mixed dataset and uni-lingual English dataset both.The devised system, uses psycho-linguistic features and very ba-sic linguistic features. Our multi-modal deep learning architecture contains, Deep Pyramid CNN, Pooled BiLSTM, and Disconnected RNN(with Glove and FastText embedding, both). Finally, the system takes the decision based on model averaging. We evaluated our system on English Code-Mixed TRAC 2018 dataset and uni-lingual English dataset obtained from Kaggle. Experimental results show that our proposed system outperforms all the previous approaches on English code-mixed dataset and uni-lingual English dataset. | ['Niraj Kumar', 'Anant Khandelwal'] | 2020-01-15 | null | null | null | null | ['style-generalization', 'aggression-identification'] | ['computer-vision', 'natural-language-processing'] | [-8.36431444e-01 -1.26346350e-01 9.58473384e-02 -2.17763454e-01
-3.45813572e-01 -3.25525999e-01 4.28255588e-01 2.35503271e-01
-5.76174140e-01 7.35116720e-01 3.46617699e-01 -2.52818942e-01
6.58068880e-02 -7.18126059e-01 -3.21851403e-01 -3.13628018e-01
-1.76170953e-02 3.71500291e-02 -1.12844974e-01 -5.53969860e-01
4.70934302e-01 3.16525429e-01 -1.26254308e+00 3.68243009e-01
7.78934777e-01 6.13511980e-01 2.94066314e-02 7.02663243e-01
-3.22505891e-01 1.30219448e+00 -3.05228680e-01 -8.46691728e-01
-2.72740304e-01 -2.13141069e-01 -8.78842056e-01 1.20400973e-02
1.62747517e-01 -2.12449804e-01 -3.06866080e-01 1.27922106e+00
6.19215488e-01 1.31617457e-01 7.09863186e-01 -1.19472909e+00
-1.03622389e+00 7.81427383e-01 -7.39712179e-01 1.23797543e-01
4.35838431e-01 6.09903596e-03 5.21512270e-01 -7.32415378e-01
2.58918166e-01 1.66600668e+00 1.00684071e+00 4.77777004e-01
-7.88791358e-01 -8.82176042e-01 -3.25593740e-01 1.61560759e-01
-1.37984836e+00 -3.20736200e-01 6.91724479e-01 -8.47505152e-01
1.00960517e+00 -2.09674031e-01 4.49388117e-01 1.29254961e+00
5.67646563e-01 3.08656305e-01 1.12051070e+00 -1.57129124e-01
-2.58295298e-01 6.72191978e-01 2.72275358e-01 9.48016107e-01
-1.88980162e-01 -3.46833229e-01 -2.20198706e-01 -3.01545143e-01
-6.87087625e-02 1.21344641e-01 1.34972245e-01 3.61845285e-01
-8.04750741e-01 1.00798535e+00 3.40404421e-01 7.69729555e-01
-1.73848480e-01 4.00600955e-02 9.57958519e-01 5.59616506e-01
2.48411909e-01 -4.86976057e-02 -3.42390269e-01 -5.28710663e-01
-5.90310574e-01 1.84195504e-01 7.81719148e-01 7.67955661e-01
8.13028812e-01 9.06525105e-02 2.76878387e-01 1.45882130e+00
6.36942625e-01 4.29631412e-01 1.18781340e+00 -5.68726897e-01
4.33174878e-01 9.28796351e-01 -3.53224009e-01 -1.65653467e+00
-8.27471197e-01 -2.56501079e-01 -8.04170012e-01 4.67264205e-02
-1.19149387e-02 -4.36157227e-01 -2.58832961e-01 1.78047609e+00
-3.42356972e-02 -6.92640617e-02 2.42558271e-01 2.50032187e-01
1.10870755e+00 6.61504865e-01 1.06020741e-01 3.51942144e-02
1.17980194e+00 -9.96441007e-01 -8.19533050e-01 -1.54671952e-01
9.10178065e-01 -1.01087713e+00 1.22004426e+00 4.16579545e-01
-9.27805543e-01 -5.90500414e-01 -9.95204151e-01 -2.70250589e-01
-8.11021745e-01 1.55432567e-01 2.68764228e-01 8.82790148e-01
-9.54217196e-01 4.05230045e-01 -4.35970366e-01 -6.33728027e-01
2.02106372e-01 3.28761548e-01 -6.65457129e-01 1.97859451e-01
-1.29694831e+00 9.17640626e-01 4.02927071e-01 3.01315077e-02
-5.18080294e-01 -6.84036985e-02 -9.75915372e-01 -2.54599720e-01
-8.32671300e-02 -1.01654038e-01 9.42914605e-01 -1.11063504e+00
-1.44961703e+00 1.25596583e+00 1.93425432e-01 -2.94262946e-01
4.61944044e-01 -1.42297804e-01 -8.89916956e-01 -3.70666504e-01
3.15626085e-01 4.24562067e-01 6.26858950e-01 -8.84110153e-01
-2.99123794e-01 -4.12340790e-01 3.34452420e-01 7.96613544e-02
-7.59511530e-01 3.05693030e-01 9.34859039e-04 -2.27415368e-01
-4.21047688e-01 -1.10532379e+00 1.86303526e-01 -2.92370409e-01
-4.54902381e-01 -4.33381200e-01 9.56045032e-01 -6.49313807e-01
1.60229957e+00 -2.65399885e+00 3.62377055e-02 -3.32371324e-01
2.23038167e-01 2.05775410e-01 2.06784815e-01 6.30638301e-01
4.65237908e-02 4.00064111e-01 -1.72605082e-01 -4.50169325e-01
-4.60223444e-02 3.21322054e-01 8.53497982e-02 6.41606450e-01
-2.31030479e-01 2.42687598e-01 -6.64040864e-01 -7.24840581e-01
1.24056153e-01 3.66910219e-01 -9.23775554e-01 9.58251208e-02
3.40542912e-01 3.02113354e-01 -3.01824480e-01 3.97846639e-01
6.39248073e-01 1.37576222e-01 -3.73308867e-01 1.22579813e-01
-4.41101164e-01 2.18103603e-02 -8.17544341e-01 1.64683664e+00
-1.12947488e+00 8.27891469e-01 3.42018083e-02 -6.59869850e-01
8.77363682e-01 2.23800257e-01 7.53596127e-02 -3.49649817e-01
7.83476830e-01 3.41705143e-01 2.85470486e-01 -1.34592628e+00
4.08716768e-01 -1.72810763e-01 -2.87761152e-01 2.28964254e-01
9.65378061e-02 -8.97074342e-02 8.21033791e-02 5.59884533e-02
7.26018608e-01 -2.43266344e-01 3.65934491e-01 -3.98156106e-01
1.10167897e+00 -4.55499202e-01 3.39146107e-01 2.30719253e-01
-6.08004391e-01 9.87795368e-02 8.20907533e-01 -3.59327197e-01
-7.27890790e-01 -6.62537396e-01 -3.23008537e-01 1.25360513e+00
-1.54559895e-01 -3.04125786e-01 -7.36549735e-01 -4.62143421e-01
-8.27612877e-02 6.99107528e-01 -7.18746364e-01 -3.55741411e-01
-2.67247409e-01 -7.35923409e-01 9.60530102e-01 -1.02151193e-01
1.00958133e+00 -1.19431818e+00 -3.11437771e-02 2.64422476e-01
-1.31013677e-01 -1.19208729e+00 -1.00610726e-01 -1.47194788e-01
-3.04949999e-01 -1.01043212e+00 -1.22484483e-01 -8.59796643e-01
1.82027534e-01 -2.91845679e-01 9.13586318e-01 3.13904136e-01
-1.29806548e-01 9.74862929e-03 -3.56487721e-01 -3.67679894e-01
-6.06149793e-01 1.37533203e-01 1.81343704e-01 1.92348376e-01
6.52825892e-01 -7.19581306e-01 -1.85805351e-01 -1.59830049e-01
-6.64160311e-01 -1.53551415e-01 2.09570006e-01 5.91320395e-01
-4.57252979e-01 -3.18257846e-02 6.60144210e-01 -9.57198501e-01
1.17781734e+00 -1.10183465e+00 3.22432965e-02 -3.05012524e-01
-8.20626840e-02 -2.61533320e-01 9.98942852e-01 -1.45036504e-01
-7.85297990e-01 -4.72890645e-01 -6.38015091e-01 3.12872231e-02
-2.85583645e-01 6.17161572e-01 1.44670814e-01 -4.18009562e-03
8.00933719e-01 8.43789503e-02 -1.28511637e-01 -2.98030168e-01
6.28576726e-02 1.56755435e+00 1.51517600e-01 -3.79225343e-01
4.07264769e-01 3.23076993e-01 -3.66730481e-01 -1.09569812e+00
-5.96901774e-01 -8.01558197e-02 -5.41987181e-01 -3.74630362e-01
1.36475933e+00 -8.86791050e-01 -1.02766979e+00 9.08775806e-01
-1.31650460e+00 -4.83886898e-02 8.37866724e-01 5.48148215e-01
-2.84900457e-01 2.89116234e-01 -9.61933792e-01 -7.10158467e-01
-8.22311789e-02 -1.36858797e+00 6.40111625e-01 1.05375819e-01
-2.47523710e-01 -1.23289204e+00 2.42024690e-01 6.31031632e-01
4.65555340e-01 5.22017419e-01 1.08118331e+00 -6.87672377e-01
4.18348968e-01 -1.31480619e-01 -2.88506746e-01 7.66662598e-01
1.63822532e-01 3.25522870e-01 -1.06437016e+00 -7.10319430e-02
1.65508941e-01 -8.43802571e-01 4.51691449e-01 -1.69655070e-01
1.13507640e+00 -3.11745644e-01 6.35102913e-02 3.88459146e-01
1.35268879e+00 2.56016076e-01 3.54337335e-01 3.15810144e-01
8.99885118e-01 4.73524392e-01 -3.64739373e-02 5.23658812e-01
7.28395224e-01 1.91582546e-01 3.03100139e-01 2.64509529e-01
1.92059398e-01 1.67482764e-01 7.15889573e-01 1.55587280e+00
1.37831822e-01 -9.35442969e-02 -1.20504582e+00 5.49577296e-01
-1.64815640e+00 -9.77441490e-01 -4.56135452e-01 1.53587222e+00
7.42430866e-01 1.34673238e-01 1.18728198e-01 8.15108195e-02
7.00334430e-01 4.86084074e-02 -2.87068427e-01 -1.24271750e+00
1.45260200e-01 -2.92676419e-01 3.02577823e-01 6.97256505e-01
-1.09426022e+00 8.38615060e-01 4.95701313e+00 8.92911911e-01
-1.59811306e+00 6.34877264e-01 5.44490933e-01 5.50254062e-02
7.69426301e-02 -5.48675299e-01 -5.75362682e-01 7.38611698e-01
1.24721086e+00 -2.71675199e-01 5.50827682e-01 9.84386921e-01
2.49566615e-01 -1.77409477e-03 -8.24067593e-01 1.35236359e+00
3.33462030e-01 -6.99179113e-01 -3.31012696e-01 -1.27657011e-01
5.13343036e-01 5.27454019e-01 1.07340030e-01 8.89297307e-01
2.97766089e-01 -1.06520486e+00 6.04339480e-01 2.14047819e-01
4.94555444e-01 -1.06494868e+00 8.74114513e-01 5.47755599e-01
-8.97790909e-01 -3.70273709e-01 -3.21889281e-01 -5.52512825e-01
-2.28073552e-01 2.66229510e-01 -3.15768510e-01 1.15063347e-01
6.80670142e-01 1.00448108e+00 -8.00044596e-01 4.71373439e-01
1.64238736e-01 2.83435881e-01 -4.72755805e-02 -3.57939690e-01
7.05652356e-01 -1.22909240e-01 2.73999572e-01 1.53279853e+00
2.12255865e-01 -2.68993765e-01 1.72255024e-01 6.63582563e-01
-2.20934808e-01 6.11420214e-01 -1.04118311e+00 -1.27018318e-01
2.09091291e-01 1.42142439e+00 -3.16272527e-01 -2.61905819e-01
-8.00163329e-01 6.92784071e-01 6.16160512e-01 -5.67180999e-02
-1.05006325e+00 -6.36722326e-01 4.95667905e-01 1.52916070e-02
-4.71840709e-01 -2.84492642e-01 -4.02316153e-02 -1.19998491e+00
-3.17506462e-01 -9.68742669e-01 1.42484158e-01 -6.12386823e-01
-1.40239680e+00 9.95818138e-01 -2.30674118e-01 -1.04925346e+00
-1.40448079e-01 -6.82651341e-01 -4.62320507e-01 5.93168736e-01
-1.19283092e+00 -1.16417348e+00 -1.97374448e-01 8.65085840e-01
7.10584760e-01 -5.84397733e-01 7.24949539e-01 1.11316133e+00
-8.86024296e-01 5.95529854e-01 3.02007478e-02 4.45657253e-01
7.90849745e-01 -9.93953407e-01 -3.28298241e-01 4.46475774e-01
-4.53989327e-01 8.99080336e-01 5.06820977e-01 -2.04439268e-01
-1.09205616e+00 -1.06552720e+00 8.16951513e-01 -2.30149195e-01
1.19630325e+00 -4.52300489e-01 -6.41777933e-01 6.98816597e-01
6.44237816e-01 -2.68574446e-01 1.01845598e+00 8.42352733e-02
-2.95850784e-01 3.85870188e-02 -1.34809172e+00 6.86322868e-01
6.51050925e-01 -8.39104116e-01 -5.54158449e-01 4.60846394e-01
5.11754334e-01 -1.50605842e-01 -7.29568303e-01 -1.21847622e-01
3.93057942e-01 -1.59644473e+00 4.61366683e-01 -4.86435920e-01
9.01085138e-01 -2.35897433e-02 -3.32152069e-01 -1.11397362e+00
-2.59429604e-01 -3.24886233e-01 5.45693815e-01 1.30715477e+00
3.73478472e-01 -6.72315955e-01 -3.48768570e-02 4.36226457e-01
-3.65397871e-01 -7.10855603e-01 -8.89518499e-01 -2.08874971e-01
6.19509220e-01 -4.62127626e-01 2.81411111e-01 1.40333343e+00
1.11487150e-01 4.88359720e-01 -4.93967831e-01 -3.14299613e-02
1.38779819e-01 -4.29396421e-01 7.21969068e-01 -1.01735640e+00
-1.70843586e-01 -7.55153775e-01 -7.28309810e-01 -4.17078584e-01
7.26094127e-01 -9.71954823e-01 -4.02631342e-01 -1.10408282e+00
2.10118011e-01 -1.57525748e-01 3.09924409e-02 4.59519774e-01
2.33331054e-01 4.79621559e-01 1.06694266e-01 -2.05809191e-01
-4.11857635e-01 4.27241981e-01 1.11204350e+00 -8.67529884e-02
-1.00060605e-01 -4.62639362e-01 -5.61839700e-01 1.21775389e+00
8.24430585e-01 -4.48774278e-01 -2.80835718e-01 -3.58194560e-01
7.77064562e-01 3.73208337e-02 1.49572119e-01 -1.34410775e+00
5.40883914e-02 2.29012482e-02 -8.44894871e-02 -1.96765140e-01
1.81496590e-01 -8.86775017e-01 -2.98387587e-01 7.15872288e-01
-3.17539126e-01 2.59520054e-01 3.74796599e-01 2.23723516e-01
-3.68212163e-01 -5.23971021e-01 1.11917388e+00 -2.44961172e-01
-3.60444456e-01 5.34600504e-02 -7.71248281e-01 2.28778705e-01
1.07252908e+00 -3.75288837e-02 -1.72388732e-01 -2.26205215e-01
-8.48308563e-01 1.80450112e-01 2.86754072e-01 8.02904487e-01
1.13422230e-01 -1.38168728e+00 -8.08136106e-01 1.10657111e-01
1.16662934e-01 -5.68822503e-01 4.17376369e-01 1.02334714e+00
-8.81974995e-01 3.25293355e-02 -5.74116111e-01 -2.26315260e-01
-1.22377646e+00 3.90185922e-01 5.89974761e-01 1.40956163e-01
-1.37672484e-01 7.76809871e-01 -2.42582515e-01 -9.51990068e-01
-1.68658711e-03 -3.00890893e-01 -6.86251879e-01 3.67568433e-01
3.09328526e-01 4.79827046e-01 7.14362739e-03 -1.20403326e+00
-4.66399312e-01 6.02044284e-01 -1.10587478e-01 3.23766209e-02
1.24043250e+00 -2.38263845e-01 -8.42990279e-01 1.14841461e+00
1.87178171e+00 5.34073770e-01 -1.29669785e-01 1.13586776e-01
-2.48417720e-01 -3.23934585e-01 3.30000110e-02 -3.75257313e-01
-8.46584499e-01 1.07896173e+00 5.79711556e-01 6.05228305e-01
7.27288187e-01 -4.04968172e-01 9.18675363e-01 5.32082856e-01
2.94002980e-01 -1.20650220e+00 9.17730704e-02 9.63081598e-01
9.26295638e-01 -1.45353711e+00 -3.68706435e-01 1.36641458e-01
-6.97671652e-01 1.32929146e+00 8.66583228e-01 -3.61434132e-01
1.22842658e+00 2.52679527e-01 2.75700271e-01 -2.47472048e-01
-6.29711270e-01 1.91383481e-01 -1.67228997e-01 1.67479083e-01
1.09679258e+00 8.38052928e-02 -6.29390121e-01 7.65571952e-01
-6.63411617e-01 -3.71881127e-02 1.02908099e+00 4.96528715e-01
-4.72603530e-01 -8.69762778e-01 -3.47623795e-01 3.46950561e-01
-8.81635904e-01 -1.05400771e-01 -3.26464802e-01 8.14188480e-01
7.65154958e-01 1.22467577e+00 -3.11293709e-03 -7.13906646e-01
-2.41948497e-02 1.85743973e-01 -4.63203304e-02 -6.43127441e-01
-1.13362813e+00 -3.62507910e-01 1.97573245e-01 -2.81649977e-01
-4.32117581e-01 -4.65271503e-01 -1.34571242e+00 -8.09455812e-01
-6.81824833e-02 -2.10769307e-02 6.53402269e-01 1.02636933e+00
1.81047127e-01 5.32292306e-01 6.42411172e-01 -6.21271074e-01
-2.29458764e-01 -1.38391304e+00 -5.18142402e-01 5.59886336e-01
4.73825097e-01 -7.91046798e-01 -4.96266127e-01 -8.82451609e-02] | [8.957769393920898, 10.588452339172363] |
a14e8623-7480-40a4-89c8-6cc792d4c91c | maximising-weather-forecasting-accuracy | 2301.12471 | null | https://arxiv.org/abs/2301.12471v1 | https://arxiv.org/pdf/2301.12471v1.pdf | Maximising Weather Forecasting Accuracy through the Utilisation of Graph Neural Networks and Dynamic GNNs | Weather forecasting is an essential task to tackle global climate change. Weather forecasting requires the analysis of multivariate data generated by heterogeneous meteorological sensors. These sensors comprise of ground-based sensors, radiosonde, and sensors mounted on satellites, etc., To analyze the data generated by these sensors we use Graph Neural Networks (GNNs) based weather forecasting model. GNNs are graph learning-based models which show strong empirical performance in many machine learning approaches. In this research, we investigate the performance of weather forecasting using GNNs and traditional Machine learning-based models. | ['Shreelakshmi C R', 'Surya Durbha', 'Gaganpreet Singh'] | 2023-01-29 | null | null | null | null | ['weather-forecasting'] | ['miscellaneous'] | [-1.32368773e-01 -2.30951354e-01 9.09993127e-02 -4.93238986e-01
5.39685845e-01 -2.72346228e-01 6.42884910e-01 3.63599151e-01
2.16782675e-03 9.79774356e-01 1.23366013e-01 -7.65067279e-01
-1.05016112e-01 -1.60243773e+00 -2.00804010e-01 -6.81363225e-01
-6.76657200e-01 2.22338811e-01 3.54700714e-01 -9.18062627e-01
1.02031060e-01 6.67404056e-01 -1.19635391e+00 -3.64407748e-01
7.90703952e-01 1.01126182e+00 -8.29567984e-02 1.29877973e+00
-3.85640204e-01 8.62046242e-01 -5.16047299e-01 1.98870420e-01
1.38377830e-01 -4.14549768e-01 -6.39449805e-02 -4.17781413e-01
3.24577183e-01 1.25438526e-01 -3.66336077e-01 1.04358494e+00
2.59371698e-01 6.43206716e-01 7.70119667e-01 -1.24031723e+00
-4.29756373e-01 4.32848454e-01 -2.90906966e-01 4.82044637e-01
3.24707240e-01 -1.87493175e-01 7.16884196e-01 -2.80657798e-01
2.62737572e-01 1.11226904e+00 9.04813290e-01 1.43069923e-02
-5.09410739e-01 -3.16554546e-01 1.21188387e-01 9.96181965e-02
-1.02482605e+00 -1.22550726e-02 7.86246836e-01 -5.60666978e-01
1.00734496e+00 3.76545668e-01 7.75864720e-01 3.98369551e-01
9.73945320e-01 -3.74519497e-01 1.22282469e+00 -3.77470106e-01
4.70968664e-01 -3.80755931e-01 4.90551680e-01 7.88829923e-01
7.37578213e-01 3.62464577e-01 -3.84520203e-01 -3.59752983e-01
4.42616403e-01 3.68209809e-01 -2.95054585e-01 4.73664761e-01
-5.44930577e-01 9.83997464e-01 9.76076663e-01 2.82894909e-01
-6.05429113e-01 2.27129117e-01 -3.10078426e-03 5.09482563e-01
9.92968440e-01 2.41213292e-01 -4.83644903e-01 3.00032586e-01
-8.82643819e-01 1.78080425e-01 1.06172979e+00 3.62275064e-01
7.90642083e-01 9.12530065e-01 7.03266263e-01 1.87739044e-01
8.46358001e-01 1.17297304e+00 2.23790616e-01 -1.65663034e-01
5.10842502e-01 7.12369084e-01 -1.23987369e-01 -1.81756043e+00
-1.26707757e+00 -3.23822081e-01 -1.54038882e+00 2.56549269e-01
4.76846620e-02 -9.37250078e-01 -9.68007624e-01 7.55933583e-01
5.63858271e-01 7.06729233e-01 1.80008784e-01 8.32190692e-01
1.23061883e+00 1.38772714e+00 2.35405654e-01 -2.12558880e-01
6.15313888e-01 -7.75236666e-01 -8.83232892e-01 6.54876232e-02
7.90128291e-01 -3.06590289e-01 5.31799972e-01 3.56340796e-01
-2.04100639e-01 -5.30840456e-01 -9.96923447e-01 6.55341029e-01
-1.17789578e+00 -4.36988086e-01 5.73310971e-01 6.74869835e-01
-1.36642563e+00 8.30670655e-01 -5.68892539e-01 -3.75543594e-01
-3.95191669e-01 5.95074110e-02 -1.58512250e-01 -6.94997311e-02
-1.75774944e+00 1.07264590e+00 5.86682141e-01 6.67348981e-01
-2.84236759e-01 -2.29093656e-01 -1.22300637e+00 -1.99263573e-01
-1.29478127e-01 -5.63886285e-01 4.05688822e-01 -3.31905812e-01
-1.50063694e+00 -8.82635266e-03 -8.43043476e-02 -4.80906546e-01
-1.84697971e-01 3.54224682e-01 -1.21705949e+00 -2.66856223e-01
-5.39862752e-01 -5.56315482e-01 7.96823561e-01 -1.06094062e+00
-5.13293028e-01 -4.25094217e-01 -3.02960128e-01 -3.92285585e-02
-5.27383201e-02 -2.70090163e-01 6.22968495e-01 -4.96733636e-01
1.30413264e-01 -9.53249753e-01 -7.61125445e-01 -7.11169541e-01
-4.19377804e-01 -1.37939766e-01 9.25124347e-01 -9.42879081e-01
1.23524702e+00 -1.93163657e+00 -2.66784579e-01 9.98868287e-01
4.17841405e-01 4.11260605e-01 -1.21527687e-01 6.98888481e-01
1.44297883e-01 1.77086040e-01 -3.25097710e-01 5.64074926e-02
-2.66621917e-01 7.51622021e-01 -4.16943491e-01 5.89490414e-01
-3.75849336e-01 8.65572393e-01 -6.67255521e-01 -9.42631587e-02
4.96991426e-01 3.16564322e-01 3.60181034e-01 3.50434214e-01
-5.39691746e-01 4.78519380e-01 -5.42279661e-01 5.50568342e-01
7.82997787e-01 -1.71867356e-01 1.78727254e-01 1.60030946e-01
-1.44071385e-01 -1.47162929e-01 -1.16251969e+00 8.87431026e-01
-2.90755779e-01 5.49911499e-01 -9.44024324e-02 -1.16919351e+00
1.56325507e+00 2.81500429e-01 -3.27649005e-02 -6.64858699e-01
7.72096738e-02 -4.46249619e-02 -1.22094415e-01 -7.95810044e-01
6.86798394e-01 -1.17960740e-02 1.61892384e-01 3.32800329e-01
-2.17039183e-01 -4.91456449e-01 -2.20571877e-03 -1.50393859e-01
6.63758695e-01 -1.60334870e-01 4.17492867e-01 -2.17759386e-01
2.68605143e-01 3.60036880e-01 2.88169175e-01 5.28754056e-01
7.03375936e-02 1.73647806e-01 1.59189224e-01 -1.23957884e+00
-7.51596630e-01 -4.37280327e-01 2.60433823e-01 1.15575290e+00
-3.83629203e-02 -2.83550709e-01 -8.17706212e-02 -3.32463264e-01
3.10301512e-01 8.03202212e-01 -5.79854846e-01 3.06906879e-01
-1.99548140e-01 -1.19076598e+00 4.42034543e-01 2.89448738e-01
2.66036719e-01 -7.84800947e-01 -1.03602996e-02 4.16201741e-01
2.96301037e-01 -9.35087740e-01 6.49147093e-01 -8.27896595e-02
-1.24559891e+00 -1.08547807e+00 -4.00750369e-01 -1.93642210e-02
4.49326336e-01 3.69918317e-01 1.29915726e+00 4.64239925e-01
1.53426856e-01 3.13225210e-01 -5.81504285e-01 -6.95309341e-01
-1.27190381e-01 3.73751111e-02 2.27507219e-01 1.13351122e-01
-2.38559134e-02 -9.33528900e-01 -1.93760470e-01 2.09790114e-02
-8.04985404e-01 -4.88208443e-01 1.11695528e-01 1.03849493e-01
2.01819628e-01 5.99957883e-01 2.63242513e-01 -7.87585199e-01
8.69091034e-01 -1.17326295e+00 -1.16723657e+00 3.83868873e-01
-8.35615754e-01 -2.76360303e-01 7.54899979e-01 2.00262651e-01
-7.01552987e-01 -1.41804174e-01 2.27056220e-01 -2.02335287e-02
-4.91870731e-01 1.52172935e+00 4.80821937e-01 -6.55772209e-01
9.19274986e-01 2.32383624e-01 -2.72478640e-01 -4.01586235e-01
4.47012216e-01 5.95717788e-01 2.65300483e-01 -1.33034512e-01
7.84804046e-01 2.29677916e-01 7.72313416e-01 -1.20951653e+00
-5.66631138e-01 -4.24827665e-01 -4.93640453e-01 -8.77609193e-01
9.65324759e-01 -8.44393313e-01 -5.39784491e-01 5.27577758e-01
-1.07565820e+00 -5.07041335e-01 3.83102417e-01 8.03605914e-01
4.03427660e-01 1.78841040e-01 -3.34210575e-01 -1.22863305e+00
-4.74442035e-01 -2.55097389e-01 4.00263935e-01 6.21642888e-01
-4.85834889e-02 -1.88190341e+00 8.35097253e-01 -1.73114747e-01
9.41474259e-01 1.01344097e+00 5.05659163e-01 -6.63377345e-01
-2.76440203e-01 -4.83242512e-01 -2.61215150e-01 5.81563413e-02
2.41587058e-01 6.17281258e-01 -5.41643322e-01 -9.17924792e-02
-3.14858288e-01 2.34984159e-01 9.04345214e-01 5.87412357e-01
5.04326522e-01 -3.41582239e-01 -2.65818954e-01 7.54116595e-01
1.78821468e+00 2.54451185e-01 3.34629267e-01 3.05942804e-01
1.15566421e+00 6.72687948e-01 -2.38078590e-02 2.51727730e-01
6.71590745e-01 -2.60305434e-01 7.58188248e-01 -1.42003402e-01
4.26684529e-01 1.60830766e-01 1.05749056e-01 1.29950845e+00
-8.55757356e-01 -7.59719312e-01 -1.65344012e+00 7.84856603e-02
-2.08737707e+00 -7.65330434e-01 -1.14490604e+00 1.73505592e+00
1.23808524e-02 1.13164531e-02 -1.12106204e-01 1.02035508e-01
6.63693547e-01 7.28866458e-01 -2.43722230e-01 -8.18612874e-01
-4.28422660e-01 2.78547138e-01 9.27008748e-01 6.80909455e-01
-1.09201753e+00 6.80887103e-01 7.23726749e+00 9.74212512e-02
-1.45828116e+00 3.69076543e-02 2.45609835e-01 5.20043910e-01
-5.42421699e-01 -1.74188957e-01 -6.30689025e-01 3.91121596e-01
1.50381839e+00 -1.13475904e-01 3.40757787e-01 8.60192835e-01
4.90695238e-01 -4.22539622e-01 -1.17512763e-01 5.50311208e-01
-1.65274683e-02 -1.38416374e+00 7.76557103e-02 -1.49060413e-01
9.70882535e-01 6.98288620e-01 -4.48385507e-01 -6.60735443e-02
8.65593314e-01 -1.11213231e+00 -3.10316026e-01 1.11075139e+00
2.26320371e-01 -4.39796984e-01 1.02457976e+00 2.36154199e-01
-1.53651261e+00 2.24546090e-01 -6.75584793e-01 -9.78529930e-01
1.48871034e-01 1.47454619e+00 -5.85451305e-01 1.15527952e+00
7.46171951e-01 1.13909805e+00 -3.11735153e-01 1.13073337e+00
-5.30856669e-01 1.27743661e+00 -5.87911069e-01 -4.99819815e-01
3.62910330e-01 -7.60003090e-01 4.42729145e-01 1.05735707e+00
4.77950513e-01 5.90071797e-01 3.92640531e-01 2.39818655e-02
4.22493160e-01 2.02357918e-01 -1.14050615e+00 1.51224390e-01
-3.76338698e-02 1.13086689e+00 -4.83273178e-01 -6.18223846e-01
-4.00535107e-01 1.13362141e-01 -5.23535088e-02 7.09011078e-01
-3.53092998e-01 -4.96497184e-01 3.69399339e-01 -1.18276358e-01
-3.88344415e-02 -8.73420537e-01 -4.07586396e-01 -1.06526291e+00
-3.02482605e-01 -2.14479417e-01 4.90883648e-01 -1.18280506e+00
-1.52652800e+00 8.16976845e-01 -2.47455135e-01 -1.22160399e+00
-2.73296684e-01 -7.72617340e-01 -1.49695420e+00 1.22001863e+00
-2.02921915e+00 -9.96196926e-01 -7.48953164e-01 9.34178174e-01
-4.13426101e-01 -2.38375857e-01 9.72638011e-01 -1.54746696e-01
-5.42352736e-01 -3.64344448e-01 2.92548120e-01 1.34047538e-01
2.28759795e-01 -1.19850695e+00 7.58863568e-01 1.00019550e+00
-5.79025820e-02 -1.18643060e-01 6.37290120e-01 -1.18756843e+00
-1.35848117e+00 -1.47560894e+00 1.10827541e+00 1.67251915e-01
1.02395952e+00 1.92566052e-01 -1.00633228e+00 6.98524654e-01
2.29562595e-01 3.38775605e-01 7.67859340e-01 2.19838247e-01
-2.88091451e-01 -3.67088795e-01 -8.34181428e-01 1.54650614e-01
3.99500221e-01 -1.18820310e-01 -6.73091412e-01 6.79193497e-01
5.92585504e-01 -3.02884966e-01 -1.01903594e+00 4.31409121e-01
2.02309355e-01 -8.06687534e-01 4.99937326e-01 -7.68739283e-01
2.81821936e-01 -4.64978278e-01 -2.18097284e-01 -1.99019432e+00
-4.67653751e-01 -3.13282073e-01 -1.30187407e-01 5.91405213e-01
6.11624241e-01 -1.37485492e+00 6.00389004e-01 6.95193291e-01
1.72937408e-01 -6.62686080e-02 -8.29678833e-01 -6.51943386e-01
-1.64371654e-01 -4.33888972e-01 9.46074426e-01 1.47957671e+00
6.53102668e-03 1.42298371e-01 -5.07280588e-01 7.98242629e-01
6.72461569e-01 1.85380667e-01 7.04647839e-01 -1.91279876e+00
1.70043558e-01 -2.05732033e-01 -7.50850141e-01 -2.84692496e-01
-1.32226422e-01 -2.28589162e-01 -2.65449554e-01 -1.97474182e+00
-9.82335985e-01 -4.99485224e-01 -4.42207396e-01 5.01629055e-01
-1.35740131e-01 1.28890976e-01 -7.39376023e-02 2.27413878e-01
-2.15534553e-01 4.06771600e-01 9.75929558e-01 -8.31272602e-02
-3.94144922e-01 1.95708320e-01 2.41095409e-01 6.89849496e-01
1.38800108e+00 -5.60367346e-01 -1.89839229e-01 -4.04680908e-01
1.13601458e+00 3.18461567e-01 7.56021291e-02 -1.06854439e+00
6.35742843e-01 -7.28692174e-01 4.13904607e-01 -6.18083596e-01
-1.80865392e-01 -8.00478816e-01 4.45390970e-01 5.69277346e-01
3.31932485e-01 4.79809999e-01 2.17638284e-01 5.66504419e-01
-6.37387574e-01 1.72801748e-01 3.54924761e-02 -1.92170516e-01
-1.01103425e+00 5.66085517e-01 -7.07288206e-01 -5.48537254e-01
8.53202343e-01 1.21958509e-01 -6.19747460e-01 -7.13014245e-01
-8.50228846e-01 2.99697191e-01 -1.73445746e-01 8.15049484e-02
4.45951998e-01 -1.01613748e+00 -9.37848151e-01 2.05463573e-01
-8.37297514e-02 -7.15440931e-03 1.06218137e-01 4.96127069e-01
-1.05132735e+00 2.77139395e-01 -9.64422077e-02 -3.27070862e-01
-8.44428360e-01 2.86215037e-01 6.03842854e-01 -1.99602947e-01
-1.62993386e-01 7.72029042e-01 -7.66600192e-01 -9.11279321e-01
-3.86250645e-01 -5.91418564e-01 -8.41231406e-01 1.44381836e-01
4.28224266e-01 6.27140522e-01 2.85871804e-01 -5.76796114e-01
-2.66958892e-01 8.54836345e-01 1.00906265e+00 2.30428725e-02
1.48203671e+00 -1.25945210e-01 -6.53749824e-01 7.86268115e-01
7.64830709e-01 -2.07274243e-01 -4.22789037e-01 -3.09229642e-01
-3.58659625e-02 -1.11070976e-01 4.01940495e-01 -6.18724287e-01
-1.29222810e+00 8.01136672e-01 3.15467298e-01 1.17415357e+00
1.01523244e+00 -7.64702320e-01 7.19387889e-01 7.19897866e-01
1.01684399e-01 -8.86318624e-01 -8.33166897e-01 8.58462036e-01
6.32156968e-01 -1.34598243e+00 3.08366507e-01 -2.94585973e-01
-1.94976583e-01 1.72113228e+00 2.50462025e-01 -3.77132148e-01
1.60958123e+00 1.08048722e-01 4.95595634e-01 -3.94804329e-01
-4.37455863e-01 -3.42280596e-01 6.07362807e-01 6.99003339e-01
4.00043391e-02 7.04468846e-01 5.76861240e-02 4.17492166e-02
-3.84058237e-01 9.75653529e-02 5.31318963e-01 8.69446635e-01
-7.73070455e-01 -6.25742614e-01 -8.77340376e-01 7.55962789e-01
-6.39400119e-03 -4.89651948e-01 -2.57805169e-01 3.78903806e-01
-2.66806811e-01 1.42386055e+00 1.90708324e-01 -6.98124111e-01
2.95251608e-01 -5.29253297e-02 -3.81571263e-01 -1.39276534e-01
-4.40430135e-01 -6.37452424e-01 1.80705220e-01 -3.87934148e-01
-9.00294960e-01 -1.01404257e-01 -1.06446421e+00 -9.56096947e-01
-1.77934065e-01 3.75614107e-01 1.31893098e+00 1.28795338e+00
3.37296873e-01 5.72799146e-01 9.51380074e-01 -9.44060862e-01
2.69543022e-01 -9.34435308e-01 -1.10018909e+00 -1.82298824e-01
4.99920160e-01 -3.43549877e-01 -5.81826329e-01 -1.22737728e-01] | [6.630339622497559, 2.7950057983398438] |
951ae201-a010-4b67-baaf-251275d084e9 | flexible-and-explainable-solutions-for-multi | 2109.08299 | null | https://arxiv.org/abs/2109.08299v1 | https://arxiv.org/pdf/2109.08299v1.pdf | Flexible and Explainable Solutions for Multi-Agent Path Finding Problems | The multi-agent path finding (MAPF) problem is a combinatorial search problem that aims at finding paths for multiple agents (e.g., robots) in an environment (e.g., an autonomous warehouse) such that no two agents collide with each other, and subject to some constraints on the lengths of paths. The real-world applications of MAPF require flexibility (e.g., solving variations of MAPF) as well as explainability. In this study, both of these challenges are addressed and some flexible and explainable solutions for MAPF and its variants are introduced. | ['Aysu Bogatarkan'] | 2021-09-17 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [-1.98813155e-01 3.72743428e-01 -9.45651159e-02 -1.64282426e-01
-3.61367092e-02 -1.05096614e+00 1.71648413e-01 5.92065513e-01
-1.29049987e-01 1.03105438e+00 -5.38988352e-01 -3.43471169e-01
-9.68551993e-01 -9.88093495e-01 -6.00731015e-01 -5.20447671e-01
-6.35203123e-01 1.04925466e+00 3.37057650e-01 -6.18917286e-01
3.06416392e-01 5.31028986e-01 -8.45770359e-01 -5.47651112e-01
8.45282435e-01 5.22295058e-01 8.82643819e-01 4.90031868e-01
-1.84994459e-01 1.77402318e-01 -6.12322688e-01 -1.97003379e-01
3.66905183e-01 -2.25287184e-01 -1.11396623e+00 2.34509885e-01
-6.43001497e-01 -4.62494902e-02 -2.90025115e-01 1.06998456e+00
-3.59508544e-01 2.59360373e-01 4.56399858e-01 -2.37365985e+00
-5.74135244e-01 6.66302681e-01 -6.61732912e-01 -3.05448502e-01
5.40679812e-01 6.32153079e-02 6.40951037e-01 -3.22292984e-01
8.10538411e-01 1.23883545e+00 2.08479047e-01 3.50917786e-01
-9.85852957e-01 -3.42436954e-02 4.98394907e-01 2.93146551e-01
-1.32286060e+00 9.02561769e-02 2.40747929e-01 4.73378832e-03
1.02063859e+00 3.31712306e-01 4.72006142e-01 4.94731069e-01
9.06932175e-01 3.59081566e-01 6.23616219e-01 -2.81865686e-01
4.60357219e-01 -1.07619008e-02 1.65002450e-01 4.62441534e-01
8.19212556e-01 -2.89354809e-02 -2.10928574e-01 -4.00198638e-01
8.10011208e-01 6.66358396e-02 -6.00364171e-02 -7.59796858e-01
-1.42893004e+00 1.01464677e+00 4.48207885e-01 -3.81262228e-02
-6.23543143e-01 1.27498522e-01 -5.14704287e-02 2.42663652e-01
-6.08509123e-01 9.68356252e-01 -5.82393229e-01 6.82014450e-02
2.05179885e-01 5.24289250e-01 8.40459049e-01 1.62342453e+00
1.06267369e+00 -2.55425751e-01 5.23784816e-01 1.16154842e-01
4.99004334e-01 4.71632957e-01 -2.76614249e-01 -1.23195124e+00
5.97236812e-01 7.25755870e-01 6.74582362e-01 -1.28480911e+00
-9.62741315e-01 8.69526342e-02 -6.54511333e-01 3.99018437e-01
2.58158475e-01 -1.05995320e-01 -4.55078959e-01 1.57674503e+00
4.38301325e-01 -1.17603138e-01 4.53156620e-01 9.10168886e-01
4.61984038e-01 1.03054535e+00 -3.05475682e-01 -4.86781567e-01
1.13974679e+00 -1.50202310e+00 -7.59306967e-01 -5.16431212e-01
3.84241939e-01 -4.82275069e-01 5.52441895e-01 4.23940867e-01
-1.23639655e+00 6.87548891e-02 -1.06574881e+00 3.17738473e-01
-4.39798623e-01 -6.26601219e-01 8.50321829e-01 1.53617650e-01
-1.12486207e+00 2.08292842e-01 -8.27951372e-01 -5.24626672e-01
-4.07802522e-01 5.72287440e-01 -5.77135563e-01 -3.60540450e-01
-9.07662094e-01 1.01368177e+00 5.96984148e-01 2.33887374e-01
-7.44734108e-01 2.72706896e-02 -8.35602641e-01 2.88120750e-02
1.10455465e+00 -6.37568533e-01 1.24656105e+00 -3.91793728e-01
-1.32124496e+00 1.46532744e-01 -2.45108962e-01 -2.23192498e-01
3.56221318e-01 3.47749501e-01 -2.23202839e-01 -2.51379833e-02
5.70175767e-01 4.36260313e-01 -5.80241717e-02 -1.52530634e+00
-8.73767734e-01 -3.87115717e-01 6.96511865e-01 2.07816228e-01
2.29763210e-01 -4.69199233e-02 -2.52503425e-01 6.42076507e-02
3.19730729e-01 -1.27926290e+00 -1.06416929e+00 -1.23401657e-01
-7.24403083e-01 -4.50817913e-01 6.94128752e-01 -1.04387552e-01
7.82957017e-01 -1.73093021e+00 5.10814309e-01 6.08015478e-01
2.58427918e-01 -4.50342298e-01 -5.21154761e-01 8.84936512e-01
4.06255215e-01 2.36484200e-01 -1.16285756e-01 3.50956805e-02
2.21296847e-01 8.16363990e-01 2.90706128e-01 4.32335168e-01
-4.16233838e-02 7.45060503e-01 -1.15433228e+00 -1.33459941e-01
-8.44099522e-02 -3.28049809e-01 -1.65115193e-01 -7.85287991e-02
-5.60235143e-01 4.33284849e-01 -7.07157552e-01 5.24115920e-01
8.20650697e-01 -1.54335707e-01 4.40001577e-01 6.32877111e-01
-6.65575683e-01 -1.57602623e-01 -1.58099544e+00 1.50644004e+00
-2.87286401e-01 1.86424285e-01 3.25782329e-01 -5.78105986e-01
1.02611864e+00 -3.09083164e-02 4.93333757e-01 -4.73343462e-01
-1.78902164e-01 4.77929026e-01 1.73930809e-01 -2.48278782e-01
6.81200266e-01 3.58923167e-01 -4.65438217e-01 7.70773888e-01
-6.67472124e-01 3.64308618e-02 5.78567743e-01 1.65432751e-01
1.30364132e+00 -4.42136884e-01 7.42006361e-01 -3.71971190e-01
3.05367231e-01 6.27226532e-01 9.10875916e-01 7.89535344e-01
-2.47883439e-01 4.82068174e-02 2.97142655e-01 -6.70593619e-01
-7.82716393e-01 -1.12229836e+00 5.50390244e-01 4.62714612e-01
1.31728160e+00 -5.63568592e-01 -3.65241200e-01 -4.83772844e-01
1.29889727e-01 7.63383806e-01 -2.81733304e-01 1.40027329e-01
-7.56145060e-01 -3.47995192e-01 -2.76539803e-01 1.48505941e-01
3.83599788e-01 -8.28275800e-01 -1.08770108e+00 6.38334692e-01
-2.76824653e-01 -1.14395618e+00 -3.78469974e-01 2.49514535e-01
-3.83041948e-01 -1.38210499e+00 -1.38569415e-01 -9.40468490e-01
1.01267469e+00 1.13651872e+00 7.59787560e-01 1.79087192e-01
-1.00668669e-01 4.65910494e-01 -6.39129102e-01 -3.56000453e-01
-3.08966428e-01 4.70711812e-02 1.58892915e-01 -6.07454002e-01
-1.84836939e-01 -2.03219876e-01 -3.47608477e-01 9.99775231e-01
-5.10310590e-01 1.17369868e-01 6.17597759e-01 3.66365135e-01
9.10504162e-01 1.03670406e+00 5.77464402e-01 -1.42409414e-01
9.16324914e-01 -7.29912341e-01 -8.47438037e-01 7.05446959e-01
-6.82056725e-01 -7.61443451e-02 6.75906301e-01 -2.35531360e-01
-5.43457210e-01 -2.48819515e-02 8.36938560e-01 1.59160286e-01
-1.27486914e-01 7.90633619e-01 -4.34213728e-01 -2.98904836e-01
1.15022212e-01 4.69314754e-02 9.20212939e-02 -1.25665320e-02
3.27257097e-01 9.32102501e-02 4.52961177e-01 -4.60317045e-01
8.20975065e-01 1.29915535e-01 6.42862082e-01 -3.62069160e-01
9.37253162e-02 -7.15388954e-02 -4.22867179e-01 -2.81832755e-01
6.59770906e-01 -1.14720464e-01 -1.15867126e+00 6.69151247e-02
-1.58925796e+00 -2.48601615e-01 1.09875038e-01 2.87027597e-01
-8.75791967e-01 -1.66223105e-02 5.00884489e-04 -1.02163351e+00
2.43509412e-01 -1.30491877e+00 4.26788837e-01 6.20473623e-01
-2.94954747e-01 -7.32137799e-01 1.21060852e-02 -4.62276191e-02
4.09028888e-01 6.48533404e-01 1.06793618e+00 -5.89919150e-01
-1.10500956e+00 1.76474839e-01 -4.82713133e-02 -9.70523715e-01
4.06746835e-01 -9.25504789e-02 3.55742663e-01 -3.56666505e-01
-2.43099675e-01 1.99356541e-01 -1.97975770e-01 5.18660069e-01
4.86693680e-01 -8.75057817e-01 -8.94926488e-01 2.17620060e-02
1.44929242e+00 1.14977837e+00 2.32484654e-01 1.11787450e+00
-4.50363532e-02 9.82632041e-01 1.11279917e+00 5.20839930e-01
9.76111710e-01 8.75593185e-01 8.88976812e-01 5.09096920e-01
5.79516053e-01 7.47809559e-03 -8.64984244e-02 2.90204793e-01
1.01293422e-01 -8.43642473e-01 -1.20866203e+00 4.37266022e-01
-2.47101712e+00 -5.47869325e-01 -4.47761685e-01 1.71809506e+00
3.86408046e-02 -9.80442613e-02 2.29435742e-01 -9.79504958e-02
9.43987787e-01 -2.64361441e-01 -8.60233545e-01 -7.29892790e-01
-2.32416779e-01 -6.54874384e-01 6.78838193e-01 7.13363647e-01
-7.72756159e-01 5.61039507e-01 5.92905664e+00 -1.03773028e-02
-3.75284672e-01 -9.43234935e-02 2.85872638e-01 1.64070398e-01
-3.52783859e-01 1.98676750e-01 -4.10112202e-01 1.92803979e-01
6.12522006e-01 -7.71600604e-01 1.17188656e+00 7.23600805e-01
4.22296792e-01 -3.73000205e-01 -8.90023351e-01 7.86639988e-01
-3.96196157e-01 -1.21863508e+00 -3.06510568e-01 1.85680524e-01
7.43258536e-01 -3.60454708e-01 -1.76753774e-01 -1.51487648e-01
6.90916419e-01 -9.68945026e-01 9.40732956e-01 1.60684004e-01
2.54679983e-03 -1.06314063e+00 6.71176791e-01 5.48719585e-01
-1.37114596e+00 -4.49934810e-01 -2.85875499e-01 -2.08573565e-01
8.63326788e-01 3.01674400e-02 -8.18283677e-01 1.13665128e+00
6.67805851e-01 1.34900704e-01 2.52415061e-01 1.38039839e+00
-3.16159546e-01 -6.81564569e-01 -4.51559663e-01 -4.77069318e-01
7.11285353e-01 -6.00390553e-01 6.97763979e-01 5.66632450e-01
5.58389366e-01 5.87623179e-01 7.05117106e-01 8.87628973e-01
5.39623857e-01 -1.59507856e-01 -6.49643660e-01 1.86520636e-01
8.09372842e-01 1.05144250e+00 -1.00065994e+00 3.41094047e-01
-1.98477089e-01 6.94009423e-01 1.08387798e-01 3.93508166e-01
-8.56174111e-01 -4.20220375e-01 8.86961043e-01 -3.61945480e-02
-1.96604013e-01 -8.02778780e-01 -5.15368283e-01 -2.49227211e-01
1.03982382e-01 -5.39897263e-01 2.62482166e-01 -9.34451461e-01
-8.70351791e-01 6.54354990e-01 1.58178777e-01 -9.37988997e-01
-3.10996652e-01 -3.25282991e-01 -8.35837662e-01 6.91810071e-01
-1.48199069e+00 -8.72843206e-01 -3.80421877e-01 4.85052735e-01
5.42891741e-01 -1.74422383e-01 8.75836611e-01 -4.10872549e-01
-5.96530795e-01 -5.21038622e-02 1.40914358e-02 -6.75577164e-01
-8.57810155e-02 -1.07586873e+00 4.02928829e-01 7.70226359e-01
-4.64076519e-01 7.03755021e-01 9.44192052e-01 -8.17120790e-01
-2.10051608e+00 -9.88426805e-01 6.93071663e-01 1.26033783e-01
7.54395008e-01 -1.82005297e-02 -3.94472301e-01 8.78994107e-01
3.12223971e-01 -5.27718306e-01 2.54938990e-01 -1.45286068e-01
2.97072500e-01 1.42026603e-01 -1.43027163e+00 9.05566335e-01
1.29732800e+00 6.17992342e-01 -1.74892604e-01 4.40937072e-01
9.09871280e-01 -4.82649177e-01 -4.57536727e-01 1.41494527e-01
2.46954679e-01 -6.13437533e-01 7.37860620e-01 -6.69302940e-01
-1.69898812e-02 -7.25778103e-01 -3.18226457e-01 -1.82107520e+00
-5.61388671e-01 -1.05771649e+00 1.29870594e-01 8.79992545e-01
6.63422227e-01 -1.17341185e+00 6.46084189e-01 8.50921392e-01
-3.04043829e-01 -7.62431800e-01 -1.15635693e+00 -1.39205956e+00
-3.20362300e-01 1.94094002e-01 1.01331806e+00 5.97834289e-01
4.41460401e-01 1.38244942e-01 -2.22357228e-01 8.77152503e-01
6.34354770e-01 4.67142791e-01 8.55029523e-01 -1.14271712e+00
5.25654294e-02 -3.75001311e-01 -1.48431987e-01 -9.96702433e-01
2.47266009e-01 -4.65271503e-01 3.33934963e-01 -2.26837587e+00
-1.91665627e-02 -9.79899466e-01 2.25188375e-01 5.22696793e-01
3.77502203e-01 -8.27520430e-01 6.55693352e-01 3.70192051e-01
-8.05842817e-01 3.41516525e-01 1.35268664e+00 -2.94057637e-01
-3.88201267e-01 2.56366581e-01 -7.53059566e-01 4.32022035e-01
1.17725718e+00 -4.84953910e-01 -6.84613585e-01 -7.33472168e-01
3.85263026e-01 8.45785141e-01 1.92746595e-02 -3.01891983e-01
6.28544867e-01 -1.12572157e+00 -5.04291475e-01 -5.21395385e-01
3.74182612e-01 -1.18414450e+00 6.20375097e-01 9.11825716e-01
1.07419737e-01 1.04405570e+00 2.85925884e-02 6.48318529e-01
-7.97420591e-02 -5.41815400e-01 2.86276668e-01 -3.28678578e-01
-9.39059496e-01 2.22497955e-01 -5.98323405e-01 -6.15606010e-01
1.90311384e+00 -2.85120159e-01 -1.02506721e+00 -5.87249875e-01
-4.81589258e-01 1.19398165e+00 4.51068759e-01 3.72250646e-01
8.79990101e-01 -1.25062442e+00 -5.46525657e-01 -1.79481417e-01
3.12815346e-02 4.17145133e-01 3.45144905e-02 6.56313419e-01
-5.74522853e-01 8.51614654e-01 -4.81023222e-01 -4.59748954e-02
-9.90780473e-01 1.02508354e+00 1.41698956e-01 -2.52306134e-01
-5.02519429e-01 3.65460515e-01 1.65440813e-01 -5.10602832e-01
5.43568619e-02 -1.75087780e-01 -4.95691560e-02 -5.32537699e-01
5.56265354e-01 7.59333551e-01 -5.30733347e-01 -2.06975907e-01
-8.81321847e-01 5.46090484e-01 1.12825118e-01 -1.78068206e-01
1.48678422e+00 -5.69738746e-01 -4.45370704e-01 -5.97443916e-02
4.81087536e-01 -3.28621656e-01 -8.17026973e-01 5.16569987e-03
2.81923860e-01 -5.79740226e-01 -6.14910960e-01 -9.50273216e-01
-8.38065028e-01 -6.11265516e-03 -2.35604078e-01 7.00394094e-01
9.16405439e-01 3.67456019e-01 6.15662813e-01 6.71516359e-01
1.37617481e+00 -7.28521585e-01 1.05647795e-01 5.97226799e-01
9.32909548e-01 -8.20798278e-01 -2.49546826e-01 -8.03106189e-01
-5.43211162e-01 1.35274482e+00 8.31409574e-01 1.09906428e-01
1.76565751e-01 3.55873078e-01 -1.34064212e-01 -1.67431787e-01
-7.25855529e-01 1.38518577e-02 -6.33630693e-01 1.07220018e+00
-6.77340150e-01 3.98118228e-01 -4.60808158e-01 6.20857358e-01
-2.98536956e-01 -5.36712408e-01 1.11893117e+00 1.22460079e+00
-7.61394739e-01 -9.52397823e-01 -7.26162851e-01 -1.37744248e-01
5.27545452e-01 6.05233371e-01 -5.76845467e-01 9.81631160e-01
5.52096292e-02 1.66788816e+00 5.50081991e-02 -2.28569284e-01
6.84299529e-01 -7.76593447e-01 3.33865047e-01 -3.29837501e-01
-4.47690785e-02 -3.28321993e-01 1.74459890e-01 -5.72660506e-01
-2.68894464e-01 -7.58801877e-01 -1.84638333e+00 -5.82768381e-01
-1.78836748e-01 4.43210542e-01 8.18720520e-01 7.86796331e-01
1.95132628e-01 3.92256707e-01 8.82988036e-01 -4.94355500e-01
-1.79923117e-01 -1.24381959e-01 -6.33959830e-01 -4.10212815e-01
5.35071939e-02 -8.19243371e-01 -5.01866313e-03 -6.34733617e-01] | [4.952594757080078, 1.7424193620681763] |
cb336239-d5d6-4506-a5a9-c59a3874e625 | distributed-heuristic-multi-agent-path | 2106.11365 | null | https://arxiv.org/abs/2106.11365v1 | https://arxiv.org/pdf/2106.11365v1.pdf | Distributed Heuristic Multi-Agent Path Finding with Communication | Multi-Agent Path Finding (MAPF) is essential to large-scale robotic systems. Recent methods have applied reinforcement learning (RL) to learn decentralized polices in partially observable environments. A fundamental challenge of obtaining collision-free policy is that agents need to learn cooperation to handle congested situations. This paper combines communication with deep Q-learning to provide a novel learning based method for MAPF, where agents achieve cooperation via graph convolution. To guide RL algorithm on long-horizon goal-oriented tasks, we embed the potential choices of shortest paths from single source as heuristic guidance instead of using a specific path as in most existing works. Our method treats each agent independently and trains the model from a single agent's perspective. The final trained policy is applied to each agent for decentralized execution. The whole system is distributed during training and is trained under a curriculum learning strategy. Empirical evaluation in obstacle-rich environment indicates the high success rate with low average step of our method. | ['Hang Ma', 'Yudong Luo', 'Ziyuan Ma'] | 2021-06-21 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [-3.07331234e-01 4.86440569e-01 -2.94064283e-01 6.63102269e-02
-6.35238469e-01 -4.67206031e-01 5.44947922e-01 1.23090170e-01
-8.88369203e-01 1.28341627e+00 3.26034799e-02 -5.38431883e-01
-5.37725270e-01 -9.28613305e-01 -7.69021153e-01 -8.80826354e-01
-7.59181738e-01 9.60843682e-01 4.31475699e-01 -7.20663071e-01
2.70785302e-01 4.66481358e-01 -8.55462313e-01 -3.54046077e-01
9.30010498e-01 3.57559294e-01 8.04848850e-01 1.01288986e+00
1.59531564e-01 1.06577170e+00 -5.57758570e-01 2.81682968e-01
6.20382249e-01 -1.23757564e-01 -1.00356627e+00 1.29480466e-01
-4.76556480e-01 -6.41291976e-01 -7.56645620e-01 6.59791470e-01
5.72378516e-01 4.92263645e-01 6.38155520e-01 -1.54485977e+00
-8.02549813e-03 6.45118475e-01 -5.63364625e-01 1.70101047e-01
1.68756887e-01 6.32712662e-01 8.08611214e-01 4.49137241e-02
6.98255360e-01 1.47248471e+00 7.33225793e-03 6.40693009e-01
-8.55357111e-01 -3.69460583e-01 4.39686984e-01 2.72501588e-01
-8.80808532e-01 1.11522697e-01 6.41143262e-01 -1.42740728e-02
1.19312024e+00 -5.48580587e-01 6.50117278e-01 7.08749831e-01
6.85764968e-01 7.21944332e-01 1.06982267e+00 -2.07504630e-01
4.29417074e-01 -2.02764034e-01 -6.47440434e-01 8.91699672e-01
4.01943922e-02 5.48203945e-01 -1.48690030e-01 -9.33845118e-02
7.81881869e-01 -1.50190651e-01 1.96865574e-01 -5.63517094e-01
-1.24058342e+00 1.13863981e+00 7.41781533e-01 -2.92493939e-01
-7.22967684e-01 6.40767515e-01 4.13679421e-01 8.44450474e-01
-1.66584536e-01 6.68322921e-01 -6.50090396e-01 -4.21808064e-02
1.79141220e-02 7.15129614e-01 1.02206826e+00 1.03087199e+00
1.04865956e+00 2.52695948e-01 -3.39245498e-02 5.32118678e-01
1.97308198e-01 5.59712231e-01 -1.91676140e-01 -1.55622327e+00
6.37525201e-01 2.16760457e-01 5.39489031e-01 -8.61595094e-01
-6.97788119e-01 -2.34044746e-01 -5.82693338e-01 9.33158219e-01
3.36558342e-01 -7.79736459e-01 -6.53923333e-01 1.54387712e+00
6.31921768e-01 2.03196332e-01 6.13038838e-01 9.72028315e-01
1.12132160e-02 8.64620745e-01 1.00291669e-01 -1.43456087e-01
8.66129577e-01 -1.45039845e+00 -3.73938888e-01 -2.95922339e-01
8.46279740e-01 -2.90293306e-01 5.62744975e-01 3.60807329e-01
-9.18877363e-01 -1.48031354e-01 -7.97257662e-01 4.64207023e-01
-1.88717559e-01 -3.26769859e-01 7.59631634e-01 -5.65339364e-02
-1.31091940e+00 6.97602570e-01 -7.70937860e-01 -4.48550493e-01
4.94513005e-01 7.70676196e-01 -1.67109519e-01 -3.30347121e-02
-1.12643492e+00 1.13477147e+00 6.13929451e-01 -3.19343477e-01
-1.88280904e+00 5.05166464e-02 -7.29100764e-01 -1.27471343e-01
1.13350296e+00 -7.40008175e-01 1.55515325e+00 -5.99483848e-01
-2.11570597e+00 -8.18740875e-02 3.47745210e-01 -7.94314206e-01
4.05291975e-01 -1.02905296e-01 2.72381723e-01 3.48080873e-01
2.94578046e-01 1.07049859e+00 7.08569646e-01 -1.31095815e+00
-1.17544019e+00 2.23032922e-01 7.28079319e-01 9.53890145e-01
3.38475220e-02 -4.86049235e-01 7.89488256e-02 9.06687453e-02
-4.25331861e-01 -1.08542609e+00 -1.10616314e+00 -3.25941950e-01
-1.18181609e-01 -7.30601668e-01 8.62795949e-01 -1.13537110e-01
5.10712922e-01 -1.58815181e+00 4.48588431e-01 2.20187843e-01
2.38362789e-01 2.36154646e-02 -5.02757788e-01 1.08315945e+00
7.52149522e-01 -2.43905991e-01 1.15980834e-01 -8.51408020e-02
1.14364184e-01 8.24946940e-01 -2.35233411e-01 5.24641752e-01
2.05290377e-01 7.29247570e-01 -1.39984131e+00 -4.08678114e-01
2.74993151e-01 6.23893738e-02 -7.15823829e-01 4.67061609e-01
-5.38731813e-01 7.91003406e-01 -1.11616862e+00 1.85449570e-01
3.51244599e-01 8.73123184e-02 2.08361208e-01 7.68627286e-01
-4.41213518e-01 1.20362230e-01 -1.16490841e+00 1.84486866e+00
-5.51690221e-01 2.75986314e-01 4.11306947e-01 -1.27790940e+00
9.42548096e-01 1.60233021e-01 6.39919221e-01 -8.42207193e-01
1.00917950e-01 1.32332876e-01 1.82948232e-01 -5.20601273e-01
5.20547628e-01 1.70184359e-01 -3.73357207e-01 5.12703478e-01
-5.91391921e-02 -4.19790804e-01 1.58422023e-01 4.08846885e-01
1.50478518e+00 3.24455053e-01 2.91011155e-01 -2.44034916e-01
3.68707418e-01 6.10790372e-01 5.77741563e-01 1.22975779e+00
-5.94918489e-01 -4.55253392e-01 6.49027705e-01 -6.83828712e-01
-1.04980588e+00 -8.29590261e-01 8.37727070e-01 1.24863875e+00
5.67059517e-01 -1.10784367e-01 -3.48588079e-01 -7.97360778e-01
1.05464444e-01 4.62586194e-01 -4.14416701e-01 6.39867559e-02
-9.60491300e-01 -2.07149893e-01 -7.67204957e-03 -1.19171971e-02
5.70034623e-01 -1.53033066e+00 -1.16500759e+00 6.99763238e-01
1.84231624e-01 -1.00571215e+00 -4.29832727e-01 3.60912502e-01
-4.67757851e-01 -1.32983744e+00 -4.22084481e-01 -1.03011298e+00
6.14453018e-01 5.92444956e-01 6.82244897e-01 7.93003291e-02
7.58516639e-02 6.11096799e-01 -4.23849642e-01 -4.41605777e-01
-4.97435421e-01 3.06782633e-01 2.58234739e-01 -5.70870101e-01
-2.43183330e-01 -5.13635993e-01 -7.95166731e-01 2.57921487e-01
-3.40483367e-01 7.51296580e-02 8.86910021e-01 8.72843444e-01
3.16123992e-01 4.25532579e-01 8.41243505e-01 -3.90068263e-01
1.00135660e+00 -6.95134401e-01 -9.81638730e-01 -1.20097294e-01
-5.63821316e-01 2.49145612e-01 1.02495360e+00 -4.40615892e-01
-7.77679145e-01 1.44292414e-01 3.92194688e-01 -1.50666699e-01
-2.59790361e-01 2.96033680e-01 3.06009173e-01 -2.31790245e-01
5.28206646e-01 1.95336401e-01 3.58655840e-01 2.00872600e-01
3.58124524e-01 3.55341911e-01 1.93264320e-01 -9.39997733e-01
7.64081657e-01 3.54623884e-01 3.12063396e-01 -5.15406072e-01
-3.31933081e-01 -3.32628071e-01 -3.57644409e-01 -3.07048202e-01
6.40286326e-01 -9.38751042e-01 -1.50636113e+00 1.73319817e-01
-1.13611567e+00 -1.11043787e+00 -1.20701581e-01 3.69363785e-01
-1.10254955e+00 1.28561422e-01 -3.90589386e-01 -9.83426273e-01
-1.50415808e-01 -1.20027471e+00 7.34267354e-01 5.00059485e-01
5.52991390e-01 -1.06219745e+00 4.22450602e-01 -1.09622352e-01
5.39987743e-01 2.19597176e-01 3.43735754e-01 -3.43376130e-01
-8.72381210e-01 3.68769884e-01 3.09902262e-02 -3.19393009e-01
-1.08309992e-01 -5.02043486e-01 -1.67350218e-01 -7.78291523e-01
-5.04243970e-01 -8.16601694e-01 5.18067479e-01 5.24830222e-01
6.77075088e-01 -5.90343356e-01 -6.15330398e-01 2.24612400e-01
1.52867889e+00 5.10531902e-01 3.15552652e-01 8.22176099e-01
2.92158306e-01 6.67023778e-01 1.04587603e+00 6.54148698e-01
9.85898137e-01 5.23226380e-01 1.00897074e+00 1.37262940e-01
2.70541370e-01 -2.60203630e-01 5.77811301e-01 3.02871555e-01
2.90820431e-02 -3.43141288e-01 -9.62476611e-01 5.61890125e-01
-2.42523122e+00 -9.24566388e-01 3.81402165e-01 1.84850538e+00
4.38337028e-01 2.11665049e-01 4.27861720e-01 -3.50116372e-01
4.07218337e-01 -3.06599047e-02 -7.19445646e-01 -5.77680469e-01
1.62830651e-01 -2.35332876e-01 9.28816259e-01 9.38174546e-01
-1.08235359e+00 1.48581064e+00 5.65155077e+00 6.76375806e-01
-8.15758169e-01 2.32616644e-02 2.83945352e-01 1.49482444e-01
2.14377299e-01 2.35953018e-01 -6.63783610e-01 9.37133804e-02
9.43417430e-01 -2.05798417e-01 9.79648054e-01 9.04622555e-01
7.87758231e-01 -5.25819182e-01 -5.89300871e-01 4.75796700e-01
-7.17029929e-01 -1.37851894e+00 -3.80896479e-01 3.05717945e-01
8.55321646e-01 4.87263441e-01 -2.94365764e-01 7.05780447e-01
1.33710217e+00 -9.64961410e-01 5.28202057e-01 2.51140296e-01
3.32406700e-01 -1.19974387e+00 5.90531945e-01 8.95704091e-01
-1.15266633e+00 -7.30626762e-01 -6.09764814e-01 -6.57488704e-01
2.57634908e-01 -2.85314798e-01 -1.48939955e+00 6.50193036e-01
4.02023345e-01 6.63076639e-01 2.24830315e-01 9.97823477e-01
-3.79517823e-01 3.11762989e-01 -3.34101558e-01 -5.98881543e-01
1.08637810e+00 -4.41437632e-01 7.97293782e-01 7.73814678e-01
-2.11024899e-02 1.29083142e-01 1.30703819e+00 3.57914776e-01
4.32541221e-01 -2.28512615e-01 -1.00972855e+00 1.75831154e-01
5.61563194e-01 1.28130245e+00 -7.32896388e-01 -1.03395097e-01
-2.48307899e-01 7.67834485e-01 9.33026910e-01 5.35043895e-01
-6.04324877e-01 -2.99464226e-01 6.35111630e-01 -2.77093560e-01
3.27315658e-01 -6.69176996e-01 3.00022483e-01 -3.99853289e-01
-4.54766065e-01 -5.63677907e-01 3.07756186e-01 -4.21163708e-01
-9.87042487e-01 4.59462166e-01 -2.53870431e-02 -1.08863568e+00
-4.75156456e-01 -3.29691380e-01 -8.79348755e-01 5.13150156e-01
-2.24463129e+00 -1.03608930e+00 -1.30623981e-01 8.58238637e-01
6.59743726e-01 -7.07306981e-01 8.16483200e-01 -2.85216302e-01
-2.92721629e-01 -1.87282711e-02 6.70620352e-02 -1.14974119e-02
4.06542838e-01 -1.43460131e+00 1.84303673e-04 4.76960480e-01
-5.99985003e-01 -8.97620469e-02 7.96488822e-01 -6.77698374e-01
-1.81842756e+00 -1.13185573e+00 5.89603521e-02 -1.59056354e-02
7.86602020e-01 -1.06696959e-03 -3.77807289e-01 4.47792321e-01
5.60064852e-01 -1.12343967e-01 -2.37281751e-02 -2.11031631e-01
2.62364298e-01 -1.46074638e-01 -1.02580118e+00 7.57526159e-01
8.16582441e-01 1.78969607e-01 -2.32229456e-01 5.20875990e-01
9.89883482e-01 -4.97698903e-01 -3.99390191e-01 -6.82044104e-02
-2.63982769e-02 -4.84773725e-01 5.62025249e-01 -8.97310674e-01
1.61229700e-01 -4.72965717e-01 7.94459879e-02 -1.96643817e+00
-2.92283207e-01 -1.38039267e+00 -4.50862497e-02 2.99323142e-01
1.67294115e-01 -8.04999769e-01 9.10284817e-01 -9.41415131e-02
-4.69270080e-01 -7.43786931e-01 -1.12661302e+00 -6.89168453e-01
3.92308712e-01 1.03837058e-01 3.88479471e-01 4.84512925e-01
2.61985242e-01 3.56312007e-01 -6.13669693e-01 6.20476305e-01
8.04736495e-01 -1.26261562e-01 1.27497554e+00 -7.05669403e-01
-3.92162383e-01 -3.11078429e-01 5.43467775e-02 -1.22584999e+00
4.40513998e-01 -7.74003744e-01 3.83290380e-01 -1.89320755e+00
-1.67485207e-01 -9.03694510e-01 -5.50312642e-03 4.96653050e-01
2.01909930e-01 -7.40922034e-01 2.97694087e-01 1.32264823e-01
-1.11199152e+00 7.83960283e-01 1.95054150e+00 -1.49217263e-01
-5.35598695e-01 2.02494189e-01 -5.14999747e-01 4.02390152e-01
1.34133029e+00 -3.87999475e-01 -7.58756638e-01 -5.99913359e-01
1.78362727e-01 8.18773031e-01 2.82371998e-01 -8.83642137e-01
7.44526505e-01 -9.54247653e-01 -3.73590469e-01 -3.59806508e-01
2.67929554e-01 -9.60778356e-01 -4.31371480e-01 1.11802399e+00
-2.68735439e-01 2.28661373e-01 -1.21901534e-01 1.03080916e+00
1.58555195e-01 2.53709424e-02 5.76735616e-01 -4.30973709e-01
-1.12063766e+00 6.37108862e-01 -6.64982080e-01 -2.76945299e-03
1.52370632e+00 1.24317504e-01 -3.76772523e-01 -7.85298526e-01
-5.33434033e-01 1.33124113e+00 3.11380345e-02 1.08202688e-01
7.62204111e-01 -8.30168426e-01 -9.00799870e-01 -1.80948034e-01
-3.21450055e-01 3.66036475e-01 1.89837683e-02 6.29872262e-01
-4.89325643e-01 4.06373650e-01 -5.48621953e-01 -3.14441234e-01
-6.56546831e-01 4.83000308e-01 6.56353772e-01 -5.07449925e-01
-8.17599356e-01 4.57573146e-01 -7.09645376e-02 -7.12369621e-01
3.60362083e-01 1.67735606e-01 -3.85814965e-01 -5.06786942e-01
3.05032074e-01 3.41290444e-01 -6.97720408e-01 -8.14027339e-02
-7.32301101e-02 2.03875035e-01 -1.62604392e-01 -3.41760814e-01
1.54565227e+00 -2.97832191e-01 1.85228676e-01 -1.36631027e-01
6.23261154e-01 -5.78623235e-01 -2.02778101e+00 -2.03916147e-01
-9.96070504e-02 -2.58872777e-01 -1.09482761e-02 -6.69001639e-01
-7.88887560e-01 5.15106618e-01 1.37452900e-01 2.07237512e-01
5.33477068e-01 -7.72911087e-02 6.01198554e-01 9.85090077e-01
1.04731572e+00 -1.37766254e+00 6.17347598e-01 1.04485393e+00
7.39124298e-01 -1.48104775e+00 -2.36012235e-01 7.32557708e-03
-9.94518936e-01 1.17724216e+00 9.89318371e-01 -6.92495465e-01
3.57028127e-01 3.12942147e-01 3.84685472e-02 -2.23485619e-01
-1.13223183e+00 -4.39686775e-01 -5.78715384e-01 1.10520875e+00
-5.45504630e-01 2.46912926e-01 -2.08753303e-01 -2.72753060e-01
-8.12776834e-02 -2.20122308e-01 9.66287851e-01 1.21330786e+00
-1.08988643e+00 -1.02629876e+00 -1.48648933e-01 3.04876149e-01
8.10084194e-02 5.66339135e-01 1.43831164e-01 8.93444657e-01
-1.18651152e-01 1.12010741e+00 2.95476913e-02 -1.18780226e-01
9.25882161e-02 -6.03850842e-01 3.42467934e-01 -6.38182938e-01
-3.67663234e-01 1.09577045e-01 1.33492157e-01 -8.47159982e-01
-4.14768904e-01 -5.78564703e-01 -1.79477477e+00 -4.42520797e-01
1.91227451e-01 4.19774801e-01 5.17534196e-01 1.01838827e+00
3.21340084e-01 6.73774421e-01 1.10481834e+00 -1.00793839e+00
-9.17607367e-01 -6.44739747e-01 -2.64674425e-01 -8.69872048e-02
6.31275952e-01 -8.80422652e-01 -8.11143219e-02 -6.65732920e-01] | [3.8705053329467773, 1.9294856786727905] |
9cdcdb29-d64c-4a6e-a046-7ed186f322fc | resource-allocation-algorithm-for-mec-based | null | null | https://ieeexplore.ieee.org/document/9679368 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9679368 | Resource allocation algorithm for MEC based on Deep Reinforcement Learning | In recent years, driven by the commercialization of the 6th Generation Communication Technology (6G), an increasing number of 6G devices connected to mobile networks produces computation-intensive tasks such as ultra-high-resolution video streaming, inter-active visual reality (VR) gaming, augmented reality (AR). However, the computing capacity and the capacity of battery of the 6G devices are limited. The technology of computation offloading would offload the tasks from the IoT devices to the edge network in the scenario of mobile edge computing (MEC). Not only can solve the shortage of mobile user device in energy effciency, but also deal with the tasks in low latency. IoT devices can offload computing tasks or execute them locally to finish the work. In order to find the optimal allocation rate of local computing tasks and offloading tasks, a resource allocation policy gradient (RAPG) based DDPG is considered. Finally we analyze the performance of RAPG by contrasts with different resource allocation algorithms. Numerial simulation results showed that the RAPG can achieve the best allocate rate between the BS and local, also can reduce the overall system delay of task combination with minimum energy consumption. | ['Shougang Du', 'Ying Chen', 'Xin Chen', 'Yijie Wang'] | 2022-01-20 | null | null | null | ieee-2022-1 | ['edge-computing'] | ['time-series'] | [-2.03156948e-01 -1.36808038e-01 -3.21518958e-01 4.49535072e-01
1.20606206e-01 -3.71111393e-01 -1.35739639e-01 -7.10379958e-01
-8.19945410e-02 7.64496505e-01 8.61444175e-02 -4.42502350e-01
-3.08044463e-01 -9.04885113e-01 -1.15838714e-01 -7.73118258e-01
-2.74621248e-01 2.30820373e-01 3.61410499e-01 6.58394992e-02
-2.28399746e-02 3.55380535e-01 -1.52257180e+00 -6.20185100e-02
1.23050606e+00 1.45199883e+00 8.96919906e-01 6.22850239e-01
7.68748287e-04 3.77894402e-01 -3.63516122e-01 -6.16478696e-02
2.20923871e-01 -2.84629971e-01 -4.13594991e-01 3.15453708e-02
-7.94212580e-01 -4.95154798e-01 -3.75090212e-01 8.12095284e-01
8.99000227e-01 1.46292329e-01 2.69268453e-01 -1.72202504e+00
-2.53953874e-01 1.57643750e-01 -7.81894386e-01 6.47176325e-01
1.17562927e-01 -3.40815008e-01 1.28233209e-01 -8.01474571e-01
6.84210777e-01 5.53090632e-01 2.44215295e-01 5.26053965e-01
-1.24630742e-01 -5.59785962e-01 5.41788749e-02 5.56271970e-01
-1.30727828e+00 -6.58291459e-01 4.95569259e-01 -1.35637179e-03
9.54315841e-01 6.84383750e-01 1.05832338e+00 2.79031068e-01
5.31895936e-01 1.95724130e-01 6.13753021e-01 -3.94910395e-01
2.44329438e-01 -1.99221790e-01 -3.98159891e-01 3.92037779e-01
5.69622338e-01 -2.12894738e-01 -4.73304838e-01 -1.22496538e-01
7.20828772e-01 2.46745795e-01 -4.34132040e-01 1.13635585e-01
-1.12606847e+00 -3.00315451e-02 1.43106759e-01 4.15248811e-01
-8.87860656e-01 3.39853793e-01 3.16366762e-01 5.51654436e-02
6.72909915e-01 -3.38300377e-01 -4.62996781e-01 -4.96646315e-01
-6.77603304e-01 -5.71626306e-01 6.86693013e-01 1.23705721e+00
2.69752294e-01 1.99537396e-01 -2.72880584e-01 3.23299438e-01
6.27894700e-01 6.72091246e-01 4.12659287e-01 -1.08755779e+00
7.17781782e-01 5.14825821e-01 3.51892948e-01 -7.46130407e-01
-4.50547546e-01 -5.61128855e-01 -1.10663390e+00 -2.26795092e-01
-9.71441567e-02 -7.70639181e-01 -5.42462587e-01 1.12031627e+00
5.05991280e-01 7.46788561e-01 -1.13221779e-01 1.09862399e+00
6.58175945e-01 1.07309914e+00 1.64218947e-01 -1.05189812e+00
1.37917936e+00 -7.29923904e-01 -1.10285854e+00 -2.57352926e-02
5.72816074e-01 -8.23545516e-01 5.45371830e-01 1.38815075e-01
-1.41792154e+00 -3.34613442e-01 -1.02573252e+00 3.25094402e-01
-1.40166506e-01 2.48013586e-01 6.81683421e-01 1.11497903e+00
-1.25641382e+00 -8.64735991e-03 -6.84188306e-01 -3.49041253e-01
3.45861912e-01 5.97405255e-01 2.14535475e-01 -1.76585406e-01
-9.68117714e-01 4.62597132e-01 3.12146753e-01 2.32289582e-01
-5.10831118e-01 -5.31400681e-01 -5.86292483e-02 3.39854926e-01
5.18526912e-01 -1.00640202e+00 7.01139331e-01 -8.13718259e-01
-1.49197400e+00 3.76401156e-01 3.91903631e-02 1.51396975e-01
2.84463763e-01 2.73959398e-01 -7.82116413e-01 -1.66619178e-02
-1.61087289e-01 1.36755750e-01 5.02556801e-01 -9.21437860e-01
-8.93552542e-01 -6.07317984e-01 -5.75842038e-02 7.75602579e-01
-3.35868686e-01 5.88170737e-02 -6.28586531e-01 -1.28000677e-01
2.16430366e-01 -1.14549029e+00 -1.87027529e-01 -4.76843268e-01
4.33663838e-02 -1.74727529e-01 1.37348461e+00 -5.05171955e-01
1.53435719e+00 -2.40902996e+00 1.31152004e-01 1.48929143e-02
3.70843470e-01 2.73847952e-02 4.98022676e-01 2.80914575e-01
4.18740213e-01 2.13659137e-01 4.94128883e-01 7.42973313e-02
-3.59009415e-01 2.15331107e-01 1.63978357e-02 2.66985416e-01
-1.03016686e+00 8.37385416e-01 -8.92097116e-01 -5.11555672e-01
9.74584594e-02 3.09693724e-01 -3.04636538e-01 7.48113990e-02
2.78246254e-01 3.92255872e-01 -7.97524869e-01 7.45556176e-01
9.45974648e-01 -5.88261545e-01 4.13572162e-01 -1.97456032e-01
-3.11474781e-02 -2.56783307e-01 -1.16129243e+00 1.51874185e+00
-8.32242310e-01 2.77798951e-01 5.23121178e-01 -6.78555131e-01
5.30329883e-01 7.25485444e-01 1.04741538e+00 -7.89375246e-01
2.23859742e-01 5.19575596e-01 -6.69034570e-02 -9.18665349e-01
4.08231646e-01 3.38545799e-01 1.26895100e-01 5.04847527e-01
-5.95394611e-01 4.69960034e-01 -4.48176526e-02 3.30637604e-01
9.01582360e-01 5.30634299e-02 8.45891684e-02 -3.76225322e-01
3.64611804e-01 -3.43654990e-01 6.68929458e-01 1.21682651e-01
-3.15000683e-01 -1.94437355e-01 -1.15197219e-01 -2.34594747e-01
-6.14166677e-01 -7.92965174e-01 3.43587667e-01 1.22728956e+00
9.20834959e-01 -1.18081279e-01 -5.54791451e-01 -2.61621445e-01
-3.40134382e-01 3.92573774e-01 2.02967972e-01 -2.41485983e-01
-3.86830896e-01 -1.09359145e+00 -2.29883164e-01 2.15206638e-01
9.05152917e-01 -8.76207530e-01 -1.16601992e+00 2.54136652e-01
-3.24072301e-01 -1.29962277e+00 -5.62518954e-01 -1.62008315e-01
-1.08175135e+00 -6.17611051e-01 -6.29776180e-01 -7.20578909e-01
7.24149048e-01 8.20826471e-01 1.05165505e+00 3.44528615e-01
1.05837755e-01 4.70740825e-01 -4.14650142e-01 -3.50044280e-01
3.75603974e-01 4.61464114e-02 1.74572513e-01 1.47691414e-01
-3.97228897e-02 -8.06719422e-01 -1.01270211e+00 3.87557536e-01
-4.24020588e-01 4.18341190e-01 3.37046951e-01 1.65938720e-01
6.85037434e-01 5.26483953e-01 8.92932713e-01 -6.11854255e-01
6.39926732e-01 -9.30692136e-01 -5.18784165e-01 2.75793940e-01
-7.13455498e-01 -7.17583835e-01 6.80454373e-01 -1.86444446e-01
-1.16382051e+00 -2.31709063e-01 3.67340416e-01 -3.52792144e-01
6.14247739e-01 1.17435649e-01 -6.97941244e-01 1.68156698e-02
4.58097905e-02 1.12264328e-01 -6.53588355e-01 -9.94294584e-02
1.21949181e-01 1.06677473e+00 1.35322824e-01 -2.32425660e-01
2.97913432e-01 4.27466601e-01 3.24201405e-01 -4.90137130e-01
7.85463601e-02 -1.57046303e-01 2.24560156e-01 -9.39450443e-01
9.44721878e-01 -1.03341055e+00 -1.16145957e+00 8.82628709e-02
-1.15564036e+00 -2.74497241e-01 1.16883591e-01 6.49830163e-01
-4.42708582e-01 1.49030760e-01 -4.03745949e-01 -1.05615091e+00
-4.91737068e-01 -1.17128909e+00 9.61050093e-01 5.87905467e-01
4.60804462e-01 -5.01500130e-01 -5.64219713e-01 3.13636601e-01
6.64814770e-01 -4.66720685e-02 5.43240368e-01 4.41785663e-01
-1.21615028e+00 9.45633799e-02 -5.24055541e-01 -4.53776598e-01
5.85876517e-02 -4.11811173e-01 -6.09179258e-01 -4.86714959e-01
4.45861928e-02 6.10418200e-01 -1.10267466e-02 8.51637006e-01
1.54517734e+00 -1.62465021e-01 -9.66437459e-01 8.63325179e-01
1.58163822e+00 8.79008532e-01 9.83310521e-01 -9.81337950e-02
6.77594960e-01 2.20493525e-01 8.43059719e-01 6.60853803e-01
3.93186152e-01 7.42408812e-01 7.69873321e-01 7.09090456e-02
-1.70760855e-01 9.15436521e-02 1.36646286e-01 1.23567653e+00
-7.58327365e-01 -9.71022844e-01 -5.95511973e-01 2.48447463e-01
-2.16869688e+00 -6.60205662e-01 -3.77966017e-01 2.23479772e+00
-1.22444421e-01 7.55644366e-02 1.07051097e-02 2.39828542e-01
8.49787891e-01 -1.61501631e-01 -5.54358721e-01 -2.36917377e-01
3.61750543e-01 -4.49790448e-01 1.04131556e+00 8.25931579e-02
-2.68128335e-01 5.18914461e-01 5.65659332e+00 1.00016093e+00
-1.03087580e+00 9.14983094e-01 1.00306058e+00 -3.37330401e-01
-3.14874053e-01 -2.05046013e-01 -3.28992784e-01 1.35503829e+00
1.15121973e+00 -5.28299987e-01 9.25817370e-01 8.41092288e-01
8.19579184e-01 -4.32293296e-01 -4.27587211e-01 1.72022855e+00
-4.10197556e-01 -1.20444787e+00 -5.61252773e-01 4.83780921e-01
7.50473738e-01 3.52164032e-03 -3.04431289e-01 -3.24815735e-02
-2.11395904e-01 -5.08799255e-01 3.19977373e-01 6.22085869e-01
1.17449152e+00 -1.00501442e+00 7.77190149e-01 5.58611870e-01
-1.67494881e+00 -2.60767907e-01 -4.22487915e-01 -2.04651251e-01
6.89929426e-01 1.02349055e+00 -1.49182558e-01 8.13776553e-01
8.98809969e-01 1.32414207e-01 9.80251580e-02 9.41479802e-01
3.75630915e-01 1.35072500e-01 -2.86643028e-01 -1.14273295e-01
-4.06125426e-01 -6.59565032e-01 4.65393335e-01 5.28382838e-01
9.84642088e-01 8.28584731e-01 3.08301151e-01 1.98920473e-01
-2.03389212e-01 1.33905381e-01 -5.42473853e-01 3.45882177e-02
8.98418605e-01 1.40986061e+00 -1.31490767e+00 -4.48314458e-01
-2.98394889e-01 1.18944955e+00 -3.60655695e-01 4.91493255e-01
-1.02416444e+00 -3.88927966e-01 3.85769606e-01 2.72215098e-01
-2.19968647e-01 -5.29545486e-01 -4.70771402e-01 -8.82948101e-01
2.17211679e-01 -1.47178724e-01 1.68521836e-01 -1.12098908e+00
-6.78860009e-01 4.81003731e-01 -2.24198863e-01 -1.26645470e+00
1.59708485e-01 -1.19602695e-01 -6.44648015e-01 8.71723771e-01
-1.14561355e+00 -5.69746017e-01 -8.70056331e-01 6.19572461e-01
6.41911566e-01 -1.27161950e-01 1.66547343e-01 1.03624630e+00
-5.20815372e-01 3.54194641e-02 1.54127628e-01 -7.98885584e-01
1.36049181e-01 -6.02170765e-01 -2.04294980e-01 9.10903513e-01
-6.31022751e-01 -4.17434908e-02 3.70508373e-01 -8.15930724e-01
-2.11288261e+00 -6.62772000e-01 5.40760517e-01 7.72014633e-02
1.17473111e-01 -3.09173137e-01 -1.64819017e-01 2.86203623e-01
1.32122666e-01 4.53239650e-01 4.00285214e-01 -4.76740181e-01
8.61060381e-01 -3.88044536e-01 -1.29902554e+00 6.59531236e-01
1.77016413e+00 -6.57702237e-02 6.29844546e-01 6.30074263e-01
8.19627821e-01 -5.45023024e-01 -6.47437334e-01 3.23342830e-01
4.00612384e-01 -6.95924819e-01 8.13263595e-01 -1.09356508e-01
-2.75894552e-01 -5.25581956e-01 -2.32594818e-01 -1.04969728e+00
-4.68647391e-01 -7.65140057e-01 -6.10936642e-01 9.06953156e-01
4.20884378e-02 -7.38473892e-01 1.05151033e+00 6.91925347e-01
-4.41955447e-01 -7.45380104e-01 -1.10276401e+00 -6.44033790e-01
-1.13816881e+00 -1.96744591e-01 7.50797808e-01 7.56316900e-01
1.76989734e-01 3.67579401e-01 -4.98023003e-01 2.12128356e-01
3.41145307e-01 -9.90268886e-02 3.78455102e-01 -1.32794797e+00
-2.26630300e-01 1.06002102e-02 -2.45738938e-01 -1.22464585e+00
-4.05300349e-01 -7.69292712e-01 -4.95094270e-01 -2.09485435e+00
9.03972462e-02 -8.20924997e-01 -1.68891296e-01 -8.10556933e-02
2.23571574e-03 2.18006268e-01 2.84601063e-01 4.03771609e-01
-9.74450290e-01 2.02641338e-01 1.61024785e+00 3.86328340e-01
-4.35583919e-01 2.30253369e-01 -4.17729348e-01 3.28337789e-01
7.24246740e-01 -3.33309859e-01 -9.66495156e-01 -4.80413079e-01
8.38639796e-01 7.97626078e-01 -2.69015342e-01 -8.44721794e-01
4.46337670e-01 -4.17533517e-01 3.48900437e-01 -5.38861036e-01
2.72872061e-01 -1.63702548e+00 8.99669528e-01 7.65339553e-01
9.01911914e-01 3.26467693e-01 -2.05435172e-01 6.51573539e-01
4.28938717e-01 1.14885774e-02 1.84337601e-01 -1.22889904e-02
-6.94464386e-01 6.16964936e-01 -6.45249128e-01 -3.15084994e-01
1.55063152e+00 -6.12454474e-01 -5.45366704e-01 -3.38648528e-01
-7.15536952e-01 3.55880022e-01 1.86812848e-01 1.98913887e-01
3.80069822e-01 -1.33995974e+00 -2.86485076e-01 -1.14593871e-01
-5.42482018e-01 -1.93328544e-01 8.50913823e-01 9.90254998e-01
-8.35092783e-01 3.28795940e-01 -2.53820568e-01 -4.00391042e-01
-1.20566857e+00 7.07726657e-01 3.69819552e-01 -2.54028976e-01
-2.06700727e-01 4.41540122e-01 7.66678751e-02 3.53091329e-01
-1.30983815e-01 3.93203050e-02 -2.21492395e-01 -1.35185555e-01
2.68184900e-01 1.34432685e+00 1.49108082e-01 -3.99363965e-01
-5.25036693e-01 4.56935048e-01 6.75478220e-01 3.63606177e-02
8.85623872e-01 -9.05984163e-01 -2.11411923e-01 -1.21129967e-01
7.49771178e-01 1.56493947e-01 -7.98295915e-01 4.55561817e-01
-3.54322374e-01 -7.39302099e-01 7.06499100e-01 -6.19657815e-01
-1.50723755e+00 1.98586285e-01 1.06619954e+00 6.93970740e-01
1.85752642e+00 -1.87689006e-01 1.10941994e+00 -2.65303075e-01
1.07087672e+00 -1.22214806e+00 -8.32060203e-02 -1.08803280e-01
2.27219269e-01 -6.85780942e-01 3.09940167e-02 -1.13196492e+00
-2.24589154e-01 8.49826038e-01 7.31512845e-01 2.13231698e-01
9.26141500e-01 4.66694683e-01 -5.05776227e-01 -3.02997738e-01
-6.07932270e-01 -1.18103586e-01 -2.71955281e-01 6.55686498e-01
2.41423890e-01 4.09457594e-01 -9.48094666e-01 6.40873969e-01
2.09623426e-01 4.79141831e-01 5.93531370e-01 8.33690286e-01
-5.12875021e-01 -9.23830450e-01 -3.62605721e-01 7.39003062e-01
-6.85316265e-01 1.85179189e-01 5.41345835e-01 1.96488038e-01
6.37692630e-01 1.21014047e+00 3.18899512e-01 -2.20523790e-01
8.09240807e-03 -4.22895640e-01 3.09468925e-01 -3.26459497e-01
-1.45457476e-01 1.97957620e-01 2.05117777e-01 -4.28493619e-01
-2.29354426e-01 -1.10760935e-01 -1.37557459e+00 -5.96756756e-01
-3.93457949e-01 2.56909579e-01 9.60183859e-01 8.28108132e-01
9.82254326e-01 1.04057276e+00 7.32749760e-01 -8.01990211e-01
4.53247190e-01 -5.70754051e-01 -8.75025451e-01 -2.48699874e-01
-3.55934560e-01 -5.10145903e-01 -2.04220980e-01 -1.12978898e-01] | [5.969771385192871, 1.5838299989700317] |
a66b443e-d61d-4cd3-94d3-738cefb78363 | dressing-in-order-recurrent-person-image | 2104.07021 | null | https://arxiv.org/abs/2104.07021v3 | https://arxiv.org/pdf/2104.07021v3.pdf | Dressing in Order: Recurrent Person Image Generation for Pose Transfer, Virtual Try-on and Outfit Editing | We propose a flexible person generation framework called Dressing in Order (DiOr), which supports 2D pose transfer, virtual try-on, and several fashion editing tasks. The key to DiOr is a novel recurrent generation pipeline to sequentially put garments on a person, so that trying on the same garments in different orders will result in different looks. Our system can produce dressing effects not achievable by existing work, including different interactions of garments (e.g., wearing a top tucked into the bottom or over it), as well as layering of multiple garments of the same type (e.g., jacket over shirt over t-shirt). DiOr explicitly encodes the shape and texture of each garment, enabling these elements to be edited separately. Joint training on pose transfer and inpainting helps with detail preservation and coherence of generated garments. Extensive evaluations show that DiOr outperforms other recent methods like ADGAN in terms of output quality, and handles a wide range of editing functions for which there is no direct supervision. | ['Svetlana Lazebnik', 'Daniel McKee', 'Aiyu Cui'] | 2021-04-14 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Cui_Dressing_in_Order_Recurrent_Person_Image_Generation_for_Pose_Transfer_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Cui_Dressing_in_Order_Recurrent_Person_Image_Generation_for_Pose_Transfer_ICCV_2021_paper.pdf | iccv-2021-1 | ['pose-transfer'] | ['computer-vision'] | [ 4.59862798e-01 2.24249810e-01 4.18716908e-01 -1.11717835e-01
-3.38938594e-01 -6.60294473e-01 3.80818963e-01 -4.00786281e-01
2.80872405e-01 7.16666818e-01 6.15551889e-01 3.55691254e-01
1.97994515e-01 -9.35134470e-01 -1.06017196e+00 -2.81890273e-01
2.27103621e-01 6.12476289e-01 -1.69291452e-01 -6.15182877e-01
-3.38221490e-01 3.15467954e-01 -1.33979142e+00 5.10667026e-01
5.39052427e-01 7.24729776e-01 1.62065938e-01 9.01868761e-01
3.26315641e-01 4.05561388e-01 -8.88606906e-01 -8.64189386e-01
3.93503696e-01 -6.99586511e-01 -5.35554945e-01 5.11555731e-01
7.10247576e-01 -8.18355441e-01 -2.28111535e-01 4.06760544e-01
7.89947212e-01 2.02536196e-01 5.15000820e-01 -9.29774582e-01
-1.04804838e+00 7.27028131e-01 -5.31047583e-01 -7.68727183e-01
9.00967360e-01 2.88002402e-01 6.94250405e-01 -5.57248950e-01
9.62587059e-01 1.45866001e+00 8.48286510e-01 8.83240640e-01
-1.36418724e+00 -4.21689659e-01 2.25507513e-01 -5.37958324e-01
-9.86551106e-01 -3.33946556e-01 7.95445859e-01 -1.93316624e-01
5.51188767e-01 7.67176747e-01 1.21730089e+00 1.62069476e+00
2.41266340e-01 8.71232331e-01 8.33715498e-01 -3.23182285e-01
-9.22076404e-02 -1.93383306e-01 -4.78612363e-01 7.29650915e-01
-1.01159196e-02 2.98398621e-02 -5.92045903e-01 6.25709072e-02
1.31114948e+00 2.37762332e-01 -2.10765868e-01 -1.95780382e-01
-1.41768157e+00 4.51904923e-01 4.42531466e-01 -2.40336880e-01
-5.16157210e-01 5.81446946e-01 3.77143621e-01 4.08535659e-01
5.29471993e-01 5.19322395e-01 -2.60566801e-01 2.90109264e-03
-6.86173737e-01 8.20833147e-01 7.17454970e-01 1.30377984e+00
1.39982924e-01 -2.58116692e-01 -7.04146087e-01 8.13530743e-01
-6.69538751e-02 2.91611612e-01 -2.27219671e-01 -8.99228275e-01
5.69459677e-01 3.13909948e-01 3.41223061e-01 -7.65469313e-01
-2.17609689e-01 -7.18569309e-02 -9.68971014e-01 3.21220249e-01
1.85622439e-01 -5.05145669e-01 -1.25768065e+00 1.62828815e+00
4.81160939e-01 1.71961963e-01 -2.73546815e-01 1.12559104e+00
1.09265435e+00 3.88940334e-01 -1.97900191e-01 3.13127220e-01
1.66701043e+00 -1.39772511e+00 -7.48118818e-01 -1.99028030e-01
4.99536693e-02 -1.01192987e+00 1.17854905e+00 4.56976682e-01
-1.52838182e+00 -6.42203510e-01 -7.55593240e-01 -6.15803361e-01
8.03589076e-02 1.30612373e-01 8.41228664e-01 1.53612837e-01
-9.30539012e-01 9.61266041e-01 -8.26487541e-01 -4.01932210e-01
2.42471859e-01 1.03037335e-01 -5.40213227e-01 -2.42704898e-01
-1.05318713e+00 7.31426597e-01 -1.35061175e-01 3.03081483e-01
-6.67623103e-01 -8.84453595e-01 -1.08539963e+00 -3.04869205e-01
2.37709776e-01 -1.42844462e+00 1.14572775e+00 -7.68329918e-01
-1.96299827e+00 9.08341706e-01 1.82397902e-01 -2.19823375e-01
1.31613493e+00 -8.44640613e-01 -1.21592231e-01 -6.34838343e-02
3.92336585e-02 8.59431148e-01 9.97215390e-01 -1.20922208e+00
1.07203074e-01 -3.37066025e-01 2.51785755e-01 5.92170715e-01
2.32363179e-01 -1.59396455e-02 -7.68365324e-01 -1.12808537e+00
-6.08500391e-02 -1.14801407e+00 -1.70512013e-02 3.63275945e-01
-9.29039240e-01 3.38242739e-01 6.20080650e-01 -1.36364925e+00
9.24561441e-01 -2.01761675e+00 8.29410195e-01 -6.45083711e-02
-1.15848682e-03 -2.04341412e-01 -6.21386617e-02 6.22746766e-01
9.95466635e-02 -6.98391199e-02 -1.45347267e-01 -8.98274839e-01
2.31149614e-01 1.51457235e-01 -1.50771752e-01 4.85022180e-02
1.59557834e-01 1.20701838e+00 -7.61448264e-01 -2.63645887e-01
2.48882577e-01 8.61823499e-01 -7.09199667e-01 2.68557936e-01
-3.86519730e-01 6.96115553e-01 -5.51418308e-03 6.97813988e-01
6.42074168e-01 -6.13573054e-03 5.43301217e-02 -4.15348381e-01
3.07080179e-01 2.39463076e-01 -1.25205362e+00 2.23584509e+00
-5.61466575e-01 1.83341205e-01 1.54629841e-01 -3.15384343e-02
7.01414883e-01 4.28290337e-01 1.41261846e-01 -2.96360880e-01
-4.60495334e-03 -2.18808725e-01 -4.93330777e-01 -4.05561805e-01
7.96086371e-01 -8.15146044e-02 -3.74280512e-01 4.28944647e-01
-2.36528724e-01 -4.76702631e-01 2.07485189e-03 1.52558675e-02
8.94256592e-01 8.99562299e-01 3.16426181e-03 3.11689943e-01
-4.97498184e-01 -4.19647157e-01 2.74680585e-01 5.33271074e-01
3.79035681e-01 1.32820237e+00 2.60375261e-01 -3.65400344e-01
-1.26806450e+00 -1.13468122e+00 3.56627733e-01 1.20261157e+00
3.05514216e-01 -4.54402536e-01 -8.53022814e-01 -4.51950699e-01
1.58603638e-01 6.47901356e-01 -9.22869682e-01 -6.55426234e-02
-6.85261667e-01 -2.90397614e-01 3.97718459e-01 6.27007306e-01
4.27960455e-01 -1.19976878e+00 -7.89116561e-01 3.35047275e-01
-2.69258082e-01 -7.94865906e-01 -1.15653324e+00 -2.01246381e-01
-7.02100694e-01 -6.72187150e-01 -1.26190078e+00 -7.41232514e-01
6.91714108e-01 1.23736253e-02 1.26776028e+00 -6.14549033e-03
-5.24497271e-01 2.64215142e-01 -3.92460555e-01 -2.92666286e-01
-3.82339835e-01 -1.54643711e-02 -1.60684496e-01 -1.45914182e-01
-5.52024186e-01 -8.41814220e-01 -8.89082313e-01 3.41124445e-01
-7.00770497e-01 8.54706228e-01 4.15194213e-01 7.73784101e-01
4.96764004e-01 -1.70274645e-01 8.21841359e-02 -9.05932784e-01
7.25414455e-01 -1.38491437e-01 1.40046403e-01 2.50981838e-01
1.55264392e-01 -1.58153996e-01 3.15555722e-01 -7.05544710e-01
-1.04508770e+00 -9.32297483e-02 -1.26656905e-01 -5.84843993e-01
-1.34881865e-02 -1.47734225e-01 -1.75663516e-01 4.70725536e-01
4.97529894e-01 -7.86790401e-02 1.50780648e-01 -7.42402792e-01
6.69266880e-01 1.16023019e-01 8.22238564e-01 -5.78893006e-01
6.94001853e-01 3.60846013e-01 -2.37903431e-01 -3.04813653e-01
-5.70327282e-01 2.54552126e-01 -4.72736746e-01 -1.56560674e-01
1.03827953e+00 -8.74964297e-01 -7.11499810e-01 8.21811795e-01
-1.35407138e+00 -7.31474161e-01 -6.21873856e-01 4.95128185e-02
-4.19188827e-01 -6.02657348e-02 -1.11174512e+00 -4.56827879e-01
-6.63275182e-01 -7.87600935e-01 1.66465330e+00 2.45591655e-01
-8.32161665e-01 -7.00874627e-01 -1.54442653e-01 5.72127938e-01
2.15279117e-01 9.52778816e-01 4.88411337e-01 2.46903390e-01
-3.77501667e-01 -1.54998183e-01 2.41185650e-01 1.37141958e-01
4.15587038e-01 -4.56492938e-02 -5.30275047e-01 -3.03375095e-01
-6.56672001e-01 -4.18019146e-01 7.52091229e-01 3.23391408e-01
8.87711704e-01 -5.01833200e-01 -2.52397835e-01 6.94549680e-01
8.15244317e-01 -1.37884304e-01 1.00394511e+00 -6.10841215e-02
1.08267713e+00 5.54909170e-01 4.51389760e-01 4.75915670e-01
4.72972184e-01 1.00004280e+00 3.28988969e-01 -5.11834741e-01
-6.17091179e-01 -7.48292863e-01 4.66010928e-01 4.19093877e-01
-5.20957589e-01 -3.13837588e-01 -6.38547763e-02 3.29575360e-01
-1.71263874e+00 -1.01143110e+00 2.04532713e-01 2.11779356e+00
9.96340156e-01 -4.66258712e-02 2.03695178e-01 -2.43374437e-01
6.68794572e-01 1.59584552e-01 -7.07286000e-01 -4.15825367e-01
8.14765785e-03 4.38034296e-01 1.55827075e-01 4.20947790e-01
-8.92174959e-01 8.61780643e-01 6.28262043e+00 2.26746485e-01
-8.92297745e-01 -1.61352694e-01 5.29323816e-01 -5.12781501e-01
-5.30564666e-01 -3.00616413e-01 -5.08203447e-01 5.75564325e-01
-1.03696719e-01 3.51142108e-01 7.65319169e-01 4.50416297e-01
8.40540379e-02 8.22279155e-02 -1.33259153e+00 1.03454983e+00
2.84933805e-01 -1.19385326e+00 9.55401063e-02 -1.01640806e-01
8.56780648e-01 -7.19707787e-01 1.45520389e-01 1.74715668e-01
5.15364766e-01 -1.10049176e+00 1.25249350e+00 8.01837862e-01
1.02409816e+00 -5.57734966e-01 2.90534317e-01 2.51623783e-02
-1.06811345e+00 4.12252337e-01 1.86179832e-01 -7.24917874e-02
6.45031691e-01 4.22562242e-01 -6.10007107e-01 7.92597771e-01
5.87203324e-01 5.26667118e-01 -1.31054550e-01 5.73262036e-01
-4.78852302e-01 1.53337028e-02 -4.21182901e-01 3.57469618e-01
-4.13978696e-01 -2.73937076e-01 5.93447506e-01 8.92316401e-01
4.68425602e-01 7.74183450e-03 1.64535910e-01 1.02406085e+00
-3.23725373e-01 -2.66778201e-01 -3.80147487e-01 6.74539134e-02
2.52855450e-01 1.15128136e+00 -3.79754066e-01 -3.34659040e-01
1.40991971e-01 2.08971095e+00 5.19141294e-02 2.40004897e-01
-1.04608905e+00 -2.93764055e-01 8.08276415e-01 4.94214058e-01
3.32328379e-01 -2.68653985e-02 -2.38172412e-01 -1.16447639e+00
2.96021074e-01 -8.81834090e-01 1.39488295e-01 -1.19449520e+00
-1.24684370e+00 4.93921191e-01 -8.94133896e-02 -1.08682680e+00
-1.70505807e-01 6.00413000e-03 -7.34819114e-01 9.70382750e-01
-6.91052794e-01 -1.77363288e+00 -5.58355391e-01 3.75555545e-01
6.57312810e-01 3.90976220e-01 9.09363627e-01 2.78012931e-01
-3.66922855e-01 7.06303298e-01 -5.71764290e-01 -8.49083513e-02
8.92157197e-01 -1.13311613e+00 1.02827954e+00 4.91835862e-01
-9.01243016e-02 5.47787786e-01 8.31675351e-01 -9.74839926e-01
-1.45955038e+00 -1.25193441e+00 5.76081395e-01 -4.65953022e-01
-1.76476762e-01 -6.98470950e-01 -4.95999753e-01 9.31166768e-01
2.22523525e-01 -2.88762629e-01 3.98008466e-01 4.97090742e-02
-1.71755582e-01 1.16217904e-01 -1.23635495e+00 1.01928008e+00
1.59918773e+00 -1.69980735e-01 -4.18314308e-01 4.65850264e-01
9.43256319e-01 -1.18377471e+00 -9.20883596e-01 1.42793745e-01
9.49646890e-01 -7.77853549e-01 1.16608918e+00 -5.94993532e-01
1.02099621e+00 -3.07535529e-01 1.26347572e-01 -1.77047622e+00
-4.42238927e-01 -1.16398144e+00 -2.62891144e-01 1.17212582e+00
1.12836100e-01 -3.10687870e-01 6.46845102e-01 7.56181061e-01
-1.38954625e-01 -7.58583546e-01 -1.96393430e-01 -4.02002573e-01
-4.88292426e-01 -8.96675363e-02 1.03693414e+00 7.97681034e-01
-3.42377752e-01 3.22619408e-01 -1.19334543e+00 1.38834685e-01
5.73384583e-01 3.17718714e-01 1.12167084e+00 -8.71427178e-01
-8.31674039e-01 6.92705363e-02 -1.12347230e-01 -1.10469389e+00
-3.98296386e-01 -7.65576899e-01 -4.97168005e-02 -2.06376529e+00
7.67066479e-02 -3.83786201e-01 3.69342059e-01 7.67221451e-01
-3.56987268e-01 5.51190376e-01 6.24370337e-01 -1.33631214e-01
-1.78671598e-01 5.05750000e-01 1.94974422e+00 -1.26973286e-01
-6.00410938e-01 1.81600116e-02 -9.27899241e-01 6.82143033e-01
4.84304905e-01 1.30709419e-02 -3.40805888e-01 -7.86908507e-01
1.91393286e-01 3.63813579e-01 7.97196269e-01 -8.62848878e-01
-2.47202545e-01 2.20508166e-02 8.25694919e-01 -3.27376157e-01
6.36749268e-01 -4.72558409e-01 1.09602785e+00 2.19848216e-01
-2.79998153e-01 1.09025553e-01 1.18579410e-01 3.63535225e-01
3.25311333e-01 4.59189564e-01 5.43466985e-01 -4.51148719e-01
-2.08491176e-01 2.49471575e-01 2.06508651e-01 -2.05544516e-01
9.57300305e-01 -3.84741098e-01 -6.04877248e-02 -5.89223862e-01
-1.16024125e+00 1.91566706e-01 7.66121268e-01 7.69890487e-01
5.35570621e-01 -1.59915960e+00 -8.42185259e-01 2.68633336e-01
-1.69598818e-01 5.40129364e-01 5.62408566e-01 4.69327599e-01
-6.88725352e-01 -4.67495173e-01 -3.68612587e-01 -2.85222262e-01
-1.49758184e+00 1.71625122e-01 2.60511398e-01 -4.63865429e-01
-1.20404828e+00 1.04783154e+00 2.55313635e-01 -5.74046910e-01
7.75121897e-02 -4.18327123e-01 2.96632469e-01 -4.67363745e-02
5.51027358e-01 3.43246639e-01 -8.15713853e-02 -2.06612751e-01
-2.33410913e-02 6.05933011e-01 -2.15968043e-01 -2.32520118e-01
1.46085465e+00 4.57835495e-02 -6.03590347e-02 2.85320520e-01
7.32483149e-01 1.95394382e-01 -1.71315634e+00 6.35073781e-02
-9.95410442e-01 -5.39441407e-01 -4.29814637e-01 -1.20992792e+00
-9.77486014e-01 3.53399009e-01 1.54056013e-01 -2.24405065e-01
1.02740109e+00 -5.79124764e-02 1.34047985e+00 -2.10028827e-01
6.51720643e-01 -1.02171910e+00 2.36064643e-01 1.61280319e-01
1.72609985e+00 -5.90225637e-01 1.34430781e-01 -6.28443778e-01
-1.06253731e+00 6.40997231e-01 4.89538550e-01 -2.96776295e-01
2.02513970e-02 4.80568290e-01 -2.16362312e-01 -1.02179259e-01
-6.32037818e-01 2.96600759e-01 3.30924988e-01 5.42938232e-01
3.57530445e-01 4.83688891e-01 -1.23145245e-01 5.40835440e-01
-7.27324307e-01 1.92322820e-01 1.36307701e-01 1.00034535e+00
2.05613390e-01 -1.12379432e+00 -5.11934459e-01 3.63907009e-01
-2.26548031e-01 4.16650018e-03 -4.64669287e-01 6.31191134e-01
3.80713195e-01 4.67605382e-01 5.66696674e-02 -3.00225914e-01
7.94590533e-01 -2.28722990e-02 1.05350769e+00 -6.61140084e-01
-9.52089667e-01 1.65443614e-01 3.20642442e-01 -5.85236609e-01
-1.27782822e-01 -4.60221678e-01 -9.39091325e-01 -4.82571214e-01
1.09284468e-01 -4.82552707e-01 3.48385245e-01 5.43089688e-01
5.55973589e-01 9.74697828e-01 2.77750522e-01 -1.44417584e+00
-1.99115574e-01 -9.41279292e-01 -6.90459967e-01 9.13687348e-01
3.31037611e-01 -4.61990327e-01 1.41484618e-01 4.72122043e-01] | [11.90518856048584, -0.8050904273986816] |
fad30418-068e-4301-ab8e-81918d895f59 | generalized-object-search | 2301.10121 | null | https://arxiv.org/abs/2301.10121v2 | https://arxiv.org/pdf/2301.10121v2.pdf | Generalized Object Search | Future collaborative robots must be capable of finding objects. As such a fundamental skill, we expect object search to eventually become an off-the-shelf capability for any robot, similar to e.g., object detection, SLAM, and motion planning. However, existing approaches either make unrealistic compromises (e.g., reduce the problem from 3D to 2D), resort to ad-hoc, greedy search strategies, or attempt to learn end-to-end policies in simulation that are yet to generalize across real robots and environments. This thesis argues that through using Partially Observable Markov Decision Processes (POMDPs) to model object search while exploiting structures in the human world (e.g., octrees, correlations) and in human-robot interaction (e.g., spatial language), a practical and effective system for generalized object search can be achieved. In support of this argument, I develop methods and systems for (multi-)object search in 3D environments under uncertainty due to limited field of view, occlusion, noisy, unreliable detectors, spatial correlations between objects, and possibly ambiguous spatial language (e.g., "The red car is behind Chase Bank"). Besides evaluation in simulators such as PyGame, AirSim, and AI2-THOR, I design and implement a robot-independent, environment-agnostic system for generalized object search in 3D and deploy it on the Boston Dynamics Spot robot, the Kinova MOVO robot, and the Universal Robots UR5e robotic arm, to perform object search in different environments. The system enables, for example, a Spot robot to find a toy cat hidden underneath a couch in a kitchen area in under one minute. This thesis also broadly surveys the object search literature, proposing taxonomies in object search problem settings, methods and systems. | ['Kaiyu Zheng'] | 2023-01-24 | null | null | null | null | ['motion-planning'] | ['robots'] | [-3.66459221e-01 1.30018353e-01 -1.25969276e-01 -1.30721509e-01
-4.49158847e-01 -8.28375578e-01 4.59021568e-01 1.25923917e-01
-5.30988514e-01 6.85895860e-01 -5.80751240e-01 -4.32397753e-01
-6.41961932e-01 -3.71352375e-01 -5.12576163e-01 -4.24959958e-01
-4.82991248e-01 1.23457909e+00 3.68796706e-01 -9.23218355e-02
2.37946108e-01 9.95097339e-01 -1.51522481e+00 -6.25424445e-01
3.76277745e-01 7.55153477e-01 9.95935321e-01 6.80196643e-01
3.67703408e-01 6.06488466e-01 -4.46342707e-01 3.44875485e-01
5.05891681e-01 5.64516298e-02 -1.00339115e+00 2.27761373e-01
-4.82238621e-01 -4.88896728e-01 -3.19312423e-01 8.26500952e-01
4.40273345e-01 2.85008371e-01 5.95233917e-01 -1.69920754e+00
-2.72530973e-01 3.61724287e-01 -3.03377628e-01 -2.53786109e-02
5.85464537e-01 7.29483783e-01 4.98355210e-01 -5.66215277e-01
6.25600398e-01 1.62804139e+00 3.64981532e-01 2.94821531e-01
-9.97698247e-01 -2.85123855e-01 1.70353785e-01 1.02907084e-01
-1.63612497e+00 -2.44818285e-01 1.42538145e-01 -3.73365849e-01
1.24524486e+00 1.46148339e-01 6.17100179e-01 9.86740589e-01
6.57812178e-01 8.01374197e-01 9.02666986e-01 -2.45045498e-01
5.09960949e-01 -3.52006368e-02 -2.85817474e-01 6.48420632e-01
6.07672155e-01 2.56131858e-01 -1.99664921e-01 -2.79739618e-01
8.05795014e-01 -4.75579016e-02 1.31419688e-01 -9.07726347e-01
-1.30464399e+00 5.48604310e-01 3.56715202e-01 5.76068163e-02
-6.85023546e-01 4.42082226e-01 -7.56333917e-02 2.13577166e-01
-3.85909498e-01 6.13910079e-01 -3.82774413e-01 -2.83267021e-01
-1.60798937e-01 6.55749857e-01 1.05876172e+00 1.57284868e+00
7.19254971e-01 -2.12910175e-01 2.33494997e-01 1.84444904e-01
8.23841393e-01 6.23220921e-01 1.94315940e-01 -1.44973862e+00
4.93015945e-01 4.20283854e-01 9.31005776e-01 -7.55165040e-01
-9.07063067e-01 -1.70433279e-02 -2.55708456e-01 7.19672024e-01
8.05266798e-02 -3.41496795e-01 -9.56749439e-01 1.46459174e+00
5.77261031e-01 -4.18083668e-01 3.29640895e-01 1.26422560e+00
1.62523642e-01 3.84016961e-01 2.04250272e-02 8.36092010e-02
1.29933834e+00 -8.65165353e-01 -4.25302565e-01 -6.81880653e-01
6.31989658e-01 -6.06365800e-01 3.61184001e-01 3.57734442e-01
-1.14189398e+00 -2.37565160e-01 -9.63501215e-01 1.19390383e-01
-2.24467859e-01 -1.68219745e-01 6.62424862e-01 2.23588467e-01
-9.99330580e-01 4.68321651e-01 -1.34105182e+00 -7.74756014e-01
3.25856209e-02 6.55252993e-01 -3.36628526e-01 -2.05473915e-01
-7.71551728e-01 1.54703414e+00 6.59693360e-01 2.40194514e-01
-1.47817290e+00 2.69745648e-01 -6.31583929e-01 4.36516069e-02
6.54592395e-01 -7.20128119e-01 1.59355628e+00 -4.79658276e-01
-1.43534696e+00 4.18243498e-01 2.39169642e-01 -4.88467842e-01
4.96192336e-01 -1.26509711e-01 9.03842151e-02 1.89435765e-01
4.42773163e-01 8.50834727e-01 4.23984796e-01 -1.44169688e+00
-8.19765866e-01 -4.13233101e-01 3.04530650e-01 9.72759366e-01
5.73404968e-01 6.17468357e-02 -3.39437783e-01 1.64324656e-01
4.93086189e-01 -1.41620398e+00 -8.30555201e-01 -2.42246930e-02
-2.00050473e-01 -2.22653687e-01 8.48890960e-01 -1.70693055e-01
3.97028893e-01 -2.02151704e+00 9.22364742e-02 3.38078439e-01
-2.10777506e-01 -4.08535004e-01 -1.56417325e-01 6.97004557e-01
4.40234303e-01 -5.08053899e-02 1.70865402e-01 -2.80854642e-01
2.92451948e-01 6.83270335e-01 1.87711909e-01 7.82617509e-01
-8.87397081e-02 8.00644279e-01 -1.26431823e+00 -5.92588305e-01
1.30378395e-01 -1.06638610e-01 -1.93602383e-01 -2.18449429e-01
-2.02073082e-01 4.18084949e-01 -1.04135883e+00 9.32562649e-01
3.55942309e-01 -1.11043446e-01 3.43693644e-01 7.32626975e-01
-4.30787653e-01 1.02980629e-01 -1.66228187e+00 1.59462440e+00
-3.87230635e-01 3.55863720e-01 7.22835362e-01 -6.64769471e-01
6.91480577e-01 3.19749564e-01 5.79789817e-01 -3.96404058e-01
1.16477884e-01 4.69778270e-01 8.02801773e-02 -4.22677070e-01
8.03790450e-01 1.46926656e-01 -3.91522050e-01 3.40625346e-01
-9.18236896e-02 -7.57369041e-01 -6.07406162e-02 -9.81150940e-02
1.67239714e+00 1.64281622e-01 1.91601217e-01 -3.23678315e-01
-8.53232294e-02 6.68344319e-01 3.18135649e-01 1.27520263e+00
-3.92499834e-01 2.77061850e-01 -6.11617267e-02 -2.26324439e-01
-7.85044849e-01 -1.05959713e+00 1.41505405e-01 7.05406249e-01
1.18466759e+00 1.25049517e-01 -2.98594803e-01 -2.57419139e-01
2.30201453e-01 8.59481454e-01 -3.75529528e-02 -7.88589269e-02
-5.11846423e-01 -1.63783938e-01 2.27050617e-01 2.10727990e-01
3.82895112e-01 -1.20813918e+00 -1.54659092e+00 6.01533115e-01
3.99109395e-03 -1.01877153e+00 4.96701822e-02 8.18956375e-01
-7.35058606e-01 -9.53139186e-01 -3.05187613e-01 -7.78653979e-01
6.28217101e-01 7.16041327e-01 7.57463813e-01 -5.05265556e-02
-2.60830939e-01 1.12367737e+00 -4.57733303e-01 -5.84561229e-01
-2.41839945e-01 -1.14279702e-01 5.76378226e-01 -7.35959947e-01
-2.79225390e-02 -1.27621680e-01 -4.84785646e-01 7.24289715e-01
-4.94432867e-01 -2.84516394e-01 8.74091208e-01 5.71299016e-01
4.25570965e-01 5.72117746e-01 2.72912830e-02 3.36037636e-01
6.58624411e-01 -7.05471039e-01 -8.90786111e-01 1.44267842e-01
-5.15882730e-01 -6.97565824e-02 4.05366719e-02 -6.07148409e-01
-1.04104459e+00 5.59417963e-01 5.49395263e-01 -6.13442957e-01
-2.98434973e-01 6.00647807e-01 -4.72447909e-02 -1.78475767e-01
5.80174565e-01 1.93599295e-02 1.76943824e-01 -3.07566375e-01
1.49835005e-01 5.79753280e-01 4.13555086e-01 -6.74897969e-01
6.93570256e-01 5.72671235e-01 2.59171069e-01 -6.63633287e-01
-5.35678305e-02 -6.29445255e-01 -4.98676807e-01 -2.55963027e-01
7.92250872e-01 -1.07700062e+00 -1.08526242e+00 2.63560534e-01
-1.33606303e+00 -7.12076306e-01 -3.11891466e-01 7.40329981e-01
-1.16852653e+00 1.35105789e-01 -3.76362145e-01 -1.36202574e+00
3.55739504e-01 -1.32819128e+00 1.17654300e+00 4.14111316e-01
-2.49067485e-01 -3.80729407e-01 -1.00857735e-01 1.04142435e-01
1.65381208e-01 -9.53610018e-02 3.96390557e-01 -6.81188345e-01
-1.18554854e+00 -3.33112001e-01 9.06704739e-02 -4.16037083e-01
-1.02941394e-01 -4.70440030e-01 -2.53699064e-01 -4.49692309e-01
2.48865075e-02 -2.85796136e-01 -4.48373146e-04 1.96543097e-01
3.66935223e-01 -2.79292047e-01 -1.03218436e+00 -1.56732500e-01
1.16995406e+00 7.89329112e-01 2.39821568e-01 8.02157879e-01
-3.25182639e-02 7.37384737e-01 1.40959883e+00 6.24175429e-01
6.30580902e-01 9.22375977e-01 1.05233335e+00 4.86975014e-01
4.28444833e-01 -9.25642103e-02 2.98762262e-01 -8.60922113e-02
5.30262180e-02 -5.19617677e-01 -1.15070951e+00 7.44210303e-01
-2.09040475e+00 -7.35957205e-01 5.90275303e-02 1.89755344e+00
1.72364831e-01 1.92333236e-01 -1.83746457e-01 -4.43821341e-01
6.57893240e-01 -4.88102138e-01 -7.40911305e-01 -3.23507965e-01
2.21828252e-01 -4.78394240e-01 1.07595706e+00 5.45787334e-01
-8.54284883e-01 1.08813310e+00 5.67479706e+00 4.41340178e-01
-5.82995176e-01 1.86268881e-01 -1.02741979e-02 -4.63321293e-03
3.12748849e-01 4.27809477e-01 -8.58370304e-01 2.22800553e-01
6.50460839e-01 -1.05362408e-01 5.71436048e-01 1.31321752e+00
1.80265978e-01 -8.83970380e-01 -1.15823889e+00 9.00717139e-01
-4.26327348e-01 -7.54628420e-01 -8.03399920e-01 3.69499832e-01
4.00689691e-01 3.35763007e-01 -7.98999667e-02 4.41493720e-01
1.09207702e+00 -9.10697579e-01 1.37444365e+00 3.44783038e-01
1.26973718e-01 -3.54786456e-01 8.13520193e-01 1.15373027e+00
-1.01794720e+00 -5.46449602e-01 -2.10871503e-01 -2.65627861e-01
6.74157381e-01 -2.53012925e-01 -1.26011515e+00 6.00187898e-01
9.47719157e-01 -2.34743446e-01 2.03192700e-02 1.34257209e+00
-1.17613748e-01 -3.17976773e-01 -9.72737551e-01 -5.55005848e-01
5.88418245e-01 -1.01440310e-01 9.41000342e-01 5.47714949e-01
5.13739586e-01 4.62726653e-01 5.10734022e-01 7.20640540e-01
6.28497660e-01 -6.26807213e-01 -7.61257827e-01 1.75274864e-01
6.96344733e-01 9.11050141e-01 -9.47814047e-01 -8.31842422e-02
1.23060560e-02 9.25038815e-01 -1.18828543e-01 3.51262122e-01
-7.31517911e-01 -1.29605278e-01 7.91259110e-01 -4.79685329e-02
2.66831428e-01 -7.21763968e-01 -1.18184380e-01 -4.90267485e-01
2.34002620e-03 -7.10762024e-01 -5.79192080e-02 -1.11418879e+00
-7.24964678e-01 2.95130610e-01 4.27046180e-01 -1.20066762e+00
-6.90418780e-01 -5.31549990e-01 -5.33549599e-02 7.97752142e-01
-1.10388887e+00 -8.31463516e-01 -1.15816802e-01 3.42249066e-01
6.48092151e-01 4.03178809e-03 5.28082132e-01 -2.04379275e-01
-5.79738878e-02 -3.18659455e-01 8.54674056e-02 -3.95986319e-01
2.16668382e-01 -8.39028180e-01 2.55761713e-01 4.48352993e-01
-1.54056355e-01 5.09316325e-01 1.00615942e+00 -1.04994380e+00
-1.99565136e+00 -6.21075451e-01 4.21238720e-01 -7.09135413e-01
6.89499080e-01 -2.07067847e-01 -3.42286378e-01 9.63975906e-01
-2.43396625e-01 -1.68627724e-01 -2.51979858e-01 -1.86986491e-01
4.28567857e-01 3.80827636e-01 -1.36544907e+00 8.98954451e-01
1.20489359e+00 -2.04544570e-02 -5.45233369e-01 5.23161292e-01
5.07718801e-01 -9.30232286e-01 -4.30551201e-01 3.94596994e-01
6.09952331e-01 -5.83577216e-01 9.23626482e-01 -4.06677037e-01
-4.30881053e-01 -5.48735142e-01 -2.63901055e-01 -1.18131363e+00
-4.07333970e-01 -8.98263037e-01 1.13106266e-01 6.72131002e-01
3.64900649e-01 -7.03596652e-01 7.67316401e-01 9.01526690e-01
-3.77768487e-01 -2.92021126e-01 -1.45793593e+00 -1.34013057e+00
-3.03060144e-01 -4.40211147e-01 4.52577144e-01 4.24631149e-01
4.79387194e-02 1.14908740e-01 -9.49683413e-02 7.34143794e-01
4.95616376e-01 2.58707888e-02 9.27852511e-01 -1.13505828e+00
-2.18483925e-01 -3.10534418e-01 -4.77354974e-01 -1.32535851e+00
1.06170237e-01 -3.59792978e-01 9.63216782e-01 -1.82954490e+00
-1.40936255e-01 -1.19530880e+00 1.97337344e-01 4.79943275e-01
5.50796211e-01 -6.65310442e-01 6.55760705e-01 6.09776676e-01
-1.06094217e+00 3.52767825e-01 1.16840136e+00 -9.96776298e-02
-3.70928884e-01 2.65579462e-01 -1.88563094e-01 8.64811063e-01
8.01800787e-01 -5.92535615e-01 -2.19699621e-01 -5.80754161e-01
2.20093682e-01 7.82628477e-01 4.90861952e-01 -1.20731723e+00
7.74341702e-01 -4.98301834e-01 1.25654027e-01 -6.28261805e-01
9.31592941e-01 -1.52148449e+00 5.52605927e-01 8.16124022e-01
1.10620983e-01 2.88621753e-01 2.64783621e-01 6.77526474e-01
1.53181940e-01 -8.52896154e-01 2.74888337e-01 -5.34509122e-01
-9.04910147e-01 2.60431003e-02 -9.88438547e-01 -5.15054405e-01
1.45633793e+00 -5.11229336e-01 -2.37464234e-01 -4.94471669e-01
-7.09193528e-01 9.03886557e-01 6.71649098e-01 4.98242885e-01
4.72869724e-01 -8.93093169e-01 -1.43069208e-01 -2.47567385e-01
-5.37093468e-02 4.44057941e-01 3.73505205e-02 7.79075503e-01
-5.79464138e-01 6.83467388e-01 -2.57702410e-01 -7.52223790e-01
-1.03969276e+00 6.94756866e-01 2.10580483e-01 -1.69202298e-01
-3.22468370e-01 5.50145566e-01 2.47090403e-02 -5.97603321e-01
2.36634791e-01 -2.66537189e-01 1.49488226e-01 -3.39330882e-01
-7.28240758e-02 4.00614589e-01 -3.03915322e-01 -4.71553981e-01
-5.58294177e-01 3.85434002e-01 4.14576590e-01 -4.46233034e-01
9.75506485e-01 -7.20036447e-01 2.68191904e-01 2.60972261e-01
3.30053300e-01 -4.16915685e-01 -1.37048376e+00 1.69847399e-01
3.30574274e-01 -2.44775534e-01 -2.35012248e-01 -6.81211054e-01
-1.87595546e-01 2.26310015e-01 6.78437054e-01 3.58779043e-01
5.43500245e-01 5.01064718e-01 5.50910793e-02 1.11803401e+00
1.44432127e+00 -1.21418476e+00 9.80691314e-02 7.48319030e-01
9.06449199e-01 -1.19783151e+00 1.16955630e-01 -7.18811899e-02
-7.44060397e-01 8.47032726e-01 6.59337461e-01 8.17730129e-02
3.01727176e-01 4.07634199e-01 -2.46437445e-01 -2.86399424e-01
-8.28391254e-01 -4.61752266e-01 -5.35648048e-01 7.47261465e-01
-7.90963888e-01 1.67538330e-01 1.62110731e-01 1.29822508e-01
3.18775475e-02 -1.38673648e-01 6.20913327e-01 1.57877040e+00
-8.85910153e-01 -6.29271209e-01 -8.69459808e-01 9.69985966e-04
-1.18450880e-01 3.96775216e-01 -1.82478353e-01 1.25200903e+00
1.57146621e-02 1.21399665e+00 1.39631450e-01 -3.22240815e-02
3.93896759e-01 -3.03458840e-01 5.07337868e-01 -7.60423005e-01
-2.79577076e-01 -7.72919804e-02 2.09577620e-01 -6.29271150e-01
-2.24674255e-01 -9.93951499e-01 -1.51511490e+00 2.09816933e-01
-5.19106328e-01 4.41528158e-03 1.08183360e+00 9.30817604e-01
4.57465649e-01 -3.75722423e-02 3.34651113e-01 -1.28337777e+00
-8.97792995e-01 -7.34860182e-01 -5.71869075e-01 -3.47057670e-01
2.37218544e-01 -1.13289022e+00 -3.37933749e-01 -5.09010673e-01] | [4.88897180557251, 1.0883888006210327] |
eb4deca5-6d6c-4d66-b33d-e54040ebde97 | diffusion-probabilistic-models-for-3d-point | 2103.01458 | null | https://arxiv.org/abs/2103.01458v2 | https://arxiv.org/pdf/2103.01458v2.pdf | Diffusion Probabilistic Models for 3D Point Cloud Generation | We present a probabilistic model for point cloud generation, which is fundamental for various 3D vision tasks such as shape completion, upsampling, synthesis and data augmentation. Inspired by the diffusion process in non-equilibrium thermodynamics, we view points in point clouds as particles in a thermodynamic system in contact with a heat bath, which diffuse from the original distribution to a noise distribution. Point cloud generation thus amounts to learning the reverse diffusion process that transforms the noise distribution to the distribution of a desired shape. Specifically, we propose to model the reverse diffusion process for point clouds as a Markov chain conditioned on certain shape latent. We derive the variational bound in closed form for training and provide implementations of the model. Experimental results demonstrate that our model achieves competitive performance in point cloud generation and auto-encoding. The code is available at \url{https://github.com/luost26/diffusion-point-cloud}. | ['Wei Hu', 'Shitong Luo'] | 2021-03-02 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Luo_Diffusion_Probabilistic_Models_for_3D_Point_Cloud_Generation_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Luo_Diffusion_Probabilistic_Models_for_3D_Point_Cloud_Generation_CVPR_2021_paper.pdf | cvpr-2021-1 | ['point-cloud-generation'] | ['computer-vision'] | [ 2.23497540e-01 1.56976119e-01 1.33604005e-01 -1.45423964e-01
-8.33719909e-01 -6.03974044e-01 8.62493396e-01 -1.70542032e-01
2.82361489e-02 4.19372261e-01 -6.43240884e-02 -2.17158526e-01
2.89400697e-01 -1.02386200e+00 -1.02135301e+00 -9.48270917e-01
4.10882324e-01 9.51072097e-01 -2.18414158e-01 2.74476886e-01
1.69825733e-01 9.25529420e-01 -1.39107883e+00 -1.46369606e-01
7.87900031e-01 6.66672587e-01 5.15005410e-01 8.83998752e-01
-3.31481308e-01 1.70013696e-01 -1.43627793e-01 -3.58277798e-01
3.02997023e-01 -1.97800010e-01 -5.28494418e-01 2.99156010e-01
2.25066647e-01 -2.07176536e-01 -5.81439912e-01 1.30612421e+00
3.27909052e-01 1.10825181e-01 1.30543470e+00 -1.13663459e+00
-1.12372291e+00 1.38789192e-01 -6.82267785e-01 -2.02899486e-01
-3.05637382e-02 1.88827410e-01 7.46264398e-01 -1.24888337e+00
7.42210150e-01 1.17777419e+00 5.16683340e-01 1.03556967e+00
-1.50267279e+00 -4.98414278e-01 -2.33299211e-01 -4.09520954e-01
-1.44011617e+00 -4.65544999e-01 1.01834190e+00 -6.97129309e-01
5.21036327e-01 2.03808814e-01 8.82320106e-01 1.10125375e+00
1.24931931e-01 9.06844437e-01 7.87383199e-01 -2.41333336e-01
5.53596258e-01 -1.77447125e-01 -3.69704127e-01 4.86073554e-01
3.04074675e-01 3.59768122e-02 -3.18243921e-01 -5.30892372e-01
1.29963934e+00 2.99431354e-01 -1.23794422e-01 -7.21166790e-01
-1.10295188e+00 8.52893651e-01 3.62791151e-01 -9.15906131e-02
-6.68075562e-01 5.84764779e-01 -2.61119157e-01 -8.36841911e-02
8.60531032e-01 -4.05813046e-02 -1.37198567e-01 -2.60550249e-02
-8.37807298e-01 7.04923630e-01 7.29689538e-01 1.33338070e+00
8.52175236e-01 -7.12345168e-02 2.85821198e-03 6.50679052e-01
7.85049796e-01 1.28598702e+00 -1.92856103e-01 -1.41917074e+00
1.64115459e-01 2.72445649e-01 3.57070029e-01 -4.96633559e-01
2.53054559e-01 -3.17806870e-01 -1.16330361e+00 4.51483369e-01
1.90424681e-01 -1.05000801e-01 -1.05111563e+00 1.50161088e+00
5.44424057e-01 6.71477854e-01 -1.05027176e-01 8.73916209e-01
6.07644975e-01 8.68441999e-01 -2.40698516e-01 -1.04971178e-01
9.60064292e-01 -3.58678520e-01 -4.19762850e-01 1.38511509e-01
1.44451648e-01 -8.51853549e-01 8.95656288e-01 2.23234564e-01
-1.39741468e+00 -2.80844361e-01 -7.48124063e-01 -1.91137373e-01
1.35755539e-01 -2.81295210e-01 3.93186420e-01 2.24071130e-01
-1.09436047e+00 8.72784853e-01 -1.29297876e+00 -1.68457568e-01
8.41143727e-01 5.56524843e-02 -8.65912363e-02 -1.50008172e-01
-5.14198244e-01 4.11692053e-01 -3.34358186e-01 -9.76443663e-03
-9.88103211e-01 -8.43980193e-01 -5.73672712e-01 -2.82555193e-01
-1.16243303e-01 -1.49261642e+00 1.36492252e+00 -4.31434005e-01
-1.51141179e+00 1.06932330e+00 -6.23724222e-01 -3.99558932e-01
5.77395558e-01 -1.57554030e-01 2.82890558e-01 1.07132206e-02
5.04205152e-02 6.87827826e-01 1.24763417e+00 -1.73610759e+00
-2.24346980e-01 -4.86366510e-01 -4.41160560e-01 2.19939277e-01
4.23097968e-01 -6.35805428e-01 -5.74253976e-01 -4.80082035e-01
4.87780422e-01 -1.37252223e+00 -4.69708651e-01 3.11556458e-01
-2.95332462e-01 -2.68885374e-01 7.21724093e-01 -1.28717184e-01
3.80577177e-01 -1.95607519e+00 3.50609064e-01 2.98988670e-01
5.33589482e-01 -1.66378409e-01 3.63905765e-02 4.60412949e-01
1.25851661e-01 2.20024690e-01 -6.51712656e-01 -8.35576236e-01
2.17940748e-01 3.25355291e-01 -6.65770471e-01 6.65098846e-01
2.70400465e-01 1.07263875e+00 -1.05514860e+00 -2.68651932e-01
4.42330569e-01 8.60029876e-01 -6.82527661e-01 5.86179160e-02
-5.09252131e-01 7.56437540e-01 -6.68447077e-01 5.42434216e-01
1.18530524e+00 -3.32866043e-01 -4.29259360e-01 1.52999103e-01
2.01748051e-02 -6.91570267e-02 -1.11720049e+00 1.92995167e+00
-3.88493031e-01 3.62786889e-01 1.04445487e-01 -3.96974087e-01
9.92210567e-01 2.55650312e-01 6.09960437e-01 1.57080919e-01
1.36645794e-01 2.20695779e-01 -4.41875100e-01 -1.67270988e-01
6.12783611e-01 -5.71231544e-01 2.26421639e-01 3.44136804e-01
-1.21231571e-01 -1.13771737e+00 -2.92216867e-01 2.88967460e-01
9.34255481e-01 2.71461636e-01 -3.93735975e-01 -2.32754480e-02
2.08411425e-01 -1.30955264e-01 3.99154663e-01 6.74128294e-01
1.92081273e-01 1.11764824e+00 1.07034788e-01 -1.02980770e-01
-1.58767700e+00 -1.64591444e+00 -1.78320572e-01 8.42196122e-02
4.84411418e-02 -1.71126187e-01 -7.88043439e-01 -2.06918061e-01
3.01957011e-01 8.55395973e-01 -4.25284594e-01 -9.34007093e-02
-3.11287075e-01 -3.13648462e-01 1.74098924e-01 1.34581268e-01
1.64228901e-01 -9.63716149e-01 -3.56429629e-02 -1.70806479e-02
-1.25154465e-01 -8.14955533e-01 -5.85211039e-01 -2.99259514e-01
-1.35746384e+00 -7.12559879e-01 -1.04350483e+00 -5.61752081e-01
8.93724024e-01 3.24221849e-01 1.19190073e+00 1.00556284e-01
-7.66636655e-02 6.85702264e-01 9.30031165e-02 -5.69803774e-01
-5.74353278e-01 -1.32690325e-01 1.00522146e-01 -2.55613923e-02
3.37898076e-01 -9.37883794e-01 -8.01342726e-01 -2.00176656e-01
-1.10686231e+00 1.49087399e-01 3.94159228e-01 5.24211824e-01
1.15483558e+00 -2.37129256e-01 -2.90461387e-02 -6.67186320e-01
4.51078683e-01 -5.20753682e-01 -7.85819769e-01 -2.90722877e-01
-2.33347207e-01 1.95313573e-01 9.68713760e-02 -2.93999910e-01
-8.14998746e-01 3.36741745e-01 -3.36547226e-01 -1.12031090e+00
-2.32093498e-01 -9.18533560e-03 -8.25578719e-02 -2.21673492e-03
3.96132588e-01 5.92633545e-01 2.94122189e-01 -5.38109004e-01
7.79449463e-01 2.73905963e-01 4.50711280e-01 -9.38195765e-01
1.29151773e+00 1.10509217e+00 3.47873718e-01 -1.09399188e+00
-4.55473363e-01 -3.39342147e-01 -8.10588360e-01 -1.82362765e-01
7.72022188e-01 -9.48361278e-01 -8.18933606e-01 6.06558263e-01
-1.44478440e+00 -3.82120490e-01 -9.18471634e-01 2.52489150e-01
-1.05473495e+00 3.13015133e-01 -3.84395808e-01 -1.18620074e+00
-3.50133002e-01 -8.80601883e-01 1.47541606e+00 2.32859477e-01
1.44656733e-01 -1.02251363e+00 3.56971323e-01 1.19352199e-01
1.67232558e-01 3.18949699e-01 7.16629207e-01 5.12606576e-02
-1.10750115e+00 -1.65211231e-01 -2.06143707e-02 4.94559467e-01
4.57892008e-03 2.33560249e-01 -8.80036294e-01 -1.98607266e-01
2.41549939e-01 9.78461653e-02 8.26767802e-01 7.52433717e-01
1.13159740e+00 -1.08125001e-01 -6.29114926e-01 9.44140196e-01
1.40149760e+00 -1.09557614e-01 6.63704157e-01 -3.18543583e-01
8.59533668e-01 2.93029010e-01 8.46322849e-02 6.23687625e-01
3.27047974e-01 2.26042002e-01 5.65352321e-01 1.90298662e-01
-1.98696360e-01 -5.56127667e-01 -1.85904242e-02 8.76024067e-01
-2.52398759e-01 -1.41023979e-01 -1.02217007e+00 6.93530083e-01
-1.58051980e+00 -1.00151241e+00 -4.29078370e-01 2.26108384e+00
7.15556800e-01 -9.44977328e-02 -9.36276019e-02 -2.04218283e-01
6.08499646e-01 5.18067256e-02 -7.99652159e-01 8.47673416e-02
8.21058545e-03 3.73458534e-01 5.94514608e-01 8.61987233e-01
-7.41642356e-01 8.92031610e-01 5.71802950e+00 6.43012464e-01
-9.50150371e-01 2.55721033e-01 4.21488047e-01 -2.50039041e-01
-8.63186836e-01 9.73892808e-02 -6.87804937e-01 4.67476130e-01
5.77939093e-01 -4.61631864e-01 4.79280114e-01 5.73406816e-01
3.45272154e-01 2.56680667e-01 -9.79388475e-01 1.27498925e+00
-3.77265096e-01 -1.65056443e+00 3.17022920e-01 5.32236099e-01
9.62285697e-01 4.73121822e-01 4.44944441e-01 -1.97307736e-01
5.59931874e-01 -6.77138865e-01 7.94645429e-01 1.04846525e+00
8.39932323e-01 -6.30878389e-01 1.66505814e-01 5.32779872e-01
-8.85051906e-01 6.09038472e-01 -6.18140817e-01 1.05223142e-01
2.89901584e-01 9.83865976e-01 -8.16385329e-01 3.70054096e-01
4.01846439e-01 8.60393822e-01 9.06045139e-02 9.67228651e-01
-1.43343613e-01 5.20716488e-01 -5.58305681e-01 3.21594253e-02
-6.84391633e-02 -9.36754644e-01 1.13119423e+00 7.21148193e-01
7.78961420e-01 3.40907156e-01 -2.73121089e-01 1.45129526e+00
-4.62256551e-01 -1.84942231e-01 -1.07635331e+00 4.81582247e-02
5.98286033e-01 1.07398319e+00 -6.10389292e-01 -3.16329449e-01
-1.18135229e-01 1.10563362e+00 1.06221303e-01 5.12822211e-01
-5.61463416e-01 -6.91557443e-03 1.07678771e+00 3.58102947e-01
3.59623611e-01 -6.31779075e-01 -5.34150481e-01 -1.19907212e+00
1.27142772e-01 -1.55866340e-01 -4.07650292e-01 -1.11819792e+00
-1.57082796e+00 4.07143950e-01 -1.69175729e-01 -1.41241062e+00
6.93573803e-02 -2.54107147e-01 -6.62607551e-01 1.23877084e+00
-1.35795867e+00 -1.11183751e+00 -1.49829715e-01 5.10502398e-01
4.19154584e-01 1.88549712e-01 6.58102334e-01 -6.74179569e-02
6.30066469e-02 -1.13797277e-01 4.87969935e-01 -2.37781093e-01
2.22828239e-01 -1.49803495e+00 1.10512567e+00 5.75951278e-01
1.51281193e-01 6.78650916e-01 7.97749043e-01 -9.51398313e-01
-1.67955422e+00 -1.14730752e+00 7.13076830e-01 -8.32474113e-01
5.31717718e-01 -5.20016551e-01 -1.02518630e+00 5.60840011e-01
-5.08831032e-02 1.74952984e-01 3.06484610e-01 -3.87670934e-01
5.13616018e-02 3.02575976e-01 -1.22023606e+00 6.13861918e-01
1.14792120e+00 -5.49089313e-01 -4.31678087e-01 6.47816956e-01
6.65674269e-01 -6.81298971e-01 -7.70357370e-01 -2.35892143e-02
1.95926413e-01 -5.34310758e-01 1.19988143e+00 -2.75932431e-01
5.32594979e-01 -4.02047932e-01 -2.45061487e-01 -1.34395564e+00
-4.24916923e-01 -8.82644713e-01 -2.81556875e-01 1.11096549e+00
2.66382277e-01 -5.76011717e-01 1.18183887e+00 6.27946734e-01
-4.65417206e-02 -6.93929553e-01 -1.01259589e+00 -6.99502289e-01
5.65178752e-01 -6.26821637e-01 8.21701705e-01 6.25162423e-01
-7.49888659e-01 1.44040689e-01 -1.06435537e-01 5.18428445e-01
1.30004990e+00 1.24915235e-01 8.34286094e-01 -1.40099049e+00
-5.80083802e-02 -3.74423891e-01 -9.03218165e-02 -1.59729147e+00
1.36430323e-01 -1.20272863e+00 2.01235861e-01 -1.91512370e+00
1.54396549e-01 -8.01934004e-01 2.91755855e-01 -2.92450696e-01
-2.16238163e-02 1.45442083e-01 3.14282686e-01 6.74096525e-01
-6.19251281e-02 1.06265140e+00 1.57545066e+00 -1.18135244e-01
-2.33751014e-01 4.55316931e-01 -5.27589798e-01 6.98170662e-01
6.49919391e-01 -6.69248521e-01 -3.49519551e-01 -6.47194862e-01
2.22826809e-01 1.38918787e-01 5.02479315e-01 -7.91167080e-01
2.01523483e-01 -2.61532158e-01 3.60505998e-01 -9.72883403e-01
7.77532816e-01 -8.67735445e-01 4.25275385e-01 3.82341206e-01
4.05218117e-02 7.09879585e-03 -5.49601167e-02 8.34558129e-01
2.28682935e-01 -2.54107028e-01 7.67269194e-01 -2.20032841e-01
7.56312311e-02 1.02044606e+00 -1.39265403e-01 8.18824768e-02
8.21342945e-01 2.24480983e-02 -8.92665423e-03 -5.13489306e-01
-9.54113841e-01 -2.01270115e-02 9.97886062e-01 2.37648845e-01
1.02959180e+00 -1.71377861e+00 -9.27701890e-01 3.37842286e-01
-2.02416331e-01 7.63001978e-01 2.67498970e-01 4.83891755e-01
-6.44419253e-01 8.73297602e-02 3.79295260e-01 -1.06277812e+00
-7.96296477e-01 4.24864888e-01 3.89034688e-01 2.98358798e-01
-7.99816191e-01 9.86163557e-01 2.72967279e-01 -5.84366679e-01
-2.27327988e-01 -5.61313868e-01 4.58205342e-01 -4.38817590e-01
3.00943375e-01 2.04324543e-01 -1.45601124e-01 -5.72026134e-01
-1.47865444e-01 7.83105671e-01 2.31595516e-01 -4.48404372e-01
1.42317820e+00 -2.20431797e-02 -2.52819031e-01 6.50790334e-01
1.13804293e+00 6.34137392e-02 -1.61909211e+00 -3.58544558e-01
-3.66283715e-01 -6.09400749e-01 1.48182020e-01 -5.05783334e-02
-1.07456827e+00 9.07643855e-01 3.77316713e-01 2.60105163e-01
5.10661304e-01 4.68385279e-01 8.39631617e-01 9.54305008e-02
3.49345058e-01 -5.35991073e-01 -2.46507987e-01 5.73387623e-01
1.08165538e+00 -9.44721580e-01 -6.00056797e-02 -3.32949162e-01
-3.88360023e-01 7.62601972e-01 2.44756993e-02 -6.63945675e-01
1.16303182e+00 2.95935184e-01 -1.16037101e-01 -3.94986212e-01
-8.64154935e-01 -6.59625381e-02 1.30116433e-01 8.95082533e-01
1.94327563e-01 2.78528780e-01 4.43739980e-01 -1.95588712e-02
-3.41420949e-01 1.77097112e-01 4.65524316e-01 7.70076215e-01
-2.68796831e-01 -1.13749599e+00 -4.28800851e-01 4.36076820e-01
6.56322762e-03 -1.47295445e-01 -2.26540387e-01 2.89680481e-01
-7.00903982e-02 4.97043639e-01 4.14919525e-01 3.21269743e-02
4.80969965e-01 1.66404054e-01 8.06488931e-01 -6.98200047e-01
1.73652381e-01 2.76709646e-01 -6.81135416e-01 -1.65084466e-01
-2.93833703e-01 -1.25151932e+00 -1.30810642e+00 -6.52199686e-01
-9.70764682e-02 2.39676401e-01 9.62323606e-01 5.23345351e-01
5.74157059e-01 2.87918627e-01 6.27965689e-01 -1.32445431e+00
-5.31636298e-01 -8.63858759e-01 -7.14629948e-01 3.87991041e-01
5.65113366e-01 -5.06648421e-01 -5.85351408e-01 2.86482871e-01] | [8.876977920532227, -3.650486707687378] |
5861ceaa-09a2-4f03-a610-79cf2e5b9e27 | is-this-a-joke-detecting-humor-in-spanish | 1703.09527 | null | http://arxiv.org/abs/1703.09527v1 | http://arxiv.org/pdf/1703.09527v1.pdf | Is This a Joke? Detecting Humor in Spanish Tweets | While humor has been historically studied from a psychological, cognitive and
linguistic standpoint, its study from a computational perspective is an area
yet to be explored in Computational Linguistics. There exist some previous
works, but a characterization of humor that allows its automatic recognition
and generation is far from being specified. In this work we build a
crowdsourced corpus of labeled tweets, annotated according to its humor value,
letting the annotators subjectively decide which are humorous. A humor
classifier for Spanish tweets is assembled based on supervised learning,
reaching a precision of 84% and a recall of 69%. | ['Guillermo Moncecchi', 'Matías Cubero', 'Diego Garat', 'Santiago Castro'] | 2017-03-28 | null | null | null | null | ['humor-detection'] | ['natural-language-processing'] | [-4.00558263e-01 3.14090878e-01 -2.48718604e-01 -2.95467824e-01
-1.26470178e-01 -4.62361246e-01 8.13863337e-01 5.03245950e-01
-2.72476196e-01 7.18742728e-01 5.41940928e-01 -2.00631365e-01
4.77603853e-01 -7.10034847e-01 2.65593603e-02 -4.13595706e-01
3.14708084e-01 6.12804830e-01 1.72605842e-01 -5.28946936e-01
5.29702127e-01 1.04589179e-01 -1.57330370e+00 5.02902210e-01
8.99175823e-01 4.56738293e-01 -1.64556965e-01 5.91063678e-01
-1.17972352e-01 1.63397241e+00 -6.92288160e-01 -8.20275784e-01
-1.53160602e-01 -7.42361903e-01 -1.05712819e+00 2.28197634e-01
5.25911525e-02 5.59796356e-02 -4.35030945e-02 9.18736696e-01
2.77590156e-01 3.58402468e-02 7.36663342e-01 -9.43993390e-01
-8.51757288e-01 8.14600706e-01 -2.24386021e-01 -6.01536073e-02
8.74031842e-01 2.10529864e-01 1.19870543e+00 -1.03080988e+00
8.53933513e-01 1.04819047e+00 4.53972876e-01 4.01623309e-01
-9.54439282e-01 -2.71036297e-01 -7.46370792e-01 4.98136371e-01
-1.11475432e+00 -1.50617540e-01 9.78129208e-01 -8.76511157e-01
6.48334682e-01 1.42784461e-01 6.57021046e-01 8.82641554e-01
-8.43261480e-02 8.01756561e-01 1.55370569e+00 -6.45828366e-01
4.80062366e-01 7.36859083e-01 3.17247927e-01 4.80535775e-01
6.39291108e-02 -4.26517546e-01 -6.78450942e-01 3.21945660e-02
1.87188506e-01 -4.95798171e-01 -3.40322673e-01 9.68056992e-02
-6.00192249e-01 1.26256073e+00 4.16663527e-01 8.88185978e-01
-3.84733468e-01 -5.02669454e-01 6.55841112e-01 3.75732660e-01
5.73139548e-01 6.91783130e-01 5.61769232e-02 -2.01999605e-01
-1.17205501e+00 4.13200200e-01 1.30907607e+00 7.64371216e-01
6.06354535e-01 9.25060585e-02 -1.29185505e-02 7.38505781e-01
7.96344355e-02 4.49293315e-01 6.53494895e-01 -5.43015778e-01
-2.28857640e-02 1.15163136e+00 2.81398386e-01 -1.53855193e+00
-6.28970563e-01 -8.75403956e-02 -4.17072266e-01 2.33866870e-01
6.10568225e-01 -3.40734012e-02 -2.92672485e-04 1.20942485e+00
4.70958613e-02 -4.52572554e-01 1.67201906e-02 1.29946625e+00
1.23538852e+00 5.19880891e-01 5.50355054e-02 -3.83223802e-01
1.48133647e+00 -9.07943249e-01 -8.83526266e-01 -1.33305445e-01
5.68295240e-01 -1.01855671e+00 1.18711400e+00 7.42691875e-01
-1.04369211e+00 -5.43622449e-02 -9.16460633e-01 -4.84595120e-01
-5.15254855e-01 6.50987402e-02 3.02130640e-01 8.43068361e-01
-6.61872506e-01 1.26557440e-01 -2.39639685e-01 -4.91836160e-01
1.48640767e-01 -2.26831526e-01 -2.21643001e-01 3.74486148e-01
-1.38614762e+00 1.49379241e+00 5.87764740e-01 -3.22540969e-01
-6.34453535e-01 -1.22400583e-04 -7.17188895e-01 -1.40601292e-01
1.88616931e-01 -4.78945613e-01 1.33978629e+00 -1.24452984e+00
-1.52308154e+00 1.37969863e+00 -2.69739628e-01 -8.55105937e-01
5.64805210e-01 4.14743759e-02 -4.77090001e-01 1.89951003e-01
5.25217056e-02 5.25564961e-02 6.98847055e-01 -1.30587757e+00
-3.35123897e-01 -3.42923641e-01 2.38202766e-01 2.94575188e-02
-7.04447210e-01 4.77923095e-01 1.67706385e-01 -5.27380526e-01
-1.99069440e-01 -7.59360135e-01 1.56089157e-01 -5.34289420e-01
-2.96794981e-01 -2.82817334e-01 3.03122252e-01 -7.44272649e-01
1.75643492e+00 -1.68800175e+00 5.65693453e-02 -1.71777397e-01
6.48888946e-01 3.24814588e-01 4.92537260e-01 8.38394105e-01
2.93633759e-01 -3.46230045e-02 -1.73015505e-01 -2.05597773e-01
2.90365726e-01 -1.83365837e-01 -2.47713834e-01 5.33236265e-01
-2.10631698e-01 8.17027092e-01 -1.20247257e+00 -7.55111158e-01
2.10937276e-01 2.94411570e-01 -4.70094711e-01 3.26952815e-01
-1.57180607e-01 2.81580817e-02 -4.51801151e-01 6.14824414e-01
2.97745675e-01 -2.45101228e-01 4.60427791e-01 2.48562574e-01
-3.02222192e-01 3.57733518e-01 -6.32112622e-01 7.70128608e-01
-5.20472348e-01 9.51258183e-01 -1.75457940e-01 -5.08469164e-01
1.49366915e+00 4.50560123e-01 2.62203693e-01 -4.97520447e-01
5.24698913e-01 5.97594261e-01 -3.19905989e-02 -8.34735930e-01
1.12853301e+00 -5.85698426e-01 -3.65154654e-01 4.85813081e-01
-1.02382377e-01 -2.65168786e-01 3.88401777e-01 2.03662917e-01
1.01147294e+00 -1.03512235e-01 9.16628897e-01 -4.31493610e-01
8.94314229e-01 4.62590456e-01 5.85447662e-02 3.94884109e-01
-3.08804393e-01 4.21628267e-01 6.68018818e-01 -5.90141654e-01
-1.23012400e+00 -4.08702195e-01 -5.09395063e-01 1.15194547e+00
-4.53860946e-02 -4.63296801e-01 -1.08510816e+00 -3.02260667e-01
-1.73439726e-01 1.01852894e+00 -5.13844728e-01 2.58092165e-01
-3.54126781e-01 -5.15739858e-01 4.27741259e-01 -2.43460741e-02
2.32335925e-01 -1.45771635e+00 -1.03992331e+00 2.80682743e-01
-2.40166634e-01 -1.15453732e+00 8.69448259e-02 -6.60274178e-03
-4.10982579e-01 -1.22703588e+00 -3.17279369e-01 -6.35637820e-01
2.84748882e-01 1.47853881e-01 1.42614162e+00 3.41218024e-01
1.94369763e-01 -1.09744199e-01 -9.40057874e-01 -4.62103665e-01
-8.40701759e-01 4.37266380e-02 -7.33422935e-02 -1.94302410e-01
9.89067972e-01 -3.27316970e-01 -2.78106183e-01 2.99723595e-01
-7.60625362e-01 -1.53478339e-01 1.68256477e-01 6.98763728e-01
-3.95730913e-01 -2.15216100e-01 6.95174038e-01 -1.19116890e+00
8.79552484e-01 -8.16777527e-01 -1.09162740e-01 -3.43176693e-01
-5.95622480e-01 -5.18420935e-01 1.04254913e+00 -7.88962916e-02
-8.84280145e-01 -2.36201901e-02 -1.08753659e-01 1.18448697e-01
-3.16790849e-01 6.08800650e-01 1.69473097e-01 1.63706064e-01
1.25294745e+00 2.51616593e-02 -9.28680152e-02 -2.85720557e-01
5.15146494e-01 9.59704518e-01 5.58348238e-01 -3.93958867e-01
6.58740163e-01 2.74033219e-01 -2.68518418e-01 -8.85582387e-01
-1.15910447e+00 -7.89472103e-01 -8.18620086e-01 -5.73460162e-01
8.02738488e-01 -7.43320525e-01 -8.32099736e-01 1.55233756e-01
-1.42851329e+00 -4.81225923e-02 -2.02965904e-02 2.57580489e-01
-6.77213371e-01 3.62571836e-01 -6.57092035e-01 -1.11454844e+00
-4.16486263e-01 -6.85653329e-01 3.40938956e-01 1.23919234e-01
-8.56277823e-01 -1.10130036e+00 1.98075160e-01 8.67159069e-01
1.78636000e-01 2.46519312e-01 6.02051258e-01 -1.03781593e+00
1.01639591e-01 -5.62116385e-01 -5.42596951e-02 1.12494053e-02
-4.16866034e-01 -1.86260417e-01 -1.04374075e+00 2.12579668e-01
4.44293499e-01 -6.64997995e-01 5.49834728e-01 -1.63061649e-01
4.25062090e-01 -6.74627960e-01 9.54563543e-02 -2.79621184e-01
1.41696513e+00 -3.84218954e-02 8.68939757e-01 7.60310948e-01
2.83540100e-01 1.01620686e+00 5.87094069e-01 9.94871080e-01
5.09136260e-01 5.59937835e-01 5.32625556e-01 4.35697913e-01
-3.51765081e-02 -2.53329754e-01 3.59842271e-01 8.40444088e-01
-1.98914886e-01 -3.56601089e-01 -1.16395366e+00 7.17951894e-01
-1.93570065e+00 -1.28097701e+00 -9.81837153e-01 1.88296080e+00
1.00486898e+00 2.13760093e-01 6.14009857e-01 4.43262011e-01
7.00727761e-01 2.99300432e-01 2.64599025e-01 -7.61417925e-01
-3.51346344e-01 -1.30920336e-01 -4.09106724e-02 7.85100996e-01
-6.88041806e-01 1.13060248e+00 6.36058331e+00 5.88904381e-01
-9.33965445e-01 2.65878022e-01 2.86589056e-01 2.05396086e-01
-4.04060453e-01 6.00764118e-02 -5.41139185e-01 5.16766548e-01
7.33224392e-01 -1.92613557e-01 4.15810317e-01 1.15888500e+00
6.08254313e-01 -2.66938329e-01 -6.84687316e-01 9.09990668e-01
6.76424980e-01 -8.19330573e-01 -2.40420014e-01 -1.70311421e-01
7.66075730e-01 -3.86730880e-01 -2.68266171e-01 3.62290710e-01
1.43725380e-01 -1.26412237e+00 1.05062842e+00 6.20157957e-01
7.37801939e-02 -6.44211709e-01 1.01196432e+00 7.29000628e-01
-5.04794836e-01 -2.56119706e-02 -3.62744272e-01 -7.60298312e-01
3.92216206e-01 9.12547648e-01 -1.22059953e+00 9.13424715e-02
2.53004253e-01 4.88039106e-01 -8.23224425e-01 6.92553163e-01
-6.35103166e-01 9.60026622e-01 1.96172282e-01 -9.54117179e-01
1.97791383e-01 -1.53440177e-01 7.61197507e-01 1.64685929e+00
-5.74634559e-02 7.81597719e-02 2.38983035e-01 1.20323253e+00
-4.01430763e-02 8.16963732e-01 -5.90026855e-01 -6.63498566e-02
4.28170532e-01 1.64588523e+00 -6.27754271e-01 -2.25238174e-01
-1.95514306e-01 6.55360460e-01 3.38855118e-01 -2.98841298e-01
-6.97413325e-01 -2.12034643e-01 -2.62924612e-01 5.39812386e-01
-2.52078652e-01 -2.15890661e-01 -9.92755532e-01 -1.12713397e+00
-1.12873830e-01 -7.44264424e-01 1.70835033e-01 -7.07101882e-01
-1.39232218e+00 8.20243895e-01 -1.97238177e-01 -1.06426454e+00
-3.74395967e-01 -7.07224190e-01 -7.32435465e-01 6.62562311e-01
-1.09036875e+00 -1.32538164e+00 -5.86183727e-01 9.11499783e-02
4.20200527e-01 -3.96164805e-01 6.67088389e-01 1.31388232e-01
-3.05599958e-01 7.03658462e-02 -3.53125662e-01 1.22002810e-02
6.24889970e-01 -1.44123816e+00 -2.73559541e-01 4.86754805e-01
-1.86367840e-01 4.76003766e-01 1.55171132e+00 -4.40015525e-01
-1.06893218e+00 -4.49887037e-01 1.90631938e+00 -7.92805016e-01
1.24565673e+00 6.99189454e-02 -1.16018414e+00 2.03118742e-01
5.37984490e-01 -5.70409119e-01 7.26942778e-01 2.17457443e-01
-5.84936857e-01 5.67449749e-01 -1.09316587e+00 4.73252565e-01
5.08263111e-01 -6.40834093e-01 -8.89124691e-01 3.83719355e-01
-1.15462393e-02 -1.87754527e-01 -8.88517737e-01 -1.30401284e-01
3.27573121e-01 -1.37062657e+00 3.23093921e-01 -3.08304459e-01
1.15336061e+00 -4.52363998e-01 -8.82393420e-02 -9.87790883e-01
-1.50992438e-01 -4.03123111e-01 7.69114420e-02 1.27504122e+00
3.10456455e-01 -1.63750038e-01 7.08336473e-01 3.29969943e-01
-1.73304811e-01 -4.55523551e-01 -5.52924514e-01 -5.45513868e-01
3.60814035e-01 -5.24966002e-01 2.48825207e-01 1.20586848e+00
1.10339916e+00 7.90884674e-01 -6.96781099e-01 -5.55643260e-01
2.27342516e-01 2.23166049e-01 9.11935508e-01 -1.20695257e+00
-3.77969183e-02 -8.95075798e-01 -3.98883194e-01 -5.91541529e-01
5.02268612e-01 -1.10925961e+00 1.66604534e-01 -1.31972063e+00
6.72255218e-01 1.79838300e-01 4.74737078e-01 3.75462383e-01
-2.44206982e-03 6.58038795e-01 2.92482555e-01 3.71825248e-01
-8.00934076e-01 3.89002323e-01 9.41066623e-01 9.95553881e-02
-2.23055512e-01 -5.10618746e-01 -5.89410722e-01 1.06330168e+00
1.04611111e+00 -4.03225064e-01 3.19736525e-02 3.02598476e-01
6.13052845e-01 6.05716072e-02 4.22892302e-01 -7.64922082e-01
1.49046972e-01 -3.79302323e-01 -1.77937806e-01 -3.49309504e-01
2.00096384e-01 -5.56252658e-01 -4.64637093e-02 3.74876738e-01
-2.44042009e-01 -2.69181311e-01 -2.60744661e-01 1.33756667e-01
-3.74573529e-01 -1.02656996e+00 1.11018515e+00 -2.33286411e-01
-5.41400075e-01 -3.77642423e-01 -7.39814162e-01 1.52233779e-01
9.98561263e-01 -2.83509493e-01 -3.42343807e-01 -6.89578712e-01
-5.04781485e-01 -1.41818658e-01 9.56874073e-01 2.37015277e-01
4.36906219e-01 -1.18574810e+00 -1.02968132e+00 -5.83542347e-01
3.56597096e-01 -7.40978062e-01 4.83735427e-02 8.07252586e-01
-7.29129851e-01 3.44599873e-01 -2.31968313e-01 -2.00811237e-01
-1.07919610e+00 8.77683163e-01 2.27698609e-02 -1.54512361e-01
-4.74162072e-01 8.11714232e-02 -3.68114203e-01 -3.20649207e-01
-2.86571145e-01 6.12519860e-01 -8.41176212e-01 4.54189122e-01
6.11334324e-01 8.03136587e-01 -8.80082548e-02 -1.39929175e+00
-4.84077595e-02 1.34799197e-01 4.23715055e-01 -1.50540277e-01
1.07830548e+00 -1.74807101e-01 -6.38694823e-01 8.55701149e-01
9.50098872e-01 2.84534305e-01 -1.36718705e-01 -2.20049894e-03
4.98097956e-01 -6.24899507e-01 -9.83584300e-02 -6.98487878e-01
-3.34577084e-01 6.83916211e-01 -1.53708711e-01 8.87117088e-01
8.03001881e-01 -3.58116031e-02 8.12366307e-01 2.39901319e-01
3.84667277e-01 -1.29114127e+00 4.00849551e-01 1.05358958e+00
1.11000931e+00 -1.33287978e+00 -4.11878712e-02 -3.17949504e-01
-1.17273819e+00 1.44720960e+00 4.77314502e-01 -3.48587394e-01
4.49022681e-01 1.20993711e-01 9.74843875e-02 -3.82134974e-01
-6.27707541e-01 -4.84004408e-01 4.67355192e-01 2.91676760e-01
1.18007493e+00 2.63938427e-01 -1.43272734e+00 9.23494935e-01
-8.04277897e-01 8.08278248e-02 9.97648120e-01 5.21396339e-01
-1.17070675e+00 -7.61008084e-01 -4.74747270e-01 1.16839126e-01
-5.50829530e-01 3.30266804e-02 -8.75693679e-01 4.85944331e-01
2.15014189e-01 1.38103831e+00 -4.35469568e-01 -5.04563272e-01
9.41518843e-02 7.55495951e-02 1.00497730e-01 -7.76719093e-01
-1.10661292e+00 -2.25223765e-01 6.00999057e-01 1.54542789e-01
-6.42325044e-01 -6.67362630e-01 -1.32699883e+00 -8.36529791e-01
-2.04569563e-01 3.66046518e-01 5.92721403e-01 1.05662632e+00
-4.74122882e-01 -4.86484580e-02 7.97507882e-01 -5.85536122e-01
-3.14993441e-01 -1.05291629e+00 -6.37569785e-01 8.35104525e-01
-8.29651281e-02 -3.97980183e-01 -4.66178268e-01 2.36126661e-01] | [8.908088684082031, 10.973648071289062] |
d38d835e-0eec-40ec-bb1e-c3a1245e9d84 | machine-learning-for-music-genre-multifaceted | 1911.12618 | null | https://arxiv.org/abs/1911.12618v1 | https://arxiv.org/pdf/1911.12618v1.pdf | Machine learning for music genre: multifaceted review and experimentation with audioset | Music genre classification is one of the sub-disciplines of music information retrieval (MIR) with growing popularity among researchers, mainly due to the already open challenges. Although research has been prolific in terms of number of published works, the topic still suffers from a problem in its foundations: there is no clear and formal definition of what genre is. Music categorizations are vague and unclear, suffering from human subjectivity and lack of agreement. In its first part, this paper offers a survey trying to cover the many different aspects of the matter. Its main goal is give the reader an overview of the history and the current state-of-the-art, exploring techniques and datasets used to the date, as well as identifying current challenges, such as this ambiguity of genre definitions or the introduction of human-centric approaches. The paper pays special attention to new trends in machine learning applied to the music annotation problem. Finally, we also include a music genre classification experiment that compares different machine learning models using Audioset. | ['Jaime Ramírez', 'M. Julia Flores'] | 2019-11-28 | null | null | null | null | ['genre-classification'] | ['computer-vision'] | [ 3.51276785e-01 -3.70973676e-01 -3.93900692e-01 1.84018742e-02
-7.02588558e-01 -8.26626956e-01 4.60444957e-01 1.37229979e-01
-2.55858749e-01 5.04571199e-01 3.02302867e-01 3.31173182e-01
-6.56880617e-01 -3.13846081e-01 5.74886799e-04 -7.39286840e-01
-1.85628757e-01 6.31098628e-01 -1.33440405e-01 -4.57283169e-01
7.57675171e-01 1.59188777e-01 -2.20681524e+00 6.38251841e-01
1.50964707e-01 9.73107100e-01 3.94480936e-02 5.29452324e-01
-2.57076204e-01 6.47850454e-01 -1.04376030e+00 -4.33811069e-01
9.86729190e-03 -7.74011314e-01 -1.35730493e+00 8.68595093e-02
5.42699516e-01 4.81764406e-01 8.46366510e-02 1.00119138e+00
6.58756077e-01 6.46878108e-02 7.22417057e-01 -1.26184523e+00
-3.53152603e-01 9.92849350e-01 -5.46792567e-01 2.02660665e-01
6.72015250e-01 -4.63937253e-01 1.39240825e+00 -4.68806893e-01
6.41007900e-01 9.79230583e-01 8.77278924e-01 6.13817513e-01
-1.12427235e+00 -6.12820506e-01 -5.95323630e-02 8.53265941e-01
-1.62912905e+00 -2.54417449e-01 8.43479693e-01 -7.44997144e-01
6.68272138e-01 7.80486584e-01 8.38733971e-01 1.17696500e+00
-2.85723686e-01 6.40819550e-01 1.10875762e+00 -8.70344281e-01
1.26558200e-01 6.90340102e-02 2.75731772e-01 -6.69203773e-02
-4.37848223e-03 -2.95869112e-01 -8.21242392e-01 -2.91929096e-01
5.85290372e-01 -5.09864390e-01 -2.15907350e-01 -2.16842741e-01
-1.36710966e+00 6.58070445e-01 -1.75236046e-01 1.18693566e+00
7.53171369e-02 -3.19361180e-01 1.03410602e+00 6.94173098e-01
1.79853663e-01 1.02986515e+00 -4.52865899e-01 -5.97947717e-01
-1.30063176e+00 4.97210979e-01 9.63584840e-01 5.84084928e-01
2.23972946e-02 3.88363786e-02 2.14769378e-01 1.35091412e+00
4.93187504e-03 -1.44323587e-01 7.64318466e-01 -9.40924346e-01
-5.92147671e-02 2.43291616e-01 -2.89974868e-01 -1.01696873e+00
-4.98199165e-01 -6.11525893e-01 -8.73629093e-01 1.92969769e-01
5.69947660e-01 4.84216541e-01 -7.54042491e-02 1.43246901e+00
-4.56493534e-02 -1.79858152e-02 -3.66864920e-01 9.88078177e-01
1.11554945e+00 2.16805577e-01 -1.29466221e-01 -5.52720249e-01
1.64800751e+00 -9.27541137e-01 -8.30575287e-01 2.55113512e-01
1.53841242e-01 -1.35972619e+00 1.00637317e+00 1.05972767e+00
-1.07312179e+00 -6.98704183e-01 -1.12095618e+00 -5.74697591e-02
-3.98936749e-01 2.59000480e-01 6.78692818e-01 7.45997965e-01
-7.52203584e-01 9.48717177e-01 -2.62279660e-01 -7.29475558e-01
1.39354482e-01 3.32168818e-01 -2.55324870e-01 5.04294455e-01
-8.95667315e-01 6.45652890e-01 4.64111626e-01 -3.01105291e-01
-2.52411544e-01 -7.00734258e-01 -2.80831397e-01 -3.60022157e-01
4.16501284e-01 -6.12224698e-01 1.53804231e+00 -1.23827553e+00
-1.57105756e+00 1.43712103e+00 1.46020085e-01 -2.12111652e-01
3.27321172e-01 -7.38844573e-02 -7.05345213e-01 4.85120490e-02
1.98915899e-01 3.32679927e-01 7.34158397e-01 -9.98413980e-01
-7.58514225e-01 -4.53527749e-01 -2.95787957e-02 2.58368731e-01
-3.09669971e-01 4.78087187e-01 -3.18446904e-01 -1.31334424e+00
3.06618214e-01 -9.90466535e-01 2.05686182e-01 -4.68329430e-01
-3.24375272e-01 -7.03707516e-01 4.06795144e-01 -1.72901109e-01
2.00211859e+00 -2.28694487e+00 4.71686572e-01 -2.66132176e-01
6.28148317e-02 -1.78514004e-01 1.21512316e-01 9.55991745e-01
-4.02584046e-01 1.16376273e-01 -1.54265016e-01 -1.41228288e-01
1.04201056e-01 9.05551240e-02 -6.28723383e-01 4.37664658e-01
-6.23060703e-01 3.98831785e-01 -8.26261163e-01 -4.00520444e-01
1.64269507e-01 2.61904359e-01 -1.45606384e-01 -2.19151020e-01
-5.87537289e-02 6.21425927e-01 -1.10121362e-01 7.69044161e-01
2.97576606e-01 4.08671051e-02 1.50872782e-01 -2.15912059e-01
-5.00093997e-01 6.65934145e-01 -1.64551866e+00 2.06923580e+00
-4.68257107e-02 8.13234508e-01 -6.90565333e-02 -1.19857693e+00
5.67842662e-01 7.01727808e-01 8.19294274e-01 -1.53398573e-01
2.60555059e-01 6.71134233e-01 2.22569823e-01 -5.41348100e-01
7.11638629e-01 -3.73709232e-01 -3.08343440e-01 6.35047197e-01
2.33167306e-01 -1.21006161e-01 5.38473189e-01 -3.70598465e-01
7.75119960e-01 1.81969330e-01 7.95887411e-01 -3.36900800e-01
8.62184405e-01 1.34858891e-01 1.86656043e-01 6.67289853e-01
-5.59000149e-02 8.25505078e-01 2.59281427e-01 -5.46267211e-01
-8.68869603e-01 -5.41737020e-01 -7.18878210e-01 1.40888786e+00
-2.84091495e-02 -9.65338886e-01 -7.39920616e-01 -2.80263811e-01
-2.04354584e-01 1.92788944e-01 -6.61991835e-01 4.08831954e-01
-5.84126413e-01 -7.00645030e-01 8.35081816e-01 7.18189925e-02
2.17309788e-01 -1.51463342e+00 -5.40555596e-01 1.39428496e-01
-5.58533669e-01 -7.49785185e-01 -1.09125786e-01 3.57172079e-02
-1.04873240e+00 -1.33094716e+00 -8.08701992e-01 -9.44012046e-01
-1.99579462e-01 2.27639660e-01 1.57860970e+00 3.95041704e-02
-4.00524914e-01 3.95114690e-01 -7.24154592e-01 -6.85606539e-01
-4.65802461e-01 5.64062178e-01 3.78350653e-02 -1.84655026e-01
7.73375690e-01 -7.94733763e-01 -3.71571213e-01 5.11807024e-01
-7.56581485e-01 -2.44432047e-01 2.33074635e-01 5.59381247e-01
6.11513019e-01 1.24924630e-01 8.36840868e-01 -7.43885815e-01
5.75718462e-01 -1.64595172e-01 8.28871690e-03 -1.68327138e-01
-6.68583214e-01 -4.16686058e-01 2.23525062e-01 -5.14566720e-01
-3.56450766e-01 -2.58875310e-01 -3.32277000e-01 5.88412583e-03
-6.05528593e-01 3.39862704e-01 4.35163975e-02 6.98184967e-03
9.67929304e-01 -7.11609721e-02 -3.10928136e-01 -1.28199089e+00
1.89557761e-01 1.00442266e+00 5.18153071e-01 -5.81387341e-01
3.50907117e-01 4.67676610e-01 -6.25442341e-02 -1.23736858e+00
-1.14901519e+00 -1.08738101e+00 -8.21180701e-01 -5.08058906e-01
6.47902906e-01 -4.84877050e-01 -8.07234883e-01 3.75713557e-01
-8.59153569e-01 3.15178156e-01 -6.60287619e-01 4.46497232e-01
-1.03916860e+00 4.72532779e-01 -7.13473022e-01 -7.74202347e-01
-4.54358399e-01 -9.36043024e-01 1.10344732e+00 -8.59451890e-02
-9.54197049e-01 -7.44976997e-01 4.22593504e-01 6.11181498e-01
1.08442694e-01 1.95633724e-01 9.84650016e-01 -5.85285187e-01
-3.40971239e-02 -3.12576331e-02 3.94171208e-01 2.04342186e-01
9.18005407e-02 -3.62810731e-01 -1.47975636e+00 -4.05378938e-02
1.66486576e-01 -3.12038422e-01 7.82970130e-01 3.23342770e-01
1.23564589e+00 -1.92236956e-02 -1.28299519e-01 2.10843042e-01
1.32745421e+00 1.39885038e-01 4.65487778e-01 9.07279313e-01
2.39333302e-01 7.11871207e-01 6.12041593e-01 4.24411565e-01
-1.55935839e-01 1.18290842e+00 1.98287845e-01 5.29000998e-01
-4.35319811e-01 7.54157379e-02 4.84976321e-02 1.22725821e+00
-6.83001339e-01 9.59231406e-02 -5.56874514e-01 4.26824600e-01
-1.87017190e+00 -1.28736556e+00 -4.24339920e-01 2.51938772e+00
8.33942413e-01 -3.26736271e-01 7.50605106e-01 1.34551346e+00
6.35896564e-01 1.20539531e-01 -8.12692754e-03 -3.60498369e-01
-3.77086848e-01 2.57158518e-01 -2.65036404e-01 -1.21459682e-02
-1.42812431e+00 6.83885396e-01 6.95568419e+00 1.22420049e+00
-1.13340390e+00 3.61805372e-02 -8.48066807e-02 -1.32695004e-01
4.23600972e-01 -1.59821466e-01 -4.02905852e-01 2.50559807e-01
5.34931064e-01 -4.18746116e-04 5.26023209e-01 6.14059925e-01
3.06100212e-02 2.22633984e-02 -1.13928330e+00 1.58967543e+00
6.00481033e-01 -1.17060339e+00 1.29066423e-01 9.12481397e-02
6.37862027e-01 -1.37198329e-01 -5.61263934e-02 3.13634574e-01
-8.44575882e-01 -8.64095211e-01 1.08082020e+00 3.37245733e-01
7.58891463e-01 -7.26957202e-01 6.30063236e-01 2.04991266e-01
-1.27290988e+00 -1.77953780e-01 -2.64968812e-01 -5.05112886e-01
-3.75158675e-02 2.43999094e-01 -1.72960594e-01 6.63691998e-01
1.06370974e+00 9.97490108e-01 -4.59485501e-01 1.44135368e+00
4.11565639e-02 6.06282175e-01 2.62029488e-02 -7.84350634e-02
-1.48842022e-01 -3.22825313e-01 1.02671993e+00 1.38704133e+00
1.93806827e-01 -1.97496802e-01 1.90003291e-01 3.79219621e-01
2.55831361e-01 8.48792374e-01 -3.24199021e-01 -1.78477466e-01
1.71470512e-02 1.27313077e+00 -9.57645833e-01 -1.15903750e-01
-2.47339442e-01 8.67396593e-01 -9.08662975e-02 -1.34388909e-01
-3.06821615e-01 -3.72822285e-01 6.22417212e-01 3.52644056e-01
1.05362944e-02 7.58013502e-02 -5.90353727e-01 -1.06785834e+00
-9.32416692e-02 -1.25505137e+00 9.19560194e-01 -5.50523400e-01
-1.56426466e+00 5.83051085e-01 7.63565376e-02 -1.74467278e+00
-2.38290593e-01 -6.01008832e-01 -1.40925363e-01 2.99231768e-01
-9.19997334e-01 -9.12574232e-01 -5.33982068e-02 2.72390723e-01
9.89358902e-01 -4.83400255e-01 1.39973867e+00 8.14762354e-01
-8.51952806e-02 5.62308133e-01 -6.62246123e-02 -1.33441329e-01
1.20981061e+00 -1.34903538e+00 -2.32317388e-01 -1.53858483e-01
8.84073555e-01 4.90330815e-01 9.44613874e-01 -1.02751069e-01
-1.05994928e+00 -3.71639311e-01 1.29718888e+00 -7.07135141e-01
8.73043239e-01 -6.39927760e-02 -7.11300611e-01 1.63139760e-01
4.02030408e-01 -7.97621310e-01 1.30890584e+00 6.93273008e-01
-5.15380919e-01 9.83916149e-02 -7.41112769e-01 4.46487874e-01
1.13183367e+00 -4.99564946e-01 -8.73882294e-01 4.14069355e-01
-5.45367636e-02 -1.88067541e-01 -9.24563348e-01 3.55964899e-01
1.02434647e+00 -1.00266552e+00 8.67295563e-01 -5.18310666e-01
1.67648166e-01 -3.97035182e-01 -1.77222967e-01 -8.50266099e-01
-4.74351168e-01 -7.45956898e-01 1.69586182e-01 1.22854829e+00
1.64022416e-01 1.99431237e-02 6.33123696e-01 -4.71405119e-01
-1.74382538e-01 -5.60954571e-01 -9.35333252e-01 -7.50267208e-01
9.44786295e-02 -1.02127671e+00 2.85175860e-01 1.23807597e+00
6.36950552e-01 9.45186436e-01 -3.61133307e-01 -5.70856452e-01
4.23958689e-01 5.42209208e-01 8.10608149e-01 -1.82417262e+00
-4.80838388e-01 -1.24957907e+00 -8.59724760e-01 -7.38319576e-01
-1.10580467e-01 -1.04658890e+00 -3.54165435e-01 -1.41801715e+00
3.98880243e-01 -8.94980803e-02 -4.09689099e-01 1.89777061e-01
3.77745897e-01 1.06382334e+00 3.51808339e-01 6.69920802e-01
-7.57418573e-01 -1.08383276e-01 1.02884138e+00 -2.05834419e-01
-2.86281556e-01 4.05787081e-01 -8.09011400e-01 1.00941014e+00
8.35308909e-01 -4.73924994e-01 -2.24349603e-01 3.22404690e-02
7.61012018e-01 -3.58099788e-01 7.26551116e-02 -1.13124299e+00
3.23507078e-02 2.54632592e-01 -1.21774822e-02 -6.57431006e-01
3.97083104e-01 -6.43936217e-01 3.26798946e-01 5.78307621e-02
-3.52695674e-01 4.90191802e-02 8.55646655e-02 3.31938803e-01
-5.73291183e-01 -6.92239523e-01 6.90528750e-01 -2.77378052e-01
-5.94274938e-01 -1.06297515e-01 -4.07216430e-01 1.89810529e-01
5.82210898e-01 -2.81569183e-01 3.40531290e-01 -1.88453749e-01
-1.05871511e+00 -4.31065738e-01 3.31065536e-01 9.39657867e-01
-6.98753521e-02 -1.23808265e+00 -7.00493097e-01 -1.84737861e-01
5.37874639e-01 -7.93988585e-01 4.01692837e-01 9.41784859e-01
-4.10914600e-01 4.86628830e-01 -1.32870600e-01 -6.99415207e-01
-1.79949200e+00 6.83238983e-01 1.33639798e-01 -7.95569643e-02
-7.59381235e-01 4.54222351e-01 -1.97540879e-01 -1.05471529e-01
7.72423804e-01 4.16231304e-02 -9.03542817e-01 5.23654044e-01
8.79127443e-01 7.23797262e-01 2.72364348e-01 -9.59224284e-01
-1.91475481e-01 1.00276268e+00 2.53301352e-01 -2.76798010e-01
1.05049670e+00 -6.42564148e-02 -5.31349361e-01 1.40787625e+00
7.84593701e-01 1.01803318e-01 7.00896755e-02 -2.67437808e-02
3.89718711e-01 -2.78454423e-01 -1.52401567e-01 -9.81158197e-01
-5.91721237e-01 7.82175958e-01 7.29349673e-01 6.81249321e-01
1.17124665e+00 2.20596045e-01 4.97956514e-01 3.49344641e-01
5.53577125e-01 -1.35359502e+00 -2.34094575e-01 5.41460633e-01
1.15307379e+00 -8.39383185e-01 2.79236943e-01 -4.53259587e-01
-3.06514084e-01 1.31553006e+00 -6.53355867e-02 -1.51706263e-01
8.72860849e-01 6.38919026e-02 2.56918341e-01 -2.59542584e-01
-3.47560197e-01 -2.72407293e-01 7.27337837e-01 4.35763478e-01
1.12867451e+00 6.14730045e-02 -1.05148602e+00 9.09075439e-01
-9.87143517e-01 2.47270316e-02 2.07467034e-01 7.00984716e-01
-3.88702869e-01 -1.57114565e+00 -5.39048970e-01 2.42235944e-01
-1.07817495e+00 2.39237055e-01 -8.66307080e-01 8.59961152e-01
5.92373550e-01 1.12459576e+00 -2.09354982e-01 -3.84962708e-01
4.42308158e-01 2.31447637e-01 9.79812622e-01 -6.91162109e-01
-1.04600453e+00 6.12342536e-01 5.34179769e-02 -1.43120483e-01
-9.54641998e-01 -9.41881478e-01 -7.72787631e-01 -1.10544540e-01
-3.43780100e-01 5.81955254e-01 7.79034019e-01 9.35547650e-01
-8.83957520e-02 5.48949361e-01 2.60874927e-01 -9.62264359e-01
-3.17230821e-01 -1.23177671e+00 -1.04904974e+00 4.35385704e-01
-2.16508303e-02 -6.33804560e-01 -4.66179729e-01 2.76621222e-01] | [15.934735298156738, 5.187588214874268] |
5a34e213-c921-4c39-b57d-a7b6a69818ce | investigating-annotator-bias-in-abusive | null | null | https://aclanthology.org/2021.ranlp-main.170 | https://aclanthology.org/2021.ranlp-main.170.pdf | Investigating Annotator Bias in Abusive Language Datasets | Nowadays, social media platforms use classification models to cope with hate speech and abusive language. The problem of these models is their vulnerability to bias. A prevalent form of bias in hate speech and abusive language datasets is annotator bias caused by the annotator’s subjective perception and the complexity of the annotation task. In our paper, we develop a set of methods to measure annotator bias in abusive language datasets and to identify different perspectives on abusive language. We apply these methods to four different abusive language datasets. Our proposed approach supports annotation processes of such datasets and future research addressing different perspectives on the perception of abusive language. | ['Georg Groh', 'Gerhard Hagerer', 'Christian Widmer', 'Maximilian Wich'] | null | null | https://aclanthology.org/2021.ranlp-1.170 | https://aclanthology.org/2021.ranlp-1.170.pdf | ranlp-2021-9 | ['abusive-language'] | ['natural-language-processing'] | [-3.20413053e-01 1.31469414e-01 -4.41576988e-01 -4.15561110e-01
-2.11274400e-02 -1.01708305e+00 5.43963432e-01 2.48397127e-01
-4.86100733e-01 5.82231462e-01 5.58089495e-01 -3.75583358e-02
3.98578942e-01 -2.76648045e-01 1.43801168e-01 -3.87715816e-01
5.04741430e-01 1.41217604e-01 4.86090779e-02 -4.84812915e-01
7.26063251e-01 3.01745743e-01 -8.58962536e-01 5.15165150e-01
6.86938763e-01 3.76795828e-01 -3.10483247e-01 6.60042644e-01
-4.20089364e-01 1.41650164e+00 -1.35665262e+00 -9.21864033e-01
-6.51665479e-02 -3.61791551e-01 -1.03793645e+00 5.55445701e-02
6.26354933e-01 -1.80717707e-01 -4.65341024e-02 1.53195274e+00
8.22693527e-01 -3.05808783e-01 7.14620829e-01 -1.31849325e+00
-1.19674063e+00 1.03082597e+00 -4.09047812e-01 9.44997966e-01
7.15537965e-01 8.95993859e-02 5.96472025e-01 -8.44458818e-01
6.30436480e-01 1.83906329e+00 7.98552394e-01 1.05640483e+00
-1.18820655e+00 -1.03160667e+00 1.71158314e-02 -1.86991602e-01
-1.33498454e+00 -4.95306998e-01 9.45433378e-01 -1.30011940e+00
3.63701224e-01 4.27601546e-01 4.26522553e-01 1.78806841e+00
-3.70448418e-02 3.90717477e-01 1.50068104e+00 -2.95906007e-01
-2.58718263e-02 9.33538377e-01 8.71154964e-01 4.86036777e-01
4.96450067e-01 -4.77593035e-01 -6.79190695e-01 -8.39302063e-01
-1.94409117e-01 -2.20664650e-01 -5.82985021e-02 3.36238354e-01
-5.97827494e-01 1.39815629e+00 2.04084575e-01 8.22158217e-01
1.16321698e-01 -2.18692288e-01 8.55326593e-01 2.18893781e-01
1.05430412e+00 8.16384614e-01 8.69852826e-02 6.28744364e-02
-7.58987665e-01 2.44176149e-01 9.96803641e-01 6.99518085e-01
4.34927762e-01 -3.09814252e-02 -4.00372028e-01 1.16242349e+00
5.16805053e-01 6.83942437e-01 7.57803917e-01 -5.39564967e-01
4.20143098e-01 4.35011148e-01 2.00374410e-01 -1.59557831e+00
-4.22853917e-01 -1.68726429e-01 -3.41663778e-01 -1.36601254e-01
5.35588801e-01 -3.01002473e-01 -3.78647089e-01 1.41794467e+00
2.19797596e-01 -6.28792107e-01 -3.33738804e-01 1.11434472e+00
7.43584573e-01 5.43422699e-01 6.51775479e-01 -2.75861979e-01
1.50514019e+00 -5.78171849e-01 -1.32888424e+00 -2.53032625e-01
1.04196918e+00 -1.03853977e+00 1.32123029e+00 3.40773672e-01
-6.78544998e-01 -1.94491699e-01 -8.14681649e-01 -3.93658221e-01
-5.99897325e-01 -2.87105083e-01 9.34038907e-02 1.25973046e+00
-5.33094227e-01 2.31878325e-01 2.05370292e-01 -4.65240300e-01
4.78341550e-01 -1.92777872e-01 1.77997034e-02 2.56115317e-01
-1.55329394e+00 1.48047936e+00 2.71184653e-01 -2.55780041e-01
-8.93056035e-01 -5.29467762e-01 -6.27250314e-01 -5.47938883e-01
2.65502214e-01 1.31066561e-01 1.14726627e+00 -1.59120822e+00
-9.52158928e-01 1.31552541e+00 9.04277042e-02 -2.38397375e-01
7.02577651e-01 -5.39070547e-01 -5.92570007e-01 -2.55835205e-01
3.08942497e-01 8.75104498e-03 1.12223768e+00 -1.42006850e+00
2.93578561e-02 -6.44131839e-01 2.22865976e-02 -9.30000916e-02
-8.14837456e-01 9.27854538e-01 8.24846983e-01 -9.93203461e-01
-3.48019093e-01 -9.80191886e-01 2.78072000e-01 -8.98272470e-02
-4.63271797e-01 -4.19034272e-01 1.14931023e+00 -9.99083102e-01
1.83176672e+00 -2.36896157e+00 1.37830317e-01 8.04689620e-03
7.17300892e-01 6.93902433e-01 4.75551188e-01 4.81420279e-01
2.84736454e-01 9.31383431e-01 -1.97120085e-01 -3.13422918e-01
7.05987588e-03 4.00143713e-01 -4.57299381e-01 7.76871860e-01
-7.45342225e-02 2.16552094e-01 -1.16149950e+00 -9.52245057e-01
-2.22004116e-01 5.67148685e-01 -1.51234418e-01 4.84322399e-01
2.12006658e-01 4.76195306e-01 -2.46308476e-01 5.67612529e-01
5.59971869e-01 4.59090680e-01 6.58697635e-02 -1.53932208e-02
-2.70341247e-01 4.64755446e-01 -3.97837400e-01 1.01336384e+00
-2.32022047e-01 9.81005251e-01 2.44492009e-01 -2.20378175e-01
1.10819173e+00 3.88205677e-01 -2.48445913e-01 1.04703777e-03
5.33665299e-01 4.32002276e-01 3.64698678e-01 -1.10407245e+00
6.12839222e-01 -3.60414892e-01 -2.00982794e-01 6.54842377e-01
1.06679067e-01 -1.38235241e-01 1.01757288e-01 2.43529350e-01
5.82188427e-01 -4.86336738e-01 6.58908844e-01 -7.27344990e-01
6.85972035e-01 -1.83757514e-01 3.79128307e-01 6.05253816e-01
-1.05479193e+00 -1.75194088e-02 7.51742005e-01 -9.27533388e-01
-1.09795845e+00 -2.47724190e-01 -1.08357064e-01 1.69514883e+00
-2.41373911e-01 -2.14961931e-01 -1.00650692e+00 -1.19999683e+00
1.84256434e-02 1.01948011e+00 -7.92317390e-01 -3.17460388e-01
-2.58104831e-01 -7.01931715e-01 1.20990062e+00 -1.52665526e-02
3.05457652e-01 -9.03298199e-01 -5.67803502e-01 -1.47783294e-01
-1.32317305e-01 -1.12963641e+00 -3.78747761e-01 -1.40044093e-01
-2.46976852e-01 -9.39078629e-01 -3.06074649e-01 -2.34882787e-01
5.87286770e-01 2.43732572e-01 9.74419594e-01 6.62574410e-01
1.20442688e-01 -1.08923256e-01 -7.45758533e-01 -9.11843061e-01
-1.21188772e+00 2.50253141e-01 2.16938570e-01 1.65783152e-01
8.03761065e-01 1.10714354e-01 -1.02074422e-01 5.19075930e-01
-1.09023952e+00 -4.86240447e-01 -1.73357278e-01 3.42896849e-01
-6.00958049e-01 -4.43819135e-01 1.02193475e-01 -1.20759046e+00
1.14822614e+00 -1.05671716e+00 -2.66767684e-02 -9.18114781e-02
-4.83460963e-01 -3.99186045e-01 4.24084216e-01 -8.16572309e-01
-8.93938184e-01 -5.49233198e-01 -8.64508376e-02 -1.29970789e-01
-2.61030555e-01 5.44645963e-03 1.00495063e-01 -1.72456041e-01
1.09804344e+00 -3.46831173e-01 -2.14699116e-02 -5.74717641e-01
-1.30085453e-01 1.20363522e+00 -2.50411153e-01 -4.02473569e-01
6.70837224e-01 3.99815649e-01 -6.14317238e-01 -9.54501152e-01
-1.58603334e+00 -5.36370337e-01 -8.02835345e-01 -6.74771011e-01
1.10958624e+00 -5.79452097e-01 -3.66866589e-01 5.30088782e-01
-1.75161326e+00 -6.81246370e-02 4.95679617e-01 5.25336266e-02
1.94154069e-01 5.78657389e-01 -6.55817330e-01 -1.52559960e+00
-4.33380932e-01 -9.26393688e-01 6.19712710e-01 -2.55457163e-01
-1.15153730e+00 -1.22521448e+00 3.37674409e-01 7.48186469e-01
5.56400716e-01 3.42014790e-01 6.26387954e-01 -1.31013739e+00
7.89431930e-01 -1.65162235e-01 -5.56199765e-03 6.13474965e-01
-2.15538591e-02 3.98797363e-01 -1.22582150e+00 -1.00289471e-01
4.69331264e-01 -6.32969022e-01 3.35008442e-01 -4.75455821e-01
8.14890563e-01 -8.75054538e-01 -9.56117734e-02 -1.93825781e-01
1.07209265e+00 1.39039010e-01 2.82787353e-01 2.69760072e-01
1.06362355e+00 1.09656012e+00 4.84435350e-01 4.94295299e-01
-2.20701080e-02 6.96362615e-01 3.02827954e-01 3.26953381e-01
1.47530541e-01 -3.43194664e-01 5.41369021e-01 8.50890458e-01
-3.95862721e-02 -3.23359877e-01 -1.37823796e+00 3.00229907e-01
-1.55187237e+00 -1.18548906e+00 -4.89210367e-01 1.62830377e+00
7.19586313e-01 -5.84989414e-02 4.04503733e-01 5.37984431e-01
1.05903780e+00 4.87207592e-01 6.41016522e-03 -1.23705661e+00
-6.66956156e-02 -5.99300981e-01 3.66757661e-01 9.21135426e-01
-9.57309127e-01 1.01326382e+00 6.83754444e+00 6.40378833e-01
-1.05171406e+00 8.98582220e-01 6.27087235e-01 -6.08872920e-02
-1.35341242e-01 -3.70791078e-01 -7.96154737e-01 9.73234236e-01
7.46610522e-01 7.93268532e-02 1.11254476e-01 1.10729408e+00
2.86524653e-01 5.99153489e-02 -7.58617640e-01 6.71333492e-01
8.19205999e-01 -4.63864923e-01 7.74723589e-02 5.88196777e-02
3.22887897e-01 -3.34511429e-01 1.64026096e-02 1.74112231e-01
2.90734380e-01 -9.84932780e-01 1.12022805e+00 4.40727957e-02
1.73762932e-01 -4.01115447e-01 9.26480174e-01 4.80312407e-01
6.60062283e-02 -3.47181529e-01 -4.40474838e-01 -6.23932540e-01
9.24523622e-02 4.58185911e-01 -8.94216776e-01 -5.63161552e-01
8.41083288e-01 4.68408465e-01 -8.55081439e-01 3.03320587e-02
-3.45552564e-01 8.75135541e-01 2.48614863e-01 -5.39509416e-01
1.73930496e-01 5.31795286e-02 8.64448369e-01 1.72105289e+00
-2.35051215e-01 -1.33133307e-01 8.77876133e-02 8.06898654e-01
1.24659665e-01 3.19209844e-01 -1.32065046e+00 -3.15695137e-01
7.41549313e-01 1.25687468e+00 -4.97999698e-01 -4.29992318e-01
-2.78719097e-01 8.74178708e-01 6.75327301e-01 3.55370576e-03
-9.07268286e-01 1.80378526e-01 5.91296196e-01 2.89917916e-01
-6.41738653e-01 -1.78157583e-01 -5.64474881e-01 -9.47207332e-01
-4.48039532e-01 -1.33858299e+00 5.20745933e-01 -6.70032442e-01
-1.68876660e+00 6.13194764e-01 6.12204224e-02 -5.83866119e-01
3.23560745e-01 -6.23528421e-01 -1.54467449e-01 6.04551375e-01
-8.31559896e-01 -1.15283251e+00 -2.95439899e-01 2.95247227e-01
6.45060480e-01 -3.18010509e-01 6.50373399e-01 3.88119489e-01
-6.76158488e-01 3.33286762e-01 -8.21561635e-01 5.83959520e-01
9.92595196e-01 -9.60291624e-01 -1.56452864e-01 6.18870497e-01
-1.51858151e-01 6.06653273e-01 1.22603035e+00 -7.61230111e-01
-6.85829937e-01 -6.43277645e-01 1.23535228e+00 -1.23785543e+00
1.27228248e+00 -6.33861542e-01 -1.03525341e+00 6.39260650e-01
5.29276967e-01 -1.95178390e-01 1.51770973e+00 6.73462031e-03
-9.67728257e-01 4.83738124e-01 -1.46721625e+00 3.46275866e-01
1.10048056e+00 -6.87121391e-01 -9.02807713e-01 5.39597392e-01
4.83682960e-01 -1.53191343e-01 -6.37865245e-01 -2.60481030e-01
4.99451101e-01 -9.83406305e-01 2.84698904e-01 -1.04105890e+00
7.19058692e-01 -1.17510138e-02 -1.21036164e-01 -1.38293672e+00
-4.32509899e-01 -5.76635480e-01 -3.85770909e-02 1.45067298e+00
2.19828814e-01 -5.09239674e-01 -1.55097991e-01 8.27070534e-01
2.57248491e-01 -2.33504511e-02 -7.47026920e-01 -4.70173806e-01
3.67781818e-01 -2.70839810e-01 3.34636539e-01 1.84421313e+00
1.89274877e-01 5.77155054e-01 -6.89822793e-01 7.88312629e-02
3.91592592e-01 -7.61338234e-01 7.68635631e-01 -1.30019295e+00
3.32224756e-01 -4.19899791e-01 -4.58281517e-01 -4.79919277e-02
6.15746081e-01 -5.75176239e-01 -2.89474428e-01 -5.94046175e-01
3.55114728e-01 -8.45715329e-02 1.69053108e-01 1.09037451e-01
-3.30398321e-01 6.83319330e-01 6.10450745e-01 4.67276305e-01
-5.24016201e-01 -1.00295551e-01 8.84347320e-01 -1.69543549e-01
6.30783290e-03 -4.97046143e-01 -6.58623219e-01 1.13418686e+00
9.22467470e-01 -1.05149937e+00 -6.14693165e-02 -5.40830970e-01
7.84701884e-01 -7.42501259e-01 3.99881005e-01 -5.83934128e-01
-3.54924738e-01 -3.03156406e-01 -1.21559173e-01 4.21547592e-02
1.02085784e-01 -9.16641116e-01 -1.80140376e-01 7.60133564e-01
-7.79875875e-01 3.31878096e-01 -2.10188359e-01 2.32094184e-01
-1.08801713e-02 -8.78329098e-01 1.12295711e+00 -2.71357447e-01
-5.79689480e-02 -2.73263693e-01 -1.10854411e+00 4.04056430e-01
1.17911911e+00 1.48717403e-01 -6.62284195e-01 -3.85587752e-01
-6.03258252e-01 -2.40882978e-01 6.49153054e-01 6.28365517e-01
7.88879842e-02 -1.27424479e+00 -8.83723080e-01 -3.12048972e-01
2.89441884e-01 -8.77159715e-01 -3.99355352e-01 7.78231978e-01
-5.60702026e-01 1.01285011e-01 -3.43764842e-01 -1.16126947e-01
-1.59051907e+00 8.51544201e-01 4.57796842e-01 2.05806464e-01
2.04138070e-01 5.90259671e-01 -1.32724851e-01 -2.52276510e-01
3.29663302e-03 3.51818293e-01 -6.31436348e-01 5.88615060e-01
7.19240546e-01 9.11392689e-01 -5.66334248e-01 -1.62505448e+00
-3.83318424e-01 1.42737105e-01 -1.55476421e-01 -1.04115158e-02
5.52723169e-01 -3.28724533e-01 -6.57738805e-01 1.13729131e+00
1.10940862e+00 5.26135445e-01 -2.23328590e-01 1.70071442e-02
5.24370968e-01 -8.63341928e-01 -2.17807144e-01 -7.55740345e-01
-4.90564704e-01 7.75729239e-01 3.58974338e-01 1.05357754e+00
1.35734141e-01 -8.18064138e-02 3.70039046e-01 1.42322943e-01
1.09588899e-01 -1.49691010e+00 2.59196371e-01 4.93073314e-01
1.25790286e+00 -1.55403948e+00 1.37435481e-01 -7.73750365e-01
-9.08273637e-01 9.75548923e-01 8.94497991e-01 -5.46091571e-02
6.35377884e-01 2.03183845e-01 9.04586673e-01 -6.10406339e-01
-2.96284705e-01 2.40057021e-01 3.23546439e-01 6.46524489e-01
1.26286674e+00 1.03273682e-01 -1.20514405e+00 4.59149688e-01
-2.92968839e-01 -4.85559523e-01 7.23743796e-01 7.07000017e-01
-3.13991070e-01 -5.74879467e-01 -8.70046079e-01 -7.73656070e-02
-1.05812001e+00 1.43339872e-01 -1.77329159e+00 6.41285539e-01
4.25953597e-01 1.25378072e+00 -1.38324007e-01 -4.83927697e-01
6.85270652e-02 4.30191278e-01 -8.72263312e-02 -8.55441034e-01
-1.26805496e+00 -5.08027136e-01 6.69082463e-01 -2.66502947e-01
-7.52555311e-01 -4.16473031e-01 -5.79052627e-01 -5.68012595e-01
-4.07907903e-01 1.79842293e-01 6.82468414e-01 1.05313456e+00
8.25249031e-02 -2.94794559e-01 5.81958830e-01 -5.13277352e-01
-5.83402991e-01 -1.59023190e+00 -3.49986732e-01 1.13197601e+00
3.97594571e-01 -8.35608304e-01 -7.34116197e-01 -1.59706727e-01] | [8.693800926208496, 10.533914566040039] |
7202338e-59a0-4425-ba0f-68ca34dd97ae | memseg-a-semi-supervised-method-for-image | 2205.00908 | null | https://arxiv.org/abs/2205.00908v1 | https://arxiv.org/pdf/2205.00908v1.pdf | MemSeg: A semi-supervised method for image surface defect detection using differences and commonalities | Under the semi-supervised framework, we propose an end-to-end memory-based segmentation network (MemSeg) to detect surface defects on industrial products. Considering the small intra-class variance of products in the same production line, from the perspective of differences and commonalities, MemSeg introduces artificially simulated abnormal samples and memory samples to assist the learning of the network. In the training phase, MemSeg explicitly learns the potential differences between normal and simulated abnormal images to obtain a robust classification hyperplane. At the same time, inspired by the mechanism of human memory, MemSeg uses a memory pool to store the general patterns of normal samples. By comparing the similarities and differences between input samples and memory samples in the memory pool to give effective guesses about abnormal regions; In the inference phase, MemSeg directly determines the abnormal regions of the input image in an end-to-end manner. Through experimental validation, MemSeg achieves the state-of-the-art (SOTA) performance on MVTec AD datasets with AUC scores of 99.56% and 98.84% at the image-level and pixel-level, respectively. In addition, MemSeg also has a significant advantage in inference speed benefiting from the end-to-end and straightforward network structure, which better meets the real-time requirement in industrial scenarios. | ['Hui Feng', 'Jing Liu', 'Peng Wu', 'Minghui Yang'] | 2022-05-02 | null | null | null | null | ['defect-detection'] | ['computer-vision'] | [ 3.08849633e-01 2.81077594e-01 -2.69359887e-01 -4.56368059e-01
-5.90303421e-01 -1.17994696e-01 -1.32105961e-01 1.04323275e-01
4.76858206e-02 2.39172697e-01 -7.63802409e-01 -1.94742709e-01
-2.82773703e-01 -9.71651673e-01 -8.43151927e-01 -8.44206631e-01
1.36436656e-01 5.40065110e-01 2.96293586e-01 2.95310020e-01
3.80535752e-01 3.93935144e-01 -1.45568979e+00 6.61288857e-01
1.08577788e+00 1.65079534e+00 3.03570360e-01 1.47762924e-01
4.75392342e-02 3.56053382e-01 -5.87171078e-01 -1.29413784e-01
3.33460331e-01 -1.02658533e-01 -4.31125432e-01 5.89280844e-01
5.25890529e-01 -1.83209375e-01 -3.47124517e-01 1.25480032e+00
3.33014935e-01 8.85251015e-02 4.88285929e-01 -1.02335799e+00
-4.81560320e-01 2.33397931e-01 -6.72168016e-01 -7.88955539e-02
-4.38551344e-02 5.61290622e-01 7.22446144e-01 -9.36004698e-01
5.61444044e-01 1.05440724e+00 4.59424078e-01 1.83611736e-01
-8.41298819e-01 -5.88876247e-01 3.02128077e-01 2.84435570e-01
-1.29929268e+00 8.98929760e-02 9.50455904e-01 -3.13818306e-01
7.51205146e-01 4.95792888e-02 5.75360894e-01 6.52558148e-01
5.51146686e-01 1.16051674e+00 8.03260446e-01 -1.02230117e-01
4.14419025e-01 1.60087183e-01 3.94991845e-01 8.05445790e-01
2.46410936e-01 5.05977273e-02 -4.94349241e-01 2.43362963e-01
1.02345407e+00 2.77264267e-01 -1.24432079e-01 -3.87118578e-01
-7.28382111e-01 6.07439756e-01 4.13618565e-01 -7.71948835e-03
-4.43734169e-01 -5.06569028e-01 4.53005731e-01 1.36195481e-01
3.19578439e-01 2.54008055e-01 -5.05824625e-01 2.60826677e-01
-9.25157130e-01 -4.59672604e-03 5.21909177e-01 8.95226240e-01
5.52960277e-01 5.30986004e-02 -7.28351772e-02 9.14378822e-01
3.74002427e-01 5.03289223e-01 4.98972654e-01 -7.49255419e-01
6.63407803e-01 1.01241064e+00 -2.46550664e-01 -1.14700770e+00
-3.77529263e-01 -9.86224413e-01 -1.01039100e+00 4.20259476e-01
4.51799095e-01 1.26620889e-01 -1.18251419e+00 1.18610311e+00
1.84598893e-01 1.83839485e-01 -1.65901497e-01 8.57383847e-01
4.03784573e-01 7.33503163e-01 -3.21957141e-01 -3.37291777e-01
1.26746035e+00 -1.01366389e+00 -5.57654381e-01 -7.01255858e-01
4.75301564e-01 -4.63207513e-01 1.13937283e+00 8.80631387e-01
-1.26351857e+00 -7.80433834e-01 -1.38938892e+00 4.65856642e-01
-2.82092482e-01 5.81818640e-01 2.84080893e-01 3.46796453e-01
-4.65873897e-01 7.39612162e-01 -9.43722785e-01 1.85329378e-01
6.83534086e-01 5.09771883e-01 8.38037133e-02 -1.05456114e-01
-7.65196145e-01 5.21357179e-01 4.39937830e-01 6.24880910e-01
-8.97025645e-01 -6.82122171e-01 -7.77559400e-01 -1.31975710e-01
6.41516507e-01 -3.06110650e-01 1.04332793e+00 -1.01470494e+00
-1.42429566e+00 6.40729070e-01 3.11291888e-02 -4.51631069e-01
6.17665946e-01 -1.85442775e-01 -4.67903078e-01 2.36893132e-01
-1.60106242e-01 5.47443211e-01 8.08956861e-01 -1.18971729e+00
-7.69014299e-01 -7.75124490e-01 -3.86960536e-01 -7.37537444e-02
-2.23269194e-01 -5.56664765e-01 -6.79618537e-01 -5.97681582e-01
5.51421225e-01 -8.43786001e-01 -1.34873629e-01 1.67385191e-01
-7.31915712e-01 -9.94208604e-02 9.49693382e-01 -8.04692984e-01
1.09623599e+00 -2.41276908e+00 -2.67703205e-01 7.24224269e-01
3.26519199e-02 4.26611304e-01 7.92014040e-03 -3.58711571e-01
-1.70275241e-01 -5.48391759e-01 -4.73175764e-01 -1.27548650e-01
-8.59720483e-02 -2.13216506e-02 -8.72255787e-02 4.10320312e-01
3.58222663e-01 7.74681032e-01 -6.60290182e-01 -5.04266441e-01
4.03380275e-01 -3.18965651e-02 -2.42397591e-01 2.92891830e-01
-2.52073407e-01 2.37656459e-01 -5.00632823e-01 9.54027474e-01
8.74876201e-01 -2.56302357e-01 1.36911720e-01 -3.37613285e-01
3.49229187e-01 -1.18458055e-01 -1.26612675e+00 1.49780452e+00
-4.91305918e-01 4.20241594e-01 7.80704096e-02 -1.13729453e+00
1.24282634e+00 1.16499819e-01 1.27348840e-01 -1.01667213e+00
2.61058033e-01 2.98487931e-01 -8.42159092e-02 -5.40342510e-01
1.54874995e-01 1.89248547e-01 1.56771258e-01 1.16767317e-01
3.62488329e-02 1.79468393e-01 -6.70925807e-03 -4.52038616e-01
7.17239499e-01 2.08668839e-02 -7.48800561e-02 -1.20906144e-01
5.25643528e-01 -1.82044476e-01 8.37648571e-01 4.68094945e-01
-4.35096845e-02 7.22335935e-01 3.35928172e-01 -4.51311737e-01
-7.60257840e-01 -1.41781104e+00 -2.07346737e-01 3.28877330e-01
4.92690742e-01 2.23611102e-01 -8.58016849e-01 -9.00815606e-01
2.66281981e-02 8.23792338e-01 -4.41247731e-01 -4.91074055e-01
-5.38254380e-01 -6.88164234e-01 8.88866559e-02 9.00764942e-01
8.57716858e-01 -1.18308187e+00 -6.74825370e-01 5.52178621e-02
2.37948135e-01 -1.07903767e+00 -2.92120486e-01 1.82895258e-01
-1.29911542e+00 -1.08028579e+00 -5.08952260e-01 -1.07406199e+00
1.12450457e+00 -1.25697196e-01 8.81402910e-01 -1.56208441e-01
-5.04875720e-01 -1.99280366e-01 2.38414556e-01 -2.41398513e-01
-2.86134452e-01 -1.05745554e-01 -2.29817793e-01 1.31291494e-01
2.82032162e-01 -4.52673942e-01 -7.72908092e-01 6.28316998e-01
-8.78674150e-01 -5.49548864e-03 9.97086048e-01 8.60103667e-01
9.29827452e-01 6.16113007e-01 6.59828782e-01 -8.32615912e-01
2.87731051e-01 -2.91435510e-01 -8.23260903e-01 2.70325571e-01
-9.72542286e-01 -3.41505200e-01 6.07970059e-01 -5.65385520e-01
-1.16330612e+00 1.67927116e-01 -7.86686912e-02 -6.19034231e-01
-3.14270049e-01 4.03081298e-01 -5.24717510e-01 3.47969800e-01
3.60207230e-01 2.83189744e-01 3.05887848e-01 -4.01447088e-01
3.84089760e-02 8.12590778e-01 8.17672729e-01 -2.94130057e-01
4.84740973e-01 2.78844774e-01 -2.11827040e-01 -6.30012155e-01
-7.51193464e-01 -2.90355504e-01 -5.49461424e-01 -3.08055580e-01
6.86074555e-01 -7.67173827e-01 -7.79655218e-01 8.51016462e-01
-8.18634510e-01 -2.83989161e-01 -4.07367378e-01 3.69322419e-01
-5.47632217e-01 2.80943573e-01 -9.17523324e-01 -8.67447317e-01
-4.87262726e-01 -1.21719086e+00 9.75662470e-01 3.66869479e-01
-1.47609428e-01 -6.91065490e-01 -8.07053089e-01 5.90796709e-01
-1.04906455e-01 2.16974735e-01 1.30266905e+00 -5.32735288e-01
-9.12662089e-01 -5.84639668e-01 -2.15097740e-01 7.99232960e-01
9.45776701e-02 -1.43339410e-01 -8.99100602e-01 -1.63058251e-01
2.52708614e-01 6.03390895e-02 7.09266424e-01 7.53967285e-01
1.37957489e+00 -2.46849172e-02 -3.74132484e-01 2.73945957e-01
1.18862927e+00 7.89851725e-01 9.08125639e-01 2.26125523e-01
5.33583343e-01 7.50593960e-01 1.20595860e+00 2.91678458e-01
-1.30675539e-01 4.57357079e-01 4.77986693e-01 -2.29046926e-01
-2.32483596e-02 -1.77612320e-01 3.47923666e-01 6.65819108e-01
6.21871710e-01 -2.19745219e-01 -5.15359759e-01 6.40647531e-01
-1.85956275e+00 -4.92853433e-01 5.20108081e-02 2.34278440e+00
5.89384258e-01 8.87018800e-01 -6.50128648e-02 2.42943838e-01
8.87527645e-01 -2.57442985e-02 -1.05546248e+00 -2.00088128e-01
-4.32823002e-02 8.39904025e-02 4.14693922e-01 -6.41832724e-02
-9.99229074e-01 5.03436744e-01 5.32739544e+00 1.28134561e+00
-1.08145356e+00 -2.35538960e-01 1.27471387e+00 1.64723564e-02
2.60687143e-01 -5.01645923e-01 -8.21341693e-01 6.21855021e-01
6.88494086e-01 4.85402972e-01 1.50945738e-01 1.06012583e+00
1.06780387e-01 -2.75729716e-01 -1.09641051e+00 8.79204571e-01
3.26936841e-02 -1.17935479e+00 -1.29860267e-01 -6.93484768e-02
7.18868077e-01 -3.61552417e-01 5.17925918e-01 7.59368949e-03
-4.12642777e-01 -7.39769995e-01 7.90920258e-01 4.02966410e-01
5.26842177e-01 -9.95422006e-01 1.13395596e+00 4.04226959e-01
-9.08924758e-01 -3.75339180e-01 -2.72998005e-01 1.79098949e-01
1.33259580e-01 1.13817811e+00 -7.95883298e-01 5.87583840e-01
5.00714004e-01 5.12692571e-01 -3.84768516e-01 8.07848454e-01
-1.11918762e-01 5.94367981e-01 -1.55113846e-01 2.24071071e-01
1.11066975e-01 -3.08132738e-01 2.71409869e-01 8.90794158e-01
3.34850460e-01 -3.53605002e-01 1.59652695e-01 1.26993251e+00
2.18287766e-01 -2.11841047e-01 -2.31132910e-01 -3.10375486e-02
3.85368049e-01 1.14260793e+00 -1.04078388e+00 -2.43170291e-01
-7.56527930e-02 8.10432434e-01 -1.84525385e-01 2.05308065e-01
-8.18991363e-01 -6.01185143e-01 3.62294555e-01 2.07040042e-01
5.31071424e-01 1.55182155e-02 -8.95475924e-01 -4.35184121e-01
4.17022616e-01 -8.53607178e-01 3.64847958e-01 -5.90081990e-01
-1.40222931e+00 5.09678304e-01 -1.78136155e-01 -1.58949184e+00
-1.35364652e-01 -1.05529356e+00 -7.12531269e-01 6.21547699e-01
-1.17374277e+00 -9.33097303e-01 -4.13582534e-01 1.92735136e-01
8.45781803e-01 -2.30170116e-01 3.69911820e-01 2.37038136e-01
-1.02427673e+00 8.42502058e-01 3.58296954e-03 8.74216855e-02
3.80415529e-01 -1.02158928e+00 4.38994288e-01 8.29871476e-01
-3.35840344e-01 5.28597593e-01 4.33470517e-01 -9.23924983e-01
-1.21452153e+00 -1.30164790e+00 1.85980409e-01 -4.61461954e-02
2.87589073e-01 -2.26086050e-01 -1.13241804e+00 3.63705128e-01
-2.20176071e-01 -4.09540832e-02 3.85017723e-01 -1.36846200e-01
1.10647283e-01 -3.08870584e-01 -1.37789762e+00 4.34218407e-01
8.77991855e-01 -3.04276556e-01 -4.00125384e-01 2.54729480e-01
5.12512565e-01 -7.30471194e-01 -9.40978646e-01 6.59145415e-01
4.77099031e-01 -8.09887886e-01 7.62870133e-01 -6.71319803e-03
6.47829294e-01 -4.31715488e-01 1.44937515e-01 -1.06572509e+00
-2.21439507e-02 -1.74611345e-01 -2.49740824e-01 1.24362016e+00
6.99589729e-01 -8.32686484e-01 1.32313704e+00 5.29078543e-01
-4.49142396e-01 -1.41363752e+00 -6.75411880e-01 -9.14832950e-01
-2.56767094e-01 -4.58696127e-01 5.34135222e-01 4.38063323e-01
-2.32713535e-01 -7.00700134e-02 7.00798482e-02 5.39443195e-01
5.05466640e-01 4.30083632e-01 2.96395093e-01 -1.18415964e+00
-3.65805566e-01 -3.02625120e-01 -5.00439942e-01 -1.20601463e+00
7.23562092e-02 -6.05972409e-01 3.26710284e-01 -1.23729324e+00
6.74308930e-03 -3.38960975e-01 -4.69758153e-01 1.93758965e-01
-3.59873697e-02 1.73782587e-01 -1.17121518e-01 -3.57372612e-02
-4.42582339e-01 3.11613172e-01 1.44836295e+00 -3.98790747e-01
-3.95486474e-01 4.47765529e-01 -4.63781029e-01 7.54736245e-01
7.13367283e-01 -1.30872041e-01 -5.82976282e-01 -1.33937746e-01
-1.76250517e-01 6.50826022e-02 2.86109626e-01 -1.00884330e+00
1.49986729e-01 1.25632137e-01 7.74526894e-01 -1.03368974e+00
2.41898835e-01 -7.65412986e-01 -1.00005746e-01 4.95101422e-01
-1.55262083e-01 -9.51756760e-02 2.70157874e-01 5.57098329e-01
-4.02736992e-01 -3.75376165e-01 7.80773401e-01 1.28121544e-02
-6.81499481e-01 2.30674744e-01 -8.47675279e-02 -2.57281333e-01
1.13396204e+00 -7.93139994e-01 -2.28745326e-01 -3.11059915e-02
-5.98483086e-01 3.41178536e-01 2.72492498e-01 4.20580059e-01
8.45403194e-01 -1.08032715e+00 -1.09102763e-01 7.13110924e-01
3.72382291e-02 4.76130038e-01 7.51481175e-01 8.38347673e-01
-3.95578027e-01 2.57519633e-01 -1.14780217e-01 -8.92059386e-01
-9.56878901e-01 4.18844491e-01 3.71734172e-01 -2.62465477e-01
-5.55472434e-01 7.37319648e-01 3.98026168e-01 -3.23355287e-01
4.38135982e-01 -6.09879315e-01 1.39738038e-01 -1.40393734e-01
4.97200817e-01 5.99117160e-01 4.90541369e-01 -1.11442588e-01
-1.10340320e-01 6.08223796e-01 -2.96316177e-01 3.49039078e-01
1.09788394e+00 1.34990066e-01 1.08408883e-01 4.76144046e-01
1.11235678e+00 -4.09004390e-01 -1.57300365e+00 -1.45853221e-01
6.58103898e-02 -3.79585564e-01 1.59817517e-01 -1.11683273e+00
-1.70224726e+00 9.21746552e-01 1.19265473e+00 -2.18751490e-01
1.38389993e+00 -1.82711214e-01 1.28928614e+00 2.10344255e-01
1.94175929e-01 -1.48330927e+00 2.38561437e-01 9.16616842e-02
4.97432083e-01 -1.05102086e+00 -1.03205167e-01 -8.16482306e-01
-6.71809852e-01 1.17399573e+00 9.19146419e-01 -1.43141791e-01
4.90791082e-01 3.06443900e-01 1.40900493e-01 -3.95231724e-01
-3.84154469e-01 5.45779228e-01 4.15057570e-01 5.77824414e-01
-1.62060603e-01 2.61689015e-02 1.07841529e-01 8.27703893e-01
1.33348122e-01 -5.16949408e-02 -3.92738683e-03 1.02201533e+00
-5.12165725e-01 -6.83949232e-01 -2.67836094e-01 7.16537356e-01
-1.63419425e-01 3.84562612e-01 -8.28336254e-02 6.30520523e-01
3.22870970e-01 9.68447983e-01 3.46128076e-01 -4.83835012e-01
5.83124757e-01 6.09634491e-03 3.60466301e-01 -4.25036132e-01
-4.41625625e-01 1.43427819e-01 -1.73271835e-01 -7.12052822e-01
3.36747259e-01 -5.38248062e-01 -1.74504709e+00 1.58607423e-01
-6.39205277e-01 -2.07212847e-02 5.49528003e-01 1.01043832e+00
4.91953373e-01 8.96995783e-01 7.57973969e-01 -7.69626319e-01
-7.85766244e-01 -9.33653533e-01 -8.93713772e-01 3.58760059e-01
-1.80774760e-02 -6.49857879e-01 -7.08102882e-02 -7.00943992e-02] | [7.50601053237915, 1.968276023864746] |
ee56a6c7-b98a-479b-ae33-c0d5ae46ac80 | a-white-box-analysis-of-colbert | 2012.09650 | null | https://arxiv.org/abs/2012.09650v1 | https://arxiv.org/pdf/2012.09650v1.pdf | A White Box Analysis of ColBERT | Transformer-based models are nowadays state-of-the-art in ad-hoc Information Retrieval, but their behavior is far from being understood. Recent work has claimed that BERT does not satisfy the classical IR axioms. However, we propose to dissect the matching process of ColBERT, through the analysis of term importance and exact/soft matching patterns. Even if the traditional axioms are not formally verified, our analysis reveals that ColBERT: (i) is able to capture a notion of term importance; (ii) relies on exact matches for important terms. | ['Stéphane Clinchant', 'Benjamin Piwowarski', 'Thibault Formal'] | 2020-12-17 | null | null | null | null | ['ad-hoc-information-retrieval'] | ['natural-language-processing'] | [ 1.08325869e-01 1.01346135e-01 -5.25595188e-01 -1.19710386e-01
-6.64490640e-01 -7.08279550e-01 1.14538682e+00 7.43551135e-01
-2.99740255e-01 4.73925978e-01 1.78276330e-01 -3.34118932e-01
-9.92664516e-01 -8.03995430e-01 -3.94847631e-01 -4.10944909e-01
-1.42600104e-01 1.15507138e+00 5.40406704e-01 -8.74905825e-01
4.17398095e-01 4.68004256e-01 -1.86400473e+00 3.76972228e-01
6.55428767e-01 1.20644140e+00 -2.64695942e-01 2.58125275e-01
-2.74400324e-01 1.04247594e+00 -3.95756215e-01 -5.21516025e-01
2.62523234e-01 -2.07515955e-01 -1.40506446e+00 -3.78081203e-01
3.14048231e-01 2.44027823e-02 -1.42063305e-01 1.25651419e+00
-8.47689509e-02 -8.63255858e-02 7.60034084e-01 -1.27695858e+00
-7.80839205e-01 7.94861376e-01 -2.28931475e-02 1.37046933e-01
6.30831242e-01 -5.48602998e-01 1.71371591e+00 -6.00451052e-01
9.40330148e-01 1.05165136e+00 5.54583371e-01 1.26816764e-01
-1.08191800e+00 -1.75403506e-01 4.38612737e-02 6.78547978e-01
-1.57873428e+00 -1.32644489e-01 6.09614253e-01 -2.43794158e-01
6.63740873e-01 7.37247884e-01 7.32552290e-01 6.11232638e-01
7.42179006e-02 6.54823482e-01 1.42717457e+00 -8.78003895e-01
7.23688826e-02 4.93529409e-01 4.80441362e-01 4.39500183e-01
5.18023908e-01 2.10812733e-01 -6.11377835e-01 -3.84006649e-01
1.78350672e-01 -1.43303156e-01 -1.26671299e-01 -4.98020172e-01
-7.36884892e-01 7.78165996e-01 2.59463668e-01 1.08531702e+00
-5.15138924e-01 -1.85336724e-01 2.55365849e-01 7.99444795e-01
3.83339405e-01 7.54177153e-01 -4.09913361e-01 4.51580584e-02
-1.12493932e+00 6.77533567e-01 8.25471163e-01 1.03499079e+00
7.34640956e-01 -6.25856876e-01 -3.12596112e-01 5.01071274e-01
4.97148156e-01 2.23857582e-01 3.88449878e-01 -7.31920004e-01
-2.02107385e-01 1.05901110e+00 4.03668731e-02 -1.03235841e+00
-1.15967892e-01 -4.46952343e-01 -4.77285802e-01 -1.14637926e-01
2.39428669e-01 8.30061495e-01 -5.08239329e-01 1.57596886e+00
-2.29335041e-03 -5.77001095e-01 -2.20195904e-01 7.84329832e-01
4.38154936e-01 7.55565334e-03 2.69569140e-02 -3.39391887e-01
1.45371675e+00 -4.89637464e-01 -8.46985936e-01 -6.53790385e-02
5.21493852e-01 -8.24382007e-01 7.93740034e-01 4.67646182e-01
-1.22010016e+00 -7.42701963e-02 -9.59596932e-01 -3.39182019e-01
-6.64269805e-01 -4.72762644e-01 8.69637251e-01 6.71057045e-01
-1.21066618e+00 5.87680101e-01 -2.23247275e-01 -3.47600341e-01
-1.60932004e-01 2.55572587e-01 -2.54719347e-01 -1.82775900e-01
-1.66866601e+00 1.42418039e+00 2.85191894e-01 -2.67390847e-01
-7.17512131e-01 -5.06583035e-01 -5.20888269e-01 3.40646714e-01
6.65725052e-01 -7.53760934e-01 1.31480742e+00 -8.60696554e-01
-9.61688280e-01 1.06810856e+00 -1.68729812e-01 -4.75126028e-01
4.11104590e-01 -1.62320256e-01 -6.27976060e-01 3.86372954e-01
-1.23957835e-01 1.77368909e-01 5.31187713e-01 -1.46002460e+00
-5.67487717e-01 -5.53433895e-01 6.15588486e-01 -2.84389015e-02
-4.41711366e-01 3.06913197e-01 -2.54020423e-01 -5.16132772e-01
2.65506715e-01 -7.29389787e-01 5.49655929e-02 8.99353847e-02
-2.82550216e-01 -5.59537649e-01 3.42185557e-01 -1.89532980e-01
1.60219538e+00 -1.61323035e+00 3.34282704e-02 7.62374878e-01
4.78738368e-01 2.84294933e-01 8.94255042e-02 9.96958017e-01
-5.67440167e-02 2.37018943e-01 -5.38338311e-02 -1.45737037e-01
5.02231300e-01 2.08734512e-01 -5.14465630e-01 2.88881749e-01
-2.65631527e-01 8.97591352e-01 -9.31719363e-01 -7.49724746e-01
8.99853185e-02 3.13480705e-01 -4.93627638e-01 -1.29255816e-01
-4.30965006e-01 -4.21417177e-01 -5.10016859e-01 8.15344453e-01
3.85459960e-01 -1.81765720e-01 5.47204494e-01 -2.58167654e-01
-1.76576704e-01 6.23140991e-01 -9.24170136e-01 1.33626819e+00
-4.71500270e-02 4.59535778e-01 -8.21808428e-02 -9.69863951e-01
6.93948925e-01 4.94167566e-01 8.09241295e-01 -9.92862582e-01
2.46168271e-01 4.74070132e-01 -1.13761604e-01 -3.29324335e-01
9.87232506e-01 -3.65868896e-01 -4.52731475e-02 4.41421986e-01
-9.40408111e-02 -1.10093668e-01 6.70420885e-01 3.69645923e-01
1.15000534e+00 -1.89177915e-01 4.64532435e-01 -7.05841959e-01
6.77851200e-01 2.95180291e-01 3.86434421e-02 7.90351272e-01
2.38480151e-01 2.37459943e-01 3.51375222e-01 -1.67641848e-01
-7.99717784e-01 -9.58059132e-01 -3.70825559e-01 9.93102968e-01
2.35708162e-01 -9.58974063e-01 -5.42346299e-01 -4.67121452e-01
1.04542531e-01 7.23781645e-01 -6.88114583e-01 -2.00868502e-01
-1.47167966e-01 -3.15605462e-01 5.58105111e-01 -2.08204612e-02
-8.49430077e-03 -6.53510749e-01 -4.31867808e-01 5.40764406e-02
-2.35479623e-01 -8.69994998e-01 -2.03334168e-01 1.84326097e-01
-7.13395119e-01 -1.39371026e+00 -4.36105847e-01 -2.67803580e-01
5.70722044e-01 3.10091168e-01 1.53711474e+00 6.04104638e-01
-9.94335264e-02 6.31215930e-01 -7.23750591e-01 -4.70606923e-01
-4.32059675e-01 1.22065060e-02 3.42331529e-02 -2.55322665e-01
8.05544615e-01 -5.98073363e-01 -4.49930340e-01 4.52973276e-01
-1.32682407e+00 -6.18827879e-01 7.22035885e-01 6.11984432e-01
4.11619604e-01 2.31596455e-01 1.22102238e-01 -8.98554444e-01
8.14310253e-01 -3.20748985e-01 -6.65176749e-01 6.71875656e-01
-1.57771158e+00 5.51676810e-01 1.67918652e-01 -2.62793183e-01
-6.90046668e-01 -5.66855788e-01 -1.39111979e-02 -1.85314670e-01
1.91068798e-01 9.19631064e-01 1.54188067e-01 -1.89554691e-01
6.49684012e-01 1.04809165e-01 -3.61973256e-01 -6.13292813e-01
2.67772138e-01 5.04926980e-01 2.60096699e-01 -8.91220510e-01
1.01815295e+00 5.08892477e-01 2.06206083e-01 -4.90866423e-01
-8.22660327e-01 -8.88283134e-01 -3.92399877e-01 -1.84008643e-01
3.85836244e-01 -4.45725501e-01 -9.39243257e-01 -4.47598472e-02
-9.88960266e-01 3.46375316e-01 -4.75899279e-01 1.87285691e-01
-3.55892867e-01 5.36805332e-01 -5.18779159e-01 -1.05160594e+00
-2.56601036e-01 -7.43337095e-01 9.17875648e-01 -3.14974964e-01
-5.19771039e-01 -9.93664861e-01 3.57220471e-01 3.26909155e-01
5.51551998e-01 -3.01657140e-01 1.29270506e+00 -8.91401172e-01
-5.78294873e-01 -6.27571702e-01 -5.15847206e-02 3.07198502e-02
-3.49970832e-02 -1.52840912e-01 -8.63615811e-01 -1.51445530e-02
7.02309087e-02 -1.49342082e-02 6.53740466e-01 5.27219400e-02
7.56287336e-01 -5.15355229e-01 -2.88728744e-01 -1.83862060e-01
1.53615820e+00 1.93284955e-02 8.94313037e-01 5.58229148e-01
-1.16454691e-01 7.47481763e-01 6.14571929e-01 2.09434807e-01
4.71449524e-01 1.15129316e+00 4.53208953e-01 1.95865497e-01
1.89590320e-01 -4.69583362e-01 3.16751264e-02 6.79893553e-01
-2.79704601e-01 -4.38319258e-02 -7.36186147e-01 6.28406882e-01
-1.89017344e+00 -1.13516152e+00 -3.17802280e-01 2.29718041e+00
1.07593238e+00 3.33587378e-01 5.14128059e-02 5.26398122e-01
2.86181837e-01 3.32112238e-03 2.42248833e-01 -2.95013249e-01
-3.18427891e-01 3.77456307e-01 4.31128711e-01 5.07214129e-01
-5.45153975e-01 6.41992509e-01 7.51753616e+00 7.41481066e-01
-6.50030315e-01 1.79986775e-01 -9.68748629e-02 2.95626611e-01
-9.71097410e-01 5.16836047e-01 -5.35345316e-01 6.69407770e-02
5.62421799e-01 -5.10043859e-01 3.74128878e-01 7.65903294e-01
-3.72838795e-01 -3.21524218e-02 -1.28495526e+00 4.21591848e-01
2.95376360e-01 -1.08897877e+00 4.59056318e-01 2.54764855e-01
5.06685019e-01 -4.73669440e-01 -2.25650623e-01 3.44757825e-01
3.50870490e-01 -1.06166708e+00 9.69816387e-01 6.85478508e-01
2.01583207e-01 -3.37227196e-01 7.97038853e-01 4.07399476e-01
-1.02444184e+00 5.93762286e-02 -2.18897626e-01 -7.29820281e-02
1.40973441e-02 7.77840316e-01 -4.33165401e-01 1.08374298e+00
4.38127548e-01 2.90229648e-01 -7.06186116e-01 1.01347613e+00
-3.52700084e-01 2.33271405e-01 -3.10663939e-01 -5.58626465e-02
1.61302447e-01 -1.21099330e-01 6.79249048e-01 9.01997089e-01
2.37560734e-01 -1.71120867e-01 -1.11926280e-01 8.42005610e-01
1.23079196e-01 1.24275416e-01 -4.57948267e-01 -2.72271663e-01
4.52321559e-01 9.60517287e-01 -3.64016026e-01 -3.60976428e-01
-2.41648957e-01 4.90008742e-01 6.18880093e-02 6.48254007e-02
-3.14963490e-01 -6.35099262e-02 4.50229645e-01 3.82804394e-01
1.08464053e-02 1.76903428e-04 -6.13012649e-02 -1.00255001e+00
3.33863318e-01 -9.96827364e-01 6.91928148e-01 -8.25875342e-01
-1.28586340e+00 5.34312963e-01 5.56726694e-01 -1.17743874e+00
-4.03769016e-01 -6.94122314e-01 -9.22889858e-02 5.94018340e-01
-1.56718504e+00 -1.15519881e+00 5.30056283e-02 6.76948369e-01
-3.78572233e-02 1.65080130e-01 1.13801289e+00 5.98030150e-01
9.05410498e-02 5.03299534e-01 -2.07984209e-01 -3.35836768e-01
6.13750517e-01 -1.28198326e+00 -1.67775363e-01 6.73720598e-01
4.23496485e-01 1.19339085e+00 1.24798071e+00 -4.31441069e-01
-1.46870530e+00 -3.06545794e-01 1.84532988e+00 -5.84502578e-01
8.88870776e-01 2.37679303e-01 -8.50044727e-01 5.83191097e-01
1.97924390e-01 -2.69173205e-01 4.21088219e-01 5.19164205e-01
-9.07391012e-01 -3.73582780e-01 -1.02127850e+00 4.76619363e-01
9.17310894e-01 -8.12587082e-01 -1.11766911e+00 4.36198533e-01
3.86779368e-01 2.12757755e-02 -8.49171877e-01 5.76241434e-01
7.87539601e-01 -9.55147326e-01 1.13868952e+00 -6.20809972e-01
2.35744461e-01 -2.46275634e-01 -4.22781199e-01 -7.62923777e-01
-2.76239008e-01 -4.91481543e-01 -1.41265363e-01 1.15517950e+00
4.72387463e-01 -6.25586808e-01 4.16699439e-01 7.51921415e-01
-8.48991424e-02 -5.00195503e-01 -1.05060279e+00 -1.06323969e+00
-3.26640680e-02 -6.30097806e-01 6.91225231e-01 1.03735328e+00
5.62126637e-01 2.29057819e-01 -1.52615950e-01 -1.54353172e-01
4.86966968e-01 1.01813525e-01 2.23817781e-01 -1.79766417e+00
-2.86248177e-01 -8.94944191e-01 -4.34752136e-01 -6.86699986e-01
-8.68209079e-03 -8.83076787e-01 -9.01463255e-02 -1.55822539e+00
3.44589472e-01 -5.04892111e-01 -6.34750664e-01 4.12879646e-01
1.66896284e-01 2.49777049e-01 -3.29769887e-02 4.37435716e-01
-7.77862608e-01 2.82850236e-01 7.85065711e-01 -3.61690342e-01
3.46526295e-01 -3.55925523e-02 -8.36523890e-01 5.75078428e-01
4.33234632e-01 -6.54130161e-01 -4.45726544e-01 -5.39797992e-02
1.07333148e+00 -1.02361254e-01 4.86921966e-01 -5.57807088e-01
5.66020906e-01 -2.43459150e-01 -4.28191990e-01 -4.79595006e-01
3.10824424e-01 -1.34597206e+00 3.63467246e-01 3.77531350e-01
-6.49311006e-01 2.76904136e-01 -1.97527826e-01 2.68818974e-01
-3.95748109e-01 -6.83406353e-01 3.26445192e-01 -2.36103594e-01
-5.41559041e-01 1.18721217e-01 -3.43391925e-01 -3.44729833e-02
3.22428226e-01 1.37138739e-01 -3.02451074e-01 -2.58870095e-01
-2.94226825e-01 -4.44667190e-02 5.30212939e-01 3.27172548e-01
3.82396519e-01 -1.39188313e+00 -5.45275211e-01 -2.61674553e-01
7.77873099e-01 -5.69134176e-01 -9.47716460e-02 9.84307647e-01
-1.80548474e-01 7.88929820e-01 2.06462443e-01 -2.74276823e-01
-1.48141885e+00 8.10001493e-01 1.23963550e-01 -8.69255006e-01
-3.17658275e-01 3.20133030e-01 -2.86702484e-01 -1.12888947e-01
1.98594809e-01 -1.14337996e-01 -2.65235782e-01 2.07250714e-01
5.08523703e-01 2.00879365e-01 4.13409263e-01 -5.27701616e-01
-6.41636252e-01 4.93046314e-01 -1.42977759e-01 -3.44823718e-01
1.13892961e+00 -1.99741036e-01 -6.23682439e-01 4.05467272e-01
9.12841976e-01 6.86338246e-02 2.25188032e-01 -5.55121541e-01
5.59889436e-01 -5.30775249e-01 2.01915979e-01 -9.70277309e-01
-5.67995191e-01 5.35024941e-01 3.53932887e-01 8.07768941e-01
1.15350842e+00 6.89308867e-02 5.67533493e-01 6.70416892e-01
6.63507283e-01 -1.22574830e+00 -3.08680832e-01 5.17313063e-01
1.01255524e+00 -7.77540147e-01 3.31455052e-01 -4.61648852e-01
-1.82624072e-01 8.24304104e-01 -5.37070483e-02 1.90468371e-01
6.43881440e-01 2.13200748e-01 -1.17250845e-01 -7.12127745e-01
-9.46406722e-01 -6.71229124e-01 8.86330009e-01 3.06446493e-01
5.31804800e-01 -1.50279254e-01 -9.72715378e-01 4.32259619e-01
-2.95991242e-01 2.18113422e-01 1.85979698e-02 9.96708572e-01
-4.28069592e-01 -1.52109146e+00 -3.43701184e-01 1.29115716e-01
-5.83062708e-01 -3.99706542e-01 -8.97744656e-01 9.89817679e-01
-1.18112177e-01 1.09872174e+00 -4.27221775e-01 -3.86713684e-01
4.37957823e-01 1.72959179e-01 6.40456200e-01 -3.55284482e-01
-8.16732943e-01 -2.63790727e-01 2.49120742e-01 -4.03761506e-01
-7.25135624e-01 -3.79526943e-01 -6.42656863e-01 -4.72783506e-01
-4.21489954e-01 7.69927442e-01 6.51738465e-01 1.26047289e+00
3.71789932e-02 3.25734228e-01 4.93227988e-01 -2.10225075e-01
-7.57831931e-01 -1.06959057e+00 -6.85734034e-01 6.00878596e-01
1.72000706e-01 -8.42980504e-01 -5.79647660e-01 -2.12972432e-01] | [9.493657112121582, 8.060552597045898] |
ce94971d-aba6-49bb-a62b-a1711743c328 | meta-particle-flow-for-sequential-bayesian | 1902.00640 | null | https://arxiv.org/abs/1902.00640v3 | https://arxiv.org/pdf/1902.00640v3.pdf | Particle Flow Bayes' Rule | We present a particle flow realization of Bayes' rule, where an ODE-based neural operator is used to transport particles from a prior to its posterior after a new observation. We prove that such an ODE operator exists. Its neural parameterization can be trained in a meta-learning framework, allowing this operator to reason about the effect of an individual observation on the posterior, and thus generalize across different priors, observations and to sequential Bayesian inference. We demonstrated the generalization ability of our particle flow Bayes operator in several canonical and high dimensional examples. | ['Xinshi Chen', 'Hanjun Dai', 'Le Song'] | 2019-02-02 | null | null | null | null | ['sequential-bayesian-inference'] | ['time-series'] | [ 1.42400771e-01 1.89456239e-01 1.12437466e-02 -2.36657634e-01
-2.46757284e-01 -2.57536769e-01 1.06343901e+00 -4.21183184e-02
-5.11537790e-01 7.91582048e-01 2.58573025e-01 -1.86273545e-01
-3.44922513e-01 -1.10301185e+00 -9.46337044e-01 -8.33670199e-01
-4.27648038e-01 8.37940991e-01 6.33929372e-01 2.11051524e-01
2.51067668e-01 7.38761187e-01 -1.32528377e+00 1.52324811e-02
3.22063684e-01 6.59866631e-01 1.78126782e-01 1.28063941e+00
-7.75674880e-02 6.52821958e-01 -6.11069262e-01 -3.45421433e-01
1.83202758e-01 -3.91589612e-01 -6.75951779e-01 -1.11509465e-01
4.08963084e-01 -2.60250270e-01 -3.99826676e-01 1.07525194e+00
1.13661543e-01 6.71524465e-01 1.15192556e+00 -1.03685331e+00
-6.63599789e-01 4.95814770e-01 1.11555845e-01 4.29118395e-01
1.32514000e-01 2.17613906e-01 9.44575548e-01 -5.04468858e-01
6.65022135e-01 1.57468939e+00 1.01718295e+00 6.46225870e-01
-1.43513739e+00 -1.41917869e-01 4.94575314e-02 -1.57708377e-01
-8.22879374e-01 -2.09622830e-02 1.76276267e-01 -7.44245529e-01
8.95588815e-01 -4.49270243e-03 1.11562932e+00 1.43057847e+00
7.91738570e-01 6.73455715e-01 9.20960724e-01 -3.73795271e-01
7.73621142e-01 -4.59981598e-02 3.24812561e-01 9.52442884e-01
2.13073552e-01 5.44307947e-01 -6.79808676e-01 -3.38484228e-01
7.65783370e-01 5.11162169e-03 -4.13652249e-02 -1.34732321e-01
-8.68546665e-01 9.09501374e-01 1.80487216e-01 -2.04632297e-01
-2.10318699e-01 7.07101882e-01 1.29087836e-01 -1.47183686e-01
5.44745803e-01 4.50381964e-01 -6.18488729e-01 -2.29485542e-01
-4.30640191e-01 5.89304507e-01 1.35503221e+00 6.26967609e-01
8.14128339e-01 -2.97359109e-01 -3.45128864e-01 3.13076973e-01
6.71600640e-01 8.13196957e-01 3.02861542e-01 -1.34736586e+00
-7.43172094e-02 1.11950771e-03 4.63531286e-01 -6.47294760e-01
-4.71946567e-01 -3.38175803e-01 -6.78484440e-01 4.47505891e-01
3.90025467e-01 -4.40978587e-01 -8.44097614e-01 1.68825197e+00
3.16421568e-01 6.68091536e-01 1.14360198e-01 5.29865623e-01
2.44714662e-01 9.37034845e-01 4.72066924e-02 -2.00863838e-01
1.11009157e+00 -5.89479148e-01 -6.04069948e-01 -3.49027403e-02
2.16744035e-01 -1.18269160e-01 8.02965820e-01 3.96248847e-01
-1.23766088e+00 -5.05419672e-01 -1.08550918e+00 1.20023392e-01
-5.00297070e-01 -1.92287758e-01 5.85106969e-01 7.22006202e-01
-9.97592092e-01 1.22773719e+00 -1.54738927e+00 -4.08368826e-01
5.36991298e-01 2.15619475e-01 1.72834739e-01 4.95302945e-01
-1.04267871e+00 1.02415979e+00 5.38946033e-01 5.43369986e-02
-1.22047853e+00 -1.13981545e+00 -4.59099889e-01 8.57356414e-02
-8.57234374e-02 -1.26976120e+00 1.44467533e+00 -3.82207073e-02
-2.07673359e+00 6.90289214e-02 -3.65335315e-01 -8.76275718e-01
5.19035280e-01 -4.27861392e-01 -2.88520575e-01 1.45185292e-01
2.07839347e-02 5.51200032e-01 1.13714767e+00 -1.01618016e+00
-8.93237352e-01 -1.34356961e-01 2.31280968e-01 -5.60021214e-02
5.42866699e-02 -5.10018110e-01 -3.94114405e-02 -2.74968147e-01
-9.80496928e-02 -1.00348616e+00 -3.01009029e-01 1.16407806e-02
-2.90548176e-01 -3.86796176e-01 6.62230015e-01 -2.52978373e-02
6.76838994e-01 -1.81568575e+00 3.46432716e-01 2.49638095e-01
3.61227579e-02 -1.14365280e-01 3.91537935e-01 4.04963672e-01
4.12083745e-01 2.49569044e-01 -5.70271015e-01 -5.59507012e-01
2.49000102e-01 5.37685454e-01 -7.32331097e-01 6.73613310e-01
1.42195925e-01 5.06337404e-01 -1.03724587e+00 -3.38250399e-01
3.47604543e-01 6.29874647e-01 -8.22999179e-01 1.60183728e-01
-5.90266526e-01 5.08509457e-01 -4.19345230e-01 -2.73542345e-01
5.12448251e-01 -3.95013422e-01 -1.38859138e-01 -1.53580025e-01
-3.09897494e-02 1.25481308e-01 -1.25519717e+00 1.40559554e+00
-5.12928843e-01 6.97979748e-01 -2.53822058e-01 -6.18484199e-01
5.70015311e-01 1.68480977e-01 2.50810653e-01 5.09498343e-02
1.23098344e-01 -2.48902187e-01 -1.83249474e-01 -6.39516294e-01
5.54986596e-01 -5.99471748e-01 1.74136236e-01 4.85651970e-01
2.41108686e-01 -1.99574113e-01 3.61312151e-01 2.87470579e-01
1.09592533e+00 3.59951437e-01 -5.80057390e-02 -5.51638365e-01
5.36768198e-01 -1.87588498e-01 4.34770226e-01 1.46293736e+00
1.85520455e-01 2.74247050e-01 3.68364066e-01 -6.13352895e-01
-9.09874976e-01 -1.56385577e+00 -5.95348060e-01 9.66936588e-01
4.49974090e-02 -4.18503642e-01 -5.41153789e-01 -5.67825317e-01
1.89953595e-01 9.75923955e-01 -8.89250875e-01 -2.20053807e-01
-4.76922840e-01 -1.23832941e+00 4.34522361e-01 5.66465735e-01
4.44443583e-01 -8.63975406e-01 -9.37564313e-01 3.12983513e-01
5.88736415e-01 -7.95704126e-01 -1.56315669e-01 1.89670295e-01
-9.83845770e-01 -1.01535738e+00 -2.20351323e-01 -9.57157016e-02
4.69205528e-01 -7.16674745e-01 8.18735123e-01 -3.33822489e-01
-1.65468514e-01 6.73020303e-01 1.33563399e-01 -4.38483685e-01
-9.45727766e-01 -1.11567825e-01 2.45295599e-01 -9.50149540e-03
-2.65956298e-02 -6.31892323e-01 -3.99944186e-01 3.97411138e-02
-6.75371408e-01 -2.63030708e-01 1.06896631e-01 7.32624710e-01
2.15075418e-01 2.47673824e-01 2.00274721e-01 -8.42696428e-01
7.05787241e-01 -3.32447648e-01 -1.01042140e+00 8.51294622e-02
-2.18988225e-01 5.42630434e-01 4.65524107e-01 -3.47958684e-01
-1.63743818e+00 -1.57663763e-01 -2.49763966e-01 -1.35127440e-01
-1.89010307e-01 2.83793956e-01 2.89844424e-01 2.39505067e-01
7.72773802e-01 -7.99420848e-02 -1.30795352e-02 -4.87122655e-01
6.36059582e-01 2.73534562e-02 7.19953418e-01 -1.19941556e+00
5.16939998e-01 1.11902857e+00 7.89232492e-01 -7.82608569e-01
-1.28064024e+00 -7.74355838e-03 -6.80022120e-01 -1.11143649e-01
1.25305068e+00 -5.14038503e-01 -1.32852244e+00 3.66045743e-01
-1.40881300e+00 -4.82761979e-01 -8.28527689e-01 7.67069519e-01
-7.18754947e-01 1.26970097e-01 -7.08522320e-01 -8.75906467e-01
1.54801548e-01 -9.85111773e-01 8.32141042e-01 3.39900464e-01
-1.67937368e-01 -1.68004811e+00 6.57959104e-01 -4.50174212e-01
1.30111396e-01 -1.75267413e-01 7.87539661e-01 -7.01828837e-01
-5.22909999e-01 -1.73077554e-01 5.66289388e-03 3.77140462e-01
-1.10899368e-02 3.81259501e-01 -8.32223117e-01 -2.32569143e-01
2.73131102e-01 1.80593133e-01 1.08237886e+00 6.44936979e-01
9.52912509e-01 -3.75222743e-01 -7.11838305e-01 5.56112885e-01
1.29488277e+00 7.85724670e-02 2.87406474e-01 1.63257588e-02
5.34008324e-01 1.70902789e-01 -2.55350649e-01 3.45286906e-01
1.03623501e-03 3.51820767e-01 9.61919650e-02 5.95175028e-01
-1.19294114e-01 -3.00605088e-01 3.58745068e-01 5.99073350e-01
-1.62091151e-01 -4.13692743e-01 -7.18009472e-01 2.24083543e-01
-1.93101752e+00 -1.02350092e+00 -6.11010753e-03 1.99290633e+00
6.46101892e-01 3.45985025e-01 -1.59819216e-01 -5.32445133e-01
6.46221161e-01 -1.27474755e-01 -5.96509099e-01 -1.81217462e-01
1.59048066e-02 3.52155715e-01 5.91753244e-01 1.07365251e+00
-1.09777272e+00 7.04278111e-01 8.61035728e+00 5.99562049e-01
-4.52760279e-01 3.59025896e-01 -8.33374634e-02 9.85853821e-02
-1.73164859e-01 2.33685851e-01 -1.37386608e+00 5.67161679e-01
1.20654356e+00 -1.09601088e-01 2.93838501e-01 4.22585458e-01
1.90935098e-02 -2.79926509e-01 -1.42560983e+00 4.20997113e-01
-5.13443649e-02 -1.87103796e+00 2.75838375e-01 2.15387091e-01
8.05678189e-01 7.05868155e-02 -2.18254372e-01 2.91916907e-01
1.09829402e+00 -5.79487145e-01 6.66649163e-01 1.08090580e+00
-1.04604736e-01 -5.30373812e-01 2.75962383e-01 3.15416664e-01
-8.23121607e-01 -1.41361058e-01 -5.10805190e-01 -3.51173341e-01
6.03178561e-01 5.72585881e-01 -8.56572747e-01 8.32793266e-02
6.08958662e-01 8.77069235e-01 -2.35629022e-01 9.55796003e-01
-3.71423513e-01 5.40464640e-01 -6.72567844e-01 -1.85911432e-01
1.84544120e-02 -3.84798646e-01 1.09164953e+00 1.22854745e+00
3.12029153e-01 7.05182552e-04 1.16737664e-01 1.16359949e+00
6.91284463e-02 -5.15987456e-01 -3.35031867e-01 2.40609765e-01
3.05361509e-01 8.80783081e-01 -9.27591443e-01 -6.93543136e-01
-8.46965387e-02 6.59542620e-01 3.01877558e-01 5.18030643e-01
-7.65978634e-01 -1.16484940e-01 8.16015899e-01 -2.89606780e-01
7.48425424e-01 -1.25665665e-01 -1.19071662e-01 -1.13023841e+00
-4.40158427e-01 9.60850567e-02 4.78694290e-01 -7.83394098e-01
-1.70610440e+00 4.90424149e-02 5.73219657e-01 -7.56372035e-01
-2.02435762e-01 -1.10304189e+00 -8.24276268e-01 6.90016150e-01
-7.36383080e-01 -5.14167488e-01 3.12402099e-01 1.72578216e-01
1.26004517e-01 -3.86144698e-01 6.94249570e-01 -3.65818471e-01
-4.33777839e-01 -1.42724514e-01 1.45069689e-01 -4.45662104e-02
1.08490050e-01 -1.62362051e+00 4.42466557e-01 7.98146725e-01
7.59358481e-02 7.90926754e-01 1.11320138e+00 -9.82470214e-01
-1.30188417e+00 -8.93157303e-01 2.76831031e-01 -8.39481354e-01
1.03171051e+00 -1.50078937e-01 -7.83994555e-01 1.10984719e+00
1.83593348e-01 4.36044633e-02 2.24961296e-01 3.22652876e-01
-1.28074229e-01 -7.72907771e-03 -1.05657899e+00 7.99332321e-01
9.76041794e-01 -4.96945381e-01 -9.36772943e-01 5.25608122e-01
7.62074172e-01 -6.28777444e-01 -1.00238407e+00 2.68265188e-01
6.78014994e-01 -9.15586233e-01 1.21115029e+00 -9.49601352e-01
2.30820477e-01 -4.25338477e-01 -1.53027266e-01 -1.31050932e+00
-2.53859550e-01 -8.39020193e-01 -6.19365156e-01 7.64377177e-01
4.46710616e-01 -8.82376730e-01 5.95835745e-01 4.69507575e-01
-3.04324567e-01 -5.22477269e-01 -1.06072950e+00 -8.74145210e-01
4.28025186e-01 -8.37879181e-01 5.94209373e-01 1.25689641e-01
-1.73593029e-01 2.15318188e-01 -1.94838822e-01 6.06312513e-01
8.67193878e-01 -1.84394017e-01 4.56255227e-01 -1.41863871e+00
-7.47154534e-01 -5.41623354e-01 -2.61640787e-01 -1.32675028e+00
5.09949207e-01 -1.16662836e+00 3.63288373e-01 -1.47094977e+00
1.15159549e-01 -1.21707648e-01 9.21050534e-02 -7.09571987e-02
2.50415802e-02 -5.00408448e-02 8.08651149e-02 1.21126525e-01
-3.85737181e-01 6.69918478e-01 1.13629198e+00 -4.32566553e-02
-2.56648600e-01 2.35910818e-01 4.54773717e-02 1.15101230e+00
6.65919542e-01 -7.09895372e-01 -2.30647966e-01 -1.48019150e-01
6.44472182e-01 -5.10899313e-02 7.11022258e-01 -1.09405482e+00
5.49492538e-01 2.25664452e-02 6.33367181e-01 -4.81128424e-01
5.67174137e-01 -4.76884693e-01 2.30957389e-01 6.09236181e-01
-4.60688978e-01 -1.57659173e-01 1.36962801e-01 1.11170185e+00
4.65750098e-01 -5.70663214e-01 7.29606092e-01 -2.83405095e-01
-3.47164780e-01 3.42985541e-01 -7.19845533e-01 2.89755166e-01
5.08072674e-01 2.26619437e-01 -4.06001717e-01 -1.06731258e-01
-1.28286839e+00 -9.09595937e-02 2.62939990e-01 -1.05028830e-01
4.03106838e-01 -1.17296779e+00 -4.03954655e-01 2.94319361e-01
-4.34965491e-01 -7.12157637e-02 1.75319254e-01 6.43891931e-01
-5.14986336e-01 1.65010262e-02 1.24554578e-02 -9.99690473e-01
-3.09715331e-01 3.34951162e-01 7.25638986e-01 -3.91523130e-02
-1.17447019e+00 7.63955176e-01 -1.25317660e-03 -3.07320416e-01
1.67148873e-01 -6.16949737e-01 2.33665302e-01 -1.78028002e-01
6.38868570e-01 7.60819197e-01 -4.19842839e-01 -1.52433395e-01
-1.29009798e-01 6.78221464e-01 2.11865395e-01 -7.14788020e-01
9.36975718e-01 -1.04296319e-01 -9.21198875e-02 1.07323444e+00
1.11223364e+00 -8.45308453e-02 -1.74752641e+00 4.71656360e-02
-1.98058590e-01 -2.41322070e-01 -7.93520808e-02 -5.48170924e-01
-5.85621774e-01 7.29726851e-01 3.93572837e-01 5.97390354e-01
2.00488165e-01 2.85032392e-01 4.97950524e-01 9.14096296e-01
1.03268348e-01 -8.93898487e-01 -3.63243558e-02 8.32151651e-01
5.82204521e-01 -8.28473568e-01 2.97891468e-01 -5.64745180e-02
-9.18151513e-02 1.31248879e+00 4.37253714e-02 -5.39988399e-01
1.53530025e+00 4.55609292e-01 -4.29126769e-01 -2.35586777e-01
-8.38942349e-01 1.54436519e-02 2.70634860e-01 6.24395728e-01
-1.80019781e-01 -4.32483815e-02 3.44515860e-01 8.74787271e-02
-3.83379549e-01 1.14787273e-01 6.87775195e-01 7.70197570e-01
-5.03626764e-01 -7.27925420e-01 -1.62070557e-01 3.72181296e-01
-2.67208338e-01 1.44430891e-01 4.12182271e-01 7.26110399e-01
2.48769864e-01 5.41743338e-01 5.91276824e-01 1.64602697e-01
4.96641360e-02 2.81086415e-01 8.68764937e-01 -6.23304427e-01
-1.10070936e-01 -2.09296152e-01 -1.17484607e-01 -4.71895993e-01
-7.88017154e-01 -1.04302645e+00 -1.30499494e+00 -1.66556954e-01
8.87643695e-02 2.60853082e-01 6.15937948e-01 1.39555693e+00
1.59604654e-01 6.16416812e-01 4.64201160e-02 -1.13316560e+00
-5.47758520e-01 -9.84390974e-01 -6.59354448e-01 1.49744570e-01
6.00874364e-01 -8.47981036e-01 -9.65582132e-01 3.09097797e-01] | [6.853137969970703, 3.8492517471313477] |
1d2e4b27-116a-4533-a318-147945e16a86 | milestones-in-autonomous-driving-and-1 | 2305.11239 | null | https://arxiv.org/abs/2305.11239v2 | https://arxiv.org/pdf/2305.11239v2.pdf | Milestones in Autonomous Driving and Intelligent Vehicles Part I: Control, Computing System Design, Communication, HD Map, Testing, and Human Behaviors | Interest in autonomous driving (AD) and intelligent vehicles (IVs) is growing at a rapid pace due to the convenience, safety, and economic benefits. Although a number of surveys have reviewed research achievements in this field, they are still limited in specific tasks and lack systematic summaries and research directions in the future. Our work is divided into 3 independent articles and the first part is a Survey of Surveys (SoS) for total technologies of AD and IVs that involves the history, summarizes the milestones, and provides the perspectives, ethics, and future research directions. This is the second part (Part I for this technical survey) to review the development of control, computing system design, communication, High Definition map (HD map), testing, and human behaviors in IVs. In addition, the third part (Part II for this technical survey) is to review the perception and planning sections. The objective of this paper is to involve all the sections of AD, summarize the latest technical milestones, and guide abecedarians to quickly understand the development of AD and IVs. Combining the SoS and Part II, we anticipate that this work will bring novel and diverse insights to researchers and abecedarians, and serve as a bridge between past and future. | ['Fei-Yue Wang', 'Nanning Zheng', 'Dongpu Cao', 'Jinjun Wang', 'Chen Lv', 'Siyu Teng', 'Zhongxu Hu', 'Li Li', 'Daxin Tian', 'Yang Xing', 'Chao Huang', 'Yuchen Li', 'Long Chen'] | 2023-05-12 | null | null | null | null | ['ethics'] | ['miscellaneous'] | [-1.36031479e-01 1.48390949e-01 -4.27447021e-01 -4.58510071e-01
4.44919690e-02 -6.01309538e-01 5.48129559e-01 -2.50864983e-01
-8.37965533e-02 4.57296222e-01 -1.34367660e-01 -6.89311266e-01
-1.76955126e-02 -7.52418339e-01 -5.66066086e-01 -4.31169093e-01
6.91103563e-02 -1.20611511e-01 4.87874120e-01 -5.95275879e-01
4.39488739e-01 6.36951625e-01 -2.17453027e+00 -2.19196707e-01
8.95126820e-01 1.14745057e+00 3.74526620e-01 4.06031579e-01
2.15145409e-01 7.27319837e-01 -4.87396598e-01 -8.51851478e-02
6.56681657e-02 -8.31067413e-02 -5.25617719e-01 -5.00661433e-02
2.76399195e-01 -5.18383622e-01 -5.81072569e-01 7.20314145e-01
4.39768583e-01 -1.02400385e-01 3.77859771e-01 -1.97098088e+00
-3.83795977e-01 -2.78246030e-02 -2.23026440e-01 -1.23619877e-01
1.25710502e-01 2.25652993e-01 4.20697212e-01 -7.47685611e-01
3.82460266e-01 9.24376607e-01 5.68678856e-01 5.91755688e-01
-5.09035110e-01 -9.43089068e-01 3.36288214e-01 5.33941269e-01
-1.22311127e+00 -6.59570038e-01 5.41419208e-01 -5.94026804e-01
8.88199568e-01 1.53946459e-01 1.00288963e+00 8.64239812e-01
4.99713808e-01 1.11692727e+00 1.05711067e+00 -3.13066512e-01
4.44864750e-01 6.05678022e-01 6.20861948e-01 3.76034468e-01
5.35600066e-01 4.03506547e-01 -2.98702449e-01 2.68572539e-01
1.75165311e-01 -5.04310191e-01 1.12031050e-01 -5.60857832e-01
-7.86210239e-01 7.07304537e-01 7.97196031e-02 4.05141562e-02
-3.26873958e-01 -8.67251679e-03 4.33880597e-01 7.33417422e-02
-1.83178008e-01 1.53306708e-01 -1.01690143e-01 -5.83942711e-01
-2.08042830e-01 4.35729980e-01 9.19725537e-01 1.37023771e+00
6.31454110e-01 3.65645945e-01 1.65706307e-01 6.22317195e-01
4.87794906e-01 9.70865309e-01 -1.11643888e-01 -1.33968723e+00
1.64893687e-01 1.55704379e-01 4.46336031e-01 -1.12233543e+00
-5.28631508e-01 -3.85121778e-02 -3.72950703e-01 6.60233855e-01
-3.92507643e-01 -6.67441547e-01 -6.37837946e-01 1.29543173e+00
-3.92205594e-03 -4.81796443e-01 1.55232742e-01 7.95759916e-01
1.09612465e+00 6.85525358e-01 -3.97987897e-03 -1.81580946e-01
1.27212882e+00 -1.03484070e+00 -1.17630219e+00 -6.18331552e-01
2.81120598e-01 -5.62990665e-01 8.42286706e-01 2.21779078e-01
-1.00842762e+00 -7.66087115e-01 -1.68707871e+00 1.60164237e-01
-5.57186604e-01 -6.68833852e-02 5.70136487e-01 9.74494994e-01
-1.17440319e+00 -9.60540846e-02 -9.55136061e-01 -8.64766955e-01
1.73485696e-01 1.49740204e-01 1.91665500e-01 -5.74787408e-02
-1.27381063e+00 1.45811617e+00 -1.89033002e-01 -1.05631672e-01
-5.79477251e-01 -3.88904124e-01 -9.59594965e-01 -6.53208673e-01
2.25340322e-01 -5.24693310e-01 1.54200983e+00 -2.00982228e-01
-1.71866846e+00 6.71864331e-01 -2.98541576e-01 -3.33090246e-01
1.12792276e-01 -2.31840044e-01 -8.66563678e-01 -9.17716324e-03
3.79617304e-01 7.58950889e-01 9.47205499e-02 -1.48235071e+00
-1.28318465e+00 -2.54164279e-01 1.57647535e-01 3.06620389e-01
5.31933419e-02 6.83997497e-02 -4.60033387e-01 1.84677631e-01
-7.32378736e-02 -9.11565959e-01 -1.86843216e-01 1.73420571e-02
-1.03566639e-01 -2.10119709e-01 1.36117792e+00 -5.02286246e-03
1.36488354e+00 -2.52688694e+00 -5.35641015e-01 -2.01263651e-02
1.24008611e-01 5.63548326e-01 2.13321134e-01 9.56505835e-01
5.08672714e-01 -1.64329767e-01 1.18425250e-01 1.21595986e-01
2.92578936e-01 4.93420869e-01 -4.01371986e-01 4.05893803e-01
-1.75575182e-01 7.78368235e-01 -7.13308752e-01 -1.21799409e-01
9.41005170e-01 2.32947633e-01 4.20008302e-02 -1.08415417e-01
3.07349026e-01 2.21263692e-01 -5.82891166e-01 5.73191464e-01
8.54724646e-01 2.85907775e-01 -2.37539873e-01 -1.97716892e-01
-1.04741311e+00 2.70704985e-01 -9.85030949e-01 8.55068147e-01
-2.72587180e-01 1.02046132e+00 3.83091748e-01 -9.47073042e-01
9.60834384e-01 2.74985075e-01 5.49868286e-01 -1.13240635e+00
1.31441697e-01 5.67443013e-01 -2.04504296e-01 -9.75178361e-01
6.90982997e-01 2.29985282e-01 -2.27598384e-01 1.11313678e-01
-6.42213702e-01 -4.01368201e-01 1.74989641e-01 2.53732753e-04
6.90485895e-01 -4.24222136e-03 2.15371981e-01 -2.84038097e-01
5.49208164e-01 2.81809062e-01 4.61341739e-01 6.57634199e-01
-9.23394263e-01 -2.71446956e-03 2.77931213e-01 -4.65300679e-01
-7.68506885e-01 -9.90305722e-01 -3.92656505e-01 6.86614871e-01
9.74193513e-01 -4.28127974e-01 -5.96166074e-01 -2.02545673e-01
2.22000614e-01 1.16894245e+00 -2.81640172e-01 2.48540957e-02
-3.85109603e-01 -3.80284518e-01 5.55773914e-01 9.15455103e-01
9.89519835e-01 -6.80312812e-01 -1.00660372e+00 -6.37376159e-02
-4.06594098e-01 -1.38974142e+00 2.39175349e-01 -1.07859060e-01
-4.01196510e-01 -1.01790011e+00 -1.46515667e-01 -1.09711337e+00
2.59319097e-01 9.25776541e-01 5.68850517e-01 -1.87562123e-01
3.99071276e-02 7.97081888e-01 -3.11807871e-01 -1.02978969e+00
-2.11963981e-01 -3.29355180e-01 3.37128580e-01 -4.78269398e-01
8.49903286e-01 -3.20472002e-01 -6.43600583e-01 8.50569010e-01
-2.18216181e-01 6.86569232e-03 5.40084600e-01 2.49339834e-01
3.49751890e-01 8.99395794e-02 5.30488551e-01 -3.46301526e-01
3.92184705e-01 -4.18571413e-01 -6.72904134e-01 -2.30945230e-01
-9.81176913e-01 -6.82115972e-01 2.56866574e-01 4.38292742e-01
-8.76024127e-01 -1.97240427e-01 -2.96009958e-01 3.31041992e-01
-4.91019905e-01 1.45977169e-01 -1.45704195e-01 -3.08310539e-01
6.38226628e-01 -3.93650234e-02 5.00467420e-01 -2.03518178e-02
1.59993425e-01 1.33321571e+00 5.12741923e-01 -2.64436513e-01
6.30833983e-01 6.16326869e-01 -6.14389144e-02 -1.09727466e+00
-3.85939568e-01 -5.26946664e-01 -3.58488441e-01 -7.85270274e-01
9.05953944e-01 -7.36817837e-01 -9.01510775e-01 6.60665512e-01
-9.35110688e-01 -3.65447462e-01 -1.61130801e-01 8.40184331e-01
-6.23975396e-01 1.40959010e-01 -4.80480075e-01 -1.02639103e+00
1.19479962e-01 -1.51023483e+00 9.18677449e-01 4.50572163e-01
-1.68763891e-01 -7.18985379e-01 -1.79685846e-01 5.40226698e-01
6.26328468e-01 2.26619944e-01 6.33670211e-01 1.79936603e-01
-6.01389885e-01 -4.54220802e-01 -2.54931241e-01 2.06836313e-01
-1.26067340e-01 2.89619386e-01 -1.06819665e+00 1.86567602e-03
-6.33471310e-02 -5.89460023e-02 3.65015090e-01 5.93558013e-01
6.29978836e-01 1.55287683e-01 -8.20335567e-01 2.53018171e-01
1.05366969e+00 1.07383132e+00 8.22388887e-01 8.01644981e-01
1.70584306e-01 8.74102652e-01 1.34003091e+00 1.66510746e-01
9.49609697e-01 7.54084527e-01 5.30454874e-01 -6.94911554e-02
-1.48037285e-01 -7.18732998e-02 4.83874947e-01 9.06030416e-01
-3.44330072e-01 -2.26092145e-01 -8.04285109e-01 4.89053547e-01
-1.71444595e+00 -7.81030238e-01 -5.92540264e-01 1.91301918e+00
-7.05963969e-02 -6.34904876e-02 2.68713474e-01 1.91771433e-01
9.34394419e-01 1.45109119e-02 -5.73000133e-01 -6.16526783e-01
1.48126602e-01 -4.16928589e-01 7.21677721e-01 4.67542380e-01
-1.06803644e+00 7.57645309e-01 7.89812994e+00 2.29553267e-01
-1.07393467e+00 -2.29559645e-01 1.42014965e-01 3.98633003e-01
-1.01260841e-01 -1.23002172e-01 -1.01051438e+00 3.14740419e-01
9.86880422e-01 -3.63916487e-01 4.28280272e-02 1.07018614e+00
5.47366500e-01 -3.52804720e-01 -7.78133869e-01 7.93415487e-01
3.17996182e-02 -1.11815178e+00 -4.40301299e-01 8.05318654e-02
5.73304176e-01 3.29447418e-01 2.84612000e-01 4.78383332e-01
1.21573545e-01 -5.88437378e-01 1.07577634e+00 1.18982419e-01
7.20351934e-01 -7.32115388e-01 1.07133734e+00 2.10461259e-01
-1.17702270e+00 -3.82572711e-01 -3.71729940e-01 -5.53969145e-01
4.73765165e-01 2.55472481e-01 -2.62755454e-01 5.42004943e-01
1.07227898e+00 6.10762537e-01 -6.10870346e-02 1.00533533e+00
-2.01361582e-01 3.13567102e-01 8.03367700e-03 -5.64280570e-01
2.83446074e-01 -1.99586913e-01 6.52584672e-01 1.12157643e+00
-2.97367163e-02 2.53662407e-01 1.49617106e-01 4.89684880e-01
7.78557837e-01 -2.33830661e-01 -9.32376564e-01 7.47329071e-02
7.67364144e-01 1.04981434e+00 -2.38050818e-01 -2.14936510e-01
-9.06840026e-01 2.40873411e-01 -4.59386498e-01 4.51913625e-01
-8.86475623e-01 -1.00514257e+00 9.74523544e-01 4.05450374e-01
-1.65840518e-02 -6.26525939e-01 -8.81054521e-01 -1.91730529e-01
6.19976558e-02 -3.81318212e-01 -6.69067502e-02 -8.92076612e-01
-7.09589005e-01 3.48147810e-01 4.89887416e-01 -1.48160982e+00
-1.57273665e-01 -5.93826950e-01 -3.80925119e-01 4.88778323e-01
-1.58905745e+00 -9.85071778e-01 -9.28531408e-01 -4.83914241e-02
4.30722475e-01 -2.77933329e-01 6.52720034e-01 4.62175637e-01
-5.68460643e-01 4.51167047e-01 1.30845591e-01 -2.86565155e-01
6.06279373e-01 -6.70144558e-01 5.78937769e-01 6.42482698e-01
-1.05639660e+00 5.51146984e-01 9.13141191e-01 -5.19083977e-01
-1.83988774e+00 -1.02288914e+00 7.66951680e-01 -4.07732517e-01
4.79977667e-01 -3.03817868e-01 -1.28848970e-01 8.04944813e-01
4.87647653e-01 -4.50760722e-01 4.56395000e-01 -1.38727143e-01
2.61217415e-01 -5.24366140e-01 -1.10417402e+00 8.66567373e-01
9.21689630e-01 -1.24043427e-01 -2.27388710e-01 -3.84809636e-02
5.67440629e-01 -5.68783045e-01 -5.88862956e-01 6.17450476e-01
1.02379930e+00 -1.04160643e+00 4.19773489e-01 1.46748975e-01
1.93000641e-02 -5.83439112e-01 -3.55862647e-01 -8.98844182e-01
-5.18122017e-01 -5.63589394e-01 2.60090560e-01 8.85792851e-01
4.26225305e-01 -1.19759417e+00 5.86005449e-01 7.83709049e-01
-1.07199407e+00 -6.54080927e-01 -5.58051288e-01 -9.08230305e-01
-1.78128839e-01 -8.01807582e-01 4.05752778e-01 5.56523740e-01
5.05097926e-01 4.31982011e-01 -6.37227148e-02 1.99994132e-01
5.23943186e-01 -2.32070297e-01 1.11260486e+00 -1.08771634e+00
4.30730611e-01 -3.97365421e-01 -9.83172596e-01 -1.40082729e+00
-2.67133564e-01 -3.11793327e-01 3.06873411e-01 -2.17284083e+00
-1.72955737e-01 -3.84008706e-01 2.67367452e-01 2.65458971e-01
3.41110826e-01 1.55674696e-01 -4.83383089e-02 5.40367849e-02
-4.57460791e-01 4.94874895e-01 1.17047405e+00 -1.61458924e-02
-2.84462273e-01 3.18141669e-01 -9.94175732e-01 7.73953319e-01
1.02041698e+00 3.94542128e-01 -7.73117304e-01 -3.01236272e-01
-1.13049842e-01 -1.16778746e-01 4.44071323e-01 -1.23203409e+00
5.04424930e-01 -4.30214971e-01 -6.94040135e-02 -1.12901211e+00
3.34419072e-01 -9.81091321e-01 -1.07518295e-02 5.78661799e-01
4.53100443e-01 8.08407292e-02 5.34127474e-01 3.03533137e-01
-2.31148452e-01 1.80230454e-01 8.40445518e-01 2.14672923e-01
-1.41637671e+00 -1.61407758e-02 -1.24112606e+00 -4.04638916e-01
1.69884467e+00 -8.43792677e-01 -5.89527905e-01 -5.54447711e-01
-2.47754365e-01 6.95809484e-01 3.95687401e-01 7.86909282e-01
4.86063570e-01 -1.29807019e+00 -2.97046900e-01 2.94882536e-01
3.15427274e-01 -1.76617578e-01 3.14285308e-01 1.01630557e+00
-6.30177796e-01 9.63391066e-01 -4.48499650e-01 -7.70108461e-01
-9.62890744e-01 4.34991866e-01 1.28093988e-01 7.39840925e-01
-6.66190267e-01 3.47818017e-01 3.99677217e-01 -3.53845686e-01
4.60716516e-01 5.20775765e-02 -5.03672123e-01 -4.54988211e-01
5.87054849e-01 9.04686987e-01 1.25743702e-01 -8.39983761e-01
-5.68193674e-01 8.03578854e-01 6.51997998e-02 -2.55393803e-01
8.18485558e-01 -8.69138002e-01 3.90166268e-02 6.68010652e-01
7.84609556e-01 -4.89438511e-02 -1.00459504e+00 5.25447488e-01
-4.42093760e-01 -3.07339638e-01 -5.09383380e-02 -6.53034449e-01
-7.80534565e-01 7.21509814e-01 7.13584363e-01 1.34100184e-01
1.03736663e+00 -2.12032367e-02 7.94401050e-01 1.16441704e-01
7.78363287e-01 -1.53933752e+00 -1.85975775e-01 7.48756468e-01
8.52382243e-01 -8.79162192e-01 -2.18695030e-01 -7.97226071e-01
-9.38502550e-01 9.52474654e-01 9.23143506e-01 2.94509768e-01
1.03754127e+00 6.61015093e-01 4.70411897e-01 -3.04659694e-01
-5.04846394e-01 -1.62931994e-01 -2.14314446e-01 1.33571172e+00
3.14455897e-01 1.53728008e-01 -5.50563395e-01 4.72934067e-01
-5.45955896e-01 2.73071323e-02 5.60059190e-01 1.25370538e+00
-9.59544003e-01 -9.76475656e-01 -4.16931570e-01 2.58139312e-01
1.34646773e-01 5.74440300e-01 -2.28850603e-01 1.07128739e+00
3.14653546e-01 1.77826476e+00 1.86609209e-01 -1.11836183e+00
9.84459579e-01 -3.15163434e-01 5.54139316e-02 -1.11354627e-01
1.04340769e-01 -4.31572706e-01 6.33459091e-01 -7.91329324e-01
-3.25201005e-01 -7.83832312e-01 -1.36923087e+00 -8.22654188e-01
-1.17354587e-01 8.47974494e-02 1.19856989e+00 7.25756526e-01
5.15902877e-01 6.62914634e-01 7.58496583e-01 -5.67730069e-01
-2.76258774e-02 -5.00679135e-01 -6.33254051e-01 -4.44756538e-01
2.56258160e-01 -9.93277490e-01 -8.56225491e-02 -3.14631432e-01] | [5.65463399887085, 1.0613548755645752] |
f7553ad1-be5a-4d97-a7ec-6df6494c96eb | leaving-the-nest-going-beyond-local-loss | 2305.16830 | null | https://arxiv.org/abs/2305.16830v1 | https://arxiv.org/pdf/2305.16830v1.pdf | Leaving the Nest: Going Beyond Local Loss Functions for Predict-Then-Optimize | Predict-then-Optimize is a framework for using machine learning to perform decision-making under uncertainty. The central research question it asks is, "How can the structure of a decision-making task be used to tailor ML models for that specific task?" To this end, recent work has proposed learning task-specific loss functions that capture this underlying structure. However, current approaches make restrictive assumptions about the form of these losses and their impact on ML model behavior. These assumptions both lead to approaches with high computational cost, and when they are violated in practice, poor performance. In this paper, we propose solutions to these issues, avoiding the aforementioned assumptions and utilizing the ML model's features to increase the sample efficiency of learning loss functions. We empirically show that our method achieves state-of-the-art results in four domains from the literature, often requiring an order of magnitude fewer samples than comparable methods from past work. Moreover, our approach outperforms the best existing method by nearly 200% when the localness assumption is broken. | ['Milind Tambe', 'Bryan Wilder', 'Andrew Perrault', 'Sanket Shah'] | 2023-05-26 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [ 9.43286940e-02 8.18080232e-02 -5.64399958e-01 -7.90037751e-01
-1.24788940e+00 -5.02170205e-01 4.73468035e-01 2.78422624e-01
-5.02452850e-01 9.32373405e-01 -4.55481075e-02 -6.01311326e-01
-1.61477193e-01 -4.16195214e-01 -7.37666786e-01 -5.59190691e-01
1.82964310e-01 5.87379932e-01 1.01842202e-01 1.92892224e-01
4.69470918e-01 2.87715465e-01 -1.41664064e+00 1.83903769e-01
1.04982316e+00 1.45064688e+00 2.14367285e-02 3.32941949e-01
-1.09004803e-01 8.46316993e-01 -4.38043952e-01 -7.28412688e-01
3.20897043e-01 -2.78851949e-02 -6.19462967e-01 -5.54514192e-02
3.61826390e-01 -2.90685028e-01 7.58553222e-02 1.06803429e+00
4.53707427e-01 1.95107490e-01 9.75344419e-01 -1.12771964e+00
-1.79122597e-01 7.52843738e-01 -5.74941576e-01 1.53684840e-01
1.00159002e-02 3.91101688e-02 1.24708724e+00 -7.91618228e-01
1.70707881e-01 1.42094302e+00 8.23477685e-01 4.15225714e-01
-1.50153530e+00 -7.14839637e-01 4.88195539e-01 9.00482759e-02
-1.45210958e+00 -6.67044103e-01 6.22193575e-01 -6.23433232e-01
7.67241180e-01 2.11658940e-01 -8.79439563e-02 1.13948929e+00
4.27206188e-01 9.02251959e-01 1.28503013e+00 -3.90699595e-01
5.44730723e-01 4.56276864e-01 1.56925127e-01 3.43754262e-01
4.63697791e-01 7.83848297e-03 -4.69727069e-01 -3.43022645e-01
1.93430439e-01 -2.28311628e-01 2.03905888e-02 -4.25290704e-01
-5.85831881e-01 1.09131551e+00 8.63332227e-02 5.77021092e-02
-1.69932991e-01 1.72949761e-01 4.70410407e-01 5.18137693e-01
9.60583627e-01 6.19233191e-01 -6.47267461e-01 -9.28052366e-02
-1.09556055e+00 6.02684319e-01 9.39192176e-01 7.49465883e-01
7.25538492e-01 -1.40178025e-01 -3.59000534e-01 7.58964539e-01
3.86039138e-01 6.26907796e-02 1.96961313e-01 -1.01206326e+00
7.82735944e-01 1.01093464e-01 5.17761052e-01 -7.29986072e-01
-3.66124958e-01 -6.94356501e-01 -3.35750729e-01 3.05217117e-01
6.78231776e-01 -4.15202647e-01 -7.29930758e-01 1.74218953e+00
2.15772465e-01 3.62105705e-02 -1.12500936e-01 6.31662667e-01
1.83237288e-02 3.23548913e-01 2.23958597e-01 -1.50687039e-01
1.02884173e+00 -6.76711738e-01 -6.43778145e-01 -6.72822416e-01
6.84668303e-01 -7.61227548e-01 1.36062062e+00 5.61796784e-01
-9.58809555e-01 -2.19839364e-01 -1.12320459e+00 2.38666028e-01
8.08409750e-02 6.88214228e-02 5.31920075e-01 8.36122811e-01
-5.70527196e-01 7.26032197e-01 -7.33988523e-01 -6.16700500e-02
3.84726226e-01 1.95729658e-01 2.48831898e-01 -1.92003384e-01
-1.06324053e+00 1.21992052e+00 3.32608819e-01 -5.87566271e-02
-9.26682472e-01 -9.02348578e-01 -6.96131170e-01 1.06829718e-01
7.71590173e-01 -6.68316185e-01 1.79830003e+00 -9.42943275e-01
-1.36060596e+00 5.16145825e-01 -2.44716242e-01 -7.71010995e-01
1.12159765e+00 -5.31587243e-01 -1.45687163e-01 -3.40015620e-01
-7.00997040e-02 2.16035515e-01 8.02774012e-01 -1.34266269e+00
-7.86593080e-01 -3.04317266e-01 2.24128217e-01 2.48332500e-01
-2.03352749e-01 9.50171053e-02 -2.27625817e-01 -7.00526953e-01
-2.24039897e-01 -8.13505828e-01 -4.44256961e-01 6.57806918e-02
-4.52627331e-01 -3.62351954e-01 4.33197051e-01 -3.83259624e-01
1.40923989e+00 -2.11888218e+00 -1.84037760e-01 2.18323395e-01
-4.29576561e-02 9.48239118e-02 1.57232985e-01 3.41909736e-01
3.52057099e-01 3.55573118e-01 -3.56466413e-01 -7.63714790e-01
3.37112069e-01 9.65297669e-02 -4.37547624e-01 4.77004677e-01
2.05618620e-01 5.69831133e-01 -5.59412122e-01 -4.48148549e-01
-3.19598168e-02 2.04558879e-01 -7.13341355e-01 3.09200227e-01
-5.78196108e-01 2.37470135e-01 -5.92145622e-01 5.41329086e-01
5.68121612e-01 -2.54246116e-01 3.38018388e-01 3.80226076e-02
1.98341519e-01 7.25134730e-01 -1.34895790e+00 1.26656747e+00
-5.98163009e-01 4.99409020e-01 2.07142457e-01 -1.29034567e+00
7.58517206e-01 8.62635821e-02 2.49945492e-01 -3.38071465e-01
5.67960590e-02 2.40264997e-01 5.01674339e-02 -2.59205133e-01
7.46626705e-02 -3.53415847e-01 -9.84596834e-02 2.18316346e-01
-2.77258933e-01 -1.46011218e-01 -1.01749757e-02 -3.40766966e-01
9.39748168e-01 7.52166007e-03 3.89071107e-01 -4.67482716e-01
2.48333007e-01 -2.37436190e-01 8.50308776e-01 1.00849020e+00
-1.23057038e-01 4.16511595e-01 5.82531571e-01 -3.96390080e-01
-8.53924155e-01 -1.07531333e+00 -4.50344831e-01 1.28835571e+00
-2.30316982e-01 -2.03832045e-01 -7.06481040e-01 -7.68155992e-01
3.97183269e-01 1.27050829e+00 -6.57852530e-01 -1.75823465e-01
-3.00830811e-01 -8.41740012e-01 3.97780359e-01 5.02655447e-01
2.05123723e-01 -6.97340727e-01 -4.86468375e-01 2.60182858e-01
-6.94318116e-02 -1.11716378e+00 -3.76903534e-01 2.52809644e-01
-8.38071406e-01 -8.70957732e-01 -5.36447823e-01 -2.25431740e-01
4.48732853e-01 -2.23423660e-01 1.27239752e+00 -3.42091411e-01
8.86940118e-03 1.32871971e-01 -4.93366979e-02 -7.81906128e-01
-2.15403050e-01 1.59516364e-01 1.12877384e-01 1.34088993e-01
4.46275234e-01 -4.56574321e-01 -3.92864436e-01 2.77226090e-01
-8.57642412e-01 -3.26715887e-01 7.15190589e-01 7.14279890e-01
5.39785862e-01 2.15361536e-01 8.45311880e-01 -1.26925159e+00
9.25502658e-01 -6.35230720e-01 -7.73330033e-01 4.16984618e-01
-1.25814462e+00 4.23754036e-01 5.46633005e-01 -4.88861740e-01
-1.33194411e+00 -1.08394533e-01 1.16475768e-01 -3.62295806e-01
-3.34089249e-02 7.01781154e-01 -2.51555562e-01 5.90197407e-02
6.23472273e-01 -1.24379084e-01 -1.98463276e-01 -6.46606147e-01
7.13997260e-02 5.10432959e-01 1.17824741e-01 -9.45702910e-01
6.07443750e-01 1.98990092e-01 -3.80969606e-02 -3.51164460e-01
-1.48290300e+00 -1.67272970e-01 -3.19647849e-01 1.41369566e-01
4.57836926e-01 -8.68840277e-01 -4.53477055e-01 1.79275215e-01
-8.63695443e-01 -4.01490152e-01 -1.92819178e-01 5.27447164e-01
-7.14119971e-01 1.73913002e-01 -5.86731791e-01 -1.21139657e+00
-2.15974107e-01 -1.20831239e+00 7.63530195e-01 -5.55619411e-02
-3.24126422e-01 -1.11168599e+00 -8.79321396e-02 4.47916925e-01
5.70431828e-01 1.21246934e-01 9.85068321e-01 -8.40734184e-01
-3.68946612e-01 -4.51891795e-02 -1.01115316e-01 6.32488787e-01
-8.69368613e-02 -2.21172005e-01 -1.06522548e+00 -2.98965245e-01
2.29798496e-01 -5.47637701e-01 9.44053352e-01 6.04640901e-01
1.47600091e+00 -6.00043774e-01 -2.18275368e-01 5.27408302e-01
1.37224376e+00 1.14089064e-01 1.80543825e-01 4.08125401e-01
2.44013637e-01 8.78418684e-01 9.21204329e-01 7.27256894e-01
4.54675376e-01 7.45571852e-01 3.38289469e-01 2.32633337e-01
3.67047369e-01 -3.20088089e-01 4.41521913e-01 1.59641683e-01
3.34222436e-01 -2.18464300e-01 -1.01896536e+00 4.82191682e-01
-2.04787827e+00 -7.02535272e-01 2.91443765e-01 2.41854048e+00
1.05821526e+00 6.00525022e-01 8.64454359e-03 -9.83073562e-02
4.08089638e-01 1.48779005e-01 -8.26727509e-01 -5.15859008e-01
7.00975135e-02 -2.57925526e-03 7.16125607e-01 6.81459427e-01
-1.24251091e+00 7.16549337e-01 6.91814470e+00 1.02838457e+00
-1.04934359e+00 2.53692456e-02 9.14631963e-01 -2.47341797e-01
-6.47124410e-01 1.77242681e-01 -1.02707565e+00 5.09515166e-01
9.81455684e-01 -3.53725880e-01 3.69848579e-01 9.23126876e-01
2.89270610e-01 -7.51099586e-02 -1.49162054e+00 6.99340940e-01
-6.76779896e-02 -9.94302809e-01 -1.50260746e-01 3.27755287e-02
7.17258036e-01 -3.69044729e-02 3.68716836e-01 3.52028161e-01
6.55963540e-01 -1.32242644e+00 9.09460604e-01 5.08829236e-01
5.25616288e-01 -9.21314299e-01 7.71763802e-01 6.38726532e-01
-7.31311679e-01 -2.52785861e-01 -3.61174464e-01 -1.28948912e-01
7.27045685e-02 1.06452942e+00 -9.26617026e-01 4.25410271e-01
6.33350670e-01 3.98624122e-01 -4.95360374e-01 1.12438202e+00
-1.83049470e-01 1.11086285e+00 -4.07554090e-01 1.38135394e-02
4.27858859e-01 6.33237287e-02 5.01988649e-01 1.36949098e+00
3.03335249e-01 -4.08616126e-01 3.58093053e-01 9.46303129e-01
-2.59659998e-02 4.41132635e-02 -6.16273820e-01 8.66366476e-02
7.17789412e-01 7.13042796e-01 -1.93845853e-01 -1.14881516e-01
-3.52559745e-01 4.81414735e-01 5.48598409e-01 5.39576411e-01
-6.01632237e-01 -1.85218006e-01 8.16814005e-01 3.50045353e-01
2.96098679e-01 -1.55161947e-01 -6.95281982e-01 -1.08337700e+00
3.55524838e-01 -9.87017870e-01 5.59491873e-01 -2.01500446e-01
-1.68003988e+00 1.53462276e-01 2.21996710e-01 -9.37281549e-01
-5.58485210e-01 -6.24748588e-01 -5.22398531e-01 8.44680488e-01
-1.71799648e+00 -8.64594698e-01 3.16587478e-01 3.19428712e-01
6.80474699e-01 -1.10953838e-01 6.04510367e-01 2.72019595e-01
-5.64682543e-01 8.61289263e-01 3.37828040e-01 -2.83712089e-01
9.54670608e-01 -1.21648407e+00 2.48168297e-02 6.71441436e-01
-7.31751174e-02 4.46274787e-01 1.04617476e+00 -4.23612863e-01
-9.29555237e-01 -1.01883936e+00 1.06338835e+00 -5.91626942e-01
6.03641748e-01 -5.22782683e-01 -8.78008842e-01 8.61014307e-01
-2.15286329e-01 -5.71374744e-02 7.84651339e-01 4.57278609e-01
-5.77714086e-01 -3.06592405e-01 -1.33127654e+00 4.76043969e-01
7.02222764e-01 -3.72215271e-01 -5.51973283e-01 3.16705644e-01
7.04746246e-01 -3.03437263e-01 -7.90855110e-01 5.24340868e-01
6.49898767e-01 -8.85043383e-01 8.03898394e-01 -7.97138810e-01
3.68762225e-01 -8.51195976e-02 -4.30751354e-01 -1.27060211e+00
-1.14887640e-01 -6.36233091e-01 -2.25337282e-01 1.33441985e+00
7.81834841e-01 -7.65440106e-01 6.53012216e-01 9.78356004e-01
5.95777594e-02 -9.99477684e-01 -1.06481600e+00 -9.03008819e-01
5.26936293e-01 -7.91228533e-01 3.92319649e-01 8.12000632e-01
-3.58689308e-01 1.90248415e-01 -5.14398992e-01 1.10395998e-01
9.59127843e-01 -6.20209426e-02 4.84646201e-01 -1.33844173e+00
-4.52475041e-01 -5.02730012e-01 5.80342440e-03 -9.66486096e-01
5.86716950e-01 -5.96474051e-01 4.35778871e-02 -1.18313527e+00
7.40499198e-02 -8.63699973e-01 -6.33443177e-01 3.82951230e-01
-3.29819560e-01 -4.40140098e-01 2.69514710e-01 3.11855793e-01
-4.93090332e-01 5.79322994e-01 7.61861801e-01 -1.25711024e-01
-8.16751271e-02 5.42591095e-01 -1.03261673e+00 8.93367827e-01
9.16558862e-01 -5.50224543e-01 -5.95326483e-01 -4.66076076e-01
1.95227623e-01 -1.54960960e-01 4.96836901e-02 -8.46821427e-01
1.23226315e-01 -5.83733499e-01 2.06772909e-01 -3.27933908e-01
3.44594955e-01 -8.68342757e-01 -6.43545156e-03 2.82756060e-01
-8.12055647e-01 -7.40584806e-02 1.61646828e-01 6.68608427e-01
-2.13529259e-01 -4.65943068e-01 8.76545072e-01 6.94162324e-02
-4.07602340e-01 1.50781363e-01 -2.30943069e-01 3.88642579e-01
1.03400433e+00 2.99339771e-01 1.38256699e-03 -5.75697064e-01
-4.42157924e-01 4.19906467e-01 2.45864019e-01 2.92753667e-01
3.11507523e-01 -1.14738989e+00 -8.78985941e-01 -1.21052705e-01
1.49561867e-01 8.68906602e-02 -1.40022680e-01 7.22633600e-01
-5.14112636e-02 5.33998311e-01 1.69028178e-01 -3.39786470e-01
-8.84347558e-01 4.41573471e-01 4.33590591e-01 -7.31270075e-01
-3.95911872e-01 7.03009188e-01 3.07021499e-01 -5.02746224e-01
6.46487474e-01 -3.46540660e-01 5.84428087e-02 1.09860376e-01
4.08302248e-01 4.36690778e-01 3.88066731e-02 3.09135951e-02
-3.27255726e-01 2.19984844e-01 -2.42683113e-01 -2.80194700e-01
1.13660991e+00 -1.89678892e-01 4.25559372e-01 7.54115462e-01
9.39574838e-01 1.18141174e-02 -1.58333826e+00 -6.66678965e-01
4.64352340e-01 -5.02765298e-01 7.81681389e-02 -9.58452344e-01
-7.60130286e-01 1.03949726e+00 3.80994856e-01 -8.61243010e-02
1.00708091e+00 -6.70053139e-02 3.93185765e-01 5.21375716e-01
4.69410717e-01 -1.24868274e+00 -2.26734146e-01 4.15435195e-01
7.94547439e-01 -1.34731531e+00 3.18901092e-02 -2.28082985e-01
-8.80437970e-01 7.33822286e-01 5.80530226e-01 -1.17393218e-01
7.66233861e-01 3.29335809e-01 -8.30330700e-02 2.47433245e-01
-1.06926703e+00 8.50945637e-02 5.15613794e-01 4.62469041e-01
5.04122019e-01 2.76607454e-01 -4.75941807e-01 8.78031373e-01
-1.33504957e-01 -3.58009562e-02 3.82017456e-02 9.10601616e-01
-5.31433344e-01 -1.22030067e+00 -2.18227208e-01 7.35365510e-01
-1.00616550e+00 -1.28457770e-01 -2.98423231e-01 7.40162194e-01
-1.79981172e-01 1.04615104e+00 -1.71945274e-01 -1.79241031e-01
3.53588849e-01 1.87536165e-01 2.24133089e-01 -6.00707471e-01
-2.69380212e-01 -3.60653922e-02 3.80243659e-01 -5.36834657e-01
-4.09250036e-02 -8.21745515e-01 -8.81700814e-01 -2.40294933e-01
-1.14181280e-01 1.33115917e-01 5.78468144e-01 1.05243599e+00
3.33366334e-01 2.28837192e-01 5.54190695e-01 -5.10861278e-01
-1.57557499e+00 -9.92464960e-01 -6.11747980e-01 2.60559440e-01
4.00767803e-01 -9.89288568e-01 -5.55085123e-01 -2.41585508e-01] | [8.505668640136719, 4.130135536193848] |
ef478c13-6c0a-4526-9686-57f785879ecc | dafa-diversity-aware-feature-aggregation-for | null | null | https://doi.org/10.1109/ACCESS.2022.3203399 | https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9874741 | DAFA: Diversity-Aware Feature Aggregation for Attention-Based Video Object Detection | We present a framework for attention-based video object detection using a simple yet effective external memory management algorithm. An attention mechanism has been adopted in video object detection task to enrich the features of key frames using adjacent frames. Although several recent studies utilized frame-level first-in-first-out (FIFO) memory to collect global video information, such a memory structure suffers from collection inefficiency, which results in low attention performance and high computational cost. To address this issue, we developed a novel scheme called diversity-aware feature aggregation (DAFA). Whereas other methods do not store sufficient feature information without expanding memory capacity, DAFA efficiently collects diverse features while avoiding redundancy using a simple Euclidean distance-based metric. Experimental results on the ImageNet VID dataset demonstrate that our lightweight model with global attention achieves 83.5 mAP on the ResNet-101 backbone, which exceeds the accuracy levels of most existing methods with a minimum runtime. Our method with global and local attention stages obtains 84.5 and 85.9 mAP on ResNet-101 and ResNeXt-101, respectively, thus achieving state-of-the-art performance without requiring additional post-processing methods. | ['Ki-Seok Chung', 'Si-Dong Roh'] | 2022-09-01 | null | null | null | ieee-access-2022-9 | ['video-object-detection'] | ['computer-vision'] | [-1.51346266e-01 -5.39100945e-01 -2.52363712e-01 -3.92090306e-02
-5.23810446e-01 -8.12012330e-03 2.05409840e-01 2.03803435e-01
-8.34360063e-01 6.25492275e-01 -4.64734286e-02 7.74770975e-02
-4.19733375e-02 -8.40575576e-01 -5.16183913e-01 -5.64077795e-01
-2.45097175e-01 4.23313119e-02 9.95033801e-01 1.26473397e-01
4.99670327e-01 4.93849337e-01 -1.98166251e+00 1.17396258e-01
4.03992146e-01 1.32252789e+00 7.10749805e-01 8.03960443e-01
-1.97299823e-01 1.19841135e+00 -7.81437516e-01 -1.18631236e-01
3.97202402e-01 -2.43142202e-01 -9.30042446e-01 -3.87386307e-02
6.78294122e-01 -9.01689827e-01 -6.90767467e-01 9.66368198e-01
6.34010434e-01 2.35296056e-01 2.75982060e-02 -1.27950811e+00
-7.47180879e-01 4.84801143e-01 -7.79321551e-01 1.05657232e+00
3.61532509e-01 1.14180714e-01 9.29975986e-01 -1.06761014e+00
5.48393250e-01 1.09063888e+00 3.90164584e-01 2.97702283e-01
-6.35404348e-01 -7.32808590e-01 2.21250862e-01 6.95115328e-01
-1.61813927e+00 -4.95374948e-01 2.52686411e-01 -1.84120238e-01
1.30771005e+00 3.97324026e-01 8.43754828e-01 2.04574645e-01
1.34028628e-01 8.38624537e-01 5.10176659e-01 -1.46321997e-01
2.21796781e-01 -4.22736466e-01 2.61619896e-01 8.02400887e-01
3.98468643e-01 -2.58390069e-01 -8.37428987e-01 -7.70666227e-02
1.10516357e+00 2.91256309e-01 -2.47313082e-01 5.32695698e-03
-1.29497278e+00 6.40945435e-01 7.17397809e-01 2.47225910e-01
-7.93225050e-01 3.73992890e-01 6.42993450e-01 3.29176754e-01
2.48240098e-01 8.00941661e-02 -1.22048154e-01 -2.10201129e-01
-1.14726925e+00 -3.78117487e-02 3.81893128e-01 1.31545997e+00
9.88654494e-01 2.28583500e-01 -3.54584843e-01 5.85237682e-01
3.13654765e-02 4.09638852e-01 4.85083103e-01 -1.06715000e+00
2.69780189e-01 6.18916512e-01 -1.62009060e-01 -1.15424335e+00
-3.65869492e-01 -2.80539572e-01 -7.92947829e-01 2.34426677e-01
8.37442577e-02 1.60197049e-01 -9.63671565e-01 1.49020016e+00
2.34339580e-01 4.28494513e-01 -3.55669051e-01 1.13357699e+00
9.28602159e-01 7.83458889e-01 3.15243870e-01 -3.44694585e-01
1.44334102e+00 -1.20994866e+00 -5.49931765e-01 8.02305415e-02
4.51899022e-01 -7.55116701e-01 7.40435958e-01 4.13526334e-02
-1.57451773e+00 -8.70486498e-01 -9.82231259e-01 -1.87453628e-01
-1.70724422e-01 -8.14612433e-02 5.28914690e-01 3.86365950e-01
-1.49824917e+00 2.29328930e-01 -7.68018663e-01 -5.36647260e-01
5.97395897e-01 6.21663511e-01 -2.28492245e-01 9.24554691e-02
-8.45185339e-01 5.34292459e-01 4.49636161e-01 -3.13672125e-01
-8.90315711e-01 -7.66749799e-01 -5.53336203e-01 4.05885786e-01
5.15232563e-01 -7.01403499e-01 1.02089512e+00 -9.36945379e-01
-1.20467806e+00 4.55968797e-01 -2.12792248e-01 -6.74534917e-01
1.83491170e-01 -4.38065439e-01 -2.97394514e-01 6.24648869e-01
1.94200724e-01 1.18996465e+00 8.12626958e-01 -5.43348789e-01
-1.05672550e+00 -1.53223760e-02 3.81939173e-01 8.91636759e-02
-6.16404295e-01 5.95083952e-01 -8.89794409e-01 -6.24598861e-01
3.20359990e-02 -5.98428607e-01 -2.78119951e-01 1.99225500e-01
5.76728489e-04 -2.81208813e-01 1.00982082e+00 -2.75329798e-01
1.64804590e+00 -2.05882525e+00 -3.62989008e-01 -2.23689321e-02
6.43610179e-01 6.42136395e-01 -1.65851951e-01 -7.74597516e-03
2.96723306e-01 4.80230413e-02 1.50597930e-01 -6.60277009e-02
-4.37013119e-01 -1.36585414e-01 -9.79456156e-02 3.52403879e-01
2.55325377e-01 7.02806115e-01 -1.02391779e+00 -8.96857977e-01
4.50247169e-01 5.45182765e-01 -9.76655483e-01 1.50641486e-01
9.88500863e-02 -1.64399043e-01 -4.35952723e-01 8.42967331e-01
7.28681505e-01 -5.98356307e-01 -6.45936504e-02 -2.92980939e-01
-3.97004932e-01 1.66160733e-01 -1.09555519e+00 1.57972288e+00
-7.63994455e-02 8.27906072e-01 -7.79410377e-02 -6.88835800e-01
7.64879823e-01 1.78475529e-01 8.03417146e-01 -8.47299337e-01
2.76134908e-01 1.00413285e-01 -1.40453596e-02 -3.48672330e-01
1.12025130e+00 6.21833265e-01 3.28535318e-01 2.58399040e-01
5.63159809e-02 9.38357294e-01 5.93253195e-01 3.52647662e-01
1.50003111e+00 -2.46351704e-01 4.31641877e-01 -5.16260386e-01
8.00521195e-01 1.63810793e-02 6.32760942e-01 9.58190739e-01
-5.08044720e-01 6.71684027e-01 9.70116630e-02 -6.54373765e-01
-1.02106690e+00 -7.67098308e-01 1.41297476e-02 1.48151302e+00
4.80303556e-01 -9.25650358e-01 -7.49527097e-01 -5.48949480e-01
-1.73552558e-01 7.43488371e-02 -3.75517428e-01 1.22144945e-01
-8.79914641e-01 -5.21592915e-01 3.60835046e-01 7.98192859e-01
9.88119304e-01 -1.30285072e+00 -1.32966888e+00 4.95545059e-01
-2.97264159e-01 -1.15321684e+00 -6.41887486e-01 -3.49455118e-01
-7.75861561e-01 -1.21491885e+00 -8.31856310e-01 -8.36814761e-01
5.73926210e-01 9.26840842e-01 1.28041172e+00 5.95795751e-01
-4.18696880e-01 3.55623156e-01 -3.62185448e-01 -8.40060320e-03
6.64848611e-02 1.69345021e-01 -2.76621189e-02 -1.72838867e-01
6.86676145e-01 -2.92925119e-01 -1.02543020e+00 4.43983614e-01
-8.48501325e-01 -1.00644222e-02 6.94397986e-01 5.40335000e-01
6.39292121e-01 -3.32392782e-01 6.93503797e-01 -9.77407321e-02
9.13878158e-02 -5.16845763e-01 -5.26190221e-01 8.81922394e-02
-2.10921839e-01 -2.76039630e-01 4.86524045e-01 -3.95303637e-01
-6.36664510e-01 2.79667955e-02 2.57957280e-02 -6.29163980e-01
4.93966937e-02 3.46974842e-02 9.10659656e-02 -2.22680658e-01
1.96359411e-01 4.00710970e-01 -1.70373321e-01 -2.73502976e-01
-5.85326254e-02 5.51645160e-01 5.85649669e-01 -2.57031173e-01
1.93052709e-01 5.22424996e-01 -4.61767726e-02 -1.07997787e+00
-3.99588645e-01 -7.18875110e-01 -4.31314826e-01 -3.80408585e-01
7.97712445e-01 -1.11709952e+00 -9.26266313e-01 4.78315681e-01
-1.11832595e+00 -2.67601144e-02 -1.83366030e-01 5.03354847e-01
-5.34449816e-01 4.76247877e-01 -7.51265645e-01 -5.44852734e-01
-6.66257203e-01 -1.26119423e+00 8.32421184e-01 2.12855175e-01
-1.86894968e-01 -3.46498519e-01 -3.24851781e-01 -6.86502410e-03
8.66475761e-01 -1.28534019e-01 3.90104264e-01 -4.02926952e-01
-1.35420847e+00 8.78094137e-02 -7.72682726e-01 8.59425589e-02
-8.97144675e-02 -1.09444156e-01 -7.11928010e-01 -5.31715810e-01
-2.28485316e-01 -1.65534273e-01 9.54533339e-01 4.03271377e-01
1.34457314e+00 -4.36191931e-02 -2.95654893e-01 5.82173228e-01
1.51886272e+00 2.39619657e-01 9.07581389e-01 6.56563222e-01
6.11886740e-01 -2.28051499e-01 7.15121448e-01 8.12288880e-01
3.41195345e-01 6.17046952e-01 5.87369263e-01 -2.44825929e-02
-5.46294451e-01 2.45425120e-01 4.49298650e-01 9.38877821e-01
-3.14186394e-01 -5.19850552e-01 -6.38538539e-01 8.21676016e-01
-1.80422711e+00 -1.16643429e+00 -6.74367622e-02 2.11008215e+00
5.25696874e-01 1.28988802e-01 3.51848096e-01 7.16429576e-02
1.01220787e+00 3.33128512e-01 -3.58453631e-01 -2.02484846e-01
-7.56001696e-02 8.40228647e-02 7.35901773e-01 8.65981355e-02
-1.12400651e+00 9.10439551e-01 6.56686163e+00 8.68268788e-01
-1.07020855e+00 2.71924436e-01 3.77952248e-01 -4.24376726e-01
4.55905646e-01 -2.25356638e-01 -1.05188572e+00 6.77966058e-01
9.51944709e-01 -1.95632815e-01 4.84407023e-02 9.67142820e-01
-7.50430822e-02 -2.73278207e-01 -7.91871309e-01 1.29212666e+00
2.25433484e-01 -1.52604091e+00 3.61942410e-01 2.82063223e-02
5.30114532e-01 2.67512053e-01 -2.10946262e-01 1.21655852e-01
-1.34721234e-01 -5.13210237e-01 7.44934857e-01 3.70674521e-01
9.42592561e-01 -7.58754015e-01 6.66836619e-01 -1.55017048e-01
-1.74261630e+00 -1.55384466e-01 -7.11346924e-01 -1.89154521e-01
2.83120960e-01 3.92198473e-01 -3.21250468e-01 2.01753870e-01
1.11854148e+00 6.87104285e-01 -6.67333961e-01 1.50451839e+00
4.75490302e-01 4.09611970e-01 -5.18298924e-01 -6.37490451e-02
4.05324131e-01 4.12927598e-01 7.27298796e-01 1.34994507e+00
4.93527383e-01 1.89118609e-01 2.51059979e-01 4.80193406e-01
-3.03468466e-01 1.55903712e-01 -1.26051858e-01 4.81802851e-01
6.10281348e-01 1.19571114e+00 -1.06233978e+00 -8.39206398e-01
-7.66482770e-01 8.83620739e-01 2.43875653e-01 2.06373781e-02
-1.25932205e+00 -6.17533267e-01 8.99870276e-01 3.17864358e-01
8.16765308e-01 -2.52735257e-01 2.10881591e-01 -9.48913753e-01
-9.42931771e-02 -5.28195024e-01 4.97470379e-01 -6.01695597e-01
-7.48062849e-01 9.06508982e-01 -1.12531371e-01 -1.29697764e+00
1.04313076e-01 -3.57320160e-01 -4.10561442e-01 3.17939013e-01
-1.47659016e+00 -6.50842488e-01 -8.32232594e-01 8.51312816e-01
9.35059905e-01 -2.13219523e-01 3.81327391e-01 7.94178486e-01
-6.82757795e-01 7.08659053e-01 -1.61424607e-01 9.66766775e-02
5.93340278e-01 -7.97087967e-01 4.64276642e-01 9.47350442e-01
-1.05099842e-01 5.07162750e-01 1.98779225e-01 -6.54783726e-01
-1.47562850e+00 -1.11456203e+00 6.98814154e-01 7.99555108e-02
4.43094343e-01 7.77920410e-02 -1.03697956e+00 5.19527316e-01
3.97687167e-01 4.93589044e-01 3.92425269e-01 -4.50869769e-01
-1.30607903e-01 -1.22629426e-01 -1.01164711e+00 4.32635814e-01
1.42328167e+00 -2.14261815e-01 -3.49145949e-01 -3.50852981e-02
6.44143462e-01 -2.58322030e-01 -8.01110148e-01 2.36418784e-01
4.83316183e-01 -1.24749863e+00 1.06262374e+00 -2.57087529e-01
1.67951703e-01 -6.44005358e-01 -3.08068603e-01 -4.43044364e-01
-8.86480808e-01 -6.50948763e-01 -5.24953544e-01 1.20236909e+00
-1.02030851e-01 -4.78386521e-01 7.86152959e-01 3.93761426e-01
-1.36041224e-01 -6.05526686e-01 -8.90338719e-01 -8.57961893e-01
-6.17852986e-01 -2.38706052e-01 7.51635015e-01 5.49594045e-01
-2.76820987e-01 3.07776369e-02 -3.25772315e-01 1.43109471e-01
5.88281512e-01 3.24494801e-02 7.60504246e-01 -9.04551506e-01
1.29495785e-01 -5.53983033e-01 -1.06470227e+00 -1.49888015e+00
-2.69454777e-01 -5.99795222e-01 -7.77082741e-02 -1.33869135e+00
5.83234489e-01 -4.60083663e-01 -7.53816366e-01 3.89036328e-01
-2.32171580e-01 7.40068078e-01 6.08945310e-01 3.81333023e-01
-1.54001796e+00 3.89351726e-01 8.59723151e-01 1.63379371e-01
-5.46921641e-02 -6.01181686e-01 -4.01309043e-01 7.69865632e-01
7.66396880e-01 -4.79818493e-01 -2.14313775e-01 -5.93660593e-01
-1.77182123e-01 -2.32272074e-01 5.08877277e-01 -1.59101272e+00
7.12743878e-01 1.00739365e-02 4.89213109e-01 -1.04209220e+00
3.42470586e-01 -8.66569281e-01 4.09811102e-02 5.43507636e-01
-1.67383701e-02 5.89603126e-01 2.60488153e-01 4.86973107e-01
-3.74266535e-01 -6.67316765e-02 7.25336432e-01 -8.17536712e-02
-1.48398340e+00 5.49889028e-01 -5.67786098e-01 1.18476063e-01
1.23204017e+00 -4.25674707e-01 -5.39953709e-01 -5.73941171e-02
-3.13359380e-01 1.95955262e-01 5.10613084e-01 4.60163236e-01
8.74566615e-01 -1.44728792e+00 -7.62657464e-01 4.39626783e-01
-5.02806529e-02 -2.68127114e-01 4.89075363e-01 8.27529490e-01
-8.67964983e-01 5.49362719e-01 -5.88204980e-01 -8.91315162e-01
-1.54922998e+00 8.06163192e-01 -7.51729757e-02 -5.00397123e-02
-8.27566743e-01 7.21255183e-01 3.10515851e-01 6.58631206e-01
2.55751908e-01 -1.74410120e-01 -3.34829912e-02 -2.69774208e-03
1.15269566e+00 9.74684477e-01 -7.14908689e-02 -9.02103126e-01
-6.34526372e-01 6.79576933e-01 -2.26662233e-01 5.13825655e-01
1.03874850e+00 -4.53002602e-01 -1.04184449e-01 -2.85223246e-01
1.14706910e+00 -1.76718712e-01 -1.30988371e+00 -2.66729027e-01
-1.26522854e-01 -9.10265386e-01 3.75985354e-01 -1.54612586e-01
-1.58446074e+00 6.02856398e-01 6.63240731e-01 3.69131505e-01
1.44799054e+00 -9.58458856e-02 1.22053373e+00 2.76645631e-01
5.74502766e-01 -9.81382608e-01 1.84517398e-01 5.78872025e-01
7.82098055e-01 -1.11119401e+00 1.87079698e-01 -3.90305609e-01
-2.35782832e-01 1.10498643e+00 1.02304089e+00 -2.50374079e-01
5.69894552e-01 2.95410931e-01 -2.81456620e-01 -7.35295638e-02
-8.71618092e-01 -4.75362241e-01 4.60246168e-02 3.13518107e-01
1.76492497e-01 -2.80380279e-01 -2.02022493e-01 1.44172117e-01
1.78116649e-01 1.72649652e-01 4.32778239e-01 1.22879481e+00
-9.48020995e-01 -6.19586170e-01 -3.16998392e-01 4.79835719e-01
-7.35144019e-01 -2.20075369e-01 2.75023639e-01 8.27755392e-01
2.70977598e-02 8.19405377e-01 6.04758680e-01 -4.18839902e-01
5.49232624e-02 -2.63690710e-01 3.05437028e-01 -1.66790888e-01
-6.91903591e-01 7.12735280e-02 -1.58448324e-01 -9.65433776e-01
-7.26941943e-01 -3.60308617e-01 -1.28811193e+00 -8.11721087e-01
-5.25500357e-01 -1.55932739e-01 2.97520459e-01 4.46560115e-01
9.05763805e-01 6.23175204e-01 4.03804839e-01 -1.04125261e+00
-7.42607415e-02 -7.42316306e-01 -2.81700015e-01 2.16385514e-01
4.17281985e-01 -6.87836885e-01 -1.46276787e-01 -4.30251248e-02] | [8.99653148651123, -0.09715411067008972] |
6423d5e2-ba40-4fa8-84fe-4888627e19ce | deep-extreme-multi-label-learning | 1704.03718 | null | http://arxiv.org/abs/1704.03718v4 | http://arxiv.org/pdf/1704.03718v4.pdf | Deep Extreme Multi-label Learning | Extreme multi-label learning (XML) or classification has been a practical and
important problem since the boom of big data. The main challenge lies in the
exponential label space which involves $2^L$ possible label sets especially
when the label dimension $L$ is huge, e.g., in millions for Wikipedia labels.
This paper is motivated to better explore the label space by originally
establishing an explicit label graph. In the meanwhile, deep learning has been
widely studied and used in various classification problems including
multi-label classification, however it has not been properly introduced to XML,
where the label space can be as large as in millions. In this paper, we propose
a practical deep embedding method for extreme multi-label classification, which
harvests the ideas of non-linear embedding and graph priors-based label space
modeling simultaneously. Extensive experiments on public datasets for XML show
that our method performs competitive against state-of-the-art result. | ['Wenjie Zhang', 'Hongyuan Zha', 'Xiangfeng Wang', 'Junchi Yan'] | 2017-04-12 | null | null | null | null | ['extreme-multi-label-classification'] | ['methodology'] | [ 1.80953350e-02 2.05211863e-01 -2.52218902e-01 -4.80224758e-01
-8.34874809e-01 -4.45028573e-01 1.38335899e-01 4.89160776e-01
-4.23709244e-01 4.03076023e-01 1.03160292e-01 -2.05177426e-01
-3.16375464e-01 -8.23104084e-01 -3.31130087e-01 -8.93968701e-01
2.10534539e-02 5.82769692e-01 -1.99128792e-01 -7.82167241e-02
5.89123322e-03 2.71068010e-02 -1.32479000e+00 -7.27687627e-02
6.66082919e-01 1.18737268e+00 -1.06914721e-01 1.81843609e-01
-5.46010792e-01 7.20121205e-01 -2.36946698e-02 -7.75911927e-01
3.56929213e-01 -1.61759302e-01 -8.36065352e-01 1.16215006e-01
5.49079061e-01 -1.28986612e-01 -3.34768921e-01 1.33220446e+00
7.05914378e-01 -7.35631911e-03 7.45254099e-01 -1.53689110e+00
-6.83867514e-01 8.67784619e-01 -1.17737627e+00 -3.80311787e-01
2.69137658e-02 -2.47655541e-01 1.36156178e+00 -7.52263725e-01
4.05823261e-01 1.31042540e+00 8.94335389e-01 4.99649733e-01
-1.13261294e+00 -8.94969225e-01 1.33816138e-01 2.46840626e-01
-1.44776809e+00 1.60496324e-01 1.06071246e+00 -3.72109473e-01
3.79761606e-01 9.17098150e-02 4.56027448e-01 9.14293528e-01
5.90955792e-03 7.47227073e-01 1.21780944e+00 -4.00035113e-01
1.30694076e-01 1.05431125e-01 3.65720659e-01 8.65757644e-01
3.17557715e-02 -2.49721512e-01 -1.01505339e-01 -2.54660964e-01
3.74990851e-01 2.29274377e-01 -2.27028765e-02 -3.50780398e-01
-1.15483570e+00 1.22897351e+00 3.95348430e-01 2.72127151e-01
-4.85405438e-02 5.20810485e-01 8.11220348e-01 4.59100604e-01
9.18595970e-01 2.27304408e-03 -4.91374284e-01 2.00266242e-02
-6.34916365e-01 -9.11733210e-02 6.19178593e-01 1.01587069e+00
8.01267028e-01 -3.46924156e-01 2.02498153e-01 1.13701069e+00
5.53423822e-01 1.99777439e-01 1.58181518e-01 -8.96667778e-01
6.12188697e-01 7.65679002e-01 -2.77975202e-01 -1.07390010e+00
-7.68843949e-01 -6.46178126e-01 -1.23745549e+00 -8.57385769e-02
5.65074086e-01 -2.14372322e-01 -3.94777417e-01 1.90438128e+00
6.43767059e-01 2.58811593e-01 -2.61669606e-01 6.56845391e-01
7.11122692e-01 7.09283948e-01 4.03285921e-01 -2.62776047e-01
1.67464614e+00 -1.07068300e+00 -6.50536478e-01 -1.60950422e-01
1.14358413e+00 -4.03043568e-01 1.15695548e+00 1.69290945e-01
-5.97408414e-01 -4.43792045e-01 -7.00801969e-01 -2.65435368e-01
-4.41566974e-01 -1.83472991e-01 1.07273555e+00 6.85552001e-01
-8.30114245e-01 4.64608073e-01 -3.28357786e-01 -3.85912031e-01
5.26663780e-01 4.55392897e-01 -3.43684763e-01 -4.26053822e-01
-1.43673205e+00 3.87786895e-01 5.71645558e-01 -9.93421674e-02
-3.92883867e-01 -5.07790625e-01 -7.91822314e-01 2.00190365e-01
6.58770680e-01 -5.41802466e-01 7.95194924e-01 -4.78186250e-01
-1.15347958e+00 1.01677990e+00 5.26350498e-01 9.28838626e-02
5.62059104e-01 2.59969354e-01 -4.10797626e-01 -8.99991244e-02
1.76802147e-02 7.34211087e-01 6.60176158e-01 -1.12721407e+00
-6.47581875e-01 -7.42162585e-01 2.91897625e-01 2.52163321e-01
-9.45495605e-01 -1.78217337e-01 -2.96235859e-01 -6.15222573e-01
3.26258123e-01 -1.24275243e+00 -3.55860770e-01 1.02613911e-01
-3.38749200e-01 -9.00976360e-01 6.75747871e-01 -4.43952322e-01
1.21637964e+00 -2.25972152e+00 6.17846511e-02 -2.76882462e-02
6.74989164e-01 -3.01346451e-01 -1.92801893e-01 5.39805412e-01
-5.79511374e-02 2.30641514e-01 -1.71078950e-01 -5.61015248e-01
6.05322063e-01 -4.37142335e-05 -3.25855091e-02 7.89777160e-01
-4.48794007e-01 9.02506709e-01 -9.82108414e-01 -1.00535095e+00
-1.51639625e-01 5.69256127e-01 -4.56609637e-01 9.74518061e-02
-3.45247746e-01 3.00556302e-01 -5.67685604e-01 7.01720417e-01
9.04847145e-01 -7.79812813e-01 3.13845873e-01 -3.38164449e-01
9.11198035e-02 -3.56305838e-01 -1.23065662e+00 2.11117435e+00
-4.74824399e-01 1.10623606e-01 -1.60645977e-01 -1.03417087e+00
5.56959569e-01 1.95529014e-01 9.18145955e-01 -3.87383670e-01
3.55319917e-01 1.51471600e-01 -3.84062290e-01 -4.68681902e-01
2.52788544e-01 -5.09020567e-01 -5.05805790e-01 8.71923864e-01
-1.33902147e-01 1.46604672e-01 1.64569452e-01 2.53557086e-01
9.28164005e-01 -1.60716996e-01 -1.04915630e-02 -2.17821568e-01
4.10796970e-01 -3.66336286e-01 7.33370721e-01 3.85366589e-01
-1.50783658e-01 2.27999777e-01 7.21755445e-01 -4.26597834e-01
-7.92037249e-01 -5.78389049e-01 -3.10088545e-01 1.47812176e+00
2.82898039e-01 -5.37858784e-01 -6.89428031e-01 -1.25813949e+00
1.34903610e-01 4.41107154e-01 -6.93276465e-01 -3.65609564e-02
-4.47661549e-01 -1.04030681e+00 2.94528514e-01 4.63398963e-01
5.45680165e-01 -8.61468077e-01 1.13732345e-01 2.21553460e-01
-1.45794779e-01 -8.89528036e-01 -6.74763799e-01 2.12433502e-01
-6.74060285e-01 -8.17586064e-01 -7.96775520e-01 -1.04283810e+00
6.26923740e-01 5.88999353e-02 1.07535648e+00 -5.12811542e-02
-2.61990756e-01 3.30794960e-01 -3.12684506e-01 -9.77834761e-02
-2.00895220e-01 3.48021656e-01 -4.69428375e-02 1.19398586e-01
4.36750919e-01 -5.27852416e-01 -6.62322223e-01 1.48208931e-01
-1.09126365e+00 4.53633070e-02 5.00909686e-01 9.34445381e-01
5.63770890e-01 4.46603388e-01 8.43727469e-01 -1.45872092e+00
5.42655170e-01 -7.75984466e-01 -5.03437161e-01 4.75949019e-01
-1.00031233e+00 1.66206196e-01 5.99270105e-01 -5.08892357e-01
-6.08899236e-01 -3.69756818e-02 -4.13109660e-01 -4.15330157e-02
2.66354103e-02 7.23967373e-01 -3.11267883e-01 -1.56016260e-01
1.09488718e-01 -7.49162138e-02 -1.48105860e-01 -6.48845017e-01
7.06563711e-01 9.13198233e-01 -2.71059334e-01 -4.72773761e-01
5.21335542e-01 3.83667350e-01 3.40339631e-01 -3.32729787e-01
-1.08156729e+00 -5.29692411e-01 -4.94419843e-01 -1.38019964e-01
8.82720470e-01 -8.07643235e-01 -1.02554810e+00 6.57030582e-01
-7.59019792e-01 -1.52072683e-01 -1.53091848e-01 3.88957322e-01
-5.01631856e-01 7.68362343e-01 -1.08649850e+00 -4.22969371e-01
-4.07971978e-01 -1.17724586e+00 1.09300411e+00 -1.31447718e-03
3.78819883e-01 -1.38098204e+00 3.26426715e-01 3.79747242e-01
2.71602422e-01 2.19964385e-01 1.49020052e+00 -5.26914179e-01
-3.43707860e-01 -2.01262623e-01 -6.09358490e-01 1.40125573e-01
1.44161889e-02 -5.95276833e-01 -8.97511244e-01 -7.30577409e-01
-4.65708189e-02 -7.81749606e-01 7.85262167e-01 2.65029296e-02
1.43443787e+00 -6.05647303e-02 -4.29317325e-01 5.88063359e-01
1.71580493e+00 -2.73413002e-01 2.85438925e-01 1.44679487e-01
1.08065927e+00 7.42835462e-01 6.86956167e-01 5.63150823e-01
7.44668186e-01 5.96929967e-01 7.91975260e-01 -1.60537854e-01
5.68798967e-02 -5.04789539e-02 -1.79977700e-01 1.25135982e+00
4.66887683e-01 -4.74164635e-01 -9.69956219e-01 1.74475402e-01
-1.76742697e+00 -6.06606126e-01 -8.90532508e-02 1.81280410e+00
8.96130860e-01 1.94401830e-03 7.06442520e-02 2.32215375e-01
8.90896142e-01 5.20145118e-01 -7.11285472e-01 -6.86970279e-02
1.75559178e-01 -2.52221406e-01 5.67964017e-01 1.07030652e-01
-1.24819326e+00 7.00204074e-01 5.06632900e+00 1.15730155e+00
-8.01268578e-01 4.44092989e-01 7.72422910e-01 2.40602344e-01
-3.93997073e-01 -3.48614827e-02 -9.75225151e-01 7.23218977e-01
8.67893398e-01 1.97409138e-01 2.79324293e-01 9.11341965e-01
-5.62509656e-01 3.63875210e-01 -1.10157204e+00 1.44773030e+00
5.09081297e-02 -9.36168551e-01 -1.84506804e-01 5.07681668e-01
6.30004406e-01 -3.79273407e-02 3.96738648e-01 8.00887346e-01
3.13819021e-01 -7.04441965e-01 2.53021926e-01 7.81637244e-03
1.30259287e+00 -9.85807598e-01 6.22875988e-01 5.96267223e-01
-1.31728995e+00 -3.67282540e-01 -6.36828184e-01 3.08122575e-01
1.23647258e-01 8.52833986e-01 -4.02023524e-01 3.76123488e-01
3.85631531e-01 7.33654737e-01 -6.15838468e-01 7.17420995e-01
1.97679535e-01 4.61803585e-01 -3.24167609e-01 -7.31342509e-02
4.04634088e-01 -4.13402528e-01 -1.09351017e-01 1.07470381e+00
3.91981006e-01 1.17740549e-01 5.09272814e-01 3.76158923e-01
-6.58110440e-01 6.83240652e-01 -6.25290573e-01 -2.20901921e-01
2.85052270e-01 1.56223536e+00 -1.03178549e+00 -6.14983775e-02
-7.75165498e-01 9.55317020e-01 6.79021060e-01 -9.98757128e-03
-1.10140419e+00 -3.46779913e-01 2.48906955e-01 -1.71938036e-02
1.31271426e-02 -1.00052416e-01 1.39160782e-01 -1.18782568e+00
-2.13375866e-01 -6.92069948e-01 6.60261810e-01 -1.32731423e-01
-1.62871230e+00 4.32819545e-01 -4.23828155e-01 -1.27061105e+00
1.06708787e-01 -5.78872025e-01 -5.72810136e-02 4.52627569e-01
-1.49053311e+00 -1.47973359e+00 -2.52209395e-01 3.37990373e-01
3.50136817e-01 9.98681560e-02 9.67222214e-01 8.06175828e-01
-5.80702722e-01 9.69719052e-01 6.34877861e-01 1.70709863e-01
7.28020251e-01 -1.49120796e+00 8.06805938e-02 6.40158588e-03
3.50571871e-01 1.37164146e-01 2.08431229e-01 -4.16438729e-01
-1.46424544e+00 -1.14602888e+00 9.16169107e-01 -2.08194628e-01
7.51598001e-01 -7.46252358e-01 -6.32653415e-01 6.16341650e-01
1.99255541e-01 4.83576596e-01 1.18465984e+00 2.98119217e-01
-4.92488086e-01 -3.69904816e-01 -1.20803690e+00 3.99083495e-01
1.18605375e+00 -6.39001071e-01 6.33938387e-02 9.03084338e-01
9.02795196e-01 -9.83560085e-02 -1.22906017e+00 5.08825958e-01
4.46839333e-01 -6.53202295e-01 8.62708747e-01 -4.60688680e-01
2.03424022e-01 3.18583399e-02 -3.72487426e-01 -1.21164870e+00
-5.10428548e-01 -5.85529581e-02 -7.08282515e-02 1.47937512e+00
2.07170576e-01 -6.46170139e-01 9.81457591e-01 5.99627197e-01
2.37325773e-01 -1.03873122e+00 -9.64979231e-01 -5.96764028e-01
2.50662565e-01 -2.85901427e-01 6.55687034e-01 1.42694008e+00
-1.16918972e-02 4.91728365e-01 -4.56768602e-01 -8.45019519e-02
1.01727736e+00 4.09549534e-01 4.80465740e-01 -1.54075646e+00
-5.19993246e-01 -2.50403166e-01 -4.71509993e-01 -1.16689837e+00
5.98685205e-01 -1.31311727e+00 -3.80950242e-01 -1.54389799e+00
5.50755382e-01 -1.18584597e+00 -7.03105271e-01 4.01541829e-01
-1.73057362e-01 2.92332590e-01 1.54537320e-01 1.47908732e-01
-1.06381452e+00 6.46674693e-01 1.13203394e+00 -4.82960045e-01
3.27925235e-01 -1.76360667e-01 -8.65061522e-01 5.93569875e-01
6.44874334e-01 -8.38362753e-01 -5.19265473e-01 -5.86051881e-01
8.59274209e-01 2.49021381e-01 -1.95088089e-01 -9.18499231e-01
9.07213464e-02 -4.58950736e-02 -1.56712040e-01 -5.05531609e-01
3.76929790e-01 -1.26656735e+00 1.75572038e-01 3.10933292e-01
-5.24029613e-01 4.05952260e-02 -3.52246672e-01 1.08763444e+00
-9.25183296e-02 -4.06240880e-01 7.05336511e-01 -1.81420416e-01
-5.93369007e-01 7.96701550e-01 3.81329119e-01 2.00895458e-01
1.24995637e+00 2.72771806e-01 -4.00181949e-01 -1.77365124e-01
-7.78311849e-01 4.19399321e-01 4.68425065e-01 2.14341253e-01
3.94351900e-01 -1.67179573e+00 -7.99751818e-01 -3.11738551e-02
3.37440133e-01 -3.12878862e-02 4.91683871e-01 7.87959576e-01
-3.29505235e-01 7.85193071e-02 3.89565557e-01 -5.49626946e-01
-1.14570606e+00 9.70857799e-01 -2.18012422e-01 -6.43441796e-01
-6.89417958e-01 1.01465774e+00 2.95240223e-01 -5.70578635e-01
5.58369935e-01 7.51380026e-02 -3.17066342e-01 4.10460144e-01
3.26782823e-01 4.18065727e-01 -2.91680068e-01 -5.99041820e-01
-1.93027914e-01 1.09908605e+00 -3.00346524e-01 2.73219168e-01
1.51891816e+00 -3.20631325e-01 -3.26003224e-01 6.87986553e-01
1.89835489e+00 -1.08744040e-01 -1.12251854e+00 -4.26124692e-01
1.87641680e-01 -3.18586916e-01 1.13373183e-01 -5.08000255e-01
-1.38241625e+00 1.01384699e+00 8.63316715e-01 5.65008223e-01
9.08845901e-01 1.52952701e-01 1.26446033e+00 3.57535571e-01
7.66634822e-01 -1.13900375e+00 2.04126924e-01 3.47290784e-01
3.79918218e-01 -1.53735650e+00 -1.15647040e-01 -4.89603281e-01
-4.02776837e-01 8.48008156e-01 3.89431387e-01 1.18405208e-01
1.13328958e+00 9.16884094e-02 -3.49280722e-02 -4.58200693e-01
-4.14691418e-01 1.22810304e-01 -1.35851800e-01 1.19354516e-01
4.68403846e-01 1.53909355e-01 -4.63320822e-01 4.82119977e-01
3.98403823e-01 -2.21054599e-01 4.77713756e-02 7.57200718e-01
-3.17403346e-01 -1.47490525e+00 -1.48987621e-02 7.85685897e-01
-6.42171621e-01 -1.49202004e-01 1.52032956e-01 5.27585626e-01
3.43230814e-01 7.79877841e-01 -2.71248251e-01 -4.94837910e-01
-1.28251627e-01 1.83038920e-01 4.15417016e-01 -5.39705217e-01
-3.87511015e-01 -2.62242109e-02 -2.24531442e-01 -2.66331971e-01
-2.81158239e-01 -5.43213546e-01 -1.35145891e+00 -4.53545690e-01
-6.48123205e-01 3.18231702e-01 8.95791650e-01 5.41791975e-01
1.90551996e-01 2.26991609e-01 8.93439889e-01 -4.10372913e-01
-8.91440451e-01 -9.87582088e-01 -1.19745243e+00 6.13433063e-01
-6.42215535e-02 -7.33291388e-01 -5.43047428e-01 -3.35001171e-01] | [9.600759506225586, 4.303903102874756] |
439dd7df-d3ee-40ca-bf53-6ff8d35ddd4d | unsupervised-dialogue-topic-segmentation-with | 2305.02747 | null | https://arxiv.org/abs/2305.02747v1 | https://arxiv.org/pdf/2305.02747v1.pdf | Unsupervised Dialogue Topic Segmentation with Topic-aware Utterance Representation | Dialogue Topic Segmentation (DTS) plays an essential role in a variety of dialogue modeling tasks. Previous DTS methods either focus on semantic similarity or dialogue coherence to assess topic similarity for unsupervised dialogue segmentation. However, the topic similarity cannot be fully identified via semantic similarity or dialogue coherence. In addition, the unlabeled dialogue data, which contains useful clues of utterance relationships, remains underexploited. In this paper, we propose a novel unsupervised DTS framework, which learns topic-aware utterance representations from unlabeled dialogue data through neighboring utterance matching and pseudo-segmentation. Extensive experiments on two benchmark datasets (i.e., DialSeg711 and Doc2Dial) demonstrate that our method significantly outperforms the strong baseline methods. For reproducibility, we provide our code and data at:https://github.com/AlibabaResearch/DAMO-ConvAI/tree/main/dial-start. | ['Yongbin Li', 'Fei Huang', 'Min Yang', 'Yuchuan Wu', 'Ting-En Lin', 'Rui Wang', 'Haoyu Gao'] | 2023-05-04 | null | null | null | null | ['semantic-textual-similarity', 'semantic-similarity'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.34265719e-02 5.73149204e-01 -2.23734900e-01 -9.08139944e-01
-1.00125587e+00 -6.96978271e-01 7.53069878e-01 2.47112021e-01
-7.27349147e-02 7.33560324e-01 6.62799954e-01 -2.09807694e-01
3.75310153e-01 -4.91021901e-01 -1.04289964e-01 -4.02703047e-01
4.55676019e-01 8.79873157e-01 4.85862285e-01 -5.22045076e-01
3.86355579e-01 -6.35829389e-01 -8.39566469e-01 2.91346133e-01
1.36986744e+00 7.57699430e-01 2.39374086e-01 4.75335538e-01
-5.30438304e-01 5.68957508e-01 -5.96444845e-01 -3.74144018e-01
-2.47886494e-01 -8.98061693e-01 -1.49569190e+00 2.29283452e-01
1.17653213e-01 -5.85864663e-01 -2.56290525e-01 1.10981762e+00
3.71145636e-01 4.12562191e-01 6.85180426e-01 -1.08035314e+00
-2.67470777e-01 1.06001997e+00 -3.86009246e-01 1.60236716e-01
5.76746285e-01 7.32048005e-02 1.23645973e+00 -7.73368239e-01
6.44098341e-01 1.59470117e+00 2.59233415e-01 5.46539903e-01
-1.18614733e+00 -4.27010775e-01 2.25648329e-01 5.67274354e-03
-9.85053301e-01 -4.31956023e-01 8.42636287e-01 -3.45938921e-01
5.69678664e-01 2.90438354e-01 3.47650647e-01 1.09151649e+00
-4.70149547e-01 1.54032218e+00 1.17794549e+00 -2.75556564e-01
1.21635497e-01 1.71389744e-01 8.28714073e-01 5.07521331e-01
-5.25262892e-01 -6.06660068e-01 -3.56152147e-01 -2.29876384e-01
5.83214343e-01 -2.82102704e-01 -5.43603241e-01 -5.03396988e-02
-1.06220376e+00 1.31206465e+00 1.32136807e-01 3.89192492e-01
-2.47579738e-02 -5.52101910e-01 6.91143870e-01 3.32098454e-01
7.86926270e-01 5.41247129e-01 -4.81411338e-01 -4.35806513e-01
-5.84955633e-01 3.11651289e-01 9.94866371e-01 1.08255363e+00
8.48829448e-01 -2.15174496e-01 -3.53782028e-01 1.25800157e+00
4.56852853e-01 7.70746842e-02 6.03859484e-01 -1.16558290e+00
5.28178632e-01 8.07516813e-01 -1.39460549e-01 -7.65985250e-01
-3.77666414e-01 3.29115659e-01 -6.08521938e-01 -7.85276890e-01
6.53235793e-01 -4.05880064e-01 -5.10815740e-01 1.59727848e+00
6.79520369e-01 -3.78650501e-02 3.30170214e-01 9.14732635e-01
1.59001684e+00 7.95629859e-01 6.93490803e-02 -2.35392839e-01
1.54648829e+00 -1.27992737e+00 -1.03626251e+00 1.55538199e-02
7.34495521e-01 -8.37096155e-01 1.39212847e+00 3.75035778e-02
-9.60344553e-01 -4.93998855e-01 -6.59872532e-01 -3.09567750e-01
-2.16990620e-01 -1.59702897e-01 5.38869977e-01 4.07559574e-01
-7.89268315e-01 2.51933873e-01 -7.75642872e-01 -7.04792261e-01
9.57075506e-02 3.73545773e-02 1.09494783e-01 7.14262798e-02
-1.66776645e+00 5.06447732e-01 6.12318516e-01 -1.51030838e-01
-7.66109347e-01 -4.00248796e-01 -1.02173245e+00 -1.55194670e-01
7.46078730e-01 -3.23811769e-01 1.87966132e+00 -5.89603782e-01
-1.89379525e+00 7.72158921e-01 -3.19844395e-01 -5.52009463e-01
4.89964575e-01 -3.40852290e-01 -2.22534742e-02 2.44039834e-01
2.54892468e-01 9.13076222e-01 2.86051363e-01 -1.18641174e+00
-5.91160476e-01 -3.96185875e-01 3.53315115e-01 6.69869125e-01
-2.47503087e-01 7.24099278e-02 -7.06385911e-01 -5.08179903e-01
3.01557600e-01 -8.37482035e-01 -3.50412577e-01 -7.32925832e-01
-1.01822162e+00 -8.41146350e-01 8.16151381e-01 -8.00302923e-01
1.14002311e+00 -1.86407399e+00 7.35989362e-02 -3.23013872e-01
2.80801922e-01 3.66163114e-03 1.00400396e-01 6.16140485e-01
4.28855687e-01 1.12845488e-01 -5.91471791e-01 -4.01645958e-01
1.40268013e-01 4.17562947e-02 -4.71626848e-01 4.19891477e-02
-1.59227289e-02 9.37621593e-01 -1.06904316e+00 -7.75241196e-01
4.32490893e-02 7.66940229e-03 -3.01982880e-01 6.66571677e-01
-7.55340219e-01 9.12373543e-01 -8.38709176e-01 5.17313123e-01
3.54722917e-01 -3.23741436e-01 4.47968841e-01 -1.00020722e-01
1.33301407e-01 9.80372071e-01 -6.82906449e-01 2.13949966e+00
-8.93881842e-02 7.16276407e-01 1.62909515e-02 -9.69870806e-01
1.07111120e+00 3.91133279e-01 2.51094848e-01 -5.79090655e-01
3.34641755e-01 3.77826653e-02 -1.14754446e-01 -3.66795301e-01
1.04870951e+00 2.12530106e-01 -4.00285959e-01 7.41549194e-01
1.63056239e-01 -4.49311465e-01 2.34215736e-01 6.77771926e-01
6.04228020e-01 8.69104564e-02 2.51199275e-01 -3.53603959e-01
4.19082880e-01 2.27785766e-01 6.13393366e-01 3.99900615e-01
-3.70839149e-01 6.52741611e-01 9.00031865e-01 1.79669932e-01
-5.27117431e-01 -9.79333162e-01 -2.05011144e-01 1.31637299e+00
4.46922064e-01 -6.06742978e-01 -1.05794954e+00 -1.01902676e+00
-2.65609235e-01 9.30509686e-01 -4.47570473e-01 1.83032393e-01
-4.56888974e-01 -5.15692532e-01 5.57721198e-01 2.40551919e-01
8.64010394e-01 -8.99662971e-01 -7.63578191e-02 1.92177355e-01
-7.75704682e-01 -1.04741180e+00 -6.66698933e-01 -2.20123038e-01
-8.90357792e-01 -1.07041276e+00 -7.54347324e-01 -7.50496626e-01
4.61163253e-01 4.57058996e-01 1.08188415e+00 -5.01199625e-02
1.09431468e-01 3.37940991e-01 -5.65044999e-01 -1.53727576e-01
-6.02134645e-01 3.77517402e-01 -1.70137063e-01 -3.38908464e-01
5.80720782e-01 -2.80833274e-01 -6.72320366e-01 5.83562255e-01
-4.94020283e-01 2.79810935e-01 -1.68910280e-01 1.01307130e+00
2.94263452e-01 -3.06699783e-01 7.83131957e-01 -1.48226190e+00
1.12677503e+00 -7.88837075e-01 -3.53484869e-01 2.10240111e-01
-4.57038164e-01 -1.85109153e-01 2.30479941e-01 -2.11123191e-02
-1.43929839e+00 -2.91678429e-01 -2.17153355e-01 4.62050326e-02
-4.88765985e-01 6.38757885e-01 -3.26865375e-01 7.62357712e-01
3.36807430e-01 2.40210161e-01 6.76132217e-02 -6.93456292e-01
5.78244448e-01 9.63188648e-01 4.24638748e-01 -7.87697732e-01
1.23837762e-01 7.63388649e-02 -1.07347691e+00 -9.82385993e-01
-9.67654526e-01 -7.62641788e-01 -7.14292288e-01 -2.10726604e-01
9.09990311e-01 -8.27418804e-01 -4.73570138e-01 4.90131557e-01
-1.32331777e+00 -5.74262857e-01 3.55480821e-04 3.71568203e-01
-5.18384218e-01 6.17198169e-01 -1.00818229e+00 -7.34111309e-01
-3.54614973e-01 -1.09647238e+00 9.41651225e-01 7.33875513e-01
-6.67371452e-01 -1.13369417e+00 1.87629789e-01 8.68193448e-01
4.92407121e-02 -5.22576980e-02 8.42133403e-01 -1.47143006e+00
-3.18831354e-01 1.56653017e-01 -2.08648324e-01 2.59227958e-02
3.47961307e-01 -5.85986525e-02 -1.00515866e+00 -5.53880371e-02
9.12009180e-02 -8.46329749e-01 7.94542789e-01 2.62538105e-01
8.34315598e-01 -3.61621261e-01 -2.09449396e-01 -2.35364679e-02
6.04131281e-01 3.42040449e-01 2.97033250e-01 4.71460633e-03
6.00913942e-01 1.25397420e+00 1.21719790e+00 4.54970121e-01
9.93374705e-01 6.25798345e-01 -5.59511967e-03 -2.29794141e-02
1.00761086e-01 -2.59717256e-01 2.02219516e-01 1.16554749e+00
3.25803429e-01 -3.06354344e-01 -1.14588344e+00 6.47774339e-01
-2.18436956e+00 -6.00709677e-01 -1.87861502e-01 1.76082075e+00
1.30873120e+00 1.71859831e-01 4.61694241e-01 -2.88940728e-01
9.42615271e-01 4.56096262e-01 -7.09885299e-01 -1.92524940e-01
-2.47409493e-02 -4.47437257e-01 -2.10306659e-01 7.74727404e-01
-1.24368334e+00 1.45624995e+00 4.89937830e+00 8.86475086e-01
-6.41681314e-01 2.45205909e-01 1.06446040e+00 3.72150660e-01
-4.80095774e-01 4.47783954e-02 -9.14442718e-01 5.57904363e-01
7.82680631e-01 -3.64627540e-01 -1.77321434e-01 8.53766918e-01
1.76227719e-01 -3.48935634e-01 -9.22066689e-01 4.90348816e-01
-5.38878664e-02 -1.12411964e+00 -1.82233393e-01 -2.65970200e-01
6.60706520e-01 -5.90182319e-02 -2.44477779e-01 6.06750965e-01
8.82403910e-01 -7.61453152e-01 4.07094806e-02 6.59161508e-02
3.90340060e-01 -5.38423359e-01 7.08574474e-01 2.36155406e-01
-9.95278358e-01 5.86680174e-01 -2.68275380e-01 2.40873143e-01
3.27993721e-01 2.18496710e-01 -1.33679771e+00 6.18695378e-01
5.26629329e-01 7.79799759e-01 -3.05255979e-01 5.78235686e-01
-4.65341508e-01 1.00534856e+00 -2.69000292e-01 -1.64437160e-01
4.72447425e-01 -5.12846470e-01 5.37464201e-01 1.31515574e+00
-2.72907287e-01 5.19011796e-01 5.73594034e-01 9.03904021e-01
-1.08712368e-01 3.78817618e-01 -3.78407240e-01 -2.69944400e-01
7.47427464e-01 1.26561499e+00 -1.00203836e+00 -4.04472977e-01
-3.89739454e-01 7.85838664e-01 1.00658759e-01 3.56008947e-01
-5.99999189e-01 -3.87850612e-01 6.63846850e-01 -3.39543462e-01
-5.98008707e-02 -1.96340114e-01 -1.50052041e-01 -1.11306536e+00
-3.18387836e-01 -9.44619417e-01 6.23020589e-01 -3.03178042e-01
-1.33545685e+00 7.18865156e-01 1.18110657e-01 -9.57669139e-01
-6.78854644e-01 8.10158923e-02 -9.34424818e-01 6.52754784e-01
-1.14241815e+00 -9.55939054e-01 -2.86363840e-01 4.95875955e-01
1.35521531e+00 -1.39852032e-01 7.90307879e-01 -1.93626285e-01
-7.18427360e-01 5.94040155e-01 1.46356568e-01 4.66993630e-01
9.21396852e-01 -1.42094541e+00 6.82161748e-01 3.99137139e-01
-8.39247555e-02 6.11316264e-01 8.10876787e-01 -8.38184118e-01
-8.69343042e-01 -7.11399138e-01 6.79051936e-01 -3.07806075e-01
7.29173303e-01 -4.42740113e-01 -1.44929993e+00 5.13330460e-01
9.21721578e-01 -9.17183936e-01 9.68901336e-01 5.50185323e-01
-2.39177153e-01 5.03063738e-01 -8.91558647e-01 7.47596622e-01
7.38602579e-01 -5.53257942e-01 -9.78497207e-01 5.86335957e-01
1.27500987e+00 -6.51509762e-01 -9.72021997e-01 3.21314245e-01
1.86987311e-01 -8.49694073e-01 5.85278213e-01 -6.60342991e-01
5.95369399e-01 1.07816555e-01 -1.66199267e-01 -1.16802216e+00
4.78933960e-01 -8.08434725e-01 6.86669573e-02 1.84496212e+00
5.57347953e-01 -5.43383479e-01 8.13647151e-01 7.56199658e-01
-2.82749474e-01 -8.19565296e-01 -6.68703377e-01 -4.43082422e-01
1.97822258e-01 -1.42673507e-01 5.19703507e-01 1.40401280e+00
5.30607820e-01 8.51400614e-01 -7.35398158e-02 -1.40660688e-01
2.76867270e-01 5.47787249e-01 8.80282700e-01 -1.07482135e+00
-1.27844274e-01 -4.91218597e-01 2.16046184e-01 -1.67555821e+00
4.19379681e-01 -6.07233226e-01 3.95341784e-01 -1.62175000e+00
7.95468464e-02 -4.00416523e-01 2.01321796e-01 3.14071864e-01
-6.03504598e-01 -2.43714720e-01 -9.04223043e-03 4.62295204e-01
-1.08760405e+00 1.07237041e+00 1.07790279e+00 -1.94046766e-01
-4.30724859e-01 2.24905401e-01 -5.97869217e-01 9.37559247e-01
1.17359579e+00 -2.13997871e-01 -6.48690283e-01 -1.74725428e-01
-6.02280855e-01 4.90700066e-01 -1.22888058e-01 -3.78543884e-01
3.29984188e-01 -2.14882582e-01 -2.52515435e-01 -9.03551698e-01
5.07299364e-01 -8.16679746e-02 -6.09688997e-01 1.60829589e-01
-8.75267625e-01 -2.50495404e-01 -3.66325374e-03 5.82718074e-01
-5.48030615e-01 -4.23167497e-01 4.60651875e-01 -2.24234641e-01
-8.05533767e-01 1.10247135e-01 -4.63112175e-01 6.53654456e-01
6.38158858e-01 3.92109007e-02 -5.63783705e-01 -8.51697445e-01
-5.37255347e-01 9.87509966e-01 3.97257298e-01 6.07767701e-01
4.52786505e-01 -1.07044148e+00 -8.43944311e-01 -3.55710447e-01
1.10975809e-01 4.19326127e-01 5.23305416e-01 6.77330971e-01
-6.00880906e-02 6.05006218e-01 8.16949755e-02 -7.10752845e-01
-1.46435249e+00 3.35330032e-02 2.15519313e-02 -2.66859770e-01
-4.87521410e-01 9.73288536e-01 5.88819027e-01 -8.14570606e-01
3.72007996e-01 -5.03733717e-02 -5.39170206e-01 3.80461812e-01
3.94734800e-01 1.78249210e-01 -4.23257083e-01 -5.08587658e-01
1.19867199e-03 1.04709812e-01 -6.50060892e-01 -2.90630192e-01
9.43008006e-01 -7.89575219e-01 -1.02180377e-01 8.46549392e-01
1.04513896e+00 -2.58807153e-01 -1.25008392e+00 -6.84227705e-01
4.74720508e-01 -3.10713232e-01 -3.57501715e-01 -5.94961941e-01
-7.69764423e-01 1.08595157e+00 1.13391384e-01 6.22842491e-01
7.22825348e-01 3.59591752e-01 1.02906370e+00 5.86404324e-01
1.80649571e-02 -1.22962320e+00 2.45325983e-01 9.29343700e-01
9.03919220e-01 -1.73818195e+00 -2.07905471e-01 -8.11579347e-01
-1.24089003e+00 8.57263505e-01 1.13292515e+00 2.59134203e-01
4.22240466e-01 -2.99663037e-01 3.22957128e-01 -3.11453134e-01
-7.01876163e-01 -3.19969118e-01 9.47021767e-02 3.57658297e-01
8.93487811e-01 -3.58652845e-02 -4.30068076e-01 8.94683182e-01
-4.12494212e-01 -6.24632537e-01 5.29742122e-01 7.65335798e-01
-5.93242824e-01 -1.08295119e+00 -2.07535941e-02 3.50258738e-01
-2.49745920e-01 -2.69710809e-01 -8.39868009e-01 7.01575458e-01
-7.42188752e-01 1.26375401e+00 1.29019514e-01 -3.91491383e-01
3.03176474e-02 3.56956393e-01 -1.29772350e-01 -9.36061740e-01
-5.92217028e-01 4.89755034e-01 5.55517852e-01 -2.58416474e-01
-4.13448691e-01 -7.73557067e-01 -1.58509123e+00 -1.87288046e-01
-4.02513564e-01 8.48081708e-01 3.51890385e-01 9.62263763e-01
1.81129336e-01 4.42108154e-01 7.18678057e-01 -3.28173906e-01
-3.44166160e-01 -1.38311946e+00 -4.23579901e-01 1.61060244e-01
-1.80362269e-01 -5.74476838e-01 -7.79545587e-03 -2.71443445e-02] | [12.607833862304688, 8.061488151550293] |
b6810260-9886-4f94-94d8-3ec6e499e680 | detection-and-classification-of-brain-tumors | 2208.13264 | null | https://arxiv.org/abs/2208.13264v1 | https://arxiv.org/pdf/2208.13264v1.pdf | Detection and Classification of Brain tumors Using Deep Convolutional Neural Networks | Abnormal development of tissues in the body as a result of swelling and morbid enlargement is known as a tumor. They are mainly classified as Benign and Malignant. Tumour in the brain is fatal as it may be cancerous, so it can feed on healthy cells nearby and keep increasing in size. This may affect the soft tissues, nerve cells, and small blood vessels in the brain. Hence there is a need to detect and classify them during the early stages with utmost precision. There are different sizes and locations of brain tumors which makes it difficult to understand their nature. The process of detection and classification of brain tumors can prove to be an onerous task even with advanced MRI (Magnetic Resonance Imaging) techniques due to the similarities between the healthy cells nearby and the tumor. In this paper, we have used Keras and Tensorflow to implement state-of-the-art Convolutional Neural Network (CNN) architectures, like EfficientNetB0, ResNet50, Xception, MobileNetV2, and VGG16, using Transfer Learning to detect and classify three types of brain tumors namely - Glioma, Meningioma, and Pituitary. The dataset we used consisted of 3264 2-D magnetic resonance images and 4 classes. Due to the small size of the dataset, various data augmentation techniques were used to increase the size of the dataset. Our proposed methodology not only consists of data augmentation, but also various image denoising techniques, skull stripping, cropping, and bias correction. In our proposed work EfficientNetB0 architecture performed the best giving an accuracy of 97.61%. The aim of this paper is to differentiate between normal and abnormal pixels and also classify them with better accuracy. | ['Harsh Kirty', 'Ranit Sen', 'Gopinath Balaji'] | 2022-08-28 | null | null | null | null | ['skull-stripping'] | ['medical'] | [-5.78489490e-02 2.87505776e-01 3.31899583e-01 -4.17887084e-02
3.61197352e-01 2.16387300e-04 5.19088566e-01 1.80951595e-01
-6.31542146e-01 8.42015147e-01 1.15431167e-01 -4.47615325e-01
2.25274250e-01 -8.58076453e-01 -4.32913929e-01 -8.74445140e-01
-3.07197511e-01 3.30573797e-01 5.45686603e-01 -2.37930968e-01
2.44678766e-01 7.56860912e-01 -1.32293510e+00 2.30395198e-01
7.87966013e-01 1.04038978e+00 3.47199708e-01 6.94290459e-01
-3.41017157e-01 9.93529737e-01 -5.68084836e-01 -4.65690307e-02
2.54618108e-01 -1.06962025e-01 -7.87563562e-01 1.40556842e-02
2.09933450e-03 -1.56911686e-01 -2.69523382e-01 1.18434358e+00
6.16408765e-01 -6.61628880e-03 6.58541501e-01 -1.12619686e+00
-3.72729570e-01 2.83337861e-01 -7.13939369e-01 7.88731754e-01
-2.09108666e-01 1.92012116e-02 -4.95188981e-02 -7.31724381e-01
4.49425548e-01 8.90275061e-01 6.93092823e-01 7.19479799e-01
-5.23734033e-01 -8.93150687e-01 -3.25657845e-01 4.42819655e-01
-1.10229397e+00 -2.51372516e-01 3.20555031e-01 -6.30331874e-01
9.03113425e-01 2.78209955e-01 9.28294659e-01 9.46371794e-01
8.07919204e-01 4.61294383e-01 1.44038045e+00 -2.48542115e-01
1.77658349e-01 3.54638658e-02 1.56988323e-01 6.39895082e-01
2.43752554e-01 8.40429142e-02 -1.67609289e-01 2.02451453e-01
8.04439068e-01 2.15814903e-01 -4.44393843e-01 1.18704922e-01
-1.03164434e+00 5.94261467e-01 6.01425886e-01 7.52280116e-01
-6.04891062e-01 5.61295785e-02 5.75118363e-01 2.21561387e-01
4.31691170e-01 4.69761994e-03 -4.98239130e-01 -4.06519808e-02
-7.62364030e-01 -1.58310920e-01 6.47376001e-01 4.09880966e-01
1.77285507e-01 8.06900784e-02 1.55842230e-01 6.77466094e-01
2.92997211e-01 9.33763981e-02 1.36286438e+00 -1.38106108e-01
4.14452553e-02 6.72538280e-01 -5.08424938e-01 -9.71863687e-01
-6.96117640e-01 -7.08983362e-01 -1.24232876e+00 5.68633199e-01
3.83120179e-01 -1.62427083e-01 -1.56093752e+00 1.32867706e+00
2.64871389e-01 2.40287155e-01 6.26355708e-02 7.71703541e-01
1.11613560e+00 5.24421096e-01 3.96734476e-02 4.48948145e-02
1.30012655e+00 -1.00535476e+00 -7.10340500e-01 -7.31592700e-02
7.18249023e-01 -7.85693467e-01 5.98015308e-01 2.67956316e-01
-9.84652996e-01 -2.53791541e-01 -9.69757557e-01 1.52210802e-01
-7.44376481e-01 -6.37634397e-02 7.69734800e-01 6.82223797e-01
-1.19365311e+00 4.63374674e-01 -1.16673958e+00 -6.86485648e-01
6.43542647e-01 5.57394743e-01 -5.77103913e-01 7.91634470e-02
-8.30834150e-01 1.16777563e+00 3.76015991e-01 9.53960419e-02
-7.47804046e-01 -6.27433836e-01 -3.23959917e-01 -9.02793407e-02
8.81485268e-02 -4.17680174e-01 8.14216912e-01 -1.07595623e+00
-1.05585742e+00 9.11398828e-01 1.93949446e-01 -6.82396472e-01
6.07057750e-01 4.58291709e-01 -4.99552876e-01 -1.05504431e-01
-1.91084474e-01 7.65165687e-01 5.27357876e-01 -7.48004794e-01
-7.32917666e-01 -6.96270525e-01 -2.73708522e-01 7.92895816e-03
-1.35208130e-01 5.43906167e-02 6.93479180e-02 -6.76153481e-01
3.89088184e-01 -8.64157617e-01 -2.43579835e-01 -1.96832363e-02
-3.00455004e-01 1.28223419e-01 1.15546787e+00 -9.84538734e-01
6.70351565e-01 -2.19794035e+00 -1.82951853e-01 2.64977157e-01
5.36770761e-01 2.85914481e-01 2.94977218e-01 -3.23065847e-01
-3.83780271e-01 1.09151125e-01 -7.47388154e-02 2.33414456e-01
-6.92879915e-01 2.37204432e-01 4.72536147e-01 6.81404352e-01
-2.30205134e-01 7.38051653e-01 -3.54928136e-01 -4.54321027e-01
2.18909025e-01 7.24636555e-01 -5.33431955e-02 -1.89919189e-01
3.67553979e-01 5.40990114e-01 -3.09384078e-01 9.37978685e-01
6.95820689e-01 6.48863465e-02 -5.60792148e-01 -2.39175484e-01
-1.14472605e-01 -3.02086771e-01 -8.66049707e-01 1.23316097e+00
-3.02865207e-01 8.61517131e-01 2.87039459e-01 -1.36201441e+00
8.73227477e-01 5.28927147e-01 6.15245461e-01 -6.94663763e-01
6.75632119e-01 4.52527881e-01 7.03102231e-01 -7.78011441e-01
-1.38942217e-02 -1.32227033e-01 6.95354223e-01 1.11902773e-01
-1.17122650e-01 1.99174359e-01 2.08654970e-01 -7.21841902e-02
1.23550642e+00 -5.05194783e-01 1.64167300e-01 -4.65770036e-01
6.74355209e-01 -1.46261573e-01 4.48811263e-01 1.71947703e-01
-4.44125175e-01 3.46266836e-01 4.24588293e-01 -6.39514565e-01
-1.02306843e+00 -7.79837489e-01 -4.14525956e-01 4.65035707e-01
-2.25406602e-01 3.77396971e-01 -7.88561463e-01 -3.98339361e-01
-2.06277266e-01 2.98037082e-01 -8.70851398e-01 -2.22146109e-01
-3.24905545e-01 -1.20356953e+00 4.78687555e-01 5.12443542e-01
1.08527362e+00 -1.16790426e+00 -9.40006554e-01 2.50664204e-01
1.93754777e-01 -7.92349517e-01 2.07837261e-02 4.49579298e-01
-1.16285706e+00 -1.27523017e+00 -1.02022898e+00 -9.44507539e-01
1.01886237e+00 -1.55976325e-01 7.03181267e-01 3.46156508e-01
-7.04611897e-01 -1.40827119e-01 -3.50902140e-01 -8.56103778e-01
-5.34006655e-01 5.95647004e-03 -1.30747393e-01 -1.68130532e-01
3.44232976e-01 -6.26362443e-01 -5.49321592e-01 1.18449450e-01
-9.83307242e-01 1.83628425e-01 8.99914563e-01 7.46780276e-01
2.29498446e-01 3.65620464e-01 4.43885535e-01 -7.26863384e-01
7.70246267e-01 -6.19397402e-01 -2.81726748e-01 -1.46954730e-02
-5.27622342e-01 -2.49508530e-01 4.17715847e-01 -4.21466589e-01
-6.42542601e-01 -5.49198091e-02 -1.90928787e-01 -1.60584569e-01
-3.21808994e-01 4.81421351e-01 2.04454809e-01 -4.58703607e-01
5.89434326e-01 1.64084047e-01 4.23226655e-01 -1.87509045e-01
-4.06842083e-01 8.76500010e-01 5.12511671e-01 1.71141386e-01
2.15127647e-01 4.99933481e-01 2.56500661e-01 -9.92794871e-01
-2.59423822e-01 -3.79718870e-01 -4.56660002e-01 -4.74133343e-01
1.07427979e+00 -5.65809369e-01 -3.51379246e-01 8.53847086e-01
-8.45428467e-01 -2.29782239e-01 -5.17715625e-02 6.43697739e-01
8.09963644e-02 8.42859074e-02 -7.39954770e-01 -4.95469391e-01
-7.27641284e-01 -1.35813928e+00 2.54371554e-01 3.27380240e-01
8.66393894e-02 -1.01513076e+00 -3.37595493e-01 2.57028520e-01
9.11709607e-01 5.16075373e-01 1.01712847e+00 -7.38904059e-01
-2.71760941e-01 -2.96584457e-01 -1.86912268e-01 4.62626696e-01
2.36877099e-01 1.09737128e-01 -7.92413950e-01 -2.13683829e-01
1.78849876e-01 -6.54886896e-03 8.76152396e-01 7.24803150e-01
1.26428795e+00 -1.02856159e-02 -4.61922467e-01 6.95237398e-01
1.31115294e+00 7.08104014e-01 9.00130510e-01 6.16165817e-01
4.79887128e-01 4.65417475e-01 1.45112336e-01 1.40583575e-01
-1.39111802e-01 1.45532647e-02 7.73825228e-01 -4.89021838e-01
-3.53283167e-01 6.34710312e-01 3.00633032e-02 9.46606815e-01
-4.40523714e-01 -4.32277694e-02 -1.13951135e+00 5.90934455e-01
-1.34939086e+00 -6.70322716e-01 -4.20898676e-01 2.02681088e+00
6.55552328e-01 1.68840885e-01 -1.40161380e-01 5.37823558e-01
7.25705028e-01 -5.36323369e-01 -4.67305988e-01 -4.24526542e-01
-4.02351953e-02 4.30107921e-01 9.92548168e-01 1.32132962e-01
-9.54733074e-01 4.70284671e-01 5.76666021e+00 5.75419366e-01
-1.91727722e+00 3.10586333e-01 9.76715446e-01 3.28959636e-02
3.76529843e-01 -5.25775135e-01 -4.28383827e-01 6.78145170e-01
8.00358117e-01 4.97973673e-02 2.36777425e-01 6.74269438e-01
9.67813581e-02 -3.99087965e-01 -5.63038170e-01 9.07467842e-01
-4.37599868e-02 -1.12972355e+00 -3.97410005e-01 1.37306839e-01
5.17463386e-01 3.75570804e-01 -6.63486794e-02 2.63363361e-01
-2.42400512e-01 -1.32966697e+00 3.07107419e-01 4.40434098e-01
4.59302604e-01 -7.61640370e-01 1.20930552e+00 3.41238886e-01
-7.33289480e-01 -2.47943223e-01 -3.79718274e-01 -1.13722663e-02
-1.76620349e-01 6.81590796e-01 -1.08757091e+00 -3.83316502e-02
9.94183362e-01 4.18019474e-01 -5.76921642e-01 1.39935052e+00
1.58227533e-01 4.90896612e-01 -3.98925066e-01 -1.37658909e-01
3.11864108e-01 7.16519654e-02 3.77079070e-01 9.46820796e-01
5.59555709e-01 3.37932929e-02 -1.37300253e-01 6.72374427e-01
7.22236633e-02 4.14942026e-01 -4.81108546e-01 1.42406508e-01
-2.72132922e-02 1.47442055e+00 -1.28970337e+00 -3.75056505e-01
-4.45679873e-01 6.47580564e-01 -2.30557203e-01 -5.09838089e-02
-7.58799493e-01 -2.93164194e-01 1.98523492e-01 3.96664798e-01
-1.13656349e-01 -2.48570051e-02 -3.68853360e-01 -8.49813819e-01
-2.58244812e-01 -6.24795735e-01 6.26057312e-02 -6.62830949e-01
-8.12237442e-01 8.49030614e-01 -2.26266071e-01 -6.85751736e-01
2.21893802e-01 -7.96250403e-01 -7.83421397e-01 8.48905563e-01
-1.26119268e+00 -8.53567302e-01 -9.16355550e-01 7.22223699e-01
5.52052200e-01 -4.25475717e-01 6.68477237e-01 4.98315930e-01
-5.99142075e-01 2.74576604e-01 -8.46910924e-02 1.49756834e-01
3.93675685e-01 -1.16184604e+00 -4.25100848e-02 7.42161095e-01
-5.69186211e-01 1.73571616e-01 6.18169129e-01 -7.15216875e-01
-6.71111286e-01 -1.13989842e+00 5.24317145e-01 4.59421307e-01
5.90367377e-01 -1.80123940e-01 -9.87739086e-01 4.89072651e-01
2.62604624e-01 3.96060437e-01 3.89741123e-01 -4.37498391e-01
3.99008989e-01 -3.29544321e-02 -1.59124863e+00 4.63630825e-01
4.86968726e-01 1.03814207e-01 -3.84879440e-01 5.45021713e-01
1.90629438e-01 -4.68376249e-01 -7.99062371e-01 4.68011469e-01
3.74136090e-01 -1.03637505e+00 7.62109637e-01 -2.54037917e-01
3.30309957e-01 -6.90107569e-02 1.83180064e-01 -1.44097269e+00
-1.78141847e-01 5.76882921e-02 1.70866147e-01 7.53064156e-01
4.86343563e-01 -8.83047462e-01 1.01584673e+00 5.94103098e-01
-3.45324516e-01 -1.00989664e+00 -8.83342445e-01 -6.15765750e-01
9.35020670e-02 -3.41433026e-02 4.91665155e-01 1.03110480e+00
-3.04431826e-01 8.74645263e-03 2.18584225e-01 -6.61511496e-02
3.72630984e-01 -7.32347488e-01 2.53941804e-01 -1.35942566e+00
1.58515722e-01 -5.03677130e-01 -1.01440334e+00 -1.60808250e-01
-1.81819528e-01 -9.34103549e-01 -3.62375051e-01 -1.57748055e+00
4.25774418e-02 -5.98821938e-01 -3.61838222e-01 6.96834683e-01
3.30296159e-01 2.91772902e-01 -1.55052677e-01 6.59944862e-02
2.71465451e-01 1.01650909e-01 1.54968154e+00 -1.49257451e-01
-2.48914976e-02 3.17653641e-02 -2.56221980e-01 8.32397163e-01
1.11549866e+00 -2.29126588e-01 -4.19447333e-01 -2.92491645e-01
-1.00465797e-01 -9.00727361e-02 3.30997974e-01 -1.38722730e+00
4.34734404e-01 2.19287828e-01 7.71027863e-01 -6.12510562e-01
1.15357250e-01 -9.85101402e-01 1.82098031e-01 9.00002360e-01
2.07574219e-02 2.91950643e-01 1.97894529e-01 2.15845052e-02
-2.02341959e-01 -4.30546761e-01 1.07670772e+00 -5.95440567e-01
-8.03783298e-01 3.45922142e-01 -7.70531774e-01 -3.80237728e-01
1.55405724e+00 -5.55803537e-01 -2.30389267e-01 -1.78307787e-01
-9.46176708e-01 -9.75518376e-02 1.95787683e-01 1.64906025e-01
5.41043103e-01 -9.13878739e-01 -5.85979700e-01 2.98369110e-01
-2.73953259e-01 2.89588034e-01 1.56859159e-01 1.43591881e+00
-1.11411679e+00 2.64021903e-01 -9.00950730e-01 -5.30889094e-01
-1.54515350e+00 2.52188772e-01 7.96197593e-01 -8.63993615e-02
-6.81178868e-01 1.01973653e+00 -5.58346361e-02 -1.85348913e-01
4.58037436e-01 -7.86742687e-01 -6.03714824e-01 -2.43358910e-01
5.18827856e-01 4.52762276e-01 4.50101912e-01 -5.42047679e-01
-2.47798830e-01 1.39616355e-01 -2.86680073e-01 3.20917457e-01
1.29265726e+00 2.30872005e-01 -5.65090835e-01 5.57346232e-02
1.13816941e+00 -3.94664913e-01 -5.81209719e-01 2.63952464e-01
-1.51301011e-01 -2.58503467e-01 7.23731518e-01 -1.01043570e+00
-1.78329110e+00 8.18743885e-01 1.06859899e+00 2.27708697e-01
1.24904513e+00 -2.94646084e-01 9.00333881e-01 3.94840958e-03
1.98449969e-01 -1.01568961e+00 -2.51329333e-01 4.74705547e-01
6.47023976e-01 -1.24549699e+00 -2.13157997e-01 -3.23203236e-01
-2.54784584e-01 1.31939197e+00 7.67562985e-01 -1.81426257e-01
8.57847929e-01 6.86524987e-01 1.75165519e-01 -3.65373135e-01
-6.47391140e-01 9.97352973e-02 -5.30542135e-02 6.97679877e-01
7.38407075e-01 3.10690068e-02 -5.45890331e-01 3.62142414e-01
-3.46969068e-01 2.44933769e-01 8.04443896e-01 1.13517416e+00
-5.39085925e-01 -7.67755747e-01 -5.95990896e-01 9.79915142e-01
-9.86436546e-01 -5.50789908e-02 -1.61824420e-01 9.66317773e-01
5.92230380e-01 6.27385020e-01 2.02003166e-01 -2.35431671e-01
7.69586861e-02 -1.92552522e-01 4.80778009e-01 -2.64555126e-01
-7.01272070e-01 -9.17803124e-02 -2.06777886e-01 -2.63976187e-01
-3.93455148e-01 -5.57814658e-01 -1.50876307e+00 -2.29875401e-01
-3.51845711e-01 -1.23977974e-01 1.35622001e+00 9.39279735e-01
-8.00573602e-02 1.01482093e+00 2.10290521e-01 -4.77813005e-01
-2.94189632e-01 -1.38878608e+00 -7.85715282e-01 3.16044211e-01
1.69363409e-01 -7.60609150e-01 -2.89834201e-01 3.93160648e-04] | [14.827262878417969, -2.593012571334839] |
65576461-dad5-41e9-83f9-c8844c5dcb23 | multi-scale-geometry-aware-transformer-for-3d | 2304.05694 | null | https://arxiv.org/abs/2304.05694v1 | https://arxiv.org/pdf/2304.05694v1.pdf | Multi-scale Geometry-aware Transformer for 3D Point Cloud Classification | Self-attention modules have demonstrated remarkable capabilities in capturing long-range relationships and improving the performance of point cloud tasks. However, point cloud objects are typically characterized by complex, disordered, and non-Euclidean spatial structures with multiple scales, and their behavior is often dynamic and unpredictable. The current self-attention modules mostly rely on dot product multiplication and dimension alignment among query-key-value features, which cannot adequately capture the multi-scale non-Euclidean structures of point cloud objects. To address these problems, this paper proposes a self-attention plug-in module with its variants, Multi-scale Geometry-aware Transformer (MGT). MGT processes point cloud data with multi-scale local and global geometric information in the following three aspects. At first, the MGT divides point cloud data into patches with multiple scales. Secondly, a local feature extractor based on sphere mapping is proposed to explore the geometry inner each patch and generate a fixed-length representation for each patch. Thirdly, the fixed-length representations are fed into a novel geodesic-based self-attention to capture the global non-Euclidean geometry between patches. Finally, all the modules are integrated into the framework of MGT with an end-to-end training scheme. Experimental results demonstrate that the MGT vastly increases the capability of capturing multi-scale geometry using the self-attention mechanism and achieves strong competitive performance on mainstream point cloud benchmarks. | ['Xuan Tang', 'Arafat Al-Jawari', 'Jian Yang', 'Zhengyu Li', 'Shing-Ho Jonathan Lin', 'Muyu Wang', 'Xian Wei'] | 2023-04-12 | null | null | null | null | ['3d-point-cloud-classification', 'point-cloud-classification'] | ['computer-vision', 'computer-vision'] | [-4.84560788e-01 -3.43767732e-01 2.51981080e-01 -2.81542033e-01
-7.37200260e-01 -3.69626731e-01 3.89927477e-01 3.36812995e-02
1.00602381e-01 6.24002144e-02 -1.11234911e-01 -4.84044757e-03
-2.98541337e-01 -1.05166507e+00 -1.05278146e+00 -7.10903466e-01
-1.41795576e-01 6.11388087e-01 4.00180459e-01 -4.21264857e-01
3.47278178e-01 9.93664384e-01 -1.66010845e+00 1.14853740e-01
9.84275401e-01 1.21231246e+00 4.11092579e-01 5.43477893e-01
-3.10679704e-01 1.83909968e-01 -2.59893358e-01 -1.19074367e-01
3.45135629e-01 1.84689105e-01 -4.99838918e-01 -6.65891543e-02
5.67230999e-01 -2.02724621e-01 -5.80435753e-01 9.19199109e-01
4.64741439e-01 1.63522914e-01 4.22377676e-01 -1.21263337e+00
-1.00596082e+00 2.51262151e-02 -7.24474072e-01 4.37818766e-01
6.51935786e-02 2.84175396e-01 1.02230215e+00 -1.40427125e+00
2.75817811e-01 1.36343241e+00 6.99759543e-01 2.85892785e-02
-7.57259071e-01 -8.28973472e-01 4.58736479e-01 2.17954740e-01
-1.72624540e+00 3.49778943e-02 1.08810484e+00 -2.32891545e-01
1.23096943e+00 2.33703420e-01 8.37806523e-01 5.91319978e-01
1.86725512e-01 5.29634774e-01 6.27234042e-01 1.65752381e-01
-7.16869682e-02 -2.35001072e-01 -1.38995528e-01 5.50909638e-01
3.43880616e-02 -4.92139906e-02 -2.73831695e-01 -1.83449864e-01
1.34644067e+00 5.83863258e-01 -5.77534810e-02 -4.80924904e-01
-1.43097866e+00 8.14686358e-01 1.14055693e+00 4.60076034e-01
-5.74141383e-01 1.89730287e-01 3.02392133e-02 1.29846871e-01
7.17723250e-01 4.52634782e-01 -5.24935007e-01 4.88695456e-03
-6.42453134e-01 3.28442276e-01 2.53883779e-01 1.45936263e+00
1.09174168e+00 -2.07425639e-01 -1.74999341e-01 6.91581190e-01
3.01084965e-01 7.95400143e-01 5.19164503e-01 -4.96325970e-01
5.44080257e-01 1.02701950e+00 -1.15581028e-01 -1.45346475e+00
-4.50726509e-01 -6.46087110e-01 -8.99963439e-01 -1.07338047e-02
-2.63934016e-01 3.42017740e-01 -6.97680295e-01 1.34034419e+00
6.92406237e-01 5.86249650e-01 -1.91330701e-01 1.07755947e+00
1.05734849e+00 8.32447052e-01 -1.66325256e-01 1.07331658e-02
1.17096186e+00 -1.05517769e+00 -1.12420462e-01 1.78308822e-02
2.07952231e-01 -6.83193088e-01 1.13377607e+00 -2.27247104e-01
-1.14341152e+00 -7.45359719e-01 -9.50732052e-01 -3.99931967e-01
-4.97816056e-01 -2.97124416e-01 5.88647068e-01 -1.48625687e-01
-8.64832461e-01 6.88200057e-01 -8.52736235e-01 -1.65014803e-01
6.35877490e-01 4.28804040e-01 -9.72561985e-02 -7.37499073e-02
-8.04288924e-01 1.50447041e-01 -8.88122618e-02 1.15692325e-01
-3.67118448e-01 -1.09842181e+00 -8.71891558e-01 3.46672714e-01
8.76848772e-02 -8.04441869e-01 9.69787061e-01 -6.01530075e-01
-1.23421621e+00 5.53440571e-01 -3.31160218e-01 1.29767403e-01
3.74104351e-01 -1.78310156e-01 -3.06191027e-01 1.27645388e-01
5.50303161e-01 5.85886836e-01 9.12431180e-01 -1.20158029e+00
-5.44572115e-01 -6.92909002e-01 -6.01227582e-02 5.71874797e-01
-9.01530012e-02 -1.13259032e-01 -8.22630763e-01 -7.10721970e-01
5.53863466e-01 -9.10634458e-01 -1.70160621e-01 -1.58557937e-01
-2.23198131e-01 -4.90961909e-01 9.85718012e-01 8.37385133e-02
1.00474572e+00 -2.27593374e+00 1.84167679e-02 1.48481429e-01
3.80612761e-01 -9.11060199e-02 -3.31790119e-01 2.82650679e-01
-2.58976996e-01 1.22243010e-01 -1.92523569e-01 -3.34054798e-01
-1.98083505e-01 1.78714618e-01 -5.45089483e-01 5.00814915e-01
6.14564061e-01 1.32645154e+00 -1.04836679e+00 -4.92200613e-01
5.42217493e-01 5.65464914e-01 -5.55906475e-01 2.85079211e-01
-2.31001839e-01 4.35694307e-01 -9.15436566e-01 8.91314864e-01
1.07252479e+00 -6.67836130e-01 -7.05004811e-01 -1.17455319e-01
-3.69841725e-01 -4.37343679e-02 -8.90408337e-01 1.88424850e+00
-5.30787706e-01 1.10110268e-01 -3.79235864e-01 -4.52118576e-01
1.06565666e+00 -1.72861770e-01 9.44083929e-01 -7.11398184e-01
7.21781328e-02 1.64884910e-01 -1.76817775e-01 -3.33166480e-01
5.02579391e-01 9.96409729e-02 5.41720688e-02 9.65450332e-02
-1.42040178e-01 -3.06682050e-01 -2.64641553e-01 6.81575984e-02
9.96667147e-01 -6.38285726e-02 -9.68597978e-02 -2.44472235e-01
6.81813180e-01 -1.16431350e-02 4.90820378e-01 3.78783435e-01
1.14249110e-01 1.20297039e+00 8.02626163e-02 -7.73820043e-01
-1.04590142e+00 -1.01435030e+00 -3.19660217e-01 9.18968499e-01
8.03777158e-01 -4.63744700e-01 -4.05330032e-01 -6.06897891e-01
4.35112327e-01 2.74456978e-01 -7.63467312e-01 -1.72114030e-01
-6.24727368e-01 -6.74300849e-01 -4.62325588e-02 7.50469267e-01
5.76689899e-01 -1.22672129e+00 -4.66081798e-01 3.29815388e-01
1.79728165e-01 -1.11221993e+00 -6.62721574e-01 -2.16665879e-01
-9.86405492e-01 -1.03910947e+00 -6.61444128e-01 -6.22039855e-01
7.20736027e-01 9.90203440e-01 1.21703112e+00 3.66891801e-01
-1.29176110e-01 3.02524447e-01 -4.16222155e-01 -5.49327672e-01
4.19866443e-01 2.68517435e-01 -4.38311473e-02 3.15550953e-01
5.14604151e-01 -9.12041783e-01 -9.57530081e-01 5.61535716e-01
-7.27525830e-01 -1.01381689e-01 8.26988757e-01 6.51079357e-01
1.24013138e+00 -7.21636638e-02 3.71851206e-01 -4.32059020e-01
2.40612596e-01 -8.22132945e-01 -4.63281333e-01 6.93679154e-02
-5.35588503e-01 -2.02312037e-01 6.74150527e-01 -3.63484830e-01
-5.47940135e-01 -7.05846697e-02 -9.71749723e-02 -1.09187233e+00
2.04242840e-02 2.19162494e-01 -3.71133119e-01 -5.13316810e-01
2.22437531e-01 4.03874248e-01 -1.76714107e-01 -4.72216070e-01
2.61590272e-01 4.39438730e-01 5.35800815e-01 -6.25077784e-01
1.15317798e+00 6.29808307e-01 -5.47206402e-02 -8.47376168e-01
-8.03678393e-01 -6.55788243e-01 -7.70539105e-01 -3.72943804e-02
8.29240263e-01 -1.02396560e+00 -6.88332379e-01 5.70332170e-01
-1.15636671e+00 -1.10282592e-01 -4.80665714e-01 2.20363662e-01
-5.47954857e-01 2.61448652e-01 -3.33023548e-01 -4.65657830e-01
-5.89598536e-01 -1.17995620e+00 1.89809275e+00 5.14236808e-01
3.79185468e-01 -7.48676956e-01 9.65432748e-02 1.11979418e-01
4.44392622e-01 2.24556640e-01 7.43356884e-01 -3.86549294e-01
-1.17040181e+00 -2.39976212e-01 -6.65552080e-01 -1.52738720e-01
1.25948399e-01 -4.47826646e-02 -7.25362420e-01 -3.98000896e-01
-1.35674560e-02 2.20574383e-02 4.42953825e-01 2.14425594e-01
1.71713126e+00 2.31285323e-03 -3.40180457e-01 1.24275529e+00
1.61524093e+00 -8.97847116e-02 4.31919724e-01 3.99837404e-01
1.27979338e+00 1.05882131e-01 5.95984876e-01 3.92750800e-01
8.42041314e-01 6.16074800e-01 7.99932599e-01 -1.15053177e-01
7.59511441e-02 -4.08050537e-01 -1.66371956e-01 1.15300381e+00
-1.32956013e-01 7.95750991e-02 -9.31093454e-01 6.00785792e-01
-1.98126996e+00 -8.02624047e-01 -2.17124179e-01 2.03884578e+00
1.52084053e-01 -6.28620619e-03 -1.04057141e-01 -4.44978893e-01
5.95057428e-01 2.69076675e-01 -1.13089073e+00 4.14619260e-02
-2.53829092e-01 1.69326052e-01 5.81728399e-01 6.33403212e-02
-1.03454614e+00 1.07906592e+00 5.09489107e+00 7.55606771e-01
-1.40770555e+00 4.02381569e-02 3.86419445e-01 -2.45437548e-01
-3.76677990e-01 -1.74975023e-01 -6.08736813e-01 6.53618872e-01
4.92344111e-01 -2.88860649e-01 4.31019902e-01 1.09530985e+00
-1.07262500e-01 5.02867460e-01 -7.35484779e-01 1.29729462e+00
7.26108847e-04 -1.38708687e+00 9.85457525e-02 2.01680392e-01
8.20075989e-01 7.17664599e-01 1.89348817e-01 3.83961946e-01
7.56970420e-02 -9.22334254e-01 7.49022424e-01 5.73414147e-01
9.24760103e-01 -1.00083756e+00 6.61543906e-01 2.50899911e-01
-1.63843977e+00 -1.39296547e-01 -7.95002699e-01 1.49379387e-01
-1.23645127e-01 4.41726178e-01 -2.73938835e-01 9.16463733e-01
1.14389813e+00 1.06667387e+00 -7.98979580e-01 1.09090292e+00
2.55665332e-01 8.67023095e-02 -5.46191394e-01 2.19757147e-02
5.78464687e-01 -4.99446392e-01 7.83474803e-01 8.52711976e-01
6.20581210e-01 2.68402159e-01 8.54221508e-02 1.03730083e+00
2.78989673e-02 3.16001415e-01 -6.97689652e-01 3.12357783e-01
5.34974992e-01 1.38507569e+00 -6.78270698e-01 -2.73965687e-01
-6.51165605e-01 8.00546348e-01 5.19694805e-01 2.08669901e-01
-8.44248176e-01 -2.78925389e-01 1.05648744e+00 3.29910547e-01
5.70490897e-01 -3.58732849e-01 -4.28399473e-01 -1.20792603e+00
4.12500918e-01 -5.80120087e-01 1.96439132e-01 -1.06033683e+00
-1.60272801e+00 1.02861619e+00 -1.64323449e-01 -1.53350461e+00
2.37527013e-01 -3.30511272e-01 -1.02246058e+00 9.74263787e-01
-1.65562034e+00 -1.52837348e+00 -9.67211425e-01 8.23656619e-01
6.49878263e-01 -1.79403469e-01 6.04893684e-01 1.12122834e-01
-5.59482932e-01 5.35295486e-01 1.33826137e-01 -2.92564686e-02
3.38442594e-01 -1.36501229e+00 9.73628283e-01 5.21132112e-01
1.12179756e-01 8.07131588e-01 1.17217779e-01 -7.70472348e-01
-1.65052295e+00 -1.57271993e+00 4.25844252e-01 -6.27758920e-01
6.78301334e-01 -4.38909292e-01 -1.35219300e+00 4.31258351e-01
-3.01292539e-01 6.87673032e-01 2.69058615e-01 -2.37201527e-02
-3.60775888e-01 -1.41434684e-01 -8.62895072e-01 3.86038542e-01
1.43213654e+00 -5.47457516e-01 -4.91704702e-01 6.17628813e-01
1.08255208e+00 -7.68136621e-01 -1.08727825e+00 7.46173680e-01
1.46518365e-01 -8.97708535e-01 1.16205716e+00 -5.53681910e-01
4.30482447e-01 -4.72114027e-01 -6.64378777e-02 -1.13782370e+00
-8.69993508e-01 -5.89129686e-01 -1.41450942e-01 1.00519240e+00
1.15199290e-01 -6.49199486e-01 9.40840840e-01 4.28936690e-01
-5.06932199e-01 -1.26939523e+00 -1.04051900e+00 -7.03232586e-01
2.81747371e-01 -4.02622342e-01 1.46689403e+00 8.97694409e-01
-5.38147271e-01 3.54421169e-01 1.78274035e-01 5.53000033e-01
5.65242708e-01 8.81476939e-01 9.17979062e-01 -1.35261297e+00
1.23006433e-01 -6.18836701e-01 -7.33552396e-01 -1.37351131e+00
1.83019433e-02 -9.01741683e-01 -2.29395658e-01 -1.29096639e+00
-4.04878519e-02 -7.54529595e-01 -3.84219259e-01 -4.47103120e-02
-3.03442001e-01 7.13909343e-02 1.20090805e-01 5.49954712e-01
-7.57676780e-01 1.10058856e+00 1.37864733e+00 -6.98051155e-02
-2.72305280e-01 -2.01389454e-02 -6.33791625e-01 6.15922630e-01
6.02585137e-01 -3.14411312e-01 -2.99404770e-01 -6.94231033e-01
1.31684914e-01 -2.11657599e-01 5.18972933e-01 -1.23018026e+00
4.75574553e-01 -1.64755121e-01 5.06063461e-01 -1.08996713e+00
2.95680404e-01 -1.07841635e+00 -7.22588822e-02 -2.00258896e-01
2.39259213e-01 5.60198605e-01 1.03319108e-01 8.27828109e-01
-2.58326471e-01 4.47132200e-01 7.41932571e-01 -1.23528630e-01
-5.26433408e-01 1.27633142e+00 5.93441188e-01 -6.36410788e-02
1.14004910e+00 -2.74094105e-01 -4.00440730e-02 1.45513779e-02
-4.07607257e-01 5.47868013e-01 8.04956436e-01 8.92602086e-01
8.09798002e-01 -1.71829009e+00 -5.96266508e-01 6.36259139e-01
4.91314530e-01 1.02722454e+00 6.62286758e-01 5.86661339e-01
-5.85015833e-01 3.32736671e-01 -1.59614637e-01 -1.06141424e+00
-7.96503961e-01 8.03284585e-01 3.54807466e-01 1.99965648e-02
-1.04396796e+00 8.74478340e-01 5.33411801e-01 -4.88869578e-01
-2.30356410e-01 -5.38219154e-01 -2.19733194e-02 -3.04847717e-01
4.06211853e-01 1.05254769e-01 1.58690304e-01 -1.02538562e+00
-3.62640858e-01 1.28761613e+00 -1.29444063e-01 4.25681353e-01
1.58079398e+00 -7.21275732e-02 -1.70037925e-01 4.47994441e-01
1.39068627e+00 1.09669985e-02 -1.39643717e+00 -5.18687010e-01
-4.01229054e-01 -7.91042566e-01 1.25322238e-01 -2.07327217e-01
-1.35485995e+00 8.12577665e-01 3.98837358e-01 3.00920784e-01
9.65645075e-01 1.47799969e-01 1.08481228e+00 2.08893746e-01
4.59130287e-01 -6.62766218e-01 8.34790617e-03 6.58416629e-01
1.24736750e+00 -1.38111699e+00 -1.76711738e-01 -4.23485875e-01
-4.67084527e-01 9.34860349e-01 9.84130859e-01 -5.87859809e-01
9.30242538e-01 3.41613515e-04 -1.66774601e-01 -7.14022458e-01
-6.51028037e-01 -1.91109866e-01 5.25008857e-01 3.81727397e-01
3.22386669e-03 7.02364789e-03 4.86177474e-01 6.50406837e-01
-4.79479283e-01 -3.35178733e-01 -5.98903038e-02 6.15781009e-01
-2.28836000e-01 -6.39073014e-01 -3.37855786e-01 5.23431778e-01
3.99394110e-02 -1.33403480e-01 -2.76391476e-01 7.02225268e-01
2.81500995e-01 5.38035631e-01 5.21722794e-01 -5.74086249e-01
6.73326194e-01 -4.01291609e-01 1.02820650e-01 -5.78737497e-01
-5.99529922e-01 1.52502097e-02 -8.72547507e-01 -9.59149241e-01
1.50290020e-02 -7.77762532e-01 -1.29567754e+00 -2.91819960e-01
-1.00776486e-01 1.71994925e-01 6.39835060e-01 6.90609455e-01
1.00204945e+00 5.84442675e-01 1.07476187e+00 -1.11049926e+00
-3.35898727e-01 -9.45746005e-01 -3.07109743e-01 5.60870528e-01
4.41187054e-01 -8.68250906e-01 -2.29355931e-01 -5.51069319e-01] | [7.935511589050293, -3.472709894180298] |
0e6cb3ba-bfe6-4fad-96da-54aa41d2a70d | low-resource-spoken-language-identification | 2106.00052 | null | https://arxiv.org/abs/2106.00052v1 | https://arxiv.org/pdf/2106.00052v1.pdf | Low-Resource Spoken Language Identification Using Self-Attentive Pooling and Deep 1D Time-Channel Separable Convolutions | This memo describes NTR/TSU winning submission for Low Resource ASR challenge at Dialog2021 conference, language identification track. Spoken Language Identification (LID) is an important step in a multilingual Automated Speech Recognition (ASR) system pipeline. Traditionally, the ASR task requires large volumes of labeled data that are unattainable for most of the world's languages, including most of the languages of Russia. In this memo, we show that a convolutional neural network with a Self-Attentive Pooling layer shows promising results in low-resource setting for the language identification task and set up a SOTA for the Low Resource ASR challenge dataset. Additionally, we compare the structure of confusion matrices for this and significantly more diverse VoxForge dataset and state and substantiate the hypothesis that whenever the dataset is diverse enough so that the other classification factors, like gender, age etc. are well-averaged, the confusion matrix for LID system bears the language similarity measure. | ['Nikolay Mikhaylovskiy', 'Roman Bedyakin'] | 2021-05-31 | null | null | null | null | ['spoken-language-identification'] | ['speech'] | [-3.91552180e-01 -2.55744923e-02 -1.06698208e-01 -5.12605608e-01
-8.29195261e-01 -6.98773623e-01 8.65211785e-01 -1.14582507e-02
-9.08145905e-01 7.26960123e-01 5.35087526e-01 -5.62378883e-01
3.08269352e-01 5.44795096e-02 -2.70573169e-01 -3.77836436e-01
1.88530758e-01 8.42566192e-01 -2.91396290e-01 -4.75297332e-01
-1.79961011e-01 5.32904208e-01 -1.07723773e+00 2.01336041e-01
4.74179417e-01 6.68155015e-01 -1.71255335e-01 8.35829258e-01
-6.65900558e-02 9.08838868e-01 -6.36433363e-01 -3.74021918e-01
2.25303307e-01 -9.89397317e-02 -1.18754113e+00 -3.82570893e-01
7.44161487e-01 -1.68519676e-01 -8.52729857e-01 8.75140488e-01
9.53279674e-01 2.00950474e-01 5.48761010e-01 -8.65767360e-01
-6.45099640e-01 1.22702098e+00 -3.52674723e-02 4.74014819e-01
3.22469711e-01 3.75430249e-02 7.80207813e-01 -9.62431729e-01
6.52600288e-01 1.61841297e+00 6.62181139e-01 8.55404258e-01
-8.56518924e-01 -7.37014830e-01 -2.95925420e-02 -7.68579915e-02
-1.57920980e+00 -1.03814816e+00 2.55856454e-01 -3.45376164e-01
1.13646877e+00 3.64341766e-01 9.79573727e-02 1.33782041e+00
-5.15673816e-01 9.04693902e-01 9.64666784e-01 -3.58757049e-01
6.13550246e-02 5.78414559e-01 5.23784161e-01 4.43746954e-01
-8.96667391e-02 -2.56119400e-01 -7.63612211e-01 -1.47046298e-01
2.02002391e-01 -3.93123060e-01 -5.09885512e-02 2.71866173e-01
-1.38051522e+00 6.42143786e-01 -4.54415232e-02 7.00733721e-01
1.18384413e-01 -1.60471559e-01 7.66797423e-01 3.07224900e-01
6.08898997e-01 5.86520731e-01 -5.78386188e-01 -1.96637154e-01
-9.67815518e-01 -2.09285483e-01 7.92221308e-01 6.21716499e-01
4.06913579e-01 3.46972644e-01 -3.92285973e-01 1.43289685e+00
2.12598443e-01 4.79144365e-01 9.86755371e-01 -5.02283096e-01
4.96241361e-01 5.05238414e-01 -1.88100532e-01 -1.98386267e-01
-3.93941462e-01 -3.14090639e-01 -8.49038064e-01 -3.47306252e-01
6.29529178e-01 -3.90077025e-01 -9.28874969e-01 1.72688711e+00
-1.80031970e-01 2.39620376e-02 4.58548248e-01 6.96242571e-01
1.22786236e+00 4.00179029e-01 1.94213107e-01 1.46355271e-01
1.45093310e+00 -7.72810757e-01 -5.18825352e-01 -3.90190482e-01
8.41101706e-01 -7.57038534e-01 8.53560567e-01 -4.76541854e-02
-7.49881446e-01 -5.24386585e-01 -1.04133356e+00 -3.01103473e-01
-7.02259779e-01 6.06638968e-01 4.53270912e-01 9.54726279e-01
-1.43266380e+00 1.04669444e-01 -4.17958707e-01 -9.89487529e-01
8.97935107e-02 5.19501269e-01 -8.48696291e-01 4.65382226e-02
-1.35475075e+00 1.05838621e+00 4.42387342e-01 1.45893693e-01
-7.71681845e-01 -6.37543976e-01 -7.99186766e-01 -2.00204536e-01
-4.43182178e-02 6.42931983e-02 1.17411101e+00 -7.72585213e-01
-1.56378078e+00 1.50679862e+00 -2.54514277e-01 -7.04259276e-01
6.08202934e-01 -1.79193050e-01 -6.46859109e-01 -2.95470178e-01
-1.90804884e-01 6.39824808e-01 3.43478650e-01 -5.99622428e-01
-3.67273003e-01 -6.02311969e-01 -3.06930900e-01 1.56082988e-01
-3.68909478e-01 7.27614641e-01 -2.75981903e-01 -3.01275104e-01
-2.08395422e-01 -9.23285067e-01 -4.28285450e-02 -7.31882989e-01
-4.70203042e-01 -6.22211099e-01 6.67695642e-01 -1.25110090e+00
1.13879561e+00 -2.55384302e+00 -1.35286450e-01 7.08431005e-02
2.21113428e-01 5.90610981e-01 -8.12290534e-02 2.48876050e-01
-2.20826730e-01 3.43496323e-01 8.08243677e-02 -6.60286307e-01
1.37316678e-02 -1.64655164e-01 -2.90601581e-01 5.45543611e-01
1.15023725e-01 8.08231831e-01 -5.14402151e-01 -1.54626951e-01
1.61723003e-01 6.22385859e-01 2.18469314e-02 1.92620486e-01
2.51704663e-01 3.87935549e-01 1.01108290e-01 4.14428085e-01
4.48938042e-01 1.33493334e-01 3.09559945e-02 3.80118862e-02
-3.62365693e-01 5.90455174e-01 -8.90976310e-01 1.41825199e+00
-4.90195990e-01 1.05012202e+00 2.26909637e-01 -6.87377393e-01
1.35281467e+00 5.52845120e-01 5.67798615e-02 -6.44465327e-01
3.54605585e-01 5.62734187e-01 2.42060214e-01 -3.95324044e-02
7.25615621e-01 2.95627534e-01 -1.25278294e-01 2.89689720e-01
4.43636626e-01 7.88712204e-02 -1.19130619e-01 2.50801295e-01
9.54721808e-01 -5.92354119e-01 2.22145930e-01 -5.33033490e-01
8.83099735e-01 -3.82394612e-01 3.61139566e-01 8.99497390e-01
-6.99699998e-01 8.14047813e-01 2.25350574e-01 -4.52056766e-01
-1.25979924e+00 -8.34624469e-01 -2.81557560e-01 1.21177423e+00
-6.65112376e-01 -2.36652777e-01 -8.16207826e-01 -4.14857566e-01
-1.14834405e-01 5.36957622e-01 -5.24385750e-01 -1.37864918e-01
-5.48044026e-01 -8.11041772e-01 1.41519177e+00 3.08019936e-01
4.99821454e-01 -9.46961045e-01 3.26102853e-01 4.65120673e-02
-1.20168505e-03 -1.45211101e+00 -6.07380092e-01 2.47675955e-01
-1.58745393e-01 -6.50215626e-01 -1.04209399e+00 -9.97760236e-01
1.45303085e-01 -1.54094532e-01 9.66993868e-01 -1.43989608e-01
-2.57569939e-01 3.49443913e-01 -4.74291556e-02 -3.42741430e-01
-7.22815335e-01 5.97059190e-01 7.00006485e-01 3.10483873e-01
8.61610293e-01 -3.50174941e-02 -2.10376382e-01 1.93205222e-01
-1.02033988e-01 -3.34447533e-01 1.76671281e-01 5.82694829e-01
-1.93092297e-03 -5.82382917e-01 6.89230144e-01 -5.55314481e-01
5.74261785e-01 -3.64084184e-01 -4.76797044e-01 6.47243023e-01
-1.51427582e-01 4.11255024e-02 4.61277068e-01 -4.43452448e-01
-7.68009841e-01 1.28851920e-01 -2.42215902e-01 -2.52994299e-01
-3.16738755e-01 9.87301767e-02 -2.70983040e-01 2.25967392e-01
5.67259908e-01 4.01598454e-01 -9.00907442e-02 -6.14571095e-01
2.52817988e-01 1.54805589e+00 7.02563465e-01 -3.93932968e-01
3.98566782e-01 1.37737766e-01 -7.05887318e-01 -1.41571164e+00
-3.58058453e-01 -6.92152500e-01 -7.46450245e-01 2.25080010e-02
9.74936247e-01 -1.27914739e+00 -8.67650330e-01 8.09396386e-01
-1.13116872e+00 -2.68194646e-01 1.20393047e-02 6.96092129e-01
-7.48722628e-02 -9.25585851e-02 -6.03251040e-01 -1.23979247e+00
-5.59140325e-01 -1.30653667e+00 8.10358107e-01 2.41361380e-01
-5.99988818e-01 -9.11252260e-01 7.65966177e-02 4.55046177e-01
6.03136480e-01 -5.63986063e-01 5.60438752e-01 -1.70000160e+00
-2.30770502e-02 -2.68382192e-01 -2.78614491e-01 5.01727581e-01
3.86182107e-02 3.82244550e-02 -1.44546247e+00 -3.64744186e-01
-5.51282704e-01 -4.87987727e-01 8.94329131e-01 1.48474231e-01
6.68151498e-01 -4.86017987e-02 2.28893813e-02 4.22273189e-01
6.89529359e-01 2.17280939e-01 2.84052253e-01 6.79843128e-02
9.91462588e-01 7.96536863e-01 -3.47411394e-01 1.15386672e-01
4.76515025e-01 7.28137136e-01 -6.64514422e-01 -3.55102010e-02
-2.60439575e-01 -1.96115837e-01 5.76607585e-01 1.05957901e+00
3.25546086e-01 -1.06524058e-01 -1.43042219e+00 7.82462120e-01
-1.54349995e+00 -6.78285837e-01 1.42389491e-01 2.31241393e+00
8.25634599e-01 -3.29048812e-01 3.91837507e-01 -1.45725265e-01
9.05525088e-01 1.79264545e-01 -3.95696431e-01 -6.94837511e-01
-8.27207863e-01 1.17553383e-01 6.72284484e-01 6.99225008e-01
-1.01977468e+00 1.22508597e+00 6.83120441e+00 8.62176180e-01
-1.42319655e+00 1.29289016e-01 1.01738310e+00 -9.27029550e-02
5.10945134e-02 -4.46401358e-01 -1.26043987e+00 3.16331059e-01
1.49395227e+00 -3.91619742e-01 6.87180996e-01 7.37450838e-01
1.42658412e-01 9.40333456e-02 -1.22234201e+00 1.30685627e+00
4.03738678e-01 -1.00656497e+00 -1.95255399e-01 -1.52057335e-01
2.89766788e-01 8.51439595e-01 3.46605241e-01 4.60070670e-01
7.02268362e-01 -1.54344845e+00 5.29251814e-01 3.36105786e-02
8.62880528e-01 -7.27573395e-01 8.41711760e-01 2.91681848e-02
-8.24108779e-01 -3.56666297e-02 -1.89025074e-01 2.47247323e-01
-1.45227075e-01 1.15927875e-01 -1.19065309e+00 9.44775343e-02
6.93470716e-01 5.88567197e-01 -9.10772264e-01 6.29902482e-01
1.53390095e-01 8.78503084e-01 -4.80913430e-01 -1.25555024e-01
2.15699732e-01 1.03236996e-01 4.17306989e-01 1.34601319e+00
-1.96045302e-02 -3.75048012e-01 -6.05248362e-02 3.97746444e-01
-4.74954277e-01 4.63050336e-01 -6.12136781e-01 -4.76077348e-01
5.82114637e-01 1.08439696e+00 -3.73199433e-01 -3.10115367e-01
-2.62080789e-01 8.44559848e-01 5.17563105e-01 3.24382454e-01
-1.66686177e-02 9.36880521e-03 9.89873469e-01 -2.10653096e-01
-2.27542192e-01 -2.46620372e-01 -3.18764418e-01 -1.21494341e+00
-3.48227739e-01 -1.05854011e+00 3.32601666e-01 -5.18230200e-01
-1.38606381e+00 9.53913867e-01 -4.33661073e-01 -6.93187952e-01
-4.33195114e-01 -7.40045726e-01 -3.81296217e-01 1.19490135e+00
-1.21220541e+00 -1.28239691e+00 2.55194843e-01 6.00307465e-01
6.27918482e-01 -1.08724856e+00 9.35018420e-01 5.47038853e-01
-1.01280868e+00 1.23738921e+00 2.71449000e-01 8.32825601e-01
9.51206088e-01 -1.03949893e+00 6.82482779e-01 8.43302488e-01
2.96013027e-01 7.75733650e-01 3.79205376e-01 -3.96218389e-01
-1.08482575e+00 -9.44036722e-01 1.27964807e+00 -7.52454460e-01
8.89158249e-01 -9.18779254e-01 -8.43347609e-01 6.97201431e-01
3.76501888e-01 -1.23338275e-01 6.42916679e-01 4.36310261e-01
-7.56345332e-01 -6.72052726e-02 -9.62259591e-01 5.45334041e-01
7.36068487e-01 -1.23187053e+00 -3.79158467e-01 3.80109936e-01
6.95989430e-01 -1.64041370e-01 -6.72859550e-01 6.85864361e-04
5.68951249e-01 -4.68387097e-01 8.13532293e-01 -9.59351420e-01
1.91876497e-02 -2.76848041e-02 -4.95993376e-01 -1.20178747e+00
6.66036010e-02 -7.75007606e-01 4.44465101e-01 1.71088767e+00
7.09852517e-01 -6.43544614e-01 4.86158043e-01 6.87264085e-01
1.12782769e-01 -1.10739015e-01 -1.27839851e+00 -8.16396058e-01
3.15997571e-01 -3.45721990e-01 4.79449004e-01 1.13640809e+00
-1.08554170e-01 5.98266125e-01 -4.34752166e-01 1.07989684e-01
2.66609758e-01 -8.85996163e-01 5.44397116e-01 -1.19391894e+00
1.16256401e-01 -3.93926620e-01 -5.29693246e-01 -2.67100424e-01
7.34426200e-01 -1.12207508e+00 -4.13057245e-02 -1.04859102e+00
1.46118149e-01 -3.51372570e-01 -4.39909965e-01 5.41200101e-01
-3.77979577e-02 2.35064343e-01 2.21921429e-01 2.74477564e-02
-6.19804323e-01 1.69503853e-01 2.79664248e-01 -2.93648124e-01
-1.81615919e-01 -2.08244801e-01 -5.49320698e-01 4.71162647e-01
8.24704170e-01 -1.23275362e-01 -1.05779562e-02 -5.26933312e-01
-1.61879674e-01 -2.16843471e-01 5.32222167e-02 -7.96843648e-01
2.68241674e-01 3.40151072e-01 2.32383415e-01 -3.87423694e-01
3.42819601e-01 -3.52057457e-01 -1.07467897e-01 2.24384993e-01
-6.70361519e-01 7.53207505e-02 3.26546371e-01 -2.68780917e-01
-2.70211309e-01 1.53680652e-01 9.03675079e-01 -9.32688341e-02
-6.72908843e-01 3.01911831e-01 -5.88466644e-01 3.70594114e-01
4.31218088e-01 -1.26497615e-02 -4.38448578e-01 -2.77427435e-01
-5.94228327e-01 2.37219036e-01 2.02466950e-01 1.01914644e+00
1.18262069e-02 -1.13965476e+00 -1.21396506e+00 3.57148290e-01
4.04951394e-01 -6.52559578e-01 3.01520377e-01 6.56282902e-01
-2.92863369e-01 7.80563414e-01 -8.19674432e-02 -3.62346739e-01
-1.48451269e+00 1.17795773e-01 5.72768986e-01 -1.03199184e-01
-1.49441317e-01 1.13655639e+00 -7.22501203e-02 -9.98046815e-01
5.17507315e-01 1.25001863e-01 -4.84577745e-01 4.14501637e-01
9.49605048e-01 2.87791401e-01 3.03831249e-01 -1.16382277e+00
-7.66409099e-01 1.22845411e-01 -4.25631344e-01 -4.04178441e-01
1.19775832e+00 -2.58747011e-01 -1.59988284e-01 8.37597013e-01
1.38175833e+00 5.07465340e-02 -5.02623439e-01 -2.63265133e-01
7.12746009e-02 1.67143747e-01 1.52208835e-01 -7.97125816e-01
-7.14383483e-01 1.04717827e+00 9.26287770e-01 -1.36744052e-01
3.86422992e-01 1.48878917e-01 4.51161534e-01 5.94621181e-01
1.17168818e-02 -1.29241669e+00 -5.43340564e-01 1.10990942e+00
8.38161528e-01 -1.49991012e+00 -4.97165233e-01 8.62439722e-02
-6.74820006e-01 7.88565457e-01 6.46619320e-01 2.81788826e-01
4.92059022e-01 2.56037444e-01 5.16777217e-01 2.64069229e-01
-5.25138617e-01 -1.40783921e-01 3.11847925e-01 6.48286462e-01
7.23724842e-01 2.96457499e-01 -1.18306451e-01 6.65308833e-01
-6.77041590e-01 -5.15688539e-01 4.20142353e-01 8.52734968e-02
3.40496190e-02 -8.38793755e-01 -2.93178141e-01 1.99658304e-01
-5.46018362e-01 -4.22335595e-01 -9.35205460e-01 5.10658741e-01
-2.75840521e-01 9.97300208e-01 4.03984576e-01 -4.56793547e-01
4.36896123e-02 5.48948407e-01 5.28841391e-02 -6.51867211e-01
-9.53665257e-01 -3.91424298e-01 4.89417672e-01 -1.86703280e-01
6.68042079e-02 -6.25017703e-01 -9.21649635e-01 -4.06956226e-01
2.60611713e-01 1.43064424e-01 9.30977523e-01 1.06543994e+00
2.81643897e-01 1.64475098e-01 4.80403304e-01 -4.36226726e-01
-3.37879688e-01 -1.27719080e+00 -6.27653956e-01 3.21355432e-01
4.90792215e-01 -2.63284951e-01 -4.20180589e-01 -3.79563063e-01] | [14.161019325256348, 6.677275657653809] |
f41f584d-0b30-4b66-b97c-b1b86119ea7a | roweisposes-including-eigenposes-supervised | 2006.15736 | null | https://arxiv.org/abs/2006.15736v1 | https://arxiv.org/pdf/2006.15736v1.pdf | Roweisposes, Including Eigenposes, Supervised Eigenposes, and Fisherposes, for 3D Action Recognition | Human action recognition is one of the important fields of computer vision and machine learning. Although various methods have been proposed for 3D action recognition, some of which are basic and some use deep learning, the need of basic methods based on generalized eigenvalue problem is sensed for action recognition. This need is especially sensed because of having similar basic methods in the field of face recognition such as eigenfaces and Fisherfaces. In this paper, we propose Roweisposes which uses Roweis discriminant analysis for generalized subspace learning. This method includes Fisherposes, eigenposes, supervised eigenposes, and double supervised eigenposes as its special cases. Roweisposes is a family of infinite number of action recongition methods which learn a discriminative subspace for embedding the body poses. Experiments on the TST, UTKinect, and UCFKinect datasets verify the effectiveness of the proposed method for action recognition. | ['Benyamin Ghojogh', 'Fakhri Karray', 'Mark Crowley'] | 2020-06-28 | null | null | null | null | ['3d-human-action-recognition'] | ['computer-vision'] | [ 9.61446390e-02 -1.54192105e-01 -2.78938621e-01 -2.58014709e-01
-2.56262273e-01 -6.87603205e-02 4.84612733e-01 -1.16864741e+00
-2.11155429e-01 3.69672060e-01 6.99052989e-01 8.60775039e-02
-3.49752486e-01 -2.29248002e-01 -1.75511256e-01 -9.28067207e-01
-8.22952092e-02 3.17610681e-01 -2.36991718e-01 -2.31600851e-01
2.33448848e-01 6.59321964e-01 -1.40818739e+00 2.92491674e-01
4.35612470e-01 8.89613688e-01 -5.17752945e-01 4.19456452e-01
-9.80849117e-02 6.77530885e-01 -2.74538457e-01 -3.04000020e-01
5.08581340e-01 -6.54975355e-01 -6.40027940e-01 4.85939920e-01
2.89595693e-01 -6.51727319e-01 -5.73384464e-01 8.02355945e-01
6.71738982e-01 2.28288978e-01 1.17937052e+00 -1.52851653e+00
-4.43887532e-01 2.13948533e-01 -8.71708751e-01 2.50074446e-01
6.58074975e-01 -1.87004104e-01 5.29886246e-01 -9.28524971e-01
1.52459979e-01 1.63524413e+00 6.86159849e-01 7.73228705e-01
-7.22255409e-01 -7.18133092e-01 -2.55589247e-01 7.19719291e-01
-1.33954442e+00 -5.50431311e-01 9.20401573e-01 -5.81896663e-01
1.02423489e+00 1.28424779e-01 5.15005291e-01 1.24363232e+00
2.38092601e-01 1.21896780e+00 1.16765594e+00 -5.25724530e-01
7.72808641e-02 -2.75134653e-01 2.28957713e-01 6.92405641e-01
-1.30414009e-01 -7.07677975e-02 -6.46044552e-01 -6.61470592e-02
8.99277329e-01 2.74513364e-01 -2.68685427e-02 -5.24825990e-01
-1.18312895e+00 5.35786211e-01 8.42982437e-03 2.60155350e-01
-2.93338925e-01 -2.32006133e-01 3.70667130e-01 1.01686381e-02
1.85760334e-01 -2.05517560e-01 -1.31571263e-01 -4.65470046e-01
-5.05549669e-01 -1.99024513e-01 6.39877141e-01 3.89990509e-01
4.19593602e-01 4.15719062e-01 1.77328270e-02 8.70831907e-01
5.78148782e-01 9.26378608e-01 8.76467347e-01 -8.33833992e-01
3.42622042e-01 9.99550402e-01 -1.35904700e-02 -1.35720944e+00
-4.23415244e-01 1.36440784e-01 -1.17183423e+00 3.89329344e-01
1.87450677e-01 1.15943840e-02 -9.34340417e-01 1.28744555e+00
4.42855507e-01 4.67119515e-01 2.24540293e-01 1.00263333e+00
7.33887672e-01 3.40984166e-01 -7.86037743e-02 -1.75132915e-01
8.88035953e-01 -5.95758259e-01 -7.89738417e-01 -2.76243258e-02
3.93989265e-01 -5.06454647e-01 6.62550151e-01 4.66035783e-01
-3.02078843e-01 -6.49444103e-01 -1.04169416e+00 2.55015850e-01
-2.17660978e-01 7.05118179e-01 8.16666424e-01 1.03970683e+00
-6.68862700e-01 5.71995914e-01 -1.04964590e+00 -5.53642333e-01
1.78155467e-01 5.01915455e-01 -8.22391152e-01 3.20454597e-01
-9.33267057e-01 7.93238580e-01 3.33985925e-01 3.29168677e-01
-6.83249891e-01 1.75552860e-01 -8.29218149e-01 -2.32792869e-01
1.34220496e-01 -1.86448500e-01 8.21025729e-01 -9.43145752e-01
-1.80044460e+00 8.45192194e-01 1.18155643e-01 -8.05834532e-02
3.61972958e-01 -3.17521840e-01 -7.02921808e-01 2.40020290e-01
-1.34275839e-01 1.90807238e-01 1.10697448e+00 -5.49511433e-01
-8.95694196e-02 -1.01508629e+00 -2.53248066e-01 3.50819021e-01
-6.83047473e-01 4.46929075e-02 -1.19009204e-02 -5.34636080e-01
6.19422674e-01 -9.71344590e-01 1.36943758e-01 -1.97446704e-01
-3.97191137e-01 -5.41540921e-01 1.10091805e+00 -8.62087131e-01
1.01614618e+00 -2.49208403e+00 5.24117470e-01 1.52122736e-01
-6.54202700e-02 3.93938363e-01 -2.69541722e-02 4.97634053e-01
-5.07830203e-01 -2.23303229e-01 8.19731057e-02 -5.23085743e-02
1.29270822e-01 4.22569066e-01 -1.89116880e-01 7.21074879e-01
-1.17315196e-01 5.88915646e-01 -3.70926499e-01 -5.48956931e-01
4.86540049e-01 4.98993516e-01 -3.08804959e-01 5.63713610e-02
4.70985025e-01 3.71159434e-01 -8.72923970e-01 7.26732135e-01
4.60843593e-01 6.53818473e-02 1.36997536e-01 -6.75259769e-01
1.06112979e-01 -3.46339494e-01 -1.71958387e+00 1.46571827e+00
1.87838554e-01 1.18223084e-02 -1.19639494e-01 -1.38569164e+00
9.07518268e-01 3.87903243e-01 9.88665700e-01 -6.65143356e-02
3.34942341e-01 -2.51517482e-02 -2.29341477e-01 -7.84108520e-01
-2.06390712e-02 -2.56844401e-01 1.65191948e-01 5.14989018e-01
2.71566570e-01 5.29488683e-01 -6.29046112e-02 7.01662377e-02
8.91661644e-01 5.38129091e-01 4.38114405e-01 -5.45235686e-02
8.49046886e-01 -3.87744337e-01 7.43763208e-01 7.20749944e-02
-4.98906016e-01 4.39672142e-01 2.93596774e-01 -5.28435051e-01
-5.47209680e-01 -9.44795370e-01 -5.06726429e-02 6.47381783e-01
4.16859053e-02 -3.51282835e-01 -5.70248246e-01 -9.94600475e-01
8.46313238e-02 2.40315139e-01 -6.20951951e-01 -4.52070385e-01
-3.74603242e-01 -5.70391595e-01 6.06744230e-01 8.74240696e-01
1.07739782e+00 -1.05338383e+00 -5.38361967e-01 -2.64003813e-01
-1.67465866e-01 -8.88287544e-01 -4.36043084e-01 -2.99369276e-01
-8.57233226e-01 -1.42240143e+00 -9.04623985e-01 -5.28437972e-01
6.28588140e-01 3.79143775e-01 3.81852418e-01 -3.52691531e-01
-3.72378260e-01 1.03594685e+00 -5.95741153e-01 -1.40397534e-01
-1.71856552e-01 -5.39391637e-01 7.97712743e-01 6.21125519e-01
7.55824924e-01 -7.18334079e-01 -4.12409663e-01 5.01625538e-01
-7.41354406e-01 -1.91508844e-01 9.12497699e-01 8.43191803e-01
4.39743519e-01 -9.64604393e-02 1.64735928e-01 -2.17580706e-01
3.43203396e-01 -1.57430664e-01 -7.88273811e-02 4.51267302e-01
-3.02938640e-01 2.34761611e-01 1.76634938e-01 -3.71885777e-01
-1.13376558e+00 4.50026035e-01 -8.77332389e-02 -8.58963013e-01
-5.12272120e-01 4.71729845e-01 -5.19629419e-01 -1.70902818e-01
7.37989128e-01 5.25072396e-01 4.65076923e-01 -9.07215536e-01
-4.27744538e-02 9.87583876e-01 1.67013839e-01 -3.12753797e-01
6.53033376e-01 5.36533952e-01 2.44232327e-01 -1.24416637e+00
-4.69613373e-01 -5.63053727e-01 -1.08083391e+00 -7.11860120e-01
1.27567267e+00 -6.99372947e-01 -9.49394107e-01 1.05101693e+00
-7.21046805e-01 1.75799519e-01 6.82277903e-02 9.53082502e-01
-6.33538246e-01 1.04470146e+00 -3.45571756e-01 -1.08610010e+00
-1.38475567e-01 -9.22600865e-01 1.13381743e+00 2.89615095e-01
4.05871719e-02 -5.67263663e-01 1.85727343e-01 6.75648153e-01
-1.59458935e-01 4.32820946e-01 6.15176737e-01 -5.26340365e-01
-7.59580955e-02 -4.03455019e-01 2.06854314e-01 7.89798796e-01
4.43318278e-01 -1.39388084e-01 -6.10505402e-01 -2.45874524e-01
4.85614650e-02 -5.70998371e-01 7.06341326e-01 1.94249615e-01
9.33545530e-01 -2.50575840e-01 -3.49819034e-01 5.51258504e-01
9.14315403e-01 5.27260184e-01 9.00315821e-01 6.71432912e-02
7.05809176e-01 2.60935456e-01 4.35648292e-01 6.16544187e-01
7.34963790e-02 7.08889365e-01 2.33639136e-01 1.60867661e-01
6.40354380e-02 -2.34831497e-01 8.97692025e-01 9.28929985e-01
-6.24239445e-01 4.02601004e-01 -8.21481466e-01 -3.58498096e-02
-1.87271738e+00 -1.39815187e+00 5.86515516e-02 1.97739577e+00
3.75934362e-01 -3.96482050e-01 3.27005893e-01 5.59298277e-01
4.86739486e-01 2.53491223e-01 -6.24379098e-01 -4.32668440e-02
7.42892316e-03 -2.93883327e-02 5.26711680e-02 -4.73531298e-02
-1.58083165e+00 7.12342083e-01 6.14770126e+00 6.93833768e-01
-1.01379669e+00 -1.88390747e-01 2.54387498e-01 4.07288194e-01
4.34636444e-01 -3.30720186e-01 -8.23611498e-01 1.79054096e-01
3.55396330e-01 1.85022086e-01 3.90519887e-01 1.06678176e+00
-1.34194970e-01 -1.57474935e-01 -9.06134009e-01 1.68728602e+00
6.20549738e-01 -6.10442877e-01 4.21543159e-02 9.75013748e-02
4.60081309e-01 -4.65945572e-01 2.40136310e-03 4.62669909e-01
-5.46968170e-03 -8.36218059e-01 1.34452879e-01 7.66879201e-01
6.22744024e-01 -6.27206624e-01 3.75551760e-01 4.42343354e-01
-1.07291389e+00 -2.22106174e-01 -5.05727887e-01 -1.96160346e-01
-7.07900301e-02 1.56794399e-01 -5.10896981e-01 7.31553376e-01
7.56003678e-01 1.33797896e+00 -5.38450658e-01 7.45211422e-01
-8.17758664e-02 5.65721333e-01 -3.17120463e-01 -1.75260678e-02
-9.53042656e-02 -5.56276023e-01 6.87684834e-01 6.73290968e-01
3.18828076e-01 5.79484642e-01 3.75411570e-01 3.46366197e-01
6.11007392e-01 2.53666312e-01 -7.23272443e-01 -4.21731472e-01
-3.06730032e-01 1.06692672e+00 -5.55451512e-01 -9.80488881e-02
-4.02405381e-01 1.25601888e+00 -2.21421346e-01 3.26301068e-01
-7.51035571e-01 -1.55240849e-01 6.55487716e-01 -1.80673584e-01
1.55871361e-01 -5.51991403e-01 1.72876880e-01 -1.46981370e+00
-8.89551565e-02 -1.14897406e+00 6.31740153e-01 -6.92667246e-01
-1.21804214e+00 1.68338105e-01 4.14876193e-01 -1.54342818e+00
-3.26087207e-01 -1.14138377e+00 -5.37013412e-01 5.22757113e-01
-4.89519447e-01 -1.12708569e+00 -2.62027025e-01 1.21215916e+00
4.77224857e-01 -7.19055414e-01 9.04280365e-01 2.07576662e-01
-8.26282501e-01 4.87196445e-01 1.25776276e-01 5.02017200e-01
5.49896002e-01 -9.58363831e-01 -2.32665926e-01 6.87213719e-01
5.15602171e-01 4.45247978e-01 4.49718833e-01 -7.75924861e-01
-1.84669340e+00 -5.12660146e-01 3.98332804e-01 -3.31575304e-01
2.95617849e-01 4.26293723e-02 -4.08025742e-01 6.59776449e-01
-6.97920695e-02 -3.38724971e-01 8.83453786e-01 1.84767663e-01
-3.07526439e-01 -2.76637346e-01 -8.91417682e-01 4.13174629e-01
1.26492143e+00 -4.62370574e-01 -8.54231536e-01 5.24753988e-01
-3.60517859e-01 -1.07127227e-01 -6.70536757e-01 7.32291520e-01
8.26596022e-01 -1.10863197e+00 1.05976307e+00 -1.00236464e+00
8.94256588e-03 -4.06391174e-01 -3.20781738e-01 -1.03230643e+00
-4.52383697e-01 -1.85352817e-01 -7.74621591e-02 1.03921068e+00
-2.37502426e-01 -6.64932609e-01 1.08630919e+00 5.33471107e-01
9.50012803e-02 -4.48710024e-01 -1.24252570e+00 -1.10174489e+00
-3.06832999e-01 -8.39026570e-02 1.28098384e-01 1.04181886e+00
1.90611362e-01 3.97493154e-01 -7.30346501e-01 -8.34260136e-02
7.95185745e-01 -3.92061882e-02 9.68672097e-01 -1.13386130e+00
-4.30336893e-01 -1.66632816e-01 -1.14972889e+00 -1.00225031e+00
3.58745515e-01 -7.47822106e-01 -3.24261963e-01 -1.53934336e+00
2.95092285e-01 3.32992405e-01 -2.62069464e-01 8.38799894e-01
1.67645022e-01 1.22681689e-02 2.65942048e-02 2.68697232e-01
-4.80608612e-01 9.78477597e-01 1.11431122e+00 -3.40552747e-01
2.33120658e-02 1.92967847e-01 -7.36581758e-02 7.96752095e-01
7.25798905e-01 2.63607353e-02 -5.69557011e-01 -1.19273648e-01
-2.73833722e-01 5.92214465e-02 4.67070788e-01 -1.38601494e+00
-9.58071351e-02 -2.65391409e-01 7.98662245e-01 -5.86842895e-01
5.32471776e-01 -8.24461937e-01 3.27369362e-01 4.84844387e-01
1.11216538e-01 -1.46542579e-01 -8.97822529e-02 6.23590350e-01
-2.87310451e-01 5.22145741e-02 5.11579633e-01 -1.94638059e-01
-1.24168003e+00 3.57321501e-01 -3.75669271e-01 -1.91845566e-01
1.11363649e+00 -7.03044832e-01 1.74183935e-01 -4.47587609e-01
-1.10854316e+00 -1.94801167e-01 -8.91378000e-02 5.47359228e-01
8.84836078e-01 -1.71680009e+00 -5.05499423e-01 5.02598822e-01
-7.44030923e-02 -8.38011026e-01 5.19868791e-01 1.12497449e+00
-1.70201227e-01 4.24768746e-01 -6.67473912e-01 -5.29801905e-01
-1.52170789e+00 4.53774214e-01 5.03986895e-01 6.60314783e-02
-5.72584867e-01 7.15581119e-01 1.52931049e-01 -3.31868917e-01
3.78754646e-01 2.25908145e-01 -6.93750501e-01 -1.18713230e-02
5.70661426e-01 6.98943615e-01 -2.35666350e-01 -1.27309203e+00
-7.64743865e-01 9.32352066e-01 2.62419403e-01 7.28160050e-03
1.16757584e+00 2.47429058e-01 -1.07570648e-01 3.69532287e-01
1.20648181e+00 -3.68469238e-01 -1.01974583e+00 1.03539549e-01
-9.07908902e-02 -5.64062536e-01 -1.59771159e-01 -6.99532628e-01
-1.05078852e+00 1.09602892e+00 1.11235666e+00 -2.56083846e-01
1.24087811e+00 -2.69957244e-01 4.24977839e-01 6.61287487e-01
7.40773261e-01 -1.22787488e+00 4.59448516e-01 4.52098101e-01
1.32284272e+00 -1.08271348e+00 5.91483563e-02 -2.51396775e-01
-7.92107522e-01 1.29836047e+00 8.71233046e-01 -2.05852613e-01
1.11962152e+00 -1.28681451e-01 1.40981153e-01 -1.08544305e-01
-4.08544689e-02 -1.91073254e-01 4.53126252e-01 6.21048570e-01
1.81788802e-01 -2.17249244e-02 -5.79323471e-01 7.73506701e-01
1.70098782e-01 1.08360954e-01 9.23993718e-03 9.54918444e-01
-3.30918938e-01 -1.16622746e+00 -5.97026765e-01 3.88329804e-01
-1.27384886e-01 5.39940119e-01 -7.17376888e-01 5.55132210e-01
-5.61202466e-02 8.45739543e-01 -3.65809739e-01 -1.05032766e+00
4.35615420e-01 5.06935179e-01 7.13786781e-01 -2.04318732e-01
1.24160074e-01 1.55087626e-02 -8.40499550e-02 -8.00578594e-01
-7.92494476e-01 -8.25345218e-01 -1.17139435e+00 8.34112912e-02
-1.77865297e-01 7.55817443e-02 3.74215811e-01 1.12028348e+00
3.18450511e-01 -5.84055223e-02 6.21558070e-01 -1.00856304e+00
-1.04169846e+00 -1.14069045e+00 -1.08831978e+00 6.12893939e-01
-1.92820087e-01 -1.21425140e+00 -3.82280588e-01 2.01966867e-01] | [7.847659111022949, 0.3877544105052948] |
1db3c544-d4bc-4890-861f-0036fe77577b | an-improved-relevance-feedback-in-cbir | 2006.11821 | null | https://arxiv.org/abs/2006.11821v2 | https://arxiv.org/pdf/2006.11821v2.pdf | An Improved Relevance Feedback in CBIR | Relevance Feedback in Content-Based Image Retrieval is a method where the feedback of the performance is being used to improve itself. Prior works use feature re-weighting and classification techniques as the Relevance Feedback methods. This paper shows a novel addition to the prior methods to further improve the retrieval accuracy. In addition to all of these, the paper also shows a novel idea to even improve the 0-th iteration retrieval accuracy from the information of Relevance Feedback. | ['Subhadip Maji', 'Smarajit Bose'] | 2020-06-21 | null | null | null | null | ['content-based-image-retrieval'] | ['computer-vision'] | [ 3.28162819e-01 -5.14990866e-01 -4.58023280e-01 -2.48612761e-01
-9.02975202e-01 -1.44003928e-01 6.06008470e-01 2.90179610e-01
-4.11633462e-01 6.18753612e-01 3.32018018e-01 4.14251648e-02
-4.24869359e-01 -6.22992635e-01 -4.07975093e-02 -5.13698339e-01
-9.79197025e-03 4.44235690e-02 5.47968984e-01 -6.61432624e-01
1.27773428e+00 5.95716536e-01 -1.97740090e+00 5.15921712e-01
5.25397003e-01 9.23693836e-01 3.49591434e-01 1.00241077e+00
-2.99317956e-01 5.98464310e-01 -7.40936637e-01 5.41022196e-02
4.74937975e-01 -4.75950748e-01 -9.54402149e-01 5.87868281e-02
4.74429995e-01 -5.73512256e-01 -3.65915269e-01 9.17411149e-01
5.86291611e-01 3.60353440e-01 6.53230309e-01 -1.04126072e+00
-1.03421092e+00 -6.65004775e-02 -6.96766317e-01 6.18533432e-01
9.46701586e-01 -5.39321482e-01 8.98548365e-01 -1.38930428e+00
5.74598670e-01 1.21033192e+00 1.75905332e-01 3.15527141e-01
-3.99461061e-01 -6.02918983e-01 1.94120646e-01 1.00227404e+00
-1.47602975e+00 -5.48267625e-02 7.48782277e-01 -2.13887706e-01
9.49387133e-01 4.61187989e-01 9.04933989e-01 5.94430640e-02
2.98746198e-01 8.24166954e-01 8.93767118e-01 -1.19104409e+00
3.23493816e-02 1.29327789e-01 5.67981005e-01 4.57127720e-01
7.99890459e-02 2.25824863e-01 -6.33073747e-01 -4.97472048e-01
4.35958534e-01 2.06534326e-01 -2.56933898e-01 1.14542477e-01
-6.21900260e-01 8.93118978e-01 5.50870180e-01 6.55183494e-01
-5.22185743e-01 1.48172393e-01 1.81551978e-01 7.23103166e-01
4.34332281e-01 6.22457206e-01 -2.23158523e-01 2.49438047e-01
-1.02149212e+00 2.39384845e-01 4.98592496e-01 6.91247940e-01
1.21304059e+00 -3.56716990e-01 -6.85906231e-01 7.29032576e-01
6.83244228e-01 4.65949297e-01 9.85782981e-01 -7.55305469e-01
-2.84713119e-01 6.31976902e-01 2.66708672e-01 -1.19989443e+00
2.13547260e-01 -1.41452834e-01 -9.71641857e-03 3.70086461e-01
-3.23118478e-01 5.35298824e-01 -1.41266787e+00 7.57404387e-01
2.54594743e-01 3.83056793e-03 -3.63993824e-01 9.47432280e-01
1.05588746e+00 7.35247016e-01 -1.95439026e-01 -3.56510758e-01
1.08882010e+00 -1.13184988e+00 -1.14321911e+00 3.27775836e-01
4.77329075e-01 -1.68925083e+00 7.90291667e-01 5.17451286e-01
-8.62231553e-01 -6.27854049e-01 -1.28813505e+00 2.90400684e-01
-7.06894994e-01 -1.38759866e-01 8.52691114e-01 6.85993135e-01
-1.58515894e+00 6.31416440e-01 -4.70749214e-02 -4.11510289e-01
-1.03908807e-01 5.20086229e-01 -1.24635831e-01 -1.67677879e-01
-1.41587520e+00 1.26877272e+00 2.55673856e-01 -2.57519007e-01
-2.36386046e-01 -5.84277093e-01 -3.82858008e-01 -1.61039561e-01
7.78999627e-02 -3.98958027e-01 1.15343988e+00 -8.83573055e-01
-1.50930500e+00 5.61801970e-01 -2.81967342e-01 2.21444294e-02
1.27438858e-01 -3.62545937e-01 -2.81688720e-01 4.84372586e-01
-6.49106801e-02 7.67889082e-01 1.19617617e+00 -1.16113639e+00
-1.06182086e+00 -2.86115170e-01 -2.15778574e-02 5.23312330e-01
-5.95855117e-01 1.22797407e-01 -6.53232932e-01 -4.85187650e-01
2.78088748e-01 -7.14520454e-01 -2.51694143e-01 -6.79838657e-02
4.00496840e-01 -6.56056523e-01 1.22749054e+00 -4.32806343e-01
1.57121599e+00 -1.77010190e+00 -4.56781060e-01 6.08235776e-01
-4.50959839e-02 6.05332971e-01 -4.47517604e-01 8.28594148e-01
-2.70314887e-02 4.25974786e-01 3.69067013e-01 3.46117973e-01
-5.18884599e-01 5.11000492e-02 -1.99322835e-01 3.18819314e-01
3.04713361e-02 9.31233287e-01 -8.75785768e-01 -7.95219958e-01
4.19729531e-01 6.09174252e-01 -3.60947877e-01 -8.64660461e-03
5.13544142e-01 3.22234109e-02 -5.15792787e-01 7.35724747e-01
5.46976745e-01 -2.27114469e-01 -2.98112243e-01 -1.54725816e-02
6.35831282e-02 -4.04839888e-02 -1.16923368e+00 1.46451771e+00
-3.23464721e-01 6.88249469e-01 -4.36336845e-01 -7.82452166e-01
8.92065763e-01 2.85480738e-01 8.69146585e-01 -1.01552331e+00
3.43497135e-02 1.90327629e-01 -2.28967279e-01 -6.22014523e-01
1.03219736e+00 1.26943022e-01 4.74221170e-01 4.14706171e-01
-1.46959141e-01 -3.20544392e-01 4.93636668e-01 4.50432003e-01
9.74997103e-01 -1.36335455e-02 7.04448044e-01 -2.11766697e-02
8.03545713e-01 1.97724536e-01 1.30563259e-01 9.78480697e-01
-2.10144952e-01 7.45349467e-01 -5.35930753e-01 -2.88041800e-01
-7.94294238e-01 -5.87345064e-01 -3.58157843e-01 1.15604544e+00
4.64910716e-01 -5.84557414e-01 -2.73994923e-01 -5.81828892e-01
8.41593072e-02 -2.27933116e-02 -9.34578478e-01 -2.22001478e-01
-2.74611443e-01 -5.67535400e-01 -1.47610798e-01 2.13209406e-01
4.26076263e-01 -1.01083446e+00 -2.94645727e-01 2.48759031e-01
-1.64028779e-01 -1.63188338e-01 -6.27417505e-01 -1.12839088e-01
-1.01357913e+00 -8.75324190e-01 -1.16696131e+00 -7.40401387e-01
8.12416494e-01 9.55293298e-01 9.87060606e-01 1.14038634e+00
-5.13936877e-01 8.35874259e-01 -1.13314807e+00 -6.64600730e-01
-1.74109221e-01 -2.27790788e-01 -2.88743109e-01 -3.54892284e-01
6.03790939e-01 -7.74868280e-02 -8.98634315e-01 2.48823822e-01
-9.21183467e-01 -8.00342619e-01 6.94279313e-01 1.11087584e+00
7.87237048e-01 3.36666740e-02 5.81304669e-01 -6.47479057e-01
1.13557982e+00 -1.40271083e-01 -3.31055641e-01 6.79567695e-01
-1.54841733e+00 2.04656143e-02 -1.68941110e-01 -4.99301314e-01
-7.63539612e-01 -3.74194793e-02 1.66124761e-01 -2.45053440e-01
3.43220174e-01 5.99050283e-01 6.38956487e-01 -8.70753825e-01
8.41669559e-01 1.56269729e-01 5.46558797e-02 -2.47949615e-01
2.28424311e-01 7.90540755e-01 -1.73644081e-01 -1.04493342e-01
5.23069620e-01 1.91135198e-01 1.47749603e-01 -4.96967733e-01
-5.69349527e-01 -1.42738068e+00 -6.54167712e-01 -6.52122736e-01
1.37783542e-01 -5.06340146e-01 -5.54997146e-01 8.30740854e-02
-8.98408413e-01 3.29901159e-01 -4.23284322e-01 7.22073376e-01
-2.47065932e-01 7.78551042e-01 -4.81183112e-01 -9.78695929e-01
-7.61884749e-01 -1.05054510e+00 8.84300888e-01 4.28419322e-01
-9.51773748e-02 -8.54804456e-01 3.70476663e-01 1.13569938e-01
6.77146375e-01 -5.36393464e-01 5.44222236e-01 -5.17870009e-01
-3.53504956e-01 -8.38335931e-01 -2.68958151e-01 1.90699011e-01
3.35493922e-01 1.60427108e-01 -9.25582528e-01 -2.91604668e-01
1.55152261e-01 7.79746026e-02 1.00149989e+00 3.89437258e-01
6.69081092e-01 -1.29931763e-01 -3.34044993e-01 -1.58846214e-01
1.67120814e+00 7.14809358e-01 1.21755922e+00 5.51676691e-01
-1.09984167e-01 4.19684082e-01 1.39296639e+00 3.69018048e-01
-1.65533215e-01 5.96980095e-01 1.54305771e-01 -6.67454600e-02
-4.16640759e-01 2.25888744e-01 5.65714501e-02 1.02129900e+00
-9.93764326e-02 -4.35811952e-02 -5.42779922e-01 5.31999052e-01
-1.88082600e+00 -1.03126037e+00 -2.28401318e-01 2.31399322e+00
8.33405197e-01 -3.62699300e-01 -2.32082307e-01 5.32460213e-01
8.03896308e-01 -3.36171627e-01 -2.24079639e-01 -5.73934257e-01
3.02460849e-01 3.79581124e-01 4.79700804e-01 7.70339966e-01
-9.01939154e-01 7.27747500e-01 8.61357594e+00 1.05076921e+00
-1.13241303e+00 1.00863706e-02 1.80444613e-01 2.33088180e-01
-3.51593316e-01 2.47716635e-01 -8.62076998e-01 7.54615366e-02
4.54838425e-01 -4.08697039e-01 1.92600042e-01 6.89138353e-01
-2.03988716e-01 -8.48097980e-01 -6.71669066e-01 1.12639689e+00
4.85527039e-01 -1.12864876e+00 5.67336977e-01 1.16923265e-01
8.84381771e-01 -4.10480678e-01 1.48280784e-01 2.36718044e-01
1.15237586e-01 -6.42497182e-01 9.58387777e-02 9.58956420e-01
3.97578001e-01 -6.84722722e-01 1.22003627e+00 -3.00444867e-02
-1.06132376e+00 3.74473184e-02 -5.48638046e-01 -2.27381498e-01
-1.42984986e-01 5.66573024e-01 -1.18429434e+00 7.49817729e-01
7.12024868e-01 4.11949694e-01 -9.04051185e-01 1.69697297e+00
-1.55624509e-01 3.16807389e-01 -1.60055086e-01 -3.00090790e-01
7.26893395e-02 1.56209424e-01 4.15896684e-01 1.12680864e+00
3.40605676e-01 2.01699704e-01 2.59355128e-01 -1.39557630e-01
3.28961939e-01 8.23710620e-01 -6.16319895e-01 2.69562095e-01
3.03357154e-01 1.35300434e+00 -7.53180802e-01 -7.28119433e-01
-2.61444062e-01 8.90493929e-01 -2.60510027e-01 2.01743484e-01
-6.72314465e-02 -7.98562109e-01 7.71475285e-02 7.46659189e-02
2.32700184e-01 2.68401474e-01 -2.71539509e-01 -5.55572808e-01
-1.66280299e-01 -3.55593204e-01 4.94465202e-01 -9.15020764e-01
-1.40447664e+00 3.85147363e-01 2.38900676e-01 -1.61203241e+00
-3.91591311e-01 -1.97097912e-01 -2.65696108e-01 1.11913824e+00
-1.97386682e+00 -7.37993538e-01 -3.03440571e-01 5.80579638e-01
5.35776675e-01 -3.31276298e-01 8.77014697e-01 3.03774089e-01
3.99637341e-01 5.70892990e-01 5.26689515e-02 -4.15081233e-01
1.23653293e+00 -1.05668306e+00 -1.93889603e-01 6.06795669e-01
9.07087326e-02 7.46900201e-01 6.19286597e-01 -8.95264268e-01
-1.07908964e+00 -4.38814849e-01 1.02069545e+00 -2.72570550e-01
3.93340290e-01 6.52072668e-01 -8.92703414e-01 -1.56476811e-01
3.21520478e-01 -1.64598927e-01 8.65041375e-01 1.28137827e-01
-6.80606738e-02 -2.79292256e-01 -1.36640716e+00 4.48242962e-01
5.93161941e-01 -4.00259525e-01 -8.24041665e-01 3.19136083e-01
4.07004893e-01 -3.39054391e-02 -8.26987028e-01 7.25172937e-01
8.05420935e-01 -6.85072720e-01 1.08661616e+00 -1.32053837e-01
3.43889818e-02 -4.25592095e-01 -2.62489051e-01 -1.07813931e+00
-7.50550508e-01 -3.03445905e-01 1.12501778e-01 8.05560052e-01
1.62549525e-01 -4.53346580e-01 8.00818443e-01 4.64147359e-01
1.38418972e-01 -5.80731094e-01 -7.48607039e-01 -6.65380359e-01
-1.36164933e-01 -7.99924135e-02 5.02186298e-01 5.91495097e-01
3.39291424e-01 1.84660405e-01 -2.82032311e-01 -5.04191458e-01
3.24778080e-01 -2.19353032e-03 4.65587348e-01 -1.41266072e+00
1.81162894e-01 -5.16459763e-01 -8.53554308e-01 -8.71827543e-01
-4.70330566e-01 -7.37922132e-01 2.81935811e-01 -1.89753795e+00
4.19378251e-01 -2.76755095e-01 -8.68035376e-01 3.86322230e-01
-5.32290578e-01 7.63352036e-01 2.84476757e-01 6.34505630e-01
-7.28452563e-01 1.40094712e-01 1.80648804e+00 -2.97339499e-01
-3.05991858e-01 1.48181096e-02 -6.05309069e-01 3.37942243e-01
6.43755257e-01 -6.55985951e-01 -6.02367520e-01 1.08202703e-01
3.84599894e-01 -7.57836327e-02 -5.55606484e-02 -8.42712820e-01
5.67119062e-01 -1.46157905e-01 5.83998799e-01 -1.09001195e+00
2.83814490e-01 -8.96566451e-01 -2.15316504e-01 5.89088857e-01
-4.68585968e-01 2.95474023e-01 -1.02454357e-01 6.84178650e-01
-6.21548057e-01 -9.26994860e-01 3.42919409e-01 -4.28768992e-01
-1.16117895e+00 6.99242875e-02 -7.67592549e-01 -6.61727190e-01
8.43802035e-01 -6.81599915e-01 4.85710055e-02 -7.16621161e-01
-7.21786499e-01 2.13698447e-01 1.59101799e-01 4.01780188e-01
1.10767937e+00 -1.53330314e+00 -6.42033637e-01 -5.12878783e-03
3.68323863e-01 -7.62883306e-01 1.56025946e-01 6.79917932e-01
-3.70170325e-01 6.00874007e-01 -3.17455798e-01 -5.08791029e-01
-1.79074907e+00 7.06301808e-01 -3.91964354e-02 -3.92807603e-01
-3.24751347e-01 7.65639603e-01 -2.43777424e-01 2.07903817e-01
3.43488157e-01 4.10758346e-01 -8.99941564e-01 3.24926049e-01
1.05281281e+00 3.52442265e-01 5.10893703e-01 -2.69590437e-01
-3.10986310e-01 8.15798163e-01 -6.11674011e-01 -5.59950233e-01
9.40855920e-01 -4.57819283e-01 -2.52788693e-01 3.47216189e-01
1.15584314e+00 -1.19153701e-01 -4.27851826e-01 -4.00409192e-01
1.11652851e-01 -1.09513652e+00 3.56189311e-01 -1.00521719e+00
-8.47136438e-01 5.69771647e-01 1.13399982e+00 1.78547069e-01
1.40745437e+00 -3.01365316e-01 3.32178712e-01 7.96316862e-01
6.41495347e-01 -1.67699587e+00 3.77477378e-01 5.72973788e-01
1.24954975e+00 -1.56807864e+00 6.88889980e-01 -3.06937695e-01
-2.06159934e-01 1.36948407e+00 2.74024785e-01 -5.45457542e-01
1.30342531e+00 6.14144839e-02 3.86572868e-01 -1.51479080e-01
-7.98842371e-01 -5.14727771e-01 7.66228914e-01 7.85849094e-01
5.98042548e-01 -4.17433709e-01 -1.35045505e+00 -1.36767626e-01
2.86990047e-01 4.02770281e-01 3.44461530e-01 1.26570559e+00
-8.88990283e-01 -1.63147008e+00 -7.23123193e-01 6.65975451e-01
-6.13612413e-01 -3.09430301e-01 -6.93241000e-01 5.34999549e-01
-3.16388935e-01 1.39737844e+00 -3.05436015e-01 -6.03794098e-01
1.64375752e-01 -3.34112681e-02 6.60552442e-01 -7.11563468e-01
-1.04802847e+00 1.99056208e-01 -3.88246417e-01 -4.10283506e-01
-8.36967409e-01 -3.83005112e-01 -1.07943058e+00 -1.17807366e-01
-1.13021815e+00 5.61607182e-01 1.11942184e+00 7.56200731e-01
1.67760253e-01 3.95774573e-01 9.37041163e-01 -7.79166758e-01
-3.29757363e-01 -1.29829073e+00 -4.17829901e-01 3.84063452e-01
1.96581349e-01 -7.46299088e-01 -3.60948592e-01 -9.19883922e-02] | [10.792852401733398, 0.1187131404876709] |
23c941b4-9741-4c36-81f5-5a9c53da5761 | time-varying-transition-matrices-with-multi | 2306.11772 | null | https://arxiv.org/abs/2306.11772v1 | https://arxiv.org/pdf/2306.11772v1.pdf | Time-Varying Transition Matrices with Multi-task Gaussian Processes | In this paper, we present a kernel-based, multi-task Gaussian Process (GP) model for approximating the underlying function of an individual's mobility state using a time-inhomogeneous Markov Process with two states: moves and pauses. Our approach accounts for the correlations between the transition probabilities by creating a covariance matrix over the tasks. We also introduce time-variability by assuming that an individual's transition probabilities vary over time in response to exogenous variables. We enforce the stochasticity and non-negativity constraints of probabilities in a Markov process through the incorporation of a set of constraint points in the GP. We also discuss opportunities to speed up GP estimation and inference in this context by exploiting Toeplitz and Kronecker product structures. Our numerical experiments demonstrate the ability of our formulation to enforce the desired constraints while learning the functional form of transition probabilities. | ['Ekin Ugurel'] | 2023-06-20 | null | null | null | null | ['gaussian-processes'] | ['methodology'] | [ 4.10970449e-02 -1.09362036e-01 -1.61072463e-01 -2.36232653e-01
-3.82200718e-01 -3.62827271e-01 9.15415108e-01 -8.38317052e-02
-4.86401558e-01 8.91404092e-01 1.28170028e-01 -3.19265753e-01
-2.78511703e-01 -5.53215504e-01 -5.32376230e-01 -9.27957296e-01
-4.70331073e-01 8.60063672e-01 2.45732799e-01 2.53512055e-01
-2.69553475e-02 4.53118205e-01 -1.03364408e+00 -3.27134967e-01
7.14132726e-01 5.55043638e-01 2.02138573e-01 1.03455567e+00
1.57843739e-01 7.66522348e-01 -5.24184942e-01 -3.72078747e-01
8.49673152e-02 -8.29570740e-02 -4.67742294e-01 2.55820483e-01
-4.71696556e-01 1.60636678e-01 -4.62032020e-01 8.20724845e-01
1.43024117e-01 4.95036781e-01 1.25817955e+00 -1.59359741e+00
-5.10879457e-01 4.41175364e-02 -4.96197611e-01 2.69020021e-01
1.86181858e-01 1.35590419e-01 8.90207410e-01 -3.07570159e-01
1.51918709e-01 1.26637363e+00 7.79126644e-01 2.83809245e-01
-1.64100444e+00 -2.34264180e-01 3.38778764e-01 -1.00184716e-01
-1.57329142e+00 -3.05148780e-01 4.87693876e-01 -8.16247582e-01
9.56564188e-01 6.35210499e-02 4.70403492e-01 1.40027583e+00
9.04094160e-01 6.78559303e-01 1.10042393e+00 -2.82055765e-01
5.28579950e-01 -1.84186488e-01 2.85107970e-01 6.41569197e-01
1.17603414e-01 4.33342420e-02 -3.53841186e-01 -8.77915025e-01
9.93293524e-01 1.33225799e-01 7.81919509e-02 -5.70446134e-01
-9.74725008e-01 9.30212736e-01 -4.51687813e-01 -3.55761275e-02
-7.27665126e-01 5.49031377e-01 1.70919091e-01 1.02976095e-02
5.58713078e-01 -2.50698835e-01 -4.00788546e-01 -4.74712759e-01
-9.85252440e-01 3.96388769e-01 1.17444253e+00 1.04640651e+00
7.59593606e-01 -3.78237627e-02 -6.63595438e-01 3.78807366e-01
4.49180633e-01 8.00383806e-01 8.03387612e-02 -1.15069902e+00
4.11780477e-01 -3.07634205e-01 7.55624890e-01 -5.58245003e-01
-4.71956670e-01 -3.52408886e-01 -6.55608237e-01 -7.77074322e-02
8.13413203e-01 -5.22983551e-01 -1.04222190e+00 1.98321760e+00
1.82922885e-01 4.31553811e-01 -9.95532498e-02 2.85804093e-01
-4.96093303e-01 6.28497899e-01 4.98377353e-01 -3.22692871e-01
1.29314744e+00 -5.53399622e-01 -8.78197372e-01 -1.98166911e-02
4.61123288e-01 -5.25922298e-01 7.97112525e-01 2.54686087e-01
-1.17166138e+00 -2.88623065e-01 -1.85806215e-01 3.03425252e-01
4.00676690e-02 1.34533882e-01 8.27889323e-01 8.43349099e-01
-1.22181427e+00 5.69739878e-01 -1.52998006e+00 -3.17928195e-01
5.44869937e-02 5.26295543e-01 4.81651956e-03 2.80900508e-01
-1.09862101e+00 8.23854506e-01 -1.40615236e-02 2.63435934e-02
-9.94042099e-01 -5.13855040e-01 -6.30457938e-01 1.60297409e-01
1.71675831e-01 -9.24614549e-01 1.40774751e+00 -5.54395735e-01
-1.60995817e+00 2.95136333e-01 -4.98819709e-01 -4.94253635e-01
6.16875112e-01 3.69108282e-02 -4.04340059e-01 -2.72351112e-02
1.32651880e-01 4.69913259e-02 1.03048706e+00 -9.23466921e-01
-5.61723650e-01 -9.20266435e-02 -1.34650707e-01 4.68074143e-01
1.35271206e-01 2.27369051e-02 -6.36995435e-01 -4.59456235e-01
-2.98118293e-01 -1.50009489e+00 -2.83387303e-01 -3.38435471e-01
-4.57712859e-01 -1.63786277e-01 4.67311651e-01 -7.46175289e-01
1.13717389e+00 -2.05090308e+00 2.71269560e-01 4.48659360e-01
-8.97457078e-02 -2.72061467e-01 -9.26591363e-03 8.27292264e-01
3.38011473e-01 -3.91519479e-02 -2.81129301e-01 -8.80120695e-01
3.20881665e-01 7.54806459e-01 -8.67305547e-02 7.55104542e-01
8.18880498e-02 6.71619773e-01 -9.29717660e-01 -2.78410673e-01
-8.65837112e-02 7.49974310e-01 -3.87056261e-01 -1.68379620e-01
-3.51014771e-02 6.34821057e-01 -7.01501489e-01 2.86232620e-01
3.44816387e-01 -2.26993456e-01 2.03460023e-01 5.19827366e-01
-7.13813826e-02 1.91580296e-01 -1.21051431e+00 1.16851544e+00
-6.24842405e-01 2.63952881e-01 2.32848942e-01 -7.58814394e-01
3.48798871e-01 5.65001309e-01 5.60217679e-01 1.30615262e-02
-2.49873623e-02 -2.33067378e-01 1.96623638e-01 -2.03082636e-01
5.09661913e-01 -4.50998902e-01 -2.50961304e-01 5.56663573e-01
9.93793756e-02 5.90308830e-02 2.13997066e-01 2.71713808e-02
1.22977781e+00 2.34216988e-01 3.59610468e-02 -4.64153409e-01
2.16666356e-01 -4.80109781e-01 6.30190551e-01 1.06280661e+00
-2.30323881e-01 1.27739711e-02 7.92413354e-01 -1.61171332e-02
-1.05052590e+00 -1.45193946e+00 -1.89800069e-01 1.01911366e+00
-4.91674334e-01 -1.22744450e-02 -4.85897571e-01 -1.17381960e-01
8.00441131e-02 7.22514927e-01 -8.47366512e-01 -7.27121681e-02
-3.33106160e-01 -1.26179314e+00 5.11007369e-01 4.48466778e-01
1.95841923e-01 -6.46308839e-01 -4.20961648e-01 4.18263167e-01
-9.35221314e-02 -9.38312709e-01 -6.47043884e-01 5.45713067e-01
-8.23117197e-01 -5.19958854e-01 -8.74923646e-01 -2.99305826e-01
5.82251489e-01 -2.16888145e-01 7.09587097e-01 -6.79617524e-01
1.39304250e-01 8.50887537e-01 1.78206310e-01 -2.22772032e-01
-2.78070033e-01 7.51465932e-02 1.94379508e-01 2.58416831e-01
9.62580740e-02 -6.64288759e-01 -3.56656313e-01 2.12973073e-01
-8.21303427e-01 -2.04677165e-01 2.62975693e-01 7.75821686e-01
4.64886457e-01 4.18168843e-01 -1.27602238e-02 -7.47607768e-01
1.05227864e+00 -6.02924407e-01 -6.71921313e-01 2.91870981e-01
-3.67955983e-01 4.57135916e-01 1.46049082e-01 -9.16659176e-01
-1.42775953e+00 5.97100183e-02 3.61246675e-01 -4.27056164e-01
-4.99422066e-02 4.74613339e-01 -5.89542054e-02 1.17236443e-01
9.36290249e-02 3.82441610e-01 -9.24123079e-02 -2.31554359e-01
4.16138679e-01 2.41744772e-01 3.51613104e-01 -1.06891346e+00
6.37265503e-01 6.64475620e-01 4.97599989e-01 -7.79531360e-01
-4.83170003e-01 -4.78429615e-01 -6.51643813e-01 1.73026063e-02
9.20755327e-01 -1.13548529e+00 -1.06497955e+00 6.57524049e-01
-1.01855075e+00 -9.23401773e-01 -3.82837206e-01 7.13124812e-01
-1.04837108e+00 3.89633715e-01 -1.08109748e+00 -1.41054678e+00
3.29233646e-01 -1.05623746e+00 1.05656242e+00 2.65033431e-02
-3.20781231e-01 -1.85789335e+00 3.52749228e-01 -1.28733993e-01
3.70262623e-01 -2.10901946e-01 7.72853374e-01 -4.95066851e-01
-5.23884118e-01 -2.26663470e-01 8.73113126e-02 5.96494079e-02
2.78986935e-02 7.67212063e-02 -5.07788479e-01 -1.20941296e-01
1.31331146e-01 4.38706875e-01 5.56757092e-01 9.30978417e-01
6.72696650e-01 -2.58536816e-01 -6.04715824e-01 4.95573580e-01
1.17211723e+00 3.71104538e-01 4.07168925e-01 1.04079254e-01
6.78230524e-01 5.56777537e-01 2.26626396e-01 7.02870846e-01
7.55715609e-01 6.03211105e-01 1.99910998e-02 3.73302072e-01
5.37042558e-01 -3.24506223e-01 3.86624396e-01 4.15049732e-01
-3.61974984e-01 -5.21347642e-01 -1.18627667e+00 5.81855834e-01
-2.07287455e+00 -1.15324533e+00 -4.54960972e-01 2.47428465e+00
5.88124812e-01 1.18958481e-01 2.86831081e-01 -3.73998284e-01
6.98841512e-01 -1.20545939e-01 -3.78256112e-01 -3.73550296e-01
9.64405239e-02 2.12968126e-01 1.15846014e+00 9.41302538e-01
-1.12740541e+00 7.99582481e-01 7.53597593e+00 5.30555129e-01
-5.51144004e-01 3.90228480e-01 2.83788294e-01 1.36262685e-01
-1.76511958e-01 1.68305799e-01 -1.00156987e+00 7.40084291e-01
1.34577990e+00 -1.77348405e-01 5.03145635e-01 2.82442331e-01
5.56666255e-01 -3.65066499e-01 -8.66201460e-01 5.19406497e-01
-3.72231245e-01 -4.22209173e-01 -5.05313516e-01 6.03262722e-01
7.13171542e-01 7.40209073e-02 3.12213749e-01 2.85580158e-01
1.13283956e+00 -5.86627185e-01 6.97197974e-01 9.88681376e-01
1.92211956e-01 -7.72003889e-01 2.83173233e-01 5.18953502e-01
-1.10189319e+00 -1.53611571e-01 -4.12229784e-02 -4.67002571e-01
5.98428667e-01 7.98997104e-01 -6.36349738e-01 2.18575239e-01
5.28982393e-02 3.14522892e-01 -8.09977353e-02 1.00083148e+00
-3.20631355e-01 7.65764296e-01 -5.12377977e-01 2.18346015e-01
2.52747685e-01 -7.52802551e-01 5.92817187e-01 1.16813362e+00
5.02751291e-01 -1.78510815e-01 3.34732205e-01 7.66923130e-01
4.65985566e-01 -3.75867426e-01 -4.04329449e-01 -5.76308556e-02
3.33104640e-01 9.04564857e-01 -7.96231985e-01 -3.07782590e-01
-5.25658906e-01 1.41297030e+00 1.73488677e-01 1.10335219e+00
-9.40338492e-01 3.19597535e-02 9.90800440e-01 1.69908106e-02
5.33748090e-01 -8.85720432e-01 5.60971126e-02 -1.21979916e+00
2.40862160e-03 -2.12910429e-01 1.50174633e-01 -5.44440866e-01
-1.37409854e+00 2.10212514e-01 3.52571905e-01 -6.20720327e-01
-4.84627843e-01 -3.65416914e-01 -6.94301963e-01 1.60310209e+00
-1.35955691e+00 -1.06804526e+00 6.13995790e-01 7.88443863e-01
9.46286544e-02 2.56515205e-01 6.73490167e-01 1.90660432e-02
-5.68898499e-01 -3.99356708e-02 3.36410552e-01 -2.28962138e-01
3.09504032e-01 -1.49549913e+00 7.74716675e-01 7.45969415e-01
-1.36589274e-01 8.73576760e-01 1.04360020e+00 -1.05126441e+00
-1.03575945e+00 -8.64900470e-01 1.05708277e+00 -6.97451890e-01
1.26315093e+00 -5.02676785e-01 -8.71908247e-01 1.22852063e+00
-3.98207977e-02 -6.49261773e-02 7.94713795e-01 3.30042005e-01
-8.13952312e-02 5.73853314e-01 -8.09037864e-01 5.22092044e-01
7.73857594e-01 -7.53027141e-01 -2.41095737e-01 4.72232074e-01
2.88771033e-01 -1.98507503e-01 -8.45951021e-01 -1.26475155e-01
4.90803331e-01 -5.20865619e-01 9.02859330e-01 -5.45139134e-01
-4.17440951e-01 -2.40040734e-01 -3.29457819e-02 -1.54256022e+00
-7.43912935e-01 -9.05135930e-01 -2.85115480e-01 1.03703642e+00
4.24389482e-01 -7.72171259e-01 7.82845140e-01 1.16043198e+00
1.84874043e-01 -3.99269134e-01 -1.08391428e+00 -1.09139061e+00
1.74224123e-01 -5.84795892e-01 2.39286840e-01 6.20452881e-01
2.52012685e-02 1.53947055e-01 -7.55039871e-01 5.48914850e-01
7.18204677e-01 -5.27347744e-01 3.60673726e-01 -1.35395789e+00
-7.48082757e-01 -1.24594681e-01 -2.68982291e-01 -1.07942677e+00
4.90065157e-01 -4.66613948e-01 1.17951043e-01 -1.26774406e+00
2.99360156e-01 -5.50966740e-01 -1.23298422e-01 1.60939828e-01
-2.10120663e-01 -4.32589263e-01 1.16049469e-01 3.56277436e-01
-5.35973132e-01 5.60916483e-01 9.47448790e-01 3.96689981e-01
-3.62463534e-01 6.24502897e-01 -2.10144281e-01 5.73106349e-01
8.21310759e-01 -5.99305987e-01 -6.61185324e-01 -1.69663623e-01
1.24206901e-01 4.88903195e-01 3.15902144e-01 -8.89230311e-01
2.53639042e-01 -3.79183501e-01 4.72349487e-02 -3.54127735e-01
8.87548625e-01 -7.05616891e-01 7.21388757e-01 3.97441924e-01
-2.17857182e-01 4.21407789e-01 3.78403463e-03 1.15876436e+00
2.32557148e-01 1.49844680e-02 4.49161738e-01 -1.15517927e-02
-6.51207417e-02 4.20404792e-01 -1.11828649e+00 -8.51054639e-02
9.37785566e-01 2.64059249e-02 1.64379597e-01 -7.37340391e-01
-1.32795298e+00 3.37905318e-01 4.69946921e-01 -5.40517271e-02
1.84764750e-02 -1.01421082e+00 -1.81866154e-01 4.75341566e-02
-3.04578036e-01 -5.73881567e-01 3.65699977e-01 1.11293232e+00
-1.07814588e-01 4.38411057e-01 -1.03869922e-01 -3.78309727e-01
-1.06526494e+00 5.95227420e-01 3.25458646e-01 -8.43523681e-01
-1.74822256e-01 5.35351694e-01 1.86579078e-01 -2.55685449e-01
8.37068632e-02 -3.05149496e-01 2.78978258e-01 -3.91280204e-02
1.83665529e-01 5.13595641e-01 -4.18825924e-01 -7.93439746e-01
-1.55636385e-01 1.35904863e-01 5.29323369e-02 -8.67523372e-01
8.68252754e-01 -5.64165652e-01 1.62919402e-01 1.09768271e+00
7.94684529e-01 -1.21060200e-01 -1.84887516e+00 -2.52708733e-01
5.56727499e-02 -1.03048906e-01 -5.10431752e-02 -4.16014075e-01
-4.27832574e-01 8.29942107e-01 2.52120793e-01 2.32124105e-01
6.09644115e-01 -1.05646014e-01 3.52105647e-01 -1.61556748e-03
4.66238201e-01 -9.76339936e-01 -3.20876271e-01 6.95044816e-01
7.34852478e-02 -6.65954053e-01 -2.77900726e-01 -4.78334486e-01
-8.30834270e-01 6.74970448e-01 -1.66176751e-01 1.01638176e-02
1.03252268e+00 3.90811086e-01 -3.15020740e-01 3.85823995e-02
-8.97629738e-01 -3.69090945e-01 9.21818465e-02 1.01894414e+00
3.94129783e-01 6.00535095e-01 -3.16655755e-01 2.41848320e-01
1.24190360e-01 3.19884792e-02 5.98947287e-01 9.87433195e-01
-2.19967872e-01 -1.42495012e+00 -4.08303231e-01 2.90350616e-01
-6.38329506e-01 -1.94653958e-01 2.66764224e-01 7.27845728e-01
-1.54337689e-01 7.80969203e-01 2.35775456e-01 1.84737116e-01
1.27630889e-01 5.12095153e-01 4.95366067e-01 -6.84535146e-01
-1.39930040e-01 2.59398311e-01 9.36937705e-02 -1.26951918e-01
-1.88741893e-01 -1.39208043e+00 -9.38835263e-01 -3.87207299e-01
-6.42787069e-02 2.35703140e-01 6.50329351e-01 1.05342984e+00
2.82842457e-01 4.30963606e-01 2.06568226e-01 -7.72630930e-01
-6.91935778e-01 -8.21988285e-01 -1.09204900e+00 1.13710545e-01
1.91655785e-01 -8.81028652e-01 -3.53743792e-01 2.45340109e-01] | [6.9413604736328125, 3.9000885486602783] |
ed1466f5-34db-4edf-92f3-a9b38c9be34e | one-model-for-the-learning-of-language | 1711.06301 | null | http://arxiv.org/abs/1711.06301v2 | http://arxiv.org/pdf/1711.06301v2.pdf | One Model for the Learning of Language | A major target of linguistics and cognitive science has been to understand
what class of learning systems can acquire the key structures of natural
language. Until recently, the computational requirements of language have been
used to argue that learning is impossible without a highly constrained
hypothesis space. Here, we describe a learning system that is maximally
unconstrained, operating over the space of all computations, and is able to
acquire several of the key structures present natural language from positive
evidence alone. The model successfully acquires regular (e.g. $(ab)^n$),
context-free (e.g. $a^n b^n$, $x x^R$), and context-sensitive (e.g.
$a^nb^nc^n$, $a^nb^mc^nd^m$, $xx$) formal languages. Our approach develops the
concept of factorized programs in Bayesian program induction in order to help
manage the complexity of representation. We show in learning, the model
predicts several phenomena empirically observed in human grammar acquisition
experiments. | ['Yuan Yang'] | 2017-11-16 | null | null | null | null | ['program-induction'] | ['computer-code'] | [ 2.45046139e-01 4.48102802e-01 1.09371059e-02 -4.17777449e-01
-5.26496291e-01 -6.33807898e-01 7.79203355e-01 2.97637910e-01
-5.29020190e-01 7.58775234e-01 -4.16400731e-01 -1.16932678e+00
-2.91164249e-01 -1.14939213e+00 -8.37412894e-01 -5.22279561e-01
-6.70826793e-01 4.49544042e-01 4.38866347e-01 -5.42509794e-01
5.12286365e-01 2.53834903e-01 -1.84796059e+00 5.55881411e-02
7.51495242e-01 5.84590971e-01 5.29447615e-01 7.54201472e-01
-3.92532319e-01 7.75379300e-01 -1.14432931e-01 -3.93365383e-01
1.85307749e-02 -5.31615138e-01 -1.05431926e+00 -4.36205208e-01
2.11336702e-01 -7.30616823e-02 2.59036832e-02 1.35346997e+00
-1.76038265e-01 8.28206092e-02 6.86832011e-01 -9.23117280e-01
-4.12334412e-01 1.16624177e+00 -1.74940124e-01 1.57114744e-01
6.27486706e-01 5.18501587e-02 1.41499209e+00 -7.69084513e-01
6.21057689e-01 1.50013614e+00 1.98718816e-01 7.03139246e-01
-1.51376629e+00 -4.92588609e-01 1.25938311e-01 -2.86560785e-02
-1.47646201e+00 -2.66841024e-01 5.40696442e-01 -4.71006960e-01
1.30623221e+00 2.93874592e-01 7.71900356e-01 4.63763624e-01
2.55218923e-01 7.92984307e-01 1.62915039e+00 -1.32841909e+00
2.49118134e-01 1.92602739e-01 6.14383996e-01 1.45165193e+00
3.50892603e-01 6.23641491e-01 -8.12042117e-01 -1.47909284e-01
6.12152100e-01 -5.53735137e-01 -2.95010787e-02 -2.06115663e-01
-7.40365267e-01 1.06783175e+00 -9.55916271e-02 7.23450363e-01
1.37651354e-01 2.58305669e-01 1.55544162e-01 4.28718239e-01
-2.54316360e-01 5.91994762e-01 -6.72910094e-01 -5.46958707e-02
-5.81856787e-01 4.61245775e-01 8.04510057e-01 1.11474669e+00
1.06162190e+00 4.82669994e-02 7.91602015e-01 5.03904402e-01
7.20369160e-01 8.52432370e-01 3.56031805e-01 -9.45679069e-01
1.32229701e-01 5.47774494e-01 -2.33834460e-01 -5.87705553e-01
-2.92487055e-01 5.41090816e-02 -2.78460026e-01 5.14792621e-01
4.71655577e-01 2.26766005e-01 -7.28785634e-01 2.21514440e+00
-9.15829316e-02 -5.37609935e-01 1.82687238e-01 1.43490568e-01
4.44868118e-01 6.72309935e-01 4.17091697e-01 -6.16601706e-01
1.41481006e+00 6.76023811e-02 -4.88391101e-01 -5.19531369e-01
9.23111975e-01 -4.38425124e-01 1.54846990e+00 5.73116541e-01
-1.41931152e+00 -3.76252145e-01 -9.31915164e-01 1.54620782e-01
-4.36191350e-01 -2.56230384e-01 1.36947286e+00 1.09599471e+00
-1.18725276e+00 4.18082893e-01 -8.22146416e-01 -2.32945129e-01
-4.20005061e-02 5.36249697e-01 -3.41277301e-01 -2.82108724e-01
-1.22202253e+00 9.82473373e-01 7.59449482e-01 -7.86200389e-02
-1.22322536e+00 -3.28792296e-02 -1.12695992e+00 -1.64014354e-01
6.07013822e-01 -3.19752812e-01 1.31312823e+00 -9.60035324e-01
-1.29935873e+00 1.14148891e+00 -3.79641056e-01 -3.64434987e-01
-1.32553264e-01 7.89228640e-03 -2.87063032e-01 6.63305074e-02
-2.75397338e-02 4.21550184e-01 4.98773158e-01 -1.22348523e+00
-6.98524237e-01 -6.37564540e-01 4.46584374e-01 -9.35715213e-02
3.07503104e-01 5.22846997e-01 2.29517177e-01 -3.07502538e-01
6.04516089e-01 -9.03682113e-01 -1.61345199e-01 -5.83656549e-01
-2.15077370e-01 -5.03307223e-01 -2.62403302e-02 -3.36541533e-01
1.19083798e+00 -2.05439615e+00 1.81706756e-01 7.22100139e-01
1.38334125e-01 -1.29394546e-01 2.64936864e-01 2.40183160e-01
-1.07361570e-01 5.76180339e-01 -5.65433919e-01 3.62907708e-01
3.65210593e-01 6.44075453e-01 -3.62887084e-01 1.92751691e-01
1.54532745e-01 6.09007955e-01 -6.94349647e-01 -7.22213864e-01
-3.98880430e-02 -3.53421792e-02 -9.46720719e-01 3.35862428e-01
-7.51977086e-01 -6.02916442e-02 -6.23112321e-01 6.51298285e-01
2.39777178e-01 -1.56182155e-01 6.40673757e-01 6.36280239e-01
-2.68419176e-01 5.88868380e-01 -1.28714073e+00 1.48405981e+00
-2.84751952e-01 5.29930532e-01 2.27469504e-01 -1.09883285e+00
7.71639168e-01 1.91106722e-01 -2.66635478e-01 -5.89765608e-01
1.49342954e-01 5.71928620e-01 6.93796456e-01 -4.08640414e-01
3.85786831e-01 -6.65771484e-01 -3.14177603e-01 6.99026942e-01
1.44305393e-01 -6.87911332e-01 2.79943317e-01 4.23160017e-01
9.62828815e-01 8.14365670e-02 4.11339134e-01 -7.14233696e-01
6.37944162e-01 1.07880794e-02 5.71821332e-01 1.06052160e+00
4.71062288e-02 -3.26771110e-01 6.53378665e-01 -4.44612086e-01
-8.61257970e-01 -1.11812150e+00 -2.75984257e-01 1.71867108e+00
-9.39170271e-02 -4.67731595e-01 -7.52684534e-01 -2.40772232e-01
-5.56463182e-01 1.19053745e+00 -4.19587165e-01 -1.18840346e-02
-1.03768885e+00 -8.65072966e-01 5.78877389e-01 3.40547144e-01
3.29753220e-01 -1.43854392e+00 -9.44079816e-01 2.27679506e-01
1.27238378e-01 -4.47210580e-01 3.32779318e-01 7.35016108e-01
-1.05684173e+00 -1.16545558e+00 2.68808782e-01 -8.79172862e-01
8.26369584e-01 -4.84493136e-01 1.28177595e+00 6.94663048e-01
-3.92738402e-01 4.58278328e-01 -1.50806919e-01 -5.71271896e-01
-7.56572962e-01 -1.24088295e-01 1.17539376e-01 -7.33640313e-01
5.27601063e-01 -4.98159885e-01 2.21010402e-01 -1.52505443e-01
-1.14554799e+00 -4.04295534e-01 4.75255162e-01 9.51915622e-01
3.07481319e-01 2.04594299e-01 1.65965572e-01 -1.05102694e+00
5.14486790e-01 1.78730115e-02 -9.78219986e-01 4.50728983e-01
-7.82752216e-01 6.11927569e-01 3.72010350e-01 -3.19526881e-01
-1.03141260e+00 3.25990166e-03 -1.96390599e-01 6.51428282e-01
-2.31802002e-01 1.05809510e+00 -5.21738946e-01 8.86119977e-02
9.63108003e-01 5.44630229e-01 -2.05384061e-01 -8.97055119e-02
4.07936454e-01 1.64967015e-01 2.64872342e-01 -1.52701306e+00
7.20514297e-01 5.84574938e-02 1.09316327e-01 -9.03589845e-01
-3.19586933e-01 1.88726157e-01 -6.92372978e-01 6.15230240e-02
6.06619120e-01 -6.84815645e-01 -6.56751990e-01 8.09434578e-02
-1.05354095e+00 -5.23933649e-01 -1.93570599e-01 5.79536259e-01
-8.96320522e-01 3.96630973e-01 -5.20172238e-01 -1.16114414e+00
5.49403280e-02 -1.05608356e+00 4.93426144e-01 -2.93324366e-02
-3.53361577e-01 -7.35783815e-01 6.84105232e-02 -2.22581372e-01
1.86229602e-01 -2.49661431e-01 1.95698082e+00 -8.34621966e-01
-8.04612696e-01 7.14435428e-02 1.72235951e-01 3.31272066e-01
-4.28544790e-01 3.26566458e-01 -6.83186352e-01 -2.26687584e-02
1.25970170e-01 -5.46688259e-01 4.98310477e-01 2.42947355e-01
5.57820380e-01 -1.72553495e-01 -2.62100518e-01 8.67177844e-02
1.47411168e+00 7.11229861e-01 7.54584491e-01 1.71006769e-01
2.27403771e-02 6.75431609e-01 3.50843310e-01 -7.02327490e-02
3.54656100e-01 6.77979290e-02 3.44578326e-02 5.97220957e-01
4.75652486e-01 -8.65279064e-02 5.35790503e-01 9.69666183e-01
-3.46845001e-01 3.25416684e-01 -1.39505732e+00 6.12129450e-01
-1.35961819e+00 -9.21133876e-01 1.19226761e-01 2.14551449e+00
1.34382927e+00 6.76980734e-01 -1.31361678e-01 3.66763949e-01
5.89272738e-01 -1.45008340e-01 2.27429271e-02 -7.03246593e-01
-4.41513509e-02 1.14882565e+00 -1.56225804e-02 9.86451626e-01
-5.72640300e-01 1.14329898e+00 7.17366505e+00 5.25742710e-01
-6.27957344e-01 -1.37293428e-01 2.47061759e-01 5.11198878e-01
-7.25138009e-01 5.46753705e-01 -7.50428379e-01 -4.02750261e-02
1.13120544e+00 -1.87562421e-01 6.07111692e-01 1.00947106e+00
-4.96931672e-01 -5.13497710e-01 -1.52470589e+00 6.68543458e-01
-5.08107245e-02 -1.08631575e+00 -6.57152608e-02 -2.59584002e-02
2.34784141e-01 -2.62734354e-01 -1.55140713e-01 7.13152587e-01
9.14282084e-01 -1.33355367e+00 9.14897978e-01 1.20239362e-01
5.73152840e-01 -8.78348529e-01 4.80362296e-01 8.58090699e-01
-9.69811916e-01 -5.25545422e-03 -3.80657792e-01 -3.66496265e-01
-1.58939570e-01 4.11380708e-01 -4.22240466e-01 5.71171492e-02
5.32933950e-01 -2.06198320e-01 -5.04751384e-01 3.02953213e-01
-6.71307027e-01 6.50751472e-01 -5.35891652e-01 -4.68926966e-01
1.86043847e-02 -2.47830987e-01 3.26526463e-01 1.16911781e+00
1.59330174e-01 8.41487825e-01 2.64518827e-01 9.73816335e-01
3.71518075e-01 1.32630214e-01 -8.56322944e-01 -1.72779262e-01
3.46905082e-01 6.60477459e-01 -8.49738002e-01 -5.34179568e-01
-2.58378953e-01 -3.57099697e-02 2.65319079e-01 -3.42743360e-02
-3.62558961e-01 -5.86683810e-01 4.59628068e-02 9.02111605e-02
1.49356678e-01 -5.43546379e-01 1.38865680e-01 -1.24477005e+00
-2.14276388e-01 -1.29250646e+00 3.76215279e-01 -5.04208744e-01
-7.40495086e-01 6.07268572e-01 5.01035690e-01 -1.73488781e-01
-7.30836928e-01 -1.04391778e+00 -2.69075692e-01 1.09779489e+00
-1.10229099e+00 -7.44021535e-01 5.72668970e-01 8.36476386e-01
1.03712797e-01 -3.97849292e-01 1.24202859e+00 -2.24294990e-01
-6.55302331e-02 2.15768754e-01 -5.52060246e-01 2.41234496e-01
-3.74737307e-02 -1.52433360e+00 -1.34183355e-02 9.92737830e-01
4.39205915e-01 1.45034730e+00 9.30476248e-01 -5.42891622e-01
-1.66535509e+00 -2.43187770e-01 1.27017784e+00 -4.04213369e-01
6.29198730e-01 -3.03297728e-01 -9.50619578e-01 8.98763359e-01
1.93699539e-01 -5.60949370e-02 8.68770063e-01 4.09353107e-01
-8.87018204e-01 5.27520329e-02 -1.05524838e+00 7.05364466e-01
9.80265021e-01 -9.27676976e-01 -1.27932477e+00 4.67274450e-02
5.20698071e-01 -1.11832455e-01 -5.60994685e-01 2.28964910e-01
6.10132098e-01 -1.06645036e+00 5.38544655e-01 -7.60471046e-01
2.99929172e-01 -2.54666179e-01 -5.79520762e-01 -7.46875226e-01
5.63986935e-02 -6.80205822e-01 1.83561519e-01 7.08890975e-01
7.09729552e-01 -7.77945161e-01 5.70128679e-01 9.22269046e-01
-8.03019181e-02 -2.27730069e-02 -1.16558862e+00 -4.77933139e-01
5.70105016e-01 -9.29051936e-01 3.26110721e-01 7.19629288e-01
4.40588057e-01 2.50280589e-01 2.05232173e-01 2.93549094e-02
7.12899029e-01 3.94681364e-01 4.51990366e-01 -1.44016075e+00
-7.04635561e-01 -4.65131015e-01 -1.65029004e-01 -8.13783228e-01
4.92597103e-01 -1.15450191e+00 3.71500283e-01 -8.53514254e-01
2.27413207e-01 -6.93474114e-01 -1.46409661e-01 4.17033672e-01
1.39233351e-01 -4.01699662e-01 -4.20479476e-03 -6.14593811e-02
-2.90456355e-01 -4.98754904e-02 7.52796769e-01 2.69864071e-02
1.31706089e-01 -2.68516481e-01 -8.14532816e-01 1.27768898e+00
7.19465852e-01 -7.56415963e-01 -5.39729953e-01 -2.84405202e-01
1.18029118e+00 1.35635495e-01 2.91946650e-01 -5.23609936e-01
6.14344440e-02 -4.44563538e-01 -4.53350358e-02 -2.15453833e-01
-1.73867941e-01 -7.23323226e-01 -5.22942953e-02 7.81011820e-01
-6.55354857e-01 3.69755387e-01 1.27674252e-01 2.37528756e-01
-1.23916820e-01 -8.46357405e-01 6.84828997e-01 -8.00929070e-01
-7.96162546e-01 -2.11937666e-01 -5.47026873e-01 1.94287404e-01
5.90175569e-01 1.75973430e-01 -1.71151772e-01 1.05436426e-02
-9.52852070e-01 -1.90875947e-01 1.93176419e-01 -1.60201594e-01
6.45801067e-01 -8.20432603e-01 -3.65493447e-01 2.93809801e-01
1.40816808e-01 8.25861618e-02 -1.41790912e-01 2.34364912e-01
-7.57581055e-01 3.69615793e-01 -2.03167900e-01 -5.82890093e-01
-1.14226866e+00 4.93575513e-01 1.33861706e-01 -3.50424737e-01
-1.08847626e-01 9.91828144e-01 -9.74910706e-02 -6.13366008e-01
1.42822236e-01 -5.78752100e-01 6.50184229e-02 -3.28590691e-01
4.98097956e-01 -1.98966369e-01 -2.52509564e-01 -3.96474332e-01
-3.73117834e-01 4.58754838e-01 -1.78244546e-01 -6.19616508e-01
1.18043733e+00 1.21116638e-01 -9.06701624e-01 7.14552283e-01
6.77877605e-01 2.59148240e-01 -4.09178197e-01 -3.44143748e-01
4.56661612e-01 -2.85325587e-01 -3.48071873e-01 -6.05330944e-01
-3.19375694e-01 9.72489059e-01 4.16315347e-01 5.34324348e-01
8.50396574e-01 4.50888544e-01 -9.52551365e-02 9.59306538e-01
1.06240129e+00 -1.17011821e+00 -1.73964202e-01 6.95881844e-01
6.29175782e-01 -9.58722711e-01 1.63222194e-01 -3.82210433e-01
-1.85352430e-01 1.24292195e+00 2.84365296e-01 -2.27821562e-02
7.47009993e-01 3.22091639e-01 -2.60428518e-01 -4.44490999e-01
-9.70356166e-01 -3.99298042e-01 -7.92556107e-02 4.62739348e-01
6.80219471e-01 3.05472344e-01 -6.76162004e-01 6.34825170e-01
-6.07644498e-01 -2.92200536e-01 5.73079884e-01 1.46779716e+00
-1.02800083e+00 -1.51456344e+00 -4.04657573e-01 1.73180148e-01
-6.80602312e-01 -4.76090342e-01 -1.93989128e-01 1.22078013e+00
2.72315979e-01 8.60389590e-01 -2.68165350e-01 -4.39512841e-02
-1.49561977e-03 6.65506601e-01 1.02466452e+00 -9.70896184e-01
-5.44391692e-01 2.52702683e-01 4.90654968e-02 -2.26302952e-01
-7.82235622e-01 -7.86350012e-01 -1.63850343e+00 -2.35938698e-01
-3.73446256e-01 4.65876639e-01 6.22673750e-01 1.07891202e+00
-4.24981147e-01 -2.62052119e-02 1.28411323e-01 -2.80604027e-02
-7.85247564e-01 -8.09804320e-01 -9.01832640e-01 9.03409943e-02
6.85370266e-02 -4.93379116e-01 -6.58287704e-01 3.47619534e-01] | [8.884347915649414, 7.164343357086182] |
251ce38f-4680-4f02-8794-5e37aa229331 | low-variance-gradient-estimation-in-unrolled | 2304.11153 | null | https://arxiv.org/abs/2304.11153v1 | https://arxiv.org/pdf/2304.11153v1.pdf | Low-Variance Gradient Estimation in Unrolled Computation Graphs with ES-Single | We propose an evolution strategies-based algorithm for estimating gradients in unrolled computation graphs, called ES-Single. Similarly to the recently-proposed Persistent Evolution Strategies (PES), ES-Single is unbiased, and overcomes chaos arising from recursive function applications by smoothing the meta-loss landscape. ES-Single samples a single perturbation per particle, that is kept fixed over the course of an inner problem (e.g., perturbations are not re-sampled for each partial unroll). Compared to PES, ES-Single is simpler to implement and has lower variance: the variance of ES-Single is constant with respect to the number of truncated unrolls, removing a key barrier in applying ES to long inner problems using short truncations. We show that ES-Single is unbiased for quadratic inner problems, and demonstrate empirically that its variance can be substantially lower than that of PES. ES-Single consistently outperforms PES on a variety of tasks, including a synthetic benchmark task, hyperparameter optimization, training recurrent neural networks, and training learned optimizers. | ['Kevin Swersky', 'Zico Kolter', 'Paul Vicol'] | 2023-04-21 | null | null | null | null | ['hyperparameter-optimization'] | ['methodology'] | [ 4.67783697e-02 2.70427018e-02 6.46856949e-02 3.61165434e-01
-6.43955290e-01 -4.57280546e-01 2.76120424e-01 2.18608305e-01
-5.25887609e-01 9.15657699e-01 -6.33693933e-02 -2.88071111e-02
-1.87512845e-01 -6.22523904e-01 -8.14956784e-01 -1.03953373e+00
-1.44773066e-01 4.54443932e-01 2.86300421e-01 -2.70022005e-01
5.10904431e-01 5.34987450e-01 -1.28208411e+00 -3.92998993e-01
8.33384216e-01 8.94231141e-01 -3.05580720e-02 8.22371304e-01
2.07312390e-01 5.68245053e-01 -7.21600831e-01 -3.37774068e-01
1.46520706e-02 -5.63400865e-01 -4.91378427e-01 -2.62078553e-01
2.90717726e-04 3.59933048e-01 -4.54069912e-01 1.22109425e+00
8.20143938e-01 5.68635345e-01 6.95340574e-01 -9.41448808e-01
-5.11804223e-01 7.10571051e-01 -5.85586309e-01 2.95143366e-01
-3.24548423e-01 1.99814409e-01 9.13965344e-01 -8.30615997e-01
4.55335855e-01 1.06908739e+00 1.32579875e+00 4.50218409e-01
-1.51866853e+00 -2.10889846e-01 2.48775288e-01 -2.98617482e-01
-1.30847597e+00 -3.74759227e-01 7.22595811e-01 -1.74946070e-01
1.05964768e+00 5.82578003e-01 7.27690935e-01 6.71343863e-01
6.34224772e-01 6.50108635e-01 5.67093670e-01 -3.73995274e-01
6.67384982e-01 -1.08781159e-01 1.87410414e-01 8.03218186e-01
4.38998759e-01 1.25008702e-01 -4.78515923e-01 -5.10250390e-01
7.13098109e-01 -4.24176902e-01 -5.60560107e-01 -4.04792547e-01
-1.08422148e+00 9.08471406e-01 3.04057747e-01 2.01988503e-01
-3.99975866e-01 5.27969122e-01 5.29275775e-01 3.86946946e-01
7.01050460e-01 1.02057743e+00 -3.11237663e-01 -5.29898047e-01
-9.59250093e-01 3.89572859e-01 9.01396036e-01 6.16162717e-01
5.12386799e-01 3.84520203e-01 -4.61378366e-01 9.37732816e-01
-2.35993460e-01 5.07896662e-01 7.74853826e-01 -1.00608766e+00
1.83120608e-01 3.33460361e-01 2.00298473e-01 -9.61845994e-01
-6.06658518e-01 -9.18968260e-01 -1.05508041e+00 1.66444406e-01
2.21159831e-01 -4.78191942e-01 -6.95444167e-01 1.76662338e+00
2.44133085e-01 1.11948028e-01 -1.78286150e-01 6.21157050e-01
3.18786263e-01 7.87378848e-01 -6.07105076e-01 -5.91580689e-01
7.88123548e-01 -1.29476857e+00 -4.81174678e-01 -2.05832571e-01
4.66467261e-01 -4.05488700e-01 1.08439660e+00 3.65799844e-01
-1.45276189e+00 -4.56965305e-02 -1.13134205e+00 2.80861765e-01
-5.18107899e-02 1.04942322e-01 3.65621001e-01 6.36671901e-01
-1.22577536e+00 1.33110023e+00 -1.09067690e+00 1.90144569e-01
1.87224954e-01 3.97174776e-01 2.20512867e-01 6.89885259e-01
-7.95428216e-01 7.66913772e-01 1.46702588e-01 1.48468912e-01
-6.63700759e-01 -9.76954281e-01 -5.86173236e-01 2.85385549e-01
3.29024762e-01 -6.97463036e-01 1.06357384e+00 -6.80545449e-01
-2.16174889e+00 2.21207440e-01 -4.68683690e-01 -6.53901994e-01
6.07877195e-01 -2.42975298e-02 5.35810068e-02 -3.12659472e-01
-3.25551957e-01 7.85033926e-02 1.25567281e+00 -9.19177115e-01
6.71110302e-02 -3.31747591e-01 -3.86050731e-01 3.04778516e-01
-5.13803661e-01 -2.40213960e-01 -4.49930549e-01 -7.52351522e-01
9.14188176e-02 -1.31800747e+00 -5.54307163e-01 -4.12855327e-01
-4.08314168e-01 -3.34010497e-02 5.15513003e-01 -3.42909664e-01
1.69612134e+00 -1.97171330e+00 7.64609039e-01 2.61755407e-01
3.48492533e-01 4.18011725e-01 -1.22631960e-01 4.00679231e-01
1.16464987e-01 1.95239216e-01 -6.50274813e-01 -5.39348185e-01
5.75404279e-02 4.13856767e-02 -3.62664908e-01 4.28432047e-01
-1.27818272e-01 1.06923926e+00 -8.81713450e-01 8.82147849e-02
-1.86684698e-01 5.31931102e-01 -6.04025662e-01 -7.45545626e-02
-3.61988634e-01 1.43936858e-01 -3.70948404e-01 2.16358438e-01
4.72334206e-01 -7.06338108e-01 3.76227535e-02 9.67793539e-02
-1.51163980e-01 2.51823515e-01 -1.16408968e+00 1.24382043e+00
-5.78302741e-01 7.83480227e-01 -3.73607650e-02 -8.88684750e-01
8.27912092e-01 -1.51929474e-02 3.21994185e-01 -4.11519587e-01
4.09058854e-02 2.06492543e-01 -2.98697948e-01 1.50225740e-02
6.85100734e-01 3.39469492e-01 2.16876075e-01 7.01482296e-01
-1.48308218e-01 -2.70458549e-01 3.83207917e-01 8.87376815e-02
1.32217789e+00 -1.89271063e-01 2.14910120e-01 -4.19908345e-01
4.66687083e-01 -2.90067852e-01 6.97526217e-01 1.05031168e+00
-1.80524420e-02 6.02617741e-01 6.68304741e-01 -2.82950729e-01
-1.13993084e+00 -8.03044379e-01 -2.32023269e-01 9.06115413e-01
3.14525276e-01 -7.44034290e-01 -8.92929375e-01 -5.25553107e-01
2.12822497e-01 7.47786999e-01 -7.62232840e-01 -3.02961528e-01
-7.94716358e-01 -1.28366792e+00 5.29465795e-01 1.50893748e-01
3.15243930e-01 -1.03497279e+00 -8.07210445e-01 4.92278278e-01
2.13909522e-01 -4.47175801e-01 -8.48588526e-01 2.17886209e-01
-1.04643166e+00 -7.91261315e-01 -9.96939600e-01 -4.21462655e-01
5.91867864e-01 -1.69520706e-01 1.27567542e+00 -9.78486612e-03
-3.48661155e-01 2.43562087e-01 1.93375424e-01 -1.22815631e-01
-3.20404470e-01 3.36201757e-01 -7.96144307e-02 -6.95584789e-02
-2.56862700e-01 -6.86277270e-01 -6.55482590e-01 3.83260816e-01
-4.64140266e-01 -3.02260637e-01 2.31668949e-01 1.34515011e+00
7.55876541e-01 2.85280924e-02 3.74072433e-01 -5.77582896e-01
1.23448086e+00 -3.08203906e-01 -8.48786116e-01 4.98268098e-01
-8.73320818e-01 6.11005843e-01 6.70891225e-01 -7.54069448e-01
-6.89925134e-01 -4.39168513e-01 1.33621961e-01 -5.42634428e-01
7.12535083e-01 4.06014800e-01 7.28613913e-01 -5.64725518e-01
6.59843743e-01 3.40092838e-01 1.42598331e-01 -3.29119086e-01
1.46618888e-01 1.79171965e-01 3.67029637e-01 -5.71780980e-01
4.09011900e-01 2.31183216e-01 2.42807418e-01 -1.04568350e+00
-5.78917384e-01 6.22344390e-02 2.14221068e-02 -1.68392807e-01
3.39511931e-01 -2.63885289e-01 -1.02930558e+00 6.06519401e-01
-9.53420937e-01 -6.54816329e-01 -6.28070235e-01 2.17260510e-01
-4.80533719e-01 2.04135701e-01 -7.85439193e-01 -1.00567615e+00
-7.65473008e-01 -1.20295417e+00 9.24673259e-01 3.57082635e-01
-1.83358237e-01 -1.23316610e+00 4.72135812e-01 -4.57584113e-01
8.26869905e-01 1.80938154e-01 8.12345266e-01 -1.59935087e-01
-2.24848568e-01 -8.58255401e-02 3.06944221e-01 2.89775282e-01
-7.83996060e-02 2.83603609e-01 -3.21376681e-01 -5.80276012e-01
3.50475580e-01 -3.42394412e-01 1.04374528e+00 7.36669123e-01
1.13863075e+00 -4.51538265e-01 -3.75855416e-01 8.59306395e-01
1.16209400e+00 9.47815701e-02 5.01469791e-01 3.68152946e-01
3.59373212e-01 3.84830195e-03 1.67865455e-01 4.97383088e-01
-7.79274013e-03 6.07331038e-01 1.17833883e-01 4.11216766e-02
4.65086140e-02 8.66401866e-02 4.41630393e-01 9.95862663e-01
-1.10796541e-01 -1.67071417e-01 -6.56583965e-01 3.20879519e-01
-2.08631396e+00 -9.67661023e-01 1.09634273e-01 2.45544338e+00
8.60684872e-01 1.46682039e-01 1.50686279e-01 -7.67065436e-02
7.91970313e-01 2.85269260e-01 -1.18607783e+00 -5.21319330e-01
-3.67002249e-01 2.73363441e-01 7.07009017e-01 7.27030337e-01
-7.96903074e-01 7.48743594e-01 7.12869072e+00 9.05800641e-01
-1.29059815e+00 2.78351475e-02 5.94606161e-01 -5.03119886e-01
-4.32391256e-01 -9.71779823e-02 -7.86721647e-01 7.19050288e-01
8.35444391e-01 -5.70531368e-01 9.78254437e-01 6.97047651e-01
-8.97073150e-02 -1.17616979e-02 -8.29260826e-01 9.61268485e-01
5.53446775e-03 -1.60988247e+00 -3.40948582e-01 -1.24130294e-01
1.09605110e+00 4.27262127e-01 4.13503319e-01 1.61973223e-01
3.72721434e-01 -9.69144464e-01 4.56303954e-01 4.49741989e-01
4.65066314e-01 -9.21111286e-01 5.09684861e-01 2.95702010e-01
-1.01966047e+00 -1.82436824e-01 -4.97849435e-01 2.40140140e-01
2.04898328e-01 9.10828769e-01 -3.67721349e-01 3.49619448e-01
5.66316962e-01 5.52482963e-01 -4.80979860e-01 1.13881552e+00
-1.42976595e-03 6.83196008e-01 -7.15825319e-01 -4.85053003e-01
1.71473756e-01 -6.22415841e-01 1.20727539e+00 9.31552172e-01
3.62284929e-01 -1.75459921e-01 -2.58214295e-01 9.46718872e-01
-1.27705857e-01 -8.25255457e-03 -2.95707017e-01 -2.28004426e-01
6.32806122e-01 9.32572901e-01 -6.05513692e-01 -1.73634276e-01
2.23900124e-01 1.04589176e+00 6.16723418e-01 6.42186284e-01
-7.67768681e-01 -8.95267487e-01 5.00674367e-01 -2.69905150e-01
6.38312101e-01 -2.10153550e-01 -3.40872228e-01 -1.17600715e+00
5.88236824e-02 -7.83230662e-01 2.27796629e-01 -3.88094276e-01
-1.09412241e+00 9.07250106e-01 -2.80250907e-01 -7.19399333e-01
-3.35892230e-01 -2.96643049e-01 -8.05565417e-01 7.35946238e-01
-1.08430529e+00 -1.82993367e-01 -3.81038734e-03 1.86712697e-01
3.74929667e-01 -1.67576566e-01 6.66262209e-01 -2.07388669e-01
-1.07644451e+00 7.45321095e-01 6.96344733e-01 -4.54736441e-01
3.22253227e-01 -1.15443873e+00 7.40119755e-01 8.16269338e-01
6.88282177e-02 7.65571535e-01 8.95309091e-01 -6.50861740e-01
-1.53239310e+00 -7.25970268e-01 5.48350036e-01 -4.37756963e-02
7.47830987e-01 -2.16092274e-01 -1.01114893e+00 4.89688605e-01
1.75163709e-02 -3.91344838e-02 2.01919109e-01 2.03624427e-01
5.34347780e-02 -1.98203381e-02 -9.50390160e-01 8.89454186e-01
1.04860914e+00 -2.80042917e-01 -2.34889656e-01 4.47386473e-01
5.60239971e-01 -7.02886045e-01 -9.12544310e-01 5.69214821e-01
4.06863362e-01 -1.03953779e+00 9.50549364e-01 -3.79271179e-01
1.06796948e-02 -4.41125222e-02 3.01766217e-01 -1.73990738e+00
-3.19941789e-01 -1.39700806e+00 -6.71011508e-01 6.48991287e-01
5.73489487e-01 -1.25151145e+00 7.88701713e-01 4.27680999e-01
-6.73670769e-02 -1.27877390e+00 -9.51650620e-01 -1.01333380e+00
2.30412453e-01 1.68980971e-01 6.03247285e-01 6.75734520e-01
-9.24269035e-02 2.94234186e-01 -3.91513079e-01 5.55928331e-03
8.06054533e-01 1.71266362e-01 5.89322209e-01 -1.05193019e+00
-8.30769360e-01 -9.44525838e-01 -5.65823261e-03 -1.09738934e+00
9.97511074e-02 -6.60569072e-01 -1.82941973e-01 -1.02235603e+00
-1.44134145e-02 -4.93717521e-01 -2.44988739e-01 8.75391886e-02
-4.74657446e-01 5.75120263e-02 4.57999893e-02 3.36938947e-01
-3.57824773e-01 1.08878160e+00 1.15298092e+00 5.55074178e-02
-5.80253303e-01 2.20272183e-01 -3.43317986e-01 6.34007394e-01
7.79368818e-01 -5.63602746e-01 -3.62440437e-01 -2.30155870e-01
5.05576849e-01 1.57736037e-02 -7.71420076e-02 -8.55607748e-01
4.26441312e-01 5.36740609e-02 4.48918864e-02 -4.19945866e-01
2.96402991e-01 5.37702218e-02 2.28351757e-01 5.90433478e-01
-4.69761312e-01 4.30000603e-01 2.86631912e-01 5.36639333e-01
-6.80438131e-02 -4.59951460e-01 8.06739569e-01 5.31369075e-02
4.92251478e-02 2.70633221e-01 -1.23184152e-01 4.31446135e-01
7.55310476e-01 5.33692539e-02 -4.01399106e-01 -3.09550554e-01
-6.76961362e-01 4.35826868e-01 5.99778652e-01 -5.90132140e-02
5.22499859e-01 -1.07703972e+00 -5.09718418e-01 4.83450927e-02
-5.71535468e-01 -5.17233275e-02 1.08526222e-01 9.33900118e-01
-6.11208320e-01 2.32699364e-01 3.40190560e-01 -6.28185511e-01
-1.17112970e+00 3.96444052e-01 6.06407046e-01 -6.37261391e-01
-8.73878062e-01 1.07566631e+00 -1.78001314e-01 -3.54877770e-01
3.78893942e-01 -6.68782890e-02 1.72681645e-01 -2.63060838e-01
5.29187918e-01 7.50128865e-01 1.67008415e-01 2.90790852e-03
-1.76626548e-01 7.45461345e-01 -4.44475636e-02 -2.68359691e-01
1.44616187e+00 1.03283271e-01 -4.02285010e-01 7.42227197e-01
1.20918393e+00 -7.59118609e-03 -1.41121209e+00 -1.26067877e-01
-1.74084336e-01 -1.58952430e-01 2.83084214e-01 -4.92792040e-01
-1.03441584e+00 6.14433706e-01 2.34895378e-01 2.72361815e-01
9.30320680e-01 -4.73115951e-01 7.40328133e-01 6.76800072e-01
2.72924274e-01 -1.17350507e+00 1.65246323e-01 7.93793321e-01
1.08436215e+00 -7.01235652e-01 8.14401209e-02 -1.35876969e-01
-5.48685491e-01 1.03270197e+00 1.78060010e-01 -2.60336697e-01
5.47983587e-01 4.85360414e-01 -5.04115999e-01 -1.89050615e-01
-9.64490294e-01 3.11844587e-01 3.65401357e-01 5.70919691e-03
1.17727540e-01 -1.70675337e-01 -3.33455980e-01 4.75283079e-02
-2.88962364e-01 -2.50573486e-01 2.70034492e-01 8.24021459e-01
-3.31930250e-01 -9.03984487e-01 -1.68566674e-01 5.21677196e-01
-2.75128007e-01 -2.35708788e-01 -2.08435237e-01 4.99320626e-01
-6.11232460e-01 5.00786841e-01 1.72907948e-01 -2.38379866e-01
5.30478060e-02 -2.54966896e-02 6.50320351e-01 -1.12591751e-01
-8.56952429e-01 -2.17446119e-01 1.26221369e-03 -6.50515914e-01
1.26493126e-01 -6.49564624e-01 -9.42368031e-01 -4.65640545e-01
-4.70595032e-01 5.55244088e-01 5.94452143e-01 6.75085843e-01
7.84352064e-01 7.30067194e-01 7.89541304e-01 -1.06989217e+00
-1.20182335e+00 -7.41063654e-01 -3.56219441e-01 -2.62496620e-02
4.89831686e-01 -6.24670982e-01 -9.18731570e-01 -6.97802544e-01] | [7.202095985412598, 3.8753445148468018] |
aa53d0b4-eb68-4238-a2a3-4b9f92c8eae0 | a-unified-transformer-framework-for-group | 2203.04708 | null | https://arxiv.org/abs/2203.04708v2 | https://arxiv.org/pdf/2203.04708v2.pdf | A Unified Transformer Framework for Group-based Segmentation: Co-Segmentation, Co-Saliency Detection and Video Salient Object Detection | Humans tend to mine objects by learning from a group of images or several frames of video since we live in a dynamic world. In the computer vision area, many researches focus on co-segmentation (CoS), co-saliency detection (CoSD) and video salient object detection (VSOD) to discover the co-occurrent objects. However, previous approaches design different networks on these similar tasks separately, and they are difficult to apply to each other, which lowers the upper bound of the transferability of deep learning frameworks. Besides, they fail to take full advantage of the cues among inter- and intra-feature within a group of images. In this paper, we introduce a unified framework to tackle these issues, term as UFO (Unified Framework for Co-Object Segmentation). Specifically, we first introduce a transformer block, which views the image feature as a patch token and then captures their long-range dependencies through the self-attention mechanism. This can help the network to excavate the patch structured similarities among the relevant objects. Furthermore, we propose an intra-MLP learning module to produce self-mask to enhance the network to avoid partial activation. Extensive experiments on four CoS benchmarks (PASCAL, iCoseg, Internet and MSRC), three CoSD benchmarks (Cosal2015, CoSOD3k, and CocA) and four VSOD benchmarks (DAVIS16, FBMS, ViSal and SegV2) show that our method outperforms other state-of-the-arts on three different tasks in both accuracy and speed by using the same network architecture , which can reach 140 FPS in real-time. | ['Qingyao Wu', 'Guosheng Lin', 'Ruizhou Sun', 'Jingliang Deng', 'Yukun Su'] | 2022-03-09 | null | null | null | null | ['co-saliency-detection', 'video-salient-object-detection'] | ['computer-vision', 'computer-vision'] | [ 2.18013838e-01 -1.93486050e-01 -2.08227813e-01 -2.41867140e-01
-2.92916268e-01 -1.51854038e-01 4.00434285e-01 2.34145224e-02
-6.08895779e-01 4.73768860e-01 -3.46513242e-02 1.23243898e-01
-4.13842909e-02 -5.53915620e-01 -8.99558783e-01 -4.87512946e-01
6.17386661e-02 -8.73287469e-02 1.09168470e+00 -2.15733066e-01
2.96434402e-01 2.55955070e-01 -1.83034670e+00 4.18025523e-01
9.09063458e-01 1.11450887e+00 7.22573280e-01 3.32252920e-01
-8.49745721e-02 7.89454997e-01 -5.11091530e-01 -1.26054019e-01
1.76499367e-01 -3.49267632e-01 -9.37033713e-01 1.67514831e-01
3.94117147e-01 -2.61589527e-01 -3.61595988e-01 1.33797908e+00
4.04480070e-01 1.50449187e-01 4.21807319e-01 -1.36784852e+00
-6.00554228e-01 7.03198910e-01 -9.44510341e-01 7.48992324e-01
1.46076173e-01 1.54903129e-01 8.87270451e-01 -9.67072248e-01
6.88318491e-01 1.49283433e+00 4.83284593e-01 4.61001575e-01
-6.97329462e-01 -5.11132717e-01 5.20072162e-01 5.81298351e-01
-1.23312616e+00 -1.52397335e-01 7.96433628e-01 -1.68387875e-01
7.80962408e-01 2.18389735e-01 7.48908877e-01 9.47305560e-01
2.18381025e-02 1.47712362e+00 8.76339972e-01 -1.51315406e-01
3.08307596e-02 4.00345474e-02 3.08169454e-01 6.23350203e-01
3.35789382e-01 -9.10373777e-02 -5.54042578e-01 3.38249356e-01
1.00933075e+00 3.45600814e-01 -3.08597833e-01 -2.90197581e-01
-1.31559694e+00 6.22702420e-01 7.24220693e-01 5.48464119e-01
-3.53562564e-01 -4.06487798e-03 4.05953616e-01 3.54111455e-02
2.16422483e-01 9.78443250e-02 -3.72948289e-01 1.13324113e-01
-7.92276621e-01 2.17988685e-01 5.08874536e-01 9.82298136e-01
7.88014293e-01 -3.63712013e-03 -2.85969377e-01 7.08612263e-01
3.06235105e-01 2.29674041e-01 6.75954640e-01 -7.38360941e-01
5.45316875e-01 7.25043237e-01 1.03057981e-01 -1.25196922e+00
-3.92411679e-01 -4.37486857e-01 -7.77316868e-01 -4.06684652e-02
2.27050155e-01 -2.83893291e-02 -1.15690565e+00 1.49683595e+00
4.19563204e-01 5.57658732e-01 -1.26607567e-02 1.37444758e+00
1.19187272e+00 6.86580658e-01 2.10286736e-01 -1.39647812e-01
1.34341168e+00 -1.53415847e+00 -5.70921361e-01 -5.56416392e-01
2.87970424e-01 -6.47408485e-01 1.06349528e+00 1.59606814e-01
-1.06879056e+00 -1.05342364e+00 -1.11972809e+00 -1.43657357e-01
-4.67675596e-01 -1.01101547e-01 6.06506169e-01 1.68225858e-02
-8.25999320e-01 5.41057706e-01 -7.63286352e-01 -3.42775166e-01
7.72897422e-01 3.30370039e-01 5.90500236e-03 4.38706428e-02
-1.24276197e+00 5.06979167e-01 8.06395769e-01 2.45023236e-01
-1.07447338e+00 -4.23443705e-01 -7.56371558e-01 2.06481755e-01
6.40378296e-01 -4.28232551e-01 9.57093239e-01 -1.60710204e+00
-1.06132555e+00 8.47129941e-01 -1.12949568e-03 -5.95504642e-01
5.20523548e-01 -5.42977095e-01 -5.00720263e-01 3.85750175e-01
2.81650484e-01 1.06357944e+00 9.32358801e-01 -1.21435201e+00
-1.12926817e+00 -1.66741684e-01 2.34254822e-01 4.26163465e-01
-2.17195913e-01 2.16721281e-01 -8.92491937e-01 -6.54823005e-01
5.43824509e-02 -5.99711299e-01 -2.37616658e-01 -1.47943228e-01
-5.25533557e-01 -5.22836924e-01 1.23506927e+00 -4.10657018e-01
1.12498844e+00 -2.36532712e+00 1.41947865e-01 -2.76943922e-01
1.73994005e-01 6.90660298e-01 -1.97318614e-01 -9.13718939e-02
-4.12692875e-02 2.93625072e-02 -3.00232202e-01 -2.26992607e-01
-2.97545791e-01 1.27798483e-01 -9.92054194e-02 2.83798277e-01
3.54348511e-01 1.17408609e+00 -1.04147804e+00 -8.55047703e-01
2.39497185e-01 3.43107402e-01 -3.01764697e-01 1.38710782e-01
-2.03466520e-01 5.52492067e-02 -6.40732825e-01 7.19110191e-01
7.30216563e-01 -4.63802993e-01 -1.71298072e-01 -2.60291994e-01
-1.31386817e-01 -6.75272942e-02 -1.40571511e+00 1.67425025e+00
9.89940092e-02 3.60629559e-01 9.07875896e-02 -1.21197951e+00
8.34723890e-01 -2.90079750e-02 4.02994275e-01 -7.91544020e-01
2.87562609e-01 2.95148015e-01 -9.57983583e-02 -7.47041166e-01
4.67349440e-01 4.88869220e-01 2.64526039e-01 1.54231176e-01
2.19308823e-01 4.40327078e-01 4.33701217e-01 1.93925679e-01
6.53833628e-01 8.29542801e-02 1.96385562e-01 -3.78888547e-01
6.57812119e-01 -1.43762097e-01 7.99356520e-01 6.84474051e-01
-5.93914866e-01 7.26535678e-01 4.65211123e-01 -5.49690366e-01
-6.88775063e-01 -9.03398752e-01 4.98444997e-02 1.08132899e+00
1.16705775e+00 -2.03127593e-01 -8.00997734e-01 -1.10032153e+00
-4.34395187e-02 4.20884848e-01 -6.97959840e-01 -2.39272490e-01
-7.10803330e-01 -7.90549338e-01 1.18745744e-01 6.86902761e-01
9.85132217e-01 -1.67177379e+00 -9.77949321e-01 2.19498649e-01
-2.33457074e-01 -1.30533862e+00 -6.60144985e-01 8.19531381e-02
-6.40663564e-01 -1.13401341e+00 -7.88191199e-01 -1.25687277e+00
4.65884209e-01 8.55623722e-01 9.91149724e-01 1.86276436e-01
-2.53285855e-01 1.86814647e-02 -3.26546460e-01 -4.91874188e-01
2.04456419e-01 8.97536203e-02 -8.69483054e-02 2.83072412e-01
6.41198337e-01 -4.04823482e-01 -7.42471933e-01 5.91694295e-01
-9.48998094e-01 3.21517766e-01 6.57241821e-01 7.20216870e-01
7.71348059e-01 -1.60475284e-01 7.12892592e-01 -7.32863426e-01
3.07468057e-01 -5.70413709e-01 -4.69669819e-01 2.20143363e-01
-3.75776082e-01 -3.40176404e-01 4.29409415e-01 -4.06652451e-01
-9.58504260e-01 1.15680531e-01 1.58310473e-01 -7.27645934e-01
-2.89834023e-01 3.65416825e-01 -2.14741752e-01 6.97608665e-02
3.44990313e-01 3.85637969e-01 -2.65752196e-01 -4.73880142e-01
8.34840387e-02 6.01678073e-01 7.89001644e-01 -2.04811886e-01
6.28020525e-01 4.91161078e-01 -3.41420740e-01 -6.06663346e-01
-1.19404995e+00 -6.18877888e-01 -6.71541691e-01 -2.21395507e-01
1.15048122e+00 -1.04106867e+00 -4.51087028e-01 6.05054200e-01
-1.17910767e+00 -1.31247640e-01 -2.33933628e-01 3.83572340e-01
-3.33697885e-01 4.19605881e-01 -5.26790261e-01 -5.39429009e-01
-4.53644902e-01 -1.22216558e+00 1.14523959e+00 9.42223907e-01
2.67497480e-01 -6.43532395e-01 -4.36459720e-01 1.91801056e-01
2.40132928e-01 2.63574213e-01 3.58105928e-01 -7.01241612e-01
-8.66697729e-01 2.86282539e-01 -7.74544418e-01 3.76452744e-01
4.29563448e-02 -1.10723041e-01 -7.89714813e-01 -2.34869853e-01
-5.62139861e-02 -3.73891324e-01 1.14937961e+00 4.90613401e-01
1.50303876e+00 -1.09237127e-01 -5.60019612e-01 6.68917537e-01
1.41388047e+00 3.06280017e-01 5.93307674e-01 4.68058348e-01
8.01414609e-01 5.42602420e-01 8.30728710e-01 1.72201127e-01
3.31807971e-01 5.44158101e-01 7.59050190e-01 -4.22620744e-01
-1.30136520e-01 -1.68772340e-01 3.12996566e-01 6.71516776e-01
-2.21325040e-01 -1.53540716e-01 -5.65685451e-01 8.62910032e-01
-2.26099300e+00 -9.25469220e-01 -1.88466728e-01 1.75242555e+00
6.01758897e-01 4.61332291e-01 2.72100985e-01 -9.32178199e-02
1.10369825e+00 3.40403467e-01 -7.53773630e-01 5.82420314e-03
-3.98130983e-01 -2.00420216e-01 3.96596730e-01 3.17772515e-02
-1.52407110e+00 1.11073935e+00 5.50775766e+00 1.03106344e+00
-1.16285622e+00 1.52532399e-01 9.21950161e-01 1.09785877e-01
5.22679202e-02 -2.33697921e-01 -9.56894398e-01 6.68118775e-01
4.28176701e-01 -1.86970048e-02 1.77356482e-01 1.10378122e+00
2.42294073e-02 -2.46521696e-01 -8.73070359e-01 1.11374402e+00
1.98959112e-01 -1.37911880e+00 1.27706844e-02 -4.11651611e-01
8.45659494e-01 2.52349615e-01 -2.28170082e-02 2.44459257e-01
-3.54080908e-02 -8.65093052e-01 8.71488452e-01 3.57148826e-01
4.03120100e-01 -8.34642172e-01 9.30472434e-01 3.25079888e-01
-1.32933700e+00 -6.52261674e-02 -6.26973867e-01 2.02554733e-01
1.50199786e-01 4.15190160e-01 -4.18086857e-01 6.92948759e-01
1.27536321e+00 1.12285089e+00 -7.57053554e-01 1.12449765e+00
-2.06409961e-01 5.39079607e-01 -2.48700500e-01 -2.21912876e-01
6.69700921e-01 -3.39187533e-02 5.74804425e-01 1.24523079e+00
4.23092172e-02 1.44507019e-02 3.26411158e-01 9.68923092e-01
-7.71887824e-02 -2.48345397e-02 -8.61853138e-02 1.75718829e-01
2.68619448e-01 1.44367218e+00 -1.20946836e+00 -5.59542537e-01
-5.53565025e-01 1.01297426e+00 2.05834076e-01 3.37611556e-01
-1.08588076e+00 -6.10559642e-01 3.63362789e-01 4.10719663e-02
8.29002321e-01 3.89739231e-04 -6.72421679e-02 -1.17809045e+00
1.54487729e-01 -8.31812799e-01 5.36017597e-01 -6.97680593e-01
-1.09680855e+00 6.53155506e-01 2.92585380e-02 -1.18111885e+00
2.71932036e-01 -3.20831865e-01 -7.49280453e-01 5.00296235e-01
-1.80711913e+00 -9.79294121e-01 -5.74185431e-01 7.83984184e-01
9.82679963e-01 -7.56959617e-02 2.59797927e-02 3.29159826e-01
-8.08590233e-01 3.34544718e-01 -1.40108287e-01 2.23653361e-01
4.87279534e-01 -9.94060338e-01 3.91448408e-01 1.07918108e+00
3.19012552e-01 2.86468238e-01 3.63254547e-01 -6.01016283e-01
-1.00863206e+00 -1.26248980e+00 5.46165228e-01 -1.04441464e-01
3.46000373e-01 -2.76031017e-01 -1.15490079e+00 4.98866051e-01
2.81623870e-01 4.52477098e-01 -9.49894264e-02 -3.36011052e-01
-6.80526299e-03 -4.07225825e-02 -9.10777211e-01 5.09762585e-01
1.38885021e+00 -1.53016791e-01 -7.77626932e-01 3.04040760e-01
1.22597528e+00 -3.48756373e-01 -3.34643334e-01 5.27749062e-01
1.91014826e-01 -1.25642085e+00 1.04508233e+00 -6.17704391e-01
5.97858608e-01 -6.86595380e-01 -3.13901603e-02 -8.87156904e-01
-3.47853839e-01 -5.88802218e-01 -3.29791069e-01 1.35767198e+00
9.23163518e-02 -3.56062084e-01 7.14487135e-01 3.99104208e-02
-3.12476754e-01 -1.05384767e+00 -6.85009956e-01 -6.76153123e-01
-4.72260028e-01 -1.74912184e-01 7.00385749e-01 8.50248873e-01
-4.07842278e-01 3.64434868e-01 -3.28108042e-01 9.63258892e-02
5.03096044e-01 3.56569916e-01 5.44758379e-01 -1.07985067e+00
-1.80177510e-01 -4.88976717e-01 -4.76433963e-01 -1.27786171e+00
-1.66270822e-01 -6.86843991e-01 7.41767213e-02 -1.42227387e+00
4.28326964e-01 -2.75811225e-01 -7.06613719e-01 5.42256296e-01
-6.51833713e-01 2.17099816e-01 4.18576419e-01 2.54820436e-01
-1.29235697e+00 6.04989707e-01 1.44960558e+00 -1.46648496e-01
-2.84598589e-01 -1.00652106e-01 -7.49001086e-01 9.97876883e-01
5.62164009e-01 -4.24743295e-01 -2.84453988e-01 -5.89324355e-01
-2.75404334e-01 -2.59028852e-01 5.58725536e-01 -1.30259228e+00
3.12336534e-01 -3.27410668e-01 4.60029513e-01 -8.81679296e-01
3.80352326e-02 -5.79817176e-01 -2.58490771e-01 5.09995043e-01
-1.65639624e-01 2.33550947e-02 1.55125007e-01 6.79972768e-01
-4.82996434e-01 -2.31205612e-01 9.70034540e-01 -1.65799990e-01
-1.46079862e+00 5.19456208e-01 2.29640603e-02 3.98656487e-01
1.23645890e+00 -2.61056602e-01 -2.86143422e-01 -4.96159531e-02
-4.07996267e-01 5.66817164e-01 1.39016718e-01 7.42452979e-01
8.21181417e-01 -1.29327083e+00 -5.53881288e-01 1.91505104e-01
-1.35936709e-02 4.42770720e-01 5.21199167e-01 9.36077535e-01
-4.79642838e-01 2.62942791e-01 -4.78586823e-01 -8.74472678e-01
-1.15737724e+00 8.17687035e-01 3.27142328e-01 -1.53561100e-01
-6.71725631e-01 1.06336582e+00 6.00553930e-01 1.10007443e-01
3.47688586e-01 -2.57581353e-01 -4.82032776e-01 5.46195172e-02
6.24912322e-01 1.84931442e-01 -2.22249702e-01 -7.55730748e-01
-4.56485391e-01 5.76804042e-01 -2.76627630e-01 4.96752232e-01
1.35151184e+00 -2.61262089e-01 1.13501595e-02 2.31385484e-01
1.16976452e+00 -4.68071759e-01 -1.60036087e+00 -4.91705835e-01
4.50821146e-02 -4.38013583e-01 -1.27103999e-01 -4.61933255e-01
-1.39844966e+00 7.74761379e-01 6.10444844e-01 2.65137702e-01
1.18112123e+00 1.16162598e-01 1.00899088e+00 2.20725968e-01
1.38080671e-01 -1.33959949e+00 3.93568933e-01 3.87829095e-01
8.14172149e-01 -1.36033976e+00 -9.89948958e-02 -5.44801950e-01
-7.77811289e-01 9.43887889e-01 1.13748169e+00 -3.62520188e-01
6.83319330e-01 -2.57703606e-02 -1.19841307e-01 -1.56145290e-01
-4.93654847e-01 -4.02543545e-01 4.16073233e-01 4.24623907e-01
3.62094305e-02 -2.13822931e-01 -3.17819446e-01 7.41740942e-01
2.30492696e-01 -5.71955275e-03 4.08243537e-01 9.27975118e-01
-7.69315779e-01 -5.31737804e-01 -1.86329558e-01 4.60176408e-01
-6.12606645e-01 8.88019651e-02 -2.59376884e-01 9.71321821e-01
4.67557967e-01 7.92031229e-01 1.10190719e-01 -4.29596514e-01
1.84604555e-01 -3.95256221e-01 -1.65812373e-02 -4.97452766e-01
-4.57763493e-01 2.12786913e-01 -2.12801293e-01 -7.35973597e-01
-9.54251170e-01 -6.40645206e-01 -1.33612752e+00 9.66591835e-02
-4.26155865e-01 7.74844736e-02 1.52160063e-01 9.98685896e-01
4.29028869e-01 7.80610323e-01 5.03472149e-01 -8.82334590e-01
-1.79035738e-01 -7.64873087e-01 -6.19183362e-01 6.35023713e-01
1.52102962e-01 -8.44904780e-01 -1.89848423e-01 9.54389870e-02] | [9.610576629638672, -0.2885572016239166] |
e04afb92-52ad-4cf1-b8c5-71ca13ea4fa2 | scientific-language-models-for-biomedical | 2106.09700 | null | https://arxiv.org/abs/2106.09700v2 | https://arxiv.org/pdf/2106.09700v2.pdf | Scientific Language Models for Biomedical Knowledge Base Completion: An Empirical Study | Biomedical knowledge graphs (KGs) hold rich information on entities such as diseases, drugs, and genes. Predicting missing links in these graphs can boost many important applications, such as drug design and repurposing. Recent work has shown that general-domain language models (LMs) can serve as "soft" KGs, and that they can be fine-tuned for the task of KG completion. In this work, we study scientific LMs for KG completion, exploring whether we can tap into their latent knowledge to enhance biomedical link prediction. We evaluate several domain-specific LMs, fine-tuning them on datasets centered on drugs and diseases that we represent as KGs and enrich with textual entity descriptions. We integrate the LM-based models with KG embedding models, using a router method that learns to assign each input example to either type of model and provides a substantial boost in performance. Finally, we demonstrate the advantage of LM models in the inductive setting with novel scientific entities. Our datasets and code are made publicly available. | ['Tom Hope', 'Hannaneh Hajishirzi', 'Noah A. Smith', 'Iz Beltagy', 'David Wadden', 'Rahul Nadkarni'] | 2021-06-17 | null | https://openreview.net/forum?id=4Exq_UvWKY8 | https://openreview.net/pdf?id=4Exq_UvWKY8 | akbc-2021-10 | ['knowledge-base-completion', 'knowledge-base-completion'] | ['graphs', 'knowledge-base'] | [-3.13303508e-02 7.64530361e-01 -8.99813414e-01 -2.19919384e-01
-7.00906336e-01 -5.71869671e-01 3.74150306e-01 6.97477162e-01
-1.48695797e-01 9.54353988e-01 6.98721588e-01 -5.54529488e-01
-4.37186867e-01 -8.76676679e-01 -1.05966389e+00 -3.19199711e-01
-4.56041038e-01 5.94294786e-01 1.70016084e-02 6.25130236e-02
-3.48007381e-01 2.21913293e-01 -5.52981794e-01 3.80385906e-01
1.08649504e+00 2.98941284e-01 -1.59614429e-01 3.01744461e-01
-1.79276422e-01 9.48257387e-01 -1.81512237e-01 -7.59861171e-01
-3.32077503e-01 1.42746747e-01 -9.51598883e-01 -6.66122675e-01
2.57138431e-01 1.75455466e-01 -6.17027640e-01 7.39967465e-01
5.52239478e-01 -2.87837327e-01 8.47999156e-01 -1.27367926e+00
-1.10680759e+00 1.31741917e+00 -4.13049936e-01 -5.00694476e-02
3.59533817e-01 -2.17969641e-02 1.55622506e+00 -1.04745185e+00
1.17950690e+00 1.30370402e+00 9.74613428e-01 7.23686993e-01
-1.51269948e+00 -6.65759444e-01 6.01517260e-02 2.84020036e-01
-1.41803777e+00 -3.48409057e-01 5.93617260e-01 -6.41538739e-01
1.13351500e+00 3.39548364e-02 3.89289975e-01 1.26273119e+00
8.68794098e-02 8.96587670e-01 5.80239713e-01 -2.11388230e-01
3.04258801e-02 3.74953389e-01 3.12712461e-01 9.35988843e-01
6.36744320e-01 -3.06003600e-01 -7.87350059e-01 -5.50929725e-01
2.93859780e-01 -1.46152914e-01 -5.70400119e-01 -3.55693012e-01
-1.29559112e+00 8.79333138e-01 6.04138017e-01 1.71816751e-01
-2.88133174e-01 1.96031734e-01 3.76915455e-01 1.33846030e-01
7.67772496e-01 9.71650720e-01 -9.26479995e-01 2.71057010e-01
-6.43925726e-01 7.20981285e-02 1.10159755e+00 1.00091219e+00
5.51587462e-01 -4.94778067e-01 -3.29572469e-01 8.29678595e-01
2.13559985e-01 1.78758875e-01 2.69880861e-01 -3.49956781e-01
4.45280075e-01 7.93431938e-01 -1.58673987e-01 -9.50316012e-01
-7.56392539e-01 -8.48131180e-01 -7.63953149e-01 -6.36888504e-01
4.42726985e-02 -1.95413679e-01 -7.40055919e-01 1.99711692e+00
4.97726411e-01 6.52105331e-01 2.09317058e-01 3.19876909e-01
1.35105479e+00 4.18669969e-01 7.06711054e-01 7.03868642e-02
1.29932666e+00 -8.17140102e-01 -7.45350778e-01 3.10945679e-02
1.34545159e+00 -2.72900850e-01 8.68827462e-01 4.98068146e-02
-8.53672862e-01 -6.00256063e-02 -6.78551614e-01 -2.88840294e-01
-7.49849081e-01 1.45778269e-01 8.24379444e-01 1.64701283e-01
-1.11370409e+00 9.71808195e-01 -8.02045763e-01 -3.22217137e-01
7.23091900e-01 3.43807161e-01 -6.22572541e-01 -1.97294787e-01
-1.72036684e+00 7.57357597e-01 3.96432817e-01 -2.87580281e-01
-5.41269362e-01 -1.51567852e+00 -9.98640418e-01 2.38368288e-01
3.71959507e-01 -1.30208242e+00 4.68185008e-01 -1.02317229e-01
-8.95699501e-01 8.81441176e-01 -2.07003638e-01 -4.48255002e-01
1.25257194e-01 -3.85671966e-02 -5.45698643e-01 1.71636953e-03
1.85730368e-01 7.10052311e-01 3.38844717e-01 -8.24083209e-01
-5.07860407e-02 -3.14048588e-01 1.90861672e-01 -2.52635837e-01
-8.64202201e-01 -3.66298884e-01 -4.93310481e-01 -6.36798263e-01
-2.61637330e-01 -8.03827047e-01 -3.09226722e-01 -1.49087310e-02
-1.04261410e+00 -4.36210722e-01 2.78664351e-01 -7.12428808e-01
1.52666140e+00 -1.83259082e+00 4.76720989e-01 3.85008514e-01
6.87398314e-01 9.71001685e-02 -3.19903702e-01 6.78347290e-01
-2.48607770e-01 5.74051142e-01 -7.02979565e-02 -2.63833851e-01
8.61694440e-02 2.89071858e-01 -1.45750955e-01 1.10776857e-01
5.44448853e-01 1.62743282e+00 -1.20600092e+00 -3.76508802e-01
-4.36618209e-01 6.24162674e-01 -7.33765841e-01 -8.23634788e-02
-4.79261488e-01 2.45713010e-01 -6.80752933e-01 6.51385665e-01
3.28679651e-01 -1.01191092e+00 5.40427089e-01 -7.71729708e-01
4.57993895e-01 4.69471633e-01 -6.77941084e-01 1.63484287e+00
-2.47819766e-01 3.17616880e-01 -3.75379831e-01 -8.88076425e-01
4.99803364e-01 2.81795770e-01 6.42433047e-01 -1.19250484e-01
-5.41385770e-01 7.64662772e-02 -2.35099599e-01 -5.86222589e-01
2.54586428e-01 -7.33958185e-02 1.49243191e-01 3.03995550e-01
3.10184926e-01 4.42235678e-01 7.51199275e-02 7.87995279e-01
1.49920404e+00 -2.46696528e-02 3.78018469e-01 -1.70878127e-01
3.00647736e-01 2.74691209e-02 2.96900898e-01 5.30249178e-01
4.90241021e-01 -1.58242553e-01 8.58416855e-01 -1.17389053e-01
-7.98677802e-01 -8.32254469e-01 -3.79956186e-01 9.50991392e-01
-3.41555953e-01 -8.63380313e-01 -2.80930609e-01 -9.15504992e-01
6.52076542e-01 4.95843798e-01 -8.91153514e-01 -5.40865421e-01
-2.00223878e-01 -1.09811246e+00 8.56703579e-01 6.60130084e-01
-3.33090782e-01 -5.78410208e-01 4.29160506e-01 2.03446448e-01
2.31097750e-02 -1.25535357e+00 -3.43038738e-01 2.36502230e-01
-8.64499450e-01 -1.28627324e+00 -6.54131413e-01 -6.86575174e-01
9.02096152e-01 -3.54241997e-01 1.41137159e+00 -1.95877224e-01
-3.56012315e-01 4.44771290e-01 -2.71273583e-01 -3.85139585e-01
-4.35054660e-01 5.56804359e-01 1.62262872e-01 -2.57301211e-01
4.23499435e-01 -6.42407894e-01 -5.91833472e-01 -1.70270935e-01
-8.93199503e-01 1.86896771e-01 7.02211499e-01 9.61330593e-01
5.28325140e-01 -4.02235597e-01 1.13558686e+00 -1.68649161e+00
7.64583766e-01 -9.55563009e-01 -1.93168193e-01 5.06444156e-01
-8.93251181e-01 5.76382577e-01 4.47903425e-01 -3.06363344e-01
-4.37291026e-01 -1.53153270e-01 -3.18423718e-01 -2.38120526e-01
4.15996432e-01 1.39365554e+00 -1.31738633e-01 -2.14633375e-01
8.07165682e-01 -1.55732557e-01 -2.77235448e-01 -7.24753618e-01
8.16352069e-01 3.54452640e-01 2.58612841e-01 -6.07525527e-01
7.65368104e-01 2.17426583e-01 3.31591487e-01 -3.57218057e-01
-9.69861925e-01 -6.08039081e-01 -4.85531539e-01 4.13980395e-01
6.06749833e-01 -1.18214238e+00 -7.49457598e-01 -1.79590628e-01
-9.72731769e-01 -2.72581667e-01 -1.96080700e-01 4.18734699e-01
-1.23517551e-01 3.43317538e-01 -8.84740829e-01 -6.74640015e-02
-3.68277401e-01 -6.03085279e-01 1.05673587e+00 -1.94380358e-01
-2.88920879e-01 -1.73293316e+00 3.39538336e-01 9.24691856e-02
1.40132919e-01 3.55717242e-01 1.67856085e+00 -9.11797583e-01
-4.67989087e-01 -7.71532953e-02 -2.60831565e-01 -4.30978239e-02
3.25205058e-01 -2.42403865e-01 -1.01396453e+00 -1.95174858e-01
-9.89820123e-01 -2.60800540e-01 1.20395613e+00 3.45320612e-01
1.23139024e+00 -7.03204215e-01 -1.04862440e+00 7.02108324e-01
1.33125162e+00 -5.11663556e-01 3.06496710e-01 -7.67324790e-02
1.11748385e+00 4.54614073e-01 1.58190310e-01 3.14633876e-01
6.78608119e-01 6.76767647e-01 3.43711637e-02 -4.19270098e-01
-1.41037658e-01 -6.96316957e-01 2.89458990e-01 9.55091357e-01
7.11579472e-02 -2.62548298e-01 -1.05268812e+00 6.65355206e-01
-1.87253737e+00 -6.79380357e-01 -1.45806715e-01 1.75629532e+00
1.68836832e+00 -8.45394582e-02 -1.11190543e-01 -4.16052669e-01
2.99418479e-01 -1.65155381e-01 -7.34598994e-01 7.64914006e-02
-2.89904565e-01 5.35211384e-01 5.50474346e-01 6.00021362e-01
-1.05033231e+00 9.96127725e-01 6.35999918e+00 8.65381420e-01
-9.33991671e-01 1.09199055e-01 5.00557125e-01 1.37369171e-01
-8.91402125e-01 -1.53983027e-01 -9.73276854e-01 2.38024592e-01
1.07457972e+00 -5.34515560e-01 -4.78586555e-02 7.88303316e-01
4.19325531e-02 5.59107423e-01 -1.58174205e+00 6.18423283e-01
-2.15420768e-01 -1.96593773e+00 4.66817558e-01 1.11791022e-01
8.13084543e-01 1.55251220e-01 1.43952072e-01 4.33276474e-01
6.08828485e-01 -1.18299699e+00 -1.60124093e-01 6.35909796e-01
9.77358043e-01 -3.51112276e-01 5.94115674e-01 2.19068043e-02
-9.70136166e-01 2.27859423e-01 -4.76128191e-01 5.59879661e-01
-4.43314463e-02 1.03882074e+00 -1.43856931e+00 1.03463411e+00
1.73672304e-01 1.41552401e+00 -6.56651080e-01 8.36803257e-01
-4.68519717e-01 7.71943569e-01 -1.21948205e-01 1.17877319e-01
-3.55290137e-02 2.39397734e-01 4.62639838e-01 1.49313939e+00
2.06396192e-01 -1.42203998e-02 7.30925500e-02 1.13284898e+00
-6.81157529e-01 3.95787269e-01 -7.82037139e-01 -6.00284994e-01
4.29308295e-01 1.21588755e+00 -1.26495704e-01 -4.27813083e-01
-4.08690870e-01 6.98368788e-01 6.03684664e-01 5.47705412e-01
-6.31091535e-01 -3.13518792e-01 7.70381093e-01 2.02251524e-01
3.14419791e-02 1.02529891e-01 -2.98304856e-02 -1.39349639e+00
-3.62228721e-01 -5.31324089e-01 6.97922528e-01 -6.99665487e-01
-1.73263538e+00 2.79604673e-01 -1.13336951e-01 -8.39864850e-01
-1.06849536e-01 -8.32899094e-01 -8.93086046e-02 7.97487915e-01
-1.96708226e+00 -1.48916626e+00 1.46540254e-01 4.57290828e-01
-2.09431112e-01 9.67073962e-02 1.06039679e+00 7.20112145e-01
-5.87199628e-01 9.88814771e-01 3.02929521e-01 2.67923087e-01
1.04588604e+00 -1.38523901e+00 3.99749011e-01 1.75381348e-01
2.56903559e-01 1.12181795e+00 4.32604283e-01 -9.77053642e-01
-1.48376369e+00 -1.67143118e+00 1.30692387e+00 -8.26253116e-01
1.02158916e+00 -4.07567590e-01 -1.07988560e+00 9.56723809e-01
-2.89557129e-01 9.82982442e-02 1.24815905e+00 8.69700909e-01
-5.84166050e-01 1.76420316e-01 -8.19758654e-01 4.60345954e-01
1.35630643e+00 -7.33452916e-01 -2.39584163e-01 7.59834945e-01
1.03574133e+00 -1.73524559e-01 -1.57262909e+00 4.99634176e-01
3.32772702e-01 8.70096385e-02 1.28438473e+00 -1.43156874e+00
6.54860795e-01 -1.23457238e-01 3.50494236e-01 -1.59828341e+00
-5.94350874e-01 -3.85359317e-01 -4.53851521e-01 1.08674812e+00
1.07570219e+00 -6.97925866e-01 6.56713307e-01 5.37297428e-01
-1.82335123e-01 -1.00424647e+00 -5.98727047e-01 -6.39611781e-01
1.73957691e-01 -1.15895554e-01 5.44758022e-01 1.54772794e+00
6.17396891e-01 5.52677870e-01 -2.62164652e-01 3.10588747e-01
3.88854951e-01 4.48464453e-02 5.03579974e-01 -1.50480127e+00
-5.74337363e-01 -2.29069039e-01 -4.42042232e-01 -7.64852583e-01
5.12469709e-01 -1.63389111e+00 -6.15895510e-01 -1.77949142e+00
6.36828065e-01 -7.38352060e-01 -6.17316663e-01 1.14823425e+00
-6.56306505e-01 5.01248389e-02 -4.22191679e-01 8.90486985e-02
-5.73554337e-01 4.03423220e-01 9.70406055e-01 -5.28406084e-01
-2.86597933e-04 -4.16175246e-01 -9.70353007e-01 4.02847767e-01
3.73432249e-01 -5.53698242e-01 -4.64682519e-01 -3.39345247e-01
1.00038981e+00 -1.68987110e-01 3.13098460e-01 -3.65187109e-01
4.22544062e-01 1.31142363e-01 2.08424002e-01 -4.98541892e-02
8.93136933e-02 -4.83543664e-01 3.57424915e-01 4.64056492e-01
-7.71613359e-01 -4.02846396e-01 3.96780521e-01 9.15196478e-01
-1.19192667e-01 6.49529248e-02 2.17333108e-01 -2.65937410e-02
-4.95522320e-01 6.72157228e-01 2.41180718e-01 2.27582946e-01
7.35181093e-01 3.10332358e-01 -6.39092684e-01 -2.44431958e-01
-1.19632924e+00 5.08460462e-01 1.70595631e-01 2.11254597e-01
6.15265429e-01 -1.38610435e+00 -8.32489669e-01 -1.49185896e-01
5.76604247e-01 -2.72620022e-01 7.83581585e-02 1.09456456e+00
9.28185601e-03 6.67555034e-01 2.80214399e-01 -2.72319227e-01
-1.21481109e+00 7.43444860e-01 6.25078529e-02 -7.45690584e-01
-5.05623281e-01 1.00091779e+00 4.73220229e-01 -5.54147482e-01
1.39795795e-01 -4.74950701e-01 -4.05905098e-01 1.15401201e-01
2.73072034e-01 -8.23862758e-03 -1.23543022e-02 2.97757182e-02
-5.85927963e-01 3.80522728e-01 -2.51937330e-01 4.72599119e-01
1.70181584e+00 3.21980476e-01 -3.51356238e-01 2.82812595e-01
1.27379727e+00 3.25623989e-01 -4.26952720e-01 -5.80226064e-01
3.61017346e-01 8.69895145e-02 -1.86423779e-01 -1.12060940e+00
-9.33095574e-01 7.69393981e-01 -8.99792165e-02 -2.06224784e-01
7.58529961e-01 4.72346157e-01 5.48001230e-01 4.66583937e-01
2.28658408e-01 -5.11313021e-01 -9.00635943e-02 2.35681832e-01
7.09548712e-01 -9.92507577e-01 8.45510736e-02 -8.45853031e-01
-4.46419030e-01 9.76265311e-01 1.49537370e-01 3.43899965e-01
7.87543893e-01 2.95964122e-01 -4.30255830e-01 -5.81042826e-01
-1.07560837e+00 -9.30414423e-02 7.57422388e-01 6.19205058e-01
6.70015693e-01 2.40788952e-01 -2.01157063e-01 1.00709283e+00
1.63095161e-01 4.23572332e-01 1.73196822e-01 5.15727997e-01
4.79516424e-02 -1.46532965e+00 3.47666711e-01 7.81608343e-01
-5.25236011e-01 -8.26586485e-01 -6.11452997e-01 4.98086035e-01
1.57587733e-02 3.86269480e-01 -5.25374293e-01 -5.33544242e-01
2.25894585e-01 3.58159423e-01 3.59981596e-01 -8.96302402e-01
-3.58327508e-01 -4.19262409e-01 4.58292544e-01 -4.02481377e-01
-3.02040637e-01 -2.71353215e-01 -1.14311147e+00 -3.17072898e-01
-3.00634861e-01 3.55963111e-01 2.39096850e-01 5.95812619e-01
1.04195821e+00 6.88280284e-01 1.95139498e-01 7.07792640e-02
-2.82799035e-01 -7.58241773e-01 -4.62732673e-01 4.14667904e-01
1.89344317e-01 -4.72147793e-01 -4.49808054e-02 2.27361634e-01] | [8.552440643310547, 7.943466663360596] |
4b9c7bf6-7c31-455e-b783-2086729d4636 | retinal-image-segmentation-with-small | 2303.05110 | null | https://arxiv.org/abs/2303.05110v1 | https://arxiv.org/pdf/2303.05110v1.pdf | Retinal Image Segmentation with Small Datasets | Many eye diseases like Diabetic Macular Edema (DME), Age-related Macular Degeneration (AMD), and Glaucoma manifest in the retina, can cause irreversible blindness or severely impair the central version. The Optical Coherence Tomography (OCT), a 3D scan of the retina with high qualitative information about the retinal morphology, can be used to diagnose and monitor changes in the retinal anatomy. Many Deep Learning (DL) methods have shared the success of developing an automated tool to monitor pathological changes in the retina. However, the success of these methods depend mainly on large datasets. To address the challenge from very small and limited datasets, we proposed a DL architecture termed CoNet (Coherent Network) for joint segmentation of layers and fluids in retinal OCT images on very small datasets (less than a hundred training samples). The proposed model was evaluated on the publicly available Duke DME dataset consisting of 110 B-Scans from 10 patients suffering from DME. Experimental results show that the proposed model outperformed both the human experts' annotation and the current state-of-the-art architectures by a clear margin with a mean Dice Score of 88% when trained on 55 images without any data augmentation. | ['Yongmin Li', 'Zidong Wang', 'Alina Miron', 'Nchongmaje Ndipenoch'] | 2023-03-09 | null | null | null | null | ['anatomy'] | ['miscellaneous'] | [-1.26124904e-01 -8.60683993e-02 1.81658253e-01 -2.97996789e-01
-3.50527138e-01 -1.98273763e-01 3.32260281e-02 -2.66148031e-01
-5.43936133e-01 8.91631842e-01 1.07346572e-01 -4.37687337e-01
-9.07042176e-02 -3.50257933e-01 -1.98431775e-01 -6.17625654e-01
-1.59389645e-01 3.24916482e-01 4.50033933e-01 4.17613834e-01
3.26538920e-01 7.17890978e-01 -1.51404369e+00 3.97451460e-01
1.42153215e+00 1.26094723e+00 1.38155743e-01 7.09699988e-01
9.68488827e-02 4.69061792e-01 -2.65465081e-02 -3.53528380e-01
4.79161233e-01 -3.43034565e-01 -6.70473874e-01 1.55604258e-01
1.12407494e+00 -6.54029131e-01 5.33784665e-02 1.23222876e+00
9.47564363e-01 -6.07685030e-01 6.25030756e-01 -3.43848437e-01
-4.74085897e-01 -3.27803552e-01 -7.48105705e-01 6.38444781e-01
-3.44117135e-01 4.89641458e-01 3.90519828e-01 -8.75560105e-01
4.65944827e-01 1.00206387e+00 6.42579257e-01 4.47534353e-01
-1.07808483e+00 -4.10794824e-01 -3.92808646e-01 4.43429023e-01
-9.27553713e-01 -6.62619054e-01 3.35587002e-02 -1.02688634e+00
8.26728523e-01 -6.81275725e-02 1.20802200e+00 4.90942031e-01
2.91993022e-01 3.54725003e-01 1.78803670e+00 -2.60512382e-01
8.19566250e-02 -3.26638073e-01 9.22692716e-02 8.89165819e-01
7.24287391e-01 2.41038665e-01 1.45169422e-01 6.39704708e-03
8.26872885e-01 -5.05795538e-01 -4.23608124e-01 -2.37427339e-01
-9.54986393e-01 5.03266037e-01 6.25728071e-01 -2.06208169e-01
-5.21820307e-01 -1.97323427e-01 3.51924300e-01 1.44203946e-01
5.54176867e-01 5.47850251e-01 -4.27969158e-01 3.57873924e-02
-6.79552078e-01 -9.68351737e-02 3.25318754e-01 4.81791683e-02
4.10950720e-01 -2.34935060e-01 -3.06801379e-01 8.84503067e-01
5.30935824e-01 2.85741270e-01 5.18284261e-01 -1.13963366e+00
1.03599094e-01 9.47098434e-01 3.43959123e-01 -3.21349531e-01
-5.22214413e-01 -5.05524039e-01 -1.07507074e+00 9.71157193e-01
6.54509187e-01 -4.69609082e-01 -1.33546901e+00 1.02089369e+00
2.23694265e-01 2.29660794e-01 -5.26408553e-01 1.25931346e+00
8.76718521e-01 -8.86847377e-02 -2.69955397e-01 -4.00513142e-01
1.16419077e+00 -6.95978642e-01 -4.99391407e-01 -3.45258862e-01
4.84284431e-01 -1.03411388e+00 7.90117204e-01 4.85844016e-01
-1.20251846e+00 -4.75633383e-01 -8.70389938e-01 -1.90721050e-01
3.04817073e-02 8.66659462e-01 4.94721204e-01 5.99358976e-01
-1.45375025e+00 3.50583971e-01 -1.04583681e+00 -6.08965337e-01
1.28151882e+00 4.42920327e-01 -4.90678757e-01 -2.81185269e-01
-2.24662095e-01 1.05374110e+00 -6.70344159e-02 3.74812484e-01
-5.48758566e-01 -7.39014566e-01 -2.50155151e-01 -3.83960158e-01
-6.00166209e-02 -1.33905065e+00 8.26210678e-01 -6.45792305e-01
-1.53028369e+00 1.42401743e+00 -3.79290611e-01 -5.87009549e-01
6.23137295e-01 -6.42759502e-01 -2.44862869e-01 4.33856308e-01
-4.53598015e-02 7.03081012e-01 6.79028869e-01 -7.98716605e-01
-7.16152668e-01 -8.58525872e-01 -3.00541013e-01 -2.43523777e-01
1.70906216e-01 2.53540933e-01 -3.28232080e-01 2.67925635e-02
1.36915758e-01 -8.66741538e-01 -2.50716627e-01 6.69874668e-01
-6.92013025e-01 3.25371996e-02 4.12228316e-01 -7.94955850e-01
8.40625942e-01 -1.88470614e+00 -2.16507807e-01 -1.81239352e-01
8.61432076e-01 1.14382601e+00 -2.45812863e-01 -3.81560087e-01
-1.20993122e-01 1.57780841e-01 3.13970707e-02 -6.82613328e-02
-7.90630698e-01 -1.64004132e-01 3.01032096e-01 6.78647816e-01
2.66419381e-01 9.46208775e-01 -6.58921659e-01 -3.91340494e-01
1.32062435e-01 4.42131490e-01 -4.10818666e-01 1.74774095e-01
-1.81713685e-01 6.57335401e-01 -3.12111229e-01 7.19007492e-01
7.40291357e-01 -7.03969240e-01 7.57048503e-02 -3.43634337e-01
-2.80731648e-01 2.88579203e-02 -6.60267472e-01 1.28366196e+00
1.41178414e-01 1.16404963e+00 -9.64060724e-02 -5.35977066e-01
6.29198134e-01 1.09795868e-01 2.61189878e-01 -6.09433413e-01
3.20261061e-01 5.65805912e-01 8.43071461e-01 -1.02084708e+00
-4.55827922e-01 7.23306388e-02 1.11076081e+00 1.66057423e-01
-3.72756958e-01 3.19356889e-01 2.11194947e-01 -4.81673360e-01
1.05200732e+00 -5.77878505e-02 5.26381969e-01 1.47943288e-01
3.89683098e-01 -1.98210061e-01 5.57311118e-01 3.17540944e-01
-5.64384937e-01 8.10531795e-01 7.82859504e-01 -8.49651575e-01
-1.21056795e+00 -9.08928931e-01 -5.31135738e-01 -2.77441651e-01
-1.29750088e-01 5.58226444e-02 -6.33843660e-01 -3.85175675e-01
2.47541536e-02 -2.50934869e-01 -5.72127461e-01 3.91148418e-01
-2.48411998e-01 -1.14569163e+00 2.32078508e-01 2.01645315e-01
1.15203011e+00 -5.46870887e-01 -7.17684388e-01 -3.72355315e-03
1.56375185e-01 -1.19168353e+00 8.08290839e-02 -7.13881791e-01
-1.19174051e+00 -1.67866123e+00 -9.68017817e-01 -8.40515494e-01
7.36557126e-01 1.45380646e-01 9.21576858e-01 -1.71462253e-01
-9.39222574e-01 -1.62797228e-01 5.60560264e-02 -3.98328006e-01
-1.67113677e-01 -3.51233333e-01 -1.10257290e-01 3.35214168e-01
5.45552373e-01 -7.69142509e-01 -1.29585063e+00 3.23026955e-01
-2.29731962e-01 2.37847921e-02 1.33119833e+00 6.04763091e-01
7.31176794e-01 -2.28994325e-01 3.59142154e-01 -6.92853391e-01
5.96463382e-01 -1.89847630e-02 -9.19332802e-01 1.39756083e-01
-8.14147174e-01 -3.70046526e-01 -7.13117095e-03 -1.68711171e-01
-6.56903863e-01 -2.65286028e-01 3.07254732e-01 -3.57753724e-01
-3.83356363e-01 4.84915912e-01 7.64502212e-02 -4.96285051e-01
8.84857297e-01 -2.80081093e-01 6.01743937e-01 -7.51087606e-01
1.46621063e-01 8.53523374e-01 5.69898844e-01 3.14928517e-02
2.18709156e-01 6.82481766e-01 5.37455320e-01 -6.63847148e-01
-1.01762915e+00 -4.92143869e-01 -6.93853021e-01 -5.83970994e-02
1.00388861e+00 -1.00552714e+00 -7.57163703e-01 1.01770914e+00
-1.17743731e+00 -3.54169369e-01 8.59900117e-02 9.22803462e-01
-2.71474510e-01 5.15636802e-01 -4.23314959e-01 -3.74793828e-01
-3.94135475e-01 -1.07391810e+00 6.35813236e-01 4.46854025e-01
1.50470674e-01 -6.38011336e-01 2.41225347e-01 5.47172844e-01
4.08858746e-01 3.61865342e-01 1.37385440e+00 -1.32411703e-01
-8.11933935e-01 -1.18855081e-01 -9.67304289e-01 9.69420016e-01
3.23968649e-01 2.50506759e-01 -8.88520718e-01 -1.69058084e-01
-3.31981272e-01 -3.12506676e-01 1.09557998e+00 1.03415287e+00
1.13107884e+00 4.47657593e-02 -3.34649742e-01 7.96955884e-01
1.47770000e+00 1.10389605e-01 1.14417493e+00 3.02277952e-01
5.08966744e-01 5.68275690e-01 2.71127135e-01 2.73016721e-01
4.04533371e-02 5.09602964e-01 6.93557978e-01 -3.98085535e-01
-8.35794151e-01 5.12901604e-01 -1.82133391e-01 3.58131677e-01
-7.16613412e-01 5.08229546e-02 -1.03165174e+00 7.24280596e-01
-1.53871107e+00 -5.96064448e-01 -6.22314394e-01 2.30945134e+00
8.62464309e-01 -1.37496069e-01 2.11263690e-02 -4.08919901e-01
7.59689033e-01 -4.20855224e-01 -8.84257972e-01 1.08857136e-02
-2.60421544e-01 4.02991980e-01 3.50082070e-01 2.55116850e-01
-1.15666604e+00 6.97772920e-01 6.52141142e+00 1.26616403e-01
-1.33182228e+00 -5.42768575e-02 6.66203082e-01 -3.76107544e-01
4.06205654e-01 -2.71008313e-01 -6.30321562e-01 5.43307841e-01
6.65969610e-01 2.03710571e-01 2.15223432e-01 1.27069175e-01
8.84242833e-01 -3.18225116e-01 -9.42422688e-01 1.18504858e+00
-2.33580008e-01 -1.42233944e+00 9.63017866e-02 5.55948138e-01
9.13007975e-01 6.58463717e-01 3.10635835e-01 -5.61718762e-01
-8.14539790e-02 -1.46671200e+00 -4.43133503e-01 1.21277857e+00
1.30859637e+00 -1.81167543e-01 1.10287011e+00 -2.23926798e-01
-4.24381316e-01 -7.95418695e-02 -4.68207985e-01 -1.19116820e-01
-1.69446856e-01 9.30607259e-01 -9.65981007e-01 -9.42529663e-02
7.57313311e-01 1.10524797e+00 -9.32364881e-01 2.34961867e+00
-2.65342742e-01 5.59820354e-01 2.27142014e-02 4.30692136e-01
2.33697295e-02 -5.76261878e-01 8.68066072e-01 3.94798189e-01
5.47272325e-01 4.53823172e-02 -3.74833912e-01 1.05345249e+00
5.29192127e-02 7.07167834e-02 -2.93440670e-01 -1.08355261e-01
2.27112815e-01 1.21210980e+00 -1.25332817e-01 7.65455738e-02
-6.60787404e-01 4.03588742e-01 9.64350551e-02 7.22338617e-01
-8.47165883e-02 -1.55922994e-01 9.17543054e-01 6.73158288e-01
2.03390956e-01 -3.17687169e-02 -4.58571941e-01 -8.02488565e-01
3.21495116e-01 -7.61105955e-01 -1.25608504e-01 -1.24509561e+00
-1.45330429e+00 6.14486158e-01 -6.89197958e-01 -1.43872249e+00
8.48639235e-02 -1.03446066e+00 -5.61918855e-01 1.42790902e+00
-1.97908831e+00 -1.07115853e+00 -7.65075028e-01 4.58698362e-01
-1.96135908e-01 -7.76510060e-01 6.74663126e-01 2.98387855e-01
-7.37770677e-01 2.37851366e-02 1.46502644e-01 1.22814953e-01
1.00212669e+00 -1.22189701e+00 3.51090729e-02 8.88453722e-01
-6.54763758e-01 6.10864460e-01 2.03825116e-01 -6.22956753e-01
-5.89603841e-01 -1.30072927e+00 7.38732219e-01 -2.82374501e-01
5.17812908e-01 6.66561306e-01 -7.38210082e-01 6.19249701e-01
1.98604807e-01 6.51984394e-01 8.33370447e-01 -1.39381409e-01
-8.98201391e-02 -3.50364983e-01 -1.10408020e+00 4.96104777e-01
7.18622804e-01 -2.28938848e-01 -4.95354980e-01 1.80922374e-01
2.04251945e-01 -3.14085305e-01 -8.43855917e-01 5.95335722e-01
6.98346257e-01 -1.52077246e+00 6.09881341e-01 -8.25859308e-01
5.56518018e-01 -3.58199954e-01 1.54478893e-01 -1.11608446e+00
-1.47514805e-01 -7.54000366e-01 -6.89540654e-02 4.71628040e-01
4.76271771e-02 -9.17127311e-01 7.60542393e-01 3.55353147e-01
-4.01335388e-01 -9.73287225e-01 -8.90384555e-01 -3.78524125e-01
6.58604056e-02 9.87551063e-02 1.12089157e-01 4.23665345e-01
-7.75587678e-01 1.72110081e-01 1.18957587e-01 3.64546835e-01
7.90430725e-01 4.59815282e-03 6.64485395e-01 -1.88318229e+00
1.84895366e-01 -5.96127868e-01 -9.78862345e-01 -7.99669802e-01
-1.67117983e-01 -6.88589334e-01 -5.97999513e-01 -1.92811346e+00
3.11054558e-01 -2.70577490e-01 -1.59388661e-01 3.23954016e-01
-1.05016835e-01 5.28252959e-01 -4.24140394e-01 4.30714339e-01
-7.72059932e-02 2.04416841e-01 1.92295110e+00 -8.17910358e-02
-3.34139645e-01 3.10347557e-01 -6.84955120e-01 9.39111531e-01
6.59570754e-01 5.69392703e-02 -3.81657958e-01 -5.36212504e-01
3.78817379e-01 -2.54432142e-01 8.97429287e-01 -1.01462913e+00
1.93313837e-01 2.93945700e-01 4.56720084e-01 -4.23575699e-01
9.88906324e-02 -3.19168478e-01 -2.30116561e-01 3.88327837e-01
4.18833941e-02 -7.05337167e-01 1.71112701e-01 3.54978949e-01
-3.03298444e-01 3.43811810e-01 1.37148738e+00 -1.51832581e-01
-5.19592464e-01 6.92917526e-01 -1.08803853e-01 1.92200735e-01
9.10327435e-01 -3.69105697e-01 -9.54584479e-01 1.71370432e-01
-9.12024498e-01 1.02890223e-01 4.63753015e-01 -1.53636247e-01
9.42474663e-01 -8.85570645e-01 -1.10560131e+00 2.97372222e-01
9.47994590e-02 2.14476228e-01 1.12775564e-01 1.66167688e+00
-9.30819750e-01 6.05274737e-01 -7.50679970e-01 -8.46199632e-01
-1.44109261e+00 -1.64291054e-01 1.18263459e+00 9.21840966e-02
-7.35681117e-01 8.73279035e-01 -1.02420077e-02 3.01456213e-01
9.76087898e-02 -6.74330294e-01 -5.58077753e-01 1.23139821e-01
6.08772099e-01 6.11732423e-01 1.91801727e-01 -1.46510959e-01
-7.11493660e-04 1.09137762e+00 -4.80395973e-01 4.04706866e-01
1.28951621e+00 -3.23198766e-01 -8.03771257e-01 1.13397002e-01
8.37845385e-01 -1.54628918e-01 -1.14983869e+00 -3.17929953e-01
-4.03333277e-01 -7.67905593e-01 6.30159855e-01 -1.28266728e+00
-1.09336245e+00 1.15691245e+00 1.42585206e+00 -1.71200335e-02
1.08799350e+00 -2.75041223e-01 8.55431497e-01 1.45524934e-01
2.70040125e-01 -4.01179820e-01 -5.10981120e-02 3.88279371e-02
7.27177441e-01 -1.58038843e+00 6.26579225e-02 -4.64657366e-01
-2.67456293e-01 1.23377705e+00 5.80149293e-01 -6.28270954e-02
7.98196375e-01 -3.71558309e-01 4.85487938e-01 -4.13916141e-01
-5.62938511e-01 -7.16396153e-01 8.61650467e-01 9.55689132e-01
4.75123256e-01 -2.86416560e-02 -3.81964982e-01 2.06517041e-01
9.77910757e-02 6.14613354e-01 7.81142831e-01 1.92199379e-01
-6.88864470e-01 -9.09035325e-01 1.74654067e-01 1.05114770e+00
-6.64536953e-01 -2.49726221e-01 -5.22852063e-01 7.32187212e-01
4.63953078e-01 8.10426354e-01 2.62687594e-01 2.09937245e-01
-5.65231144e-02 -2.57351935e-01 7.37044334e-01 -8.73266757e-01
1.36716226e-02 2.25832015e-01 3.02772880e-01 -8.73068810e-01
-7.66223967e-01 -4.75089908e-01 -7.37132907e-01 8.83317217e-02
2.24728644e-01 -6.09033465e-01 3.51772726e-01 8.50737095e-01
6.56724811e-01 3.26364666e-01 4.25928026e-01 -3.44781667e-01
-1.95646569e-01 -1.33862889e+00 -1.03100038e+00 -1.77753106e-01
7.72641361e-01 -6.03537917e-01 -2.92135298e-01 1.22464627e-01] | [15.814611434936523, -3.9891960620880127] |
38554476-3106-4022-b3cb-d8f076004539 | maximum-state-entropy-exploration-using | 2306.14808 | null | https://arxiv.org/abs/2306.14808v1 | https://arxiv.org/pdf/2306.14808v1.pdf | Maximum State Entropy Exploration using Predecessor and Successor Representations | Animals have a developed ability to explore that aids them in important tasks such as locating food, exploring for shelter, and finding misplaced items. These exploration skills necessarily track where they have been so that they can plan for finding items with relative efficiency. Contemporary exploration algorithms often learn a less efficient exploration strategy because they either condition only on the current state or simply rely on making random open-loop exploratory moves. In this work, we propose $\eta\psi$-Learning, a method to learn efficient exploratory policies by conditioning on past episodic experience to make the next exploratory move. Specifically, $\eta\psi$-Learning learns an exploration policy that maximizes the entropy of the state visitation distribution of a single trajectory. Furthermore, we demonstrate how variants of the predecessor representation and successor representations can be combined to predict the state visitation entropy. Our experiments demonstrate the efficacy of $\eta\psi$-Learning to strategically explore the environment and maximize the state coverage with limited samples. | ['Glen Berseth', 'Irina Rish', 'Lucas Lehnert', 'Arnav Kumar Jain'] | 2023-06-26 | null | null | null | null | ['efficient-exploration'] | ['methodology'] | [-3.72279063e-02 8.23953468e-03 -6.27032161e-01 -3.50189716e-01
-2.74959177e-01 -4.67912108e-01 2.50207216e-01 4.92373556e-01
-7.94285595e-01 9.13370252e-01 4.46062610e-02 -4.94675636e-01
-5.02703607e-01 -1.07228732e+00 -7.92259276e-01 -5.69642603e-01
-1.01266706e+00 4.94540393e-01 -1.45081192e-01 -1.04441062e-01
4.76825476e-01 4.26435530e-01 -1.57119405e+00 -1.13205656e-01
7.71043718e-01 8.16815615e-01 6.38841271e-01 5.02089918e-01
-1.57991663e-01 5.93512177e-01 -5.53495139e-02 3.16901594e-01
3.35464746e-01 -5.57309210e-01 -8.78997803e-01 -3.56250077e-01
-6.01023376e-01 -4.46354449e-01 -6.07082471e-02 8.14838409e-01
3.63491885e-02 7.87703156e-01 5.52521706e-01 -1.01406395e+00
-1.69393584e-01 1.04868770e+00 -3.44811261e-01 6.05216026e-01
3.43516916e-01 4.32044506e-01 9.75203872e-01 -1.62131310e-01
6.05897725e-01 8.81217241e-01 3.50644886e-01 3.55710685e-01
-1.38663781e+00 -7.60200441e-01 5.96233487e-01 1.86475500e-01
-9.48639035e-01 -1.71340570e-01 6.20768487e-01 -2.70180702e-01
1.32797647e+00 6.43868581e-04 1.22812057e+00 8.32084537e-01
4.09310281e-01 9.88092780e-01 1.22542131e+00 -2.16580048e-01
7.76766896e-01 -2.41890132e-01 -1.84865832e-01 7.60942101e-01
2.35539988e-01 7.87796378e-01 -6.65369391e-01 -1.15122840e-01
6.59941137e-01 2.32888311e-01 5.71384095e-02 -6.42858148e-01
-1.00616777e+00 8.49263430e-01 6.49813712e-01 1.40505239e-01
-4.99974251e-01 3.74872416e-01 7.35598952e-02 4.63911027e-01
-1.86645880e-01 1.21502674e+00 -4.64088589e-01 -4.99572158e-01
-8.45759630e-01 4.89389598e-01 8.29794586e-01 8.71431351e-01
1.14926887e+00 3.12550142e-02 8.73058513e-02 2.90929019e-01
3.68773907e-01 5.33099055e-01 5.31899750e-01 -1.10568595e+00
3.66806030e-01 6.00714505e-01 4.25936520e-01 -4.37861890e-01
-5.26156843e-01 -4.05206889e-01 -1.19840965e-01 4.06949490e-01
5.35698906e-02 -2.91742027e-01 -8.84745836e-01 2.07905984e+00
8.98601264e-02 -1.83378801e-01 1.77004170e-02 4.58943725e-01
-3.00581157e-01 6.93326950e-01 4.05092478e-01 -1.96983516e-01
6.55518711e-01 -6.57359660e-01 -2.95728296e-01 -7.09914923e-01
7.50676751e-01 7.00613186e-02 9.62985814e-01 3.69778067e-01
-1.25826120e+00 -2.49878690e-01 -9.90042090e-01 3.86676282e-01
-4.19232398e-01 -2.68170327e-01 1.04294515e+00 3.65520567e-01
-8.57056260e-01 1.21300137e+00 -1.52153587e+00 -2.97582448e-01
5.58921456e-01 3.59121144e-01 -1.60829036e-03 2.00356722e-01
-8.09123635e-01 1.07552195e+00 7.66833067e-01 -7.92184249e-02
-1.35304010e+00 -2.90678173e-01 -8.48493874e-01 2.64307976e-01
4.63085264e-01 -2.47175947e-01 1.28374565e+00 -2.99229860e-01
-1.26157570e+00 4.30299252e-01 -8.77816454e-02 -9.21008766e-01
1.18605345e-01 -2.50244766e-01 -8.39220732e-02 -1.09931670e-01
2.67073929e-01 9.09727156e-01 2.99664199e-01 -7.64902830e-01
-7.38604605e-01 -3.28469485e-01 -7.61908814e-02 4.93789524e-01
-8.23867396e-02 -6.57886982e-01 6.82728961e-02 -3.15094441e-01
3.41818362e-01 -1.03727162e+00 -7.32197702e-01 -9.29920301e-02
-1.56818524e-01 -5.80199920e-02 2.74791241e-01 -1.13035344e-01
1.34453642e+00 -1.86551404e+00 1.77838191e-01 5.31268060e-01
-3.57354075e-01 -1.45238817e-01 -1.01577796e-01 7.57104933e-01
4.50488091e-01 3.06643751e-02 -1.59769371e-01 -4.51722089e-03
9.94083956e-02 4.31772739e-01 -4.88518387e-01 2.78109103e-01
-2.64601916e-01 7.65173495e-01 -1.11698568e+00 -1.66569814e-01
1.24117754e-01 -2.34640658e-01 -7.82891691e-01 2.13360205e-01
-6.35623395e-01 4.61546689e-01 -8.04277837e-01 7.29794979e-01
7.81420097e-02 -3.02702785e-01 4.77154464e-01 7.31533587e-01
-4.99625504e-01 8.06922734e-01 -9.87350702e-01 1.60245812e+00
-4.70419586e-01 3.00810099e-01 -1.87800810e-01 -8.63225400e-01
7.44119167e-01 -2.97176689e-01 5.69997072e-01 -9.20142770e-01
1.20683566e-01 3.14745098e-01 6.24298714e-02 -1.99824467e-01
5.42265773e-01 -3.00737638e-02 -2.00580776e-01 9.19968009e-01
-1.13504402e-01 -1.01614453e-01 3.33488017e-01 -1.44783169e-01
1.17099166e+00 5.93044698e-01 7.42724657e-01 -4.20153350e-01
-6.89594299e-02 1.95802599e-01 5.78580320e-01 1.18050623e+00
-2.27457479e-01 -3.05169642e-01 3.86745840e-01 -5.13427019e-01
-8.06195736e-01 -1.28550792e+00 1.88765523e-03 1.32308495e+00
3.69471490e-01 -1.57822505e-01 -4.56651330e-01 -4.06288445e-01
1.03614032e-01 1.17682838e+00 -7.09652066e-01 -3.60181928e-01
-7.01356173e-01 -3.82877469e-01 1.26283675e-01 8.47332299e-01
5.12143910e-01 -1.63610709e+00 -1.75465107e+00 3.57939482e-01
4.22494747e-02 -8.58242065e-02 -2.85364658e-01 1.19727302e+00
-1.23185396e+00 -7.82291889e-01 -2.49925420e-01 -5.57261705e-01
6.92198575e-01 5.58302328e-02 8.57108951e-01 9.63769481e-02
-1.81687906e-01 3.32114518e-01 -2.24992037e-01 -2.52290875e-01
7.67329261e-02 3.74977738e-01 3.11221570e-01 -9.24014509e-01
5.28523088e-01 -9.25367117e-01 -9.03513610e-01 1.72741920e-01
-4.43016380e-01 -2.96138022e-02 6.32352412e-01 8.58602107e-01
9.65559065e-01 2.69849122e-01 6.52818799e-01 -5.02566099e-01
4.76930976e-01 -8.16145837e-01 -8.18038821e-01 2.50789553e-01
-1.00392890e+00 8.35232794e-01 3.59867036e-01 -4.32055920e-01
-8.60169172e-01 1.03686176e-01 -1.18038028e-01 -1.90673128e-01
-8.44806433e-03 6.15207493e-01 1.66470677e-01 2.69277155e-01
6.37529969e-01 5.30925751e-01 -2.66248077e-01 -4.86993253e-01
3.12753290e-01 -4.74454537e-02 3.36701781e-01 -8.85229468e-01
3.26772690e-01 2.79845685e-01 -2.34315637e-02 -4.69747663e-01
-5.35541713e-01 -3.28631222e-01 -3.22173923e-01 -3.23939845e-02
5.23835421e-01 -6.72011733e-01 -1.15605259e+00 -7.77666178e-03
-2.36147806e-01 -9.64700699e-01 -6.78526819e-01 6.91959143e-01
-1.14440751e+00 -7.88322762e-02 -1.18641019e-01 -1.18753660e+00
-8.43205526e-02 -9.77080762e-01 5.24278581e-01 6.41078532e-01
-5.08425832e-01 -7.51063287e-01 3.32547247e-01 -3.26606452e-01
4.73336488e-01 4.95554246e-02 8.36720467e-01 -6.48141861e-01
-9.91350293e-01 2.48391330e-01 4.39347267e-01 -5.42847455e-01
7.70886336e-03 -6.65930033e-01 -3.37094635e-01 -6.18552446e-01
-1.96535230e-01 -5.07544100e-01 1.03946733e+00 4.75838423e-01
1.08069956e+00 -6.24847710e-01 -6.80410922e-01 7.35736072e-01
1.29548931e+00 7.53010154e-01 2.93941438e-01 7.89481580e-01
-3.06538194e-01 4.81486350e-01 8.66453528e-01 7.75137246e-01
2.77991325e-01 2.19192654e-01 6.13752842e-01 8.46013367e-01
6.36449873e-01 -9.88122880e-01 3.45211565e-01 1.01029038e-01
1.10576324e-01 -2.08973244e-01 -7.71777272e-01 9.73285496e-01
-1.69152308e+00 -1.22289538e+00 9.68861938e-01 2.30667853e+00
8.81237209e-01 4.13402677e-01 1.69640735e-01 -2.78562486e-01
4.96188365e-02 1.95729151e-01 -1.32686734e+00 -5.69613874e-01
3.37274551e-01 3.16206634e-01 7.91671515e-01 4.93498504e-01
-8.16288769e-01 9.49192703e-01 7.05513763e+00 4.22563851e-01
-9.12739158e-01 -3.59611034e-01 4.90106136e-01 -4.82783258e-01
-6.35163248e-01 1.34892792e-01 -1.04571950e+00 5.26653588e-01
8.88151884e-01 -2.09758446e-01 8.95784795e-01 1.40043008e+00
-9.23144147e-02 -7.48661101e-01 -1.23214376e+00 4.95374739e-01
-4.71987009e-01 -1.21987128e+00 -3.52833003e-01 3.15006733e-01
7.29055882e-01 1.47896841e-01 4.19668794e-01 5.26709199e-01
9.47116435e-01 -1.13528454e+00 7.19246030e-01 5.57880223e-01
3.86647224e-01 -8.64721358e-01 1.11822575e-01 8.72045755e-01
-1.30660236e+00 -6.75354719e-01 -3.18544477e-01 -3.04736286e-01
3.20785224e-01 -5.53969890e-02 -1.03275490e+00 -1.27654195e-01
8.23757291e-01 3.77059519e-01 -2.24849045e-01 1.14528263e+00
-3.62074256e-01 3.61858875e-01 -7.11698592e-01 -6.93527758e-01
4.98473495e-01 -3.74615103e-01 4.97646272e-01 6.78281069e-01
7.42602289e-01 3.16493273e-01 2.49629974e-01 1.27419734e+00
2.68892139e-01 -2.22577944e-01 -7.96448469e-01 -4.77933377e-01
7.36016572e-01 5.89384198e-01 -9.66738224e-01 1.10809453e-01
3.04115772e-01 4.74018812e-01 5.95598996e-01 8.99768993e-02
-3.27457845e-01 -2.41130963e-01 6.64670825e-01 8.77958909e-02
6.62619174e-01 -5.09209692e-01 -2.60297865e-01 -8.41599762e-01
-4.02113080e-01 -4.84423518e-01 4.92593378e-01 -5.40336967e-01
-6.41059220e-01 4.03420955e-01 3.01508337e-01 -9.13674533e-01
-6.78933799e-01 -3.49117368e-01 -6.19972229e-01 7.40088165e-01
-1.45951617e+00 -4.14478511e-01 3.49227428e-01 3.58035356e-01
4.80393350e-01 -1.07407823e-01 7.19126642e-01 -5.66299677e-01
-2.93781579e-01 4.09388959e-01 2.50088513e-01 -3.20842326e-01
4.24262956e-02 -1.31773055e+00 3.90967101e-01 7.14916527e-01
1.32982627e-01 9.64981318e-01 8.58567417e-01 -1.05025780e+00
-1.29183853e+00 -6.38131320e-01 4.80835974e-01 -9.13167819e-02
3.70093524e-01 2.64077773e-03 -7.60537326e-01 1.05585229e+00
-1.49221480e-01 -5.20317197e-01 6.56167209e-01 4.87686008e-01
-4.83189411e-02 1.24786466e-01 -1.11329043e+00 8.66549909e-01
1.15414083e+00 -3.04829717e-01 -7.37299085e-01 -2.08284169e-01
3.68696272e-01 -4.05695140e-01 -5.41159332e-01 1.53643414e-01
6.78749561e-01 -9.58275616e-01 7.69915938e-01 -7.17868865e-01
7.64533132e-02 -8.61980841e-02 -2.22419560e-01 -1.30055547e+00
-2.98743784e-01 -7.44572461e-01 -3.95744175e-01 3.57244879e-01
4.85316247e-01 -5.45800269e-01 9.79836881e-01 6.16407573e-01
-1.16471827e-01 -1.02856088e+00 -9.61431444e-01 -7.70933032e-01
-4.62934077e-02 -1.44370869e-01 7.93353856e-01 2.24064246e-01
4.39342290e-01 -3.68750662e-01 -3.56547356e-01 -3.48817408e-02
6.93691492e-01 6.65016711e-01 3.87767106e-01 -8.63916636e-01
-4.48171556e-01 -4.58176404e-01 1.17227659e-01 -1.44673002e+00
-9.17674042e-03 -7.18019366e-01 4.39151973e-01 -1.32954216e+00
-3.73825105e-03 -8.32203388e-01 -5.62273800e-01 7.23509848e-01
1.59487575e-01 -6.36252642e-01 -7.57967383e-02 6.99694008e-02
-8.03015888e-01 5.39856315e-01 1.03244972e+00 1.39551669e-01
-7.51922190e-01 2.96589762e-01 -8.67126822e-01 6.18068755e-01
9.45098579e-01 -4.73790467e-01 -7.22022831e-01 -1.95942760e-01
8.77287924e-01 4.35874373e-01 -1.03615681e-02 -1.19273961e+00
3.85073513e-01 -7.67912626e-01 5.40711403e-01 -9.03934240e-01
3.58795166e-01 -5.46132863e-01 2.34465405e-01 8.88197005e-01
-9.47446406e-01 1.83993980e-01 2.54625827e-01 9.78188992e-01
3.28297168e-01 -4.40972537e-01 6.92983389e-01 -4.15770233e-01
-8.92392814e-01 3.14715117e-01 -7.28844166e-01 -3.36644705e-04
1.13889003e+00 -3.31352174e-01 1.24068223e-02 -2.90890366e-01
-9.05144513e-01 8.35678101e-01 6.28612101e-01 1.85368638e-02
5.26542544e-01 -1.00832093e+00 1.28786668e-01 4.14907008e-01
-1.27156511e-01 -1.07136637e-01 3.54473233e-01 3.99026275e-01
-3.44159156e-01 6.17972612e-01 -7.61708617e-01 -1.48667127e-01
-5.65776944e-01 7.33630359e-01 3.92958879e-01 -5.44327140e-01
-5.68175197e-01 9.58104789e-01 -1.83339894e-01 -2.57395595e-01
2.66981423e-01 -5.29289722e-01 -1.58349976e-01 -1.46277383e-01
4.50104237e-01 5.56165159e-01 -4.73325044e-01 6.13455698e-02
-2.21075162e-01 1.59391865e-01 -1.84046447e-01 -5.14713109e-01
1.42960536e+00 -3.12181562e-01 2.18295664e-01 4.30060774e-01
6.87265992e-01 -3.69315922e-01 -1.75988853e+00 -1.16423577e-01
3.73891816e-02 -5.52532911e-01 -6.95696399e-02 -8.97236764e-01
-4.99134898e-01 7.08007097e-01 4.29310322e-01 5.28692603e-02
9.34350967e-01 -1.35455308e-02 7.09498405e-01 1.05586147e+00
1.01070225e+00 -1.36800170e+00 1.74466074e-01 4.47117180e-01
3.44022423e-01 -1.00755489e+00 1.76893130e-01 6.19095981e-01
-7.19533801e-01 7.97679484e-01 7.59167492e-01 -1.12083733e-01
6.27935946e-01 1.88125774e-01 -5.30353665e-01 -1.44937396e-01
-1.03044319e+00 -3.39093089e-01 -2.10757732e-01 5.57363033e-01
-2.91840043e-02 1.84781775e-01 -5.19699529e-02 3.10145348e-01
-4.80487347e-01 1.94844101e-02 2.00912848e-01 1.38128686e+00
-1.17768669e+00 -1.06879044e+00 3.09854411e-02 6.82415128e-01
-6.03151917e-02 1.41891837e-01 5.80090135e-02 6.01925015e-01
-1.55277951e-02 5.83645046e-01 1.64882481e-01 -1.82329953e-01
-1.26472175e-01 2.05721319e-01 5.88644624e-01 -5.49725890e-01
-4.33777004e-01 -1.61611840e-01 -1.44398063e-01 -8.97862434e-01
-6.24042340e-02 -1.16217029e+00 -1.51784217e+00 -3.25398654e-01
-1.00364387e-01 5.52898288e-01 4.60072249e-01 8.34702730e-01
3.44054312e-01 1.92946732e-01 6.07571185e-01 -6.69935226e-01
-8.97823572e-01 -5.74243188e-01 -6.90829337e-01 -2.11547300e-01
3.79407346e-01 -8.57602954e-01 -2.40745619e-01 -4.59383756e-01] | [3.966669797897339, 1.6859065294265747] |
bc959165-164f-45da-85e9-b56d70d2751f | signed-link-representation-in-continuous-time | 2207.03408 | null | https://arxiv.org/abs/2207.03408v3 | https://arxiv.org/pdf/2207.03408v3.pdf | Representation Learning in Continuous-Time Dynamic Signed Networks | Signed networks allow us to model conflicting relationships and interactions, such as friend/enemy and support/oppose. These signed interactions happen in real-time. Modeling such dynamics of signed networks is crucial to understanding the evolution of polarization in the network and enabling effective prediction of the signed structure (i.e., link signs and signed weights) in the future. However, existing works have modeled either (static) signed networks or dynamic (unsigned) networks but not dynamic signed networks. Since both sign and dynamics inform the graph structure in different ways, it is non-trivial to model how to combine the two features. In this work, we propose a new Graph Neural Network (GNN)-based approach to model dynamic signed networks, named SEMBA: Signed link's Evolution using Memory modules and Balanced Aggregation. Here, the idea is to incorporate the signs of temporal interactions using separate modules guided by balance theory and to evolve the embeddings from a higher-order neighborhood. Experiments on 4 real-world datasets and 4 different tasks demonstrate that SEMBA consistently and significantly outperforms the baselines by up to $80\%$ on the tasks of predicting signs of future links while matching the state-of-the-art performance on predicting the existence of these links in the future. We find that this improvement is due specifically to the superior performance of SEMBA on the minority negative class. | ['Kartik Sharma', 'Srijan Kumar', 'Yeon Chang Lee', 'Anand Kumar M', 'Mohit Raghavendra'] | 2022-07-07 | null | null | null | null | ['link-sign-prediction'] | ['graphs'] | [-8.13672766e-02 2.61936992e-01 -6.28848732e-01 -3.56346250e-01
7.89563835e-01 -6.29461467e-01 9.06728506e-01 2.78322488e-01
3.97853591e-02 7.46847510e-01 2.43971631e-01 -3.23202193e-01
-6.83981776e-01 -1.17601228e+00 -7.45028019e-01 -4.95161384e-01
-9.60572898e-01 7.31383324e-01 4.78700697e-01 -7.31149971e-01
2.02211499e-01 4.43878412e-01 -1.20801198e+00 6.55443668e-02
7.82916605e-01 7.97904432e-01 -4.99639988e-01 6.79054976e-01
-6.82765543e-02 8.18262696e-01 -2.30761662e-01 -7.90132344e-01
2.02510104e-01 -3.23886603e-01 -6.71567082e-01 -3.03445339e-01
4.40138072e-01 1.60937712e-01 -5.52610874e-01 9.93950784e-01
2.32171252e-01 -1.86452106e-01 6.73506737e-01 -1.56808913e+00
-8.49137008e-01 1.14300060e+00 -6.76913679e-01 4.63132083e-01
4.17518131e-02 1.40840635e-01 1.50961614e+00 -4.53046322e-01
1.10148489e+00 1.35231566e+00 8.14743578e-01 3.11822414e-01
-1.29097164e+00 -5.78755498e-01 5.54903626e-01 4.84393686e-01
-8.80556226e-01 -1.36892393e-01 1.20771289e+00 -5.63534915e-01
7.96654284e-01 7.66489878e-02 1.06054318e+00 1.27560353e+00
2.60025173e-01 2.97786504e-01 8.70525658e-01 -7.28821456e-02
-1.34984657e-01 -1.62740588e-01 5.96327007e-01 9.08609450e-01
5.99699676e-01 4.29678738e-01 -8.10104370e-01 -1.68128133e-01
3.76152813e-01 -3.14587772e-01 -3.15152615e-01 -7.65242636e-01
-1.09365296e+00 9.07667041e-01 1.18211305e+00 4.73818302e-01
-2.46216342e-01 3.22555900e-01 1.95484594e-01 5.23907959e-01
5.84826767e-01 4.29951817e-01 -5.11243343e-01 7.56463502e-03
-4.54004258e-01 4.94179726e-02 1.14000082e+00 2.17375368e-01
6.41775489e-01 -1.60727501e-01 1.55259609e-01 6.82138503e-01
6.04373336e-01 1.94586322e-01 1.07488766e-01 -6.46279931e-01
4.71254826e-01 9.43407655e-01 -3.33541930e-01 -1.76417589e+00
-6.35195553e-01 -9.47382927e-01 -8.88671994e-01 -4.41289246e-02
4.17543501e-01 -6.48195073e-02 -6.96218967e-01 2.18473125e+00
2.09188566e-01 4.66489524e-01 -4.23012078e-01 4.58346128e-01
6.11187279e-01 5.41261733e-01 -3.32334071e-01 -7.06370249e-02
7.09964037e-01 -8.75663102e-01 -5.31309485e-01 -4.00795788e-01
4.54527736e-01 -2.06410930e-01 4.11134183e-01 7.36144632e-02
-9.17005241e-01 -1.00996211e-01 -1.29343605e+00 3.45267266e-01
-7.79042482e-01 -5.94251812e-01 9.30354834e-01 3.99408191e-01
-1.36774838e+00 1.10368717e+00 -7.01980472e-01 -5.01200736e-01
3.69485527e-01 6.43649876e-01 -2.09453881e-01 1.67615488e-01
-1.54713106e+00 8.35740924e-01 2.14398012e-01 4.35630322e-01
-3.05203140e-01 -6.48070335e-01 -6.59863770e-01 7.10240677e-02
1.66222706e-01 -6.74032927e-01 4.32068765e-01 -9.67117310e-01
-9.84245539e-01 8.45608532e-01 3.33938450e-02 -6.19780838e-01
6.80959523e-01 3.01944882e-01 -7.36342430e-01 -1.05004489e-01
-3.34946841e-01 5.87272227e-01 7.39858389e-01 -1.29237473e+00
-1.93827868e-01 -4.23618466e-01 2.34005198e-01 -1.06781520e-01
-9.23026741e-01 -4.41079319e-01 -2.52486736e-01 -3.82430345e-01
1.23781860e-01 -1.11169541e+00 9.66157243e-02 2.77865767e-01
-3.48776788e-01 -2.76424617e-01 9.68841791e-01 -3.61476570e-01
1.64870930e+00 -1.72744334e+00 5.31786263e-01 5.94258130e-01
6.27321184e-01 4.32028770e-01 -4.04329777e-01 5.22241294e-01
-2.42746904e-01 3.20920020e-01 -1.54799923e-01 -4.15917598e-02
-5.55596091e-02 4.51420099e-01 -2.23925322e-01 2.49228776e-01
1.50089502e-01 1.14374995e+00 -1.07258356e+00 -2.26954043e-01
-2.90730923e-01 6.63784325e-01 -6.52883291e-01 -2.12883264e-01
-9.99593213e-02 1.95774660e-01 -7.43731484e-02 5.25831401e-01
5.96794784e-01 -5.23260295e-01 8.81089568e-01 -2.44035333e-01
1.51895493e-01 3.16024125e-01 -1.02760553e+00 8.59808147e-01
-2.51238085e-02 8.27130020e-01 1.48581594e-01 -1.14577222e+00
9.97018099e-01 -2.81008869e-01 4.12607074e-01 -7.13728547e-01
1.22267134e-01 4.32129979e-01 6.66502655e-01 -1.94801241e-01
2.34590590e-01 2.26753831e-01 3.31218243e-01 5.11963308e-01
-3.33520435e-02 3.29770386e-01 6.37746990e-01 5.95075905e-01
1.37271202e+00 -3.24600369e-01 -3.92354690e-02 -3.52466911e-01
4.50458318e-01 -4.41492409e-01 5.27314603e-01 4.18669313e-01
-3.12410623e-01 -1.82080176e-02 1.14740694e+00 -4.79201287e-01
-5.90315700e-01 -1.21139944e+00 6.61735386e-02 9.45981264e-01
2.05257818e-01 -3.17475498e-01 -9.27501842e-02 -8.70079696e-01
5.22227347e-01 1.70184419e-01 -1.00313127e+00 -6.40044391e-01
-9.35004294e-01 -9.95723188e-01 2.22387254e-01 4.06198353e-01
2.42663309e-01 -9.41509008e-01 2.06338972e-01 1.97256386e-01
8.22638627e-03 -8.06567013e-01 -3.43685478e-01 8.17079544e-02
-1.02788448e+00 -1.51684284e+00 -1.67243242e-01 -7.75972366e-01
6.78861916e-01 9.77310166e-02 1.29604197e+00 4.53209698e-01
-1.00535363e-01 2.13915810e-01 -8.97615999e-02 1.11838095e-01
-1.78350717e-01 9.55525637e-02 7.26086348e-02 2.31960565e-01
-1.27420381e-01 -1.07184315e+00 -7.36619115e-01 2.78379649e-01
-6.51308775e-01 -2.29168326e-01 4.18217301e-01 9.49137449e-01
-1.45744815e-01 1.05599038e-01 5.72095394e-01 -1.00511146e+00
9.34075475e-01 -6.68186903e-01 -2.92735130e-01 3.27224880e-01
-1.09994197e+00 1.69318587e-01 4.22880322e-01 -6.21489882e-01
-5.90388358e-01 -6.65981531e-01 3.08966488e-01 -2.55674124e-02
8.52654457e-01 8.17178547e-01 2.14217260e-01 -2.88245857e-01
4.75406319e-01 -3.41667652e-01 1.03781417e-01 -1.42122790e-01
4.27105099e-01 1.18105233e-01 2.09340125e-01 -5.36105156e-01
9.54834223e-01 6.26451552e-01 2.77631998e-01 -4.09467936e-01
-6.53963804e-01 -3.70738991e-02 -7.62084365e-01 -4.72830087e-01
2.19556049e-01 -4.77367789e-01 -6.92921877e-01 6.51251912e-01
-1.08016264e+00 -4.00739282e-01 -3.78674716e-02 2.58711241e-02
5.84284589e-02 1.96878389e-01 -7.39951611e-01 -5.22669315e-01
-2.36510351e-01 -6.42413855e-01 4.17575955e-01 1.65823534e-01
-3.61026973e-01 -1.63481307e+00 3.82809818e-01 1.53365135e-01
6.38553977e-01 5.32762766e-01 1.18121147e+00 -4.53229308e-01
-6.61961913e-01 -2.22018585e-01 -3.71056259e-01 9.15037468e-02
-6.44296780e-02 4.90030497e-01 -4.04338330e-01 -1.98182270e-01
-7.00408161e-01 -3.15461936e-03 1.27876246e+00 1.83274463e-01
5.51509380e-01 -4.10159796e-01 -6.30023301e-01 5.06588876e-01
1.05328691e+00 -9.61583853e-02 4.22975391e-01 8.52736458e-02
7.19787419e-01 7.90226638e-01 1.72499582e-01 2.88899571e-01
8.06949496e-01 6.25180244e-01 9.10624146e-01 1.68159515e-01
-3.29919189e-01 -4.01469797e-01 5.10606825e-01 1.14750493e+00
-1.00361235e-01 -5.17675877e-01 -9.66307938e-01 5.96116543e-01
-2.09848142e+00 -1.11099315e+00 -5.69286108e-01 1.85137331e+00
7.90700614e-01 7.56546378e-01 9.97570753e-02 4.29567471e-02
8.97899389e-01 7.34573126e-01 -6.13881350e-01 -4.75222230e-01
-3.65759373e-01 7.15735555e-02 3.68080646e-01 7.80062675e-01
-8.84642363e-01 6.15178883e-01 5.81056118e+00 1.76745340e-01
-1.41872847e+00 -2.23545626e-01 5.63003659e-01 -6.75787553e-02
-5.35960913e-01 2.23713815e-01 -6.98109090e-01 5.42117298e-01
8.14449787e-01 -1.56975418e-01 6.76364422e-01 2.70458698e-01
2.12336294e-02 2.72648185e-01 -9.74302471e-01 5.34705698e-01
1.49166331e-01 -1.44446993e+00 3.53170559e-03 3.79999369e-01
8.35230708e-01 2.23256797e-01 2.02515855e-01 3.12470734e-01
5.75793326e-01 -7.13680923e-01 4.87180591e-01 6.98155582e-01
2.84682423e-01 -3.62929106e-01 6.14928246e-01 1.80204168e-01
-1.24959910e+00 -4.82678652e-01 1.50341526e-01 -1.88390180e-01
2.16306165e-01 9.35240626e-01 -6.72794342e-01 3.99232626e-01
4.89502609e-01 1.39164984e+00 -8.73292267e-01 8.74526858e-01
-4.08241600e-01 5.60205877e-01 -2.79852569e-01 -2.30314255e-01
9.26971287e-02 -3.37583691e-01 7.31367946e-01 8.59663367e-01
2.49300495e-01 -3.76297057e-01 -1.60204098e-01 7.84540534e-01
-5.07004440e-01 -3.70491236e-01 -6.27773225e-01 -5.23801744e-01
4.51195836e-01 1.27986336e+00 -6.42491162e-01 -1.83702558e-02
-3.49801816e-02 6.25627756e-01 6.70864701e-01 3.39375138e-01
-8.75007629e-01 -1.42648339e-01 7.41705835e-01 5.56620657e-01
2.88257390e-01 -4.14498031e-01 -9.34657305e-02 -1.16982889e+00
3.28605175e-02 -4.59533542e-01 5.91132760e-01 -4.68488753e-01
-1.59280658e+00 6.09119058e-01 -3.22833329e-01 -8.48785162e-01
-1.54227450e-01 -7.39795446e-01 -9.48921323e-01 3.25399548e-01
-1.74361205e+00 -1.21766949e+00 -2.49478742e-01 2.66254902e-01
-3.28051269e-01 1.65165246e-01 3.54786426e-01 4.31641161e-01
-7.00607777e-01 3.39747697e-01 -3.38313654e-02 1.86599240e-01
6.99339926e-01 -1.19536495e+00 3.00706536e-01 6.75492525e-01
2.02253312e-01 7.52667904e-01 6.19468272e-01 -8.92489195e-01
-1.31546521e+00 -9.03574347e-01 1.32864809e+00 -4.27527934e-01
1.40510106e+00 -4.24369007e-01 -7.20092475e-01 5.81212640e-01
1.55129567e-01 4.38858569e-01 3.63213271e-01 5.44432044e-01
-7.07555473e-01 -4.37978417e-01 -7.93949008e-01 6.41619027e-01
1.82516372e+00 -3.98841470e-01 -2.79661983e-01 1.63406610e-01
5.49514771e-01 3.38427387e-02 -6.38895512e-01 8.33043575e-01
7.20640898e-01 -1.16563284e+00 1.16314518e+00 -7.48592257e-01
5.40568233e-01 -8.60249326e-02 1.01910658e-01 -1.66590500e+00
-4.53728080e-01 -4.73703980e-01 -8.05818498e-01 1.22171962e+00
5.23955882e-01 -1.05526757e+00 8.36355567e-01 2.09170744e-01
2.58995980e-01 -1.07820475e+00 -8.28053355e-01 -7.49107242e-01
-5.62379695e-02 8.90591182e-03 5.48436284e-01 1.41520953e+00
-3.58687826e-02 4.29519862e-01 -3.19768637e-01 -1.72449201e-02
5.59987783e-01 1.58140868e-01 4.35359895e-01 -1.87820697e+00
-3.30512673e-01 -9.93950307e-01 -7.35249043e-01 -6.57191932e-01
4.11273479e-01 -1.29564691e+00 -6.64628804e-01 -1.61938667e+00
7.50370622e-02 -6.25606000e-01 -5.41241825e-01 3.08325827e-01
-7.48414472e-02 1.95555776e-01 1.30109549e-01 2.71364152e-02
-6.47104740e-01 6.32021368e-01 1.18907797e+00 -5.22269249e-01
-2.07288176e-01 -1.07073903e-01 -7.88374662e-01 6.39526367e-01
6.56732202e-01 -3.23296130e-01 -4.04698402e-01 -4.14782435e-01
8.77079546e-01 -2.38810867e-01 4.54662979e-01 -7.52558053e-01
3.64811629e-01 5.62810563e-02 1.10750437e-01 -3.19365859e-01
3.30686361e-01 -6.92109644e-01 6.19666763e-02 7.51543343e-01
-2.61623263e-01 8.64439309e-02 -2.48440146e-01 9.48361099e-01
-1.34222925e-01 2.53324091e-01 5.80303848e-01 2.43192405e-01
-5.06493330e-01 5.21087587e-01 4.04449999e-01 1.55416891e-01
9.45570350e-01 -1.39868468e-01 -1.04132903e+00 -5.94043314e-01
-9.05122817e-01 5.98391294e-01 3.04416567e-01 7.54346490e-01
4.52050239e-01 -1.37756193e+00 -4.43511754e-01 1.45431712e-01
-1.12181351e-01 -6.65857971e-01 4.12596650e-02 1.06611776e+00
-2.71789581e-01 9.35761854e-02 -2.38798961e-01 -3.47600430e-01
-1.30979836e+00 1.27522707e-01 4.45623547e-01 -8.18752527e-01
6.32895529e-02 1.08898318e+00 -3.79725635e-01 -7.75029838e-01
1.38312310e-01 1.09902602e-02 -3.79788756e-01 7.19017923e-01
6.94499258e-03 5.49507737e-01 -1.25048280e-01 -6.38258815e-01
-4.92046237e-01 7.33137608e-01 3.42726558e-02 1.98169500e-01
1.78220999e+00 4.67278399e-02 -6.62463367e-01 4.89510566e-01
1.35028124e+00 7.63516277e-02 -9.52521861e-01 -3.65952343e-01
2.75329053e-01 -3.24292451e-01 -2.25678056e-01 -9.10445988e-01
-1.48806548e+00 9.17792499e-01 3.42193544e-01 5.70203543e-01
7.25654840e-01 7.07979575e-02 7.23021269e-01 1.78255051e-01
2.85701215e-01 -8.19883227e-01 4.44491535e-01 7.73404360e-01
9.72353458e-01 -1.14423692e+00 1.86703548e-01 -5.25430202e-01
-2.91476578e-01 1.14980292e+00 7.46713459e-01 -1.37769073e-01
1.28756261e+00 -6.97656199e-02 -2.53792077e-01 -5.77607870e-01
-1.10376465e+00 4.36245650e-02 4.35021490e-01 4.43137646e-01
3.88280421e-01 1.50313944e-01 -5.09506464e-01 1.54228240e-01
-1.40355289e-01 -3.46675605e-01 3.85757983e-01 7.41320372e-01
-2.58294940e-01 -1.52480507e+00 8.25558677e-02 7.76167691e-01
1.98576879e-03 4.63575087e-02 -1.06606257e+00 7.11966634e-01
1.89456701e-01 8.03171635e-01 1.56769678e-01 -6.42679811e-01
2.28969842e-01 -1.22623788e-02 4.35240239e-01 -3.01544905e-01
-5.32775462e-01 -4.61149275e-01 3.31526518e-01 -5.25856256e-01
-4.49352413e-01 -6.32903039e-01 -1.05984819e+00 -6.79429710e-01
-2.80328602e-01 -1.49119899e-01 4.57508951e-01 5.07121086e-01
5.77840269e-01 6.76476479e-01 7.27789640e-01 -6.85791016e-01
-3.74527961e-01 -7.02045143e-01 -5.91414630e-01 6.27168477e-01
2.95703411e-01 -1.08274603e+00 -7.50439346e-01 -4.71514225e-01] | [7.183151721954346, 6.170722484588623] |
ddc22dfd-99c0-4823-95c8-7db385ccf27f | visual-analysis-of-discrimination-in-machine | 2007.15182 | null | https://arxiv.org/abs/2007.15182v2 | https://arxiv.org/pdf/2007.15182v2.pdf | Visual Analysis of Discrimination in Machine Learning | The growing use of automated decision-making in critical applications, such as crime prediction and college admission, has raised questions about fairness in machine learning. How can we decide whether different treatments are reasonable or discriminatory? In this paper, we investigate discrimination in machine learning from a visual analytics perspective and propose an interactive visualization tool, DiscriLens, to support a more comprehensive analysis. To reveal detailed information on algorithmic discrimination, DiscriLens identifies a collection of potentially discriminatory itemsets based on causal modeling and classification rules mining. By combining an extended Euler diagram with a matrix-based visualization, we develop a novel set visualization to facilitate the exploration and interpretation of discriminatory itemsets. A user study shows that users can interpret the visually encoded information in DiscriLens quickly and accurately. Use cases demonstrate that DiscriLens provides informative guidance in understanding and reducing algorithmic discrimination. | ['Zhutian Chen', 'Shixia Liu', 'Qianwen Wang', 'Zhenhua Xu', 'Yong Wang', 'Huamin Qu'] | 2020-07-30 | null | null | null | null | ['crime-prediction'] | ['miscellaneous'] | [ 1.25483215e-01 5.06234281e-02 -4.99119371e-01 -5.67691803e-01
-7.14249304e-03 -7.39905894e-01 4.44777995e-01 9.28533256e-01
-2.30068251e-01 6.96384847e-01 3.57554674e-01 -1.20802522e+00
-6.35986507e-01 -6.27163947e-01 2.29616642e-01 -2.32210115e-01
-2.11562783e-01 2.88389862e-01 -1.93528712e-01 4.28581201e-02
8.72608960e-01 8.37130308e-01 -1.75974095e+00 5.84456146e-01
1.54470789e+00 5.36066115e-01 -5.46949089e-01 3.56002182e-01
-3.34816009e-01 6.61687076e-01 -8.71961832e-01 -7.43566155e-01
1.77610904e-01 -4.23419744e-01 -6.57672405e-01 -4.44908798e-01
5.25778174e-01 -1.72446102e-01 5.09848654e-01 1.01058829e+00
3.06947738e-01 -2.67177578e-02 8.98646832e-01 -1.72996151e+00
-5.69787502e-01 7.52210081e-01 -7.85910845e-01 4.58125532e-01
3.86768639e-01 1.93542391e-01 8.88061106e-01 -5.17628312e-01
4.84197944e-01 1.61118793e+00 4.18630660e-01 2.36187205e-01
-1.50949609e+00 -1.25165093e+00 1.18496098e-01 3.59361947e-01
-1.17889845e+00 -1.11910641e-01 5.93743145e-01 -9.75680113e-01
5.24550200e-01 1.14461410e+00 9.72048521e-01 6.28796041e-01
1.02397151e-01 5.15322804e-01 1.52290606e+00 -6.17047250e-01
3.52351785e-01 1.85950220e-01 4.68217909e-01 8.35675299e-01
9.98876214e-01 1.22066155e-01 -6.08357966e-01 -5.60344040e-01
5.56251109e-01 1.63366631e-01 9.99004990e-02 -3.93525183e-01
-7.10572064e-01 1.19713902e+00 2.74884701e-01 2.11993000e-04
1.21708773e-01 -3.30725402e-01 3.70197266e-01 1.42248303e-01
5.38252413e-01 7.57642806e-01 8.04332867e-02 -1.34596154e-01
-1.03264260e+00 4.38165158e-01 5.37389576e-01 2.78133363e-01
5.21678627e-01 -1.32868379e-01 -4.19560581e-01 5.92172563e-01
3.77159387e-01 2.85478979e-01 1.61046460e-01 -9.31360960e-01
1.92071691e-01 1.20831144e+00 1.76421016e-01 -1.31687343e+00
-5.81826150e-01 -3.09911847e-01 -6.59554541e-01 9.24266934e-01
6.60280585e-01 -2.41006371e-02 -6.56060576e-01 1.16342938e+00
4.32036966e-01 -2.94503152e-01 -4.40931559e-01 9.72085357e-01
7.08769977e-01 4.60623987e-02 7.72001445e-01 -2.07783967e-01
1.25763416e+00 -1.41477361e-01 -7.89797008e-01 2.70249788e-02
1.03913069e+00 -4.29871619e-01 1.20427454e+00 5.28443336e-01
-7.23729908e-01 -7.85941854e-02 -1.12489879e+00 -2.36425161e-01
-6.08090699e-01 1.25829093e-02 8.05102646e-01 1.17282081e+00
-4.50461894e-01 8.57455194e-01 -6.87740684e-01 -4.97441627e-02
1.16598165e+00 -2.15966254e-02 -8.46232176e-02 6.24618158e-02
-1.02639937e+00 8.53482842e-01 2.52756983e-01 -5.15262544e-01
-1.88417897e-01 -1.19584775e+00 -8.31240594e-01 3.58571678e-01
2.43628815e-01 -5.67404985e-01 9.16102111e-01 -7.93301523e-01
-8.09790671e-01 7.29259253e-01 2.11275741e-01 -2.14493230e-01
9.02447224e-01 -1.77060977e-01 -3.24544668e-01 -1.39552370e-01
1.18806586e-01 2.75920659e-01 2.82818317e-01 -1.16531122e+00
-8.07024300e-01 -6.97981060e-01 -5.61317243e-02 5.87200895e-02
-4.23791736e-01 3.67367983e-01 4.06978995e-01 -8.48085225e-01
-2.10964352e-01 -5.36380529e-01 -2.94912368e-01 2.83375621e-01
-5.91810584e-01 -2.54996002e-01 9.98832941e-01 -4.99065608e-01
1.98494446e+00 -1.99464643e+00 -4.75024432e-01 6.83690012e-01
5.32654226e-01 3.05656254e-01 4.79374856e-01 1.61817566e-01
-3.99439573e-01 8.73317122e-01 -1.23987846e-01 -2.89680250e-02
2.50742674e-01 -4.30750400e-02 -2.06351399e-01 2.96345532e-01
-4.75703478e-02 3.95974129e-01 -1.04279089e+00 -6.90690935e-01
3.32766473e-01 1.76429927e-01 -5.26536703e-01 -2.54152231e-02
1.35123149e-01 1.44726068e-01 -2.87964374e-01 6.05884135e-01
7.92093754e-01 -2.57261366e-01 5.33646703e-01 1.80799127e-01
-5.48841298e-01 3.79714400e-01 -1.20120096e+00 1.11625314e+00
2.85419561e-02 9.00170624e-01 -1.65911198e-01 -6.44656241e-01
9.22730029e-01 -2.91772544e-01 -2.76355725e-02 -6.89136267e-01
1.79583952e-02 -1.02775790e-01 1.91244632e-01 -5.64525008e-01
4.25047964e-01 -1.18756779e-01 -8.20556805e-02 9.78697598e-01
-5.91706216e-01 2.68747002e-01 4.17039067e-01 3.82523119e-01
7.85136759e-01 -3.07955831e-01 8.09920132e-01 -6.11122668e-01
4.38945480e-02 2.76342988e-01 5.87555647e-01 7.32294619e-01
-4.45776135e-02 1.96057767e-01 8.45802069e-01 -7.62083471e-01
-6.57566547e-01 -1.09255600e+00 -2.63778538e-01 1.22621143e+00
-9.60635021e-02 -7.07408965e-01 -5.65938830e-01 -8.54388356e-01
4.61054534e-01 1.12360048e+00 -8.52219403e-01 5.70474155e-02
-2.04535737e-03 -7.59499133e-01 3.02577525e-01 3.01786453e-01
2.22646385e-01 -6.42768979e-01 -1.42673767e+00 -4.73474354e-01
1.90515324e-01 -1.66578174e-01 -1.33037135e-01 -8.57777819e-02
-8.97289217e-01 -1.42968047e+00 4.43927292e-03 -1.50114475e-02
6.53489709e-01 9.61973891e-02 1.23621202e+00 3.46162915e-01
-9.05307651e-01 -1.29927248e-01 -2.22566143e-01 -1.05539644e+00
-3.89836103e-01 -1.90292642e-01 -1.42398477e-01 -3.52490842e-01
4.64782000e-01 -4.63718891e-01 -7.44322956e-01 2.36114070e-01
-7.19790876e-01 5.32437146e-01 1.76268533e-01 3.48461628e-01
1.16718128e-01 -1.29787579e-01 4.29755449e-01 -1.34258723e+00
1.23404086e+00 -5.17496586e-01 -6.71548486e-01 4.29639041e-01
-1.17421937e+00 1.89845681e-01 3.85290205e-01 -3.03819180e-01
-1.03042805e+00 -2.73545444e-01 4.07818139e-01 -2.83540487e-01
-2.51715481e-01 4.21555936e-01 -2.82230657e-02 1.87670350e-01
1.02740002e+00 -6.58140361e-01 -1.26353234e-01 -4.73915100e-01
3.87338579e-01 6.10153973e-01 2.77792532e-02 -4.12974358e-01
3.68113458e-01 2.68484563e-01 7.92366937e-02 -2.67758369e-01
-3.84531558e-01 -6.37535080e-02 -6.66442335e-01 -4.78423893e-01
5.69743037e-01 -2.69915998e-01 -1.39680660e+00 -2.64600962e-01
-7.33614981e-01 -2.46119305e-01 -1.87664211e-01 2.96864778e-01
-6.11406267e-02 1.02974497e-01 1.49981547e-02 -1.13429844e+00
-2.22491562e-01 -8.63870442e-01 4.56642926e-01 5.41276336e-01
-1.01461482e+00 -9.42684829e-01 6.70664012e-02 3.15501541e-01
8.69756863e-02 7.08525300e-01 1.48301804e+00 -8.92941177e-01
-2.11476550e-01 6.99428841e-02 -2.24117190e-01 -4.03568655e-01
2.22135529e-01 4.57571894e-01 -8.08535635e-01 1.33626372e-01
-8.90560448e-01 -2.35375781e-02 4.93141562e-01 1.77019879e-01
1.60816610e+00 -7.55770981e-01 -5.22310495e-01 3.86423647e-01
1.08624744e+00 5.37618399e-01 4.64214712e-01 3.59694541e-01
5.36023200e-01 1.00583613e+00 8.21194470e-01 7.94903100e-01
1.93598762e-01 3.52131844e-01 4.08314943e-01 -4.62837577e-01
3.90124708e-01 -2.89339244e-01 -4.21018392e-01 -1.06003553e-01
-3.03000063e-01 1.01248808e-01 -1.25811982e+00 5.13131656e-02
-1.84239054e+00 -1.14719141e+00 -3.65020126e-01 2.29594088e+00
7.42742598e-01 1.30960837e-01 4.80628133e-01 4.89826769e-01
7.30215967e-01 -3.52792069e-02 -5.46454787e-01 -1.12836468e+00
3.74251693e-01 7.31702894e-02 2.98618257e-01 4.62087691e-01
-9.62447166e-01 4.54288304e-01 6.69307089e+00 8.94351125e-01
-1.14353359e+00 -2.35233620e-01 1.19643831e+00 -3.58314514e-01
-6.75662577e-01 -5.63126691e-02 -2.17930958e-01 4.03286368e-01
4.49248046e-01 -6.88112557e-01 -1.20530948e-01 7.31570184e-01
7.39035606e-01 -3.44342977e-01 -1.12316370e+00 1.04148686e+00
-3.92178982e-01 -1.78429472e+00 4.47978407e-01 1.98030084e-01
4.39345449e-01 -8.60556066e-01 2.29552209e-01 -1.02109820e-01
8.84921908e-01 -1.48967803e+00 6.29725814e-01 5.64800918e-01
9.54977274e-01 -1.15244377e+00 1.61880508e-01 2.16694146e-01
-6.51075363e-01 -3.91146988e-01 1.33751929e-01 -6.00332618e-01
-3.59014899e-01 6.94779217e-01 -1.15194428e+00 4.64614868e-01
7.61819482e-01 5.42678773e-01 -7.81364977e-01 1.10509372e+00
-2.22586200e-01 5.33950508e-01 1.88542426e-01 -8.63939077e-02
-3.43268365e-01 -2.80359924e-01 2.49869421e-01 1.66191840e+00
1.60226628e-01 4.18041110e-01 -8.29390958e-02 1.17167759e+00
2.09797382e-01 2.52023935e-01 -4.82924879e-01 -7.59436712e-02
7.24581897e-01 1.34180307e+00 -1.00886869e+00 -2.74982721e-01
-3.59865511e-03 3.63199949e-01 2.35608816e-01 1.54079825e-01
-4.20676738e-01 -3.24223101e-01 1.00853455e+00 4.50143784e-01
-4.88768667e-01 -1.25786752e-01 -1.20421517e+00 -7.01996744e-01
-5.78456044e-01 -9.55150843e-01 1.02172065e+00 -6.94192588e-01
-1.00612354e+00 4.83932532e-02 3.41725886e-01 -1.01569581e+00
-1.21113271e-01 -4.00232077e-01 -8.31050098e-01 9.69656587e-01
-7.86151409e-01 -6.73191905e-01 -2.99209565e-01 2.84290701e-01
1.16205126e-01 1.33680940e-01 9.31408703e-01 5.43961041e-02
-7.61173368e-01 6.09847426e-01 -5.50710484e-02 1.32314963e-02
6.98019445e-01 -1.47268856e+00 8.91158804e-02 7.54461348e-01
-3.39871645e-03 6.82338119e-01 9.19690728e-01 -8.32074821e-01
-4.96804655e-01 -7.58706033e-01 6.26674950e-01 -3.52212578e-01
3.08851361e-01 -3.47934842e-01 -9.41244900e-01 2.41325140e-01
2.40250565e-02 -2.55601078e-01 1.66006780e+00 3.90636235e-01
-5.38583875e-01 1.99747458e-01 -1.40889692e+00 8.40435505e-01
1.12869740e+00 -1.39559627e-01 -5.10701954e-01 -1.29094705e-01
3.18958908e-02 -5.09720929e-02 -6.96890056e-01 2.18624994e-01
8.97980809e-01 -1.35225892e+00 8.42919350e-01 -1.08036387e+00
5.05463898e-01 2.64258403e-02 4.32323158e-01 -1.33623219e+00
-4.19077128e-01 -4.46832329e-01 2.01598436e-01 1.14748287e+00
4.28390592e-01 -4.43796575e-01 7.55667508e-01 1.16800010e+00
3.05527210e-01 -7.06801295e-01 -6.29569650e-01 -3.65580320e-01
2.14082934e-02 -3.76385212e-01 8.08638275e-01 1.59427404e+00
6.29100621e-01 5.88034913e-02 -8.85239542e-02 -1.08120851e-01
6.59085095e-01 4.75230336e-01 4.65535223e-01 -1.95118225e+00
2.56162614e-01 -7.46748388e-01 -3.98552060e-01 -4.68020290e-02
-1.19539797e-01 -9.70073462e-01 -8.83757234e-01 -1.55428624e+00
4.34392393e-01 -8.27320993e-01 -9.46461037e-02 8.18587124e-01
-3.49596709e-01 1.91397607e-01 4.87490892e-01 6.13818876e-02
-3.00402850e-01 -2.02088822e-02 9.28810358e-01 -2.44946536e-02
-3.49653214e-01 -1.70624435e-01 -1.21339262e+00 9.09889758e-01
8.45255017e-01 -6.89875066e-01 -4.43554848e-01 1.78909989e-03
4.59025294e-01 -2.56956369e-01 3.62950772e-01 -4.94683892e-01
-9.14078727e-02 -6.91836953e-01 9.13183272e-01 -3.56795549e-01
-2.26396248e-01 -5.08379519e-01 1.97727755e-01 8.11914980e-01
-6.92392170e-01 2.50658661e-01 5.27004838e-01 2.48785913e-01
2.59953946e-01 -2.17469390e-02 5.90498447e-01 2.08274752e-01
-3.41648310e-01 -1.15130171e-01 -5.83606422e-01 2.73991702e-03
1.06532037e+00 -3.68845880e-01 -6.87883973e-01 -2.94498473e-01
-8.34254801e-01 4.54698056e-01 3.72816622e-01 3.58714104e-01
5.30159771e-01 -1.28944683e+00 -5.59829652e-01 2.65201122e-01
7.82616064e-02 -3.93568188e-01 1.93454683e-01 4.10050064e-01
-5.78307867e-01 1.93738431e-01 -7.35469878e-01 -2.91241616e-01
-2.04293513e+00 4.74632531e-01 1.61904469e-02 -2.06952721e-01
-3.24483305e-01 4.76638794e-01 2.47759387e-01 9.33919773e-02
-1.80015992e-02 -2.86280274e-01 -7.66077161e-01 5.04572093e-01
7.93006301e-01 1.10009778e+00 -5.77830942e-03 -8.27481300e-02
-5.66438317e-01 -6.89503318e-03 3.25863548e-02 -6.87800869e-02
1.13179195e+00 7.52692297e-02 5.43291494e-02 4.14410830e-01
6.20692790e-01 2.99939245e-01 -1.19693398e+00 3.95770192e-01
3.00061077e-01 -1.24463701e+00 -3.02633524e-01 -1.38133955e+00
-5.85442722e-01 9.80344594e-01 7.36204922e-01 6.42409027e-01
1.05490196e+00 -3.61207247e-01 -3.55120838e-01 -8.02979898e-03
-8.84228870e-02 -9.13651407e-01 -2.66988903e-01 -9.97546837e-02
8.71348083e-01 -1.22063696e+00 4.87848401e-01 -5.91881752e-01
-6.52427137e-01 1.33596885e+00 5.03320694e-01 2.58176595e-01
4.61155862e-01 4.93631512e-01 2.19707802e-01 -3.09459060e-01
-6.55261159e-01 1.56718314e-01 4.11694050e-01 5.32415330e-01
6.02869391e-01 5.64075887e-01 -8.34679425e-01 7.40956426e-01
-5.40337205e-01 -6.56053051e-02 4.34937239e-01 8.56299043e-01
-4.51226771e-01 -1.06175780e+00 -8.38699341e-01 8.65561008e-01
-4.63914156e-01 -3.90893891e-02 -8.17633152e-01 7.78898716e-01
4.37441677e-01 9.53979671e-01 4.70523745e-01 -3.00464392e-01
1.29866421e-01 -2.26529706e-02 2.74571329e-01 -5.91853440e-01
-6.77302420e-01 -3.61184001e-01 2.73240834e-01 -4.80569899e-01
-8.64120722e-02 -5.89894950e-01 -1.26082957e+00 -7.15758681e-01
6.77311048e-02 3.86253178e-01 7.06485808e-01 7.24013865e-01
5.33907950e-01 4.82610822e-01 7.86860585e-02 -4.06039566e-01
-5.31721339e-02 -6.10479951e-01 -5.06365836e-01 3.98006469e-01
2.26416796e-01 -8.01835597e-01 -1.92641512e-01 -1.95649028e-01] | [8.807538986206055, 5.47583532333374] |
15384655-6bfa-48bf-a84b-a0cedacf561a | learning-to-speak-fluently-in-a-foreign | 1907.04448 | null | https://arxiv.org/abs/1907.04448v2 | https://arxiv.org/pdf/1907.04448v2.pdf | Learning to Speak Fluently in a Foreign Language: Multilingual Speech Synthesis and Cross-Language Voice Cloning | We present a multispeaker, multilingual text-to-speech (TTS) synthesis model based on Tacotron that is able to produce high quality speech in multiple languages. Moreover, the model is able to transfer voices across languages, e.g. synthesize fluent Spanish speech using an English speaker's voice, without training on any bilingual or parallel examples. Such transfer works across distantly related languages, e.g. English and Mandarin. Critical to achieving this result are: 1. using a phonemic input representation to encourage sharing of model capacity across languages, and 2. incorporating an adversarial loss term to encourage the model to disentangle its representation of speaker identity (which is perfectly correlated with language in the training data) from the speech content. Further scaling up the model by training on multiple speakers of each language, and incorporating an autoencoding input to help stabilize attention during training, results in a model which can be used to consistently synthesize intelligible speech for training speakers in all languages seen during training, and in native or foreign accents. | ['Bhuvana Ramabhadran', 'RJ Skerry-Ryan', 'Ye Jia', 'Ron J. Weiss', 'Andrew Rosenberg', 'Yonghui Wu', 'Zhifeng Chen', 'Yu Zhang', 'Heiga Zen'] | 2019-07-09 | null | null | null | null | ['voice-cloning'] | ['speech'] | [ 8.62506330e-02 3.47354650e-01 -7.02318177e-03 -1.93099484e-01
-1.02474046e+00 -1.00802469e+00 6.60617948e-01 -5.69840431e-01
-2.25875333e-01 6.42428219e-01 6.29859984e-01 -7.67367601e-01
5.74082494e-01 -4.34482217e-01 -7.80033886e-01 -4.61549550e-01
2.27190226e-01 5.24085104e-01 -1.50732681e-01 -3.54177803e-01
-5.07558465e-01 4.68462199e-01 -1.13593972e+00 3.83508921e-01
7.72918403e-01 4.10339355e-01 5.19662380e-01 9.61276412e-01
-8.98777470e-02 9.78528142e-01 -8.40020120e-01 -4.72712010e-01
3.95966977e-01 -7.99932241e-01 -8.06150258e-01 1.09651662e-01
5.78232348e-01 -2.42430583e-01 -4.05289531e-01 7.51932323e-01
5.76431572e-01 8.67110416e-02 6.67224348e-01 -8.61344635e-01
-8.16242456e-01 9.46985364e-01 -9.88409296e-02 1.11482590e-01
2.86604822e-01 3.61264139e-01 9.44130480e-01 -1.01636243e+00
5.87300122e-01 1.56658018e+00 4.26700026e-01 9.86396194e-01
-1.42839956e+00 -7.93926477e-01 1.04055062e-01 -2.69450694e-01
-1.26616657e+00 -1.07522476e+00 4.59801883e-01 -2.79040575e-01
9.13268745e-01 4.55079079e-01 4.52072412e-01 1.49163055e+00
7.06432983e-02 7.02776372e-01 1.24412346e+00 -6.48615599e-01
-1.26375929e-01 4.95029300e-01 -5.62697947e-01 3.73798043e-01
-5.28471589e-01 4.41382706e-01 -6.71356082e-01 2.21062779e-01
5.82131505e-01 -7.01558173e-01 -4.96625125e-01 1.81018505e-02
-1.25500965e+00 5.74588835e-01 2.25955293e-01 4.60877508e-01
-3.03841770e-01 -2.88932770e-03 4.25101519e-01 7.65113652e-01
3.14017057e-01 5.90323448e-01 -3.53770316e-01 3.56294513e-02
-9.39735293e-01 2.81486288e-02 7.07830191e-01 1.09796858e+00
4.46065634e-01 7.62771130e-01 -1.28217489e-01 1.11359179e+00
2.27368608e-01 9.73606527e-01 7.85369396e-01 -9.67427075e-01
7.13605046e-01 -3.02891612e-01 -8.05386752e-02 -2.45943636e-01
7.13400990e-02 -4.66413498e-01 -5.38512588e-01 4.29132134e-01
3.89591843e-01 -4.62527782e-01 -9.88656759e-01 2.30225396e+00
2.76748594e-02 6.41992539e-02 7.42106616e-01 7.19102681e-01
5.54337144e-01 1.00633419e+00 5.00415154e-02 -6.00900427e-02
1.27077913e+00 -1.09441876e+00 -7.47931123e-01 -5.16012013e-01
3.47253472e-01 -1.10922873e+00 1.35427523e+00 1.56885192e-01
-1.40659380e+00 -8.07354510e-01 -1.03177989e+00 -1.82207137e-01
-2.60130584e-01 1.69852730e-02 7.65600502e-02 7.87697077e-01
-1.31763196e+00 2.45387137e-01 -7.83909678e-01 -2.14995861e-01
-4.83427793e-02 3.33966434e-01 -5.11870682e-01 -4.98954505e-02
-1.34210801e+00 1.27524817e+00 3.32543820e-01 -1.59299046e-01
-1.21297777e+00 -6.11945629e-01 -9.45225179e-01 3.79631668e-02
-2.61013895e-01 -7.04006195e-01 1.50900674e+00 -1.61963201e+00
-1.92633986e+00 6.47431910e-01 -2.26801395e-01 -4.12063867e-01
5.18802702e-01 -7.54619762e-03 -8.17136586e-01 -2.96856891e-02
6.72789440e-02 8.43750715e-01 9.53341126e-01 -1.33827555e+00
-5.52289605e-01 -6.23976486e-03 -3.11480194e-01 7.16912210e-01
-3.47671509e-01 2.78819114e-01 -2.48541310e-01 -8.29101205e-01
-2.14829579e-01 -1.06826830e+00 1.59916818e-01 -3.11916232e-01
-5.14856100e-01 1.62480980e-01 5.50476551e-01 -1.19574106e+00
7.75411129e-01 -2.18783450e+00 4.01055187e-01 1.07216582e-01
-3.22556585e-01 2.68262088e-01 -4.99131650e-01 4.84564126e-01
-4.05240357e-02 6.71456084e-02 -5.21465614e-02 -6.90743566e-01
-2.03560397e-01 2.76536047e-01 -5.86560071e-01 2.22022623e-01
3.73604268e-01 8.02083671e-01 -7.38631427e-01 -4.97898944e-02
1.31199947e-02 5.74040294e-01 -5.34341872e-01 4.12617981e-01
-1.08738840e-01 8.49997818e-01 5.64582311e-02 5.38078547e-02
1.47981763e-01 3.64467502e-01 1.55023947e-01 1.67360187e-01
-1.02093831e-01 8.21907699e-01 -8.83816302e-01 1.46764886e+00
-1.17608345e+00 6.96122706e-01 5.90945065e-01 -3.87586206e-01
7.63455272e-01 9.49522495e-01 -1.10798731e-01 -5.52473485e-01
1.50674628e-02 5.82394242e-01 4.64760065e-01 -9.69956890e-02
3.61795306e-01 -6.65757120e-01 6.34845495e-02 5.60824811e-01
3.45141351e-01 -5.70985794e-01 -1.85643300e-01 3.09718177e-02
6.29159272e-01 -2.29331791e-01 -1.48342028e-01 -4.21426386e-01
5.13259351e-01 -2.77558953e-01 3.24668676e-01 5.56143343e-01
1.47894874e-01 6.37201548e-01 -1.07895367e-01 9.54014286e-02
-1.24962223e+00 -1.34933424e+00 2.77747419e-02 1.22045445e+00
-3.48361105e-01 -7.05874041e-02 -7.71142006e-01 -3.37919652e-01
-2.26146370e-01 1.27590346e+00 -2.62183398e-01 -3.03914428e-01
-8.15100014e-01 -7.61933327e-02 8.81808102e-01 4.45981860e-01
7.83445090e-02 -1.31001782e+00 9.19044390e-02 4.39622939e-01
-3.02176476e-01 -1.08809185e+00 -9.97736990e-01 3.54491562e-01
-5.16140580e-01 -4.32275653e-01 -9.34938788e-01 -1.17060423e+00
3.55358452e-01 -9.74321514e-02 1.03553891e+00 -3.13159227e-01
3.39928299e-01 2.82762170e-01 -4.07865271e-02 -3.56849909e-01
-1.46715522e+00 2.28399560e-01 4.30843115e-01 8.13103393e-02
-1.76489875e-01 -5.64731419e-01 3.23820189e-02 2.10624725e-01
-7.37480640e-01 2.62139469e-01 3.15172195e-01 8.80628943e-01
1.20912604e-01 -1.68492302e-01 7.75865614e-01 -6.82659924e-01
6.79303646e-01 -4.35832232e-01 -4.05806094e-01 2.44219139e-01
-2.05375418e-01 9.62174237e-02 1.19754827e+00 -7.42573559e-01
-1.00242400e+00 -1.85755551e-01 -5.43012261e-01 -5.20865321e-01
-2.96370685e-02 2.01774031e-01 -3.38754386e-01 3.73982310e-01
7.51250982e-01 3.80195677e-01 1.62815794e-01 -4.17700499e-01
6.77277505e-01 1.02725661e+00 7.97252893e-01 -5.63516319e-01
9.36832845e-01 -1.25499547e-01 -5.85772872e-01 -9.69475925e-01
-4.59988356e-01 -1.02627296e-02 -4.28292036e-01 6.35494739e-02
8.41768205e-01 -1.28497994e+00 -3.29911143e-01 6.44726932e-01
-1.28393316e+00 -6.90074027e-01 -4.07397121e-01 8.22498679e-01
-6.87476814e-01 -3.23403776e-01 -8.15366983e-01 -7.01962829e-01
-1.97524816e-01 -1.37085664e+00 7.84114718e-01 -1.21469107e-02
-3.67838979e-01 -1.17692661e+00 9.82652418e-03 3.36082995e-01
7.22481310e-01 -4.78130281e-01 9.12262261e-01 -8.67120862e-01
-4.40787047e-01 2.55971849e-01 4.05516595e-01 1.02082932e+00
5.67353368e-01 -1.94696203e-01 -1.16709483e+00 -5.48973143e-01
1.76578552e-01 -5.96475899e-01 4.61809933e-01 5.94284795e-02
4.28262174e-01 -6.38026059e-01 2.77834181e-02 6.01610124e-01
7.72919118e-01 3.14146370e-01 3.20676744e-01 -2.40128726e-01
7.59536803e-01 7.71539807e-01 -2.09415004e-01 -4.44983780e-01
4.58701223e-01 6.30822837e-01 -1.30638003e-01 -2.95796067e-01
-6.89271092e-01 -5.77548623e-01 9.33810651e-01 1.53496468e+00
4.26293761e-01 -3.33297521e-01 -8.04829121e-01 6.67494893e-01
-1.04040480e+00 -1.04426563e+00 3.13546568e-01 2.23921371e+00
1.26804221e+00 7.04561397e-02 8.46741498e-02 -1.20434739e-01
7.87068188e-01 9.38457102e-02 -5.18473864e-01 -8.31736326e-01
-2.72550553e-01 4.65388507e-01 3.80872816e-01 1.21986866e+00
-7.83430755e-01 1.42350185e+00 6.99163675e+00 7.03262269e-01
-1.56184888e+00 1.66658968e-01 5.26671588e-01 -1.06758118e-01
-6.97672546e-01 -1.38600022e-01 -6.22834742e-01 3.38465989e-01
1.38496923e+00 -5.35684168e-01 1.14353967e+00 4.02824074e-01
2.51680493e-01 5.37908196e-01 -1.15982461e+00 4.72479343e-01
1.55028671e-01 -9.38762069e-01 1.56375393e-01 -1.44525245e-01
8.65796208e-01 2.99560755e-01 2.77702719e-01 4.39506561e-01
7.56443977e-01 -1.42383099e+00 1.17861676e+00 1.23586379e-01
1.17912090e+00 -7.20255494e-01 2.04886153e-01 5.43646812e-01
-9.01084423e-01 2.02142388e-01 5.89200994e-03 2.07431316e-01
1.80166915e-01 -4.44954298e-02 -1.17285001e+00 4.24502313e-01
2.17064261e-01 2.51928449e-01 -3.15569162e-01 3.57040912e-01
-3.60307425e-01 9.43402708e-01 -3.42747301e-01 2.71044016e-01
1.83035016e-01 3.36057134e-02 6.66191578e-01 1.06737888e+00
5.55568933e-01 -2.52519816e-01 2.47008115e-01 7.54691243e-01
-2.71732002e-01 2.41275862e-01 -8.39460254e-01 -9.70505774e-02
6.46585464e-01 6.92194104e-01 -4.22445275e-02 -2.41017282e-01
-4.69444811e-01 1.27617216e+00 2.93680102e-01 5.85717857e-01
-3.69976759e-01 -1.61743909e-01 6.35508597e-01 1.21862702e-01
2.25975618e-01 -2.81870037e-01 -2.64950544e-02 -9.62158680e-01
-1.56464681e-01 -1.53587532e+00 -2.44246095e-01 -9.05927837e-01
-1.29130697e+00 1.24423003e+00 -3.49042118e-01 -9.65043604e-01
-8.06307673e-01 -4.89118546e-01 -5.32505035e-01 1.83937526e+00
-1.37014771e+00 -1.36759210e+00 6.00520074e-01 7.37266004e-01
5.67442238e-01 -6.38736963e-01 1.00262845e+00 3.44302177e-01
-2.35253289e-01 8.57940912e-01 4.46342528e-02 1.99494362e-01
8.74477208e-01 -1.28822124e+00 7.90591419e-01 8.57718170e-01
5.51887155e-01 8.43594193e-01 8.01574826e-01 -3.35786939e-01
-1.03593194e+00 -1.17080402e+00 1.11774623e+00 -4.77878749e-01
6.70326352e-01 -6.06422126e-01 -9.50021088e-01 1.06556094e+00
6.47199571e-01 -2.11668193e-01 6.61632538e-01 9.74221155e-02
-4.86903608e-01 -1.25413075e-01 -8.87152016e-01 1.04937851e+00
8.05005431e-01 -1.12720263e+00 -5.98617673e-01 2.69077182e-01
8.93729925e-01 -5.28577507e-01 -5.96367061e-01 2.86955573e-02
4.53108102e-01 -6.32713258e-01 6.98981404e-01 -7.45428979e-01
1.94809556e-01 -2.56724238e-01 -2.60392845e-01 -1.98221457e+00
-1.62981704e-01 -9.67737019e-01 3.66084993e-01 1.49320304e+00
7.97210991e-01 -7.17985630e-01 9.09583643e-02 1.97524220e-01
-4.48319256e-01 -2.10242376e-01 -1.12977457e+00 -1.06992853e+00
8.00826788e-01 -3.61562639e-01 6.28317118e-01 9.08275843e-01
-2.00865120e-01 7.83160508e-01 -5.87790012e-01 3.68953019e-01
-1.83375850e-02 -2.23101810e-01 6.78324640e-01 -6.74313903e-01
-5.33746541e-01 -4.67126161e-01 1.69406086e-01 -1.05927765e+00
6.03324056e-01 -1.38946986e+00 2.88375497e-01 -1.19379759e+00
-4.08498794e-01 -6.28943026e-01 5.49548678e-03 5.06489158e-01
-1.97732002e-01 1.02167211e-01 4.85543936e-01 1.23108052e-01
3.05259675e-01 6.69852257e-01 1.48652387e+00 -1.43806897e-02
-1.47871688e-01 1.16096921e-01 -8.51662397e-01 4.80688006e-01
8.08021903e-01 -5.32087207e-01 -6.14353538e-01 -6.60806835e-01
-4.78056103e-01 3.88470203e-01 9.30070132e-02 -1.02385485e+00
1.27869561e-01 -7.44614825e-02 1.50924504e-01 1.59732595e-01
5.05107045e-01 -8.25354397e-01 3.23710442e-01 3.39162588e-01
-5.92340589e-01 2.10181311e-01 5.14662206e-01 -8.67668353e-03
-3.36185157e-01 -2.54291296e-01 9.99051154e-01 -1.01950392e-01
-1.03887558e-01 2.18229983e-02 -6.04166210e-01 2.96770692e-01
5.87527812e-01 1.15858227e-01 -1.68553874e-01 -7.36136377e-01
-7.45111704e-01 8.75205621e-02 6.14251792e-01 8.58292878e-01
3.30252014e-02 -1.52991259e+00 -1.13265860e+00 6.80586696e-01
-9.75465626e-02 -4.69547451e-01 -3.22531015e-02 1.87676921e-01
-3.43273252e-01 4.19261426e-01 -1.24205045e-01 -3.06926191e-01
-1.04339123e+00 4.42503065e-01 6.04024291e-01 -1.25190765e-01
-3.02443802e-01 8.22834373e-01 2.40512908e-01 -8.84201109e-01
1.55107975e-01 2.98285820e-02 1.79154813e-01 -1.45490214e-01
4.24081832e-01 -1.75670996e-01 4.35554646e-02 -9.88293648e-01
-2.92315990e-01 2.52984494e-01 3.11419312e-02 -1.00841522e+00
8.68428886e-01 -2.23337904e-01 2.55716503e-01 7.98657238e-01
1.22591794e+00 8.71935010e-01 -1.25040114e+00 -2.10173219e-01
-4.11219686e-01 -3.08171026e-02 1.99185163e-02 -1.07786560e+00
-9.50544953e-01 9.65241015e-01 2.65073568e-01 1.66076794e-01
9.19957578e-01 -9.35220942e-02 7.64664531e-01 3.89383920e-02
2.56114423e-01 -7.95805275e-01 -1.44627094e-01 7.88006067e-01
1.16258979e+00 -8.68322432e-01 -6.49532735e-01 -8.28069150e-02
-1.03114367e+00 9.17656898e-01 3.05831283e-01 1.02440216e-01
4.70140874e-01 5.28221071e-01 7.40924299e-01 3.97835910e-01
-8.44814181e-01 6.80118650e-02 4.87906098e-01 6.32432878e-01
6.69129610e-01 2.49498531e-01 1.79332942e-01 2.12365016e-01
-8.85393858e-01 -5.33006728e-01 4.20802563e-01 3.90595824e-01
-1.28811046e-01 -1.49160433e+00 -5.31776905e-01 9.26192477e-02
-5.97518981e-01 -5.63573718e-01 -2.53674984e-01 4.91593659e-01
2.05811858e-02 1.18925178e+00 1.83779776e-01 -1.78591296e-01
2.48476192e-01 5.44244051e-01 3.85622650e-01 -9.06570077e-01
-9.46821868e-01 1.27514780e-01 2.40257680e-01 -1.55325541e-02
3.15567218e-02 -7.25537539e-01 -9.30553079e-01 -2.97198892e-01
-1.06372572e-01 2.37598091e-01 6.76872551e-01 8.10016394e-01
9.99598503e-02 6.11645699e-01 9.15984988e-01 -6.31634712e-01
-7.19218194e-01 -1.00538182e+00 -5.61583996e-01 2.69398153e-01
7.92535782e-01 -3.43178436e-02 -4.83948290e-01 2.74256110e-01] | [14.729727745056152, 6.8204216957092285] |
5e78ba77-3193-433d-9688-a3573e1164f5 | hybrid-approach-to-identify-druglikeness | 2208.06362 | null | https://arxiv.org/abs/2208.06362v3 | https://arxiv.org/pdf/2208.06362v3.pdf | Hybrid Approach to Identify Druglikeness Leading Compounds against COVID-19 3CL Protease | SARS-COV-2 is a positive single-strand RNA-based macromolecule that has caused the death of more than 6.3 million people since June 2022. Moreover, by disturbing global supply chains through lockdown, the virus has indirectly caused devastating damage to the global economy. It is vital to design and develop drugs for this virus and its various variants. In this paper, we developed an in-silico study-based hybrid framework to repurpose existing therapeutic agents in finding drug-like bioactive molecules that would cure Covid-19. We employed the Lipinski rules on the retrieved molecules from the ChEMBL database and found 133 drug-likeness bioactive molecules against SARS coronavirus 3CL Protease. Based on standard IC50, the dataset was divided into three classes active, inactive, and intermediate. Our comparative analysis demonstrated that the proposed Extra Tree Regressor (ETR) based QSAR model has improved prediction results related to the bioactivity of chemical compounds as compared to Gradient Boosting, XGBoost, Support Vector, Decision Tree, and Random Forest based regressor models. ADMET analysis is carried out to identify thirteen bioactive molecules with ChEMBL IDs 187460, 190743, 222234, 222628, 222735, 222769, 222840, 222893, 225515, 358279, 363535, 365134 and 426898. These molecules are highly suitable drug candidates for SARS-COV-2 3CL Protease. In the next step, the efficacy of bioactive molecules is computed in terms of binding affinity using molecular docking and then shortlisted six bioactive molecules with ChEMBL IDs 187460, 222769, 225515, 358279, 363535, and 365134. These molecules can be suitable drug candidates for SARS-COV-2. It is anticipated that the pharmacologist/drug manufacturer would further investigate these six molecules to find suitable drug candidates for SARS-COV-2. They can adopt these promising compounds for their downstream drug development stages. | ['Abdul Majid', 'Imra Aqeel'] | 2022-08-03 | null | null | null | null | ['molecular-docking'] | ['medical'] | [ 1.37480184e-01 -4.78805840e-01 -3.05214792e-01 9.42913350e-03
-3.30330878e-01 -6.65234268e-01 2.69050092e-01 5.57626605e-01
-1.67977855e-01 1.54828310e+00 4.37801257e-02 -9.15644467e-01
-1.53339013e-01 -6.70828104e-01 -4.63685900e-01 -7.72465587e-01
-3.82412106e-01 3.59258503e-01 -9.87285003e-02 -3.18569034e-01
3.66284221e-01 8.59324574e-01 -7.46456683e-01 3.37507091e-02
1.40137529e+00 5.43600678e-01 2.26451054e-01 3.74328405e-01
2.58651078e-01 5.29844046e-01 -5.24095237e-01 -1.75896510e-01
1.20766342e-01 -6.16728485e-01 -3.17823917e-01 -5.82081378e-01
-2.07653254e-01 -1.33978322e-01 5.84482074e-01 6.77295089e-01
4.32861775e-01 -1.39713719e-01 1.07977378e+00 -8.25989544e-01
-3.02804083e-01 -1.95200905e-01 -5.69858015e-01 -2.85473149e-02
6.01261139e-01 9.88879427e-02 6.93984151e-01 -1.06836736e+00
9.27312434e-01 9.83076394e-01 6.22090757e-01 4.28589940e-01
-1.01677108e+00 -1.04116714e+00 -1.40022650e-01 4.04714853e-01
-1.25038683e+00 -2.46575772e-04 3.49345148e-01 -7.32057691e-01
1.40847957e+00 4.90714967e-01 7.84599960e-01 7.53865004e-01
9.76971924e-01 1.54415682e-01 1.19605136e+00 8.21970478e-02
4.03613597e-01 5.74575961e-02 1.91948712e-01 8.54125023e-01
6.67825162e-01 -5.58055099e-03 -4.23142284e-01 -7.63676345e-01
3.46496075e-01 3.51783812e-01 -4.20570999e-01 -1.98683128e-01
-5.80959857e-01 1.11388981e+00 4.61292267e-01 6.72826096e-02
-6.91782236e-01 -4.05220658e-01 1.80349812e-01 2.58087330e-02
2.06815258e-01 5.35404205e-01 -1.02761114e+00 2.70018071e-01
-3.40131164e-01 3.95837873e-01 5.02151370e-01 3.01673144e-01
7.22980142e-01 -1.41937528e-02 2.79270768e-01 3.18025142e-01
7.90186346e-01 6.93604112e-01 -4.99000289e-02 -7.70079568e-02
7.01107681e-02 6.95314109e-01 4.04541552e-01 -1.01933026e+00
-8.43950629e-01 -1.94891632e-01 -7.18017578e-01 -6.03729151e-02
4.74310741e-02 -2.35717624e-01 -8.32780719e-01 1.55032933e+00
4.29799050e-01 -4.95680645e-02 1.95509747e-01 7.77639210e-01
8.40400457e-01 1.08466542e+00 6.01447225e-01 -1.07838464e+00
1.44744515e+00 -5.50579011e-01 -4.23191667e-01 5.04089177e-01
3.23507935e-01 -8.95027280e-01 5.23744643e-01 5.75900316e-01
-4.06501621e-01 -1.34690240e-01 -1.23817313e+00 6.68894291e-01
-3.28164905e-01 -2.11998969e-01 9.52824056e-01 7.61354864e-01
-5.02774000e-01 5.37718356e-01 -8.62276852e-01 -4.27556485e-01
2.22932443e-01 5.92153907e-01 -3.18375349e-01 -3.61689515e-02
-1.25525033e+00 1.09834957e+00 1.43379226e-01 -1.92110445e-02
-8.84936273e-01 -6.83499634e-01 -4.38945860e-01 -3.06365103e-01
-1.02735825e-01 -8.13398302e-01 4.18246776e-01 -3.88317555e-01
-1.44129467e+00 2.33642071e-01 -2.50069559e-01 -4.01787877e-01
-6.48943037e-02 -2.17519458e-02 -6.04212105e-01 1.51233912e-01
1.18191317e-01 2.79664636e-01 1.97636977e-01 -1.02505016e+00
-3.70346338e-01 -8.85736883e-01 -2.69357234e-01 2.42574424e-01
2.88234502e-01 4.58614707e-01 3.20430905e-01 -5.50656319e-01
-1.33603543e-01 -9.36943650e-01 -4.22817558e-01 -7.24247396e-01
-6.61213875e-01 -1.37619331e-01 5.24993479e-01 -5.90847850e-01
9.15927649e-01 -1.37563586e+00 -7.07773045e-02 4.71951455e-01
-4.53392044e-02 4.92670357e-01 -1.38125252e-02 1.07682836e+00
-7.36007839e-02 3.41962606e-01 -9.37042236e-02 8.45928431e-01
-8.03166270e-01 -3.83129120e-01 -2.37803698e-01 6.64685190e-01
9.95781496e-02 6.20757163e-01 -7.74292111e-01 5.80362231e-03
3.98359857e-02 6.31833017e-01 -4.55118179e-01 1.66137800e-01
-4.56617355e-01 8.30847859e-01 -1.09405649e+00 1.03152156e+00
8.53807330e-01 -2.97421291e-02 5.14038205e-01 -1.91234231e-01
-2.09831774e-01 1.39711633e-01 -4.62126076e-01 1.09182334e+00
1.51176974e-02 7.56640360e-02 -3.72449547e-01 -5.16223490e-01
1.06329262e+00 3.91290963e-01 7.78358221e-01 -5.67177236e-01
9.24698040e-02 2.79918313e-01 1.50867611e-01 -3.87058198e-01
-4.20963652e-02 -5.31459033e-01 2.99932063e-01 -1.53137237e-01
-4.42983478e-01 2.80306429e-01 8.37846026e-02 -9.22919810e-03
1.12953019e+00 2.40114510e-01 7.21911073e-01 -3.68479580e-01
8.16863894e-01 4.04651821e-01 7.52137542e-01 8.20695907e-02
-1.01669550e-01 -1.73383951e-01 2.21549049e-01 -6.47636354e-01
-5.49491584e-01 -9.13767695e-01 -5.60702860e-01 6.68271601e-01
1.95650190e-01 -4.68738616e-01 -4.90436703e-01 -3.29428196e-01
-1.94356382e-01 5.01300454e-01 -2.52013534e-01 9.89232659e-02
-3.23461622e-01 -1.00642300e+00 3.51844490e-01 -3.38458061e-01
2.90475637e-01 -9.59282100e-01 -3.11845273e-01 6.04869545e-01
1.60778821e-01 -2.95800358e-01 7.52887800e-02 3.25752527e-01
-1.00425720e+00 -1.56844258e+00 -4.37850803e-01 -4.60486263e-01
3.70254308e-01 1.46721080e-01 4.35921639e-01 1.26753435e-01
-1.87259972e-01 -6.37601972e-01 -4.01650906e-01 -8.23013902e-01
-4.08551365e-01 -3.29508305e-01 4.28499997e-01 -5.15366852e-01
4.23141301e-01 -4.92372066e-01 -9.45050359e-01 2.22735479e-01
-1.86698839e-01 -1.44072115e-01 4.77250129e-01 6.89634264e-01
5.98133922e-01 -3.24710160e-01 1.02359867e+00 -9.86914694e-01
5.28399348e-01 -8.11362028e-01 -6.54870391e-01 2.25954980e-01
-9.44626093e-01 -1.93753779e-01 8.85477483e-01 -8.11651424e-02
-9.88943517e-01 2.03627095e-01 -3.52385402e-01 4.37296003e-01
2.04770844e-02 7.61882126e-01 -3.68506938e-01 4.16901894e-02
6.25101745e-01 1.61553081e-02 -1.49341613e-01 -5.60085297e-01
-6.41368553e-02 6.92647159e-01 -5.62194549e-02 -3.20989817e-01
6.13291562e-01 1.44633487e-01 1.52733609e-01 -8.93409073e-01
-2.06782147e-01 -5.01893640e-01 2.69134849e-01 2.62666140e-02
1.06762147e+00 -9.84857202e-01 -1.26903832e+00 2.41765290e-01
-1.24930930e+00 2.58784086e-01 8.07249367e-01 8.31462860e-01
-1.84583992e-01 3.08139682e-01 -6.29067421e-01 -6.80965781e-01
-9.86862123e-01 -1.16134632e+00 3.69591564e-01 2.14827597e-01
-2.67110407e-01 -5.65788209e-01 5.82055151e-01 6.15971029e-01
1.91493005e-01 6.49665177e-01 1.47109163e+00 -1.05580044e+00
-6.66845322e-01 6.49131238e-02 1.63287844e-03 -6.31560236e-02
2.65139341e-01 3.29773933e-01 -3.99647683e-01 -2.88192064e-01
-1.69413626e-01 -7.07446858e-02 5.46044767e-01 5.88806868e-01
3.55188876e-01 -4.89750445e-01 -5.46059787e-01 6.50606334e-01
1.63700056e+00 1.39795041e+00 8.31104100e-01 4.77523237e-01
5.32644689e-01 2.85474300e-01 9.71591771e-01 5.21488667e-01
7.34348670e-02 5.52645922e-01 6.07791722e-01 -1.36306524e-01
6.89503551e-01 -1.73942998e-01 4.06774342e-01 3.43742967e-01
-6.49285316e-01 -3.28781962e-01 -8.50694597e-01 2.81046815e-02
-1.31235874e+00 -7.94376314e-01 -6.14466190e-01 2.08056283e+00
1.09344435e+00 -3.08754563e-01 8.65204036e-02 -3.50724757e-01
3.12072039e-01 -3.25849444e-01 -6.13267720e-01 -8.10097873e-01
2.87228040e-02 4.26752955e-01 5.37756264e-01 3.93565118e-01
-9.02751148e-01 9.37598348e-01 4.60822105e+00 7.53759384e-01
-1.08822274e+00 -4.15966988e-01 6.53350294e-01 4.18602586e-01
-1.98125154e-01 5.37651241e-01 -7.31954277e-01 3.73385787e-01
1.03207052e+00 -1.65299460e-01 2.37402767e-01 6.22070312e-01
7.58514941e-01 -2.82259285e-01 -7.67252803e-01 5.27715921e-01
-3.11530828e-01 -1.62912273e+00 1.47705361e-01 2.35271484e-01
9.06935871e-01 1.82935707e-02 -3.41596514e-01 -1.30678937e-01
1.65712476e-01 -1.17412007e+00 6.21058978e-03 5.06005049e-01
5.42286575e-01 -1.04030085e+00 8.65198910e-01 2.29144096e-01
-7.82749951e-01 2.17858136e-01 -3.86158019e-01 1.54771328e-01
4.55155037e-02 4.39706832e-01 -1.25668311e+00 6.89772725e-01
3.29981714e-01 3.92023891e-01 -2.36651778e-01 9.06239331e-01
-2.84831136e-01 5.18592715e-01 5.33797368e-02 -4.28157717e-01
2.32802462e-02 -5.50726175e-01 3.99229437e-01 8.33127081e-01
-1.18135093e-02 4.80553538e-01 1.62380099e-01 3.58341455e-01
1.80482581e-01 7.86426246e-01 -5.48348665e-01 -1.19124122e-01
5.16923547e-01 8.81842434e-01 -6.20561957e-01 -2.25385502e-01
-2.50273675e-01 5.64854324e-01 -4.41187263e-01 5.04908681e-01
-7.11041331e-01 -3.54529947e-01 9.88765180e-01 1.27928957e-01
1.95603341e-01 1.33331329e-01 -1.99471917e-02 -7.63071716e-01
-6.87932968e-01 -1.25807405e+00 2.05718130e-01 -4.55328286e-01
-1.08854067e+00 6.18951261e-01 -1.13506071e-01 -9.83581603e-01
-3.82827059e-03 -4.59451914e-01 -4.10769850e-01 1.06495178e+00
-1.00254142e+00 -9.57979560e-01 2.30905965e-01 3.46929461e-01
2.47643083e-01 -4.25554723e-01 1.15682733e+00 -1.60349652e-01
-4.73639220e-01 -7.78354108e-02 5.97833037e-01 -5.93825877e-01
6.20431960e-01 -6.64409816e-01 -1.59020141e-01 1.51277289e-01
-5.42282164e-01 1.31168926e+00 8.56269002e-01 -1.13720930e+00
-1.45853686e+00 -1.01763487e+00 9.12219405e-01 -7.52179846e-02
6.00586176e-01 -7.67473578e-02 -5.47979951e-01 1.92724720e-01
1.53506607e-01 -6.70932055e-01 1.23104751e+00 -1.69503540e-01
-1.50817141e-01 3.63398679e-02 -1.29369390e+00 5.81180990e-01
3.72100949e-01 -4.16807495e-02 -1.98097363e-01 6.43149495e-01
6.81490958e-01 -6.47530258e-02 -1.05075085e+00 6.32327855e-01
8.16667378e-01 -8.75236154e-01 9.66582716e-01 -9.72905517e-01
2.03559846e-01 -6.91516995e-01 3.42585221e-02 -1.01164913e+00
-2.23484755e-01 -5.83466589e-01 3.02262247e-01 6.53883815e-01
7.50647306e-01 -8.04585755e-01 6.88302755e-01 1.50603771e-01
7.65083507e-02 -1.13853037e+00 -8.35743785e-01 -7.54927278e-01
7.04872385e-02 1.40679598e-01 4.56633806e-01 9.26555872e-01
-2.80844439e-02 9.14538145e-01 -5.71429908e-01 -7.76506364e-02
2.83531994e-01 3.49819005e-01 6.28585696e-01 -1.30649912e+00
-1.82389170e-01 6.40814775e-04 -1.62587658e-01 -4.86373395e-01
-3.55803490e-01 -5.65657496e-01 -5.59421837e-01 -1.66011882e+00
5.71810126e-01 -4.89575893e-01 -2.86305666e-01 3.09939891e-01
4.65332158e-02 -3.00666243e-02 -6.97675794e-02 2.92768478e-01
-5.33673279e-02 2.93656021e-01 1.19476509e+00 -1.04484968e-01
-6.32092714e-01 2.21445128e-01 -8.22633445e-01 5.91620445e-01
9.04830217e-01 -5.23674905e-01 -5.11209249e-01 6.25815511e-01
4.84321803e-01 3.51097763e-01 -1.05266415e-01 -6.52102232e-02
-3.01867247e-01 -8.15672398e-01 6.07525945e-01 -1.00252330e+00
1.74309269e-01 -8.02711546e-01 8.53105485e-01 1.05917490e+00
3.43657225e-01 -8.02916214e-02 1.00080922e-01 5.45919001e-01
1.94915593e-01 -1.65513843e-01 5.67102075e-01 1.29507124e-01
-2.31355101e-01 2.54645020e-01 -9.10510719e-01 -3.72892380e-01
1.15829432e+00 -5.14304101e-01 -3.15006852e-01 6.84101731e-02
-6.79260850e-01 -1.20501392e-01 4.75041926e-01 1.77478895e-01
6.72893703e-01 -8.41021061e-01 -6.89428747e-01 -1.00712717e-01
1.90797821e-01 -4.68681931e-01 1.69543341e-01 7.82150567e-01
-1.11698914e+00 9.29403186e-01 -3.86637658e-01 -2.43544847e-01
-1.56626010e+00 7.99850941e-01 1.45440847e-01 -3.49611670e-01
5.59410192e-02 6.29934549e-01 3.06562841e-01 5.49676567e-02
-8.34015906e-02 -1.74641311e-01 -6.07568204e-01 1.31647080e-01
2.90651858e-01 5.17085195e-01 2.28174463e-01 -7.33909428e-01
-1.14543402e+00 5.09411752e-01 -1.20641507e-01 7.97082901e-01
1.53712499e+00 4.33756709e-01 -2.86346942e-01 -1.87648833e-01
1.21102083e+00 4.03130680e-01 -5.17169237e-01 3.90462399e-01
1.28774181e-01 -2.35782340e-01 -5.50478458e-01 -1.30994928e+00
-3.30687940e-01 2.14076594e-01 8.27072740e-01 -4.32240099e-01
1.04282999e+00 -1.11377046e-01 5.97428977e-01 3.76476526e-01
6.89257920e-01 -5.37076712e-01 -2.07138419e-01 3.24195087e-01
9.65106606e-01 -9.78491187e-01 3.59411895e-01 -4.43833470e-01
-5.00846744e-01 1.02050912e+00 2.19721049e-01 1.08609445e-01
3.75953794e-01 -2.11378574e-01 -1.19558759e-01 -4.85838622e-01
-1.01825607e+00 3.40307832e-01 2.45164111e-01 6.49223685e-01
8.44314516e-01 3.23110700e-01 -1.31252623e+00 6.61914527e-01
1.26054004e-01 -1.16912603e-01 4.44588810e-01 7.74697423e-01
-6.08554959e-01 -1.50675535e+00 -4.34843749e-01 3.16891402e-01
-6.75621510e-01 -5.52775025e-01 -7.50197291e-01 7.31042802e-01
2.70559609e-01 1.10434020e+00 -6.05592191e-01 -2.94972599e-01
1.69621885e-01 -1.04797691e-01 3.31644088e-01 -4.13559943e-01
-7.44415164e-01 6.81033254e-01 3.95127505e-01 -5.64070744e-03
-4.02795136e-01 -3.83228838e-01 -1.79101276e+00 -3.29105258e-01
-7.74811149e-01 5.49735308e-01 8.37329924e-01 7.64339924e-01
4.56999570e-01 -4.74564433e-02 9.67106283e-01 -1.64559111e-01
-1.67163879e-01 -9.26913798e-01 -5.19495666e-01 -1.68753684e-01
4.26119715e-02 -5.39358377e-01 2.90514790e-02 1.13382325e-01] | [4.672873497009277, 5.146955966949463] |
bf7ba012-88c3-457f-bd8f-d3772d9f06f3 | clustering-without-knowing-how-to-application | 2209.10267 | null | https://arxiv.org/abs/2209.10267v3 | https://arxiv.org/pdf/2209.10267v3.pdf | Clustering Without Knowing How To: Application and Evaluation | Crowdsourcing allows running simple human intelligence tasks on a large crowd of workers, enabling solving problems for which it is difficult to formulate an algorithm or train a machine learning model in reasonable time. One of such problems is data clustering by an under-specified criterion that is simple for humans, but difficult for machines. In this demonstration paper, we build a crowdsourced system for image clustering and release its code under a free license at https://github.com/Toloka/crowdclustering. Our experiments on two different image datasets, dresses from Zalando's FEIDEGGER and shoes from the Toloka Shoes Dataset, confirm that one can yield meaningful clusters with no machine learning algorithms purely with crowdsourcing. | ['Dmitry Ustalov', 'Daniil Fedulov', 'Daniil Likhobaba'] | 2022-09-21 | null | null | null | null | ['image-clustering'] | ['computer-vision'] | [-1.49426237e-01 8.80711600e-02 2.78297424e-01 -3.15723568e-01
-3.92501980e-01 -8.05724621e-01 3.74422282e-01 1.24906406e-01
-5.80770016e-01 6.42108560e-01 -1.19769603e-01 -1.39509022e-01
1.49196967e-01 -4.46841359e-01 -6.65228188e-01 -5.79974413e-01
3.62735987e-01 1.02523303e+00 5.99472463e-01 -5.88136435e-01
4.93735746e-02 3.03893685e-01 -1.73057580e+00 3.13447624e-01
7.71725953e-01 3.94955069e-01 2.81344742e-01 1.01138306e+00
5.78697696e-02 9.67536628e-01 -6.96084857e-01 -7.12514281e-01
7.99264312e-01 -3.41162980e-01 -9.26812291e-01 5.22392273e-01
3.06234151e-01 2.22334312e-03 -9.94414464e-02 9.46914554e-01
5.30262589e-01 1.83997780e-01 3.69192630e-01 -1.60248876e+00
-7.97889531e-01 2.52946228e-01 -3.70422333e-01 -2.53797323e-01
6.70432627e-01 5.36349714e-01 6.11749828e-01 -5.86816788e-01
7.53307164e-01 1.14318514e+00 6.96707606e-01 5.91740668e-01
-9.58288729e-01 -2.56579161e-01 -1.87901437e-01 -3.25131267e-02
-1.56255603e+00 -3.64298552e-01 4.00049597e-01 -8.19562912e-01
5.00088155e-01 4.17018145e-01 5.04765749e-01 6.38164103e-01
-2.80147374e-01 7.63873398e-01 1.31473601e+00 -5.16925156e-01
5.92037320e-01 2.12726340e-01 3.48077677e-02 7.61168957e-01
2.67167479e-01 -3.94386172e-01 -4.30960625e-01 -4.06815052e-01
5.48781991e-01 2.15529203e-01 -1.57014374e-02 -2.86026001e-01
-1.14496183e+00 8.77189815e-01 3.22605789e-01 6.87732548e-02
-2.12511778e-01 1.59400418e-01 2.95659512e-01 2.48231694e-01
5.67835927e-01 4.44339395e-01 -2.82950103e-01 4.95685451e-02
-6.04888916e-01 6.71761334e-01 9.88813937e-01 1.08674407e+00
1.07864118e+00 -5.09752035e-01 1.53755084e-01 6.90620840e-01
4.98989560e-02 6.92932725e-01 3.23813587e-01 -1.28899312e+00
2.51386583e-01 7.97783315e-01 7.75438368e-01 -1.11210835e+00
-4.41246539e-01 6.80652738e-01 -4.21768636e-01 4.38374400e-01
9.76129889e-01 -4.91053879e-01 -6.27599716e-01 9.18382049e-01
7.36595452e-01 -2.01001406e-01 -3.51535797e-01 1.34105253e+00
8.88934076e-01 1.47965372e-01 -9.88731086e-02 2.12838352e-01
1.43488681e+00 -1.13632286e+00 -5.24727285e-01 -1.77500516e-01
8.11480522e-01 -8.62098992e-01 1.36709690e+00 5.72414875e-01
-1.03588986e+00 -5.35094202e-01 -6.11665130e-01 -4.09964532e-01
-5.31229019e-01 1.96176350e-01 8.82184684e-01 7.33820558e-01
-1.11222696e+00 5.45283437e-01 -6.53775275e-01 -8.10519338e-01
3.19334656e-01 6.33682728e-01 -5.37481010e-01 -1.05215013e-01
-8.01073730e-01 7.70831406e-01 1.97279498e-01 4.61726449e-02
-8.11594784e-01 -4.93410788e-03 -7.16270387e-01 -6.01685286e-01
5.47801912e-01 -5.49658716e-01 1.34611809e+00 -1.17060626e+00
-1.32990241e+00 1.36935461e+00 -2.91269243e-01 -3.92000675e-01
8.59513521e-01 -9.22423154e-02 -1.55224398e-01 2.83402801e-01
3.74671340e-01 5.92627585e-01 7.87900627e-01 -1.50967920e+00
-6.93192184e-01 -4.49492276e-01 2.54494816e-01 8.51463079e-02
-1.76884905e-02 4.38906044e-01 -4.84353244e-01 -1.94837719e-01
-3.43322754e-01 -1.46938646e+00 -6.62463009e-01 1.41330464e-02
-4.30955470e-01 -4.14380312e-01 2.24117726e-01 -4.51030970e-01
7.23339140e-01 -2.00435543e+00 -1.64516255e-01 2.20776461e-02
4.81067598e-01 3.86525840e-01 -3.50245759e-02 4.92904574e-01
3.55915308e-01 1.92879856e-01 -9.23517942e-02 -3.06778044e-01
1.71517923e-01 3.10770512e-01 1.52747929e-01 7.43503630e-01
6.59617856e-02 9.44241583e-01 -1.25259471e+00 -5.29965162e-01
1.63284868e-01 -1.01078048e-01 -2.47724429e-01 1.02310248e-01
-1.86002508e-01 4.94653046e-01 -1.73746213e-01 6.16136134e-01
6.70655966e-01 -4.63958055e-01 2.96287030e-01 4.37280029e-01
5.88569902e-02 -3.92058045e-01 -1.50861871e+00 1.50223374e+00
-4.03626636e-02 5.97375274e-01 2.89697260e-01 -4.20675457e-01
7.89557219e-01 2.15097159e-01 3.31738025e-01 -1.94272935e-01
1.86902821e-01 2.37268850e-01 -3.14894527e-01 -9.72514570e-01
7.34507501e-01 -5.75345643e-02 -2.29260802e-01 6.97955430e-01
-1.77195564e-01 -3.03491384e-01 4.13787752e-01 1.51652038e-01
1.26590490e+00 -1.04078293e-01 1.37934178e-01 -3.02009374e-01
8.85877237e-02 8.37293029e-01 3.34677547e-01 8.66347671e-01
-4.70472753e-01 8.99862528e-01 1.45844743e-01 -8.82507145e-01
-1.22693610e+00 -8.15458596e-01 3.22928280e-01 1.53012061e+00
4.79356110e-01 -2.32622966e-01 -1.20502806e+00 -5.84832311e-01
2.25534067e-02 -7.09782988e-02 -6.48484707e-01 4.60695624e-01
-3.66853923e-01 -4.36584681e-01 6.36956453e-01 1.63880885e-01
3.22512895e-01 -9.85334933e-01 -5.32981813e-01 -1.37504444e-01
-2.62457132e-01 -1.21430814e+00 -4.30910677e-01 -2.92918682e-01
-2.64179170e-01 -1.29360271e+00 -8.15773845e-01 -8.49023819e-01
8.48520935e-01 8.68755221e-01 1.27791321e+00 6.24069452e-01
-5.79044998e-01 3.83089036e-01 -6.22372866e-01 -8.29127729e-01
-2.27369338e-01 6.35615960e-02 3.02124053e-01 -1.95453484e-02
9.37676370e-01 -1.01108760e-01 -5.58216214e-01 7.88840055e-01
-9.17264283e-01 -2.12870955e-01 -2.29830705e-02 3.86581242e-01
1.93873405e-01 6.33844808e-02 3.05609405e-01 -9.02642369e-01
7.38477707e-01 -5.23235977e-01 -5.42917430e-01 2.23516867e-01
-1.28482923e-01 -4.61403400e-01 5.89755774e-01 -3.10213298e-01
-6.54405713e-01 6.12050772e-01 3.07574093e-01 -1.88162446e-01
-7.71801412e-01 -2.95385092e-01 1.13556229e-01 -9.65520218e-02
1.23844206e+00 -2.61583358e-01 1.11803040e-01 -2.91169912e-01
5.35520911e-01 1.02980506e+00 4.96378899e-01 -6.73945010e-01
8.07169676e-01 1.05091012e+00 -5.07467568e-01 -9.31209445e-01
-7.15129912e-01 -8.93566787e-01 -9.42848086e-01 -2.93554902e-01
1.19582593e+00 -1.15103507e+00 -1.15129292e+00 4.35550898e-01
-1.14356172e+00 -6.45059586e-01 -2.64665484e-01 1.23825386e-01
-3.42558742e-01 5.27588546e-01 -4.86058116e-01 -1.03279209e+00
3.79190184e-02 -9.26386952e-01 1.24733627e+00 3.27488691e-01
-5.13359189e-01 -8.65625560e-01 2.08034739e-02 1.03881693e+00
3.59979854e-03 3.98382694e-01 -1.41392544e-01 -7.96553552e-01
-3.76897246e-01 -4.50283706e-01 6.10206649e-02 1.56890169e-01
-1.54057905e-01 2.38775000e-01 -1.12428725e+00 -6.02863841e-02
-2.00187653e-01 -6.64048851e-01 3.61758590e-01 3.20074037e-02
8.46093714e-01 -2.63329118e-01 -1.31744832e-01 7.58849904e-02
1.21801364e+00 -1.94608584e-01 5.59851825e-01 2.74230212e-01
7.66197383e-01 8.78084958e-01 6.99394882e-01 6.51145518e-01
6.82921350e-01 4.72042322e-01 8.51734057e-02 -2.53243536e-01
2.12262541e-01 -6.81727901e-02 1.22408286e-01 4.98568743e-01
-6.72130704e-01 -1.08994819e-01 -1.18634558e+00 8.07127237e-01
-2.41250992e+00 -7.83260703e-01 -7.44397223e-01 1.97094643e+00
6.89456999e-01 -4.36279595e-01 6.04348242e-01 -2.19515637e-02
1.07325256e+00 -3.16584349e-01 -5.89600466e-02 -2.80790299e-01
-3.89719419e-02 -1.03721321e-01 5.25544405e-01 6.24748409e-01
-1.11989391e+00 1.12039971e+00 6.24717236e+00 7.49834299e-01
-5.21015227e-01 3.59728783e-01 3.62702906e-01 -1.49039492e-01
-1.17358817e-02 1.09489590e-01 -6.08621180e-01 6.67696297e-01
6.75994575e-01 3.16117890e-02 8.42979550e-01 8.74426484e-01
4.89217490e-01 -5.51850796e-01 -9.12084281e-01 9.47640836e-01
3.76735665e-02 -1.23270547e+00 -5.11112869e-01 1.04097530e-01
9.19595301e-01 6.49570897e-02 -4.73687947e-01 4.92380820e-02
8.30730319e-01 -1.17335570e+00 7.40690231e-01 4.43181217e-01
2.91309714e-01 -5.76442838e-01 7.41442323e-01 7.24086583e-01
-7.10341394e-01 -1.14981860e-01 -6.78095818e-01 -5.97005844e-01
7.12090433e-02 5.90500295e-01 -1.11224747e+00 1.80615515e-01
8.24899137e-01 4.27634940e-02 -9.03280199e-01 9.57689643e-01
-4.10751641e-01 4.23744708e-01 -3.25962812e-01 -4.78284031e-01
2.30573304e-02 -2.95427948e-01 2.45664403e-01 1.23578846e+00
-1.24173559e-01 3.94987524e-01 5.81894040e-01 5.81256688e-01
-6.78644404e-02 9.90434885e-02 -8.75576079e-01 2.37281397e-01
4.44068968e-01 1.44215941e+00 -9.77138162e-01 -4.32905704e-01
-1.48579419e-01 1.23290896e+00 3.44598889e-01 3.31208050e-01
-7.09468305e-01 -3.41726601e-01 3.75889182e-01 4.48863238e-01
-1.79373726e-01 -3.15057427e-01 -4.68865842e-01 -1.17659235e+00
2.33520105e-01 -1.07479143e+00 1.43286794e-01 -1.03208172e+00
-1.35986137e+00 4.83506829e-01 -2.32545152e-01 -1.05586517e+00
-1.19674385e-01 -7.38113284e-01 -4.55865502e-01 6.26143515e-01
-9.81646836e-01 -1.27040184e+00 -6.53374434e-01 9.35828567e-01
2.26939112e-01 -1.87781423e-01 5.90362370e-01 1.00888677e-01
-3.82403016e-01 1.59837753e-01 -4.35794741e-02 3.82358193e-01
8.62820625e-01 -1.37885368e+00 3.82410109e-01 5.98725855e-01
7.92511776e-02 6.91936731e-01 9.15363193e-01 -6.84937894e-01
-1.32867372e+00 -8.89986932e-01 9.83935058e-01 -1.17978930e+00
6.51491523e-01 -8.11565578e-01 -6.60837114e-01 4.81634080e-01
3.38435322e-01 3.29648435e-01 8.84338558e-01 -1.46065857e-02
-7.42950141e-02 2.96339393e-01 -1.27842033e+00 5.80598414e-01
1.11882401e+00 -6.03646219e-01 -3.71086717e-01 1.25856173e+00
4.92691278e-01 -2.40460485e-01 -8.74463141e-01 -3.17667335e-01
1.84950069e-01 -1.02672505e+00 5.80691576e-01 -7.14282632e-01
4.78815049e-01 -9.47029412e-01 -1.29905507e-01 -1.05022061e+00
-1.37936801e-01 -1.03029430e+00 4.48144495e-01 1.13374305e+00
3.99857521e-01 -4.51530367e-01 7.67961740e-01 1.25095236e+00
2.88300544e-01 -3.89653258e-02 -5.08324862e-01 -9.66470718e-01
-3.04400865e-02 -3.93505991e-01 6.02590084e-01 1.17029595e+00
1.94084764e-01 2.33861506e-01 -3.57531756e-01 1.42453626e-01
4.90760416e-01 -2.45253602e-03 1.59304023e+00 -1.22829580e+00
-2.39277899e-01 1.02627367e-01 -4.59652096e-01 -4.62465346e-01
2.11587861e-01 -6.97395146e-01 4.26824689e-01 -1.70103240e+00
2.15995714e-01 -4.93495733e-01 6.98803961e-01 6.93341553e-01
-3.72284591e-01 6.23622060e-01 4.70688164e-01 5.06824911e-01
-1.36895156e+00 -5.20871207e-02 1.15953124e+00 1.82755757e-02
-9.03742090e-02 -1.00350529e-01 -6.63426995e-01 9.99842346e-01
8.76184225e-01 -5.59929907e-01 -1.21634468e-01 -6.60298347e-01
4.09389317e-01 -3.12712491e-01 6.67738855e-01 -8.89604867e-01
7.20246673e-01 -3.08963001e-01 1.28841355e-01 4.36263643e-02
5.82671463e-02 -6.87776268e-01 2.11309731e-01 2.29880884e-01
-1.33840829e-01 9.93407592e-02 -1.07142091e-01 5.14223576e-01
2.48080213e-02 -5.06227374e-01 5.93686461e-01 -7.23182917e-01
-8.57772410e-01 -1.12109920e-02 -4.80560094e-01 2.41058350e-01
1.37123370e+00 -1.95144325e-01 -5.65299511e-01 -3.03826690e-01
-8.04174960e-01 3.62897366e-01 1.21303988e+00 2.39154264e-01
2.53975213e-01 -1.13830554e+00 -9.21418428e-01 -2.41691217e-01
2.69633889e-01 1.53289393e-01 8.04475471e-02 6.03069663e-01
-1.07962990e+00 -1.53439835e-01 -7.41823483e-03 -6.48193479e-01
-1.34236014e+00 8.24349165e-01 2.86999624e-02 1.44067734e-01
-4.46011335e-01 7.29436994e-01 -1.08764291e-01 -1.01385438e+00
-1.18237108e-01 -7.91016966e-02 2.46041998e-01 -8.73519033e-02
7.62950003e-01 6.58409357e-01 5.49233742e-02 -7.39877701e-01
-3.55523437e-01 3.47468376e-01 5.45663059e-01 -1.64418921e-01
1.05818558e+00 -3.33592445e-01 -4.33969557e-01 3.93891215e-01
7.85735071e-01 9.62342843e-02 -1.01927626e+00 9.74174961e-02
1.66445643e-01 -6.36477113e-01 -5.94075561e-01 -5.14888227e-01
-5.86459100e-01 7.10048079e-01 4.03857231e-01 6.71655715e-01
7.36524880e-01 1.67491570e-01 8.13664258e-01 8.16170216e-01
7.59688437e-01 -1.57535839e+00 -6.98005687e-03 3.49332780e-01
6.81463003e-01 -1.82256305e+00 5.93600199e-02 -4.16592002e-01
-1.26957548e+00 8.40587199e-01 3.84551913e-01 -5.00988841e-01
5.37853599e-01 3.38619262e-01 4.83215839e-01 -4.73005027e-01
-4.56138462e-01 -6.54474616e-01 -3.49091440e-02 8.93057704e-01
2.92870760e-01 4.27518189e-01 -1.84532553e-01 5.73857725e-01
-2.51410604e-01 3.65860820e-01 7.57992506e-01 1.07938361e+00
-6.74512148e-01 -7.88779438e-01 -8.63785982e-01 1.15609519e-01
-4.03460950e-01 2.26916060e-01 -9.86174464e-01 6.70520723e-01
5.88732362e-01 1.68909132e+00 -1.01510271e-01 -4.23042446e-01
3.66792023e-01 -2.31291220e-01 2.07059801e-01 -7.88493097e-01
-1.05491614e+00 -1.54054374e-01 3.14890742e-01 -5.55395722e-01
-7.87337065e-01 -5.90387106e-01 -1.29123604e+00 -5.71707606e-01
-2.14984715e-01 2.28283525e-01 5.94638407e-01 9.75166619e-01
4.22244132e-01 -2.94443816e-01 6.23838365e-01 -1.11222291e+00
-4.92255092e-02 -7.63218105e-01 -8.31315398e-01 9.37504530e-01
-2.99625583e-02 -2.47557357e-01 -1.96575165e-01 5.55019855e-01] | [9.707919120788574, 4.745421409606934] |
1d38958f-6ced-4596-9146-4383ec2dcb4e | reprogramming-audio-driven-talking-face | 2306.16003 | null | https://arxiv.org/abs/2306.16003v1 | https://arxiv.org/pdf/2306.16003v1.pdf | Reprogramming Audio-driven Talking Face Synthesis into Text-driven | In this paper, we propose a method to reprogram pre-trained audio-driven talking face synthesis models to be able to operate with text inputs. As the audio-driven talking face synthesis model takes speech audio as inputs, in order to generate a talking avatar with the desired speech content, speech recording needs to be performed in advance. However, this is burdensome to record audio for every video to be generated. In order to alleviate this problem, we propose a novel method that embeds input text into the learned audio latent space of the pre-trained audio-driven model. To this end, we design a Text-to-Audio Embedding Module (TAEM) which is guided to learn to map a given text input to the audio latent features. Moreover, to model the speaker characteristics lying in the audio features, we propose to inject visual speaker embedding into the TAEM, which is obtained from a single face image. After training, we can synthesize talking face videos with either text or speech audio. | ['Yong Man Ro', 'Se Jin Park', 'Minsu Kim', 'Jeongsoo Choi'] | 2023-06-28 | null | null | null | null | ['face-generation'] | ['computer-vision'] | [ 6.02213860e-01 4.08804655e-01 1.02788381e-01 -5.19370139e-01
-7.02808142e-01 -3.48733664e-01 5.82979441e-01 -7.64460862e-01
1.98459104e-01 3.33207816e-01 4.23184723e-01 3.52215208e-02
2.50007212e-01 -7.39202857e-01 -9.24726605e-01 -8.32457125e-01
4.29652631e-01 2.11785927e-01 -3.50631416e-01 1.04261041e-01
-1.20686457e-01 3.23811471e-01 -1.88293529e+00 6.04858637e-01
4.52984154e-01 9.64420080e-01 3.35482866e-01 7.96052158e-01
-3.61228496e-01 8.07769835e-01 -5.83122015e-01 -2.98258990e-01
-8.56685266e-02 -7.96917498e-01 -4.87082094e-01 5.29566050e-01
3.88376921e-01 -6.16883457e-01 -4.44038689e-01 8.82961333e-01
3.63631189e-01 3.62001024e-02 9.82884109e-01 -1.43752575e+00
-6.71017885e-01 9.41188216e-01 -1.95780799e-01 -3.10376704e-01
5.99169254e-01 1.81124851e-01 7.67892718e-01 -1.14376271e+00
6.57391429e-01 1.61304843e+00 8.36438034e-03 9.79170501e-01
-1.28605080e+00 -8.97120178e-01 1.55427545e-01 6.67832345e-02
-1.38509154e+00 -1.07020473e+00 1.16858327e+00 -5.53930759e-01
3.18611443e-01 1.47553563e-01 6.57158732e-01 1.43040943e+00
-7.28128031e-02 6.99737489e-01 6.34691894e-01 -3.96070212e-01
2.14897677e-01 4.70524371e-01 -6.35198236e-01 5.10103762e-01
-5.88219166e-01 -9.75597203e-02 -8.57903600e-01 5.52519672e-02
7.03370035e-01 -7.73974210e-02 -2.94196188e-01 -4.00090307e-01
-9.20299232e-01 8.27450395e-01 1.60481691e-01 2.42160276e-01
-2.61762887e-01 3.08909059e-01 3.12123418e-01 4.00630921e-01
4.07197088e-01 -2.07891837e-01 2.10194379e-01 5.30682355e-02
-1.01128900e+00 -1.21019684e-01 5.37330449e-01 7.26244450e-01
6.73291087e-01 4.30027336e-01 -1.09147727e-01 7.65305340e-01
7.06408501e-01 5.97597778e-01 5.16497314e-01 -8.77581120e-01
4.07745689e-01 3.90504062e-01 -7.92612582e-02 -1.03180969e+00
1.47006899e-01 6.23144489e-03 -8.74068737e-01 4.75331843e-02
-8.13271776e-02 1.06857404e-01 -8.74450088e-01 1.76002216e+00
4.09562171e-01 4.06755567e-01 3.47555339e-01 9.04293358e-01
9.01371479e-01 1.24645019e+00 -1.64017320e-01 -5.10523438e-01
1.21250641e+00 -8.00658703e-01 -1.03892148e+00 -1.25599587e-02
2.76694477e-01 -7.08904326e-01 1.32191133e+00 3.02061051e-01
-1.08059072e+00 -8.07620943e-01 -1.00864542e+00 2.27027424e-02
-8.12893454e-03 5.91361940e-01 -1.19351543e-01 5.49849689e-01
-9.82684851e-01 -7.47392997e-02 -6.02887332e-01 -1.39076024e-01
-4.50191647e-03 2.90578455e-01 -5.53550184e-01 3.86039317e-02
-1.07850766e+00 2.79783845e-01 3.17318112e-01 2.01690480e-01
-1.59583533e+00 -4.17348891e-01 -1.08814323e+00 3.90015632e-01
3.72152001e-01 -5.46899199e-01 1.19277096e+00 -1.50930202e+00
-2.19909549e+00 6.49571240e-01 -3.97261798e-01 -1.62570462e-01
4.21181321e-01 8.23652968e-02 -4.20859307e-01 4.91702110e-01
-1.79713264e-01 8.35981131e-01 1.74811900e+00 -1.45897031e+00
-1.84620440e-01 -1.70957565e-01 -6.39176443e-02 8.29190984e-02
-7.45444655e-01 8.65478739e-02 -4.86563295e-01 -6.86459959e-01
2.55428813e-03 -8.59869659e-01 3.89241427e-01 7.41739050e-02
-3.32547277e-01 -4.95277159e-02 1.06915855e+00 -5.70866883e-01
8.89086545e-01 -2.61395144e+00 4.08503205e-01 1.17490955e-01
8.63902122e-02 2.61834194e-03 -2.87321448e-01 3.83194745e-01
-1.16276436e-01 -1.12768650e-01 -1.16698153e-01 -7.52791464e-01
-9.03460756e-02 5.26671521e-02 -7.88055718e-01 2.69240141e-01
4.06273186e-01 4.35645521e-01 -5.66259325e-01 -6.39053524e-01
2.44433388e-01 1.06221330e+00 -7.80618131e-01 5.67207634e-01
-3.85157555e-01 7.66936779e-01 -4.14050668e-01 2.85144776e-01
6.15338802e-01 8.07679221e-02 1.77611366e-01 -1.29740641e-01
-3.84449847e-02 2.20664665e-01 -1.13283312e+00 1.65236485e+00
-7.47771144e-01 4.93668139e-01 4.41848606e-01 -9.14499342e-01
1.04572248e+00 9.28599715e-01 4.42135692e-01 -2.74961710e-01
3.46273094e-01 4.24303040e-02 -4.24508899e-01 -6.42086565e-01
1.46112025e-01 -3.17470461e-01 1.32373393e-01 5.29030681e-01
4.43194956e-01 -5.00895828e-02 -2.29237020e-01 7.53312930e-02
5.42153060e-01 2.78561041e-02 -3.02642107e-01 3.10474366e-01
1.07744408e+00 -5.52433074e-01 2.68629164e-01 -1.44312801e-02
2.99074441e-01 5.55875063e-01 5.03758013e-01 -1.17052972e-01
-8.35959792e-01 -1.02189767e+00 1.28300279e-01 1.31952083e+00
-2.51883179e-01 -4.78939652e-01 -1.07009220e+00 -5.26172996e-01
-4.00507540e-01 3.95646870e-01 -5.80220222e-01 -4.28190559e-01
-5.49617469e-01 1.31608650e-01 4.21733320e-01 3.22261184e-01
2.58484274e-01 -1.11364591e+00 -1.67044267e-01 2.37164408e-01
-3.29070747e-01 -1.15867162e+00 -7.38706708e-01 -2.66568214e-01
-5.11001110e-01 -4.59404081e-01 -9.85055149e-01 -1.02025723e+00
7.19016731e-01 6.05134964e-02 4.45020705e-01 -3.66104811e-01
-1.36535212e-01 4.12177801e-01 -3.02078098e-01 -1.55672818e-01
-7.73521125e-01 -4.78371084e-02 2.82690823e-01 9.17321742e-01
2.11232845e-02 -6.69807494e-01 -3.24192643e-01 3.85540456e-01
-1.12727046e+00 3.34642977e-01 4.67989802e-01 8.32297504e-01
3.97936493e-01 -8.25734213e-02 7.10658491e-01 -2.79700518e-01
4.18854058e-01 -2.56705105e-01 -5.61028719e-01 2.38800496e-01
4.27401029e-02 1.05430903e-02 9.86992180e-01 -7.57323384e-01
-1.10994244e+00 6.01009727e-01 -2.56206602e-01 -1.04149687e+00
-1.85975090e-01 3.62719595e-01 -6.96317613e-01 2.37299383e-01
3.54885101e-01 5.49869299e-01 2.93921083e-01 -4.48244691e-01
5.91470659e-01 1.27933407e+00 6.02849483e-01 -3.13278675e-01
8.11183095e-01 4.56813574e-01 -3.25458944e-01 -1.08508015e+00
-3.28759670e-01 -5.56310825e-02 -4.30738568e-01 -4.56575245e-01
1.07521999e+00 -9.75732088e-01 -9.33603168e-01 2.26608142e-01
-1.20219934e+00 -2.11259723e-01 -4.15225066e-02 5.40221334e-01
-9.84195888e-01 1.02805920e-01 -3.88474077e-01 -1.00413156e+00
-2.16360092e-01 -1.36672974e+00 1.22861242e+00 -9.43718031e-02
6.93709357e-03 -5.55062592e-01 7.52876699e-02 3.07492554e-01
2.66585737e-01 -1.09240137e-01 6.87128246e-01 -2.52584189e-01
-6.04118884e-01 4.96243685e-02 3.05538088e-05 4.01599497e-01
4.00305539e-01 3.43629539e-01 -1.43160415e+00 -4.23353016e-01
2.78904498e-01 -4.56754357e-01 7.07864404e-01 2.40114346e-01
1.24364364e+00 -5.97014487e-01 -1.86472654e-01 7.06422389e-01
8.10157180e-01 1.84850693e-01 3.67770404e-01 -2.57067651e-01
5.64731419e-01 8.86065602e-01 4.10370201e-01 4.19962943e-01
2.61958003e-01 8.52777123e-01 3.31744671e-01 1.82030916e-01
-2.08839297e-01 -7.50492930e-01 1.04499459e+00 9.64478433e-01
3.40283334e-01 -2.22270906e-01 -5.37269533e-01 4.68733668e-01
-1.45801079e+00 -9.51005757e-01 5.51594257e-01 1.92899680e+00
8.41477513e-01 -5.59334941e-02 7.29996189e-02 2.93790460e-01
8.24527383e-01 8.29302073e-02 -4.68897283e-01 -3.24027121e-01
2.65657067e-01 -1.05372973e-01 -3.96154076e-01 5.91229916e-01
-7.42999256e-01 1.04255009e+00 5.73940086e+00 7.56668866e-01
-1.76986647e+00 -2.82533877e-02 2.63121217e-01 -2.15653732e-01
-5.74951649e-01 -1.29719704e-01 -7.28388071e-01 5.55779457e-01
1.15991688e+00 -2.67092258e-01 6.30097926e-01 8.65836442e-01
4.73892301e-01 4.56460357e-01 -1.36687303e+00 1.26527107e+00
3.07018161e-01 -1.04318595e+00 6.26938403e-01 1.70408953e-02
3.59557271e-01 -8.10502529e-01 4.40319151e-01 2.69120932e-01
-2.81836927e-01 -1.22738934e+00 7.83614516e-01 4.08422798e-01
1.29378319e+00 -6.98442638e-01 1.68724567e-01 4.36470747e-01
-1.30681777e+00 -1.49312168e-01 -3.52649003e-01 3.37592393e-01
2.26012737e-01 2.20657229e-01 -1.11380839e+00 2.72067755e-01
5.88379860e-01 5.56539178e-01 -6.76771998e-02 4.50916678e-01
-1.47849649e-01 6.77253425e-01 -1.49109364e-01 4.11475986e-01
-3.86892334e-02 -3.94702703e-02 4.49623495e-01 8.25853229e-01
7.18107581e-01 -1.26607880e-01 1.30441517e-01 9.26388323e-01
-3.26401651e-01 2.50685364e-01 -9.08090949e-01 -3.98291796e-01
2.60415345e-01 1.20190597e+00 -3.83580744e-01 -3.41309041e-01
-1.95436105e-01 1.12095892e+00 -1.59939319e-01 3.24842036e-01
-9.15116251e-01 -3.55072320e-01 5.62174678e-01 2.30617151e-01
1.74533829e-01 -6.34781048e-02 3.02526563e-01 -1.14287436e+00
3.25822793e-02 -9.17909682e-01 -3.91015485e-02 -9.94336307e-01
-8.49603117e-01 9.08919692e-01 -2.04019383e-01 -1.39666271e+00
-6.50200069e-01 -3.92263383e-01 -5.79789937e-01 7.81582415e-01
-1.32133090e+00 -1.31219947e+00 -3.61179829e-01 9.99595344e-01
7.64253378e-01 -3.07659686e-01 7.95119166e-01 3.98499280e-01
-3.30980271e-01 8.61210942e-01 -2.93955892e-01 1.23519689e-01
1.00540531e+00 -6.45634055e-01 1.66246459e-01 3.69948596e-01
2.19200552e-01 5.57958603e-01 7.17191815e-01 -4.00680959e-01
-1.44601536e+00 -1.14356279e+00 6.67880952e-01 -3.94221954e-02
4.13187206e-01 -7.32421696e-01 -8.54041278e-01 8.25537443e-01
2.39456326e-01 -1.60808221e-01 7.19462812e-01 -4.51286912e-01
-2.45290846e-01 -3.79644275e-01 -9.93747056e-01 5.74263930e-01
7.13352859e-01 -1.01092088e+00 -5.03719985e-01 1.11340493e-01
9.03072596e-01 -9.62493569e-02 -6.32806838e-01 2.11218968e-01
5.88127494e-01 -4.94587183e-01 7.80072689e-01 -2.86614060e-01
4.78680402e-01 -4.49640155e-01 -1.24287963e-01 -1.22462416e+00
2.93388907e-02 -7.46366739e-01 -5.13243675e-02 1.57747495e+00
2.32310802e-01 -4.29798722e-01 7.65640199e-01 3.71348374e-02
1.03880979e-01 -2.13953763e-01 -1.00696802e+00 -4.95262235e-01
-1.67428359e-01 -1.74198449e-01 8.28257024e-01 7.82802403e-01
1.29438322e-02 4.78953183e-01 -7.54665315e-01 3.72931153e-01
3.61155719e-01 5.18884175e-02 9.88572001e-01 -1.06803322e+00
-2.56865442e-01 4.68956679e-02 -2.94769138e-01 -1.14902043e+00
5.19478023e-01 -8.98971796e-01 1.87566325e-01 -1.22276413e+00
4.06645015e-02 -1.57055482e-01 4.10456881e-02 3.45821202e-01
2.76253134e-01 2.88824379e-01 2.95827478e-01 5.51065095e-02
-8.21536183e-02 1.04166257e+00 1.30205190e+00 -3.58214796e-01
-3.80871803e-01 -8.05037469e-03 -3.87598753e-01 5.35648346e-01
5.82637787e-01 -4.96758968e-01 -7.35684216e-01 -3.67651403e-01
8.41189250e-02 6.98811114e-01 3.87139589e-01 -9.27690983e-01
2.40508690e-01 -8.48425776e-02 3.78704548e-01 -5.13959765e-01
8.32750142e-01 -1.01659632e+00 2.53490210e-01 1.58929542e-01
-7.29626358e-01 -2.93224096e-01 -5.97519502e-02 3.64771217e-01
-5.86457431e-01 -1.04627475e-01 6.64527416e-01 1.72162294e-01
1.58320423e-02 3.19850504e-01 -7.28127599e-01 -5.19611299e-01
1.01396656e+00 -1.53845996e-01 7.42302611e-02 -8.21920514e-01
-1.14285207e+00 1.06242545e-01 3.46996039e-01 6.14005983e-01
8.52653027e-01 -1.52726817e+00 -5.26658833e-01 7.14381218e-01
2.24934265e-01 -1.91911325e-01 6.33152902e-01 5.70394754e-01
-6.31337762e-02 2.93264061e-01 -3.76477331e-01 -7.78714776e-01
-1.43395782e+00 8.68896544e-01 1.53980121e-01 4.12125766e-01
-5.37644148e-01 5.74209929e-01 4.66008693e-01 -4.18743610e-01
5.28662801e-01 -2.10152566e-01 -3.39779794e-01 2.19414666e-01
6.85570002e-01 -1.03967324e-01 -1.77185431e-01 -9.11180377e-01
-6.07928075e-02 8.15998733e-01 1.99621201e-01 -6.10345185e-01
1.08851016e+00 -1.56326622e-01 2.34653521e-02 6.15421116e-01
1.40802765e+00 1.80006698e-01 -1.33517528e+00 -1.28974646e-01
-5.64818144e-01 -3.31130594e-01 -1.33455265e-03 -2.00969711e-01
-1.12924695e+00 1.40355432e+00 6.92448020e-01 4.55601737e-02
1.15273535e+00 7.32767060e-02 7.18238235e-01 3.65246683e-01
1.07796267e-01 -8.58975530e-01 3.79128724e-01 1.46085009e-01
1.07503712e+00 -8.87178600e-01 -7.09927201e-01 -4.90010023e-01
-6.84471309e-01 1.26305401e+00 4.42950964e-01 1.97555780e-01
6.80003226e-01 2.14335129e-01 9.64022055e-02 7.07395896e-02
-8.75583410e-01 8.59338194e-02 2.33755261e-01 3.97165120e-01
1.11853972e-01 -1.68381527e-01 2.20803350e-01 4.71375495e-01
-5.34064353e-01 2.89924234e-01 4.73356277e-01 5.83149672e-01
-4.74486232e-01 -1.07161343e+00 -5.55222034e-01 -1.57661214e-01
-2.94265211e-01 2.61092752e-01 -5.44396996e-01 5.21542467e-02
-5.61761484e-02 1.04302728e+00 1.17817372e-01 -5.99596858e-01
7.19788298e-02 4.99788046e-01 2.74696648e-01 -6.64053917e-01
-1.91872060e-01 3.50091964e-01 -3.44140381e-01 -3.73594165e-01
-3.22832793e-01 -2.86477357e-01 -1.20252633e+00 -5.16177379e-02
-1.29316568e-01 2.89945334e-01 8.25867295e-01 6.93024814e-01
2.47556984e-01 5.82739532e-01 1.05750930e+00 -1.09136391e+00
-3.53238791e-01 -9.96762097e-01 -4.76202846e-01 1.10166028e-01
6.49019659e-01 -5.87874472e-01 -5.65778196e-01 5.83733976e-01] | [13.238214492797852, -0.3803410828113556] |
dab3b074-e71d-4a66-8c9b-179bfa37cd77 | video-restoration-with-a-deep-plug-and-play | 2209.02854 | null | https://arxiv.org/abs/2209.02854v2 | https://arxiv.org/pdf/2209.02854v2.pdf | Video Restoration with a Deep Plug-and-Play Prior | This paper presents a novel method for restoring digital videos via a Deep Plug-and-Play (PnP) approach. Under a Bayesian formalism, the method consists in using a deep convolutional denoising network in place of the proximal operator of the prior in an alternating optimization scheme. We distinguish ourselves from prior PnP work by directly applying that method to restore a digital video from a degraded video observation. This way, a network trained once for denoising can be repurposed for other video restoration tasks. Our experiments in video deblurring, super-resolution, and interpolation of random missing pixels all show a clear benefit to using a network specifically designed for video denoising, as it yields better restoration performance and better temporal stability than a single image network with similar denoising performance using the same PnP formulation. Moreover, our method compares favorably to applying a different state-of-the-art PnP scheme separately on each frame of the sequence. This opens new perspectives in the field of video restoration. | ['Andrés Almansa', 'Matias Tassano', 'Julie Delon', 'Antoine Monod'] | 2022-09-06 | null | null | null | null | ['video-denoising', 'video-restoration'] | ['computer-vision', 'computer-vision'] | [ 4.65544015e-01 3.51104699e-02 4.02702779e-01 -6.68771416e-02
-8.17989349e-01 -1.54233590e-01 7.52322316e-01 -4.11033243e-01
-4.79430616e-01 7.55598068e-01 3.71211976e-01 -1.27666533e-01
-1.35691732e-01 -6.09644294e-01 -1.03909922e+00 -1.16922617e+00
2.15656936e-01 3.42940688e-02 2.52772987e-01 -2.67931521e-01
1.49283946e-01 4.49667275e-01 -1.49133229e+00 2.82992363e-01
7.59114385e-01 8.67576540e-01 5.26798606e-01 6.46669865e-01
4.04870629e-01 9.07779872e-01 -3.93972844e-01 -3.53613406e-01
4.50534552e-01 -5.79605103e-01 -6.09881222e-01 3.17732424e-01
6.95793152e-01 -7.56402612e-01 -8.69434178e-01 1.14596188e+00
5.25777519e-01 3.45625907e-01 2.41435662e-01 -5.16419411e-01
-5.28852165e-01 4.61893380e-01 -3.57722163e-01 2.30364606e-01
5.67185581e-01 4.67811376e-02 5.51286757e-01 -7.11141944e-01
7.15504766e-01 1.16888213e+00 8.00748825e-01 5.71192086e-01
-1.59260833e+00 -1.31630033e-01 -1.44162565e-01 4.76281196e-01
-9.97072995e-01 -7.12508321e-01 6.73968494e-01 -3.04357618e-01
6.96376801e-01 2.67364122e-02 4.20020282e-01 1.23299253e+00
1.72285005e-01 6.45260870e-01 1.00544405e+00 -3.81277889e-01
1.88611418e-01 -3.69489580e-01 -2.06481591e-01 3.39804649e-01
8.15897901e-03 2.69559562e-01 -5.49191236e-01 1.69135779e-01
9.38242495e-01 2.01585386e-02 -7.52035141e-01 -3.40494603e-01
-1.20195115e+00 4.04680401e-01 2.59614974e-01 3.94935489e-01
-6.74617469e-01 3.79273951e-01 4.41590399e-01 3.89878780e-01
6.74599767e-01 -7.95227066e-02 -6.26253784e-02 6.63199723e-02
-1.60154355e+00 3.25777054e-01 7.31221139e-01 2.32953146e-01
6.46193862e-01 3.95306021e-01 -2.40342751e-01 6.59079611e-01
2.08761066e-01 1.60881504e-01 3.11767966e-01 -1.61286128e+00
1.48836151e-01 -3.94762456e-01 2.51083404e-01 -8.26308608e-01
-7.66202360e-02 -4.28861171e-01 -1.10034549e+00 6.88928068e-01
4.69483256e-01 2.42998358e-02 -8.99576724e-01 1.59762633e+00
1.00421198e-01 4.75616157e-01 2.20135838e-01 1.13582551e+00
4.99580234e-01 9.10196304e-01 -2.94023156e-01 -3.59820634e-01
1.15718615e+00 -8.68200839e-01 -8.53480816e-01 -9.81457811e-03
-1.66155562e-01 -7.87660360e-01 4.62315142e-01 9.14366782e-01
-1.35370684e+00 -6.78998232e-01 -1.15467024e+00 -2.68975139e-01
2.35597908e-01 6.30803704e-02 6.54958114e-02 4.44919735e-01
-1.58034551e+00 1.29753292e+00 -9.85763371e-01 -3.77810180e-01
3.91799688e-01 2.14176267e-01 -6.50631249e-01 -3.96862328e-01
-9.38275754e-01 9.40385878e-01 3.05765688e-01 4.42068368e-01
-1.44370365e+00 -5.87929666e-01 -8.19568336e-01 1.84691578e-01
2.92281300e-01 -8.32727611e-01 9.44728255e-01 -1.18870485e+00
-1.93192792e+00 7.21766829e-01 -1.86459407e-01 -6.71656132e-01
8.65295231e-01 -5.23418844e-01 -2.05494538e-01 4.08421069e-01
-1.42501771e-01 3.80569458e-01 1.35937285e+00 -1.38314700e+00
-2.43755892e-01 -1.12009436e-01 1.42935798e-01 -8.31283256e-02
-3.07531119e-03 -7.25061540e-03 -4.53619391e-01 -7.92774260e-01
1.45800829e-01 -4.30814981e-01 -2.64896780e-01 8.70267674e-02
-1.50680557e-01 2.68372566e-01 8.00994933e-01 -1.44480348e+00
8.91176462e-01 -2.21874046e+00 6.66749954e-01 -7.10838586e-02
1.61621094e-01 4.56125647e-01 -3.57617468e-01 5.42013764e-01
-3.88568074e-01 -2.09949568e-01 -6.08336389e-01 -6.85806990e-01
-2.41346911e-01 3.34162533e-01 -2.66384244e-01 7.27016449e-01
-3.49475555e-02 4.86956775e-01 -8.98841381e-01 3.94834168e-02
3.83795083e-01 1.05937910e+00 -6.69290304e-01 2.68938154e-01
-7.35379383e-02 8.03460062e-01 6.23123012e-02 2.33602732e-01
9.84427631e-01 1.12353384e-01 4.27475631e-01 -3.69146109e-01
-2.27215648e-01 8.96812081e-02 -1.11382627e+00 1.96244657e+00
-4.38757330e-01 8.96134496e-01 6.70045912e-01 -1.43606603e+00
7.25043476e-01 6.10046089e-01 4.06826049e-01 -5.33429265e-01
7.25786248e-03 3.21011305e-01 -4.18728173e-01 -7.15334296e-01
5.42450786e-01 -3.70396525e-01 7.16450751e-01 1.12144262e-01
3.58832747e-01 1.22165293e-01 4.07790214e-01 1.06703453e-01
1.27060163e+00 7.30690598e-01 -1.54788634e-02 -3.18087518e-01
6.17116153e-01 -5.73658705e-01 4.43458468e-01 7.37123132e-01
5.16528971e-02 1.11097479e+00 5.93213797e-01 -4.23565745e-01
-1.23856783e+00 -1.02229464e+00 -9.15757343e-02 6.76761270e-01
-1.00985356e-01 -1.57848716e-01 -9.54602897e-01 -2.76498258e-01
-3.65743399e-01 5.35179853e-01 -3.99993062e-01 1.58957839e-01
-8.35404873e-01 -9.09852445e-01 2.56470591e-01 1.75207406e-01
6.28745437e-01 -9.46098089e-01 -3.25084418e-01 3.45307738e-01
-5.21459579e-01 -1.23854017e+00 -5.76757006e-02 1.23435080e-01
-1.07692921e+00 -9.16424990e-01 -1.12213182e+00 -9.00181532e-01
3.96468878e-01 3.53800178e-01 1.02023268e+00 8.41111839e-02
1.48612350e-01 3.64219517e-01 -2.61790842e-01 5.38474321e-01
-7.12597609e-01 -3.32973182e-01 -7.33532384e-03 4.17050868e-01
-2.84424156e-01 -1.17324173e+00 -7.01015174e-01 -1.08127240e-02
-1.31128073e+00 -1.08330272e-01 3.58887076e-01 9.44809020e-01
2.80063510e-01 2.04664409e-01 2.52431810e-01 -4.96046454e-01
5.23364842e-01 -4.44923282e-01 -7.76157439e-01 -4.27731313e-02
-4.41930383e-01 6.22473396e-02 7.01318979e-01 -1.81994170e-01
-1.35007167e+00 1.54619426e-01 -4.49609220e-01 -6.20627344e-01
-1.47536606e-01 3.92894447e-01 -1.58650577e-01 -1.17513984e-01
6.10047400e-01 4.12927866e-01 3.49199057e-01 -1.02502513e+00
4.03941542e-01 3.46155941e-01 1.13901103e+00 -4.70198810e-01
8.12659442e-01 7.80169070e-01 1.02016903e-01 -7.64892280e-01
-4.91472244e-01 -3.06827933e-01 -5.95535815e-01 -2.40234256e-01
8.63749087e-01 -1.05881536e+00 -7.15236366e-01 7.84258962e-01
-1.42959654e+00 -3.51683199e-01 -3.02353084e-01 4.78959799e-01
-6.76043570e-01 1.05568647e+00 -9.98635888e-01 -5.20896912e-01
-9.33003351e-02 -1.21338665e+00 9.27318096e-01 1.46344841e-01
3.49190682e-01 -9.75652754e-01 1.73566774e-01 4.73681092e-01
6.08799338e-01 6.75921366e-02 4.13746983e-01 -1.54396236e-01
-8.07482481e-01 -2.16377035e-01 -2.66949296e-01 1.07366788e+00
-8.18610117e-02 -1.71784446e-01 -1.12083578e+00 -4.48818296e-01
5.01832664e-01 1.68913677e-02 1.44269276e+00 7.58348465e-01
9.10112023e-01 -2.80367851e-01 1.30104914e-01 8.89167726e-01
1.70398891e+00 -2.68557012e-01 1.27571094e+00 4.76166010e-01
5.83848178e-01 4.39100146e-01 2.95891091e-02 4.15514588e-01
7.06376284e-02 7.79737711e-01 7.84782112e-01 -3.87050281e-03
-3.76671672e-01 6.70082644e-02 8.00173640e-01 4.86529320e-01
-5.16538382e-01 -3.78853440e-01 -5.45554280e-01 6.63905323e-01
-1.95665503e+00 -1.16761708e+00 -1.42634273e-01 2.30920935e+00
7.69302905e-01 -2.18176425e-01 -2.79522508e-01 9.68944952e-02
7.51569510e-01 4.65237796e-01 -6.57002479e-02 -6.83993325e-02
-4.76951212e-01 1.95891663e-01 3.13895255e-01 7.64894724e-01
-1.12785137e+00 5.44454038e-01 6.29333496e+00 9.59016979e-01
-9.51129556e-01 3.58066171e-01 4.89282846e-01 2.54302304e-02
-1.00320302e-01 1.84743196e-01 -3.06793630e-01 4.10665214e-01
1.00432634e+00 3.40547204e-01 8.88165712e-01 2.45476156e-01
6.45781636e-01 -2.68149137e-01 -1.00030792e+00 9.02648032e-01
2.11227208e-01 -1.42731631e+00 -3.26376804e-03 -5.72538003e-02
6.96330547e-01 6.41185194e-02 -1.27372503e-01 -2.68548191e-01
-3.89740728e-02 -8.13875437e-01 8.07037592e-01 9.30653095e-01
4.94173139e-01 -3.78944635e-01 8.62277806e-01 1.90384209e-01
-6.65078878e-01 7.93222487e-02 -4.63634908e-01 -1.10004492e-01
6.82247996e-01 1.04373813e+00 -1.95580386e-02 8.06150079e-01
8.39136600e-01 1.01995206e+00 -7.66610578e-02 1.18162966e+00
-4.13538158e-01 6.16090059e-01 -9.16021243e-02 1.04542530e+00
-2.41166800e-02 -6.10134840e-01 9.94142830e-01 1.18252075e+00
5.76812208e-01 -9.09025446e-02 -1.83944896e-01 7.32514858e-01
-1.24569252e-01 -3.41652513e-01 -3.27621192e-01 3.45235914e-01
-2.63958335e-01 1.18632412e+00 -4.84033346e-01 -5.22931576e-01
-3.18582237e-01 1.32093978e+00 -1.03439353e-01 6.49740160e-01
-5.74819088e-01 2.50286274e-02 5.17759025e-01 1.50361955e-01
8.06105852e-01 -1.21589296e-01 2.67623607e-02 -1.50127900e+00
2.90597528e-01 -9.62840438e-01 -5.05042113e-02 -1.06978369e+00
-1.24102724e+00 7.49839783e-01 -1.24704301e-01 -1.30508053e+00
-2.88886487e-01 -5.39961100e-01 -5.25159836e-01 1.03081417e+00
-1.84966564e+00 -8.79101157e-01 -3.73330384e-01 5.92576504e-01
4.79101866e-01 1.37011841e-01 4.95055705e-01 5.08059978e-01
-4.62428927e-01 5.86684700e-03 5.66493571e-01 -1.48695201e-01
7.87150800e-01 -1.03627801e+00 2.85521895e-01 1.53561676e+00
-2.70017803e-01 3.14103425e-01 1.15240765e+00 -5.38085580e-01
-1.16835630e+00 -8.03950548e-01 7.44698286e-01 6.29016617e-03
5.57237685e-01 6.19424060e-02 -1.11515594e+00 6.05943561e-01
8.41301024e-01 2.62829307e-02 -6.02811016e-02 -3.47852618e-01
-2.54288495e-01 -2.97209978e-01 -1.08854508e+00 4.14596796e-01
8.08881700e-01 -5.04357517e-01 -4.70269859e-01 3.60852003e-01
5.27384281e-01 -3.92866731e-01 -8.20033967e-01 2.29155272e-01
4.82280433e-01 -1.38658023e+00 1.26670837e+00 -4.45179939e-02
9.03568864e-01 -5.01964390e-01 -2.26679727e-01 -1.29791152e+00
-2.26783365e-01 -1.05763006e+00 -1.69642597e-01 1.13221920e+00
-1.59005389e-01 -6.03112757e-01 4.21399951e-01 2.28015214e-01
-4.00593936e-01 -2.47154430e-01 -1.04492688e+00 -7.44691312e-01
-1.17969751e-01 -3.13381761e-01 4.06867266e-02 5.44141173e-01
-4.10214067e-01 -1.26038074e-01 -9.95930791e-01 1.99376225e-01
1.11414468e+00 -3.41137022e-01 4.74216312e-01 -9.83758986e-01
-5.35190761e-01 -2.16065556e-01 -3.80625337e-01 -1.37907648e+00
2.05041349e-01 -6.85926080e-01 3.82409513e-01 -1.51725924e+00
1.26785278e-01 2.71254808e-01 -1.46532446e-01 2.60463834e-01
1.98709145e-02 5.34224331e-01 3.25206667e-01 2.62705177e-01
-1.65915012e-01 6.08412325e-01 1.11989391e+00 -2.83600464e-02
1.15151025e-01 -1.45299703e-01 -3.65650594e-01 6.20451152e-01
4.17853922e-01 -4.53069150e-01 2.84412820e-02 -6.16318882e-01
6.77352548e-02 3.88097465e-01 8.69531989e-01 -9.91080940e-01
3.23983967e-01 3.84834558e-01 1.94679141e-01 -1.82392687e-01
4.55189437e-01 -8.04579318e-01 3.99690986e-01 3.64738762e-01
-5.88745736e-02 -2.78268456e-01 6.74840733e-02 7.19792962e-01
-4.52281564e-01 -4.80335921e-01 9.26569998e-01 -2.17481658e-01
-6.88094616e-01 -1.47539839e-01 -6.66188121e-01 -4.53593373e-01
3.06010693e-01 -2.98403770e-01 -2.62865514e-01 -5.20585537e-01
-1.15351331e+00 -3.34525198e-01 5.77481985e-01 3.96436490e-02
5.98019361e-01 -9.91835177e-01 -1.04708612e+00 1.32108495e-01
-6.32446885e-01 -1.90859795e-01 5.82256496e-01 1.23056793e+00
-8.32648695e-01 -1.87927801e-02 -3.74819040e-01 -6.12361848e-01
-1.11555278e+00 5.40661395e-01 6.14733458e-01 -3.36149186e-01
-1.00061870e+00 5.26459336e-01 1.80059433e-01 -4.51487526e-02
1.34333953e-01 -3.49103771e-02 -1.08069822e-01 2.29140297e-02
7.46064901e-01 6.17486238e-01 2.96097755e-01 -5.54297447e-01
-1.06496029e-01 4.84506220e-01 1.91872820e-01 -3.45841289e-01
1.71619904e+00 -4.51111764e-01 -6.02820933e-01 3.68968286e-02
1.01054943e+00 1.40223755e-02 -1.76219726e+00 -2.55345047e-01
-2.82800436e-01 -5.30632496e-01 4.24336463e-01 -6.40208125e-01
-1.46721923e+00 6.08850002e-01 6.42567396e-01 1.55690476e-01
1.53795469e+00 -3.02198052e-01 6.26537502e-01 9.27124992e-02
1.63608506e-01 -7.97004879e-01 -1.85548395e-01 3.39771539e-01
1.11830747e+00 -1.11240757e+00 1.53056815e-01 -2.33828053e-01
-1.91456601e-01 1.45437622e+00 -1.61318973e-01 -5.21052480e-01
6.00121617e-01 1.03583410e-01 -2.85697319e-02 8.66162106e-02
-4.76417840e-01 -7.19549134e-02 6.95770234e-02 6.02560937e-01
3.38935018e-01 -3.10552388e-01 -3.55920345e-01 1.54604524e-01
3.97343874e-01 3.77125621e-01 9.63519871e-01 6.25517964e-01
-2.28810534e-01 -1.20403326e+00 -5.50844789e-01 -1.45911396e-01
-6.38396084e-01 -3.99149537e-01 3.42825532e-01 5.00588357e-01
-8.04303139e-02 1.02789640e+00 -2.12372363e-01 -1.75425798e-01
4.53688726e-02 -4.39143553e-02 5.93764544e-01 -1.54745385e-01
-5.52237988e-01 4.68482733e-01 5.23493141e-02 -9.37607586e-01
-9.47862685e-01 -8.61679971e-01 -6.97291553e-01 -3.94286960e-01
5.08016013e-02 -1.01548150e-01 6.34055376e-01 1.09768879e+00
2.19692424e-01 6.16040349e-01 4.14834410e-01 -1.58953106e+00
-4.06301409e-01 -9.22788262e-01 -6.57281041e-01 4.09972101e-01
7.69667804e-01 -2.58114964e-01 -5.65625429e-01 5.22183537e-01] | [11.535454750061035, -2.335871934890747] |
6b179541-6900-44db-a346-c4f85aa0c0ba | an-eeg-channel-selection-framework-for-driver | 2304.14920 | null | https://arxiv.org/abs/2304.14920v1 | https://arxiv.org/pdf/2304.14920v1.pdf | An EEG Channel Selection Framework for Driver Drowsiness Detection via Interpretability Guidance | Drowsy driving has a crucial influence on driving safety, creating an urgent demand for driver drowsiness detection. Electroencephalogram (EEG) signal can accurately reflect the mental fatigue state and thus has been widely studied in drowsiness monitoring. However, the raw EEG data is inherently noisy and redundant, which is neglected by existing works that just use single-channel EEG data or full-head channel EEG data for model training, resulting in limited performance of driver drowsiness detection. In this paper, we are the first to propose an Interpretability-guided Channel Selection (ICS) framework for the driver drowsiness detection task. Specifically, we design a two-stage training strategy to progressively select the key contributing channels with the guidance of interpretability. We first train a teacher network in the first stage using full-head channel EEG data. Then we apply the class activation mapping (CAM) to the trained teacher model to highlight the high-contributing EEG channels and further propose a channel voting scheme to select the top N contributing EEG channels. Finally, we train a student network with the selected channels of EEG data in the second stage for driver drowsiness detection. Experiments are designed on a public dataset, and the results demonstrate that our method is highly applicable and can significantly improve the performance of cross-subject driver drowsiness detection. | ['Yang Liu', 'Liming Zhai', 'Chenyu Liu', 'Jiaping Xiao', 'Ziyu Jia', 'Dan Lin', 'Xinliang Zhou'] | 2023-04-26 | null | null | null | null | ['eeg', 'eeg'] | ['methodology', 'time-series'] | [ 4.32090819e-01 -1.78818062e-01 1.07264556e-01 -5.48083782e-01
-5.65753996e-01 -2.70315826e-01 4.63536754e-02 -5.20522557e-02
-3.80105942e-01 6.21723652e-01 7.19528587e-04 -3.48965377e-01
-4.97734189e-01 -3.81331325e-01 -4.18146133e-01 -7.85376251e-01
3.78234506e-01 -1.39154896e-01 1.33173153e-01 -2.68139869e-01
3.60293597e-01 4.36203301e-01 -1.80231893e+00 1.36293247e-01
1.33535337e+00 1.21526980e+00 2.31576055e-01 4.70471442e-01
3.69730890e-01 3.67398143e-01 -8.11706662e-01 -1.16671845e-02
-8.97222012e-02 -4.61229086e-01 -2.27736756e-01 8.76379833e-02
-6.86847195e-02 -5.69792986e-02 -4.40728933e-01 1.08580017e+00
9.01511550e-01 9.84458104e-02 2.73951918e-01 -1.62185943e+00
-2.51508296e-01 1.34659171e-01 -3.39471459e-01 8.04384232e-01
1.24445990e-01 3.62780362e-01 4.82639700e-01 -9.32413936e-01
-1.90221518e-01 7.06503153e-01 2.85705000e-01 5.61871231e-01
-6.73927307e-01 -1.39518404e+00 2.78291345e-01 7.06990600e-01
-1.53352427e+00 -5.24133503e-01 9.92599845e-01 -4.36803758e-01
8.55985820e-01 3.11992764e-01 9.66133893e-01 8.24327528e-01
4.71956313e-01 6.64168537e-01 1.14901114e+00 2.70891190e-02
8.98955688e-02 2.62654066e-01 5.82500994e-01 2.46826231e-01
1.66440710e-01 1.94214374e-01 -8.21959436e-01 -4.03756052e-02
6.06552437e-02 2.83542961e-01 -4.77786690e-01 1.80073842e-01
-1.21642244e+00 5.99803746e-01 4.60334331e-01 -1.21845476e-01
-5.45321822e-01 -1.56648666e-01 3.61120433e-01 3.90450299e-01
4.74722594e-01 4.77613091e-01 -2.10642695e-01 -2.29939342e-01
-7.40465522e-01 1.69987038e-01 2.39720985e-01 7.85374045e-01
8.42280149e-01 -1.63948581e-01 -3.33379060e-01 7.19705462e-01
-8.13750178e-03 4.49915290e-01 4.82784569e-01 -4.23405856e-01
3.68515879e-01 7.94780791e-01 -6.77768737e-02 -8.83902788e-01
-4.67580855e-01 -4.89324421e-01 -6.40801966e-01 -7.10958838e-02
-3.99528414e-01 -2.95260370e-01 -7.13691175e-01 1.28753138e+00
3.67155783e-02 4.72416312e-01 6.36705309e-02 1.08194578e+00
8.92110586e-01 5.35684288e-01 2.25634888e-01 -2.53041863e-01
1.58749425e+00 -5.19276738e-01 -8.26530516e-01 -4.51594383e-01
3.85671586e-01 -3.54939163e-01 1.08875823e+00 7.19556749e-01
-7.26218462e-01 -8.95080149e-01 -1.33500063e+00 2.41431653e-01
-3.84466290e-01 3.26252222e-01 5.89631498e-01 5.53945422e-01
-8.26570213e-01 -1.49878740e-01 -6.12262845e-01 7.17248172e-02
6.18880570e-01 5.98984241e-01 -3.63573283e-01 -2.12122396e-01
-1.56770027e+00 1.02156937e+00 5.44381440e-01 5.88447094e-01
-1.06626344e+00 -6.70514286e-01 -7.01627254e-01 -4.42258716e-02
4.10879135e-01 -2.94762671e-01 8.62108529e-01 -6.49330974e-01
-1.21032345e+00 4.71191406e-01 -3.55528295e-01 -2.29888380e-01
1.81442425e-01 1.80031136e-02 -9.40497696e-01 -1.62830979e-01
-1.63840026e-01 8.09466600e-01 9.55931008e-01 -9.92012382e-01
-9.39200103e-01 -5.12810171e-01 -1.02363698e-01 2.33542383e-01
-5.00162721e-01 8.33182260e-02 -2.43786544e-01 -1.90315470e-01
3.88797261e-02 -9.57773566e-01 1.54127374e-01 -6.56158507e-01
-3.93873125e-01 -7.71299005e-01 7.15131581e-01 -6.83667004e-01
1.59679711e+00 -2.45651460e+00 -1.15037732e-01 4.43422914e-01
5.90235412e-01 2.93001980e-01 -1.10115707e-01 -1.51588723e-01
-3.09200794e-01 -4.09637123e-01 -2.59431362e-01 -7.51866847e-02
-4.90559898e-02 9.63705685e-03 -1.33109704e-01 4.31567252e-01
6.80186391e-01 8.77341628e-01 -7.85607159e-01 -1.57877222e-01
2.44528383e-01 5.31139672e-01 -3.01874131e-01 5.41003883e-01
5.74721277e-01 3.97249103e-01 -3.51079166e-01 4.43788081e-01
7.72760332e-01 3.34956437e-01 -4.95490104e-01 -1.67684183e-01
-2.53730774e-01 3.68161529e-01 -9.02834356e-01 1.36940002e+00
-3.69268656e-01 8.62316191e-01 -3.42474580e-01 -9.20110285e-01
1.17338622e+00 2.59565800e-01 7.73363411e-02 -9.65938032e-01
2.69640565e-01 1.24966325e-02 5.92588425e-01 -8.94230962e-01
2.20487580e-01 -1.26677481e-02 -8.43061060e-02 2.48113155e-01
-2.21888840e-01 -2.51387316e-03 -4.01527286e-02 3.89982685e-02
9.66884911e-01 -3.54270607e-01 -1.03403844e-01 -2.54259497e-01
6.57666206e-01 -1.72076479e-01 6.16173446e-01 3.31460059e-01
-4.10761386e-01 2.67018735e-01 5.70619822e-01 -3.65437627e-01
-4.02612954e-01 -6.22062802e-01 -2.61950761e-01 1.06847847e+00
4.52256680e-01 -3.02552849e-01 -8.92240763e-01 -4.86668497e-01
1.78957619e-02 9.33435380e-01 -7.21261442e-01 -9.63493943e-01
-2.69627512e-01 -8.94113541e-01 3.78518850e-01 7.52790034e-01
4.77272421e-01 -1.23497736e+00 -8.87580514e-01 6.09403439e-02
-1.29186124e-01 -9.09488201e-01 -3.51867169e-01 4.01403487e-01
-4.13442075e-01 -1.07957590e+00 -4.43309844e-01 -6.77045465e-01
7.82040477e-01 5.81706285e-01 5.83613634e-01 2.88702488e-01
-2.41325215e-01 -1.16955146e-01 -8.31490830e-02 -1.12853456e+00
6.48333051e-04 2.26275250e-01 9.12672356e-02 3.41623485e-01
1.14791071e+00 -1.36938408e-01 -8.35279942e-01 5.18559813e-01
-7.45645881e-01 7.80287758e-02 9.05830026e-01 7.00431824e-01
3.56776625e-01 3.81046176e-01 1.05775535e+00 -5.68690836e-01
1.15555155e+00 -4.27654922e-01 -5.33181131e-01 9.26176235e-02
-6.45793378e-01 -2.45740980e-01 4.20471162e-01 -3.19575399e-01
-9.45708871e-01 7.47480467e-02 -2.38105491e-01 -2.49913737e-01
-5.72005987e-01 4.70449001e-01 -5.47983408e-01 -1.04305215e-01
4.83675361e-01 4.38729525e-01 -7.03464821e-02 -1.16255358e-01
-2.94564694e-01 1.22809982e+00 6.26570702e-01 8.65087211e-02
5.13819337e-01 -2.30533537e-02 -4.44062471e-01 -6.90752983e-01
-5.37285149e-01 -7.17166722e-01 -6.06591403e-01 -3.13697785e-01
8.66874814e-01 -1.08542049e+00 -9.99798536e-01 3.37159872e-01
-1.00376081e+00 -2.78533492e-02 3.55756432e-01 7.83668220e-01
-5.14372922e-02 -6.29174486e-02 6.34939000e-02 -8.99950743e-01
-2.72237808e-01 -1.47839165e+00 9.02743638e-01 2.09831491e-01
-1.47960588e-01 -3.79803628e-01 -2.92880386e-01 4.73682225e-01
3.28080028e-01 -3.60379368e-02 8.28947127e-01 -5.49842954e-01
-1.57386974e-01 -6.13315761e-01 -3.30673665e-01 4.85881865e-01
8.17767158e-03 -4.90065038e-01 -1.38582742e+00 -2.20780537e-01
7.65119269e-02 -1.55935183e-01 9.31990921e-01 3.76500726e-01
1.59616840e+00 4.10759300e-01 -3.98332179e-01 4.95025873e-01
8.20847154e-01 5.27100980e-01 8.72653365e-01 6.20754659e-02
6.85530126e-01 7.85959601e-01 6.31007910e-01 2.59641677e-01
5.34405887e-01 3.15486044e-01 4.03084815e-01 -4.00819361e-01
2.01226398e-01 -1.04939230e-01 4.47707415e-01 7.88379014e-01
-6.45747408e-02 -2.13909626e-01 -8.14642608e-01 4.75803405e-01
-1.61858726e+00 -5.99582970e-01 -1.24297157e-01 2.10257053e+00
4.63837802e-01 3.82036120e-01 1.22921847e-01 4.95678812e-01
7.12553382e-01 -2.77679324e-01 -8.97996962e-01 -2.87468791e-01
9.20069218e-02 1.33845910e-01 3.35989982e-01 1.25754282e-01
-8.62777054e-01 6.01585627e-01 5.61436939e+00 5.65365493e-01
-1.15150201e+00 2.00801373e-01 4.95473027e-01 -3.12837064e-01
-2.82738686e-01 -2.70565897e-01 -9.03773427e-01 8.33740890e-01
1.07547820e+00 -1.42317533e-01 4.11052912e-01 6.56344712e-01
5.81005871e-01 -1.12356246e-01 -8.94732594e-01 1.35045123e+00
4.11821634e-01 -7.07462668e-01 -3.39855582e-01 7.50337867e-03
3.98330867e-01 -8.19435567e-02 -1.15204742e-02 3.23347837e-01
-5.27028024e-01 -1.23191512e+00 5.28172016e-01 5.58813393e-01
8.58462095e-01 -1.16953909e+00 1.04761481e+00 5.15311897e-01
-1.05417371e+00 -4.34447825e-01 -6.84103429e-01 -3.84140722e-02
-1.23378366e-01 4.32852626e-01 -6.26991928e-01 2.13026270e-01
7.42873847e-01 8.08293045e-01 -8.05381417e-01 1.20925069e+00
-4.10130501e-01 7.60676801e-01 3.78672518e-02 -1.63106620e-01
1.29754499e-01 1.10666074e-01 2.09805414e-01 9.89294946e-01
3.72844785e-01 4.36673671e-01 -6.83620051e-02 7.00341761e-01
3.02320588e-02 -2.07775682e-01 -4.39748853e-01 3.57315838e-01
4.37030047e-01 1.34080052e+00 -4.46967900e-01 -2.69351780e-01
-4.97338295e-01 7.25397110e-01 1.19736874e-02 3.65230352e-01
-9.39471960e-01 -9.71041083e-01 7.73000896e-01 7.52649829e-02
-2.21899807e-01 1.58010006e-01 -2.12732613e-01 -7.63515532e-01
-3.07849851e-02 -7.79786408e-01 3.14620018e-01 -8.58534575e-01
-1.03210771e+00 8.69841039e-01 4.14795466e-02 -1.25494397e+00
2.05704361e-01 -5.32401145e-01 -9.26254570e-01 1.23903167e+00
-2.02951312e+00 -6.45889282e-01 -9.71322834e-01 8.08343589e-01
6.26384854e-01 -1.84219599e-01 4.84775722e-01 5.01843870e-01
-1.03985941e+00 6.84500992e-01 -6.09112263e-01 -1.12146676e-01
8.44113588e-01 -9.08841491e-01 1.71280384e-01 8.24349046e-01
-4.71114039e-01 6.32961094e-01 4.79723334e-01 -3.78346056e-01
-1.41371930e+00 -1.19309449e+00 8.69432569e-01 -4.93540078e-01
2.36355454e-01 -4.98568743e-01 -1.11304820e+00 3.62327754e-01
3.32372129e-01 -1.13924265e-01 9.48029041e-01 -1.84270695e-01
3.07439029e-01 -4.75850433e-01 -7.69235551e-01 1.88170597e-01
9.07104790e-01 -5.25518715e-01 -7.93385446e-01 1.15919508e-01
3.56241316e-01 -1.80557400e-01 -5.91045260e-01 2.85226732e-01
5.57446837e-01 -6.49210453e-01 5.39704561e-01 -5.12077808e-01
-1.67994779e-02 -5.42102098e-01 4.86177176e-01 -1.84407437e+00
-4.90383774e-01 -5.87664366e-01 2.54507095e-01 9.15624082e-01
4.05901283e-01 -9.65200722e-01 3.85445833e-01 7.59229302e-01
-6.82146907e-01 -7.41774499e-01 -6.14969552e-01 -2.13156328e-01
-4.16461915e-01 -7.06049204e-01 1.01366973e+00 5.04231811e-01
2.49761209e-01 3.47501844e-01 -2.37271726e-01 3.15548450e-01
2.59698600e-01 -1.40146956e-01 6.10257030e-01 -1.53911960e+00
3.23298156e-01 -3.51873994e-01 -5.39069712e-01 -8.69897306e-01
4.83816862e-01 -8.92427742e-01 5.36669433e-01 -1.30767441e+00
2.18259811e-01 -2.49400720e-01 -1.02315879e+00 5.17318726e-01
-7.28501618e-01 4.49117631e-01 -3.02113920e-01 -3.55984755e-02
-4.53953475e-01 3.72125268e-01 1.02736330e+00 -1.14654116e-01
-4.48084027e-01 2.09412500e-01 -1.15161085e+00 4.31025594e-01
9.02406037e-01 -4.50158030e-01 -8.53827477e-01 -3.65550697e-01
1.78613961e-01 -1.47698164e-01 5.01753986e-01 -1.03508329e+00
4.43007231e-01 2.70648133e-02 7.33103693e-01 -6.62359536e-01
1.26408473e-01 -6.90645397e-01 -1.45610109e-01 4.43909705e-01
-5.02187192e-01 3.47781867e-01 4.36267406e-01 5.17978132e-01
-3.12355161e-01 1.40783712e-01 5.08953214e-01 4.26319689e-01
-1.05056143e+00 4.54468578e-01 -5.78950346e-01 -2.96350837e-01
1.10916615e+00 -4.12816733e-01 -1.44280255e-01 -1.35482982e-01
-4.25792128e-01 5.62323749e-01 -1.57863528e-01 7.27334797e-01
1.12897336e+00 -1.21053302e+00 -8.42658699e-01 1.01181662e+00
7.06048071e-01 -2.44028807e-01 4.33207750e-01 1.26119661e+00
5.70406690e-02 4.62681502e-01 -3.43563765e-01 -7.15938807e-01
-1.11799812e+00 4.86749738e-01 2.33582541e-01 5.83254993e-01
-4.43812579e-01 8.69579017e-01 2.94675857e-01 4.62667681e-02
1.64410412e-01 -3.61510664e-01 -5.51123500e-01 1.55700430e-01
7.84493029e-01 4.28502768e-01 6.02461636e-01 -6.78236306e-01
-6.95340514e-01 4.27641153e-01 4.67471927e-02 3.85441273e-01
1.01369703e+00 -8.05131122e-02 6.70232698e-02 1.73834324e-01
1.08169484e+00 -4.89311457e-01 -1.25021029e+00 2.09187806e-01
-2.54319012e-01 -4.62864578e-01 5.81299603e-01 -7.19250321e-01
-1.08601463e+00 1.38866138e+00 9.91361141e-01 8.75643715e-02
1.48980701e+00 -1.38069630e-01 9.61870432e-01 3.63340288e-01
-1.85150467e-02 -1.05193889e+00 -2.20604286e-01 -3.91326658e-03
7.15203226e-01 -1.31638074e+00 -2.79840708e-01 -1.73810571e-01
-1.04985416e+00 9.32317495e-01 7.66677082e-01 -8.34214501e-03
7.11856365e-01 1.69070750e-01 1.63795099e-01 -4.39088494e-01
-8.24777305e-01 -2.58292407e-01 6.14853323e-01 5.92341363e-01
2.22662419e-01 1.87067732e-01 -1.87628448e-01 8.47335994e-01
-3.15255344e-01 4.30356003e-02 1.62162587e-01 5.67274988e-01
-5.11537135e-01 -5.95135093e-01 -2.13130295e-01 8.49837601e-01
-2.27869153e-02 -2.41480041e-02 -3.65322620e-01 4.39129680e-01
5.78156173e-01 1.25663984e+00 2.32168242e-01 -7.68584669e-01
8.13442171e-01 3.29778343e-01 2.02392321e-02 -4.84717309e-01
-6.30477428e-01 9.97775421e-03 -3.32641602e-01 -3.85283411e-01
-2.36310288e-01 -3.44506413e-01 -1.20208931e+00 -1.08836696e-01
-5.30490458e-01 3.25722784e-01 5.53571880e-01 1.01618946e+00
7.21962571e-01 8.76658320e-01 8.15206587e-01 -6.24126613e-01
-1.73476234e-01 -1.06868362e+00 -6.09406173e-01 2.69100994e-01
5.27136624e-01 -1.02133834e+00 -4.55725044e-01 -1.37805909e-01] | [13.133111953735352, 3.2142913341522217] |
bab03106-3979-44c7-907f-5c36bfa8c1d0 | radar-robust-ai-text-detection-via | 2307.03838 | null | https://arxiv.org/abs/2307.03838v1 | https://arxiv.org/pdf/2307.03838v1.pdf | RADAR: Robust AI-Text Detection via Adversarial Learning | Recent advances in large language models (LLMs) and the intensifying popularity of ChatGPT-like applications have blurred the boundary of high-quality text generation between humans and machines. However, in addition to the anticipated revolutionary changes to our technology and society, the difficulty of distinguishing LLM-generated texts (AI-text) from human-generated texts poses new challenges of misuse and fairness, such as fake content generation, plagiarism, and false accusation of innocent writers. While existing works show that current AI-text detectors are not robust to LLM-based paraphrasing, this paper aims to bridge this gap by proposing a new framework called RADAR, which jointly trains a Robust AI-text Detector via Adversarial leaRning. RADAR is based on adversarial training of a paraphraser and a detector. The paraphraser's goal is to generate realistic contents to evade AI-text detection. RADAR uses the feedback from the detector to update the paraphraser, and vice versa. Evaluated with 8 different LLMs (Pythia, Dolly 2.0, Palmyra, Camel, GPT-J, Dolly 1.0, LLaMA, and Vicuna) across 4 datasets, experimental results show that RADAR significantly outperforms existing AI-text detection methods, especially when paraphrasing is in place. We also identify the strong transferability of RADAR from instruction-tuned LLMs to other LLMs, and evaluate the improved capability of RADAR via GPT-3.5. | ['Tsung-Yi Ho', 'Pin-Yu Chen', 'Xiaomeng Hu'] | 2023-07-07 | null | null | null | null | ['fairness', 'fairness', 'text-generation'] | ['computer-vision', 'miscellaneous', 'natural-language-processing'] | [ 3.01824003e-01 1.02573685e-01 8.17786679e-02 8.01538378e-02
-7.54876256e-01 -7.53067315e-01 1.25619698e+00 -8.47916305e-02
-2.75535583e-01 6.53958380e-01 2.79050440e-01 -5.15962660e-01
5.72750211e-01 -7.53042161e-01 -6.97391987e-01 -1.51698798e-01
3.87668788e-01 7.16912925e-01 1.48609474e-01 -6.17105246e-01
5.89074850e-01 2.53523797e-01 -9.82533634e-01 7.07277536e-01
1.18975067e+00 3.27393651e-01 -1.32720560e-01 1.06391954e+00
-2.63044864e-01 1.41450000e+00 -1.35744131e+00 -9.48712707e-01
2.86133528e-01 -5.76242864e-01 -8.42037320e-01 -1.35836601e-01
5.71376920e-01 -5.64754009e-01 -6.06370270e-01 1.04607105e+00
6.19283736e-01 -2.89325118e-01 7.09474087e-01 -1.80334318e+00
-1.22246933e+00 9.50584471e-01 -5.36661983e-01 1.31079799e-03
5.86188853e-01 6.47220075e-01 7.70938933e-01 -7.51776338e-01
5.41434228e-01 1.77763045e+00 5.88749349e-01 9.25703526e-01
-9.41878498e-01 -1.04437566e+00 -3.87449980e-01 -7.35927671e-02
-1.04893517e+00 -4.18650240e-01 6.32275403e-01 -4.30328518e-01
9.13819015e-01 3.37211579e-01 1.78699493e-01 1.92500174e+00
5.90544343e-01 1.15525961e+00 1.05966067e+00 -4.52917308e-01
9.81613696e-02 4.95454520e-01 -2.86188871e-01 5.31700492e-01
3.60196114e-01 2.20590338e-01 -5.54280400e-01 -5.69945216e-01
3.91719908e-01 -3.48904014e-01 -3.71729173e-02 1.06653251e-01
-1.18059921e+00 1.23729324e+00 1.74073875e-01 2.38725364e-01
3.45592350e-02 8.87435898e-02 5.97018600e-01 6.63540959e-01
4.20853138e-01 9.11584735e-01 2.23834693e-01 -2.34785095e-01
-1.08467853e+00 5.54973602e-01 1.06479597e+00 1.13607156e+00
1.42444462e-01 3.15827191e-01 -4.45469648e-01 6.20993853e-01
1.66415974e-01 9.79552269e-01 1.01187003e+00 -6.64237440e-01
9.23006892e-01 4.72247332e-01 2.76368499e-01 -1.30365551e+00
6.07836014e-03 -1.41237393e-01 -8.54734302e-01 1.32704139e-01
3.63039553e-01 -1.79394379e-01 -4.77133572e-01 1.30688381e+00
-9.89345163e-02 1.02442466e-01 3.25003862e-01 6.96704268e-01
6.62454545e-01 7.55121410e-01 -9.39366743e-02 2.68412411e-01
1.04541910e+00 -1.15110600e+00 -5.40757418e-01 -6.71154797e-01
7.68846333e-01 -1.04583740e+00 1.25366604e+00 3.59667212e-01
-8.84673178e-01 -3.71995658e-01 -1.07491446e+00 5.94346672e-02
-4.88192946e-01 -2.00295582e-01 1.96962655e-01 9.89776790e-01
-7.94275761e-01 5.30380130e-01 -1.08293645e-01 -2.52926230e-01
5.64512789e-01 -1.21213607e-01 -1.79479241e-01 1.05149157e-01
-1.72074842e+00 1.11845028e+00 2.83541948e-01 -2.21222118e-01
-9.70630884e-01 -4.09915715e-01 -6.31364107e-01 -2.67599493e-01
1.91212982e-01 -5.72080255e-01 1.62437606e+00 -1.16664100e+00
-1.58168852e+00 1.07503080e+00 3.47884178e-01 -8.25647414e-01
1.31038713e+00 -3.45805138e-01 -5.22982478e-01 1.75359890e-01
3.27726066e-01 3.78818929e-01 1.46385169e+00 -1.18066466e+00
-4.39444184e-01 -1.86008196e-02 -1.79178566e-01 1.28958151e-01
-3.94058913e-01 2.11210310e-01 3.13910455e-01 -9.08785284e-01
-6.95913792e-01 -8.03306699e-01 1.00502610e-01 -2.37934649e-01
-8.31201434e-01 -2.10238978e-01 1.10801959e+00 -4.78598952e-01
1.14896226e+00 -1.90441096e+00 -2.48080850e-01 -1.23603314e-01
4.00974989e-01 8.15173984e-01 -4.16887999e-01 7.98588037e-01
1.11240581e-01 4.37704057e-01 -3.41240317e-03 -4.28677171e-01
9.01818722e-02 2.33179349e-02 -9.38123584e-01 5.08351803e-01
2.54072875e-01 1.27455723e+00 -1.24782145e+00 -6.53414369e-01
1.63069636e-01 -1.40011787e-01 -3.23009223e-01 3.11122835e-01
-3.18006933e-01 3.73055451e-02 -4.49969053e-01 4.67675090e-01
5.62923014e-01 -2.59079784e-01 -2.36068487e-01 5.98996401e-01
1.04392022e-01 2.07478374e-01 -7.84085870e-01 1.07577646e+00
-3.66833001e-01 7.86247313e-01 -9.73976776e-03 -6.82326078e-01
8.67581129e-01 2.40634650e-01 -1.50798962e-01 -7.18393922e-01
3.61440331e-01 2.51922846e-01 -1.88944131e-01 -5.52451968e-01
1.03124285e+00 -1.23224929e-01 -4.90614206e-01 9.17984664e-01
-2.71061361e-01 -6.28519893e-01 -5.46473488e-02 7.45973408e-01
1.10320807e+00 -3.73351783e-01 1.94848046e-01 1.76995948e-01
5.34034610e-01 1.31330058e-01 -1.38515860e-01 1.40944254e+00
-3.55498284e-01 4.23002511e-01 4.59455013e-01 -6.20184354e-02
-1.18431008e+00 -1.11988127e+00 2.87536293e-01 1.11317062e+00
8.50297362e-02 -2.44643971e-01 -7.50811756e-01 -1.02160931e+00
2.13009313e-01 1.28743529e+00 -3.91238958e-01 -7.02211022e-01
-6.61736667e-01 -4.17770326e-01 1.39024127e+00 1.48629457e-01
5.96514046e-01 -1.46246648e+00 -5.06797075e-01 1.81350216e-01
-4.52102184e-01 -1.19861662e+00 -7.11694002e-01 -5.52956164e-01
-4.82015520e-01 -7.34050214e-01 -5.56251645e-01 -5.64066350e-01
3.80524725e-01 3.85248065e-01 1.22019482e+00 2.30852157e-01
-1.73647985e-01 1.23642817e-01 -5.44580400e-01 -6.04611039e-01
-1.75720620e+00 7.49507174e-02 3.55674438e-02 -3.13906997e-01
2.11413980e-01 -9.07669589e-02 -1.80148616e-01 3.89808565e-01
-9.94477689e-01 1.97713882e-01 6.31064951e-01 9.95538592e-01
-6.22745216e-01 -1.72522128e-01 8.90857697e-01 -1.08695900e+00
1.20366693e+00 -4.73319948e-01 -3.92072946e-01 1.21015899e-01
-5.25307655e-01 -1.13325730e-01 1.11473966e+00 -7.42229700e-01
-9.65433478e-01 -6.75582230e-01 -1.04291469e-01 -4.14346427e-01
2.89105028e-02 3.98777705e-03 1.16732761e-01 9.15871635e-02
1.15822446e+00 4.73377109e-01 3.18320692e-02 7.60742128e-02
6.73151612e-01 1.38062346e+00 6.96070910e-01 -4.95772451e-01
1.27965140e+00 2.81456381e-01 -6.32964790e-01 -6.81936264e-01
-7.36944795e-01 -1.84294730e-01 -7.93921202e-02 -1.54388145e-01
3.57114047e-01 -7.27110088e-01 -5.43246686e-01 8.97106886e-01
-1.45895362e+00 -2.50401676e-01 -2.07319036e-01 -1.08849131e-01
-4.41852689e-01 8.99967253e-01 -8.50655794e-01 -6.91816032e-01
-8.63073647e-01 -9.34977531e-01 9.95709062e-01 -3.34716402e-02
-6.78231120e-01 -8.46415043e-01 5.19334450e-02 7.84187853e-01
5.18498778e-01 2.05174852e-02 8.26251268e-01 -1.15222514e+00
-2.38055825e-01 -7.84496069e-01 -2.23433331e-01 4.67882782e-01
-1.67480465e-02 -1.13551028e-03 -9.14353788e-01 -3.57297897e-01
1.09133676e-01 -7.26963937e-01 4.32540745e-01 -4.18942690e-01
6.29389942e-01 -9.00148451e-01 -1.76038906e-01 5.76678216e-02
7.71252394e-01 -9.70928650e-03 7.12913692e-01 4.10153627e-01
6.41521752e-01 3.93503457e-01 5.04716277e-01 4.81929898e-01
1.18369401e-01 2.27068856e-01 4.03809458e-01 1.84002444e-01
-6.70261011e-02 -8.01218212e-01 8.32775712e-01 6.06524348e-01
6.42533243e-01 -8.28295350e-01 -7.55873322e-01 2.34344095e-01
-1.69015753e+00 -1.37146235e+00 -2.20110372e-01 2.07078242e+00
1.11613572e+00 4.93762255e-01 -7.48826051e-03 -1.82770677e-02
7.88764536e-01 3.39555979e-01 -4.65351373e-01 -8.18715870e-01
-1.80835977e-01 3.42188403e-02 3.85106176e-01 3.37178260e-01
-8.53254795e-01 1.34813178e+00 6.32790709e+00 1.06621921e+00
-1.01373804e+00 3.50498743e-02 4.45893496e-01 4.66149636e-02
-2.46682674e-01 -1.82426184e-01 -7.16337502e-01 8.40340495e-01
1.05736816e+00 -7.12528944e-01 3.82180572e-01 8.95842910e-01
2.21838757e-01 1.89189687e-01 -1.08275509e+00 7.31039464e-01
4.98022914e-01 -1.33454943e+00 4.47988331e-01 -1.54191494e-01
7.29998827e-01 7.21030533e-02 1.28819212e-01 7.88338780e-01
8.68580520e-01 -1.16979885e+00 8.90211165e-01 1.53144179e-02
6.01852238e-01 -5.86692154e-01 7.22694278e-01 9.26957846e-01
-3.77736390e-01 3.37290540e-02 -3.36070061e-01 -5.62044568e-02
-9.47280675e-02 2.62936682e-01 -1.37437141e+00 2.63909847e-01
8.34087208e-02 4.58480060e-01 -7.55832672e-01 4.89169568e-01
-4.37311947e-01 5.34701407e-01 3.65366563e-02 -4.79998440e-01
3.32645059e-01 2.50902236e-01 9.16803956e-01 1.44936526e+00
2.38956492e-02 -4.15491372e-01 9.17475298e-02 1.17864335e+00
-5.79891980e-01 1.55621686e-03 -8.09103310e-01 -5.45703888e-01
5.78074276e-01 1.24956536e+00 -1.91798151e-01 -5.93860388e-01
-1.60460189e-01 1.42234206e+00 2.61393398e-01 7.31467362e-03
-1.10152793e+00 -4.42915171e-01 2.92064607e-01 1.52667820e-01
-2.12576747e-01 1.24482006e-01 -2.09945783e-01 -1.30944335e+00
-1.80085540e-01 -1.60018635e+00 1.83289409e-01 -9.33082938e-01
-1.68497574e+00 5.30488133e-01 -2.39483565e-01 -1.08134937e+00
-6.10970855e-01 -3.35404813e-01 -8.99096429e-01 8.72426629e-01
-1.12059736e+00 -1.26293826e+00 -2.64955759e-02 4.73832101e-01
8.50261927e-01 -5.24878383e-01 6.47558331e-01 -4.91068810e-02
-3.63637954e-01 1.05849957e+00 2.62646973e-01 4.98966932e-01
1.04564357e+00 -1.14047217e+00 1.15583754e+00 8.72073829e-01
3.03301327e-02 4.25587058e-01 1.00043058e+00 -9.88490820e-01
-1.33032179e+00 -1.19173193e+00 9.78517175e-01 -9.14409637e-01
1.18537605e+00 -7.35880673e-01 -9.15567994e-01 5.93922436e-01
3.13057244e-01 -4.59443390e-01 2.99896449e-01 -5.81153631e-01
-6.50127709e-01 5.62122643e-01 -1.29643822e+00 9.18195784e-01
7.05208540e-01 -6.78059757e-01 -9.86448228e-01 6.12540483e-01
8.48889768e-01 -4.65383530e-01 -2.85769135e-01 -2.16482535e-01
4.10598993e-01 -8.66168737e-01 8.80750537e-01 -8.21346223e-01
9.14162040e-01 1.80519998e-01 1.92624643e-01 -1.35937154e+00
-3.79370637e-02 -1.08428502e+00 -1.46624371e-01 1.35173392e+00
7.64175951e-02 -7.98040569e-01 6.56107128e-01 4.02698845e-01
8.41234997e-02 -8.14672410e-02 -7.74884343e-01 -9.81092811e-01
5.94910979e-01 -3.10609132e-01 3.48987222e-01 1.19853783e+00
2.54280180e-01 5.98412454e-01 -8.30337048e-01 1.01466691e-02
6.33930206e-01 8.54621902e-02 1.17667913e+00 -9.12026882e-01
-4.45996135e-01 -6.50921404e-01 -1.17028400e-01 -1.02783573e+00
3.88784051e-01 -1.11251009e+00 6.21058941e-02 -1.10265875e+00
1.20465368e-01 -8.75056684e-02 4.20590162e-01 3.07516158e-01
-2.62149900e-01 1.19926892e-01 5.09984493e-01 4.50599611e-01
-5.21751165e-01 4.89359945e-01 1.20975089e+00 -3.84995103e-01
-3.72207770e-03 1.98117375e-01 -8.02317977e-01 7.21962988e-01
8.86420548e-01 -7.22692192e-01 -1.72301948e-01 -2.20651925e-01
2.83338457e-01 7.70175159e-02 6.44495428e-01 -8.68497610e-01
1.98611245e-01 -1.88894495e-01 1.78393811e-01 -4.06380147e-01
-8.63560364e-02 -4.46527362e-01 -4.71538186e-01 6.93680406e-01
-6.19076550e-01 2.13351957e-02 2.03912675e-01 6.05411530e-01
2.58084029e-01 -5.93503594e-01 8.80461454e-01 -3.07798743e-01
-3.99403721e-01 -1.18699431e-01 -9.03555930e-01 6.96092665e-01
8.73631299e-01 -1.45456985e-01 -9.01392937e-01 -6.75434172e-01
3.00728437e-03 3.45094204e-01 4.80571717e-01 9.98000205e-01
6.75372005e-01 -1.02545166e+00 -1.14009845e+00 2.06049249e-01
3.02858770e-01 -4.50806290e-01 -6.28557205e-02 2.27219507e-01
-5.03469765e-01 2.55322397e-01 -1.47084370e-01 -3.17967713e-01
-1.29125059e+00 6.05833173e-01 2.71948695e-01 -5.32025278e-01
-5.56369185e-01 6.17307484e-01 1.43729731e-01 -4.70929384e-01
1.21693894e-01 1.03516474e-01 2.84342676e-01 -3.73134792e-01
6.40061915e-01 4.37341869e-01 -1.66981921e-01 -6.13003850e-01
-4.92272601e-02 -2.99306870e-01 -5.10020971e-01 -1.10134900e-01
6.75709128e-01 2.23291311e-02 4.48745601e-02 3.77487898e-01
9.98402417e-01 2.49822244e-01 -5.41404068e-01 -2.70039856e-01
-5.11551388e-02 -5.78014672e-01 -3.74903858e-01 -1.02451456e+00
-4.03336942e-01 8.52159262e-01 9.33108032e-02 3.86504531e-01
4.07817781e-01 -1.79648638e-01 1.19141424e+00 5.16781628e-01
4.30434942e-01 -9.84946966e-01 7.87749648e-01 7.27465093e-01
1.14460409e+00 -1.27755904e+00 -9.87586305e-02 5.26896082e-02
-9.98867571e-01 8.82801831e-01 8.28644872e-01 -8.50861818e-02
-1.17397480e-01 2.61277646e-01 1.58424288e-01 2.41679698e-02
-7.78219938e-01 6.27732873e-01 -8.93289447e-02 4.64776397e-01
2.77453005e-01 -1.29171996e-03 1.67256277e-02 4.24348682e-01
-6.98244274e-01 -6.17610440e-02 1.01587689e+00 9.70768809e-01
-4.61789459e-01 -9.59122658e-01 -7.31686711e-01 4.68462110e-01
-4.77881461e-01 -2.93498427e-01 -9.51243281e-01 8.16928566e-01
-2.40106568e-01 1.24387395e+00 -1.39123887e-01 -4.36773211e-01
1.99903741e-01 2.77584959e-02 9.63602737e-02 -5.99966347e-01
-1.19659555e+00 -4.09337461e-01 1.80291414e-01 -2.41584361e-01
2.58680642e-01 -3.69515389e-01 -1.14187992e+00 -9.54165339e-01
-5.10801792e-01 1.10771567e-01 2.74869710e-01 7.64836609e-01
3.18336368e-01 -3.41426171e-02 7.14895546e-01 -3.98211449e-01
-1.29719436e+00 -1.04416966e+00 -2.62556463e-01 7.61912048e-01
2.37570599e-01 -5.77794835e-02 -6.48647666e-01 -1.56428799e-01] | [6.304945945739746, 8.260127067565918] |
4871415a-6443-465f-9cd2-e24bf78723bd | talk-to-edit-fine-grained-facial-editing-via | 2109.04425 | null | https://arxiv.org/abs/2109.04425v1 | https://arxiv.org/pdf/2109.04425v1.pdf | Talk-to-Edit: Fine-Grained Facial Editing via Dialog | Facial editing is an important task in vision and graphics with numerous applications. However, existing works are incapable to deliver a continuous and fine-grained editing mode (e.g., editing a slightly smiling face to a big laughing one) with natural interactions with users. In this work, we propose Talk-to-Edit, an interactive facial editing framework that performs fine-grained attribute manipulation through dialog between the user and the system. Our key insight is to model a continual "semantic field" in the GAN latent space. 1) Unlike previous works that regard the editing as traversing straight lines in the latent space, here the fine-grained editing is formulated as finding a curving trajectory that respects fine-grained attribute landscape on the semantic field. 2) The curvature at each step is location-specific and determined by the input image as well as the users' language requests. 3) To engage the users in a meaningful dialog, our system generates language feedback by considering both the user request and the current state of the semantic field. We also contribute CelebA-Dialog, a visual-language facial editing dataset to facilitate large-scale study. Specifically, each image has manually annotated fine-grained attribute annotations as well as template-based textual descriptions in natural language. Extensive quantitative and qualitative experiments demonstrate the superiority of our framework in terms of 1) the smoothness of fine-grained editing, 2) the identity/attribute preservation, and 3) the visual photorealism and dialog fluency. Notably, user study validates that our overall system is consistently favored by around 80% of the participants. Our project page is https://www.mmlab-ntu.com/project/talkedit/. | ['Ziwei Liu', 'Chen Change Loy', 'Xingang Pan', 'Ziqi Huang', 'Yuming Jiang'] | 2021-09-09 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Jiang_Talk-To-Edit_Fine-Grained_Facial_Editing_via_Dialog_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Jiang_Talk-To-Edit_Fine-Grained_Facial_Editing_via_Dialog_ICCV_2021_paper.pdf | iccv-2021-1 | ['facial-editing'] | ['computer-vision'] | [ 1.48043454e-01 1.34618089e-01 1.20659016e-01 -7.60043085e-01
-2.92179227e-01 -7.68975735e-01 7.98037827e-01 -4.62657064e-01
3.97139974e-02 3.75682652e-01 4.45904434e-01 7.61506259e-02
2.01722965e-01 -5.22115290e-01 -4.99023020e-01 -3.69726628e-01
5.98856568e-01 3.31169009e-01 -2.19064727e-01 -3.43470722e-01
1.88448623e-01 5.13522565e-01 -1.43646896e+00 4.77405846e-01
1.00630903e+00 8.24066341e-01 1.44632772e-01 5.34488738e-01
-2.06570357e-01 3.51411521e-01 -2.85529912e-01 -6.84003055e-01
3.04098785e-01 -5.52513778e-01 -7.69430757e-01 4.40423012e-01
7.86579430e-01 -4.28751230e-01 1.22510120e-01 1.24750888e+00
3.23735148e-01 1.48442790e-01 5.40477872e-01 -1.40730572e+00
-9.21357930e-01 4.75771576e-02 -7.02212989e-01 -6.33044302e-01
6.05728030e-01 3.80758703e-01 7.60614872e-01 -1.09581244e+00
8.52577806e-01 1.56473160e+00 4.17811006e-01 8.33725214e-01
-1.23400438e+00 -7.38584280e-01 4.00826693e-01 -2.15626702e-01
-1.42550099e+00 -6.81630313e-01 8.70217144e-01 -7.89021671e-01
3.69390249e-01 5.07386804e-01 7.73869038e-01 1.13633358e+00
-2.60258466e-03 3.99112493e-01 1.12420607e+00 -3.42009872e-01
3.85705545e-03 4.63499933e-01 -3.05097252e-01 8.73038828e-01
-5.55717468e-01 6.28120732e-03 -7.24572539e-01 -9.58369002e-02
1.06528735e+00 -1.00377910e-01 -7.47927576e-02 -3.85577857e-01
-1.08175075e+00 6.90613449e-01 1.87010795e-01 3.84712331e-02
-3.78496915e-01 4.82422374e-02 2.04760090e-01 2.67211497e-01
6.32644117e-01 1.59779221e-01 -1.81667417e-01 -2.17792362e-01
-8.06149960e-01 1.90188333e-01 5.90281725e-01 1.16423595e+00
8.42339158e-01 -8.20325464e-02 -4.86287147e-01 9.79140997e-01
3.17609072e-01 4.37113255e-01 3.73145379e-02 -1.20457375e+00
1.56401508e-02 6.81435704e-01 2.48243734e-01 -1.01334822e+00
3.12532596e-02 3.42108570e-02 -7.99968958e-01 6.23003840e-01
4.02676344e-01 -1.81404069e-01 -7.83349097e-01 2.01465631e+00
5.60699999e-01 -1.91835105e-01 -3.83881181e-01 9.46320415e-01
6.92515135e-01 4.10709083e-01 3.01560789e-01 -2.60817498e-01
1.53722453e+00 -8.74611914e-01 -1.02825820e+00 -4.52105440e-02
1.65861577e-01 -1.08841133e+00 1.77809703e+00 1.18412413e-01
-1.25665140e+00 -6.88944042e-01 -4.77725178e-01 -3.25167328e-01
-1.07086353e-01 3.87979060e-01 6.51711047e-01 5.70347548e-01
-1.34244049e+00 2.11483985e-01 -4.44101900e-01 -5.93266845e-01
4.40273374e-01 1.34875640e-01 -4.89771307e-01 1.83311477e-01
-8.88186157e-01 5.88506639e-01 -2.00752392e-01 2.22846538e-01
-6.09240651e-01 -9.09305334e-01 -6.87284946e-01 -1.07994303e-01
2.73254365e-01 -9.45571065e-01 1.35980082e+00 -1.53781915e+00
-1.82139599e+00 1.26104665e+00 -5.93966603e-01 3.54600191e-01
1.09374952e+00 -2.22519740e-01 -4.50853333e-02 -1.38695911e-01
1.05495006e-01 1.07461870e+00 9.76804733e-01 -1.40553629e+00
-3.98466080e-01 -4.38529611e-01 2.79637247e-01 5.91620266e-01
-3.42838883e-01 1.11177303e-01 -7.48733103e-01 -8.35560322e-01
-2.00781494e-01 -9.30339098e-01 1.50548611e-02 6.74906254e-01
-5.50425470e-01 -2.05298349e-01 9.80430007e-01 -5.44140399e-01
1.04577434e+00 -2.40705109e+00 7.59022012e-02 1.83149189e-01
3.74477088e-01 -1.24518573e-02 -5.96283115e-02 3.72333437e-01
-3.32695805e-02 2.20210090e-01 2.74389461e-02 -7.19487548e-01
1.25650674e-01 -2.44294792e-01 -2.93489486e-01 3.21839422e-01
-3.09576932e-02 1.11182845e+00 -7.71441579e-01 -4.88387018e-01
3.39221925e-01 7.35717952e-01 -5.64285219e-01 3.63706768e-01
-2.41994306e-01 7.42831349e-01 -4.54719633e-01 7.15742648e-01
7.30845571e-01 -1.53049678e-01 2.53455341e-02 -5.67494869e-01
-2.57236660e-01 -3.91047031e-01 -1.01546216e+00 1.71493268e+00
-5.11286020e-01 5.95043302e-01 2.85184145e-01 -3.35717537e-02
9.49463665e-01 1.46141410e-01 2.90641189e-01 -6.00607097e-01
-7.06337616e-02 -2.03118443e-01 -4.68884975e-01 -4.72939610e-01
5.64706743e-01 -1.96138248e-01 3.47953439e-02 5.17446518e-01
-1.78860456e-01 -1.76552519e-01 -1.66458905e-01 3.72790903e-01
3.18019986e-01 2.17609271e-01 2.74100155e-01 -3.22355688e-01
3.09560955e-01 -3.66836369e-01 3.20556045e-01 4.48314935e-01
-2.45318100e-01 6.72925830e-01 5.23902595e-01 -2.50464231e-01
-8.58104765e-01 -1.03286660e+00 6.01920076e-02 1.49463427e+00
2.24275574e-01 -4.42625910e-01 -1.18321538e+00 -4.64755803e-01
-1.17633946e-01 6.42164052e-01 -8.78260911e-01 2.76115015e-02
-1.34601727e-01 1.50168026e-02 3.10229391e-01 2.89900094e-01
6.72722101e-01 -1.32772553e+00 -4.74626958e-01 -2.72105068e-01
-3.00350249e-01 -9.40717399e-01 -1.34499454e+00 -8.30184877e-01
-4.64835167e-01 -7.78093338e-01 -5.24577558e-01 -6.77643836e-01
1.03711951e+00 1.48308575e-01 8.96339655e-01 6.41494617e-02
-2.56762564e-01 6.60155535e-01 -1.29497632e-01 -2.41820246e-01
-3.12088192e-01 -3.57120693e-01 -6.53658062e-02 5.83206594e-01
8.38271081e-02 -5.10233223e-01 -7.09677815e-01 5.67785859e-01
-7.33315527e-01 6.35186553e-01 1.72137171e-01 5.97308874e-01
5.91614425e-01 -3.92397642e-01 3.29984516e-01 -1.04480672e+00
1.04880500e+00 1.68396328e-02 -5.09660423e-01 3.73728424e-01
-4.49946314e-01 -3.28416407e-01 2.86120266e-01 -5.60302317e-01
-1.51103103e+00 2.81897694e-01 -2.96111852e-02 -4.54120994e-01
-3.84010136e-01 -6.51178658e-02 -4.43669438e-01 -1.41222194e-01
5.49108267e-01 6.37150332e-02 1.93019181e-01 -1.66380599e-01
7.72244573e-01 5.60435951e-01 6.23638272e-01 -7.05607951e-01
8.30534875e-01 7.05774248e-01 -4.28105950e-01 -8.39782178e-01
-5.50031304e-01 -1.01821706e-01 -8.03764284e-01 -7.36726582e-01
9.99104798e-01 -7.03220010e-01 -1.27873826e+00 6.84131503e-01
-1.23054492e+00 -5.28449655e-01 -3.87840927e-01 -5.04112840e-02
-6.66432500e-01 1.72434971e-01 -3.47626120e-01 -8.39047611e-01
-2.98788965e-01 -1.27215576e+00 1.42686951e+00 4.20809031e-01
-5.73731601e-01 -1.07729840e+00 -1.91144109e-01 3.52601826e-01
5.24145842e-01 3.92437935e-01 7.27261305e-01 1.31870925e-01
-4.85386997e-01 8.05102736e-02 -3.17673028e-01 2.61001512e-02
4.83203322e-01 3.61588717e-01 -1.16527629e+00 -1.00543506e-01
-1.96526453e-01 -3.36338460e-01 2.46265277e-01 4.17727858e-01
1.27244794e+00 -4.08406585e-01 -1.22577906e-01 7.31831133e-01
8.68100226e-01 1.13525756e-01 6.40034258e-01 -1.65089816e-01
8.89604032e-01 9.09201145e-01 6.78257942e-01 5.39916933e-01
4.41321909e-01 8.69972169e-01 2.22846732e-01 -4.20479715e-01
-1.63035974e-01 -6.98839903e-01 3.13225776e-01 2.71320969e-01
-1.63561702e-01 2.22069286e-02 -6.01274252e-01 2.81647384e-01
-1.77213490e+00 -9.21651602e-01 7.19372928e-02 2.05546713e+00
9.93335843e-01 -3.66211802e-01 5.92425056e-02 -6.08497262e-01
7.77328134e-01 1.62633210e-01 -6.76899910e-01 -5.39232135e-01
9.60781202e-02 -1.13436922e-01 -7.26079643e-02 9.58218694e-01
-7.67188489e-01 1.42391765e+00 5.38835001e+00 7.43822575e-01
-1.34490407e+00 6.31461442e-02 8.47467124e-01 -8.49814489e-02
-6.13589168e-01 -8.13770890e-02 -6.26435935e-01 4.42676246e-01
1.23106495e-01 -2.65970677e-01 7.19882727e-01 5.79368949e-01
7.67536819e-01 3.27844732e-02 -1.19212627e+00 1.14856887e+00
-5.91099449e-02 -1.38551807e+00 2.80418664e-01 9.20970142e-02
6.26631856e-01 -5.67241907e-01 3.46987307e-01 -1.02684144e-02
2.27483898e-01 -1.27514911e+00 9.91847336e-01 8.39788258e-01
1.63649368e+00 -5.09836018e-01 8.64518154e-03 1.65717840e-01
-1.24938405e+00 4.03896123e-01 7.41695687e-02 1.27303302e-01
3.86548847e-01 2.53511548e-01 -6.86078608e-01 8.63941908e-02
7.00615466e-01 5.01663923e-01 -3.85239065e-01 2.76752055e-01
-3.11750740e-01 1.40958384e-01 -6.47375779e-03 2.10314155e-01
-7.05819279e-02 -6.76958382e-01 4.88299817e-01 1.24957108e+00
1.15785532e-01 3.25311542e-01 1.30204082e-01 1.18079603e+00
-2.53332049e-01 2.28204191e-01 -5.26721060e-01 -5.54654635e-02
5.34512579e-01 1.40345764e+00 -4.31832731e-01 -5.99501207e-02
-1.33091897e-01 1.37588453e+00 2.99795449e-01 7.05826402e-01
-7.29437470e-01 -1.13029569e-01 1.03095138e+00 2.90840805e-01
-1.18438117e-01 -1.33206114e-01 -4.74802643e-01 -9.36359286e-01
1.09728612e-01 -8.58380795e-01 7.37527534e-02 -1.11093378e+00
-1.23432159e+00 7.63334036e-01 -9.41154361e-02 -9.26317036e-01
-1.93695933e-01 -1.85480550e-01 -7.02986658e-01 1.21145856e+00
-1.08784914e+00 -1.63193405e+00 -8.90717506e-01 9.35323417e-01
7.97056735e-01 1.31057538e-02 7.93037772e-01 1.45874545e-01
-2.89527386e-01 8.92452896e-01 -5.67196012e-01 -2.82856319e-02
1.00714266e+00 -9.89821970e-01 3.96358281e-01 4.98920470e-01
-3.10528055e-02 8.52063239e-01 6.92617118e-01 -8.20699513e-01
-1.24710321e+00 -1.13152254e+00 8.18779945e-01 -5.01314759e-01
4.19438809e-01 -7.02424109e-01 -8.26406062e-01 8.46955478e-01
3.46828401e-01 -7.40751103e-02 5.13399601e-01 2.80229561e-02
-3.36198986e-01 -5.39582446e-02 -1.35031366e+00 9.17651772e-01
1.33642721e+00 -7.25033820e-01 -5.08642197e-02 3.29664588e-01
5.86671531e-01 -6.64533377e-01 -6.10634625e-01 1.82223976e-01
9.02719140e-01 -9.98386681e-01 6.92019284e-01 -5.84214568e-01
4.00578022e-01 -3.41795504e-01 1.07830693e-03 -1.32289588e+00
-3.35903674e-01 -1.03807139e+00 3.55129421e-01 1.54442859e+00
1.52636901e-01 -5.20440102e-01 6.08885288e-01 9.97232854e-01
1.27592474e-01 -7.75284946e-01 -4.92792577e-01 -7.55012706e-02
-2.31758446e-01 -2.91576564e-01 5.80928743e-01 1.10486829e+00
-2.58195013e-01 2.57047594e-01 -6.10810399e-01 6.44201636e-02
4.56268817e-01 2.39358805e-02 1.08505225e+00 -1.03392494e+00
8.95017758e-03 -5.57637393e-01 2.17397958e-02 -1.03081322e+00
2.07415462e-01 -6.47561193e-01 -1.03500269e-01 -1.41230190e+00
2.85876721e-01 -4.88036513e-01 3.54802579e-01 5.51122069e-01
-6.80451691e-02 2.38674715e-01 2.77160823e-01 2.39071921e-01
-3.57301354e-01 7.07463145e-01 1.54828465e+00 1.87387139e-01
-4.15205151e-01 -5.76853491e-02 -9.54322219e-01 8.32446814e-01
5.98166764e-01 -1.47510541e-03 -7.23772705e-01 -4.64223206e-01
1.10625364e-01 5.25849648e-02 5.70262372e-01 -2.40541562e-01
2.17824385e-01 -6.41840160e-01 3.26286644e-01 -1.69025287e-01
5.93144953e-01 -6.29569590e-01 5.50876737e-01 -1.05323642e-02
-5.14070868e-01 1.02801181e-01 1.46730959e-01 3.34235847e-01
-5.75090311e-02 2.93507218e-01 9.94967282e-01 -7.53719881e-02
-6.76940322e-01 6.49510860e-01 1.79782845e-02 2.76875100e-03
1.14793873e+00 -4.59153503e-01 -1.93138272e-01 -8.54788959e-01
-8.40215862e-01 2.10078225e-01 8.45219195e-01 5.98719776e-01
6.64909422e-01 -1.40283859e+00 -7.35517085e-01 4.54166114e-01
1.86489314e-01 -1.68669790e-01 5.91650486e-01 6.29387438e-01
-2.38111109e-01 -8.07786286e-02 -3.71206313e-01 -7.32951999e-01
-1.58019507e+00 1.02848187e-02 6.11317813e-01 3.22314650e-01
-6.08308375e-01 9.20917213e-01 8.61556828e-01 -4.63469654e-01
2.10278675e-01 1.70532353e-02 -1.54431621e-02 8.84874538e-02
5.02163827e-01 6.84073940e-02 -2.33702794e-01 -7.81978130e-01
-1.26611471e-01 8.14905584e-01 -1.97147448e-02 -2.89764434e-01
9.83752489e-01 -4.62884665e-01 -2.71551549e-01 3.74779612e-01
9.40466404e-01 3.86294067e-01 -1.63093448e+00 -9.72384214e-02
-5.30326784e-01 -8.35191250e-01 -3.36675584e-01 -9.51387048e-01
-1.05582750e+00 7.44514883e-01 5.66349566e-01 -1.56004727e-01
1.21754372e+00 1.46263078e-01 3.62151742e-01 -1.52174681e-01
1.61640108e-01 -1.00812328e+00 2.75432467e-01 3.28490853e-01
1.41755009e+00 -1.08190811e+00 -3.78408611e-01 -6.60520077e-01
-1.22189689e+00 9.12512779e-01 8.87867332e-01 3.30324292e-01
5.93384922e-01 1.67813659e-01 4.73389089e-01 -3.63068223e-01
-6.37066066e-01 1.47560328e-01 4.62852687e-01 5.74776113e-01
5.38787723e-01 3.27898473e-01 -1.45911891e-02 4.16572690e-01
-5.69490731e-01 3.40395086e-02 3.05559009e-01 3.53979290e-01
-1.29777357e-01 -9.54515338e-01 -1.11192562e-01 2.10371718e-01
4.90351170e-02 8.24666396e-03 -6.91564262e-01 6.31637752e-01
1.47000730e-01 8.61572087e-01 3.51347141e-02 -2.05662668e-01
5.46863675e-01 -4.22739163e-02 3.49910051e-01 -6.33714795e-01
-5.04843354e-01 6.27958626e-02 -1.20731652e-01 -9.14254546e-01
-1.02589272e-01 -5.97013652e-01 -1.11469495e+00 -5.31389534e-01
1.88306704e-01 -1.86829507e-01 7.10066319e-01 7.64953852e-01
5.68763912e-01 2.89259970e-01 6.37737572e-01 -8.27359676e-01
-2.09497362e-01 -8.47874165e-01 -4.21989024e-01 7.10434675e-01
2.76348323e-01 -6.38558805e-01 -8.06747898e-02 4.94762629e-01] | [12.647252082824707, -0.3278326094150543] |
6e37e18a-a8aa-4ab7-9680-756c9dfd5c07 | meta-prediction-model-for-distillation-aware | 2305.16948 | null | https://arxiv.org/abs/2305.16948v1 | https://arxiv.org/pdf/2305.16948v1.pdf | Meta-prediction Model for Distillation-Aware NAS on Unseen Datasets | Distillation-aware Neural Architecture Search (DaNAS) aims to search for an optimal student architecture that obtains the best performance and/or efficiency when distilling the knowledge from a given teacher model. Previous DaNAS methods have mostly tackled the search for the neural architecture for fixed datasets and the teacher, which are not generalized well on a new task consisting of an unseen dataset and an unseen teacher, thus need to perform a costly search for any new combination of the datasets and the teachers. For standard NAS tasks without KD, meta-learning-based computationally efficient NAS methods have been proposed, which learn the generalized search process over multiple tasks (datasets) and transfer the knowledge obtained over those tasks to a new task. However, since they assume learning from scratch without KD from a teacher, they might not be ideal for DaNAS scenarios. To eliminate the excessive computational cost of DaNAS methods and the sub-optimality of rapid NAS methods, we propose a distillation-aware meta accuracy prediction model, DaSS (Distillation-aware Student Search), which can predict a given architecture's final performances on a dataset when performing KD with a given teacher, without having actually to train it on the target task. The experimental results demonstrate that our proposed meta-prediction model successfully generalizes to multiple unseen datasets for DaNAS tasks, largely outperforming existing meta-NAS methods and rapid NAS baselines. Code is available at https://github.com/CownowAn/DaSS | ['Sung Ju Hwang', 'Minseon Kim', 'Sohyun An', 'Hayeon Lee'] | 2023-05-26 | null | null | null | null | ['architecture-search'] | ['methodology'] | [-7.68491402e-02 1.99988182e-03 -5.25777452e-02 -3.90247077e-01
-9.38399971e-01 -5.46405315e-01 3.43214631e-01 -4.16473262e-02
-6.43942952e-01 5.45932829e-01 -3.81279737e-01 -3.41829211e-01
-3.16734612e-01 -6.35452628e-01 -9.32722211e-01 -7.96252728e-01
4.03417438e-01 1.01163411e+00 2.38580406e-01 -1.19611189e-01
-6.05266280e-02 2.69594073e-01 -1.78905690e+00 2.00162902e-01
1.05674720e+00 1.16099465e+00 6.10435307e-01 5.99452615e-01
-4.02630456e-02 4.31504458e-01 -6.21468663e-01 -6.39071643e-01
3.17504406e-01 -2.97896743e-01 -8.54666889e-01 -4.75977600e-01
8.46933842e-01 -3.76034677e-01 -2.22474933e-01 9.01706874e-01
7.15671122e-01 4.22418356e-01 5.40685058e-01 -1.23612857e+00
-7.55926728e-01 7.83553958e-01 -3.46912891e-01 3.36315811e-01
-2.89660245e-01 2.93409109e-01 1.03624201e+00 -1.17614794e+00
1.08712219e-01 1.11186850e+00 5.34882605e-01 8.26869607e-01
-1.19347513e+00 -8.92068088e-01 3.64728808e-01 4.27006364e-01
-1.37886846e+00 -3.08544785e-01 7.82743573e-01 -2.87194610e-01
9.15844381e-01 2.63553768e-01 5.43852031e-01 1.23993874e+00
-3.07630956e-01 1.18899798e+00 8.55806351e-01 -4.19465423e-01
4.23182130e-01 2.79671103e-01 2.90916502e-01 5.91865957e-01
7.04834089e-02 -1.16549758e-02 -5.36012828e-01 -1.11687683e-01
2.00140521e-01 -1.91782645e-04 -1.96891114e-01 -5.06377578e-01
-1.04084551e+00 7.16773212e-01 5.92910051e-01 1.66613549e-01
-4.39372629e-01 1.26556978e-01 2.39229321e-01 6.64959610e-01
4.23965752e-01 8.18826973e-01 -1.08054531e+00 -1.53628245e-01
-9.78594959e-01 3.96375537e-01 5.67652762e-01 7.97183096e-01
8.56496513e-01 2.69056380e-01 -2.29525462e-01 8.64839673e-01
4.46031615e-03 3.22661281e-01 1.04417062e+00 -7.43289948e-01
5.02723992e-01 9.49376345e-01 -1.02794468e-01 -4.10682768e-01
-2.23493695e-01 -8.96514416e-01 -6.62804067e-01 1.57644093e-01
3.23615164e-01 -2.27949440e-01 -1.11148036e+00 1.96349072e+00
6.59633636e-01 7.72533178e-01 1.84084371e-01 7.51310587e-01
1.06121361e+00 6.24701321e-01 -9.07855555e-02 -1.42500117e-01
1.16230679e+00 -1.27639151e+00 -9.90058854e-02 -3.76243770e-01
9.05958652e-01 -3.82972986e-01 1.37042975e+00 7.17627227e-01
-1.28341436e+00 -7.17755795e-01 -9.13539290e-01 -1.41051158e-01
-3.51623446e-01 2.08771050e-01 2.16579944e-01 3.39786172e-01
-1.14322257e+00 4.47957695e-01 -7.56582260e-01 7.68646449e-02
3.86623293e-01 6.29041076e-01 1.50236011e-01 -1.33575022e-01
-9.50383425e-01 9.04278815e-01 6.17511809e-01 1.07350528e-01
-1.39245260e+00 -1.15558350e+00 -3.33908588e-01 3.46287519e-01
7.68586814e-01 -8.64087582e-01 1.66504347e+00 -1.16481566e+00
-1.52538455e+00 5.45233250e-01 1.52367800e-01 -6.01248980e-01
4.64733571e-01 -3.47000927e-01 3.44476104e-02 -4.68283504e-01
-3.69356215e-01 7.80107081e-01 7.82288790e-01 -1.02097416e+00
-6.09736860e-01 -4.88863200e-01 1.36677787e-01 5.22257268e-01
-7.47776985e-01 -5.16282856e-01 -3.98295850e-01 -5.65520227e-01
-3.88973318e-02 -9.80823696e-01 -1.88044831e-01 -3.31330597e-01
-1.61504999e-01 -7.79482365e-01 7.78675616e-01 -2.44849563e-01
1.33410501e+00 -2.05258298e+00 4.45360273e-01 -1.56706339e-03
2.34063029e-01 7.60723233e-01 -7.18097985e-01 -5.65308407e-02
-1.78058036e-02 -1.55959606e-01 1.47861779e-01 -5.17680585e-01
-1.84554502e-01 2.44627669e-01 -2.39501953e-01 -1.15546629e-01
-8.74216929e-02 9.38601434e-01 -8.64851892e-01 -1.27228409e-01
-9.77888256e-02 2.71885008e-01 -5.88965297e-01 5.32776654e-01
-5.64347744e-01 3.03593993e-01 -5.19075871e-01 5.25427341e-01
3.96479815e-01 -4.88147736e-01 1.18585736e-01 -7.04807639e-02
1.86820090e-01 3.54028314e-01 -1.17685854e+00 1.80041587e+00
-7.66610920e-01 1.92972988e-01 -6.88151047e-02 -1.34636116e+00
1.00739062e+00 1.74351469e-01 2.21458480e-01 -8.01281571e-01
6.69899583e-02 4.24007982e-01 1.71060920e-01 -3.10866475e-01
2.04916045e-01 1.01556949e-01 1.02029748e-01 4.87785637e-01
4.20805514e-01 -2.16317102e-02 -2.08510593e-01 -1.36800334e-01
1.12132466e+00 -4.80163209e-02 1.23080932e-01 -1.79193452e-01
4.74349916e-01 1.64303914e-01 6.18597925e-01 8.12301636e-01
4.65116138e-03 1.84447005e-01 -8.64103585e-02 -8.89543653e-01
-9.25429523e-01 -9.60237384e-01 1.10835887e-01 1.57879174e+00
-1.36949643e-01 -1.59161821e-01 -7.28666961e-01 -1.01109898e+00
-1.05410451e-02 8.07311416e-01 -7.05498755e-01 -4.10158217e-01
-6.31836712e-01 -5.64024866e-01 4.21662480e-01 5.28884590e-01
4.58832294e-01 -8.85473549e-01 -7.14676201e-01 1.78190589e-01
5.61633930e-02 -6.71670914e-01 -4.53364879e-01 4.32773829e-01
-1.06086791e+00 -9.75345016e-01 -7.36064970e-01 -9.04986203e-01
5.80477297e-01 2.36882061e-01 1.07066727e+00 3.33510756e-01
-2.51167357e-01 3.10134053e-01 2.66590863e-02 -5.28242409e-01
-3.02121609e-01 5.67853153e-01 3.14949334e-01 -8.87578651e-02
5.82652926e-01 -6.63659513e-01 -6.52537882e-01 3.33341390e-01
-6.59625232e-01 3.04817855e-01 6.72450542e-01 1.09740353e+00
6.31525636e-01 1.73641345e-03 8.49979401e-01 -7.17990577e-01
5.47235847e-01 -6.16224468e-01 -8.00609291e-01 6.25111639e-01
-1.15642488e+00 2.49953568e-01 7.63005376e-01 -9.18725669e-01
-9.21181619e-01 9.81567428e-02 1.27419353e-01 -9.79569256e-01
-1.31350040e-01 6.21755123e-01 6.71025738e-02 1.18959226e-01
8.74393046e-01 3.63280058e-01 8.07961002e-02 -8.66458297e-01
1.85622439e-01 5.33900321e-01 5.74841022e-01 -8.55403721e-01
7.25161076e-01 -1.87311605e-01 -1.61867499e-01 -2.87090153e-01
-1.25883842e+00 -3.73192549e-01 -7.07160234e-01 2.65999854e-01
4.62002397e-01 -1.11491477e+00 -6.64070129e-01 4.81912225e-01
-9.41107988e-01 -7.37410367e-01 -4.62486476e-01 4.47941929e-01
-4.97747868e-01 -1.48550212e-01 -2.09067553e-01 -5.84299147e-01
-5.14287412e-01 -1.43868661e+00 5.35918176e-01 2.77501553e-01
-3.53420936e-02 -1.08648217e+00 2.45614886e-01 6.60748005e-01
5.68907261e-01 -3.07084769e-01 1.20181715e+00 -1.25510633e+00
-5.66892982e-01 1.53024226e-01 1.83505505e-01 4.68659997e-01
-6.93244115e-02 -3.03365797e-01 -1.04492784e+00 -5.47609925e-01
-5.67377321e-02 -9.23291564e-01 8.11514378e-01 2.12922066e-01
1.42658079e+00 -6.88152552e-01 -2.45546937e-01 6.43299639e-01
1.31563747e+00 3.32510740e-01 -2.70753559e-02 6.30298615e-01
5.84327281e-01 2.99022883e-01 4.99441445e-01 2.14755312e-01
4.60376024e-01 8.09004307e-01 6.42492533e-01 1.50262788e-01
-2.93794930e-01 -2.80247122e-01 3.81583184e-01 9.81781423e-01
6.24308847e-02 -2.65927464e-01 -1.16933322e+00 8.49288464e-01
-1.95606697e+00 -4.95927721e-01 2.65269965e-01 2.23766279e+00
1.05879700e+00 -1.83338597e-01 1.78477809e-01 -2.46694565e-01
3.51227283e-01 -2.38375783e-01 -1.31370413e+00 -4.35246080e-01
2.53011346e-01 3.62846941e-01 -9.75159258e-02 1.90847769e-01
-8.01307738e-01 8.89182031e-01 5.35567236e+00 9.83119667e-01
-1.13759899e+00 3.95532697e-01 7.04762936e-01 -4.88169163e-01
-3.19393545e-01 -2.25485355e-01 -1.10907984e+00 2.54058927e-01
1.27900720e+00 -2.06690207e-01 5.69077969e-01 1.24019659e+00
-2.27893576e-01 2.95815855e-01 -1.43209231e+00 8.97593677e-01
-2.24880166e-02 -1.22864616e+00 2.32549638e-01 -9.00490507e-02
9.30073082e-01 2.58500040e-01 5.19334078e-01 9.13447857e-01
5.69029927e-01 -9.74833846e-01 5.34923375e-01 2.35676065e-01
3.50579411e-01 -5.56640446e-01 5.95648706e-01 8.93232524e-01
-8.03365290e-01 -4.97681171e-01 -4.39822555e-01 1.22751422e-01
-6.11791790e-01 2.30704427e-01 -1.30700755e+00 2.98922271e-01
7.82808900e-01 2.40913093e-01 -9.08386946e-01 1.02348793e+00
-8.13110992e-02 7.54285157e-01 -4.10044760e-01 -1.48120284e-01
4.66364712e-01 3.84491272e-02 3.69854748e-01 7.25303173e-01
7.95755506e-01 -4.54914896e-03 3.21137667e-01 8.62969041e-01
-3.58806610e-01 7.63044134e-02 -3.41986150e-01 2.12054327e-01
7.60621727e-01 1.07933903e+00 -1.66858941e-01 -3.48521680e-01
-3.77453119e-01 7.59910285e-01 8.00270498e-01 3.84157181e-01
-5.95920444e-01 1.74641088e-01 6.32514536e-01 6.08044155e-02
3.00207376e-01 1.63307041e-01 -2.51664668e-01 -1.05324447e+00
-2.28472203e-02 -1.15473402e+00 7.97264397e-01 -7.97175646e-01
-1.18739378e+00 6.41303778e-01 1.12095214e-01 -1.09898412e+00
-4.56163257e-01 -3.78660619e-01 -7.32556820e-01 7.44666934e-01
-1.59834301e+00 -1.13185406e+00 -2.57431775e-01 7.85764992e-01
9.73661900e-01 -7.02351928e-01 6.68250263e-01 9.33916420e-02
-7.87575960e-01 9.68769193e-01 2.72717506e-01 -1.98222488e-01
7.03162253e-01 -1.35406876e+00 1.98177144e-01 5.68187833e-01
3.78046274e-01 3.82087141e-01 4.28296685e-01 -2.80422419e-01
-1.56059408e+00 -1.11464858e+00 5.76874614e-01 -4.78595942e-01
5.36632657e-01 -1.65988013e-01 -1.27749979e+00 7.79349566e-01
3.66845019e-02 -5.17932465e-03 5.86848080e-01 4.40769404e-01
-3.51815313e-01 -3.98593545e-01 -8.78784359e-01 4.18143034e-01
9.29176927e-01 -6.73048720e-02 -7.60163963e-01 2.75631726e-01
8.40128362e-01 -5.69239497e-01 -6.86698854e-01 5.70089757e-01
4.10506338e-01 -6.59608066e-01 1.01053011e+00 -9.52437162e-01
1.57666534e-01 -8.92398730e-02 -4.10084836e-02 -1.75129342e+00
-2.25679368e-01 -2.08083257e-01 -5.00989735e-01 1.13716328e+00
6.14345551e-01 -5.73021114e-01 9.35815036e-01 8.21429908e-01
-3.38130653e-01 -1.22853994e+00 -1.12231266e+00 -9.83897448e-01
3.73295784e-01 -2.91171461e-01 1.17579830e+00 1.16468227e+00
-5.69674492e-01 4.98488128e-01 -1.35054290e-01 3.97850275e-01
4.27637011e-01 3.43870372e-01 1.05097234e+00 -1.60138929e+00
-5.16513169e-01 -5.04502773e-01 -3.03389654e-02 -1.11606538e+00
3.16915721e-01 -1.05606902e+00 -5.84872700e-02 -1.08628905e+00
2.87976205e-01 -9.99549747e-01 -6.18008375e-01 9.07159507e-01
-3.50332499e-01 -2.00924203e-01 1.27591416e-01 3.52097005e-01
-6.15744829e-01 7.03112006e-01 1.20259738e+00 -3.14318389e-01
-4.32494223e-01 3.28611881e-01 -6.49199843e-01 6.31697595e-01
7.29170620e-01 -6.34701192e-01 -9.58206475e-01 -7.75498331e-01
2.91626066e-01 -2.76029902e-03 4.08899099e-01 -8.86751354e-01
6.34121656e-01 -2.78789610e-01 2.91310579e-01 -3.49547178e-01
4.69845861e-01 -8.87170196e-01 1.18037842e-01 4.90525484e-01
-4.90498692e-01 2.69560069e-01 2.72469461e-01 5.03852427e-01
-6.74003959e-02 -6.62818432e-01 6.91889584e-01 -7.02665821e-02
-6.95464075e-01 5.22725403e-01 3.19183439e-01 1.13939047e-01
1.14739072e+00 -1.26699120e-01 -5.58224738e-01 -3.26465368e-02
-8.90592456e-01 6.79841936e-01 1.87305883e-01 4.98421073e-01
6.19103193e-01 -1.17978847e+00 -6.92125440e-01 2.22008109e-01
2.70330207e-03 6.43178523e-01 3.73704523e-01 5.79167604e-01
4.27443534e-02 3.21330637e-01 -1.30616471e-01 -6.60171688e-01
-1.30503047e+00 7.44260848e-01 4.22450513e-01 -5.38914204e-01
-2.91459173e-01 1.24061561e+00 3.89119148e-01 -8.54625285e-01
4.99615878e-01 -3.71977210e-01 2.36987807e-02 1.82347652e-02
5.25192559e-01 3.78438145e-01 3.71936888e-01 -1.22785971e-01
3.42967585e-02 3.04690719e-01 -4.47685748e-01 2.62309939e-01
1.54929650e+00 1.53392494e-01 4.16357756e-01 5.23987412e-01
9.33630764e-01 -5.48366725e-01 -1.37761223e+00 -7.36911118e-01
3.50824706e-02 -1.28898397e-01 3.36562544e-01 -1.15290809e+00
-1.26239777e+00 1.04459476e+00 7.36071348e-01 -1.32032990e-01
1.32927454e+00 9.75933149e-02 7.37991810e-01 8.80172312e-01
2.89597869e-01 -8.76206636e-01 4.78672743e-01 4.05812651e-01
8.35384429e-01 -1.28332615e+00 -1.23529233e-01 2.09046200e-01
-4.55748111e-01 1.08616567e+00 1.35429442e+00 2.24864930e-01
6.05350137e-01 -4.95705083e-02 -1.15127951e-01 -1.83136433e-01
-1.46004760e+00 5.41666560e-02 5.66557109e-01 3.45030487e-01
2.49924064e-02 -3.76477353e-02 3.49079043e-01 8.09150696e-01
-3.03707272e-01 -4.22051549e-03 2.27230921e-01 5.62653065e-01
-5.41668952e-01 -1.11966801e+00 -6.99020699e-02 5.23881912e-01
-1.55479237e-01 -7.10469633e-02 -3.61111194e-01 6.42570078e-01
1.38087422e-01 4.87645537e-01 5.72173521e-02 -4.54202861e-01
5.08107722e-01 6.05631828e-01 2.90716529e-01 -8.17585945e-01
-9.14623082e-01 -2.25600943e-01 -1.93746313e-01 -3.14665943e-01
-7.02453554e-02 -4.91648823e-01 -1.04268992e+00 -1.09544754e-01
-4.32143509e-01 4.07464892e-01 6.80567086e-01 8.94203484e-01
5.88604510e-01 3.74294847e-01 5.85379541e-01 -4.71036673e-01
-1.12958932e+00 -1.04482734e+00 -2.00472146e-01 5.09951971e-02
4.55281258e-01 -7.58646309e-01 -3.52500558e-01 -2.05472484e-01] | [9.442245483398438, 3.312537908554077] |
a9965d2d-0cfa-4e7a-981e-adcfc9f88a3c | analyzing-the-robustness-of-unsupervised | 2110.03509 | null | https://arxiv.org/abs/2110.03509v5 | https://arxiv.org/pdf/2110.03509v5.pdf | Analyzing the Robustness of Unsupervised Speech Recognition | Unsupervised speech recognition (unsupervised ASR) aims to learn the ASR system with non-parallel speech and text corpus only. Wav2vec-U has shown promising results in unsupervised ASR by self-supervised speech representations coupled with Generative Adversarial Network (GAN) training, but the robustness of the unsupervised ASR framework is unknown. In this work, we further analyze the training robustness of unsupervised ASR on the domain mismatch scenarios in which the domains of unpaired speech and text are different. Three domain mismatch scenarios include: (1) using speech and text from different datasets, (2) utilizing noisy/spontaneous speech, and (3) adjusting the amount of speech and text data. We also quantify the degree of the domain mismatch by calculating the JS-divergence of phoneme n-gram between the transcription of speech and text. This metric correlates with the performance highly. Experimental results show that domain mismatch leads to inferior performance, but a self-supervised model pre-trained on the targeted speech domain can extract better representation to alleviate the performance drop. | ['Yu Tsao', 'Hung-Yi Lee', 'Da-Rong Liu', 'Chan-Jan Hsu', 'Guan-Ting Lin'] | 2021-10-07 | null | null | null | null | ['unsupervised-speech-recognition'] | ['speech'] | [ 4.70626682e-01 2.76798606e-01 1.71796039e-01 -3.95940602e-01
-9.15780246e-01 -5.38439870e-01 7.68967688e-01 -4.37950492e-01
-3.08684617e-01 5.15780687e-01 7.43404150e-01 -3.60513628e-01
4.21432883e-01 -4.80799556e-01 -5.36693871e-01 -8.25367332e-01
2.90046096e-01 4.85607505e-01 -2.04521388e-01 -4.21567947e-01
-2.95417935e-01 3.55010688e-01 -1.28422940e+00 2.46958107e-01
8.04610133e-01 8.43220115e-01 2.25689277e-01 9.21441853e-01
-3.14708352e-01 9.71480191e-01 -1.10008085e+00 -2.90211976e-01
4.19779450e-01 -8.61712396e-01 -5.50942540e-01 3.70516568e-01
3.39411236e-02 -1.04988478e-01 -9.69779193e-01 1.05865395e+00
8.03589344e-01 3.44449937e-01 6.87329769e-01 -7.68212616e-01
-7.79682815e-01 9.28676307e-01 2.94840597e-02 2.45800331e-01
1.26218736e-01 2.65635699e-01 7.89454937e-01 -9.58470345e-01
3.76610160e-01 1.23378456e+00 2.17251524e-01 9.71930623e-01
-1.06811666e+00 -7.00591445e-01 -1.25089213e-01 -8.74781013e-02
-1.53688979e+00 -1.18551266e+00 9.30737317e-01 -2.88977206e-01
1.11984503e+00 3.22469145e-01 -7.81438649e-02 1.72957683e+00
-2.95797110e-01 7.00863302e-01 9.54946578e-01 -5.28335094e-01
5.32795250e-01 2.74490118e-01 -1.10267840e-01 3.20868306e-02
-4.83424246e-01 4.68409121e-01 -6.84408009e-01 7.21198991e-02
3.78219634e-01 -2.24229380e-01 -1.56841159e-01 2.06315354e-01
-1.04551089e+00 8.59310091e-01 -4.95701097e-02 4.71049756e-01
-2.32980758e-01 -2.27854967e-01 5.63910961e-01 8.29609036e-01
5.82145989e-01 3.97729516e-01 -4.98649240e-01 -3.66338074e-01
-1.11384225e+00 -3.76053005e-01 6.32173061e-01 1.13747275e+00
5.08522332e-01 1.05140972e+00 -2.13718936e-01 1.25275195e+00
1.54512838e-01 8.02314341e-01 1.16335821e+00 -4.19252872e-01
9.17239070e-01 1.32460266e-01 -3.24227482e-01 -4.44197029e-01
2.30485082e-01 -4.75091577e-01 -1.09934568e+00 -1.60291672e-01
1.73817635e-01 -3.92670780e-01 -1.31518531e+00 1.62609649e+00
-3.07227433e-01 2.84466475e-01 8.23574126e-01 7.10792422e-01
9.71414268e-01 8.96934927e-01 -2.54950583e-01 -2.78614223e-01
7.96765149e-01 -1.19159544e+00 -1.14060295e+00 -5.08838952e-01
6.39558315e-01 -7.56258428e-01 1.09396696e+00 8.49048607e-03
-8.91433299e-01 -5.97051322e-01 -1.04598069e+00 2.56117821e-01
-5.63257635e-01 3.07283521e-01 -3.04486930e-01 1.00845337e+00
-1.10096931e+00 2.69601345e-01 -6.01090312e-01 -3.46495926e-01
9.62350741e-02 1.51089758e-01 -5.04885316e-01 -6.94364891e-04
-1.43469775e+00 7.87035406e-01 4.71942902e-01 -1.89360250e-02
-1.13877094e+00 -4.10734862e-01 -9.53462303e-01 1.04412116e-01
1.55825332e-01 -1.00619853e-01 9.87376750e-01 -1.33618069e+00
-2.12377429e+00 6.33620739e-01 -1.96430206e-01 -8.01174819e-01
4.75363970e-01 -4.27631848e-02 -8.25317144e-01 -1.39975131e-01
-3.58202964e-01 2.55736560e-01 1.13851535e+00 -1.12596476e+00
-1.16290629e-01 -4.41065729e-01 -5.11297703e-01 3.10643256e-01
-4.79469538e-01 1.35005638e-01 -1.18845046e-01 -1.07845557e+00
2.27687862e-02 -8.56564820e-01 -1.48249805e-01 -8.17637026e-01
-5.90370297e-01 2.29798704e-02 9.92580056e-01 -1.09775817e+00
1.08009005e+00 -2.55181623e+00 6.80371374e-02 1.49889708e-01
-1.97693840e-01 5.31566203e-01 -3.32346648e-01 6.18806779e-01
-2.13720649e-01 3.17449868e-01 -3.47972870e-01 -5.33067107e-01
-1.20320849e-01 5.10630548e-01 -7.45014071e-01 2.74059623e-01
3.79135996e-01 6.89210713e-01 -7.64807940e-01 -1.15894012e-01
2.41097003e-01 4.84073758e-01 -1.54901490e-01 7.08804846e-01
6.43956065e-02 6.01718068e-01 -1.62977293e-01 4.54948008e-01
5.12802780e-01 2.49017030e-01 1.72541454e-01 7.14566410e-02
2.80813962e-01 8.40801656e-01 -9.15787458e-01 1.50922096e+00
-5.51525533e-01 8.25325608e-01 -8.96034110e-03 -1.16407931e+00
1.34483743e+00 7.06675231e-01 5.41569069e-02 -8.62597764e-01
2.44901683e-02 1.71928987e-01 1.98881701e-01 -2.29326218e-01
4.17992651e-01 -1.59471586e-01 5.10289967e-02 3.22579116e-01
5.19903958e-01 -2.92020768e-01 -2.22092539e-01 7.80398324e-02
1.34006310e+00 -4.86838877e-01 1.41548246e-01 6.77222088e-02
3.87400419e-01 -5.19664764e-01 2.89932311e-01 7.97919512e-01
-1.52107388e-01 9.38008010e-01 7.44435191e-02 1.82429567e-01
-1.26006818e+00 -1.28603959e+00 1.06267296e-01 9.08246219e-01
-2.58550584e-01 -1.54884592e-01 -6.88518763e-01 -7.87734985e-01
-3.07460487e-01 1.16410685e+00 -4.62919652e-01 -3.44876200e-01
-4.60225075e-01 -3.36789519e-01 1.03863800e+00 5.45326710e-01
3.60295922e-01 -1.08053315e+00 3.47675979e-01 1.21290505e-01
-1.43452615e-01 -1.37720680e+00 -5.42875350e-01 4.83485609e-01
-7.05215931e-01 -3.13340038e-01 -9.04196978e-01 -6.77955329e-01
6.26584351e-01 5.62959090e-02 8.76854539e-01 -2.72853911e-01
4.17350709e-01 3.23893487e-01 -8.99476528e-01 -1.51471555e-01
-1.19477117e+00 1.47021294e-01 4.17502016e-01 3.02631170e-01
1.57235995e-01 -6.67035818e-01 -1.34487003e-01 5.26725411e-01
-1.04333639e+00 -3.26425672e-01 4.22191113e-01 9.76195812e-01
3.38193357e-01 1.95955858e-01 7.58262992e-01 -7.79502630e-01
6.82498693e-01 -5.23299575e-01 -1.19267695e-01 2.50366688e-01
-4.58108097e-01 1.36257917e-01 1.02078855e+00 -6.48470163e-01
-1.06526804e+00 -1.77856773e-01 -5.82701325e-01 -8.27489316e-01
-3.41048270e-01 3.63705188e-01 -5.48514009e-01 4.72813994e-01
7.26927936e-01 8.05047095e-01 -6.64162403e-03 -4.88342494e-01
4.02058870e-01 1.42119610e+00 4.19581294e-01 -2.76458532e-01
1.01042521e+00 7.79668167e-02 -8.85589480e-01 -1.35986924e+00
-4.93459642e-01 -4.59226847e-01 -4.40954775e-01 4.68918234e-02
7.21015334e-01 -9.88764644e-01 2.52390742e-01 4.28229183e-01
-1.09602523e+00 -5.39863706e-01 -5.05162239e-01 5.86776078e-01
-5.39293885e-01 3.12498629e-01 -4.57488030e-01 -1.06108022e+00
-4.59168643e-01 -1.22526348e+00 8.62455547e-01 -1.50241017e-01
-1.00570098e-01 -8.73165190e-01 3.15034762e-02 4.35951710e-01
6.21803224e-01 -3.27233225e-01 6.08672798e-01 -1.43658161e+00
-3.07106972e-02 -2.12036163e-01 1.96989149e-01 1.17411959e+00
4.88872886e-01 -1.76351801e-01 -1.30789876e+00 -4.07609254e-01
2.07578853e-01 -1.99717849e-01 5.73453426e-01 7.99851343e-02
9.78725851e-01 -6.59829319e-01 1.58104584e-01 5.67973018e-01
9.95349824e-01 5.97505629e-01 1.12098849e+00 -2.76958533e-02
7.47596681e-01 5.65874577e-01 2.70417929e-01 6.77062646e-02
-2.09141850e-01 6.33165658e-01 -6.46349508e-03 -6.72452524e-02
-3.03299248e-01 -4.80681807e-01 1.00237906e+00 1.69366050e+00
3.41738135e-01 -6.88271880e-01 -9.03933346e-01 6.00155234e-01
-1.39062417e+00 -1.01164603e+00 5.26093662e-01 2.23869777e+00
8.38492930e-01 1.32639334e-01 -5.31068482e-02 2.55972594e-01
8.43659222e-01 4.17787939e-01 -5.41107416e-01 -4.38871175e-01
-4.65930074e-01 3.34714651e-01 5.13140023e-01 3.86099726e-01
-7.46010303e-01 1.03025985e+00 6.40036583e+00 1.20651472e+00
-1.33700621e+00 3.10747623e-01 5.81133068e-01 3.26939262e-02
-4.05204058e-01 -1.69813752e-01 -4.87712979e-01 8.80319118e-01
1.49771941e+00 -1.67502165e-01 5.90700448e-01 7.44385481e-01
2.42655665e-01 6.63656533e-01 -1.15087116e+00 9.78865147e-01
2.11612344e-01 -9.18059647e-01 1.47906572e-01 -8.47718790e-02
6.61489785e-01 3.74481738e-01 1.18624486e-01 5.93106747e-01
4.65516567e-01 -1.25128257e+00 5.89614809e-01 2.22657606e-01
1.07992959e+00 -6.06805503e-01 8.97553861e-01 4.67950255e-01
-8.36570382e-01 7.38048106e-02 -2.39946365e-01 4.08278197e-01
-1.14932274e-02 4.21692818e-01 -1.05744934e+00 4.78749603e-01
4.97175455e-01 4.96879011e-01 -3.36764961e-01 3.18054557e-01
-2.03455895e-01 1.28863299e+00 -2.66324192e-01 1.00916140e-01
6.50197491e-02 -3.36348772e-01 8.68350506e-01 1.42690635e+00
3.51342618e-01 -1.46024883e-01 -2.18623877e-01 7.57093489e-01
-2.65950739e-01 1.32500276e-01 -8.63526881e-01 -7.18622029e-01
8.17427278e-01 5.42155087e-01 -2.24692658e-01 -2.50659436e-01
-3.47004235e-01 1.19766843e+00 2.53186464e-01 7.99511194e-01
-4.52741504e-01 -3.52746546e-01 8.13538849e-01 8.84719789e-02
3.66012335e-01 -4.15544808e-01 -2.04908296e-01 -1.23040020e+00
-5.21516018e-02 -1.10618055e+00 -3.64759727e-03 -8.52836311e-01
-1.49484265e+00 8.94520879e-01 -4.71232712e-01 -1.30290246e+00
-5.82017779e-01 -3.14182162e-01 -5.93260050e-01 9.12225246e-01
-1.28154218e+00 -8.90666306e-01 1.46463931e-01 5.76085448e-01
1.07426465e+00 -9.63273764e-01 9.31620181e-01 2.11457133e-01
-6.10111475e-01 1.23675370e+00 5.77551007e-01 6.64262474e-01
5.47508717e-01 -1.13451743e+00 7.81037509e-01 9.76541042e-01
4.25922453e-01 4.98028368e-01 6.67713821e-01 -4.75507319e-01
-1.31655943e+00 -1.27005088e+00 6.49282515e-01 -3.37380648e-01
6.13455594e-01 -7.04099834e-01 -1.03978550e+00 6.46356285e-01
2.91137755e-01 -5.71882911e-02 8.38267624e-01 -9.51168835e-02
-4.44130510e-01 -1.37536451e-01 -1.13218176e+00 6.71512127e-01
8.98527324e-01 -1.07154357e+00 -7.87246704e-01 6.98892251e-02
1.22612166e+00 -2.60090023e-01 -9.02618647e-01 3.41566563e-01
3.64525318e-02 -6.68542027e-01 7.94733226e-01 -6.04274869e-01
4.26329076e-01 -1.04811497e-01 -7.08578587e-01 -1.68805075e+00
1.61442310e-01 -7.58236051e-01 -1.10216305e-01 1.63763297e+00
6.01205587e-01 -7.50633657e-01 5.38882077e-01 1.24269091e-01
-2.30205849e-01 -3.27022374e-01 -1.02694976e+00 -1.18435788e+00
1.48759067e-01 -4.37904656e-01 6.32704854e-01 1.07750535e+00
-3.43439430e-01 4.50444132e-01 -4.09727782e-01 2.61814415e-01
2.12468982e-01 -5.99449337e-01 8.00763547e-01 -6.23935759e-01
-2.25832701e-01 -1.49452835e-01 -5.29144526e-01 -1.07492065e+00
4.67767239e-01 -1.00028515e+00 3.26768667e-01 -1.08366263e+00
-3.24080378e-01 -2.51577616e-01 -4.48829800e-01 3.10705215e-01
1.45907784e-02 -2.58978546e-01 7.45336413e-02 2.81416088e-01
-1.44402996e-01 1.00744462e+00 7.37590551e-01 -3.81343663e-01
-3.12571406e-01 6.68102652e-02 -4.53253031e-01 2.98948318e-01
9.96207774e-01 -5.66419721e-01 -5.75985253e-01 -4.27771419e-01
-3.66152734e-01 2.29908392e-01 -6.61186427e-02 -1.03227949e+00
1.10189691e-01 7.76301250e-02 1.66067071e-02 -4.94068444e-01
3.16920727e-01 -8.04862320e-01 -1.17258906e-01 -8.12414065e-02
-7.05116272e-01 -2.56833613e-01 1.25674844e-01 6.04617655e-01
-6.08964860e-01 -3.15209508e-01 8.07570815e-01 3.30861621e-02
-4.15434510e-01 7.20158443e-02 -6.61188602e-01 3.06362301e-01
5.32521248e-01 -2.40315169e-01 -1.07099131e-01 -7.91267753e-01
-7.12656140e-01 -2.43300423e-01 2.83622026e-01 6.59636974e-01
8.31136465e-01 -1.33643556e+00 -1.01614392e+00 5.50965309e-01
2.46651307e-01 -1.38853565e-01 1.66753083e-01 2.59921253e-01
-2.42458731e-02 1.64062843e-01 2.68453628e-01 -5.09977996e-01
-1.12069631e+00 3.71733755e-01 4.21156287e-01 -2.16584235e-01
-2.96208084e-01 7.51705229e-01 1.67667449e-01 -8.90495420e-01
4.12460804e-01 6.60871267e-02 -8.79130065e-02 -2.26561129e-01
3.31131130e-01 2.78241873e-01 3.38275790e-01 -8.66941333e-01
-2.46971816e-01 5.24238497e-02 -2.17146203e-01 -3.92359138e-01
1.14049506e+00 -3.02535087e-01 3.38888973e-01 5.82857192e-01
1.37296140e+00 2.83191651e-01 -1.00499094e+00 -4.70871806e-01
-1.41244859e-01 -1.43930078e-01 1.25105977e-01 -6.32178009e-01
-1.18362379e+00 9.32381034e-01 6.27696574e-01 3.54764909e-01
1.04023612e+00 -1.86293367e-02 8.95083249e-01 3.40994596e-01
7.23620877e-02 -1.28756285e+00 2.13841781e-01 9.26450193e-01
9.07636940e-01 -1.14630151e+00 -4.91485536e-01 -1.22296453e-01
-9.93116319e-01 7.92201936e-01 4.72783357e-01 -1.28163714e-02
6.19933486e-01 5.60269177e-01 4.07533348e-01 2.08870098e-01
-6.82158589e-01 -3.91372174e-01 2.48340085e-01 9.29413676e-01
3.04451793e-01 1.43607348e-01 2.04526275e-01 7.17536807e-01
-6.55557752e-01 -5.21328568e-01 4.06861633e-01 6.21639669e-01
-3.97857912e-02 -9.73909795e-01 -3.16313863e-01 4.05043840e-01
-3.98756772e-01 -3.82531613e-01 -7.23457754e-01 3.36494833e-01
-3.89046729e-01 1.37171257e+00 1.64650500e-01 -9.03154910e-01
3.94788533e-01 3.00588459e-01 -1.94450915e-02 -7.88365364e-01
-4.94149953e-01 -7.39896251e-03 1.69268087e-01 -1.91692069e-01
-5.80520369e-02 -2.32375249e-01 -1.01501942e+00 5.42772412e-02
-4.09526944e-01 2.30351359e-01 8.77785563e-01 1.09932375e+00
3.71365607e-01 5.78891397e-01 1.20972490e+00 -4.78104144e-01
-8.88097167e-01 -1.37563002e+00 -6.53696179e-01 6.31494284e-01
5.46896577e-01 -3.04574639e-01 -9.26712632e-01 2.79639333e-01] | [14.635765075683594, 6.525681018829346] |
b08755ab-a03f-45fb-b14c-33e8804a4484 | sapphire-approaches-for-enhanced-concept-to | 2108.06643 | null | https://arxiv.org/abs/2108.06643v2 | https://arxiv.org/pdf/2108.06643v2.pdf | SAPPHIRE: Approaches for Enhanced Concept-to-Text Generation | We motivate and propose a suite of simple but effective improvements for concept-to-text generation called SAPPHIRE: Set Augmentation and Post-hoc PHrase Infilling and REcombination. We demonstrate their effectiveness on generative commonsense reasoning, a.k.a. the CommonGen task, through experiments using both BART and T5 models. Through extensive automatic and human evaluation, we show that SAPPHIRE noticeably improves model performance. An in-depth qualitative analysis illustrates that SAPPHIRE effectively addresses many issues of the baseline model generations, including lack of commonsense, insufficient specificity, and poor fluency. | ['Varun Gangal', 'Eduard Hovy', 'Chaitanya Narisetty', 'Jessica Huynh', 'Steven Y. Feng'] | 2021-08-15 | null | https://aclanthology.org/2021.inlg-1.21 | https://aclanthology.org/2021.inlg-1.21.pdf | inlg-acl-2021-8 | ['concept-to-text-generation'] | ['natural-language-processing'] | [ 5.51090002e-01 7.67140985e-01 -1.50273129e-01 -2.32240826e-01
-9.70934033e-01 -7.20173597e-01 1.10097766e+00 -5.42509463e-03
-1.48007825e-01 1.14033675e+00 8.54830921e-01 -3.60910654e-01
-3.74789760e-02 -5.92340887e-01 -5.50493777e-01 -4.36910242e-02
5.68875015e-01 1.13802719e+00 -2.11063355e-01 -7.91517675e-01
3.80390733e-01 -8.16202238e-02 -1.21922040e+00 4.68644798e-01
1.30368376e+00 3.08707625e-01 -5.57308756e-02 4.32839394e-01
-1.94214880e-01 1.23863709e+00 -1.06641865e+00 -1.08373725e+00
-2.55223691e-01 -8.53385091e-01 -1.19543707e+00 -3.33593972e-02
3.11510772e-01 -3.23322088e-01 3.69244553e-02 8.56562257e-01
6.34159803e-01 2.34570950e-01 9.39989567e-01 -1.25614440e+00
-1.37916815e+00 1.72034621e+00 -1.71524614e-01 2.88799882e-01
6.09146893e-01 3.41326416e-01 1.35555208e+00 -9.13585961e-01
8.13770950e-01 1.64073575e+00 1.03999114e+00 9.46514070e-01
-1.58289480e+00 -7.77203619e-01 -9.69511569e-02 1.62922457e-01
-1.10845196e+00 -5.58943927e-01 8.23746681e-01 -3.36565763e-01
1.22794664e+00 2.90609121e-01 8.24143708e-01 1.67098129e+00
-1.43989801e-01 7.55651653e-01 1.36563134e+00 -6.16973221e-01
2.40395084e-01 3.01042765e-01 1.03327870e-01 5.42791247e-01
3.05989921e-01 1.36078492e-01 -7.31485248e-01 -2.56699890e-01
8.13232183e-01 -7.32972443e-01 -1.75506249e-01 2.23917186e-01
-1.11687446e+00 1.11171675e+00 3.08435798e-01 1.93529367e-01
-1.86751336e-01 4.52653140e-01 3.78960222e-01 9.47923884e-02
4.69839364e-01 1.17857397e+00 -4.13911700e-01 -4.02872831e-01
-9.83777404e-01 7.33477175e-01 8.17791581e-01 1.26683295e+00
1.31233290e-01 2.54494995e-01 -7.66872048e-01 9.81886923e-01
9.29495469e-02 5.25021374e-01 6.63988411e-01 -1.12188852e+00
5.31938553e-01 3.08105826e-01 7.20303729e-02 -7.36018538e-01
-1.41355276e-01 -6.23253167e-01 -5.06982386e-01 -5.16922593e-01
5.49410954e-02 -2.40584508e-01 -1.16475260e+00 2.09489250e+00
-2.12245449e-01 -1.89208955e-01 2.42263794e-01 4.50333387e-01
1.01218665e+00 5.09891152e-01 6.02871597e-01 -7.78708681e-02
1.23248589e+00 -9.61586952e-01 -9.32678998e-01 -6.51076317e-01
4.89122003e-01 -5.00963092e-01 1.50416267e+00 8.68411884e-02
-1.47357750e+00 -1.90914333e-01 -7.77296126e-01 -2.87672937e-01
-3.83433878e-01 2.69468725e-02 1.07724929e+00 6.93170607e-01
-1.07858634e+00 5.39530754e-01 -3.32121938e-01 -3.08198482e-01
4.48067814e-01 -2.33515769e-01 1.55946627e-01 -4.31680083e-02
-1.71437705e+00 1.38617659e+00 8.28064442e-01 -1.97691858e-01
-9.15136397e-01 -1.03263164e+00 -1.02136183e+00 1.33690953e-01
3.56614053e-01 -1.29781246e+00 1.64098287e+00 -4.68279064e-01
-1.48034477e+00 7.83189058e-01 8.13967809e-02 -6.78584933e-01
5.20459652e-01 -3.52966636e-01 -2.48769656e-01 -1.00726120e-01
3.99491876e-01 1.04497838e+00 4.42522913e-01 -1.38751352e+00
-2.73792684e-01 6.93489015e-02 -6.13028780e-02 4.58359897e-01
-1.89947113e-01 1.41802624e-01 1.15337528e-01 -1.12892234e+00
-3.17921899e-02 -7.18837678e-01 -5.94867580e-02 -9.12991047e-01
-7.88293362e-01 -3.96111548e-01 1.18120741e-02 -8.96528661e-01
1.24670994e+00 -1.43600583e+00 1.58084303e-01 -3.54804024e-02
1.79314747e-01 -1.80365771e-01 -4.08202410e-02 4.05283004e-01
-8.50118548e-02 6.45496488e-01 -5.03192544e-01 -3.82423043e-01
3.39251727e-01 2.09770679e-01 -7.22496986e-01 -5.46584308e-01
3.77288401e-01 1.54161453e+00 -1.14315546e+00 -6.67918980e-01
-5.12789972e-02 4.35767621e-01 -7.55615294e-01 -2.46137261e-01
-6.70762300e-01 7.51161724e-02 3.18573043e-02 5.97363412e-01
6.43537492e-02 -2.43971676e-01 4.83144335e-02 4.60459255e-02
3.78564358e-01 1.00142026e+00 -4.30557728e-01 1.72121513e+00
-3.42384338e-01 5.67751527e-01 -6.77936494e-01 -1.47594795e-01
6.92732871e-01 2.98807949e-01 -3.75216424e-01 -4.43268061e-01
1.19321145e-01 2.93246537e-01 1.77353635e-01 -1.00018144e-01
9.92588878e-01 -9.57917750e-01 -1.52931660e-01 4.69317466e-01
3.52526635e-01 -1.01765728e+00 4.81206179e-01 6.50958419e-01
1.01160491e+00 2.48689801e-01 3.35448086e-01 -3.89170378e-01
1.73482932e-02 4.22912568e-01 4.05707628e-01 9.61813152e-01
1.77902937e-01 5.10651588e-01 3.83347690e-01 2.47641921e-01
-1.11070645e+00 -1.07766724e+00 3.76555473e-02 1.05925930e+00
-3.15274179e-01 -7.67017186e-01 -7.66472161e-01 -5.96072376e-01
-2.84050610e-02 1.87460732e+00 -8.00759792e-01 -3.20269406e-01
-3.33516866e-01 -6.42359316e-01 1.07836962e+00 7.72789299e-01
3.45347226e-01 -1.58564079e+00 -4.46343154e-01 4.95640934e-01
-6.23187482e-01 -1.12987244e+00 -1.92889139e-01 5.56801073e-02
-8.64498615e-01 -5.09616375e-01 -5.45480251e-01 -5.15543878e-01
2.54549652e-01 -4.49788272e-02 1.79162014e+00 -1.46341875e-01
-4.30677384e-02 1.66486815e-01 -5.05695343e-01 -6.35200918e-01
-9.22997117e-01 -5.31411842e-02 -3.53165895e-01 -1.09043717e+00
2.93861896e-01 -7.31321037e-01 1.12000499e-02 -2.49470279e-01
-6.42008364e-01 4.43002552e-01 4.83205169e-01 1.04020846e+00
2.28289321e-01 -1.11405971e-02 7.79733002e-01 -1.20375907e+00
1.42560172e+00 -2.63793737e-01 1.53003737e-01 3.83385599e-01
-8.00384283e-01 -7.08652614e-03 3.23577493e-01 -1.62742704e-01
-1.33489287e+00 -7.21475422e-01 -3.95892151e-02 -1.40264168e-01
1.92552909e-01 7.01321423e-01 4.65964079e-02 5.12304664e-01
1.01238358e+00 2.70460069e-01 -1.99278489e-01 -2.23422185e-01
9.38356757e-01 3.09882790e-01 7.75794804e-01 -9.23149943e-01
7.22436965e-01 -4.40377258e-02 -4.26377326e-01 -5.08583665e-01
-1.21731138e+00 1.97310075e-01 -1.17398344e-01 1.20805286e-01
5.69603682e-01 -9.08138454e-01 -6.79537430e-02 6.31257072e-02
-1.57679355e+00 -5.73313296e-01 -7.66398907e-01 9.85722896e-03
-7.41838455e-01 1.28721993e-03 -7.67609000e-01 -6.67748392e-01
-6.65962338e-01 -6.65397346e-01 8.40731144e-01 4.93314005e-02
-9.96321738e-01 -1.26330113e+00 8.42445791e-02 7.16080546e-01
4.29104954e-01 1.58652678e-01 1.42175066e+00 -8.08569610e-01
-9.01472420e-02 2.70906001e-01 -3.42645459e-02 3.57713252e-01
-5.27576022e-02 1.26822097e-02 -8.06268871e-01 3.32544565e-01
-1.24803975e-01 -7.15336800e-01 9.83692408e-01 2.61851370e-01
7.90355086e-01 -6.83809280e-01 -1.35377169e-01 4.17059392e-01
1.04361165e+00 1.41045526e-01 6.37611687e-01 2.14802071e-01
6.60761654e-01 4.55289066e-01 5.13527632e-01 4.52407449e-01
6.34223700e-01 3.89809102e-01 -1.09780543e-01 1.52028427e-01
-4.92839634e-01 -8.35844040e-01 3.74498963e-01 7.03948736e-01
-9.13658515e-02 -3.27786207e-01 -1.13487446e+00 8.02062094e-01
-1.65172994e+00 -1.27227259e+00 7.64626414e-02 1.44365323e+00
1.57446563e+00 3.10210019e-01 4.38076518e-02 1.25874445e-01
7.36842215e-01 -2.60859653e-02 -1.97318330e-01 -6.22081041e-01
-4.19932276e-01 6.54309332e-01 -5.54861985e-02 6.13572419e-01
-6.36490285e-01 1.69000494e+00 7.64075899e+00 9.83081222e-01
-2.68991917e-01 3.22264433e-01 5.33960700e-01 -1.18540034e-01
-1.08298111e+00 1.54839739e-01 -6.47704065e-01 2.43273929e-01
7.41339624e-01 -5.61836541e-01 5.38097680e-01 7.78470516e-01
-2.30996087e-01 2.94630915e-01 -1.21798801e+00 6.52004659e-01
2.87602603e-01 -1.59463334e+00 6.23520732e-01 -3.68668616e-01
1.04462457e+00 -3.99997294e-01 -8.40678290e-02 8.06201577e-01
8.77665639e-01 -1.26551688e+00 1.10505223e+00 1.38205364e-01
7.72829771e-01 -8.64677191e-01 5.67848146e-01 8.50444585e-02
-4.13571566e-01 -8.27751867e-03 -4.23804447e-02 -1.94475070e-01
4.85040784e-01 5.84357202e-01 -1.29307044e+00 2.83484101e-01
1.97291061e-01 4.11046863e-01 -7.54492760e-01 4.11486238e-01
-1.11953712e+00 8.50433588e-01 -2.66440231e-02 -2.19684005e-01
1.57313049e-01 1.33188576e-01 7.69226193e-01 1.56992614e+00
7.36882389e-02 3.49838644e-01 -8.01944882e-02 1.71444118e+00
-1.35237351e-01 -1.99059442e-01 -5.48028588e-01 -5.18795073e-01
9.60341752e-01 8.65107536e-01 -5.73334038e-01 -7.14478672e-01
4.30930704e-01 9.48423803e-01 3.03747416e-01 2.86993295e-01
-1.02974188e+00 -4.30564493e-01 2.32051641e-01 -7.38624707e-02
6.92821592e-02 -4.69870828e-02 -8.53321433e-01 -8.34336638e-01
-2.94186860e-01 -1.16587186e+00 2.66437292e-01 -1.16252649e+00
-1.49375403e+00 5.31651080e-01 3.39014620e-01 -2.91259527e-01
-7.59438515e-01 -1.41941264e-01 -6.07516706e-01 8.59512508e-01
-1.05646753e+00 -1.44546008e+00 -1.48409978e-01 3.71633977e-01
5.82599580e-01 -1.55506149e-01 9.14401054e-01 -3.65304351e-01
-4.42480147e-01 7.01290667e-01 -5.81049144e-01 -7.22381845e-02
4.17137206e-01 -1.63485539e+00 1.07398999e+00 8.12250078e-01
1.76963821e-01 1.03302956e+00 1.08816516e+00 -1.08131826e+00
-6.47973061e-01 -9.19446945e-01 1.10588992e+00 -8.40596616e-01
8.42633069e-01 -3.53172988e-01 -5.23087859e-01 1.04856277e+00
4.11284834e-01 -9.33155894e-01 6.93165064e-01 4.17383820e-01
-5.47539055e-01 5.50151527e-01 -1.18011248e+00 1.07967281e+00
1.40593874e+00 -5.52850127e-01 -1.42874014e+00 6.18517399e-01
1.25389111e+00 -3.69525582e-01 -6.18491173e-01 3.11356246e-01
2.55601227e-01 -7.06911683e-01 7.76754200e-01 -8.99330139e-01
1.15389407e+00 1.64695144e-01 -7.73706585e-02 -1.80437565e+00
-5.80649018e-01 -1.19758379e+00 -2.36755326e-01 1.58125043e+00
6.92869127e-01 -4.57421333e-01 3.67953956e-01 8.39537680e-01
-4.19257611e-01 -4.23886359e-01 -7.32096195e-01 -8.96126270e-01
3.97368759e-01 -4.30727690e-01 7.32763112e-01 1.20733821e+00
3.70343447e-01 1.08331633e+00 3.88689786e-02 -5.32488763e-01
5.32768190e-01 -1.28619537e-01 3.85228634e-01 -1.04564500e+00
-5.10695219e-01 -5.96809328e-01 1.24047906e-03 -5.63388288e-01
4.03017640e-01 -1.07179749e+00 2.82216907e-01 -1.92184222e+00
5.33926725e-01 -4.59722400e-01 -2.19552591e-02 8.06343138e-01
-4.70281303e-01 2.13344395e-01 2.58090496e-01 -4.78192680e-02
-1.24120936e-01 8.13840687e-01 1.32037425e+00 -1.28419057e-01
-4.52254713e-01 -5.23345351e-01 -1.49987817e+00 6.21919692e-01
1.01550984e+00 -4.68097657e-01 -7.20534503e-01 -4.60259169e-01
4.13306415e-01 -1.89295962e-01 6.73657417e-01 -7.53018558e-01
-1.18275493e-01 -3.23865682e-01 2.83809096e-01 -4.24408704e-01
5.51768243e-01 6.41648173e-02 6.00859188e-02 2.17292041e-01
-7.58107185e-01 2.78160810e-01 4.75668162e-01 2.18727261e-01
1.04915112e-01 -3.42908621e-01 3.95838141e-01 -3.94296616e-01
-5.31086028e-01 -5.31271577e-01 9.09073558e-03 7.83368945e-01
5.68025827e-01 -1.61891028e-01 -7.77043045e-01 -5.34660161e-01
-4.22642678e-01 4.43122797e-02 2.61013955e-01 3.60751688e-01
6.62815273e-01 -1.39879799e+00 -1.08283639e+00 -3.16662997e-01
1.55672848e-01 -2.01773584e-01 -6.78070784e-02 4.99379635e-01
-1.87920496e-01 6.74080729e-01 -1.38508081e-01 -1.09406307e-01
-8.92543316e-01 2.46368065e-01 1.62260488e-01 -3.20888609e-01
-6.47219956e-01 1.32352865e+00 -2.38506675e-01 -4.70097572e-01
-5.09712435e-02 -3.53223950e-01 -3.27095352e-02 -5.98054752e-02
2.36988604e-01 5.04164279e-01 -1.55190304e-01 -2.59049505e-01
-2.26772025e-01 -2.01261342e-01 -3.97044897e-01 -1.07597101e+00
1.21367025e+00 2.18863711e-01 -2.20070481e-01 4.61343944e-01
2.23246232e-01 -4.12664823e-02 -7.38255799e-01 1.76709607e-01
-7.16428235e-02 -8.78416672e-02 4.68931273e-02 -1.80312634e+00
-2.31557265e-01 5.80399930e-01 -4.07123357e-01 7.92374387e-02
6.89888179e-01 9.73078683e-02 7.20331609e-01 4.85062093e-01
2.79487222e-01 -1.19515133e+00 1.05792873e-01 9.31659520e-01
1.35142708e+00 -6.95986271e-01 9.95555446e-02 -4.57822442e-01
-1.17762029e+00 5.25354624e-01 7.45674849e-01 2.84410894e-01
-5.74375615e-02 1.15450129e-01 -1.71927392e-01 -3.86849642e-01
-1.00916064e+00 -1.06747858e-01 7.84649178e-02 6.64606631e-01
7.75467277e-01 3.02681148e-01 -5.02241135e-01 9.81446028e-01
-1.27319384e+00 6.43752422e-03 6.75104082e-01 8.82532835e-01
-2.68563628e-01 -8.38034391e-01 -1.99253082e-01 4.11425978e-01
-9.60291103e-02 -8.72969210e-01 -1.10589111e+00 1.07957876e+00
2.53823493e-02 1.13696647e+00 -8.88511240e-02 -3.44479442e-01
1.58587396e-01 5.57534277e-01 9.86810625e-01 -9.84044671e-01
-9.10647452e-01 -2.72437990e-01 6.38622582e-01 -1.16678625e-01
-5.27956411e-02 -5.65654397e-01 -1.31799543e+00 -3.88145566e-01
-5.19000530e-01 1.73334748e-01 5.53044498e-01 9.94127274e-01
2.98774332e-01 7.35649765e-01 -3.71512435e-02 -3.74943823e-01
-1.04651916e+00 -1.49261272e+00 -3.05418581e-01 5.55160046e-01
-1.08894326e-01 -6.69739425e-01 -5.02620220e-01 3.45790642e-03] | [11.289871215820312, 8.754678726196289] |
b0ecd14a-de2d-466f-9e9c-ff3f38983e0d | multisubs-a-large-scale-multimodal-and | 2103.01910 | null | https://arxiv.org/abs/2103.01910v3 | https://arxiv.org/pdf/2103.01910v3.pdf | MultiSubs: A Large-scale Multimodal and Multilingual Dataset | This paper introduces a large-scale multimodal and multilingual dataset that aims to facilitate research on grounding words to images in their contextual usage in language. The dataset consists of images selected to unambiguously illustrate concepts expressed in sentences from movie subtitles. The dataset is a valuable resource as (i) the images are aligned to text fragments rather than whole sentences; (ii) multiple images are possible for a text fragment and a sentence; (iii) the sentences are free-form and real-world like; (iv) the parallel texts are multilingual. We set up a fill-in-the-blank game for humans to evaluate the quality of the automatic image selection process of our dataset. We show the utility of the dataset on two automatic tasks: (i) fill-in-the-blank; (ii) lexical translation. Results of the human evaluation and automatic models demonstrate that images can be a useful complement to the textual context. The dataset will benefit research on visual grounding of words especially in the context of free-form sentences, and can be obtained from https://doi.org/10.5281/zenodo.5034604 under a Creative Commons licence. | ['Lucia Specia', 'Chiraag Lala', 'Josiel Figueiredo', 'Pranava Madhyastha', 'Josiah Wang'] | 2021-03-02 | null | https://aclanthology.org/2022.lrec-1.730 | https://aclanthology.org/2022.lrec-1.730.pdf | lrec-2022-6 | ['multimodal-lexical-translation', 'multimodal-text-prediction'] | ['natural-language-processing', 'natural-language-processing'] | [ 1.93840712e-01 -4.17434685e-02 -1.76648259e-01 -4.22319680e-01
-8.02802980e-01 -7.66701698e-01 9.77303624e-01 3.99139941e-01
-6.95879519e-01 6.59710169e-01 4.61277068e-01 -4.33655024e-01
1.98838532e-01 -4.03716296e-01 -8.18690479e-01 -3.08880687e-01
2.76127040e-01 5.67372501e-01 2.62199581e-01 -4.90999609e-01
2.34257698e-01 3.68646026e-01 -1.53680742e+00 7.57889867e-01
4.79680330e-01 6.40117705e-01 8.03224981e-01 4.78047520e-01
-3.58720839e-01 6.21442497e-01 -4.84687239e-01 -5.17363966e-01
1.09326981e-01 -5.27308166e-01 -7.88106561e-01 4.14880782e-01
8.38201761e-01 -1.11551121e-01 -5.82764521e-02 1.06851399e+00
3.41878146e-01 -1.38565496e-01 4.17048126e-01 -1.14773595e+00
-7.61036873e-01 3.85582924e-01 -4.98535216e-01 4.44699861e-02
7.55402982e-01 1.70720756e-01 8.82363856e-01 -1.07593346e+00
1.21537495e+00 1.16904056e+00 1.28392816e-01 3.00144017e-01
-9.82405424e-01 -3.23649585e-01 7.79208764e-02 1.37751594e-01
-1.49648964e+00 -5.13865769e-01 3.46075594e-01 -6.55531406e-01
9.04707193e-01 3.75894725e-01 6.55593216e-01 1.10970902e+00
1.45673871e-01 5.77723384e-01 1.29245114e+00 -9.59143639e-01
-1.81198612e-01 5.90477765e-01 -2.77997702e-01 7.82348156e-01
2.03880459e-01 -4.35600579e-02 -7.49891996e-01 2.49948978e-01
7.85355926e-01 -2.91185021e-01 -1.32465988e-01 -1.81934059e-01
-1.50765491e+00 8.05050850e-01 2.66679108e-01 6.00077868e-01
-1.64116442e-01 -9.49105546e-02 4.65696961e-01 3.67452234e-01
3.09602588e-01 2.42731035e-01 -1.31653771e-02 4.02195118e-02
-9.18288112e-01 3.24782759e-01 5.12538671e-01 9.76978898e-01
7.91785121e-01 -4.42699522e-01 1.56552251e-02 9.00033891e-01
2.71119088e-01 6.91394031e-01 4.57556158e-01 -7.10361779e-01
5.81528544e-01 4.27122831e-01 1.48964554e-01 -1.09763336e+00
-2.59275705e-01 -3.09697147e-02 -3.25509220e-01 1.20713383e-01
5.29455066e-01 1.14013568e-01 -7.07584321e-01 1.54623997e+00
1.38082862e-01 -5.99881053e-01 5.17471172e-02 9.94329453e-01
9.56999481e-01 7.08806992e-01 1.48191139e-01 -2.11320832e-01
1.82759464e+00 -7.07690060e-01 -8.57192814e-01 -3.60202521e-01
5.75787961e-01 -1.32350302e+00 1.44295180e+00 1.46445453e-01
-1.06801212e+00 -4.68644381e-01 -9.76777494e-01 -2.43746296e-01
-6.70288444e-01 3.26046020e-01 2.85265326e-01 4.55535889e-01
-1.09186029e+00 -3.39016654e-02 -3.48725438e-01 -8.53634834e-01
1.26897722e-01 -1.69337839e-01 -7.04352200e-01 -9.59022865e-02
-1.08652616e+00 1.10229492e+00 4.69938219e-01 -9.33052376e-02
-4.74459678e-01 -8.67071226e-02 -9.75456476e-01 -5.78014612e-01
5.23264587e-01 -4.97353762e-01 1.01999664e+00 -1.39245081e+00
-8.44094157e-01 1.77718627e+00 -3.26392174e-01 -2.28158683e-01
4.76478428e-01 -6.17141724e-02 -4.22854722e-01 6.66761637e-01
6.28181279e-01 9.64753091e-01 7.05178022e-01 -1.34991658e+00
-7.84023106e-01 -3.34067136e-01 2.22901866e-01 3.11606407e-01
-1.58005729e-01 5.29719114e-01 -6.39531374e-01 -8.18761170e-01
-1.12515964e-01 -9.75907803e-01 2.78829128e-01 7.59188309e-02
-2.66243815e-01 1.22423850e-01 5.66684484e-01 -8.33988607e-01
1.01978600e+00 -1.94798136e+00 -3.70708406e-02 -1.40764564e-01
-1.29893526e-01 -7.05842823e-02 -3.82294506e-02 9.91130114e-01
5.80267236e-02 3.41302544e-01 5.86254112e-02 -1.69185385e-01
2.69301198e-02 1.52334094e-01 -1.42259732e-01 4.44984883e-01
2.47685146e-02 9.06584799e-01 -7.11220920e-01 -9.06705260e-01
2.02930152e-01 2.14219317e-01 7.36320615e-02 -6.85799122e-02
-1.66988447e-01 4.08252388e-01 -3.76122706e-02 5.59100330e-01
1.84707269e-01 2.28768084e-02 1.07222818e-01 -4.46128100e-01
-5.04041612e-01 -1.51361758e-02 -9.99809086e-01 1.76730931e+00
-4.32380617e-01 1.04582596e+00 -2.01733802e-02 -4.81540203e-01
5.47108889e-01 3.76158237e-01 3.71679217e-02 -9.59561467e-01
1.06902845e-01 2.77847052e-01 -8.22572485e-02 -8.61384988e-01
8.12495530e-01 -2.71975338e-01 -2.29147434e-01 3.98837477e-01
-1.73300896e-02 -4.65990275e-01 7.90335536e-01 5.05825162e-01
2.95180440e-01 2.13282958e-01 3.26570243e-01 -4.01618838e-01
3.35287988e-01 1.68729141e-01 -5.57866730e-02 6.46112919e-01
-4.31873947e-02 5.78048527e-01 5.43093741e-01 -3.36910903e-01
-1.30033505e+00 -9.13643658e-01 -3.06028187e-01 1.04190445e+00
2.28734672e-01 -4.68428642e-01 -7.05112457e-01 -2.71418631e-01
-4.31559861e-01 8.97631645e-01 -6.61530077e-01 4.71122503e-01
-1.79082185e-01 -1.18167773e-01 3.98029059e-01 2.64173865e-01
2.09077075e-01 -1.10985482e+00 -8.65148544e-01 -1.49449453e-01
-6.02430582e-01 -1.53802836e+00 -6.62903607e-01 -8.84296149e-02
-3.50099206e-01 -1.08082604e+00 -6.82737529e-01 -1.07888520e+00
7.23248065e-01 3.16845447e-01 1.08720195e+00 7.00334087e-02
-1.61763772e-01 6.75564408e-01 -6.46239817e-01 -4.51662511e-01
-5.68188250e-01 -4.16334867e-01 -1.54679259e-02 -1.31346911e-01
2.94388443e-01 -2.49385815e-02 -2.74148196e-01 2.96501815e-01
-1.22668779e+00 5.12575686e-01 4.69684720e-01 6.16398752e-01
6.35230541e-01 -2.45808691e-01 1.50394633e-01 -7.23533154e-01
5.88389158e-01 -1.68539986e-01 -5.21373689e-01 4.23652858e-01
-1.44795209e-01 -1.98997691e-01 -5.36397807e-02 -3.12650114e-01
-8.38560641e-01 -2.21233759e-02 1.31255761e-01 -1.36744529e-01
-2.38268733e-01 5.85087240e-01 3.34247984e-02 1.81842014e-01
6.66750550e-01 5.77630587e-02 3.27323154e-02 -4.00024615e-02
6.12321854e-01 6.68175876e-01 6.13065124e-01 -6.66997910e-01
5.22023976e-01 5.50753236e-01 -1.10663407e-01 -1.23347664e+00
-5.92288911e-01 -3.48125160e-01 -8.14788818e-01 -6.89226091e-01
1.16967726e+00 -1.02140963e+00 -4.88714799e-02 2.41220117e-01
-1.07795763e+00 -2.78358668e-01 -1.24324612e-01 4.55712497e-01
-4.15743411e-01 4.15974617e-01 -3.35024148e-01 -6.16929233e-01
5.17968573e-02 -1.17623460e+00 1.17977369e+00 1.20667890e-01
-4.69889611e-01 -1.00307584e+00 -2.76906669e-01 4.54245299e-01
5.52716777e-02 4.46538478e-01 8.09408128e-01 -5.24317741e-01
-2.46736899e-01 -2.37194851e-01 -2.88866311e-01 4.88154292e-02
5.14080515e-03 2.76333243e-01 -7.31422663e-01 -1.15115419e-01
-2.07344055e-01 -6.87604308e-01 4.71395314e-01 1.36233687e-01
3.87918651e-01 -2.73314059e-01 -3.30074392e-02 -1.87057462e-02
1.42637420e+00 1.54505044e-01 6.96982622e-01 5.91076195e-01
4.80058432e-01 1.03258359e+00 6.91028416e-01 8.47511441e-02
3.50673139e-01 1.04950976e+00 2.27786347e-01 -3.01401466e-01
-1.91073611e-01 -3.07345420e-01 3.28501672e-01 7.60728717e-01
1.35181531e-01 -1.75122142e-01 -1.18044436e+00 7.00742602e-01
-1.64007866e+00 -9.28637803e-01 -4.33809668e-01 2.10477686e+00
7.99656749e-01 4.16897051e-02 2.37863213e-01 -2.12380216e-01
8.61045539e-01 3.51580791e-02 2.91505575e-01 -4.64762002e-01
-5.82719088e-01 -9.09857526e-02 1.61632329e-01 6.31880403e-01
-9.46902394e-01 1.02084386e+00 5.33888865e+00 8.17289412e-01
-1.20280683e+00 1.96847379e-01 4.86905426e-01 -1.96129475e-02
-3.00472796e-01 8.56050104e-02 -6.58661783e-01 3.80434185e-01
6.16825044e-01 6.80527091e-02 1.26903668e-01 3.95993352e-01
4.12913620e-01 -5.11725247e-01 -9.73712385e-01 8.54511380e-01
4.86793816e-01 -1.28019643e+00 1.72623336e-01 -1.05206810e-01
5.63531518e-01 -8.03590640e-02 9.44407284e-02 -1.49173781e-01
-1.83073983e-01 -9.69735742e-01 1.41025209e+00 3.45327973e-01
1.18897784e+00 -4.23158437e-01 6.88062966e-01 1.14125140e-01
-9.51197386e-01 3.27329546e-01 -2.44201407e-01 -1.10438457e-02
4.35863137e-01 1.18022271e-01 -7.08981156e-01 4.52886462e-01
6.56091094e-01 3.27720314e-01 -9.55147862e-01 7.12211192e-01
-4.36749220e-01 1.18110545e-01 -8.56006369e-02 -3.06961387e-01
3.92177671e-01 -1.79886609e-01 5.55724084e-01 1.59900105e+00
1.96493670e-01 -9.70654264e-02 2.54916877e-01 5.56738019e-01
1.02586746e-01 5.70405185e-01 -7.17907250e-01 -4.02986795e-01
3.75621825e-01 1.16362762e+00 -1.18587613e+00 -1.62488326e-01
-6.92421317e-01 9.88888562e-01 5.91195785e-02 2.74254531e-01
-5.88825166e-01 -3.46497744e-01 -6.84387311e-02 5.34444928e-01
2.93651342e-01 -1.13875665e-01 -6.58159656e-03 -1.10414040e+00
2.36885041e-01 -1.09524596e+00 2.40607455e-01 -1.47122109e+00
-1.10541272e+00 8.16891670e-01 4.10701454e-01 -1.17620039e+00
-1.80131629e-01 -7.17993617e-01 -1.82396874e-01 7.63257682e-01
-9.50399578e-01 -1.57386100e+00 -1.67599067e-01 5.01520455e-01
6.21780872e-01 -1.49547249e-01 7.67205298e-01 1.67354330e-01
-1.48964405e-01 1.79861590e-01 -3.61624137e-02 6.52944073e-02
8.74678671e-01 -9.71535981e-01 -1.63758308e-01 9.40462351e-01
6.53910816e-01 7.42275178e-01 9.30846393e-01 -6.38556242e-01
-9.37680721e-01 -6.08429730e-01 1.23610795e+00 -4.59134042e-01
9.61482942e-01 -4.68645364e-01 -5.58544040e-01 6.73311472e-01
8.49260032e-01 -5.76832294e-01 7.43552446e-01 -1.30019113e-01
-2.50792295e-01 1.44751668e-01 -9.75470483e-01 7.27694035e-01
5.35532594e-01 -8.28881443e-01 -7.39453733e-01 6.70376897e-01
3.98501784e-01 -5.87798059e-01 -5.41646302e-01 -3.38576846e-02
8.59931767e-01 -8.52313757e-01 7.10093856e-01 -4.02786106e-01
9.57419395e-01 -3.72029454e-01 -4.07434911e-01 -8.29348385e-01
1.63326934e-02 -4.06262964e-01 6.42450452e-01 1.29397070e+00
5.99153876e-01 -3.43805403e-01 1.14936307e-01 3.76301289e-01
-6.45388812e-02 -5.62489927e-01 -9.21482325e-01 -6.06449783e-01
7.17897862e-02 -5.59558809e-01 1.17846154e-01 9.81949806e-01
8.30728412e-02 4.85434502e-01 -2.17962205e-01 -1.59660146e-01
1.67436123e-01 -1.86759587e-02 6.85111761e-01 -7.48517573e-01
-1.18136987e-01 -5.21035552e-01 -4.53330904e-01 -7.20936179e-01
-4.96642152e-03 -7.51393497e-01 -1.23932116e-01 -1.66820049e+00
3.54546010e-01 -5.16066253e-02 2.78038502e-01 5.10965765e-01
-1.03362598e-01 4.93444473e-01 5.96456945e-01 3.95018399e-01
-6.55557454e-01 -5.61217731e-03 1.21566057e+00 -4.45654765e-02
8.85504857e-02 -5.81547320e-01 -7.00488806e-01 7.03897536e-01
6.47823393e-01 -3.56225938e-01 -4.04269725e-01 -4.49547023e-01
4.84064221e-01 2.28327159e-02 4.38879132e-01 -4.69536275e-01
5.42796776e-02 -2.83053339e-01 2.67704964e-01 -4.84702736e-01
4.00994807e-01 -7.46220767e-01 1.76071942e-01 3.38161856e-01
-2.90246695e-01 4.65458661e-01 4.99839187e-01 2.69933969e-01
-2.21224949e-01 -5.58886707e-01 7.00553536e-01 -4.58751291e-01
-9.57405448e-01 -4.24593091e-01 -5.41051626e-01 2.43933484e-01
1.06443501e+00 -3.64500463e-01 -4.79983091e-01 -6.54550791e-01
-6.45390093e-01 1.25146434e-01 8.93918931e-01 5.31538367e-01
5.17416894e-01 -1.29914224e+00 -8.60715270e-01 -2.19451234e-01
5.17330527e-01 -4.73015487e-01 4.19830391e-03 8.68269026e-01
-9.76196766e-01 5.75515211e-01 -3.37490976e-01 -6.18685901e-01
-1.49763620e+00 5.47229946e-01 -9.14331805e-03 6.56541213e-02
-4.78987455e-01 7.50808418e-01 3.12131971e-01 -4.75331284e-02
1.13103412e-01 6.59204787e-03 -1.77173406e-01 4.53412801e-01
4.89071727e-01 -1.08972467e-01 -1.55955553e-01 -1.45556593e+00
-2.95530111e-01 3.98384631e-01 3.46933343e-02 -7.96868265e-01
1.11601150e+00 -5.13609946e-01 -4.41643268e-01 8.96620095e-01
1.07556057e+00 3.65311325e-01 -5.81623197e-01 -8.52176175e-02
-7.97160491e-02 -5.95224917e-01 -2.54471213e-01 -9.99663711e-01
-5.60097277e-01 7.26631999e-01 5.13045669e-01 2.33848348e-01
9.20994461e-01 3.10587227e-01 4.30769652e-01 1.07076138e-01
2.56833166e-01 -1.13505709e+00 1.37497738e-01 3.20954293e-01
1.38954854e+00 -1.20968866e+00 2.77781904e-01 -3.18971336e-01
-1.10856116e+00 1.25096464e+00 2.48665765e-01 3.19073647e-01
1.09513916e-01 -1.36907184e-02 4.20199007e-01 -4.15751308e-01
-4.72599596e-01 -3.36512476e-01 4.99428064e-01 5.86933851e-01
7.33656406e-01 3.33474539e-02 -7.43649483e-01 3.06920707e-01
-3.36667866e-01 -3.69238913e-01 5.77872157e-01 9.29552615e-01
-2.39464924e-01 -1.14219701e+00 -5.80375791e-01 1.03752822e-01
-4.87778842e-01 -2.67931223e-01 -7.73229063e-01 1.19364309e+00
3.63124460e-01 1.11961770e+00 -4.48167101e-02 4.66663763e-02
2.30918944e-01 1.15125649e-01 7.40865290e-01 -6.02547646e-01
-4.05493170e-01 3.17448318e-01 4.53985482e-01 -1.95565179e-01
-6.93892777e-01 -8.15252900e-01 -9.26165402e-01 -3.99700105e-02
-1.75269410e-01 5.23496084e-02 9.16979671e-01 9.39810395e-01
8.76609460e-02 2.66766638e-01 -1.72516087e-03 -8.76448452e-01
1.87057346e-01 -9.18383718e-01 -4.62290347e-01 6.42552912e-01
4.12876159e-02 -5.14634788e-01 -1.25395253e-01 6.18343174e-01] | [11.198681831359863, 1.3080523014068604] |
3fb9eb4a-25fd-4157-8cac-68fd01208e9e | a-method-to-disambiguate-a-word-by-using | null | null | https://aclanthology.org/2021.icon-main.47 | https://aclanthology.org/2021.icon-main.47.pdf | A Method to Disambiguate a Word by Using Restricted Boltzmann Machine | Finding the correct sense of a word is of great importance in many textual data related applications such as information retrieval, text mining and natural language processing. We have proposed one novel Word Sense Disambiguation (WSD) method according to its context. Based on collocation extraction score, the proposed method extracts three different features for each sense definition of a target word. These features create a feature vector and all the feature vectors create a sense matrix. Here, Restricted Boltzmann Machine (RBM) is used to enhance the sense matrix. Comparison of the proposed WSD method is made with current state-of-the-art systems using SENSEVAL and Sem Eval datasets. The proposed WSD method shows the practical implementation by applying on query-based text summary. For evaluation on query-based text summary, the proposed WSD method uses DUC datasets containing news-wire articles. Finally, the experimental analysis shows that our proposed WSD method performs better as compared to the current systems. | ['Bhogeswar Borah', 'Nazreena Rahman'] | null | null | null | null | icon-2021-12 | ['word-sense-disambiguation'] | ['natural-language-processing'] | [ 3.10776144e-01 -2.45208800e-01 -1.45191342e-01 -1.12282075e-01
-7.38305211e-01 -5.88421702e-01 8.64205122e-01 8.39434266e-01
-1.17578876e+00 9.40204024e-01 6.07820451e-01 -1.76010013e-01
-2.53203422e-01 -6.79206908e-01 2.65105013e-02 -3.79491329e-01
2.89005727e-01 4.16252613e-01 4.65243816e-01 -6.85853601e-01
9.17283058e-01 1.12696871e-01 -1.28541017e+00 1.28003523e-01
8.83985043e-01 7.42975354e-01 9.75377917e-01 2.59892762e-01
-6.71837568e-01 1.42625257e-01 -7.33561635e-01 3.29072773e-02
1.56355500e-01 -1.86194241e-01 -1.07303607e+00 -4.25752521e-01
-5.50976805e-02 4.87352729e-01 2.01106980e-01 1.20311165e+00
5.26336133e-01 7.13783860e-01 7.44360149e-01 -7.87449419e-01
-6.09048009e-01 9.04540181e-01 -8.32995534e-01 5.02906740e-01
5.93758702e-01 -7.23086596e-01 1.23302054e+00 -8.76602471e-01
6.83506727e-01 1.21544731e+00 4.20383722e-01 3.13026696e-01
-9.78412211e-01 -6.33856535e-01 7.94748962e-02 4.03595716e-01
-1.39103734e+00 2.55612358e-02 8.78053844e-01 -3.10059845e-01
1.46703303e+00 2.73283929e-01 7.67829418e-01 7.68994570e-01
4.46857154e-01 8.05528164e-01 1.34363008e+00 -8.03571820e-01
4.88081396e-01 2.00889692e-01 1.04677916e+00 3.32494617e-01
4.66720492e-01 -4.50579494e-01 -5.49125433e-01 -4.81888473e-01
9.35700014e-02 -1.59264132e-01 -1.69415653e-01 6.01741746e-02
-1.20696723e+00 8.01433802e-01 2.84444317e-02 9.35662746e-01
-6.32853389e-01 -2.70978212e-01 4.13917065e-01 6.96662068e-02
5.45584857e-01 6.01666808e-01 -5.45967340e-01 -2.01261133e-01
-1.06369269e+00 5.04531205e-01 6.78175807e-01 7.44284034e-01
8.04100990e-01 -3.48455220e-01 -1.48620158e-01 1.11248577e+00
4.49382365e-01 5.19612670e-01 1.23613441e+00 8.35065320e-02
2.79630452e-01 7.89210558e-01 4.51508194e-01 -1.42200768e+00
-6.57265782e-01 -3.52503031e-01 -7.18365908e-01 -3.31109196e-01
-4.09106553e-01 -2.03944221e-01 -8.16403687e-01 1.54155254e+00
4.42781836e-01 3.78311187e-01 5.76323867e-01 7.97992885e-01
9.58889961e-01 1.01451838e+00 2.17854857e-01 -4.23465133e-01
1.76232922e+00 -2.93107599e-01 -7.51759350e-01 -5.59770226e-01
4.16744083e-01 -1.17138898e+00 9.88085389e-01 5.74113667e-01
-3.92691374e-01 -3.46006602e-01 -1.48174298e+00 2.39049003e-01
-7.62151003e-01 1.30692497e-01 4.70755398e-01 5.44217885e-01
-7.20503330e-01 3.63061935e-01 -4.60389972e-01 -1.08845592e+00
-1.19229354e-01 2.73455113e-01 -4.02672708e-01 3.06771100e-01
-1.66194749e+00 1.10759354e+00 1.18176770e+00 -3.12249213e-01
-6.11415841e-02 -3.12563807e-01 -8.23144019e-01 -1.25825524e-01
3.37389618e-01 -5.56577504e-01 1.08634126e+00 -5.66589057e-01
-1.10467851e+00 8.99333358e-01 -2.58822739e-01 -5.79026520e-01
-3.64998639e-01 -3.35173398e-01 -7.06866145e-01 -1.83958054e-01
2.99763471e-01 1.90308467e-01 3.51936966e-01 -1.04453266e+00
-8.77912223e-01 -4.58576471e-01 -2.16867253e-01 4.67924714e-01
-2.80181944e-01 4.51084748e-02 -1.33548751e-01 -8.31147969e-01
2.05758467e-01 -8.84872913e-01 -9.66872424e-02 -1.03429937e+00
-5.18205166e-01 -6.63997889e-01 7.05613792e-01 -3.40258062e-01
1.68016803e+00 -1.87708390e+00 -9.39791352e-02 3.37494284e-01
-4.00951989e-02 4.34446990e-01 3.70637141e-02 7.25262403e-01
-8.06061774e-02 1.17377914e-01 -2.02445611e-01 2.52226670e-03
-4.37366143e-02 1.86686441e-01 -1.86853051e-01 2.67521124e-02
-1.81778535e-01 1.80163950e-01 -9.99229133e-01 -6.32118702e-01
1.67248741e-01 1.76418498e-01 -1.98584855e-01 2.72762757e-02
-7.37490878e-02 -2.87088811e-01 -9.10265386e-01 -6.78382516e-02
5.21089017e-01 4.23137136e-02 2.59685874e-01 -3.63515496e-01
-2.85406560e-02 2.75035381e-01 -1.49810183e+00 1.82238173e+00
-6.82006299e-01 2.37237424e-01 -5.33252239e-01 -8.47446024e-01
1.22584939e+00 2.34123498e-01 3.27487141e-01 -5.88490546e-01
2.80437410e-01 8.33743736e-02 -9.05547887e-02 -5.88147998e-01
1.17876911e+00 -2.98413217e-01 -2.72205740e-01 4.10855263e-01
3.15405309e-01 2.79750265e-02 3.32787693e-01 3.66395712e-01
9.78317440e-01 -1.50181293e-01 1.14450622e+00 -8.15124035e-01
7.16136098e-01 3.09509546e-01 7.07534492e-01 5.98460972e-01
-8.14479887e-02 8.31123739e-02 -2.16483437e-02 -5.63288666e-02
-7.20595241e-01 -8.32434177e-01 9.97308120e-02 8.91254246e-01
3.84413123e-01 -7.03530610e-01 -7.02885866e-01 -5.56562781e-01
-1.45857573e-01 1.07152593e+00 -6.57326043e-01 1.04210906e-01
-1.97589025e-01 -8.10736179e-01 3.19795758e-01 7.50140920e-02
6.19183421e-01 -1.16311610e+00 -6.73765242e-01 4.72275168e-01
-1.50592148e-01 -8.08823347e-01 -4.59433794e-01 1.05975173e-01
-5.49083412e-01 -8.35623443e-01 -3.85143906e-01 -7.70476043e-01
-7.79910907e-02 3.90731096e-01 1.01313102e+00 -1.80445462e-01
-1.76211625e-01 5.44519275e-02 -9.06679571e-01 -8.41668487e-01
-2.71034956e-01 4.18005258e-01 4.16811764e-01 -9.14138258e-02
1.00427771e+00 -5.57194948e-01 -4.98407483e-01 -2.73881853e-01
-1.12375629e+00 -1.07593350e-01 6.00895941e-01 8.57613564e-01
6.07178330e-01 9.18055624e-02 8.64965498e-01 -1.11280560e+00
1.36782897e+00 -7.15077400e-01 -5.18273950e-01 3.60721201e-01
-1.03187978e+00 6.20574653e-01 1.19519256e-01 -3.27783376e-01
-1.37354493e+00 -1.40894905e-01 -2.26906717e-01 4.10142243e-01
-1.88730821e-01 9.80815947e-01 5.73338792e-02 6.43662810e-01
7.81198442e-01 3.12627763e-01 -4.86797422e-01 -7.05712616e-01
2.82826215e-01 1.03413284e+00 1.44150361e-01 -5.70103526e-01
2.76568770e-01 -4.52889409e-03 -3.09005439e-01 -1.13934612e+00
-9.05188560e-01 -1.06342888e+00 -5.07454276e-01 4.39156704e-02
1.11166787e+00 -3.55802745e-01 -4.31515455e-01 8.64696801e-02
-1.26899266e+00 5.70769846e-01 3.05959404e-01 7.86451221e-01
-2.49058709e-01 3.52245241e-01 -8.27623457e-02 -1.09751940e+00
-9.86202180e-01 -8.93836737e-01 8.44217241e-01 3.48683894e-01
-7.13369548e-01 -9.41264153e-01 5.46002924e-01 1.55614048e-01
3.39843005e-01 1.22788377e-01 1.14102900e+00 -1.50067329e+00
3.38656276e-01 -6.60517812e-02 1.60019234e-01 8.11588019e-02
3.82465899e-01 -4.73332375e-01 -6.73188984e-01 5.13243824e-02
-8.37968290e-03 -5.14346063e-02 9.01670873e-01 1.53331473e-01
4.15029049e-01 -4.29956466e-01 -5.47695756e-01 -1.98459029e-01
1.79092848e+00 5.69786370e-01 1.95956022e-01 6.20854378e-01
3.66317987e-01 2.03497455e-01 1.06508625e+00 4.89443213e-01
1.39641747e-01 4.25489426e-01 -1.59040332e-01 1.36201456e-01
3.69921327e-01 -1.32516742e-01 1.79326013e-01 1.16702890e+00
3.44295651e-01 -5.12264431e-01 -9.86209571e-01 7.37328827e-01
-1.97349918e+00 -9.70203638e-01 2.21460797e-02 1.99868667e+00
9.76604939e-01 3.48674685e-01 -9.78843495e-03 2.27363095e-01
7.73424447e-01 4.74370748e-01 -1.05980739e-01 -6.96921825e-01
-4.94634956e-02 5.25562584e-01 4.66281980e-01 6.53689742e-01
-8.95414710e-01 1.30731916e+00 5.00958824e+00 1.28938198e+00
-1.03056729e+00 2.42043376e-01 -1.00232348e-01 2.88812429e-01
-3.85923296e-01 -2.96169496e-03 -7.16826618e-01 2.00436488e-01
5.63936591e-01 -6.92859888e-01 7.74227232e-02 8.40494454e-01
1.92991644e-01 -5.99336922e-01 -5.76187551e-01 1.25750339e+00
2.61888981e-01 -1.14160860e+00 3.12886000e-01 -3.38559985e-01
4.85036761e-01 1.83742449e-01 -5.20700336e-01 1.51375666e-01
3.25632602e-01 -6.15272820e-01 4.50309455e-01 6.54600561e-01
2.68024147e-01 -1.12716210e+00 1.14710474e+00 3.26810837e-01
-1.05121613e+00 4.40801829e-01 -4.45091605e-01 -9.13883820e-02
-5.44347428e-02 8.05364490e-01 -1.25716460e+00 6.85389996e-01
4.40035999e-01 4.84309733e-01 -3.87645811e-01 5.66216707e-01
1.64007574e-01 8.40876341e-01 -2.78648674e-01 -8.29457581e-01
3.94660234e-01 -1.41337454e-01 8.71551812e-01 1.61563230e+00
4.01283801e-01 4.61825132e-01 2.65675455e-01 7.07367778e-01
2.02111676e-01 8.72117877e-01 -6.23929143e-01 -1.89572230e-01
7.12272823e-01 1.26519299e+00 -8.29769373e-01 -6.08182251e-01
-4.30337600e-02 9.65197206e-01 5.88078871e-02 8.84900019e-02
-1.70958504e-01 -7.59806991e-01 6.71514332e-01 -2.68344015e-01
2.38030441e-02 -2.44535640e-01 -1.10776551e-01 -9.30662513e-01
-8.07853416e-02 -5.98299205e-01 5.87904394e-01 -7.95336962e-01
-1.39891422e+00 9.45443392e-01 3.41768235e-01 -9.94049370e-01
-1.85301185e-01 -4.95290011e-01 -3.11778367e-01 8.89871120e-01
-1.16829038e+00 -8.39645565e-01 1.46458186e-02 4.14215744e-01
6.57231688e-01 -4.62151170e-01 9.79756594e-01 -1.88111160e-02
-2.20197275e-01 2.40330040e-01 -1.21340714e-02 3.81049290e-02
8.17433715e-01 -1.30229378e+00 2.81367272e-01 7.88340271e-01
5.23155034e-01 1.08602083e+00 1.16306043e+00 -9.67800379e-01
-1.07443416e+00 -6.09219313e-01 1.16557157e+00 -1.00404814e-01
5.95832288e-01 -1.01981707e-01 -7.52537370e-01 3.43110442e-01
7.63417780e-01 -6.91822886e-01 1.07953262e+00 2.23583162e-01
-1.64627716e-01 -2.02800602e-01 -1.31145990e+00 4.64596897e-01
5.72805882e-01 -2.04663023e-01 -1.61013114e+00 3.07156779e-02
7.59194493e-01 -6.31772205e-02 -6.12596571e-01 8.14336017e-02
4.92979705e-01 -8.02802086e-01 6.33958876e-01 -5.23717701e-01
-3.57417725e-02 -5.15183568e-01 -3.97739649e-01 -1.82641613e+00
-3.52495909e-01 -3.71276319e-01 3.57068717e-01 1.56969059e+00
5.67320645e-01 -5.19611061e-01 1.56852901e-01 2.18123496e-01
2.10157245e-01 -7.29906738e-01 -7.78677166e-01 -5.35464287e-01
-1.12596259e-01 -6.41512811e-01 8.70136023e-01 1.24201334e+00
4.76417065e-01 9.43401098e-01 -1.07254565e-01 -6.68198243e-02
1.81863040e-01 4.28740382e-02 2.39542201e-01 -1.36112857e+00
-1.95889071e-01 -2.36716673e-01 -6.56651258e-01 -7.02260375e-01
1.65752709e-01 -9.23147023e-01 5.24611119e-03 -1.73713160e+00
3.58838767e-01 -3.24174672e-01 -9.19384718e-01 4.03146535e-01
-6.35432899e-01 -1.54704675e-01 1.49344847e-01 -3.11909556e-01
-4.36178297e-01 2.86135823e-01 6.77987576e-01 -2.48027623e-01
-4.01322573e-01 -3.31025720e-01 -8.16427112e-01 6.29253447e-01
9.36114311e-01 -8.14884126e-01 -4.15301353e-01 -6.52717650e-02
5.90153813e-01 -1.36554703e-01 -2.09106401e-01 -8.52857649e-01
1.11805759e-01 -5.32226920e-01 9.74552557e-02 -7.82607853e-01
1.67597055e-01 -6.76766753e-01 -1.62177846e-01 2.93018371e-01
-3.17579925e-01 5.44857442e-01 2.57210076e-01 5.33152103e-01
-2.23745748e-01 -4.79540527e-01 5.32236040e-01 -1.58669487e-01
-1.18746889e+00 -2.97299236e-01 -5.29981434e-01 2.16964975e-01
8.10447156e-01 -4.09933701e-02 6.46086782e-02 -1.20677069e-01
-3.17115575e-01 4.22655717e-02 1.49920836e-01 8.68158042e-01
6.18623972e-01 -1.35388541e+00 -7.71268249e-01 -2.65221417e-01
5.66634893e-01 -4.50886071e-01 8.58730599e-02 3.08140010e-01
-1.97075784e-01 4.68097061e-01 -1.62014455e-01 -2.03261048e-01
-1.58838403e+00 5.79630971e-01 -2.01468527e-01 -4.24527854e-01
-2.13301703e-01 6.64203584e-01 -2.33767524e-01 -1.99661136e-01
-3.01958591e-01 -4.50815797e-01 -7.91398942e-01 1.62489384e-01
5.76600969e-01 1.58421487e-01 2.84568012e-01 -7.12149024e-01
-6.24381602e-01 5.89397907e-01 -1.31842598e-01 -6.66836500e-01
1.24469137e+00 -1.62067875e-01 -2.56286770e-01 7.69741893e-01
9.75612104e-01 4.16011035e-01 1.61492288e-01 -4.01992708e-01
5.72854042e-01 -2.09203139e-01 1.78822517e-01 -9.08294976e-01
-3.83650571e-01 4.89240110e-01 7.26678669e-01 7.49370158e-02
1.12483132e+00 -1.93767786e-01 9.63737190e-01 7.86759436e-01
5.00377715e-01 -1.41296732e+00 -5.10070741e-01 7.66458452e-01
8.31832290e-01 -1.10157979e+00 3.95469628e-02 4.35558632e-02
-8.09361458e-01 9.39823925e-01 3.06680113e-01 -2.71501064e-01
9.68157530e-01 7.66650289e-02 1.58968747e-01 -1.31422594e-01
-5.80203295e-01 -3.79257441e-01 5.14326930e-01 4.10151660e-01
6.36511266e-01 1.33636877e-01 -1.44791698e+00 1.00095367e+00
-4.87573773e-01 -7.76553154e-02 4.55849707e-01 8.41757715e-01
-8.62553835e-01 -1.32800734e+00 -3.58933359e-01 6.41618788e-01
-5.13809681e-01 -3.74789834e-01 -5.68283260e-01 6.78820550e-01
2.59148419e-01 1.28298759e+00 -2.36823022e-01 -7.37648010e-01
5.79595454e-02 3.31251740e-01 2.77039945e-01 -8.08371425e-01
-7.05763340e-01 -1.15568545e-02 3.55591923e-01 -2.80265957e-01
-6.78848743e-01 -3.41790169e-01 -1.55925608e+00 2.19233334e-01
-5.68153679e-01 8.21253657e-01 8.65254223e-01 1.37553263e+00
3.68424237e-01 2.68117577e-01 3.14606488e-01 -2.81432629e-01
-3.25147837e-01 -1.41137755e+00 -7.32399464e-01 6.92159235e-01
1.70142189e-01 -8.64690840e-01 1.34993732e-01 -1.24593057e-01] | [10.159637451171875, 9.034948348999023] |
7c20e540-8c8f-4b34-a0f8-2e77368474e6 | deep-learning-for-joint-acoustic-echo-and | 2305.01637 | null | https://arxiv.org/abs/2305.01637v2 | https://arxiv.org/pdf/2305.01637v2.pdf | Deep Learning for Joint Acoustic Echo and Acoustic Howling Suppression in Hybrid Meetings | Hybrid meetings have become increasingly necessary during the post-COVID period and also brought new challenges for solving audio-related problems. In particular, the interplay between acoustic echo and acoustic howling in a hybrid meeting makes the joint suppression of them difficult. This paper proposes a deep learning approach to tackle this problem by formulating a recurrent feedback suppression process as an instantaneous speech separation task using the teacher-forced training strategy. Specifically, a self-attentive recurrent neural network is utilized to extract the target speech from microphone recordings with accessible and learned reference signals, thus suppressing acoustic echo and acoustic howling simultaneously. Different combinations of input signals and loss functions have been investigated for performance improvement. Experimental results demonstrate the effectiveness of the proposed method for suppressing echo and howling jointly in hybrid meetings. | ['Dong Yu', 'Meng Yu', 'Hao Zhang'] | 2023-05-02 | null | null | null | null | ['speech-separation'] | ['speech'] | [ 3.36430818e-01 4.05319547e-03 4.96818393e-01 -1.41554708e-02
-1.21839607e+00 -2.06120118e-01 3.38077992e-01 -2.91476727e-01
-2.62524217e-01 4.45470363e-01 5.25827467e-01 5.98433688e-02
-2.99577981e-01 -4.49564196e-02 -4.05177742e-01 -9.81188774e-01
1.02955028e-02 -2.07396582e-01 -1.26485556e-01 -2.85539269e-01
-4.75130975e-02 3.04802269e-01 -1.71532106e+00 2.02978805e-01
1.05918491e+00 7.55299628e-01 5.34731746e-01 8.83250773e-01
2.07304731e-01 5.78199029e-01 -1.08735728e+00 2.09488526e-01
2.59756029e-01 -6.17208898e-01 3.40718627e-02 1.57540902e-01
4.83049929e-01 -5.23966402e-02 -4.20185566e-01 9.66371179e-01
1.02393603e+00 5.54728568e-01 5.16600907e-01 -8.40563655e-01
-1.03252567e-01 8.15741479e-01 -7.18290389e-01 4.69063193e-01
2.80666143e-01 9.80205163e-02 7.33957410e-01 -1.07737434e+00
-1.86370566e-01 1.35181868e+00 7.27488518e-01 4.51363683e-01
-1.21271884e+00 -1.18121314e+00 3.59676689e-01 2.02480420e-01
-1.36699438e+00 -9.19918716e-01 1.21246183e+00 -2.60996759e-01
9.40817893e-01 4.46828544e-01 4.61035848e-01 1.16853356e+00
9.36095715e-02 6.98258400e-01 8.03719044e-01 -6.26765013e-01
-1.59649640e-01 1.86049879e-01 1.35893866e-01 4.62230183e-02
-3.25589508e-01 1.66272566e-01 -7.79208958e-01 -9.19646919e-02
4.42731798e-01 -1.18806519e-01 -6.04280233e-01 2.84256727e-01
-7.18565226e-01 3.87056231e-01 1.91057816e-01 5.51976800e-01
-5.01021743e-01 1.55150436e-03 3.71936798e-01 4.56440032e-01
6.66904569e-01 4.54820782e-01 -3.10064465e-01 8.89323428e-02
-1.05272675e+00 8.74682665e-02 5.41336715e-01 5.12445688e-01
3.52758139e-01 7.41140187e-01 -1.57743737e-01 1.39126682e+00
2.65791804e-01 6.49243832e-01 5.04064679e-01 -7.89095879e-01
5.79379439e-01 -1.31197408e-01 1.87091812e-01 -1.25318873e+00
-4.58806068e-01 -9.20828581e-01 -1.08332372e+00 5.72489724e-02
4.29927893e-02 -5.10068536e-01 -8.85500789e-01 1.86760283e+00
3.24185431e-01 7.44162619e-01 3.40510607e-02 1.02448761e+00
7.48000920e-01 1.10382783e+00 -2.90020227e-01 -7.55133331e-01
8.85851145e-01 -1.13936484e+00 -1.32287955e+00 -2.83529967e-01
-8.37913379e-02 -9.81820703e-01 9.54769790e-01 7.18669116e-01
-1.15862823e+00 -8.27909589e-01 -1.12401581e+00 3.50414604e-01
2.40491569e-01 1.19843058e-01 -1.82777748e-01 4.24185097e-01
-9.99572337e-01 1.89848453e-01 -7.42377877e-01 1.79586411e-01
-3.26935560e-01 4.28782582e-01 2.58545913e-02 4.10975784e-01
-1.29263151e+00 5.09548366e-01 -1.63760751e-01 8.41653466e-01
-1.02852881e+00 -6.81541920e-01 -7.43089259e-01 3.05675119e-01
4.24608767e-01 -1.59987003e-01 1.38177180e+00 -1.05374074e+00
-1.85837376e+00 2.05013886e-01 -1.70588911e-01 -3.36180866e-01
3.33181411e-01 -8.13358724e-01 -8.70728910e-01 -7.25534707e-02
-5.51399216e-02 1.31746344e-02 1.54230893e+00 -1.22589612e+00
-5.06426215e-01 -7.71329701e-02 -4.85843062e-01 6.21208370e-01
-6.15803421e-01 3.33459722e-03 -1.75918549e-01 -9.76416171e-01
2.27061197e-01 -7.83433855e-01 -2.24603117e-01 -7.29149997e-01
-4.48325038e-01 -2.94992011e-02 8.24062467e-01 -9.28462386e-01
1.47844481e+00 -2.47047782e+00 3.87872070e-01 1.92905426e-01
1.27709702e-01 6.05321825e-01 -3.49063873e-01 4.98481542e-01
-2.08816603e-01 -2.57730603e-01 -4.12130430e-02 -5.35786629e-01
-3.30481201e-01 -3.64767462e-01 -6.60542428e-01 2.89783806e-01
5.45305237e-02 1.40357941e-01 -6.06758714e-01 -1.43643275e-01
2.69498944e-01 8.78203928e-01 -3.99155438e-01 6.76547110e-01
1.83090612e-01 6.52491391e-01 -2.27154735e-02 1.36151895e-01
7.20807791e-01 4.39691484e-01 1.45704687e-01 -4.69245017e-02
-3.56867522e-01 4.66722280e-01 -1.32426262e+00 1.37263215e+00
-8.19634736e-01 7.05376089e-01 1.06595170e+00 -9.67639506e-01
1.20576394e+00 6.17984116e-01 1.61755040e-01 -8.34226668e-01
5.55341430e-02 3.27105552e-01 8.54971707e-02 -4.74131763e-01
3.56475711e-01 -4.47072148e-01 8.07303563e-02 1.84630565e-02
6.63278624e-02 -3.54960859e-01 -3.03090572e-01 -1.73638463e-01
1.04330659e+00 -3.27475011e-01 -8.85363370e-02 -5.84326051e-02
9.27704811e-01 -7.47216523e-01 7.80864835e-01 8.15183580e-01
-1.29893422e-01 6.20495796e-01 -1.15670279e-01 4.89588873e-03
-5.82710564e-01 -9.73838985e-01 2.20424160e-01 1.23221719e+00
9.07111689e-02 -2.65205830e-01 -5.94340324e-01 -1.82902113e-01
-2.37027824e-01 5.81476152e-01 -1.48433581e-01 -3.92221361e-01
-1.08433378e+00 -2.55517483e-01 4.42946404e-01 1.97876349e-01
1.71675220e-01 -1.19750953e+00 -1.62285924e-01 4.60128188e-01
-4.00483757e-01 -8.72397482e-01 -8.70926082e-01 5.78649759e-01
-4.98689860e-01 -4.73151237e-01 -8.22005332e-01 -1.00179839e+00
2.10361183e-01 7.47663140e-01 6.44524813e-01 -1.35168716e-01
-1.23763546e-01 2.64279723e-01 -7.66007006e-02 -3.87422860e-01
-3.67996812e-01 7.82304928e-02 4.10260320e-01 3.65953237e-01
-1.42565280e-01 -9.26650524e-01 -4.28345501e-01 4.02360409e-01
-7.60181725e-01 -1.99151844e-01 4.92889315e-01 1.06753540e+00
2.58147597e-01 3.76167357e-01 1.02660620e+00 -3.98634851e-01
1.14520693e+00 -2.78527588e-01 -5.10382116e-01 -6.29997998e-03
4.91778776e-02 -3.22465688e-01 9.06930983e-01 -6.76386833e-01
-1.43700767e+00 -8.91150683e-02 -2.40316018e-01 -6.27034187e-01
9.47984830e-02 4.38667446e-01 -3.05998951e-01 2.59953082e-01
4.01631147e-01 2.52016127e-01 -8.00666437e-02 -6.15539551e-01
-1.16242338e-02 1.05520618e+00 5.02391994e-01 -1.40821293e-01
8.26619148e-01 -8.28488469e-02 -5.61275482e-01 -1.34332848e+00
-9.30644095e-01 -6.58507049e-01 -1.03835903e-01 -3.59927297e-01
5.08212149e-01 -1.05692387e+00 -6.46948099e-01 7.70170987e-01
-1.15115654e+00 -2.20567331e-01 -1.45019382e-01 7.38901258e-01
-2.86881596e-01 2.42436916e-01 -5.64739645e-01 -1.36950421e+00
-5.46392918e-01 -1.08074296e+00 7.98376739e-01 4.62639540e-01
-1.77675411e-01 -5.46356499e-01 3.45500052e-01 3.42504799e-01
5.34740925e-01 -3.70734334e-01 4.37946767e-01 -4.31983262e-01
-3.12348366e-01 -1.08185805e-01 1.97037727e-01 6.74061477e-01
4.22897935e-01 -3.43163073e-01 -1.28996646e+00 -3.25479239e-01
6.55274808e-01 -2.55343229e-01 8.49314451e-01 6.41660810e-01
7.79395342e-01 -4.39682215e-01 -1.08290002e-01 4.35676515e-01
8.19009542e-01 5.32505751e-01 3.13758433e-01 -1.44268155e-01
5.43119669e-01 5.90820134e-01 5.44841647e-01 4.89107132e-01
-2.16445923e-01 7.46787846e-01 1.95366561e-01 -1.99442700e-01
-2.71740705e-01 -5.27686886e-02 5.85412621e-01 1.75991416e+00
2.77368337e-01 -3.43605876e-01 -6.67611301e-01 5.47893822e-01
-1.78200400e+00 -8.42242122e-01 -1.72469001e-02 2.07413340e+00
9.32818890e-01 8.30075294e-02 -8.75546336e-02 4.19764042e-01
9.86464858e-01 4.64470685e-01 -5.33620477e-01 -4.21703815e-01
-9.98223647e-02 3.61225218e-01 -1.31002944e-02 7.85958529e-01
-1.03641331e+00 6.27832592e-01 6.41342926e+00 1.04027867e+00
-1.31071138e+00 1.90399557e-01 3.47247005e-01 -3.18017721e-01
-1.01552568e-01 -6.24415755e-01 -7.95705914e-01 2.42011830e-01
1.03430355e+00 -3.82952974e-03 4.87127751e-01 3.76580268e-01
5.55401266e-01 8.55677947e-02 -6.57978833e-01 1.06150174e+00
1.82136014e-01 -7.31819093e-01 -2.98149705e-01 -4.35086042e-01
5.23768425e-01 -2.26425141e-01 4.10120130e-01 4.63788539e-01
-4.71268175e-03 -1.02210510e+00 7.07530379e-01 6.12020731e-01
5.47261059e-01 -8.65307271e-01 4.72146600e-01 4.50447083e-01
-1.28881073e+00 -4.15664971e-01 9.02910084e-02 -2.60258794e-01
1.56032115e-01 7.31521845e-01 -8.28027010e-01 4.47731316e-01
5.92259526e-01 6.24933660e-01 1.32674560e-01 1.08690798e+00
-3.39294344e-01 9.08141434e-01 -4.72048432e-01 2.18821689e-01
8.75438303e-02 -2.09128529e-01 1.16882169e+00 1.30100608e+00
3.95443857e-01 1.20704491e-02 2.13636279e-01 5.90076029e-01
4.42935601e-02 -3.33509520e-02 -5.14394641e-01 1.60153076e-01
6.22414112e-01 1.09856093e+00 -2.41136938e-01 1.22226790e-01
3.01933126e-03 6.16716444e-01 3.53579931e-02 7.34981596e-01
-8.44782710e-01 -6.35114014e-01 7.17718780e-01 5.08501157e-02
2.90180594e-01 -2.69928604e-01 7.06993090e-03 -9.05881166e-01
1.64891854e-01 -1.08660984e+00 -7.53030330e-02 -6.21219814e-01
-1.07282352e+00 8.17715347e-01 -3.17078471e-01 -1.25522208e+00
-3.56279314e-01 6.05418440e-03 -7.78109252e-01 9.63538527e-01
-1.35349298e+00 -6.83913529e-01 -1.09221287e-01 5.03778696e-01
9.59157825e-01 -2.15096235e-01 7.82072306e-01 6.20824099e-01
-9.20481563e-01 7.20481098e-01 2.57745117e-01 -3.40205610e-01
8.26660633e-01 -9.01586652e-01 -5.86017594e-02 8.79757047e-01
1.20341837e-01 5.93496561e-01 1.05291355e+00 -4.00525212e-01
-1.13243079e+00 -9.50939596e-01 5.79442859e-01 1.95881918e-01
5.78686476e-01 -7.40619123e-01 -1.04066110e+00 3.21049422e-01
5.53549588e-01 -3.83397698e-01 4.64960724e-01 1.03990749e-01
-1.17786810e-01 -5.85315049e-01 -5.62181056e-01 4.89205688e-01
8.08142543e-01 -6.35561109e-01 -7.38166809e-01 8.17025751e-02
9.81598794e-01 -2.81139076e-01 -2.09111139e-01 3.89247835e-01
3.07689279e-01 -7.76113987e-01 9.02975202e-01 -6.55185133e-02
1.83184985e-02 -2.89950430e-01 -1.86504461e-02 -1.70727503e+00
-1.75475910e-01 -1.35522711e+00 -4.54903133e-02 1.53599787e+00
3.28452766e-01 -4.61394191e-01 4.60471272e-01 -1.09727317e-02
-5.40893674e-01 -3.26294839e-01 -1.05102754e+00 -6.34977281e-01
-3.73860985e-01 -3.22695911e-01 -6.51534647e-03 6.12118423e-01
-7.63480663e-02 7.25113571e-01 -9.19016719e-01 3.63243312e-01
4.38975364e-01 -3.18582833e-01 6.94612920e-01 -9.79289114e-01
-4.32542950e-01 -2.95566112e-01 -4.96529462e-03 -1.29858792e+00
2.92532980e-01 -4.09341663e-01 6.77131534e-01 -1.21306062e+00
-3.46254557e-01 -2.26128042e-01 -6.59936666e-01 3.57607268e-02
-2.88306534e-01 -7.94675797e-02 2.84134358e-01 -5.04633375e-02
-3.59628052e-01 1.08896315e+00 9.61814702e-01 -2.89817572e-01
-7.03393281e-01 3.71227741e-01 -5.19619942e-01 7.96456635e-01
5.64417243e-01 -5.66104829e-01 -6.17005348e-01 -4.30303514e-01
-6.23805635e-02 6.12789392e-01 1.06237270e-02 -1.21586907e+00
4.66665506e-01 1.79140404e-01 -1.93325728e-01 -8.14127624e-01
8.23988676e-01 -9.23688352e-01 1.55565843e-01 3.00723970e-01
-5.47776163e-01 -3.88961047e-01 5.06088316e-01 7.60028958e-01
-6.14338994e-01 8.52563232e-02 8.36200058e-01 2.99691379e-01
2.24525128e-02 -3.51781517e-01 -7.39669025e-01 2.07506698e-02
6.19717121e-01 4.43923771e-02 2.16391370e-01 -7.01426148e-01
-9.83827055e-01 2.30270386e-01 -5.47379255e-01 4.92839724e-01
7.94181108e-01 -1.06839669e+00 -8.38543892e-01 4.38258529e-01
-4.86147076e-01 -1.51790112e-01 6.16372705e-01 8.59084308e-01
1.67430535e-01 2.67572224e-01 1.05287701e-01 -5.27634859e-01
-1.51762891e+00 1.72259748e-01 6.34186566e-01 -3.21050406e-01
-6.82944179e-01 1.23423791e+00 7.15392828e-01 -4.72964168e-01
9.95655715e-01 -2.40948901e-01 -4.62044775e-01 1.66601300e-01
6.81528211e-01 4.71417606e-01 2.31210902e-01 -3.62053871e-01
-1.74645230e-01 4.16987240e-01 -2.49303862e-01 -3.35889071e-01
1.52084684e+00 -4.40645665e-01 3.64344865e-02 8.22583318e-01
9.46038544e-01 4.17149514e-01 -1.16492438e+00 -3.90085220e-01
-1.47159562e-01 -3.05391550e-01 2.29288399e-01 -7.73831367e-01
-1.18157506e+00 9.12647128e-01 6.51863098e-01 3.82015675e-01
1.52695894e+00 -4.85508651e-01 8.66214335e-01 1.47447586e-01
-1.00891605e-01 -1.03369653e+00 4.67185229e-01 6.43177211e-01
1.24189579e+00 -9.25920367e-01 -4.48490232e-01 -3.59354764e-01
-3.38387847e-01 8.46920609e-01 6.68017685e-01 -2.58609623e-01
8.23091626e-01 4.74987775e-01 3.39254946e-01 2.18138084e-01
-8.47872615e-01 1.94465831e-01 3.06975633e-01 4.78484482e-01
5.60473442e-01 -5.77768981e-02 -2.13135272e-01 8.64125788e-01
-1.05867177e-01 -4.42879707e-01 4.24465388e-01 6.44468069e-01
-4.65326935e-01 -8.97946477e-01 -8.10657144e-01 2.90854037e-01
-5.65589845e-01 -2.09577367e-01 -4.67836589e-01 2.65088439e-01
-1.72865599e-01 1.42676222e+00 -1.04335122e-01 -5.83764195e-01
6.43645048e-01 4.65886928e-02 -2.13454496e-02 -7.10639536e-01
-1.04812706e+00 1.10984397e+00 8.32127407e-02 -2.35103413e-01
-3.26200157e-01 -5.41468859e-01 -1.04905725e+00 2.70987064e-01
-7.89894700e-01 5.42920232e-01 5.39100945e-01 7.82594979e-01
3.86495262e-01 1.19970667e+00 1.03573656e+00 -1.21418357e+00
-8.85456204e-01 -1.33014941e+00 -7.23145902e-01 9.85466242e-02
9.24446523e-01 -5.32042980e-01 -8.18794131e-01 -1.95013419e-01] | [15.017820358276367, 5.886640548706055] |
96022027-4ba4-463a-9fcd-e7384665da09 | to-scene-a-large-scale-dataset-for | 2203.09440 | null | https://arxiv.org/abs/2203.09440v3 | https://arxiv.org/pdf/2203.09440v3.pdf | TO-Scene: A Large-scale Dataset for Understanding 3D Tabletop Scenes | Many basic indoor activities such as eating or writing are always conducted upon different tabletops (e.g., coffee tables, writing desks). It is indispensable to understanding tabletop scenes in 3D indoor scene parsing applications. Unfortunately, it is hard to meet this demand by directly deploying data-driven algorithms, since 3D tabletop scenes are rarely available in current datasets. To remedy this defect, we introduce TO-Scene, a large-scale dataset focusing on tabletop scenes, which contains 20,740 scenes with three variants. To acquire the data, we design an efficient and scalable framework, where a crowdsourcing UI is developed to transfer CAD objects from ModelNet and ShapeNet onto tables from ScanNet, then the output tabletop scenes are simulated into real scans and annotated automatically. Further, a tabletop-aware learning strategy is proposed for better perceiving the small-sized tabletop instances. Notably, we also provide a real scanned test set TO-Real to verify the practical value of TO-Scene. Experiments show that the algorithms trained on TO-Scene indeed work on the realistic test data, and our proposed tabletop-aware learning strategy greatly improves the state-of-the-art results on both 3D semantic segmentation and object detection tasks. Dataset and code are available at https://github.com/GAP-LAB-CUHK-SZ/TO-Scene. | ['Xiaoguang Han', 'Haolin Liu', 'Pei Chen', 'Mutian Xu'] | 2022-03-17 | null | null | null | null | ['scene-parsing'] | ['computer-vision'] | [ 3.23622108e-01 -2.92748243e-01 -1.57018661e-01 -4.71202970e-01
-9.28109288e-01 -7.36569345e-01 4.82721142e-02 -2.02192530e-01
6.29698783e-02 1.10808305e-01 -2.55318373e-01 -1.74942210e-01
6.50567114e-02 -7.08703578e-01 -1.13784468e+00 -3.75474095e-01
2.78744221e-01 7.26643980e-01 6.49816394e-01 -3.47939059e-02
6.15928359e-02 2.16199234e-01 -1.51691115e+00 6.52728379e-01
9.10397232e-01 1.15310645e+00 8.87034237e-01 5.94115973e-01
-1.64740067e-02 4.41679060e-01 -4.16600913e-01 -3.48783880e-01
5.38720548e-01 -1.42699182e-01 -3.13334823e-01 7.20939636e-01
5.32163799e-01 -6.51133299e-01 -5.42841479e-02 1.06196642e+00
5.57885468e-01 1.16557389e-01 2.42184773e-01 -1.22498512e+00
-5.67296743e-01 2.31358260e-01 -5.02769113e-01 -2.04503641e-01
8.34023058e-01 2.33681038e-01 6.91343546e-01 -9.16277885e-01
4.07741100e-01 1.22205102e+00 4.88597661e-01 2.26822510e-01
-7.38140821e-01 -6.17311776e-01 3.35289270e-01 -8.33899304e-02
-1.40301895e+00 -2.40229473e-01 7.69566417e-01 -2.35526979e-01
5.81094205e-01 5.24437785e-01 7.50852287e-01 1.31011117e+00
-2.03056321e-01 1.51050961e+00 1.09981358e+00 -2.10446492e-01
2.51056463e-01 1.14506111e-01 -2.72633821e-01 6.75402403e-01
2.33343437e-01 -2.37061024e-01 -4.40692544e-01 2.76449353e-01
1.14240897e+00 4.43513244e-01 -3.34839880e-01 -7.58756936e-01
-1.29415083e+00 2.88388640e-01 4.74438459e-01 5.72117381e-02
-1.15931869e-01 -1.27673760e-01 2.21546128e-01 -1.22818589e-01
3.49966139e-01 5.37476167e-02 -7.90094793e-01 -1.91880152e-01
-6.62707090e-01 2.41387784e-01 5.93251824e-01 1.72975624e+00
4.22438741e-01 -4.31913584e-01 9.19857323e-02 7.58265793e-01
4.39944386e-01 7.56184816e-01 3.63953948e-01 -7.11867571e-01
1.20515275e+00 8.78576398e-01 3.89951587e-01 -6.91148579e-01
-3.32146734e-01 4.28326204e-02 -3.71438295e-01 -2.63629645e-01
4.93089557e-01 4.66660783e-02 -1.14706254e+00 8.04501653e-01
6.40360832e-01 4.08100069e-01 -4.10013884e-01 1.27705169e+00
9.79464829e-01 4.47537482e-01 -1.25095695e-01 2.42885277e-01
1.43985295e+00 -1.34225690e+00 -6.00474119e-01 -3.74672771e-01
6.32662475e-01 -8.47767770e-01 1.84869218e+00 6.35259390e-01
-1.05418527e+00 -8.01746190e-01 -9.71774399e-01 -3.39160681e-01
-4.35453206e-01 6.42469287e-01 6.49741292e-01 7.39602864e-01
-4.38652098e-01 2.60342181e-01 -1.11559439e+00 -4.49734360e-01
7.61110365e-01 2.30592072e-01 -1.30693778e-01 -5.39753675e-01
-6.42424881e-01 2.47988060e-01 4.41457480e-01 2.17988610e-01
-9.01085019e-01 -3.77652973e-01 -9.80745733e-01 -2.17385218e-01
9.68526304e-01 -4.19267178e-01 1.42041755e+00 -5.86539507e-01
-1.57124603e+00 9.01573420e-01 -1.03233606e-01 1.51510075e-01
6.61980212e-01 -6.72076464e-01 -2.98857063e-01 -3.29929753e-03
3.02728295e-01 4.34980243e-01 6.31786168e-01 -1.67253506e+00
-6.52996540e-01 -6.59667969e-01 1.67617559e-01 5.16956866e-01
1.66840162e-02 -2.65189171e-01 -1.24929261e+00 -2.48476446e-01
5.03467858e-01 -1.04583418e+00 -1.04988627e-01 1.89787429e-02
-8.75097573e-01 -6.87503163e-03 6.88556910e-01 -4.95580256e-01
8.71350646e-01 -2.16861224e+00 -3.10752064e-01 1.42529607e-02
-1.72941759e-01 2.04380855e-01 2.34358013e-02 2.63044566e-01
2.79300004e-01 -1.00084364e-01 -2.40889579e-01 -5.38527966e-01
2.05931932e-01 3.42794538e-01 -1.92472190e-01 2.12816134e-01
-9.79239568e-02 1.02986765e+00 -9.21503305e-01 -5.74954987e-01
7.44727790e-01 1.35753095e-01 -4.39404905e-01 1.85830563e-01
-4.61749613e-01 5.68704784e-01 -9.99452829e-01 1.31707108e+00
9.12445068e-01 -3.65781367e-01 1.94534048e-01 -1.90532599e-02
2.18197688e-01 2.10573807e-01 -1.53242552e+00 2.41965842e+00
-4.38853323e-01 1.36719465e-01 -6.73362389e-02 -7.04006433e-01
6.59991622e-01 8.63973200e-02 2.74812698e-01 -9.39055026e-01
2.22910345e-01 2.03456581e-01 -3.53285581e-01 -9.89372253e-01
3.97620410e-01 3.87938976e-01 -1.59441710e-01 6.61824122e-02
-3.95538568e-01 -5.69513321e-01 -8.02921206e-02 -2.27332801e-01
7.43818820e-01 7.73805320e-01 1.05983257e-01 1.33214980e-01
2.57142842e-01 3.00299913e-01 3.39382768e-01 7.46515155e-01
-3.50435466e-01 1.10806394e+00 5.27219027e-02 -4.59362030e-01
-8.56055796e-01 -1.26064134e+00 -2.38264382e-01 1.14518869e+00
9.02102470e-01 -2.08690941e-01 -1.06836295e+00 -7.47707665e-01
-7.50198364e-02 5.63686550e-01 -3.49187672e-01 4.97153133e-01
-3.80420744e-01 -5.52175343e-01 1.41427606e-01 7.35459626e-01
8.68404269e-01 -1.15499377e+00 -7.32544959e-01 2.07624882e-02
-2.46641219e-01 -1.60155487e+00 -6.40384793e-01 1.04779124e-01
-8.62667084e-01 -1.40586495e+00 -3.93322557e-01 -9.11900640e-01
7.11435258e-01 7.04234958e-01 1.20166183e+00 -3.14755440e-02
-2.29318142e-01 2.97211528e-01 -5.56878805e-01 -5.39042234e-01
2.40772590e-01 1.91469923e-01 -1.93963543e-01 -1.99029028e-01
5.50653577e-01 -3.40751588e-01 -7.88310647e-01 8.03617716e-01
-9.11308527e-01 2.73897767e-01 4.86739784e-01 2.94901043e-01
1.03302944e+00 7.76560307e-02 -1.90398898e-02 -9.68289614e-01
2.79930115e-01 -2.46783197e-01 -5.89753807e-01 2.61124074e-01
1.32106468e-01 -5.13692856e-01 5.01347840e-01 -5.05149007e-01
-9.98845398e-01 4.85691547e-01 -5.67386951e-03 -4.61059153e-01
-7.03140795e-01 4.78855781e-02 -7.97012091e-01 3.01305979e-01
5.60776055e-01 1.46789894e-01 -6.21352732e-01 -5.22216320e-01
3.19431424e-01 1.02509475e+00 3.14735502e-01 -6.80450141e-01
7.07412601e-01 6.96933508e-01 -5.49928069e-01 -6.75076306e-01
-1.14999270e+00 -6.65305734e-01 -8.81158412e-01 -6.28093258e-02
1.14637041e+00 -1.19352388e+00 -5.78571022e-01 3.91928881e-01
-8.63819540e-01 -9.17890012e-01 -6.52252370e-03 4.92954880e-01
-6.45141363e-01 3.81519981e-02 -4.42450315e-01 -7.06481636e-01
1.00790471e-01 -1.39511156e+00 1.94959390e+00 1.88818648e-01
1.72415420e-01 -5.66698313e-01 -2.98094809e-01 9.46887314e-01
-2.45251015e-01 3.19769382e-01 3.68373245e-01 -4.03816283e-01
-1.08535111e+00 -2.80342191e-01 -3.79600585e-01 6.63602799e-02
2.74654180e-01 -4.13899958e-01 -1.19033122e+00 2.77534202e-02
1.03046045e-01 -4.69225138e-01 2.84161091e-01 4.01079953e-01
1.82285368e+00 2.81147301e-01 -4.90058154e-01 6.87372565e-01
1.29376721e+00 3.85712415e-01 6.76735818e-01 3.40073138e-01
1.16980636e+00 3.15487772e-01 1.12302291e+00 3.74468207e-01
6.56951964e-01 8.29617500e-01 5.11516809e-01 1.42370127e-02
5.70925996e-02 -5.78829944e-01 6.93813562e-02 7.71419048e-01
-8.87287036e-03 -5.23765564e-01 -8.79680932e-01 5.19465268e-01
-1.92306924e+00 -4.39071298e-01 -4.59378719e-01 2.26476455e+00
5.39540946e-01 3.32910895e-01 -4.96631339e-02 -1.18772589e-01
7.13648736e-01 -6.75473129e-03 -7.73129582e-01 2.65125901e-01
1.90131500e-01 9.20071974e-02 4.47139204e-01 2.01831445e-01
-1.33887434e+00 1.08708119e+00 4.52473450e+00 9.32130396e-01
-8.57477725e-01 2.68761396e-01 6.59569979e-01 -2.46866018e-01
1.15633965e-01 -4.77843612e-01 -6.05093658e-01 6.80969179e-01
2.73638308e-01 6.52208924e-01 6.42405212e-01 1.38285112e+00
4.74177420e-01 -2.60179520e-01 -1.33120561e+00 1.53673983e+00
1.37827292e-01 -8.25239897e-01 -2.75164425e-01 -2.60401163e-02
9.41435575e-01 1.80830956e-01 -8.27959254e-02 3.37758273e-01
1.09225631e-01 -9.56568599e-01 7.77465940e-01 9.53824967e-02
8.10662389e-01 -3.93554479e-01 5.81502676e-01 6.66513264e-01
-1.39456153e+00 1.43454328e-01 -4.77961868e-01 4.12453823e-02
1.36080682e-01 4.69909132e-01 -7.74183035e-01 5.35733998e-01
9.90203857e-01 7.15169609e-01 -6.34974003e-01 1.11639345e+00
-3.12049747e-01 3.30940485e-01 -4.49447155e-01 -3.65950614e-02
1.04766130e-01 -2.25053012e-01 1.95564050e-03 7.97088325e-01
5.05423248e-01 2.10813075e-01 6.25914454e-01 7.18541622e-01
-1.65754661e-01 -6.11544102e-02 -6.18618667e-01 2.19767660e-01
5.17560959e-01 1.27544534e+00 -1.12352288e+00 -3.37116241e-01
-3.85942310e-01 1.43365216e+00 7.38948584e-03 4.40051168e-01
-1.15235662e+00 -1.44009590e-01 4.86440033e-01 3.80009979e-01
4.18515414e-01 -3.41216505e-01 -4.34959084e-01 -1.40515089e+00
6.24211669e-01 -8.49612772e-01 -1.36741877e-01 -1.30595160e+00
-1.24589550e+00 4.32595462e-01 -1.95672512e-02 -1.58835268e+00
2.90789247e-01 -8.85612845e-01 -2.99852192e-01 3.65084052e-01
-1.39850855e+00 -1.34350157e+00 -9.64719057e-01 8.18755269e-01
1.21995163e+00 2.86388338e-01 4.83472019e-01 5.19878507e-01
-5.33349752e-01 4.23967481e-01 -4.76878956e-02 7.55429268e-02
5.78598261e-01 -1.35300100e+00 5.75568020e-01 4.43938047e-01
4.32024449e-01 2.98160613e-01 8.94309208e-02 -6.83005571e-01
-1.69851708e+00 -1.24941766e+00 9.39903632e-02 -1.14969063e+00
3.54008913e-01 -1.04288054e+00 -6.84689283e-01 7.99803197e-01
-3.10195595e-01 2.90992141e-01 4.04025078e-01 -2.58568615e-01
1.91661775e-01 1.38015777e-01 -1.13892031e+00 6.94904208e-01
1.68836570e+00 -4.95666921e-01 -3.40752602e-01 8.78988802e-01
8.41918886e-01 -1.32499135e+00 -7.33306229e-01 2.61975497e-01
4.96026874e-01 -9.89460409e-01 1.02772737e+00 -2.04878479e-01
5.83881378e-01 -5.44797122e-01 -4.12153304e-01 -9.34641600e-01
2.37409323e-01 -3.67626548e-01 -5.82838245e-02 8.73048127e-01
2.77071387e-01 -3.29421073e-01 1.26373637e+00 7.01573610e-01
-5.23364067e-01 -8.82631719e-01 -6.96751535e-01 -6.64367318e-01
-5.25761724e-01 -9.66647267e-01 9.36770558e-01 9.45869029e-01
-4.70117778e-01 8.75238776e-02 -2.06292998e-02 4.45124835e-01
3.70097727e-01 2.70139962e-01 1.23995996e+00 -1.07357931e+00
-2.64480472e-01 1.08635940e-01 -1.91299275e-01 -1.66377556e+00
-4.09978747e-01 -5.98795354e-01 2.74580151e-01 -1.72649908e+00
1.18119664e-01 -7.19968081e-01 1.53575651e-02 3.15721005e-01
-3.49325299e-01 4.13966060e-01 2.41397858e-01 2.88893014e-01
-1.12389910e+00 4.42211956e-01 1.88213491e+00 6.11174963e-02
-5.69264948e-01 1.49729982e-01 -4.27819222e-01 1.00033033e+00
6.76868975e-01 -3.05833131e-01 -7.44803786e-01 -8.09903860e-01
-3.60595249e-02 2.08439566e-02 4.59455281e-01 -1.15616620e+00
-4.93040197e-02 -1.48302212e-01 6.67113066e-01 -8.90804946e-01
4.38152283e-01 -1.13211143e+00 -4.95798439e-02 2.89953984e-02
4.18082476e-02 -2.05536738e-01 8.67754519e-02 6.19130790e-01
6.23765662e-02 -9.60302651e-02 1.30671054e-01 -3.69263709e-01
-7.89595902e-01 4.48441982e-01 3.85837972e-01 2.55606826e-02
1.13305736e+00 -6.37016177e-01 -1.52747154e-01 1.18353523e-01
-7.15346038e-01 4.45946544e-01 6.13197982e-01 6.56978071e-01
6.56424940e-01 -1.25777435e+00 -1.25457108e-01 3.11113596e-01
3.06519449e-01 9.81910288e-01 3.72801632e-01 6.86535418e-01
-8.15797508e-01 4.14066643e-01 1.64826706e-01 -1.01126158e+00
-1.04976118e+00 4.05088693e-01 1.40248671e-01 -8.84004161e-02
-6.79504693e-01 7.25752950e-01 5.55587351e-01 -9.36668813e-01
2.14460373e-01 -1.09911108e+00 3.68198127e-01 -4.02671605e-01
3.32946092e-01 1.40731558e-01 3.12728733e-01 -1.49793252e-01
-3.12220156e-01 6.83295131e-01 1.86863273e-01 1.87380686e-01
1.10424161e+00 -1.63591757e-01 3.94194156e-01 4.62435931e-01
9.03258860e-01 -1.78250987e-02 -1.78683877e+00 9.07203630e-02
-2.53126800e-01 -8.28927040e-01 -2.42679656e-01 -8.65841746e-01
-9.16340292e-01 7.43107140e-01 7.48857200e-01 1.03909500e-01
1.02486527e+00 2.09520265e-01 1.06837344e+00 6.53347254e-01
9.18088913e-01 -1.20806098e+00 2.43405655e-01 1.85574114e-01
7.18946218e-01 -1.66923714e+00 -1.13279529e-01 -9.33265448e-01
-8.64050210e-01 7.61683643e-01 9.03528512e-01 -9.18396637e-02
3.28225613e-01 6.74133599e-02 9.47368797e-03 -4.35731590e-01
1.66407861e-02 -2.26978004e-01 2.77269006e-01 7.74477720e-01
1.76227868e-01 4.02832568e-01 3.63608897e-01 8.78661275e-01
-3.40825289e-01 1.50885984e-01 2.77835995e-01 9.98679280e-01
-2.13128790e-01 -9.05510783e-01 -5.76093435e-01 4.52447057e-01
-6.43290728e-02 7.59373009e-02 -2.64187545e-01 1.02592325e+00
6.03351355e-01 8.60766888e-01 -6.11418160e-03 -2.48066604e-01
7.08735287e-01 -2.54395723e-01 5.14503181e-01 -9.78005707e-01
-2.80183643e-01 2.92550981e-01 -4.63485382e-02 -9.38981473e-01
-4.27593440e-01 -5.91306269e-01 -1.27196252e+00 2.19108135e-01
-4.78986859e-01 -3.65432799e-01 7.81145394e-01 9.05199468e-01
3.48553240e-01 6.06736422e-01 4.53629971e-01 -1.33107746e+00
-8.68127197e-02 -8.47849548e-01 -6.70645893e-01 6.49258912e-01
-9.91338342e-02 -7.61151731e-01 2.35844385e-02 2.71742970e-01] | [7.076288223266602, -1.9184362888336182] |
7805e721-d649-4503-b124-48018f699634 | sesql-yet-another-large-scale-session-level | 2208.12711 | null | https://arxiv.org/abs/2208.12711v1 | https://arxiv.org/pdf/2208.12711v1.pdf | SeSQL: Yet Another Large-scale Session-level Chinese Text-to-SQL Dataset | As the first session-level Chinese dataset, CHASE contains two separate parts, i.e., 2,003 sessions manually constructed from scratch (CHASE-C), and 3,456 sessions translated from English SParC (CHASE-T). We find the two parts are highly discrepant and incompatible as training and evaluation data. In this work, we present SeSQL, yet another large-scale session-level text-to-SQL dataset in Chinese, consisting of 5,028 sessions all manually constructed from scratch. In order to guarantee data quality, we adopt an iterative annotation workflow to facilitate intense and in-time review of previous-round natural language (NL) questions and SQL queries. Moreover, by completing all context-dependent NL questions, we obtain 27,012 context-independent question/SQL pairs, allowing SeSQL to be used as the largest dataset for single-round multi-DB text-to-SQL parsing. We conduct benchmark session-level text-to-SQL parsing experiments on SeSQL by employing three competitive session-level parsers, and present detailed analysis. | ['Min Zhang', 'Hua Wu', 'Xinyan Xiao', 'Fukang Yan', 'Chenhui Dou', 'Zeyang Liu', 'Zhenghua Li', 'Lijie Wang', 'Saihao Huang'] | 2022-08-26 | null | null | null | null | ['text-to-sql'] | ['computer-code'] | [-2.04306558e-01 -3.13912891e-02 1.09609868e-02 -9.98230517e-01
-1.85472846e+00 -1.07185721e+00 -3.23799923e-02 3.78518760e-01
-6.35862529e-01 7.85621107e-01 3.73558521e-01 -6.62220418e-01
8.54663923e-02 -8.81458640e-01 -1.10465777e+00 1.63633779e-01
2.47945815e-01 7.94263184e-01 5.26931167e-01 -3.65295589e-01
1.25375345e-01 -7.95813352e-02 -9.44927692e-01 8.77073646e-01
1.29247963e+00 7.08985865e-01 2.48082608e-01 6.87931478e-01
-8.04503143e-01 8.99278581e-01 -8.12143743e-01 -1.04139626e+00
-2.44446173e-01 -3.87514919e-01 -1.06988132e+00 -4.82085109e-01
4.15751070e-01 -1.22275606e-01 1.95945442e-01 6.19031608e-01
4.64535773e-01 -4.26972628e-01 -3.02485228e-01 -9.01037037e-01
-6.30593538e-01 1.00837696e+00 1.94841735e-02 -8.56615901e-02
7.19835937e-01 2.95231063e-02 1.30536640e+00 -8.53522718e-01
8.72013807e-01 8.94715667e-01 4.97606874e-01 6.10089481e-01
-8.91366005e-01 -5.61150432e-01 -3.40195447e-02 5.13626896e-02
-1.05624127e+00 -4.05902207e-01 4.18864369e-01 1.92262903e-02
1.19041359e+00 5.94039381e-01 3.56650680e-01 1.11210954e+00
-2.22387671e-01 8.97617400e-01 1.19715405e+00 -5.23494363e-01
2.18421951e-01 4.38555866e-01 2.73834616e-01 5.65140963e-01
1.52437016e-01 -3.70325387e-01 -5.17026544e-01 -1.45314127e-01
2.36886680e-01 -4.40901726e-01 8.10632855e-02 2.96245748e-03
-1.13153803e+00 8.16641688e-01 1.52388364e-02 2.57087052e-01
-1.57319024e-01 -3.16478461e-01 6.07049763e-01 4.34145838e-01
2.82138283e-03 4.89313602e-01 -1.04334760e+00 -5.85129976e-01
-3.13153714e-01 4.36843008e-01 1.28197455e+00 1.79912102e+00
7.19288349e-01 -4.64754283e-01 -2.71233678e-01 1.00332403e+00
-1.15635619e-01 5.93772411e-01 1.93324983e-01 -8.09202909e-01
1.25584996e+00 8.21117222e-01 -1.30828634e-01 -2.55447596e-01
-3.37197363e-01 5.47325192e-03 -3.73325795e-01 -9.48174715e-01
3.67063969e-01 -2.45199978e-01 -5.11915267e-01 1.65248108e+00
1.64257854e-01 -6.92311466e-01 3.85874838e-01 6.30613387e-01
1.31646442e+00 5.11028051e-01 4.00243789e-01 -2.87435129e-02
1.83590090e+00 -1.08769882e+00 -7.60788321e-01 -3.80921453e-01
9.59138036e-01 -8.93558741e-01 1.78400314e+00 3.18951249e-01
-1.29279578e+00 -4.13362265e-01 -5.05882800e-01 -5.19515336e-01
-4.88960862e-01 8.20365027e-02 6.45908594e-01 8.09116006e-01
-8.17785978e-01 -2.63272792e-01 -7.16676950e-01 -5.68181932e-01
-6.21222034e-02 -1.69678733e-01 -3.38370740e-01 -2.42974937e-01
-1.27919078e+00 6.30498111e-01 5.81479549e-01 -1.27915755e-01
-3.87740612e-01 -9.18667078e-01 -7.53588915e-01 -2.38683522e-02
1.01067936e+00 -1.12319961e-01 1.36734462e+00 4.77853976e-02
-1.21055841e+00 1.02988458e+00 -5.14005721e-01 -7.85711035e-02
3.15072834e-01 -3.84138227e-01 -7.74820566e-01 -9.63362772e-03
4.14894402e-01 3.71899664e-01 -2.09882766e-01 -9.61483657e-01
-7.70393610e-01 -4.56265420e-01 1.15519106e-01 1.05971619e-01
1.31779760e-02 4.77117211e-01 -1.39013350e+00 -4.89242166e-01
1.98015794e-01 -7.18821466e-01 9.69394743e-02 -6.21859372e-01
-6.45415187e-01 -3.78498852e-01 2.78675824e-01 -1.11986470e+00
1.65222490e+00 -2.13498759e+00 -3.53463948e-01 -1.13172516e-01
-1.18522964e-01 -1.22609725e-02 -2.62529165e-01 8.47994030e-01
2.08541512e-01 5.60198188e-01 -2.63354003e-01 -2.16173857e-01
1.27648637e-01 5.03971457e-01 -3.57591003e-01 -5.28398275e-01
4.92079824e-01 1.31509423e+00 -5.43422222e-01 -9.69901383e-01
-3.35091919e-01 -3.64577442e-01 -8.68992031e-01 6.89172685e-01
-6.43184483e-01 2.85848171e-01 -8.47850978e-01 9.47698712e-01
7.38870561e-01 -8.54808018e-02 4.24221694e-01 -1.31566048e-01
-1.93715781e-01 9.37096417e-01 -7.86136031e-01 2.12871480e+00
-6.35590255e-01 1.29642799e-01 5.43551371e-02 -5.74711978e-01
9.04989004e-01 1.75989881e-01 3.19177181e-01 -1.43414533e+00
-5.71623564e-01 5.41450560e-01 -4.12261218e-01 -9.71592963e-01
6.67954743e-01 1.63848475e-01 -1.10193658e+00 2.96512395e-01
1.49511635e-01 -3.57388020e-01 5.97976565e-01 5.44859648e-01
1.31563282e+00 -1.69490259e-02 1.07606031e-01 -2.52288207e-02
5.45103610e-01 5.33422351e-01 8.32125723e-01 9.08539891e-01
-3.89037281e-02 4.29177791e-01 9.56348360e-01 -2.23914742e-01
-1.09240913e+00 -1.18679667e+00 -2.50050902e-01 1.13432729e+00
-2.79854178e-01 -7.16487348e-01 -8.86546314e-01 -1.08905041e+00
-1.58592552e-01 9.73792791e-01 -2.49868080e-01 3.71398211e-01
-1.08114302e+00 -4.95537192e-01 1.01970446e+00 5.22679567e-01
6.91507936e-01 -1.31207860e+00 -2.89638638e-01 5.85809052e-01
-6.28504694e-01 -1.62651265e+00 -5.25539458e-01 3.91125709e-01
-5.09001434e-01 -1.29545701e+00 -6.66346459e-04 -7.93501556e-01
3.45103778e-02 -1.80634752e-01 1.80373204e+00 5.07358946e-02
-1.69310659e-01 1.68683171e-01 -6.97276175e-01 -3.83758456e-01
-4.63283718e-01 3.15655082e-01 -7.51017570e-01 -7.52503932e-01
8.14569414e-01 -6.97873384e-02 -2.30319038e-01 4.40984696e-01
-9.46443141e-01 1.28554165e-01 7.12424278e-01 5.68598449e-01
7.10047126e-01 -2.56711215e-01 7.02317834e-01 -1.47134852e+00
5.97614884e-01 -3.84694964e-01 -7.88759708e-01 1.02985513e+00
-4.84157026e-01 2.69099791e-02 8.60154271e-01 1.36369437e-01
-1.26958179e+00 -3.18609864e-01 -6.81033492e-01 4.28642929e-01
-2.50624776e-01 9.80611324e-01 -5.08024335e-01 5.13969362e-01
4.89003003e-01 3.63143981e-01 -3.66045028e-01 -7.42391646e-01
4.47424501e-01 7.96663582e-01 7.95401633e-01 -9.96495724e-01
5.01096547e-01 -2.20169321e-01 -9.60164011e-01 -4.38660979e-01
-1.09861767e+00 -4.47000742e-01 -4.04517323e-01 2.61915177e-01
1.13197958e+00 -9.86721098e-01 -9.04958487e-01 4.10833597e-01
-1.06842768e+00 -5.34571946e-01 -1.61858514e-01 1.07303604e-01
-2.18911320e-01 5.33817448e-02 -8.05495322e-01 -4.63302404e-01
-3.51676166e-01 -9.27063167e-01 1.13738251e+00 2.27307066e-01
1.59958795e-01 -6.67190075e-01 7.80297145e-02 8.98407221e-01
3.34685117e-01 -1.47192270e-01 1.26212966e+00 -1.08724535e+00
-6.80591702e-01 -2.35634357e-01 -3.78224760e-01 2.12286621e-01
-9.17669684e-02 -4.02829200e-02 -4.86639857e-01 -1.01671144e-02
1.80916302e-02 -9.04542804e-01 1.98138654e-01 -2.77147293e-01
1.45965576e+00 -2.30939716e-01 3.04804742e-01 5.09073377e-01
1.75602317e+00 4.35954303e-01 7.16810048e-01 3.57325047e-01
5.43219030e-01 7.68856704e-01 8.35663438e-01 2.10306704e-01
1.10686088e+00 5.57916045e-01 2.39086542e-02 9.99443457e-02
2.92860806e-01 -6.15826011e-01 1.77940652e-01 1.39905751e+00
6.61315203e-01 -1.90317050e-01 -1.02853632e+00 4.59754229e-01
-1.41302741e+00 -4.59865957e-01 -4.51672137e-01 2.01652288e+00
1.23473668e+00 3.82094055e-01 -6.94343522e-02 -5.75975239e-01
3.35547686e-01 4.85412963e-02 -5.86566031e-01 -4.56240356e-01
-3.64120036e-01 5.95879912e-01 3.46812457e-01 1.07326277e-01
-8.39509547e-01 1.22948265e+00 5.75606155e+00 7.28551984e-01
-7.39042163e-01 1.12179786e-01 4.38544959e-01 6.12206869e-02
-7.65999556e-01 2.21957192e-01 -1.13894498e+00 7.10644484e-01
1.29443014e+00 -1.18618198e-01 3.48759592e-01 7.64105022e-01
-4.90569621e-02 -2.73618758e-01 -1.06779289e+00 5.06374955e-01
-1.81418166e-01 -1.39109826e+00 -1.85613930e-01 -3.11572999e-01
3.74811679e-01 1.05836362e-01 -4.95601535e-01 1.16105044e+00
4.03358251e-01 -6.93184972e-01 6.19455457e-01 2.38217905e-01
9.37358379e-01 -4.53782529e-01 8.38843584e-01 4.49833632e-01
-1.32007372e+00 6.50507361e-02 -2.90876031e-01 7.86598802e-01
3.28303814e-01 3.98847193e-01 -4.77833062e-01 1.12343371e+00
1.17306995e+00 1.63465068e-01 -1.09887993e+00 4.79027599e-01
-1.06747948e-01 8.38423073e-01 -3.87898028e-01 -3.98464203e-01
2.16440260e-01 -2.53458142e-01 -9.82722640e-02 1.26350105e+00
1.41971782e-01 3.50200295e-01 7.24476650e-02 8.37926805e-01
-3.27455491e-01 3.22054178e-01 -1.45300984e-01 -1.85115486e-01
9.24920738e-01 1.24574733e+00 -1.08866006e-01 -5.76536298e-01
-8.81215572e-01 7.98350573e-01 6.94040537e-01 2.93958306e-01
-1.00902259e+00 -5.03638089e-01 1.47184938e-01 -1.96062580e-01
1.72590747e-01 -4.52596433e-02 -3.48308593e-01 -1.40796447e+00
8.09324682e-01 -1.35829246e+00 7.93652296e-01 -7.00554013e-01
-1.36592233e+00 7.10414410e-01 -1.10526413e-01 -7.83605874e-01
-1.56939626e-01 -3.68619174e-01 -5.60280383e-01 9.74148691e-01
-1.52222002e+00 -1.15177202e+00 -5.45879126e-01 6.77606285e-01
6.13245428e-01 -4.58465796e-03 9.48053539e-01 8.20263445e-01
-8.55448246e-01 8.62117052e-01 -1.49658635e-01 4.05202955e-01
9.98104870e-01 -1.47933853e+00 6.66229665e-01 6.93998754e-01
-3.13427113e-02 8.60771060e-01 1.76452398e-01 -7.71897435e-01
-1.88656104e+00 -9.89736736e-01 1.41463292e+00 -7.90012836e-01
7.37354875e-01 -9.07155812e-01 -1.38721907e+00 8.19227874e-01
2.65839249e-02 -1.68993965e-01 7.40317643e-01 3.20536196e-01
-3.53512585e-01 -2.97954172e-01 -9.11146581e-01 3.40329081e-01
9.36049640e-01 -9.01007652e-01 -6.42429888e-01 3.02384555e-01
1.35230863e+00 -8.41612160e-01 -1.14855206e+00 4.73701924e-01
2.94648081e-01 -9.92644310e-01 5.60105145e-01 -8.51988494e-01
5.53198755e-01 2.54936844e-01 -3.66096467e-01 -5.03104806e-01
3.22708249e-01 -3.44368696e-01 1.75403655e-01 1.58010280e+00
8.53376448e-01 -5.53507090e-01 1.00717032e+00 8.42128575e-01
-5.09572625e-01 -8.63654852e-01 -8.07989419e-01 -5.81505358e-01
2.48396903e-01 -7.47142851e-01 9.58395958e-01 8.71775985e-01
-2.58482844e-01 2.89466381e-01 -5.32656200e-02 -1.81287769e-02
2.61528701e-01 2.64509946e-01 1.03254187e+00 -5.97974539e-01
-2.56110698e-01 2.53587142e-02 6.59475207e-01 -1.21765947e+00
-2.20470607e-01 -6.81903124e-01 1.88844517e-01 -1.43097186e+00
2.01938435e-01 -9.82552528e-01 6.19457811e-02 4.80751663e-01
-4.11260784e-01 -4.66866910e-01 4.31727991e-03 7.55723491e-02
-8.21842790e-01 3.33360463e-01 1.04578567e+00 3.78236979e-01
-2.06973493e-01 -9.61942971e-02 -1.00089705e+00 5.47360927e-02
4.89725053e-01 -4.29520309e-01 -1.53558716e-01 -7.73682117e-01
4.58454043e-01 6.25101566e-01 2.97565218e-02 -6.42781556e-01
3.75845224e-01 -2.91093320e-01 -1.47492930e-01 -1.06374002e+00
-1.38927102e-01 -6.49137735e-01 -8.70750695e-02 1.55024752e-01
-6.10646546e-01 3.90558541e-01 2.05086932e-01 1.84228703e-01
-7.03788996e-01 -4.34577674e-01 2.14129314e-01 -2.91441023e-01
-8.83820832e-01 2.76826799e-01 5.48521697e-04 8.94965410e-01
7.33678341e-01 1.42979935e-01 -7.26224422e-01 -1.33960336e-01
-4.38782811e-01 8.12025666e-01 1.02165371e-01 5.97236395e-01
3.19186002e-01 -1.00409591e+00 -8.08639407e-01 3.47219974e-01
6.34232998e-01 3.58497560e-01 6.39529467e-01 5.13425350e-01
-8.12767804e-01 9.44823325e-01 7.37927482e-02 -4.24413651e-01
-9.68740761e-01 2.71026641e-01 -2.02072740e-01 -7.22519457e-01
-4.57261205e-01 8.09276760e-01 -2.68348038e-01 -1.24488759e+00
2.78676510e-01 -3.31495792e-01 1.32621497e-01 -2.28854880e-01
1.81637049e-01 -2.00108260e-01 4.39343840e-01 1.49108514e-01
-5.61278343e-01 2.11859956e-01 -1.96805820e-01 -2.55762249e-01
1.34225237e+00 -1.26919225e-01 -4.00756508e-01 2.82910317e-01
1.38154852e+00 4.57531691e-01 -7.14248419e-01 -3.98992777e-01
6.02144718e-01 -2.21582413e-01 -7.11081505e-01 -1.30413473e+00
-7.70275295e-01 6.30301535e-01 1.36840731e-01 2.73099661e-01
1.29608309e+00 3.02133501e-01 1.12085247e+00 7.04712808e-01
5.55790007e-01 -1.25194669e+00 -1.47361066e-02 6.94929838e-01
8.59378994e-01 -1.41677487e+00 -3.43430340e-01 -6.01233780e-01
-8.93669248e-01 9.04520214e-01 1.14370072e+00 4.94725436e-01
2.53349811e-01 3.00996333e-01 3.01099449e-01 -3.23511094e-01
-1.06529140e+00 7.74975345e-02 1.35043308e-01 2.24073768e-01
6.29502535e-01 -7.72532076e-02 -4.09451813e-01 1.11548567e+00
-5.99839211e-01 -8.93532932e-02 3.04170638e-01 1.40862465e+00
-9.55349114e-03 -1.58014762e+00 7.65241086e-02 5.66611290e-01
-6.21242523e-01 -3.35407168e-01 -2.94594616e-01 1.16344678e+00
-2.62025356e-01 9.86135006e-01 1.32859305e-01 -2.10701257e-01
9.60536063e-01 2.70079255e-01 4.14927378e-02 -7.64726520e-01
-1.04789937e+00 -5.61070926e-02 6.61545694e-01 -7.71710575e-01
8.97804201e-02 -4.69232649e-01 -1.39299667e+00 -3.30282032e-01
-1.67653281e-02 5.71505606e-01 9.26254630e-01 6.46734893e-01
4.91396517e-01 3.33694577e-01 5.86602807e-01 7.16015875e-01
-6.31297469e-01 -8.47148776e-01 -1.81429222e-01 5.30960739e-01
-1.99023604e-01 1.99882612e-01 3.96316685e-02 -8.90833884e-02] | [9.93319034576416, 7.880223751068115] |
3d45dec7-aa77-4da5-8568-695aa93446e0 | layout-aware-dreamer-for-embodied-referring | 2212.00171 | null | https://arxiv.org/abs/2212.00171v2 | https://arxiv.org/pdf/2212.00171v2.pdf | Layout-aware Dreamer for Embodied Referring Expression Grounding | In this work, we study the problem of Embodied Referring Expression Grounding, where an agent needs to navigate in a previously unseen environment and localize a remote object described by a concise high-level natural language instruction. When facing such a situation, a human tends to imagine what the destination may look like and to explore the environment based on prior knowledge of the environmental layout, such as the fact that a bathroom is more likely to be found near a bedroom than a kitchen. We have designed an autonomous agent called Layout-aware Dreamer (LAD), including two novel modules, that is, the Layout Learner and the Goal Dreamer to mimic this cognitive decision process. The Layout Learner learns to infer the room category distribution of neighboring unexplored areas along the path for coarse layout estimation, which effectively introduces layout common sense of room-to-room transitions to our agent. To learn an effective exploration of the environment, the Goal Dreamer imagines the destination beforehand. Our agent achieves new state-of-the-art performance on the public leaderboard of the REVERIE dataset in challenging unseen test environments with improvement in navigation success (SR) by 4.02% and remote grounding success (RGS) by 3.43% compared to the previous state-of-the-art. The code is released at https://github.com/zehao-wang/LAD | ['Marie-Francine Moens', 'Tinne Tuytelaars', 'Zehao Wang', 'Mingxiao Li'] | 2022-11-30 | null | null | null | null | ['referring-expression', 'common-sense-reasoning'] | ['computer-vision', 'reasoning'] | [-6.92970008e-02 2.77782351e-01 3.86019021e-01 -4.77101207e-01
-7.47590303e-01 -5.97397566e-01 3.67658824e-01 4.38254535e-01
-5.19847393e-01 5.83574176e-01 2.94905156e-01 -3.53998989e-01
-9.17496085e-02 -9.75854218e-01 -1.12748659e+00 -8.19698870e-01
-7.80836493e-02 7.21746564e-01 -1.19284809e-01 -5.07203758e-01
2.62431830e-01 3.32408249e-01 -1.57464957e+00 1.01062588e-01
7.98274517e-01 6.21952891e-01 9.71397996e-01 5.87020457e-01
7.41612092e-02 9.28571999e-01 -6.22270942e-01 3.88597667e-01
7.84842297e-02 -3.23118716e-01 -9.69128907e-01 6.90835044e-02
2.82874674e-01 -2.04895154e-01 -4.25178736e-01 9.12920654e-01
4.32770371e-01 1.03949940e+00 3.00526440e-01 -1.15896177e+00
-6.10135376e-01 7.10214794e-01 -4.10219967e-01 1.69926167e-01
6.88448191e-01 3.27904075e-01 6.07301295e-01 -6.60856366e-01
4.48000461e-01 1.25123847e+00 1.04507752e-01 4.22561198e-01
-8.63546014e-01 -3.79598558e-01 7.91314721e-01 3.22349250e-01
-1.57608461e+00 -2.86073357e-01 7.22248614e-01 -1.43529475e-01
8.21829081e-01 1.68092400e-01 5.84537268e-01 1.17867959e+00
2.55798757e-01 8.32388639e-01 1.16363657e+00 -4.57460314e-01
8.08089972e-01 -1.64511483e-02 2.05289032e-02 1.13091612e+00
-5.48774190e-02 -2.24550758e-02 -7.19720721e-01 2.16203198e-01
5.71376801e-01 7.11957589e-02 -5.15390635e-01 -5.33355772e-01
-1.30885804e+00 6.28966868e-01 9.35394049e-01 3.13528568e-01
-4.62117285e-01 7.58077204e-02 -1.74864709e-01 -7.54397362e-02
-4.94941929e-03 8.30793142e-01 -3.03745605e-02 -2.51844287e-01
-3.85250658e-01 3.33780736e-01 7.89785028e-01 1.12551320e+00
8.75804245e-01 -2.42039070e-01 -7.34579712e-02 3.33628029e-01
1.56549469e-01 4.20143247e-01 3.26952815e-01 -1.07689917e+00
6.70228899e-01 6.45289183e-01 3.91391665e-01 -7.40806162e-01
-6.59854412e-01 -4.67176944e-01 -5.63607395e-01 2.77226835e-01
3.73967141e-01 -1.03667974e-01 -1.06098080e+00 1.91480899e+00
4.70971704e-01 2.02906117e-01 3.31788540e-01 1.04485488e+00
8.40937018e-01 7.78247833e-01 -2.55157966e-02 3.99412930e-01
1.45897627e+00 -1.49907160e+00 -6.53132737e-01 -9.34235275e-01
8.13257158e-01 -2.43039802e-01 1.59142959e+00 3.48082691e-01
-7.39439309e-01 -6.60693884e-01 -1.11491537e+00 -5.63401096e-02
-6.72607660e-01 4.71138135e-02 5.30704856e-01 1.28695726e-01
-1.05145264e+00 2.59012818e-01 -1.06089365e+00 -6.48630440e-01
2.78368026e-01 5.78503646e-02 -6.03486001e-01 -4.50130373e-01
-7.99530566e-01 8.52369487e-01 3.53772342e-01 3.67335081e-01
-1.55164933e+00 -3.99633884e-01 -1.32159579e+00 2.02575624e-02
6.91511810e-01 -7.96928465e-01 1.21349072e+00 -2.75715530e-01
-1.41205990e+00 7.25469708e-01 -4.00466532e-01 -2.22129256e-01
2.59765148e-01 -4.76638585e-01 -1.97484747e-01 -1.35343388e-01
4.92535681e-01 6.35715485e-01 9.49071646e-02 -1.50634038e+00
-6.14999950e-01 -5.43971598e-01 5.32377362e-01 7.92948484e-01
3.26357871e-01 -8.48232687e-01 -4.51098830e-01 -1.82365682e-02
4.93351549e-01 -1.16706955e+00 -4.48489368e-01 -3.48822594e-01
-5.96043289e-01 -3.00063491e-01 4.59375769e-01 -4.16057140e-01
9.32735860e-01 -2.26149631e+00 1.50458515e-01 -4.44248244e-02
1.56971425e-01 -4.27893996e-01 -1.04967862e-01 5.58289111e-01
1.86286107e-01 -2.41887048e-01 -1.43180072e-01 -5.96356750e-01
1.19375817e-01 3.43140811e-01 -3.85406524e-01 4.92691576e-01
-5.47595978e-01 7.04238534e-01 -1.39446557e+00 -1.45123646e-01
4.00030345e-01 3.18803757e-01 -4.16111439e-01 5.09134352e-01
-1.58740699e-01 8.99304807e-01 -7.24776328e-01 4.39364702e-01
4.48035806e-01 -2.77761132e-01 1.19770423e-01 3.06111842e-01
-3.52084011e-01 5.35279930e-01 -1.20366704e+00 2.47811937e+00
-8.34250629e-01 6.30961180e-01 -2.26916801e-02 -4.45621610e-01
6.95660830e-01 -1.31501377e-01 -2.64604092e-01 -9.49498296e-01
-1.77391935e-02 -1.06832825e-01 -2.48950481e-01 -6.56030536e-01
5.99709034e-01 2.20941290e-01 -3.26387465e-01 2.52913088e-01
-3.48679483e-01 -6.03426099e-02 -1.53838977e-01 3.00887793e-01
1.25913131e+00 4.57310706e-01 3.05954129e-01 -4.58498448e-01
1.82688549e-01 2.85902262e-01 2.29998201e-01 1.00915766e+00
-6.11366443e-02 3.35675955e-01 1.19637378e-01 -5.76384425e-01
-3.90050620e-01 -1.37467289e+00 4.52561915e-01 1.37228417e+00
8.37888420e-01 -4.07584816e-01 -8.93296123e-01 -4.04970825e-01
-5.05575359e-01 1.40497696e+00 -8.95736158e-01 -1.52225792e-01
-8.28314066e-01 -1.67877465e-01 1.59351844e-02 6.59817874e-01
8.29827607e-01 -1.18336117e+00 -1.19864905e+00 -2.01871656e-02
-5.94934523e-01 -9.95677710e-01 -4.11665201e-01 5.00494123e-01
-4.15193677e-01 -9.50224042e-01 -1.08875431e-01 -9.24484551e-01
1.08545196e+00 4.51188087e-01 1.23052609e+00 -2.35272553e-02
-1.90087512e-01 3.86377752e-01 -2.57418752e-01 -3.68171960e-01
6.42683953e-02 1.71683729e-01 4.34576906e-02 -4.14089650e-01
2.35249713e-01 -3.71334910e-01 -8.35222661e-01 3.01044941e-01
-3.26576501e-01 4.41729784e-01 4.88467723e-01 6.60912335e-01
7.50361502e-01 3.94764453e-01 -4.87121046e-02 -6.07443631e-01
4.08715367e-01 -4.78107363e-01 -6.00693047e-01 9.49919075e-02
-2.92763382e-01 4.03374463e-01 5.28163373e-01 -2.34896675e-01
-1.09266031e+00 2.42754389e-02 4.19630446e-02 -7.28773773e-02
-7.08155096e-01 3.01984787e-01 -5.00910699e-01 3.85210514e-01
5.87427318e-01 2.57692337e-01 -5.26645482e-01 -4.16899055e-01
3.65384191e-01 1.78614452e-01 7.29308248e-01 -9.58350360e-01
5.80269456e-01 4.47897434e-01 4.28205952e-02 -6.03655636e-01
-1.14156950e+00 -5.12112081e-01 -5.20534515e-01 -1.15997903e-01
1.12293649e+00 -1.00646985e+00 -1.09952998e+00 8.46237913e-02
-8.64496529e-01 -1.16699326e+00 -2.00421125e-01 3.01122695e-01
-6.15531147e-01 -4.18344826e-01 -1.10295676e-01 -8.08849037e-01
2.80007780e-01 -1.27878773e+00 1.17428374e+00 4.96427298e-01
-1.35792226e-01 -1.04433286e+00 4.90439944e-02 4.66717333e-01
2.27878600e-01 5.11216700e-01 8.41541290e-01 -4.01194602e-01
-9.27332819e-01 2.23620936e-01 2.64284402e-01 -4.49412733e-01
2.39419624e-01 -9.21960890e-01 -1.02388501e+00 -3.72385025e-01
-1.22799920e-02 -9.83667299e-02 6.10447943e-01 1.52675733e-01
1.12898815e+00 -3.36534500e-01 -6.95801973e-01 5.33153474e-01
1.43300521e+00 4.49104697e-01 5.28183103e-01 8.41893792e-01
7.08902597e-01 6.75240040e-01 9.29583967e-01 3.59652966e-01
9.72106516e-01 6.08053029e-01 8.70341718e-01 -4.35063839e-02
8.86341780e-02 -6.70620024e-01 1.80896699e-01 3.53362411e-01
1.87086001e-01 -7.22124219e-01 -1.11616492e+00 7.53692985e-01
-1.90538526e+00 -7.04175591e-01 3.34367603e-01 2.06315041e+00
3.84624094e-01 7.97954202e-02 -3.89485598e-01 -4.12992656e-01
3.32657188e-01 3.92886221e-01 -6.50652468e-01 -8.18527117e-02
1.74160510e-01 -2.03149572e-01 3.35890114e-01 8.85916829e-01
-9.13124561e-01 1.17223465e+00 4.47429895e+00 4.02896792e-01
-6.58202171e-01 -6.84893224e-04 6.60179079e-01 -1.06699921e-01
-3.16830464e-02 -7.10646212e-02 -9.06393170e-01 1.46677241e-01
7.91836619e-01 3.22376788e-01 8.60228717e-01 9.49821234e-01
2.13482469e-01 -5.80852747e-01 -1.42249691e+00 9.85467434e-01
9.48665589e-02 -9.90480423e-01 -3.10830861e-01 -1.24044567e-01
5.61659932e-01 -4.74403538e-02 3.62269908e-01 5.22686124e-01
5.61590850e-01 -1.25675428e+00 9.00810599e-01 6.36278570e-01
1.96835726e-01 -7.18137085e-01 6.49640799e-01 8.06311548e-01
-1.28817952e+00 -1.71014294e-01 -1.07895568e-01 -3.15281302e-01
6.04421571e-02 -4.69478108e-02 -1.13825321e+00 3.46770376e-01
9.42108452e-01 1.20411374e-01 -4.64899391e-01 7.67501056e-01
-6.99339747e-01 1.31725892e-01 -2.19831303e-01 -1.70379296e-01
7.02094197e-01 -2.23398089e-01 5.50787985e-01 6.16116107e-01
4.89932299e-01 5.12358010e-01 4.28586930e-01 9.26531196e-01
6.20321259e-02 -2.88946033e-01 -7.35470593e-01 4.61082935e-01
5.87353885e-01 1.14005244e+00 -7.43591011e-01 -1.06899060e-01
1.45396844e-01 1.09022224e+00 6.03361547e-01 6.24183178e-01
-8.04043412e-01 -3.12292308e-01 6.17943704e-01 3.24123949e-01
1.72530249e-01 -4.40494150e-01 -4.42912765e-02 -6.90192282e-01
-7.00907130e-03 -6.29319191e-01 1.58958927e-01 -1.32303929e+00
-5.82806766e-01 9.35410917e-01 5.05174808e-02 -9.63749588e-01
-5.92671782e-02 -4.66420859e-01 -7.01944292e-01 8.33453000e-01
-1.36446393e+00 -8.50487411e-01 -1.01334989e+00 2.90847570e-01
9.36533868e-01 -4.87411469e-02 1.02843893e+00 -2.52669245e-01
-5.05208671e-01 1.80685699e-01 -1.59629047e-01 3.94915044e-03
5.84465206e-01 -1.61303473e+00 4.00353163e-01 7.96925187e-01
2.35311255e-01 9.50574875e-01 1.09262538e+00 -4.46694046e-01
-1.50001097e+00 -1.11299896e+00 6.12806261e-01 -7.86237478e-01
3.08281690e-01 -6.90963626e-01 -7.37726092e-01 1.10248053e+00
2.61896580e-01 -9.71369818e-02 4.88546133e-01 1.33334130e-01
-1.34268314e-01 4.69616503e-02 -1.00252330e+00 1.06479156e+00
1.29848087e+00 -3.27573985e-01 -7.60480285e-01 4.86277819e-01
7.73017287e-01 -9.71463621e-01 -2.43044600e-01 -1.42208919e-01
7.15093464e-02 -1.04226518e+00 7.48902142e-01 -2.95784861e-01
1.36940986e-01 -6.65804863e-01 -4.85734314e-01 -1.73446095e+00
-2.83136040e-01 -5.69107294e-01 1.12860590e-01 7.53962934e-01
3.42839688e-01 -6.20050311e-01 8.51354837e-01 5.97455621e-01
-5.72260737e-01 -9.11604226e-01 -8.00437570e-01 -5.92994928e-01
-3.29138964e-01 -1.62155688e-01 9.57550585e-01 6.15923107e-01
-6.30814657e-02 4.89006639e-01 5.98893054e-02 8.27169538e-01
5.30376792e-01 3.68430853e-01 8.83262932e-01 -8.11001897e-01
-1.21298514e-01 -1.01200648e-01 -5.36141768e-02 -1.59381974e+00
4.56090003e-01 -7.03260958e-01 6.55083179e-01 -2.16675568e+00
-3.15251807e-03 -5.18852830e-01 -3.28437060e-01 6.05977118e-01
-3.54538485e-02 -3.40508997e-01 5.76480106e-02 -2.20901296e-01
-1.18637490e+00 5.55980325e-01 1.38613403e+00 -2.79519916e-01
-5.40049493e-01 -9.38012823e-02 -7.86757231e-01 8.32010865e-01
8.90690327e-01 -3.98910314e-01 -7.18693852e-01 -7.44593441e-01
2.71154016e-01 2.61230350e-01 5.43591797e-01 -1.23140049e+00
6.82382524e-01 -1.36043832e-01 3.22451025e-01 -7.17403650e-01
7.78952003e-01 -8.11166763e-01 4.08702455e-02 4.17939901e-01
-5.13649225e-01 4.45530981e-01 5.22452056e-01 6.98995948e-01
9.64697972e-02 -7.75128007e-02 2.57559299e-01 -4.69523489e-01
-1.31593847e+00 5.33677218e-03 -3.19412112e-01 8.66524056e-02
1.05107152e+00 -1.69106141e-01 -6.86891258e-01 -3.50120604e-01
-7.98774362e-01 6.42692089e-01 5.98545969e-01 4.90884006e-01
8.00728083e-01 -1.06166959e+00 -3.83469105e-01 2.05313683e-01
3.35690647e-01 4.54293787e-01 6.18842900e-01 5.98738253e-01
-5.99164724e-01 3.69528264e-01 -1.58411637e-01 -5.12067020e-01
-9.93054569e-01 6.92014873e-01 4.54325527e-01 -1.57882512e-01
-9.84861553e-01 1.13060355e+00 7.72519171e-01 -6.26736939e-01
4.22665238e-01 -5.06257713e-01 -1.84645116e-01 -5.09041190e-01
6.00303292e-01 4.37328100e-01 -5.55991270e-02 -5.04718482e-01
-5.50488651e-01 4.11668897e-01 1.75983623e-01 -1.05052106e-01
1.10374820e+00 -3.63322318e-01 -3.01792435e-02 8.29075515e-01
9.23386872e-01 2.16361489e-02 -1.33931410e+00 -5.35658523e-02
-2.58470982e-01 -3.64255279e-01 7.07682148e-02 -1.07990873e+00
-5.26196778e-01 7.37968504e-01 7.40195036e-01 -1.13943353e-01
9.38271463e-01 2.97268301e-01 4.70443994e-01 9.09030914e-01
9.16911185e-01 -8.37200642e-01 3.05092752e-01 5.13148844e-01
9.41032171e-01 -1.38913751e+00 -1.77430674e-01 -7.37727508e-02
-7.75529861e-01 6.73204005e-01 1.01254094e+00 1.17595963e-01
3.17631274e-01 1.44885272e-01 1.23233825e-01 -4.61167544e-01
-7.38070965e-01 -8.42764154e-02 1.01211645e-01 6.84988499e-01
8.13409090e-02 4.15833533e-01 7.23039746e-01 4.59234715e-01
-7.37860501e-01 -6.38978124e-01 5.32699347e-01 1.03748906e+00
-6.44395709e-01 -4.04670447e-01 -4.79891807e-01 -1.43883914e-01
2.10950628e-01 -3.13106626e-02 -9.91436467e-03 8.86987865e-01
2.25781798e-01 1.20199335e+00 3.56055260e-01 -1.57633588e-01
5.08989036e-01 -1.46542102e-01 4.77591664e-01 -7.92030513e-01
-3.46779138e-01 -2.13354900e-01 6.88231811e-02 -9.30955708e-01
-1.25949860e-01 -4.62056786e-01 -1.83814669e+00 -1.82235450e-01
1.80629998e-01 4.82970327e-01 6.35519266e-01 8.97035837e-01
4.15373415e-01 1.08735645e+00 2.27221906e-01 -9.41461444e-01
-7.12964386e-02 -6.93513751e-01 -4.93285269e-01 7.12319911e-02
6.44346952e-01 -8.46634150e-01 -2.59570956e-01 -1.82631373e-01] | [4.491841793060303, 0.5150647163391113] |
877b03bd-13c1-41fe-918d-0076f8a3998d | performance-comparison-of-large-language | 2307.02288 | null | https://arxiv.org/abs/2307.02288v2 | https://arxiv.org/pdf/2307.02288v2.pdf | Performance Comparison of Large Language Models on VNHSGE English Dataset: OpenAI ChatGPT, Microsoft Bing Chat, and Google Bard | This paper presents a performance comparison of three large language models (LLMs), namely OpenAI ChatGPT, Microsoft Bing Chat, and Google Bard, on the VNHSGE English dataset. The results show that BingChat is better than ChatGPT and Bard. Therefore, BingChat and Bard can replace ChatGPT while ChatGPT is not yet officially available in Vietnam. The results also indicate that ChatGPT, Bing Chat, and Bard outperform Vietnamese students in English language proficiency. The findings of this study contribute to the understanding of the potential of LLMs in English language education. The remarkable performance of ChatGPT, Bing Chat, and Bard demonstrates their potential as effective tools for teaching and learning English at the high school level. | ['Xuan-Quy Dao'] | 2023-07-05 | null | null | null | null | ['question-answering'] | ['natural-language-processing'] | [-1.1608789e+00 -1.9392239e-01 -4.4754732e-01 1.0566199e-01
-9.7775859e-01 -6.0358751e-01 3.9871374e-01 4.9235240e-01
-6.4108521e-01 8.8256723e-01 1.4305198e-01 -1.1938204e+00
-7.1559615e-02 -8.0176228e-01 -3.6174998e-01 -2.9618427e-01
5.0726789e-01 2.8320915e-01 3.3785290e-01 -7.0123452e-01
2.3807980e-01 2.3688364e-01 -1.1732959e+00 1.3785973e-02
1.5203277e+00 -2.5178960e-02 6.5839672e-01 7.1281803e-01
-4.8059484e-01 8.9583057e-01 -7.9675686e-01 -8.9900160e-01
-3.2934651e-01 -2.5910899e-01 -8.8353014e-01 -6.1389965e-01
3.0532134e-01 -4.2715365e-01 -2.3495886e-01 7.7462584e-01
6.2662995e-01 4.9835697e-01 2.2975260e-01 -1.1598458e+00
-1.1233392e+00 9.3242472e-01 -2.8775123e-01 2.8159317e-01
6.4054728e-01 -1.6397445e-01 8.9296037e-01 -8.4595829e-01
6.6348040e-01 1.3038666e+00 6.6565484e-01 4.1063058e-01
-6.8543512e-01 -9.3766576e-01 -9.7773537e-02 -6.3792840e-02
-1.4743208e+00 2.3464876e-01 6.2928379e-02 -5.7968789e-01
1.0115514e+00 1.5672755e-01 1.1430825e+00 1.0444726e+00
6.0807556e-01 8.4046245e-01 1.3734317e+00 -5.4622477e-01
1.4761550e-02 4.4449845e-01 2.5458691e-01 6.9845045e-01
2.2632936e-01 -4.0587124e-01 -6.2733078e-01 -2.7267210e-02
6.6813338e-01 -2.3408471e-01 1.7342976e-01 4.5805538e-01
-9.0911746e-01 1.2375009e+00 2.4546561e-01 6.4294469e-01
-4.3136351e-02 -9.6145876e-02 3.1839046e-01 6.3950568e-01
8.2180840e-01 1.9661549e-01 -3.3057597e-01 -1.1987164e+00
-4.5165482e-01 1.4452754e-01 9.4988173e-01 1.2695115e+00
2.5409475e-01 1.5127939e-01 2.7632669e-01 1.1390765e+00
4.9486330e-01 3.9475507e-01 6.7950016e-01 -5.8296698e-01
6.3540864e-01 5.9330970e-01 -5.3156793e-01 -7.6837552e-01
1.3488410e-01 2.0215967e-01 -2.0908664e-01 -4.3065783e-01
5.6978995e-01 -4.3786883e-01 -6.0562491e-01 1.0581532e+00
2.4776955e-01 9.2913285e-02 2.1043581e-01 3.4254831e-01
1.4495492e+00 1.1417345e+00 5.7413012e-01 2.8693423e-01
1.1203532e+00 -1.2057662e+00 -9.8545688e-01 -1.8302253e-01
1.4578365e+00 -1.0860429e+00 1.3736736e+00 3.4557360e-01
-1.1577847e+00 -6.8272299e-01 -5.5441463e-01 -4.0689310e-01
-8.9587718e-01 -1.4100420e-01 8.8950223e-01 1.3356440e+00
-1.2452980e+00 8.5353725e-02 -7.4815071e-01 -5.4779220e-01
-1.7701678e-01 1.8425710e-01 -5.4833692e-01 -4.3431550e-01
-1.1202813e+00 1.0892817e+00 5.1123101e-01 -7.0628479e-02
-6.4721406e-01 -5.8124119e-01 -1.0774786e+00 8.4976479e-02
9.8467588e-02 3.7487417e-01 1.5215772e+00 -1.3100056e-01
-1.6940351e+00 8.7138546e-01 1.6486092e-01 -1.4612867e-01
4.1543561e-01 -2.2959055e-01 -3.5310993e-01 -1.3814980e-01
2.6223969e-01 3.9169222e-01 -3.5240912e-01 -5.1047325e-01
-6.3418728e-01 7.6654572e-03 5.0000608e-02 2.4946102e-01
-4.8605907e-01 4.9936333e-01 -3.7770295e-01 -3.3857661e-01
-2.8274752e-02 -7.1548110e-01 -2.6309544e-01 -6.4522779e-01
-2.7380192e-01 -8.5280585e-01 8.8605523e-01 -1.0070000e+00
1.6316880e+00 -2.1393385e+00 -4.1586959e-01 2.1559025e-01
2.3461743e-01 4.1024351e-01 -1.1400803e-01 8.8144726e-01
5.0582880e-01 7.2572654e-01 7.3529512e-01 -4.8125897e-02
2.0710935e-01 5.9438461e-01 3.4423465e-01 1.6471724e-01
-5.7603496e-01 9.2535990e-01 -1.0070249e+00 -6.7651612e-01
4.7165313e-01 3.4008592e-01 -6.2747043e-01 1.4656338e-01
1.6310515e-01 3.6187845e-01 -3.6989933e-01 6.1088187e-01
3.3149320e-01 3.5378803e-02 1.5977861e-01 1.0635426e+00
-8.5313010e-01 7.2496134e-01 -7.8293556e-01 1.3506795e+00
-6.6609108e-01 8.3992893e-01 -2.9163193e-02 -5.6925982e-01
1.1499878e+00 7.6404870e-01 9.4750330e-02 -7.8838158e-01
2.7478784e-01 7.0953113e-01 3.6020273e-01 -4.0661514e-01
7.7831829e-01 3.8985032e-01 -1.4739515e-01 4.8622912e-01
3.0900723e-01 -6.5452820e-01 4.3347251e-01 4.4532019e-01
3.3508646e-01 -8.5503854e-02 2.5802433e-01 -5.4936433e-01
2.6739603e-01 -3.0021942e-01 2.4052225e-01 9.1006827e-01
-3.4474593e-01 -1.1138085e-01 5.0436455e-01 1.2178630e-01
-7.8937721e-01 -9.2448360e-01 -2.0010283e-02 1.6197618e+00
-1.5086414e-01 -8.5720038e-01 -8.4788859e-01 -4.4008079e-01
-2.8000563e-01 8.3583975e-01 6.9082007e-02 2.6461568e-01
-3.4929323e-01 -1.8122941e-02 4.4425258e-01 2.6839888e-01
7.7977294e-01 -7.5303173e-01 -2.6433298e-02 2.9328361e-01
-2.4997187e-01 -1.1289181e+00 -6.8304145e-01 -2.8432338e-02
-6.1811316e-01 -3.3913809e-01 -9.9936455e-01 -1.3587276e+00
1.6444536e-01 3.0688033e-01 7.5518870e-01 4.0069932e-01
9.1997683e-02 5.0484723e-01 -5.1318067e-01 -7.1035713e-01
-8.1092715e-01 5.4039776e-01 -3.0019745e-02 -9.5725161e-01
7.6445502e-01 -3.5115427e-01 7.8826964e-02 1.5043461e-01
-2.9118139e-01 3.5698396e-01 1.8358149e-01 6.1376065e-01
-1.9532578e-02 -8.5931823e-02 6.7811781e-01 -8.6605948e-01
8.3169687e-01 -6.6431099e-01 -6.6298145e-01 4.4917036e-02
-4.5796737e-01 -7.6834923e-01 3.3705211e-01 -5.7187903e-01
-1.1712953e+00 -9.1543132e-01 -9.2507261e-01 6.4012192e-02
-5.5344231e-02 1.1382828e+00 2.5945032e-02 -4.1434404e-01
2.2240986e-01 -6.5215938e-02 -1.6657880e-01 -6.7116493e-01
-8.8615194e-02 1.1057580e+00 1.5796918e-02 -7.6738727e-01
3.2862946e-01 -5.1637334e-01 -7.8313965e-01 -1.4009149e+00
-5.3293753e-01 -6.6219079e-01 -3.6722967e-01 -4.1067129e-01
1.1600059e+00 -1.3789241e+00 -1.0368429e+00 5.6212795e-01
-6.5890455e-01 -8.7222999e-01 1.4134492e-01 1.1103613e+00
1.7202778e-02 -2.9576041e-02 -1.4161952e+00 -8.8987046e-01
2.4551645e-01 -1.3933043e+00 5.5441189e-01 8.2468343e-01
-8.3144262e-02 -1.6135966e+00 -1.3897458e-01 9.6246547e-01
2.6277211e-01 -2.1823424e-01 9.7593391e-01 -1.2274795e+00
-3.9490402e-01 -2.1483397e-01 1.6860031e-01 3.7238222e-01
-1.5447998e-01 3.8215363e-01 -4.4629046e-01 -3.0392438e-01
-3.3523440e-01 -6.1273491e-01 -2.0301449e-01 2.5935599e-01
5.3693300e-01 -3.3471441e-01 -2.3247027e-03 2.0014839e-01
1.4226451e+00 7.6229596e-01 3.0992731e-01 7.4358314e-01
5.6270623e-01 1.8188393e-01 6.0879993e-01 1.9765727e-02
8.8244236e-01 2.1423593e-01 -2.8499806e-01 6.3246742e-02
7.5971887e-02 -7.1801138e-01 6.1812127e-01 1.9571661e+00
-2.3983243e-01 -3.2225674e-01 -1.4252317e+00 6.1618799e-01
-1.3568231e+00 -5.3047013e-01 -5.9108365e-01 1.7947739e+00
7.9551947e-01 1.0787812e-01 1.4428821e-01 -1.3003691e-01
3.4223294e-01 -1.4032476e-01 3.7405592e-01 -1.3375338e+00
1.1318814e-01 5.0654727e-01 2.9962134e-01 6.8646860e-01
-4.4190750e-01 1.3078943e+00 6.5500488e+00 9.3617839e-01
-9.3728864e-01 3.9129159e-01 6.0046637e-01 5.3410941e-01
-2.5637200e-01 1.0284640e-02 -1.2220076e+00 2.9800704e-01
1.6197845e+00 -6.3602459e-01 1.8204446e-01 8.8935423e-01
1.5259241e-01 -3.6527869e-01 -5.9730488e-01 6.1660630e-01
-2.3576908e-01 -1.3156835e+00 -1.8076105e-01 3.7527049e-01
1.0554459e+00 1.1313355e-01 -3.2124799e-02 9.2638999e-01
9.0195882e-01 -1.0297985e+00 5.4016656e-01 -2.0286286e-01
3.9525712e-01 -8.4371603e-01 7.9390788e-01 6.8955863e-01
-1.0472867e+00 -4.5934208e-02 -4.6204615e-01 -5.3288251e-01
-2.0076050e-01 -2.0472954e-01 -9.6655089e-01 6.2177873e-01
7.8954762e-01 4.9314460e-01 -6.9857132e-01 1.0816811e+00
-3.7389195e-01 1.3356920e+00 -1.8022551e-01 -6.4291698e-01
9.0645248e-01 -6.3393110e-01 1.8434739e-01 1.1730934e+00
6.4237744e-01 1.7556851e-01 6.8127799e-01 3.4606108e-01
1.8942143e-01 6.4004326e-01 -6.1789149e-01 -6.5460646e-01
9.0352160e-01 8.5685682e-01 -7.4912882e-01 -2.0925884e-01
-1.0859694e+00 4.9186113e-01 6.2792264e-02 2.6952428e-01
-7.0338273e-01 -6.0582221e-01 6.4500809e-01 6.0089909e-02
-1.8187939e-01 -6.0532296e-01 -9.1063939e-02 -8.7062269e-01
-6.3222122e-01 -1.0907998e+00 5.0564885e-01 -9.1680592e-01
-9.2819232e-01 2.2914255e-01 -6.5943874e-02 -7.1239227e-01
-1.3029614e-01 -6.9930053e-01 -7.3289645e-01 1.1315958e+00
-9.7411400e-01 -9.3386394e-01 6.8806738e-02 7.0800513e-01
6.8198949e-01 -8.4519565e-02 8.0340052e-01 4.5935273e-01
-8.7133229e-01 7.3926014e-01 5.3145260e-01 5.0988734e-01
2.4526477e-01 -1.2121911e+00 2.9774213e-01 5.6293255e-01
1.7875334e-02 6.8514919e-01 4.4068956e-01 -8.2257408e-01
-1.1951770e+00 -4.8954132e-01 1.4159236e+00 -2.7363145e-01
9.3017763e-01 -6.0469669e-01 -8.0460763e-01 1.2287253e+00
7.2238696e-01 -8.3665514e-01 9.8285770e-01 2.8587645e-01
1.7954558e-01 3.2757074e-01 -9.3081909e-01 8.1692994e-01
4.7992009e-01 -7.3731327e-01 -6.2120003e-01 6.5906894e-01
8.8740307e-01 -7.8445315e-01 -1.4247814e+00 -5.4111034e-01
2.9887193e-01 -8.4763753e-01 7.5074100e-01 -6.9393486e-01
4.5957544e-01 7.3400915e-01 8.4736712e-02 -1.3532140e+00
1.7014082e-01 -8.0921477e-01 5.4059964e-01 1.6025289e+00
3.9963171e-01 -9.9193996e-01 6.7441362e-01 8.5549533e-01
-5.6417006e-01 -5.3883195e-01 -1.1561503e+00 -8.0578822e-01
9.0719062e-01 -4.2856565e-01 4.5319918e-01 1.2504889e+00
3.3356261e-01 1.5628662e-02 -2.2652198e-01 -2.0678969e-01
-4.6912447e-02 -4.4290367e-01 7.8094459e-01 -1.1389774e+00
3.1785077e-01 -2.6201129e-01 -3.3470720e-01 -7.1110582e-01
2.1107818e-01 -8.6013305e-01 -2.2616281e-01 -1.8740661e+00
-2.1062930e-01 -4.0767112e-01 3.8274407e-01 5.3101611e-01
-1.5403296e-01 5.3003240e-02 4.1043708e-01 -3.4346443e-01
-1.8622079e-01 1.8999954e-01 1.8236157e+00 2.1742202e-01
-3.6786082e-01 -6.2797301e-02 -6.5398282e-01 8.6849093e-01
8.4231210e-01 -2.0726828e-01 -3.5986993e-01 -3.9934742e-01
-6.4638242e-02 3.7048432e-01 -1.4010192e-01 -8.4302163e-01
2.8709143e-01 -5.8283812e-01 4.2108014e-02 -2.6932648e-01
1.8782556e-01 -5.2198505e-01 -2.2209404e-01 4.5848057e-01
-1.7340371e-01 6.3936442e-01 4.4319236e-01 -3.3384550e-01
-4.0555483e-01 -3.2596955e-01 5.9051132e-01 -1.4300221e-01
-7.0460939e-01 1.0437529e-01 -1.2430321e+00 2.2649010e-01
1.0705363e+00 -4.9098146e-01 -5.5904406e-01 -8.3542103e-01
-6.2815970e-01 4.9508896e-01 -2.3173596e-01 6.3342667e-01
5.5643082e-01 -1.1688172e+00 -5.4086071e-01 3.0448225e-01
-3.9955246e-01 -1.3645191e-01 4.6678612e-01 9.1044122e-01
-9.8094779e-01 8.4931308e-01 -2.1884380e-01 -2.5637382e-01
-1.4376872e+00 2.5348367e-02 -6.2088943e-03 -2.9495734e-01
-5.9292066e-01 1.1042880e+00 -1.8707070e-01 -1.1463666e+00
3.7933508e-01 -3.8169578e-01 -1.6175757e-01 -8.8289946e-02
4.1554129e-01 6.0972691e-01 -1.4455742e-01 -7.3604530e-01
2.2115867e-01 -1.1153016e-01 -1.2160510e-01 -1.7198727e-01
1.1024846e+00 -3.6805433e-01 -1.2069688e-01 8.5170817e-01
9.8510969e-01 5.4470891e-01 -1.8427126e-01 2.1530418e-01
2.3461862e-01 -6.2953109e-01 -1.8140115e-01 -7.1761352e-01
-7.2545671e-01 9.0396774e-01 2.5160193e-01 2.5720039e-01
5.2010137e-01 -2.1519471e-02 8.4000140e-01 3.5290849e-01
4.1643089e-01 -1.2337424e+00 -7.3754944e-02 1.1054647e+00
3.4732604e-01 -9.7276258e-01 -5.7136291e-01 -2.8740296e-01
-6.2165481e-01 9.4066936e-01 1.2695667e+00 3.2786044e-01
8.1672263e-01 1.4308804e-01 3.6707193e-01 1.0487893e-01
-7.3753673e-01 -1.1302155e-02 -1.2923335e-02 3.7786758e-01
1.1115804e+00 2.2947656e-01 -7.4115860e-01 4.7486624e-01
-9.2762285e-01 -5.8385283e-02 1.0363183e+00 1.1211611e+00
-5.4305291e-01 -1.0530324e+00 -5.4540849e-01 3.8176215e-01
-6.8934500e-01 -3.9920846e-01 -1.7413610e-01 1.4695210e+00
-6.0556144e-02 1.0671929e+00 -7.4988358e-02 -2.7072376e-01
1.4131296e-01 5.4668897e-01 3.9942968e-03 -8.1043351e-01
-1.0951997e+00 2.5955346e-01 2.0439091e-01 -1.3161197e-02
1.5604722e-01 -2.6616544e-01 -1.0580133e+00 -1.1387376e+00
-6.4496857e-01 1.1262450e+00 8.2055825e-01 8.7570602e-01
-2.0355037e-01 2.6836896e-01 9.6342480e-03 -1.6001616e-01
-1.7251046e-01 -1.0484189e+00 -8.4517086e-01 -3.9498803e-01
-3.0205041e-02 -3.1734154e-01 -1.9927843e-01 -6.9013268e-01] | [10.96613597869873, 9.130141258239746] |
2b236dd1-9438-459e-ad6f-44eb855ae001 | the-way-to-my-heart-is-through-contrastive-1 | 2111.09748 | null | https://arxiv.org/abs/2111.09748v1 | https://arxiv.org/pdf/2111.09748v1.pdf | The Way to my Heart is through Contrastive Learning: Remote Photoplethysmography from Unlabelled Video | The ability to reliably estimate physiological signals from video is a powerful tool in low-cost, pre-clinical health monitoring. In this work we propose a new approach to remote photoplethysmography (rPPG) - the measurement of blood volume changes from observations of a person's face or skin. Similar to current state-of-the-art methods for rPPG, we apply neural networks to learn deep representations with invariance to nuisance image variation. In contrast to such methods, we employ a fully self-supervised training approach, which has no reliance on expensive ground truth physiological training data. Our proposed method uses contrastive learning with a weak prior over the frequency and temporal smoothness of the target signal of interest. We evaluate our approach on four rPPG datasets, showing that comparable or better results can be achieved compared to recent supervised deep learning methods but without using any annotation. In addition, we incorporate a learned saliency resampling module into both our unsupervised approach and supervised baseline. We show that by allowing the model to learn where to sample the input image, we can reduce the need for hand-engineered features while providing some interpretability into the model's behavior and possible failure modes. We release code for our complete training and evaluation pipeline to encourage reproducible progress in this exciting new direction. | ['Simon Stent', 'John Gideon'] | 2021-11-18 | the-way-to-my-heart-is-through-contrastive | http://openaccess.thecvf.com//content/ICCV2021/html/Gideon_The_Way_to_My_Heart_Is_Through_Contrastive_Learning_Remote_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Gideon_The_Way_to_My_Heart_Is_Through_Contrastive_Learning_Remote_ICCV_2021_paper.pdf | iccv-2021-1 | ['image-variation'] | ['computer-vision'] | [ 4.68435138e-01 4.01272655e-01 3.89396623e-02 -5.88929176e-01
-7.29398906e-01 -1.19060606e-01 3.56573731e-01 -6.37504235e-02
-3.99257153e-01 7.27266371e-01 4.37006801e-01 2.18982950e-01
3.04280072e-01 -2.99086392e-01 -6.48886859e-01 -7.47717083e-01
-1.51364848e-01 -3.11316717e-02 1.45453870e-01 1.05797522e-01
2.08026037e-01 3.02503973e-01 -1.22754049e+00 2.99288660e-01
5.81819594e-01 1.09308696e+00 -1.62262604e-01 7.83967555e-01
4.35531616e-01 8.02736521e-01 -4.88846749e-01 6.90843239e-02
3.24148238e-01 -7.44799316e-01 -6.71513855e-01 2.09048353e-02
6.49109483e-01 -5.06253481e-01 -4.29714918e-01 7.11116433e-01
1.06543851e+00 -1.99095570e-02 4.11944836e-01 -8.74505222e-01
-3.11773837e-01 1.53184205e-01 -5.70809007e-01 5.95152318e-01
3.20014447e-01 5.96521258e-01 5.13536274e-01 -6.43767357e-01
5.62765241e-01 7.95898557e-01 9.13265646e-01 8.42597783e-01
-1.65738547e+00 -3.57116222e-01 -2.15575248e-01 9.44105685e-02
-1.04299295e+00 -9.64620948e-01 7.99811780e-01 -4.35739487e-01
9.20680404e-01 2.24185094e-01 7.08954275e-01 1.08822894e+00
2.12350488e-01 6.40976191e-01 1.39758492e+00 -3.95957798e-01
2.40529016e-01 2.89925098e-01 -1.34776860e-01 6.04267061e-01
-1.39152154e-01 2.07641721e-01 -6.91829979e-01 -1.27963081e-01
9.19143200e-01 -9.07790139e-02 -7.37713337e-01 -5.09869039e-01
-1.10053837e+00 4.47268575e-01 3.42706561e-01 2.63501227e-01
-5.36025763e-01 4.01619852e-01 6.16377413e-01 -9.48988367e-03
5.74394166e-01 5.10065138e-01 -4.40401644e-01 -2.45151252e-01
-1.31224692e+00 -1.42294645e-01 8.55645001e-01 4.24346298e-01
3.51497322e-01 1.24340482e-01 -4.04705346e-01 6.28065348e-01
3.88471372e-02 1.23737454e-01 7.41880357e-01 -1.22530532e+00
-1.83684841e-01 2.53569663e-01 1.50940701e-01 -8.48416746e-01
-6.41134143e-01 -3.66201401e-01 -4.45685387e-01 3.58735532e-01
4.08644736e-01 -3.58987540e-01 -8.80156875e-01 1.58826733e+00
2.92237192e-01 8.43323350e-01 -2.30197981e-01 1.13769412e+00
8.67906511e-01 2.24497870e-01 2.75502782e-02 -3.82030934e-01
1.18405688e+00 -6.87030375e-01 -5.49245954e-01 -1.55472681e-01
3.29000831e-01 -1.92963749e-01 1.09797633e+00 5.01378715e-01
-1.20945930e+00 -4.33933258e-01 -9.79156137e-01 -1.04645953e-01
-3.36899273e-02 5.21759093e-02 5.84288061e-01 7.96323180e-01
-1.29595280e+00 1.09582448e+00 -1.08976793e+00 -4.04952466e-01
6.48894727e-01 4.36676949e-01 -2.67849743e-01 1.04334310e-01
-1.07635534e+00 1.08463240e+00 -8.30546394e-02 2.18404025e-01
-8.66613865e-01 -1.25954080e+00 -8.10162723e-01 7.69673614e-03
2.23193184e-01 -6.72936022e-01 1.08330107e+00 -1.08657742e+00
-1.89855862e+00 1.14517212e+00 -2.18301728e-01 -7.06803620e-01
6.42475963e-01 -1.64869025e-01 -1.95593327e-01 7.99948335e-01
-2.97172785e-01 7.95863390e-01 1.04909289e+00 -7.85594642e-01
1.76105678e-01 -3.73130627e-02 -2.86266088e-01 2.94502843e-02
-1.32372543e-01 7.59048164e-02 -1.53397098e-01 -3.15653592e-01
-2.56961554e-01 -7.58925915e-01 -1.23529904e-01 5.09687245e-01
-7.92645887e-02 2.50465304e-01 5.29141009e-01 -6.59076869e-01
7.60889530e-01 -2.02767253e+00 -1.98445216e-01 1.63685549e-02
4.82724845e-01 3.46528679e-01 -5.19476794e-02 4.00765724e-02
-2.07402349e-01 -1.29291892e-01 -3.10486764e-01 -3.50456059e-01
-2.26158395e-01 -1.19909130e-01 -6.39555678e-02 9.13910568e-01
3.37072998e-01 1.15769398e+00 -9.26375747e-01 -3.95322829e-01
5.96688628e-01 8.04530621e-01 -4.38138902e-01 2.15300396e-01
1.11472271e-01 8.72996390e-01 8.93333852e-02 4.87243026e-01
4.25879240e-01 -3.55406970e-01 2.64210194e-01 -3.57176214e-01
9.75864753e-02 5.53699255e-01 -7.44638205e-01 1.82583761e+00
-4.03201103e-01 7.75501490e-01 1.28160045e-01 -1.29393899e+00
8.03678095e-01 6.14464104e-01 7.71820247e-01 -6.87557876e-01
1.63362250e-01 -8.61546304e-03 1.06375981e-02 -7.68312573e-01
-2.08046749e-01 -5.93159199e-01 4.26144719e-01 4.24760520e-01
3.11611325e-01 -5.71216531e-02 -2.63138980e-01 -5.73686499e-04
1.13859892e+00 4.74362075e-01 3.74155164e-01 -5.39937735e-01
3.66152912e-01 -2.85341918e-01 5.33671200e-01 6.43167377e-01
-6.78994358e-01 1.05298042e+00 5.46959341e-01 -5.05829811e-01
-7.94402659e-01 -9.18474495e-01 -3.81075680e-01 7.24948704e-01
-1.06549501e-01 -1.22602984e-01 -7.51888812e-01 -5.88379383e-01
-2.92871855e-02 4.30767030e-01 -9.00114775e-01 -2.48407617e-01
-5.10375977e-01 -9.05462027e-01 5.00707507e-01 6.22657478e-01
2.73360491e-01 -1.26494634e+00 -1.07817662e+00 2.39769518e-01
-8.71830583e-02 -1.15996039e+00 -3.83532763e-01 -9.72364563e-03
-1.01634300e+00 -1.02011883e+00 -8.35245192e-01 -3.66003305e-01
5.38390160e-01 -1.40018597e-01 1.24275148e+00 -2.10797489e-01
-9.08786654e-01 8.59223008e-01 4.22461554e-02 -5.66510499e-01
-6.98337629e-02 -2.91633219e-01 7.42396414e-02 9.35487226e-02
5.80045760e-01 -8.12543154e-01 -1.18234622e+00 4.01607808e-03
-5.16589224e-01 -5.54960780e-02 3.66590649e-01 6.77097678e-01
5.24789691e-01 -6.65685117e-01 7.54210174e-01 -9.71652567e-01
5.16533673e-01 -2.86086440e-01 -4.22214717e-01 -1.27894551e-01
-5.77827275e-01 8.26514792e-03 5.02425015e-01 -4.52403724e-01
-7.95777857e-01 2.39561990e-01 1.19960427e-01 -7.03798950e-01
-3.41805398e-01 2.91201979e-01 4.07091677e-01 -3.28105658e-01
1.01208782e+00 3.47121656e-01 4.33795691e-01 -1.64195225e-01
1.84696227e-01 3.81220222e-01 8.80324602e-01 -3.36471081e-01
2.92163879e-01 7.51834333e-01 1.26384288e-01 -8.57639432e-01
-6.49173200e-01 -4.96055186e-01 -8.50084543e-01 -3.48726004e-01
7.33284056e-01 -9.56772208e-01 -8.37240160e-01 3.43250930e-01
-7.43867695e-01 -5.99981964e-01 -5.24723053e-01 6.34104848e-01
-7.54667878e-01 5.22013366e-01 -6.05955005e-01 -8.05952668e-01
-5.13997853e-01 -7.52680004e-01 1.05918849e+00 2.25566238e-01
-4.69921887e-01 -1.31884158e+00 2.76873916e-01 2.92896241e-01
7.95909762e-01 7.28068113e-01 2.64429212e-01 -4.91217762e-01
-3.42288196e-01 -2.03683794e-01 -1.82559669e-01 6.16251528e-01
2.46355370e-01 -3.89377862e-01 -1.52469170e+00 -3.37515146e-01
4.22945499e-01 -6.30804121e-01 8.74589205e-01 9.15041089e-01
1.28165615e+00 -1.03204727e-01 -2.24733818e-02 7.81357706e-01
1.27697825e+00 -3.02344143e-01 9.22907948e-01 -6.40113503e-02
5.17827809e-01 6.66378915e-01 1.53818265e-01 4.28447992e-01
1.28615424e-01 3.93869996e-01 6.89406246e-02 -5.05783081e-01
-2.43869990e-01 2.10295599e-02 4.01665837e-01 2.76222646e-01
-1.69204950e-01 3.41589361e-01 -6.15549386e-01 5.79738021e-01
-1.50372338e+00 -8.63423169e-01 1.16443984e-01 2.45615673e+00
1.06716251e+00 -7.16480389e-02 2.24517182e-01 -3.56321037e-02
5.62296450e-01 1.38711870e-01 -8.21003854e-01 -3.59988332e-01
3.33792359e-01 5.82427680e-01 5.23808599e-01 4.76991385e-01
-1.06116724e+00 5.49787760e-01 7.20292091e+00 -3.00496183e-02
-1.72479033e+00 1.07427411e-01 7.42397070e-01 -5.86581767e-01
8.56548175e-02 -3.01633716e-01 -4.19522405e-01 5.56753337e-01
1.32089818e+00 -5.46009578e-02 4.09266561e-01 4.90523368e-01
7.31468141e-01 -2.93271840e-01 -1.45700586e+00 1.09978700e+00
4.93312925e-01 -1.36836767e+00 -7.81308591e-01 -1.47613570e-01
2.87495553e-01 1.32517859e-01 -1.05916090e-01 1.03606787e-02
-1.62439868e-01 -1.18041062e+00 6.94815814e-02 7.97788203e-01
1.05995309e+00 -1.41717136e-01 4.63815719e-01 7.09642889e-03
-6.28627658e-01 2.95446038e-01 -2.84655690e-01 -3.33089046e-02
9.29931831e-03 8.06612194e-01 -8.99497867e-01 1.07598498e-01
5.65199554e-01 9.37225997e-01 -4.23202813e-01 1.15427852e+00
-2.18656644e-01 8.49962592e-01 -3.09221625e-01 2.49102026e-01
-2.36671805e-01 1.63190827e-01 5.15888214e-01 1.24896121e+00
-8.36259797e-02 1.81863502e-01 -8.40140681e-04 1.06060648e+00
-3.38154621e-02 8.64248946e-02 -4.57773745e-01 1.31091520e-01
-1.67930294e-02 1.28694773e+00 -4.68972772e-01 -5.86213529e-01
-4.91382509e-01 1.04219651e+00 9.87017825e-02 2.59180725e-01
-6.85900748e-01 -4.19029862e-01 2.34458432e-01 4.72927362e-01
1.49945304e-01 1.29624993e-01 -3.10406774e-01 -1.42686021e+00
1.22696564e-01 -6.49495602e-01 3.35640192e-01 -9.71514463e-01
-1.24656105e+00 3.49459529e-01 -8.59495923e-02 -9.95807350e-01
-3.35455358e-01 -4.78000075e-01 -7.64167607e-01 1.13457036e+00
-1.93107569e+00 -7.11241484e-01 -4.41853493e-01 5.31968832e-01
2.45699286e-01 3.05567682e-01 8.50052595e-01 8.99808928e-02
-3.93784314e-01 3.79676282e-01 -3.95160228e-01 4.57474701e-02
1.12116230e+00 -1.34556878e+00 1.62022084e-01 7.94667244e-01
-1.26140118e-01 5.48960805e-01 7.60466039e-01 -3.80975574e-01
-1.23723888e+00 -8.64671290e-01 5.51478684e-01 -5.58817327e-01
5.05420268e-01 -4.62095767e-01 -1.03351510e+00 6.40401840e-01
3.12708467e-01 7.43287444e-01 7.92743802e-01 -6.86697364e-02
-1.45344466e-01 -2.93427140e-01 -1.45878780e+00 1.44515052e-01
6.64507449e-01 -5.20629346e-01 -7.50874758e-01 3.12872440e-01
2.35409960e-01 -5.35381734e-01 -9.29224133e-01 2.79340297e-01
7.07927108e-01 -1.03897822e+00 7.63969064e-01 -4.20134991e-01
4.60006088e-01 -9.30904672e-02 4.79237258e-01 -1.32563305e+00
-1.69988811e-01 -8.67563426e-01 -2.97943026e-01 8.37190568e-01
1.37132987e-01 -7.84673691e-01 1.05245304e+00 1.09570801e+00
-7.27158934e-02 -6.54857516e-01 -7.78183877e-01 -4.30459887e-01
-1.67219918e-02 -2.02821687e-01 -6.03040755e-02 1.00610328e+00
5.22304535e-01 1.05006300e-01 -5.85841298e-01 1.05540968e-01
7.67101049e-01 6.66587651e-02 4.43118960e-01 -1.10817885e+00
-5.48166394e-01 -3.70818600e-02 -6.62186861e-01 -6.95054352e-01
5.66988774e-02 -7.60385990e-01 1.17125012e-01 -1.27123034e+00
2.38448694e-01 4.32239026e-02 -6.76141620e-01 6.45772517e-01
-1.42266512e-01 6.64459467e-01 -1.37992963e-01 6.49753287e-02
-4.27557439e-01 2.94737756e-01 1.09938228e+00 1.95267424e-01
-4.43467438e-01 -2.93863475e-01 -7.45252550e-01 4.75247830e-01
8.75665665e-01 -4.47280526e-01 -4.77460086e-01 -2.91100647e-02
-3.64811271e-01 1.88421354e-01 6.98458433e-01 -1.08060968e+00
2.04342261e-01 1.69515207e-01 7.10997403e-01 6.69429228e-02
2.69228727e-01 -4.14248794e-01 -2.72682577e-01 4.90969002e-01
-5.37300527e-01 -5.04282296e-01 4.04051751e-01 3.20029676e-01
-2.90657263e-02 1.49821891e-02 1.23389888e+00 -5.22957504e-01
-4.13733751e-01 1.45183995e-01 -3.20022702e-01 2.22936437e-01
7.27341413e-01 -2.10988253e-01 -1.77811578e-01 -5.54708540e-01
-1.03649139e+00 1.02445461e-01 3.63540351e-01 7.53154904e-02
6.03612661e-01 -8.70601416e-01 -6.95008039e-01 3.11416894e-01
-7.72559643e-02 -3.28922242e-01 2.90551156e-01 1.38149476e+00
-4.18846905e-01 2.92734981e-01 -4.43335950e-01 -7.73855746e-01
-1.01824319e+00 4.41529244e-01 9.21244979e-01 9.01645049e-02
-9.65554535e-01 4.94072378e-01 1.34315550e-01 -1.74410790e-01
6.14318848e-02 -3.32014292e-01 -1.32712439e-01 -2.33769163e-01
8.20410013e-01 1.94094643e-01 1.88487664e-01 -2.47933686e-01
-3.37485671e-01 3.14573109e-01 -7.28735328e-02 8.70911870e-03
1.45596802e+00 -1.09151624e-01 4.61122729e-02 5.05431056e-01
1.16286802e+00 -1.58679485e-01 -1.69804525e+00 -7.99179003e-02
-2.08570033e-01 -5.23801148e-01 4.27397043e-01 -1.05133784e+00
-1.15434086e+00 1.17793024e+00 9.36898232e-01 -2.02610865e-01
1.18437624e+00 -2.57026017e-01 5.54951072e-01 1.21106701e-02
5.63700199e-02 -1.09929097e+00 -6.21849322e-04 -2.60924548e-01
7.35920906e-01 -1.33009672e+00 2.99604297e-01 -3.53949815e-01
-7.43907988e-01 1.06362748e+00 4.26319182e-01 -5.18983364e-01
6.16513431e-01 3.96235168e-01 3.70122552e-01 -2.80100882e-01
-8.65540147e-01 8.29733759e-02 3.57223272e-01 7.74882376e-01
6.82453156e-01 -3.49703610e-01 -2.23205969e-01 1.94168299e-01
1.94542408e-01 7.46679962e-01 8.17730963e-01 8.22361827e-01
-2.79362530e-01 -7.59854972e-01 7.43873045e-02 5.78591585e-01
-8.64592075e-01 -2.56699562e-01 -7.12492093e-02 5.07278860e-01
-2.50981748e-01 5.58735728e-01 4.37116772e-02 1.10725679e-01
2.16941804e-01 4.30409104e-01 6.89910293e-01 -8.17839086e-01
-4.76168454e-01 2.62585402e-01 -6.49410039e-02 -9.02185023e-01
-7.03488171e-01 -8.07249010e-01 -1.05011392e+00 2.34968185e-01
3.22077088e-02 -2.73349136e-01 5.24907231e-01 8.82498324e-01
6.17713749e-01 4.01550263e-01 5.03474772e-01 -1.14484668e+00
-2.43541881e-01 -9.10514534e-01 -5.87489009e-01 4.96042848e-01
5.84037721e-01 -4.76172119e-01 -4.75712895e-01 3.15361083e-01] | [13.896225929260254, 2.7822697162628174] |
49a700f5-d2d4-4d71-9533-f77dd767a726 | restoration-of-the-jpeg-maximum-lossy | 2306.12757 | null | https://arxiv.org/abs/2306.12757v1 | https://arxiv.org/pdf/2306.12757v1.pdf | Restoration of the JPEG Maximum Lossy Compressed Face Images with Hourglass Block based on Early Stopping Discriminator | When a JPEG image is compressed using the loss compression method with a high compression rate, a blocking phenomenon can occur in the image, making it necessary to restore the image to its original quality. In particular, restoring compressed images that are unrecognizable presents an innovative challenge. Therefore, this paper aims to address the restoration of JPEG images that have suffered significant loss due to maximum compression using a GAN-based net-work method. The generator in this network is based on the U-Net architecture and features a newly presented hourglass structure that can preserve the charac-teristics of deep layers. Additionally, the network incorporates two loss functions, LF Loss and HF Loss, to generate natural and high-performance images. HF Loss uses a pretrained VGG-16 network and is configured using a specific layer that best represents features, which can enhance performance for the high-frequency region. LF Loss, on the other hand, is used to handle the low-frequency region. These two loss functions facilitate the generation of images by the generator that can deceive the discriminator while accurately generating both high and low-frequency regions. The results show that the blocking phe-nomenon in lost compressed images was removed, and recognizable identities were generated. This study represents a significant improvement over previous research in terms of image restoration performance. | ['Sungyoung Kim', 'Jongwook Si'] | 2023-06-22 | null | null | null | null | ['image-restoration', 'blocking'] | ['computer-vision', 'natural-language-processing'] | [ 5.72280705e-01 7.67771080e-02 1.82755709e-01 -1.00068688e-01
-4.18769479e-01 -1.51554972e-01 3.56505901e-01 -2.40607619e-01
-1.52385697e-01 7.42130876e-01 2.65321523e-01 -1.72426328e-02
2.07861468e-01 -1.27584577e+00 -9.41992164e-01 -7.65048265e-01
-2.46205673e-01 -2.22818732e-01 5.56400195e-02 -5.33884466e-01
2.71782368e-01 5.08383393e-01 -1.68337107e+00 7.52174437e-01
9.32399213e-01 1.24100888e+00 3.58297110e-01 5.90919495e-01
2.50031203e-01 9.45775092e-01 -9.52962756e-01 -5.95316112e-01
5.66996098e-01 -6.94621325e-01 -5.11623085e-01 -1.08611040e-01
6.46595001e-01 -8.83265674e-01 -6.18147016e-01 1.17168605e+00
5.93896151e-01 -9.85899009e-03 4.66161579e-01 -1.12961245e+00
-6.87690020e-01 4.50302690e-01 -3.81578058e-01 -3.92535999e-02
4.20670092e-01 2.63426304e-01 7.33677566e-01 -3.48601967e-01
5.58670878e-01 1.24946129e+00 8.00546527e-01 6.24949872e-01
-1.13055742e+00 -8.43364477e-01 -5.56639314e-01 3.31842870e-01
-1.24718702e+00 -2.42641464e-01 9.07237232e-01 -1.36704504e-01
9.64191854e-01 3.77617240e-01 7.33556330e-01 1.14980066e+00
6.86953485e-01 3.99436831e-01 8.36983323e-01 -3.53610367e-01
4.01227176e-02 -2.24406660e-01 -7.14042544e-01 2.96519458e-01
1.96610779e-01 4.60157841e-01 -5.36586285e-01 2.47344404e-01
8.51766407e-01 -3.21511626e-02 -7.06870198e-01 9.27101746e-02
-6.50584638e-01 7.84096658e-01 9.46537256e-01 2.94344246e-01
-5.17117798e-01 2.67967433e-01 2.83316404e-01 4.63229209e-01
1.24319963e-01 5.02708018e-01 3.77480388e-01 6.33675382e-02
-1.45052433e+00 3.40892188e-02 4.87187088e-01 7.16824293e-01
6.45054519e-01 5.56623757e-01 -2.37649724e-01 8.37744057e-01
1.30580753e-01 2.54160166e-01 6.50102913e-01 -9.15471613e-01
4.07889336e-01 4.39214885e-01 -3.04079682e-01 -1.39079976e+00
3.39646637e-02 -5.53104818e-01 -1.48302364e+00 7.05905020e-01
-7.23068714e-02 2.00505108e-01 -9.64756966e-01 1.67511964e+00
-2.35878795e-01 6.12277053e-02 2.93845922e-01 9.59554911e-01
7.45902598e-01 8.80439818e-01 -1.58146322e-01 1.01179361e-01
1.16834891e+00 -8.49788308e-01 -7.17278779e-01 -2.38478929e-01
-2.65888702e-02 -9.19554591e-01 1.02618992e+00 4.88154143e-01
-1.22592092e+00 -9.67801929e-01 -1.60362864e+00 -2.40243286e-01
-1.02123655e-01 -1.32168934e-01 2.53201336e-01 5.39835215e-01
-1.27267396e+00 8.97465467e-01 -3.02152216e-01 -2.32845154e-02
5.91123104e-01 2.33597964e-01 -5.07108867e-01 -2.34534636e-01
-1.47469711e+00 6.33761704e-01 6.92998707e-01 1.83798417e-01
-9.53223705e-01 -5.36500931e-01 -8.00082982e-01 3.76049280e-01
-2.89146632e-01 -5.46995819e-01 5.67413926e-01 -1.39131379e+00
-1.26018214e+00 6.93818748e-01 5.64068079e-01 -7.55452693e-01
8.90265703e-01 1.31633088e-01 -5.97276628e-01 6.32755101e-01
-1.98068190e-02 9.35672879e-01 1.10583305e+00 -1.40923727e+00
-3.44682604e-01 -2.48375069e-02 -8.12864304e-02 5.34021333e-02
-5.00788689e-01 -5.06907880e-01 -1.84277758e-01 -1.14410281e+00
5.15845371e-03 -5.32032907e-01 3.31314981e-01 5.75108081e-02
-4.02040690e-01 5.80402553e-01 1.03052914e+00 -1.19745874e+00
1.10219693e+00 -2.41398501e+00 -2.23556668e-01 2.80187249e-01
2.20106006e-01 4.86981601e-01 -2.69628763e-01 6.52797520e-01
-1.81691200e-01 2.97995478e-01 -5.13442338e-01 -1.77228287e-01
-4.13391232e-01 1.64579138e-01 -3.73234719e-01 3.11573058e-01
8.00634325e-02 6.76612854e-01 -6.56039000e-01 -2.71726131e-01
2.65122861e-01 8.28500867e-01 -7.38438725e-01 3.61840218e-01
2.17635045e-03 7.42776841e-02 1.81706175e-01 3.73787820e-01
1.23612416e+00 1.35967761e-01 7.39938989e-02 -6.66395187e-01
6.56828731e-02 2.92484090e-02 -8.08054388e-01 1.47631454e+00
-3.87445033e-01 6.46327674e-01 1.41901627e-01 -8.10617805e-01
1.16772747e+00 2.14583158e-01 1.04580149e-01 -9.94503319e-01
1.23779908e-01 3.23790640e-01 -7.76817724e-02 -4.42097247e-01
8.04577172e-01 -2.51204163e-01 2.25954831e-01 6.93281218e-02
-2.43299343e-02 -2.60080457e-01 4.98835705e-02 3.39478642e-01
1.03091955e+00 -5.00601791e-02 -7.26908147e-02 -5.07625937e-02
4.93592113e-01 -3.77983779e-01 4.26594853e-01 6.84130132e-01
1.55272648e-01 1.25091112e+00 4.78918642e-01 -2.55527943e-01
-1.38103712e+00 -1.07437718e+00 6.31714836e-02 2.85473496e-01
3.75009239e-01 -3.37519914e-01 -7.85330594e-01 -3.10804904e-01
-1.48557693e-01 7.63855100e-01 -3.73075247e-01 -7.80663788e-01
-4.17432040e-01 -4.85006005e-01 8.75098944e-01 2.37104997e-01
1.35860264e+00 -1.35953081e+00 -6.87057257e-01 2.08668053e-01
-5.41078746e-01 -9.40908015e-01 -4.07006741e-01 -7.70645142e-02
-8.28091383e-01 -1.05585575e+00 -7.68143654e-01 -9.41658199e-01
8.02260101e-01 1.76659480e-01 9.57884848e-01 6.33817315e-01
-4.43681568e-01 -1.56021401e-01 -5.57340145e-01 4.12393250e-02
-8.74699593e-01 -1.66694164e-01 -4.28308934e-01 9.43696126e-02
-8.30963850e-02 -7.76022851e-01 -8.45010519e-01 1.18677631e-01
-1.51166761e+00 2.80053228e-01 6.46068692e-01 1.08276236e+00
3.64403754e-01 6.53239965e-01 4.51952219e-01 -6.48658931e-01
5.50493538e-01 -2.34101728e-01 -2.16134220e-01 6.44775433e-03
-5.70807993e-01 -1.49581179e-01 1.09818804e+00 -2.09562764e-01
-9.96571839e-01 -3.04052144e-01 -4.43134069e-01 -3.65092784e-01
8.94906074e-02 1.65686741e-01 -3.53668541e-01 -3.81062686e-01
6.19781911e-01 4.93427038e-01 3.48226726e-01 -3.94725561e-01
8.80675577e-03 7.43690014e-01 8.80052507e-01 -1.34089485e-01
7.83381939e-01 4.10638362e-01 -1.87631752e-02 -7.43116021e-01
-1.43354625e-01 2.19116330e-01 -6.11407496e-02 -4.23648834e-01
6.04589522e-01 -8.80279660e-01 -4.70903635e-01 1.04583693e+00
-1.00468612e+00 -2.28322372e-01 -5.08481860e-01 2.80832291e-01
-4.69941497e-01 5.47182083e-01 -9.64042366e-01 -4.08591419e-01
-5.83642542e-01 -1.12803972e+00 7.41771698e-01 3.32293749e-01
8.15677047e-02 -6.13904417e-01 -4.78538096e-01 3.62524718e-01
7.89254248e-01 5.15166163e-01 1.08228254e+00 8.71919543e-02
-6.06498599e-01 -1.98701516e-01 -2.49944180e-01 1.12593412e+00
1.40763015e-01 -1.49872124e-01 -9.82830763e-01 -6.85499489e-01
3.88146490e-01 -2.78576374e-01 1.24349260e+00 2.11048305e-01
1.20036554e+00 -7.19368517e-01 2.46514753e-01 1.15287828e+00
1.58588207e+00 3.13566625e-01 1.71574867e+00 5.77805400e-01
5.61913252e-01 2.62355506e-01 1.12456426e-01 2.65271783e-01
4.28174995e-02 5.14211893e-01 9.70845520e-01 -5.08462787e-01
-8.00182939e-01 -5.73882818e-01 5.15049577e-01 4.98404980e-01
-6.78539053e-02 -5.78330994e-01 -3.71912509e-01 1.59304306e-01
-1.27592766e+00 -1.26900446e+00 3.92973088e-02 2.29950070e+00
1.00164843e+00 2.48763248e-01 -3.95004123e-01 4.32934314e-01
8.83484483e-01 1.85563356e-01 -3.25846463e-01 -3.92527014e-01
-6.26348913e-01 2.37033859e-01 5.34591973e-01 3.83644253e-01
-8.19231629e-01 6.73287213e-01 6.12270737e+00 1.06498480e+00
-1.47325933e+00 -2.59577572e-01 8.64559829e-01 1.97979510e-02
-3.36462498e-01 -6.14203624e-02 -2.60118008e-01 8.03947687e-01
5.58859289e-01 2.30198540e-03 5.52499533e-01 5.79387486e-01
1.14256524e-01 -1.66926309e-01 -6.48572087e-01 9.30654764e-01
4.44747597e-01 -1.30029428e+00 5.68640053e-01 1.61336362e-01
6.60971642e-01 -5.63291073e-01 3.03721547e-01 -7.96975940e-02
-1.22178517e-01 -1.20152056e+00 9.19453025e-01 5.29825151e-01
1.23136091e+00 -9.63672519e-01 9.80034471e-01 4.18295562e-02
-8.78812730e-01 -1.48121804e-01 -6.13684893e-01 1.04952231e-01
3.15802604e-01 8.90949667e-01 -7.19102204e-01 6.39582098e-01
8.02164435e-01 6.77627385e-01 -5.44918954e-01 1.05772007e+00
-3.97737414e-01 4.99613136e-01 -3.97948585e-02 7.30232239e-01
4.98235561e-02 -1.84677288e-01 5.69876432e-01 1.00858712e+00
8.27189565e-01 -3.11382353e-01 -1.66522592e-01 9.70087886e-01
-4.46289659e-01 -2.22043231e-01 -5.86458087e-01 1.97027713e-01
1.61751226e-01 1.00644588e+00 -6.57346368e-01 -2.46643469e-01
1.01343818e-01 1.36254025e+00 -2.41122708e-01 2.26817280e-01
-7.50538945e-01 -6.02348685e-01 3.74057084e-01 5.05873084e-01
3.37240696e-01 1.00928165e-01 -7.62115419e-02 -9.32268262e-01
1.19146354e-01 -1.18447280e+00 2.02127501e-01 -9.33778524e-01
-1.20205522e+00 9.05225813e-01 -3.51981521e-01 -1.49147296e+00
-2.08620027e-01 -2.08461300e-01 -4.42628980e-01 9.54178452e-01
-1.58966029e+00 -1.16294229e+00 -6.32550657e-01 6.68737531e-01
2.70664036e-01 -3.81834924e-01 7.25609720e-01 5.52629352e-01
-1.40127763e-01 7.71505356e-01 1.08324725e-03 1.07785828e-01
6.42055333e-01 -8.59436095e-01 2.48942941e-01 1.20420098e+00
-3.04858834e-01 3.06711406e-01 5.53683281e-01 -9.00814354e-01
-1.06386769e+00 -1.14674699e+00 7.13268340e-01 6.72684312e-01
-2.70154625e-01 -3.29760343e-01 -9.96024072e-01 3.01808417e-01
3.34188432e-01 -2.25798592e-01 4.25438583e-01 -8.70297492e-01
-4.41187054e-01 -4.27963585e-01 -1.83839703e+00 3.94044578e-01
6.82318032e-01 -5.58430016e-01 -1.67258874e-01 3.28577589e-03
4.85730022e-01 -3.62392604e-01 -6.66067898e-01 3.19840431e-01
5.27330637e-01 -1.45437253e+00 1.14934731e+00 1.62348568e-01
1.07237387e+00 -3.04503769e-01 -1.93292007e-01 -1.37368083e+00
-4.21502769e-01 -5.59525728e-01 3.05902977e-02 1.38955903e+00
-3.24089155e-02 -5.57329834e-01 5.32646477e-01 7.88124651e-02
-1.11505032e-01 -3.27616155e-01 -9.48801816e-01 -8.69687915e-01
-2.05580413e-01 5.01663685e-02 7.20428944e-01 7.32577801e-01
-4.20539677e-01 -3.11023816e-02 -8.25985134e-01 -7.45618790e-02
6.63141608e-01 -2.06303582e-01 4.56905723e-01 -7.98361599e-01
-3.00385982e-01 -3.78283709e-01 -7.32499003e-01 -9.18871641e-01
-1.95003703e-01 -9.68590736e-01 3.76216099e-02 -1.45381868e+00
-1.05892219e-01 -3.22813421e-01 -9.91012156e-02 4.33044910e-01
1.03126109e-01 8.19320858e-01 4.86549914e-01 3.47847193e-01
2.86028415e-01 7.11500108e-01 1.38619089e+00 -4.23714697e-01
1.34249881e-01 -2.49618366e-01 -8.16869736e-01 3.21167558e-01
9.36649323e-01 -4.37583327e-01 -4.40686852e-01 -5.48319459e-01
1.69760779e-01 -1.36603326e-01 5.97249091e-01 -1.50559795e+00
8.64007473e-02 3.44259888e-01 7.76794851e-01 -2.42660075e-01
3.62895697e-01 -9.82897699e-01 5.53769469e-01 7.59580791e-01
-2.94669360e-01 3.39397346e-03 2.10646361e-01 2.30586797e-01
-7.01960027e-01 -3.04445535e-01 1.35361648e+00 -3.33305717e-01
-6.22581840e-01 6.82043238e-03 -2.36197665e-01 -1.53998196e-01
9.38221693e-01 -5.37875056e-01 -3.70039403e-01 -7.87239432e-01
-1.41723484e-01 -1.90726891e-01 8.66574526e-01 3.46932352e-01
1.06986845e+00 -1.44498014e+00 -8.40766191e-01 6.72610700e-01
-3.49648833e-01 -1.89910084e-01 6.65048361e-01 1.61716208e-01
-1.21242476e+00 -3.40449721e-01 -8.71234000e-01 -2.60271668e-01
-1.03704429e+00 4.06352431e-01 3.84060681e-01 -1.88536525e-01
-9.01030898e-01 7.26500869e-01 7.33467564e-02 5.98497763e-02
1.04545668e-01 1.40947670e-01 -2.04782188e-01 1.22823911e-02
7.14853227e-01 4.04029310e-01 1.97365254e-01 -7.29053140e-01
-9.20129195e-02 3.41924787e-01 2.70439908e-02 3.30411121e-02
1.39432657e+00 -6.91531971e-02 -4.21080977e-01 -1.61510512e-01
1.39936733e+00 -1.21198505e-01 -1.23566127e+00 3.54325920e-01
-8.70725930e-01 -8.41357052e-01 3.03009957e-01 -9.75779355e-01
-1.72941530e+00 8.06394041e-01 7.80394018e-01 2.94665128e-01
1.77110815e+00 -7.21990108e-01 1.13812697e+00 -4.36450958e-01
1.61787137e-01 -1.09445488e+00 2.31969818e-01 2.57026523e-01
1.20843577e+00 -6.76609695e-01 -7.88956881e-03 -4.14664298e-01
-3.73492718e-01 1.45531142e+00 6.93797648e-01 -2.28226542e-01
2.48545647e-01 3.46003383e-01 -6.05908856e-02 9.55520198e-02
-2.00777441e-01 4.04288679e-01 -2.15480383e-02 7.47354031e-01
1.79683104e-01 -4.26381305e-02 -4.11874920e-01 2.76175410e-01
-7.24300563e-01 1.00151524e-01 7.76076317e-01 7.84019411e-01
-1.87771767e-01 -1.02715933e+00 -5.52540362e-01 5.63879251e-01
-5.28151572e-01 -3.81339431e-01 -2.18757298e-02 4.53326672e-01
5.22584438e-01 1.00029492e+00 1.90035328e-01 -6.91796005e-01
2.22409621e-01 -3.55057567e-01 3.22430521e-01 -3.21159996e-02
-9.09576654e-01 -1.10841177e-01 -2.67185986e-01 -6.00277126e-01
-1.81844413e-01 7.23887905e-02 -1.07771981e+00 -7.24717438e-01
-1.32054403e-01 -6.29394725e-02 6.37657166e-01 2.38138214e-01
3.46191525e-01 7.82465279e-01 7.64065862e-01 -8.49865198e-01
-4.20015067e-01 -7.47352898e-01 -8.62918556e-01 8.90754461e-01
4.36017305e-01 -2.25852370e-01 -6.69468999e-01 2.35018685e-01] | [11.352043151855469, -1.6113883256912231] |
7e7071bd-3c05-406b-b9f4-46b940f706cb | rovist-learning-robust-metrics-for-visual-2 | null | null | https://aclanthology.org/2022.findings-naacl.206 | https://aclanthology.org/2022.findings-naacl.206.pdf | RoViST: Learning Robust Metrics for Visual Storytelling | Visual storytelling (VST) is the task of generating a story paragraph that describes a given image sequence. Most existing storytelling approaches have evaluated their models using traditional natural language generation metrics like BLEU or CIDEr. However, such metrics based on n-gram matching tend to have poor correlation with human evaluation scores and do not explicitly consider other criteria necessary for storytelling such as sentence structure or topic coherence. Moreover, a single score is not enough to assess a story as it does not inform us about what specific errors were made by the model. In this paper, we propose 3 evaluation metrics sets that analyses which aspects we would look for in a good story: 1) visual grounding, 2) coherence, and 3) non-redundancy. We measure the reliability of our metric sets by analysing its correlation with human judgement scores on a sample of machine stories obtained from 4 state-of-the-arts models trained on the Visual Storytelling Dataset (VIST). Our metric sets outperforms other metrics on human correlation, and could be served as a learning based evaluation metric set that is complementary to existing rule-based metrics. | ['Josiah Poon', 'Caren Han', 'Eileen Wang'] | null | null | null | null | findings-naacl-2022-7 | ['visual-storytelling'] | ['natural-language-processing'] | [ 1.46267235e-01 3.65485936e-01 4.25431505e-02 -1.97880760e-01
-7.02314973e-01 -6.04516923e-01 1.39625001e+00 6.69809937e-01
-1.13458961e-01 8.22450459e-01 7.82551050e-01 -1.57841772e-01
-1.56685337e-01 -8.36102068e-01 -6.18433475e-01 -3.71446401e-01
2.09157273e-01 5.07796586e-01 4.80689555e-01 -2.87469000e-01
6.27880454e-01 -1.22887447e-01 -1.84854424e+00 7.90135145e-01
7.28672028e-01 6.91360354e-01 4.22287762e-01 6.39225721e-01
-3.81520927e-01 1.54038346e+00 -9.30822194e-01 -4.81421351e-01
-2.48844996e-01 -9.90042865e-01 -9.60231245e-01 4.12497446e-02
4.64179993e-01 -2.49198433e-02 3.97448950e-02 7.66266525e-01
2.64794976e-01 -1.04735129e-01 9.95448649e-01 -1.20159745e+00
-5.56086361e-01 9.46283519e-01 -1.39485925e-01 -3.97277363e-02
1.04968500e+00 2.74417788e-01 1.21503246e+00 -7.19828427e-01
1.26639962e+00 1.18417454e+00 4.72960293e-01 2.54302174e-01
-9.83459890e-01 -1.01088434e-01 -3.49991888e-01 5.02169490e-01
-9.57277775e-01 -2.01411754e-01 5.78576803e-01 -9.59132254e-01
8.55451584e-01 3.87191534e-01 7.87440836e-01 1.25873280e+00
1.34596705e-01 7.14922309e-01 1.40400982e+00 -5.54781616e-01
3.85010123e-01 2.50797212e-01 -2.19087571e-01 3.83399218e-01
1.52829379e-01 2.46794075e-02 -8.39619339e-01 1.44609183e-01
6.21067941e-01 -7.39577830e-01 -1.59019485e-01 -4.66354162e-01
-1.68969214e+00 9.14283514e-01 1.81943074e-01 7.42489874e-01
-4.05859917e-01 1.19239837e-01 6.30233645e-01 3.26805979e-01
2.08255187e-01 8.83768439e-01 1.96741760e-01 -4.09287214e-01
-1.28815985e+00 7.24790037e-01 6.21603727e-01 7.13194728e-01
3.53751540e-01 -3.89206521e-02 -8.54063094e-01 6.75324619e-01
1.46440148e-01 1.50140241e-01 6.60295844e-01 -8.20069432e-01
2.88469404e-01 7.55267203e-01 1.35601595e-01 -1.08927929e+00
-1.42440155e-01 -2.06635535e-01 -4.86165792e-01 5.32937169e-01
4.62194324e-01 1.71271056e-01 -5.43177605e-01 1.52726591e+00
-1.11554630e-01 -2.90947288e-01 4.67820615e-02 9.74174500e-01
1.43020606e+00 7.30211616e-01 -1.80112153e-01 -2.68211007e-01
1.22179198e+00 -9.23822343e-01 -7.56495893e-01 -1.94339290e-01
5.39408147e-01 -1.08763456e+00 1.33812761e+00 6.37731194e-01
-1.17468894e+00 -7.10051775e-01 -1.35221326e+00 9.93800014e-02
-4.32513565e-01 -2.78904792e-02 3.05296481e-01 4.34764415e-01
-1.13772559e+00 7.36507177e-01 -6.92778304e-02 -6.60994828e-01
-1.18022442e-01 -4.95881110e-01 -2.90421605e-01 1.45916119e-01
-1.00898147e+00 1.28955007e+00 6.62588716e-01 -6.13811910e-01
-1.00602925e+00 -2.25271255e-01 -6.20570779e-01 -1.12280875e-01
3.37858886e-01 -6.97077990e-01 1.32269871e+00 -8.04398894e-01
-1.19415414e+00 1.15232134e+00 4.64809984e-02 -6.00444138e-01
8.63464177e-01 1.86348483e-02 -2.17667267e-01 1.97018817e-01
2.86234945e-01 7.21131086e-01 5.45342386e-01 -1.61907244e+00
-3.99634600e-01 1.66886836e-01 2.00695783e-01 1.37385592e-01
2.07243185e-03 1.27933770e-01 -8.64486769e-02 -8.58321011e-01
-1.72552690e-01 -4.67851043e-01 1.28380015e-01 -2.75526792e-01
-4.11782026e-01 -2.39652634e-01 4.89622444e-01 -7.27754712e-01
1.38555324e+00 -1.75777960e+00 7.03429207e-02 1.00052748e-02
2.38060635e-02 1.07525513e-01 -1.30677491e-01 9.04504538e-01
3.43305543e-02 1.72057852e-01 -1.89732298e-01 -1.53852075e-01
1.23810112e-01 -1.56879485e-01 -3.15592498e-01 -2.11297765e-01
1.17488600e-01 6.76672101e-01 -1.23605061e+00 -9.54806507e-01
3.09935927e-01 4.14321214e-01 -1.89836696e-01 1.37147710e-01
-5.92023373e-01 3.04014087e-01 -5.08525781e-02 1.57365113e-01
5.81327500e-03 -1.14808269e-01 5.86865097e-02 7.69361332e-02
-1.10284112e-01 4.93787885e-01 -9.39127803e-01 1.71915352e+00
-2.72533804e-01 1.08959126e+00 -9.66843486e-01 -4.77287740e-01
1.24521792e+00 5.54503143e-01 9.47277769e-02 -8.76116455e-01
1.49066627e-01 1.18885428e-01 -9.12420452e-02 -6.56307042e-01
8.17578673e-01 -6.21071905e-02 -1.07383989e-01 5.99638224e-01
3.00336897e-01 -5.90283096e-01 7.53667235e-01 3.70592415e-01
1.22808731e+00 3.90959054e-01 6.34784162e-01 -1.87847286e-01
3.12912434e-01 2.45060265e-01 -8.09212551e-02 8.58676791e-01
1.95996508e-01 1.10084999e+00 8.64557207e-01 -3.44286054e-01
-1.37232697e+00 -1.10190952e+00 2.40647957e-01 5.48610210e-01
-4.70757075e-02 -9.29854453e-01 -9.37387884e-01 -5.27694046e-01
-5.00442743e-01 1.33749890e+00 -7.26570666e-01 9.27507579e-02
-4.22085822e-02 -2.71544456e-01 4.30427253e-01 1.10978946e-01
3.08791786e-01 -1.33328736e+00 -1.13434339e+00 1.92199931e-01
-5.51798940e-01 -9.77620065e-01 -9.61303711e-02 -3.11977357e-01
-6.00278080e-01 -1.01099575e+00 -5.93898654e-01 -4.64216888e-01
3.26529145e-01 1.93378553e-01 1.54120481e+00 8.44584256e-02
1.22289911e-01 1.85270578e-01 -1.02487338e+00 -4.70820397e-01
-1.04567289e+00 -2.20091552e-01 -3.24752599e-01 -1.06415041e-01
8.70528966e-02 -4.91707116e-01 -3.30049455e-01 3.04573983e-01
-1.08184314e+00 5.19182324e-01 6.37142956e-01 6.16267264e-01
2.81020850e-01 -5.58210090e-02 2.79947072e-01 -6.89767480e-01
1.03656244e+00 -2.21700475e-01 -6.44393861e-02 4.70202804e-01
-5.76075792e-01 1.79365322e-01 3.23880106e-01 -3.28133613e-01
-9.35766160e-01 -2.90033221e-01 1.35179982e-01 -9.83146504e-02
-2.11965591e-01 6.35856211e-01 1.77387372e-01 3.78284365e-01
1.01931787e+00 2.41467670e-01 -3.17267865e-01 -4.20360044e-02
3.68699312e-01 3.08892429e-01 5.51368952e-01 -3.46242577e-01
7.97991455e-01 1.10270604e-01 -1.85350686e-01 -6.76211774e-01
-6.18798196e-01 -3.06854457e-01 -3.46432686e-01 -9.79256094e-01
9.41263378e-01 -5.69806695e-01 -4.07142580e-01 1.30600363e-01
-1.54087412e+00 -2.06128478e-01 -4.91678506e-01 4.15245652e-01
-9.85871136e-01 2.90227264e-01 -1.54501557e-01 -9.50620294e-01
-1.37802258e-01 -7.84795702e-01 9.54933822e-01 -2.04574745e-02
-1.15661991e+00 -7.10797429e-01 4.26897049e-01 4.19840693e-01
3.96359116e-01 8.75662029e-01 8.79058957e-01 -3.86474073e-01
-3.27769846e-01 -3.09885025e-01 -1.21706322e-01 6.30399734e-02
-2.11367488e-01 1.46614596e-01 -8.82848680e-01 1.96689114e-01
-8.48351046e-02 -4.70422208e-01 7.68827200e-01 2.89404690e-01
5.51963389e-01 -4.23193693e-01 7.81283230e-02 -2.13944882e-01
1.45531225e+00 2.88889349e-01 1.00976920e+00 6.69498444e-01
4.00159448e-01 9.65888560e-01 8.27820718e-01 4.74165559e-01
3.11888754e-01 8.50870788e-01 5.13674438e-01 1.20605305e-01
-4.52992797e-01 -7.40720570e-01 5.46041906e-01 6.76908970e-01
-3.92592818e-01 -5.62475681e-01 -1.11952627e+00 6.67053759e-01
-2.08869314e+00 -1.30714869e+00 -4.86288130e-01 2.17262435e+00
5.86627841e-01 5.69810033e-01 2.83274263e-01 6.31167769e-01
5.80177724e-01 4.19777870e-01 2.65829831e-01 -6.68651462e-01
-5.30970991e-01 -9.74091142e-02 -2.92491913e-02 3.51573080e-01
-6.85829520e-01 7.17212617e-01 6.31038189e+00 1.05544615e+00
-6.88051343e-01 1.71317011e-01 4.38856214e-01 -1.24674551e-01
-5.42954385e-01 2.02375934e-01 -1.34443462e-01 3.73894602e-01
6.93359554e-01 -3.21661979e-01 1.28328189e-01 7.68288553e-01
3.86826962e-01 -4.81395662e-01 -1.18869019e+00 9.59929228e-01
4.75485176e-01 -1.41814709e+00 4.29328978e-01 -1.17971987e-01
8.54348958e-01 -4.03904229e-01 -2.47319490e-01 7.06188455e-02
4.09402907e-01 -1.13483214e+00 1.48508298e+00 8.64273787e-01
5.20250499e-01 -4.89955366e-01 9.20427501e-01 2.71560282e-01
-7.88494229e-01 4.52476233e-01 5.13744205e-02 -2.29483396e-01
4.11436260e-01 4.86811012e-01 -1.15070951e+00 5.99029064e-01
4.75016624e-01 3.58402908e-01 -9.16464210e-01 1.24024439e+00
-6.60145819e-01 5.07662833e-01 2.56424636e-01 -5.39506078e-01
2.74444282e-01 1.23968616e-01 8.74681294e-01 1.41639328e+00
6.64647341e-01 -3.16894412e-01 -3.30429345e-01 9.79514241e-01
4.00186151e-01 4.97547865e-01 -8.60773444e-01 -1.80137411e-01
2.34136239e-01 9.24947679e-01 -9.92217183e-01 -3.93765032e-01
-1.45570293e-01 8.91042471e-01 -4.76080477e-02 -1.20349474e-01
-7.07819819e-01 -2.11689785e-01 5.99889383e-02 4.94785190e-01
1.22981302e-01 -7.30215162e-02 -5.71459949e-01 -7.31179178e-01
1.52297825e-01 -9.33638394e-01 5.80617860e-02 -1.53505886e+00
-9.55123007e-01 9.07039165e-01 1.71979472e-01 -1.60064209e+00
-5.97276032e-01 -2.46912763e-01 -8.11336935e-01 3.60143214e-01
-8.02361012e-01 -1.18285739e+00 -4.62640941e-01 9.42717046e-02
5.88635743e-01 -1.86802864e-01 7.33000219e-01 -1.79239556e-01
1.70563027e-01 1.42680123e-01 -3.47677618e-01 -1.79948494e-01
7.47583508e-01 -1.36937106e+00 4.74129856e-01 9.70456898e-01
4.85130221e-01 1.26946747e-01 1.47339678e+00 -6.06871247e-01
-5.43718517e-01 -6.61882579e-01 1.40092945e+00 -6.59520745e-01
5.75679719e-01 -1.01809293e-01 -7.61778235e-01 1.43122196e-01
5.80353916e-01 -1.00689781e+00 5.23586929e-01 -7.57121518e-02
-4.15977538e-01 3.06811213e-01 -8.23330104e-01 6.25221074e-01
1.01622856e+00 -3.26917082e-01 -8.91318619e-01 3.26616347e-01
5.71929216e-01 -1.27880752e-01 -6.05490565e-01 1.34901404e-01
5.78799605e-01 -1.42535508e+00 6.46767318e-01 -3.64669681e-01
1.35787308e+00 -4.80829686e-01 -1.12575762e-01 -1.32941091e+00
-2.88828492e-01 -4.67458338e-01 2.28243530e-01 1.41729057e+00
4.90064204e-01 -2.48068701e-02 4.44355339e-01 3.15472633e-02
-1.13941438e-01 -3.52236241e-01 -7.87326992e-01 -9.57080841e-01
-2.16747150e-01 -6.28826201e-01 7.32923150e-01 1.04445624e+00
3.57849777e-01 6.27034068e-01 -5.70356011e-01 -5.85333467e-01
3.77864778e-01 -2.88619280e-01 1.02658629e+00 -1.21346307e+00
-3.35927755e-01 -9.17239249e-01 -6.79153085e-01 -2.08869472e-01
-2.93293834e-01 -7.89226413e-01 1.46360368e-01 -2.20211887e+00
3.89510781e-01 2.00451970e-01 7.68797845e-02 2.36707494e-01
2.60422118e-02 2.11069405e-01 4.84155923e-01 3.08929414e-01
-8.12131345e-01 3.48657191e-01 1.14456105e+00 -1.89803779e-01
2.27433518e-02 -5.34634531e-01 -5.58999538e-01 6.53160214e-01
7.22166121e-01 -3.88876826e-01 -4.35075432e-01 -3.29278201e-01
8.00479710e-01 1.84659690e-01 6.03013098e-01 -1.52328873e+00
1.12599500e-01 -1.17145695e-01 8.84140879e-02 -4.83181596e-01
1.98839515e-01 -4.48703796e-01 5.48680186e-01 3.57596993e-01
-4.16672081e-01 6.17514737e-02 -2.02130839e-01 1.36802405e-01
-5.24151683e-01 -4.45804924e-01 4.28403646e-01 -2.73782492e-01
-7.93338299e-01 -4.50073451e-01 -6.54795349e-01 -1.10855848e-01
1.02176154e+00 -6.06224775e-01 -4.78798658e-01 -8.59379053e-01
-4.25000429e-01 -1.32608740e-02 7.53379703e-01 7.24164784e-01
6.90009832e-01 -1.58512723e+00 -1.24486387e+00 -5.64772308e-01
6.19586289e-01 -4.78292525e-01 3.45496880e-03 4.73679960e-01
-7.87585855e-01 4.79251921e-01 -5.53025723e-01 -5.02771676e-01
-1.37498546e+00 4.62358028e-01 -1.47793636e-01 -6.61557674e-01
-2.86396980e-01 4.77468431e-01 -2.33289208e-02 1.47594035e-01
-3.89764085e-02 -1.62547767e-01 -6.49274230e-01 3.78691405e-01
6.81441486e-01 3.50990593e-01 -3.80495489e-02 -7.79646397e-01
-1.13635190e-01 4.62412536e-01 2.97163278e-01 -7.14202523e-01
1.12749219e+00 1.28397584e-01 6.33796602e-02 9.11604643e-01
6.54147029e-01 -2.82116890e-01 -8.09584379e-01 -1.19870529e-01
2.58981913e-01 -4.26479727e-01 -1.53997123e-01 -9.35955524e-01
-4.34538931e-01 7.52818644e-01 3.56728613e-01 7.68836915e-01
8.35907102e-01 1.99110046e-01 4.12295580e-01 1.81425288e-01
4.44724858e-01 -1.28906357e+00 5.09521902e-01 4.42991346e-01
1.56145108e+00 -1.20288455e+00 -4.87693101e-02 -1.83175907e-01
-1.01436400e+00 1.10874844e+00 3.25312197e-01 6.67365342e-02
-3.59700955e-02 -6.05764613e-02 -3.98816578e-02 -3.31728667e-01
-8.45381618e-01 -4.38846141e-01 5.62815547e-01 6.28972292e-01
7.32875288e-01 2.16207832e-01 -8.17140818e-01 3.70415717e-01
-9.59143937e-01 9.40136835e-02 7.01112509e-01 7.12443531e-01
-6.20933294e-01 -8.80659580e-01 -5.20485103e-01 3.44673425e-01
-1.03839971e-01 1.71749622e-01 -9.77231383e-01 7.70964265e-01
2.70922452e-01 1.23687875e+00 -1.35327920e-01 -7.34608531e-01
4.03858602e-01 1.35344435e-02 7.35066354e-01 -6.36651456e-01
-6.80302262e-01 -2.93153644e-01 5.53799212e-01 -3.31749290e-01
-6.94449067e-01 -8.89214754e-01 -7.40583420e-01 -4.70165014e-01
2.58502197e-02 9.50867385e-02 6.75443053e-01 1.03104973e+00
-2.14649737e-01 5.95015824e-01 1.59979120e-01 -6.78953707e-01
2.21944287e-01 -1.28928292e+00 -2.17560470e-01 7.10722744e-01
-1.11150846e-01 -4.80763346e-01 -1.30661398e-01 4.20820147e-01] | [11.75562572479248, 8.791535377502441] |
34cd6c43-657d-4c85-9fc7-2b209fdc7ab3 | discover-the-mysteries-of-the-maya-selected | 2208.03163 | null | https://arxiv.org/abs/2208.03163v2 | https://arxiv.org/pdf/2208.03163v2.pdf | Discover the Mysteries of the Maya: Selected Contributions from the Machine Learning Challenge & The Discovery Challenge Workshop at ECML PKDD 2021 | The volume contains selected contributions from the Machine Learning Challenge "Discover the Mysteries of the Maya", presented at the Discovery Challenge Track of The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2021). Remote sensing has greatly accelerated traditional archaeological landscape surveys in the forested regions of the ancient Maya. Typical exploration and discovery attempts, beside focusing on whole ancient cities, focus also on individual buildings and structures. Recently, there have been several successful attempts of utilizing machine learning for identifying ancient Maya settlements. These attempts, while relevant, focus on narrow areas and rely on high-quality aerial laser scanning (ALS) data which covers only a fraction of the region where ancient Maya were once settled. Satellite image data, on the other hand, produced by the European Space Agency's (ESA) Sentinel missions, is abundant and, more importantly, publicly available. The "Discover the Mysteries of the Maya" challenge aimed at locating and identifying ancient Maya architectures (buildings, aguadas, and platforms) by performing integrated image segmentation of different types of satellite imagery (from Sentinel-1 and Sentinel-2) data and ALS (lidar) data. | ['Žiga Kokalj', 'Ivica Dimitrovski', 'Ana Kostovska', 'Nikola Simidjievski', 'Dragi Kocev'] | 2022-08-05 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 1.46960586e-01 -7.16189817e-02 -6.25042170e-02 -2.22237706e-02
-9.33948815e-01 -4.01528299e-01 8.92530203e-01 1.72339112e-01
-5.96615434e-01 7.65449464e-01 -2.58339614e-01 -5.54006457e-01
-3.80658984e-01 -1.20486939e+00 -3.96790236e-01 -6.70177877e-01
-4.54257041e-01 7.95541644e-01 1.26987129e-01 -5.61566770e-01
8.39954019e-02 9.46019709e-01 -1.74256623e+00 3.19706023e-01
5.04169405e-01 8.86658907e-01 5.37717342e-01 3.48807067e-01
-2.30551749e-01 1.21236883e-01 -1.40501514e-01 1.54325455e-01
4.96780872e-01 2.78855693e-02 -9.21387970e-01 -2.02488288e-01
3.33938599e-01 -1.67095542e-01 -6.21779338e-02 7.09600210e-01
1.93208992e-01 -2.03159809e-01 5.92495859e-01 -7.35720634e-01
-5.53673133e-02 4.42112923e-01 -6.98065162e-01 4.75379914e-01
-6.89437389e-02 5.35252057e-02 1.20851302e+00 -9.87917244e-01
6.85513616e-01 1.08289242e+00 1.11311591e+00 -2.54799873e-01
-1.12251198e+00 -4.64448512e-01 -3.94318700e-02 4.62452143e-01
-1.98110020e+00 -5.13166547e-01 3.68438840e-01 -7.47156620e-01
9.82160449e-01 3.22230995e-01 8.98683012e-01 5.45318604e-01
-1.88019816e-02 7.36651003e-01 1.20817232e+00 -5.09421468e-01
4.36954528e-01 -2.89480239e-01 -3.67398053e-01 6.61108971e-01
4.34023619e-01 1.21292926e-01 -3.94091994e-01 -4.04977351e-01
5.49649835e-01 6.90544471e-02 -1.30166292e-01 -2.83454340e-02
-1.05657518e+00 9.75485384e-01 5.00102282e-01 4.41714048e-01
-7.91576982e-01 -1.18318103e-01 2.35090971e-01 2.40655333e-01
3.84813935e-01 2.58749217e-01 -6.30484700e-01 4.52647991e-02
-1.36664379e+00 2.27168992e-01 4.53242451e-01 4.14863795e-01
1.29645956e+00 -2.45690793e-02 8.45510244e-01 8.57224643e-01
5.61118364e-01 7.58532643e-01 1.88210011e-01 -6.92304790e-01
3.66621435e-01 7.31249332e-01 2.39000961e-01 -7.63375282e-01
-6.82607234e-01 -1.74440414e-01 -7.37513781e-01 4.50906843e-01
2.36772835e-01 -6.36425614e-02 -9.67586696e-01 8.14929962e-01
4.48822826e-01 -2.12606281e-01 -1.04034217e-02 7.91965604e-01
4.59516913e-01 4.73201513e-01 1.56461984e-01 -4.87115793e-03
1.28575754e+00 -6.22666813e-02 -1.00191981e-02 -4.07435566e-01
6.46049142e-01 -4.55065161e-01 7.93209195e-01 1.64206848e-01
-3.92721593e-01 -3.80728513e-01 -1.08235657e+00 4.04216468e-01
-8.84384692e-01 1.29934013e-01 9.98410165e-01 5.98501444e-01
-8.60392630e-01 5.23009300e-01 -8.46633017e-01 -8.29481363e-01
8.39189529e-01 1.00504458e-01 -4.12190765e-01 -2.35616136e-02
-1.21335018e+00 9.25970852e-01 3.96531641e-01 3.87503833e-01
-7.37365961e-01 -7.61388838e-01 -6.64944351e-01 -3.95583063e-01
2.58492202e-01 -2.67679274e-01 7.08373189e-01 -6.89535916e-01
-7.91778147e-01 1.33590376e+00 3.88649374e-01 -6.36473596e-01
3.22164625e-01 -2.58080572e-01 -5.43902755e-01 -7.56801758e-03
7.60435283e-01 4.96656656e-01 1.25174448e-02 -1.10395062e+00
-1.16681004e+00 -9.24131393e-01 -2.20941678e-01 -2.24243492e-01
2.14957640e-01 1.59646541e-01 1.75580367e-01 -5.09727836e-01
4.98995125e-01 -8.48324776e-01 -4.66354221e-01 -7.39479288e-02
-1.55288264e-01 -3.47213969e-02 7.12324679e-01 -8.74824464e-01
9.18157756e-01 -1.95911503e+00 -1.98833421e-01 4.83764589e-01
-1.78096384e-01 2.24207014e-01 2.03328058e-01 7.47463703e-01
9.27630886e-02 3.44614983e-01 -7.10558057e-01 3.13512683e-01
-2.16481164e-01 5.93152285e-01 -2.80085802e-01 5.86075068e-01
-1.37430221e-01 5.83528459e-01 -7.13133276e-01 -4.18716073e-01
2.80420721e-01 1.88060239e-01 2.11000010e-01 -5.82632065e-01
-3.96388531e-01 2.02800110e-01 -6.34180963e-01 1.28432417e+00
7.45085418e-01 3.74560542e-02 2.05976218e-01 4.54842895e-02
-8.51856947e-01 4.02953364e-02 -9.86261129e-01 1.64829290e+00
-4.16438222e-01 6.88970804e-01 4.26625103e-01 -1.00052774e+00
8.84100378e-01 5.18932343e-02 7.58069158e-01 -8.44152987e-01
-2.46255443e-01 6.37408793e-01 -1.40956894e-01 -7.39551723e-01
2.41739333e-01 -5.00587404e-01 1.24107018e-01 3.58986527e-01
-5.40834248e-01 -4.47592279e-03 -9.24830511e-02 -1.92849830e-01
1.11929262e+00 4.34558988e-01 4.49527055e-01 -5.01869500e-01
3.42474639e-01 7.49902189e-01 4.59372401e-01 5.12432098e-01
1.58265531e-02 5.18769741e-01 -2.65087515e-01 -1.06935585e+00
-1.22173047e+00 -1.27040362e+00 -6.76992416e-01 8.96097243e-01
-2.07024083e-01 -3.19238871e-01 -1.68639183e-01 -2.22974628e-01
4.50869888e-01 4.92043376e-01 -6.72440469e-01 5.13555706e-01
-4.30797845e-01 -1.26866078e+00 8.61253738e-01 2.88581520e-01
7.88005531e-01 -8.65202010e-01 -9.62534070e-01 3.43165427e-01
-3.32579136e-01 -7.76601672e-01 6.89978898e-01 3.22491497e-01
-9.50596988e-01 -1.44073582e+00 -4.09499288e-01 -2.41935253e-01
3.43379229e-02 1.55832380e-01 9.26665902e-01 -9.58886743e-02
-8.73632610e-01 3.86470318e-01 -3.40549260e-01 -5.41228473e-01
4.91338857e-02 4.73767854e-02 -7.45588616e-02 1.18294820e-01
4.61060852e-01 -9.65468228e-01 -3.33122849e-01 2.15012178e-01
-3.48024607e-01 -2.00357065e-01 1.10334456e+00 2.68641144e-01
9.63606000e-01 3.31438601e-01 6.62430227e-01 -6.36365533e-01
-2.80298114e-01 -9.89105403e-01 -5.52062094e-01 2.77377456e-01
-5.03991663e-01 -3.64901900e-01 1.03087321e-01 3.28654081e-01
-8.40938687e-01 3.23808849e-01 -2.13490412e-01 3.74691859e-02
-6.26937628e-01 9.22307253e-01 -2.05612749e-01 5.47887459e-02
9.55312490e-01 2.42692277e-01 -2.27739755e-02 -8.70209575e-01
2.07444370e-01 1.11859250e+00 6.12359226e-01 -6.20269775e-01
8.42759013e-01 1.16372156e+00 3.97523008e-02 -1.72551870e+00
-6.63687766e-01 -8.79620969e-01 -1.19833446e+00 -2.24503204e-01
6.77127481e-01 -1.06775916e+00 -1.84005097e-01 2.03660041e-01
-7.70111680e-01 -2.65030831e-01 -2.44357333e-01 3.28013510e-01
-3.35082203e-01 1.02654070e-01 1.79552987e-01 -1.13760066e+00
-3.17733943e-01 -2.77770251e-01 1.02934301e+00 1.38744256e-02
-2.92204291e-01 -7.35886097e-01 4.44369823e-01 5.76633275e-01
4.06309366e-01 8.84579360e-01 8.88743699e-01 -3.49382043e-01
-5.00786006e-01 -9.26244184e-02 -2.09549159e-01 1.00068472e-01
4.05266404e-01 -6.35518087e-03 -6.93108618e-01 1.32024698e-02
-3.19726974e-01 1.02694564e-01 1.20525539e+00 1.32207230e-01
2.05996975e-01 -4.66405414e-02 -6.79804206e-01 2.84521639e-01
1.63798177e+00 2.43131742e-01 7.92026639e-01 1.22954440e+00
3.61896366e-01 7.48711884e-01 6.22251153e-01 6.26743436e-01
3.47911298e-01 4.38881993e-01 6.43839300e-01 8.71427730e-02
-5.40461168e-02 2.87593663e-01 1.04360774e-01 -2.15746835e-02
-6.90525532e-01 5.92961252e-01 -1.66441011e+00 1.13712788e+00
-1.74880135e+00 -1.10506284e+00 -6.61806464e-01 2.25779366e+00
5.03859699e-01 -2.07160294e-01 1.75736621e-01 8.37958232e-02
3.53528261e-01 1.51859537e-01 -4.50026572e-01 2.03122705e-01
-5.58747470e-01 5.31346738e-01 8.62894714e-01 3.39726090e-01
-1.34070802e+00 9.01927471e-01 6.42808914e+00 6.20189607e-01
-9.79814649e-01 1.41858444e-01 2.36112401e-02 -1.03231147e-01
-2.55808532e-01 1.72227457e-01 -6.89342201e-01 3.25503141e-01
1.13328505e+00 6.17016945e-03 2.62621820e-01 9.18118536e-01
5.20943403e-01 -4.68209088e-01 -3.31317395e-01 7.11324334e-01
-4.45767432e-01 -1.69175673e+00 -1.57931864e-01 4.96242464e-01
5.37951648e-01 1.04849792e+00 -4.51703668e-01 1.26502588e-01
2.47594714e-01 -8.84151936e-01 6.60964549e-01 7.22039163e-01
6.68365479e-01 -7.70356059e-01 5.55033922e-01 2.70250440e-01
-1.36933529e+00 -3.53337228e-01 -3.10968786e-01 -2.83156991e-01
-7.10242253e-04 6.90349638e-01 -6.42557085e-01 8.43409061e-01
1.33045828e+00 6.75244868e-01 -7.73224652e-01 1.10309005e+00
-2.01893106e-01 6.20667875e-01 -7.95235991e-01 3.26842189e-01
6.39648080e-01 -4.04367536e-01 4.73658055e-01 9.18870866e-01
2.55458564e-01 2.12950975e-01 2.00496703e-01 6.72222972e-01
2.67915726e-01 7.24929422e-02 -7.53148198e-01 1.06937900e-01
5.23118496e-01 1.08933210e+00 -6.89501166e-01 1.06180079e-01
-3.21430176e-01 4.11108792e-01 -1.73264727e-01 9.09991041e-02
-2.97046453e-01 -1.68852806e-01 7.65809715e-01 7.83871651e-01
2.65331566e-01 -4.26140755e-01 -3.80474389e-01 -5.12714207e-01
-7.88640156e-02 -4.62205470e-01 5.70040464e-01 -6.60214245e-01
-1.08119833e+00 -5.45240520e-03 9.13555771e-02 -1.01661551e+00
6.57294095e-02 -5.22679448e-01 -6.54325366e-01 7.97530115e-01
-1.46145809e+00 -1.69265056e+00 -2.70897627e-01 3.55087221e-01
3.13532084e-01 -3.84881377e-01 9.93523896e-01 2.54864067e-01
-2.02688649e-01 -3.98534387e-01 8.25340331e-01 -5.32989688e-02
9.14133713e-02 -1.00829494e+00 3.81392121e-01 3.76262575e-01
2.14840397e-01 2.05709130e-01 3.50562662e-01 -9.42420423e-01
-1.22241092e+00 -1.31912518e+00 7.67361343e-01 -1.43304691e-01
8.81404996e-01 5.25224917e-02 -8.05969656e-01 9.09556568e-01
-2.54676551e-01 -4.81188744e-01 8.59910905e-01 3.71417373e-01
-2.01672196e-01 -2.76051164e-01 -1.11618567e+00 1.79557040e-01
7.42887855e-01 -5.43587685e-01 -5.66830277e-01 5.88897705e-01
-2.10513353e-01 4.74702060e-01 -9.41966951e-01 5.11340261e-01
8.90414834e-01 -8.78686607e-01 1.06625402e+00 -4.68497038e-01
2.46429101e-01 -4.30244595e-01 -7.90496528e-01 -8.64682674e-01
-1.57261088e-01 -1.52690262e-01 5.38235366e-01 8.65044057e-01
3.74393314e-01 -3.30329865e-01 1.08895087e+00 -1.36071807e-02
-2.20375210e-01 -3.25757861e-01 -1.39702034e+00 -1.11383247e+00
2.45082408e-01 -5.34092069e-01 3.09746623e-01 1.03233612e+00
-6.97410941e-01 1.26088560e-01 -1.03530571e-01 6.00593030e-01
9.27316666e-01 2.84804046e-01 7.18571067e-01 -1.95188987e+00
1.52278677e-01 -3.79880607e-01 -6.97969139e-01 -1.76568031e-01
-3.19022298e-01 -8.67911696e-01 -1.01443939e-01 -1.82820368e+00
-3.49880189e-01 -7.52099276e-01 -7.24587077e-03 9.44073915e-01
5.13623834e-01 2.58681834e-01 -1.93490088e-01 8.53665769e-01
-1.21137075e-01 4.25662011e-01 2.60449260e-01 -4.74471480e-01
-7.34082088e-02 1.54484332e-01 -3.59401137e-01 9.07128394e-01
8.96121502e-01 -4.06025022e-01 1.93290994e-01 -4.35929060e-01
3.92657310e-01 -3.69900137e-01 7.20546246e-01 -1.25991917e+00
1.80104136e-01 -4.33525801e-01 1.90884829e-01 -1.18645215e+00
3.91097993e-01 -8.84027004e-01 7.76449084e-01 7.22039282e-01
5.68324327e-01 -5.16398132e-01 3.09096098e-01 5.49152076e-01
-6.81687444e-02 -1.66837230e-01 1.06205559e+00 -5.47119439e-01
-1.24393570e+00 2.03608066e-01 -7.81120956e-01 -3.66548568e-01
1.02715921e+00 -4.57611203e-01 -3.25798281e-02 1.58514872e-01
-8.85408282e-01 2.22381562e-01 5.05972087e-01 1.62887216e-01
3.72908920e-01 -1.08013868e+00 -1.12258673e+00 1.43121853e-01
1.98487580e-01 1.18662357e-01 2.94860750e-01 7.67056882e-01
-1.10079801e+00 4.44805056e-01 -5.09587705e-01 -4.95132476e-01
-1.03056502e+00 9.60857570e-02 5.25937617e-01 6.50608018e-02
-8.49718809e-01 6.71773791e-01 -4.45225835e-01 -5.87910950e-01
-4.00658578e-01 3.47002238e-01 -2.15209097e-01 6.34278476e-01
6.47689044e-01 6.45283937e-01 8.75687003e-02 -9.67429876e-01
-7.10465670e-01 4.67135966e-01 1.86199605e-01 -2.15157215e-02
2.03657985e+00 -2.19641596e-01 -3.12005132e-01 6.48784399e-01
7.28254199e-01 -2.64772177e-01 -6.36496902e-01 -3.42798769e-01
6.80581331e-01 -2.45704457e-01 2.55378783e-01 -1.04178143e+00
-8.72946084e-01 8.04962873e-01 9.18784738e-01 1.49416566e-01
7.02348769e-01 2.85082251e-01 4.75241631e-01 8.69159937e-01
6.91787124e-01 -1.33293664e+00 -6.53187215e-01 1.87544942e-01
1.06362176e+00 -1.14316559e+00 4.01503652e-01 -6.47562891e-02
-4.25030999e-02 1.20199001e+00 1.07863963e-01 2.82045037e-01
8.89572322e-01 4.58451621e-02 -4.76301014e-02 -5.55613339e-01
-1.75920904e-01 -7.25221932e-01 -4.39313442e-01 9.50178921e-01
-1.16701379e-01 2.99103260e-01 -1.44722700e-01 5.69652021e-01
-2.71342516e-01 -1.02076747e-01 2.14928344e-01 1.38071454e+00
-1.08884454e+00 -9.40178037e-01 -9.50614214e-01 6.23421848e-01
-5.44984974e-02 -3.42871845e-01 -5.94512701e-01 1.33397043e+00
5.05625904e-01 7.07802892e-01 1.06721930e-01 -2.53117085e-01
3.92558843e-01 1.49886504e-01 2.38460183e-01 -5.42387545e-01
-3.21304232e-01 -6.43612966e-02 3.06777149e-01 -3.23710412e-01
-5.50453126e-01 -1.28660250e+00 -1.28854084e+00 -5.02582550e-01
4.15161923e-02 1.89261287e-01 1.14017105e+00 1.10196733e+00
1.66583106e-01 -1.87451951e-02 4.61788893e-01 -1.00777292e+00
-7.53441527e-02 -9.65651035e-01 -1.20712829e+00 -3.84149581e-01
-1.88111812e-01 -6.49758458e-01 -1.71409830e-01 2.95237843e-02] | [9.425747871398926, -1.4595485925674438] |
0f184ac8-4d15-4a4e-9195-9b223d6d6f23 | iterative-span-selection-self-emergence-of | null | null | https://aclanthology.org/2022.coling-1.478 | https://aclanthology.org/2022.coling-1.478.pdf | Iterative Span Selection: Self-Emergence of Resolving Orders in Semantic Role Labeling | Semantic Role Labeling (SRL) is the task of labeling semantic arguments for marked semantic predicates. Semantic arguments and their predicates are related in various distinct manners, of which certain semantic arguments are a necessity while others serve as an auxiliary to their predicates. To consider such roles and relations of the arguments in the labeling order, we introduce iterative argument identification (IAI), which combines global decoding and iterative identification for the semantic arguments. In experiments, we first realize that the model with random argument labeling orders outperforms other heuristic orders such as the conventional left-to-right labeling order. Combined with simple reinforcement learning, the proposed model spontaneously learns the optimized labeling orders that are different from existing heuristic orders. The proposed model with the IAI algorithm achieves competitive or outperforming results from the existing models in the standard benchmark datasets of span-based SRL: CoNLL-2005 and CoNLL-2012. | ['Satoshi Sekine', 'Kentaro Inui', 'Hiroki Ouchi', 'Shuhei Kurita'] | null | null | null | null | coling-2022-10 | ['semantic-role-labeling'] | ['natural-language-processing'] | [ 6.37223303e-01 5.93967438e-01 -7.62810290e-01 -6.25873685e-01
-6.42727733e-01 -1.07919061e+00 6.49341166e-01 4.45344567e-01
-3.92441988e-01 9.55671012e-01 6.31403625e-01 -3.90041322e-01
-5.39771795e-01 -6.39586151e-01 -7.34138429e-01 -4.03318554e-01
5.14686815e-02 1.05603790e+00 6.89010561e-01 -4.30355340e-01
4.56303477e-01 1.53648928e-02 -1.78666532e+00 7.21901417e-01
8.63792300e-01 1.02841032e+00 2.51838773e-01 2.02364564e-01
-6.59621894e-01 1.32162035e+00 -7.05622077e-01 -3.13550651e-01
-1.17121488e-01 -4.23313409e-01 -1.40021193e+00 -3.18413019e-01
-4.83740494e-03 1.37370080e-01 4.92297076e-02 1.03700042e+00
2.60181993e-01 1.16987452e-01 3.14841509e-01 -1.44321465e+00
-6.62324905e-01 1.35368001e+00 -1.72231138e-01 6.81151599e-02
6.72298133e-01 -5.22901177e-01 1.74393666e+00 -2.69036412e-01
6.68759584e-01 1.79762900e+00 3.31349164e-01 6.54155731e-01
-7.96810448e-01 -3.83500278e-01 6.64741814e-01 6.60485148e-01
-6.32468283e-01 -2.08963335e-01 9.91909802e-01 -8.92477576e-03
9.45206642e-01 1.62428737e-01 3.64388913e-01 7.44670570e-01
-6.10077858e-01 1.18003261e+00 1.07298338e+00 -5.00124037e-01
2.14551687e-01 -3.63290101e-01 6.87121570e-01 4.66725230e-01
2.00931206e-01 -1.11364193e-01 -5.17621219e-01 -4.59632635e-01
3.04579049e-01 -5.33171237e-01 9.85624641e-02 -2.94356406e-01
-1.14013577e+00 9.42669749e-01 1.75228447e-01 1.39228150e-01
-9.42405984e-02 5.35330117e-01 7.62417436e-01 4.12487358e-01
1.76739722e-01 4.99773890e-01 -9.66060042e-01 -2.60701105e-02
-2.60456234e-01 3.65478307e-01 7.96480119e-01 1.05880618e+00
4.55265909e-01 -2.62081683e-01 -3.77206802e-01 1.06152034e+00
4.52142537e-01 1.78800553e-01 3.36287022e-01 -1.28457654e+00
6.11888766e-01 1.08527589e+00 4.62793916e-01 -4.49541122e-01
-4.55489725e-01 -1.23255901e-01 2.18224283e-02 -3.37026179e-01
5.85803270e-01 2.63065994e-01 -6.87437475e-01 2.17801738e+00
6.01430714e-01 2.14194626e-01 5.77228248e-01 6.47907317e-01
1.20142257e+00 4.73653287e-01 7.31218517e-01 -2.32017681e-01
1.75118971e+00 -1.06747973e+00 -8.10114384e-01 -3.40759009e-01
7.39001036e-01 -5.16381741e-01 1.26342034e+00 3.79348323e-02
-8.87587070e-01 -2.36100167e-01 -8.59766662e-01 -2.42890209e-01
-3.81677210e-01 1.66723952e-01 1.04256701e+00 3.57693344e-01
-6.22275114e-01 5.62596202e-01 -1.20119341e-01 -1.21474795e-01
4.74688381e-01 1.94282979e-01 -1.92392468e-01 9.21427831e-02
-1.85310376e+00 5.52480817e-01 1.02627158e+00 -1.16386242e-01
-9.39457834e-01 -5.88505983e-01 -1.08549058e+00 1.25073686e-01
8.52954865e-01 -3.68742287e-01 1.46625173e+00 -7.78714955e-01
-1.22616363e+00 1.25632489e+00 -3.18639189e-01 -6.90699041e-01
3.35860163e-01 -4.51001614e-01 -4.49396372e-01 3.31500679e-01
6.05785370e-01 5.92191994e-01 7.38732159e-01 -1.38189602e+00
-1.13244104e+00 -1.99279949e-01 7.15763748e-01 5.83491862e-01
8.48028436e-02 2.92295039e-01 -1.77050769e-01 -3.91939461e-01
4.45600897e-01 -6.74238145e-01 -9.72793773e-02 -5.82839608e-01
-5.26829720e-01 -1.09148383e+00 8.05287302e-01 -3.01256597e-01
1.22326159e+00 -1.95143688e+00 -7.05561787e-02 1.26063414e-02
-1.03009894e-01 1.52587637e-01 -1.00842431e-01 3.81673872e-01
-4.53008026e-01 3.72956060e-02 -1.45813107e-01 2.21952572e-01
3.03154796e-01 6.51429057e-01 -5.62786281e-01 1.57694280e-01
-2.03100175e-01 9.88546968e-01 -1.47165406e+00 -7.21239150e-01
-3.89011502e-02 -3.63359392e-01 -3.78424793e-01 2.93698847e-01
-8.91919613e-01 5.07998407e-01 -7.14282691e-01 7.97297955e-01
4.81501967e-01 -5.02649307e-01 7.62136221e-01 -3.32746834e-01
2.40434557e-01 1.01790881e+00 -1.14888811e+00 1.48251784e+00
-9.74581987e-02 -1.40098363e-01 -7.43182600e-02 -1.32065475e+00
6.13155246e-01 4.21596885e-01 4.33365196e-01 -7.10019767e-01
1.06422320e-01 4.93640453e-01 1.50075153e-01 -3.55304450e-01
1.44037470e-01 1.98672470e-02 -6.17095053e-01 5.32634854e-01
-1.76797897e-01 6.62288070e-02 7.82968163e-01 1.43742844e-01
9.43671644e-01 5.20516336e-01 5.98387659e-01 -5.21139801e-01
1.28197157e+00 1.77401572e-01 9.59254146e-01 8.92788529e-01
-2.49197096e-01 -1.27367690e-01 9.55086291e-01 -5.44502378e-01
-7.34254301e-01 -9.12709117e-01 -2.67476141e-02 1.53516567e+00
6.77862942e-01 -3.92844021e-01 -6.42248809e-01 -1.43541372e+00
1.30382597e-01 8.56673360e-01 -3.81289184e-01 1.11030146e-01
-1.04602027e+00 -4.88120794e-01 7.90915728e-01 6.09582305e-01
6.95922792e-01 -1.64506578e+00 -7.32759953e-01 3.35861176e-01
-3.21590453e-01 -1.38361478e+00 -3.42922024e-02 5.09190619e-01
-6.80255353e-01 -1.77041638e+00 1.93586215e-01 -1.11909533e+00
6.40778542e-01 -1.93144843e-01 1.41796803e+00 3.85071516e-01
2.91774720e-01 1.27729937e-01 -7.40094781e-01 -2.49779910e-01
-4.01473492e-01 -1.10201262e-01 -3.32961380e-01 -3.18416625e-01
3.67932647e-01 -4.73290324e-01 -4.60221529e-01 3.30467701e-01
-6.59185767e-01 -6.70070527e-03 1.79020703e-01 1.03343678e+00
7.23538876e-01 3.26558024e-01 7.89152205e-01 -1.61542308e+00
6.90672636e-01 -4.11537886e-01 -5.60943961e-01 6.65908098e-01
-5.19525707e-01 4.60894495e-01 8.22793841e-01 -1.79540440e-01
-1.36327887e+00 -1.22945540e-01 -2.62477428e-01 3.05409491e-01
-1.01325102e-01 1.88846141e-01 -6.82349682e-01 3.90711486e-01
3.08887690e-01 2.63411924e-03 -4.20435011e-01 -5.06455421e-01
5.94610333e-01 2.48378396e-01 4.41084892e-01 -1.57086468e+00
4.33010727e-01 4.07084852e-01 2.65051313e-02 -6.47168681e-02
-1.66284394e+00 -3.95851165e-01 -2.73188502e-01 2.07248777e-01
7.14401186e-01 -7.77568817e-01 -9.26164329e-01 3.13622743e-01
-1.27955437e+00 -3.03357631e-01 -4.86484170e-01 3.97561044e-02
-8.26560676e-01 4.61906105e-01 -8.27483177e-01 -5.55360973e-01
-2.55660117e-01 -9.73338604e-01 1.14525831e+00 1.83758482e-01
-4.24190134e-01 -1.05610669e+00 -5.00000834e-01 4.23892647e-01
-9.31959450e-02 -1.61992908e-01 1.68576777e+00 -1.28169501e+00
-1.45547748e-01 2.93519765e-01 -3.41824204e-01 2.82397151e-01
2.18469948e-01 -6.96266532e-01 -6.16214037e-01 4.85490933e-02
-3.42330150e-02 -6.02749169e-01 6.97815061e-01 2.24128202e-01
1.37010145e+00 -3.48863900e-01 -3.93492907e-01 2.64879942e-01
1.20933199e+00 6.33492887e-01 5.58338940e-01 7.12077260e-01
5.83007753e-01 8.54990840e-01 1.27440035e+00 1.69059038e-01
4.14809436e-01 2.16651857e-01 6.09038174e-01 1.73769265e-01
-1.49613708e-01 -6.97268844e-01 -4.97851670e-02 1.66058719e-01
1.12080842e-01 -3.07691574e-01 -5.54874182e-01 4.31743771e-01
-2.29567790e+00 -9.93679583e-01 -2.27528185e-01 1.80966198e+00
9.27531183e-01 4.13954943e-01 -9.65483636e-02 4.54021633e-01
1.01135445e+00 5.44046283e-01 -6.51552498e-01 -3.12319249e-01
-4.06394780e-01 1.45768091e-01 5.29034436e-01 5.60846329e-01
-1.30015337e+00 1.48175728e+00 6.69915676e+00 1.03126860e+00
-2.21005857e-01 4.11166072e-01 4.01709408e-01 5.33959925e-01
-7.80490160e-01 6.08552992e-01 -1.07665908e+00 4.18210149e-01
4.15603101e-01 3.94951217e-02 4.59547192e-01 1.05858076e+00
-1.96775451e-01 -5.93463518e-02 -1.21703815e+00 6.51534975e-01
-2.51138717e-01 -1.32142830e+00 3.38271707e-01 -6.30932450e-01
4.68249083e-01 -4.95400190e-01 -3.74951601e-01 3.85337532e-01
8.03759575e-01 -7.54116058e-01 1.11128366e+00 -1.55913875e-01
6.32417560e-01 -5.78820229e-01 4.99284476e-01 3.18623751e-01
-1.43657100e+00 -6.08223736e-01 -2.63338447e-01 -1.06017269e-01
2.73548663e-01 3.34560215e-01 -4.91730422e-01 6.27885342e-01
4.55435485e-01 8.12297165e-01 -1.11141287e-01 3.75407696e-01
-1.39506710e+00 7.57286429e-01 -8.90928060e-02 -1.58117205e-01
2.36554682e-01 1.27955712e-02 5.61231256e-01 6.85146749e-01
-1.12870589e-01 8.37820545e-02 2.81791717e-01 5.12834728e-01
-2.20537201e-01 1.30544975e-01 -2.28965461e-01 -1.74004152e-01
1.03480327e+00 8.83066475e-01 -9.80247855e-01 -6.72560692e-01
-2.16745451e-01 7.07965672e-01 3.74957383e-01 1.62373617e-01
-1.01843989e+00 -2.00408567e-02 5.28969526e-01 1.30920336e-01
1.10750459e-02 2.38959864e-01 -1.63634554e-01 -6.77706063e-01
-6.11327700e-02 -9.45031703e-01 1.30234218e+00 -6.72361493e-01
-1.20603871e+00 4.03554857e-01 4.68853742e-01 -9.76668239e-01
-2.10213557e-01 -7.86887050e-01 -3.30909640e-01 5.13738334e-01
-1.94647086e+00 -1.30692697e+00 1.89947769e-01 6.46881223e-01
7.48759806e-01 -9.67806205e-02 7.66324639e-01 1.71838954e-01
-1.16192633e-02 3.06491375e-01 -5.59151947e-01 -9.86542553e-02
3.73914689e-01 -1.30423272e+00 2.46455371e-01 6.58478558e-01
-1.26582891e-01 7.37440765e-01 6.40349090e-01 -8.72299135e-01
-8.90815377e-01 -7.68096268e-01 1.22021127e+00 -2.84046024e-01
7.42587626e-01 -9.12050903e-02 -5.95815361e-01 8.85711193e-01
8.82521123e-02 -6.57692775e-02 4.98006254e-01 1.95051059e-01
-4.86402214e-01 -1.34044224e-02 -1.15873861e+00 4.30897743e-01
1.75045562e+00 -4.05406594e-01 -1.30005169e+00 5.14097631e-01
1.28305411e+00 -4.48233128e-01 -2.72087663e-01 7.61327326e-01
2.83154309e-01 -7.26439416e-01 1.24696934e+00 -8.62291217e-01
2.92763799e-01 -7.39723861e-01 -3.67093831e-01 -8.96711290e-01
-2.57330209e-01 -4.81915176e-01 -8.14664662e-02 1.27241719e+00
4.90246594e-01 -7.60236800e-01 8.36651027e-01 3.17393601e-01
-2.78028101e-01 -4.19005841e-01 -9.38132942e-01 -7.70862997e-01
-2.32651815e-01 -5.24813831e-01 1.11396229e+00 8.39573324e-01
-1.92512095e-01 6.25061989e-01 -1.42379671e-01 1.13803737e-01
7.09394276e-01 6.87674701e-01 9.87814814e-02 -1.50337446e+00
-3.37614924e-01 -2.87214994e-01 2.13543802e-01 -1.15593505e+00
9.78445888e-01 -1.19535851e+00 -6.83709010e-02 -1.81544065e+00
-7.27791563e-02 -9.03442025e-01 -6.14041686e-01 9.07445610e-01
-4.01848644e-01 -3.49329710e-01 -1.86371818e-01 1.93032891e-01
-1.03124499e+00 3.21257651e-01 1.25645888e+00 -1.39719486e-01
-4.05517817e-02 -1.35213837e-01 -9.89904225e-01 1.08483934e+00
6.77251101e-01 -6.71723664e-01 -9.97446358e-01 -3.56263012e-01
6.45006835e-01 1.26965418e-01 7.63409119e-03 -5.50916672e-01
1.02935664e-01 -3.46873045e-01 -1.85522050e-01 -6.82531297e-01
3.08671743e-02 -7.65669525e-01 -1.64616778e-01 4.83185112e-01
-8.14903021e-01 -4.66080494e-02 -2.67781585e-01 4.64351296e-01
-4.01640713e-01 -7.13979959e-01 5.01524389e-01 -5.33479035e-01
-1.43646371e+00 -5.66907413e-02 7.71921203e-02 6.41490042e-01
9.80323732e-01 -3.57521474e-02 -6.22801065e-01 5.56413643e-02
-8.02010894e-01 5.13288558e-01 5.95632754e-03 4.94826347e-01
4.50328678e-01 -1.24000084e+00 -4.66616303e-01 -4.63688672e-02
2.04037130e-01 1.75892845e-01 -1.19417030e-02 3.29653710e-01
-3.43267024e-01 3.72966439e-01 2.21004989e-02 -2.09354952e-01
-1.28562856e+00 5.88839173e-01 1.91508792e-02 -8.19055796e-01
-4.36592489e-01 1.09865594e+00 3.19044679e-01 -7.70441711e-01
2.16959745e-01 -4.80980426e-02 -7.29478180e-01 3.39877978e-02
1.95270300e-01 7.64333904e-02 -2.78653949e-01 -5.75154841e-01
-5.83220840e-01 3.82297069e-01 1.32239237e-01 1.27361327e-01
1.08298254e+00 -2.94177104e-02 -5.84590614e-01 2.11693406e-01
6.02069855e-01 1.33651897e-01 -8.54786336e-01 -2.93598503e-01
7.25425959e-01 -2.70725131e-01 -7.15592206e-01 -8.02305162e-01
-4.87421900e-01 3.89535517e-01 -6.74990118e-02 2.62715757e-01
1.06461751e+00 5.26328683e-01 7.45794117e-01 2.79190600e-01
6.61318779e-01 -1.41001713e+00 7.51454085e-02 8.96682143e-01
6.87089682e-01 -7.46012509e-01 -1.45877674e-01 -1.13409138e+00
-5.75639904e-01 8.62249792e-01 7.12332726e-01 -8.87869447e-02
2.74616271e-01 1.04435787e-01 -1.61618412e-01 -3.40309918e-01
-7.52774715e-01 -6.32295012e-01 -3.37152518e-02 5.22278488e-01
4.44783628e-01 1.07168205e-01 -1.18881297e+00 8.12158585e-01
-2.12107629e-01 -3.67366761e-01 -1.21394962e-01 1.05417752e+00
-5.52502871e-01 -1.73183930e+00 -1.88510701e-01 1.54771879e-01
-6.44783258e-01 -3.20894301e-01 -2.22720250e-01 8.62794220e-01
4.14773256e-01 1.17422116e+00 -1.85905010e-01 1.29195303e-01
3.93511444e-01 3.38534355e-01 5.40202260e-01 -7.99935639e-01
-4.99364108e-01 -2.01433629e-01 9.98951793e-01 -6.33131087e-01
-7.30482161e-01 -4.48230892e-01 -2.17599130e+00 4.04725730e-01
4.61159162e-02 7.05066800e-01 3.63373488e-01 1.23472905e+00
-4.88252677e-02 5.79483211e-01 4.05436844e-01 -2.77219955e-02
-6.72262073e-01 -5.37599623e-01 -5.95348537e-01 9.49665904e-01
-2.16913596e-01 -1.06809247e+00 -3.45937014e-01 -1.02218166e-01] | [10.21533489227295, 9.275347709655762] |
5af8ab51-6b37-4db5-832b-b8434cf82c18 | read-look-or-listen-what-s-needed-for-solving | 2307.04532 | null | https://arxiv.org/abs/2307.04532v1 | https://arxiv.org/pdf/2307.04532v1.pdf | Read, Look or Listen? What's Needed for Solving a Multimodal Dataset | The prevalence of large-scale multimodal datasets presents unique challenges in assessing dataset quality. We propose a two-step method to analyze multimodal datasets, which leverages a small seed of human annotation to map each multimodal instance to the modalities required to process it. Our method sheds light on the importance of different modalities in datasets, as well as the relationship between them. We apply our approach to TVQA, a video question-answering dataset, and discover that most questions can be answered using a single modality, without a substantial bias towards any specific modality. Moreover, we find that more than 70% of the questions are solvable using several different single-modality strategies, e.g., by either looking at the video or listening to the audio, highlighting the limited integration of multiple modalities in TVQA. We leverage our annotation and analyze the MERLOT Reserve, finding that it struggles with image-based questions compared to text and audio, but also with auditory speaker identification. Based on our observations, we introduce a new test set that necessitates multiple modalities, observing a dramatic drop in model performance. Our methodology provides valuable insights into multimodal datasets and highlights the need for the development of more robust models. | ['Roy Schwartz', 'Yonatan Bitton', 'Netta Madvil'] | 2023-07-06 | null | null | null | null | ['video-question-answering', 'question-answering', 'speaker-identification'] | ['computer-vision', 'natural-language-processing', 'speech'] | [ 3.70960295e-01 2.07224451e-02 9.43848714e-02 -2.82096863e-01
-1.52962697e+00 -1.27248228e+00 7.64011025e-01 1.49301276e-01
-4.91431534e-01 3.77552629e-01 5.41715264e-01 -1.99555084e-01
-2.63742894e-01 -2.81370789e-01 -6.68745637e-01 -4.68817711e-01
3.24542850e-01 5.90330899e-01 2.34584972e-01 -3.13242525e-01
1.02018498e-01 -5.17636091e-02 -1.69957399e+00 6.75238550e-01
3.91679794e-01 1.08112359e+00 -1.10102035e-01 1.00006068e+00
-1.88328251e-01 1.05659854e+00 -6.27221286e-01 -7.58335829e-01
-5.46591617e-02 -4.54579771e-01 -1.15852678e+00 1.52180493e-01
9.03255701e-01 -3.76684874e-01 -2.06987962e-01 5.79996288e-01
6.35637820e-01 -4.38769273e-02 5.86567581e-01 -1.48739469e+00
-2.44531691e-01 4.34427708e-01 -1.60995871e-01 2.66461253e-01
1.07610273e+00 3.00854713e-01 1.32053792e+00 -8.11864495e-01
8.21320951e-01 1.24396884e+00 6.19943082e-01 4.31911081e-01
-1.32777667e+00 -2.97981918e-01 -1.83819965e-01 2.04151154e-01
-1.14367628e+00 -9.29007947e-01 5.03826797e-01 -5.63194513e-01
5.75504184e-01 5.21960318e-01 9.20876935e-02 1.27061188e+00
-4.20694262e-01 8.39498401e-01 1.21050262e+00 -4.57512438e-01
1.30562767e-01 2.51252621e-01 5.20297252e-02 4.49297905e-01
-2.80539840e-01 -3.42833579e-01 -1.00861382e+00 -3.46556783e-01
2.48290449e-01 -4.51653153e-01 -2.81979114e-01 -2.50648260e-01
-1.55278409e+00 5.66909969e-01 -2.04041258e-01 2.81015992e-01
-1.09516606e-01 -1.42147496e-01 4.33224946e-01 5.34406424e-01
-1.64931570e-03 6.31817222e-01 -5.39107323e-01 -5.15921652e-01
-8.85986865e-01 1.05599910e-01 1.05533302e+00 7.25982785e-01
7.34490454e-01 -5.71603239e-01 -2.95415550e-01 8.10917556e-01
2.15638489e-01 5.15944421e-01 1.78175583e-01 -1.38393152e+00
6.26509249e-01 6.54236913e-01 1.48588747e-01 -9.60211456e-01
-4.36632335e-01 1.39521763e-01 -1.48125559e-01 -3.57169926e-01
8.50956619e-01 -4.74360771e-02 -5.24389088e-01 1.73023939e+00
2.91385919e-01 -3.11410338e-01 -6.25263378e-02 8.37598264e-01
1.15838933e+00 2.62438804e-01 1.52427733e-01 -1.51564423e-02
1.60623598e+00 -6.29764259e-01 -5.93851089e-01 3.71901840e-02
4.87796724e-01 -9.40210998e-01 1.42982364e+00 4.47950304e-01
-1.18066406e+00 -3.82072330e-01 -4.70243126e-01 -1.70543998e-01
-2.96807289e-01 -3.66406664e-02 2.97850013e-01 7.98309863e-01
-1.16823125e+00 8.36946815e-02 -4.37392563e-01 -8.18759918e-01
9.99425799e-02 1.76599741e-01 -5.96714437e-01 -2.62864113e-01
-8.57391715e-01 7.70276666e-01 1.15690574e-01 -2.31500208e-01
-9.88713861e-01 -4.47903752e-01 -7.05885291e-01 -1.59893259e-01
6.29983962e-01 -4.92212981e-01 1.30100811e+00 -8.70141208e-01
-1.14772296e+00 9.06706631e-01 -4.63486940e-01 6.05601072e-02
4.25255865e-01 -2.13498604e-02 -5.27695417e-01 8.05000007e-01
3.98433544e-02 8.33885074e-01 7.95207024e-01 -1.40407228e+00
-7.52677560e-01 -3.15095007e-01 4.70115960e-01 2.03206480e-01
-4.72390860e-01 3.06371361e-01 -6.96636200e-01 -1.85374290e-01
6.77251369e-02 -9.22987998e-01 3.55050206e-01 -2.50311255e-01
-2.96876848e-01 -1.66262537e-01 6.50882185e-01 -7.33014822e-01
1.31651831e+00 -2.17336988e+00 4.03277785e-01 1.74623221e-01
4.09621358e-01 -4.32075948e-01 -4.67392415e-01 7.46500552e-01
2.18408242e-01 3.68968487e-01 -1.45362079e-01 -5.47553182e-01
2.17269123e-01 3.01756591e-01 -3.56052577e-01 1.70703098e-01
3.43064845e-01 9.61450458e-01 -8.55680823e-01 -7.77315378e-01
-5.61150163e-02 2.63892025e-01 -4.68486756e-01 2.17584223e-01
-1.88018426e-01 6.20438278e-01 -1.69381872e-02 1.15245450e+00
2.94859827e-01 -3.96558583e-01 2.22209826e-01 -4.92911488e-01
-1.03468141e-02 2.28972122e-01 -9.83879387e-01 1.71144748e+00
-1.07584812e-01 8.45074236e-01 2.75122702e-01 -6.34416401e-01
3.27472508e-01 5.78248441e-01 4.43524301e-01 -7.76106894e-01
7.20459744e-02 2.17406437e-01 -1.30582899e-01 -1.03191304e+00
6.03518903e-01 -2.36845873e-02 -1.64632544e-01 6.98537946e-01
4.38039511e-01 -2.04090148e-01 5.07486641e-01 4.77122426e-01
1.23431802e+00 -1.60738811e-01 -1.86220884e-01 1.86229780e-01
2.97592014e-01 2.33854592e-01 -5.91505915e-02 1.01695740e+00
-3.48921508e-01 8.37982178e-01 6.58263385e-01 -1.49370050e-02
-8.60554993e-01 -1.06321108e+00 -2.24916656e-02 1.65334380e+00
-8.12202618e-02 -5.85671544e-01 -5.69581985e-01 -8.01057935e-01
-1.43731967e-01 2.34840199e-01 -6.62930131e-01 6.01759329e-02
-2.15540349e-01 -6.16401672e-01 9.53656137e-01 1.37956366e-01
5.65593094e-02 -7.67104208e-01 -5.87687373e-01 -1.99457675e-01
-8.22752178e-01 -1.41519034e+00 -1.27297580e-01 -7.89517015e-02
-6.50487006e-01 -1.32540596e+00 -5.29133916e-01 -3.77933621e-01
3.01728517e-01 2.55504727e-01 1.47702897e+00 4.96592745e-02
1.64987147e-01 1.50596797e+00 -6.65978789e-01 -1.85159910e-02
-4.43224967e-01 2.24676922e-01 -2.21355051e-01 6.18406869e-02
3.16108495e-01 -1.61764905e-01 -4.70548898e-01 6.83275342e-01
-1.15081251e+00 -3.31101000e-01 5.97719789e-01 5.49416959e-01
2.21769378e-01 -1.21597305e-01 4.76151019e-01 -4.52190906e-01
7.75667846e-01 -8.17753553e-01 6.22681854e-03 4.87538785e-01
-6.80319145e-02 -1.84641391e-01 2.23166831e-02 -4.84393209e-01
-6.81628406e-01 4.23104875e-02 5.08467667e-02 -1.04365386e-01
-4.11909372e-01 5.93960583e-01 -1.32423311e-01 -2.42394358e-01
6.32537603e-01 3.14491503e-02 9.62470192e-03 -4.15961772e-01
4.36730236e-01 5.28282523e-01 6.00199997e-01 -7.88483799e-01
7.06443369e-01 5.72247684e-01 -2.70818770e-01 -9.09025908e-01
-6.89003766e-01 -5.61536074e-01 -5.15886724e-01 -6.24447167e-01
8.25361371e-01 -8.10173810e-01 -1.09055924e+00 1.14522472e-01
-8.93901289e-01 -1.62729919e-01 -9.94109437e-02 3.74944180e-01
-4.03269678e-01 5.72643876e-01 -5.73406518e-01 -8.73530984e-01
1.87551424e-01 -1.18335140e+00 1.35819566e+00 -3.31591964e-02
-5.59609354e-01 -8.09465051e-01 1.71553358e-01 1.13346839e+00
3.26698989e-01 2.00707242e-02 9.50100958e-01 -8.18841577e-01
-5.38037598e-01 -1.77897006e-01 -3.25863212e-01 5.50524704e-02
-4.26542908e-02 1.11803047e-01 -1.26792920e+00 -2.84925252e-01
-1.31566182e-01 -9.38091040e-01 6.56440556e-01 -1.38565272e-01
7.18411982e-01 -6.02606684e-02 2.04000697e-01 -4.88269180e-02
1.04918921e+00 -9.30001438e-02 4.74669605e-01 3.21643502e-01
6.01976514e-01 9.68156874e-01 3.60836834e-01 2.70726532e-01
9.27133441e-01 4.89528656e-01 4.61278439e-01 1.29059866e-01
8.98945332e-02 -2.14310646e-01 3.69892955e-01 8.30227196e-01
1.29055023e-01 -3.07478279e-01 -1.17882967e+00 6.95711136e-01
-1.62475002e+00 -9.29164350e-01 -1.93193883e-01 2.11973023e+00
8.17502320e-01 -1.64878026e-01 5.29242814e-01 3.05958241e-01
3.57711136e-01 4.34126817e-02 -2.06858546e-01 5.40943146e-02
-4.52509612e-01 3.02674510e-02 -1.23838140e-02 2.93622106e-01
-9.59144950e-01 5.21947980e-01 7.90318203e+00 5.65635920e-01
-9.42119241e-01 1.15930840e-01 4.00502145e-01 -2.42574528e-01
-6.51341677e-01 6.25211652e-03 -3.35624635e-01 3.92759502e-01
1.08773792e+00 3.15310210e-01 5.75175285e-01 2.22868204e-01
-7.88058788e-02 -5.09639144e-01 -1.38461864e+00 9.71314251e-01
3.90499443e-01 -9.62807357e-01 1.58583775e-01 -1.13253206e-01
3.88133496e-01 -7.00521320e-02 8.90200585e-02 2.68698573e-01
-1.63676552e-02 -1.11826909e+00 8.73914599e-01 5.53208768e-01
5.58989942e-01 -3.53467166e-01 5.87838531e-01 1.85291663e-01
-9.46257889e-01 -1.40859067e-01 1.13993295e-01 1.99083947e-02
2.87618712e-02 1.23320714e-01 -8.40706706e-01 5.20588458e-01
9.17639792e-01 2.34497249e-01 -1.12818944e+00 9.14192557e-01
2.74248004e-01 7.55189896e-01 -4.17124420e-01 2.27052644e-01
-1.94831312e-01 2.33544692e-01 4.82983321e-01 1.18135095e+00
2.60165751e-01 -5.56492321e-02 3.64918150e-02 3.99697185e-01
-1.08411819e-01 2.00935051e-01 -4.93350297e-01 -4.27995980e-01
4.81441855e-01 1.29868972e+00 -4.89493877e-01 -2.36359894e-01
-6.42556906e-01 6.82285011e-01 8.84485021e-02 4.51501697e-01
-5.60390770e-01 -2.59096306e-02 1.04497559e-01 -4.57176715e-02
5.01891002e-02 -2.72288322e-01 -8.63328800e-02 -1.27092409e+00
1.76333323e-01 -1.46282625e+00 8.06237161e-01 -1.14596748e+00
-1.31977189e+00 4.38031584e-01 2.07254663e-02 -1.07759619e+00
-1.88590467e-01 -4.06311542e-01 -1.67001918e-01 4.82917726e-01
-1.42652905e+00 -1.10932004e+00 -5.83779693e-01 9.68021274e-01
2.07724378e-01 -7.08435755e-03 6.57450557e-01 5.78188300e-01
-3.47987950e-01 5.81106722e-01 -2.64770597e-01 9.89140477e-03
1.26836801e+00 -1.03089237e+00 -3.04208755e-01 3.82446557e-01
3.10196996e-01 5.88805854e-01 7.34490812e-01 -2.46667385e-01
-1.82220566e+00 -3.56251270e-01 9.40486252e-01 -1.25041306e+00
9.53990221e-01 -2.65480012e-01 -8.46344411e-01 4.92029458e-01
4.95623052e-01 -4.52559978e-01 1.12623060e+00 3.11665088e-01
-6.24373138e-01 1.26463771e-01 -1.11664534e+00 3.74424487e-01
7.63774335e-01 -1.13986921e+00 -7.79448271e-01 7.00688064e-02
6.82040811e-01 -2.41242871e-01 -8.70101690e-01 4.05502975e-01
7.53101170e-01 -1.11978173e+00 7.09778070e-01 -8.00607145e-01
6.54609323e-01 -1.95455492e-01 -6.09021783e-01 -8.77397418e-01
2.32677236e-01 -4.34198707e-01 -4.88548316e-02 1.50422072e+00
6.10202134e-01 -3.37537587e-01 5.51985443e-01 9.05234218e-01
2.18428195e-01 -2.17545986e-01 -1.01060617e+00 -3.32560748e-01
-2.60956109e-01 -7.13441789e-01 3.74153078e-01 1.19649184e+00
9.29008946e-02 3.25750291e-01 -3.02659214e-01 9.27071720e-02
3.27317804e-01 -1.29871026e-01 1.00610709e+00 -1.23286092e+00
-2.61062026e-01 -2.54098713e-01 -3.03154737e-01 -8.15851331e-01
-5.49410395e-02 -5.91902673e-01 -1.15245946e-01 -1.28925812e+00
4.95400876e-01 -4.60859947e-02 -2.17346802e-01 4.25375879e-01
-7.48227313e-02 6.76170707e-01 3.05809379e-01 3.65899831e-01
-1.15467799e+00 1.33084387e-01 9.83468771e-01 -1.33971319e-01
9.46381092e-02 -3.50303322e-01 -8.85170281e-01 6.40644848e-01
4.73995715e-01 -2.66639560e-01 -2.12862238e-01 -6.45972192e-01
8.55743349e-01 8.77370909e-02 6.18018866e-01 -8.73200238e-01
4.02933121e-01 3.04434337e-02 1.63527995e-01 -5.46657622e-01
5.75872064e-01 -1.00226855e+00 -3.55479307e-02 -1.12451769e-01
-6.01590872e-01 3.08336258e-01 2.80459404e-01 4.75548953e-01
-4.18974817e-01 -1.09699316e-01 1.93026528e-01 -1.30850181e-01
-6.75417721e-01 -3.33265632e-01 -4.28049386e-01 4.97327745e-01
5.94578683e-01 -1.19739845e-01 -6.33394122e-01 -8.16048443e-01
-9.09760058e-01 5.03112674e-01 5.39970279e-01 4.69961166e-01
3.41717720e-01 -1.27981055e+00 -5.84245384e-01 -1.59648895e-01
5.23240685e-01 -5.53323328e-01 3.56758416e-01 1.15793812e+00
-2.02100798e-02 3.15892994e-01 -5.01413122e-02 -9.54666674e-01
-1.40417910e+00 1.44686833e-01 1.59672290e-01 1.55767977e-01
2.29306787e-01 6.47209048e-01 -1.29943401e-01 -6.01564348e-01
3.93019617e-01 -3.64873596e-02 -2.68290251e-01 7.49681890e-01
3.89316499e-01 4.63149756e-01 3.31776887e-02 -8.82947981e-01
-4.02144939e-01 5.37128448e-01 3.15447778e-01 -5.68127036e-01
1.03851140e+00 -5.67578852e-01 -3.01004112e-01 7.70657122e-01
1.15717137e+00 3.00548613e-01 -8.12986612e-01 -2.19908237e-01
1.57222658e-01 -4.09979999e-01 -3.62127781e-01 -9.90741551e-01
-7.19126463e-01 7.57179439e-01 5.23421586e-01 6.27537549e-01
1.16763282e+00 4.38153118e-01 5.30281782e-01 5.62752306e-01
2.92754799e-01 -1.12073028e+00 4.25350130e-01 5.15533566e-01
7.67443001e-01 -1.70068109e+00 -2.20461756e-01 -6.87048733e-02
-8.06133449e-01 9.79897678e-01 2.90754110e-01 5.75516284e-01
2.14378119e-01 -5.73574454e-02 3.27859759e-01 -4.15887833e-01
-9.13416207e-01 -3.18892360e-01 5.09502709e-01 3.37877125e-01
4.16603416e-01 -2.26794302e-01 1.65027753e-01 4.54814017e-01
-1.39738038e-01 -1.38001248e-01 5.52945971e-01 1.07898474e+00
-2.30022594e-01 -9.93142188e-01 -6.51564896e-01 3.47385347e-01
-4.56080168e-01 -9.35549289e-02 -9.32847142e-01 7.80239344e-01
-5.53613268e-02 1.51017070e+00 -8.46466646e-02 -5.23612857e-01
4.35941249e-01 5.35319030e-01 5.00561178e-01 -3.89237612e-01
-6.94117844e-01 1.16686992e-01 3.19792569e-01 -7.20806241e-01
-7.97389090e-01 -7.87484229e-01 -5.89671254e-01 -2.04170972e-01
-8.10426176e-02 1.37915343e-01 6.76143348e-01 1.06246412e+00
4.62323695e-01 1.03837423e-01 4.05382425e-01 -6.90291584e-01
-2.21464604e-01 -7.67098665e-01 -9.80128795e-02 6.28906667e-01
5.96221209e-01 -3.98598105e-01 -6.83501303e-01 2.90380865e-01] | [10.694540977478027, 1.5864102840423584] |
c1570595-f9ea-45c6-bf5f-3fe57ad296c4 | semi-supervised-semantic-segmentation-methods | 2212.13486 | null | https://arxiv.org/abs/2212.13486v2 | https://arxiv.org/pdf/2212.13486v2.pdf | Semi-Supervised Semantic Segmentation Methods for UW-OCTA Diabetic Retinopathy Grade Assessment | People with diabetes are more likely to develop diabetic retinopathy (DR) than healthy people. However, DR is the leading cause of blindness. At present, the diagnosis of diabetic retinopathy mainly relies on the experienced clinician to recognize the fine features in color fundus images. This is a time-consuming task. Therefore, in this paper, to promote the development of UW-OCTA DR automatic detection, we propose a novel semi-supervised semantic segmentation method for UW-OCTA DR image grade assessment. This method, first, uses the MAE algorithm to perform semi-supervised pre-training on the UW-OCTA DR grade assessment dataset to mine the supervised information in the UW-OCTA images, thereby alleviating the need for labeled data. Secondly, to more fully mine the lesion features of each region in the UW-OCTA image, this paper constructs a cross-algorithm ensemble DR tissue segmentation algorithm by deploying three algorithms with different visual feature processing strategies. The algorithm contains three sub-algorithms, namely pre-trained MAE, ConvNeXt, and SegFormer. Based on the initials of these three sub-algorithms, the algorithm can be named MCS-DRNet. Finally, we use the MCS-DRNet algorithm as an inspector to check and revise the results of the preliminary evaluation of the DR grade evaluation algorithm. The experimental results show that the mean dice similarity coefficient of MCS-DRNet v1 and v2 are 0.5161 and 0.5544, respectively. The quadratic weighted kappa of the DR grading evaluation is 0.7559. Our code will be released soon. | ['Zeyu Ding', 'Hizmawati Madzin', 'Zhuoyi Tan'] | 2022-12-27 | null | null | null | null | ['semi-supervised-semantic-segmentation'] | ['computer-vision'] | [ 1.23809300e-01 -6.28317799e-03 -7.11049065e-02 -6.43244147e-01
-6.47341669e-01 -3.72302234e-01 -7.68468678e-02 -1.91983998e-01
-4.90155965e-01 6.85170949e-01 2.12680191e-01 -5.10872483e-01
-2.16570616e-01 -7.46344984e-01 5.88143878e-02 -9.10395265e-01
2.39546224e-01 1.39954463e-01 2.75691390e-01 2.37354681e-01
5.10604143e-01 5.35120785e-01 -1.63815999e+00 3.82115692e-01
1.62431026e+00 1.01965296e+00 2.52017260e-01 7.59612083e-01
-1.65818870e-01 7.28033423e-01 -2.06207722e-01 -2.99980551e-01
3.76132578e-01 -7.27528811e-01 -6.96219504e-01 3.93386185e-01
6.19885147e-01 -7.18366563e-01 -1.34364655e-02 1.25067699e+00
8.46182585e-01 -8.98039341e-02 7.88454413e-01 -4.37955618e-01
-4.57505852e-01 2.46640593e-01 -7.78695583e-01 6.18509293e-01
-3.58833149e-02 2.99157083e-01 5.79610825e-01 -5.53051651e-01
4.90063429e-01 9.15909111e-01 4.59698170e-01 3.77480149e-01
-1.02379131e+00 -4.83420432e-01 -2.05790520e-01 5.33457279e-01
-1.25206292e+00 -3.48514020e-01 4.61255819e-01 -8.58891606e-01
3.86937588e-01 1.19492933e-01 8.51588726e-01 1.75053537e-01
1.30458593e-01 7.57952869e-01 1.74150717e+00 -2.54337937e-01
3.35476935e-01 -6.31875619e-02 4.39438701e-01 8.33270967e-01
4.84203368e-01 1.62965044e-01 3.95613968e-01 1.40292451e-01
9.65355814e-01 -2.56138861e-01 -4.38962877e-01 -1.75723702e-01
-6.99978530e-01 7.09224284e-01 3.81199539e-01 2.65983164e-01
-3.43607277e-01 -3.69178921e-01 1.72970638e-01 1.19794644e-01
4.11205828e-01 8.26820135e-02 -1.73535675e-01 1.07306734e-01
-9.31887209e-01 -2.09263042e-01 2.62306154e-01 3.50408137e-01
4.79032546e-01 -3.19219112e-01 -1.80255130e-01 1.03629827e+00
5.41861057e-01 3.47713500e-01 4.88361239e-01 -1.03961837e+00
3.43759470e-02 1.12892735e+00 -8.32792670e-02 -6.20761216e-01
-3.59561235e-01 -5.17818034e-01 -9.83628511e-01 7.61903822e-01
5.60080528e-01 -2.74084777e-01 -1.59343672e+00 8.78603697e-01
2.02857271e-01 -1.46723166e-01 5.21595068e-02 1.24102521e+00
9.97330189e-01 2.64577448e-01 2.10550934e-01 -3.52262944e-01
1.28696978e+00 -7.14960456e-01 -5.89633703e-01 1.54811069e-01
6.18595183e-01 -6.92418456e-01 8.41049850e-01 4.76608276e-01
-1.04866457e+00 -6.37939274e-01 -8.32010925e-01 -1.16193734e-01
5.34449033e-02 8.91698003e-01 6.80981338e-01 6.92627430e-01
-1.21846771e+00 1.75862759e-01 -6.40872657e-01 -5.77653050e-01
7.53867567e-01 2.84036964e-01 -2.00153798e-01 -4.84286457e-01
-6.06116712e-01 9.92655754e-01 1.86245218e-01 3.99448246e-01
-3.66216987e-01 -5.58599174e-01 -4.21153426e-01 -5.18488288e-01
-7.56438300e-02 -9.07959282e-01 8.69486451e-01 -1.02673042e+00
-1.21853220e+00 1.31547701e+00 -3.55701327e-01 -1.35524243e-01
6.30524814e-01 1.48530006e-01 -6.05653703e-01 5.84213614e-01
2.48039868e-02 4.12018538e-01 4.29085016e-01 -1.39423192e+00
-1.21475196e+00 -7.57585406e-01 1.22270584e-02 1.47653535e-01
4.80717391e-01 1.65762559e-01 -2.70969003e-01 -4.50844139e-01
2.79549688e-01 -4.28774029e-01 -3.55919152e-01 2.91219711e-01
-2.85939395e-01 -3.81203413e-01 3.16169888e-01 -8.61388385e-01
1.19266498e+00 -2.17555928e+00 -3.38647813e-01 4.49253201e-01
7.04905629e-01 6.14535689e-01 5.99014647e-02 -5.30298173e-01
-1.13508493e-01 1.06920794e-01 -3.25697690e-01 1.91137627e-01
-5.37959278e-01 4.28664312e-02 3.71271908e-01 5.26662529e-01
4.18382958e-02 5.50384641e-01 -6.75903678e-01 -7.20237017e-01
2.27146924e-01 2.50996113e-01 -2.21145183e-01 2.20743522e-01
2.83055812e-01 4.16979283e-01 -5.74274004e-01 8.06145430e-01
8.98590446e-01 -1.68768674e-01 -1.93678468e-01 -6.20487154e-01
-4.80848849e-01 -2.95411319e-01 -1.07775295e+00 1.21625173e+00
5.61531857e-02 5.55946767e-01 -8.99205580e-02 -8.63053501e-01
1.07014930e+00 1.88745588e-01 4.47169334e-01 -6.66017950e-01
3.34450513e-01 5.29138386e-01 3.67598563e-01 -1.01585817e+00
-1.72974885e-01 -1.27017349e-01 8.54063332e-01 1.15619965e-01
-1.94256827e-01 2.62458682e-01 6.99778378e-01 -1.18374400e-01
6.61274016e-01 5.20147681e-02 4.74095911e-01 -8.54751095e-02
7.88336098e-01 1.61404565e-01 6.53480411e-01 3.21861893e-01
-5.49108803e-01 7.88509607e-01 5.97179472e-01 -4.87411022e-01
-7.82926798e-01 -1.03958964e+00 -6.80863202e-01 9.48745981e-02
2.75695771e-01 2.38949805e-01 -6.62215412e-01 -7.39648759e-01
-3.13501954e-02 5.15056193e-01 -5.92259526e-01 1.88028261e-01
-2.42308587e-01 -1.12688971e+00 2.09652483e-01 4.66424018e-01
9.17848349e-01 -5.23929834e-01 -4.24922466e-01 -6.74693361e-02
1.38441951e-03 -6.16826594e-01 -2.82700807e-01 -4.53630507e-01
-1.02688825e+00 -1.47819591e+00 -1.19864237e+00 -9.88564312e-01
9.87556219e-01 3.51079941e-01 7.32185006e-01 1.38981596e-01
-6.16628230e-01 9.92606208e-02 -3.46600890e-01 -9.15158242e-02
-1.85404122e-01 -4.35867310e-01 -3.89285326e-01 3.52441251e-01
7.25031912e-01 -5.65319896e-01 -1.19706047e+00 3.81151855e-01
-5.07462382e-01 1.52701288e-02 1.24473262e+00 3.70247394e-01
9.11294937e-01 5.04119955e-02 2.83111423e-01 -7.41140664e-01
3.50974232e-01 3.12981457e-02 -6.54829979e-01 3.25853229e-01
-8.46931159e-01 -4.21351045e-01 -5.24587817e-02 -1.51435927e-01
-1.11468315e+00 1.22539043e-01 3.17063890e-02 -5.84723009e-03
-4.96704668e-01 3.93136799e-01 -8.87425914e-02 -2.34117150e-01
5.78092515e-01 1.47573009e-01 3.00703079e-01 -5.67925215e-01
2.48105317e-01 1.18033099e+00 5.50517261e-01 2.00173277e-02
4.48237419e-01 3.81030500e-01 -1.65172577e-01 -7.50299811e-01
-9.13608193e-01 -7.98284173e-01 -4.92770344e-01 -4.32543516e-01
1.40832627e+00 -9.02436256e-01 -4.60016042e-01 7.64864981e-01
-8.73833597e-01 -2.27202103e-01 -1.15292380e-02 1.02183473e+00
-3.36683631e-01 5.87322533e-01 -3.45459759e-01 -5.83578587e-01
-4.66461092e-01 -1.20273805e+00 5.05266786e-01 7.17185557e-01
1.19979739e-01 -8.59909058e-01 1.39536306e-01 8.55467618e-01
2.70341218e-01 2.42413133e-01 1.17611659e+00 -2.23112985e-01
-4.38523173e-01 -8.40589628e-02 -9.67986524e-01 7.86652386e-01
2.81012744e-01 3.52876544e-01 -7.48243272e-01 2.91368384e-02
-2.27268592e-01 8.21593851e-02 1.10450161e+00 9.83626664e-01
1.03165960e+00 1.31142020e-01 -1.16040461e-01 7.89343536e-01
1.64744937e+00 6.21703088e-01 1.05224085e+00 3.02744061e-01
5.15132129e-01 6.50628090e-01 6.23379052e-01 1.59685820e-01
5.87363958e-01 1.32490546e-01 2.77292907e-01 -6.55240774e-01
-5.58586895e-01 2.52981663e-01 -1.79204531e-02 4.49789405e-01
-7.87335932e-01 3.61433402e-02 -8.81110847e-01 6.76316082e-01
-1.42319214e+00 -7.71395206e-01 -7.67325938e-01 2.00237989e+00
7.90007353e-01 -9.57139358e-02 1.25754207e-01 -1.04954466e-01
8.72091591e-01 -3.20273310e-01 -4.99595135e-01 -2.34385297e-01
-1.96575075e-01 2.71602660e-01 4.98304367e-01 3.14107031e-01
-1.03163004e+00 5.05183995e-01 5.79506016e+00 4.81644899e-01
-1.08185327e+00 -1.58810675e-01 7.97061205e-01 1.62977904e-01
3.63883283e-03 -6.32288679e-02 -7.26023257e-01 4.94721681e-01
2.71005809e-01 -2.29042992e-02 2.05200449e-01 3.92595917e-01
7.97237813e-01 -7.17782795e-01 -7.10463524e-01 1.05583954e+00
9.22262520e-02 -9.06206369e-01 2.24451169e-01 1.33684918e-01
7.29827344e-01 -1.03101417e-01 -7.72827864e-02 -8.66182819e-02
1.38896808e-01 -1.04526055e+00 -1.23834379e-01 1.03816378e+00
1.02196538e+00 -6.28683746e-01 1.32018638e+00 -1.81335986e-01
-7.69852638e-01 -2.40550954e-02 -4.64659989e-01 3.21848154e-01
1.41755817e-02 9.96464252e-01 -3.91015977e-01 4.31896329e-01
6.21084809e-01 7.99531341e-01 -7.64402449e-01 1.82684100e+00
-3.77727449e-01 5.99413991e-01 1.51003942e-01 3.87754232e-01
2.17528138e-02 -6.54486239e-01 5.13138831e-01 8.40272605e-01
4.25946295e-01 4.88094300e-01 -4.08521183e-02 8.45362663e-01
4.04541582e-01 3.73755991e-01 -9.49436948e-02 1.67985991e-01
-1.45154983e-01 1.25474572e+00 -5.23725629e-01 -2.97568500e-01
-4.71254468e-01 5.76796055e-01 -3.44944715e-01 6.27793729e-01
-3.40944529e-01 -5.92595279e-01 4.35975671e-01 2.46829167e-01
1.26309559e-01 2.71417294e-02 -6.55068696e-01 -8.57259572e-01
-6.26171678e-02 -7.21230209e-01 4.47924882e-01 -1.05269563e+00
-1.34725058e+00 5.14181077e-01 -4.13181454e-01 -1.29070592e+00
2.15318754e-01 -6.40701592e-01 -7.13768244e-01 1.22805679e+00
-1.85241592e+00 -1.00686026e+00 -7.55297542e-01 5.52202463e-01
1.20015621e-01 -3.85028660e-01 4.59044099e-01 2.91624069e-01
-8.03840518e-01 3.08634937e-01 -4.95245419e-02 3.00202578e-01
7.91852117e-01 -1.43946779e+00 -5.27096450e-01 9.42147493e-01
-7.46177435e-01 5.28865993e-01 3.03416103e-01 -7.14744806e-01
-5.21840036e-01 -1.18016303e+00 7.47834444e-01 7.40631223e-02
3.52009714e-01 6.46277964e-01 -7.75430620e-01 2.95290291e-01
-7.34022409e-02 1.38537541e-01 8.51603508e-01 -2.90132791e-01
5.57313226e-02 -4.11664814e-01 -1.53987837e+00 5.25222480e-01
8.10626268e-01 -1.44696444e-01 -7.84757018e-01 -3.28445323e-02
1.15438342e-01 -1.75614893e-01 -1.19883442e+00 6.21120393e-01
6.46704197e-01 -1.10666656e+00 5.71173429e-01 -2.42254525e-01
5.43594003e-01 -7.36883163e-01 9.50582884e-03 -9.95285869e-01
-2.21435830e-01 -2.04188973e-01 3.04283738e-01 1.10100019e+00
4.26384032e-01 -8.62035155e-01 5.76865196e-01 4.43515092e-01
-2.18841761e-01 -6.57186091e-01 -6.58262074e-01 -2.29257777e-01
-7.39571005e-02 3.02612353e-02 1.03959531e-01 7.99011528e-01
-5.33080757e-01 -7.17873313e-03 4.17106301e-01 3.65120083e-01
1.00069344e+00 2.07729146e-01 4.14683342e-01 -1.35750842e+00
3.24392505e-02 -6.40212238e-01 -7.13677526e-01 -8.05226147e-01
-5.13318419e-01 -8.85963142e-01 -2.69369990e-01 -2.26616764e+00
4.24591988e-01 -3.54604155e-01 -3.66189212e-01 4.20759529e-01
-2.63110340e-01 4.30547655e-01 -3.63873929e-01 2.57341117e-01
-1.20895974e-01 1.70351177e-01 1.82863390e+00 -4.84171137e-02
-5.57872832e-01 3.93004537e-01 -8.92280281e-01 7.77690887e-01
1.02179110e+00 -2.18222998e-02 -5.07191300e-01 -1.93229869e-01
-7.79559761e-02 -2.19288602e-01 4.65921700e-01 -8.84547293e-01
2.13776845e-02 -1.84280291e-01 5.11631906e-01 -8.62207353e-01
-4.20184821e-01 -4.55736309e-01 -2.82591611e-01 3.75643820e-01
-1.24001041e-01 -5.90308845e-01 -1.43917054e-01 2.05329075e-01
-3.26990366e-01 -2.02774450e-01 1.27201772e+00 -2.33078897e-01
-8.57414544e-01 3.40795904e-01 -5.55673420e-01 -5.16959876e-02
1.04222429e+00 -7.63851583e-01 -5.55597007e-01 3.74984257e-02
-9.78814125e-01 4.22759175e-01 3.86158675e-01 -2.14733034e-01
7.28216052e-01 -7.99495220e-01 -1.00521350e+00 9.12976824e-03
2.04735875e-01 1.44105077e-01 5.50407469e-01 1.47650361e+00
-8.33778441e-01 1.94597840e-01 -3.86808515e-01 -5.82786679e-01
-1.52027607e+00 1.77394018e-01 7.08121717e-01 5.18956259e-02
-6.73410892e-01 6.03069544e-01 1.67809576e-01 1.64290980e-01
5.33069558e-02 -5.22189200e-01 -8.63127232e-01 2.45659292e-01
7.48356819e-01 6.22988403e-01 1.39938192e-02 -4.23538297e-01
1.48042506e-02 1.14262962e+00 -3.25650901e-01 2.66674966e-01
1.14895427e+00 -3.80872488e-01 -6.59441829e-01 7.06537142e-02
8.56361806e-01 -8.04718211e-03 -9.36644852e-01 -1.95515249e-02
-1.66555554e-01 -4.54539448e-01 6.96251452e-01 -1.24706638e+00
-1.33438158e+00 7.45994806e-01 1.46497345e+00 -8.94795805e-02
1.40205276e+00 -1.68457359e-01 5.88151038e-01 -1.50945097e-01
2.09812090e-01 -8.23466480e-01 -5.11942625e-01 -4.84871045e-02
5.50180137e-01 -1.06559408e+00 4.60940413e-02 -5.98612666e-01
-6.55800939e-01 1.24304342e+00 5.33597887e-01 -1.02182679e-01
4.82887208e-01 -2.75613129e-01 6.66189849e-01 -6.09177798e-02
-1.50727332e-01 -7.63986886e-01 6.85851514e-01 9.34958816e-01
4.04722810e-01 5.91639429e-02 -1.11099613e+00 4.09340292e-01
1.08677588e-01 5.83839357e-01 6.43374681e-01 4.04162079e-01
-8.46619189e-01 -8.98974657e-01 -6.96825841e-03 8.66370618e-01
-4.20113325e-01 1.04165211e-01 -3.46610695e-01 5.86978972e-01
4.84923244e-01 9.90560234e-01 1.01209655e-01 -1.18642934e-01
3.20027262e-01 -2.64656335e-01 3.76011997e-01 -3.36930364e-01
-1.81883201e-01 1.63583934e-01 1.93252265e-01 -5.75314105e-01
-8.36644530e-01 -4.47666854e-01 -1.26845944e+00 -3.81558761e-02
7.35117272e-02 -2.55905658e-01 5.25896251e-01 8.67063761e-01
1.28977627e-01 3.22792113e-01 6.44416034e-01 -1.98990002e-01
-2.93571651e-01 -9.19477940e-01 -9.11919057e-01 -3.97729352e-02
3.49577814e-01 -4.79966938e-01 -4.95583385e-01 1.71456575e-01] | [15.833372116088867, -3.997908115386963] |
ff243df3-2f66-4386-8cbe-8af3e8374379 | linear-stochastic-bandits-over-a-bit | 2203.01198 | null | https://arxiv.org/abs/2203.01198v1 | https://arxiv.org/pdf/2203.01198v1.pdf | Linear Stochastic Bandits over a Bit-Constrained Channel | One of the primary challenges in large-scale distributed learning stems from stringent communication constraints. While several recent works address this challenge for static optimization problems, sequential decision-making under uncertainty has remained much less explored in this regard. Motivated by this gap, we introduce a new linear stochastic bandit formulation over a bit-constrained channel. Specifically, in our setup, an agent interacting with an environment transmits encoded estimates of an unknown model parameter to a server over a communication channel of finite capacity. The goal of the server is to take actions based on these estimates to minimize cumulative regret. To this end, we develop a novel and general algorithmic framework that hinges on two main components: (i) an adaptive encoding mechanism that exploits statistical concentration bounds, and (ii) a decision-making principle based on confidence sets that account for encoding errors. As our main result, we prove that when the unknown model is $d$-dimensional, a channel capacity of $O(d)$ bits suffices to achieve order-optimal regret. To demonstrate the generality of our approach, we then show that the same result continues to hold for non-linear observation models satisfying standard regularity conditions. Finally, we establish that for the simpler unstructured multi-armed bandit problem, $1$ bit channel-capacity is sufficient for achieving optimal regret bounds. Overall, our work takes a significant first step towards paving the way for statistical decision-making over finite-capacity channels. | ['George J. Pappas', 'Hamed Hassani', 'Aritra Mitra'] | 2022-03-02 | null | null | null | null | ['decision-making-under-uncertainty', 'decision-making-under-uncertainty'] | ['medical', 'reasoning'] | [ 3.21416527e-01 3.69153291e-01 -4.40541893e-01 -2.55466938e-01
-1.35700345e+00 -7.08227992e-01 -1.74534082e-01 3.88829172e-01
-4.59665626e-01 1.12440705e+00 -2.08915770e-02 -5.26041687e-01
-5.63357472e-01 -7.25575209e-01 -1.14452386e+00 -1.07509840e+00
-3.96052659e-01 5.14096022e-01 -4.37802672e-01 2.78656334e-01
2.03001723e-01 3.09119940e-01 -7.82620251e-01 -2.29156017e-01
6.22105956e-01 1.56832874e+00 1.83303148e-01 8.59381258e-01
4.87822257e-02 6.11440361e-01 -3.51952672e-01 -4.13482785e-01
5.05837440e-01 -6.28403664e-01 -6.80399239e-01 3.87439281e-01
-1.73620209e-01 -4.64277446e-01 -3.31945866e-01 1.33396435e+00
5.66679418e-01 -1.23606913e-01 5.28700709e-01 -8.67867827e-01
-1.05248116e-01 1.14262033e+00 -5.24265230e-01 -1.03314325e-01
5.57148345e-02 -2.08934873e-01 1.18965590e+00 2.48098504e-02
2.74578363e-01 1.01543367e+00 5.82951427e-01 3.51269692e-01
-1.17873800e+00 -6.59939587e-01 3.69721860e-01 -1.41374275e-01
-1.09903967e+00 -4.45384026e-01 7.02525198e-01 -2.46523097e-01
3.47218215e-01 2.56206065e-01 5.99729478e-01 7.20051765e-01
6.19051680e-02 1.00924933e+00 1.20673490e+00 -6.92890704e-01
7.59188235e-01 6.27108589e-02 1.61688715e-01 5.95769942e-01
6.38448894e-01 8.27510357e-02 -5.49485683e-01 -3.57841372e-01
5.99854767e-01 7.78663903e-02 -3.82176727e-01 -5.60109496e-01
-6.90226674e-01 1.15085149e+00 8.97239074e-02 1.06147438e-01
-4.26166564e-01 5.65144777e-01 3.15162428e-02 6.25171959e-01
6.33700132e-01 1.40127808e-01 -4.39618617e-01 -2.95327514e-01
-7.21942782e-01 2.18778715e-01 1.17749679e+00 9.84517813e-01
5.85420668e-01 -1.82849109e-01 -1.30391419e-01 4.98405486e-01
3.36555153e-01 9.28783596e-01 -3.09171617e-01 -9.42699969e-01
9.99820650e-01 -2.70110965e-01 5.95743001e-01 -7.78361917e-01
-1.95047021e-01 -9.26552355e-01 -7.97006011e-01 -2.79655978e-02
5.96464694e-01 -7.84191310e-01 -5.99575996e-01 1.92501104e+00
2.01089874e-01 -1.86566010e-01 2.64015794e-01 8.43158543e-01
-4.43662673e-01 5.78838408e-01 -5.46519876e-01 -7.47889042e-01
7.70265281e-01 -7.11131752e-01 -6.98856235e-01 -2.69231588e-01
7.58587420e-01 -2.84753859e-01 2.90144235e-01 5.56506157e-01
-1.38468134e+00 4.38243896e-01 -1.01685786e+00 5.82357526e-01
2.98535436e-01 -3.00199896e-01 9.77326393e-01 1.06621182e+00
-7.65621126e-01 3.33735377e-01 -7.82826662e-01 -1.50165021e-01
4.40435976e-01 2.83741385e-01 4.09017019e-02 -4.92921382e-01
-7.37870038e-01 2.65130252e-01 1.78116560e-01 2.12821245e-01
-9.80415881e-01 -2.68127382e-01 -5.89200556e-01 1.41917706e-01
7.86895692e-01 -7.95778155e-01 1.54885352e+00 -8.47646713e-01
-1.58394468e+00 2.48945236e-01 -2.68863797e-01 -8.90485287e-01
8.40404332e-01 -1.21645950e-01 4.01364833e-01 1.61166862e-01
-9.16085541e-02 -4.12318110e-01 6.93493366e-01 -1.24846029e+00
-7.93161035e-01 -7.53780365e-01 3.71458173e-01 2.53984749e-01
-2.82056928e-01 -1.85172096e-01 -4.44422156e-01 -4.12031919e-01
1.65317699e-01 -1.10764229e+00 -6.64233387e-01 -1.04896016e-01
-6.60907984e-01 2.42643014e-01 -2.61768773e-02 -2.08335951e-01
1.15746415e+00 -1.92111611e+00 2.35034645e-01 6.37766540e-01
4.99496162e-02 -3.58719647e-01 1.35093376e-01 6.61577463e-01
5.04808545e-01 1.64976537e-01 -2.93467611e-01 -7.94353843e-01
2.68537998e-01 1.02545120e-01 -3.84411842e-01 6.56362712e-01
-5.87308407e-01 6.23280942e-01 -5.87153614e-01 -9.15467292e-02
-1.68768734e-01 -5.82194328e-02 -7.35031843e-01 1.21771179e-01
-4.03652042e-01 3.04463923e-01 -1.05318260e+00 5.80138505e-01
6.86099887e-01 -5.02262414e-01 4.44379300e-01 4.27867055e-01
6.49602190e-02 4.62647602e-02 -1.41927135e+00 1.68098629e+00
-6.15120113e-01 1.70945808e-01 9.38477218e-01 -1.52008808e+00
2.10359529e-01 2.11845070e-01 4.94742572e-01 -3.90988678e-01
2.95585126e-01 5.29751897e-01 -3.29026192e-01 -3.46875876e-01
1.10314246e-02 -4.13607091e-01 -3.28637928e-01 7.55608439e-01
-3.31690252e-01 3.96298133e-02 -2.24418744e-01 2.12625772e-01
1.45561087e+00 -6.73840106e-01 1.69144109e-01 4.26996425e-02
-7.44238719e-02 -3.23210418e-01 5.45064211e-01 1.48044884e+00
2.17173640e-02 1.52816102e-01 6.31237447e-01 -3.36292833e-02
-8.21244895e-01 -7.94045389e-01 -4.53228457e-03 7.89934278e-01
8.99742320e-02 -2.45778244e-02 -6.37277424e-01 -5.08750737e-01
4.37607437e-01 5.88989973e-01 -8.43814433e-01 3.48455220e-01
1.25027895e-01 -7.60530114e-01 1.36312246e-01 1.60070851e-01
4.83901501e-01 -2.03163207e-01 -3.13209295e-01 4.99526441e-01
6.53751418e-02 -9.26901400e-01 -3.73961180e-01 6.48435593e-01
-8.35003793e-01 -8.18477750e-01 -9.39006388e-01 -1.85198575e-01
6.68201327e-01 4.54308063e-01 5.69503367e-01 -3.11287045e-01
1.46806300e-01 8.92818093e-01 -3.55756015e-01 -7.33614445e-01
5.04850335e-02 3.12661305e-02 -1.08197160e-01 1.00234620e-01
-1.43930867e-01 -4.77033168e-01 -7.15698540e-01 -6.96401671e-03
-9.03969705e-01 -1.05333842e-01 8.75802696e-01 6.35294616e-01
5.59439421e-01 3.01770866e-01 3.10459256e-01 -9.53031778e-01
5.51550508e-01 -6.62631869e-01 -1.05572569e+00 3.05276424e-01
-5.93468308e-01 2.72730142e-01 5.57600915e-01 -8.07874128e-02
-7.85004437e-01 1.82597771e-01 1.06018461e-01 -4.79795784e-02
3.03611606e-01 8.65911007e-01 -1.54452935e-01 -2.41131708e-01
2.42972359e-01 2.65094042e-01 -7.42702708e-02 -3.68521214e-01
2.01483026e-01 7.75913954e-01 -9.82836559e-02 -9.29477334e-01
7.01564729e-01 6.44915581e-01 2.00693056e-01 -5.72472751e-01
-1.21920907e+00 -2.07918644e-01 -4.61050235e-02 5.77091984e-02
3.70517999e-01 -9.09680545e-01 -1.05032027e+00 3.75379026e-01
-1.01412511e+00 -6.40692234e-01 -1.45969093e-01 7.22336709e-01
-8.92020524e-01 2.74384528e-01 -4.82243061e-01 -1.76440895e+00
-1.55459642e-01 -8.57363343e-01 7.44328737e-01 -4.25651148e-02
5.70648730e-01 -1.00683033e+00 1.24223270e-01 5.27941287e-01
5.34513772e-01 6.22551367e-02 5.33378601e-01 -5.44936061e-01
-1.00626957e+00 -2.99715191e-01 -7.23680332e-02 1.77615926e-01
-2.24668369e-01 -7.47007489e-01 -5.41463256e-01 -4.40080881e-01
1.60077021e-01 -4.94112402e-01 9.97272074e-01 7.80475914e-01
1.37406540e+00 -6.36742771e-01 -4.29126829e-01 6.28494143e-01
1.54445434e+00 2.57925659e-01 6.71249256e-02 3.71258229e-01
1.25843734e-01 2.60659397e-01 4.26513880e-01 1.14170063e+00
4.24574047e-01 5.12913644e-01 6.23386860e-01 2.79368252e-01
6.41586304e-01 -1.41066611e-01 1.38114721e-01 3.31246614e-01
1.14554808e-01 -6.43873692e-01 -6.06482208e-01 5.36032379e-01
-2.03353357e+00 -8.78903568e-01 3.57118338e-01 2.74904728e+00
9.22790766e-01 1.63288966e-01 1.25682980e-01 1.01438388e-01
6.22609735e-01 3.05588488e-02 -9.43891287e-01 -2.80288815e-01
-1.49105504e-01 -5.10694385e-02 1.28046477e+00 7.02973902e-01
-8.99708092e-01 4.45117652e-01 5.88520241e+00 9.21882868e-01
-9.72190738e-01 2.13181973e-01 9.16204154e-01 -4.37462568e-01
-4.76284385e-01 -1.00021139e-01 -8.73010814e-01 4.93673384e-01
9.00344908e-01 -2.11678550e-01 8.21464658e-01 9.00628984e-01
3.24015081e-01 -4.09775883e-01 -9.87821519e-01 1.11051345e+00
-1.32608056e-01 -1.30592084e+00 -4.72237915e-01 7.09492505e-01
8.76584291e-01 3.37347798e-02 2.11755082e-01 -1.37864649e-01
7.94057131e-01 -9.25842047e-01 8.01771760e-01 3.69881690e-01
7.01646209e-01 -9.46005464e-01 7.79230714e-01 4.89922374e-01
-7.31566668e-01 -4.69605446e-01 -3.87104094e-01 -3.89368087e-01
3.16502899e-01 1.19081533e+00 -4.57228988e-01 6.09166026e-01
2.64982194e-01 1.91015154e-01 4.45139259e-01 1.34957230e+00
-2.43516490e-02 8.28288317e-01 -7.40259588e-01 -2.99301505e-01
4.85673040e-01 -1.60089016e-01 3.72950286e-01 1.13367891e+00
5.79441786e-01 3.88308614e-01 2.22124860e-01 2.49304801e-01
-2.68659323e-01 1.38265833e-01 -5.74502945e-01 -2.02999875e-01
6.56543612e-01 7.15144575e-01 -5.05350292e-01 -7.57707208e-02
-3.78920496e-01 9.15393531e-01 4.73513782e-01 5.10030329e-01
-6.27481699e-01 -1.94998622e-01 6.88762665e-01 -3.76582712e-01
7.47748315e-01 -4.51334536e-01 -5.79431176e-01 -1.03025270e+00
3.16758305e-01 -5.50329387e-01 3.37870955e-01 -1.74427386e-02
-1.25055301e+00 -9.39869285e-02 -4.23565120e-01 -6.56324148e-01
-2.11779267e-01 -2.65559793e-01 -2.04758406e-01 7.21312106e-01
-1.70359361e+00 -7.70250618e-01 2.14028880e-01 6.06463850e-01
2.70980626e-01 2.34903187e-01 7.90570021e-01 -7.36114234e-02
-6.53495789e-01 6.47986472e-01 1.00609827e+00 -2.44124187e-03
1.89260408e-01 -1.23402154e+00 -3.74327898e-01 7.28998601e-01
3.64387743e-02 4.24228877e-01 7.93867230e-01 -4.21911955e-01
-2.05336833e+00 -8.53023171e-01 4.44860607e-01 6.35232702e-02
8.43998253e-01 -3.54888022e-01 -1.23682499e-01 7.04364657e-01
-1.65993422e-01 3.87352481e-02 9.78669763e-01 3.36733669e-01
-2.74223626e-01 -2.99283534e-01 -1.13934767e+00 4.91341799e-02
1.03115845e+00 -2.47847989e-01 1.48941755e-01 6.86774790e-01
4.27257746e-01 -4.58211750e-01 -6.32536471e-01 -2.32000444e-02
6.63827598e-01 -7.82262981e-01 4.76367712e-01 -5.89588404e-01
1.94692552e-01 2.02248499e-01 -6.90608621e-01 -1.31180179e+00
3.37638403e-03 -1.30032349e+00 -1.77821085e-01 7.72772431e-01
6.78870261e-01 -8.14112306e-01 1.06084192e+00 9.05150831e-01
1.80426121e-01 -8.02809119e-01 -1.23056626e+00 -8.65451217e-01
2.44155094e-01 -9.38907981e-01 2.77415246e-01 4.37665910e-01
1.40207872e-01 1.01944499e-01 -6.99810565e-01 4.56049979e-01
1.01083958e+00 2.82466143e-01 7.55405128e-01 -8.58896136e-01
-9.76948321e-01 -1.49618745e-01 4.94044311e-02 -1.86464119e+00
1.93557329e-02 -4.03450251e-01 2.23072112e-01 -1.35692286e+00
5.00188172e-01 -9.67141569e-01 -3.67010593e-01 1.57426611e-01
2.02363595e-01 -1.19362518e-01 4.05250609e-01 8.72229114e-02
-1.13962424e+00 4.67660278e-01 1.04294121e+00 -1.13333620e-01
-9.77198333e-02 7.72106647e-01 -1.11455703e+00 4.08942193e-01
6.34153724e-01 -4.50097322e-01 -3.36794049e-01 -6.33646131e-01
5.27558923e-01 8.26055706e-01 -2.36285459e-02 -7.29729295e-01
3.80786300e-01 -3.47199589e-01 -1.19348414e-01 -2.94477940e-01
3.49416852e-01 -1.03992832e+00 -8.12532082e-02 6.22502148e-01
-6.42574668e-01 -6.50642633e-01 -2.48866126e-01 1.27089286e+00
2.01348141e-01 -4.27328318e-01 5.52531004e-01 1.43158743e-02
1.11633539e-01 4.68480915e-01 -3.59716028e-01 3.45677458e-04
1.11286795e+00 1.65438041e-01 -2.02948488e-02 -1.10736227e+00
-7.00353265e-01 3.89493078e-01 2.76027769e-01 -4.28304195e-01
2.53462851e-01 -8.81783485e-01 -6.64513648e-01 -1.73516661e-01
-1.74249426e-01 1.80183980e-03 2.81105429e-01 8.87988508e-01
-2.61434931e-02 9.27782953e-01 6.84617281e-01 -3.39618057e-01
-9.96924460e-01 4.56169575e-01 1.83000386e-01 -3.91316056e-01
-9.27650481e-02 1.08081079e+00 -3.12925652e-02 1.99553102e-01
6.67155683e-01 -3.20655644e-01 5.46426833e-01 -1.99232802e-01
6.29739642e-01 1.39700145e-01 3.49066556e-02 6.04292080e-02
5.32750860e-02 3.28162491e-01 -1.51915967e-01 -6.83206916e-01
1.47601068e+00 -6.56669557e-01 2.49784514e-01 4.56933707e-01
1.05622494e+00 4.39806670e-01 -1.56235313e+00 -7.30170190e-01
-1.17835827e-01 -5.58692634e-01 2.77533919e-01 -8.72403860e-01
-1.02441835e+00 6.07371032e-01 3.20936322e-01 5.72556853e-01
1.06785870e+00 1.71908692e-01 6.43276930e-01 7.34377265e-01
1.09323978e+00 -1.01237118e+00 -3.09175193e-01 5.01727700e-01
4.61723983e-01 -1.26176834e+00 -2.95393467e-01 -2.07793698e-01
-2.04052895e-01 1.07027090e+00 -3.11244279e-01 2.25817338e-01
6.05506003e-01 3.40392590e-01 -3.85119289e-01 1.22917593e-01
-7.76379049e-01 -4.46595609e-01 -3.60389143e-01 3.33177090e-01
2.23014489e-01 3.32178384e-01 -3.33891422e-01 7.44571209e-01
-6.07760064e-02 8.02149698e-02 4.17253852e-01 9.55292165e-01
-9.16831195e-01 -1.09237432e+00 -5.48464417e-01 6.12022042e-01
-9.36305285e-01 -1.69237211e-01 -1.79456636e-01 3.01337212e-01
-4.63469505e-01 1.21368456e+00 -2.23333463e-01 5.07888570e-02
-1.07892357e-01 -2.61933148e-01 6.13680840e-01 -5.49400091e-01
1.05482459e-01 2.77222663e-01 1.45462900e-01 -5.42645514e-01
-2.86887527e-01 -8.02100539e-01 -7.92325675e-01 -4.32502419e-01
-4.38896835e-01 6.72047496e-01 1.09926367e+00 1.15792918e+00
2.71895826e-01 1.25501320e-01 1.00645888e+00 -5.32297432e-01
-1.22269762e+00 -7.63752520e-01 -1.08410037e+00 -2.83883065e-01
5.79948187e-01 -3.37375641e-01 -6.50324404e-01 -4.60315704e-01] | [4.605367660522461, 3.3791158199310303] |
e984dfd6-d6c9-4d3e-9e87-e76a09edbb0b | meme-generating-rnn-model-explanations-via-1 | null | null | https://openreview.net/forum?id=0beaSUVK_n4 | https://openreview.net/pdf?id=0beaSUVK_n4 | MEME: Generating RNN Model Explanations via Model Extraction | Recurrent Neural Networks (RNNs) have achieved remarkable performance on a range of tasks. A key step to further empowering RNN-based approaches is improving their explainability and interpretability. In this work we present MEME: a model extraction approach capable of approximating RNNs with interpretable models represented by human-understandable concepts and their interactions. We demonstrate how MEME can be applied to two multivariate, continuous data case studies: Room Occupation Prediction, and In-Hospital Mortality Prediction. Using these case-studies, we show how our extracted models can be used to interpret RNNs both locally and globally, by approximating RNN decision-making via interpretable concept interactions. | ['Anonymous'] | 2020-10-15 | null | null | null | neurips-workshop-hamlets-2020-12 | ['occupation-prediction'] | ['natural-language-processing'] | [ 2.42962018e-01 1.01087260e+00 -3.44056077e-02 -7.20934451e-01
-2.17408583e-01 -2.91366894e-02 3.86349648e-01 8.09989274e-02
9.98057500e-02 9.10423696e-01 9.47937131e-01 -7.51994908e-01
-3.38167965e-01 -7.14746714e-01 -5.35219491e-01 1.18040890e-01
1.76957443e-01 9.64173794e-01 -9.19418275e-01 -1.41752958e-01
-2.17519969e-01 4.79285240e-01 -1.39957106e+00 8.80004883e-01
6.43648684e-01 5.68134725e-01 -1.78745881e-01 8.29188764e-01
-2.93008894e-01 1.99797153e+00 -8.06316376e-01 -2.68793106e-01
-3.49746287e-01 -5.36664665e-01 -1.09276688e+00 -4.43427593e-01
-1.39561504e-01 -1.79658487e-01 -4.78816688e-01 3.71632397e-01
1.65354479e-02 3.72005165e-01 1.07661366e+00 -1.03296900e+00
-9.89682138e-01 1.16190326e+00 4.29633647e-01 -1.02529414e-01
7.13696852e-02 -1.35133237e-01 1.05521560e+00 -7.84942865e-01
3.61539155e-01 1.61286163e+00 9.65493321e-01 1.00188076e+00
-1.49658823e+00 -3.74974698e-01 1.29569501e-01 2.69521754e-02
-1.09942985e+00 -1.90733582e-01 4.85135376e-01 -2.80570984e-01
1.59267259e+00 6.94330692e-01 7.14654624e-01 1.58107233e+00
3.69850755e-01 7.62606800e-01 8.51141751e-01 -3.41246754e-01
2.79507756e-01 -3.35199460e-02 4.67525512e-01 7.11241245e-01
1.83882281e-01 1.14259169e-01 -4.43053097e-01 -1.72453552e-01
9.49088097e-01 8.64349842e-01 2.19549108e-02 6.37634337e-01
-1.24409807e+00 7.06037641e-01 8.10339272e-01 2.97510356e-01
-7.75981128e-01 8.53952467e-01 3.15435559e-01 3.56703401e-02
8.22806358e-01 1.06470895e+00 -5.51101744e-01 -1.77434027e-01
-8.96500051e-01 -5.67705631e-02 1.01858985e+00 9.74277496e-01
3.73073637e-01 3.04911971e-01 -3.63427550e-01 7.23088503e-01
2.31934041e-01 4.47994262e-01 5.82038164e-01 -8.64669681e-01
2.07882822e-01 8.63214076e-01 -1.00059792e-01 -9.55708206e-01
-9.47729409e-01 -4.81522381e-01 -1.54580772e+00 -3.06876630e-01
-2.83110470e-01 -9.27645937e-02 -1.10944259e+00 1.57845259e+00
-6.15916789e-01 2.65059203e-01 4.02832747e-01 3.67647260e-01
9.93172526e-01 7.54338741e-01 4.54183370e-01 7.54563510e-02
1.39515436e+00 -1.18714249e+00 -1.05817389e+00 -4.16253686e-01
9.07699108e-01 -4.78105322e-02 7.04376280e-01 1.37015328e-01
-9.53148186e-01 -6.97895706e-01 -6.41635537e-01 -2.40452126e-01
-4.80008751e-01 1.54341221e-01 6.25636280e-01 -1.26647791e-02
-1.24679875e+00 7.36545086e-01 -1.12876678e+00 -2.60156780e-01
6.08210742e-01 5.63491464e-01 -8.36558491e-02 1.38288110e-01
-1.24484873e+00 1.26766634e+00 4.55605268e-01 4.99760866e-01
-7.24026859e-01 -7.07285285e-01 -8.96267593e-01 4.66217667e-01
-4.74176668e-02 -1.53587294e+00 1.37520182e+00 -9.48549807e-01
-9.36077058e-01 3.05940509e-01 -5.46685100e-01 -1.07418633e+00
1.31974384e-01 -3.24984372e-01 -6.61228240e-01 -2.79449612e-01
-2.51360208e-01 7.47888207e-01 4.30415362e-01 -1.26027286e+00
-2.22207323e-01 -7.94328526e-02 3.93261341e-03 1.05232306e-01
-3.86075526e-01 -2.36820802e-01 5.27035668e-02 -9.01042461e-01
2.76873633e-02 -8.25067103e-01 -5.45122504e-01 -3.50858361e-01
-1.00433934e+00 -1.57140717e-01 4.72484052e-01 -7.81511486e-01
1.54508531e+00 -1.62261760e+00 3.56422782e-01 3.56424719e-01
1.08774734e+00 8.89811292e-02 -8.47393796e-02 2.78701395e-01
-4.02313054e-01 6.89636946e-01 -4.22401428e-01 -9.14447665e-01
-3.74928005e-02 8.95748913e-01 -4.99922037e-01 -3.95454168e-01
5.39321005e-01 1.48744965e+00 -9.81247365e-01 -2.10701495e-01
4.10825729e-01 8.33034992e-01 -3.72560054e-01 4.37344283e-01
-2.25486442e-01 2.00758740e-01 -4.81929690e-01 5.08840263e-01
-1.85465470e-01 -5.44454336e-01 3.48340452e-01 3.99319874e-03
5.54823279e-01 4.85205203e-01 -4.43778306e-01 1.10796225e+00
-1.06911027e+00 1.30792284e+00 -7.75573492e-01 -8.49611163e-01
9.84740674e-01 5.50179064e-01 1.17509685e-01 -2.24061400e-01
-1.48750497e-02 4.24387306e-02 -1.98421195e-01 -5.30985475e-01
6.10921919e-01 -5.73757589e-01 2.81205624e-02 7.53653169e-01
-2.50052720e-01 1.35068879e-01 -4.70996946e-01 6.90339729e-02
1.39517415e+00 -1.06785879e-01 5.89295506e-01 -3.24173748e-01
2.34554648e-01 -9.74212140e-02 2.53182411e-01 9.84399259e-01
4.16958779e-01 8.29012454e-01 4.27030087e-01 -1.12044454e+00
-1.14472592e+00 -8.53887081e-01 2.01938540e-01 8.77305925e-01
-6.38514042e-01 -5.41860163e-01 -8.89511168e-01 -5.99635959e-01
-1.08749010e-01 1.41937935e+00 -1.28121960e+00 -4.43981916e-01
-6.09306931e-01 -5.62032282e-01 9.09678221e-01 1.32655334e+00
-5.88719808e-02 -1.49430978e+00 -6.14807427e-01 4.59341258e-01
-4.76783484e-01 -9.56087232e-01 -2.34849583e-02 5.23194373e-01
-1.31176805e+00 -6.96979702e-01 -2.59611547e-01 -4.00354862e-01
1.10846722e+00 -1.60903558e-01 1.87856591e+00 3.77237797e-01
-9.29853842e-02 3.74437511e-01 -1.11826889e-01 -8.56342733e-01
-1.03140223e+00 3.27583969e-01 3.64089847e-01 -3.70726705e-01
6.15022480e-01 -5.45163691e-01 -3.96541506e-01 1.00383937e-01
-1.14996636e+00 7.93432295e-01 6.35061860e-01 1.00193632e+00
4.93483841e-01 -3.88653576e-01 4.83141154e-01 -1.28789628e+00
8.81655812e-01 -6.75061941e-01 2.93482900e-01 5.14865398e-01
-7.25028813e-01 5.94732463e-01 9.06979859e-01 -1.59130096e-01
-8.65376055e-01 -2.57680327e-01 -2.01528221e-02 -4.86228734e-01
-2.87739009e-01 8.19169402e-01 3.20919514e-01 8.34729433e-01
8.97431135e-01 -1.22092143e-02 -9.82912257e-02 -3.74753982e-01
4.03588802e-01 7.00600505e-01 6.23453319e-01 -4.44044828e-01
2.53958613e-01 3.97202760e-01 1.98695183e-01 -4.48178142e-01
-1.08462906e+00 -3.86466831e-02 -6.31742835e-01 -3.51997279e-02
8.87712061e-01 -8.21820498e-01 -7.05636382e-01 -3.30259889e-01
-1.67608750e+00 -3.35745633e-01 -6.26866817e-01 3.30195695e-01
-7.19147980e-01 -4.57391381e-01 -6.75827146e-01 -1.11464870e+00
-6.62891567e-01 -7.63819456e-01 1.03520238e+00 -4.55507301e-02
-1.20405233e+00 -1.62355864e+00 -1.88458897e-02 3.11596066e-01
5.25240719e-01 4.14043009e-01 1.19233334e+00 -1.12568879e+00
-3.34197879e-01 -2.52632916e-01 -1.98626280e-01 3.63723099e-01
3.49865735e-01 -2.51354456e-01 -1.23677206e+00 3.25639874e-01
-1.12855464e-01 -2.23511321e-04 8.28498662e-01 6.06426060e-01
1.63104784e+00 -8.08855176e-01 -5.00264704e-01 5.27486563e-01
1.03634417e+00 -6.28248751e-02 6.97117209e-01 1.61190331e-02
1.16406274e+00 7.17173934e-01 4.57359590e-02 2.44472608e-01
3.61220270e-01 2.64441997e-01 3.94792140e-01 -6.68798089e-01
-3.17316055e-02 -5.37490189e-01 1.18447244e-01 1.01196957e+00
-7.79770911e-01 -5.08073509e-01 -1.41416776e+00 2.25743935e-01
-2.44593596e+00 -9.77831006e-01 -6.58506677e-02 1.35461724e+00
4.95907962e-01 4.67504710e-02 -3.13214868e-01 -1.64344430e-01
3.92740279e-01 -1.92618787e-01 -6.86257780e-01 -9.28124487e-01
-1.54143408e-01 2.15366676e-01 1.98024973e-01 4.43520069e-01
-6.95418298e-01 7.05610275e-01 7.32109833e+00 3.53389204e-01
-4.86700684e-01 -4.26074192e-02 1.41871881e+00 -1.56221017e-01
-6.64122403e-01 -3.59567434e-01 -2.82893986e-01 -8.10432360e-02
1.78249407e+00 9.40735042e-02 5.74386954e-01 8.99380624e-01
6.89254045e-01 6.53086782e-01 -1.75336921e+00 1.02587819e+00
5.22987405e-03 -1.82176554e+00 6.73933506e-01 -9.66883078e-02
6.84500396e-01 -2.76508778e-02 -4.18149233e-02 4.86511618e-01
5.12713373e-01 -2.19954586e+00 5.36299109e-01 1.40065873e+00
7.49842644e-01 -7.39396453e-01 1.10453582e+00 3.24872553e-01
-1.01747656e+00 -3.02812040e-01 -3.55281532e-01 -3.87745380e-01
-8.51911157e-02 3.68794024e-01 -1.20690846e+00 4.26669627e-01
4.82761472e-01 1.09000230e+00 -3.97024274e-01 1.26935855e-01
-3.44184309e-01 6.96236253e-01 -4.80737351e-02 -1.70857638e-01
1.91336721e-01 1.31179377e-01 3.70273501e-01 1.54065835e+00
2.21474245e-01 2.64490724e-01 -5.40566981e-01 1.58091629e+00
-3.53063732e-01 -3.42248142e-01 -8.61768603e-01 -1.10653147e-01
2.76679605e-01 9.97044265e-01 -4.16635990e-01 -6.53919399e-01
-1.98481418e-02 9.45693195e-01 5.52825749e-01 6.75753236e-01
-8.14846396e-01 -1.17547937e-01 8.27320755e-01 1.37853667e-01
-1.76490635e-01 4.79653142e-02 -7.77829111e-01 -1.00029981e+00
-3.13244015e-01 -9.08032000e-01 3.03081989e-01 -1.39264321e+00
-1.33212411e+00 1.34025764e+00 1.01507954e-01 -9.26366389e-01
-9.39362466e-01 -8.80516112e-01 -6.19928598e-01 9.51594472e-01
-9.78731453e-01 -1.49073720e+00 -3.00478399e-01 1.04085468e-01
7.31603384e-01 -1.80820018e-01 1.45324993e+00 -2.30086699e-01
-8.47827196e-01 3.16545516e-01 4.95458990e-02 1.94352075e-01
-8.86086896e-02 -1.36708748e+00 9.18326914e-01 2.91856259e-01
2.93650091e-01 1.16431904e+00 7.68606007e-01 -4.02800977e-01
-7.94340014e-01 -1.60642910e+00 1.19450831e+00 -1.18049324e+00
4.01764929e-01 -3.93694580e-01 -9.36303794e-01 1.29868007e+00
-7.75992796e-02 -2.55934596e-01 1.02131510e+00 5.16251326e-01
-1.30417362e-01 2.05486387e-01 -7.87737906e-01 9.19052362e-01
1.19320345e+00 -9.24073398e-01 -8.71786594e-01 4.00840372e-01
9.41101074e-01 -2.74918169e-01 -9.50098634e-01 4.71284628e-01
3.97664487e-01 -7.58108139e-01 8.97842050e-01 -1.48847353e+00
1.01116979e+00 -9.96795669e-03 2.40714755e-02 -1.43452883e+00
-3.37177038e-01 -6.27780080e-01 -8.41805041e-01 6.78320229e-01
9.29985464e-01 -5.23133159e-01 5.80108285e-01 1.30946112e+00
-2.72682786e-01 -1.19975495e+00 -5.14207184e-01 -2.21585482e-01
-1.36956319e-01 -8.32955658e-01 1.08503044e+00 7.25332081e-01
9.51112881e-02 5.58544397e-01 -6.49347961e-01 -9.48696584e-02
1.18987702e-01 -2.72522599e-01 3.28059673e-01 -1.49486768e+00
-3.29965293e-01 -4.73176330e-01 -3.91072810e-01 -7.94714928e-01
5.35056889e-01 -8.22160244e-01 2.19948008e-03 -2.02973700e+00
4.14776385e-01 -9.71210226e-02 -6.99549973e-01 9.38062012e-01
-1.35232389e-01 3.79286543e-03 1.16761409e-01 3.78334165e-01
-4.54618424e-01 3.71305406e-01 7.59830773e-01 -3.02765071e-01
-1.74222887e-01 1.31466761e-01 -9.14458811e-01 9.10395741e-01
8.60735178e-01 -6.53824270e-01 -4.43548828e-01 -5.55947185e-01
6.27927661e-01 1.44641697e-02 5.18364429e-01 -8.34775388e-01
1.17896214e-01 -1.52413249e-01 5.69490790e-01 -3.90796870e-01
4.40991491e-01 -8.75905156e-01 5.03842711e-01 5.59621155e-01
-1.00313890e+00 5.23844361e-01 5.84399939e-01 7.15045571e-01
-2.17694283e-01 3.27171907e-02 4.56215814e-02 -1.88429505e-01
-2.90977806e-01 4.65670936e-02 -7.85118520e-01 -1.66071996e-01
5.76225042e-01 2.17538569e-02 -2.37715334e-01 -8.90124142e-01
-9.44385946e-01 4.44343574e-02 -1.18844189e-01 5.41928172e-01
1.14147878e+00 -1.23553836e+00 -7.75021970e-01 3.91894504e-02
4.14396860e-02 1.90421551e-01 1.44217238e-02 3.35268021e-01
-7.03999996e-01 7.86265850e-01 -1.96854454e-02 -3.65553141e-01
-1.05550563e+00 3.37416977e-01 8.63893032e-01 -5.93203068e-01
-6.62043273e-01 5.22929847e-01 2.49375775e-01 -6.83878660e-01
1.90668210e-01 -1.21736753e+00 -2.30173901e-01 -4.55159724e-01
5.81163824e-01 4.21191543e-01 -2.80690938e-01 -3.10200274e-01
-1.32075340e-01 -3.08412053e-02 1.16104178e-01 1.43074557e-01
1.58465242e+00 8.60069469e-02 -2.84844249e-01 8.84130120e-01
9.41566646e-01 -7.56790042e-01 -7.08597839e-01 5.35765067e-02
2.61985153e-01 4.26434994e-01 -1.33048952e-01 -9.23150837e-01
-6.77257061e-01 9.91689205e-01 1.93271458e-01 3.34024370e-01
1.13514125e+00 7.59452656e-02 7.52080083e-01 9.95105743e-01
-1.87097073e-01 -7.60453522e-01 -1.77067712e-01 6.50374532e-01
9.81345057e-01 -1.01710761e+00 -1.49789348e-01 -5.28422091e-03
-7.24745631e-01 1.48420393e+00 3.87242168e-01 1.38761759e-01
5.06926358e-01 1.34034827e-01 8.37244019e-02 -4.17323798e-01
-1.16818929e+00 2.88036764e-01 8.13688517e-01 4.36110914e-01
6.47946477e-01 6.40133560e-01 6.81569278e-01 1.03264046e+00
-3.59759688e-01 4.29624841e-02 4.75502640e-01 4.35303569e-01
-1.81798171e-02 -6.56121850e-01 -2.04832673e-01 9.45104003e-01
-1.14061005e-01 -6.16201937e-01 -6.34327233e-01 6.63326859e-01
-1.08902425e-01 8.15847039e-01 4.22426432e-01 -7.37688541e-01
4.00476515e-01 2.22674385e-01 -2.31822222e-01 -8.86685073e-01
-8.91139150e-01 -6.91604853e-01 4.37750429e-01 -5.27952313e-01
-4.51415479e-01 -1.70315951e-01 -1.65297329e+00 -5.24183929e-01
1.25292122e-01 5.07538803e-02 4.01505291e-01 1.27263772e+00
7.02716827e-01 1.41059935e+00 1.36437103e-01 -5.29664993e-01
-3.43072891e-01 -1.25334251e+00 -1.71061754e-01 5.07361472e-01
5.61668456e-01 -7.56614879e-02 -2.89051265e-01 8.82866010e-02] | [8.334943771362305, 6.058230400085449] |
93d7f7b6-c300-463f-9d24-0c69b326435c | hintedbt-augmenting-back-translation-with | 2109.04443 | null | https://arxiv.org/abs/2109.04443v1 | https://arxiv.org/pdf/2109.04443v1.pdf | HintedBT: Augmenting Back-Translation with Quality and Transliteration Hints | Back-translation (BT) of target monolingual corpora is a widely used data augmentation strategy for neural machine translation (NMT), especially for low-resource language pairs. To improve effectiveness of the available BT data, we introduce HintedBT -- a family of techniques which provides hints (through tags) to the encoder and decoder. First, we propose a novel method of using both high and low quality BT data by providing hints (as source tags on the encoder) to the model about the quality of each source-target pair. We don't filter out low quality data but instead show that these hints enable the model to learn effectively from noisy data. Second, we address the problem of predicting whether a source token needs to be translated or transliterated to the target language, which is common in cross-script translation tasks (i.e., where source and target do not share the written script). For such cases, we propose training the model with additional hints (as target tags on the decoder) that provide information about the operation required on the source (translation or both translation and transliteration). We conduct experiments and detailed analyses on standard WMT benchmarks for three cross-script low/medium-resource language pairs: {Hindi,Gujarati,Tamil}-to-English. Our methods compare favorably with five strong and well established baselines. We show that using these hints, both separately and together, significantly improves translation quality and leads to state-of-the-art performance in all three language pairs in corresponding bilingual settings. | ['Aravindan Raghuveer', 'Abhirut Gupta', 'Melvin Johnson', 'Sahana Ramnath'] | 2021-09-09 | null | https://aclanthology.org/2021.emnlp-main.129 | https://aclanthology.org/2021.emnlp-main.129.pdf | emnlp-2021-11 | ['transliteration'] | ['natural-language-processing'] | [ 2.92160422e-01 -7.16807917e-02 -5.02812564e-01 -5.40318668e-01
-1.34111381e+00 -9.20788109e-01 8.26679468e-01 -1.46212295e-01
-5.23643255e-01 9.02438819e-01 4.06373233e-01 -8.92266452e-01
5.21854222e-01 -3.24005544e-01 -1.07379901e+00 -4.04657632e-01
3.57004493e-01 8.21977854e-01 -3.33504081e-01 -5.34613431e-01
-1.46330908e-01 1.49246916e-01 -6.92940652e-01 5.63660860e-01
9.98196781e-01 3.52305651e-01 2.81743348e-01 5.19042790e-01
-5.20326868e-02 5.91177464e-01 -5.10681808e-01 -9.40993369e-01
6.12137556e-01 -5.75526536e-01 -1.11721873e+00 -1.31264493e-01
5.11196256e-01 -3.06533277e-01 -2.64891773e-01 8.72714639e-01
5.42595923e-01 -3.05199444e-01 4.21634167e-01 -1.01139998e+00
-7.40639985e-01 1.15922058e+00 -4.67351019e-01 3.81520480e-01
1.52190149e-01 2.60033518e-01 1.13128376e+00 -1.26729369e+00
7.79994488e-01 1.06220269e+00 6.62382245e-01 7.24840939e-01
-1.33078039e+00 -6.12975299e-01 -7.92553127e-02 5.61274700e-02
-1.03908968e+00 -1.00508678e+00 3.55610788e-01 -3.23172987e-01
1.33458078e+00 3.69314432e-01 1.78987578e-01 1.43688369e+00
7.77069181e-02 8.65812659e-01 1.34300244e+00 -7.34616816e-01
-2.62057602e-01 2.74885118e-01 -3.44405770e-02 5.88622272e-01
-2.20780611e-01 1.41072378e-01 -7.48290658e-01 -1.15643321e-02
3.91813666e-01 -4.34623182e-01 -3.55803579e-01 8.69599134e-02
-1.76984465e+00 5.62570393e-01 4.05614302e-02 5.43932855e-01
-1.51065677e-01 -2.30702199e-03 5.74972391e-01 8.52254331e-01
5.29296219e-01 5.50530910e-01 -9.72504318e-01 -4.97048706e-01
-9.98413086e-01 -1.47850394e-01 7.16387272e-01 1.35853648e+00
8.96010280e-01 3.74999940e-02 -5.50549269e-01 9.72507954e-01
-1.24971583e-01 6.27909422e-01 5.38362980e-01 -4.09004986e-01
1.33600044e+00 3.64919960e-01 1.28887920e-02 -1.49157733e-01
3.54281142e-02 -4.22053128e-01 -7.21056819e-01 -3.99044216e-01
5.14803112e-01 -2.53962547e-01 -9.18016374e-01 1.95525646e+00
5.75131038e-04 -1.67084917e-01 1.88334972e-01 8.45695257e-01
3.87585491e-01 7.57826686e-01 -3.47684205e-01 -2.85154104e-01
1.15201628e+00 -1.38435853e+00 -6.32615626e-01 -6.39546394e-01
1.07130170e+00 -1.25921512e+00 1.50152779e+00 5.04868366e-02
-1.15328908e+00 -5.07552564e-01 -7.54626215e-01 -2.81572074e-01
-2.41769657e-01 3.84026229e-01 2.94429094e-01 4.17421669e-01
-1.15113032e+00 7.59736061e-01 -9.72045600e-01 -4.75039065e-01
-5.32425148e-03 4.56863940e-01 -5.90111554e-01 -3.68358016e-01
-1.24888563e+00 1.32718110e+00 1.18587576e-01 7.39869103e-02
-8.20617676e-01 -5.38711786e-01 -7.64917135e-01 -2.70388186e-01
7.12741613e-02 -4.95414048e-01 1.51979494e+00 -1.34407163e+00
-1.61048126e+00 1.01226652e+00 -4.26075131e-01 -2.87886947e-01
5.88153899e-01 -4.74169493e-01 -3.42508912e-01 -3.31508487e-01
3.15892816e-01 6.09048128e-01 6.83939934e-01 -9.74588156e-01
-5.06980002e-01 -1.64778560e-01 -1.10747240e-01 3.87463897e-01
-3.44685197e-01 5.34569502e-01 -6.62917495e-01 -6.99931562e-01
-9.54596624e-02 -1.12081254e+00 1.38776740e-02 -5.01378000e-01
-6.82407677e-01 4.13314952e-03 4.04382199e-01 -1.14297044e+00
1.03718960e+00 -1.91067934e+00 4.09956306e-01 -3.24369185e-02
-2.89483905e-01 3.29033703e-01 -4.34499890e-01 6.44633532e-01
-1.34841263e-01 1.76721767e-01 -2.07898229e-01 -6.28722191e-01
2.95442082e-02 5.60202062e-01 -2.86313564e-01 3.91848683e-01
4.30824399e-01 1.18135929e+00 -8.21405411e-01 -3.75287294e-01
-1.34212866e-01 1.96595743e-01 -2.98250407e-01 4.01153326e-01
-1.66247249e-01 8.22576046e-01 2.77523287e-02 5.26636302e-01
4.32800114e-01 1.59787595e-01 3.35946023e-01 -9.38772708e-02
-1.64058670e-01 1.12578464e+00 -6.15893602e-01 1.79515243e+00
-7.82324255e-01 5.85039377e-01 -1.85635373e-01 -6.97203994e-01
7.10958660e-01 6.16910875e-01 -3.12048569e-02 -8.62761617e-01
4.91744243e-02 6.48737729e-01 2.44372278e-01 -2.92975932e-01
4.99637038e-01 -1.87819138e-01 -1.61594138e-01 7.60765254e-01
3.24280143e-01 1.15674019e-01 3.24572235e-01 -2.50673052e-02
1.01394629e+00 4.26204562e-01 1.51625857e-01 -3.60004067e-01
2.80755222e-01 8.49271119e-02 6.61598146e-01 6.75017715e-01
1.22718163e-01 5.23855925e-01 3.47329795e-01 -2.00903296e-01
-1.62078357e+00 -7.31930792e-01 2.97882766e-01 1.35186303e+00
-4.30219442e-01 -4.96127278e-01 -8.07641268e-01 -9.69050705e-01
-4.25761729e-01 7.19363332e-01 -3.64910781e-01 -1.25779599e-01
-1.20811343e+00 -7.51908839e-01 7.41724312e-01 5.33155859e-01
2.48738021e-01 -8.90563488e-01 1.22309409e-01 3.00632626e-01
-7.82187641e-01 -1.27476823e+00 -1.00618267e+00 7.03234613e-01
-9.82834458e-01 -4.81180310e-01 -6.64676070e-01 -9.56496596e-01
6.14348173e-01 9.91507173e-02 1.44955051e+00 7.08809644e-02
4.68031645e-01 -2.65096307e-01 -3.70183468e-01 7.45659918e-02
-9.67015207e-01 4.82103586e-01 2.76351690e-01 -1.64160803e-01
3.07076097e-01 -3.03564370e-01 -9.35024023e-02 4.91042376e-01
-4.67808545e-01 4.97566998e-01 7.62459338e-01 1.09041119e+00
4.91728097e-01 -5.30039728e-01 1.39986068e-01 -9.42261338e-01
4.54800487e-01 -2.87943989e-01 -5.03445506e-01 4.79189306e-01
-5.54290891e-01 3.38606030e-01 9.83408034e-01 -5.52230716e-01
-7.18714774e-01 -6.89914152e-02 -2.68815845e-01 -2.31957510e-01
5.44525981e-02 4.48006451e-01 -3.86000484e-01 4.31133434e-02
6.79005921e-01 4.77515727e-01 -3.72748762e-01 -8.15993726e-01
3.78037184e-01 9.46619511e-01 5.41051805e-01 -9.09361839e-01
9.41216826e-01 -1.92262828e-01 -3.93378496e-01 -1.23631299e-01
-7.13503480e-01 -3.73509347e-01 -9.20560539e-01 3.29841077e-01
3.95948917e-01 -1.01404977e+00 -1.45557061e-01 3.12381506e-01
-1.32801652e+00 -6.55280113e-01 -8.15285146e-02 5.36484361e-01
-6.34455144e-01 1.89622343e-01 -1.10681820e+00 -3.68270725e-01
-6.06765330e-01 -1.51319659e+00 1.12424195e+00 -3.55934262e-01
-2.88638711e-01 -9.12086248e-01 -5.95474765e-02 4.62623656e-01
3.67678076e-01 -3.66570741e-01 1.23269916e+00 -8.49783659e-01
-4.68032509e-01 1.43834174e-01 -3.85367051e-02 6.50156379e-01
1.79148495e-01 -2.37456456e-01 -7.63127506e-01 -4.58844811e-01
-3.59531976e-02 -4.17313904e-01 4.78313565e-01 -2.62670636e-01
4.69199359e-01 -7.28223145e-01 -1.12852283e-01 6.04273260e-01
1.24587440e+00 -1.57538950e-01 4.42117393e-01 3.88850749e-01
8.14161360e-01 4.66124535e-01 5.65531552e-01 -1.14720911e-01
4.74972606e-01 1.03278482e+00 -7.72304833e-02 -3.34089786e-01
-2.73013353e-01 -3.92284036e-01 9.26066160e-01 1.51638210e+00
5.07566705e-03 -2.52071679e-01 -1.00695586e+00 6.36208475e-01
-1.69266641e+00 -6.06435418e-01 -2.65572071e-01 2.40946460e+00
1.58860588e+00 1.03605658e-01 4.61548157e-02 1.25127314e-02
5.45120060e-01 -2.29645729e-01 -1.88347593e-01 -7.04826355e-01
-2.44768053e-01 3.77303988e-01 7.01441765e-01 6.83132470e-01
-8.03986430e-01 1.46627343e+00 6.12386274e+00 7.39386678e-01
-1.18488204e+00 5.43563545e-01 5.64345062e-01 7.93379471e-02
-2.69758791e-01 1.58578515e-01 -9.94678676e-01 4.04615968e-01
1.31990206e+00 1.05138294e-01 8.55259418e-01 5.05923331e-01
1.86894104e-01 2.39579558e-01 -1.77604806e+00 6.59678400e-01
-5.23557104e-02 -1.02946615e+00 -2.79854331e-02 6.75379112e-02
7.89586544e-01 5.03023446e-01 -7.13803545e-02 5.28225422e-01
4.95693713e-01 -8.52691472e-01 9.99863386e-01 -2.85803191e-02
1.13736951e+00 -5.96348107e-01 7.99355984e-01 5.88631392e-01
-8.09289813e-01 3.38743150e-01 -3.56700331e-01 8.12970623e-02
3.69551890e-02 3.71783376e-01 -1.16532063e+00 6.89972878e-01
4.07385886e-01 6.38035774e-01 -5.00765026e-01 4.85736012e-01
-5.51853597e-01 9.82518315e-01 -2.37519175e-01 2.36480668e-01
3.84554207e-01 -6.89322874e-02 4.16633427e-01 1.49560738e+00
3.65145534e-01 -4.70172316e-01 1.03424406e-02 6.14643455e-01
-4.51859146e-01 3.52791667e-01 -5.93206108e-01 -2.05721721e-01
3.56411129e-01 9.79658663e-01 -2.04196826e-01 -4.61192697e-01
-6.39537394e-01 1.51035488e+00 6.96143866e-01 2.82504559e-01
-6.98582351e-01 -2.71689743e-02 7.56511331e-01 -2.70382632e-02
1.40749335e-01 -3.12953472e-01 -2.38325372e-01 -1.48238790e+00
3.23108584e-01 -1.56972158e+00 2.70862486e-02 -6.25475705e-01
-1.15786719e+00 8.72989297e-01 -3.27559888e-01 -1.19384611e+00
-4.42250699e-01 -4.56732184e-01 -2.95158446e-01 1.39954138e+00
-1.66900849e+00 -1.43046868e+00 4.60323513e-01 6.33840322e-01
6.84576750e-01 -2.31714234e-01 9.62276280e-01 6.38843000e-01
-5.31594038e-01 1.03319561e+00 3.15077245e-01 4.62422252e-01
1.04554808e+00 -1.19674230e+00 1.05270970e+00 1.20453775e+00
5.69308460e-01 8.39197636e-01 6.37493730e-01 -5.94401896e-01
-1.62358797e+00 -1.21161306e+00 1.72219515e+00 -6.78473592e-01
7.42241621e-01 -8.85914743e-01 -7.62600541e-01 1.09639120e+00
4.46040630e-01 -1.19381800e-01 5.59063077e-01 3.15179229e-01
-4.46432859e-01 2.59085689e-02 -7.48949111e-01 7.89120793e-01
1.01695299e+00 -7.25113928e-01 -6.45306706e-01 6.49949133e-01
8.75417292e-01 -5.65792680e-01 -7.62397945e-01 2.08807275e-01
3.25283945e-01 -5.52778900e-01 6.43750489e-01 -8.70692551e-01
5.27329504e-01 -9.91535634e-02 -4.12306249e-01 -1.66177058e+00
-2.18733355e-01 -9.44596648e-01 1.38391405e-01 1.40838206e+00
8.82262588e-01 -3.01163197e-01 4.47832257e-01 2.26388633e-01
-4.23552185e-01 -6.96917415e-01 -9.87065017e-01 -1.01001394e+00
3.99631500e-01 -3.66061628e-01 6.89131021e-01 1.09528005e+00
1.42357588e-01 6.99317753e-01 -8.12355161e-01 -2.28650849e-02
1.08702570e-01 5.85928410e-02 8.16328883e-01 -6.49252892e-01
-5.46838582e-01 -3.06119412e-01 7.52462223e-02 -1.19788611e+00
2.93655574e-01 -1.37022436e+00 2.74878621e-01 -1.31215823e+00
2.76114345e-01 -4.93510753e-01 -1.28368586e-01 8.97870481e-01
-3.58823329e-01 4.45352405e-01 1.21232636e-01 5.11606276e-01
-1.95659935e-01 3.41357619e-01 1.11201870e+00 2.91553065e-02
1.87277161e-02 3.88328396e-02 -6.27070963e-01 2.64411151e-01
7.31682539e-01 -8.04717660e-01 5.06119281e-02 -1.20144987e+00
3.05424750e-01 1.97809041e-01 -3.75124402e-02 -6.51473403e-01
-5.78727014e-02 -1.02973051e-01 3.34285609e-02 -2.29449049e-01
1.57215342e-01 -7.28164196e-01 -8.05114731e-02 2.69290924e-01
-4.53815043e-01 5.83928466e-01 1.08945526e-01 -8.48175138e-02
-1.14841476e-01 -1.59332275e-01 7.26210058e-01 -1.60072044e-01
-1.80116713e-01 2.59130180e-01 -3.04251194e-01 2.25090817e-01
3.34101558e-01 1.49981290e-01 -3.56310904e-01 -2.17827916e-01
-4.57118869e-01 5.88296615e-02 6.02435529e-01 6.38171375e-01
1.19467504e-01 -1.45463336e+00 -9.65754330e-01 3.29091817e-01
4.24178839e-01 -4.69791710e-01 -5.36904156e-01 1.17759705e+00
-2.94744611e-01 5.67532301e-01 -1.76233530e-01 -3.57032657e-01
-1.19316363e+00 4.18401390e-01 3.06519955e-01 -4.84079301e-01
-3.50343466e-01 8.65982890e-01 -1.88580796e-01 -1.00044620e+00
9.07985717e-02 -3.87547016e-01 4.60321605e-01 -3.37787390e-01
2.36922309e-01 1.82402711e-02 6.22218966e-01 -9.57218945e-01
-3.27275872e-01 2.15589374e-01 -3.79176855e-01 -5.05054057e-01
1.17909205e+00 -1.99555829e-01 -3.60676438e-01 5.66211879e-01
1.25112557e+00 2.75175005e-01 -8.70320559e-01 -7.93423176e-01
3.08686644e-01 -3.34281772e-01 -2.33723387e-01 -1.24069226e+00
-8.31021369e-01 1.18165708e+00 2.67827451e-01 -4.87383813e-01
8.73159349e-01 -7.84924328e-02 1.01792037e+00 4.49843884e-01
4.32698011e-01 -1.08137095e+00 -1.79548368e-01 9.46986735e-01
7.20098436e-01 -1.38847876e+00 -5.25303006e-01 -1.72222227e-01
-6.78829730e-01 9.48685229e-01 5.65701365e-01 4.09586549e-01
-3.15527506e-02 4.54755455e-01 4.44728941e-01 3.73621553e-01
-9.55569446e-01 -7.23056048e-02 4.13483679e-01 4.30880308e-01
9.04256284e-01 1.07134573e-01 -2.73630500e-01 3.94404948e-01
-5.31113625e-01 -2.53722340e-01 3.51759166e-01 9.50737059e-01
-6.71245232e-02 -1.70347166e+00 -2.50959426e-01 3.71212751e-01
-7.45884120e-01 -9.09110725e-01 -5.65661550e-01 5.69845974e-01
1.04122952e-01 9.57826495e-01 -2.20533311e-01 -4.57949936e-01
2.80163556e-01 4.13027734e-01 5.89348197e-01 -8.14601004e-01
-1.15099251e+00 1.92613602e-01 3.25916350e-01 -4.64553505e-01
1.65174790e-02 -6.78029895e-01 -8.53566885e-01 -4.27161038e-01
-3.87191474e-01 2.21797168e-01 8.70707273e-01 1.23319185e+00
2.21727476e-01 1.33509189e-01 6.50191069e-01 -4.60116506e-01
-7.61000872e-01 -1.31473291e+00 -2.01079473e-02 3.74618232e-01
3.64294887e-01 -1.08200073e-01 -1.30891977e-02 2.31390879e-01] | [11.598931312561035, 10.302660942077637] |
aaa94c78-40a1-46a2-878a-d5c9265c36f1 | threat-detection-in-self-driving-vehicles | 2209.02438 | null | https://arxiv.org/abs/2209.02438v1 | https://arxiv.org/pdf/2209.02438v1.pdf | Threat Detection In Self-Driving Vehicles Using Computer Vision | On-road obstacle detection is an important field of research that falls in the scope of intelligent transportation infrastructure systems. The use of vision-based approaches results in an accurate and cost-effective solution to such systems. In this research paper, we propose a threat detection mechanism for autonomous self-driving cars using dashcam videos to ensure the presence of any unwanted obstacle on the road that falls within its visual range. This information can assist the vehicle's program to en route safely. There are four major components, namely, YOLO to identify the objects, advanced lane detection algorithm, multi regression model to measure the distance of the object from the camera, the two-second rule for measuring the safety, and limiting speed. In addition, we have used the Car Crash Dataset(CCD) for calculating the accuracy of the model. The YOLO algorithm gives an accuracy of around 93%. The final accuracy of our proposed Threat Detection Model (TDM) is 82.65%. | ['Poonam Saini', 'Taresh Sharma', 'Akshay Singh', 'Param Patil', 'Aaryan Jagetia', 'Umang Goenka'] | 2022-09-06 | null | null | null | null | ['lane-detection'] | ['computer-vision'] | [-2.06286281e-01 -3.75809491e-01 -2.92777712e-03 -1.92047372e-01
-3.49678993e-01 -2.73939013e-01 3.83195758e-01 -1.44271255e-01
-6.58617198e-01 4.74214941e-01 -2.87248909e-01 -7.54529238e-01
-6.68215752e-02 -1.03593373e+00 -4.82880503e-01 -4.26844239e-01
4.15402025e-01 -1.14759997e-01 1.13346303e+00 -5.24753928e-01
7.62930334e-01 7.42475390e-01 -1.84876049e+00 -2.28002951e-01
9.70047712e-01 9.57253218e-01 2.04957917e-01 8.84435713e-01
-2.92253178e-02 6.58318281e-01 -2.02474102e-01 -5.54897590e-03
2.91639805e-01 1.37976393e-01 -1.87179029e-01 -2.88659722e-01
4.32672322e-01 -8.49912226e-01 -4.18155342e-01 8.68482649e-01
1.89691395e-01 1.65287346e-01 8.81003261e-01 -1.63207293e+00
7.75946975e-02 -2.94969022e-01 -6.80949390e-01 5.16704679e-01
1.57455504e-01 1.56984895e-01 1.04313351e-01 -8.19197893e-01
3.06640387e-01 1.11994195e+00 4.36377198e-01 1.62756845e-01
-3.82793248e-01 -8.46249819e-01 -4.06792730e-01 9.18790579e-01
-1.68610704e+00 -4.54265565e-01 7.12046981e-01 -6.54787958e-01
6.52566731e-01 7.43519962e-02 1.25417426e-01 2.75587976e-01
5.32749712e-01 7.48110712e-02 1.10527599e+00 -4.95311975e-01
1.25011176e-01 4.31819946e-01 6.53320253e-01 5.97734153e-01
8.39393735e-01 4.39143658e-01 1.04133025e-01 2.64741331e-01
1.99209049e-01 -1.14956908e-01 2.44129330e-01 -1.03049524e-01
-3.40012670e-01 8.56990218e-01 1.55152842e-01 9.74959061e-02
-3.12814713e-01 -1.29734576e-01 3.70290548e-01 -9.35222879e-02
-5.07683679e-02 -4.93122250e-01 3.34693417e-02 -2.94927567e-01
-5.15435100e-01 1.69230640e-01 2.15015665e-01 9.10565138e-01
8.44939232e-01 1.05791412e-01 2.77136207e-01 4.08496618e-01
5.58519304e-01 8.03237379e-01 -1.18049935e-01 -9.98295724e-01
6.22905314e-01 6.26583993e-01 1.73061356e-01 -1.14280868e+00
-2.18955517e-01 2.02738181e-01 -3.45744491e-01 1.02490103e+00
3.95337254e-01 -2.75041550e-01 -6.08634233e-01 9.77976561e-01
5.70000529e-01 3.22146565e-01 7.69618303e-02 7.05930114e-01
7.34226048e-01 8.38792801e-01 1.45541951e-01 6.55491799e-02
1.19210434e+00 -6.36011064e-01 -7.99419641e-01 -6.52031600e-02
7.08976030e-01 -7.72785544e-01 6.20424390e-01 3.73205036e-01
-5.42550802e-01 -7.29972720e-01 -1.23390472e+00 6.56351969e-02
-7.83261359e-01 2.36731276e-01 6.55713603e-02 8.72256041e-01
-6.85417831e-01 -5.01237176e-02 -2.60552168e-01 -4.14871216e-01
1.43509626e-01 9.43542644e-02 -3.57955039e-01 -2.97490001e-01
-1.15496230e+00 1.44831467e+00 2.63356715e-01 1.37762919e-01
-8.61699462e-01 -2.61797011e-01 -8.16642642e-01 -2.30859965e-01
4.63313371e-01 -1.49690220e-02 6.34153187e-01 -4.88003790e-01
-1.06091845e+00 6.63848817e-01 -2.70603657e-01 -3.57316971e-01
5.93468308e-01 -3.67508709e-01 -7.80572355e-01 2.23255783e-01
3.44088286e-01 3.35967928e-01 4.62458253e-01 -1.16392946e+00
-1.27619970e+00 -3.91896755e-01 -1.56888738e-01 1.27836779e-01
1.66713998e-01 2.84837663e-01 -2.51162916e-01 3.29550445e-01
-3.39027762e-01 -8.15352321e-01 -1.14550360e-01 1.94717031e-02
-2.14691117e-01 -4.06943321e-01 1.48266876e+00 -7.28994429e-01
1.33006775e+00 -2.10059786e+00 -8.46438348e-01 3.49521995e-01
9.55376681e-03 8.74466658e-01 1.83788702e-01 2.00765297e-01
2.34012932e-01 -1.37924969e-01 1.40392840e-01 3.46521914e-01
-3.69348437e-01 6.66611046e-02 -1.05150059e-01 3.92393500e-01
1.21362291e-01 2.28619203e-01 -4.00995135e-01 -5.58415711e-01
8.82436633e-01 4.11253035e-01 1.99493319e-02 2.18080163e-01
5.00985146e-01 -1.35732949e-01 -4.55471635e-01 4.30838972e-01
1.18313801e+00 1.01763022e+00 -6.44693375e-01 2.31609531e-02
-1.06190550e+00 -1.66723683e-01 -1.44478500e+00 4.02594715e-01
-2.36137778e-01 1.01504159e+00 1.26840146e-02 -6.20177805e-01
1.10907352e+00 6.42156452e-02 1.91615418e-01 -9.32526410e-01
3.61060351e-01 2.40420133e-01 -5.38896360e-02 -9.69194412e-01
6.11801863e-01 4.22317743e-01 2.49975219e-01 -2.49778941e-01
-6.57690346e-01 2.16539457e-01 4.06614751e-01 3.15898955e-02
8.89695048e-01 -1.14852814e-02 8.46169963e-02 4.96496866e-03
1.02296674e+00 4.91566539e-01 3.56538951e-01 5.22572398e-01
-8.04208040e-01 1.29856646e-01 1.95282936e-01 -4.78740394e-01
-1.06782579e+00 -8.01491499e-01 -8.57420713e-02 4.33700651e-01
6.44020796e-01 1.29646897e-01 -9.06451404e-01 -3.61281812e-01
1.95301883e-02 9.47018683e-01 -1.85612977e-01 -2.93096542e-01
-6.04166269e-01 -1.18857034e-01 4.92845148e-01 3.32965314e-01
8.35360587e-01 -6.80647671e-01 -7.69613683e-01 7.07430020e-02
-5.22405952e-02 -1.34119642e+00 -3.71011309e-02 -4.51724470e-01
-3.65169466e-01 -1.43982804e+00 -6.44615367e-02 -6.64061129e-01
5.21401167e-01 1.09825778e+00 7.20674917e-02 1.67012736e-01
-5.08411825e-01 2.01924860e-01 -1.39228091e-01 -8.29640210e-01
-4.94824320e-01 -3.59560281e-01 -8.56391191e-02 -5.25895841e-02
8.07529032e-01 1.42800689e-01 -5.20063579e-01 5.77101409e-01
-2.42368862e-01 -2.17593327e-01 4.34346318e-01 -2.28820592e-01
1.52677163e-01 4.20281023e-01 5.33133805e-01 -1.22650921e-01
4.24368471e-01 -6.93725586e-01 -1.05700576e+00 -1.22736409e-01
-6.44599319e-01 -4.05297637e-01 5.28065622e-01 -6.97150156e-02
-9.85754192e-01 2.96181180e-02 -3.09893876e-01 -7.65501559e-02
-8.45385969e-01 -9.19683054e-02 -2.11291283e-01 -2.32994199e-01
4.85040039e-01 6.36355877e-02 2.50165552e-01 -7.81347156e-02
1.53722391e-01 1.04721749e+00 4.64078575e-01 2.09343970e-01
8.70470405e-01 5.87287188e-01 4.62362140e-01 -1.32725978e+00
-2.57028162e-01 -8.97080302e-01 -6.33029759e-01 -9.91097987e-01
1.06375134e+00 -7.15185881e-01 -1.14124405e+00 4.94830877e-01
-1.09892023e+00 1.52516007e-01 5.51519394e-01 7.35545218e-01
-1.20921507e-01 6.11161172e-01 2.03745859e-03 -1.34928155e+00
-2.38182530e-01 -9.49255705e-01 5.15099227e-01 5.04423320e-01
2.03870982e-01 -6.17194533e-01 -6.19417876e-02 6.70106292e-01
3.45527619e-01 2.35859171e-01 5.04897594e-01 -1.12017810e-01
-6.69147611e-01 -5.15897036e-01 -4.93906528e-01 3.88381958e-01
-1.43283114e-01 5.04252732e-01 -7.57123470e-01 3.05617332e-01
-2.77633190e-01 2.78801113e-01 7.75466084e-01 4.63195145e-01
4.85930532e-01 2.38351077e-01 -4.70171660e-01 8.57994631e-02
1.61066568e+00 7.65794218e-01 1.10522091e+00 7.35422194e-01
5.44925213e-01 7.00884879e-01 1.35275722e+00 6.03745542e-02
7.32833683e-01 7.15402544e-01 7.25202978e-01 -1.47109106e-02
-2.04956949e-01 -2.13737879e-02 5.18005967e-01 2.92214692e-01
-9.58809033e-02 -7.27958232e-02 -1.00207508e+00 4.65814501e-01
-1.70902896e+00 -1.37345958e+00 -1.29134464e+00 2.30739284e+00
-1.84370130e-01 4.99612451e-01 3.03603113e-01 5.79867661e-01
1.03255534e+00 -4.82838154e-01 -3.85887995e-02 -9.23031032e-01
3.11977714e-01 -4.74064171e-01 1.03432024e+00 9.51364636e-01
-1.16260600e+00 8.88004839e-01 5.93324804e+00 8.01829994e-01
-1.08917749e+00 -2.55522616e-02 1.61876410e-01 5.71223319e-01
2.00534686e-01 4.92058992e-02 -1.31460929e+00 6.29750907e-01
1.09815168e+00 -4.52534147e-02 1.44076198e-01 1.01172411e+00
1.03707933e+00 -9.04554486e-01 -1.15476020e-01 6.70329452e-01
1.97009563e-01 -9.10423160e-01 -2.27854818e-01 4.34359647e-02
7.59737268e-02 -1.63311824e-01 -2.64549524e-01 2.61124462e-01
-7.47379139e-02 -5.07862329e-01 5.75443327e-01 5.48115253e-01
5.83399415e-01 -1.09457743e+00 8.38493288e-01 6.91614151e-01
-1.29330432e+00 -2.50499278e-01 -4.92504150e-01 -3.25487405e-01
2.87392795e-01 8.74674544e-02 -9.94332314e-01 2.61155933e-01
5.12067437e-01 1.10912815e-01 -5.82452059e-01 1.26571667e+00
-5.81359630e-03 5.27349114e-01 -3.76371175e-01 -3.22844148e-01
3.49372774e-01 -5.14440954e-01 5.15660882e-01 9.58643079e-01
3.99830103e-01 7.64542222e-02 -1.80322547e-02 4.63230282e-01
7.14968920e-01 1.27182350e-01 -1.17300761e+00 8.12684476e-01
4.77289200e-01 1.15226924e+00 -3.97780865e-01 -1.58111960e-01
-6.68046176e-01 2.65039742e-01 2.20210925e-02 1.36768386e-01
-1.13910007e+00 -8.47756624e-01 6.77932680e-01 4.79974389e-01
2.17137486e-01 -5.18911481e-01 -3.88560891e-01 -1.43234685e-01
7.89184496e-02 -5.24850003e-02 1.10760191e-02 -1.06604040e+00
-2.53632665e-01 1.83298215e-01 1.72072843e-01 -1.24596786e+00
1.59337297e-01 -8.58117282e-01 -8.54888320e-01 7.22561538e-01
-1.76419675e+00 -1.00879872e+00 -7.32475162e-01 4.47994798e-01
5.77456713e-01 -2.73241222e-01 1.97722629e-01 6.45722270e-01
-7.32715726e-01 1.96040243e-01 1.03527941e-01 1.53879046e-01
3.17725092e-01 -5.78451812e-01 -8.44566897e-02 1.29662895e+00
-7.90534019e-01 -2.09689066e-02 7.40374804e-01 -8.66159499e-01
-1.17052794e+00 -1.04329896e+00 8.72792959e-01 -5.03914058e-01
3.51422995e-01 -1.15090748e-02 -6.98708653e-01 1.47959650e-01
-5.03130406e-02 -2.08372593e-01 1.50901780e-01 -6.47651851e-01
1.33153245e-01 -4.59072948e-01 -1.28405547e+00 7.05244064e-01
4.56066310e-01 -4.84766513e-02 -4.11841303e-01 -2.45600060e-01
2.76902139e-01 -7.95214809e-03 -2.15190053e-01 2.09300131e-01
6.86897933e-01 -9.78530824e-01 7.92481542e-01 -1.60172015e-01
6.16909713e-02 -6.90365732e-01 4.26456369e-02 -5.71821570e-01
-2.59534538e-01 -6.00041710e-02 2.38772124e-01 1.23336649e+00
2.55810589e-01 -6.43202484e-01 6.38100922e-01 7.43675709e-01
-3.22832167e-01 -2.67192125e-01 -9.82025623e-01 -5.93457520e-01
-1.85055643e-01 -8.07192028e-01 1.08767495e-01 3.19355190e-01
-2.91112304e-01 2.00619593e-01 -2.61871397e-01 6.95014834e-01
8.20056975e-01 -6.63343430e-01 1.07944775e+00 -1.16340518e+00
9.59112048e-01 -2.57503629e-01 -8.19474399e-01 -5.39692938e-01
-2.41072532e-02 -2.23989800e-01 -3.92969958e-02 -1.64488721e+00
-4.65243123e-02 -3.40283304e-01 1.17127165e-01 5.53676523e-02
-1.29616514e-01 -1.88976116e-02 -1.42437816e-02 -9.25426558e-02
-4.09417361e-01 1.18836135e-01 7.70024002e-01 1.41911194e-01
-2.77409256e-02 2.08729848e-01 -2.92747170e-01 7.15511799e-01
1.12439048e+00 -5.51471055e-01 -2.89925814e-01 -1.82306916e-02
-1.68468148e-01 -6.82396069e-02 4.81099635e-01 -1.31129050e+00
4.60417807e-01 -4.57143098e-01 1.35357091e-02 -1.19898081e+00
2.72899777e-01 -1.40406704e+00 -1.40080929e-01 8.29098642e-01
2.32592866e-01 1.62169069e-01 3.61376852e-01 2.80004621e-01
1.01329852e-02 -5.28609157e-01 8.70702267e-01 3.30271751e-01
-1.26251519e+00 -2.85581648e-02 -1.07817924e+00 -2.71855921e-01
1.79592967e+00 -8.89272094e-01 -6.60189569e-01 -2.20165446e-01
-2.14587967e-03 3.64657491e-01 8.64354447e-02 6.78754449e-01
9.25865233e-01 -1.11065817e+00 -6.37271106e-01 3.35280120e-01
1.27121493e-01 -4.79478091e-01 3.14785182e-01 6.32253289e-01
-1.02575314e+00 6.97388172e-01 -4.62317109e-01 -4.57306206e-01
-1.79077399e+00 7.09222794e-01 2.16082275e-01 5.20999908e-01
-2.55349457e-01 -1.21163845e-01 -3.08736444e-01 1.46127030e-01
-1.87082011e-02 4.98243123e-02 -8.19969594e-01 -2.52191752e-01
7.55214930e-01 1.28465569e+00 -1.54180929e-01 -1.23666000e+00
-5.98111987e-01 1.12868500e+00 2.06045628e-01 -1.27121508e-01
6.73451841e-01 -5.28519511e-01 2.62846351e-01 2.03992352e-01
9.26188588e-01 1.56383529e-01 -9.80777502e-01 3.28929842e-01
-5.44381440e-02 -4.90182787e-01 3.43055218e-01 -5.02724290e-01
-3.77160788e-01 8.79736722e-01 1.06172252e+00 8.47398192e-02
6.47544086e-01 -5.12239039e-01 7.03950822e-01 2.44958669e-01
3.43795002e-01 -1.55016434e+00 -3.21293980e-01 4.84756231e-01
4.41188991e-01 -1.52457976e+00 -1.19000196e-01 -7.67414689e-01
-5.78520477e-01 1.22618949e+00 1.03039646e+00 -8.89915451e-02
6.97059989e-01 2.99822509e-01 2.22605407e-01 5.47385439e-02
-4.44347233e-01 -5.53093016e-01 -2.64389366e-02 1.00104320e+00
-1.87836394e-01 -1.46276783e-02 -5.81742048e-01 9.04605240e-02
2.60171622e-01 1.84502676e-01 8.03448677e-01 8.34610999e-01
-1.44920695e+00 -4.61174667e-01 -8.48381639e-01 2.01622024e-01
-1.75839499e-01 3.90569389e-01 -7.28043467e-02 1.00307286e+00
4.74960953e-01 1.54491913e+00 8.18616003e-02 -3.80551249e-01
6.81745112e-01 -2.74524927e-01 -1.46705255e-01 -4.57127169e-02
1.04857162e-01 -3.88368249e-01 2.94807196e-01 -2.73572087e-01
8.79609138e-02 -5.21645427e-01 -1.40353227e+00 -5.79476893e-01
-3.51328731e-01 -1.99498981e-02 1.05822742e+00 9.67397094e-01
2.45412260e-01 1.76888987e-01 1.02333450e+00 -3.94411594e-01
-9.93449539e-02 -5.75982273e-01 -1.90231979e-01 2.24438936e-01
3.06395531e-01 -8.27681363e-01 -3.98957253e-01 -2.79169798e-01] | [7.923628330230713, -1.125646948814392] |
ce634d01-515b-4100-8fe3-d43db7c2f909 | enforcing-almost-sure-reachability-in-pomdps | 2007.00085 | null | https://arxiv.org/abs/2007.00085v3 | https://arxiv.org/pdf/2007.00085v3.pdf | Enforcing Almost-Sure Reachability in POMDPs | Partially-Observable Markov Decision Processes (POMDPs) are a well-known stochastic model for sequential decision making under limited information. We consider the EXPTIME-hard problem of synthesising policies that almost-surely reach some goal state without ever visiting a bad state. In particular, we are interested in computing the winning region, that is, the set of system configurations from which a policy exists that satisfies the reachability specification. A direct application of such a winning region is the safe exploration of POMDPs by, for instance, restricting the behavior of a reinforcement learning agent to the region. We present two algorithms: A novel SAT-based iterative approach and a decision-diagram based alternative. The empirical evaluation demonstrates the feasibility and efficacy of the approaches. | ['Sanjit A. Seshia', 'Nils Jansen', 'Sebastian Junges'] | 2020-06-30 | null | null | null | null | ['safe-exploration'] | ['robots'] | [ 2.05088124e-01 6.16997004e-01 -1.18281305e-01 7.50186071e-02
-7.02727735e-01 -7.69687295e-01 6.31702363e-01 1.26142040e-01
-4.17751461e-01 1.15561330e+00 -1.64050702e-02 -8.79627228e-01
-4.76483434e-01 -9.51973021e-01 -4.68216360e-01 -9.18149352e-01
-2.83591568e-01 8.67806435e-01 5.13649464e-01 5.17479442e-02
3.99877168e-02 5.74509263e-01 -1.25300431e+00 -2.88251698e-01
4.67218369e-01 8.15634191e-01 1.40994951e-01 7.86880136e-01
4.72161770e-02 8.17625046e-01 -3.40125442e-01 2.53733367e-01
3.84050488e-01 -6.01718903e-01 -1.11261451e+00 2.17755884e-01
-7.85422266e-01 -4.68251467e-01 -9.27000940e-02 1.31733131e+00
1.24464639e-01 2.57269770e-01 2.52909064e-01 -1.49562061e+00
3.18483531e-01 9.07780647e-01 -1.80794135e-01 3.64910625e-02
5.92285633e-01 6.13116562e-01 7.60287344e-01 2.11496368e-01
6.74444795e-01 1.38796270e+00 -2.05296814e-01 6.38142288e-01
-1.37575996e+00 -1.92734465e-01 4.25326675e-01 1.25720158e-01
-1.33337593e+00 -8.04643929e-02 1.67748541e-01 -6.92019463e-02
9.46660101e-01 3.19872886e-01 7.41152406e-01 9.24645007e-01
6.48591042e-01 6.77012742e-01 1.77994800e+00 -6.17950916e-01
1.14938164e+00 -2.40645230e-01 -1.58425644e-01 3.52174431e-01
3.89420629e-01 6.32315159e-01 2.03099158e-02 -5.01017988e-01
6.14691436e-01 2.69808667e-03 1.93867698e-01 -4.45363075e-01
-1.01201320e+00 5.84107697e-01 -5.05719721e-01 -6.59988001e-02
-6.68546796e-01 3.75046670e-01 5.82907088e-02 6.44031942e-01
-2.38048911e-01 4.59559977e-01 -1.76594466e-01 -3.84962887e-01
-6.14378750e-01 5.87320387e-01 1.28146672e+00 8.31803977e-01
4.54760522e-01 8.00028443e-02 -4.10973907e-01 -5.80997884e-01
3.24643403e-01 6.83301985e-01 -2.68228631e-02 -1.24638200e+00
2.88262993e-01 2.59572476e-01 8.77989829e-01 4.59560491e-02
-3.10448885e-01 9.55608338e-02 -1.63863942e-01 8.35324347e-01
5.12080193e-01 -5.98535717e-01 -7.45534658e-01 1.70407856e+00
5.13894677e-01 -2.48436667e-02 3.85590374e-01 7.27365136e-01
-3.04759294e-01 8.32195759e-01 -6.59452900e-02 -8.58546793e-01
9.04485047e-01 -5.01983941e-01 -6.54152751e-01 -4.48790677e-02
1.19412303e-01 -1.27153486e-01 5.72567880e-01 6.71762824e-01
-1.37169647e+00 2.42313355e-01 -1.15025365e+00 8.30936074e-01
2.91540653e-01 -6.97822928e-01 4.73474860e-01 6.72713757e-01
-9.79280293e-01 7.98653781e-01 -1.45837533e+00 -1.90772071e-01
-1.93853334e-01 5.40454865e-01 3.96379493e-02 -1.31585449e-01
-8.63310754e-01 9.38843191e-01 7.07555354e-01 -1.40209407e-01
-1.92985690e+00 2.51398802e-01 -4.87468809e-01 2.42940083e-01
1.37397540e+00 -3.39297175e-01 1.73714924e+00 -5.83054721e-01
-2.08906150e+00 3.21802646e-01 1.09758601e-02 -8.66569161e-01
7.60387301e-01 2.28190184e-01 -2.00795397e-01 1.98995993e-01
-1.02938101e-01 1.82101447e-02 8.32094371e-01 -9.11311686e-01
-7.67980337e-01 -3.42812240e-01 6.91524506e-01 4.55866516e-01
2.58014053e-01 1.15310416e-01 -1.11393377e-01 4.21274275e-01
2.33011439e-01 -1.35298860e+00 -1.08979976e+00 -5.27943909e-01
-7.41452336e-01 -2.46162623e-01 1.02681428e-01 -2.31602974e-02
1.09653878e+00 -1.77627766e+00 1.79879978e-01 5.07554293e-01
-1.15857147e-01 -1.56759098e-01 2.04544626e-02 9.91872609e-01
2.80349553e-01 8.20935816e-02 -3.95053700e-02 1.57652602e-01
3.32607508e-01 4.27294791e-01 -5.45647919e-01 5.42552233e-01
1.29756434e-02 4.52270299e-01 -8.41718137e-01 -4.91737723e-01
2.43525833e-01 -6.03851438e-01 -2.65535146e-01 2.75171757e-01
-7.99115896e-01 6.46228015e-01 -8.99097264e-01 2.87498891e-01
4.36311990e-01 -3.90496403e-02 7.86924124e-01 9.99005854e-01
-4.91655201e-01 4.99187648e-01 -1.72735572e+00 1.03282249e+00
-2.06557825e-01 -1.08078748e-01 8.36296678e-02 -4.95252550e-01
4.12153453e-01 4.74014074e-01 4.59635347e-01 -4.20959055e-01
4.06448126e-01 -3.17347310e-02 5.59660569e-02 -5.95366433e-02
4.03219014e-01 -5.26397824e-01 -4.84869033e-01 1.03374541e+00
-3.20581168e-01 -1.90584958e-01 1.29429549e-01 -9.50725600e-02
1.42382300e+00 2.17350632e-01 9.25604641e-01 -3.46881747e-01
3.64928454e-01 1.28590181e-01 6.33233249e-01 1.24811423e+00
-2.79507488e-01 8.80340871e-04 1.01656044e+00 -1.08526595e-01
-6.23833477e-01 -1.01584578e+00 4.48911816e-01 5.12705445e-01
4.83973384e-01 -3.27385217e-01 -7.45385647e-01 -4.84539658e-01
-2.52563894e-01 1.14376473e+00 -4.24221009e-01 -1.10623818e-02
-3.79286677e-01 -1.16982564e-01 8.25117975e-02 4.00224589e-02
3.30245733e-01 -1.14946675e+00 -1.38145339e+00 4.16784018e-01
1.30635664e-01 -8.21945071e-01 7.93017820e-02 5.49072742e-01
-8.60051036e-01 -8.32972467e-01 -2.08869740e-01 5.01041524e-02
5.53747773e-01 -9.61034819e-02 5.88375330e-01 -3.08270931e-01
4.38901126e-01 4.48950320e-01 -3.28737423e-02 -4.99488413e-01
-8.53480279e-01 -1.09203845e-01 2.64570296e-01 -3.85080814e-01
5.99171706e-02 -2.65225530e-01 -2.80025333e-01 3.19758236e-01
-8.14548135e-01 7.91092515e-02 3.43889803e-01 2.93585479e-01
1.02988327e+00 6.93014979e-01 5.39988950e-02 -4.94949758e-01
8.84843111e-01 -3.16087604e-01 -1.29715073e+00 4.08342808e-01
-6.01014674e-01 6.43313229e-01 7.85664320e-01 -2.71881193e-01
-9.96537149e-01 3.46641392e-01 1.83704481e-01 -6.88965172e-02
-4.42781359e-01 3.56131405e-01 -5.54874659e-01 3.09793830e-01
2.81171113e-01 2.94386804e-01 -8.35967138e-02 -6.52053505e-02
8.14810172e-02 3.32157582e-01 2.16971591e-01 -9.45652723e-01
6.94003224e-01 5.31052768e-01 5.68842053e-01 -4.23198849e-01
-4.11986560e-01 -2.16172531e-01 -5.35206776e-03 -3.34718347e-01
5.54099619e-01 -3.98481876e-01 -1.29096186e+00 8.93279165e-02
-5.37333190e-01 -6.64240658e-01 -7.64954925e-01 4.03785676e-01
-1.17520535e+00 2.55039185e-01 -2.80624628e-01 -1.57759607e+00
1.40110627e-01 -1.21570706e+00 5.76841235e-01 4.55796808e-01
-1.42724127e-01 -4.63745326e-01 4.05012429e-01 -3.21880013e-01
-3.58859077e-02 5.75254738e-01 6.07646227e-01 -7.84957170e-01
-8.36115062e-01 -5.21591268e-02 5.17827868e-01 -1.77789927e-01
-2.81081855e-01 -1.03915170e-01 -5.87544501e-01 -5.37999868e-01
2.62894481e-01 -1.26202092e-01 1.67809442e-01 5.11516809e-01
7.09778011e-01 -6.34698451e-01 -4.22903568e-01 -1.62627369e-01
1.63616180e+00 8.11959565e-01 5.36283851e-01 4.88857239e-01
-2.72326022e-01 3.45322162e-01 1.01173496e+00 9.28552508e-01
9.44718122e-02 5.63207388e-01 6.84402168e-01 5.96836627e-01
7.51980186e-01 -5.01943111e-01 5.93467236e-01 -5.40785119e-02
-2.11604029e-01 -3.05466354e-01 -9.02495146e-01 4.51712817e-01
-1.98044896e+00 -1.02440107e+00 2.89771378e-01 2.64801121e+00
7.11584628e-01 6.97033584e-01 3.15838635e-01 2.05546066e-01
4.99049604e-01 -2.58105159e-01 -8.59796584e-01 -7.12730467e-01
4.13717419e-01 2.94868022e-01 9.56997931e-01 6.60644114e-01
-7.78730094e-01 8.38184357e-01 6.41484165e+00 5.49771369e-01
-6.97849572e-01 5.54203056e-02 1.55259922e-01 -1.86187312e-01
-2.29995072e-01 6.59413695e-01 -9.59981561e-01 1.96791887e-01
1.44678783e+00 -6.51966095e-01 7.56180286e-01 9.59463656e-01
5.73122263e-01 -7.90539920e-01 -1.12449944e+00 3.33590925e-01
-9.34971273e-01 -1.10229456e+00 -4.91537422e-01 4.52845901e-01
8.62136126e-01 -3.16055834e-01 -2.65491277e-01 1.97497070e-01
1.38148856e+00 -8.36601615e-01 1.21302724e+00 3.25458378e-01
5.99467874e-01 -1.11587560e+00 5.78686059e-01 8.05153131e-01
-9.45419014e-01 -4.77320582e-01 -1.55081421e-01 -6.03844762e-01
5.36425710e-01 2.29699567e-01 -9.39030349e-01 5.97717524e-01
3.55552465e-01 -2.92742312e-01 1.66982383e-01 1.08437884e+00
-7.40858257e-01 7.76476800e-01 -6.31645083e-01 -4.82152969e-01
5.95110953e-01 -3.75196129e-01 8.13221693e-01 5.38494170e-01
2.27093354e-01 4.22799081e-01 5.89694560e-01 7.70013332e-01
7.10708678e-01 -6.41224265e-01 -5.86356699e-01 -2.12811127e-01
5.08694410e-01 7.53801584e-01 -1.16679716e+00 -2.33932927e-01
1.00778975e-01 6.55767322e-01 -3.93796563e-02 3.31286460e-01
-8.38716209e-01 -3.10368165e-02 6.13412797e-01 -2.56917328e-01
3.71868700e-01 -3.58316749e-01 -2.20009148e-01 -6.58904433e-01
-2.64405310e-01 -8.84265721e-01 4.46478933e-01 -4.49502110e-01
-6.41830742e-01 3.41663927e-01 3.75421941e-01 -1.07044470e+00
-8.63807917e-01 -2.66759042e-02 -5.25830209e-01 7.14281321e-01
-1.00745797e+00 -4.06593800e-01 2.80975193e-01 6.73936844e-01
2.22435698e-01 2.71015584e-01 7.92929769e-01 -8.61437976e-01
-4.27826494e-01 -1.85214117e-01 2.01249763e-01 -6.79530740e-01
-2.10265607e-01 -1.48823273e+00 2.60248363e-01 1.10433078e+00
-1.98501199e-01 3.28948677e-01 1.33395767e+00 -6.16779327e-01
-1.81129789e+00 -6.81049705e-01 4.08312738e-01 -6.80531710e-02
6.35617495e-01 1.23588182e-03 -3.27157140e-01 8.60680759e-01
4.28513661e-02 -5.52958965e-01 4.31335308e-02 -2.16559410e-01
4.55346525e-01 1.64789364e-01 -1.21662366e+00 1.00013566e+00
8.19512248e-01 -4.05649990e-02 -6.43025815e-01 1.10243917e-01
6.69200957e-01 -4.64990526e-01 -4.36129123e-01 -6.60577370e-03
2.14908391e-01 -8.55511010e-01 4.31110263e-01 -7.79726267e-01
9.79770645e-02 -5.85899115e-01 -2.37622987e-02 -1.18648088e+00
-2.11789042e-01 -1.50188196e+00 -2.55075336e-01 5.43643594e-01
2.73736060e-01 -6.37470961e-01 7.55862772e-01 8.34450126e-01
2.11739704e-01 -5.82302988e-01 -1.45154798e+00 -1.19213772e+00
2.14813389e-02 -4.37241822e-01 8.65363121e-01 1.58821598e-01
4.97278690e-01 -2.16839779e-02 -2.41792813e-01 5.03518224e-01
5.72511375e-01 4.35423076e-01 5.13747811e-01 -8.05854976e-01
-7.48271942e-01 -1.22533269e-01 8.04782212e-02 -9.21425819e-01
5.95959798e-02 -2.56478488e-01 6.26459241e-01 -1.56468511e+00
1.47930905e-01 -3.79119635e-01 -2.45449871e-01 5.01477659e-01
2.86808103e-01 -8.89688671e-01 4.59285170e-01 1.48447230e-01
-1.10670424e+00 4.08455819e-01 1.33175361e+00 2.01326638e-01
-5.44685960e-01 7.22907782e-01 -4.83286828e-01 5.80210984e-01
9.11932826e-01 -6.61940098e-01 -5.25840282e-01 2.51911134e-01
4.81885254e-01 1.28474569e+00 -5.41273132e-02 -9.74470317e-01
1.77075744e-01 -1.21300566e+00 -5.64501226e-01 -5.53166091e-01
7.80729726e-02 -8.87430549e-01 5.59429646e-01 1.20284665e+00
-5.95547259e-01 5.86106777e-02 3.57841663e-02 8.67374420e-01
2.01399416e-01 -7.03350961e-01 4.71523046e-01 -2.48070046e-01
-7.83250332e-01 1.84207425e-01 -1.14102650e+00 -4.92846556e-02
1.45866287e+00 -1.31286472e-01 1.47426113e-01 -7.78874457e-01
-9.16613460e-01 4.05158788e-01 4.32453036e-01 -1.51398689e-01
4.07982469e-01 -7.40994096e-01 -2.53501534e-01 -1.82670057e-01
-2.11570784e-01 -1.23922676e-01 -1.21554442e-01 7.10124016e-01
-1.31344706e-01 4.81859624e-01 -3.82941484e-01 -2.41845444e-01
-1.10912061e+00 8.15925956e-01 3.32242578e-01 -7.99824774e-01
-6.76603615e-01 6.23667315e-02 -1.59000441e-01 -2.87566148e-02
6.29419237e-02 -4.77977067e-01 1.19273644e-02 -6.03820980e-01
6.81542814e-01 6.16469622e-01 -1.82865500e-01 1.50780529e-01
-3.35824162e-01 -2.95071334e-01 2.83807427e-01 -1.15893853e+00
9.55979288e-01 -3.25519800e-01 4.93921302e-02 5.13807356e-01
1.44894734e-01 -2.88403302e-01 -1.71510911e+00 -2.17783645e-01
3.33653897e-01 -1.59150779e-01 -6.24481477e-02 -6.86001718e-01
-2.88367480e-01 4.85144913e-01 3.24208438e-01 6.56954050e-01
9.31358874e-01 -2.85421964e-02 2.73601294e-01 6.96553946e-01
1.21972752e+00 -1.22444916e+00 -3.79092753e-01 6.61437273e-01
2.44549856e-01 -4.66969460e-01 -6.44973442e-02 3.42245437e-02
-8.56821835e-01 9.48041677e-01 2.93183446e-01 -2.61725243e-02
1.69354454e-01 6.86979771e-01 -5.59652746e-01 1.88341200e-01
-1.19528234e+00 -5.53019464e-01 -5.32770872e-01 5.62154233e-01
-6.88029230e-01 6.91715419e-01 -4.23662186e-01 5.30900121e-01
-7.00537637e-02 3.76790941e-01 1.01597786e+00 1.62392759e+00
-8.15724850e-01 -9.61404562e-01 -6.34597480e-01 1.23958953e-01
-4.02390808e-01 3.74525607e-01 -3.83078963e-01 6.02598250e-01
-4.99716341e-01 1.31845093e+00 -2.50427991e-01 -6.84766173e-02
1.31453842e-01 4.48284559e-02 6.92897856e-01 -7.61624455e-01
-1.75231621e-01 4.69752438e-02 3.22430879e-01 -9.17684197e-01
-1.27225786e-01 -9.94328558e-01 -1.46144378e+00 -4.28597897e-01
-1.81293309e-01 2.73716986e-01 5.55592895e-01 1.37880743e+00
-1.15719803e-01 3.36444527e-01 7.68151581e-01 -3.13897699e-01
-1.45515263e+00 -5.73305011e-01 -7.43823230e-01 -4.20505077e-01
1.77064046e-01 -3.70081931e-01 -2.85202235e-01 -4.75501657e-01] | [4.347409725189209, 2.1167449951171875] |
4e6206d9-6bb1-47c6-a21c-7d5a1fd30568 | generative-modeling-of-high-resolution-global | 2210.12504 | null | https://arxiv.org/abs/2210.12504v1 | https://arxiv.org/pdf/2210.12504v1.pdf | Generative Modeling of High-resolution Global Precipitation Forecasts | Forecasting global precipitation patterns and, in particular, extreme precipitation events is of critical importance to preparing for and adapting to climate change. Making accurate high-resolution precipitation forecasts using traditional physical models remains a major challenge in operational weather forecasting as they incur substantial computational costs and struggle to achieve sufficient forecast skill. Recently, deep-learning-based models have shown great promise in closing the gap with numerical weather prediction (NWP) models in terms of precipitation forecast skill, opening up exciting new avenues for precipitation modeling. However, it is challenging for these deep learning models to fully resolve the fine-scale structures of precipitation phenomena and adequately characterize the extremes of the long-tailed precipitation distribution. In this work, we present several improvements to the architecture and training process of a current state-of-the art deep learning precipitation model (FourCastNet) using a novel generative adversarial network (GAN) to better capture fine scales and extremes. Our improvements achieve superior performance in capturing the extreme percentiles of global precipitation, while comparable to state-of-the-art NWP models in terms of forecast skill at 1--2 day lead times. Together, these improvements set a new state-of-the-art in global precipitation forecasting. | ['Peter Harrington', 'Shashank Subramanian', 'James Duncan'] | 2022-10-22 | null | null | null | null | ['weather-forecasting'] | ['miscellaneous'] | [-4.06316638e-01 -2.22696483e-01 2.90379137e-01 -4.81155217e-01
-6.37239873e-01 -5.23190200e-01 7.44624794e-01 -3.43777649e-02
1.63072664e-02 1.24610877e+00 2.33714536e-01 -7.92955697e-01
2.57772833e-01 -1.29784596e+00 -5.41713953e-01 -9.29238796e-01
-4.20409709e-01 5.44933796e-01 -2.91248679e-01 -8.26504350e-01
-2.32448608e-01 9.13040519e-01 -1.30413127e+00 -3.70727271e-01
1.17507386e+00 6.92710221e-01 3.66571057e-03 8.25448751e-01
-4.15274054e-01 5.86688817e-01 -4.22897398e-01 2.28439152e-01
2.81092376e-01 -3.97038490e-01 -3.83409441e-01 -3.67093682e-01
6.06438518e-01 -4.56102759e-01 -1.14782162e-01 6.19991422e-01
5.96404254e-01 2.85252273e-01 7.24297583e-01 -8.51193190e-01
-5.27244627e-01 2.93083251e-01 -6.16406083e-01 4.43271697e-01
-3.93615007e-01 3.45668852e-01 7.71138489e-01 -8.04418802e-01
9.24069658e-02 1.03464592e+00 1.09617352e+00 3.66359949e-01
-1.28354275e+00 -7.39088237e-01 2.46912524e-01 -4.87044990e-01
-1.03138900e+00 -3.55663002e-01 3.27584922e-01 -8.26293707e-01
9.87559080e-01 2.66589820e-01 7.76999891e-01 7.80554354e-01
5.96496820e-01 8.73573199e-02 1.42917287e+00 2.40527317e-02
2.71847874e-01 -3.17990094e-01 -5.29608410e-03 2.34240815e-01
1.74711928e-01 5.96110225e-01 -2.33642235e-01 -2.13118032e-01
6.86203480e-01 8.22649747e-02 -3.16882938e-01 4.88786876e-01
-6.00766659e-01 1.21124828e+00 7.61290491e-01 8.58268514e-02
-8.02513540e-01 1.83940098e-01 -1.29890993e-01 2.13282734e-01
1.32526183e+00 5.03707230e-01 -6.47894800e-01 -1.08964145e-01
-1.63246799e+00 9.73092735e-01 7.54765391e-01 1.87620610e-01
7.46695995e-01 9.83329654e-01 -7.98286051e-02 5.20329654e-01
3.08636874e-01 1.55467331e+00 -2.00020634e-02 -6.97720349e-01
2.62597114e-01 -4.31402810e-02 6.99265361e-01 -1.04279780e+00
-7.61207104e-01 -7.69354284e-01 -1.24418700e+00 7.24167407e-01
2.97086656e-01 -7.59524584e-01 -1.12069798e+00 1.66373515e+00
2.35131204e-01 3.92420799e-01 2.10668966e-01 1.01666093e+00
6.37676358e-01 1.41454470e+00 3.76386881e-01 6.56750351e-02
1.08596039e+00 -7.81978846e-01 -5.38336635e-01 -6.04073763e-01
3.70222740e-02 -7.08659768e-01 8.16236079e-01 -4.58568573e-01
-8.25637460e-01 -6.94875002e-01 -8.77296329e-01 3.81990314e-01
-6.47791266e-01 -2.95848995e-01 6.68332338e-01 3.65224868e-01
-1.22667921e+00 9.04228091e-01 -8.76720905e-01 -1.92624554e-01
6.58194646e-02 -1.43160015e-01 9.12573710e-02 4.37213242e-01
-1.78009999e+00 9.82195973e-01 -1.01415902e-01 5.23010850e-01
-9.08129871e-01 -1.38427615e+00 -9.64392602e-01 5.41630983e-01
-7.11252511e-01 -6.70731723e-01 1.11954737e+00 -5.06899595e-01
-1.41223538e+00 4.51281011e-01 -2.61343181e-01 -7.26674974e-01
2.98260510e-01 -2.36827195e-01 -6.87597811e-01 -2.26501003e-01
1.12855203e-01 8.00180852e-01 8.44411433e-01 -1.21810818e+00
-5.96650898e-01 -9.35737118e-02 -3.57769877e-01 2.57996261e-01
3.02535594e-01 -1.23562969e-01 5.41052520e-01 -7.38163054e-01
-2.85693794e-01 -1.05790162e+00 -6.17588401e-01 -2.60135736e-02
1.40013754e-01 2.19363958e-01 8.59928250e-01 -8.98235202e-01
7.79668093e-01 -1.95288539e+00 -2.77720034e-01 1.64469644e-01
2.88543820e-01 5.25514781e-01 -8.70008618e-02 5.09725332e-01
6.78953528e-02 2.96665430e-01 -6.54091895e-01 -6.13833070e-01
8.55972171e-02 5.33165336e-01 -1.37454331e+00 4.15524036e-01
3.99011850e-01 8.46007526e-01 -5.86330533e-01 2.61251211e-01
4.70344961e-01 6.26930833e-01 -2.97256857e-02 6.86941683e-01
-6.01275861e-01 1.01808274e+00 -2.17217088e-01 2.42616117e-01
1.43489814e+00 -2.13911176e-01 -2.06416681e-01 7.28335738e-01
-4.96880770e-01 2.24722728e-01 -8.72642159e-01 6.85161114e-01
-6.63067043e-01 6.98006511e-01 2.41556004e-01 -5.12851000e-01
1.19193661e+00 2.22560331e-01 -4.78693247e-02 -6.64126456e-01
-3.33600432e-01 3.30562413e-01 -1.37387455e-01 -1.96597323e-01
6.90681815e-01 -6.85638368e-01 3.12206727e-02 4.42045033e-01
-4.08270866e-01 -6.47684753e-01 -1.69281736e-01 -1.74169049e-01
1.85664237e-01 5.22790011e-03 -1.05823144e-01 -6.66495085e-01
1.33513764e-01 3.57998908e-02 4.89202142e-01 9.05406713e-01
-2.04375256e-02 8.86564672e-01 2.90682346e-01 -1.13120830e+00
-1.09430826e+00 -1.10925043e+00 -4.07696873e-01 8.70211184e-01
-4.54584301e-01 1.18495815e-01 -2.65105218e-01 -1.31939575e-01
5.87536573e-01 7.02497661e-01 -8.52307975e-01 4.16686058e-01
-3.05681109e-01 -1.35475564e+00 8.31099868e-01 5.87863445e-01
6.16955161e-01 -1.08254409e+00 -4.51651305e-01 4.21613306e-01
-3.81146297e-02 -9.11784768e-01 1.90235913e-01 1.54761836e-01
-9.47243869e-01 -5.02276003e-01 -9.28690612e-01 -1.60684809e-01
1.85960025e-01 -9.12554786e-02 1.65332818e+00 -2.03526050e-01
2.43802458e-01 -2.24135891e-01 -5.20959198e-02 -8.29147220e-01
-3.40673685e-01 2.50196189e-01 1.84364785e-02 -4.10000414e-01
1.70406982e-01 -8.01826477e-01 -7.82314062e-01 -1.26605913e-01
-7.08720446e-01 -7.70050436e-02 1.23513855e-01 6.44473433e-01
4.23902303e-01 -1.59939930e-01 7.25232244e-01 -7.51125276e-01
6.35070145e-01 -9.64293718e-01 -1.01437759e+00 -2.13713363e-01
-6.70584977e-01 3.60159948e-02 8.64938557e-01 3.47438127e-01
-1.36715770e+00 -3.72616112e-01 -6.03795230e-01 -1.31352425e-01
-4.43955451e-01 1.00198972e+00 7.53912151e-01 -1.25521138e-01
5.54629683e-01 3.56105506e-01 -8.38348493e-02 -5.26083052e-01
2.29315475e-01 3.65698636e-01 4.75972146e-01 -5.19420207e-01
9.18314040e-01 5.53826571e-01 2.94865817e-01 -8.57717633e-01
-1.10858285e+00 -1.60716206e-01 -1.65878385e-01 -8.66007954e-02
8.59563470e-01 -1.39076483e+00 -2.23950818e-01 1.00569880e+00
-8.20441186e-01 -7.48294294e-01 -2.15940803e-01 2.28626832e-01
-1.14771426e-02 -9.43513513e-02 -6.22616053e-01 -8.50465715e-01
-8.26343060e-01 -7.34667957e-01 1.01198542e+00 6.07537568e-01
-1.59307599e-01 -1.42233407e+00 8.79395247e-01 -3.43445957e-01
1.45341170e+00 7.58220434e-01 7.93161213e-01 2.06165090e-02
-2.34191313e-01 2.98239067e-02 -2.93402344e-01 2.81999651e-02
9.27393213e-02 1.99743330e-01 -1.22890723e+00 -2.72506177e-01
-5.83282001e-02 -2.59303153e-01 1.21329618e+00 8.51846635e-01
8.18960130e-01 -1.98483810e-01 -6.03766739e-02 1.10268545e+00
1.41072202e+00 -1.80588126e-01 5.54748535e-01 3.30043763e-01
4.85960871e-01 1.89300328e-01 5.39824441e-02 7.61187613e-01
3.90003592e-01 4.08740602e-02 3.99759650e-01 -7.51298130e-01
9.46070626e-02 -6.05768748e-02 -6.14172481e-02 3.88494194e-01
-1.05059341e-01 -1.98586673e-01 -1.36571360e+00 6.77017212e-01
-1.58763289e+00 -1.12774110e+00 -3.43298227e-01 1.91232824e+00
7.26458251e-01 -3.92430164e-02 -3.87124270e-01 -5.45892537e-01
2.37010583e-01 1.06915355e+00 -6.06269360e-01 -7.88286507e-01
-3.81850243e-01 7.09585249e-01 8.11356246e-01 8.52282643e-01
-1.41592944e+00 9.14035559e-01 7.11856365e+00 1.88823089e-01
-1.87549794e+00 7.50189275e-02 9.66534674e-01 1.89262941e-01
-3.95231605e-01 -2.20057756e-01 -1.17259693e+00 5.58009923e-01
1.48019075e+00 1.30740866e-01 3.47815871e-01 7.12225258e-01
7.85439193e-01 -1.63606524e-01 -3.52432430e-01 3.11916918e-01
-5.01520932e-01 -1.74098480e+00 8.24627057e-02 -1.11657724e-01
1.28033984e+00 7.61392832e-01 2.64925987e-01 5.02948105e-01
7.99979329e-01 -1.50714695e+00 2.82891691e-01 8.75293732e-01
7.42902339e-01 -7.32497156e-01 8.47275853e-01 2.19617799e-01
-1.31838822e+00 3.57531160e-01 -7.47923434e-01 -7.49936759e-01
3.32668394e-01 1.24828899e+00 -2.73956031e-01 4.03048724e-01
8.37018967e-01 7.59727836e-01 -1.15964860e-01 6.32555544e-01
-4.01792765e-01 1.27665269e+00 -5.44960856e-01 5.47742784e-01
7.99813628e-01 -2.11634457e-01 3.95019203e-01 1.21138060e+00
5.25930703e-01 3.27959210e-01 2.29498178e-01 1.03707516e+00
-1.08608216e-01 -2.94672251e-01 -6.10441148e-01 2.60772854e-01
3.55669320e-01 1.09500515e+00 -2.03775719e-01 -4.41784501e-01
-2.88071454e-01 4.06319559e-01 3.45330834e-01 6.09583318e-01
-8.10464323e-01 -7.84254447e-02 1.51944280e+00 -9.52860788e-02
4.33491826e-01 -5.83652377e-01 -5.97585499e-01 -1.12752426e+00
-3.05416763e-01 -6.41868114e-01 7.66205639e-02 -8.23752224e-01
-1.31775105e+00 6.99085236e-01 -3.70235860e-01 -1.01835287e+00
-3.60631317e-01 -4.37704593e-01 -1.26319063e+00 1.94758916e+00
-2.41221094e+00 -1.10707557e+00 -5.31394482e-01 1.99660793e-01
1.80909216e-01 1.18469901e-01 1.28326094e+00 -1.37622267e-01
-1.76908076e-01 1.99964885e-02 9.73546207e-01 -8.62328485e-02
7.21460581e-01 -1.39389873e+00 1.25842047e+00 9.40445960e-01
-4.39149559e-01 3.13447982e-01 9.13868368e-01 -6.59389377e-01
-9.66698289e-01 -1.66766393e+00 1.16171193e+00 -7.02742636e-02
5.96100688e-01 -1.08239606e-01 -1.14763474e+00 6.56575501e-01
3.40034723e-01 2.94353783e-01 4.70081359e-01 2.93056071e-01
-2.26898476e-01 -2.54454941e-01 -1.05330288e+00 3.92993391e-01
3.90473306e-02 -4.48781341e-01 -5.54666460e-01 4.03271765e-01
4.89588231e-01 -6.85478330e-01 -8.53610933e-01 6.22438252e-01
4.86868203e-01 -8.27633739e-01 6.14823580e-01 -6.42413259e-01
9.59704101e-01 -3.52464050e-01 4.61879745e-02 -2.13410306e+00
-6.55407786e-01 -3.36810261e-01 -6.43874258e-02 9.36780035e-01
5.01383364e-01 -9.08158123e-01 5.66313148e-01 7.27672100e-01
-1.41570820e-02 -5.61332107e-01 -8.39550495e-01 -5.39040446e-01
9.33900416e-01 -7.29196072e-02 9.53212917e-01 1.07485056e+00
-4.78011221e-01 -1.74713567e-01 -6.97970986e-01 8.80851269e-01
6.02598488e-01 7.64680922e-01 5.34226179e-01 -1.43299282e+00
-8.21948238e-03 -4.26906258e-01 1.52826741e-01 -7.51873374e-01
2.83913404e-01 -3.62555593e-01 5.40081598e-02 -1.55085027e+00
-3.20611566e-01 -6.05561674e-01 -2.89894730e-01 5.61645150e-01
-5.17435253e-01 3.12856674e-01 7.49034286e-02 3.13760698e-01
3.72590661e-01 9.71587002e-01 9.59060788e-01 -1.01342894e-01
-3.49360779e-02 -1.00265101e-01 -2.78071553e-01 4.84364450e-01
1.10948515e+00 -4.80342925e-01 4.97353449e-02 -6.98914886e-01
2.71079093e-01 2.18822911e-01 2.36240283e-01 -1.32207656e+00
-2.08833978e-01 -5.14945090e-01 9.24294889e-01 -5.59056818e-01
1.16958573e-01 -3.96136820e-01 2.15239555e-01 3.54782075e-01
-8.03303421e-02 2.41019070e-01 9.08757865e-01 1.88402280e-01
-4.00246471e-01 4.06311333e-01 9.78882074e-01 -1.99292302e-01
-7.94578433e-01 4.86639291e-01 -9.47517395e-01 1.97531044e-01
6.33984327e-01 5.39600551e-01 -6.09061062e-01 -6.49049461e-01
-7.06823766e-01 5.75662136e-01 3.95223111e-01 2.51833200e-01
1.85625911e-01 -8.25013041e-01 -1.33412814e+00 5.60734212e-01
-2.24662051e-01 -1.86640881e-02 5.55470884e-01 2.67024815e-01
-9.84514654e-01 3.71430397e-01 -3.47149909e-01 -3.26835752e-01
-5.28886139e-01 -8.34073406e-03 1.01735866e+00 -5.96223533e-01
-6.71185374e-01 8.54049206e-01 2.26492256e-01 -9.34813082e-01
-5.01349270e-01 -4.71622407e-01 -1.59562066e-01 4.20917198e-02
7.01303899e-01 5.01753129e-02 1.56577066e-01 -3.81406039e-01
-7.24595487e-02 5.05313337e-01 3.16331983e-01 -2.58907359e-02
1.55142283e+00 -1.50120571e-01 -9.67228040e-02 4.71149325e-01
6.45298839e-01 -1.08043335e-01 -1.70927334e+00 -5.94747625e-02
-6.77571356e-01 -2.31413543e-01 4.60659534e-01 -1.05497813e+00
-1.46159124e+00 1.12353289e+00 6.28462493e-01 9.70456377e-02
8.21750402e-01 -5.02523184e-01 1.06743181e+00 2.58433700e-01
-1.53454483e-01 -5.45937717e-01 -8.92057598e-01 1.23853934e+00
9.56606328e-01 -1.51282358e+00 -7.95957670e-02 2.43087336e-01
-4.33462083e-01 1.16457701e+00 4.06256109e-01 -3.97500008e-01
1.23788381e+00 6.73368752e-01 7.05785453e-01 -2.08546057e-01
-5.72335780e-01 -4.23737094e-02 1.61736179e-02 2.03963697e-01
4.90217209e-01 7.16900229e-01 8.21652785e-02 2.44636938e-01
-3.99369627e-01 -7.79493377e-02 4.36327785e-01 6.99659407e-01
-5.67634225e-01 -5.80573022e-01 -7.08580256e-01 4.54498649e-01
-4.31087136e-01 -8.04003894e-01 3.26939195e-01 3.91770244e-01
-2.40414187e-01 7.81956732e-01 5.11413217e-01 1.53853253e-01
-5.95858544e-02 3.14977705e-01 -3.30667615e-01 -1.86664402e-01
-6.08919322e-01 -4.13204074e-01 -7.61112943e-02 -2.15395063e-01
-2.55475402e-01 -6.32992446e-01 -9.01768267e-01 -1.04356360e+00
3.27989280e-01 4.55588490e-01 6.00124002e-01 1.09150863e+00
6.03677750e-01 4.99981046e-01 7.94509232e-01 -1.51559222e+00
-5.71847677e-01 -1.15852726e+00 -9.72802579e-01 6.33032024e-02
8.53179634e-01 -4.35470402e-01 -5.00457168e-01 -1.46490484e-01] | [6.5774712562561035, 2.950648069381714] |
22c9096d-eccd-4a60-b8ef-8885396e3789 | learning-answer-generation-using-supervision | 2305.15344 | null | https://arxiv.org/abs/2305.15344v1 | https://arxiv.org/pdf/2305.15344v1.pdf | Learning Answer Generation using Supervision from Automatic Question Answering Evaluators | Recent studies show that sentence-level extractive QA, i.e., based on Answer Sentence Selection (AS2), is outperformed by Generation-based QA (GenQA) models, which generate answers using the top-k answer sentences ranked by AS2 models (a la retrieval-augmented generation style). In this paper, we propose a novel training paradigm for GenQA using supervision from automatic QA evaluation models (GAVA). Specifically, we propose three strategies to transfer knowledge from these QA evaluation models to a GenQA model: (i) augmenting training data with answers generated by the GenQA model and labelled by GAVA (either statically, before training, or (ii) dynamically, at every training epoch); and (iii) using the GAVA score for weighting the generator loss during the learning of the GenQA model. We evaluate our proposed methods on two academic and one industrial dataset, obtaining a significant improvement in answering accuracy over the previous state of the art. | ['Alessandro Moschitti', 'Rik Koncel-Kedziorski', 'Siddhant Garg', 'Matteo Gabburo'] | 2023-05-24 | null | null | null | null | ['answer-generation'] | ['natural-language-processing'] | [ 4.12111670e-01 4.65129822e-01 5.00291169e-01 -3.65236938e-01
-1.70313466e+00 -8.03667128e-01 7.56876171e-01 1.24263786e-01
-2.60856867e-01 1.13565052e+00 3.07932436e-01 -4.98113722e-01
1.82477962e-02 -1.19783688e+00 -7.06780255e-01 -6.01200402e-01
5.14880121e-01 1.23761714e+00 3.12814415e-01 -6.59915388e-01
3.44005644e-01 -9.46727023e-03 -1.45567966e+00 6.78242922e-01
1.65976000e+00 9.28232253e-01 2.72171181e-02 9.45480883e-01
-4.32858646e-01 1.13024879e+00 -1.01533127e+00 -9.99490499e-01
8.26700255e-02 -9.76412117e-01 -1.31967342e+00 -4.08749878e-01
5.46415687e-01 -1.32357761e-01 2.66494215e-01 8.04186165e-01
6.56892002e-01 1.51720559e-02 6.38202369e-01 -8.99336219e-01
-7.48349488e-01 6.33288741e-01 1.77615970e-01 -4.89212908e-02
6.94497049e-01 4.38905835e-01 1.18004513e+00 -8.90875459e-01
7.11732626e-01 1.31919587e+00 2.33241171e-01 9.86573935e-01
-1.06421280e+00 -1.52233362e-01 -2.39547074e-01 2.21698478e-01
-9.54929590e-01 -4.45148915e-01 7.14048564e-01 -3.01630586e-01
1.00111699e+00 3.73360693e-01 1.64340556e-01 8.22920859e-01
1.93876490e-01 9.93962169e-01 1.20488501e+00 -7.91652977e-01
3.57123315e-01 7.16395080e-02 2.92934299e-01 7.61949420e-01
-1.60707176e-01 -1.39941797e-01 -6.84647143e-01 -4.99151438e-01
-1.52282164e-01 -8.98926616e-01 -2.92930782e-01 -2.22207725e-01
-1.04187739e+00 1.15236032e+00 2.46607736e-01 -7.10785389e-02
-7.87547886e-01 -9.06324685e-02 3.61414909e-01 6.65919662e-01
3.62005681e-01 1.05888689e+00 -5.14841616e-01 2.68958490e-02
-8.32021654e-01 7.06157088e-01 9.06881630e-01 7.25430071e-01
7.81869352e-01 -4.64654900e-02 -1.03721356e+00 6.58451021e-01
2.74009138e-01 6.86845601e-01 6.52820289e-01 -8.20648670e-01
1.03525877e+00 7.55474925e-01 2.68058896e-01 -3.86020750e-01
1.48346527e-02 -4.22837555e-01 -4.60666448e-01 -1.21324286e-01
3.96767229e-01 -5.51140070e-01 -9.85918641e-01 1.78258944e+00
3.32176626e-01 -4.98787373e-01 5.84405601e-01 6.80262685e-01
1.01172733e+00 9.80764389e-01 2.35687688e-01 -2.65768200e-01
1.27181602e+00 -1.35369229e+00 -8.56429398e-01 7.39693060e-04
9.01450694e-01 -6.11357749e-01 1.37298894e+00 6.72645986e-01
-1.54484129e+00 -8.18033278e-01 -9.24223781e-01 -9.92183909e-02
-2.23343030e-01 3.53481382e-01 2.96833873e-01 8.50374818e-01
-1.38580501e+00 2.62787759e-01 -3.04696441e-01 -5.51688150e-02
-4.97899624e-03 2.56278008e-01 -4.84229550e-02 -1.47759944e-01
-1.71546197e+00 1.07683754e+00 3.42324346e-01 1.38032222e-02
-1.22677565e+00 -3.12986225e-01 -9.07020390e-01 1.49377629e-01
1.71829820e-01 -1.23192072e+00 1.52676630e+00 -8.45270336e-01
-1.94286764e+00 4.58520830e-01 -3.03978354e-01 -7.67971396e-01
3.58538806e-01 -4.88125205e-01 -1.09172821e-01 4.18162763e-01
1.71065867e-01 7.52322555e-01 7.72676945e-01 -1.39152718e+00
-3.87196660e-01 -3.30483645e-01 5.16376615e-01 5.66083312e-01
-1.99590951e-01 -1.45506352e-01 1.28405288e-01 -6.82469979e-02
-8.49951580e-02 -6.69299066e-01 -1.94974452e-01 -8.69666874e-01
-4.05227095e-01 -7.29353189e-01 1.22835964e-01 -1.02718556e+00
1.31638253e+00 -1.47933269e+00 3.82238328e-01 1.58141866e-01
-2.79467732e-01 5.81529379e-01 -5.27077734e-01 7.44889796e-01
3.49023312e-01 1.43873766e-02 -4.56813335e-01 -2.81202376e-01
8.24232027e-02 -3.48065011e-02 -6.35526180e-01 -4.79351163e-01
6.82331979e-01 1.11490333e+00 -1.22404921e+00 -5.50884962e-01
-1.40053615e-01 -6.68336824e-02 -5.17266631e-01 7.30800986e-01
-6.05339706e-01 4.93042529e-01 -4.64312285e-01 4.73601758e-01
4.80655700e-01 -1.79969743e-02 1.44175559e-01 1.45124448e-02
2.43068144e-01 9.33972955e-01 -7.58491039e-01 1.71924353e+00
-7.46449292e-01 -4.55319397e-02 -3.32420021e-01 -5.99747837e-01
1.19235539e+00 6.33364201e-01 -2.17154592e-01 -8.90675068e-01
-1.78000882e-01 4.82376069e-01 -6.69790385e-03 -3.86815637e-01
7.93882310e-01 -6.92064241e-02 -1.91867232e-01 5.53159118e-01
6.88083351e-01 -4.86945242e-01 7.46353805e-01 4.81262296e-01
1.14044738e+00 2.65341967e-01 -1.81661337e-03 -1.33645415e-01
1.25949609e+00 3.06990802e-01 1.99550569e-01 1.02500188e+00
8.69385302e-02 8.17820132e-01 4.13399786e-01 -1.19744211e-01
-6.63911402e-01 -1.28653073e+00 3.35386604e-01 9.14413214e-01
-2.46819705e-01 -2.91720301e-01 -1.25859261e+00 -1.21498168e+00
-4.35845524e-01 1.22887802e+00 -3.63971949e-01 -3.45481306e-01
-8.41495335e-01 -5.89385211e-01 5.95545888e-01 3.79168838e-01
5.98210156e-01 -1.56014788e+00 -3.94033939e-01 2.94816196e-01
-6.48846686e-01 -6.28319204e-01 -1.85325205e-01 5.92836291e-02
-9.39551890e-01 -6.73673332e-01 -6.94246948e-01 -5.34679294e-01
6.61438048e-01 -3.70386422e-01 1.69978344e+00 1.53842703e-01
4.89597052e-01 4.05597746e-01 -7.20082521e-01 -3.64176720e-01
-8.65931630e-01 5.28214633e-01 -3.33833963e-01 -1.11556880e-03
1.97768107e-01 4.08930564e-03 -4.15050238e-01 -1.44355386e-01
-8.86287451e-01 4.03231345e-02 8.86469662e-01 9.65181410e-01
4.14441198e-01 -4.34220046e-01 1.34087157e+00 -1.19142652e+00
1.31913829e+00 -3.87719810e-01 -3.75245124e-01 7.44696617e-01
-6.66055620e-01 2.96144158e-01 8.20335329e-01 1.64704248e-01
-1.56063890e+00 -4.57618982e-02 -6.61763072e-01 2.14351133e-01
-1.31872624e-01 6.40225053e-01 -3.50671321e-01 1.46499142e-01
9.74614203e-01 3.71592075e-01 -2.29488358e-01 -2.06550032e-01
6.47758722e-01 6.38700008e-01 4.28038567e-01 -8.51900160e-01
8.40413213e-01 -1.04087509e-01 -2.57837564e-01 -2.98615769e-02
-1.11993885e+00 -2.35732734e-01 -5.33520460e-01 -1.30425885e-01
9.84822512e-01 -6.73779428e-01 -2.73551166e-01 2.03549579e-01
-1.42977095e+00 -2.10379198e-01 -4.95994002e-01 2.86619276e-01
-7.14863122e-01 1.39160708e-01 -7.79172838e-01 -1.04273677e+00
-1.12208581e+00 -1.08076966e+00 1.12486625e+00 3.66494805e-01
-1.63822174e-01 -8.17311227e-01 5.29486358e-01 9.83411908e-01
4.89960164e-01 -9.88702998e-02 1.25442350e+00 -8.32436562e-01
-5.01534462e-01 -1.94363743e-01 1.43374816e-01 5.77623963e-01
-5.21473736e-02 -3.93431664e-01 -1.05890191e+00 -7.37944171e-02
8.00907537e-02 -9.04720843e-01 6.66661441e-01 -8.70631561e-02
7.36455739e-01 -3.69060904e-01 2.59196341e-01 -2.05124855e-01
1.20054924e+00 2.51793087e-01 8.12413931e-01 2.68065959e-01
2.77509570e-01 7.60020614e-01 8.61409128e-01 -1.96136668e-01
5.28511465e-01 5.41592538e-01 3.67533565e-01 2.50584334e-01
-2.77082771e-01 -4.81809646e-01 6.38902664e-01 1.19329941e+00
4.12717015e-02 -6.51169002e-01 -7.95937479e-01 8.08851361e-01
-1.76744735e+00 -7.05412984e-01 -3.65917176e-01 2.28341937e+00
1.13810861e+00 1.04908817e-01 -1.23541564e-01 9.01079252e-02
3.18598926e-01 -1.21871814e-01 -6.74594939e-02 -8.75743687e-01
-2.52525777e-01 8.73463511e-01 -2.47311518e-01 7.48302162e-01
-8.10810566e-01 1.00564301e+00 5.54134989e+00 7.96579421e-01
-5.40391326e-01 1.98983967e-01 5.99436283e-01 4.87940758e-01
-7.39216864e-01 1.71752140e-01 -7.77338624e-01 2.38564059e-01
1.27259469e+00 -2.20940188e-01 1.92938477e-01 6.89022601e-01
-2.22327977e-01 -1.58491340e-02 -9.98708367e-01 2.31229097e-01
4.34266239e-01 -1.18897426e+00 7.97651291e-01 -5.00457346e-01
9.49973583e-01 -4.63163882e-01 3.41927707e-02 8.67610216e-01
4.85482574e-01 -8.77154589e-01 5.27987421e-01 5.43732345e-01
2.75060892e-01 -7.42200434e-01 1.37973762e+00 4.02789801e-01
-6.69668794e-01 1.01260252e-01 -3.26674938e-01 5.81297651e-02
5.46421528e-01 3.89369130e-01 -9.98388588e-01 1.27200937e+00
2.85271645e-01 -2.42888093e-01 -8.21442902e-01 8.72373998e-01
-9.36937690e-01 1.09376216e+00 2.08798409e-01 -1.99742347e-01
3.70524138e-01 -2.36283630e-01 3.79422605e-01 8.83623719e-01
3.33252192e-01 6.86357841e-02 -3.05141062e-01 8.24714839e-01
-2.04087600e-01 3.48113596e-01 -2.90336728e-01 1.39190197e-01
3.29472393e-01 1.30676115e+00 -5.09408005e-02 -5.25208890e-01
-6.07195683e-02 1.12666941e+00 4.85015273e-01 2.14935005e-01
-3.83468717e-01 -5.25104463e-01 -2.30229169e-01 -2.38447428e-01
2.26955220e-01 1.47918001e-01 -1.23448074e-01 -9.24983799e-01
1.91613466e-01 -1.13666356e+00 6.33964479e-01 -1.09675503e+00
-1.44041312e+00 1.00624931e+00 -7.66474456e-02 -1.05485880e+00
-9.26386476e-01 -3.46701682e-01 -7.06889689e-01 1.24948704e+00
-1.62574446e+00 -1.02755070e+00 -7.99046084e-03 4.37148452e-01
6.61907613e-01 -1.45966023e-01 1.08450305e+00 2.20269039e-01
2.09260806e-02 7.12960243e-01 -2.66642123e-01 -1.00753615e-02
6.16847575e-01 -1.54999411e+00 4.41404909e-01 7.96602309e-01
3.58202189e-01 6.48358524e-01 5.32087147e-01 -5.01547992e-01
-1.23120821e+00 -1.00876296e+00 1.70005035e+00 -9.11702275e-01
2.17206001e-01 -3.09517801e-01 -8.67949069e-01 3.71464610e-01
6.65195405e-01 -7.13042021e-01 5.22029877e-01 -1.14464372e-01
3.89667191e-02 -1.31029531e-01 -1.36775935e+00 4.01460469e-01
5.38625479e-01 -3.64545733e-01 -1.05119967e+00 5.26666105e-01
9.77551699e-01 -4.57840502e-01 -5.48368871e-01 6.67208016e-01
5.62781170e-02 -5.94586670e-01 7.19814181e-01 -8.21884453e-01
6.61961019e-01 -5.24522424e-01 -7.66170397e-03 -1.65230274e+00
3.32550034e-02 -4.84518945e-01 -3.65365654e-01 1.32669258e+00
8.05427015e-01 -4.76063758e-01 6.43103361e-01 3.58407021e-01
-4.00101095e-01 -8.40768397e-01 -9.06344354e-01 -5.80826581e-01
3.81752253e-01 8.38330835e-02 8.28970790e-01 4.54404175e-01
-3.66522044e-01 8.80931854e-01 -1.65586337e-01 5.48848324e-02
2.65797615e-01 1.59445181e-01 7.75101066e-01 -9.72464204e-01
-3.23867470e-01 1.98158727e-04 -2.15566717e-02 -9.36949253e-01
-5.05671538e-02 -7.30749726e-01 3.89185220e-01 -1.85211599e+00
-4.14886745e-03 -2.77717680e-01 -1.68879047e-01 1.53454825e-01
-7.14538217e-01 2.32495964e-01 2.47395486e-01 -3.16497907e-02
-7.67449141e-01 8.50580513e-01 1.28932345e+00 -8.60651731e-02
-1.13567360e-01 9.88077596e-02 -6.04712427e-01 2.21705243e-01
7.85543740e-01 -4.18328792e-01 -7.25775659e-01 -7.35000789e-01
4.87707287e-01 3.47574383e-01 2.69280393e-02 -1.05756259e+00
1.54699266e-01 1.93322822e-01 3.57035398e-02 -6.94015682e-01
3.18466783e-01 -1.41368315e-01 -4.29199785e-01 4.37765926e-01
-7.03981578e-01 3.39099765e-01 -1.92151871e-02 2.27086261e-01
-5.50932944e-01 -8.48245382e-01 1.90537214e-01 -3.14672410e-01
-2.32340515e-01 -1.43411905e-01 -2.67317384e-01 3.44065160e-01
3.96701872e-01 2.71654814e-01 -6.82515383e-01 -6.40922070e-01
-5.38316250e-01 4.02544141e-01 -1.68087289e-01 1.41644403e-01
6.56650722e-01 -1.30140209e+00 -1.14854825e+00 -1.56760320e-01
2.25507274e-01 1.44571051e-01 4.28694099e-01 4.95924622e-01
-5.05977631e-01 7.38353014e-01 6.37783334e-02 -3.24897140e-01
-9.82718110e-01 2.23592743e-01 3.71200949e-01 -1.18849432e+00
1.64755195e-01 9.92956579e-01 -3.83239985e-02 -1.14304972e+00
-1.77475169e-01 2.22714245e-01 -6.81612730e-01 -1.62741482e-01
3.30485791e-01 1.61445618e-01 6.78914487e-01 -4.40000176e-01
-4.69293725e-03 1.47233784e-01 -9.72158909e-02 -5.99095225e-01
7.33144403e-01 1.38737500e-01 -1.42607465e-01 1.83001295e-01
5.54582119e-01 1.01575166e-01 -5.73330045e-01 -1.52239874e-01
1.17481813e-01 -9.09409374e-02 -4.73775357e-01 -1.45472240e+00
-5.86472273e-01 1.10185409e+00 4.55669165e-01 2.07390040e-01
1.18337440e+00 -2.36572400e-02 1.01632357e+00 6.72753870e-01
5.73413908e-01 -1.01634526e+00 3.07637364e-01 6.02830648e-01
1.07548797e+00 -9.03098106e-01 -4.10640895e-01 -1.76652387e-01
-1.00715613e+00 9.44546998e-01 9.80356574e-01 -1.23572268e-01
-1.41202614e-01 -6.12997174e-01 3.86266440e-01 -2.75299549e-01
-1.09787595e+00 -3.25943738e-01 4.77204859e-01 5.47331214e-01
5.57619214e-01 -3.36586349e-02 -8.58209789e-01 6.67464256e-01
-5.64761281e-01 -9.06122550e-02 3.85226011e-01 8.57779801e-01
-3.50244105e-01 -1.50951672e+00 -2.38769144e-01 4.97383833e-01
-2.86349475e-01 -4.62898791e-01 -7.96711266e-01 4.24773335e-01
6.46408424e-02 1.49513674e+00 -3.22751850e-01 -3.62909973e-01
4.95934635e-01 6.59977376e-01 5.61390519e-01 -9.24802899e-01
-1.22444868e+00 -7.65135169e-01 5.98002672e-01 -2.00180694e-01
-2.05249652e-01 -6.39005601e-01 -9.50384438e-01 3.89826596e-01
-5.59735656e-01 9.91661966e-01 5.57489157e-01 1.12731922e+00
4.23686951e-01 4.91348177e-01 6.82937682e-01 -1.25561729e-01
-1.15949512e+00 -1.37539899e+00 1.10559717e-01 4.74590659e-01
8.62419698e-03 -2.70749718e-01 -3.80835146e-01 -4.31390144e-02] | [11.47999095916748, 8.208985328674316] |
a67d7066-d1cd-4f9c-bc56-e9fdda5d6bc4 | cyberloc-towards-accurate-long-term-visual | 2301.02403 | null | https://arxiv.org/abs/2301.02403v1 | https://arxiv.org/pdf/2301.02403v1.pdf | CyberLoc: Towards Accurate Long-term Visual Localization | This technical report introduces CyberLoc, an image-based visual localization pipeline for robust and accurate long-term pose estimation under challenging conditions. The proposed method comprises four modules connected in a sequence. First, a mapping module is applied to build accurate 3D maps of the scene, one map for each reference sequence if there exist multiple reference sequences under different conditions. Second, a single-image-based localization pipeline (retrieval--matching--PnP) is performed to estimate 6-DoF camera poses for each query image, one for each 3D map. Third, a consensus set maximization module is proposed to filter out outlier 6-DoF camera poses, and outputs one 6-DoF camera pose for a query. Finally, a robust pose refinement module is proposed to optimize 6-DoF query poses, taking candidate global 6-DoF camera poses and their corresponding global 2D-3D matches, sparse 2D-2D feature matches between consecutive query images and SLAM poses of the query sequence as input. Experiments on the 4seasons dataset show that our method achieves high accuracy and robustness. In particular, our approach wins the localization challenge of ECCV 2022 workshop on Map-based Localization for Autonomous Driving (MLAD-ECCV2022). | ['Jiangwei Li', 'Wei Luo', 'Yuxiang Wen', 'Yangdong Liu', 'Miao Jia', 'Qichao Xu', 'Xiao Liang', 'Yukai Lin', 'Liu Liu'] | 2023-01-06 | null | null | null | null | ['image-based-localization', 'visual-localization'] | ['computer-vision', 'computer-vision'] | [-4.45755422e-01 -3.70771587e-01 -5.67083210e-02 -4.59389120e-01
-1.21992838e+00 -9.34531033e-01 5.49163222e-01 4.86913435e-02
-5.44627547e-01 1.91106483e-01 -2.74190336e-01 1.66485682e-02
-9.49907750e-02 -3.91816080e-01 -9.17444170e-01 -4.70320582e-01
-1.11300960e-01 7.98298836e-01 6.33724332e-01 -1.57362893e-01
6.64361656e-01 1.00131762e+00 -1.54363561e+00 -5.68594992e-01
4.86363143e-01 1.12343276e+00 5.86800158e-01 8.34560335e-01
2.22103268e-01 3.19425493e-01 -2.29178354e-01 -1.54783159e-01
5.86629927e-01 1.11770704e-01 -3.17832738e-01 7.99749792e-02
9.25109744e-01 -3.60936552e-01 -4.72221047e-01 1.21847022e+00
3.58401090e-01 2.44443402e-01 2.58447051e-01 -1.49631429e+00
2.83002406e-01 -3.84753257e-01 -6.22146130e-01 -1.54515132e-01
5.57690084e-01 3.29842031e-01 5.62434494e-01 -1.16561627e+00
7.42704630e-01 9.81524050e-01 8.99995685e-01 -3.20479386e-02
-8.41144621e-01 -6.01847708e-01 -1.13619283e-01 4.21747148e-01
-1.87878537e+00 -3.61761838e-01 6.11606359e-01 -5.96270204e-01
8.67182136e-01 7.57744089e-02 7.09997058e-01 5.66135228e-01
4.55990851e-01 1.12149335e-01 5.58873951e-01 -3.15161049e-02
3.13290805e-01 -1.43019184e-01 -2.09678829e-01 7.63653874e-01
2.60088950e-01 1.01598017e-01 -8.73317897e-01 -1.97326601e-01
4.76025432e-01 1.82621494e-01 7.29709119e-02 -1.03938329e+00
-1.62984681e+00 5.70759952e-01 7.18507648e-01 -1.72248021e-01
-4.31110114e-01 3.70662421e-01 7.42881000e-02 1.34984195e-01
8.07777941e-02 3.83610129e-01 -2.11596608e-01 -1.21130489e-01
-9.04475927e-01 3.80084723e-01 5.46656728e-01 1.52464664e+00
1.50061953e+00 -2.61851996e-01 1.23965785e-01 3.91691595e-01
5.21833420e-01 1.12760198e+00 6.38549700e-02 -1.28319013e+00
5.22526085e-01 4.74176049e-01 5.98869205e-01 -1.38620698e+00
-4.77168828e-01 -2.77140021e-01 -4.52403873e-01 7.52265146e-03
-1.21566109e-01 2.74082065e-01 -6.89985275e-01 1.23852897e+00
6.23577237e-01 3.00886989e-01 4.00719829e-02 1.24105513e+00
5.90743303e-01 4.70357329e-01 -6.34724617e-01 1.29547864e-01
1.16410363e+00 -9.12544429e-01 -4.01486248e-01 -9.40222502e-01
4.44810838e-01 -9.29974198e-01 2.27483645e-01 -4.42334935e-02
-4.38190043e-01 -6.19180024e-01 -1.29103458e+00 5.91324940e-02
-3.46394688e-01 4.20525253e-01 4.61499766e-02 1.69837356e-01
-1.26359046e+00 -1.20176338e-01 -8.00881863e-01 -6.55913472e-01
-1.23785742e-01 2.62260616e-01 -7.85106599e-01 -4.40595239e-01
-8.01923990e-01 1.25607157e+00 4.76442009e-01 2.87835002e-01
-1.22079086e+00 -4.23675388e-01 -1.22736502e+00 -4.95298088e-01
3.53198975e-01 -4.61124986e-01 9.79708791e-01 -2.20285073e-01
-1.05826259e+00 1.02513802e+00 -5.74817717e-01 -4.83317465e-01
4.15019125e-01 -2.54786819e-01 -2.07693443e-01 2.12888673e-01
5.85062027e-01 9.36260939e-01 7.99256384e-01 -1.46666932e+00
-7.79605567e-01 -6.48124278e-01 -3.30277860e-01 4.60473299e-01
6.26206279e-01 -3.46989483e-01 -1.17962861e+00 -7.57069467e-03
1.12126565e+00 -1.26536787e+00 -4.25048918e-01 -5.78946657e-02
-9.14232135e-02 2.35946968e-01 8.68533731e-01 -5.75417042e-01
4.93268520e-01 -2.25170445e+00 3.63075361e-02 4.36554611e-01
6.30533025e-02 -3.12726825e-01 -1.90476850e-01 5.23888409e-01
3.26544285e-01 -5.25759339e-01 5.18754423e-02 -7.42389321e-01
4.46261279e-02 3.12631935e-01 -1.93025902e-01 9.93530929e-01
2.67427545e-02 8.81371498e-01 -8.54885578e-01 -2.68439859e-01
6.48939133e-01 2.18018577e-01 -2.02798948e-01 2.45477811e-01
-1.49843711e-02 6.50597572e-01 -3.99223089e-01 7.43205905e-01
1.22055173e+00 1.75344989e-01 -2.20798910e-01 -2.97779053e-01
-6.48631752e-01 -6.07195571e-02 -1.28124547e+00 2.14125419e+00
-2.91119963e-01 7.19173372e-01 2.11158723e-01 -6.12171054e-01
1.25805449e+00 -2.28268027e-01 5.76488435e-01 -7.30069757e-01
2.26032715e-02 4.64857489e-01 -5.42721093e-01 -3.28625232e-01
9.52402294e-01 5.28471291e-01 -4.56173837e-01 -8.73762295e-02
1.04477033e-01 -8.80626082e-01 -3.96595523e-02 3.71167153e-01
1.12559080e+00 2.38480940e-01 1.04561530e-01 -2.04354718e-01
7.17563331e-01 4.04395729e-01 6.72506392e-01 7.41277814e-01
-3.70055109e-01 8.66414249e-01 1.24328230e-02 -6.08229995e-01
-1.21310353e+00 -9.94116127e-01 1.08298793e-01 4.74719614e-01
1.01552904e+00 -3.72820735e-01 -3.48121285e-01 -3.85164797e-01
2.80866206e-01 2.82699078e-01 -3.11571628e-01 -5.03783412e-02
-4.35524791e-01 -5.96781522e-02 2.28484467e-01 1.72994629e-01
6.75959647e-01 -3.20954919e-01 -1.00269055e+00 4.62861825e-03
-4.11189049e-01 -1.34722018e+00 -6.05238557e-01 2.44918600e-01
-5.82360387e-01 -1.17961955e+00 -3.53899330e-01 -8.30066085e-01
6.59392715e-01 8.47290337e-01 7.70250559e-01 -1.75654382e-01
-1.38023674e-01 5.45152009e-01 -2.44309798e-01 -1.23130813e-01
-1.21216610e-01 -2.77444929e-01 4.08776939e-01 1.06001481e-01
4.41103369e-01 -1.07226618e-01 -5.07827461e-01 7.94354200e-01
-1.02339402e-01 -1.77097797e-01 6.19652867e-01 5.27115941e-01
1.19279122e+00 -3.90575945e-01 -5.95309623e-02 5.94717450e-02
-1.94818780e-01 -3.14744323e-01 -1.42803907e+00 1.08474314e-01
-3.47063154e-01 -1.01806037e-01 5.07859811e-02 5.23383282e-02
-4.78501976e-01 9.97648716e-01 3.30755413e-02 -9.26423967e-01
-1.00250095e-01 4.45384890e-01 -1.33057564e-01 -4.95115399e-01
6.40532255e-01 5.89050651e-01 1.27357453e-01 -2.42162943e-01
3.38780135e-01 6.67288244e-01 1.20671320e+00 -1.77371889e-01
1.38013709e+00 5.98588467e-01 1.85195342e-01 -6.99444294e-01
-5.93823314e-01 -1.27754676e+00 -1.19946229e+00 -4.61135745e-01
1.10318828e+00 -1.63436317e+00 -5.81795990e-01 5.26340008e-01
-1.29023170e+00 3.06245405e-02 2.11975515e-01 8.25574040e-01
-6.81226313e-01 2.99372435e-01 1.20632388e-01 -6.78486943e-01
-2.35133134e-02 -1.62346745e+00 1.55570138e+00 2.57742614e-01
9.80582014e-02 -4.28218931e-01 3.73656541e-01 3.65817666e-01
2.96668708e-01 1.79315656e-01 -2.49325082e-01 -3.17688048e-01
-1.23139226e+00 -5.57439983e-01 -1.17643237e-01 -3.47043201e-02
-2.84187704e-01 -3.23974997e-01 -7.80728579e-01 -6.85083032e-01
-1.85626909e-01 -2.06641212e-01 5.81627667e-01 1.41528592e-01
3.73171806e-01 1.13817334e-01 -5.90268552e-01 1.00955689e+00
1.54039383e+00 1.31588608e-01 4.51374114e-01 4.73666817e-01
7.11405456e-01 2.87662417e-01 1.21661496e+00 3.26546431e-01
8.27045083e-01 8.46965492e-01 9.88007128e-01 1.56781867e-01
2.37787470e-01 -4.84669209e-01 3.31047356e-01 6.96886182e-01
4.15968746e-01 2.81855613e-01 -9.69293058e-01 6.59631729e-01
-1.96046197e+00 -5.89522302e-01 -3.67250651e-01 2.46183991e+00
-9.69552472e-02 7.56115699e-03 -2.87092030e-01 -5.45428336e-01
7.50580728e-01 3.30212384e-01 -7.00638890e-01 2.71576971e-01
-4.69839275e-02 -4.29755330e-01 1.25454831e+00 6.93803191e-01
-1.09764898e+00 1.14282346e+00 5.44591999e+00 3.58988434e-01
-9.58424330e-01 9.69319940e-02 -1.34941801e-01 2.81362921e-01
-9.84943435e-02 5.34429371e-01 -1.05050492e+00 2.34950170e-01
6.39363706e-01 6.87986314e-02 2.90716022e-01 1.12270010e+00
9.54896733e-02 -7.07040966e-01 -8.33789468e-01 1.52888703e+00
4.19287562e-01 -1.53770912e+00 -4.98995811e-01 1.74357951e-01
7.22302139e-01 9.14998770e-01 -3.73187035e-01 -1.86810475e-02
2.66313441e-02 -4.87254053e-01 1.24676073e+00 5.77192307e-01
8.31590295e-01 -9.17266488e-01 8.13340187e-01 6.82893276e-01
-1.49588954e+00 -4.39918693e-03 -5.90501726e-01 2.34931022e-01
1.44217059e-01 4.36441660e-01 -1.02509427e+00 8.96857917e-01
9.48005080e-01 9.26250100e-01 -9.18491244e-01 1.47613192e+00
-2.27921888e-01 -1.93964168e-01 -5.62851846e-01 6.74636811e-02
2.13723376e-01 -2.34652132e-01 7.99591422e-01 8.77349496e-01
7.29093671e-01 -1.77346066e-01 5.12676835e-01 7.42127359e-01
2.93374717e-01 -4.91509318e-01 -9.32908714e-01 6.50096357e-01
1.04819858e+00 1.52232718e+00 -6.28689528e-01 -1.43401846e-01
-7.63117671e-02 1.09596837e+00 4.03048135e-02 1.39016643e-01
-7.96829343e-01 -4.25522804e-01 8.54365528e-01 -1.17759071e-01
2.51038581e-01 -8.44244778e-01 7.09022731e-02 -1.02648926e+00
9.19149667e-02 -4.21644121e-01 3.53878029e-02 -1.15786290e+00
-6.91135049e-01 4.15358782e-01 -6.85088243e-03 -1.73770881e+00
-3.23838681e-01 -2.70876467e-01 -2.81781465e-01 1.00042152e+00
-1.33716679e+00 -1.18140888e+00 -8.34261239e-01 5.16296625e-01
3.98430794e-01 -1.89047351e-01 4.17024404e-01 1.26433238e-01
-9.85417888e-02 1.17295846e-01 5.50989844e-02 -1.78019926e-02
8.84543777e-01 -9.37605500e-01 7.20705986e-01 1.23352504e+00
1.21981241e-01 4.08365816e-01 5.82632482e-01 -7.30118871e-01
-2.17028952e+00 -1.30532610e+00 1.00186884e+00 -8.10680091e-01
4.65159625e-01 -7.02203870e-01 -4.44027454e-01 7.98975050e-01
-7.82694891e-02 2.43279651e-01 -2.33691424e-01 -4.76345628e-01
-2.16463491e-01 -4.35334444e-01 -9.66296792e-01 1.94974765e-01
8.82142007e-01 -7.25747049e-01 -2.89138615e-01 4.75304723e-01
7.87098706e-01 -1.06186187e+00 -7.29416132e-01 3.92620593e-01
3.98888439e-01 -1.00618994e+00 1.05344355e+00 6.04081094e-01
-5.17143667e-01 -1.26913834e+00 -5.37766099e-01 -1.12060809e+00
-1.16095319e-01 -5.91273487e-01 2.73657918e-01 8.48281622e-01
4.41817120e-02 -5.74770808e-01 7.28486121e-01 8.26861337e-03
-4.89454985e-01 6.92103803e-02 -1.34252059e+00 -8.48715127e-01
-7.64690876e-01 -6.68597817e-01 4.82926309e-01 6.37139440e-01
-5.71147978e-01 1.96254343e-01 -3.27991635e-01 9.52345729e-01
8.95563722e-01 1.23938628e-01 1.50395322e+00 -9.98071492e-01
1.79950804e-01 2.10914966e-02 -1.04755843e+00 -1.35234809e+00
1.51147678e-01 -7.35191703e-01 7.52768755e-01 -1.51050353e+00
-2.92646643e-02 -3.56675386e-01 3.23146611e-01 2.20487520e-01
2.19784558e-01 3.85421962e-01 1.59643948e-01 4.99688148e-01
-1.01733661e+00 3.94398451e-01 6.52333558e-01 -1.11314662e-01
1.25090797e-02 -2.52407677e-02 3.48445438e-02 3.74065459e-01
3.62105906e-01 -5.29600024e-01 -1.62940100e-01 -5.97027779e-01
2.38484427e-01 3.37251395e-01 6.93042934e-01 -1.51096582e+00
8.39525223e-01 -8.63247514e-02 4.47483778e-01 -1.66729176e+00
5.35034478e-01 -1.01144695e+00 4.91043717e-01 5.97081184e-01
2.78313488e-01 5.02106369e-01 1.25803560e-01 6.77327693e-01
-4.99606967e-01 -1.61844924e-01 6.78155839e-01 -9.78792533e-02
-1.44023037e+00 5.28295219e-01 -2.59947270e-01 -2.43721485e-01
1.22068894e+00 -3.43698263e-01 -1.80773169e-01 -3.96119386e-01
-2.03831717e-01 7.49924541e-01 9.14057732e-01 6.92386508e-01
8.66954505e-01 -1.47211635e+00 -6.50167882e-01 5.79923272e-01
8.86369586e-01 5.26137471e-01 2.50497073e-01 1.01851022e+00
-1.07514560e+00 3.68043363e-01 -8.09490606e-02 -1.48628604e+00
-1.00920486e+00 4.47525352e-01 4.75590706e-01 2.70680219e-01
-3.89180154e-01 7.89260030e-01 -1.45323977e-01 -9.36761618e-01
1.89468443e-01 -1.32717371e-01 2.67192513e-01 -1.15179963e-01
5.57832241e-01 1.35195091e-01 3.48896056e-01 -1.46321261e+00
-1.01253378e+00 1.17275500e+00 4.77646172e-01 -2.89601237e-01
1.02161741e+00 -6.31256104e-01 -1.74325004e-01 2.13975996e-01
1.47621477e+00 5.42264879e-02 -1.61582112e+00 -6.09174110e-02
2.52374738e-01 -7.06029296e-01 -3.78334001e-02 -2.47009531e-01
-8.28532696e-01 6.40558720e-01 8.25209677e-01 -4.55956459e-01
7.47958422e-01 1.99933842e-01 4.04402941e-01 6.06254458e-01
8.96610677e-01 -1.00505126e+00 7.32760737e-03 1.08239627e+00
9.34305727e-01 -1.40912449e+00 2.24470869e-02 -7.61219040e-02
-5.82302630e-01 1.10743940e+00 6.29132867e-01 -1.35571554e-01
3.06019217e-01 5.84118068e-02 3.41804028e-01 -1.80196181e-01
-2.63891280e-01 -2.26056620e-01 3.24708343e-01 7.27481067e-01
-4.82288510e-01 -1.01315953e-01 3.24430704e-01 -1.53699309e-01
-2.11314797e-01 -4.93112504e-01 2.66778678e-01 9.01437879e-01
-6.75331056e-01 -7.17441440e-01 -7.42027223e-01 -2.36530885e-01
4.96326536e-01 2.05363333e-01 -4.28970158e-01 6.72926605e-01
2.43014231e-01 8.96751821e-01 2.68368483e-01 -8.95565152e-01
4.64947075e-01 -2.00766280e-01 3.24134946e-01 -3.63749057e-01
-2.15888858e-01 1.74515292e-01 -1.07413366e-01 -1.18341982e+00
-1.82318345e-01 -7.73096502e-01 -1.28355348e+00 -3.11522894e-02
-2.94294477e-01 1.16002709e-01 1.33485425e+00 6.77831709e-01
7.07170606e-01 -2.13211924e-01 8.37382615e-01 -1.31005287e+00
-3.61732692e-01 -6.17182314e-01 -4.12290394e-01 3.12859053e-03
6.98766351e-01 -6.82826519e-01 -3.64105016e-01 -1.19524322e-01] | [7.366966247558594, -2.1978719234466553] |
72d1a3a6-4bba-4226-bed7-a491cc9b98f2 | how-good-is-a-video-summary-a-new | 2101.10514 | null | https://arxiv.org/abs/2101.10514v1 | https://arxiv.org/pdf/2101.10514v1.pdf | How Good is a Video Summary? A New Benchmarking Dataset and Evaluation Framework Towards Realistic Video Summarization | Automatic video summarization is still an unsolved problem due to several challenges. The currently available datasets either have very short videos or have few long videos of only a particular type. We introduce a new benchmarking video dataset called VISIOCITY (VIdeo SummarIzatiOn based on Continuity, Intent and DiversiTY) which comprises of longer videos across six different categories with dense concept annotations capable of supporting different flavors of video summarization and other vision problems. For long videos, human reference summaries necessary for supervised video summarization techniques are difficult to obtain. We explore strategies to automatically generate multiple reference summaries from indirect ground truth present in VISIOCITY. We show that these summaries are at par with human summaries. We also present a study of different desired characteristics of a good summary and demonstrate how it is normal to have two good summaries with different characteristics. Thus we argue that evaluating a summary against one or more human summaries and using a single measure has its shortcomings. We propose an evaluation framework for better quantitative assessment of summary quality which is closer to human judgment. Lastly, we present insights into how a model can be enhanced to yield better summaries. Sepcifically, when multiple diverse ground truth summaries can exist, learning from them individually and using a combination of loss functions measuring different characteristics is better than learning from a single combined (oracle) ground truth summary using a single loss function. We demonstrate the effectiveness of doing so as compared to some of the representative state of the art techniques tested on VISIOCITY. We release VISIOCITY as a benchmarking dataset and invite researchers to test the effectiveness of their video summarization algorithms on VISIOCITY. | ['Ganesh Ramakrishnan', 'Rishabh Iyer', 'Anshul Tomar', 'Suraj Kothawade', 'Vishal Kaushal'] | 2021-01-26 | null | null | null | null | ['supervised-video-summarization'] | ['computer-vision'] | [ 2.97039598e-01 1.01983033e-01 -3.51958066e-01 -2.55824536e-01
-1.28355849e+00 -7.33843803e-01 6.95705831e-01 4.34171438e-01
-2.31441513e-01 9.27799702e-01 8.93336356e-01 2.23467171e-01
-6.00090511e-02 -3.85133326e-01 -7.41933703e-01 -6.40307903e-01
-8.31857771e-02 3.42183501e-01 2.95653164e-01 1.54618099e-02
5.36511183e-01 1.50776863e-01 -1.77911890e+00 6.36338055e-01
1.01583016e+00 7.83769548e-01 2.19974995e-01 1.20326722e+00
2.86309302e-01 8.46299946e-01 -1.09365773e+00 -4.72860992e-01
2.40909189e-01 -7.66911805e-01 -7.08176613e-01 5.59403241e-01
1.33384693e+00 -5.60912728e-01 -3.54094565e-01 8.62045705e-01
6.26555443e-01 3.60598892e-01 9.63624001e-01 -1.44845498e+00
-4.76053894e-01 6.77661180e-01 -2.09054053e-01 2.92163849e-01
8.51673424e-01 3.13071102e-01 1.11909854e+00 -5.27187943e-01
8.97008359e-01 1.07262015e+00 6.21021926e-01 3.27276379e-01
-1.05659044e+00 -2.53173560e-01 2.48123690e-01 2.08640844e-01
-9.70341980e-01 -7.46921837e-01 4.26793128e-01 -4.75790352e-01
7.56860793e-01 5.99364042e-01 5.21343172e-01 1.37052560e+00
4.19797972e-02 9.61225748e-01 6.30686164e-01 -8.58376473e-02
1.92106903e-01 1.13677859e-01 2.13304967e-01 7.32190490e-01
6.04926169e-01 -3.23087484e-01 -6.38843417e-01 -2.39770815e-01
1.97520167e-01 -2.66443402e-01 -7.24819899e-01 -3.75950426e-01
-1.60268521e+00 7.67223060e-01 -6.77411333e-02 2.15317175e-01
-4.37797040e-01 6.69719577e-02 8.89426768e-01 3.56667340e-01
3.82182240e-01 8.08603585e-01 1.32737175e-01 -3.40517849e-01
-1.60373569e+00 6.47982955e-01 9.96657312e-01 1.07706261e+00
4.51103926e-01 2.46450812e-01 -8.45805824e-01 6.13234639e-01
-3.29363316e-01 4.22849298e-01 4.19306397e-01 -1.52256155e+00
5.79897225e-01 2.39653841e-01 2.70917267e-01 -1.29169130e+00
-1.39261350e-01 -3.58509183e-01 -7.55192280e-01 -1.05987012e-01
2.53722340e-01 -1.09724425e-01 -5.91226161e-01 1.49537885e+00
-2.59269446e-01 1.92547977e-01 1.73453674e-01 8.28423619e-01
1.38313532e+00 9.47221100e-01 -2.65576452e-01 -4.69580024e-01
9.56764221e-01 -1.27139425e+00 -6.89666927e-01 8.77530128e-02
5.28228879e-01 -6.18389845e-01 1.06420290e+00 6.79466009e-01
-1.38983715e+00 -6.56717598e-01 -1.31484079e+00 -1.19603453e-02
-5.01558967e-02 3.45987886e-01 1.89764664e-01 4.41058904e-01
-1.26666629e+00 7.39916503e-01 -5.62206864e-01 -6.66935384e-01
3.80014390e-01 3.76962504e-04 -5.22941589e-01 -1.60875157e-01
-6.78178310e-01 7.03106880e-01 6.09696865e-01 -4.97901946e-01
-9.58169699e-01 -4.94436115e-01 -1.06147683e+00 1.10498868e-01
6.49448454e-01 -9.67758775e-01 1.23451436e+00 -1.02498388e+00
-1.15262794e+00 6.76679373e-01 -1.38699055e-01 -5.80518425e-01
8.16687405e-01 -2.48341575e-01 -1.95166513e-01 7.46976554e-01
3.50166261e-01 8.05242479e-01 8.04970503e-01 -1.40288973e+00
-8.40465009e-01 1.11338392e-01 2.43420318e-01 3.53421777e-01
-2.85476774e-01 -1.61849737e-01 -4.92586941e-01 -9.31721747e-01
-5.11752009e-01 -7.15886712e-01 8.68196711e-02 -3.93469572e-01
-6.77197278e-01 -6.51391223e-02 7.31058419e-01 -8.48420322e-01
1.56669617e+00 -1.72299433e+00 4.62576956e-01 -2.57989258e-01
3.26085150e-01 3.19558531e-01 -3.94610584e-01 7.13750541e-01
2.11392641e-01 3.33622754e-01 -2.28270352e-01 -4.30319697e-01
-4.60113771e-02 -1.65768921e-01 -2.57594287e-01 3.08006257e-01
3.42531726e-02 6.81188941e-01 -1.19165421e+00 -5.62531054e-01
-1.08899195e-02 1.58192724e-01 -5.87131977e-01 1.27516061e-01
-2.72717506e-01 2.20826551e-01 -1.89306200e-01 5.19969404e-01
2.51953185e-01 -8.11776146e-02 -6.67630555e-03 -4.83794421e-01
-5.46495290e-03 -1.40411174e-02 -1.03101730e+00 1.63497317e+00
-1.08111985e-01 1.07546055e+00 -3.08004111e-01 -9.51704502e-01
6.87229097e-01 4.92798507e-01 4.21096385e-01 -1.73678473e-01
1.04784615e-01 1.84184626e-01 -3.89865547e-01 -7.20587909e-01
1.14214361e+00 1.27772361e-01 -9.86941382e-02 5.51981926e-01
2.92112112e-01 -3.08348387e-01 9.06949282e-01 6.06108785e-01
1.34517479e+00 6.11920506e-02 4.28743720e-01 -2.27040052e-02
4.59361345e-01 2.98822433e-01 3.36140364e-01 1.15086293e+00
-2.68760473e-01 1.21902919e+00 8.10974300e-01 -5.93828745e-02
-1.23124921e+00 -8.96510363e-01 3.70794684e-01 7.58769751e-01
1.18427135e-01 -9.05667603e-01 -6.48861647e-01 -7.66151309e-01
-2.06864342e-01 8.42230499e-01 -4.38544393e-01 -8.18031058e-02
-4.59630787e-01 -4.21797752e-01 4.79628950e-01 3.93307567e-01
5.10389209e-01 -1.01648700e+00 -7.41985977e-01 -6.93383580e-03
-6.18328273e-01 -1.24279273e+00 -7.56675184e-01 -4.18866724e-01
-8.55327547e-01 -1.14479733e+00 -1.04943526e+00 -5.43267608e-01
3.78425926e-01 6.44618392e-01 1.41533160e+00 8.10385793e-02
2.50153691e-01 9.28789437e-01 -6.43472314e-01 -1.03800967e-01
-8.38428974e-01 1.60226643e-01 3.52385901e-02 -9.37037617e-02
-1.96308404e-01 -3.45237225e-01 -5.40291488e-01 1.24633402e-01
-1.15919363e+00 1.42083332e-01 4.02820289e-01 6.72987998e-01
3.17374319e-01 7.32518807e-02 8.93351316e-01 -9.13007438e-01
8.70748580e-01 -5.66751957e-01 1.07131051e-02 3.21905911e-01
-1.12413421e-01 -1.82060048e-01 5.62945127e-01 -3.37596267e-01
-8.07623565e-01 -2.87465036e-01 2.26959258e-01 -5.97508848e-01
-3.27504128e-02 5.03206491e-01 6.89704611e-04 5.13188064e-01
7.90728569e-01 2.75745809e-01 -3.26005034e-02 -4.44602631e-02
3.29711109e-01 5.48327029e-01 8.79297078e-01 -3.90422106e-01
4.48999494e-01 4.44027901e-01 -3.93655717e-01 -1.09013832e+00
-8.74346435e-01 -6.75421417e-01 -3.48944992e-01 -4.44746643e-01
7.19163358e-01 -9.41551864e-01 -3.48643690e-01 2.15191394e-01
-1.19427991e+00 -2.64787257e-01 -4.21232343e-01 3.47938478e-01
-9.41797256e-01 8.32828999e-01 -3.23724896e-01 -4.40302640e-01
-3.68857890e-01 -1.07945168e+00 1.14434707e+00 1.28084734e-01
-4.44551170e-01 -9.28690732e-01 1.68885604e-01 4.21146810e-01
7.06531256e-02 6.14575863e-01 4.02391911e-01 -7.85274446e-01
-4.84374166e-01 -2.56332666e-01 -6.34602085e-02 6.77865148e-01
2.05803126e-01 4.31781381e-01 -5.69047689e-01 -5.28720260e-01
-1.69680297e-01 -3.15265000e-01 1.23482668e+00 5.20191908e-01
1.01602757e+00 -7.09622622e-01 -8.91278982e-02 2.83513188e-01
1.38922632e+00 -1.23300087e-02 7.13109612e-01 3.83163273e-01
6.61061704e-01 6.69755936e-01 9.25817668e-01 3.82325619e-01
4.54801500e-01 5.44975936e-01 1.88001201e-01 4.18712497e-01
-3.84766936e-01 -1.46304578e-01 7.78869867e-01 8.74085903e-01
-3.36849421e-01 -9.16061640e-01 -6.07771635e-01 8.95799458e-01
-2.03847480e+00 -1.54551363e+00 5.76532679e-03 2.23701930e+00
5.45431256e-01 9.65665728e-02 6.36804461e-01 4.76240292e-02
8.35301995e-01 4.72033322e-01 -2.96296477e-01 -4.38350648e-01
-3.05385917e-01 -3.38473141e-01 2.43194014e-01 2.54132926e-01
-1.14344776e+00 5.77801943e-01 6.61408806e+00 9.37470138e-01
-9.82135832e-01 -9.69041809e-02 5.72337389e-01 -2.32960433e-01
-3.57335299e-01 -2.26372451e-01 -5.32554567e-01 6.65243328e-01
8.49297643e-01 -4.13053721e-01 -4.37442884e-02 6.00034773e-01
4.13844079e-01 -4.40301150e-01 -1.38594162e+00 1.12852669e+00
7.30184495e-01 -1.76181316e+00 4.59561527e-01 -1.91833854e-01
1.20708275e+00 -1.42568707e-01 -1.55464575e-01 2.99068153e-01
5.07967211e-02 -8.42167497e-01 9.06159282e-01 6.71070397e-01
6.95047379e-01 -5.65197766e-01 8.44453514e-01 2.64678389e-01
-8.44828427e-01 6.18420392e-02 -2.33362004e-01 1.71547502e-01
2.59594172e-01 3.14850181e-01 -6.72627449e-01 9.59036410e-01
4.92589861e-01 1.08474255e+00 -9.02194440e-01 1.40314543e+00
4.29481752e-02 5.13203323e-01 3.08130756e-02 9.10482854e-02
3.15017939e-01 -1.11908838e-01 1.16260195e+00 1.52650189e+00
6.99963689e-01 -4.47703451e-02 3.12675267e-01 3.60920697e-01
-9.68783945e-02 8.63205343e-02 -7.86203921e-01 -2.19903335e-01
4.13938552e-01 1.11057043e+00 -7.90504217e-01 -9.26883101e-01
-3.36842269e-01 8.99316013e-01 3.39336917e-02 4.06019807e-01
-8.55141044e-01 -2.84841955e-01 2.10237056e-01 1.18380435e-01
4.27782357e-01 1.57498028e-02 -2.98542045e-02 -1.55117595e+00
1.59047507e-02 -1.20500648e+00 5.12558818e-01 -1.03952301e+00
-1.07383156e+00 8.39289486e-01 4.54289377e-01 -1.55882967e+00
-3.77314210e-01 -6.53252453e-02 -9.00924325e-01 9.30930674e-02
-1.20152509e+00 -8.32095683e-01 -6.84776485e-01 2.25716874e-01
1.03642941e+00 -2.88579106e-01 2.93637037e-01 7.49266669e-02
-4.90191460e-01 4.20085549e-01 1.10151852e-02 -2.93367833e-01
1.00185299e+00 -1.29916108e+00 1.06684022e-01 1.00189400e+00
1.40243128e-01 9.86753479e-02 1.23017490e+00 -7.52878845e-01
-9.47166443e-01 -1.23761630e+00 6.72041535e-01 -4.37624037e-01
4.42599207e-01 1.82564050e-01 -8.65012467e-01 7.75105655e-01
5.51645815e-01 -5.96149445e-01 4.82633114e-01 -3.39130282e-01
-8.78456235e-02 6.00757413e-02 -9.94463265e-01 6.74154818e-01
1.08201873e+00 -1.21909119e-01 -7.43197680e-01 3.67050022e-01
6.93860352e-01 -2.13490322e-01 -6.77803993e-01 3.27258915e-01
4.43526745e-01 -1.41303313e+00 7.91017294e-01 -4.02738035e-01
9.81093347e-01 -3.01111907e-01 -1.38716236e-01 -1.54277921e+00
4.55475375e-02 -7.10190594e-01 -2.82840103e-01 1.51267123e+00
9.46944114e-03 -2.38890857e-01 6.39821053e-01 8.62702578e-02
-4.51263666e-01 -4.74742293e-01 -3.11908394e-01 -9.99187827e-01
-2.86590070e-01 -1.31176651e-01 2.58917719e-01 7.10954428e-01
-4.19330820e-02 4.78317410e-01 -5.61088860e-01 -2.30534524e-01
5.75769186e-01 1.24727979e-01 1.12962317e+00 -9.02983785e-01
-1.00309849e-01 -5.77178180e-01 -6.16698623e-01 -9.19871449e-01
8.84437934e-02 -8.21654916e-01 1.06007658e-01 -1.93956017e+00
5.96662879e-01 3.62306600e-03 1.22618832e-01 8.52237344e-02
-2.77748317e-01 4.47388113e-01 4.82551634e-01 1.86126739e-01
-1.21767104e+00 5.10850728e-01 1.24781406e+00 -3.25067788e-01
-2.43473396e-01 -7.64367655e-02 -8.56390476e-01 6.62542701e-01
7.03756392e-01 -3.24455947e-01 -5.50052285e-01 -2.32343212e-01
2.00528800e-01 3.77064079e-01 4.12453085e-01 -1.30040157e+00
2.37603098e-01 8.48476738e-02 -1.16219446e-01 -8.77971292e-01
1.63851202e-01 -2.08970919e-01 2.74987102e-01 1.47828460e-01
-5.10593474e-01 1.07458569e-01 -1.04917148e-02 7.88720250e-01
-5.00651121e-01 -4.09095317e-01 4.50893283e-01 -2.36814052e-01
-7.31470883e-01 1.65451169e-01 -4.55938518e-01 2.20168367e-01
9.51563835e-01 -6.12903595e-01 -5.02121031e-01 -9.36550081e-01
-5.75589120e-01 3.79515231e-01 7.78171182e-01 3.82945091e-01
6.96910858e-01 -1.35445368e+00 -1.22985387e+00 -5.68079531e-01
2.11280704e-01 -2.83196837e-01 3.09896201e-01 8.80285203e-01
-8.53049994e-01 3.83100808e-01 -3.91439736e-01 -7.73202777e-01
-1.64841235e+00 4.80747312e-01 -6.76530376e-02 -2.12309107e-01
-6.94436967e-01 3.53007704e-01 2.82047719e-01 5.83943650e-02
3.49722028e-01 -3.56997401e-01 -4.79347676e-01 4.97767955e-01
6.89908087e-01 8.58282387e-01 -3.19688022e-02 -8.22236717e-01
-2.29304228e-02 4.07989323e-01 8.07661116e-02 -6.34787902e-02
1.35070860e+00 -3.88711512e-01 2.11928084e-01 3.91332388e-01
1.16662526e+00 1.74112514e-01 -1.42797041e+00 9.73026082e-02
-1.28019601e-01 -5.15733004e-01 -2.82585204e-01 -5.95154941e-01
-9.66671407e-01 4.84399945e-01 -7.90491793e-03 4.26856548e-01
1.14478552e+00 -1.03279352e-02 7.72289217e-01 4.10555601e-01
1.25251085e-01 -9.11250949e-01 5.86857557e-01 1.62910536e-01
1.28894818e+00 -1.45896959e+00 5.49873352e-01 -2.45424852e-01
-1.08594048e+00 1.18337154e+00 2.52835095e-01 -4.02232051e-01
-8.35254714e-02 -1.80647671e-01 -2.66615599e-01 -2.39641353e-01
-9.84084487e-01 -1.29936144e-01 5.82406402e-01 5.63441753e-01
1.63773715e-01 -9.23702195e-02 -4.34897274e-01 3.94540697e-01
-2.98852384e-01 -1.04888186e-01 1.27186465e+00 6.99834406e-01
-5.91935575e-01 -6.85793340e-01 -3.14942420e-01 7.51891732e-01
-5.28756499e-01 2.30758741e-01 -3.43565106e-01 8.82987261e-01
-2.31627390e-01 1.25516129e+00 -6.43039048e-02 -2.93542475e-01
1.81735754e-01 -3.41695368e-01 5.97575784e-01 -8.08698535e-01
-4.86788243e-01 -7.58127794e-02 5.05843997e-01 -5.42582154e-01
-9.02037263e-01 -8.52258205e-01 -7.21166134e-01 -5.03721595e-01
-9.16322172e-02 2.10330620e-01 2.48025656e-01 7.60221779e-01
2.12517142e-01 6.02036536e-01 3.87537092e-01 -1.22545016e+00
-3.46602648e-01 -7.58145094e-01 -3.40145290e-01 4.40460026e-01
5.56656897e-01 -4.31994677e-01 -4.57645774e-01 4.91534889e-01] | [10.478824615478516, 0.47169992327690125] |
ac3e2ac7-9ae5-4303-846d-a53bfbe217c3 | creating-corpora-for-research-in-feedback | null | null | https://aclanthology.org/2020.lrec-1.42 | https://aclanthology.org/2020.lrec-1.42.pdf | Creating Corpora for Research in Feedback Comment Generation | In this paper, we report on datasets that we created for research in feedback comment generation {---} a task of automatically generating feedback comments such as a hint or an explanatory note for writing learning. There has been almost no such corpus open to the public and accordingly there has been a very limited amount of work on this task. In this paper, we first discuss the principle and guidelines for feedback comment annotation. Then, we describe two corpora that we have manually annotated with feedback comments (approximately 50,000 general comments and 6,700 on preposition use). A part of the annotation results is now available on the web, which will facilitate research in feedback comment generation | ["Shin{'}ichiro Ishikawa", 'Ryo Nagata', 'Kentaro Inui'] | 2020-05-01 | null | null | null | lrec-2020-5 | ['comment-generation'] | ['natural-language-processing'] | [ 3.33009750e-01 8.29967976e-01 -2.02539936e-01 -4.39462930e-01
-1.14899313e+00 -6.37495279e-01 7.43054688e-01 6.07944429e-01
-3.11864942e-01 1.26268601e+00 1.11787677e+00 -8.19975376e-01
5.28831959e-01 -1.74809694e-01 -3.46926451e-01 -1.77739888e-01
5.53872347e-01 2.56052136e-01 2.39457741e-01 -4.64495122e-01
9.32292461e-01 -3.25898677e-01 -1.39022446e+00 9.40682948e-01
8.86572719e-01 4.22082007e-01 2.57594496e-01 1.07186866e+00
-2.00362846e-01 9.03253675e-01 -1.20677173e+00 -7.62738645e-01
-2.86698043e-01 -9.91953850e-01 -1.30041587e+00 1.21889353e-01
5.76466918e-01 -2.74863631e-01 3.50336492e-01 6.06262743e-01
7.19774246e-01 1.05109014e-01 8.42307210e-01 -9.83154058e-01
-1.17009234e+00 1.18456578e+00 -7.83700868e-02 3.51680845e-01
9.54251587e-01 -1.25793546e-01 1.41902590e+00 -1.24571824e+00
9.29802120e-01 8.71209681e-01 3.09693992e-01 1.22314954e+00
-8.45318317e-01 -2.56052554e-01 -7.36857802e-02 -1.07891085e-02
-3.23688656e-01 -3.81211847e-01 6.13107204e-01 -7.67986894e-01
8.14142346e-01 4.73630875e-01 6.12458467e-01 1.32882595e+00
-2.18739346e-01 1.01029682e+00 8.33464563e-01 -9.75058675e-01
-1.22932382e-01 8.70292425e-01 3.02957237e-01 4.06223893e-01
-7.55246580e-02 -1.97594851e-01 -8.52171421e-01 -2.97208458e-01
3.52759540e-01 -4.37863171e-01 -5.69902182e-01 4.20942158e-01
-9.85597253e-01 9.77966547e-01 3.33898723e-01 2.42628813e-01
8.08253884e-02 -9.27548409e-02 6.27493322e-01 5.89361846e-01
9.32994127e-01 6.59727991e-01 -4.31471497e-01 -3.82690579e-01
-7.00426459e-01 7.09841430e-01 1.23924506e+00 1.03968942e+00
6.40429199e-01 -1.46162495e-01 -5.61012626e-01 1.01375926e+00
4.72636580e-01 3.03103417e-01 6.01557732e-01 -7.63767660e-01
9.39093292e-01 5.64065099e-01 7.00965822e-01 -8.66064608e-01
-1.05320983e-01 2.37446036e-02 -2.07677782e-01 -1.02034725e-01
3.71159852e-01 -8.82679880e-01 -2.84450322e-01 1.01800501e+00
-2.43469968e-01 -6.49271727e-01 7.41440952e-02 7.53873348e-01
1.56947052e+00 5.96522748e-01 -3.13293010e-01 -4.41542953e-01
9.86917973e-01 -1.28256989e+00 -8.77024293e-01 9.84478444e-02
8.47939789e-01 -1.36465287e+00 1.44483674e+00 1.61602348e-01
-1.33122241e+00 -4.51088279e-01 -7.84251750e-01 -1.09558076e-01
-6.17098697e-02 4.00826246e-01 6.90367877e-01 5.77697992e-01
-1.36226571e+00 2.83010095e-01 -1.57417491e-01 -5.26964068e-01
6.68535978e-02 -1.94180220e-01 -1.07794683e-02 2.46130779e-01
-1.11929572e+00 7.21561611e-01 -1.93871543e-01 -3.01278323e-01
-2.42877156e-01 -7.71803021e-01 -6.46808624e-01 -4.30004537e-01
5.45266531e-02 -5.12822509e-01 2.24081826e+00 -1.04334962e+00
-1.37540901e+00 8.00941706e-01 -4.24822271e-01 -3.26265156e-01
4.68602538e-01 -3.51874650e-01 -2.84800619e-01 8.91048089e-02
2.66803533e-01 8.46658051e-01 4.52932209e-01 -1.21476305e+00
-6.64687693e-01 8.36784020e-02 4.01133746e-01 4.27552998e-01
-4.56939429e-01 7.84020483e-01 2.08597943e-01 -7.97246397e-01
-6.62405074e-01 -7.22438395e-01 -1.61425903e-01 -4.71735835e-01
-3.72785896e-01 -8.46053302e-01 3.44601870e-01 -7.47795522e-01
1.63976121e+00 -1.70758259e+00 -2.13353112e-01 -2.96891272e-01
2.55167577e-03 9.95368361e-02 8.50658566e-02 1.16153562e+00
6.68112412e-02 5.98360419e-01 -1.89086244e-01 -2.73236185e-01
2.10672002e-02 -2.64034867e-01 -7.20468998e-01 -1.64565846e-01
2.06543028e-01 9.63525236e-01 -1.34347844e+00 -2.12932363e-01
-5.00407219e-01 1.65850520e-02 -6.94612324e-01 4.61614847e-01
-5.96526980e-01 4.52692181e-01 -3.80452663e-01 3.36059064e-01
-5.34118004e-02 -3.38322133e-01 -3.91235560e-01 6.86517656e-01
-5.44080019e-01 1.10826552e+00 -4.63059455e-01 1.34461093e+00
-3.76047581e-01 9.82143879e-01 -2.16450989e-01 -1.39798045e-01
1.16681087e+00 7.86194384e-01 6.68282881e-02 -2.53313541e-01
-1.82689819e-02 4.89775896e-01 1.04241513e-01 -5.71198046e-01
1.22974718e+00 -5.80617860e-02 5.38234040e-02 1.40977192e+00
-1.00650825e-01 -3.36359620e-01 8.39055777e-01 6.10389411e-01
1.13790917e+00 -1.37761965e-01 1.34277537e-01 -7.91977495e-02
6.18765533e-01 2.30522975e-01 -6.34438396e-02 7.16769695e-01
2.00374663e-01 9.87060189e-01 3.52297425e-01 -1.28292337e-01
-1.05069160e+00 -4.38633323e-01 2.79504061e-02 1.20093811e+00
-6.30333424e-01 -1.10127926e+00 -7.07591474e-01 -7.87802756e-01
-7.61113390e-02 7.96846867e-01 -7.19222009e-01 4.21416402e-01
-4.11564976e-01 -1.58771008e-01 1.63336366e-01 6.27924204e-01
-7.96820894e-02 -1.74184334e+00 -4.82173771e-01 1.23567179e-01
-4.72054362e-01 -2.42842019e-01 -8.23283911e-01 2.06917509e-01
-7.35762298e-01 -8.26784730e-01 -9.60646570e-01 -7.34904706e-01
8.19088101e-01 2.53603846e-01 1.35398364e+00 6.85758293e-01
1.99665725e-01 4.10808653e-01 -1.09942293e+00 -1.07388985e+00
-8.04615080e-01 2.18880787e-01 -4.78213012e-01 -6.41527593e-01
4.19347584e-01 -7.14480653e-02 -2.31450796e-01 -4.74054068e-02
-6.53529704e-01 9.49075967e-02 3.63658041e-01 8.96334052e-01
-2.44595319e-01 -9.76411819e-01 7.85645306e-01 -1.53440547e+00
1.57875049e+00 -4.88645345e-01 8.45653638e-02 -6.11217022e-02
-5.39556682e-01 -6.66678473e-02 3.64837140e-01 3.19181569e-03
-1.31408429e+00 -3.87681067e-01 -4.04200822e-01 5.13257682e-01
-9.53500122e-02 8.92356634e-01 5.33520222e-01 3.14307541e-01
1.26411986e+00 -4.10528809e-01 -2.83800542e-01 -4.45124090e-01
3.09577912e-01 1.17702806e+00 2.20265508e-01 -6.36116982e-01
5.53720176e-01 -2.16400936e-01 -7.28787959e-01 -7.60870457e-01
-1.27253807e+00 -8.27487946e-01 -6.30654454e-01 -6.25845730e-01
5.56656897e-01 -6.61152661e-01 -1.42211795e-01 -1.77536495e-02
-1.33713603e+00 -6.65684164e-01 -5.91950297e-01 1.02744736e-01
-6.11760497e-01 1.25762388e-01 -7.93392420e-01 -1.05822992e+00
-5.61271369e-01 -7.11610198e-01 6.45266354e-01 4.31465358e-01
-1.01902759e+00 -1.17846513e+00 3.40251863e-01 5.33522010e-01
3.58096391e-01 -7.59467557e-02 5.61859429e-01 -5.35581529e-01
-1.82862550e-01 1.38888121e-01 -3.90425883e-02 3.72099161e-01
-5.34254946e-02 3.73721451e-01 -8.55173290e-01 1.04662202e-01
-2.55288243e-01 -1.06774426e+00 9.26823497e-01 3.23933475e-02
8.07992280e-01 -7.69018710e-01 4.81665321e-02 -3.12973082e-01
8.43621433e-01 -8.56696442e-02 3.97007078e-01 1.65702045e-01
3.69729310e-01 9.54034507e-01 9.08704519e-01 7.17802405e-01
5.13907850e-01 3.35632801e-01 1.26579985e-01 2.34144390e-01
-2.52301961e-01 -5.73481679e-01 5.30246615e-01 1.43816888e+00
-1.56209961e-01 -5.02578795e-01 -7.31424153e-01 8.27509999e-01
-1.91344631e+00 -1.06964910e+00 -7.90684879e-01 1.69828832e+00
1.33168554e+00 6.48324937e-02 2.82220542e-01 1.86695829e-01
4.53015864e-01 6.69806823e-02 7.30019435e-02 -1.06230152e+00
1.29165262e-01 3.30201179e-01 4.58718948e-02 7.66327262e-01
-6.05794311e-01 4.93472576e-01 6.95482874e+00 1.09511958e-02
-6.84556544e-01 -1.20768799e-02 4.89575922e-01 3.88345569e-01
-1.04185450e+00 1.61155298e-01 -1.04783583e+00 2.16256157e-01
1.15707183e+00 -5.98001719e-01 -1.90525696e-01 7.99108386e-01
5.53769350e-01 -2.35873148e-01 -1.19754529e+00 5.02469957e-01
4.39089417e-01 -1.41066718e+00 1.68652475e-01 -2.21059680e-01
1.04207158e+00 -2.45112181e-01 -5.23664020e-02 1.82953909e-01
7.08487391e-01 -6.90265477e-01 9.49252248e-01 3.87605786e-01
7.16604292e-01 -4.41243738e-01 8.29022408e-01 3.99190485e-01
-4.58760470e-01 -5.32575995e-02 -4.56190705e-01 -7.66889453e-01
2.70462722e-01 4.57273602e-01 -1.34238029e+00 3.91974837e-01
5.49980521e-01 1.00664949e+00 -7.91639805e-01 1.26850832e+00
-1.05603957e+00 9.87308681e-01 4.49381202e-01 -8.44568193e-01
8.47643688e-02 -5.68162985e-02 2.96736807e-01 1.34222484e+00
2.28295535e-01 8.14267769e-02 2.47330353e-01 5.03409386e-01
-3.00794184e-01 6.61413252e-01 -4.08360898e-01 -9.20466110e-02
4.27543014e-01 1.24597931e+00 -5.41856170e-01 -2.18704924e-01
-5.14592052e-01 7.92799771e-01 4.90566045e-01 3.24944466e-01
-2.34097496e-01 -4.21384156e-01 2.53912687e-01 3.60442340e-01
-1.43777058e-01 1.51845858e-01 -4.70762610e-01 -1.03707469e+00
1.36009127e-01 -8.44614267e-01 2.41403729e-01 -1.15661430e+00
-1.47560751e+00 4.25635368e-01 -1.73017845e-01 -1.06962824e+00
-7.52717674e-01 -4.60858583e-01 -9.53386724e-01 1.14510000e+00
-1.14344060e+00 -5.36520541e-01 -2.61591971e-01 1.27741396e-01
9.85322177e-01 -1.89214032e-02 9.89589632e-01 7.87339881e-02
4.71708216e-02 5.08941531e-01 -3.37352276e-01 1.22736841e-01
1.40542507e+00 -1.79142296e+00 3.21936160e-01 5.52690387e-01
1.91154197e-01 8.08541536e-01 9.17070031e-01 -7.11485028e-01
-5.06139040e-01 -6.85653865e-01 1.83984065e+00 -9.02281106e-01
8.68512213e-01 -1.37778908e-01 -6.47897959e-01 6.53514683e-01
1.00854576e+00 -8.20577383e-01 1.15099728e+00 2.57799655e-01
1.40206814e-02 5.04641593e-01 -7.00855255e-01 7.05905735e-01
7.49312103e-01 -4.49260086e-01 -8.81439865e-01 7.36633718e-01
9.79000211e-01 -5.73178411e-01 -4.99480128e-01 -6.33910596e-02
2.14191899e-01 -9.60625887e-01 2.34191611e-01 -6.56783938e-01
1.11896944e+00 -1.36714056e-01 3.40541005e-01 -1.75079858e+00
-1.88769042e-01 -9.28009868e-01 4.10537198e-02 1.71950662e+00
1.14715135e+00 -4.63494658e-02 7.39872515e-01 6.24618769e-01
-8.04278493e-01 -7.10790157e-01 -4.72010449e-02 -6.98384121e-02
2.28571787e-01 -2.38146842e-01 1.40722647e-01 7.23386884e-01
8.73693883e-01 9.45925951e-01 -1.11853711e-01 -9.18547273e-01
-2.15779379e-01 -6.83835745e-02 9.10839438e-01 -1.39109659e+00
-1.90585032e-01 -4.39363062e-01 4.67849225e-01 -1.29951358e+00
3.86969112e-02 -1.16520321e+00 2.66433954e-01 -1.99286091e+00
4.26471055e-01 -2.75619149e-01 3.55898112e-01 5.28370321e-01
-5.81098557e-01 2.26001203e-01 9.61970165e-02 2.76837975e-01
-4.04251605e-01 2.93617487e-01 1.45831275e+00 -4.41320566e-03
-2.79839575e-01 4.47198838e-01 -1.31392360e+00 4.39876527e-01
7.35418856e-01 -2.35001773e-01 -5.71936607e-01 -4.05783921e-01
7.26473987e-01 9.84976739e-02 -1.95511162e-01 -5.07592499e-01
1.28205121e-01 -1.11246124e-01 2.34594524e-01 -8.13886821e-01
-1.64063364e-01 -9.99868438e-02 -5.39737225e-01 7.90736675e-02
-1.12092113e+00 4.44302350e-01 -1.66749015e-01 2.83741653e-01
-4.01207924e-01 -9.12602305e-01 1.48582444e-01 -4.17729497e-01
-7.92441964e-02 -3.29075120e-02 -8.79570007e-01 3.07668865e-01
5.39234757e-01 3.14247794e-02 -6.51341021e-01 -9.98846114e-01
-4.48209226e-01 1.91145211e-01 5.03572702e-01 5.89977205e-01
6.30558074e-01 -1.22662103e+00 -1.09619963e+00 -2.25136191e-01
5.27243018e-01 -1.05779760e-01 -3.63988847e-01 5.29786527e-01
-2.71490842e-01 6.86685026e-01 -1.42545968e-01 -1.04815438e-01
-1.45235801e+00 1.32049084e-01 -4.32180971e-01 -2.55933940e-01
-4.41174865e-01 1.20233929e+00 -4.05329615e-01 -4.68795568e-01
1.57849029e-01 -3.95383805e-01 -8.50793540e-01 4.95151371e-01
1.12832153e+00 1.80680230e-01 2.28355959e-01 -4.33793575e-01
3.53813350e-01 -3.54845375e-01 -8.50002095e-02 -5.83783805e-01
1.17205834e+00 -2.56110817e-01 -2.36697569e-01 1.11094677e+00
7.18679249e-01 5.14490128e-01 -8.97641838e-01 3.85030685e-03
3.86164993e-01 -4.29781258e-01 -6.51870668e-01 -8.96329880e-01
-3.26676309e-01 8.70210350e-01 -3.52335155e-01 8.44681025e-01
5.62825799e-01 1.69367999e-01 5.72874546e-01 3.68354410e-01
1.90808102e-02 -1.23395073e+00 5.60616016e-01 1.10617423e+00
1.42446136e+00 -1.15541577e+00 -1.47036225e-01 -3.34099442e-01
-9.47601557e-01 1.41211832e+00 7.55624712e-01 1.62541196e-01
4.62793738e-01 1.85832083e-01 5.28395057e-01 -1.69631645e-01
-1.35794556e+00 2.50938418e-03 3.32543612e-01 8.96579742e-01
1.62398756e+00 -7.62285292e-02 -1.09977317e+00 4.49506044e-01
-1.02204192e+00 -8.48061070e-02 1.41035032e+00 8.50206077e-01
-7.79136300e-01 -1.62251461e+00 -2.50601768e-01 8.07760119e-01
-6.86818838e-01 -3.88244450e-01 -1.23992276e+00 2.94103742e-01
-5.00954568e-01 1.45012522e+00 -1.77342266e-01 -1.30969957e-01
2.39208639e-01 2.44500771e-01 3.66908044e-01 -1.60116506e+00
-1.27197850e+00 -3.60198677e-01 7.34162152e-01 1.25108421e-01
-7.28640139e-01 -1.10759199e+00 -1.18930161e+00 -1.78686351e-01
-3.51671487e-01 7.44173527e-01 5.10378778e-01 8.10859084e-01
-1.96595684e-01 4.56404537e-01 3.56976151e-01 -7.53836751e-01
-6.17021441e-01 -1.66201854e+00 -3.71741802e-01 4.81272012e-01
2.07957476e-01 -2.12930530e-01 -3.34418625e-01 5.42034864e-01] | [11.833080291748047, 9.043704986572266] |
75733332-38a7-4cf8-8a6f-c4eba0197099 | high-throughput-cotton-phenotyping-big-data | 2305.05423 | null | https://arxiv.org/abs/2305.05423v1 | https://arxiv.org/pdf/2305.05423v1.pdf | High-throughput Cotton Phenotyping Big Data Pipeline Lambda Architecture Computer Vision Deep Neural Networks | In this study, we propose a big data pipeline for cotton bloom detection using a Lambda architecture, which enables real-time and batch processing of data. Our proposed approach leverages Azure resources such as Data Factory, Event Grids, Rest APIs, and Databricks. This work is the first to develop and demonstrate the implementation of such a pipeline for plant phenotyping through Azure's cloud computing service. The proposed pipeline consists of data preprocessing, object detection using a YOLOv5 neural network model trained through Azure AutoML, and visualization of object detection bounding boxes on output images. The trained model achieves a mean Average Precision (mAP) score of 0.96, demonstrating its high performance for cotton bloom classification. We evaluate our Lambda architecture pipeline using 9000 images yielding an optimized runtime of 34 minutes. The results illustrate the scalability of the proposed pipeline as a solution for deep learning object detection, with the potential for further expansion through additional Azure processing cores. This work advances the scientific research field by providing a new method for cotton bloom detection on a large dataset and demonstrates the potential of utilizing cloud computing resources, specifically Azure, for efficient and accurate big data processing in precision agriculture. | ['Glen Rains', 'Javad Mohammadpour Velni', 'Alireza Ebrahimi', 'Amanda Issac'] | 2023-05-09 | null | null | null | null | ['plant-phenotyping', 'automl'] | ['computer-vision', 'methodology'] | [-3.49133343e-01 -3.94924939e-01 4.01858717e-01 -5.86880110e-02
-7.27978572e-02 -7.85804868e-01 2.40865201e-01 8.83895755e-01
4.05820049e-02 -1.59341414e-02 -6.31303668e-01 -4.31238502e-01
3.74015234e-02 -1.53533065e+00 -7.10076928e-01 -9.49711621e-01
-2.89555848e-01 3.20248425e-01 4.24823076e-01 -3.16276476e-02
-6.18776083e-02 1.10332513e+00 -1.83271611e+00 5.55891931e-01
5.66511095e-01 1.24803340e+00 6.09286904e-01 1.12121892e+00
1.52189568e-01 6.16291165e-01 -9.10413623e-01 -7.73327500e-02
3.28001589e-01 5.82406282e-01 -1.28019348e-01 -2.77793974e-01
3.59545350e-01 -6.56202674e-01 1.35766506e-01 7.27806211e-01
6.31605506e-01 -5.12158155e-01 1.20911531e-01 -1.38553488e+00
-6.55806124e-01 4.72860456e-01 -7.08382905e-01 2.92080998e-01
-1.15851395e-01 7.29913354e-01 6.27099216e-01 -7.49383926e-01
3.72514725e-01 9.36807275e-01 7.95855939e-01 -2.83334911e-01
-9.93748367e-01 -8.64070594e-01 -4.01028514e-01 1.87806353e-01
-1.31032395e+00 -5.70436269e-02 -1.37430802e-01 -5.52554846e-01
1.21613944e+00 4.52356398e-01 8.09138358e-01 3.91954869e-01
6.70046508e-01 3.90523076e-01 9.05644238e-01 -2.15830520e-01
4.95481312e-01 -1.76177993e-01 5.71415842e-01 8.54508758e-01
6.85493767e-01 1.92192808e-01 -5.26822388e-01 -8.23974609e-02
7.12876499e-01 4.91060704e-01 3.28973472e-01 8.12390298e-02
-9.25950110e-01 3.94529164e-01 7.66124666e-01 1.49453133e-01
-9.77113068e-01 9.40897465e-02 6.37959301e-01 1.82612285e-01
5.13244808e-01 3.16995054e-01 -6.25055671e-01 2.82859713e-01
-6.88384056e-01 1.02833547e-01 9.35860872e-01 1.36185813e+00
6.23825729e-01 5.58464043e-02 -5.98477900e-01 1.69768959e-01
2.05303833e-01 7.70826280e-01 -1.82949781e-01 -6.49850070e-01
-2.97439426e-01 1.08237183e+00 2.90498763e-01 -9.46597815e-01
-7.32954741e-01 -1.70127600e-01 -8.08771014e-01 5.97243369e-01
3.21500003e-01 -2.83505112e-01 -6.60509288e-01 9.87889469e-01
6.72352076e-01 3.74335468e-01 1.17113978e-01 5.96546471e-01
1.24728918e+00 7.67974734e-01 3.52096051e-01 1.42627716e-01
1.93222249e+00 -4.89068538e-01 -6.05508029e-01 5.02503216e-01
5.51057279e-01 -9.73936856e-01 1.09497774e+00 6.61813796e-01
-1.08811355e+00 -5.34640729e-01 -1.09058452e+00 -1.33856729e-01
-9.54363227e-01 6.33937061e-01 1.09294665e+00 7.22716212e-01
-9.30974722e-01 4.37087953e-01 -1.46221840e+00 -7.83301532e-01
7.34982193e-01 3.87843251e-01 -3.21453996e-02 1.50898263e-01
-2.31043383e-01 3.18899661e-01 8.03609252e-01 3.35417658e-01
-1.46633983e+00 -1.14462304e+00 -2.80830771e-01 6.97778881e-01
1.24384798e-01 -2.77739018e-01 1.12385023e+00 -6.96090683e-02
-1.33550429e+00 8.25688362e-01 3.37007880e-01 -4.61626440e-01
-1.80571839e-01 -3.65787625e-01 -2.12530807e-01 5.92801981e-02
-2.95859665e-01 5.01401842e-01 4.28933084e-01 -6.86123729e-01
-1.17215955e+00 -7.27069974e-01 1.78863719e-01 -5.96533537e-01
-5.79311252e-01 4.64061171e-01 4.28678654e-02 -2.20705062e-01
-1.25851348e-01 -7.39389956e-01 -3.01551889e-03 3.94023389e-01
-1.64372638e-01 -2.22880840e-01 1.15884650e+00 -5.03000557e-01
6.30938470e-01 -2.03138089e+00 -7.52866387e-01 -6.31111190e-02
3.97584140e-01 5.65384269e-01 -5.70833646e-02 6.17447615e-01
2.40180239e-01 -2.87227958e-01 4.07549560e-01 2.89111167e-01
-5.11870869e-02 1.03493407e-01 -3.77916515e-01 4.34403419e-01
5.40426612e-01 6.94021583e-01 -7.85443962e-01 -6.36950731e-02
4.21770602e-01 7.98587441e-01 -4.36909795e-01 5.17137766e-01
-5.99768043e-01 1.43484008e-02 -2.34970346e-01 1.07976317e+00
1.42147899e+00 -4.84861553e-01 1.26199007e-01 -5.12713194e-01
-9.57783759e-01 -2.79006451e-01 -1.27815139e+00 1.21078265e+00
-4.24034059e-01 5.32309532e-01 4.26825583e-01 -5.10085344e-01
1.17378795e+00 3.56580913e-01 5.79926670e-01 -4.08721045e-02
1.96095437e-01 -8.56773108e-02 -2.87889063e-01 -6.21357262e-01
3.89661312e-01 8.07738483e-01 5.18620968e-01 4.01347011e-01
1.96165249e-01 6.35224804e-02 5.27233183e-01 2.70981751e-02
1.55686617e+00 2.83232570e-01 2.77349472e-01 -3.72969985e-01
3.77621502e-02 4.44096088e-01 3.81994277e-01 7.47451484e-01
-1.73125908e-01 -1.18672671e-02 5.34760714e-01 -8.48092377e-01
-9.58331704e-01 -8.55972648e-01 -4.71956134e-01 1.50450337e+00
-7.38439411e-02 -7.67537177e-01 -4.65804428e-01 -2.34324843e-01
1.09534033e-01 3.81187469e-01 -3.02272856e-01 2.75503546e-01
-3.05790901e-01 -1.44283473e+00 5.67842722e-01 5.35202742e-01
6.84427857e-01 -1.39003778e+00 -1.42292094e+00 3.06933910e-01
5.59087515e-01 -1.09089124e+00 6.74596906e-01 5.34468472e-01
-6.73317075e-01 -1.15737426e+00 -3.69335748e-02 -7.11067855e-01
3.64865631e-01 5.05366683e-01 1.31318986e+00 1.40175387e-01
-9.51050043e-01 -3.02365124e-01 -4.31148887e-01 -1.19702280e+00
-2.42779627e-01 -8.78653750e-02 -4.70430702e-01 -5.26427090e-01
1.07168770e+00 -5.10939360e-01 -8.44350278e-01 1.66452583e-02
-1.00664186e+00 5.93199320e-02 4.97042269e-01 5.03117383e-01
5.70510745e-01 8.90895873e-02 3.41317356e-01 -9.02179122e-01
2.85758585e-01 -7.16515124e-01 -1.65538418e+00 3.99604708e-01
-5.30175626e-01 -3.96043509e-01 6.27261758e-01 8.50833505e-02
-1.02780569e+00 3.91977251e-01 -1.01881474e-01 -1.50035411e-01
-5.89629054e-01 2.78234988e-01 7.38992617e-02 -2.44561657e-01
8.00325513e-01 -2.96175450e-01 -2.70618320e-01 -5.02238393e-01
2.22129852e-01 8.76541913e-01 6.73303485e-01 -2.76150048e-01
3.24301898e-01 6.09851718e-01 3.01223874e-01 -8.11695755e-01
-3.85654598e-01 -6.43837690e-01 -7.74968386e-01 -7.02730194e-02
9.01972055e-01 -1.11547935e+00 -1.78122222e+00 6.35278285e-01
-1.34673429e+00 -3.05030882e-01 -2.17624754e-01 2.03192800e-01
1.41615253e-02 -1.45133778e-01 -9.29820836e-01 -5.51231027e-01
-1.10678589e+00 -8.41212332e-01 1.38835192e+00 4.60755706e-01
5.93517482e-01 -1.57317445e-01 -6.42583519e-02 -2.50337899e-01
5.92805207e-01 5.87733448e-01 5.75882137e-01 -4.77088749e-01
-1.11490226e+00 -5.29474854e-01 -9.94230151e-01 -1.44203126e-01
4.61168177e-02 8.76177371e-01 -1.49055171e+00 -3.36852968e-01
-2.45958164e-01 -2.76600242e-01 5.94213605e-01 5.15292108e-01
1.51678455e+00 7.03042373e-02 -4.29648042e-01 9.97424066e-01
1.71458781e+00 7.75720105e-02 4.31839615e-01 2.50525773e-01
6.12370133e-01 1.76070407e-01 7.00018287e-01 1.18139946e+00
1.09569691e-01 1.87901944e-01 1.23118281e+00 -3.62143785e-01
1.09207660e-01 3.79549891e-01 8.79509225e-02 3.34999591e-01
-7.06346706e-02 -2.92694271e-01 -1.17516923e+00 2.27259099e-01
-1.80164671e+00 -1.02763999e+00 -4.37989742e-01 1.97121787e+00
4.54512656e-01 -3.14120770e-01 -5.48782460e-02 -1.04714027e-02
4.47425455e-01 -1.84207186e-01 -6.18893325e-01 -3.26740742e-01
-2.50232387e-02 5.47288060e-01 5.95491290e-01 -8.96639898e-02
-1.36995506e+00 1.13697863e+00 5.67547512e+00 2.93908685e-01
-1.26780808e+00 1.81787491e-01 2.87161708e-01 -3.85422558e-02
6.29666030e-01 -1.47569492e-01 -1.27421629e+00 1.40439913e-01
1.19428289e+00 5.61701972e-03 1.56135157e-01 1.13320589e+00
4.17051643e-01 3.64936106e-02 -8.49493742e-01 4.60800350e-01
-4.05168980e-01 -1.65795243e+00 -3.78894717e-01 -1.08908877e-01
4.12766367e-01 4.43373352e-01 -3.83566976e-01 1.08578280e-01
8.70686829e-01 -5.72876930e-01 1.86506018e-01 1.42591372e-01
5.43132365e-01 -6.53605163e-01 6.85507238e-01 3.36546361e-01
-1.37718272e+00 -4.12886143e-01 -1.04166651e+00 -1.10287413e-01
-4.83332634e-01 8.90298665e-01 -1.46218371e+00 3.25452179e-01
1.17005622e+00 6.70123279e-01 -9.37679112e-01 1.27432525e+00
1.85155243e-01 7.00914383e-01 -7.13014543e-01 7.11880438e-03
-1.82791427e-02 -7.87068680e-02 9.87615734e-02 1.60108352e+00
5.63112020e-01 -2.99932044e-02 3.28176051e-01 8.36196899e-01
-5.43141104e-02 1.14639424e-01 -6.95864975e-01 -6.64436892e-02
7.23571956e-01 2.00183249e+00 -9.60273147e-01 -4.48315650e-01
-4.86198157e-01 3.63117784e-01 7.28004724e-02 -2.04351038e-01
-6.32277727e-01 -7.28024304e-01 5.27870417e-01 -1.44685209e-01
4.06436980e-01 -1.37699753e-01 -5.39627910e-01 -7.41429627e-01
-3.29921395e-01 -8.72085273e-01 5.58918774e-01 -9.88526404e-01
-9.67492938e-01 5.51220119e-01 -2.77873576e-01 -8.36976111e-01
1.34755671e-01 -9.91188228e-01 -6.79604232e-01 6.94436610e-01
-1.22829735e+00 -1.83238780e+00 -1.39177263e+00 4.00625795e-01
3.94594133e-01 -3.14825445e-01 1.30219030e+00 3.25520307e-01
-1.07257164e+00 9.44972932e-02 4.37711924e-01 -3.36158782e-01
5.23713410e-01 -1.47765219e+00 5.33214509e-01 1.21189380e+00
-7.65518472e-02 3.21347892e-01 4.95516002e-01 -7.14375079e-01
-1.98537743e+00 -1.48643804e+00 3.61490577e-01 -3.30247879e-01
4.95199084e-01 -5.80923319e-01 -7.29535162e-01 7.92179644e-01
4.13821906e-01 3.74761611e-01 6.25278592e-01 -2.31384393e-02
-1.65936306e-01 -6.64396882e-01 -1.44733405e+00 2.13796139e-01
4.25477117e-01 3.37853283e-03 4.38544631e-01 1.18765128e+00
7.60556400e-01 -4.18542832e-01 -1.31277788e+00 3.32913905e-01
5.46665967e-01 -8.73731673e-01 8.77622485e-01 -4.96176451e-01
1.96108595e-01 -4.19723064e-01 -1.65034831e-01 -7.41416693e-01
-4.70214725e-01 -7.60918379e-01 -3.90763491e-01 1.32155788e+00
-1.94700554e-01 -1.01337321e-01 7.23067999e-01 2.18020126e-01
-4.13748205e-01 -3.46310139e-01 -1.83388501e-01 -4.11762059e-01
-4.02344763e-01 -3.11749071e-01 1.24048686e+00 5.30200183e-01
-3.36027533e-01 -1.52942866e-01 2.20638558e-01 1.18495667e+00
5.61168134e-01 4.95280266e-01 9.23883677e-01 -1.29301453e+00
-7.54911527e-02 4.46993858e-02 -5.55280805e-01 -3.38613987e-01
-5.07894158e-01 -6.79782867e-01 5.17658377e-03 -1.12331212e+00
1.53536469e-01 -5.32021344e-01 -2.52532989e-01 8.18229735e-01
2.09656302e-02 4.18420672e-01 2.12966725e-01 2.64281705e-02
-2.67018855e-01 -1.91529974e-01 7.57891953e-01 5.10485955e-02
-3.88759732e-01 6.71605542e-02 -2.33226463e-01 7.54045844e-01
8.27689111e-01 -4.50706601e-01 1.11943640e-01 -7.86182165e-01
3.10736150e-01 -3.47154528e-01 6.19303763e-01 -1.30878127e+00
3.11238855e-01 -6.10614792e-02 8.22066009e-01 -9.91819382e-01
8.37872252e-02 -9.79545474e-01 1.51498377e-01 8.14549029e-01
1.49536148e-01 4.98241305e-01 6.89979792e-01 1.94712073e-01
4.10512269e-01 1.23102114e-01 8.19725871e-01 -4.92489897e-02
-7.38622785e-01 2.78914332e-01 -2.56156802e-01 -7.37292588e-01
1.40761173e+00 3.23033959e-01 -8.46524000e-01 6.75308108e-01
-4.88193691e-01 8.67801532e-02 2.71913916e-01 1.61540672e-01
2.77698457e-01 -7.62369037e-01 -9.31051016e-01 8.45841110e-01
2.06650451e-01 9.80806351e-02 -1.27172023e-01 6.67887866e-01
-1.41551220e+00 4.18216914e-01 -9.93414879e-01 -1.05347383e+00
-1.57182312e+00 7.40464926e-01 1.90572627e-02 -1.37615234e-01
-9.98763025e-01 6.35372877e-01 -1.55469747e-02 -1.59799531e-01
5.19229233e-01 -7.54725635e-01 -3.54232848e-01 -1.31740883e-01
8.99517059e-01 6.43400311e-01 6.20550573e-01 1.34747967e-01
-2.18868002e-01 -3.95180136e-02 5.60211688e-02 6.29726768e-01
1.81146348e+00 4.71509248e-01 -4.66342539e-01 5.21902815e-02
1.96605459e-01 -3.89278024e-01 -1.03582489e+00 3.53013366e-01
9.29062814e-02 -6.25897586e-01 3.31453711e-01 -8.90461743e-01
-1.16062665e+00 8.62238050e-01 1.54140985e+00 6.10290527e-01
1.49171138e+00 -3.21890712e-01 3.77085298e-01 8.85701895e-01
7.98329189e-02 -3.88624519e-01 -2.43108407e-01 4.42314208e-01
6.14083469e-01 -1.25660193e+00 -1.33155048e-01 -3.41513902e-01
6.08249269e-02 1.24896967e+00 8.90873611e-01 -2.89105535e-01
8.09524119e-01 9.57839429e-01 8.36851597e-02 -7.76369333e-01
-1.10836387e+00 -1.20582536e-01 -6.21608317e-01 7.97336817e-01
7.98753440e-01 2.42874876e-01 3.56389806e-02 2.50206381e-01
1.29221171e-01 6.67604983e-01 4.58255321e-01 1.22186720e+00
-6.49057269e-01 -8.00255418e-01 -5.02242863e-01 7.34401166e-01
-5.80236435e-01 -1.03578612e-01 6.27767593e-02 3.82880211e-01
4.32394207e-01 8.15549016e-01 4.05865759e-01 1.09668538e-01
3.11489433e-01 -2.27510229e-01 6.66159317e-02 -8.06929410e-01
-1.33443344e+00 4.36189249e-02 -2.73779750e-01 -6.66185737e-01
1.09116510e-01 -1.43011749e-01 -1.01909482e+00 -7.04026222e-01
-4.58217502e-01 -4.68242168e-01 1.18178737e+00 3.62461269e-01
9.39659476e-01 6.68332696e-01 4.45568413e-01 -8.34149420e-01
-4.63518947e-01 -1.00959718e+00 -6.71304166e-01 -4.35228422e-02
6.35389984e-02 -1.62036553e-01 3.74357790e-01 3.47621918e-01] | [9.155543327331543, -1.5072975158691406] |
6468b941-bc90-4758-8c25-21e8bf201da2 | view-volume-network-for-semantic-scene | 1806.05361 | null | http://arxiv.org/abs/1806.05361v1 | http://arxiv.org/pdf/1806.05361v1.pdf | View-volume Network for Semantic Scene Completion from a Single Depth Image | We introduce a View-Volume convolutional neural network (VVNet) for inferring
the occupancy and semantic labels of a volumetric 3D scene from a single depth
image. The VVNet concatenates a 2D view CNN and a 3D volume CNN with a
differentiable projection layer. Given a single RGBD image, our method extracts
the detailed geometric features from the input depth image with a 2D view CNN
and then projects the features into a 3D volume according to the input depth
map via a projection layer. After that, we learn the 3D context information of
the scene with a 3D volume CNN for computing the result volumetric occupancy
and semantic labels. With combined 2D and 3D representations, the VVNet
efficiently reduces the computational cost, enables feature extraction from
multi-channel high resolution inputs, and thus significantly improves the
result accuracy. We validate our method and demonstrate its efficiency and
effectiveness on both synthetic SUNCG and real NYU dataset. | ['Yu-Xiao Guo', 'Xin Tong'] | 2018-06-14 | null | null | null | null | ['3d-semantic-scene-completion'] | ['computer-vision'] | [ 1.07130013e-01 6.52741864e-02 8.28989875e-03 -6.86740994e-01
-5.07465482e-01 -4.65842366e-01 4.82825071e-01 -3.08011323e-01
-2.47996584e-01 2.31074452e-01 2.54091918e-01 -1.43557817e-01
2.86850750e-01 -1.39723647e+00 -9.43954885e-01 -4.03619111e-01
1.23087220e-01 5.28213680e-01 1.57329425e-01 3.04944426e-01
1.42216653e-01 9.04864430e-01 -1.60249412e+00 3.14023942e-01
3.77030345e-03 1.70205832e+00 2.19910711e-01 6.36341989e-01
-4.68430042e-01 5.96158981e-01 -1.05705582e-01 3.31707895e-01
5.85281193e-01 1.28522933e-01 -7.67933846e-01 4.65845019e-01
4.58989233e-01 -8.86643529e-01 -5.55201232e-01 6.09237850e-01
4.18690115e-01 6.29827846e-03 6.37677133e-01 -7.22218990e-01
-3.61773223e-01 -2.69920588e-01 -5.92977703e-01 -2.34104842e-02
4.35221851e-01 7.10496530e-02 8.06168258e-01 -1.28567290e+00
7.40885794e-01 1.32397139e+00 4.02871847e-01 2.42088035e-01
-1.06077480e+00 -4.80656385e-01 2.08553448e-01 -5.43495178e-01
-1.22927046e+00 6.20321743e-02 9.85500336e-01 -5.92914402e-01
1.25172877e+00 -4.59787697e-02 1.28720307e+00 6.99180186e-01
3.41246314e-02 5.73573709e-01 1.22436190e+00 1.51541755e-01
4.06682044e-01 -3.09550703e-01 -2.08130375e-01 1.19013929e+00
-1.97464213e-01 1.43406878e-03 -4.79158103e-01 5.66680990e-02
1.32346499e+00 6.29770398e-01 -1.46948248e-01 -8.93760145e-01
-1.07878101e+00 7.91202366e-01 8.56840253e-01 -3.84903610e-01
-2.89933026e-01 4.98943031e-01 1.37636572e-01 -1.08995609e-01
7.18262255e-01 1.21999651e-01 -6.25534058e-01 5.70841953e-02
-5.36163807e-01 7.09910840e-02 4.99244899e-01 9.34925437e-01
1.23137116e+00 -2.08876431e-01 1.58766657e-01 3.87663662e-01
5.54873347e-01 9.85744417e-01 -9.88624990e-02 -1.24549377e+00
6.72107160e-01 1.11616957e+00 -6.50040656e-02 -6.90478146e-01
-6.86132133e-01 -2.15656906e-01 -5.31800270e-01 5.92319310e-01
9.61882398e-02 1.20155126e-01 -1.19476306e+00 1.19198620e+00
4.84829932e-01 -5.22746332e-02 -5.62973917e-02 1.10523951e+00
1.00166893e+00 6.05046809e-01 -5.55071473e-01 1.77268028e-01
1.06915069e+00 -6.66967988e-01 -7.45911002e-02 -2.13862106e-01
5.57742894e-01 -1.85844541e-01 1.07192564e+00 2.23422423e-01
-1.00327778e+00 -3.33133787e-01 -1.20713162e+00 -6.72529578e-01
-3.39495778e-01 -1.00277625e-01 6.97324872e-01 2.70055324e-01
-9.82899487e-01 3.65576476e-01 -1.07634437e+00 -1.80610895e-01
7.62017131e-01 5.09262025e-01 -4.67264533e-01 -4.30854321e-01
-7.04781890e-01 3.93287927e-01 1.28082037e-01 -8.42356402e-03
-1.04974830e+00 -8.37083519e-01 -1.24848592e+00 -4.81771603e-02
6.52664304e-02 -8.83231997e-01 9.76020634e-01 -3.11133206e-01
-1.61985350e+00 1.07514465e+00 -1.07370742e-01 1.42395094e-01
4.50228900e-01 -1.09418333e-01 2.76092321e-01 2.01014906e-01
1.64115444e-01 7.27734447e-01 4.45825756e-01 -1.37723053e+00
-4.97569412e-01 -9.51073885e-01 3.74467462e-01 4.91656840e-01
1.70233667e-01 -8.13984632e-01 -7.25804865e-01 1.60996750e-01
9.05068696e-01 -5.32972038e-01 -1.43077180e-01 4.88178104e-01
-4.97800797e-01 2.82027483e-01 8.63660872e-01 -3.62617373e-01
4.08738762e-01 -2.24144077e+00 2.29735762e-01 3.68538946e-01
6.13003850e-01 -4.88043606e-01 1.48099825e-01 -9.03909057e-02
3.48072827e-01 -4.98638339e-02 -3.80454242e-01 -5.39514720e-01
-3.38104278e-01 3.94145191e-01 -2.31278554e-01 7.48235464e-01
7.51719996e-02 1.13583207e+00 -9.97155726e-01 -1.52988389e-01
6.95565522e-01 8.57223511e-01 -6.69398785e-01 4.94602829e-01
-4.19421375e-01 5.79912961e-01 -7.81802714e-01 6.12675965e-01
7.24490643e-01 -5.02296031e-01 -1.73320040e-01 -1.47374049e-01
-2.54245341e-01 4.19146985e-01 -7.55293727e-01 2.28583646e+00
-8.03677201e-01 5.50422549e-01 -1.02274820e-01 -5.99293351e-01
1.15069485e+00 -1.60077125e-01 6.89310730e-01 -1.04945779e+00
2.76262462e-01 8.23508725e-02 -8.44028533e-01 -5.02480149e-01
3.21044415e-01 -1.48770019e-01 -3.52066755e-01 4.20272052e-01
-5.94318099e-03 -6.75699413e-01 -6.87949121e-01 -5.17218839e-04
1.05610323e+00 4.34616238e-01 1.88472301e-01 1.38834357e-01
3.02887946e-01 -2.39434332e-01 1.67850405e-01 1.85347140e-01
2.09729671e-01 8.71925950e-01 6.84677958e-01 -9.95443702e-01
-1.29240084e+00 -1.61725795e+00 -1.36780858e-01 3.62088025e-01
3.70558709e-01 -7.31491521e-02 -3.89293134e-01 -5.80232501e-01
2.48589531e-01 9.03137177e-02 -7.94924140e-01 1.50939107e-01
-4.25037354e-01 -2.15084627e-01 6.12577759e-02 6.88154936e-01
7.34382927e-01 -6.36859000e-01 -1.19709063e+00 -1.58118941e-02
4.53846492e-02 -1.37150013e+00 -1.74066782e-01 2.74114221e-01
-1.13777578e+00 -1.13740957e+00 -3.85995120e-01 -6.06382847e-01
8.78942907e-01 2.23603427e-01 9.93739724e-01 -1.29834965e-01
-3.06068301e-01 3.50172341e-01 -3.99261340e-02 -2.53908843e-01
2.10427523e-01 -1.54606700e-02 -1.29361480e-01 -1.16320252e-01
2.18783259e-01 -9.08992231e-01 -8.87401283e-01 -3.90100032e-02
-6.15515053e-01 5.25740504e-01 2.22744167e-01 5.02439499e-01
1.29022861e+00 -3.87855738e-01 -2.06153229e-01 -9.02026653e-01
-2.16000587e-01 -2.10491568e-01 -1.09379220e+00 -2.96045303e-01
-2.46715322e-01 2.69659236e-02 4.10853356e-01 2.03736156e-01
-7.32401609e-01 6.21924460e-01 -7.34660849e-02 -8.95411551e-01
-8.64228979e-02 2.01025605e-01 -3.14998060e-01 9.43285003e-02
3.96820545e-01 2.87412107e-01 -5.06718606e-02 -4.08831149e-01
5.35180807e-01 5.40001273e-01 2.81006515e-01 -3.38899076e-01
5.86750448e-01 1.28726840e+00 2.44556978e-01 -5.69837987e-01
-1.14670813e+00 -3.66766155e-01 -1.22171748e+00 -3.04480404e-01
1.29111922e+00 -1.28599346e+00 -9.03119624e-01 3.97258610e-01
-1.22901058e+00 -6.06736362e-01 -4.17400211e-01 5.07142544e-01
-7.52874970e-01 -6.24482408e-02 -5.19945681e-01 -5.56747854e-01
-1.79524332e-01 -1.40411639e+00 1.67060935e+00 6.96632564e-02
2.55220830e-01 -8.18829596e-01 -4.26456667e-02 2.15778872e-01
2.80319341e-02 7.28350997e-01 1.02442348e+00 2.89126754e-01
-1.14644384e+00 -1.96430683e-02 -5.61632156e-01 1.37714043e-01
-9.30935517e-03 -2.78274715e-01 -1.46865153e+00 4.65272404e-02
-1.19255148e-02 -3.93802494e-01 7.57407546e-01 6.09813452e-01
1.45396185e+00 1.42354995e-01 -1.74258739e-01 1.41131818e+00
1.76377010e+00 -1.70273874e-02 3.37048978e-01 3.75999212e-02
1.09379280e+00 3.21405768e-01 1.98371366e-01 6.47959590e-01
6.49117112e-01 2.82242507e-01 1.06961429e+00 -3.18776965e-01
-4.76631001e-02 -5.98232448e-01 -4.16755565e-02 9.43915963e-01
3.16959582e-02 2.00169310e-01 -8.86588097e-01 3.23049575e-01
-1.44454563e+00 -4.14155275e-01 1.68464646e-01 2.16219974e+00
2.03884870e-01 1.14450872e-01 -1.55478850e-01 -1.22596055e-01
1.56053960e-01 4.28038627e-01 -1.10789609e+00 -3.08675140e-01
7.17429370e-02 1.51670411e-01 7.46370852e-01 6.11359358e-01
-8.37485969e-01 7.35335648e-01 5.90480709e+00 -5.70613109e-02
-1.27035964e+00 3.91236916e-02 8.78768444e-01 -5.94356239e-01
-7.87148774e-01 -2.50361234e-01 -5.64309120e-01 3.33335716e-03
4.25280124e-01 3.04598927e-01 5.96723020e-01 9.92029250e-01
-7.40913227e-02 -1.38274789e-01 -1.36571395e+00 1.37592196e+00
9.19063911e-02 -1.58236456e+00 7.17100278e-02 6.29694462e-01
7.81563401e-01 7.06357300e-01 3.48933488e-02 3.35579030e-02
1.69620097e-01 -1.22931862e+00 7.87275791e-01 5.25093555e-01
1.25462401e+00 -7.96311617e-01 4.89690632e-01 4.07215595e-01
-1.34671116e+00 5.50178252e-02 -5.26841283e-01 -4.78241682e-01
2.31694013e-01 6.73997819e-01 -5.64432979e-01 3.06488007e-01
7.72286892e-01 8.93302321e-01 -1.15858391e-01 5.38660347e-01
-1.74236014e-01 -1.56922698e-01 -4.14189219e-01 1.08018503e-01
2.96603918e-01 -3.55604947e-01 1.04868427e-01 6.19621217e-01
3.08391929e-01 3.87471199e-01 2.99164206e-01 1.27448225e+00
-3.29895377e-01 -2.15319410e-01 -1.12207663e+00 3.61862034e-01
3.13236624e-01 1.18642533e+00 -9.60779488e-01 -2.50304937e-01
-5.70496678e-01 9.82595444e-01 6.59235775e-01 4.31283653e-01
-5.77035427e-01 -9.11688507e-02 8.32397938e-01 1.58267975e-01
3.76502156e-01 -5.07384479e-01 -7.34026372e-01 -1.11862433e+00
9.32163447e-02 1.89479694e-01 -8.24675336e-02 -1.09987175e+00
-9.57139611e-01 5.54347515e-01 -1.31189451e-01 -1.21819198e+00
1.17107794e-01 -8.93394172e-01 -1.25010088e-01 1.06926858e+00
-1.66421461e+00 -8.06182861e-01 -8.20739448e-01 6.22991383e-01
3.66987288e-01 2.56720275e-01 7.93588161e-01 -2.94262413e-02
-1.59673616e-01 -1.37793362e-01 -2.26918787e-01 3.62009734e-01
-1.12530753e-01 -1.16750991e+00 6.83430254e-01 1.15982763e-01
-8.01633373e-02 1.16456322e-01 -1.78245202e-01 -4.26610112e-01
-1.85964477e+00 -1.37759161e+00 2.99245089e-01 -7.56036162e-01
1.22913100e-01 -8.10313046e-01 -4.79783326e-01 7.21718609e-01
-4.62650210e-01 7.10080326e-01 5.01973867e-01 -2.23678708e-01
-6.15331173e-01 -3.92913586e-03 -1.25633132e+00 2.04636425e-01
1.56482196e+00 -1.05829751e+00 -2.12170660e-01 2.92870879e-01
1.15574026e+00 -1.16426814e+00 -1.09899187e+00 2.94035375e-01
6.65436625e-01 -1.03547800e+00 1.15767241e+00 -2.21476525e-01
7.48359621e-01 -2.94274032e-01 -7.34010398e-01 -9.65136111e-01
-4.03438658e-02 3.43958944e-01 -1.02588035e-01 3.57121706e-01
3.72083515e-01 -4.04542536e-01 1.01726961e+00 6.18008137e-01
-2.17100233e-01 -1.30187035e+00 -1.08157659e+00 -2.30056524e-01
-1.12161778e-01 -8.92570853e-01 8.18509758e-01 5.85724652e-01
-3.53989422e-01 3.90351683e-01 1.53682232e-01 4.54612523e-01
7.71643519e-01 5.68031132e-01 7.95659840e-01 -1.23091543e+00
-4.29507606e-02 -9.98107344e-02 -7.59004474e-01 -1.43404877e+00
1.40258491e-01 -1.10003102e+00 -7.28508085e-02 -1.86612630e+00
1.89013481e-01 -3.12204748e-01 7.44490139e-03 2.06976995e-01
3.94563287e-01 4.59862024e-01 -9.77637805e-03 8.73920918e-02
-3.85652095e-01 7.25331128e-01 1.52427971e+00 -1.75591245e-01
-2.40105003e-01 -3.16682011e-01 -1.98330477e-01 7.17557251e-01
4.62763131e-01 -2.68864810e-01 -4.84258533e-01 -7.85171092e-01
3.41238439e-01 3.83066297e-01 4.70336586e-01 -9.17908549e-01
-2.29731366e-01 -9.85826999e-02 9.67132568e-01 -1.03884327e+00
7.89949477e-01 -8.21631372e-01 -2.15694889e-01 3.21912289e-01
-7.25388378e-02 -1.06232107e-01 1.37848069e-03 5.60055673e-01
1.67718217e-01 6.32830381e-01 6.08397543e-01 -5.11968374e-01
-6.49558723e-01 9.76880431e-01 1.77140832e-01 -1.67732194e-01
8.53793979e-01 -4.54184145e-01 1.46316774e-02 -1.97821885e-01
-6.72951221e-01 2.81809807e-01 9.21298206e-01 2.99983144e-01
1.08980262e+00 -1.40369141e+00 -1.31240949e-01 7.94389844e-01
2.32902735e-01 1.03732526e+00 2.52987832e-01 4.71430749e-01
-9.27249908e-01 4.11674410e-01 -2.54310489e-01 -1.21284819e+00
-6.51055634e-01 4.26767498e-01 7.50487089e-01 1.45577818e-01
-1.25182033e+00 8.53344858e-01 8.56638908e-01 -9.45732832e-01
2.28566855e-01 -8.13769519e-01 1.27702905e-02 -4.40995187e-01
4.45530325e-01 2.78789783e-03 6.00131638e-02 -6.33704364e-01
-3.06322604e-01 9.68662560e-01 3.93881589e-01 -3.47295821e-01
1.49925399e+00 -1.63299665e-01 -1.67785175e-02 8.51072848e-01
1.85603130e+00 -4.03280199e-01 -1.90664256e+00 -3.38365942e-01
-5.70245087e-01 -6.65943265e-01 3.07423890e-01 -2.25960746e-01
-1.46049070e+00 1.30342984e+00 5.40749669e-01 -3.41732115e-01
8.89230967e-01 3.52090210e-01 6.83936357e-01 3.23990643e-01
6.82518303e-01 -7.45496154e-01 2.98429430e-01 7.27853417e-01
7.29076087e-01 -1.06677699e+00 2.16028005e-01 -3.73599648e-01
-1.97539657e-01 1.15848005e+00 6.83867097e-01 -3.43551397e-01
9.94958639e-01 2.97440797e-01 1.14181682e-01 -8.18833232e-01
-3.97435099e-01 1.10336235e-02 -7.64732286e-02 5.26432872e-01
1.14553414e-01 2.21978039e-01 6.08338952e-01 1.48227915e-01
-2.90197462e-01 5.73557988e-02 3.94949615e-01 7.91113973e-01
-4.02685851e-01 -4.49591488e-01 -8.78795609e-02 5.47149718e-01
1.36842817e-01 1.09681420e-01 -1.91286787e-01 6.46507204e-01
2.28137448e-01 1.92232892e-01 7.51948953e-01 -7.38446295e-01
5.87061107e-01 1.94341608e-03 6.65583014e-01 -9.47654605e-01
-4.14759964e-02 -6.80527696e-03 -3.51212263e-01 -1.16469812e+00
-3.45224619e-01 -3.39611709e-01 -1.76408076e+00 -1.07991062e-01
1.67122856e-01 -5.23263812e-01 1.09103012e+00 9.66035366e-01
4.04954851e-01 6.28690720e-01 1.03688765e+00 -1.29472148e+00
5.83481044e-02 -5.33157766e-01 -7.56251097e-01 1.09311484e-01
5.47158659e-01 -7.72719443e-01 -1.79743886e-01 -2.27444932e-01] | [8.468034744262695, -2.9555981159210205] |
fa140bb7-00ab-4a77-acd3-0eaad7c91678 | single-shot-neural-relighting-and-svbrdf | null | null | https://www.ecva.net/papers/eccv_2020/papers_ECCV/html/3151_ECCV_2020_paper.php | https://www.ecva.net/papers/eccv_2020/papers_ECCV/papers/123640086.pdf | Single-Shot Neural Relighting and SVBRDF Estimation | We present a novel physically-motivated deep network for joint shape and material estimation, as well as relighting under novel illumination conditions, using a single image captured by a mobile phone camera. Our physically-based modeling leverages a deep cascaded architecture trained on a large-scale synthetic dataset that consists of complex shapes with microfacet SVBRDF. In contrast to prior works that train rendering layers subsequent to inverse rendering, we propose deep feature sharing and joint training that transfer insights across both tasks, to achieve significant improvements in both reconstruction and relighting. We demonstrate in extensive qualitative and quantitative experiments that our network generalizes very well to real images, achieving high-quality shape and material estimation, as well as image-based relighting. Code, models and data will be publicly released. | ['Shen Sang', 'Manmohan Chandraker'] | null | null | null | null | eccv-2020-8 | ['svbrdf-estimation'] | ['computer-vision'] | [ 2.16151416e-01 1.00248996e-02 4.57882911e-01 -3.25455666e-01
-9.11747575e-01 -6.92198575e-01 8.25926125e-01 -3.73594046e-01
1.42373279e-01 6.84763432e-01 3.65820140e-01 -2.34319374e-01
1.43325299e-01 -9.54799592e-01 -1.18550313e+00 -2.50620842e-01
6.57553300e-02 5.80292583e-01 -1.82050452e-01 -1.38153479e-01
1.37715310e-01 8.42795491e-01 -1.31793320e+00 3.33779097e-01
7.21690416e-01 1.07657194e+00 -2.31364831e-01 9.06874299e-01
1.99511722e-01 5.58971286e-01 -7.64324665e-02 -7.43778825e-01
5.46895087e-01 -8.64026919e-02 -7.11763799e-01 3.95533681e-01
8.54065716e-01 -1.04097140e+00 -4.29494500e-01 4.59356248e-01
4.23143744e-01 1.48371860e-01 7.89809883e-01 -8.46323371e-01
-1.12616873e+00 4.44234023e-03 -6.47305071e-01 -5.40322781e-01
4.97563303e-01 5.22300541e-01 8.08355570e-01 -1.24888575e+00
6.56660557e-01 1.42583096e+00 1.13536453e+00 3.21997762e-01
-1.50887775e+00 -4.73580122e-01 -1.96126074e-01 -4.54497218e-01
-1.11104476e+00 -5.13704538e-01 8.24770153e-01 -4.07665640e-01
8.42837036e-01 9.97679830e-02 8.33182454e-01 1.13169789e+00
1.83641911e-01 5.71366847e-01 9.79911029e-01 -1.03324093e-01
-2.97760721e-02 -2.59561926e-01 -5.61392605e-01 8.20876539e-01
1.61811516e-01 3.99971098e-01 -3.14215094e-01 -1.63959876e-01
1.49938452e+00 2.36196406e-02 -2.13964999e-01 -6.11590147e-01
-1.19956052e+00 6.16447151e-01 4.93339032e-01 -3.72127950e-01
-3.02374005e-01 8.57990503e-01 -2.90405396e-02 -2.98957154e-02
7.53828287e-01 4.06350255e-01 -3.37800205e-01 -9.35144350e-02
-8.20680499e-01 4.49718237e-01 6.85658455e-01 9.34390008e-01
8.65217984e-01 3.45331907e-01 9.70004499e-02 7.67253160e-01
4.22728837e-01 8.57129395e-01 -3.66201639e-01 -1.73548436e+00
1.11538805e-01 2.58329064e-01 5.27165711e-01 -9.97364581e-01
-1.78629294e-01 -5.61849713e-01 -6.01704001e-01 7.65650809e-01
1.45938545e-01 -1.80452779e-01 -9.72703040e-01 1.56345606e+00
3.12924087e-01 4.25908357e-01 -1.87651768e-01 8.76343727e-01
9.41409349e-01 5.23679376e-01 -2.10732445e-01 4.34242755e-01
1.02967691e+00 -1.04246712e+00 -2.63785928e-01 -4.15287092e-02
2.26148114e-01 -9.64140534e-01 1.28547001e+00 5.99658787e-01
-1.65121651e+00 -3.24924618e-01 -8.83642435e-01 -6.99671686e-01
9.30073112e-02 7.50617264e-03 9.75454271e-01 3.11715901e-01
-1.37274075e+00 9.40072894e-01 -9.33215082e-01 -1.54109567e-01
8.64377618e-01 1.80497184e-01 -3.49852562e-01 -1.34813040e-01
-5.99893391e-01 4.88844365e-01 -4.70341623e-01 4.71646562e-02
-1.21618307e+00 -1.06417775e+00 -8.99578512e-01 -8.73304680e-02
-1.40679896e-01 -1.16823232e+00 1.27908778e+00 -9.75271344e-01
-1.73298216e+00 9.94253933e-01 7.74215981e-02 -4.13409948e-01
8.34303141e-01 -3.10834289e-01 4.94188964e-02 2.42576182e-01
-1.87645525e-01 7.81527042e-01 7.70456314e-01 -1.83447647e+00
2.20263191e-02 -5.23215160e-02 1.58636510e-01 1.68418914e-01
-7.70098865e-02 -3.13875794e-01 -3.50990295e-01 -8.05066526e-01
-7.47879520e-02 -6.62325859e-01 -1.00111738e-01 8.82898808e-01
-3.98387402e-01 5.29073536e-01 8.47234070e-01 -8.67984235e-01
3.19507986e-01 -1.87029600e+00 5.90373427e-02 -4.82595861e-02
2.80301958e-01 -2.71359347e-02 -5.12217462e-01 4.82878447e-01
-3.32779288e-02 2.35182673e-01 -4.91765708e-01 -1.14907396e+00
1.14006653e-01 1.21803936e-02 -6.60546362e-01 4.55486238e-01
3.04073364e-01 1.36613679e+00 -7.06041455e-01 -1.60368085e-01
4.95101452e-01 1.11658740e+00 -8.92849505e-01 2.56621748e-01
-3.31941783e-01 8.69441867e-01 -1.40360892e-01 7.28403151e-01
1.09871137e+00 -3.45407337e-01 -1.12285204e-01 -4.20088828e-01
5.79173788e-02 1.79630756e-01 -7.70391703e-01 2.02696538e+00
-9.18250680e-01 6.70402348e-01 3.64511013e-01 -5.36032200e-01
7.20798612e-01 1.17947459e-01 6.66538358e-01 -6.79039776e-01
9.74289849e-02 2.33270243e-01 -6.20284021e-01 -1.80973992e-01
5.75173140e-01 -2.08782226e-01 2.47036308e-01 5.19023776e-01
-1.56484798e-01 -8.23448658e-01 -4.87262875e-01 3.31983745e-01
9.94471788e-01 8.94913375e-01 -2.66915500e-01 -1.32113099e-01
-1.41387647e-02 -1.85135007e-01 1.66029140e-01 4.39748496e-01
4.38560396e-01 1.31162024e+00 1.81629226e-01 -5.90436101e-01
-1.51972830e+00 -1.61816406e+00 -2.02054769e-01 6.59991324e-01
2.78233051e-01 -1.35619357e-01 -7.01103330e-01 -2.92510062e-01
3.12432706e-01 6.33723974e-01 -7.10662067e-01 -1.72814410e-02
-7.76813149e-01 -2.54176289e-01 2.39197269e-01 5.22941053e-01
6.11163318e-01 -9.49943423e-01 -2.59144604e-01 7.30750114e-02
1.42544925e-01 -1.25724828e+00 -4.58733171e-01 -4.49089885e-01
-7.73309290e-01 -1.00281692e+00 -7.80277729e-01 -4.91480798e-01
4.54402298e-01 3.62359613e-01 1.61533964e+00 5.09486437e-01
-2.39478543e-01 6.41905725e-01 6.42250776e-02 -1.31448731e-01
-5.56759655e-01 -1.63359895e-01 -4.60469365e-01 -2.81951111e-02
-7.60400891e-01 -1.06334662e+00 -1.21662354e+00 2.08222285e-01
-8.61078978e-01 4.87114996e-01 2.91638821e-01 5.95555902e-01
4.80779946e-01 -5.59745252e-01 1.59214720e-01 -6.26345694e-01
3.41418087e-01 -2.68126190e-01 -7.29327619e-01 -7.85389543e-03
-2.73695379e-01 -1.15387060e-01 7.30900049e-01 -9.37609598e-02
-1.23737776e+00 -1.11278467e-01 -4.46319312e-01 -5.84373116e-01
-6.35204986e-02 -1.33609716e-02 5.61611503e-02 -5.43928385e-01
5.53091764e-01 6.44687042e-02 9.11319703e-02 -4.98025268e-01
4.13217038e-01 2.99508423e-01 6.51091754e-01 -7.58334219e-01
1.07701206e+00 1.09506524e+00 2.36001849e-01 -4.89627510e-01
-7.10912466e-01 3.84519875e-01 -3.92753333e-01 -8.07500407e-02
6.96005106e-01 -1.20802426e+00 -1.01985884e+00 7.01198041e-01
-1.19277763e+00 -1.01954496e+00 -5.49676538e-01 2.08248109e-01
-8.54977548e-01 4.37693328e-01 -7.27384508e-01 -5.20512640e-01
-3.98974955e-01 -1.14372194e+00 1.86139083e+00 -1.22048885e-01
-5.89223020e-03 -1.19657063e+00 -8.00933037e-03 5.43922961e-01
5.69138110e-01 6.83381915e-01 7.40042984e-01 4.22468692e-01
-1.09251511e+00 1.21960089e-01 -6.40938342e-01 2.28261992e-01
1.36262089e-01 3.29286546e-01 -1.08012426e+00 -3.11864555e-01
-2.52127469e-01 -5.96844435e-01 9.51861143e-01 3.96033764e-01
1.34902310e+00 -2.21649870e-01 -1.87796086e-01 1.31806791e+00
1.42520988e+00 -3.19073260e-01 8.14279556e-01 -1.05279900e-01
1.15239251e+00 3.39888424e-01 1.25168547e-01 4.70047265e-01
6.24471486e-01 7.58562624e-01 8.28524351e-01 -6.06206179e-01
-8.13282549e-01 -4.42718595e-01 9.42578614e-02 5.12494624e-01
-1.46655500e-01 -3.24405879e-01 -6.89451337e-01 3.99302691e-01
-1.44447243e+00 -5.47238290e-01 -9.00932029e-02 2.08733392e+00
6.61891162e-01 -1.75256893e-01 -1.93886310e-02 -3.14861149e-01
2.28628278e-01 1.64777204e-01 -8.23060751e-01 -3.52662265e-01
-1.13286078e-01 3.72614622e-01 3.49613577e-01 9.42997515e-01
-7.86200225e-01 9.33604658e-01 7.10243225e+00 6.03042066e-01
-1.26735771e+00 1.55165717e-01 9.17992890e-01 -1.66840240e-01
-1.31124556e+00 3.03247944e-02 -7.36455470e-02 7.67918751e-02
6.04403913e-01 4.52624857e-01 9.99076545e-01 4.30300295e-01
3.55514437e-01 2.50667464e-02 -8.98305237e-01 1.09207535e+00
-2.92399190e-02 -1.77234793e+00 3.38435233e-01 3.19171011e-01
1.07116079e+00 1.93866774e-01 3.60129565e-01 -8.26957524e-02
7.64279604e-01 -1.31471395e+00 9.80231106e-01 6.77853286e-01
1.23987997e+00 -5.94257414e-01 -3.85682359e-02 -4.70126420e-02
-1.20803738e+00 1.82686299e-01 -1.33535475e-01 3.40812374e-03
3.10251743e-01 6.59171999e-01 -4.07854259e-01 5.53713500e-01
4.34652835e-01 1.10398972e+00 -2.97524303e-01 7.05293596e-01
-1.13501161e-01 3.61608922e-01 -2.36418456e-01 6.03193521e-01
-5.62345833e-02 -3.73661518e-01 4.44600672e-01 1.02153885e+00
4.47404474e-01 -7.23997504e-02 -8.05597529e-02 1.46680820e+00
-4.88764733e-01 -1.96749702e-01 -7.06435204e-01 8.05236325e-02
2.54918009e-01 1.40837991e+00 -6.69224143e-01 -2.31442064e-01
-2.55914181e-01 1.16961277e+00 4.67599422e-01 6.30996406e-01
-9.64985788e-01 -6.67373044e-03 8.83707881e-01 4.54072982e-01
2.78870821e-01 -6.89576507e-01 -6.85810208e-01 -1.41539979e+00
-9.73998085e-02 -4.12764043e-01 -4.99302059e-01 -1.29624557e+00
-1.29731166e+00 4.69076663e-01 -2.80894428e-01 -1.15979683e+00
4.17009592e-02 -6.53408229e-01 -7.34615982e-01 8.57114077e-01
-1.66991436e+00 -1.78782940e+00 -6.06845319e-01 4.18905944e-01
2.24891737e-01 2.60306120e-01 7.77223170e-01 1.07631259e-01
-1.92857534e-01 4.19999868e-01 1.27003193e-01 3.49464342e-02
3.89184564e-01 -1.15158236e+00 1.04326975e+00 6.08217061e-01
4.67295572e-02 5.58881223e-01 5.98192871e-01 -5.87356687e-01
-1.74898553e+00 -1.22461724e+00 -4.74343114e-02 -5.87307036e-01
1.73364148e-01 -6.53281033e-01 -8.18853855e-01 7.88041592e-01
4.40471381e-01 4.06570613e-01 1.43141642e-01 -2.84227997e-01
-5.19845009e-01 -3.62663306e-02 -1.46110773e+00 6.97495282e-01
1.68091273e+00 -6.43766999e-01 2.63939705e-02 5.91091514e-01
8.39772046e-01 -7.14441001e-01 -1.07373023e+00 4.13446307e-01
9.22081769e-01 -1.36343932e+00 1.37119377e+00 -2.08449304e-01
8.84933114e-01 -7.78301582e-02 -2.30708137e-01 -1.33245599e+00
-1.83744967e-01 -8.64060700e-01 -1.40469536e-01 8.51291895e-01
1.67819694e-01 -7.03899324e-01 8.10627878e-01 4.88394320e-01
-4.25329328e-01 -9.75453854e-01 -5.83063602e-01 -6.26972437e-01
4.03339624e-01 -5.44372916e-01 9.53206539e-01 6.64798498e-01
-8.33186090e-01 -4.17555384e-02 -3.90523285e-01 -1.70229916e-02
8.17718685e-01 4.73102152e-01 9.88162220e-01 -1.08649516e+00
-4.23287570e-01 -2.05730230e-01 -2.56530121e-02 -1.24180937e+00
2.50619501e-01 -8.21504354e-01 -1.39148965e-01 -1.87043107e+00
1.34063646e-01 -5.41335285e-01 3.38544995e-01 2.97631592e-01
2.97670424e-01 1.05753255e+00 -6.76411167e-02 1.81475669e-01
-2.87215918e-01 9.50315475e-01 1.80436778e+00 9.45909545e-02
1.27597526e-01 -3.83798897e-01 -7.75644660e-01 8.09442580e-01
3.75234872e-01 -3.22551951e-02 -4.58531708e-01 -1.00103259e+00
5.32024443e-01 1.05828546e-01 9.01195884e-01 -8.88536453e-01
-3.65815848e-01 -1.11438401e-01 9.06257331e-01 -4.12963778e-01
9.68357444e-01 -7.11534381e-01 4.17476147e-01 1.85909241e-01
-2.46068224e-01 -1.54790908e-01 3.79727155e-01 4.46391732e-01
2.65241683e-01 3.52319688e-01 9.62103128e-01 -1.64971501e-01
-1.30071044e-01 6.36372983e-01 2.70663381e-01 8.39301944e-02
6.95117235e-01 -4.87817079e-01 -2.51014918e-01 -9.17160571e-01
-5.67725956e-01 -2.20641971e-01 1.21477437e+00 1.48430347e-01
9.78246033e-01 -1.56296265e+00 -8.16821635e-01 3.85514796e-01
-2.14776516e-01 3.40624571e-01 2.93366313e-01 4.99336660e-01
-9.74475265e-01 -2.03553557e-01 -6.79351985e-02 -6.33230031e-01
-6.79780841e-01 1.50769994e-01 6.76202655e-01 2.64094286e-02
-7.42242515e-01 6.75217927e-01 5.31985879e-01 -8.71841609e-01
-2.89792508e-01 -4.42962557e-01 5.51344693e-01 -6.84545934e-01
2.44479194e-01 1.90126777e-01 1.46944404e-01 -5.99425018e-01
-9.85345840e-02 1.04815340e+00 3.72952491e-01 -3.03792179e-01
1.62598979e+00 -2.78809547e-01 -1.36240069e-02 4.90047298e-02
1.44486344e+00 3.82064342e-01 -2.15159488e+00 1.86690141e-03
-8.88750970e-01 -6.51339054e-01 5.97143881e-02 -9.26805019e-01
-1.46159172e+00 9.28641260e-01 2.91044295e-01 -2.03209013e-01
9.04446065e-01 -2.62759894e-01 1.23798645e+00 2.00807407e-01
4.30796355e-01 -5.51263690e-01 4.00326103e-01 4.51623291e-01
1.36190772e+00 -1.25137138e+00 2.30176389e-01 -5.88079989e-01
-3.11583012e-01 1.07042599e+00 3.28647882e-01 -6.19392872e-01
8.46843958e-01 5.46546817e-01 -1.17010221e-01 -2.95759976e-01
-5.79593778e-01 3.56103450e-01 3.27434033e-01 6.41840219e-01
4.82031643e-01 -5.44164218e-02 4.88024354e-01 9.65335071e-02
-2.86042809e-01 1.26852095e-01 4.86678034e-01 7.11962819e-01
3.54273580e-02 -8.75325918e-01 -2.65266389e-01 2.39711285e-01
-1.79529831e-01 -1.80468380e-01 -3.34664613e-01 6.29658341e-01
-5.03077805e-02 6.43431425e-01 2.93264806e-01 -2.28987753e-01
1.76172689e-01 -6.09228730e-01 8.51164758e-01 -5.94474316e-01
-5.67276955e-01 2.14501447e-03 1.23291463e-01 -8.89358342e-01
-2.48300523e-01 -3.84977549e-01 -1.34391999e+00 -6.29073560e-01
1.00789741e-01 -4.97372478e-01 7.29339480e-01 7.21563995e-01
7.33426511e-01 4.44538742e-01 8.06107700e-01 -1.51624405e+00
-1.64404929e-01 -7.00189114e-01 -5.19086480e-01 7.54358530e-01
5.52288830e-01 -6.50191724e-01 -3.47185642e-01 -5.58088906e-03] | [9.299371719360352, -3.2532169818878174] |
08f962d5-5f6d-45cf-b5b8-7a6598f25ddd | re-iqa-unsupervised-learning-for-image | 2304.00451 | null | https://arxiv.org/abs/2304.00451v2 | https://arxiv.org/pdf/2304.00451v2.pdf | Re-IQA: Unsupervised Learning for Image Quality Assessment in the Wild | Automatic Perceptual Image Quality Assessment is a challenging problem that impacts billions of internet, and social media users daily. To advance research in this field, we propose a Mixture of Experts approach to train two separate encoders to learn high-level content and low-level image quality features in an unsupervised setting. The unique novelty of our approach is its ability to generate low-level representations of image quality that are complementary to high-level features representing image content. We refer to the framework used to train the two encoders as Re-IQA. For Image Quality Assessment in the Wild, we deploy the complementary low and high-level image representations obtained from the Re-IQA framework to train a linear regression model, which is used to map the image representations to the ground truth quality scores, refer Figure 1. Our method achieves state-of-the-art performance on multiple large-scale image quality assessment databases containing both real and synthetic distortions, demonstrating how deep neural networks can be trained in an unsupervised setting to produce perceptually relevant representations. We conclude from our experiments that the low and high-level features obtained are indeed complementary and positively impact the performance of the linear regressor. A public release of all the codes associated with this work will be made available on GitHub. | ['Alan C. Bovik', 'Sandeep Mishra', 'Avinab Saha'] | 2023-04-02 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Saha_Re-IQA_Unsupervised_Learning_for_Image_Quality_Assessment_in_the_Wild_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Saha_Re-IQA_Unsupervised_Learning_for_Image_Quality_Assessment_in_the_Wild_CVPR_2023_paper.pdf | cvpr-2023-1 | ['image-quality-assessment'] | ['computer-vision'] | [ 4.36892480e-01 -1.32873684e-01 1.30389914e-01 -4.57969397e-01
-1.13920939e+00 -5.59166670e-01 5.12331307e-01 2.06212327e-01
-3.70857775e-01 2.80179083e-01 3.35370719e-01 -8.64794552e-02
5.17569073e-02 -9.22779977e-01 -9.53394294e-01 -3.79075646e-01
-3.10565867e-02 -2.11417735e-01 3.79306786e-02 -9.06594992e-02
2.53168017e-01 3.10042799e-01 -1.74383068e+00 5.45971692e-01
1.00250459e+00 1.38830948e+00 1.29390329e-01 9.99909639e-01
4.53639984e-01 7.82504082e-01 -6.62390947e-01 -7.06329226e-01
5.29961228e-01 -3.74494344e-01 -5.92375636e-01 3.11436564e-01
6.00300968e-01 -5.14369667e-01 -4.11663920e-01 1.22775137e+00
5.24069667e-01 -2.55190413e-02 6.78317666e-01 -1.28275585e+00
-9.13101077e-01 -3.33404727e-02 -3.81274492e-01 1.69812590e-01
5.05812407e-01 5.41589916e-01 1.10247779e+00 -8.56397867e-01
3.38637292e-01 1.29757464e+00 5.26985109e-01 5.07780574e-02
-1.37057376e+00 -4.33339089e-01 -1.10680208e-01 2.98602372e-01
-1.26551437e+00 -6.53358757e-01 6.31857216e-01 -5.48090994e-01
8.39278162e-01 8.14580545e-03 5.08754730e-01 9.66072619e-01
2.48584807e-01 4.54931647e-01 1.42013144e+00 -5.30443728e-01
2.42581218e-01 1.55459151e-01 -4.19495821e-01 7.31806397e-01
5.78620471e-02 2.81495363e-01 -5.44677258e-01 6.03617914e-02
8.83329153e-01 -2.38640293e-01 -1.53473467e-01 -3.75922203e-01
-1.13230419e+00 8.41483057e-01 6.80397987e-01 9.91951227e-02
-5.98358393e-01 2.43707985e-01 8.71344656e-02 5.07748246e-01
5.33418179e-01 4.78308886e-01 -1.14422724e-01 -1.08044013e-01
-9.28981662e-01 1.63034678e-01 4.54982489e-01 5.01539171e-01
8.11361074e-01 -8.65063071e-02 -2.71894991e-01 1.01231921e+00
2.33268782e-01 7.14828253e-01 3.53874713e-01 -1.42883813e+00
2.69548774e-01 4.18915898e-01 2.99026430e-01 -1.28999043e+00
-3.49589735e-02 -5.03482103e-01 -6.08599901e-01 7.29944110e-01
3.10283154e-01 4.58801389e-02 -9.60734427e-01 1.71041989e+00
1.46432724e-02 4.65595536e-02 1.46889538e-01 8.87966156e-01
7.18556702e-01 8.45830917e-01 1.22226164e-01 1.61580727e-01
1.33783770e+00 -8.79658163e-01 -4.35233802e-01 -1.63909510e-01
6.53467476e-02 -8.52945924e-01 1.30942535e+00 5.11119306e-01
-1.52492476e+00 -9.45442855e-01 -1.21518314e+00 -1.08573787e-01
-3.47838372e-01 3.23700368e-01 1.50082350e-01 5.80789506e-01
-1.32291079e+00 8.28943789e-01 -5.68539619e-01 -2.34351844e-01
6.62032783e-01 1.54738903e-01 -3.68860453e-01 -2.94115752e-01
-1.13354778e+00 9.08045053e-01 7.37967789e-02 -1.60418078e-01
-1.13655150e+00 -5.56083798e-01 -8.70249689e-01 4.86843884e-02
2.64736444e-01 -6.45677745e-01 1.24662149e+00 -1.43844235e+00
-1.49762917e+00 1.02442646e+00 -2.90447213e-02 -2.42708772e-01
3.44071329e-01 -2.33452782e-01 -4.42511499e-01 5.49032271e-01
2.51674205e-01 8.35755467e-01 1.14485323e+00 -1.53649354e+00
-6.42912686e-01 -6.41340986e-02 3.02831084e-01 1.50650814e-01
-4.22219366e-01 9.97091755e-02 -6.57410502e-01 -7.32974231e-01
-1.68328494e-01 -5.34385681e-01 -8.16098601e-02 3.49942446e-01
-1.47643268e-01 1.54469892e-01 2.44644850e-01 -8.27676594e-01
9.05357182e-01 -2.12285256e+00 2.61109084e-01 2.40467325e-01
1.77895129e-01 2.45247036e-01 -5.61316490e-01 3.32598120e-01
-3.28405648e-02 2.32198700e-01 -4.28650200e-01 -3.06926817e-01
7.85022080e-02 -9.76198390e-02 -5.31527586e-02 3.94827068e-01
6.17636800e-01 9.68782127e-01 -1.07834339e+00 -2.22793788e-01
3.82716626e-01 5.77976227e-01 -5.10968983e-01 6.63443983e-01
-2.25807745e-02 3.57613176e-01 8.70772302e-02 4.34003472e-01
5.98528564e-01 -3.74354094e-01 -3.09424922e-02 -5.74977636e-01
1.42115816e-01 1.90384895e-01 -1.01449358e+00 1.65721059e+00
-7.93558717e-01 7.27647424e-01 -1.49743617e-01 -8.01453471e-01
7.01221764e-01 3.73899400e-01 4.09675270e-01 -1.24362612e+00
-2.21298859e-02 1.57590553e-01 -5.01008891e-02 -4.84800965e-01
3.21277171e-01 -8.69911537e-02 -7.63757899e-02 4.58171755e-01
3.84358913e-01 -3.21929604e-01 2.05517158e-01 2.59919703e-01
1.03577971e+00 1.34714589e-01 2.03503117e-01 3.21264528e-02
6.16395712e-01 -3.64619762e-01 3.29992801e-01 7.19294786e-01
-1.94124982e-01 8.40469480e-01 4.83498484e-01 -1.79823697e-01
-1.44518352e+00 -1.52109432e+00 -5.97406104e-02 9.71745133e-01
3.47989649e-02 -3.98783386e-01 -7.61267364e-01 -6.20019436e-01
-1.05345361e-01 3.50795239e-01 -7.06822276e-01 -2.75513053e-01
-2.13444337e-01 -4.60759431e-01 2.38981724e-01 3.02076399e-01
5.12219489e-01 -1.19251442e+00 -5.53828478e-01 4.27592546e-02
-3.05080175e-01 -1.29953432e+00 -2.40448326e-01 -5.31618223e-02
-4.18664038e-01 -1.05808842e+00 -8.78685951e-01 -4.79591340e-01
7.05748439e-01 2.05628663e-01 1.56542206e+00 2.51957566e-01
-3.05851787e-01 5.21258354e-01 -4.99192595e-01 -3.31791282e-01
-6.77854419e-01 -3.54349524e-01 -1.68221787e-01 2.71029860e-01
-3.59109007e-02 -6.91949964e-01 -1.04036546e+00 1.94704220e-01
-1.16080523e+00 1.52952243e-02 6.81501329e-01 6.23896599e-01
5.54293513e-01 3.13697010e-01 5.22050321e-01 -5.26510537e-01
7.37849653e-01 -4.06424373e-01 -5.45704663e-01 -1.54665206e-02
-5.53141415e-01 1.71679929e-02 6.83807969e-01 -1.15535706e-01
-8.39597225e-01 -1.68381095e-01 -3.39178383e-01 -2.91645557e-01
-3.22592586e-01 3.68319571e-01 -2.40953267e-01 -1.38588890e-01
7.09476471e-01 3.37923020e-02 -1.49938300e-01 -3.69753212e-01
6.38824105e-01 7.08260357e-01 8.30195546e-01 -4.54973400e-01
1.04161108e+00 4.04787660e-01 -5.29789142e-02 -4.69171792e-01
-9.10355389e-01 -2.68791020e-01 -5.76857269e-01 -4.93607461e-01
9.14751172e-01 -1.20088303e+00 -3.54425520e-01 6.79410338e-01
-8.96327972e-01 -5.50006986e-01 -4.97509837e-01 1.50635734e-01
-7.38703907e-01 3.62332344e-01 -5.18535316e-01 -6.43465996e-01
-2.20119208e-01 -1.33397293e+00 1.27178776e+00 1.22697882e-01
3.82527597e-02 -8.69349718e-01 4.93913796e-03 2.86549121e-01
4.63038027e-01 3.93884301e-01 8.33961070e-01 1.09154806e-01
-5.01217484e-01 -1.39520347e-01 -5.52149534e-01 8.56219828e-01
1.37167305e-01 3.26822735e-02 -1.12220693e+00 -4.49928969e-01
-1.38516098e-01 -7.89734125e-01 8.71645689e-01 3.73065084e-01
1.22846913e+00 -1.76025033e-01 3.52418303e-01 6.13059700e-01
1.63842607e+00 -7.72419721e-02 9.80065167e-01 3.38729769e-01
4.76137996e-01 5.49444675e-01 5.02964199e-01 3.27285528e-01
3.55693042e-01 8.93433332e-01 4.23183113e-01 -5.31888485e-01
-3.94870102e-01 -3.67101133e-01 4.82398868e-01 7.19915748e-01
9.68831033e-02 -2.92644054e-01 -8.13371360e-01 5.54343581e-01
-1.50628912e+00 -8.72777462e-01 2.53788143e-01 2.24303818e+00
9.04757738e-01 1.44120991e-01 2.81685829e-01 3.35899919e-01
4.70717728e-01 1.43822014e-01 -4.52214092e-01 -4.26214010e-01
2.23930795e-02 3.42443973e-01 2.29696557e-01 4.23081160e-01
-1.05412972e+00 6.43242657e-01 6.55412531e+00 6.07189953e-01
-1.05562079e+00 1.39709860e-01 9.87741649e-01 -9.83827561e-02
-3.53809863e-01 -2.44965449e-01 9.32043195e-02 5.60836315e-01
1.29695702e+00 -5.79390042e-02 6.40162826e-01 5.30062854e-01
4.55128342e-01 -1.69020057e-01 -1.11322796e+00 1.10578883e+00
1.57167330e-01 -1.11724997e+00 1.04179285e-01 1.66791957e-02
8.82904112e-01 -4.60248254e-02 5.39479554e-01 5.72279617e-02
4.36601907e-01 -1.31537747e+00 8.83482337e-01 6.15551054e-01
1.22935379e+00 -7.08276987e-01 5.95853567e-01 -6.24134429e-02
-1.04731905e+00 -2.09859937e-01 -3.50870460e-01 -2.69955792e-03
1.17186680e-01 7.21662641e-01 -3.72372329e-01 6.13949060e-01
8.80187273e-01 7.65009642e-01 -1.06052363e+00 1.04151857e+00
-3.17386895e-01 5.19221425e-01 1.47070229e-01 6.42498076e-01
-1.95406489e-02 -3.99106853e-02 1.94027275e-01 1.11375129e+00
3.45537066e-01 -8.96240026e-02 -3.38936038e-03 9.94077981e-01
-3.39440137e-01 9.78416353e-02 -4.03179854e-01 -1.27758346e-02
1.81734428e-01 1.27150059e+00 -5.08571684e-01 -4.63597178e-01
-5.06899953e-01 1.10802317e+00 2.80580938e-01 3.98727834e-01
-7.08139360e-01 -3.80718708e-01 5.76462805e-01 6.71802610e-02
2.87000984e-01 -3.27153727e-02 -2.29048908e-01 -1.18995869e+00
1.80417106e-01 -1.39399624e+00 2.03535073e-02 -1.17337191e+00
-1.46639788e+00 7.99408555e-01 -1.06087431e-01 -1.41888607e+00
-2.84836978e-01 -6.76137924e-01 -6.20521784e-01 1.05339825e+00
-1.72818935e+00 -8.99262905e-01 -6.22225881e-01 5.86627662e-01
3.51093739e-01 -6.77110627e-02 7.37712801e-01 2.83775687e-01
-2.93038368e-01 6.37057006e-01 9.71582755e-02 2.87514746e-01
7.91144550e-01 -1.29965103e+00 5.47129929e-01 1.11236560e+00
3.52861583e-01 1.84085578e-01 7.02601135e-01 -2.61983097e-01
-1.07847631e+00 -1.03090966e+00 5.86115301e-01 -5.02449036e-01
3.78548175e-01 -3.23186010e-01 -8.13932180e-01 1.95824414e-01
3.81793052e-01 2.44896859e-01 6.41208351e-01 -2.23995194e-01
-5.22245049e-01 -2.65330523e-01 -1.21712899e+00 3.29438806e-01
6.93681538e-01 -8.63714635e-01 -3.97323966e-01 1.67865500e-01
6.59642875e-01 -7.72269517e-02 -1.10249782e+00 3.02792042e-01
5.45170367e-01 -1.34155703e+00 1.14152956e+00 -2.09211111e-01
1.12661231e+00 -2.65360594e-01 -3.85278672e-01 -1.50812972e+00
-3.58289182e-01 -1.92387894e-01 -7.23031685e-02 1.10382187e+00
3.57807100e-01 -4.04520005e-01 4.75058347e-01 2.85641283e-01
1.60735294e-01 -8.06442857e-01 -6.92876101e-01 -6.26302302e-01
2.33902320e-01 -4.43775415e-01 5.87821901e-01 4.29928154e-01
-3.60935301e-01 1.98452443e-01 -4.06549245e-01 2.13432550e-01
7.53588438e-01 6.53122813e-02 8.28690708e-01 -1.02672720e+00
-5.75844944e-01 -3.71680945e-01 -5.14895916e-01 -9.06338573e-01
-8.71169418e-02 -7.13265359e-01 1.40686631e-01 -1.55992842e+00
3.88186216e-01 -3.67594928e-01 -5.88054240e-01 1.53582066e-01
-3.22536260e-01 9.18525279e-01 4.46077347e-01 1.29711524e-01
-7.14460254e-01 5.41721880e-01 1.25859094e+00 -2.26776108e-01
9.19940546e-02 -3.55414957e-01 -8.44122469e-01 5.32294452e-01
7.41471767e-01 -3.43347013e-01 -4.59017575e-01 -5.16581357e-01
3.73709649e-01 -6.42778575e-02 5.93147278e-01 -1.33006716e+00
-2.45468438e-01 -9.31290090e-02 6.82677269e-01 8.56272802e-02
3.49241436e-01 -7.04479992e-01 -9.32348818e-02 1.85055867e-01
-5.83373666e-01 5.23957461e-02 1.16825357e-01 4.16754693e-01
-4.14453238e-01 -1.35111153e-01 9.53922629e-01 -2.60043234e-01
-6.69966221e-01 3.45572203e-01 -1.76448300e-01 5.86330108e-02
6.52621627e-01 -1.78243518e-02 -1.84400618e-01 -8.69427383e-01
-4.91078526e-01 -1.76767394e-01 7.68994033e-01 4.75331634e-01
8.06414008e-01 -1.51586092e+00 -1.04104185e+00 1.97519913e-01
2.96504736e-01 -4.33578074e-01 2.25751758e-01 4.58457023e-01
-3.32355529e-01 9.19622257e-02 -5.79303086e-01 -5.48952520e-01
-8.97237539e-01 6.21079028e-01 3.64549696e-01 -3.07957441e-01
-4.10585612e-01 6.44937336e-01 1.62777930e-01 -1.76674575e-01
8.88203159e-02 -1.18264727e-01 -1.15224488e-01 -2.09026471e-01
6.68886960e-01 1.50517836e-01 1.29971102e-01 -1.00028098e+00
-8.93388540e-02 5.45277297e-01 2.56269038e-01 -3.89961094e-01
1.39116693e+00 -3.60208750e-01 4.03691232e-02 3.86279166e-01
1.51813364e+00 9.22347698e-03 -1.53752995e+00 -2.22605631e-01
-2.93029189e-01 -8.77662599e-01 3.27876002e-01 -1.13774729e+00
-1.18037152e+00 1.04601681e+00 1.06256270e+00 2.16652438e-01
1.63599527e+00 -1.64993092e-01 6.44202471e-01 -1.13211591e-02
2.23954469e-01 -9.39772666e-01 6.43833518e-01 -6.42151898e-03
1.12061381e+00 -1.54870272e+00 -5.01533039e-03 -2.43842632e-01
-5.19603014e-01 9.24962044e-01 3.66950125e-01 -3.77528131e-01
3.18626851e-01 1.38343111e-01 2.92910486e-01 -6.84796507e-03
-7.58721709e-01 -3.03998888e-01 5.85782170e-01 8.09292078e-01
5.42274773e-01 -3.19295526e-02 1.27390027e-01 2.75264621e-01
-1.63312897e-01 5.65811060e-02 5.89328408e-01 5.86530983e-01
-2.71143258e-01 -1.31496501e+00 -3.84011418e-01 4.74167466e-01
-5.44513762e-01 -1.28376871e-01 -1.48676783e-01 2.23751798e-01
2.66209155e-01 1.31336665e+00 1.04764998e-01 -5.08416891e-01
4.88317668e-01 -2.62927622e-01 4.52548236e-01 -4.20223027e-01
-4.75289524e-01 -1.92915052e-01 -4.38370034e-02 -9.48443174e-01
-4.07789230e-01 -4.10956472e-01 -6.98029637e-01 -2.28565350e-01
2.08319098e-01 -1.43475771e-01 6.28856182e-01 6.61396086e-01
3.13248158e-01 6.74777389e-01 8.40118289e-01 -1.06081665e+00
-3.59180301e-01 -7.99993575e-01 -3.06391746e-01 1.06303585e+00
5.16389012e-01 -5.96125901e-01 -2.86411703e-01 3.36686879e-01] | [11.797201156616211, -1.7734094858169556] |
226edcc0-abf0-4462-93a2-1008a52dc5c6 | an-unsupervised-attentive-adversarial | 2202.09635 | null | https://arxiv.org/abs/2202.09635v2 | https://arxiv.org/pdf/2202.09635v2.pdf | Deep Single Image Deraining using An Asymetric Cycle Generative and Adversarial Framework | In reality, rain and fog are often present at the same time, which can greatly reduce the clarity and quality of the scene image. However, most unsupervised single image deraining methods mainly focus on rain streak removal by disregarding the fog, which leads to low-quality deraining performance. In addition, the samples are rather homogeneous generated by these methods and lack diversity, resulting in poor results in the face of complex rain scenes. To address the above issues, we propose a novel Asymetric Cycle Generative and Adversarial framework (ACGF) for single image deraining that trains on both synthetic and real rainy images while simultaneously capturing both rain streaks and fog features. ACGF consists of a Rain-fog2Clean (R2C) transformation block and a Clean2Rain-fog (C2R) transformation block. The former consists of parallel rain removal path and rain-fog feature extraction path by the rain and derain-fog network and the attention rain-fog feature extraction network (ARFE) , while the latter only contains a synthetic rain transformation path. In rain-fog feature extraction path, to better characterize the rain-fog fusion feature, we employ an ARFE to exploit the self-similarity of global and local rain-fog information by learning the spatial feature correlations. Moreover, to improve the translational capacity of C2R and the diversity of models, we design a rain-fog feature decoupling and reorganization network (RFDR) by embedding a rainy image degradation model and a mixed discriminator to preserve richer texture details in synthetic rain conversion path. Extensive experiments on benchmark rain-fog and rain datasets show that ACGF outperforms state-of-the-art deraining methods. We also conduct defogging performance evaluation experiments to further demonstrate the effectiveness of ACGF. | ['Zixiang Xiong', 'Tao Lu', 'Cheng Chen', 'Rui Jiang', 'Wei Liu'] | 2022-02-19 | null | null | null | null | ['single-image-deraining'] | ['computer-vision'] | [ 6.34058937e-02 -3.87180537e-01 6.06691897e-01 -2.94883847e-01
-3.87903154e-01 -4.63216752e-01 3.85332793e-01 -7.01099098e-01
-2.79130284e-02 7.93216228e-01 6.03279173e-02 2.07688436e-02
2.37013429e-01 -1.02505934e+00 -7.03243732e-01 -1.41199434e+00
2.44347498e-01 -1.38226807e-01 6.09050989e-02 -5.51118612e-01
-3.37635994e-01 4.17649120e-01 -1.60055709e+00 7.27760270e-02
1.62389672e+00 6.87373519e-01 5.71813464e-01 5.83413363e-01
-3.03070224e-03 8.78736198e-01 -6.30689085e-01 -1.55731544e-01
5.13758063e-01 -8.52421105e-01 1.64378777e-01 8.39114040e-02
5.86466253e-01 -4.76009667e-01 -5.71725726e-01 1.16488993e+00
6.57212198e-01 9.04491469e-02 3.90210748e-01 -8.30953896e-01
-9.40896690e-01 -4.19831313e-02 -5.96107244e-01 4.40042138e-01
1.49406623e-02 5.05535662e-01 5.02631247e-01 -1.16706443e+00
3.27195019e-01 1.27476680e+00 4.86398965e-01 2.66249865e-01
-8.73831868e-01 -7.58118927e-01 2.42536351e-01 6.64921030e-02
-1.24478424e+00 -2.46014923e-01 7.83190906e-01 -1.17273018e-01
2.61877507e-01 5.04804313e-01 9.31119382e-01 7.38323927e-01
4.59523052e-01 6.34809554e-01 1.58053660e+00 -1.26189262e-01
-4.64680344e-02 4.44220863e-02 8.53912309e-02 7.01994658e-01
5.93611360e-01 4.57473516e-01 -1.36341676e-01 3.46328497e-01
7.87659705e-01 4.31237608e-01 -9.34928834e-01 1.95807278e-01
-6.03207767e-01 7.58271635e-01 8.80613148e-01 1.29444405e-01
-3.59948128e-01 -2.59330720e-01 -3.72342378e-01 5.86593270e-01
6.77406788e-01 2.50061542e-01 9.21259522e-02 4.90167439e-01
-1.05524981e+00 6.96258187e-01 7.26185739e-01 7.45706439e-01
1.03911459e+00 6.34356320e-01 -4.25575346e-01 6.46994710e-01
2.64899880e-01 1.50033069e+00 1.52037323e-01 -5.04201293e-01
4.14009482e-01 1.81619555e-01 3.50358844e-01 -1.03503489e+00
-1.87852561e-01 -7.02826023e-01 -1.17123938e+00 5.38959444e-01
5.99977560e-02 -1.38630241e-01 -1.36429691e+00 1.36808419e+00
2.94430166e-01 4.55155104e-01 2.95221567e-01 1.51172745e+00
1.05253053e+00 9.42300260e-01 -1.12449937e-01 -4.25322384e-01
1.21578705e+00 -9.91190076e-01 -9.58917141e-01 -4.73312974e-01
3.38611752e-02 -8.52702737e-01 1.07825279e+00 1.64434448e-01
-9.56770658e-01 -7.59060860e-01 -1.17746592e+00 5.99627458e-02
-2.96808302e-01 -1.64419085e-01 4.73441482e-01 5.41435599e-01
-6.90910161e-01 3.38125467e-01 -5.27808666e-01 3.07746255e-03
3.23423058e-01 -1.75631762e-01 -3.70789096e-02 -9.26879466e-01
-1.43547595e+00 8.51719797e-01 1.93611816e-01 9.82901692e-01
-1.15758824e+00 -7.21457183e-01 -6.73401296e-01 -4.77938168e-02
-6.84487671e-02 -7.49232769e-01 2.45982781e-01 -1.20061278e+00
-1.18367863e+00 3.88281345e-01 -7.31285438e-02 -4.91002053e-02
4.18410569e-01 -5.22421777e-01 -8.63982320e-01 1.56987116e-01
-8.48954543e-02 1.07783712e-01 1.44512451e+00 -1.67593181e+00
-5.50350428e-01 -2.69438535e-01 1.52539372e-01 4.95486319e-01
3.45927238e-01 -3.03405106e-01 -2.32812777e-01 -9.30634558e-01
4.60266806e-02 -7.89395034e-01 -1.03793278e-01 -3.15899909e-01
-1.81886777e-01 7.75058389e-01 1.17325866e+00 -1.07225251e+00
1.12108707e+00 -2.22308183e+00 5.87490313e-02 6.83653653e-02
3.50586683e-01 6.09479904e-01 -4.35512096e-01 2.32327692e-02
-2.18559489e-01 -2.64439940e-01 -5.48417628e-01 -2.38039285e-01
-3.60687286e-01 7.13437438e-01 -6.87024653e-01 6.59997046e-01
5.44881582e-01 9.78883624e-01 -8.85597467e-01 -3.49726737e-01
2.45569885e-01 6.22235239e-01 -3.54334950e-01 7.08458006e-01
-1.56281024e-01 6.94628119e-01 -4.95804012e-01 9.70327675e-01
1.49707103e+00 2.10933104e-01 -2.64550716e-01 -2.38736480e-01
-5.40550873e-02 -3.72458398e-01 -1.04674029e+00 9.82822061e-01
-5.78297794e-01 2.50317991e-01 3.42482835e-01 -6.22061014e-01
1.00924098e+00 5.60466945e-02 -1.17130607e-01 -8.60876381e-01
4.14771549e-02 3.30303282e-01 -1.19156808e-01 -7.24338889e-01
3.23741883e-01 -5.16983867e-01 4.44224507e-01 8.15629512e-02
-7.27474466e-02 -3.98507684e-01 -1.23915359e-01 6.24550395e-02
7.68157303e-01 6.58315644e-02 -2.88513005e-01 -1.33217514e-01
4.37323689e-01 -3.34278315e-01 7.33857214e-01 7.36604631e-01
-9.65097845e-02 1.19119823e+00 -1.40128424e-02 -4.10620749e-01
-7.37352729e-01 -1.23000395e+00 2.91880090e-02 6.15731061e-01
6.26045942e-01 2.54588485e-01 -5.77320993e-01 -5.36033452e-01
-4.44745123e-02 6.56647325e-01 -7.48170793e-01 -2.74378061e-01
-7.53501892e-01 -1.40697873e+00 2.97624052e-01 5.32184951e-02
1.08143044e+00 -1.20104408e+00 -1.03115141e-01 1.05190016e-01
-4.40068036e-01 -1.13319731e+00 -4.78921771e-01 -5.98939694e-03
-6.58647358e-01 -1.03249228e+00 -6.87070251e-01 -6.07508659e-01
6.26217484e-01 8.52512062e-01 1.19695723e+00 2.55902410e-01
-2.38225788e-01 -1.22045800e-01 -8.38152349e-01 -5.06199539e-01
-1.03245243e-01 -4.30640459e-01 -3.06106001e-01 3.26099306e-01
-9.42294747e-02 -1.00964868e+00 -9.86854196e-01 1.84937194e-01
-1.32629180e+00 -1.04143098e-01 7.71147847e-01 1.01778281e+00
5.19753993e-01 2.26827040e-02 2.63186961e-01 -9.22715902e-01
3.67034703e-01 -6.40482068e-01 -4.86440331e-01 1.18823417e-01
-6.17065907e-01 -3.20439219e-01 6.94120824e-01 -2.52457678e-01
-1.37980890e+00 -2.59409755e-01 -1.43279126e-02 -7.38030136e-01
1.13782339e-01 4.35845554e-01 -6.03695273e-01 -3.30778688e-01
5.17101228e-01 6.74253166e-01 -1.31751940e-01 -2.96515971e-01
4.74103034e-01 2.93516964e-01 6.72589719e-01 -2.21547298e-02
1.59615386e+00 6.51862741e-01 -9.30464417e-02 -8.40198457e-01
-1.05553508e+00 -1.76779777e-01 -9.96097550e-02 -2.51677543e-01
8.32106709e-01 -1.35459352e+00 -1.44435197e-01 8.57937157e-01
-9.23032820e-01 -2.70945758e-01 -3.41914147e-01 3.65666807e-01
-5.01127653e-02 3.89841646e-01 -4.88872409e-01 -8.34487081e-01
-6.20528579e-01 -8.69273782e-01 8.50158095e-01 6.71169341e-01
6.90982044e-01 -6.64881527e-01 8.64325538e-02 3.47715795e-01
6.77268982e-01 4.98429298e-01 5.82918942e-01 1.12409331e-01
-9.00952280e-01 4.05408032e-02 -4.09813523e-01 6.62235498e-01
3.24267119e-01 -2.35825405e-01 -1.10390913e+00 -5.59612036e-01
4.54867721e-01 3.53002436e-02 1.25018322e+00 3.36112052e-01
5.16993582e-01 -3.95629376e-01 1.56298816e-01 1.17137682e+00
1.75125051e+00 3.17260600e-03 1.25230038e+00 2.68383771e-01
1.08579946e+00 2.77902931e-01 8.03456008e-01 1.67911112e-01
1.51056767e-01 2.70094663e-01 7.39491403e-01 -7.12746441e-01
-5.01782537e-01 8.32111016e-02 5.74357271e-01 7.67205179e-01
-4.87648606e-01 -6.03005528e-01 -2.04501852e-01 5.32012165e-01
-1.67838240e+00 -1.08073521e+00 -1.00631736e-01 1.91972017e+00
6.42876446e-01 -1.48343205e-01 -4.57252592e-01 -1.49913386e-01
4.96142596e-01 6.96021020e-01 -4.96125191e-01 6.89485818e-02
-8.21655929e-01 5.91257215e-01 6.35825694e-01 6.20672345e-01
-9.58306134e-01 9.50350404e-01 4.81896257e+00 7.47750342e-01
-1.26372874e+00 2.64614642e-01 3.47888350e-01 -5.25248498e-02
-4.68232155e-01 2.70964447e-02 -6.07672036e-01 7.83268929e-01
3.71824831e-01 2.76070625e-01 7.05392778e-01 3.42258781e-01
5.52896261e-01 -6.05962835e-02 -1.75341830e-01 8.79165590e-01
2.19178945e-01 -7.77315259e-01 3.14002752e-01 -2.59572297e-01
8.42580855e-01 8.61464292e-02 6.71447217e-02 3.54465187e-01
3.74363989e-01 -1.10701334e+00 4.97247458e-01 1.06040323e+00
7.96376765e-01 -6.38078690e-01 1.00497723e+00 -4.80186231e-02
-1.14584470e+00 -4.35724258e-02 -5.08018732e-01 1.64434195e-01
3.47809009e-02 1.20996487e+00 -1.31211102e-01 1.24578702e+00
8.80440831e-01 7.16397822e-01 -5.05833447e-01 9.27651525e-01
-7.11808443e-01 7.37530708e-01 -2.45031983e-01 6.78915560e-01
1.72427356e-01 -8.22076321e-01 9.29729402e-01 1.30785120e+00
2.82837778e-01 5.48188448e-01 1.15308367e-01 9.42836642e-01
7.23579600e-02 -2.68410087e-01 -4.73643541e-01 2.36687690e-01
2.19150737e-01 1.34438884e+00 -3.50005955e-01 -2.40809262e-01
-2.64440507e-01 1.12723207e+00 -1.87803954e-01 8.02124441e-01
-9.84058499e-01 -5.82720935e-01 9.42560196e-01 1.24495231e-01
6.27761662e-01 -1.11780606e-01 9.24336016e-02 -1.48950744e+00
2.08613217e-01 -9.97829735e-01 -1.48727074e-01 -9.97895300e-01
-1.39728355e+00 7.98159242e-01 -3.78020823e-01 -1.45741904e+00
2.91452527e-01 -2.06110418e-01 -1.11098385e+00 1.33924758e+00
-2.34286070e+00 -1.45431817e+00 -1.18073368e+00 8.86777520e-01
3.69448692e-01 1.16246760e-01 4.18988436e-01 4.27551985e-01
-6.39656186e-01 4.30246592e-01 4.23286855e-01 -8.83239508e-02
8.99283469e-01 -1.10916209e+00 1.23508431e-01 1.46325445e+00
-3.12323898e-01 1.58816308e-01 9.47368979e-01 -7.17812061e-01
-1.42942524e+00 -1.90424275e+00 2.84621477e-01 -3.17407399e-02
1.81557074e-01 -2.71291912e-01 -1.36172390e+00 4.28664327e-01
4.49355654e-02 4.83411312e-01 -4.17122617e-02 -4.83367741e-01
-3.39954346e-01 -5.05535483e-01 -1.09512937e+00 5.65792501e-01
7.85853148e-01 -2.66936958e-01 -4.79660541e-01 4.31111187e-01
8.52903605e-01 -4.64358360e-01 -5.24192214e-01 5.60223401e-01
2.88829625e-01 -1.24678969e+00 8.63859832e-01 -4.06755917e-02
6.61166370e-01 -7.44072855e-01 -2.80596972e-01 -1.58288074e+00
-4.76472318e-01 -4.71393913e-01 -2.70173252e-01 1.28249538e+00
-5.63519746e-02 -9.47459280e-01 3.01085323e-01 -1.50854290e-01
-3.47545385e-01 -5.73826790e-01 -5.86874306e-01 -7.04459310e-01
-6.17053546e-02 2.42881924e-01 6.81151450e-01 1.03884709e+00
-9.46071923e-01 6.01776876e-02 -6.66178882e-01 7.55709291e-01
7.51129508e-01 6.74509883e-01 7.60053039e-01 -8.75089526e-01
-3.21837366e-01 5.01171052e-02 -8.55440870e-02 -7.62268424e-01
-1.85740560e-01 -5.38869679e-01 4.78597075e-01 -1.36914098e+00
4.61628214e-02 -3.54405552e-01 -3.55383992e-01 2.10006803e-01
-8.98534417e-01 4.00465995e-01 2.77298301e-01 6.24443054e-01
-1.66341275e-01 1.06200254e+00 1.71677589e+00 -1.68703377e-01
-2.93359071e-01 -4.76178378e-02 -5.99381089e-01 5.76245844e-01
6.63664997e-01 -4.94402528e-01 -2.81531066e-01 -5.77125609e-01
-1.76656529e-01 -6.26716912e-02 5.20518601e-01 -1.12637091e+00
-2.00173929e-01 -2.64073372e-01 6.16913617e-01 -3.66576731e-01
3.02935809e-01 -5.91219366e-01 1.62590250e-01 2.51993477e-01
6.15591645e-01 -2.07295552e-01 4.69733551e-02 8.03490162e-01
-6.37854695e-01 2.95542121e-01 1.09331143e+00 -2.52770096e-01
-4.83539909e-01 5.15510798e-01 -1.10238522e-01 -3.61237377e-02
6.60502195e-01 -2.31036171e-01 -6.68828547e-01 -4.47631449e-01
-4.67723548e-01 3.10975701e-01 4.42055047e-01 3.15565735e-01
7.84456372e-01 -8.97022188e-01 -1.17046952e+00 4.87930000e-01
-1.22639313e-01 1.79651514e-01 7.61452138e-01 8.96960258e-01
-6.62497282e-01 -2.73056120e-01 -1.97536528e-01 -1.58438399e-01
-1.00549281e+00 4.91740644e-01 5.70210099e-01 -4.97742414e-01
-8.33955228e-01 6.27455056e-01 8.87741923e-01 -2.47850999e-01
-4.17179406e-01 -8.86888430e-02 -2.66897738e-01 -9.47319046e-02
6.23349190e-01 1.42341673e-01 1.90803796e-01 -4.42677081e-01
8.41457695e-02 5.52643418e-01 -3.01583633e-02 3.51908177e-01
1.31887662e+00 -3.87240052e-01 -2.58790433e-01 1.56429127e-01
6.80720747e-01 1.41084373e-01 -1.22897923e+00 -1.92631841e-01
-1.00060248e+00 -7.44663477e-01 2.23151788e-01 -7.37579703e-01
-1.74212742e+00 6.45039260e-01 1.01638603e+00 4.56406502e-03
1.68304205e+00 -5.13540924e-01 8.94661367e-01 1.16285190e-01
-2.77171917e-02 -4.55848098e-01 -1.87321529e-01 6.71662867e-01
1.02177477e+00 -1.11350012e+00 1.80273786e-01 -5.84083080e-01
-5.11015058e-01 8.47287536e-01 5.27311981e-01 -6.57215238e-01
7.45213270e-01 1.77164227e-01 2.95478731e-01 -2.70243675e-01
-2.35088602e-01 -4.31433529e-01 -2.96776742e-02 5.86703241e-01
-2.16903731e-01 7.22299190e-03 -2.27812648e-01 6.09990895e-01
-5.81427366e-02 -7.94154257e-02 6.23693883e-01 9.50172901e-01
-5.53153932e-01 -6.18004262e-01 -7.22326577e-01 3.59915942e-01
-1.50129035e-01 -6.21410131e-01 1.25182748e-01 8.31032634e-01
5.37020862e-01 1.10225248e+00 -1.11219347e-01 -4.29106921e-01
4.96539623e-01 -3.27265888e-01 4.16820228e-01 -2.54402131e-01
-5.86511612e-01 2.02966526e-01 -3.44135582e-01 -4.90251929e-01
-5.14851689e-01 -2.71464765e-01 -6.42471552e-01 -3.04721653e-01
-4.90386635e-01 1.87453300e-01 1.41044974e-01 9.34936166e-01
2.38925949e-01 7.20061898e-01 1.14077795e+00 -1.00916421e+00
-3.58164459e-02 -1.17284858e+00 -1.34188211e+00 2.88543373e-01
6.39759541e-01 -6.54953361e-01 -8.02206576e-01 1.33558616e-01] | [10.936129570007324, -3.1900570392608643] |
8e4de4d2-8934-4cd5-b503-90637b8a6b61 | beyond-single-receptive-field-a-receptive | 2207.10278 | null | https://arxiv.org/abs/2207.10278v1 | https://arxiv.org/pdf/2207.10278v1.pdf | Beyond single receptive field: A receptive field fusion-and-stratification network for airborne laser scanning point cloud classification | The classification of airborne laser scanning (ALS) point clouds is a critical task of remote sensing and photogrammetry fields. Although recent deep learning-based methods have achieved satisfactory performance, they have ignored the unicity of the receptive field, which makes the ALS point cloud classification remain challenging for the distinguishment of the areas with complex structures and extreme scale variations. In this article, for the objective of configuring multi-receptive field features, we propose a novel receptive field fusion-and-stratification network (RFFS-Net). With a novel dilated graph convolution (DGConv) and its extension annular dilated convolution (ADConv) as basic building blocks, the receptive field fusion process is implemented with the dilated and annular graph fusion (DAGFusion) module, which obtains multi-receptive field feature representation through capturing dilated and annular graphs with various receptive regions. The stratification of the receptive fields with point sets of different resolutions as the calculation bases is performed with Multi-level Decoders nested in RFFS-Net and driven by the multi-level receptive field aggregation loss (MRFALoss) to drive the network to learn in the direction of the supervision labels with different resolutions. With receptive field fusion-and-stratification, RFFS-Net is more adaptable to the classification of regions with complex structures and extreme scale variations in large-scale ALS point clouds. Evaluated on the ISPRS Vaihingen 3D dataset, our RFFS-Net significantly outperforms the baseline approach by 5.3% on mF1 and 5.4% on mIoU, accomplishing an overall accuracy of 82.1%, an mF1 of 71.6%, and an mIoU of 58.2%. Furthermore, experiments on the LASDU dataset and the 2019 IEEE-GRSS Data Fusion Contest dataset show that RFFS-Net achieves a new state-of-the-art classification performance. | ['Martin Weinmann', 'Kun fu', 'Xiaonan Lu', 'Xian Sun', 'Wenhui Diao', 'Kaiqiang Chen', 'Yongqiang Mao'] | 2022-07-21 | null | null | null | null | ['point-cloud-classification'] | ['computer-vision'] | [ 7.79714137e-02 -1.99893773e-01 2.80113101e-01 -5.90230882e-01
-6.21192813e-01 -4.44617301e-01 4.08043265e-01 -7.11804256e-02
-2.91401982e-01 4.16398168e-01 -4.11214769e-01 -2.88359404e-01
-2.67291278e-01 -1.23898077e+00 -9.48989153e-01 -7.73926258e-01
-1.88363865e-01 3.73097688e-01 1.70154467e-01 -3.70536715e-01
-6.25272319e-02 1.21856868e+00 -1.92628729e+00 6.09239399e-01
9.88490880e-01 1.44836080e+00 5.59780478e-01 4.67733562e-01
-2.77704865e-01 5.18221498e-01 -3.29316556e-01 8.37067962e-02
5.64180017e-01 3.31938684e-01 -6.83831811e-01 -1.10609055e-01
1.08587730e+00 -1.97287679e-01 -9.84669402e-02 1.10164225e+00
6.35180235e-01 8.56164470e-02 5.46448827e-01 -9.45577502e-01
-5.99821627e-01 3.87619883e-01 -6.48572981e-01 2.52795160e-01
-3.05867672e-01 2.86209937e-02 8.65379632e-01 -1.07892156e+00
1.24093860e-01 1.48081470e+00 7.84840345e-01 2.15638846e-01
-1.19992614e+00 -9.64835584e-01 2.49201268e-01 -3.80575582e-02
-1.86026299e+00 5.83480224e-02 4.35761184e-01 -6.76786065e-01
1.09992218e+00 3.18600476e-01 6.23680472e-01 2.35788837e-01
3.02788824e-01 2.73605198e-01 1.23526251e+00 -4.36773412e-02
1.75354064e-01 -4.01168972e-01 -6.37734607e-02 5.11795938e-01
1.59453511e-01 1.55290499e-01 2.61204150e-02 7.62147307e-02
1.01503682e+00 3.68565857e-01 -3.46936643e-01 1.88323230e-01
-1.04574311e+00 8.32745194e-01 1.34211814e+00 2.77745157e-01
-6.42996907e-01 1.64910704e-01 -1.57890499e-01 2.00260520e-01
5.87626040e-01 1.35852993e-01 -3.87299448e-01 8.28551054e-01
-1.04431224e+00 2.27006361e-01 3.77974600e-01 8.64950478e-01
1.35341036e+00 1.03396401e-01 -1.56521976e-01 8.67313862e-01
4.75741684e-01 1.04013824e+00 -1.29047170e-01 -6.73619986e-01
5.31594455e-01 1.03660560e+00 -3.37375104e-01 -8.02339911e-01
-5.58080077e-01 -6.04278326e-01 -1.19605064e+00 6.68068647e-01
-1.27186194e-01 1.21688731e-02 -1.34143102e+00 1.31483388e+00
3.60919088e-01 1.78332582e-01 1.32543594e-01 9.86952543e-01
1.14932859e+00 8.59688878e-01 1.30138099e-01 2.20306277e-01
1.24381316e+00 -3.75325233e-01 -1.34484321e-01 -3.88917685e-01
4.18658674e-01 -5.19624770e-01 7.95259833e-01 -3.28356028e-02
-7.30681956e-01 -8.99720371e-01 -1.05595565e+00 1.10567793e-01
-4.85615164e-01 3.24044347e-01 5.67193389e-01 1.48876876e-01
-1.33270431e+00 6.18461549e-01 -6.27832472e-01 -3.45646352e-01
8.05875182e-01 6.70352697e-01 -4.25159484e-01 -2.90259090e-03
-1.00993586e+00 5.81515193e-01 4.63229120e-01 5.06408513e-01
-5.91484547e-01 -8.80745590e-01 -6.88078880e-01 3.07089984e-01
-9.99947544e-03 -6.50303006e-01 7.14702964e-01 -8.39175224e-01
-9.89835560e-01 6.95257604e-01 3.41991872e-01 -3.55392367e-01
1.03300132e-01 -2.89708581e-02 -5.07939756e-01 1.14108518e-01
2.33821511e-01 1.12541342e+00 8.78816605e-01 -1.26300418e+00
-9.18127358e-01 -5.81514955e-01 2.49341968e-02 4.21098232e-01
5.37859350e-02 -2.34922484e-01 5.30896895e-02 -4.86969262e-01
4.19112772e-01 -7.89346576e-01 -1.23407796e-01 -1.12721607e-01
-2.95583140e-02 -3.26946050e-01 8.85948122e-01 -3.39054197e-01
1.01952982e+00 -2.34370613e+00 -2.15179369e-01 4.01560634e-01
3.02969873e-01 4.22620684e-01 -2.32477650e-01 1.04696237e-01
-2.62638807e-01 1.23337828e-01 -4.97235239e-01 5.14957458e-02
-2.65940577e-01 3.48524332e-01 -4.24930990e-01 4.69134063e-01
5.31778812e-01 7.65533626e-01 -6.22664869e-01 -3.28886390e-01
6.53850198e-01 6.67511046e-01 -3.92977089e-01 2.04534277e-01
-1.01288103e-01 3.24300230e-01 -5.98783731e-01 9.30333495e-01
1.29512954e+00 -2.82552421e-01 -1.51327729e-01 -4.09088761e-01
-4.48709220e-01 -1.53408097e-02 -1.35139406e+00 1.32850337e+00
-4.44297552e-01 2.91526943e-01 3.52017999e-01 -6.95164621e-01
1.30332661e+00 -3.47818211e-02 4.17600214e-01 -5.50532699e-01
1.20991036e-01 3.56956333e-01 -4.01804075e-02 -9.69930440e-02
2.01022401e-01 -2.49724865e-01 1.28506243e-01 -2.18297213e-01
1.82207197e-01 -2.79699028e-01 -2.42847517e-01 -2.00471163e-01
9.82051671e-01 -1.11899011e-01 -8.70870203e-02 -5.50226569e-01
7.27628708e-01 -8.48411247e-02 3.90850455e-01 7.14578271e-01
1.10230900e-01 8.34029198e-01 -2.21027479e-01 -7.63519287e-01
-6.94144189e-01 -1.10023904e+00 -6.30826235e-01 8.84277642e-01
1.21716186e-01 -4.50168326e-02 -5.61904132e-01 -6.65162563e-01
3.34035009e-01 3.74408275e-01 -3.93666804e-01 1.47190049e-01
-5.34048736e-01 -6.57273829e-01 5.41113198e-01 6.31391048e-01
1.04954171e+00 -1.06538713e+00 -4.49248791e-01 1.22311838e-01
-5.71351033e-03 -1.25039303e+00 9.94721651e-02 3.01813185e-01
-1.03667986e+00 -8.77400279e-01 -3.36783230e-01 -8.23068202e-01
6.29297614e-01 3.87357414e-01 1.05189109e+00 5.14519513e-02
-2.77565680e-02 -8.69527906e-02 -2.40957454e-01 -4.34078544e-01
-1.72551021e-01 2.65191317e-01 -3.23249549e-01 2.23104313e-01
4.45906639e-01 -6.31704748e-01 -5.47994792e-01 3.71056676e-01
-1.00812173e+00 -2.57363766e-02 7.80737340e-01 7.22107410e-01
8.94253731e-01 -5.51776886e-02 3.53733569e-01 -4.73120004e-01
3.13220285e-02 -3.64711702e-01 -9.52076674e-01 1.82370484e-01
-5.89636862e-01 -1.35755673e-01 6.68149948e-01 1.40221119e-01
-8.24175596e-01 2.87386507e-01 -2.84194559e-01 -7.30467319e-01
-4.75280613e-01 3.87102693e-01 -2.00737059e-01 -5.55804729e-01
8.66640925e-01 -4.15212624e-02 -1.23582013e-01 -3.75532746e-01
6.71055689e-02 9.39091623e-01 3.48028839e-01 -2.57977635e-01
8.74887407e-01 6.65614009e-01 2.35448062e-01 -8.61986876e-01
-7.05851316e-01 -5.12022793e-01 -6.64124429e-01 -3.30567956e-01
1.04631400e+00 -1.38926435e+00 -6.01720393e-01 5.51028907e-01
-1.12652528e+00 -1.75153881e-01 -4.54729080e-01 5.07629395e-01
-1.75674662e-01 5.63074276e-02 -2.12917551e-01 -6.25581443e-01
-6.17267728e-01 -1.18231094e+00 1.53520417e+00 2.64690846e-01
4.47072417e-01 -5.63096344e-01 -2.45509595e-01 1.62626252e-01
4.74914223e-01 3.03727359e-01 7.67653465e-01 -4.00195926e-01
-8.95832300e-01 -8.07014148e-05 -5.57779551e-01 8.05100858e-01
5.05487584e-02 -3.89273614e-02 -1.34824586e+00 -2.76315719e-01
-2.12976068e-01 -1.10588476e-01 1.11494696e+00 5.32050550e-01
1.31281865e+00 1.85489804e-02 -2.25123167e-01 1.17933631e+00
1.70014894e+00 7.23994747e-02 8.00282419e-01 1.44355312e-01
1.00533772e+00 3.63452703e-01 4.65979278e-01 2.41757140e-01
4.06099439e-01 2.26453006e-01 1.06286538e+00 -3.27817827e-01
-1.65881395e-01 1.17041610e-01 1.41482010e-01 3.85727406e-01
-5.51496506e-01 -2.08791927e-01 -1.06653035e+00 4.55701292e-01
-1.73525763e+00 -9.04586554e-01 -1.71365812e-01 2.08357430e+00
2.55843550e-01 -2.50745684e-01 -4.44678307e-01 -7.61271194e-02
1.04678881e+00 4.79753613e-01 -3.82475048e-01 -2.16844022e-01
-4.28899646e-01 6.18908286e-01 1.03018296e+00 4.66115594e-01
-1.45875633e+00 1.06370842e+00 5.01484632e+00 9.42032099e-01
-1.56051385e+00 -7.01870471e-02 3.06505501e-01 2.78159142e-01
-1.19431458e-01 -1.57657936e-01 -1.06468689e+00 2.96996802e-01
6.01801157e-01 7.21340954e-01 5.17339766e-01 7.61408806e-01
-2.02101424e-01 1.88714668e-01 -6.82610929e-01 1.12609994e+00
-2.74213672e-01 -1.40969956e+00 2.23952204e-01 1.29327178e-01
7.13759959e-01 9.59099710e-01 -1.11679256e-01 2.82472342e-01
1.89868674e-01 -1.16171074e+00 7.00331688e-01 3.88940215e-01
1.16802776e+00 -5.69831908e-01 8.92503262e-01 2.68081725e-01
-1.63343680e+00 -3.24080974e-01 -7.17467308e-01 1.29483407e-02
-2.15700597e-01 6.99561059e-01 -8.17871392e-01 8.68313074e-01
1.10753119e+00 8.11747134e-01 -4.45437372e-01 7.21165836e-01
-1.34728476e-01 5.04984856e-01 -7.37412274e-01 2.83377111e-01
3.64771187e-01 -2.96339273e-01 4.80151713e-01 1.22742105e+00
5.66911161e-01 1.82187170e-01 2.66544431e-01 9.76331592e-01
-1.67815968e-01 -7.82397538e-02 -6.03580892e-01 1.64644584e-01
5.33501089e-01 1.43893325e+00 -5.44652164e-01 -1.50217339e-01
-2.47580290e-01 3.57177824e-01 3.15712452e-01 3.30632925e-01
-7.75668323e-01 -2.37321183e-01 8.50847006e-01 3.89101237e-01
6.43764496e-01 -2.48738136e-02 -3.73203456e-01 -9.63454664e-01
-2.12903041e-02 -5.24354696e-01 4.00325984e-01 -8.70671272e-01
-1.33421445e+00 1.05004621e+00 6.62651360e-02 -1.48928928e+00
2.67110169e-01 -8.06719303e-01 -4.92329031e-01 1.26208127e+00
-1.92233968e+00 -1.63255680e+00 -8.79771292e-01 6.70428634e-01
1.77389830e-01 -2.14237109e-01 7.12969542e-01 4.73351121e-01
-1.40244067e-01 1.43405050e-01 -6.16584606e-02 2.12192580e-01
3.15527260e-01 -1.18550611e+00 4.41890031e-01 8.15719843e-01
-1.64772332e-01 1.50601603e-02 -7.24249855e-02 -7.04704821e-01
-1.15125072e+00 -1.82554150e+00 7.64540732e-01 -3.01585626e-02
3.08682561e-01 -2.79314369e-01 -1.12006557e+00 5.15254080e-01
-1.28673598e-01 5.77713013e-01 2.84525245e-01 -2.99410105e-01
-2.35187814e-01 -6.60510182e-01 -1.21624005e+00 2.95866765e-02
1.10978127e+00 -5.25456905e-01 -4.44578707e-01 3.53721231e-01
4.48020428e-01 -4.75002587e-01 -9.99289930e-01 1.20986152e+00
2.77880490e-01 -8.76582384e-01 1.09815550e+00 -5.61324880e-02
2.15916365e-01 -8.28582108e-01 -6.70224428e-01 -1.13766909e+00
-8.86673272e-01 2.85803210e-02 3.92536372e-01 9.62464511e-01
1.82570353e-01 -8.73639464e-01 3.91348034e-01 -1.31694391e-01
-5.91171682e-01 -6.48072183e-01 -1.06787825e+00 -6.16765738e-01
1.67319756e-02 -4.97116596e-01 1.05439293e+00 9.04182494e-01
-9.80553508e-01 4.37063128e-01 3.00106704e-01 8.30849349e-01
5.37711024e-01 3.55392069e-01 5.75187624e-01 -1.71365309e+00
1.72609791e-01 -4.15845692e-01 -7.05321372e-01 -8.62163424e-01
6.60780668e-02 -1.28063560e+00 2.19944362e-02 -1.74733734e+00
-3.97196412e-01 -7.23693192e-01 -4.08689976e-01 7.44971335e-01
1.22014351e-01 3.15142483e-01 2.98904270e-01 4.07428056e-01
-1.44017130e-01 5.70933104e-01 1.27575779e+00 -4.23706204e-01
-9.99026746e-02 -7.05594139e-04 -3.89802039e-01 6.82270765e-01
6.77099109e-01 -3.30608964e-01 -5.56474440e-02 -7.05024600e-01
2.00072363e-01 -3.95790696e-01 7.89102614e-01 -1.46916735e+00
1.34951323e-01 -4.81492328e-03 6.22504830e-01 -1.09109437e+00
2.24333555e-01 -1.03399205e+00 3.62623364e-01 2.33869329e-01
2.09366038e-01 -3.15695405e-02 3.70250434e-01 2.89738715e-01
-3.70956272e-01 1.58362210e-01 9.25073445e-01 -1.92406811e-02
-9.00926352e-01 6.34238482e-01 2.04952480e-03 -2.95991123e-01
8.49788785e-01 -4.25038189e-01 -4.37557518e-01 1.97444662e-01
-6.35443866e-01 3.25409949e-01 3.33901286e-01 2.42833048e-01
7.95953214e-01 -1.26444829e+00 -1.00038159e+00 8.16046834e-01
2.26882339e-01 9.11861956e-01 4.97592062e-01 6.20534003e-01
-8.27578723e-01 1.64798498e-02 -3.01288903e-01 -1.17774415e+00
-9.61795688e-01 1.00814536e-01 7.15901375e-01 -5.54886349e-02
-5.09067953e-01 9.40549612e-01 5.61654866e-01 -7.58445382e-01
-1.52810439e-01 -7.74323881e-01 -4.50170696e-01 1.48428051e-04
2.35660493e-01 2.85644680e-01 5.77894747e-01 -8.82812083e-01
-5.72924316e-01 8.50805461e-01 2.45111957e-01 4.67969060e-01
1.38201046e+00 2.13676706e-01 -3.60985160e-01 4.61142920e-02
1.00652826e+00 -3.83915842e-01 -1.28251004e+00 -2.33434364e-01
-4.77210134e-01 -3.22423726e-01 4.97522175e-01 -7.99616933e-01
-1.30469406e+00 1.05605710e+00 1.10122406e+00 1.72069669e-01
1.38527262e+00 1.56310629e-02 3.90961736e-01 3.53092521e-01
3.26257110e-01 -8.18263173e-01 -7.25325167e-01 9.23394740e-01
1.04785633e+00 -1.24539983e+00 -1.38986766e-01 -5.12987852e-01
-4.27881092e-01 1.23375607e+00 6.79078341e-01 -4.31857914e-01
9.44919825e-01 3.19653302e-01 7.43346587e-02 -5.05638897e-01
-2.94285208e-01 -4.91023809e-01 5.82789063e-01 4.11557496e-01
1.00161947e-01 3.42206150e-01 1.52906165e-01 3.00484329e-01
-9.55791771e-02 5.95737882e-02 -1.10572115e-01 8.57531309e-01
-7.39695072e-01 -5.77352941e-01 -5.06161332e-01 6.10020041e-01
-2.11788177e-01 -3.01487476e-01 -4.86840308e-02 6.15712106e-01
6.51119590e-01 7.61086166e-01 5.90924978e-01 -7.66974092e-01
5.78564703e-01 -2.89723694e-01 2.47001573e-01 -7.77980804e-01
-1.01236701e+00 2.09281877e-01 -2.61722654e-01 -5.05905211e-01
-6.48102999e-01 -3.65589768e-01 -1.53535783e+00 -2.72879571e-01
-4.37971503e-01 5.93706705e-02 7.35412300e-01 8.58029366e-01
6.58068836e-01 6.57171667e-01 8.45657408e-01 -1.08659708e+00
-5.77916741e-01 -1.01694155e+00 -7.71852612e-01 1.96744315e-02
5.46227157e-01 -6.03469431e-01 -3.16977531e-01 -2.11894706e-01] | [7.996367931365967, -2.978407621383667] |
afb414d9-cf1c-4807-92cd-3148003b5327 | semantic-based-pre-training-for-dialogue | 2209.09146 | null | https://arxiv.org/abs/2209.09146v1 | https://arxiv.org/pdf/2209.09146v1.pdf | Semantic-based Pre-training for Dialogue Understanding | Pre-trained language models have made great progress on dialogue tasks. However, these models are typically trained on surface dialogue text, thus are proven to be weak in understanding the main semantic meaning of a dialogue context. We investigate Abstract Meaning Representation (AMR) as explicit semantic knowledge for pre-training models to capture the core semantic information in dialogues during pre-training. In particular, we propose a semantic-based pre-training framework that extends the standard pre-training framework (Devlin et al., 2019) by three tasks for learning 1) core semantic units, 2) semantic relations and 3) the overall semantic representation according to AMR graphs. Experiments on the understanding of both chit-chats and task-oriented dialogues show the superiority of our model. To our knowledge, we are the first to leverage a deep semantic representation for dialogue pre-training. | ['Yue Zhang', 'Linfeng Song', 'Xuefeng Bai'] | 2022-09-19 | null | https://aclanthology.org/2022.coling-1.49 | https://aclanthology.org/2022.coling-1.49.pdf | coling-2022-10 | ['dialogue-understanding'] | ['natural-language-processing'] | [ 5.40734828e-01 1.05734634e+00 2.68934853e-02 -6.20938480e-01
-2.28138074e-01 -5.16521335e-01 1.07094276e+00 2.75003880e-01
-2.76997596e-01 7.39352763e-01 7.12072194e-01 -5.19283414e-01
2.01794356e-01 -9.59956229e-01 -2.45898604e-01 8.89538750e-02
2.75949568e-01 7.79277265e-01 2.50293881e-01 -9.96737659e-01
2.73558378e-01 -2.81790972e-01 -1.02549362e+00 7.76519597e-01
8.72205019e-01 7.86640525e-01 2.54691064e-01 6.25801742e-01
-7.28098094e-01 1.18301618e+00 -8.51676464e-01 -6.05102122e-01
-4.37857509e-01 -9.92978513e-01 -1.82819998e+00 2.18699258e-02
-1.46698654e-01 -2.10083455e-01 -3.01823020e-01 9.35710192e-01
1.14104862e-03 3.54655623e-01 6.14724576e-01 -1.02060759e+00
-5.81017494e-01 1.08865118e+00 1.39490306e-01 -1.17832623e-01
8.10698032e-01 -7.62657002e-02 1.40048707e+00 -7.04525173e-01
6.04498565e-01 1.81950629e+00 6.89519465e-01 1.20315921e+00
-1.05699539e+00 -3.44015136e-02 2.80767441e-01 2.85540909e-01
-6.66232884e-01 -4.75368232e-01 7.36318827e-01 -3.17333527e-02
1.47441804e+00 1.67507231e-01 5.98854899e-01 1.38154554e+00
-3.85751396e-01 1.07350087e+00 1.12447000e+00 -6.12465620e-01
1.13186672e-01 -4.17924710e-02 6.45006478e-01 5.99745929e-01
-2.85863787e-01 -4.90607232e-01 -6.74465358e-01 -1.25527233e-01
8.06461871e-01 -2.99917758e-01 -2.66394645e-01 -2.15842947e-01
-8.45529556e-01 1.29007149e+00 2.61139095e-01 5.51961958e-01
-7.30857765e-03 3.04486305e-02 7.59484828e-01 7.39718735e-01
5.91578424e-01 7.85169542e-01 -8.35039616e-01 -4.97735471e-01
-1.45144582e-01 2.18248487e-01 1.53587723e+00 8.14225495e-01
5.41637301e-01 -2.30543762e-01 -2.98187435e-01 1.19208634e+00
3.22342604e-01 3.66668478e-02 7.07872331e-01 -1.00487816e+00
5.42461038e-01 9.33950067e-01 -1.86648190e-01 -5.92086256e-01
-5.54963171e-01 1.12993345e-01 -5.20837903e-01 -4.88304228e-01
6.06615365e-01 -3.94048005e-01 -4.23801363e-01 1.90100169e+00
-3.32055837e-02 1.32121578e-01 8.21586072e-01 6.98010087e-01
1.23817456e+00 4.66904104e-01 3.28006297e-01 6.49587959e-02
1.66004241e+00 -1.12021422e+00 -8.78131390e-01 -4.40600932e-01
1.11384463e+00 -3.20022285e-01 1.28242600e+00 -2.61374135e-02
-8.95055532e-01 -3.87018055e-01 -8.89841497e-01 -2.75541991e-01
-6.04118407e-01 -1.72929898e-01 9.71873343e-01 5.52483022e-01
-9.60022569e-01 5.85543394e-01 -4.30132061e-01 -7.66370714e-01
2.81571507e-01 -3.33508290e-02 -1.48505479e-01 -2.64657140e-02
-1.79432857e+00 1.21447825e+00 6.76060915e-01 -1.56111270e-01
-9.16353703e-01 -6.22875452e-01 -1.32721424e+00 1.77496165e-01
7.89024293e-01 -9.19451416e-01 1.59930313e+00 -1.08839905e+00
-2.22379112e+00 1.04483426e+00 -8.38793516e-02 -7.49080718e-01
1.34582996e-01 -4.05249327e-01 7.78533369e-02 1.74361974e-01
-3.81795503e-02 6.10534608e-01 5.23358405e-01 -1.27712262e+00
-4.33404326e-01 -4.13982362e-01 8.65198135e-01 7.04974771e-01
-2.57923156e-01 1.01078101e-01 -4.22299793e-03 -5.61120331e-01
-8.29042420e-02 -7.77043283e-01 -2.10636601e-01 -5.80016315e-01
-3.90762597e-01 -7.04665601e-01 6.89275384e-01 -6.96534991e-01
9.28294063e-01 -1.78366446e+00 3.02515030e-01 -2.62106866e-01
1.64274439e-01 3.98466259e-01 -1.99514970e-01 7.60485828e-01
4.60371263e-02 8.31639208e-03 -3.96969587e-01 -4.44845349e-01
2.96080351e-01 6.91977978e-01 -4.46888328e-01 -2.83137918e-01
3.12857360e-01 1.20787585e+00 -1.32156599e+00 -1.72131643e-01
3.65069568e-01 1.87729910e-01 -5.26273906e-01 7.71245003e-01
-7.51663804e-01 5.58361232e-01 -6.30617321e-01 9.23284516e-02
9.46161058e-03 -3.42405766e-01 6.61067963e-01 6.20348193e-02
5.85494041e-01 1.13545513e+00 -5.49441814e-01 2.24758458e+00
-9.83263493e-01 4.75606650e-01 9.96584669e-02 -1.38900506e+00
8.77960861e-01 6.55168355e-01 -1.11535965e-02 -4.66329426e-01
2.03919291e-01 -4.64879535e-02 3.79039757e-02 -2.88990498e-01
8.18132043e-01 -5.36973357e-01 -3.55804324e-01 7.33305633e-01
5.86113453e-01 -3.93804967e-01 1.04423203e-01 4.62578803e-01
8.78958225e-01 1.99234962e-01 4.62589532e-01 -2.71553308e-01
8.27230811e-01 1.22411698e-01 1.25236968e-02 6.49093270e-01
5.42603694e-02 3.93649280e-01 9.96723533e-01 -2.73825973e-01
-5.40939867e-01 -6.97955608e-01 1.21800385e-01 1.48988986e+00
2.18161345e-01 -9.34278071e-01 -1.09474766e+00 -1.21827173e+00
-2.30218112e-01 1.18171680e+00 -5.63979864e-01 -2.25850374e-01
-6.42886877e-01 -3.35859358e-01 7.10237026e-01 4.81064558e-01
6.87410235e-01 -1.29788518e+00 -5.50710618e-01 2.23135501e-01
-3.86648864e-01 -1.53373611e+00 5.41737825e-02 1.53221900e-03
-8.89986873e-01 -1.38164926e+00 -3.60537052e-01 -5.28516829e-01
4.60793436e-01 1.05668277e-01 1.51356375e+00 3.99028033e-01
2.59219468e-01 6.95177913e-01 -8.32708597e-01 -3.36732537e-01
-8.66913259e-01 2.07977727e-01 -4.45579141e-01 -1.59843057e-01
4.46227551e-01 -3.56098652e-01 -1.88402444e-01 1.72299519e-01
-6.23507977e-01 4.06039447e-01 4.63515967e-02 1.08897114e+00
-2.10420236e-01 -2.34958753e-01 7.71644294e-01 -1.40744841e+00
1.26266706e+00 -4.95329350e-01 1.49532750e-01 2.91007906e-01
-3.42197388e-01 2.09312230e-01 7.25916326e-01 1.77075356e-01
-1.51113296e+00 -6.71615541e-01 -4.53092605e-01 2.50384718e-01
-4.85999256e-01 5.00301659e-01 -1.71550423e-01 3.79817426e-01
4.30620581e-01 7.19534457e-02 1.72656715e-01 -6.19584739e-01
7.67049789e-01 5.59736669e-01 4.41344023e-01 -1.16907334e+00
2.72541881e-01 1.97499365e-01 -2.17589945e-01 -9.77444530e-01
-1.38367558e+00 -5.20698607e-01 -7.02358067e-01 3.02375734e-01
1.10852015e+00 -7.39178300e-01 -6.67627513e-01 2.63080776e-01
-1.41749167e+00 -7.16858447e-01 -4.03414845e-01 5.32899611e-02
-8.26092184e-01 6.84110224e-01 -8.23024154e-01 -6.98694170e-01
-4.30955172e-01 -7.01313078e-01 9.87405419e-01 1.11504667e-01
-6.27594352e-01 -1.70281196e+00 -1.12966314e-01 6.06143296e-01
4.89900738e-01 -1.21593021e-01 1.14736688e+00 -1.43957233e+00
1.62852958e-01 3.29401016e-01 -3.32770884e-01 3.78035605e-01
2.86754787e-01 -7.26564944e-01 -1.15631115e+00 -2.04791781e-02
2.58517355e-01 -9.09591854e-01 7.21186459e-01 -8.49622786e-02
1.05319965e+00 -4.36699390e-01 -1.11134574e-01 1.01122767e-01
8.17764699e-01 -4.63301390e-02 5.73342323e-01 2.30716109e-01
6.17188454e-01 1.00468278e+00 6.79286063e-01 3.27779233e-01
8.34042192e-01 6.69550836e-01 2.13372812e-01 -1.13729082e-01
-2.52458364e-01 -5.05643427e-01 3.57052475e-01 9.01108384e-01
-5.01285680e-02 -2.26531215e-02 -8.33774328e-01 4.62110281e-01
-2.15684915e+00 -6.83949351e-01 1.33372769e-01 1.62510586e+00
1.35981345e+00 1.11190096e-01 -2.74816062e-02 -3.09762508e-01
5.03952861e-01 4.79105741e-01 -2.29437441e-01 -6.38035238e-01
1.25153363e-01 4.92821366e-01 -1.70733422e-01 7.94079661e-01
-9.70847189e-01 1.56025600e+00 5.59409666e+00 7.11680114e-01
-3.74620289e-01 1.93714947e-01 3.68538618e-01 8.16657543e-01
-3.28519285e-01 4.58562672e-01 -6.58703685e-01 7.81130418e-02
9.52296734e-01 -1.34266138e-01 3.11286062e-01 7.65408635e-01
-2.33551040e-01 5.71845807e-02 -1.35136700e+00 7.20828354e-01
1.88092574e-01 -1.26627815e+00 4.79028404e-01 -4.00351584e-01
3.31993639e-01 -3.72170687e-01 -5.67526937e-01 7.75846243e-01
7.16883302e-01 -1.30416679e+00 7.60456920e-02 1.98364794e-01
4.87239301e-01 -4.14307058e-01 1.04772151e+00 3.22419137e-01
-8.61604214e-01 1.84575424e-01 -3.41347486e-01 -4.56048369e-01
1.91360086e-01 -3.31760570e-02 -1.14535093e+00 8.81899118e-01
8.52882117e-02 1.03948390e+00 -4.91319805e-01 -1.41960382e-01
-8.69867444e-01 6.34154260e-01 -3.60306613e-02 -4.55027595e-02
5.24010241e-01 -1.70404300e-01 4.15520787e-01 1.39075184e+00
-2.84616381e-01 3.26729298e-01 3.19162965e-01 8.91196311e-01
-1.83188409e-01 1.39832068e-02 -5.31446755e-01 -3.38687509e-01
3.08273077e-01 1.05756772e+00 -5.82895398e-01 -6.24947727e-01
-6.60137951e-01 1.00424516e+00 3.71634364e-01 2.17990890e-01
-2.66981959e-01 -3.40783298e-01 7.94136584e-01 -3.30067575e-01
-1.03235401e-01 -2.25073740e-01 -3.17675203e-01 -1.38661253e+00
-4.46187615e-01 -9.46015775e-01 6.92588925e-01 -6.13246143e-01
-1.34371603e+00 4.20769811e-01 1.83836013e-01 -4.05000955e-01
-5.98699987e-01 -9.38675940e-01 -8.36504400e-01 8.36571097e-01
-1.60835874e+00 -1.39761722e+00 -1.29304156e-01 5.96727431e-01
1.17839921e+00 -3.52239996e-01 1.39982891e+00 -5.17887115e-01
-1.62539184e-01 3.04979801e-01 -5.09380519e-01 4.82409954e-01
4.28067595e-01 -1.62813807e+00 6.41362131e-01 2.67304361e-01
2.52298802e-01 8.11819732e-01 5.94064236e-01 -4.02557284e-01
-1.32318389e+00 -5.09263933e-01 9.08430576e-01 -8.13571751e-01
8.94952297e-01 -4.34652656e-01 -1.17305887e+00 8.18219066e-01
7.03770280e-01 -7.55027890e-01 9.12565410e-01 6.87568605e-01
-3.36670637e-01 6.38887823e-01 -8.23406279e-01 5.67072630e-01
1.22269225e+00 -8.66041958e-01 -1.63535821e+00 3.58755738e-01
1.02786207e+00 -4.76502657e-01 -1.03237689e+00 3.33598137e-01
1.34737521e-01 -8.57555807e-01 9.76946533e-01 -1.05100155e+00
6.58140659e-01 1.83563620e-01 -3.07623576e-02 -1.54911256e+00
4.14163381e-01 -8.18547249e-01 -6.88080043e-02 1.23901498e+00
3.59817326e-01 -5.09006977e-01 5.64939559e-01 5.85142076e-01
-4.07714874e-01 -5.86890638e-01 -6.74567342e-01 -5.52717865e-01
3.50123614e-01 -4.58515048e-01 4.73624617e-01 1.22161794e+00
7.34132290e-01 1.31028175e+00 -2.95296218e-02 -3.83532107e-01
3.23116630e-01 1.95614293e-01 9.36707497e-01 -1.35883057e+00
-3.43614668e-01 -5.72556198e-01 -1.90668937e-03 -1.40215862e+00
8.71157408e-01 -1.04836845e+00 -6.67797700e-02 -1.93776071e+00
7.08499774e-02 -3.25304627e-01 -2.89054681e-02 7.46311545e-01
-5.02704859e-01 -3.35164845e-01 1.39159948e-01 -4.62102741e-02
-9.12393928e-01 9.34464455e-01 1.25471675e+00 4.53164689e-02
7.96968956e-03 -1.13826901e-01 -8.20880711e-01 8.69953454e-01
8.74898553e-01 -8.42140168e-02 -8.13720286e-01 -3.36299866e-01
1.55790091e-01 6.01402745e-02 2.44229585e-01 -3.46498460e-01
-4.40353751e-02 -2.08311543e-01 -3.28725845e-01 -3.12526710e-02
4.11972046e-01 -5.25005519e-01 -6.97342098e-01 2.76512027e-01
-8.34237993e-01 -4.66031373e-01 1.13050088e-01 3.68858546e-01
-5.27392924e-01 -5.62772632e-01 5.28340876e-01 -5.31139076e-01
-1.05143654e+00 -2.91575044e-01 -3.29288721e-01 7.31850207e-01
3.88085395e-01 -3.95137779e-02 -4.28688526e-01 -7.24864960e-01
-8.58109772e-01 4.51543897e-01 3.18691730e-01 6.76938474e-01
5.09463191e-01 -9.35043335e-01 -6.06624007e-01 -1.10547148e-01
1.31638110e-01 1.69385850e-01 4.07851040e-02 4.32514906e-01
-2.14816913e-01 5.61686456e-01 -1.91983566e-01 -3.36765736e-01
-1.23325682e+00 2.54890174e-01 3.26620549e-01 -6.23484969e-01
-8.80953670e-01 7.49877274e-01 5.73650002e-01 -8.78267705e-01
8.03007856e-02 -2.66963929e-01 -6.18126988e-01 -5.52507862e-02
4.43879277e-01 5.13457991e-02 -2.09764838e-01 -5.33527136e-01
-1.32463187e-01 3.00273240e-01 -4.49521802e-02 -1.73325881e-01
1.25488675e+00 -3.39176923e-01 -2.58102745e-01 5.04197419e-01
9.08237159e-01 -5.10401130e-01 -9.40369785e-01 -5.80166459e-01
5.87648273e-01 -1.74609393e-01 -4.28971708e-01 -8.21893454e-01
-5.32957792e-01 1.17675626e+00 -3.26995969e-01 5.45558453e-01
6.24965191e-01 3.21599960e-01 9.41847622e-01 8.06509197e-01
3.16696048e-01 -1.23580360e+00 5.01085997e-01 1.40866899e+00
1.05443192e+00 -1.07899010e+00 -2.93616235e-01 -6.27857208e-01
-1.03377819e+00 1.42588151e+00 7.32821226e-01 4.97469008e-02
3.16919774e-01 -4.53833155e-02 1.50811687e-01 -5.14984131e-01
-9.69722629e-01 -5.30554235e-01 6.69325516e-03 7.09111452e-01
8.15785348e-01 2.43306458e-02 -2.98502922e-01 9.62053239e-01
-3.59662831e-01 -1.90276995e-01 4.65669096e-01 9.93593335e-01
-3.48101228e-01 -1.29798937e+00 2.22081169e-01 9.11261514e-02
-3.83418322e-01 -3.81718993e-01 -1.04748297e+00 6.51511490e-01
-6.03562474e-01 1.26730013e+00 -1.03901312e-01 -1.88445434e-01
5.28792977e-01 6.84800386e-01 6.38212979e-01 -1.27922821e+00
-8.97171795e-01 -5.23726046e-01 9.90867078e-01 -5.20017862e-01
-6.50504291e-01 -6.06599376e-02 -1.60290420e+00 -8.77543986e-02
-8.68590549e-02 6.81421697e-01 3.41807961e-01 1.41119099e+00
1.77247047e-01 6.26520753e-01 3.67284924e-01 -4.10026163e-01
-7.48827040e-01 -1.37657464e+00 -2.88619041e-01 9.03148055e-01
-1.34473443e-01 -6.56556189e-01 -2.61882901e-01 8.05296563e-03] | [12.42521858215332, 8.150345802307129] |
45e1307b-28aa-4440-a2d5-910c75ae488a | are-training-trajectories-of-deep-single | 2306.08744 | null | https://arxiv.org/abs/2306.08744v1 | https://arxiv.org/pdf/2306.08744v1.pdf | Are training trajectories of deep single-spike and deep ReLU network equivalent? | Communication by binary and sparse spikes is a key factor for the energy efficiency of biological brains. However, training deep spiking neural networks (SNNs) with backpropagation is harder than with artificial neural networks (ANNs), which is puzzling given that recent theoretical results provide exact mapping algorithms from ReLU to time-to-first-spike (TTFS) SNNs. Building upon these results, we analyze in theory and in simulation the learning dynamics of TTFS-SNNs. Our analysis highlights that even when an SNN can be mapped exactly to a ReLU network, it cannot always be robustly trained by gradient descent. The reason for that is the emergence of a specific instance of the vanishing-or-exploding gradient problem leading to a bias in the gradient descent trajectory in comparison with the equivalent ANN. After identifying this issue we derive a generic solution for the network initialization and SNN parameterization which guarantees that the SNN can be trained as robustly as its ANN counterpart. Our theoretical findings are illustrated in practice on image classification datasets. Our method achieves the same accuracy as deep ConvNets on CIFAR10 and enables fine-tuning on the much larger PLACES365 dataset without loss of accuracy compared to the ANN. We argue that the combined perspective of conversion and fine-tuning with robust gradient descent in SNN will be decisive to optimize SNNs for hardware implementations needing low latency and resilience to noise and quantization. | ['Wulfram Gerstner', 'Angeliki Pantazi', 'Giovanni Cherubini', 'Guillaume Bellec', 'Stanisław Woźniak', 'Ana Stanojevic'] | 2023-06-14 | null | null | null | null | ['quantization'] | ['methodology'] | [ 5.08750737e-01 9.03152395e-03 3.40263307e-01 -2.44182289e-01
1.99513555e-01 -4.50224996e-01 4.21572089e-01 -8.08460042e-02
-8.28790486e-01 1.04652333e+00 -4.31760281e-01 -2.47423723e-01
-3.50369602e-01 -8.01560998e-01 -1.08167756e+00 -1.14862800e+00
-1.19108364e-01 2.95526415e-01 3.00440997e-01 -2.68825293e-01
3.18687320e-01 8.30319881e-01 -1.47888982e+00 -8.45721364e-03
4.28051084e-01 1.32394528e+00 4.49906200e-01 6.14435017e-01
-1.12984180e-01 6.66989207e-01 -6.27373099e-01 -2.05105424e-01
5.16665161e-01 -5.67966104e-01 -4.62609202e-01 -4.52976227e-01
1.35293230e-01 1.01486407e-01 -3.18394542e-01 1.07595384e+00
5.91435015e-01 -1.82638466e-01 6.43354058e-01 -1.17186356e+00
-4.11497056e-01 9.08356369e-01 -1.24125525e-01 4.01275694e-01
-5.21150172e-01 1.11679047e-01 4.21094000e-01 -6.38605535e-01
5.81524372e-01 7.42368162e-01 9.34890270e-01 8.14900875e-01
-1.39816236e+00 -7.63846099e-01 -2.66040474e-01 2.54427865e-02
-1.34883761e+00 -6.22291982e-01 4.15235460e-01 -4.86806706e-02
1.25084567e+00 -6.50145784e-02 1.13136137e+00 1.14052868e+00
4.91501749e-01 1.99295655e-01 1.10852873e+00 -3.03042501e-01
7.15361297e-01 -7.77507871e-02 1.51911587e-01 4.04692650e-01
6.30351603e-01 6.82233125e-02 -6.36284351e-01 2.94233441e-01
1.15215003e+00 -5.54099083e-02 -3.12914878e-01 -1.61542967e-01
-1.22997046e+00 4.43377763e-01 8.88644814e-01 5.17254889e-01
-5.20386636e-01 9.87402558e-01 2.13208497e-01 4.46879625e-01
-2.92928845e-01 5.51076353e-01 -4.25328344e-01 -1.03101850e-01
-1.04045451e+00 -6.60998300e-02 8.46572638e-01 6.87437057e-01
7.39793181e-01 4.22587365e-01 1.34968236e-01 5.00415623e-01
-4.51563746e-02 5.74179351e-01 6.91368103e-01 -1.00291157e+00
3.93725149e-02 5.26447952e-01 -2.52977759e-01 -5.35835147e-01
-6.87766850e-01 -9.04934704e-01 -1.24325776e+00 6.30534053e-01
7.38969088e-01 -3.32498997e-01 -8.15766096e-01 1.89084363e+00
-3.52007419e-01 3.13644886e-01 2.66407967e-01 7.71778286e-01
4.12809074e-01 6.88962221e-01 -2.56112516e-01 -2.68440366e-01
9.86004591e-01 -4.55724061e-01 -5.69064617e-01 -6.02668583e-01
4.13232595e-01 -1.27509326e-01 6.57855511e-01 2.56867349e-01
-1.22279036e+00 -3.23158294e-01 -1.39391899e+00 3.53910923e-02
-5.33979714e-01 -9.12358165e-02 6.89531267e-01 6.71053052e-01
-1.51472569e+00 9.74415898e-01 -9.69765961e-01 -5.84529459e-01
5.82983434e-01 8.78751218e-01 -8.29420984e-02 4.19673920e-01
-8.76061440e-01 9.75873053e-01 6.16132915e-01 2.85898179e-01
-6.17265105e-01 -5.83682179e-01 -1.53016001e-01 1.62219957e-01
-4.26221520e-01 -7.14994490e-01 9.24531162e-01 -1.22963178e+00
-1.69179022e+00 6.92398608e-01 -9.39201936e-02 -1.14558029e+00
3.56754780e-01 4.88714576e-01 -8.18347707e-02 -7.64808878e-02
-5.78836977e-01 9.86205339e-01 6.86916351e-01 -9.50912714e-01
-3.95478189e-01 -3.79176289e-01 -2.21198693e-01 -1.91058382e-01
-5.16943514e-01 -3.75932664e-01 8.29859972e-02 -4.69108492e-01
4.63622153e-01 -8.31266820e-01 -1.90979481e-01 1.62101001e-01
-1.18467420e-01 1.66380703e-02 6.78656340e-01 8.65155905e-02
8.72395337e-01 -2.01802778e+00 1.11720525e-01 3.10423553e-01
1.10427640e-01 1.57388315e-01 -1.03536263e-01 9.53517854e-02
-5.69249727e-02 -2.23090470e-01 -3.17873001e-01 -9.27429944e-02
-1.55034378e-01 4.94000882e-01 -4.76541221e-01 6.01345241e-01
2.36216247e-01 1.18866599e+00 -6.75100505e-01 -2.24452950e-02
-6.02383316e-02 8.50359321e-01 -4.93467748e-01 -2.53989309e-01
8.88412148e-02 2.94611692e-01 -1.06660545e-01 2.99588233e-01
4.42268491e-01 -3.68380457e-01 3.92607413e-02 -2.54427373e-01
-4.00594383e-01 2.88459748e-01 -9.37882900e-01 1.54266632e+00
-4.48388398e-01 1.04889882e+00 6.16161637e-02 -1.36817992e+00
1.15970981e+00 8.58828947e-02 2.38225073e-01 -1.01598132e+00
4.61869329e-01 8.02891076e-01 3.40460271e-01 1.14953421e-01
-1.55904397e-01 9.20445565e-03 2.23735243e-01 4.30246830e-01
5.52624106e-01 6.82632253e-02 1.95409328e-01 -1.07531585e-01
1.21523035e+00 -1.41006872e-01 -6.18419461e-02 -8.20923150e-01
1.02534406e-01 -2.96855003e-01 2.73496419e-01 8.71303618e-01
-2.87511982e-02 3.91370595e-01 4.95690197e-01 -4.66312885e-01
-1.13485670e+00 -1.06990743e+00 -5.49950302e-01 7.49607205e-01
1.20024092e-01 4.34293561e-02 -1.16371715e+00 9.34997201e-02
-2.88001299e-01 4.32729244e-01 -7.53493726e-01 -2.76880741e-01
-7.07819223e-01 -1.08360422e+00 8.20241511e-01 3.45352501e-01
7.25236297e-01 -1.33549714e+00 -1.41456485e+00 5.47930062e-01
3.25826168e-01 -1.03567219e+00 1.80298239e-01 1.41078317e+00
-1.10032094e+00 -6.49212241e-01 -6.31571233e-01 -9.04071748e-01
8.44063044e-01 -3.47381592e-01 9.60693598e-01 1.07076749e-01
-3.21456343e-01 -6.68446124e-02 5.45601696e-02 -4.25896257e-01
-2.03800723e-01 2.34675288e-01 1.70032084e-01 -2.37931460e-01
4.10218507e-01 -1.18724525e+00 -7.82447278e-01 8.82193893e-02
-8.16306889e-01 4.67349812e-02 6.44251525e-01 7.29818821e-01
7.72546887e-01 9.03321430e-03 7.60376096e-01 -4.25374925e-01
3.35815072e-01 -2.26513207e-01 -9.35699463e-01 1.00168390e-02
-8.20087135e-01 6.26606405e-01 8.80664229e-01 -4.35757339e-01
-3.92247766e-01 3.93755227e-01 -2.80623287e-01 8.56248755e-03
1.08211851e-02 2.77890302e-02 3.09667736e-01 -7.49365032e-01
8.94130588e-01 5.72252989e-01 -2.30040297e-01 1.41839581e-02
-9.37959403e-02 2.03008965e-01 8.19166362e-01 -2.39989683e-01
6.97312415e-01 5.52313030e-01 3.72975916e-01 -7.45633006e-01
-3.15340042e-01 2.42559075e-01 -4.49339181e-01 -2.99035728e-01
7.54798651e-01 -5.20889640e-01 -1.04000938e+00 5.86470962e-01
-1.32434130e+00 -7.12668419e-01 -5.91494858e-01 2.20206663e-01
-6.24433696e-01 -3.41384947e-01 -7.21514404e-01 -7.22427726e-01
-4.44666833e-01 -9.37806129e-01 4.08778518e-01 4.97805804e-01
4.22917679e-03 -9.71140087e-01 -1.28646672e-01 -5.39338470e-01
9.10403550e-01 9.32834595e-02 7.83006251e-01 -5.34824908e-01
-5.07122517e-01 5.35276979e-02 -2.67182082e-01 2.62310505e-01
-2.98302501e-01 2.06461176e-03 -1.11232877e+00 -2.78930992e-01
3.20440322e-01 -3.65631729e-01 1.13425517e+00 7.54982114e-01
1.03291965e+00 -3.35825086e-01 -4.18517232e-01 1.05243242e+00
1.92714286e+00 1.10685259e-01 6.57686055e-01 4.05954719e-01
2.66181558e-01 2.25070700e-01 -5.51520407e-01 3.59570235e-01
-3.03145442e-02 3.43097121e-01 7.27254093e-01 1.70466751e-01
-2.83972621e-01 6.05715774e-02 3.23374093e-01 8.61108959e-01
-2.31788710e-01 -6.47293180e-02 -8.28977883e-01 3.96149844e-01
-1.50346446e+00 -8.28011692e-01 -9.82985124e-02 2.31172895e+00
8.94081295e-01 4.74758744e-01 -2.93347210e-01 4.05133843e-01
5.21567702e-01 -2.97456115e-01 -8.18858206e-01 -5.27972460e-01
-6.19065046e-01 5.73259890e-01 1.10604811e+00 2.04251096e-01
-5.37468076e-01 5.39615452e-01 6.52375841e+00 6.03304982e-01
-1.58796263e+00 9.24852192e-02 6.57761514e-01 -3.07044446e-01
-1.40397966e-01 -3.21538955e-01 -1.05565715e+00 5.41573465e-01
1.42451298e+00 -4.32192273e-02 1.08780801e+00 4.60776925e-01
-1.34665579e-01 3.09837889e-02 -1.24522674e+00 1.29024577e+00
-4.25152868e-01 -1.66814625e+00 -1.38074979e-01 -4.41881977e-02
7.21538961e-01 5.45058191e-01 1.91728049e-03 1.49770707e-01
7.98057914e-02 -1.25778294e+00 9.25538778e-01 3.99803519e-01
5.73054254e-01 -7.00056374e-01 5.99897683e-01 4.24997032e-01
-9.86544073e-01 -3.13890964e-01 -7.73758888e-01 -2.93527842e-01
-1.01127326e-01 8.93459976e-01 -4.86388057e-01 -1.51226297e-01
7.72141516e-01 4.73051250e-01 -4.50999111e-01 1.20901370e+00
1.42488971e-01 4.47880000e-01 -8.37054789e-01 -4.85405952e-01
2.53115505e-01 -1.75116643e-01 1.33912846e-01 1.30056787e+00
7.63031840e-01 3.65395136e-02 -7.45246410e-01 1.19238687e+00
-4.59088236e-01 -2.96994001e-01 -4.35188860e-01 -2.32383281e-01
7.21012056e-01 1.31276298e+00 -1.33737385e+00 -2.89890170e-02
-1.43175591e-02 1.01024604e+00 4.07761782e-01 3.83128077e-01
-6.44544542e-01 -6.86080694e-01 4.28641230e-01 1.22300036e-01
5.03644168e-01 -2.22163454e-01 -8.78027380e-01 -6.31532967e-01
-2.25200020e-02 -1.75536305e-01 -3.36593360e-01 -6.34822249e-01
-8.50551188e-01 9.92241144e-01 -6.48481011e-01 -7.75810003e-01
-2.34858587e-01 -9.32920754e-01 -4.63284552e-01 5.96602917e-01
-1.58297503e+00 -4.19389307e-01 -3.30397755e-01 7.48550653e-01
6.66288882e-02 8.66289511e-02 9.22603846e-01 3.63378286e-01
-4.41366374e-01 5.62278390e-01 2.96236038e-01 -1.91643126e-02
1.05786264e-01 -1.02736115e+00 4.20845836e-01 7.82109678e-01
1.88437030e-01 5.05821347e-01 7.90274680e-01 -1.23310566e-01
-1.57803440e+00 -1.06133425e+00 7.10367322e-01 -1.42525077e-01
7.70910084e-01 -6.37560010e-01 -8.37947845e-01 4.76081133e-01
1.90559015e-01 2.96329021e-01 3.99092324e-02 -6.08347774e-01
-1.10212065e-01 -3.63909245e-01 -1.09180117e+00 5.15680671e-01
1.29054594e+00 -2.61011809e-01 -1.89652234e-01 2.04111263e-01
3.08661878e-01 -1.48302421e-01 -5.78229368e-01 2.59113520e-01
5.04509509e-01 -1.19103014e+00 7.99844980e-01 4.68883775e-02
1.88237146e-01 -2.41586387e-01 -1.93953559e-01 -1.23909628e+00
-1.18482150e-01 -6.27083957e-01 -1.11800037e-01 7.91751683e-01
4.84812707e-01 -9.06583250e-01 8.06699634e-01 1.67103440e-01
-4.49469984e-02 -8.12454104e-01 -1.38289845e+00 -1.05339766e+00
1.88925743e-01 -3.24716777e-01 2.84194618e-01 5.33312261e-01
-1.34007365e-01 9.97798592e-02 1.77544296e-01 8.78096819e-02
7.04716682e-01 -3.08053523e-01 -2.12005321e-02 -1.47795343e+00
-8.82039219e-02 -8.64656866e-01 -7.03507364e-01 -8.43974948e-01
-4.11972888e-02 -9.55133438e-01 2.80403495e-01 -1.38801920e+00
-1.07630730e-01 -6.19632721e-01 -6.40429378e-01 5.07614195e-01
6.35822058e-01 7.24656701e-01 -7.72259161e-02 1.78020418e-01
-2.91325331e-01 8.43842030e-02 8.28037500e-01 1.52138337e-01
-3.85664888e-02 -1.39266536e-01 -2.77549595e-01 6.05092108e-01
8.80214930e-01 -7.50034630e-01 -3.23837012e-01 -5.13012767e-01
7.76405632e-01 -2.48112082e-01 5.31227589e-01 -1.75917077e+00
8.61038566e-01 4.01412010e-01 5.69762945e-01 -7.60473907e-02
2.49918222e-01 -9.96844530e-01 3.11117083e-01 9.60229337e-01
-3.59607220e-01 -3.41764241e-02 2.53127426e-01 2.48028025e-01
1.30244210e-01 -4.96802896e-01 1.10058486e+00 -1.96329113e-02
-4.95543897e-01 2.02570215e-01 -5.99036157e-01 -4.49795872e-02
9.06755686e-01 -5.83396435e-01 -4.34557974e-01 -1.89184062e-02
-4.94459569e-01 -1.69523984e-01 3.72357488e-01 -8.50219503e-02
4.99306321e-01 -1.23605502e+00 -5.02111495e-01 5.73878706e-01
-3.24733317e-01 -1.81069538e-01 -1.83613732e-01 9.26043332e-01
-6.22793734e-01 4.82271761e-01 -7.52404511e-01 -6.48768127e-01
-4.78064775e-01 3.07670027e-01 9.00406361e-01 1.02863722e-01
-4.19836313e-01 1.10822499e+00 -1.58644497e-01 -1.65424541e-01
4.99116421e-01 -5.72673857e-01 1.36065111e-01 -1.88791037e-01
4.32676256e-01 1.44707680e-01 3.93179595e-01 -1.04507066e-01
-4.92854357e-01 6.48291707e-01 2.73932785e-01 -8.98936018e-03
1.63389373e+00 5.02806418e-02 -4.11776811e-01 5.72168529e-01
1.12602901e+00 -4.77691233e-01 -1.60669231e+00 5.37674800e-02
4.35140207e-02 4.16232467e-01 2.10499674e-01 -7.58485556e-01
-1.48132503e+00 1.09404635e+00 9.63192821e-01 2.78731257e-01
1.24320519e+00 -1.95266306e-01 6.96628153e-01 9.73071873e-01
5.17803669e-01 -1.10293818e+00 -9.37660411e-02 6.18808925e-01
7.08308578e-01 -7.73208439e-01 -5.10232091e-01 2.28273064e-01
1.53033051e-03 1.59604359e+00 3.75232458e-01 -7.11236894e-01
7.23014772e-01 1.03735912e+00 -1.89316943e-01 -2.59966943e-02
-7.80478477e-01 2.35628888e-01 -3.44635755e-01 5.02599359e-01
1.92023993e-01 -1.56110078e-01 -7.39471689e-02 5.44600308e-01
-4.00756925e-01 4.26998377e-01 6.25762701e-01 7.35177159e-01
-7.72382617e-01 -7.88615823e-01 4.62032184e-02 4.67616618e-01
-3.97948503e-01 -3.51940960e-01 3.74590419e-02 4.99599993e-01
5.88919632e-02 4.62991267e-01 4.64603096e-01 -2.94088215e-01
5.42704538e-02 1.48223743e-01 8.81248653e-01 -5.61448187e-02
-8.17456424e-01 -2.56777316e-01 -5.00837028e-01 -5.54247916e-01
-2.00419739e-01 -1.98074311e-01 -1.92353368e+00 -4.77750152e-01
-1.73710361e-01 -5.02072684e-02 1.21167827e+00 1.03649533e+00
2.61676788e-01 8.78675401e-01 2.91060656e-01 -1.11470771e+00
-4.24420327e-01 -5.39399087e-01 -6.54384077e-01 -8.47542733e-02
4.93159533e-01 -3.05216730e-01 -5.37700534e-01 -1.06260402e-03] | [8.164657592773438, 2.579181432723999] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.