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 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
405751a7-cb45-461b-ad09-e5718f515d4c | a-cloud-edge-terminal-collaborative-system | 2107.05078 | null | https://arxiv.org/abs/2107.05078v1 | https://arxiv.org/pdf/2107.05078v1.pdf | A Cloud-Edge-Terminal Collaborative System for Temperature Measurement in COVID-19 Prevention | To prevent the spread of coronavirus disease 2019 (COVID-19), preliminary temperature measurement and mask detection in public areas are conducted. However, the existing temperature measurement methods face the problems of safety and deployment. In this paper, to realize safe and accurate temperature measurement even when a person's face is partially obscured, we propose a cloud-edge-terminal collaborative system with a lightweight infrared temperature measurement model. A binocular camera with an RGB lens and a thermal lens is utilized to simultaneously capture image pairs. Then, a mobile detection model based on a multi-task cascaded convolutional network (MTCNN) is proposed to realize face alignment and mask detection on the RGB images. For accurate temperature measurement, we transform the facial landmarks on the RGB images to the thermal images by an affine transformation and select a more accurate temperature measurement area on the forehead. The collected information is uploaded to the cloud in real time for COVID-19 prevention. Experiments show that the detection model is only 6.1M and the average detection speed is 257ms. At a distance of 1m, the error of indoor temperature measurement is about 3%. That is, the proposed system can realize real-time temperature measurement in public areas. | ['Zhiyong Bu', 'Bin Zhou', 'Qingwen Liu', 'Wen Fang', 'Hao Li', 'Zheyi Ma'] | 2021-07-11 | null | null | null | null | ['face-alignment'] | ['computer-vision'] | [-8.99493098e-02 -8.43860984e-01 3.89177918e-01 -3.90500337e-01
-2.96902567e-01 -5.34478009e-01 -1.35812105e-03 -5.00200927e-01
-6.96027040e-01 1.17036372e-01 -6.93617165e-01 -3.98297280e-01
1.71437487e-01 -6.41363323e-01 -3.46298218e-01 -1.13114214e+00
4.41789478e-01 1.00052558e-01 -1.53164595e-01 1.83945894e-01
-3.49078327e-01 5.90647161e-01 -1.34702766e+00 3.77240255e-02
9.91456807e-01 1.19245505e+00 2.69549459e-01 5.51878572e-01
3.71604472e-01 -2.13583007e-01 -8.55003059e-01 -1.53786406e-01
3.46492052e-01 2.18257327e-02 -1.63824320e-01 -3.31926681e-02
2.75749058e-01 -1.04770839e+00 2.76827008e-01 9.12711084e-01
5.97029328e-01 -4.06517357e-01 3.42369527e-01 -1.60281777e+00
-2.83590555e-01 -3.79794449e-01 -8.25324059e-01 -8.43222290e-02
-4.77049723e-02 1.73382044e-01 -1.66550145e-01 -4.82615501e-01
5.54354228e-02 9.30860043e-01 8.48077774e-01 6.08208179e-01
-6.93602800e-01 -1.06407464e+00 -1.72771960e-01 2.16483518e-01
-2.08079362e+00 -2.17844903e-01 7.31869221e-01 -2.96641916e-01
5.97048879e-01 5.40570140e-01 8.66933227e-01 6.58795774e-01
4.69915628e-01 5.55619858e-02 1.13736212e+00 2.28251587e-03
-7.21819177e-02 3.64248425e-01 -4.91289943e-01 7.92083502e-01
6.00276411e-01 3.83369857e-03 1.88792750e-01 -2.36803308e-01
6.94190025e-01 8.39326859e-01 -3.23711187e-01 3.37084144e-01
-8.25045764e-01 4.54564214e-01 3.93008024e-01 4.12371457e-01
-1.71006680e-01 -2.54244898e-02 1.35129556e-01 3.95649299e-02
6.34081602e-01 -2.67204076e-01 -4.64412421e-01 2.47694254e-02
-9.79279280e-01 -2.76677579e-01 1.44137248e-01 9.58881438e-01
7.35758603e-01 -2.52883047e-01 6.02633096e-02 3.25472742e-01
6.91471517e-01 1.57795477e+00 6.81594685e-02 -4.67365563e-01
2.80716300e-01 6.69945657e-01 3.45729113e-01 -1.01908791e+00
-6.87377334e-01 -1.10321455e-02 -1.05206418e+00 2.38517642e-01
2.06082344e-01 -7.70306826e-01 -7.95604765e-01 1.17843914e+00
7.55114198e-01 6.25718415e-01 -2.82126874e-01 1.22449625e+00
5.85495591e-01 6.27362311e-01 -2.77940124e-01 -6.55947745e-01
1.56907451e+00 -5.23862422e-01 -8.21125031e-01 3.58061939e-02
7.07373738e-01 -9.36266541e-01 6.37860298e-01 3.08734357e-01
-3.04559588e-01 -5.12992740e-01 -1.03513288e+00 2.26612628e-01
-2.81766951e-01 6.23294413e-01 2.21643075e-01 1.25002813e+00
-1.11873531e+00 -5.84585927e-02 -7.57520795e-01 -3.26930791e-01
2.16407478e-01 4.28866774e-01 1.05019845e-01 -5.83842583e-02
-9.90614951e-01 7.22723961e-01 -1.27492622e-01 8.08095098e-01
-7.19735324e-01 -4.78861511e-01 -5.10529637e-01 -3.74840856e-01
7.27354363e-02 -6.02529764e-01 9.74815309e-01 -6.95419729e-01
-1.46045876e+00 8.73547316e-01 -3.67752492e-01 1.13436505e-01
4.45503384e-01 -3.27522248e-01 -6.37864888e-01 2.19806165e-01
-2.30305403e-01 1.24399185e-01 1.04438066e+00 -8.50394189e-01
-5.94508290e-01 -7.15530336e-01 -3.11859548e-01 1.17998302e-01
-4.64506924e-01 4.51131821e-01 -4.04543698e-01 1.51082769e-01
-1.47636592e-01 -1.10439324e+00 -1.04288682e-01 1.64495453e-01
-1.99556991e-01 -3.34095280e-03 1.46410728e+00 -7.34048665e-01
8.77069652e-01 -2.11284542e+00 -5.54498672e-01 3.51019710e-01
1.51722401e-01 5.56145549e-01 7.66403154e-02 -2.17636690e-01
3.26916277e-01 4.31154035e-02 6.30046660e-03 -4.51399714e-01
-4.89109427e-01 -5.12229614e-02 1.51817977e-01 9.66515303e-01
-3.32489848e-01 8.53222251e-01 -4.51222152e-01 -6.85509861e-01
6.46744668e-01 9.51055110e-01 -9.54296067e-02 4.56560999e-01
1.80880263e-01 5.44101417e-01 -4.97564554e-01 6.85556471e-01
1.58305228e+00 -4.61641215e-02 -7.51256049e-02 -2.35888660e-01
-3.82711470e-01 -4.15470421e-01 -9.94131804e-01 1.34620464e+00
-5.78265548e-01 3.90218079e-01 4.07831758e-01 -1.93401709e-01
9.89949703e-01 5.68205357e-01 7.85835028e-01 -5.15610218e-01
6.36667669e-01 -3.09616271e-02 -4.60814178e-01 -7.82690346e-01
1.84124514e-01 3.17389965e-02 2.13572219e-01 4.91786182e-01
-6.90935612e-01 2.94332914e-02 -6.63336575e-01 -3.46557915e-01
5.26919603e-01 -3.36538218e-02 -3.94594729e-01 1.67030990e-02
5.21564543e-01 -2.07196146e-01 7.69629180e-01 1.71300903e-01
-5.99880993e-01 5.01458466e-01 -3.92070651e-01 -6.12228453e-01
-8.13540876e-01 -8.99425089e-01 -1.02134936e-01 5.34752011e-01
5.07828116e-01 -2.59919256e-01 -1.14195549e+00 -3.53430748e-01
-2.06856489e-01 -2.26706564e-02 -2.57042438e-01 -2.27430776e-01
-5.49165487e-01 -8.89153719e-01 5.03159940e-01 4.14083213e-01
1.11282122e+00 -5.09102702e-01 -9.42322552e-01 -2.49004915e-01
-4.73152429e-01 -1.18407798e+00 -6.07893884e-01 -5.86277425e-01
-5.83298385e-01 -1.20176625e+00 -5.49132943e-01 -7.27657080e-01
8.81172061e-01 7.80158043e-01 2.94569522e-01 4.81178254e-01
-6.00956917e-01 2.10178599e-01 -1.46991968e-01 -7.15228498e-01
2.20550179e-01 -2.76001364e-01 3.27663392e-01 1.88637659e-01
8.72261703e-01 1.47848548e-02 -1.04259109e+00 5.69666862e-01
-5.35098612e-01 -5.34521900e-02 2.49981284e-01 2.53746092e-01
2.52200484e-01 1.43431619e-01 -7.46511146e-02 -1.15418971e-01
3.14112931e-01 -3.47941518e-02 -1.19362617e+00 3.99637222e-01
-6.63112462e-01 -7.23101199e-01 4.99042332e-01 -3.25993061e-01
-1.05974030e+00 2.22419798e-01 6.18195608e-02 -7.74068773e-01
-2.10211992e-01 -1.24817736e-01 -2.64639676e-01 -2.75845706e-01
3.56013685e-01 1.57896623e-01 1.96833625e-01 -2.86642492e-01
1.40008107e-01 1.19500613e+00 1.42808646e-01 1.00171283e-01
1.14081240e+00 7.29545414e-01 -1.00332290e-01 -9.88388538e-01
-1.29466116e-01 -5.55542767e-01 -6.04714632e-01 -6.92257941e-01
1.32502508e+00 -1.17365706e+00 -1.52353013e+00 1.03481936e+00
-1.33438540e+00 -1.88581795e-02 8.27334762e-01 7.57348895e-01
7.70994872e-02 2.78863192e-01 -6.62203789e-01 -1.13323462e+00
-9.42569613e-01 -1.06998539e+00 1.41319978e+00 5.65230846e-01
3.38489980e-01 -6.27701581e-01 -1.97624728e-01 6.83615208e-01
6.34769082e-01 9.94338989e-02 2.48979583e-01 2.92546481e-01
-6.57266200e-01 -2.14251548e-01 -1.56701133e-01 1.59499362e-01
5.70590198e-01 3.16311061e-01 -1.29655468e+00 -5.50108790e-01
5.11260569e-01 1.54825315e-01 4.53425139e-01 5.46540380e-01
1.22658706e+00 -1.18453607e-01 -7.99255133e-01 9.62252378e-01
1.50492275e+00 5.98119557e-01 4.72552717e-01 9.01118889e-02
8.76624227e-01 2.15620100e-01 8.05590510e-01 4.65829849e-01
3.47007036e-01 4.68146443e-01 7.32070625e-01 -5.45271099e-01
5.76923013e-01 1.90194681e-01 1.91617623e-01 6.09625220e-01
-2.54039317e-01 -7.39788637e-02 -7.93622732e-01 7.37451017e-02
-1.43107378e+00 -7.46268690e-01 -2.95704514e-01 2.32554388e+00
3.65838349e-01 -4.11596596e-01 1.53001468e-03 -2.02458873e-01
1.07825804e+00 -4.66514044e-02 -4.17870998e-01 -3.37959915e-01
3.51183802e-01 -2.35472769e-01 6.98815107e-01 4.26503956e-01
-1.24857867e+00 5.04696548e-01 5.56656981e+00 4.13200706e-01
-1.54940951e+00 2.84764290e-01 8.08236659e-01 -4.29528296e-01
-5.16598206e-03 -5.15333176e-01 -8.47628593e-01 6.66894078e-01
8.60225558e-01 2.86716372e-01 4.77017105e-01 8.77155066e-01
6.18178785e-01 7.89169967e-03 -4.94203329e-01 1.46126485e+00
1.58242241e-01 -8.12201977e-01 -5.75293660e-01 1.62557021e-01
2.86893308e-01 3.01421378e-02 2.27846652e-01 -2.12255955e-01
-3.71478200e-01 -8.31370831e-01 7.29082823e-02 3.80977809e-01
1.16210341e+00 -9.65117574e-01 8.60503197e-01 4.10421848e-01
-1.49154413e+00 1.09563291e-01 -5.48140168e-01 3.24803707e-03
4.47691567e-02 7.23042667e-01 -9.94401097e-01 4.34205323e-01
1.06994677e+00 1.85077772e-01 -2.84455985e-01 8.89487445e-01
-1.60908490e-01 3.85763794e-01 -7.83389866e-01 -4.96883929e-01
-4.99216877e-02 -7.29731858e-01 6.95751533e-02 9.40679789e-01
6.54221237e-01 1.12832636e-01 1.13992125e-01 7.14906991e-01
1.01792358e-01 8.58932212e-02 -6.24346316e-01 7.14789212e-01
6.15726233e-01 1.72098815e+00 -5.48982918e-01 -1.83019444e-01
-3.39862049e-01 1.25581706e+00 -2.68145144e-01 4.47021544e-01
-1.36784744e+00 -4.88131225e-01 1.05121827e+00 -1.48119569e-01
1.71673253e-01 -3.68117958e-01 -1.68769836e-01 -1.00459731e+00
3.61581892e-01 -2.65721381e-01 -2.15757154e-02 -8.42700064e-01
-5.65719783e-01 5.06473601e-01 -3.07812810e-01 -1.38217318e+00
2.95559913e-01 -6.28181338e-01 -9.05687511e-01 9.35896754e-01
-1.33123589e+00 -1.31083453e+00 -7.63709307e-01 8.77082288e-01
-1.23486049e-01 2.51425833e-01 1.00640631e+00 5.30306160e-01
-9.99899924e-01 7.61917889e-01 3.68902028e-01 3.04309398e-01
7.27379143e-01 -5.37886858e-01 2.22521171e-01 9.51449931e-01
-4.88168925e-01 6.80874050e-01 9.99659523e-02 -7.39165187e-01
-1.58249474e+00 -1.50718486e+00 6.49780869e-01 -2.06222758e-01
-1.11378901e-01 -8.43807876e-01 -3.94120246e-01 5.41695356e-01
3.55215043e-01 1.44656643e-01 5.47276735e-01 -2.64889836e-01
-9.31566730e-02 -8.10503483e-01 -1.64876258e+00 4.80473846e-01
6.81294322e-01 -5.71291924e-01 1.09062023e-01 6.58329546e-01
7.47101545e-01 -4.41311479e-01 -6.57863677e-01 2.21491456e-01
6.70209289e-01 -7.05655992e-01 6.18165374e-01 1.83035135e-01
-4.20187265e-01 -6.92621350e-01 3.20028782e-01 -1.09037745e+00
-1.79929271e-01 -7.12078214e-01 4.38973278e-01 1.08029795e+00
1.57902047e-01 -9.64702606e-01 8.15516472e-01 6.46295547e-01
2.16316625e-01 -5.04941821e-01 -1.20011103e+00 -6.56190157e-01
-4.28712875e-01 -4.04826730e-01 9.89341199e-01 6.69688880e-01
-1.37174487e-01 3.60618047e-02 -6.09210432e-01 7.12445021e-01
6.22258127e-01 2.56959289e-01 6.41323984e-01 -1.08665729e+00
5.29038131e-01 2.30313212e-01 -1.87181950e-01 -8.00661147e-01
-3.21523517e-01 -3.07185084e-01 7.58282244e-02 -1.29087520e+00
1.23192050e-01 -3.02385539e-01 -1.68743908e-01 3.22327584e-01
-4.25446630e-02 2.87358493e-01 3.32367346e-02 -6.33194819e-02
-3.92964751e-01 3.05710375e-01 1.09355986e+00 -1.67737588e-01
-1.59108654e-01 1.04436316e-01 -1.58215463e-01 5.86911798e-01
1.16930056e+00 -2.72468299e-01 -4.84043628e-01 -7.76005745e-01
3.77300531e-02 -1.13692522e-01 4.43294317e-01 -1.09746492e+00
3.58578086e-01 -2.85854697e-01 8.48856509e-01 -7.74644256e-01
6.23186648e-01 -1.48082280e+00 1.93175748e-01 1.01655424e+00
5.95048070e-01 2.89202899e-01 2.46295229e-01 1.78650081e-01
4.37338650e-01 2.78149307e-01 5.46516716e-01 1.22157931e-01
-1.79190606e-01 7.07260370e-01 -6.05748415e-01 -6.28764212e-01
1.52024949e+00 -2.35004202e-01 -6.70517445e-01 -1.74955040e-01
-7.12301135e-02 3.21110874e-01 5.23020327e-01 4.26961720e-01
1.05950379e+00 -1.41093194e+00 -6.68895602e-01 6.65689766e-01
5.39525226e-02 5.00806756e-02 5.90736866e-01 1.04311490e+00
-7.34992027e-01 3.93900007e-01 1.37196034e-01 -1.04812777e+00
-1.88962412e+00 6.63418233e-01 8.33282351e-01 4.81629342e-01
-2.91807979e-01 7.09922493e-01 -2.15952620e-01 -2.39764184e-01
-3.60137261e-02 -5.98132610e-01 -6.75106123e-02 -2.56270707e-01
8.29384625e-01 3.44973207e-01 6.26500174e-02 -6.58644617e-01
-8.92203033e-01 1.12722397e+00 8.64817873e-02 2.25895390e-01
8.98782313e-01 -3.48928273e-01 -3.92120957e-01 -9.91390795e-02
1.49821889e+00 -1.43745854e-01 -8.84507239e-01 1.90706670e-01
-6.73192441e-01 -6.21920645e-01 -1.35993868e-01 -4.96377945e-01
-1.38310552e+00 8.56522977e-01 1.52343464e+00 -2.28269413e-01
1.48196995e+00 -4.07424837e-01 1.02399719e+00 2.49870360e-01
5.15488565e-01 -1.14716041e+00 -3.17330897e-01 1.25708887e-02
2.92106003e-01 -1.26771045e+00 -8.77050757e-02 -4.68086571e-01
-2.61286378e-01 9.72204983e-01 9.40872729e-01 3.08477223e-01
8.67391407e-01 5.58088243e-01 6.44407272e-01 -1.50950089e-01
-2.46403977e-01 5.76244816e-02 -1.64939716e-01 7.19646454e-01
6.14529056e-03 2.79981166e-01 -5.61378859e-02 5.10624377e-03
2.68965960e-02 1.87466945e-02 1.66870415e-01 7.80544937e-01
-4.55418646e-01 -5.07930815e-01 -8.82074952e-01 4.17201579e-01
-3.49319845e-01 7.49234483e-02 -6.26978353e-02 2.87774146e-01
6.55315518e-01 1.62443018e+00 2.77660847e-01 -9.94521797e-01
-3.30147780e-02 -8.32414776e-02 2.48114079e-01 -5.47592007e-02
-5.21867752e-01 2.14489922e-01 -3.07635844e-01 -6.68638527e-01
-3.65698159e-01 -3.73175293e-01 -1.16085923e+00 -5.75203955e-01
-4.60941255e-01 8.51086453e-02 9.88657415e-01 9.97474194e-01
4.19152439e-01 1.24085322e-01 1.17530000e+00 -4.41284060e-01
3.78962068e-05 -9.16962504e-01 -6.30362988e-01 3.25657753e-03
4.82384354e-01 -2.27933273e-01 -2.66439855e-01 -2.10964352e-01] | [7.071809768676758, 0.2769826054573059] |
0ca1218a-932d-4100-8ecd-1578138e53d2 | generalized-wasserstein-dice-loss-test-time | 2112.13054 | null | https://arxiv.org/abs/2112.13054v1 | https://arxiv.org/pdf/2112.13054v1.pdf | Generalized Wasserstein Dice Loss, Test-time Augmentation, and Transformers for the BraTS 2021 challenge | Brain tumor segmentation from multiple Magnetic Resonance Imaging (MRI) modalities is a challenging task in medical image computation. The main challenges lie in the generalizability to a variety of scanners and imaging protocols. In this paper, we explore strategies to increase model robustness without increasing inference time. Towards this aim, we explore finding a robust ensemble from models trained using different losses, optimizers, and train-validation data split. Importantly, we explore the inclusion of a transformer in the bottleneck of the U-Net architecture. While we find transformer in the bottleneck performs slightly worse than the baseline U-Net in average, the generalized Wasserstein Dice loss consistently produces superior results. Further, we adopt an efficient test time augmentation strategy for faster and robust inference. Our final ensemble of seven 3D U-Nets with test-time augmentation produces an average dice score of 89.4% and an average Hausdorff 95% distance of 10.0 mm when evaluated on the BraTS 2021 testing dataset. Our code and trained models are publicly available at https://github.com/LucasFidon/TRABIT_BraTS2021. | ['Tom Vercauteren', 'Sébastien Ourselin', 'Johannes C. Paetzold', 'Ivan Ezhov', 'Suprosanna Shit', 'Lucas Fidon'] | 2021-12-24 | null | null | null | null | ['brain-tumor-segmentation'] | ['medical'] | [ 1.94087014e-01 1.21374011e-01 -8.31468925e-02 -5.31611443e-01
-1.31266713e+00 -4.43054885e-01 3.55275542e-01 1.95478454e-01
-6.63459301e-01 8.24017525e-01 8.11353847e-02 -4.89793122e-01
-2.23567382e-01 -3.77209723e-01 -6.23246908e-01 -7.92518377e-01
-2.79950172e-01 4.24222499e-01 1.32355526e-01 3.66958678e-01
1.36640579e-01 5.10711491e-01 -6.16555274e-01 5.28477430e-02
9.67318475e-01 1.17353189e+00 1.60509020e-01 7.48689055e-01
4.23066467e-01 7.83521712e-01 -2.23044455e-01 -3.63105536e-01
4.02473778e-01 -3.60915661e-01 -1.02713108e+00 -9.15597826e-02
5.63069999e-01 -5.71459591e-01 -3.40771258e-01 1.11863697e+00
8.83800030e-01 2.20383972e-01 8.27058733e-01 -1.00520301e+00
-4.91979301e-01 7.07459688e-01 -3.85087073e-01 5.35577595e-01
-2.23059580e-01 4.61501181e-01 7.10535288e-01 -5.03619015e-01
7.52480090e-01 7.94069350e-01 8.83916020e-01 6.01707399e-01
-1.36448085e+00 -5.60271740e-01 -2.19361424e-01 1.66654408e-01
-1.26902342e+00 -4.39000160e-01 2.32743472e-01 -5.50064981e-01
8.47659051e-01 1.55590713e-01 2.99775362e-01 9.88041520e-01
6.44490302e-01 7.67499030e-01 1.26685619e+00 -7.23195896e-02
4.89871092e-02 -1.38557762e-01 1.81968972e-01 8.57451618e-01
9.21269730e-02 -7.91364759e-02 1.26076147e-01 -4.37939540e-02
9.01504457e-01 -4.91173640e-02 -4.29532945e-01 -3.22291851e-01
-1.30150115e+00 7.53293514e-01 6.78049505e-01 3.53180259e-01
-3.00822228e-01 2.37355411e-01 6.13754272e-01 1.34543806e-01
6.66107357e-01 5.25636077e-01 -3.58244359e-01 -5.16068889e-03
-1.06314695e+00 6.12589754e-02 5.08395135e-01 6.39474034e-01
3.85634005e-02 -2.24655047e-01 -2.95283467e-01 9.00485992e-01
-7.08361119e-02 1.20598733e-01 5.55623889e-01 -1.23521364e+00
2.69023627e-01 1.33135468e-01 -4.07090724e-01 -5.33403814e-01
-8.06031883e-01 -8.51852357e-01 -9.35564160e-01 2.11791933e-01
6.16522908e-01 -3.65499079e-01 -1.16935289e+00 1.67017841e+00
2.08800286e-01 1.83294386e-01 -3.71462673e-01 8.45142722e-01
5.41031480e-01 -4.66886200e-02 2.19371449e-02 -8.05918053e-02
1.17522109e+00 -1.06057632e+00 -4.84030366e-01 3.79449837e-02
1.10637522e+00 -6.20324790e-01 8.64586353e-01 2.25644246e-01
-1.32470191e+00 -9.07341912e-02 -8.96167696e-01 -1.92854375e-01
-1.04607880e-01 -6.85737133e-02 4.09571409e-01 5.19585669e-01
-1.20202589e+00 9.60745156e-01 -1.25034404e+00 -9.03834701e-02
8.30928802e-01 4.63961720e-01 -5.18375993e-01 -2.16914788e-01
-8.80506814e-01 1.19429755e+00 2.15657651e-01 9.25802346e-03
-8.13569188e-01 -1.03412116e+00 -8.23046386e-01 -1.69097438e-01
1.66591793e-01 -7.87051857e-01 1.27369380e+00 -4.96525943e-01
-1.24385846e+00 8.21789086e-01 6.16468526e-02 -7.45760143e-01
1.07165003e+00 2.13039681e-01 5.44345118e-02 2.60627836e-01
1.71789885e-01 8.88644457e-01 2.00403750e-01 -8.05793583e-01
-1.83500111e-01 -6.20470166e-01 -3.79948735e-01 9.99097675e-02
-2.57539377e-02 -4.95005697e-02 -2.36670792e-01 -6.52648270e-01
1.50895491e-01 -1.02995133e+00 -4.42131877e-01 1.03669539e-01
-4.10749376e-01 1.04501784e-01 2.89920390e-01 -1.06545091e+00
8.87358963e-01 -1.74493480e+00 -2.92002913e-02 2.46413484e-01
4.26339984e-01 -3.92867904e-03 -9.46706608e-02 -3.53983343e-01
-1.56078026e-01 1.86866954e-01 -7.42310643e-01 -3.48908037e-01
-2.46494204e-01 -7.97001645e-02 4.20167983e-01 7.15418875e-01
1.26830921e-01 1.02636027e+00 -8.31002235e-01 -6.19195163e-01
8.05480927e-02 5.04198968e-01 -7.62851477e-01 -2.62874693e-01
2.18227133e-01 7.85092294e-01 -2.72240162e-01 7.23639309e-01
6.00837350e-01 -4.21179712e-01 1.29936375e-02 -1.90303162e-01
3.80079225e-02 2.43080318e-01 -6.72858357e-01 1.82309079e+00
-3.62085313e-01 5.74450850e-01 1.61988169e-01 -1.03880596e+00
5.36463022e-01 3.14559042e-01 8.77379179e-01 -7.04938591e-01
3.41205478e-01 4.92586613e-01 5.42425632e-01 -4.23089236e-01
-1.07897408e-01 -3.06705534e-01 3.54599983e-01 4.88142967e-01
1.92620412e-01 -2.39627719e-01 2.57228881e-01 -9.06874612e-02
1.47446406e+00 -6.69531301e-02 4.23049144e-02 -5.96850932e-01
1.94285214e-01 -1.28864214e-01 4.89731044e-01 7.18844771e-01
-7.97937453e-01 8.09241474e-01 5.65704226e-01 -2.65352368e-01
-1.21560442e+00 -1.17243171e+00 -8.69266212e-01 5.20278215e-01
-4.07662779e-01 8.01706091e-02 -8.93195868e-01 -8.77148628e-01
-4.48590778e-02 6.84383392e-01 -7.77033925e-01 9.15880501e-03
-4.83953029e-01 -1.10158312e+00 7.33207941e-01 6.33980274e-01
3.93074870e-01 -7.09399760e-01 -5.22029996e-01 7.77923837e-02
-2.43591204e-01 -1.06869888e+00 -6.89756274e-01 4.16098535e-01
-1.23981464e+00 -1.01396680e+00 -1.09435332e+00 -6.54552519e-01
7.75802195e-01 -3.12159985e-01 1.02257979e+00 -1.55788735e-01
-4.26432699e-01 9.87911448e-02 -6.14531673e-02 -2.92835683e-01
-2.60191143e-01 3.28254610e-01 -6.15808405e-02 -5.55643678e-01
3.87385637e-02 -3.76672745e-01 -8.70174825e-01 3.67830217e-01
-9.31268990e-01 3.89076658e-02 5.25405526e-01 1.03348958e+00
9.11657691e-01 -3.07967931e-01 4.37618881e-01 -8.21782947e-01
4.67969388e-01 -4.83069658e-01 -3.51992548e-01 2.38452062e-01
-6.80065572e-01 8.34000707e-02 4.33730453e-01 -3.18623781e-01
-6.49498463e-01 -1.53383628e-01 -3.78500670e-01 -4.15290087e-01
3.26821208e-02 4.08767432e-01 3.67669582e-01 -2.66055793e-01
6.73409224e-01 -7.89677203e-02 4.94815260e-01 -1.74119577e-01
1.07962533e-03 3.44267488e-01 3.29615593e-01 -4.58687127e-01
2.87383854e-01 5.50527632e-01 1.61769629e-01 -5.65127134e-01
-7.85414040e-01 -1.88453779e-01 -7.18990862e-01 -1.92455441e-01
1.12874603e+00 -5.65875292e-01 -3.65431249e-01 4.38559622e-01
-8.64971220e-01 -8.73721719e-01 -2.21720278e-01 8.25697660e-01
-7.00760067e-01 1.58768192e-01 -8.74282122e-01 -2.47089699e-01
-6.53675020e-01 -1.81400883e+00 7.94258177e-01 -5.33174388e-02
-2.33037278e-01 -1.23073113e+00 -1.68152358e-02 5.49399197e-01
5.38366258e-01 4.96163428e-01 7.58947849e-01 -8.87816966e-01
-3.58910888e-01 -1.61930487e-01 -3.37740481e-01 6.39252782e-01
2.49780893e-01 -2.87588120e-01 -7.74602890e-01 -3.30419987e-01
1.07713781e-01 -3.60034794e-01 9.27389979e-01 1.11441314e+00
1.72262537e+00 -8.69195983e-02 -1.25016466e-01 9.37693298e-01
1.24059641e+00 2.82290637e-01 4.57736254e-01 3.58327329e-01
5.55904448e-01 4.03947175e-01 7.93197751e-02 2.64953915e-02
2.52087742e-01 3.95906985e-01 2.79633492e-01 -1.30366422e-02
-2.59312451e-01 2.55729616e-01 9.87162814e-02 9.63263214e-01
-6.55334210e-03 4.08305191e-02 -1.31459951e+00 5.60332596e-01
-1.47916269e+00 -7.84088135e-01 1.16971105e-01 2.18458009e+00
9.20881391e-01 2.38720492e-01 7.00504258e-02 -1.72352508e-01
6.14548266e-01 -1.43915489e-01 -7.92582273e-01 -2.02814966e-01
4.64051515e-02 2.11292550e-01 9.32951927e-01 7.74071634e-01
-1.02084184e+00 4.86017406e-01 6.22178221e+00 7.49541819e-01
-1.20981944e+00 4.11769241e-01 1.39995015e+00 -4.37551528e-01
-2.09023148e-01 -3.04564387e-01 -3.53742391e-01 4.59913522e-01
9.98662829e-01 -4.97980043e-02 3.24808836e-01 4.42928165e-01
2.80342668e-01 -1.05792701e-01 -9.94929075e-01 7.17077255e-01
6.79127648e-02 -1.27566969e+00 -3.57006520e-01 2.91965693e-01
8.15526247e-01 7.30677605e-01 1.97857216e-01 -2.03748397e-03
1.69538051e-01 -1.27497399e+00 4.76878107e-01 5.51543176e-01
9.98218119e-01 -5.91020346e-01 9.67625320e-01 5.77185163e-03
-6.13482416e-01 1.73344925e-01 -1.46128133e-01 3.67906451e-01
1.81582212e-01 7.04991579e-01 -1.03854823e+00 3.91793251e-01
7.15900183e-01 5.29348016e-01 -5.23492157e-01 1.33722043e+00
2.20594287e-01 5.32249510e-01 -4.14906263e-01 2.13390648e-01
4.08760726e-01 -3.69744688e-01 4.76896077e-01 1.17467630e+00
3.44388217e-01 5.78999668e-02 -3.78816761e-02 7.10311174e-01
-2.58747607e-01 1.49821371e-01 -5.32266200e-01 2.32669964e-01
1.19248442e-01 1.26619470e+00 -1.00715888e+00 -2.13837177e-01
-2.54263192e-01 6.95533574e-01 2.74364680e-01 1.28659636e-01
-8.79941344e-01 -2.78368801e-01 3.79925042e-01 2.03817278e-01
-8.27773958e-02 -6.68554455e-02 -6.90491140e-01 -1.05665648e+00
3.80181223e-02 -7.94437766e-01 4.05330628e-01 -5.70985794e-01
-1.36345172e+00 6.51538968e-01 -3.52785997e-02 -7.96611726e-01
-1.32321060e-01 -8.94306302e-01 -6.50472939e-01 8.89871418e-01
-1.29323959e+00 -8.04527402e-01 -7.96348602e-02 4.02123392e-01
3.37503344e-01 1.36051238e-01 5.40656328e-01 4.15450126e-01
-6.69934809e-01 7.33795106e-01 3.52625191e-01 3.52050543e-01
7.14837134e-01 -1.33537149e+00 1.89750716e-01 7.98642159e-01
-3.02816093e-01 6.53566182e-01 5.85005045e-01 -5.40910542e-01
-8.90916824e-01 -9.72360313e-01 5.64285398e-01 -4.93306458e-01
9.39342678e-01 8.69398490e-02 -8.84789169e-01 9.80868936e-01
1.50466740e-01 2.61823505e-01 7.79987395e-01 -7.19796121e-02
-3.17286029e-02 9.47726220e-02 -1.48637140e+00 6.63434625e-01
8.77198100e-01 -3.02523196e-01 -2.77562290e-01 5.24980307e-01
3.91805351e-01 -6.03750348e-01 -1.47778821e+00 6.98871613e-01
6.45344257e-01 -7.80330777e-01 7.68596947e-01 -5.26556373e-01
6.47966266e-01 1.19059935e-01 -3.46955433e-02 -1.45508075e+00
-2.42574140e-01 -2.36152411e-01 3.07112515e-01 6.06019735e-01
7.23108470e-01 -7.81923115e-01 6.82700157e-01 7.80311644e-01
-3.77582699e-01 -1.27622342e+00 -1.12699914e+00 -6.73898518e-01
8.32182527e-01 -5.60212135e-01 1.47338465e-01 8.79696786e-01
-3.96981947e-02 -1.03676282e-01 1.21703446e-01 -3.08755994e-01
9.22053635e-01 -4.59612340e-01 1.80064306e-01 -9.48991835e-01
-2.24846378e-01 -8.57772350e-01 -3.24654847e-01 -5.43780208e-01
1.56212837e-01 -1.40002739e+00 -5.95787354e-02 -1.65694284e+00
5.22822082e-01 -4.45970863e-01 -5.00740170e-01 4.88917232e-01
-1.37301274e-02 2.18630433e-01 5.45497760e-02 1.10024504e-01
-3.47973436e-01 4.20227945e-01 1.66703999e+00 -2.13650763e-01
-3.42693529e-03 -5.32263070e-02 -5.46931803e-01 6.03480399e-01
1.24388993e+00 -4.24447328e-01 -2.79763222e-01 -6.98451936e-01
-3.03772628e-01 6.18683696e-02 4.20849234e-01 -1.01950943e+00
2.55152553e-01 2.60346144e-01 5.64180136e-01 -2.69601226e-01
9.49124172e-02 -4.70535964e-01 -9.68758240e-02 8.23999047e-01
-5.09492695e-01 3.33548784e-01 3.59815568e-01 1.03585562e-02
1.01798072e-01 -3.19156915e-01 1.21330595e+00 -2.27393344e-01
7.53970668e-02 7.02931285e-01 -9.87990573e-02 2.13869661e-01
8.91782880e-01 -9.98948812e-02 -2.87547737e-01 -5.28024621e-02
-8.85777593e-01 3.71708184e-01 3.89567584e-01 -3.28835323e-02
3.88198972e-01 -1.24480915e+00 -8.96274865e-01 -1.30210772e-01
-1.70191497e-01 2.27855351e-02 4.20503408e-01 1.63652825e+00
-7.93181062e-01 4.59414363e-01 -4.05935705e-01 -6.32669032e-01
-1.00643909e+00 2.41292696e-02 8.75321090e-01 -5.82838655e-01
-6.63193464e-01 8.92397106e-01 8.14039707e-02 -7.03429937e-01
8.49669874e-02 -4.76452857e-01 3.03520173e-01 -1.04662791e-01
3.60656857e-01 6.02241576e-01 3.06106299e-01 -4.45510924e-01
-4.44636911e-01 2.31046662e-01 -3.25561315e-01 -2.84926325e-01
1.45445788e+00 1.40866101e-01 -1.52522922e-01 2.19850093e-01
1.61013091e+00 -4.66141552e-01 -1.29286587e+00 -1.96078103e-02
3.10201403e-02 -5.89286163e-02 4.51380908e-01 -9.23980594e-01
-1.36761284e+00 6.89205527e-01 8.48278463e-01 -1.07774381e-02
8.65141213e-01 -3.28482203e-02 7.57962167e-01 1.55113786e-01
5.45645691e-02 -1.10170138e+00 -2.67450184e-01 4.97775704e-01
7.19636858e-01 -1.54028583e+00 -1.88813508e-02 9.75019410e-02
-5.70716083e-01 9.06767547e-01 4.22391564e-01 -1.26745030e-01
7.66312182e-01 4.67502087e-01 1.10907875e-01 -2.44622633e-01
-5.14084280e-01 -9.33522242e-04 2.73764372e-01 3.65825832e-01
7.56881773e-01 1.14335008e-01 -4.87804800e-01 2.32543588e-01
-3.18137020e-01 9.95798185e-02 4.05118793e-01 6.60712838e-01
-1.70563444e-01 -7.90066004e-01 -1.42906189e-01 1.04873860e+00
-1.00241411e+00 -1.74069688e-01 1.06111452e-01 6.97083890e-01
-1.80453017e-01 5.24393559e-01 8.57259482e-02 -9.38170925e-02
8.12103599e-03 1.47069842e-01 7.26180553e-01 -3.02304476e-01
-4.54876781e-01 -1.70178544e-02 -2.47077808e-01 -6.14454806e-01
-2.27756813e-01 -1.01669407e+00 -1.17570782e+00 -3.51523399e-01
-2.72017688e-01 -9.93499607e-02 7.56021798e-01 8.57373595e-01
1.70760408e-01 8.07036042e-01 2.42746145e-01 -6.71506941e-01
-6.93328202e-01 -1.06647730e+00 -4.18913543e-01 2.74143219e-01
3.55077296e-01 -5.49124181e-01 -4.29917723e-01 -2.53961086e-01] | [14.421968460083008, -2.3668324947357178] |
d45bee1d-775f-4406-b02b-cec9ca4bc8f1 | multi-kernel-filtering-an-extension-of | 1908.06307 | null | https://arxiv.org/abs/1908.06307v4 | https://arxiv.org/pdf/1908.06307v4.pdf | Multi-Kernel Filtering for Nonstationary Noise: An Extension of Bilateral Filtering Using Image Context | Bilateral filtering (BF) is one of the most classical denoising filters, however, the manually initialized filtering kernel hampers its adaptivity across images with various characteristics. To deal with image variation (i.e., non-stationary noise), in this paper, we propose multi-kernel filter (MKF) which adapts filtering kernels to specific image characteristics automatically. The design of MKF takes inspiration from adaptive mechanisms of human vision that make full use of information in a visual context. More specifically, for simulating the visual context and its adaptive function, we construct the image context based on which we simulate the contextual impact on filtering kernels. We first design a hierarchically clustering algorithm to generate a hierarchy of large to small coherent image patches, organized as a cluster tree, so that obtain multi-scale image representation. The leaf cluster and corresponding predecessor clusters are used to generate one of multiple range kernels that are capable of catering to image variation. At first, we design a hierarchically clustering framework to generate a hierarchy of large to small coherent image patches that organized as a cluster tree, so that obtain multi-scale image representation, i.e., the image context. Next, a leaf cluster is used to generate one of the multiple kernels, and two corresponding predecessor clusters are used to fine-tune the adopted kernel. Ultimately, the single spatially-invariant kernel in BF becomes multiple spatially-varying ones. We evaluate MKF on two public datasets, BSD300 and BrainWeb which are added integrally-varying noise and spatially-varying noise, respectively. Extensive experiments show that MKF outperforms state-of-the-art filters w.r.t. both mean absolute error and structural similarity. | ['Pew-Thian Yap', 'Jun Feng', 'Dinggang Shen', 'Feihong Liu'] | 2019-08-17 | null | null | null | null | ['image-variation'] | ['computer-vision'] | [ 8.72528479e-02 -5.97606778e-01 4.57182288e-01 -3.32690030e-01
-4.39364046e-01 -5.69645762e-01 2.94415325e-01 -2.61606257e-02
-4.63945210e-01 4.28963065e-01 1.26479819e-01 1.66995093e-01
-4.10715580e-01 -8.24957967e-01 -6.20533109e-01 -1.13818729e+00
1.10050291e-01 -3.19231063e-01 8.11717451e-01 -3.70692760e-02
3.39278549e-01 4.75593597e-01 -1.66367817e+00 4.43136036e-01
1.19436491e+00 9.11690593e-01 7.10686505e-01 4.34175879e-01
-4.69976254e-02 1.12444088e-01 -5.16240597e-01 -8.28518867e-02
1.98112041e-01 -5.63580096e-01 -3.34123790e-01 3.76469344e-01
3.18516999e-01 1.48776457e-01 6.52289614e-02 1.40620065e+00
6.32600546e-01 2.56828725e-01 7.38576770e-01 -9.18557227e-01
-7.85645485e-01 3.74188602e-01 -7.00404763e-01 4.77003932e-01
-1.24918513e-01 1.45179376e-01 4.92310286e-01 -9.00355756e-01
5.40305853e-01 1.24582195e+00 4.48481113e-01 3.39279711e-01
-1.56184816e+00 -5.11654437e-01 4.54044938e-01 2.30940357e-01
-1.51900327e+00 -2.37345830e-01 8.74948680e-01 -4.48168993e-01
3.31701159e-01 3.99500996e-01 7.11052179e-01 7.17486680e-01
3.77245694e-01 4.79582936e-01 1.45386028e+00 -2.39229321e-01
4.11905795e-01 5.44951111e-02 1.32354036e-01 2.44236901e-01
1.56368524e-01 -2.10997567e-01 -2.44836614e-01 -1.01717554e-01
1.04570651e+00 -1.22260593e-01 -6.72481120e-01 -3.31746608e-01
-1.10571647e+00 5.72498858e-01 5.93860447e-01 6.95303440e-01
-5.14541388e-01 -1.14890933e-01 2.11270109e-01 2.04848841e-01
1.60739392e-01 -3.10536921e-02 -3.80535096e-01 2.86731422e-01
-8.55537474e-01 1.18564673e-01 2.83996284e-01 6.74345434e-01
9.29049969e-01 -7.29641765e-02 -5.57419956e-01 1.15720057e+00
1.53933421e-01 3.11012179e-01 6.61969960e-01 -8.07389379e-01
6.51997328e-02 6.30592585e-01 -2.47381348e-02 -1.22262359e+00
-4.36269671e-01 -5.68909466e-01 -1.26111245e+00 3.05333078e-01
2.65220731e-01 7.83213228e-02 -9.87822711e-01 1.80176377e+00
4.82489526e-01 3.64845097e-01 -6.16824105e-02 9.91995513e-01
7.37638772e-01 5.18589735e-01 -5.56902215e-02 -5.13359725e-01
1.58150876e+00 -5.96746922e-01 -7.57437825e-01 -4.43875268e-02
5.36058918e-02 -8.13753664e-01 1.16792381e+00 5.11223674e-01
-1.04838061e+00 -8.31710041e-01 -7.36846030e-01 4.17998731e-01
-3.33027750e-01 2.52781332e-01 1.31562009e-01 4.06388909e-01
-1.10369658e+00 4.68733639e-01 -6.38040602e-01 -2.84648210e-01
1.90018624e-01 9.18066353e-02 -4.10669804e-01 1.11846507e-01
-1.06623828e+00 5.59523225e-01 5.54264903e-01 3.26863021e-01
-5.30415833e-01 -5.82754612e-01 -5.93882382e-01 -1.12941777e-02
2.47034177e-01 -6.67758524e-01 5.71851492e-01 -1.26107633e+00
-1.35545564e+00 6.79461598e-01 -1.58279881e-01 -1.86106771e-01
5.01672566e-01 1.15797065e-01 -6.06738031e-01 2.67573178e-01
1.08087488e-01 5.06920159e-01 1.40163827e+00 -1.55079186e+00
-6.01420045e-01 -2.92124122e-01 -1.91434428e-01 2.44184226e-01
-4.88319874e-01 2.29884274e-02 -8.60729218e-01 -9.67119694e-01
2.79566407e-01 -6.66363955e-01 -2.19017342e-01 -2.26568297e-01
-3.41464877e-01 -1.30720390e-02 7.58844435e-01 -6.22598886e-01
1.68846381e+00 -2.38314795e+00 4.19948339e-01 5.83365977e-01
1.52426943e-01 1.50193021e-01 -1.19971924e-01 3.52560654e-02
-1.87675238e-01 3.57968882e-02 -5.20358324e-01 1.50721818e-01
-2.49846399e-01 2.00697944e-01 1.09812312e-01 3.22426558e-01
7.81597421e-02 5.44935942e-01 -7.12412238e-01 -5.59413552e-01
2.26937503e-01 5.52606821e-01 -5.87971807e-01 1.42089888e-01
1.03425592e-01 6.06306791e-01 -5.36028802e-01 3.46065551e-01
1.13259554e+00 -9.18410197e-02 -1.45904109e-01 -9.04818594e-01
-2.85553515e-01 -6.74763501e-01 -1.73417902e+00 1.38472903e+00
-3.47979665e-01 1.07964166e-01 4.04147714e-01 -9.29275095e-01
1.04072714e+00 7.13600069e-02 3.87413114e-01 -4.36610669e-01
4.09832150e-02 1.59172043e-01 1.34268418e-01 -4.54614043e-01
-3.28738764e-02 3.93497720e-02 2.60178834e-01 1.22233294e-02
-5.06247133e-02 1.74850374e-02 1.97022095e-01 3.85500044e-02
1.00551116e+00 -1.10243529e-01 2.77732432e-01 -6.15729511e-01
1.07468247e+00 -4.23667938e-01 7.59730756e-01 8.05127800e-01
-2.31421381e-01 9.41015363e-01 1.59941018e-01 -4.20019060e-01
-7.70912468e-01 -1.38030481e+00 -4.29852247e-01 9.23214436e-01
4.41413134e-01 -2.34819442e-01 -1.02542329e+00 -3.57253522e-01
-7.53341988e-02 3.20443660e-01 -7.57894158e-01 -2.45721325e-01
-5.31855702e-01 -9.67624724e-01 1.47806898e-01 2.83809721e-01
1.09419775e+00 -1.19082189e+00 -6.20552480e-01 2.90359318e-01
-7.32368827e-02 -7.50896454e-01 -7.81134605e-01 1.86895747e-02
-6.57717764e-01 -9.29620981e-01 -7.93298066e-01 -8.66558611e-01
7.76241660e-01 3.54530662e-01 9.19916153e-01 1.83354840e-01
-3.94787461e-01 4.39764827e-01 -3.63184303e-01 1.09967478e-01
-2.02279061e-01 -3.01670223e-01 1.11365253e-02 6.32865489e-01
3.17496643e-03 -8.23735774e-01 -1.00020444e+00 6.33488834e-01
-1.40900862e+00 -7.48592988e-02 7.89458513e-01 9.22215283e-01
9.07979071e-01 5.17000914e-01 5.54957151e-01 -6.61612988e-01
9.00108576e-01 -2.91048318e-01 -5.43935180e-01 3.78283978e-01
-2.31887862e-01 -9.98109728e-02 6.85145676e-01 -9.01480615e-01
-1.29866195e+00 1.64131090e-01 9.33654010e-02 -4.33055997e-01
-3.25883269e-01 4.48261887e-01 -5.05437732e-01 -1.04397066e-01
8.40934277e-01 5.34226239e-01 -2.45219171e-01 -4.35042799e-01
5.42792201e-01 6.53462768e-01 6.82204485e-01 -6.20540738e-01
9.64626431e-01 4.79756713e-01 -1.82478189e-01 -8.38467658e-01
-4.20129508e-01 -3.65571439e-01 -7.52365530e-01 -3.22077453e-01
1.07221222e+00 -6.82727098e-01 -5.19390106e-01 8.11962128e-01
-9.52899933e-01 -2.21573293e-01 -5.48221767e-02 3.15032840e-01
-4.06778246e-01 3.10815454e-01 -6.04194224e-01 -5.18025994e-01
-2.99948454e-01 -1.23183370e+00 9.29768801e-01 5.59452832e-01
9.96679515e-02 -8.20640445e-01 -1.37120426e-01 -6.12274334e-02
4.81397808e-01 2.27058366e-01 8.27343345e-01 -2.15901613e-01
-4.59736496e-01 7.93703943e-02 -3.83673251e-01 5.47988534e-01
3.94645721e-01 2.97362041e-02 -8.30832005e-01 -4.51229036e-01
1.70120016e-01 3.06748360e-01 7.98829615e-01 6.70053542e-01
1.39764595e+00 -9.53401551e-02 -3.20507765e-01 7.50070930e-01
1.37228775e+00 3.58371109e-01 8.19590688e-01 5.24827302e-01
5.71263909e-01 5.40216506e-01 4.20082390e-01 2.90033251e-01
1.64587200e-02 7.17985451e-01 1.79683372e-01 -2.11548299e-01
-6.24603331e-02 1.28945664e-01 8.58073235e-02 6.54878855e-01
-1.51280925e-01 1.65803760e-01 -4.67367083e-01 4.17952508e-01
-1.81098008e+00 -9.30601597e-01 -5.91281988e-02 2.28310275e+00
9.09877002e-01 8.08549672e-02 4.44662720e-02 -1.41331088e-02
1.13356507e+00 -2.32646652e-02 -5.89686334e-01 -7.68766329e-02
-3.41943115e-01 9.04599652e-02 2.54656971e-01 2.30474085e-01
-1.21903861e+00 7.91652501e-01 4.85520411e+00 1.28972435e+00
-1.20328355e+00 -1.23843402e-02 5.15417039e-01 1.97972938e-01
-3.06859046e-01 -9.12253559e-02 -4.87081409e-01 8.09855580e-01
2.66526848e-01 -1.98270008e-01 5.62940240e-01 5.42089939e-01
4.67792302e-01 -1.14254460e-01 -6.50845766e-01 1.06809402e+00
-1.83069468e-01 -1.01550639e+00 1.95802927e-01 -1.29811972e-01
7.81574905e-01 -4.38350052e-01 5.27277030e-02 -9.05500725e-02
3.44951786e-02 -7.76729524e-01 7.50162303e-01 7.27077305e-01
4.43367928e-01 -6.85831428e-01 5.74801266e-01 3.05796295e-01
-1.54612315e+00 -1.71971723e-01 -5.79468369e-01 5.55879891e-01
5.18837534e-02 8.78296673e-01 1.05654791e-01 6.08331382e-01
1.19409680e+00 5.78642190e-01 -6.87212110e-01 1.45006657e+00
-1.24168769e-01 4.70095545e-01 -2.17655838e-01 3.17866027e-01
1.28524415e-02 -5.84357560e-01 7.13907063e-01 1.36401355e+00
3.88912737e-01 2.04466820e-01 2.43609220e-01 9.23245192e-01
1.95897564e-01 3.33988667e-01 -2.89871842e-01 6.88529909e-01
6.72983885e-01 1.51655924e+00 -9.40515518e-01 -3.75560224e-01
-3.65957439e-01 1.30196750e+00 1.81630567e-01 7.42764354e-01
-7.56160736e-01 -4.89143342e-01 6.05728149e-01 -1.79203581e-02
5.40753663e-01 1.97687447e-02 -1.25272229e-01 -1.09266126e+00
2.44024381e-01 -8.53715479e-01 4.55926090e-01 -8.04693222e-01
-1.58676302e+00 7.64984488e-01 1.10691652e-01 -1.38861036e+00
2.88944542e-01 -3.45977813e-01 -1.01716697e+00 1.11173189e+00
-1.23651612e+00 -1.10100114e+00 -7.30484605e-01 9.36581314e-01
4.06261981e-01 5.96444942e-02 4.60292757e-01 3.51862013e-01
-6.46648586e-01 4.69734013e-01 1.40913546e-01 -7.00472221e-02
8.65107358e-01 -1.12210405e+00 -7.72342384e-02 9.75070775e-01
-1.41752183e-01 8.84927034e-01 5.83111465e-01 -6.85730696e-01
-9.85951841e-01 -1.16413081e+00 1.05188809e-01 4.39176634e-02
6.07461989e-01 -2.53155887e-01 -1.43056512e+00 3.37503701e-01
1.54475989e-02 2.42881328e-01 2.99369842e-01 -4.68453377e-01
-2.53458977e-01 -4.30870086e-01 -1.30368376e+00 7.68853843e-01
9.29971695e-01 -2.97798455e-01 -5.17628610e-01 1.67851746e-02
5.21136820e-01 -4.58033115e-01 -9.25292969e-01 5.55078387e-01
4.40689743e-01 -1.23797834e+00 1.13125491e+00 -9.52713117e-02
1.87910963e-02 -9.53259051e-01 -3.80140133e-02 -1.48205757e+00
-8.34767938e-01 -4.75164503e-01 4.22984302e-01 1.47562742e+00
8.19987729e-02 -7.70760417e-01 3.62136722e-01 3.38380486e-01
-2.12742597e-01 -6.58866465e-01 -9.34318423e-01 -8.04396808e-01
2.21487936e-02 -1.40598297e-01 5.22407055e-01 8.46569061e-01
-5.24285793e-01 -2.85059810e-02 -6.00523539e-02 4.32694227e-01
7.03862131e-01 2.03031525e-01 5.95693588e-01 -1.04322970e+00
-2.51017809e-01 -5.79251409e-01 -4.78966415e-01 -8.95252049e-01
-1.54208675e-01 -5.78404546e-01 6.04149103e-02 -1.39523065e+00
1.71456411e-01 -3.59405041e-01 -4.58756477e-01 1.67865336e-01
-5.38468122e-01 4.14586931e-01 7.62359798e-02 3.15185934e-01
-2.97029495e-01 4.55244571e-01 1.40093946e+00 -1.01726159e-01
-5.92753947e-01 5.35145476e-02 -6.72915757e-01 8.18996072e-01
7.31569827e-01 -8.75624269e-02 -4.48088378e-01 -2.63491124e-01
-2.61331469e-01 -3.46623629e-01 4.49593484e-01 -1.01670229e+00
1.97084218e-01 -1.66939139e-01 6.91023231e-01 -4.23375130e-01
-2.38191318e-02 -9.01241124e-01 4.64007944e-01 3.74117941e-01
-6.05419017e-02 -3.46047357e-02 2.42020100e-01 6.00472271e-01
-3.47535819e-01 -1.31484583e-01 1.20025492e+00 -2.58582771e-01
-8.09662104e-01 2.88858354e-01 -4.80293751e-01 1.05480766e-02
1.10057724e+00 -4.30510759e-01 -1.47276014e-01 -1.27790242e-01
-9.61452067e-01 1.71642229e-01 4.75472689e-01 3.45886320e-01
7.21008360e-01 -1.53233755e+00 -6.47289038e-01 3.38268489e-01
1.12983793e-01 -2.90536378e-02 6.78688884e-01 8.94565463e-01
-2.01083079e-01 -2.95344085e-01 -2.73874611e-01 -8.03845167e-01
-1.30150330e+00 6.11625373e-01 5.51952541e-01 8.23314413e-02
-7.86607802e-01 7.35681534e-01 6.40235007e-01 -1.11752309e-01
3.01384814e-02 -4.76212263e-01 -4.74554777e-01 8.66882950e-02
6.26486957e-01 1.57122225e-01 2.27343477e-02 -6.62937164e-01
-2.89528817e-01 1.22738457e+00 1.31896704e-01 7.18684420e-02
1.09216690e+00 -3.94740999e-01 -4.65513289e-01 2.02298999e-01
1.01908827e+00 1.47425756e-01 -1.30662870e+00 -2.76521802e-01
-1.46189049e-01 -5.48477113e-01 2.42072754e-02 -6.59292877e-01
-1.26577938e+00 5.58659434e-01 9.85725880e-01 2.76209742e-01
1.79899848e+00 -8.95319581e-02 4.83549893e-01 -5.48180118e-02
2.41912186e-01 -9.84309375e-01 1.86626896e-01 2.31743351e-01
9.61748838e-01 -9.18844163e-01 -1.96593314e-01 -5.79932094e-01
-6.24373436e-01 1.17831469e+00 8.29434454e-01 -1.53698534e-01
8.33898842e-01 2.76042342e-01 1.41502813e-01 -1.28618509e-01
-2.26305038e-01 -3.07196468e-01 6.55786335e-01 7.37518191e-01
1.80502445e-01 -4.87712026e-03 -4.03155893e-01 8.21023226e-01
-3.87120880e-02 -8.52385983e-02 8.73424485e-02 5.18941104e-01
-6.84005260e-01 -9.72606301e-01 -7.31747448e-01 4.69942510e-01
-1.63930446e-01 -1.09400369e-01 -3.96962650e-02 4.86522436e-01
6.43943071e-01 9.16359544e-01 5.19459061e-02 -3.84306192e-01
6.20119333e-01 -2.68160433e-01 2.61645079e-01 -3.51470739e-01
-6.90823615e-01 3.95147324e-01 -5.18880427e-01 -4.99013811e-01
-4.17132348e-01 -4.73813623e-01 -1.21892440e+00 7.48347789e-02
-2.65277326e-01 1.39398739e-01 1.76984593e-01 7.10157514e-01
2.30983272e-01 5.90702772e-01 6.00493431e-01 -9.32914317e-01
-2.09456071e-01 -8.81138265e-01 -7.32320428e-01 7.99066961e-01
1.89024001e-01 -7.57010579e-01 -5.12474954e-01 2.36248419e-01] | [11.058582305908203, -1.4229164123535156] |
7e2833a1-1ff0-41d8-830d-384badacfd49 | real-time-p-qrs-and-t-wave-detection-by-qrs | null | null | https://www.google.com/url?sa=t&source=web&rct=j&url=http://www.ijieee.org.in/volume.php%3Fvolume_id%3D482&ved=2ahUKEwiGmLiWqvvsAhUQO3AKHYGnBGgQFjACegQIAhAB&usg=AOvVaw0g8O8vZTRpQ6bTmjcUhERw | https://www.google.com/url?sa=t&source=web&rct=j&url=https://www.digitalxplore.org/up_proc/pdf/371-15299043251-5.pdf&ved=2ahUKEwj95ZzPqfvsAhUZZ94KHdrjBf0QFjAAegQIBhAB&usg=AOvVaw073g7FKNIvvTvWjz_vgO01&cshid=1605126531345 | Real time P, QRS and T wave detection by QRS matched filter method | ECG signals have always been a key concern for heart disease analysis and heart monitoring system. That’s why
for a very long time, people are to find newer and easier processes for retrieving information from ECG. And still now some
of the methods are quite accurate and implemented successfully all over the world, but still now many people are trying to
find a more robust and simple method. My intention was to derive a simpler real time detection of P, QRS and T waves. For
this purpose, I developed such a model which uses mostly peak thresholding approaches for detecting these waves. But the
novelty of this model is that in almost all approaches people tried to filter ECG signals to avoid baseline shifts, removal of
noises. I tried to find such a filter which can make QRS peaks completely visible along with removing noises and baseline
shifts. I used to address this filter as QRS matched filter. Using this filter QRS peaks were successfully captured and then
after further processing like cross correlations, P and T waves were also successfully detected. This model was tested on
both AVEC 2016 ECG signal database for emotion recognition and MIT-BIH Arrhythmia database. | ['Abdullah Al Masud'] | 2018-07-01 | null | null | null | null | ['electrocardiography-ecg'] | ['methodology'] | [ 1.37076631e-01 -1.63239524e-01 5.75599730e-01 -3.28161389e-01
-4.66238171e-01 -5.57205677e-01 -1.91706121e-01 5.31584382e-01
-2.03426316e-01 8.04883242e-01 9.90362093e-03 8.04483891e-02
-3.33152950e-01 -4.87727702e-01 1.17917359e-01 -5.16495109e-01
-1.91818312e-01 4.06454951e-02 1.90825850e-01 -2.92302817e-01
4.57882464e-01 4.12942231e-01 -1.35757017e+00 5.88838346e-02
6.91539884e-01 5.70014536e-01 -5.64998591e-05 9.33103144e-01
4.17831242e-02 3.57671201e-01 -8.25735271e-01 -6.50227368e-02
4.92345877e-02 -8.08022380e-01 -6.33395731e-01 -1.54752254e-01
-3.89665008e-01 8.43530148e-02 2.57579625e-01 8.83774698e-01
1.08420694e+00 -1.05591957e-02 6.56671107e-01 -1.08004963e+00
1.72459573e-01 5.11884868e-01 -7.98715472e-01 5.47913015e-01
5.29633820e-01 -2.28921562e-01 4.65171427e-01 -6.43186986e-01
5.72913706e-01 6.02333128e-01 1.11443818e+00 1.04050353e-01
-1.01802862e+00 -5.83032668e-01 -6.00248456e-01 5.16728401e-01
-1.58057320e+00 -8.81760940e-02 9.67858672e-01 -1.25317514e-01
1.18737686e+00 6.27220869e-01 8.69025946e-01 4.96822059e-01
6.03278458e-01 2.82608628e-01 1.25863802e+00 -5.70442021e-01
-1.02671191e-01 5.02487600e-01 5.00943184e-01 3.59123617e-01
1.15302570e-01 -3.63248557e-01 -2.59006560e-01 -1.99855179e-01
3.35049659e-01 1.02377445e-01 -4.67104316e-01 2.82727391e-01
-1.11127138e+00 4.70127612e-01 -2.15087861e-01 8.81114483e-01
-6.67086601e-01 -2.47453973e-01 5.85505188e-01 4.96659517e-01
-2.06899606e-02 3.48033994e-01 -4.42924440e-01 -4.86900419e-01
-1.06698859e+00 2.07537368e-01 8.64550233e-01 4.19253916e-01
4.56536055e-01 1.24667026e-01 -1.67935684e-01 7.97168791e-01
1.97285131e-01 6.03954911e-01 6.71719670e-01 -5.45950472e-01
-5.20305745e-02 6.67319059e-01 -1.24902055e-01 -1.35880101e+00
-7.52344191e-01 -6.96308136e-01 -8.36974561e-01 1.54154032e-01
3.27383339e-01 -3.40204358e-01 -7.18920171e-01 1.30823767e+00
2.65011400e-01 1.23400100e-01 -7.01244644e-05 9.16334212e-01
7.52173960e-01 8.35246146e-01 -3.64860222e-02 -6.75148547e-01
1.62522590e+00 -1.07183516e-01 -1.22269452e+00 2.23002926e-01
3.02661955e-01 -1.20168674e+00 5.75951397e-01 9.31163907e-01
-9.27047908e-01 -6.46809101e-01 -1.18782127e+00 3.18585515e-01
-2.07850337e-01 -4.97621903e-03 2.67497569e-01 9.84820187e-01
-9.07442153e-01 8.79559696e-01 -5.58522880e-01 -8.35011959e-01
6.04118966e-02 3.80687296e-01 -3.48480195e-01 5.44867158e-01
-1.28596151e+00 1.13554275e+00 2.58855462e-01 3.07905197e-01
-1.52800113e-01 -2.40024716e-01 -2.99043924e-01 -2.71846473e-01
1.61281943e-01 -4.16543841e-01 7.74035811e-01 -1.22387123e+00
-1.16410494e+00 8.67501616e-01 -3.02896112e-01 -5.39658129e-01
2.07558468e-01 -8.86563435e-02 -8.00201774e-01 1.50985762e-01
-1.23739138e-01 5.06496727e-02 8.48549902e-01 -6.79780424e-01
-4.29157346e-01 -5.70715368e-01 -7.20407128e-01 2.89828926e-01
1.08101718e-01 2.01987416e-01 1.17983051e-01 -5.19571185e-01
5.81631780e-01 -7.45385945e-01 -6.60618618e-02 -5.83152890e-01
-3.60586315e-01 -4.23718430e-02 6.97650909e-01 -1.04509866e+00
1.48578131e+00 -2.14855504e+00 -8.64394158e-02 6.77594543e-01
1.08628176e-01 4.24119920e-01 3.59427750e-01 6.76065207e-01
-3.71967852e-01 1.69152945e-01 -1.32893547e-01 2.44347647e-01
-3.83284748e-01 1.62825152e-01 -1.18957393e-01 5.28279424e-01
1.73164412e-01 4.29838687e-01 -2.11778358e-01 -5.79603374e-01
2.42226481e-01 8.47193241e-01 -6.34091869e-02 -9.88332480e-02
5.39497912e-01 5.51506102e-01 -1.99251220e-01 5.80404758e-01
6.92854524e-01 5.15937984e-01 3.20798904e-02 -5.68578720e-01
-2.43539333e-01 -1.38041675e-01 -1.67838442e+00 1.28846586e+00
1.83690459e-01 6.83742344e-01 1.40357958e-02 -1.47625339e+00
1.37983775e+00 8.96375418e-01 5.38991690e-01 -5.09183884e-01
5.28979123e-01 2.89958537e-01 4.47781950e-01 -9.95534599e-01
-4.61096577e-02 -4.22155887e-01 2.92357504e-01 1.87989071e-01
-2.15892360e-01 -1.07860394e-01 2.01209728e-02 -9.84055623e-02
9.42929924e-01 1.09203175e-01 7.48989224e-01 -3.58986616e-01
8.05184484e-01 -1.25758639e-02 6.44350469e-01 5.28497994e-01
-5.40830076e-01 7.80087411e-01 3.61928403e-01 -4.54006582e-01
-6.01489067e-01 -7.87085712e-01 -1.60710812e-01 3.54623020e-01
-1.83621123e-01 -3.89326841e-01 -5.91632068e-01 -4.96536307e-02
-3.49756479e-01 2.84804583e-01 -3.22090060e-01 -2.27010194e-02
-5.96313775e-01 -6.81184530e-01 7.04359949e-01 -1.82805769e-02
5.13235033e-01 -1.36793363e+00 -1.02882886e+00 7.45828271e-01
-8.55897889e-02 -5.17668784e-01 1.46241963e-01 4.15632457e-01
-1.24762487e+00 -9.43943322e-01 -8.15752625e-01 -5.72606266e-01
1.50133282e-01 -1.06736138e-01 9.55619752e-01 2.45943129e-01
-1.08763409e+00 2.40238175e-01 -5.45382977e-01 -1.13204765e+00
-2.28993088e-01 -1.74718469e-01 -1.66377649e-01 -3.46296765e-02
3.68956655e-01 -8.39599729e-01 -6.57067180e-01 -1.86978709e-02
-5.90502381e-01 -4.32929337e-01 6.89680636e-01 3.48215461e-01
4.29531634e-01 3.56081426e-01 8.95251334e-01 -8.98983479e-01
9.54920590e-01 -2.72178322e-01 9.20948666e-03 -3.96372750e-02
-5.92997074e-01 -3.54911149e-01 5.02809286e-01 -1.43912122e-01
-6.10519826e-01 -5.60963973e-02 -5.48987210e-01 1.85021028e-01
-3.65324974e-01 3.59790921e-01 3.33730504e-02 1.93790793e-01
7.17272639e-01 1.87269032e-01 1.32530794e-01 -3.29436809e-01
-2.52972960e-01 7.97889173e-01 3.36138248e-01 -1.16501011e-01
4.97706950e-01 2.76649565e-01 1.88823283e-01 -1.22745216e+00
-2.15335414e-01 -7.83134758e-01 -3.69202495e-01 -3.73320878e-01
1.11006832e+00 -4.69082057e-01 -6.08289123e-01 3.37078363e-01
-1.20097303e+00 5.31400681e-01 1.06631875e-01 6.74248397e-01
-4.00529616e-02 5.52599251e-01 -2.02778161e-01 -1.40262759e+00
-9.59816396e-01 -6.59954250e-01 3.47133040e-01 6.13443971e-01
-6.80587351e-01 -5.93796670e-01 1.79589167e-01 -2.76590493e-02
6.90815628e-01 6.26185715e-01 6.99556768e-01 -4.50161517e-01
1.32903397e-01 -3.77441198e-01 1.78407192e-01 4.62921143e-01
4.06494677e-01 1.74527168e-01 -8.17167938e-01 9.83124897e-02
6.41814351e-01 4.58132267e-01 3.81883800e-01 4.50612694e-01
4.99379098e-01 9.45496336e-02 -4.79497761e-01 3.55892152e-01
1.60776865e+00 8.26707661e-01 1.11938262e+00 1.60177454e-01
1.88676685e-01 4.81463641e-01 5.59746504e-01 3.55952919e-01
3.55402008e-02 3.65171820e-01 -1.03117101e-01 -3.24504286e-01
1.59605071e-01 3.57223988e-01 1.29025832e-01 9.38299596e-01
-4.91885275e-01 7.02709556e-02 -8.51633787e-01 4.44486618e-01
-1.50641084e+00 -1.03578937e+00 -9.48994458e-01 2.32061219e+00
7.76513636e-01 2.77948111e-01 1.95455000e-01 9.88584101e-01
6.55133486e-01 -4.63639885e-01 -1.65757537e-01 -7.88455427e-01
-3.40277374e-01 9.15194273e-01 1.92786723e-01 2.47260049e-01
-9.45637226e-01 9.33831707e-02 5.50732088e+00 2.14954361e-01
-1.51359212e+00 4.23537344e-02 4.93203431e-01 8.40594694e-02
2.87622690e-01 7.59585053e-02 -5.14055848e-01 4.23535615e-01
9.97610211e-01 -2.09272653e-01 -5.08787371e-02 5.09725511e-01
5.95783472e-01 -5.10922968e-01 -5.82151175e-01 1.41982448e+00
7.09130317e-02 -7.63305008e-01 -4.33472335e-01 -4.19760555e-01
1.30396158e-01 -4.39581901e-01 -5.81610799e-01 2.18426928e-01
-9.57857430e-01 -1.06295669e+00 2.73609787e-01 9.53672111e-01
8.95444080e-02 -9.71512198e-01 1.04311419e+00 3.22096378e-01
-1.04083073e+00 6.17086291e-02 -3.06838334e-01 -1.98339105e-01
5.97632527e-02 8.36984158e-01 -6.90587759e-01 7.60047257e-01
7.76114523e-01 5.35976052e-01 -5.38821161e-01 1.44203138e+00
1.45512328e-01 7.52507329e-01 -4.94819462e-01 -1.55639052e-01
-2.56087184e-01 -1.90151379e-01 7.08376765e-01 1.33619726e+00
6.16736293e-01 4.04991895e-01 -3.57857376e-01 4.99071360e-01
6.90551460e-01 6.73786223e-01 -7.92909741e-01 2.14590132e-01
2.02911735e-01 1.50919294e+00 -1.15537429e+00 -1.87516734e-01
-2.06042826e-01 8.02358449e-01 -4.20456827e-01 -7.65662193e-02
-9.50819194e-01 -9.82117891e-01 7.62629835e-03 5.18450260e-01
-2.24833906e-01 3.32124010e-02 -2.56153435e-01 -6.05121851e-01
-5.48268817e-02 -1.18435144e+00 5.54593980e-01 -6.33906662e-01
-8.83620620e-01 7.26624370e-01 -6.23375922e-02 -1.08838522e+00
5.85816726e-02 -2.27286369e-01 -9.40557241e-01 1.08158290e+00
-9.35693741e-01 -4.61107671e-01 -2.47475922e-01 7.34314620e-01
3.65202487e-01 8.28900710e-02 9.81253326e-01 5.25293171e-01
-3.92644405e-01 4.23272550e-02 -5.01277745e-01 -5.33745512e-02
1.01495421e+00 -1.26224828e+00 -4.96148467e-01 7.68209934e-01
2.64289081e-01 8.56781065e-01 1.14755344e+00 -5.22756636e-01
-1.27914619e+00 -2.51657873e-01 9.30159509e-01 3.75334993e-02
-6.18506130e-03 1.27704546e-01 -9.38369572e-01 4.47504036e-02
6.44538224e-01 -4.98217463e-01 6.89018548e-01 -2.06912324e-01
4.23100233e-01 -5.90439260e-01 -1.16419697e+00 1.76262692e-01
2.57851779e-01 -6.34224107e-03 -9.79937971e-01 1.14929944e-03
-4.64499444e-01 -1.55625507e-01 -6.95033252e-01 4.92715180e-01
5.99758565e-01 -1.45822668e+00 6.45831585e-01 -1.60814449e-02
-2.56371498e-01 -6.01064205e-01 3.67862314e-01 -1.07400835e+00
-2.34655887e-02 -1.26388645e+00 6.06777966e-01 1.49871445e+00
4.60892648e-01 -8.69992495e-01 4.39390332e-01 2.79055417e-01
6.75022453e-02 -6.36560559e-01 -8.48156035e-01 -2.44027138e-01
-4.12953138e-01 -3.31786156e-01 -1.02137186e-01 9.25079644e-01
2.31366962e-01 4.94458586e-01 -4.83148575e-01 -1.36289999e-01
6.27561569e-01 -8.75549167e-02 3.61617833e-01 -1.51829576e+00
-2.22170159e-01 -1.50543004e-01 -6.47363484e-01 9.89222005e-02
-7.88969696e-01 -4.66694504e-01 -1.61748201e-01 -1.66047347e+00
9.68658775e-02 -1.75022483e-01 -4.09654707e-01 2.61481464e-01
-3.40683907e-01 5.23550272e-01 -6.75848648e-02 -2.11592205e-02
5.12924790e-02 -4.04046863e-01 9.02959704e-01 4.01709765e-01
-4.99088585e-01 1.48318991e-01 -5.73510408e-01 6.54136896e-01
9.46295917e-01 -7.66488135e-01 -6.14480376e-01 3.77425551e-01
5.56726515e-01 2.84939557e-01 1.35891557e-01 -1.43839908e+00
1.06199935e-01 4.30578947e-01 7.34483182e-01 -9.38350677e-01
2.41928712e-01 -9.00031924e-01 8.73190165e-01 5.91515243e-01
2.38659978e-01 3.23955387e-01 3.12298894e-01 2.00997204e-01
-2.84945428e-01 -4.02644575e-01 8.41132045e-01 -4.08120632e-01
-3.43636781e-01 -2.94423789e-01 -7.69602418e-01 -1.06868617e-01
1.28555226e+00 -7.17421889e-01 1.50647551e-01 -4.24580693e-01
-1.05647385e+00 2.15069205e-02 -2.37943217e-01 1.33040875e-01
5.94205499e-01 -8.42308700e-01 -7.69480705e-01 1.71354562e-01
-3.23084742e-01 -5.26577652e-01 4.82337266e-01 1.56049454e+00
-9.09764588e-01 2.33505458e-01 -6.84963226e-01 -6.16887093e-01
-2.02156258e+00 1.29283488e-01 4.31968510e-01 -3.81174460e-02
-9.72118437e-01 7.66704202e-01 -7.03274727e-01 4.63496357e-01
2.48083070e-01 -5.32592952e-01 -7.28150845e-01 3.98474395e-01
4.72344756e-01 4.55721855e-01 1.41870424e-01 -4.69496995e-01
-6.93060338e-01 9.83928919e-01 3.44446033e-01 -2.06528872e-01
1.44677353e+00 -2.83000350e-01 -3.86453748e-01 7.55064666e-01
9.11521077e-01 3.23806286e-01 -1.94841906e-01 5.47181427e-01
1.94945022e-01 -1.83097944e-01 -1.62263095e-01 -9.66539979e-01
-8.35001051e-01 1.03117383e+00 1.15975094e+00 7.16707647e-01
1.51399124e+00 -6.65678918e-01 6.50118589e-01 2.31731668e-01
1.90817431e-01 -1.24936712e+00 -3.03114057e-01 5.39777353e-02
7.89627910e-01 -9.10621941e-01 2.61131138e-01 -3.27595115e-01
-8.04129303e-01 1.52938032e+00 -1.05563670e-01 -2.91613817e-01
8.78117323e-01 4.28938866e-01 3.55048001e-01 -2.92558283e-01
-4.28372383e-01 -1.82654902e-01 1.07110940e-01 7.19664514e-01
9.51850832e-01 -1.75510868e-01 -1.17849803e+00 7.40236700e-01
-1.38947695e-01 4.07148600e-01 7.01324284e-01 1.03795958e+00
-7.23286152e-01 -1.25779140e+00 -6.89652503e-01 4.87636417e-01
-1.23088157e+00 1.58035770e-01 -3.23910743e-01 8.74549508e-01
4.46194023e-01 1.28046525e+00 -5.33632696e-01 -2.07019687e-01
5.60532272e-01 6.08791769e-01 5.01247644e-01 -3.66791040e-01
-1.07668614e+00 3.72421592e-01 1.20022245e-01 -3.71953517e-01
-4.33388263e-01 -5.19960821e-01 -1.52897859e+00 1.11391492e-01
-3.24070066e-01 5.08298755e-01 9.21458960e-01 7.05396175e-01
7.51686469e-02 6.62007511e-01 3.88113916e-01 -2.42985308e-01
-2.66023189e-01 -1.33369434e+00 -8.24592352e-01 4.69605803e-01
3.53202745e-02 -3.81942093e-01 -2.40888238e-01 2.49795884e-01] | [14.147150993347168, 3.190702438354492] |
6e03abfe-1c73-46d8-b79f-7097393a8265 | multipath-based-slam-for-non-ideal-reflective | 2304.05680 | null | https://arxiv.org/abs/2304.05680v2 | https://arxiv.org/pdf/2304.05680v2.pdf | Multipath-based SLAM for Non-Ideal Reflective Surfaces Exploiting Multiple-Measurement Data Association | Multipath-based simultaneous localization and mapping (SLAM) is a promising approach to obtain position information of transmitters and receivers as well as information regarding the propagation environments in future mobile communication systems. Usually, specular reflections of the radio signals occurring at flat surfaces are modeled by virtual anchors (VAs) that are mirror images of the physical anchors (PAs). In existing methods for multipath-based SLAM, each VA is assumed to generate only a single measurement. However, due to imperfections of the measurement equipment such as non-calibrated antennas or model mismatch due to roughness of the reflective surfaces, there are potentially multiple multipath components (MPCs) that are associated to one single VA. In this paper, we introduce a Bayesian particle-based sum-product algorithm (SPA) for multipath-based SLAM that can cope with multiple-measurements being associated to a single VA. Furthermore, we introduce a novel statistical measurement model that is strongly related to the radio signal. It introduces additional dispersion parameters into the likelihood function to capture additional MPCs-related measurements. We demonstrate that the proposed SLAM method can robustly fuse multiple measurements per VA based on numerical simulations. | ['Erik Leitinger', 'Thomas Wilding', 'Alexander Venus', 'Lukas Wielandner'] | 2023-04-12 | null | null | null | null | ['simultaneous-localization-and-mapping'] | ['computer-vision'] | [ 1.82177112e-01 -1.18285611e-01 4.88886297e-01 -2.18252331e-01
-1.10102642e+00 -6.17592514e-01 6.43286049e-01 4.07531083e-01
-3.01066130e-01 9.52498257e-01 -1.69323340e-01 -5.09239584e-02
-2.83339977e-01 -9.27816808e-01 -9.29521024e-01 -8.59497309e-01
-4.02672887e-01 6.64245009e-01 3.59656602e-01 -1.92454562e-01
9.90493298e-02 7.21950114e-01 -1.18626285e+00 -3.36996555e-01
9.00269389e-01 7.49739230e-01 5.79966724e-01 6.78413451e-01
-1.06547564e-01 5.84711917e-02 -9.17040110e-01 -5.26370816e-02
8.96777362e-02 -7.48135522e-02 1.03495434e-01 -3.78982872e-01
1.22907601e-01 4.32505831e-02 2.72436924e-02 1.10033011e+00
3.21794480e-01 -4.83401455e-02 5.70025682e-01 -1.39210308e+00
3.53758276e-01 2.50357866e-01 -5.11472702e-01 -3.55364591e-01
4.83414352e-01 -4.69190717e-01 4.19443280e-01 -7.69840598e-01
2.88591504e-01 1.09886765e+00 8.92883420e-01 -1.95499361e-01
-8.07969511e-01 -5.97963095e-01 -3.43004018e-01 -7.88079873e-02
-1.70912588e+00 -5.64658880e-01 7.43364453e-01 -1.53957441e-01
5.14291003e-02 5.49153090e-01 5.37650824e-01 9.04472709e-01
8.18368733e-01 8.47782642e-02 1.00423467e+00 -5.48443854e-01
4.35643852e-01 2.31255233e-01 -7.93334991e-02 5.12262225e-01
9.55669403e-01 9.20692757e-02 -5.96955001e-01 -6.12316370e-01
5.17541051e-01 -3.19631428e-01 -3.69998723e-01 -6.79737747e-01
-1.09123349e+00 5.15705705e-01 5.89211643e-01 2.50001192e-01
-7.95031846e-01 4.13039774e-01 -4.29571658e-01 1.00208039e-03
1.20553426e-01 2.75643021e-01 -1.01270795e-01 -5.55642601e-03
-8.56866777e-01 3.41873504e-02 8.09090674e-01 1.15218687e+00
1.00984621e+00 -2.15478260e-02 2.92328805e-01 3.31849843e-01
1.06096375e+00 1.26697111e+00 -9.28994715e-02 -7.59258330e-01
5.44679880e-01 7.43371919e-02 9.16111529e-01 -1.29670155e+00
-7.08150625e-01 -1.16236937e+00 -6.32531464e-01 2.98943464e-02
1.44960985e-01 -2.74489999e-01 -4.82654661e-01 1.51461089e+00
6.06189966e-01 7.07357585e-01 4.00345445e-01 6.53069139e-01
3.47508460e-01 5.32996833e-01 -6.05235755e-01 -3.53157848e-01
9.99488235e-01 -4.70818698e-01 -6.90830946e-01 -5.16309977e-01
6.05753779e-01 -1.13554931e+00 1.04174428e-01 4.87917244e-01
-6.40469193e-01 -3.41321081e-02 -1.42568076e+00 8.12886596e-01
7.14331344e-02 -4.94107865e-02 1.25851974e-01 9.50491905e-01
-8.45808566e-01 8.14598724e-02 -9.93487179e-01 -2.61309057e-01
-3.28393102e-01 1.36784762e-01 -8.75447690e-02 -2.87242144e-01
-1.03325450e+00 8.67458224e-01 -2.92678565e-01 5.91503263e-01
-7.03130662e-01 -4.86647338e-01 -6.77423239e-01 -3.23748767e-01
4.17432487e-02 -5.13503850e-01 8.09826791e-01 -4.38641995e-01
-1.40896881e+00 -1.81494713e-01 -6.74530566e-01 -4.67417300e-01
4.59762067e-01 -2.42152795e-01 -7.77307212e-01 9.55911353e-03
2.13472843e-01 -2.24290207e-01 5.24456322e-01 -1.98378825e+00
-4.98711050e-01 -1.90736622e-01 -2.95582801e-01 1.50036931e-01
4.68399435e-01 -3.61787587e-01 -1.78494886e-01 -6.40427843e-02
9.07786608e-01 -1.18711019e+00 -3.71909499e-01 -2.60073930e-01
-4.45002854e-01 8.61549258e-01 4.27069634e-01 -3.46935958e-01
6.10452652e-01 -2.13927102e+00 -2.52423286e-01 8.39108467e-01
-1.15986682e-01 -2.24411979e-01 -1.89694956e-01 8.63279939e-01
7.40538538e-01 -4.03127134e-01 -4.94977832e-02 -6.78654134e-01
-1.92180812e-01 3.66022706e-01 -1.46681637e-01 1.01481974e+00
-5.15382230e-01 3.66059840e-01 -1.00579977e+00 -1.00828178e-01
3.09684128e-01 7.29061186e-01 4.28988673e-02 -1.58555999e-01
2.83758909e-01 8.90125930e-01 -5.62897742e-01 4.47333604e-01
1.38279390e+00 2.76531875e-01 1.52857482e-01 -7.50997290e-02
-4.12797332e-01 2.93751270e-01 -1.66144669e+00 1.56334853e+00
-8.71148765e-01 3.88894588e-01 5.89574754e-01 -3.65697086e-01
1.12338662e+00 1.69247180e-01 3.38077813e-01 -5.81801951e-01
-9.81788784e-02 7.70815492e-01 -7.80561268e-02 1.46993250e-01
7.91964829e-01 -2.30316222e-02 -1.88124850e-01 1.15694888e-01
-3.27160418e-01 -7.55792707e-02 -5.73819757e-01 3.31385434e-02
1.26288986e+00 -8.77411664e-02 4.33700442e-01 -2.93417543e-01
7.82804251e-01 -2.11606994e-01 8.21157217e-01 9.43551302e-01
1.92854851e-01 3.38070482e-01 -2.18797043e-01 1.42028794e-01
-6.40370846e-01 -1.07128358e+00 -2.38586754e-01 2.37729382e-02
8.83444428e-01 -2.11554900e-01 -4.88427319e-02 1.31428093e-02
1.07511856e-01 7.17615545e-01 -1.53022766e-01 -4.30443697e-02
-3.65790218e-01 -7.71243393e-01 4.81116861e-01 -2.14523792e-01
2.37170443e-01 -1.25126094e-01 -4.94846612e-01 6.66697085e-01
-3.15887839e-01 -1.20944846e+00 2.64734209e-01 -2.15324201e-02
-6.73620760e-01 -9.87866342e-01 -3.18464011e-01 -1.15871757e-01
6.93864584e-01 9.26452577e-01 6.23852491e-01 4.78374884e-02
1.83908895e-01 5.70130885e-01 -5.44480324e-01 -5.99822879e-01
-6.35455430e-01 -6.02423608e-01 4.12625909e-01 5.13990343e-01
-2.67496407e-01 -5.19350886e-01 -5.29336572e-01 7.13859677e-01
-2.20001504e-01 -1.69319734e-01 6.41533732e-01 3.59469146e-01
5.40750146e-01 2.71687597e-01 2.47314423e-01 -3.44809324e-01
1.94240645e-01 -7.47752070e-01 -9.68168080e-01 1.39754117e-01
-2.50974774e-01 -9.23777670e-02 3.33456516e-01 6.35488629e-02
-1.03849220e+00 -5.59245497e-02 9.20734555e-02 1.91221729e-01
-9.63407606e-02 6.19315445e-01 -4.09254730e-01 -9.88030732e-01
4.51347560e-01 3.25985074e-01 -3.25935006e-01 -3.92241776e-01
1.55046225e-01 5.94313443e-01 2.99173564e-01 -3.36945653e-01
1.38889444e+00 8.74735832e-01 8.66529465e-01 -1.28526092e+00
-5.10354042e-01 -5.78102469e-01 -4.39338773e-01 -2.54159182e-01
1.31862849e-01 -1.09064698e+00 -3.69475543e-01 4.08744574e-01
-1.19136858e+00 1.87692106e-01 3.15842271e-01 1.03873587e+00
-2.01452628e-01 5.99038363e-01 1.21501222e-01 -1.33154249e+00
1.17469039e-02 -1.04193270e+00 1.02927959e+00 1.24396637e-01
3.59340315e-03 -1.08709335e+00 4.37461764e-01 -7.84822330e-02
6.46954775e-01 4.01797980e-01 1.63364559e-01 -3.53182733e-01
-9.17562783e-01 -6.69551253e-01 1.87095016e-01 -1.48886964e-01
2.33509451e-01 -4.54634011e-01 -7.12060213e-01 -5.53359389e-01
2.00822800e-01 6.54695451e-01 2.33569920e-01 4.60921466e-01
-7.11464658e-02 -1.03411965e-01 -8.51294398e-01 7.26959109e-01
1.52620685e+00 1.62579879e-01 6.08549833e-01 3.81307393e-01
5.91304541e-01 2.34513596e-01 1.15490305e+00 5.25022566e-01
4.40384120e-01 9.72281337e-01 9.28479195e-01 2.47476205e-01
1.26911014e-01 5.12698479e-02 2.76323706e-01 6.73122048e-01
1.23825409e-01 -6.31045520e-01 -6.15010381e-01 2.75117725e-01
-1.70238721e+00 -4.69148159e-01 -1.09527612e+00 2.84716129e+00
-1.39918521e-01 -8.64446759e-02 -7.21665740e-01 -1.86066907e-02
6.27954483e-01 1.61608696e-01 6.13522567e-02 2.11237431e-01
-5.15224822e-02 -1.52427897e-01 1.26374555e+00 1.21212959e+00
-5.91092050e-01 2.60166794e-01 4.84497070e+00 3.16264898e-01
-9.14138079e-01 2.17191041e-01 -7.73159385e-01 3.81418496e-01
-7.24304020e-01 4.50582743e-01 -9.17496562e-01 4.90446061e-01
1.01805973e+00 2.47658238e-01 9.54635143e-02 2.68498391e-01
4.23050642e-01 -6.06540501e-01 -4.87990320e-01 7.29200661e-01
-5.88021614e-02 -9.44159329e-01 -2.74349749e-01 5.31125844e-01
6.40303969e-01 3.33874375e-01 -8.80329236e-02 -3.12783301e-01
2.02543065e-02 -3.61122489e-01 7.76625574e-01 8.52949619e-01
5.70565015e-02 -6.86286628e-01 1.12605608e+00 5.26279807e-01
-1.33870268e+00 7.73903355e-02 -3.97951365e-01 6.15205765e-02
6.77575707e-01 1.13840342e+00 -1.39083993e+00 1.29021049e+00
1.13663249e-01 2.18527704e-01 -1.02416150e-01 1.76636362e+00
-3.83572191e-01 4.52694505e-01 -7.65983522e-01 -1.08680064e-02
2.13969290e-01 -3.09382349e-01 1.35693061e+00 7.32886076e-01
1.22445405e+00 -3.52399468e-01 6.11243658e-02 4.57455903e-01
4.24325287e-01 -5.97679475e-03 -4.57428753e-01 6.49500787e-01
1.15802014e+00 1.21911454e+00 -4.07362133e-01 1.19492792e-01
-3.17180306e-01 7.87249386e-01 -5.14820874e-01 5.25357485e-01
-7.17652798e-01 -1.86553210e-01 6.03622913e-01 9.34839919e-02
1.16212726e-01 -8.68620574e-01 -8.51045270e-03 -8.38424742e-01
1.02009177e-01 -2.08256677e-01 -4.00647134e-01 -5.72303832e-01
-7.92167008e-01 2.79183745e-01 -3.17931920e-01 -1.75690258e+00
1.41935162e-02 -1.96290109e-02 -5.33630967e-01 1.13508248e+00
-1.82183683e+00 -1.20293283e+00 -7.13800848e-01 3.19613546e-01
-2.45495737e-01 3.52764018e-02 9.71513927e-01 2.35608116e-01
7.63802379e-02 2.62762085e-02 9.80260730e-01 -4.43824530e-01
7.31228292e-01 -8.62174511e-01 3.07449043e-01 1.08639097e+00
3.33033979e-01 6.29599690e-01 1.33404553e+00 -9.19773042e-01
-1.64577460e+00 -1.07727516e+00 7.34349549e-01 -1.35459244e-01
5.11843204e-01 -4.13813144e-01 -5.50807595e-01 6.38057947e-01
-3.32371026e-01 4.74756844e-02 6.05118275e-01 -8.26001093e-02
1.92159578e-01 -3.04507256e-01 -1.22810459e+00 1.06025495e-01
4.64244455e-01 -9.92705971e-02 -1.28689036e-01 4.22152847e-01
2.35872731e-01 -5.77248037e-01 -5.32968760e-01 4.69255239e-01
5.83930731e-01 -9.07323241e-01 1.08923209e+00 5.27103484e-01
-8.58680785e-01 -7.28144646e-01 -6.50094390e-01 -1.52222323e+00
-6.42497912e-02 -4.93134558e-01 2.49332786e-02 1.23689020e+00
1.35304064e-01 -1.22930038e+00 5.20557821e-01 -1.51244178e-01
-2.59940147e-01 1.01215623e-01 -1.42051458e+00 -1.09890640e+00
-4.84180003e-01 -6.77173495e-01 8.32547605e-01 4.46433991e-01
-5.87922335e-01 8.69276747e-03 -4.77001429e-01 1.61943233e+00
1.43613696e+00 -1.79919526e-01 1.12699676e+00 -1.64495718e+00
-3.84831309e-01 2.56956190e-01 -5.41054308e-01 -1.04902518e+00
-2.52960712e-01 -4.43732262e-01 5.03367066e-01 -1.76059842e+00
-4.63648945e-01 -1.20614505e+00 -1.72433600e-01 -4.01733339e-01
4.52690095e-01 2.06334651e-01 -1.61812291e-01 3.16060424e-01
-3.43059421e-01 5.17939866e-01 7.26491392e-01 3.33183527e-01
-3.69139016e-01 8.58220994e-01 -6.81051016e-02 8.72184455e-01
5.05273938e-01 -7.83836961e-01 -2.13301927e-01 -5.03198981e-01
5.97880006e-01 5.34385920e-01 7.02749312e-01 -1.56968665e+00
4.07088786e-01 1.47828329e-02 8.48082080e-03 -8.92641902e-01
8.48048091e-01 -1.19098806e+00 8.25252712e-01 5.72420299e-01
4.29468006e-01 -4.72309023e-01 -5.30422144e-02 1.14125526e+00
-5.27973101e-02 -5.51311195e-01 4.93885189e-01 2.27826580e-01
-3.90772015e-01 -1.04966350e-01 -4.83234584e-01 -7.66183078e-01
9.34297085e-01 1.03265390e-01 -4.42843556e-01 -8.64412427e-01
-3.39286447e-01 9.89481360e-02 5.75681627e-01 1.94267631e-02
5.74307203e-01 -1.17776883e+00 -5.70553899e-01 -3.47487465e-03
1.89670578e-01 -1.97295621e-01 2.55288512e-01 1.01162994e+00
-5.55220723e-01 5.72319627e-01 1.27890065e-01 -6.51118755e-01
-1.30866945e+00 -1.28646269e-01 4.31580096e-01 1.55643120e-01
-3.66494775e-01 6.90901756e-01 -1.11052640e-01 -5.47201037e-01
-8.12876597e-02 -5.05088791e-02 4.39243019e-02 -3.18444312e-01
4.09289926e-01 5.22758543e-01 2.10136175e-01 -1.04770303e+00
-7.84148753e-01 8.58810723e-01 4.77946013e-01 -5.88913262e-01
9.62469339e-01 -7.73522794e-01 -1.73571870e-01 5.52218318e-01
9.08003211e-01 1.05204546e+00 -7.53682494e-01 -3.90782684e-01
-3.74830775e-02 -5.64765275e-01 1.07900258e-02 -6.56772196e-01
-2.95326561e-01 5.12950778e-01 6.19806230e-01 -2.31028609e-02
4.59419280e-01 -3.05515110e-01 5.06415665e-01 3.75554889e-01
1.51578367e+00 -5.81924379e-01 -4.50073510e-01 4.33912814e-01
5.77943385e-01 -7.65784085e-01 2.02927411e-01 -5.77312589e-01
7.71750510e-03 9.62705076e-01 -7.72939771e-02 -1.44710496e-01
5.98274767e-01 2.88042933e-01 2.03756228e-01 7.51348138e-02
-3.81522812e-02 2.26086602e-02 7.92971030e-02 6.93814039e-01
-8.60620737e-02 2.69634843e-01 -2.25386366e-01 4.46452558e-01
-3.22442949e-01 -5.04412591e-01 9.57719624e-01 1.12347817e+00
-9.04191077e-01 -1.32021427e+00 -1.10878778e+00 4.00635004e-02
1.62012950e-01 5.46565615e-02 3.09370935e-01 7.15098381e-01
-2.43518539e-02 1.13134849e+00 -1.39694154e-01 -2.47739688e-01
1.81052059e-01 -2.74108231e-01 5.12465894e-01 -3.88019502e-01
2.88543761e-01 1.61394909e-01 2.65177697e-01 -5.86889803e-01
-2.38742262e-01 -9.27645266e-01 -1.44822323e+00 1.27210155e-01
-4.58246768e-01 5.70103943e-01 1.46522141e+00 1.00721943e+00
4.22000200e-01 4.79879230e-01 9.40106988e-01 -8.70631039e-01
-4.23718184e-01 -7.41404235e-01 -1.04230571e+00 -5.18668115e-01
6.16420686e-01 -1.04008353e+00 -6.16836429e-01 -8.75699282e-01] | [6.234395503997803, 1.0369789600372314] |
d1982a0f-45eb-45b6-859c-3ec267b8e74d | stock-price-direction-prediction-by-directly | 1309.7119 | null | http://arxiv.org/abs/1309.7119v3 | http://arxiv.org/pdf/1309.7119v3.pdf | Stock price direction prediction by directly using prices data: an empirical study on the KOSPI and HSI | The prediction of a stock market direction may serve as an early
recommendation system for short-term investors and as an early financial
distress warning system for long-term shareholders. Many stock prediction
studies focus on using macroeconomic indicators, such as CPI and GDP, to train
the prediction model. However, daily data of the macroeconomic indicators are
almost impossible to obtain. Thus, those methods are difficult to be employed
in practice. In this paper, we propose a method that directly uses prices data
to predict market index direction and stock price direction. An extensive
empirical study of the proposed method is presented on the Korean Composite
Stock Price Index (KOSPI) and Hang Seng Index (HSI), as well as the individual
constituents included in the indices. The experimental results show notably
high hit ratios in predicting the movements of the individual constituents in
the KOSPI and HIS. | ['Yanshan Wang'] | 2013-09-27 | null | null | null | null | ['stock-prediction'] | ['time-series'] | [-9.09208596e-01 -3.59548032e-01 -5.19137442e-01 1.28112817e-02
-3.61297220e-01 -4.39061254e-01 6.09984994e-01 -9.37100872e-02
-1.87432379e-01 9.95629907e-01 2.92670101e-01 -7.69861400e-01
3.52077968e-02 -1.32380688e+00 7.56397471e-02 -7.45172739e-01
9.39184725e-02 6.88346475e-02 3.49264801e-01 -2.95617193e-01
9.83488023e-01 4.46835786e-01 -1.04473591e+00 -1.96031332e-01
6.30637527e-01 1.47659993e+00 -6.75062314e-02 6.21038117e-02
-2.54564643e-01 1.13455963e+00 -4.12907034e-01 -5.94769478e-01
5.20803273e-01 -3.69658291e-01 -8.80519226e-02 -1.31165132e-01
-6.23344660e-01 -6.69700563e-01 -1.28207624e-01 1.16846192e+00
2.04742163e-01 -3.23580392e-02 4.30408508e-01 -9.49648142e-01
-6.09339118e-01 9.48385477e-01 -7.21349418e-01 6.90396488e-01
2.72437464e-03 -1.22348694e-02 1.44397938e+00 -1.06793582e+00
2.26925671e-01 4.84608471e-01 4.60628271e-01 -1.67790934e-01
-5.89851797e-01 -1.02377784e+00 8.34539905e-02 3.19775224e-01
-8.53649735e-01 -1.30957514e-02 9.50800300e-01 -5.16671717e-01
6.70423567e-01 2.08489716e-01 9.68886375e-01 1.80705354e-01
5.08972943e-01 6.24444425e-01 1.09589434e+00 -1.64218903e-01
1.50045589e-01 2.24373981e-01 2.48298690e-01 -3.69450040e-02
7.12767124e-01 2.67873317e-01 -3.97592723e-01 -2.36852318e-01
9.73837078e-01 3.13856930e-01 -2.35467300e-01 5.68554699e-01
-1.29761493e+00 1.12009072e+00 3.26328307e-01 6.56749070e-01
-8.36495757e-01 -3.64368051e-01 1.70239970e-01 4.53382790e-01
8.50391746e-01 1.63942695e-01 -6.06941164e-01 -3.37126285e-01
-1.02767372e+00 3.76431823e-01 8.45799148e-01 4.01500314e-01
3.49463582e-01 4.98196691e-01 4.96184416e-02 3.02019984e-01
4.00127292e-01 6.72896087e-01 8.95426273e-01 -7.07866490e-01
7.31737554e-01 5.74752569e-01 4.97418940e-01 -1.45808554e+00
-3.58758897e-01 -4.90953326e-01 -7.30995119e-01 3.26974213e-01
4.22885120e-01 -4.38112408e-01 -1.34271681e-01 9.35431242e-01
-8.27694535e-02 3.24112028e-01 3.11949342e-01 4.81356084e-01
3.44641954e-01 1.19237149e+00 -2.79842407e-01 -8.22170675e-01
1.08781981e+00 -8.44005406e-01 -8.45475018e-01 9.83967930e-02
5.71182787e-01 -6.73699439e-01 3.32600415e-01 4.23398376e-01
-1.17535126e+00 -3.09636235e-01 -7.15895772e-01 7.59856045e-01
-3.18302304e-01 1.34359539e-01 4.86799955e-01 2.63969153e-01
-6.42834961e-01 8.92364323e-01 -8.75492394e-01 4.51847374e-01
-8.71430859e-02 9.61808860e-02 1.21004604e-01 7.26197779e-01
-1.27997267e+00 9.47359383e-01 7.61798143e-01 2.43467450e-01
1.24059223e-01 -5.54631174e-01 -4.03569758e-01 3.51491332e-01
-6.28214851e-02 2.51494423e-02 1.13824618e+00 -6.62155688e-01
-1.33720767e+00 3.62205654e-01 3.19861084e-01 -5.65473258e-01
3.98746789e-01 6.87701777e-02 -7.31784999e-01 -3.73060592e-02
-7.84819480e-03 -3.21336359e-01 1.17842287e-01 -3.96633804e-01
-1.25992417e+00 -2.34377116e-01 -1.45693153e-01 2.13906784e-02
-4.21360761e-01 3.25156361e-01 2.73573548e-01 -1.26775134e+00
4.64283198e-01 -6.32881105e-01 -2.21533254e-01 -7.72586405e-01
-2.88090169e-01 -3.12012643e-01 3.96075249e-01 -1.25247955e+00
1.78925288e+00 -1.88812900e+00 -6.51561737e-01 6.01415396e-01
-1.90722525e-01 8.00033063e-02 6.20474458e-01 5.17584383e-01
-3.63579720e-01 2.54498452e-01 6.58747777e-02 2.79520929e-01
1.24255255e-01 -7.81374350e-02 -7.43478358e-01 3.75704497e-01
1.02486216e-01 6.45935118e-01 -5.65167606e-01 -7.38963559e-02
-9.89997908e-02 -3.19007516e-01 -2.04541028e-01 1.95610628e-01
2.66003281e-01 -1.08561236e-02 -6.45900905e-01 7.86659300e-01
6.96668983e-01 -1.76281467e-01 -1.80943474e-01 1.22760706e-01
-8.90764892e-01 5.24583519e-01 -1.20728648e+00 2.97974706e-01
2.39078104e-01 3.33968103e-01 -5.52222431e-01 -1.04388785e+00
1.45348525e+00 5.13497531e-01 5.57039499e-01 -8.31189454e-01
-6.39358237e-02 8.90263855e-01 1.01630338e-01 -2.02380404e-01
5.96339643e-01 -5.18379450e-01 -4.86121215e-02 6.47010386e-01
-6.89197004e-01 3.53818566e-01 6.85515761e-01 -2.19946191e-01
4.80577350e-01 -2.69258946e-01 5.60348213e-01 -3.70734543e-01
7.55321026e-01 -2.09541842e-02 1.16685033e+00 -2.73669809e-02
1.58432629e-02 2.68050224e-01 8.84580910e-01 -4.83534396e-01
-9.59407508e-01 -6.48007333e-01 -1.30361483e-01 4.51460540e-01
-2.99878836e-01 1.43627927e-01 -2.15261981e-01 -4.31784838e-01
2.90701926e-01 8.63635838e-01 -3.04802418e-01 1.73894152e-01
-5.60373962e-01 -1.24451208e+00 -4.03596759e-02 6.84991658e-01
7.65005350e-01 -1.28628194e+00 -4.30452496e-01 5.19572139e-01
7.38799870e-02 -5.50928116e-01 -1.68982849e-01 -1.67346328e-01
-1.10988438e+00 -1.03095782e+00 -1.10588026e+00 -6.62876129e-01
4.84940320e-01 7.41738686e-03 9.96474504e-01 6.28154203e-02
7.19297528e-01 -4.31087792e-01 -3.04807097e-01 -7.77538836e-01
-1.98329568e-01 -2.26075128e-02 7.03823939e-02 2.07874581e-01
4.73340213e-01 -4.48012143e-01 -6.65871680e-01 4.55053091e-01
-5.11908114e-01 1.70400478e-02 3.26552719e-01 7.07415640e-01
2.65427381e-01 4.59458053e-01 1.26848829e+00 -5.98871529e-01
5.95600963e-01 -6.97054207e-01 -1.32368064e+00 1.57262057e-01
-1.27574337e+00 -2.95049071e-01 2.82065094e-01 7.77468784e-03
-1.45411170e+00 -5.95591962e-01 -1.83552742e-01 2.15856671e-01
4.67626572e-01 1.32863665e+00 2.97917277e-01 1.97083414e-01
-3.23712766e-01 4.91327912e-01 -4.30152528e-02 -7.91406274e-01
-5.75214565e-01 5.76528311e-01 4.80706364e-01 4.95312959e-02
1.08568466e+00 -3.06361988e-02 -2.31295079e-02 -2.52960443e-01
-6.64511621e-01 -5.18279135e-01 -6.41162694e-01 -1.14925988e-01
4.35910851e-01 -1.11018109e+00 -6.50751412e-01 9.36788976e-01
-9.25095379e-01 1.09957151e-01 4.60772775e-02 9.01993394e-01
-2.20209137e-01 1.16035938e-01 -1.10953629e+00 -1.03324127e+00
-3.52755368e-01 -7.60347486e-01 1.18443877e-01 6.45805120e-01
8.39753896e-02 -1.36027980e+00 3.91693324e-01 1.00770541e-01
6.21898293e-01 4.71854329e-01 6.92233324e-01 -1.09605837e+00
-6.23574674e-01 -5.42029023e-01 -3.31264198e-01 5.95313430e-01
2.35952258e-01 4.01078075e-01 -4.23011661e-01 -7.08412156e-02
3.45764428e-01 1.40051976e-01 5.46745956e-01 3.16124409e-01
4.56568480e-01 -6.15885496e-01 1.20863751e-01 3.56632411e-01
1.57150638e+00 9.20404196e-01 7.82859027e-01 1.06288850e+00
1.11934438e-01 4.36681509e-01 9.93625045e-01 9.97935414e-01
3.54625523e-01 1.19639367e-01 -4.75298390e-02 1.13604628e-01
7.64316678e-01 -2.49292422e-02 6.91159189e-01 1.53883386e+00
-7.54641533e-01 1.65053636e-01 -9.24712598e-01 3.94839197e-01
-1.74579895e+00 -1.33439720e+00 -3.25855374e-01 1.98790467e+00
7.60585427e-01 4.98155564e-01 2.40672469e-01 4.19786513e-01
4.36820328e-01 2.27144435e-01 -2.18473196e-01 8.15623850e-02
-2.94109404e-01 -8.02219361e-02 6.83872938e-01 -1.89403910e-02
-9.07778919e-01 3.47469747e-01 6.19012690e+00 3.69664699e-01
-1.25734961e+00 -3.74415785e-01 1.08360946e+00 2.21696645e-01
-3.76734555e-01 4.73017991e-02 -1.13375902e+00 9.75336075e-01
1.02184653e+00 -9.94591773e-01 -1.57712549e-01 9.38226163e-01
5.95157862e-01 -1.47527248e-01 -4.30573672e-01 8.46252441e-01
-5.04303157e-01 -1.68626666e+00 -3.83841991e-01 5.08229733e-01
8.16277206e-01 -1.07464954e-01 6.48418218e-02 1.58306450e-01
3.11622806e-02 -2.79216707e-01 7.84542680e-01 1.00963593e+00
7.54660666e-02 -1.08809996e+00 1.31674731e+00 5.11733055e-01
-1.18692696e+00 -3.33738536e-01 -5.61112702e-01 -4.81278092e-01
3.30356568e-01 7.43485987e-01 -2.61396319e-01 4.84993875e-01
4.95492846e-01 9.01118755e-01 -3.38112265e-01 1.17618048e+00
-1.62577853e-01 8.69959772e-01 -1.56972110e-01 -8.38985145e-02
4.20692474e-01 -9.32829738e-01 5.79335578e-02 7.14179158e-01
1.00170815e+00 5.28846204e-01 -1.12881623e-01 6.36090755e-01
-2.02283300e-02 4.09912914e-01 -4.00922626e-01 -1.63122430e-01
1.98059395e-01 8.32724750e-01 -6.17282629e-01 -5.00294745e-01
-9.66033220e-01 1.12573735e-01 -3.41001362e-01 2.69354880e-01
-5.77909946e-01 -4.47226495e-01 5.70315242e-01 2.40359128e-01
4.38243926e-01 -2.30319992e-01 -7.83508122e-01 -1.52985406e+00
6.53567985e-02 -6.24926090e-01 5.06747842e-01 -4.14400458e-01
-1.23861647e+00 1.42617613e-01 -1.23384275e-01 -1.55743706e+00
-6.34137452e-01 -6.57965302e-01 -1.36447406e+00 1.20425010e+00
-1.89186430e+00 -3.00374210e-01 1.63844734e-01 4.03274536e-01
1.37972966e-01 -6.63537800e-01 3.58594805e-01 1.87390000e-01
-1.09181130e+00 4.64281328e-02 7.77510166e-01 7.31866658e-01
6.25790432e-02 -1.24259651e+00 4.75043505e-01 1.07901907e+00
-1.13344625e-01 3.65582347e-01 2.97875464e-01 -8.65522444e-01
-6.28242433e-01 -6.14581704e-01 1.37503970e+00 1.36088639e-01
1.28227842e+00 5.45460105e-01 -1.08610916e+00 7.32366920e-01
2.78516501e-01 -2.72467852e-01 6.31694734e-01 -3.46497774e-01
2.80978739e-01 -3.78538966e-01 -9.27336574e-01 2.41312668e-01
-1.23880409e-01 -1.15554646e-01 -7.72670627e-01 1.34579599e-01
4.33742821e-01 5.69198020e-02 -1.41163480e+00 1.87646568e-01
6.21422768e-01 -1.20747507e+00 8.04181397e-01 -2.88571626e-01
2.70602703e-01 1.59197282e-02 6.81794509e-02 -1.24838209e+00
-7.14523375e-01 -3.55033994e-01 -3.96309048e-02 1.57787752e+00
6.16000473e-01 -1.21471453e+00 8.57051551e-01 1.08794475e+00
2.98913300e-01 -5.14554143e-01 -8.41927707e-01 -7.19366252e-01
1.66890308e-01 -7.58762285e-02 1.02680421e+00 9.66172874e-01
3.30276847e-01 -1.96147904e-01 -3.15206409e-01 -8.42170194e-02
5.60599506e-01 5.51143646e-01 4.49108452e-01 -1.48187649e+00
-7.66486675e-02 -9.52135026e-01 -9.77134034e-02 -6.14031672e-01
-5.93868047e-02 -3.75169367e-01 -5.66869736e-01 -1.15059566e+00
-8.52352232e-02 -4.90927845e-01 -8.84348571e-01 1.22753479e-01
-2.09337741e-01 2.78197788e-02 3.97855252e-01 7.09594786e-01
3.35860699e-01 4.73879278e-01 9.38983738e-01 1.59267873e-01
-1.73173711e-01 7.15525508e-01 -6.57825351e-01 1.03441429e+00
1.12139368e+00 -2.08909750e-01 7.89508522e-02 1.77586600e-01
5.52841902e-01 5.89219987e-01 -1.12126656e-01 -7.69293547e-01
1.96297050e-01 -9.34294045e-01 3.42351556e-01 -1.23237646e+00
-3.14591557e-01 -4.54491317e-01 2.54059970e-01 8.10473442e-01
-6.86636940e-02 6.28780007e-01 -1.73663676e-01 2.11610913e-01
-7.96450436e-01 -5.38262367e-01 5.67018032e-01 -3.01884472e-01
-6.34211063e-01 3.76124233e-01 -3.55632722e-01 -2.15854198e-01
1.13108563e+00 -1.03739642e-01 -2.95656979e-01 -4.03495759e-01
-4.11767572e-01 3.01438749e-01 1.89534441e-01 1.18949421e-01
6.59165561e-01 -1.68718684e+00 -1.01518965e+00 2.70740390e-02
-2.99685776e-01 -3.78339708e-01 -2.51037061e-01 9.69580114e-01
-9.03869629e-01 8.14992249e-01 -2.03091711e-01 2.27508858e-01
-6.94451988e-01 4.26550359e-01 2.70951152e-01 -7.33116925e-01
-4.76217479e-01 3.56449783e-01 2.12378472e-01 3.45293730e-01
-1.01573177e-01 -5.08549988e-01 -7.85512745e-01 8.07578921e-01
9.67854917e-01 5.63531876e-01 -1.13690317e-01 -7.18258977e-01
-7.87338018e-02 4.08471525e-01 3.05059612e-01 -5.56038879e-02
1.81490660e+00 -2.96298325e-01 -4.55943495e-01 7.76830852e-01
1.00666261e+00 -5.02543431e-03 -1.05677783e+00 -3.88520628e-01
9.47216272e-01 -5.35223842e-01 1.60507649e-01 -2.72274345e-01
-1.60301042e+00 5.62349081e-01 5.41728064e-02 6.28406048e-01
1.15146518e+00 -4.65304554e-01 1.28933275e+00 2.99762577e-01
3.65524381e-01 -1.31867111e+00 -2.25038901e-01 3.51188362e-01
8.75562012e-01 -1.19229329e+00 -3.77357751e-02 -1.35530941e-02
-6.26610458e-01 1.57424033e+00 1.34684876e-01 -2.21224114e-01
1.26093316e+00 2.80470233e-02 1.80156380e-01 1.30954266e-01
-8.75414670e-01 7.55526796e-02 1.95950016e-01 -1.64847802e-02
3.54185313e-01 1.48977607e-01 -6.88406229e-01 9.08111572e-01
-4.30333465e-01 -1.51976086e-02 9.81653690e-01 9.17668223e-01
-7.12834835e-01 -9.23538625e-01 -6.95696533e-01 7.78915823e-01
-1.23345184e+00 -1.46812335e-01 2.00371429e-01 7.37336278e-01
-3.96542370e-01 5.53122282e-01 4.62235600e-01 -1.38829052e-01
3.38483691e-01 1.27052218e-01 -6.24623716e-01 -2.12779418e-01
-4.90216970e-01 4.13554221e-01 8.20352733e-02 -1.02473646e-01
-5.49240947e-01 -1.13708675e+00 -1.27389693e+00 -7.67547846e-01
-4.15543586e-01 3.21522206e-01 2.78973013e-01 9.73800898e-01
-3.89800668e-01 -1.93521027e-02 1.52450919e+00 -6.73239052e-01
-9.36070204e-01 -1.00903535e+00 -1.49861085e+00 3.47693302e-02
1.81059361e-01 -4.90171283e-01 -6.02813661e-01 6.08918555e-02] | [4.528965950012207, 4.205076694488525] |
52649c37-b58b-4673-a182-6da6ff84defb | shortest-paths-in-hsi-space-for-color-texture | 1904.07429 | null | http://arxiv.org/abs/1904.07429v1 | http://arxiv.org/pdf/1904.07429v1.pdf | Shortest Paths in HSI Space for Color Texture Classification | Color texture representation is an important step in the task of texture
classification. Shortest paths was used to extract color texture features from
RGB and HSV color spaces. In this paper, we propose to use shortest paths in
the HSI space to build a texture representation for classification. In
particular, two undirected graphs are used to model the H channel and the S and
I channels respectively in order to represent a color texture image. Moreover,
the shortest paths is constructed by using four pairs of pixels according to
different scales and directions of the texture image. Experimental results on
colored Brodatz and USPTex databases reveal that our proposed method is
effective, and the highest classification accuracy rate is 96.93% in the
Brodatz database. | ['Hongyan zhang', 'Tianyu Wang', 'Lintao Zheng', 'Yongsheng Dong', 'Mingxin Jin', 'Lingfei Liang'] | 2019-04-16 | null | null | null | null | ['texture-classification'] | ['computer-vision'] | [ 1.55560598e-01 -7.91487932e-01 1.65893964e-03 -1.45841062e-01
5.34732863e-02 -2.80678332e-01 1.81177706e-01 -1.60733759e-01
-9.51634720e-02 4.67144430e-01 -3.66955549e-01 -1.08944625e-01
-4.26034629e-01 -1.34873605e+00 8.11778381e-02 -9.69739497e-01
-1.66023746e-01 -2.01696590e-01 4.24584806e-01 -2.55031437e-01
5.74401498e-01 7.09326506e-01 -1.46574724e+00 4.72348690e-01
4.56148475e-01 1.51821351e+00 -2.06015855e-02 6.05288267e-01
-3.95528108e-01 4.61295456e-01 -2.03246400e-01 -3.05698186e-01
2.20388681e-01 -5.82424998e-01 -8.35533381e-01 5.43320775e-01
-8.98910910e-02 -5.96381277e-02 -5.58638990e-01 1.14127374e+00
6.59128949e-02 2.61389881e-01 7.09763527e-01 -1.33587694e+00
-8.06469917e-01 -3.70631702e-02 -1.00110877e+00 -8.48561451e-02
3.03252101e-01 -4.99722004e-01 7.21506417e-01 -6.98853493e-01
7.31597066e-01 1.23373520e+00 2.42262468e-01 1.08049802e-01
-8.34339440e-01 -4.11907673e-01 -1.73218891e-01 7.11403966e-01
-1.63713169e+00 7.80356452e-02 1.08937895e+00 -1.23739325e-01
1.98468909e-01 6.78671300e-01 1.06750751e+00 5.29516459e-01
5.99880099e-01 5.65806031e-01 1.84182596e+00 -6.25487566e-01
-1.22459188e-01 -2.13396490e-01 2.54226834e-01 1.18610775e+00
2.72592813e-01 -7.79278055e-02 -4.76904660e-01 -9.16655064e-02
1.05006480e+00 1.56147301e-01 -9.26238000e-02 -2.82630771e-01
-1.05869150e+00 7.36890256e-01 5.27957261e-01 2.25048617e-01
-3.82050276e-02 -1.30507555e-02 1.06983200e-01 3.72226834e-01
4.45916116e-01 -1.95547342e-01 6.10808656e-02 7.87530988e-02
-3.10245365e-01 -2.99243301e-01 7.11438477e-01 9.13373113e-01
8.02483618e-01 -2.28084400e-01 2.48575613e-01 1.03195989e+00
2.91209549e-01 7.72312760e-01 4.65425029e-02 -6.84504569e-01
2.36315787e-01 5.78286290e-01 -1.27917215e-01 -1.84566116e+00
-4.92207199e-01 2.72790164e-01 -1.06928396e+00 1.53133258e-01
3.67057443e-01 3.52633446e-01 -9.80408370e-01 7.61775732e-01
3.78981113e-01 -1.10073816e-02 -1.18719086e-01 9.79319036e-01
6.63813829e-01 8.71741712e-01 -1.67688161e-01 4.43391278e-02
1.45421410e+00 -5.92708468e-01 -6.79355860e-01 3.58261943e-01
8.99257287e-02 -1.03935957e+00 1.11163163e+00 7.52934039e-01
-4.84993815e-01 -5.42952359e-01 -1.14841926e+00 -5.52819809e-04
-6.07834935e-01 3.54366779e-01 8.34823489e-01 7.38862574e-01
-7.45660543e-01 3.94588768e-01 -7.15266943e-01 -4.08528924e-01
1.23933464e-01 -2.69126650e-02 -6.31192267e-01 -5.03317773e-01
-9.85669255e-01 4.03074384e-01 3.58864456e-01 6.56364441e-01
-8.05686936e-02 4.30304378e-01 -7.31832743e-01 -1.99130610e-01
1.04771994e-01 -1.17638884e-02 1.80908650e-01 -8.24655533e-01
-1.46946168e+00 7.87710488e-01 -4.77870367e-02 4.48931962e-01
3.15641046e-01 5.14526606e-01 -8.47537756e-01 4.78773922e-01
-4.00423646e-01 1.20559059e-01 8.08917522e-01 -1.27958953e+00
-8.52593005e-01 -5.30050039e-01 -8.41063261e-02 1.75251782e-01
-3.31319332e-01 -3.28985631e-01 -9.63951766e-01 -4.55778271e-01
1.10522628e+00 -1.01939332e+00 1.03415594e-01 9.27796960e-02
-6.93265676e-01 -8.00983235e-03 1.19695115e+00 -8.62364292e-01
1.17791831e+00 -2.23047471e+00 -4.51484211e-02 1.07718515e+00
2.01679260e-01 -3.40140671e-01 -2.06160814e-01 3.33931684e-01
2.58398801e-01 1.76798150e-01 -5.25327176e-02 5.69369137e-01
-2.54983395e-01 2.11345017e-01 1.94749713e-01 5.86955369e-01
-1.36707485e-01 3.08515072e-01 -5.90009034e-01 -6.98279202e-01
1.60005599e-01 5.66442251e-01 -2.43097376e-02 -8.72765481e-02
4.42661643e-01 2.38271981e-01 -8.37979972e-01 8.38670194e-01
1.11887288e+00 1.23302355e-01 4.24603462e-01 -6.05209470e-01
-4.07154486e-02 -4.20839578e-01 -1.27813768e+00 1.35798156e+00
-3.26344341e-01 6.00728273e-01 -4.06503469e-01 -9.45667386e-01
1.40169394e+00 -5.12164570e-02 4.51059461e-01 -1.27670074e+00
2.10983202e-01 2.55542427e-01 -2.12862805e-01 -8.22840750e-01
5.07897198e-01 1.20057836e-01 -1.41807169e-01 2.41720274e-01
-3.99055958e-01 -1.67613342e-01 3.79401416e-01 -2.74911523e-01
4.87768590e-01 9.28723812e-02 5.40758073e-02 -3.40747148e-01
7.00713634e-01 1.06358409e-01 4.88108218e-01 3.75385821e-01
-1.58299343e-03 4.24287319e-01 8.02906036e-01 -8.43801916e-01
-7.92469978e-01 -8.84028137e-01 -1.92687929e-01 6.69263363e-01
7.96428502e-01 -1.62253797e-01 -6.39739037e-01 -6.79311514e-01
-6.25211447e-02 1.08154379e-01 -8.79962683e-01 -2.88929534e-03
-4.48899597e-01 -6.86020792e-01 1.12510957e-01 6.34380281e-02
1.14529705e+00 -7.74210393e-01 -3.47842127e-01 -3.94539386e-02
-3.37657303e-01 -7.70714641e-01 -3.02172869e-01 -2.89003044e-01
-8.97005379e-01 -1.58470023e+00 -5.71151197e-01 -9.33001637e-01
8.96858871e-01 6.14333153e-01 4.68423039e-01 2.74400979e-01
-6.29518270e-01 2.02552259e-01 -6.24190807e-01 9.77054983e-02
2.90045857e-01 -2.29950026e-01 -3.70942146e-01 5.45425177e-01
2.27269113e-01 5.91468029e-02 -7.38778174e-01 6.77025080e-01
-9.39812243e-01 4.04976219e-01 3.83867294e-01 9.33303833e-01
9.94184911e-01 6.48659468e-01 -2.20218837e-01 -8.62964869e-01
5.03314972e-01 -2.57195562e-01 -7.40794718e-01 7.02287912e-01
-3.15856099e-01 -1.39941871e-01 3.68192136e-01 1.75160188e-02
-1.07222152e+00 1.15242088e-02 1.24642611e-01 1.26979817e-02
9.86225059e-05 6.53739810e-01 -2.06190236e-02 -6.69824302e-01
4.34644893e-02 4.73629683e-01 -2.75766626e-02 -5.13782501e-01
6.94971979e-02 6.96199477e-01 -4.66034822e-02 -5.20608664e-01
4.74659234e-01 6.90680265e-01 4.45860058e-01 -1.11758757e+00
-2.47449294e-01 -1.67204350e-01 -7.14965045e-01 -4.11039472e-01
9.30683553e-01 -1.70797259e-01 -1.04873836e+00 7.96480596e-01
-6.52881145e-01 -1.20671272e-01 4.88472730e-01 4.88623828e-01
-3.10543448e-01 3.86456907e-01 -7.64558256e-01 -6.48702264e-01
-1.23992324e-01 -1.19492698e+00 6.57428443e-01 4.55117583e-01
4.88470823e-01 -1.18868637e+00 -3.77403378e-01 -3.31690907e-02
3.14863563e-01 4.84160095e-01 1.17718065e+00 2.52758622e-01
-5.25456548e-01 -2.45820627e-01 -6.93394721e-01 6.00722842e-02
4.64056224e-01 5.06237388e-01 -4.57797945e-01 -2.58225538e-02
-2.30250150e-01 3.33148316e-02 8.15455735e-01 1.66173428e-01
1.67021871e+00 -1.48416579e-01 -1.92521736e-01 9.35837030e-01
1.77501738e+00 7.78892100e-01 9.90317762e-01 5.32165766e-01
8.09271097e-01 6.99968934e-01 8.53352547e-01 3.08482736e-01
3.15713495e-01 4.05964017e-01 -2.74927169e-03 -4.05256271e-01
-1.19543038e-01 -4.12834175e-02 -3.30959290e-01 9.28290069e-01
-4.75192696e-01 -2.09389001e-01 -9.78003502e-01 -1.47743851e-01
-1.52052450e+00 -7.01788902e-01 -5.58866918e-01 2.16562462e+00
2.85629749e-01 -7.82119259e-02 -2.51812547e-01 5.35991728e-01
7.31953025e-01 -1.45710604e-02 -3.56167629e-02 -4.17794645e-01
-3.75114083e-01 3.30084264e-01 5.25346100e-01 4.00151730e-01
-1.12270403e+00 8.37760389e-01 6.08300352e+00 1.07237411e+00
-1.53690100e+00 -3.54460448e-01 9.59612131e-01 5.26462793e-01
-6.48668110e-02 -5.44190221e-02 -2.06185371e-01 3.59241694e-01
-3.36462930e-02 -1.23445189e-03 6.64203048e-01 3.42537582e-01
1.60308301e-01 -6.26050711e-01 -2.11997032e-01 1.31418478e+00
-3.63224931e-02 -9.42435384e-01 3.14935327e-01 -7.99857676e-02
5.57213902e-01 -6.17670178e-01 3.22524399e-01 -4.38008845e-01
-2.68444810e-02 -8.79191935e-01 4.75236565e-01 9.35094714e-01
1.17580247e+00 -9.42301631e-01 5.97731709e-01 -3.55744094e-01
-1.68439233e+00 1.08338542e-01 -6.44195795e-01 2.21390173e-01
-4.56880778e-01 4.27058786e-01 -3.09678435e-01 7.46933758e-01
7.00307250e-01 7.63668597e-01 -7.16985285e-01 1.25708199e+00
-1.17437787e-01 3.19072604e-01 -4.84056808e-02 -4.08520281e-01
3.00879776e-01 -7.80672073e-01 5.10596903e-03 1.06882155e+00
6.55812144e-01 4.58540112e-01 2.91681498e-01 2.71908849e-01
1.76360309e-01 3.28612655e-01 -5.48670530e-01 -4.35162038e-02
2.43767053e-01 1.42041159e+00 -1.37919176e+00 -1.57741904e-01
-3.26925486e-01 1.24034870e+00 -7.51616657e-02 8.37942421e-01
-6.46633267e-01 -1.07876480e+00 2.22110793e-01 -2.33563766e-01
-7.25755468e-02 -5.33036947e-01 -3.55731487e-01 -8.66809964e-01
-2.80606747e-01 -6.10953212e-01 4.03844655e-01 -9.43053544e-01
-1.09822035e+00 7.27496743e-01 -3.71535897e-01 -1.45382011e+00
7.11438954e-01 -9.37479019e-01 -2.93713510e-01 9.94098902e-01
-1.51139915e+00 -1.27372169e+00 -7.96521723e-01 1.14263594e+00
8.94084871e-02 7.10352063e-02 7.97591627e-01 9.59464461e-02
-6.79985106e-01 2.37756431e-01 3.70396256e-01 2.84094542e-01
3.66206437e-01 -9.31532741e-01 -1.15503140e-01 5.96869707e-01
-4.38974146e-03 3.78798485e-01 1.86564982e-01 -5.64323962e-01
-1.86491776e+00 -7.60672450e-01 5.47708929e-01 4.51923370e-01
4.47569728e-01 -1.12357736e-01 -5.31546474e-01 1.24506801e-01
-1.15789451e-01 1.88505411e-01 5.94985545e-01 -1.48692474e-01
-2.86838770e-01 -5.78624249e-01 -1.12795496e+00 5.79619646e-01
8.08557749e-01 -5.77040076e-01 2.30364591e-01 1.57341361e-01
-2.00801343e-01 -3.49939197e-01 -8.47300112e-01 -9.03299265e-03
9.99499977e-01 -1.00047362e+00 6.92482412e-01 -1.41203478e-01
2.83348799e-01 -5.16268432e-01 -3.41735780e-01 -1.18630707e+00
-5.14086485e-01 9.16006044e-02 6.42251611e-01 7.72141814e-01
2.46894360e-01 -7.93263674e-01 7.04215586e-01 9.83908251e-02
3.17374021e-01 -6.85424507e-01 -7.26142466e-01 -5.21975756e-01
-4.43553507e-01 -2.20351875e-01 6.64292574e-01 9.50711131e-01
7.85000548e-02 -2.52940565e-01 -3.73760343e-01 1.55277848e-02
8.68252575e-01 3.34515154e-01 4.04675692e-01 -1.18679142e+00
3.13138843e-01 -5.00492334e-01 -6.95622265e-01 -5.50262451e-01
-2.54210353e-01 -6.28826201e-01 -2.81445414e-01 -1.60908759e+00
2.26496994e-01 -1.21050107e+00 -5.89617014e-01 4.05539513e-01
-9.65914805e-04 8.44077528e-01 1.17231160e-01 1.08554788e-01
-3.07733178e-01 2.54462779e-01 1.75721002e+00 -4.58539277e-01
-6.80483282e-02 -1.65502489e-01 -3.42282534e-01 5.06802142e-01
7.26495147e-01 -1.31234571e-01 -4.37231332e-01 -5.72756469e-01
1.01449192e-01 3.40761930e-01 2.30131894e-01 -7.47660279e-01
1.23408079e-01 -5.59383392e-01 7.06284285e-01 -3.87960762e-01
4.31372553e-01 -1.03688478e+00 3.21731746e-01 5.53544760e-01
-7.71124363e-02 1.60479471e-02 5.65809850e-03 4.99023318e-01
-5.19584894e-01 2.14030638e-01 7.02195168e-01 1.04026338e-02
-1.24774837e+00 5.35511672e-01 -2.09334210e-01 -5.51644266e-01
1.05483329e+00 -4.43347037e-01 -3.68157327e-01 -1.49616599e-01
-8.13537121e-01 -2.92606503e-01 4.14860815e-01 2.95031488e-01
1.21298563e+00 -1.67819154e+00 -3.98476452e-01 6.03852272e-01
3.00535321e-01 -8.21164131e-01 4.94625837e-01 8.42723250e-01
-1.35439301e+00 1.72956288e-01 -7.75121927e-01 -5.06260157e-01
-1.46630967e+00 3.07793152e-02 3.06232125e-01 9.79549065e-02
-5.27319252e-01 7.71427572e-01 -2.31676288e-02 1.19405322e-01
-1.02897488e-01 -1.51617944e-01 -4.80590075e-01 5.12628108e-02
4.55807835e-01 6.98723078e-01 1.98605657e-01 -7.80679047e-01
-2.66900480e-01 1.18132150e+00 3.00513148e-01 -2.28505611e-01
9.59399343e-01 -3.84416938e-01 -6.31973326e-01 4.60612625e-01
1.53803968e+00 1.14202395e-01 -8.11864436e-01 -2.88254023e-02
-1.17150865e-01 -1.08312058e+00 1.93762407e-01 -6.57742560e-01
-1.57630789e+00 9.81635213e-01 8.82683873e-01 5.16466439e-01
1.61435723e+00 -6.93664312e-01 6.54728055e-01 2.76125550e-01
7.99896359e-01 -1.19657242e+00 -6.02104142e-02 5.58418572e-01
7.48237729e-01 -8.97939444e-01 -1.26075849e-01 -8.10043514e-01
-6.79509163e-01 1.87475586e+00 3.92368942e-01 4.04206384e-03
8.46189260e-01 -6.98880106e-02 2.54863113e-01 -2.52950490e-01
-7.87656978e-02 -1.09829083e-01 4.02405143e-01 5.45705557e-01
5.10779619e-01 4.03323561e-01 -5.89778244e-01 1.52114868e-01
4.68227267e-02 -2.17053235e-01 4.16314870e-01 9.66189921e-01
-4.67318207e-01 -1.02286398e+00 -6.56385720e-01 4.93965834e-01
-2.34829798e-01 1.62471220e-01 -2.74543077e-01 8.12749863e-01
2.15220917e-02 1.09666800e+00 1.49612606e-01 -8.95109177e-01
2.62844563e-01 -2.84525394e-01 7.05438197e-01 2.35302031e-01
2.96586663e-01 1.52700245e-01 -7.48983677e-03 -7.70430863e-01
-3.05421352e-01 -1.68235675e-01 -1.14995706e+00 -6.68538570e-01
-2.00712159e-01 2.59788454e-01 9.70068395e-01 5.26847720e-01
-1.12266436e-01 3.16400111e-01 1.13577318e+00 -4.96468127e-01
3.26479793e-01 -4.56684470e-01 -1.47560287e+00 5.87430298e-01
-3.69621255e-02 -7.33204603e-01 -4.43865173e-03 1.22097678e-01] | [10.388235092163086, -0.3999563455581665] |
b926f0fe-6a2e-4554-8766-36e4c4b43c4b | composing-pick-and-place-tasks-by-grounding | 2102.08094 | null | https://arxiv.org/abs/2102.08094v1 | https://arxiv.org/pdf/2102.08094v1.pdf | Composing Pick-and-Place Tasks By Grounding Language | Controlling robots to perform tasks via natural language is one of the most challenging topics in human-robot interaction. In this work, we present a robot system that follows unconstrained language instructions to pick and place arbitrary objects and effectively resolves ambiguities through dialogues. Our approach infers objects and their relationships from input images and language expressions and can place objects in accordance with the spatial relations expressed by the user. Unlike previous approaches, we consider grounding not only for the picking but also for the placement of everyday objects from language. Specifically, by grounding objects and their spatial relations, we allow specification of complex placement instructions, e.g. "place it behind the middle red bowl". Our results obtained using a real-world PR2 robot demonstrate the effectiveness of our method in understanding pick-and-place language instructions and sequentially composing them to solve tabletop manipulation tasks. Videos are available at http://speechrobot.cs.uni-freiburg.de | ['Wolfram Burgard', 'Oier Mees'] | 2021-02-16 | null | null | null | null | ['natural-language-visual-grounding'] | ['reasoning'] | [ 9.83200669e-02 3.32008988e-01 1.19403258e-01 -5.12483299e-01
-1.99290186e-01 -9.56229925e-01 4.51584101e-01 8.92015547e-02
-2.23524868e-01 7.10356534e-01 -1.39397070e-01 -2.87918597e-01
-2.78524488e-01 -4.12016422e-01 -7.17424810e-01 -2.57651061e-01
9.84691307e-02 1.05175090e+00 9.35723856e-02 -6.27586007e-01
5.29164135e-01 7.09424436e-01 -1.49296248e+00 2.93765932e-01
7.08399773e-01 3.87279689e-01 1.26044142e+00 7.65876472e-01
-2.46069461e-01 9.30441320e-01 -6.48441672e-01 2.87920475e-01
1.61354095e-01 1.00977556e-03 -1.27214015e+00 3.38576049e-01
1.83854904e-02 -5.87698102e-01 -2.48911679e-02 1.08893394e+00
-1.91555366e-01 3.10383946e-01 5.86778104e-01 -1.55501795e+00
-5.56846023e-01 9.32916343e-01 -5.59373014e-02 -6.50813460e-01
1.17980862e+00 3.12868282e-02 9.25647914e-01 -4.40899968e-01
8.23965490e-01 1.65593433e+00 4.50301133e-02 5.42667866e-01
-1.07753611e+00 -2.65964985e-01 3.61961216e-01 2.09061012e-01
-1.54885399e+00 -3.10969204e-01 4.60869849e-01 -6.93899989e-01
9.68387663e-01 2.06140101e-01 3.25285524e-01 6.70366049e-01
2.17716277e-01 7.51598239e-01 6.49790466e-01 -8.16170096e-01
1.12465061e-01 1.04436092e-01 2.28009075e-01 6.89173043e-01
1.48012564e-01 -4.45711553e-01 -3.23704123e-01 8.25647917e-03
1.35621309e+00 -6.76760748e-02 -2.40925685e-01 -8.58841717e-01
-1.63376522e+00 2.95730293e-01 3.54497761e-01 4.94125128e-01
-4.44792897e-01 3.53210747e-01 -9.05075762e-03 3.56081612e-02
-3.90761226e-01 9.39363897e-01 -5.22989213e-01 -2.58349627e-01
-1.11511594e-03 5.85302770e-01 1.04787266e+00 1.82477140e+00
6.96229696e-01 -6.66813910e-01 -1.72018576e-02 7.09668100e-01
3.16147327e-01 2.46853530e-01 1.40047222e-01 -1.55091870e+00
6.98618889e-01 5.69583535e-01 1.06006348e+00 -9.44822431e-01
-6.41839147e-01 6.89087391e-01 -2.05318049e-01 3.08672935e-01
4.88430798e-01 -1.02695763e-01 -6.64781034e-01 1.41341305e+00
3.79420549e-01 -4.97097522e-01 2.96684682e-01 1.24186325e+00
5.33108175e-01 6.31788492e-01 9.89244133e-02 1.01726241e-01
1.85493171e+00 -9.97263432e-01 -1.18712163e+00 -3.68667215e-01
7.47016549e-01 -6.48696721e-01 1.18554211e+00 6.24136090e-01
-8.51452768e-01 -6.24440908e-01 -9.15050209e-01 -6.83056593e-01
-3.19695354e-01 6.42158568e-01 8.86844814e-01 -2.74532229e-01
-1.06476080e+00 3.95049423e-01 -8.47879827e-01 -7.68371761e-01
-3.10941100e-01 7.34125912e-01 -4.75500464e-01 -6.52581453e-02
-7.17744827e-01 1.05856812e+00 6.76576376e-01 3.46762449e-01
-3.77342761e-01 1.19072318e-01 -1.13539851e+00 -9.53129977e-02
6.66973293e-01 -5.06619453e-01 2.01204515e+00 -4.27151382e-01
-2.03784299e+00 7.90488183e-01 -1.31998375e-01 -2.48238578e-01
3.92204463e-01 -6.40883803e-01 1.97587922e-01 2.14506805e-01
5.76510131e-01 1.11278725e+00 2.23721594e-01 -1.49531925e+00
-7.18302667e-01 -4.01993185e-01 6.12585366e-01 5.66861570e-01
6.11307919e-01 3.49405780e-02 -5.41671753e-01 -1.15354106e-01
6.14110708e-01 -1.30620182e+00 -2.26084694e-01 1.53818548e-01
-7.44953215e-01 -4.82842028e-01 6.10300362e-01 -3.65871459e-01
5.31349659e-01 -2.19363117e+00 5.36290705e-01 -3.97962257e-02
1.98600050e-02 -1.94155246e-01 -7.91281983e-02 5.35020947e-01
1.64874002e-01 -1.69917792e-01 3.32200006e-02 -3.26175570e-01
4.61127937e-01 6.27763569e-01 -2.60654807e-01 2.32219145e-01
2.10186720e-01 6.62349224e-01 -1.16582096e+00 -3.85276645e-01
6.43452525e-01 2.71197915e-01 -4.18271124e-01 4.68435377e-01
-6.19427145e-01 7.80563354e-01 -7.49836087e-01 3.74179482e-01
3.88679862e-01 4.13464196e-02 6.19287908e-01 1.20269835e-01
-3.24079812e-01 4.79824454e-01 -1.31831574e+00 1.91455936e+00
-5.98311782e-01 4.16280806e-01 5.82213879e-01 -5.05295455e-01
8.57599676e-01 2.95370758e-01 1.54733971e-01 -1.16851903e-01
1.06278196e-01 2.84655094e-01 -8.48183036e-02 -1.00107741e+00
8.22309792e-01 3.00101399e-01 -4.82582569e-01 1.83547243e-01
-2.59725988e-01 -1.11348271e+00 2.79594570e-01 6.93957582e-02
7.63313055e-01 6.96720302e-01 7.16234624e-01 -1.69395640e-01
3.28720927e-01 3.58949691e-01 -6.49377257e-02 6.37730241e-01
-2.08787005e-02 3.16989601e-01 3.80660325e-01 -4.40319806e-01
-7.29966998e-01 -9.82620537e-01 3.13919902e-01 1.31483877e+00
6.86025679e-01 -2.94047207e-01 -8.14710200e-01 -2.06228048e-02
-1.67913198e-01 1.12069201e+00 -1.36377558e-01 4.96909469e-01
-7.70725727e-01 2.66075552e-01 1.45504205e-03 2.76833743e-01
2.74925768e-01 -1.53235126e+00 -1.32541847e+00 1.45677164e-01
-4.77661312e-01 -1.57788968e+00 -3.76403511e-01 2.93325663e-01
-4.50105876e-01 -1.16294134e+00 -2.43829206e-01 -1.21371996e+00
1.10350442e+00 4.34517860e-01 8.47347975e-01 2.36246809e-02
-1.37185633e-01 5.77008665e-01 -4.77575123e-01 -4.94557083e-01
-4.29759204e-01 1.03608176e-01 1.85692251e-01 -3.88527989e-01
1.60003275e-01 -3.12069446e-01 -3.87387313e-02 4.74534482e-01
-5.21932423e-01 5.86913466e-01 4.18656558e-01 3.10409099e-01
2.89282441e-01 4.81114499e-02 -1.59702703e-01 -5.17928839e-01
8.88947487e-01 -3.59892473e-02 -7.10686684e-01 2.27895021e-01
2.57767677e-01 1.14252612e-01 2.76177824e-01 -4.66036409e-01
-9.10380363e-01 7.43097603e-01 4.97413099e-01 4.76116426e-02
-8.08332503e-01 1.39530143e-03 -2.75768965e-01 2.94825435e-01
2.84151226e-01 -1.88099489e-01 -1.10926561e-01 -3.03324848e-01
6.45317554e-01 8.06586742e-01 8.50390494e-01 -1.15430713e+00
3.73615295e-01 2.33296812e-01 -1.01979390e-01 -8.67734969e-01
-3.02022785e-01 -4.66042638e-01 -1.16403866e+00 -1.55127734e-01
1.04794157e+00 -5.37908435e-01 -1.50960159e+00 1.45238996e-01
-1.85649157e+00 -6.58114374e-01 1.79789290e-02 4.60938811e-01
-1.05569828e+00 5.03627211e-03 -5.54744124e-01 -1.13313913e+00
2.41397709e-01 -1.51306427e+00 1.58168256e+00 6.71386719e-02
-9.85178292e-01 -2.85682827e-01 -5.54158568e-01 1.25261649e-01
-1.08303472e-01 8.66421014e-02 8.11073780e-01 -3.42859298e-01
-7.26001322e-01 -6.07378222e-02 -1.58620626e-01 -5.36832035e-01
5.54823041e-01 -7.62903690e-02 -2.79813290e-01 2.97075927e-01
-1.63788110e-01 -1.15575023e-01 -2.68955380e-01 1.34375900e-01
8.71190369e-01 -3.94986033e-01 -6.32980287e-01 -9.23399851e-02
9.20519173e-01 7.88785219e-01 5.85726261e-01 4.37884629e-01
6.66669369e-01 1.11283386e+00 1.10282421e+00 2.11376607e-01
6.70514405e-01 1.17407691e+00 3.38803232e-01 5.14494479e-01
4.67785388e-01 -2.13539824e-01 2.40504056e-01 2.71006972e-01
-3.20253149e-02 -1.66813090e-01 -1.00363171e+00 3.07323098e-01
-2.11666107e+00 -5.01624048e-01 -2.80620068e-01 1.64076543e+00
7.14321256e-01 -1.69991598e-01 -2.32096031e-01 -1.24522395e-01
8.69989097e-01 -4.91687357e-01 -1.88327938e-01 -4.83280152e-01
5.06032646e-01 -2.57213771e-01 2.36356735e-01 1.03893745e+00
-1.06036043e+00 1.41081500e+00 5.31081343e+00 4.89942282e-02
-8.09240580e-01 -2.85140961e-01 -5.98875023e-02 1.87977239e-01
3.95616323e-01 -6.75502941e-02 -7.11215615e-01 -1.89084616e-02
6.79481849e-02 3.12359512e-01 8.49324465e-01 8.01540017e-01
3.78967583e-01 -5.33632517e-01 -1.84172797e+00 1.10488272e+00
-1.63326859e-01 -7.16415107e-01 -1.24581896e-01 -3.41124296e-01
5.98369464e-02 -5.26164532e-01 -3.11505765e-01 2.98633456e-01
6.36240244e-01 -8.66396248e-01 1.29251993e+00 4.54697400e-01
3.96835595e-01 -1.05757840e-01 3.62548292e-01 7.72907138e-01
-9.62920845e-01 -1.72605187e-01 -1.30397543e-01 -6.15054786e-01
4.19947535e-01 -1.11025132e-01 -1.43919730e+00 3.11552018e-01
5.89996219e-01 1.95279598e-01 1.48643509e-01 4.52049732e-01
-8.64087641e-01 -5.77942610e-01 -5.10072947e-01 -5.13759136e-01
1.16744064e-01 -5.66370904e-01 3.52555007e-01 8.85812700e-01
2.38899291e-01 7.28506386e-01 5.75189352e-01 1.04667699e+00
3.34380090e-01 -3.13508324e-02 -7.72194684e-01 1.31549880e-01
6.04699135e-01 7.69551158e-01 -8.16808522e-01 -3.70996058e-01
1.55020162e-01 1.30121803e+00 3.35532516e-01 4.11122173e-01
-5.96730113e-01 -5.91989100e-01 5.98885834e-01 2.51993220e-02
6.63360059e-02 -1.06791437e+00 -4.60797325e-02 -7.19947338e-01
2.68693537e-01 -7.19654262e-01 -2.34996259e-01 -1.55660939e+00
-6.06306374e-01 5.34684718e-01 5.54852247e-01 -1.13934946e+00
-4.11844730e-01 -1.06960678e+00 -9.44377184e-02 5.86562097e-01
-1.00377226e+00 -9.01363015e-01 -4.27478671e-01 1.22660957e-01
9.20549393e-01 4.74612087e-01 9.94830906e-01 -3.00689459e-01
-1.23021252e-01 -3.82505625e-01 -6.74223542e-01 4.45191152e-02
4.60193217e-01 -1.19418263e+00 2.60651201e-01 1.48151770e-01
1.18731953e-01 1.04547060e+00 8.47852290e-01 -6.58269405e-01
-1.60047376e+00 -4.25545424e-01 9.35486436e-01 -6.29234016e-01
5.44914305e-01 -7.33844697e-01 -5.88181674e-01 1.23802221e+00
2.46712998e-01 -3.41188222e-01 1.99912246e-02 -3.11161548e-01
5.68653606e-02 2.51406550e-01 -1.18173516e+00 1.11452389e+00
1.25943053e+00 -3.26665789e-01 -1.00353587e+00 7.35487819e-01
8.96184921e-01 -1.09875762e+00 -2.39207968e-01 1.59097910e-01
5.30440331e-01 -4.31629360e-01 9.02267814e-01 -4.23716724e-01
1.99892491e-01 -8.51080894e-01 -3.29735637e-01 -1.11584997e+00
-2.51550317e-01 -8.26609910e-01 4.35726047e-01 9.09427583e-01
3.83211076e-01 -5.91877580e-01 2.85869926e-01 1.09595525e+00
-1.16396390e-01 -8.10227245e-02 -5.72760940e-01 -4.53497678e-01
-5.40269315e-01 -4.25934255e-01 5.75362802e-01 6.78712785e-01
7.95132339e-01 3.69325817e-01 1.11540137e-02 7.54850149e-01
8.93937573e-02 6.04583807e-02 1.10697532e+00 -1.03912807e+00
-2.55822744e-02 -2.83454746e-01 -1.79554641e-01 -1.58123684e+00
6.08856797e-01 -4.50643599e-01 9.94314492e-01 -1.77383482e+00
-3.95605117e-01 -6.61890507e-01 6.01624608e-01 8.23572934e-01
2.99636662e-01 -3.56790334e-01 6.00023806e-01 2.45041877e-01
-7.11714864e-01 2.24959135e-01 1.29564905e+00 -1.32099956e-01
-6.37913465e-01 -1.67467028e-01 -2.37586498e-01 9.18550730e-01
9.58861232e-01 -1.51601642e-01 -3.67929786e-01 -8.10594440e-01
2.31986307e-02 2.81214058e-01 3.20375323e-01 -9.61027384e-01
6.27146721e-01 -6.44434988e-01 -9.10109133e-02 -3.30172747e-01
6.51218295e-01 -1.32886946e+00 1.73535168e-01 4.41258430e-01
-5.18071115e-01 1.89405382e-01 2.61248916e-01 2.52717827e-02
1.46921933e-01 -4.24841285e-01 7.82977417e-02 -4.63876307e-01
-1.04636264e+00 -6.20942533e-01 -7.85199642e-01 -7.76234150e-01
1.35748136e+00 1.05439099e-02 -2.18235418e-01 -4.98020649e-01
-9.80140328e-01 5.48037887e-01 4.21877652e-01 7.19886720e-01
5.75680435e-01 -9.67170656e-01 -1.64547890e-01 1.41299337e-01
8.19514319e-02 6.71605170e-01 -3.53339404e-01 3.81406546e-01
-1.12380338e+00 6.34247124e-01 -2.35170305e-01 -7.71344364e-01
-1.18636930e+00 5.94018459e-01 2.14986622e-01 2.50937372e-01
-5.44642031e-01 6.07663810e-01 4.82636780e-01 -8.37095559e-01
3.24993312e-01 -1.05599570e+00 -2.03428671e-01 -5.90963602e-01
3.02089185e-01 -6.71409965e-02 -3.85572016e-01 -5.77532530e-01
-4.86842573e-01 8.41153860e-01 1.45580336e-01 -3.44210535e-01
1.01830697e+00 -3.44672471e-01 -6.15139604e-01 7.07616806e-01
4.91195291e-01 -4.79343683e-02 -9.78250086e-01 1.47339568e-01
3.84646863e-01 -2.69915909e-01 -8.35515201e-01 -6.90422654e-01
-7.68427476e-02 5.09922743e-01 -1.07208215e-01 2.36451522e-01
5.75282693e-01 3.21983039e-01 1.35551140e-01 1.19136274e+00
1.27486646e+00 -9.13825452e-01 1.48707598e-01 8.57260585e-01
1.42994726e+00 -9.77119267e-01 1.62736196e-02 -1.23887587e+00
-7.18086183e-01 1.27128804e+00 9.43294346e-01 -4.17181849e-02
6.68698400e-02 5.68536758e-01 1.80693880e-01 -2.72495806e-01
-3.90141636e-01 -1.15654024e-03 -2.77838767e-01 6.88418686e-01
3.27986062e-01 5.24471879e-01 -1.40314326e-01 4.36192513e-01
-5.93107462e-01 5.73763028e-02 6.96794629e-01 1.42713785e+00
-5.25007546e-01 -9.29087758e-01 -8.55697334e-01 -2.57943898e-01
1.78897128e-01 3.89688313e-01 -7.20587313e-01 9.25295293e-01
2.21342370e-01 1.18396354e+00 2.54038185e-01 -9.74778384e-02
7.27624238e-01 -5.41040935e-02 7.70537734e-01 -1.12669039e+00
-2.04353094e-01 -3.11585486e-01 1.54322550e-01 -6.39878869e-01
-4.06205773e-01 -6.40527189e-01 -2.07006431e+00 3.77140015e-01
-3.79606821e-02 1.78128526e-01 8.82843018e-01 1.06175399e+00
2.63283074e-01 4.40048069e-01 1.21984221e-02 -1.72166550e+00
-3.29479069e-01 -9.69469309e-01 -2.26763740e-01 3.05528164e-01
4.25869405e-01 -8.26705694e-01 9.30389538e-02 2.70029455e-01] | [4.587829113006592, 0.7614148855209351] |
3f87e15a-1246-4449-b25c-68aedcbc68a4 | enhanced-prototypical-learning-for | 2205.11419 | null | https://arxiv.org/abs/2205.11419v1 | https://arxiv.org/pdf/2205.11419v1.pdf | Enhanced Prototypical Learning for Unsupervised Domain Adaptation in LiDAR Semantic Segmentation | Despite its importance, unsupervised domain adaptation (UDA) on LiDAR semantic segmentation is a task that has not received much attention from the research community. Only recently, a completion-based 3D method has been proposed to tackle the problem and formally set up the adaptive scenarios. However, the proposed pipeline is complex, voxel-based and requires multi-stage inference, which inhibits it for real-time inference. We propose a range image-based, effective and efficient method for solving UDA on LiDAR segmentation. The method exploits class prototypes from the source domain to pseudo label target domain pixels, which is a research direction showing good performance in UDA for natural image semantic segmentation. Applying such approaches to LiDAR scans has not been considered because of the severe domain shift and lack of pre-trained feature extractor that is unavailable in the LiDAR segmentation setup. However, we show that proper strategies, including reconstruction-based pre-training, enhanced prototypes, and selective pseudo labeling based on distance to prototypes, is sufficient enough to enable the use of prototypical approaches. We evaluate the performance of our method on the recently proposed LiDAR segmentation UDA scenarios. Our method achieves remarkable performance among contemporary methods. | ['Junmo Kim', 'JuYoung Yang', 'Eojindl Yi'] | 2022-05-23 | null | null | null | null | ['lidar-semantic-segmentation'] | ['computer-vision'] | [ 5.77721238e-01 6.78188130e-02 -1.40782475e-01 -5.88714123e-01
-6.11511350e-01 -5.25278628e-01 7.79619098e-01 1.07091248e-01
-5.29015779e-01 5.86849570e-01 -4.74974424e-01 -2.79742002e-01
-3.58821273e-01 -8.72657597e-01 -5.70380688e-01 -4.55595940e-01
1.69922128e-01 1.27058649e+00 8.01962435e-01 1.23703312e-02
4.38831449e-01 1.01596022e+00 -2.02445555e+00 -1.11458756e-01
1.09347773e+00 8.22272420e-01 6.01999581e-01 3.21716845e-01
-7.48538494e-01 -1.00408010e-01 -1.77665457e-01 4.58956100e-02
5.76521218e-01 -3.37983847e-01 -8.43919337e-01 3.99950147e-01
7.14681029e-01 -9.73288268e-02 4.51610476e-01 8.40464056e-01
5.27910352e-01 2.01827511e-01 1.06102824e+00 -9.73677576e-01
5.16645089e-02 3.31061959e-01 -5.02534032e-01 -3.01789522e-01
8.22268650e-02 1.58325918e-02 7.71483064e-01 -8.70629549e-01
9.10028458e-01 1.04501534e+00 6.46591425e-01 3.94111872e-01
-1.25207365e+00 -5.70493877e-01 -1.72625743e-02 2.51178950e-01
-1.41856956e+00 -1.10204570e-01 9.86988604e-01 -6.21825397e-01
6.21562183e-01 7.77838975e-02 7.56790042e-01 7.95167804e-01
-5.28008282e-01 6.31865561e-01 1.42372978e+00 -6.07092083e-01
7.52316654e-01 4.16344881e-01 1.66728467e-01 4.33190793e-01
2.02367410e-01 -8.71776044e-02 -1.95103660e-01 2.63416380e-01
6.67635322e-01 -3.33371669e-01 1.27223819e-01 -8.59993637e-01
-8.44655812e-01 7.09635258e-01 3.50250125e-01 3.55097413e-01
-4.19662118e-01 -2.66800731e-01 2.37199590e-01 -8.69023055e-02
4.97122318e-01 3.89837414e-01 -5.15543401e-01 1.08841069e-01
-1.42884302e+00 2.03559369e-01 5.49279869e-01 9.76953566e-01
1.05426443e+00 -1.10276721e-01 1.53557315e-01 8.66747439e-01
1.68887123e-01 5.92427492e-01 3.76311876e-02 -9.95696783e-01
1.44583181e-01 8.18507493e-01 -7.93869570e-02 -6.44113600e-01
-4.80892688e-01 -2.80456692e-01 -4.95017946e-01 4.26167428e-01
6.50252402e-01 3.24244589e-01 -1.13429987e+00 1.11893499e+00
7.97873795e-01 2.98490971e-01 1.27661796e-02 8.34114671e-01
4.74937767e-01 3.58646125e-01 7.28024393e-02 -2.27809936e-01
1.13558245e+00 -5.55179417e-01 -2.51144946e-01 -1.12503596e-01
4.42709863e-01 -7.78073609e-01 1.24895978e+00 5.31936824e-01
-7.10874677e-01 -6.85706019e-01 -9.41501677e-01 -5.68723045e-02
-5.76591194e-01 6.30047694e-02 5.92821419e-01 9.28233802e-01
-7.95119464e-01 4.78486270e-01 -6.67215049e-01 -8.02407146e-01
6.12347960e-01 4.19028908e-01 -2.29680032e-01 -2.51336098e-01
-8.29880655e-01 8.75093102e-01 4.99058038e-01 7.15031177e-02
-6.62324905e-01 -5.77434003e-01 -7.12437868e-01 -3.42171639e-01
5.42362094e-01 -6.28844500e-01 9.80470359e-01 -6.44741178e-01
-1.58077395e+00 1.07069635e+00 -7.33943433e-02 -4.90183264e-01
7.12758005e-01 1.19048934e-02 3.32598127e-02 2.25227982e-01
2.43105456e-01 1.03302646e+00 9.95426476e-01 -1.53153598e+00
-6.83107913e-01 -5.51233172e-01 -5.88186495e-02 9.78571922e-02
-1.56714112e-01 -4.02274072e-01 -3.75030130e-01 -1.49803117e-01
4.21205223e-01 -9.57284987e-01 -3.67480785e-01 5.79362847e-02
-2.19328433e-01 -2.23010615e-01 1.00479972e+00 -2.59461105e-01
7.03225315e-01 -2.03454566e+00 -1.74869314e-01 4.42906797e-01
-3.69257838e-01 4.42713529e-01 1.11327112e-01 2.44482666e-01
2.06983164e-01 -9.87668410e-02 -1.00459635e+00 -3.29868138e-01
6.10426106e-02 5.28106511e-01 -2.17566341e-01 3.76304150e-01
2.25840300e-01 4.30726349e-01 -7.15447187e-01 -9.69932079e-01
7.19310224e-01 5.13440907e-01 -6.76571071e-01 1.73256725e-01
-6.73391581e-01 7.27977693e-01 -6.66825652e-01 7.32215285e-01
1.24043810e+00 3.40165198e-01 7.36548826e-02 -2.31875449e-01
-5.67717850e-01 -8.43162555e-03 -1.58706856e+00 1.96322608e+00
-5.67638278e-01 3.50166917e-01 8.86283815e-02 -1.29917836e+00
1.16122615e+00 -1.50745019e-01 6.03931606e-01 -6.13069355e-01
3.40416990e-02 6.05603397e-01 -1.05040751e-01 -5.44082165e-01
5.93001842e-01 -2.41201818e-01 2.28091720e-02 2.20544696e-01
9.87244621e-02 -9.05259907e-01 3.01401645e-01 -3.13169770e-02
6.96074128e-01 9.02214229e-01 2.73016304e-01 -3.61190975e-01
7.54802167e-01 5.63505650e-01 4.25653219e-01 6.43148839e-01
-5.31914011e-02 8.39830697e-01 -1.78755689e-02 -3.93726528e-02
-1.02455914e+00 -1.05921435e+00 -5.96070170e-01 8.31925154e-01
2.63819307e-01 1.00345083e-01 -9.74061847e-01 -6.31214857e-01
9.34981480e-02 8.61777186e-01 -2.56299675e-01 3.33546877e-01
-6.53874218e-01 -5.61754763e-01 3.45940590e-01 3.26361179e-01
7.41997719e-01 -1.03163791e+00 -8.32854211e-01 1.49758145e-01
1.54877961e-01 -1.29445326e+00 1.25998527e-01 2.22481504e-01
-1.21657038e+00 -1.08640862e+00 -6.68472946e-01 -6.53667688e-01
6.08393073e-01 8.64046738e-02 8.52513790e-01 -2.73197234e-01
-4.08983827e-01 6.04284465e-01 -2.37502262e-01 -4.04462218e-01
-2.71852702e-01 3.48667294e-01 -2.49536112e-01 2.39768159e-03
2.10732967e-01 -7.32743979e-01 -4.04364347e-01 3.90307724e-01
-9.81768966e-01 -9.96932536e-02 6.82557583e-01 3.52527618e-01
9.73323524e-01 -6.60698563e-02 4.98820841e-01 -1.16266656e+00
1.93987265e-01 -1.34555459e-01 -8.83898437e-01 2.46985517e-02
-6.98885262e-01 1.20342568e-01 3.95393908e-01 -3.32664758e-01
-1.23324740e+00 6.87410474e-01 -3.90027463e-01 -2.38260120e-01
-8.55473757e-01 2.55182218e-02 -4.59029883e-01 -3.99334244e-02
6.01442218e-01 1.50480837e-01 -1.69960812e-01 -6.47721708e-01
6.26257002e-01 7.06131220e-01 5.78055501e-01 -6.84952974e-01
9.62615967e-01 7.46113300e-01 3.63432825e-01 -1.21862423e+00
-5.84929526e-01 -7.71931350e-01 -1.25673306e+00 -2.15624571e-01
9.70723927e-01 -6.55481696e-01 -1.35577694e-01 1.44240782e-01
-1.14863575e+00 -3.67908329e-01 -6.10896766e-01 4.72157210e-01
-7.21235156e-01 5.06384909e-01 1.06416963e-01 -9.97814238e-01
-1.27433129e-02 -1.25428915e+00 1.22408390e+00 1.36675701e-01
2.58703344e-02 -7.95103192e-01 -1.04948744e-01 5.22260129e-01
2.66274840e-01 2.33044699e-01 7.92160809e-01 -5.61977327e-01
-8.92065287e-01 1.14009097e-01 -2.18977675e-01 4.21846151e-01
-3.34232226e-02 1.44337475e-01 -1.18631089e+00 2.91221708e-01
-9.00678709e-02 -1.28354311e-01 7.17864633e-01 3.18919152e-01
1.18351424e+00 5.34756243e-01 -4.93988156e-01 4.59889621e-01
1.33263803e+00 1.14150584e-01 4.30881351e-01 3.86270165e-01
5.26305854e-01 8.43892157e-01 1.10577321e+00 2.05485567e-01
3.65088135e-01 8.14376831e-01 5.60105085e-01 -1.00713491e-01
-4.37534839e-01 -7.09413067e-02 4.61097285e-02 5.30485988e-01
-2.08074495e-01 9.68897715e-02 -1.05550766e+00 6.01928592e-01
-1.64586198e+00 -5.79982817e-01 -6.12243354e-01 2.19833803e+00
4.81025577e-01 3.57153505e-01 8.10807496e-02 3.03727925e-01
6.28388762e-01 -1.33353561e-01 -5.38522601e-01 -5.38479030e-01
5.59245460e-02 6.58348799e-01 4.70145077e-01 5.91516852e-01
-1.03380036e+00 1.22582448e+00 5.63379717e+00 1.00425196e+00
-1.04484797e+00 1.81225941e-01 -3.24696414e-02 5.58133185e-01
-3.34370792e-01 2.20546439e-01 -8.49321485e-01 2.38245785e-01
6.01234198e-01 2.91116208e-01 1.06803030e-01 1.12470150e+00
2.42160663e-01 -5.50778687e-01 -9.02438879e-01 8.02558780e-01
-1.06215432e-01 -9.56222296e-01 -6.95002079e-02 3.37435752e-02
5.99878490e-01 -1.61516711e-01 -2.49952778e-01 3.42195064e-01
-8.24661404e-02 -6.39349580e-01 5.24300635e-01 3.29175383e-01
7.05160916e-01 -5.72872043e-01 4.03564304e-01 6.20221496e-01
-1.19849837e+00 1.58877775e-01 -5.83485425e-01 -2.11499836e-02
2.97713667e-01 9.25177217e-01 -1.36296761e+00 8.09960127e-01
4.63937640e-01 4.34786618e-01 -6.47343993e-01 9.53989565e-01
-1.97978497e-01 6.55254006e-01 -7.11982250e-01 1.62343681e-01
3.51133138e-01 -5.56650281e-01 4.95458126e-01 1.38002348e+00
3.43264997e-01 -1.03920177e-01 3.31406116e-01 9.24403191e-01
2.61063367e-01 2.92977124e-01 -8.09134007e-01 2.79261500e-01
3.55954647e-01 1.41421199e+00 -1.33750570e+00 -2.47710407e-01
4.33646515e-03 7.91425884e-01 1.77203561e-03 -1.11224409e-02
-7.67442822e-01 -8.04377571e-02 3.53387654e-01 5.40839434e-01
3.70426863e-01 -5.30475259e-01 -5.60726285e-01 -6.62040114e-01
-1.12749383e-01 -2.55338669e-01 2.69544661e-01 -6.74868643e-01
-1.02858150e+00 4.14764702e-01 5.68619311e-01 -1.28344512e+00
-3.61568987e-01 -6.02109492e-01 -3.64464492e-01 5.51010907e-01
-1.68397319e+00 -1.38185143e+00 -3.87965888e-01 5.83697200e-01
7.00044692e-01 1.61006935e-02 6.38720036e-01 4.73552793e-01
-3.24144840e-01 2.99260318e-02 -1.80590361e-01 -3.82264018e-01
6.54631078e-01 -1.31888330e+00 2.95787491e-02 7.68260419e-01
1.36107162e-01 2.15171427e-01 7.03859091e-01 -6.94416821e-01
-1.07281506e+00 -1.01875246e+00 7.29286134e-01 -4.14748967e-01
2.93539733e-01 -4.61507946e-01 -9.95315909e-01 3.45231503e-01
-6.59195185e-02 3.08640227e-02 3.37358177e-01 -7.71053210e-02
4.37372876e-03 -2.58065104e-01 -1.35651863e+00 3.24348092e-01
1.32719302e+00 -3.49773556e-01 -7.77373910e-01 2.12664366e-01
4.17456925e-01 -1.86638966e-01 -6.57477140e-01 7.61461079e-01
2.83165485e-01 -1.23794222e+00 1.09194946e+00 7.62143731e-02
8.27276185e-02 -5.69994926e-01 -1.42096534e-01 -8.78289223e-01
2.71398515e-01 5.49020991e-02 4.12671238e-01 1.49032819e+00
2.23105863e-01 -6.00530982e-01 9.44693029e-01 3.34786743e-01
-4.22995865e-01 -3.98295015e-01 -1.00195324e+00 -8.13481569e-01
5.24653122e-03 -7.98277915e-01 5.21037042e-01 9.05189335e-01
-7.69200563e-01 2.32819408e-01 2.84634769e-01 2.36692131e-01
7.41354704e-01 7.02263832e-01 1.12554801e+00 -1.55133009e+00
1.01684175e-01 -3.97931010e-01 -4.13912177e-01 -9.47997510e-01
3.07884663e-01 -1.10450983e+00 1.95688471e-01 -1.73069608e+00
-2.55820870e-01 -9.63940144e-01 2.29950562e-01 1.93252370e-01
2.60786831e-01 2.30467677e-01 3.82267013e-02 1.07505411e-01
-2.81649232e-01 4.71332878e-01 1.20843506e+00 -1.13392420e-01
-2.52167910e-01 1.05550267e-01 -1.58136368e-01 8.12759399e-01
7.65938759e-01 -6.21887982e-01 -6.24623299e-01 -7.46091306e-02
-1.93020433e-01 -3.23030263e-01 2.46390611e-01 -1.20291650e+00
1.22119598e-01 -1.99424610e-01 4.74012271e-02 -9.97587562e-01
5.11573017e-01 -1.17822766e+00 1.65001631e-01 3.27466339e-01
8.28180686e-02 -4.54805970e-01 -2.73044929e-02 3.87220412e-01
-1.92366242e-01 -5.95343947e-01 9.18448031e-01 -2.54703283e-01
-1.24983704e+00 2.06360444e-01 -3.19744647e-02 -1.38779461e-01
1.06288218e+00 -8.39990079e-01 4.70124185e-01 1.83523506e-01
-6.74159825e-01 -3.13070952e-03 6.36784136e-01 3.92017737e-02
3.78463864e-01 -8.89310300e-01 -4.72667754e-01 1.94733441e-01
1.44639179e-01 7.59570658e-01 2.14241207e-01 7.41357386e-01
-6.17577434e-01 1.97757885e-01 -2.98823893e-01 -1.10988820e+00
-1.03322458e+00 4.54762757e-01 1.48969352e-01 -9.79408771e-02
-6.84236944e-01 4.54877555e-01 -1.29935602e-02 -8.19745362e-01
-4.43861857e-02 -3.74169081e-01 -2.91455626e-01 3.31918597e-01
-1.61879301e-01 4.34860677e-01 3.41852486e-01 -6.51754320e-01
-4.79873091e-01 1.13050210e+00 3.35313350e-01 -2.36794963e-01
1.18513906e+00 -1.50079265e-01 1.17276363e-01 4.86344934e-01
7.75244951e-01 2.56478600e-02 -1.16220498e+00 4.47343402e-02
3.23998928e-01 -4.78614569e-01 8.10160041e-02 -5.96646845e-01
-7.94063807e-01 1.24538827e+00 7.62902498e-01 -1.29761454e-02
1.09704590e+00 1.73725523e-02 5.89157164e-01 2.46271223e-01
6.89546227e-01 -1.37011659e+00 -1.36806622e-01 3.82454842e-01
6.54388189e-01 -1.32220387e+00 2.02291951e-01 -8.38812351e-01
-4.48992938e-01 1.04623830e+00 6.12451017e-01 -1.36230826e-01
6.59647822e-01 1.65372252e-01 3.04317158e-02 -6.41323030e-02
4.13281731e-02 -9.16212678e-01 1.18903898e-01 8.80223691e-01
4.17524800e-02 3.23544033e-02 -5.50865531e-01 9.50678140e-02
-7.60442019e-02 5.85731380e-02 2.76327729e-01 8.94243002e-01
-5.96039593e-01 -1.45920670e+00 -5.26757240e-01 1.36210844e-01
1.34643823e-01 2.89793938e-01 -2.91516989e-01 1.23789811e+00
6.80708170e-01 6.28037095e-01 1.40221760e-01 -2.13917848e-02
5.58601022e-01 3.24359447e-01 7.36574113e-01 -6.98394477e-01
-3.40904981e-01 3.48140933e-02 -1.47134602e-01 -3.04073662e-01
-9.00737464e-01 -8.39116395e-01 -1.54811251e+00 2.34053314e-01
-3.13664228e-01 -1.03511577e-02 1.13132715e+00 9.82605875e-01
1.51324466e-01 1.49468750e-01 4.86494511e-01 -1.14750397e+00
-1.46474823e-01 -8.34939778e-01 -4.40583795e-01 4.00604814e-01
-1.96648195e-01 -1.08386517e+00 -1.09292708e-01 3.11641321e-02] | [8.170758247375488, -2.6629133224487305] |
09a96c43-509a-4eb7-8a3f-9b9d6a1ec680 | maf-multimodal-alignment-framework-for-weakly | 2010.05379 | null | https://arxiv.org/abs/2010.05379v1 | https://arxiv.org/pdf/2010.05379v1.pdf | MAF: Multimodal Alignment Framework for Weakly-Supervised Phrase Grounding | Phrase localization is a task that studies the mapping from textual phrases to regions of an image. Given difficulties in annotating phrase-to-object datasets at scale, we develop a Multimodal Alignment Framework (MAF) to leverage more widely-available caption-image datasets, which can then be used as a form of weak supervision. We first present algorithms to model phrase-object relevance by leveraging fine-grained visual representations and visually-aware language representations. By adopting a contrastive objective, our method uses information in caption-image pairs to boost the performance in weakly-supervised scenarios. Experiments conducted on the widely-adopted Flickr30k dataset show a significant improvement over existing weakly-supervised methods. With the help of the visually-aware language representations, we can also improve the previous best unsupervised result by 5.56%. We conduct ablation studies to show that both our novel model and our weakly-supervised strategies significantly contribute to our strong results. | ['Zhewei Yao', 'Michael W. Mahoney', 'Sheng Shen', 'Hao Tan', 'Qinxin Wang'] | 2020-10-12 | null | https://aclanthology.org/2020.emnlp-main.159 | https://aclanthology.org/2020.emnlp-main.159.pdf | emnlp-2020-11 | ['phrase-grounding'] | ['natural-language-processing'] | [ 4.36317593e-01 1.13722667e-01 -4.07938212e-01 -3.34817499e-01
-1.47252727e+00 -8.19272220e-01 8.26900840e-01 2.13891804e-01
-5.26104271e-01 4.99880791e-01 6.25470638e-01 -1.23277150e-01
2.26927176e-01 -1.97393894e-01 -9.84105945e-01 -5.24952173e-01
1.71141550e-01 2.75897831e-01 2.96325207e-01 -7.45658204e-02
2.74861515e-01 1.10109165e-01 -1.35521567e+00 6.91746712e-01
3.44766647e-01 7.14324713e-01 3.99516165e-01 6.09859765e-01
-8.47130120e-02 7.13371694e-01 -7.18974024e-02 -4.74081516e-01
1.47321030e-01 -2.09527552e-01 -1.08441973e+00 1.95832551e-01
9.01404023e-01 -1.44792870e-01 7.01722652e-02 8.07878792e-01
3.87787253e-01 -2.00366780e-01 5.85999608e-01 -1.18423998e+00
-1.05544102e+00 6.35932565e-01 -9.48234797e-01 -5.97661994e-02
6.29750073e-01 4.17676717e-02 1.47529113e+00 -1.22242260e+00
6.26597941e-01 1.21370316e+00 4.82151777e-01 3.98800790e-01
-1.32951021e+00 -2.81035513e-01 3.32296163e-01 9.50849205e-02
-1.44432008e+00 -4.15735066e-01 7.86899984e-01 -3.68164301e-01
7.78423905e-01 3.39604437e-01 1.51049107e-01 1.12087679e+00
-3.45793188e-01 9.73662674e-01 1.21451020e+00 -8.14799190e-01
-2.09184214e-01 4.49550092e-01 3.15140188e-02 7.25465119e-01
-8.57076980e-03 -1.05581813e-01 -7.77419865e-01 -1.66503981e-01
6.56405985e-01 -3.72540593e-01 -3.40195179e-01 -6.46812379e-01
-1.53006017e+00 7.08131135e-01 7.02305198e-01 2.90705323e-01
-1.80224016e-01 4.83105600e-01 2.05902606e-01 -7.09942803e-02
4.81493264e-01 5.66939771e-01 -2.85039425e-01 8.30130372e-03
-8.54582310e-01 1.03008881e-01 4.46108341e-01 1.08481324e+00
8.98353040e-01 -5.37710667e-01 -4.82594758e-01 8.74888718e-01
7.76638806e-01 7.37776697e-01 4.22059819e-02 -9.71111715e-01
6.73836231e-01 4.93004322e-01 2.87358046e-01 -1.11945367e+00
-8.10671821e-02 -2.55306870e-01 -2.19950706e-01 -1.93367913e-01
2.54712582e-01 3.63113254e-01 -9.48570967e-01 1.66508520e+00
1.22435048e-01 2.88382601e-02 3.89850810e-02 1.06362808e+00
9.45267022e-01 5.09647667e-01 3.67001712e-01 -9.94291306e-02
1.57575059e+00 -1.42409062e+00 -5.77272952e-01 -4.28134948e-01
6.13008857e-01 -9.19281304e-01 1.43473268e+00 2.22968068e-02
-9.81756806e-01 -5.57373762e-01 -7.44622588e-01 -2.30160668e-01
-3.12521696e-01 1.96606472e-01 4.81569678e-01 2.53738493e-01
-1.37610734e+00 -1.08193092e-01 -4.62972492e-01 -6.64143145e-01
2.71408916e-01 2.34362096e-01 -5.52645028e-01 -2.66064227e-01
-8.78988206e-01 8.97165298e-01 3.02216828e-01 3.78657989e-02
-8.54648530e-01 -5.47348201e-01 -9.52462018e-01 -1.66301042e-01
4.25761104e-01 -8.42141092e-01 1.17732573e+00 -1.08493590e+00
-9.71921325e-01 1.15835702e+00 -4.31980103e-01 -2.43539244e-01
2.94055879e-01 -3.20353925e-01 -1.47477016e-02 5.94301760e-01
3.15902919e-01 1.34643209e+00 9.25918758e-01 -1.80881631e+00
-6.75922513e-01 -9.17311907e-02 2.68138677e-01 3.58710736e-01
-5.23129582e-01 2.51373172e-01 -1.07381308e+00 -5.75223088e-01
-1.30406305e-01 -1.10523808e+00 -1.35162041e-01 1.73730273e-02
-2.45408773e-01 -2.29733065e-01 6.84609652e-01 -6.63580537e-01
9.77454722e-01 -2.01466918e+00 3.31561059e-01 -7.14902952e-03
1.00775016e-02 -9.82961580e-02 -6.88962817e-01 5.05618930e-01
-5.51271550e-02 3.41406286e-01 -2.69649476e-01 -8.18437576e-01
1.34423286e-01 2.30096996e-01 -3.90329391e-01 1.62871927e-01
6.13547325e-01 1.47000611e+00 -1.06426752e+00 -8.21292579e-01
5.09188883e-02 3.86776954e-01 -4.15218264e-01 2.51077801e-01
-3.00884068e-01 3.04698914e-01 -3.88214231e-01 8.26859772e-01
3.47268373e-01 -5.54170728e-01 3.04299011e-03 -5.14234066e-01
8.06231573e-02 -1.54076248e-01 -6.06323838e-01 2.11003566e+00
-4.68012989e-01 6.48264408e-01 -1.11339666e-01 -6.46737039e-01
6.85639441e-01 3.25949132e-01 1.97123811e-01 -7.49160111e-01
-2.50694871e-01 4.83068191e-02 -5.51810145e-01 -5.69020212e-01
7.83563375e-01 1.02709465e-01 -2.69121200e-01 4.11269277e-01
1.35702845e-02 -6.23148493e-02 1.47823289e-01 5.63700497e-01
8.72344315e-01 6.04444504e-01 1.85058296e-01 -1.89387128e-01
4.99506027e-01 1.86402708e-01 6.56433078e-03 8.81416082e-01
-2.59024560e-01 1.08875263e+00 2.39419013e-01 -1.65275887e-01
-1.03345883e+00 -1.01182961e+00 7.62224048e-02 1.53835428e+00
2.49039188e-01 -6.43063486e-01 -3.91258240e-01 -1.09670186e+00
-6.50104955e-02 4.85306501e-01 -6.91760957e-01 2.17904925e-01
-3.44333529e-01 -5.26922286e-01 3.26043427e-01 7.94373095e-01
1.84722781e-01 -9.55670834e-01 -1.67044804e-01 -1.33616641e-01
-4.03684080e-01 -1.59983373e+00 -4.43066239e-01 1.75980199e-02
-4.48550433e-01 -6.21231496e-01 -9.61012840e-01 -1.03595531e+00
8.26359332e-01 5.74730098e-01 1.42194951e+00 1.90980464e-01
1.28021568e-01 1.08469415e+00 -6.66767478e-01 -7.82741532e-02
-2.74354517e-01 4.53605242e-02 -1.37090385e-01 -3.94613519e-02
2.68239260e-01 -9.63045657e-02 -4.86705273e-01 3.63208145e-01
-8.77171695e-01 3.32020193e-01 7.53180385e-01 7.31948793e-01
6.74089313e-01 -5.52787840e-01 4.67753857e-01 -6.19645357e-01
3.98421049e-01 -3.25174123e-01 -3.59499812e-01 5.81168354e-01
-4.98743832e-01 1.73576161e-01 3.51514220e-02 -3.12409878e-01
-8.97633553e-01 3.10574859e-01 2.76107490e-01 -4.15823042e-01
-2.20149010e-01 5.85900724e-01 2.86965575e-02 -2.73795575e-01
4.05767292e-01 1.48181781e-01 -2.70232826e-01 -2.45119855e-01
9.45020616e-01 6.57943964e-01 6.82142198e-01 -7.66898036e-01
9.75993335e-01 5.33816218e-01 -1.53424263e-01 -4.63226855e-01
-1.10064566e+00 -8.85923088e-01 -8.43861997e-01 -2.39978865e-01
1.02046287e+00 -1.34002686e+00 -2.55888343e-01 -1.82331041e-01
-1.18435550e+00 -2.28529230e-01 -3.97072099e-02 2.41406277e-01
-5.76906085e-01 7.15786636e-01 -3.46053600e-01 -7.57732749e-01
-1.74246684e-01 -1.11123002e+00 1.79688418e+00 -2.09699608e-02
-1.63848713e-01 -1.10587645e+00 1.30153969e-01 9.14390743e-01
2.51062483e-01 3.56236249e-02 5.21181881e-01 -4.87368703e-01
-7.33869493e-01 -6.85586855e-02 -7.28126943e-01 1.51647314e-01
-8.79853219e-02 -1.15559109e-01 -1.18471146e+00 -3.06716710e-01
-5.42535067e-01 -8.11346531e-01 1.05122113e+00 -2.29524132e-02
8.50724041e-01 -1.46665141e-01 -3.27549070e-01 1.83966652e-01
1.45963275e+00 -5.80310166e-01 3.96628469e-01 4.66798186e-01
1.10922456e+00 7.04766989e-01 8.77369106e-01 6.51484355e-02
7.49534130e-01 1.00727296e+00 4.59989965e-01 -4.68807489e-01
-3.11066478e-01 -3.50548804e-01 3.95257980e-01 8.29811335e-01
-5.84092699e-02 -3.04991543e-01 -1.12183952e+00 9.16630864e-01
-2.20474911e+00 -6.96242630e-01 -2.19722807e-01 1.74979115e+00
8.88352931e-01 -8.47678855e-02 8.78199264e-02 -4.18771088e-01
5.41224480e-01 1.12188935e-01 3.42415422e-02 -2.26136684e-01
-8.59216005e-02 -2.72136807e-01 4.79347765e-01 4.71022457e-01
-1.36444592e+00 1.17181134e+00 6.75297260e+00 8.13515723e-01
-7.98303723e-01 1.66815802e-01 3.97326678e-01 -2.81541012e-02
-5.77589869e-01 7.32344463e-02 -7.71906912e-01 9.69327167e-02
8.80613208e-01 3.90216380e-01 1.76482663e-01 6.41137064e-01
2.81866491e-01 -1.03580728e-01 -1.42763186e+00 1.10437369e+00
5.76584756e-01 -1.21284688e+00 3.46410662e-01 1.30893663e-01
8.26717436e-01 8.39275122e-02 2.42287979e-01 1.93342388e-01
1.85689986e-01 -1.05136049e+00 8.11652243e-01 5.01355529e-01
7.18310773e-01 -4.01194692e-01 8.54615033e-01 2.03060117e-02
-1.24782324e+00 1.35379195e-01 -1.88341901e-01 9.44763720e-02
2.53559023e-01 2.31027380e-01 -1.09179652e+00 8.51206422e-01
8.11341405e-01 9.94384944e-01 -1.13344264e+00 8.43798399e-01
-4.49413180e-01 6.02665722e-01 -1.56016707e-01 1.61306456e-01
4.08784568e-01 2.67721146e-01 3.90145600e-01 1.44965374e+00
-9.11945626e-02 -2.13916302e-01 4.56256986e-01 8.59216452e-01
-9.77246016e-02 3.08533847e-01 -6.35917604e-01 -2.19475508e-01
2.11145580e-01 1.57578301e+00 -6.69584513e-01 -1.35293230e-01
-7.08814800e-01 1.12307751e+00 5.13787627e-01 5.87920964e-01
-7.99731374e-01 2.54930705e-01 1.06601581e-01 -1.48394659e-01
5.50325632e-01 -2.74124056e-01 -1.70325369e-01 -1.19232285e+00
1.32041186e-01 -7.64115870e-01 3.60389471e-01 -1.34329927e+00
-1.44448674e+00 4.48361337e-01 2.10473403e-01 -1.20828247e+00
-1.25945315e-01 -6.18728161e-01 -2.03825995e-01 7.86992967e-01
-2.00529504e+00 -1.97164953e+00 -2.84093618e-01 4.23036009e-01
4.81884122e-01 1.16409458e-01 8.67500484e-01 -3.66645753e-02
-3.32480311e-01 6.05555713e-01 -3.00690055e-01 1.11760966e-01
1.05215883e+00 -1.36422789e+00 1.09617792e-01 9.42240536e-01
7.06289053e-01 6.43161118e-01 5.79396486e-01 -4.36687827e-01
-1.30104101e+00 -1.04397917e+00 9.60300624e-01 -1.14928854e+00
8.34643781e-01 -5.42378008e-01 -8.64192545e-01 6.41275108e-01
6.45593762e-01 1.44465595e-01 7.76883125e-01 2.31649265e-01
-6.51939332e-01 8.06798041e-02 -6.79345131e-01 6.38597965e-01
9.59036589e-01 -8.54272246e-01 -9.33866739e-01 3.66381466e-01
1.02653515e+00 -1.53508529e-01 -8.58576059e-01 5.07395744e-01
4.96608496e-01 -5.22369444e-01 1.10353494e+00 -5.33388853e-01
6.62332773e-01 -5.66470861e-01 -5.07179081e-01 -9.30733442e-01
-3.77177805e-01 -2.99649507e-01 4.57748324e-02 1.53359103e+00
6.85289025e-01 -5.10543361e-02 5.51579416e-01 4.86519814e-01
7.58788809e-02 -4.04653370e-01 -6.93874240e-01 -5.31783760e-01
-1.35553524e-01 -5.15481591e-01 1.06499091e-01 8.14591408e-01
1.88757747e-01 5.73783636e-01 -4.91603971e-01 3.89295876e-01
5.15625656e-01 2.99636394e-01 8.74628425e-01 -7.37051249e-01
-2.99521148e-01 -1.34037703e-01 -4.22712505e-01 -1.09859240e+00
2.94060469e-01 -8.91381979e-01 3.91936660e-01 -1.65685689e+00
8.90846312e-01 -2.42163509e-01 -6.33071005e-01 8.46350610e-01
-4.73066390e-01 9.50695395e-01 3.73813093e-01 4.27342921e-01
-1.33685434e+00 4.00464326e-01 1.10596073e+00 -2.80658752e-01
2.04376385e-01 -6.67192101e-01 -9.23241973e-01 5.99857926e-01
4.47705954e-01 -3.82247686e-01 -3.40412349e-01 -8.53502572e-01
4.15963084e-01 -3.52907568e-01 6.24031246e-01 -5.05324185e-01
1.39948055e-01 -3.35112028e-02 2.52543300e-01 -6.08192861e-01
3.95110458e-01 -9.55693960e-01 -2.91535288e-01 -4.63615842e-02
-7.19850421e-01 6.64287060e-02 2.86516428e-01 9.02172744e-01
-2.70933002e-01 -1.79295182e-01 3.00594240e-01 -9.42953452e-02
-1.00440001e+00 -8.56619403e-02 -6.17163666e-02 -8.44923854e-02
1.02023506e+00 -8.71549267e-03 -4.24400479e-01 -5.62843800e-01
-6.16511106e-01 4.67226982e-01 8.13249886e-01 6.24699175e-01
6.75294518e-01 -1.45916951e+00 -8.64382386e-01 -1.34904176e-01
8.29502285e-01 -2.79766858e-01 -1.66078448e-01 8.40815187e-01
-1.68125644e-01 6.44732475e-01 1.54734030e-01 -1.17930174e+00
-1.44291198e+00 5.99374115e-01 -1.77922517e-01 -1.93882659e-01
-2.80413210e-01 8.90176535e-01 3.72248381e-01 -2.45983124e-01
3.05166781e-01 -1.81412339e-01 -2.91714936e-01 5.30023873e-02
4.86063212e-01 -2.89673835e-01 -1.13817930e-01 -9.88579869e-01
-5.37763059e-01 8.87184322e-01 -1.50584757e-01 -4.60218996e-01
1.23642576e+00 -6.47933185e-01 -1.18886642e-01 5.00065923e-01
1.17331266e+00 1.95781603e-01 -1.21550274e+00 -4.72162843e-01
3.28418195e-01 -3.74416977e-01 8.39631781e-02 -9.76408124e-01
-7.65504956e-01 8.50020468e-01 5.75871110e-01 -9.84433666e-03
1.16792166e+00 5.99000514e-01 2.95350313e-01 3.81712854e-01
3.46604645e-01 -7.76195109e-01 5.11905134e-01 3.53500098e-01
1.01496410e+00 -1.70067513e+00 2.67029856e-03 -4.51569557e-01
-9.18236792e-01 9.33018327e-01 5.85228741e-01 2.05891043e-01
6.63288459e-02 1.06808078e-03 3.53141397e-01 -1.34250835e-01
-8.36527586e-01 -5.18762350e-01 8.02717865e-01 5.74829280e-01
5.18295348e-01 -2.44343057e-01 -1.63038090e-01 3.84315401e-01
1.94352254e-01 -1.19369075e-01 3.93384159e-01 9.13070023e-01
-3.41732383e-01 -1.13857353e+00 -4.53358233e-01 -1.67890444e-01
-3.14141303e-01 -4.50082511e-01 -6.27610803e-01 7.05523133e-01
-1.69801265e-01 1.13913941e+00 -2.96779140e-03 -2.21755937e-01
9.43588242e-02 3.43531929e-03 4.56366956e-01 -8.31507564e-01
-4.51727867e-01 3.00492585e-01 2.50889927e-01 -7.20243454e-01
-1.16109729e+00 -4.57288206e-01 -9.90063250e-01 4.47017670e-01
-3.81538481e-01 1.45728467e-02 5.17565370e-01 9.48123813e-01
3.93256158e-01 2.51650065e-01 3.76786768e-01 -8.60977232e-01
-1.00687265e-01 -8.08101952e-01 -9.04222056e-02 7.13017881e-01
2.89593458e-01 -3.99540335e-01 -2.63185769e-01 4.04437661e-01] | [10.650477409362793, 1.4909441471099854] |
48aad668-7639-495a-85aa-f8285402d16c | a-joint-model-for-dropped-pronoun-recovery | 2106.03345 | null | https://arxiv.org/abs/2106.03345v1 | https://arxiv.org/pdf/2106.03345v1.pdf | A Joint Model for Dropped Pronoun Recovery and Conversational Discourse Parsing in Chinese Conversational Speech | In this paper, we present a neural model for joint dropped pronoun recovery (DPR) and conversational discourse parsing (CDP) in Chinese conversational speech. We show that DPR and CDP are closely related, and a joint model benefits both tasks. We refer to our model as DiscProReco, and it first encodes the tokens in each utterance in a conversation with a directed Graph Convolutional Network (GCN). The token states for an utterance are then aggregated to produce a single state for each utterance. The utterance states are then fed into a biaffine classifier to construct a conversational discourse graph. A second (multi-relational) GCN is then applied to the utterance states to produce a discourse relation-augmented representation for the utterances, which are then fused together with token states in each utterance as input to a dropped pronoun recovery layer. The joint model is trained and evaluated on a new Structure Parsing-enhanced Dropped Pronoun Recovery (SPDPR) dataset that we annotated with both two types of information. Experimental results on the SPDPR dataset and other benchmarks show that DiscProReco significantly outperforms the state-of-the-art baselines of both tasks. | ['Ji-Rong Wen', 'Nianwen Xue', 'Jun Guo', 'Sheng Gao', 'Si Li', 'Jun Xu', 'Kerui Xu', 'Jingxuan Yang'] | 2021-06-07 | null | https://aclanthology.org/2021.acl-long.138 | https://aclanthology.org/2021.acl-long.138.pdf | acl-2021-5 | ['discourse-parsing'] | ['natural-language-processing'] | [ 4.47356611e-01 8.49785089e-01 -1.12871207e-01 -5.93422890e-01
-1.07032514e+00 -4.82214868e-01 7.96165884e-01 -3.62142958e-02
-8.94845054e-02 4.83671963e-01 9.05336440e-01 -3.76999795e-01
4.66416866e-01 -6.66752338e-01 -7.36777902e-01 -5.51485538e-01
-2.18810722e-01 7.70443976e-01 2.00452402e-01 -3.53557557e-01
-2.22280473e-01 4.98370863e-02 -8.92970443e-01 7.45446920e-01
6.34959757e-01 5.90366602e-01 3.34817737e-01 1.02325654e+00
-4.28761780e-01 1.57133698e+00 -7.54901528e-01 -3.66531610e-01
-3.15754205e-01 -5.00229597e-01 -1.48329663e+00 2.92353779e-01
-9.40257162e-02 -5.97365618e-01 -6.49443746e-01 7.46519148e-01
2.18817681e-01 6.43546999e-01 4.13066834e-01 -7.98641145e-01
-6.71716690e-01 1.18565989e+00 -1.93057358e-01 2.04794601e-01
7.42393136e-01 -1.74118474e-01 1.35666287e+00 -7.25105762e-01
9.32560205e-01 2.02598739e+00 1.62521988e-01 8.71586323e-01
-1.02759695e+00 -2.45862976e-01 6.06460154e-01 2.82974914e-02
-7.59405375e-01 -4.57784474e-01 9.09393072e-01 -5.34660444e-02
1.66214538e+00 3.33947271e-01 4.05342668e-01 1.26692796e+00
-3.60120609e-02 1.30161822e+00 5.67100704e-01 -2.05657870e-01
2.74347011e-02 -4.56092536e-01 6.76818192e-01 8.01517904e-01
-5.62777460e-01 -3.03192347e-01 -5.47626615e-01 4.18643141e-03
5.03089249e-01 -1.67929277e-01 -2.47122079e-01 3.55005920e-01
-9.36386764e-01 9.46671486e-01 6.32984221e-01 1.60038710e-01
-4.26433682e-01 3.70131917e-02 5.23696959e-01 3.45329940e-01
7.92786896e-01 1.25686079e-01 -1.33446693e-01 -2.26952627e-01
-2.29082018e-01 4.79919225e-01 1.36256778e+00 9.47189271e-01
7.47464359e-01 -1.56227425e-01 -4.06688750e-01 1.26056409e+00
3.08442563e-01 4.15804498e-02 3.13540667e-01 -1.20412028e+00
8.67084324e-01 7.38054276e-01 -2.45385483e-01 -6.87041402e-01
-4.81349647e-01 1.09435268e-01 -6.87467217e-01 -6.06139183e-01
2.04706416e-01 -4.66703326e-01 -6.84087455e-01 1.69526649e+00
2.88814366e-01 2.12355658e-01 5.83324075e-01 1.02035570e+00
1.43340790e+00 1.15430522e+00 1.48892954e-01 -1.85780689e-01
1.56918919e+00 -1.50278187e+00 -1.05206323e+00 -5.34379005e-01
5.33701420e-01 -4.45289642e-01 5.65214634e-01 -7.77115896e-02
-1.25920200e+00 -2.74435967e-01 -7.07189262e-01 -5.82195163e-01
5.57928951e-03 -3.12897980e-01 5.39479971e-01 -1.64060891e-01
-1.15186334e+00 3.82600069e-01 -1.17218959e+00 -8.90231282e-02
1.09773621e-01 2.06367135e-01 -2.57285684e-01 -1.13996290e-01
-1.36739123e+00 8.96365345e-01 4.32737708e-01 3.01172286e-01
-1.04947102e+00 -1.53169289e-01 -1.44667625e+00 2.07546696e-01
4.43397731e-01 -4.40030336e-01 1.91479623e+00 -8.77594590e-01
-1.93156970e+00 7.53288805e-01 -6.15668356e-01 -7.95198739e-01
8.61657131e-03 -6.61679506e-02 -2.05160245e-01 1.38634145e-01
2.65005101e-02 6.91318512e-01 4.74199563e-01 -1.19113004e+00
-7.79646397e-01 -2.22397327e-01 5.63435793e-01 6.42051876e-01
4.77112055e-01 2.49067634e-01 -7.52896905e-01 -9.41059887e-02
2.91911513e-01 -7.90795803e-01 -7.74795637e-02 -1.22480154e+00
-8.28631520e-01 -1.05727196e+00 8.53067577e-01 -1.04219449e+00
1.19307017e+00 -1.96916258e+00 8.02712619e-01 -4.32041943e-01
3.12092513e-01 3.06701171e-03 -5.06930113e-01 5.61538339e-01
-2.82805115e-02 -5.59137166e-02 -3.54204476e-01 -9.76293921e-01
-5.79726808e-02 5.64392388e-01 -3.90177071e-01 2.17678875e-01
7.86016881e-01 1.13436091e+00 -1.26493537e+00 -2.65899748e-01
8.75646248e-02 5.33934534e-01 -4.91572648e-01 8.46817613e-01
-7.06062913e-01 3.83370012e-01 -4.83798712e-01 4.55378711e-01
3.33839417e-01 -2.21852481e-01 8.29714596e-01 2.13856027e-01
5.94575591e-02 1.24551845e+00 -6.21465802e-01 1.40488636e+00
-4.81664419e-01 7.51785934e-01 2.57430643e-01 -1.03006780e+00
1.02471852e+00 5.26831985e-01 -3.09652425e-02 -4.56114978e-01
-2.07029446e-03 -3.57911438e-02 3.17699239e-02 -4.51281279e-01
8.79301608e-01 5.29873138e-03 -4.78519380e-01 3.15256596e-01
4.80509222e-01 -2.14044914e-01 1.74911737e-01 5.32767892e-01
1.25952792e+00 3.26713137e-02 1.90344766e-01 1.11688025e-01
6.78169489e-01 -2.41095364e-01 7.72405744e-01 4.31880713e-01
-2.04428762e-01 5.25885940e-01 1.21006310e+00 -3.21726590e-01
-6.04564488e-01 -8.54022205e-01 2.90637702e-01 1.39030313e+00
-6.90642139e-03 -4.88651633e-01 -9.86045599e-01 -8.73823225e-01
-7.63237327e-02 7.98779130e-01 -3.85368794e-01 3.03100199e-01
-1.23012269e+00 -3.84137332e-01 3.45208585e-01 6.02267921e-01
4.27388370e-01 -1.64241529e+00 9.23178121e-02 5.01257896e-01
-5.60976148e-01 -1.50926268e+00 -5.61295271e-01 2.74016380e-01
-4.09962893e-01 -1.02693534e+00 -2.57860988e-01 -1.23025680e+00
3.85931224e-01 1.71337426e-01 1.27621198e+00 1.25634193e-01
5.80256701e-01 2.39922822e-01 -5.10806561e-01 -7.71155283e-02
-1.02821898e+00 1.99895620e-01 -4.14378166e-01 1.32027030e-01
4.39096510e-01 -3.55145305e-01 -3.02054703e-01 -2.32569113e-01
-5.21050394e-01 2.02036336e-01 1.66197449e-01 1.05601525e+00
5.13091207e-01 -4.75655735e-01 7.62763023e-01 -1.44514930e+00
9.37651396e-01 -5.75921237e-01 -5.95127702e-01 4.32403050e-02
3.62718165e-01 3.53823649e-03 4.17295933e-01 -2.11904287e-01
-1.52223921e+00 -9.67705473e-02 -3.42474163e-01 -1.82531804e-01
-2.18710676e-02 7.57758439e-01 -2.60581404e-01 8.78517330e-01
1.93730265e-01 -6.16869330e-03 -2.69318279e-02 -4.96152252e-01
6.20311081e-01 1.02379501e+00 8.99568379e-01 -5.79116762e-01
6.32012933e-02 -5.30835129e-02 -5.11837363e-01 -1.07810271e+00
-1.18834054e+00 -4.97016758e-01 -5.07660866e-01 -1.08708426e-01
1.41861463e+00 -1.01575863e+00 -9.04878139e-01 5.62263608e-01
-1.76100838e+00 -6.99997246e-01 -8.56306478e-02 1.40882403e-01
-4.61897075e-01 3.50745231e-01 -1.23382771e+00 -1.17443359e+00
-4.22958255e-01 -1.21961439e+00 1.24772429e+00 4.46877092e-01
-1.49281636e-01 -1.23331082e+00 -5.65477982e-02 8.30538690e-01
-8.28930549e-03 2.44787350e-01 8.33923995e-01 -1.11652803e+00
-6.03148758e-01 2.92931311e-02 -1.84135780e-01 5.09592235e-01
3.91204022e-02 -2.25602463e-01 -1.08546209e+00 -1.63204089e-01
2.84741856e-02 -3.97571534e-01 1.06932366e+00 2.66708553e-01
9.00675833e-01 -4.93790835e-01 -3.10992867e-01 3.23681712e-01
6.75165355e-01 2.32777014e-01 5.33374906e-01 -1.15257651e-01
9.15122628e-01 8.28906178e-01 4.06407982e-01 -1.85122639e-01
1.10468435e+00 4.43822861e-01 4.74496484e-01 1.10116862e-01
-4.16052610e-01 -2.72872776e-01 6.19636655e-01 1.62115932e+00
-4.47635874e-02 -4.99887556e-01 -9.46518302e-01 6.83227897e-01
-2.09826875e+00 -7.86945999e-01 -8.81721079e-02 1.44090486e+00
1.02609789e+00 1.20345376e-01 -1.05772391e-01 -5.89143217e-01
8.45296383e-01 8.04919481e-01 -4.02902544e-01 -8.40674043e-01
-1.73712060e-01 1.94092751e-01 -1.40638173e-01 1.01937592e+00
-1.26446486e+00 1.33662808e+00 5.69563341e+00 2.22247586e-01
-7.04554319e-01 1.45775244e-01 6.48591578e-01 3.37120771e-01
-1.96877494e-01 1.79375365e-01 -7.85313308e-01 -1.14339180e-02
1.33678401e+00 -1.01509981e-01 7.20244229e-01 8.10626149e-01
-6.36914968e-02 4.69920272e-03 -1.30403554e+00 5.01707554e-01
1.15522012e-01 -1.39754021e+00 -1.18277408e-01 -2.79533088e-01
6.15613520e-01 4.08879906e-01 -4.84297544e-01 8.73644173e-01
8.90013635e-01 -1.08631098e+00 2.51590371e-01 2.45935276e-01
4.57769364e-01 -8.59535813e-01 9.49041843e-01 4.66460168e-01
-1.12944925e+00 8.70448574e-02 -3.31983298e-01 -2.50486612e-01
3.55153888e-01 -7.55141675e-02 -1.26823664e+00 7.25487053e-01
2.26650313e-01 8.21760356e-01 9.76083502e-02 -6.65180683e-02
-7.37839162e-01 7.62619436e-01 1.18572280e-01 -2.38931537e-01
4.70580459e-01 -2.51659840e-01 7.82825291e-01 1.50930059e+00
-4.65542197e-01 4.90582854e-01 4.17450279e-01 8.54306400e-01
-6.20465994e-01 -3.34973276e-01 -2.86773682e-01 -2.44289577e-01
7.92623520e-01 1.43184638e+00 -1.76031917e-01 -6.32943690e-01
-6.48029387e-01 9.36603606e-01 1.02581811e+00 4.53643054e-01
-3.87519777e-01 -7.71853179e-02 7.50464320e-01 -3.91306758e-01
2.50441611e-01 -1.17212415e-01 3.75900716e-01 -1.20493042e+00
-1.91435754e-01 -9.70604062e-01 5.39247751e-01 -5.16608000e-01
-1.48701656e+00 8.81755173e-01 -9.00436193e-02 -4.38458532e-01
-9.38390553e-01 -4.31891531e-01 -8.99256289e-01 1.26721811e+00
-1.69383097e+00 -1.24206173e+00 -1.46246433e-01 3.19149643e-01
1.09479082e+00 -1.98596731e-01 1.15818644e+00 -2.71524996e-01
-9.57946122e-01 1.67675674e-01 -4.32904929e-01 8.00038218e-01
2.54601240e-01 -1.70689821e+00 1.00482273e+00 7.42310345e-01
-1.58885807e-01 5.20143509e-01 2.90857017e-01 -7.93980122e-01
-1.59803104e+00 -1.31426263e+00 1.17627716e+00 -4.41066593e-01
6.81861997e-01 -6.15773320e-01 -1.44872618e+00 1.33929598e+00
6.46784484e-01 -2.26681218e-01 4.52383578e-01 4.55064952e-01
-2.23310187e-01 3.91895711e-01 -5.86522520e-01 5.13049066e-01
9.21562970e-01 -9.41076875e-01 -1.34053385e+00 4.98634785e-01
1.65227795e+00 -1.19162047e+00 -8.73810828e-01 -1.57407839e-02
1.71966478e-02 -6.28845572e-01 5.84978938e-01 -8.49363744e-01
6.23974800e-01 1.90798447e-01 -2.23879769e-01 -1.26746678e+00
-5.42439846e-03 -9.96224701e-01 -2.96441674e-01 1.41348004e+00
4.97864664e-01 -5.22891581e-01 3.52092147e-01 5.79078138e-01
-7.57916451e-01 -7.63513744e-01 -1.03776538e+00 -2.47115478e-01
1.40839502e-01 -2.66535342e-01 6.53061986e-01 8.46977651e-01
4.80678201e-01 1.15180278e+00 -3.28181684e-02 2.64447570e-01
2.04412028e-01 2.48758137e-01 6.81644678e-01 -7.82815814e-01
-3.41983348e-01 -1.38518021e-01 1.18493930e-01 -1.46437693e+00
9.14069951e-01 -1.22890007e+00 4.15753573e-01 -1.93926644e+00
2.95736313e-01 -1.48024010e-02 8.87144655e-02 4.43725109e-01
-5.62078297e-01 -4.84640211e-01 2.98885494e-01 1.03741564e-01
-7.57239103e-01 5.71159542e-01 1.23233867e+00 -2.26776123e-01
-5.98538160e-01 -1.08272567e-01 -5.50419688e-01 7.98775554e-01
3.99970531e-01 -1.17935315e-01 -2.43386049e-02 -5.20712376e-01
-4.58505034e-01 8.58048558e-01 1.06344052e-01 -2.12072298e-01
2.29356185e-01 8.84385407e-02 -2.04037115e-01 -7.95054197e-01
6.26161456e-01 -5.69229275e-02 -4.09334123e-01 8.82858038e-02
-4.93627280e-01 -2.21207231e-01 -1.09629363e-01 5.69668710e-01
-5.10132015e-01 5.67873865e-02 4.54863161e-01 -8.70157033e-02
-6.20409608e-01 -8.37894902e-03 -3.10595125e-01 1.92146719e-01
5.05432546e-01 5.45363963e-01 -6.93928719e-01 -7.43349433e-01
-8.10719609e-01 7.29977489e-01 -1.40679702e-01 7.58367777e-01
4.94176716e-01 -1.09760177e+00 -1.04724741e+00 -1.01697378e-01
-3.50620419e-01 7.98346817e-01 2.83094168e-01 3.96409601e-01
-3.16490710e-01 5.64223528e-01 3.32857490e-01 -7.11393595e-01
-1.31813371e+00 2.96246231e-01 4.68241513e-01 -6.89687967e-01
-7.99516380e-01 1.00616038e+00 4.23970610e-01 -7.53223956e-01
4.43920165e-01 -6.59883618e-01 -5.41100860e-01 1.30763546e-01
5.41971922e-01 2.41793804e-02 -7.88674429e-02 -8.48619759e-01
-4.34029698e-01 -2.63984174e-01 -3.36440980e-01 -1.94590151e-01
1.64448988e+00 -1.21499233e-01 -4.94386971e-01 4.77754176e-01
1.26670301e+00 -1.74114183e-01 -1.53592277e+00 -5.95378578e-01
2.00605795e-01 2.15889141e-01 -2.29275554e-01 -4.52846348e-01
-8.24201107e-01 7.16393232e-01 -3.11130852e-01 6.73572361e-01
8.32354426e-01 5.55895746e-01 9.31679130e-01 5.19765377e-01
1.32917808e-02 -9.39815879e-01 2.06624135e-01 1.19532537e+00
1.22156501e+00 -1.01451564e+00 -4.90140051e-01 -6.97590709e-01
-1.13686848e+00 1.16333032e+00 4.90600377e-01 -4.02898967e-01
2.44380295e-01 3.96203101e-01 9.32282507e-02 -1.91052720e-01
-1.17201459e+00 -2.61194497e-01 2.31826276e-01 6.89000726e-01
8.10813069e-01 4.25848693e-01 -1.30296722e-01 1.09138882e+00
-4.14570838e-01 -4.36981022e-01 6.04239643e-01 6.44010067e-01
-3.09157193e-01 -1.16635168e+00 7.41363764e-02 7.76535049e-02
-3.36863309e-01 -3.86130624e-02 -7.16226935e-01 7.48575807e-01
-5.00137508e-01 1.27537668e+00 5.40589690e-01 -3.69301170e-01
5.57665825e-01 4.21531707e-01 1.39221668e-01 -1.08721459e+00
-1.00456405e+00 3.19920391e-01 9.67835128e-01 -6.42417550e-01
-3.79322529e-01 -7.39065170e-01 -1.85090756e+00 -1.95068330e-01
-1.90040395e-01 1.38926849e-01 4.39999908e-01 9.86517906e-01
2.11567655e-01 1.23138654e+00 6.29355550e-01 -8.25753629e-01
-5.95945060e-01 -1.33407557e+00 -3.42773944e-01 5.04081905e-01
6.11734867e-01 -2.79652685e-01 -2.84649223e-01 6.69379439e-03] | [12.4485502243042, 7.966202735900879] |
a9cc7806-27c5-434c-8240-3d88ad2bbdb1 | detecting-aggression-and-toxicity-using-a | null | null | https://aclanthology.org/W19-3517 | https://aclanthology.org/W19-3517.pdf | Detecting Aggression and Toxicity using a Multi Dimension Capsule Network | In the era of social media, hate speech, trolling and verbal abuse have become a common issue. We present an approach to automatically classify such statements, using a new deep learning architecture. Our model comprises of a Multi Dimension Capsule Network that generates the representation of sentences which we use for classification. We further provide an analysis of our model{'}s interpretation of such statements. We compare the results of our model with state-of-art classification algorithms and demonstrate our model{'}s ability. It also has the capability to handle comments that are written in both Hindi and English, which are provided in the TRAC dataset. We also compare results on Kaggle{'}s Toxic comment classification dataset. | ['Prerna Khurana', 'Saurabh Srivastava'] | 2019-08-01 | null | null | null | ws-2019-8 | ['toxic-comment-classification'] | ['natural-language-processing'] | [-8.06744024e-02 4.53143865e-01 5.78966774e-02 -5.77886343e-01
-4.39597458e-01 -5.63484788e-01 8.52431118e-01 4.13916677e-01
-3.48554373e-01 8.67105544e-01 7.14316487e-01 -7.06983626e-01
1.11025773e-01 -5.86798370e-01 -2.88932413e-01 -2.96624929e-01
5.15543595e-02 4.01742637e-01 -1.76914379e-01 -5.85854530e-01
5.22502363e-01 2.56687135e-01 -8.03722084e-01 8.65856290e-01
4.61996794e-01 7.02789545e-01 -6.24300063e-01 7.20371842e-01
-6.90819174e-02 1.71639025e+00 -1.22476768e+00 -1.09836340e+00
-3.41201216e-01 -3.00934762e-01 -1.28894031e+00 5.26688434e-02
5.35991073e-01 -6.38429940e-01 -4.19366539e-01 1.09219503e+00
4.58840549e-01 -1.09698966e-01 1.12619054e+00 -1.27740562e+00
-1.23689091e+00 1.23263657e+00 -4.19373751e-01 4.79990363e-01
3.77781928e-01 -7.12084919e-02 9.47811246e-01 -8.43931913e-01
9.64484632e-01 1.40305579e+00 9.81093824e-01 7.83537686e-01
-9.56274509e-01 -8.45568657e-01 -7.30509460e-02 1.88005522e-01
-9.44489241e-01 -3.12658668e-01 1.04548192e+00 -8.99265110e-01
1.01993537e+00 3.49001259e-01 4.27022696e-01 1.85819387e+00
2.88745970e-01 1.18022120e+00 9.66873348e-01 -2.39483297e-01
1.09776855e-01 2.98506051e-01 6.58829570e-01 7.50843048e-01
-6.26533478e-02 -3.83749604e-01 -5.18222928e-01 -4.40476596e-01
-4.81371731e-02 -2.11203530e-01 -6.91112131e-03 2.59513736e-01
-7.10184336e-01 1.71250737e+00 4.61145341e-01 4.46617544e-01
-1.25246301e-01 2.12314010e-01 8.56439412e-01 2.07831770e-01
1.05807066e+00 4.55589712e-01 -8.39690417e-02 1.60268997e-03
-9.39041734e-01 5.04233718e-01 1.06345713e+00 6.35099888e-01
-2.57026535e-02 1.30466789e-01 -5.43717742e-02 8.27565074e-01
2.80111969e-01 2.00792313e-01 4.60374057e-01 -7.51247108e-01
8.12587023e-01 3.32188696e-01 -6.52450472e-02 -1.42218077e+00
-7.07772553e-01 -4.44075495e-01 -8.85056973e-01 1.29429791e-02
1.07685700e-01 -4.93188858e-01 -5.17804325e-01 1.29717422e+00
-3.07681203e-01 -1.67760953e-01 2.83379912e-01 5.55303097e-01
1.18577135e+00 5.10607719e-01 1.13538638e-01 2.53557973e-02
1.02468765e+00 -7.55147755e-01 -9.59399164e-01 -5.23770675e-02
9.99475658e-01 -4.64089006e-01 4.99933511e-01 3.85096639e-01
-8.59708130e-01 -2.29186434e-02 -9.66445923e-01 -2.94117838e-01
-6.06192827e-01 3.16787884e-02 6.59819245e-01 8.65618348e-01
-9.87883925e-01 6.42154217e-01 -2.91852295e-01 -3.25011343e-01
8.37742448e-01 2.75002718e-01 -3.80703956e-01 3.61384660e-01
-1.41355026e+00 1.08234823e+00 4.66072559e-01 1.26359975e-02
-6.21697426e-01 -3.92582029e-01 -9.69819844e-01 -1.48601919e-01
-1.16851434e-01 -1.49540037e-01 1.31389666e+00 -1.06242156e+00
-1.12777185e+00 1.22366619e+00 2.80812442e-01 -8.54424357e-01
5.47235727e-01 -2.31703326e-01 -7.07953572e-01 2.96265900e-01
1.32708892e-01 3.40525717e-01 8.51650894e-01 -1.20724511e+00
-6.53406382e-02 -2.37927690e-01 2.83653259e-01 -3.27429682e-01
-7.08551407e-01 8.22175205e-01 6.69118464e-01 -7.13404655e-01
-5.67581952e-01 -9.48066294e-01 1.64418742e-01 -2.26826429e-01
-1.21984208e+00 -3.39611977e-01 1.20799160e+00 -1.03374028e+00
1.66789353e+00 -2.20644569e+00 1.29101440e-01 8.38474631e-02
7.30254173e-01 5.34451604e-01 1.33564547e-01 8.11911523e-01
-4.27761555e-01 7.47901618e-01 -5.07562459e-01 -6.93252385e-01
6.41842559e-02 1.22695565e-01 -5.39687276e-01 6.54352963e-01
2.87066579e-01 7.68249214e-01 -8.20282519e-01 -3.65260661e-01
-4.07269932e-02 3.23553145e-01 -5.88422120e-01 2.37928137e-01
-1.09125152e-01 1.69141173e-01 -3.59272212e-01 1.48339599e-01
4.05099422e-01 -2.56919086e-01 1.96068794e-01 3.04232743e-02
-5.95741533e-02 4.78764355e-01 -3.90083462e-01 1.10362470e+00
-1.57341540e-01 1.09827280e+00 1.35777757e-01 -8.32631886e-01
9.87966180e-01 4.01855379e-01 -2.78395675e-02 -1.05598152e-01
6.18390620e-01 1.02355197e-01 7.19312653e-02 -7.94758141e-01
4.61136013e-01 -3.69492352e-01 -4.48204964e-01 8.89668047e-01
-1.93380062e-02 -9.54571962e-02 1.09616444e-01 6.42393708e-01
1.15913522e+00 -3.25574577e-01 4.20264721e-01 -4.49804962e-01
5.39483070e-01 -1.40143856e-01 -3.81992050e-02 5.37210226e-01
-1.67315200e-01 4.78603452e-01 1.05555475e+00 -9.21941936e-01
-1.03956950e+00 -6.52546227e-01 -5.65173887e-02 9.22951162e-01
-5.93854249e-01 -6.74988985e-01 -5.81784010e-01 -1.04561388e+00
-5.86463436e-02 1.28839171e+00 -1.24643672e+00 -1.34074822e-01
-3.91995817e-01 -7.82854974e-01 1.09182358e+00 5.52708209e-01
5.22785366e-01 -1.15485466e+00 -6.00463808e-01 8.72367918e-02
-8.36389363e-02 -9.76413727e-01 8.32731351e-02 2.35870674e-01
-1.37761116e-01 -1.07056069e+00 -1.87160134e-01 -4.94849116e-01
4.61202353e-01 -5.46848774e-01 1.10082972e+00 5.23214281e-01
1.21464089e-01 1.57862939e-02 -6.60869777e-01 -8.02571237e-01
-1.15283167e+00 3.29104692e-01 -3.10050789e-02 -8.87471214e-02
3.81191254e-01 -3.97013694e-01 -6.36298582e-02 -3.92702729e-01
-1.27144647e+00 2.42820270e-02 -2.39188466e-02 7.53903389e-01
-7.63941228e-01 -1.77129075e-01 5.86787581e-01 -1.69684744e+00
1.32250643e+00 -1.03502619e+00 1.87976912e-01 -2.47717544e-01
4.20510732e-02 -1.91273913e-01 7.66520023e-01 -1.41511992e-01
-7.58284688e-01 -4.00876760e-01 -6.82841539e-01 -2.24215582e-01
-2.68643260e-01 7.98801064e-01 4.06967789e-01 4.70121861e-01
8.79082859e-01 -2.00172350e-01 6.24844320e-02 -4.38851118e-01
1.41169488e-01 1.27399421e+00 2.88810372e-01 -1.79565549e-01
8.30589116e-01 3.04971844e-01 -3.02317977e-01 -9.84346211e-01
-1.30709767e+00 -8.76152962e-02 -8.91567349e-01 -2.05242872e-01
1.17341185e+00 -6.48112178e-01 -6.41001463e-01 6.90337121e-01
-1.74131691e+00 -9.96526405e-02 3.65620494e-01 -5.99302463e-02
-3.15935403e-01 4.50629056e-01 -1.34224534e+00 -1.05272377e+00
-5.72536647e-01 -7.59064972e-01 7.19154298e-01 -2.99080878e-01
-9.08923388e-01 -1.44711781e+00 1.71252489e-01 4.86230999e-01
3.76350909e-01 6.68755352e-01 1.20123410e+00 -1.52498817e+00
4.36304599e-01 -6.73980787e-02 -2.18863025e-01 5.52143514e-01
-1.29610702e-01 1.62034139e-01 -9.82477963e-01 -3.46112102e-01
-2.77591161e-02 -8.09829414e-01 7.98275173e-01 -4.62353408e-01
1.30480611e+00 -9.57416356e-01 3.51646990e-02 2.29060665e-01
1.11940765e+00 -9.83350500e-02 5.17577291e-01 3.76940966e-01
7.12323785e-01 6.01584613e-01 1.20535016e-01 6.03621602e-01
4.06163126e-01 4.07773733e-01 7.22430766e-01 2.95394827e-02
1.11779511e-01 -9.83612388e-02 4.19515967e-01 9.12768185e-01
1.07765339e-01 -5.51329553e-01 -1.33565640e+00 3.99502039e-01
-1.86564183e+00 -1.34977782e+00 -5.16328096e-01 1.11554909e+00
7.80052781e-01 1.76968321e-01 3.00332814e-01 3.37465584e-01
5.24222076e-01 4.87357110e-01 -1.16959907e-01 -1.23063529e+00
6.47716271e-03 3.05618197e-01 7.33222738e-02 5.49478471e-01
-1.44460464e+00 8.45185280e-01 7.05877686e+00 3.88717294e-01
-1.12143791e+00 1.47903964e-01 8.35150838e-01 1.81318775e-01
-3.01356435e-01 -8.06802392e-01 -4.42806512e-01 5.25981784e-01
1.40145910e+00 -1.96137521e-02 -6.98598474e-02 8.04230511e-01
-3.20485421e-02 3.05759490e-01 -1.10515726e+00 7.12942183e-01
8.21925759e-01 -1.66535068e+00 1.19930074e-01 -1.24648884e-02
5.20574391e-01 4.93766442e-02 1.67077824e-01 3.81990582e-01
7.11084545e-01 -1.38665354e+00 1.01458716e+00 1.16156630e-01
5.08810639e-01 -8.24032426e-01 1.17333663e+00 5.17658710e-01
-3.63254994e-02 -3.52324277e-01 -1.18920363e-01 -4.45522219e-01
-7.07168058e-02 3.71053606e-01 -9.86243486e-01 2.09041208e-01
5.72451890e-01 1.11716044e+00 -9.97231960e-01 3.28047007e-01
-5.90151072e-01 8.67789268e-01 8.28191042e-02 -4.58050191e-01
4.33812886e-01 2.10379571e-01 4.28454876e-01 1.79591203e+00
-7.52132386e-02 7.27892667e-02 8.89034122e-02 7.15123534e-01
-1.79520786e-01 1.30825881e-02 -9.71469700e-01 -2.13115454e-01
3.74302231e-02 9.56613719e-01 -4.01381880e-01 -4.93618935e-01
-3.31789255e-01 1.12573719e+00 8.02525997e-01 4.11135070e-02
-8.42618287e-01 -1.15778126e-01 3.72328848e-01 7.19285235e-02
1.56417698e-01 -1.87788993e-01 -9.52931121e-02 -1.35943258e+00
-4.12974656e-01 -1.01751900e+00 4.19268250e-01 -9.03966844e-01
-1.62083232e+00 8.82099688e-01 -7.60877784e-03 -7.87056625e-01
-5.88088989e-01 -9.28394496e-01 -6.39228404e-01 4.45620924e-01
-1.13313997e+00 -1.47121966e+00 2.95225769e-01 4.75866884e-01
4.81433362e-01 -3.24344337e-01 1.21045029e+00 2.85040677e-01
-7.00922132e-01 3.98687482e-01 -1.09328836e-01 9.93943512e-01
4.20326144e-01 -1.19011724e+00 3.67876321e-01 6.45927012e-01
2.05408707e-02 7.12365031e-01 9.53570783e-01 -6.03712678e-01
-5.94253838e-01 -9.97740030e-01 1.07396817e+00 -9.01249468e-01
1.39274430e+00 -5.95735192e-01 -7.76749671e-01 1.17937899e+00
7.10770667e-01 -4.63094234e-01 1.49625325e+00 1.62054777e-01
-8.47326279e-01 6.17464125e-01 -1.17977309e+00 4.11310941e-01
7.79460073e-01 -7.55408645e-01 -1.02495599e+00 7.01675475e-01
4.00180846e-01 -2.35554934e-01 -5.72289646e-01 -2.00140134e-01
3.18494231e-01 -1.10643125e+00 4.60193217e-01 -1.48659647e+00
1.49102080e+00 2.46652648e-01 -2.49638613e-02 -1.59099901e+00
-4.09941822e-01 -3.45365822e-01 -1.21228166e-01 1.36257565e+00
6.90310478e-01 -3.49650711e-01 4.84069675e-01 7.36221194e-01
-6.53564604e-03 -5.79229712e-01 -9.25286770e-01 -2.78237164e-01
6.96052372e-01 -5.32663703e-01 2.31937155e-01 1.49874949e+00
3.92532647e-01 8.08077633e-01 -7.09770799e-01 -1.83344111e-01
4.66886580e-01 -1.38173312e-01 4.11279738e-01 -1.29447854e+00
-2.04930142e-01 -3.98763925e-01 -5.23430467e-01 -2.98629045e-01
8.48901987e-01 -1.18593848e+00 -5.64530671e-01 -1.48817778e+00
3.37002426e-01 2.03599408e-02 4.54264283e-02 6.01060331e-01
1.40653551e-01 5.63205838e-01 5.86221635e-01 3.06106135e-02
-5.79611778e-01 1.77926421e-01 6.39089644e-01 -4.25113469e-01
4.72385675e-01 -2.77841628e-01 -8.32744718e-01 1.04507947e+00
1.00629032e+00 -5.43480098e-01 3.05280369e-02 -5.16390026e-01
7.07158327e-01 -2.26387978e-01 3.43573332e-01 -7.35787272e-01
-2.06586614e-01 7.03654662e-02 4.33112174e-01 -4.30999666e-01
3.47544402e-01 -6.02208316e-01 -3.02031338e-01 6.47273839e-01
-9.44909096e-01 9.25776437e-02 4.10638638e-02 2.42950022e-01
-1.15966760e-01 -4.87421095e-01 8.16248417e-01 -2.38859996e-01
-2.73066461e-01 -1.37682945e-01 -9.37578380e-01 -6.67567253e-02
9.56070781e-01 4.46790218e-01 -8.25950325e-01 -9.17366207e-01
-8.06349576e-01 1.40005862e-02 2.63115376e-01 6.72559500e-01
6.89643025e-01 -1.08390605e+00 -1.05880570e+00 -2.33696163e-01
1.07746951e-01 -6.87826395e-01 -3.14169794e-01 4.17349458e-01
-7.69180596e-01 2.30202600e-01 -2.69110441e-01 2.59362832e-02
-1.35537577e+00 6.97968781e-01 1.40354946e-01 -3.11335921e-01
-4.88042414e-01 7.66859233e-01 -1.94224507e-01 -5.45270979e-01
-2.63146430e-01 -4.94395643e-02 -6.89544141e-01 3.03413391e-01
8.21577430e-01 1.71970412e-01 -1.82451561e-01 -1.02800643e+00
-4.57231462e-01 -3.02699387e-01 -3.66316885e-01 1.10273957e-01
1.73145890e+00 1.53717831e-01 -4.97280419e-01 7.96923757e-01
1.57966590e+00 2.28078589e-01 -3.60717267e-01 2.23504379e-01
7.49329105e-02 -2.10593984e-01 -2.11963981e-01 -9.26140666e-01
-8.52102637e-01 1.02886164e+00 1.42302578e-02 7.37534523e-01
3.50232422e-01 1.64902627e-01 7.27749884e-01 4.53946501e-01
-6.03395998e-02 -9.61789727e-01 3.58383536e-01 1.00405562e+00
1.15891790e+00 -1.08822203e+00 9.34052318e-02 -5.69342494e-01
-7.65739858e-01 1.48984253e+00 4.08917814e-01 -2.94670045e-01
6.38408005e-01 4.29968417e-01 3.99910837e-01 -6.13562286e-01
-7.82821536e-01 3.33877414e-01 -1.13101803e-01 6.57548308e-01
7.63451636e-01 1.13253042e-01 -4.66607571e-01 6.75455332e-01
-7.29242504e-01 -4.07339573e-01 1.25898170e+00 8.08821678e-01
-7.96768069e-02 -7.83558607e-01 -2.28371948e-01 6.32942379e-01
-1.01693320e+00 -2.18547359e-01 -1.14892554e+00 7.43207753e-01
-6.88168555e-02 1.11126447e+00 -1.80982519e-02 -5.80165565e-01
-1.18248619e-01 2.72127509e-01 -1.39778033e-01 -9.29238319e-01
-1.35914445e+00 -5.78770697e-01 8.81082714e-01 -6.26198277e-02
-5.35235405e-01 -7.25045562e-01 -1.11575925e+00 -6.75713837e-01
4.68120761e-02 1.65484920e-01 5.48387468e-01 1.11760879e+00
5.68921231e-02 3.76682371e-01 3.30968022e-01 -4.97718304e-01
-5.80322444e-01 -1.17374790e+00 -5.94910860e-01 1.02440500e+00
5.97839892e-01 -3.15581769e-01 -4.88607466e-01 2.34799683e-02] | [8.784991264343262, 10.623607635498047] |
bd451a83-0c77-4fb1-b5f9-0e7bd472392f | lambeq-an-efficient-high-level-python-library | 2110.04236 | null | https://arxiv.org/abs/2110.04236v1 | https://arxiv.org/pdf/2110.04236v1.pdf | lambeq: An Efficient High-Level Python Library for Quantum NLP | We present lambeq, the first high-level Python library for Quantum Natural Language Processing (QNLP). The open-source toolkit offers a detailed hierarchy of modules and classes implementing all stages of a pipeline for converting sentences to string diagrams, tensor networks, and quantum circuits ready to be used on a quantum computer. lambeq supports syntactic parsing, rewriting and simplification of string diagrams, ansatz creation and manipulation, as well as a number of compositional models for preparing quantum-friendly representations of sentences, employing various degrees of syntax sensitivity. We present the generic architecture and describe the most important modules in detail, demonstrating the usage with illustrative examples. Further, we test the toolkit in practice by using it to perform a number of experiments on simple NLP tasks, implementing both classical and quantum pipelines. | ['Bob Coecke', 'Stephen Clark', 'Konstantinos Meichanetzidis', 'Giovanni De Felice', 'Alexis Toumi', 'Robin Lorenz', 'Anna Pearson', 'Richie Yeung', 'Ian Fan', 'Dimitri Kartsaklis'] | 2021-10-08 | null | null | null | null | ['tensor-networks'] | ['methodology'] | [ 3.69998366e-01 3.38073820e-01 3.05442989e-01 -6.30379200e-01
-8.47350478e-01 -1.03037679e+00 6.61969721e-01 4.81494457e-01
-3.18629414e-01 4.09934670e-01 2.51404852e-01 -9.31872785e-01
2.20649883e-01 -1.25671327e+00 -6.70612931e-01 -4.06605572e-01
-7.35460892e-02 4.62917596e-01 1.30930677e-01 -6.51694417e-01
3.64564419e-01 4.78988945e-01 -1.15047550e+00 6.60379767e-01
5.63668132e-01 5.29051840e-01 -3.02913994e-01 1.19676399e+00
-2.44186491e-01 8.67402494e-01 -3.02078545e-01 -1.03608334e+00
3.93964648e-01 -1.78131074e-01 -1.12356341e+00 -7.90278673e-01
3.96357626e-01 -2.82227844e-01 -6.94427848e-01 1.28380787e+00
3.16345155e-01 8.08712617e-02 3.07844639e-01 -7.76799262e-01
-6.11678600e-01 1.07282710e+00 5.20561278e-01 3.31192672e-01
8.78292322e-01 5.30831933e-01 1.41544163e+00 -3.63678783e-01
7.79804230e-01 1.42153966e+00 5.34256876e-01 5.59300303e-01
-1.74158096e+00 -4.37506050e-01 -9.55871105e-01 1.61935359e-01
-1.19316888e+00 -5.86230278e-01 2.12788805e-01 -2.54456818e-01
1.99254453e+00 9.81800035e-02 3.02480787e-01 7.18263507e-01
7.09865332e-01 1.16129003e-01 1.12218547e+00 -6.21888041e-01
2.41244629e-01 -2.65213177e-02 8.21656406e-01 1.08280277e+00
-2.53730953e-01 1.76467553e-01 -4.42691624e-01 -5.48005521e-01
4.57117230e-01 -6.09721065e-01 7.05377281e-01 1.59850448e-01
-1.39606524e+00 8.81927848e-01 4.51994210e-01 3.65159921e-02
9.03323218e-02 4.04566348e-01 7.46846318e-01 5.83371460e-01
-1.57469496e-01 6.08792067e-01 -2.34538525e-01 -2.85862148e-01
-3.50363106e-01 5.65469027e-01 1.17350745e+00 1.31898820e+00
9.79148269e-01 -4.76467818e-01 -3.94004822e-01 4.83685434e-01
2.94965170e-02 3.91075104e-01 -2.92395115e-01 -1.16026914e+00
4.42637265e-01 1.84997588e-01 -3.65506440e-01 -6.15984917e-01
-6.71600223e-01 5.21038100e-02 -6.06713176e-01 -2.55224556e-01
3.70500624e-01 4.99957614e-02 -6.49013042e-01 1.42411184e+00
8.19317549e-02 -2.95542777e-01 4.19129848e-01 5.41723371e-01
1.15471601e+00 8.61272395e-01 2.54741311e-01 1.38914242e-01
1.77559769e+00 -7.04945326e-01 -3.69397938e-01 -6.36845380e-02
1.13268733e+00 -8.44741702e-01 8.78558457e-01 3.19390893e-01
-1.28074193e+00 -2.75972575e-01 -1.02304590e+00 -9.05213058e-01
-6.75062358e-01 -3.58610183e-01 1.36212194e+00 7.83023596e-01
-1.20265937e+00 1.38864708e+00 -6.76309586e-01 -4.93092567e-01
1.05787039e-01 5.26702523e-01 -3.12884450e-01 5.81727363e-02
-1.58791256e+00 1.04185867e+00 6.53887928e-01 -7.71859884e-02
-4.61722910e-01 -4.94653940e-01 -9.45861697e-01 2.19053790e-01
-1.83540583e-01 -1.19226825e+00 1.82502854e+00 5.60256653e-02
-1.92850173e+00 1.13442671e+00 -2.20085144e-01 -8.01086247e-01
-1.60710625e-02 4.26294863e-01 -4.32602018e-01 3.31015557e-01
-1.18530607e-02 4.54780281e-01 2.35153839e-01 1.32444531e-01
-2.51304686e-01 -7.70091042e-02 7.74608970e-01 -1.35337606e-01
4.30020273e-01 6.58122897e-01 -3.01100671e-01 7.59038255e-02
1.01264521e-01 -7.51518846e-01 -4.38146710e-01 -5.09713709e-01
-6.14773333e-01 -5.57637870e-01 -1.11455619e-01 -4.25979435e-01
1.10453844e+00 -1.95359409e+00 1.01050302e-01 2.74983823e-01
1.05038136e-01 4.23458815e-02 3.30040120e-02 9.84238625e-01
-9.47828442e-02 2.30546966e-01 -4.29306746e-01 -1.53846502e-01
6.29638672e-01 2.63773531e-01 -3.21167171e-01 3.16048086e-01
5.94504595e-01 1.23481679e+00 -1.17351413e+00 -5.14830470e-01
4.60935235e-01 1.05256677e-01 -8.88077259e-01 -1.34837553e-01
-4.80180413e-01 4.40183431e-01 -2.92946428e-01 4.12052453e-01
7.53875613e-01 -2.65437514e-02 1.30177587e-01 -1.65984541e-01
-3.70880485e-01 1.27684438e+00 -1.11865914e+00 1.91712427e+00
-3.73431265e-01 5.98005176e-01 1.23691186e-01 -6.33453190e-01
3.65346313e-01 -4.54678684e-02 -3.11814159e-01 -7.71475136e-01
2.34446585e-01 2.03685954e-01 5.22906594e-02 -5.55447996e-01
8.25183392e-01 -4.48034406e-01 -6.05849087e-01 7.10122049e-01
4.49032634e-01 -8.83657396e-01 8.98404479e-01 8.77131402e-01
1.39982378e+00 6.55529788e-03 2.60434508e-01 -2.54105896e-01
6.82211339e-01 2.11567879e-01 7.41412565e-02 1.16013825e+00
-2.57057518e-01 1.76537767e-01 9.30660605e-01 -5.11415303e-01
-1.54679143e+00 -1.32402718e+00 -5.35028994e-01 1.07049191e+00
-1.87371165e-01 -1.31863189e+00 -7.95219958e-01 5.71259186e-02
-2.89882064e-01 1.10092783e+00 -1.47838444e-01 -1.27056018e-01
-4.49660689e-01 -1.05241942e+00 1.06503963e+00 1.03154786e-01
2.63620615e-01 -1.24212492e+00 -4.28491496e-02 1.75463557e-01
1.91137448e-01 -1.47976804e+00 1.24066442e-01 3.36291611e-01
-8.29654157e-01 -8.42321992e-01 4.72385913e-01 -6.10296190e-01
4.58909571e-01 -2.50649869e-01 1.18286991e+00 -2.16335580e-01
-4.26125288e-01 5.08640101e-03 -2.90206149e-02 -2.24136803e-02
-1.23185408e+00 -6.61050081e-02 5.05901240e-02 -7.27897584e-01
4.69204783e-01 -6.00220084e-01 -2.31781647e-01 -5.86095691e-01
-1.01486886e+00 2.12902337e-01 4.14912581e-01 6.82215273e-01
7.95962736e-02 -1.92532226e-01 -3.26954275e-01 -1.12693357e+00
9.00273561e-01 1.46857894e-03 -8.78565550e-01 8.50759819e-03
-6.83249384e-02 4.29134429e-01 1.11921442e+00 4.59859163e-01
-6.91090286e-01 8.88243467e-02 -6.32823467e-01 5.36131978e-01
-2.08448440e-01 5.22757709e-01 1.59749404e-01 -2.24882141e-01
8.95930529e-01 1.36903226e-01 -3.71574819e-01 -2.00288504e-01
9.89971519e-01 7.30813503e-01 5.72170317e-01 -9.20333564e-01
8.14580739e-01 1.98771089e-01 3.89140040e-01 -1.01949060e+00
-7.88841963e-01 -2.89050013e-01 -1.00264990e+00 4.96482998e-01
8.68408382e-01 -6.97971642e-01 -1.18158364e+00 3.44249040e-01
-1.49961054e+00 -1.20274402e-01 -1.99591085e-01 2.73652583e-01
-6.08588815e-01 7.29572415e-01 -1.38111198e+00 -5.83305120e-01
-5.92842758e-01 -1.39041853e+00 1.02431393e+00 2.58204401e-01
-7.44023100e-02 -8.56553257e-01 1.58405229e-01 2.79371560e-01
1.57062098e-01 -1.29088059e-01 1.35149431e+00 -5.55694759e-01
-8.27460408e-01 -2.74369866e-01 -5.40560663e-01 3.08226496e-01
-7.18510628e-01 4.00054306e-01 -1.10100698e+00 1.16525665e-01
-4.41564023e-01 -5.46226561e-01 7.60295630e-01 -3.09047997e-01
8.95193994e-01 -1.47025183e-01 2.95350235e-02 7.28111625e-01
1.27200377e+00 -4.42077219e-01 8.03283036e-01 1.47280693e-01
5.27899086e-01 3.33425134e-01 -7.37182871e-02 1.05210334e-01
5.34703314e-01 3.45866382e-01 1.15232915e-01 4.87854093e-01
3.80460441e-01 -2.45615155e-01 7.29690433e-01 1.34783804e+00
1.43648371e-01 6.58361197e-01 -8.48092973e-01 -1.58034682e-01
-1.52584302e+00 -1.10314155e+00 -5.02435267e-01 1.97105289e+00
8.35951149e-01 3.29000980e-01 -2.66512972e-04 -2.36989245e-01
4.57796693e-01 3.49855796e-03 -4.01990227e-02 -1.46010947e+00
-5.42597622e-02 1.13555992e+00 6.77634954e-01 7.13998020e-01
-1.25492966e+00 1.45001101e+00 8.02068233e+00 6.62325621e-01
-5.70714414e-01 2.45231274e-03 -1.14085237e-02 3.21971983e-01
-2.42594376e-01 8.02223027e-01 -6.10204279e-01 -3.87375355e-02
1.70547616e+00 -3.51173341e-01 1.15375268e+00 4.81227905e-01
1.27625734e-01 5.47674410e-02 -1.43042147e+00 9.46379662e-01
-6.74056113e-01 -1.81671834e+00 -3.82095985e-02 -4.70287204e-01
1.98556840e-01 6.63887620e-01 -6.23498023e-01 6.52988434e-01
5.97727001e-01 -8.93867850e-01 5.95299602e-01 4.04821336e-01
6.37047231e-01 -6.00311756e-01 5.01759052e-01 2.22790584e-01
-9.64036286e-01 -1.37357995e-01 -8.37501466e-01 -7.08620489e-01
1.32726043e-01 4.19560075e-01 -6.97519481e-01 8.39715719e-01
5.03442287e-01 6.00387454e-01 -7.34569252e-01 5.94469488e-01
-6.00304246e-01 3.33903223e-01 -6.60370052e-01 -3.90432000e-01
3.73112291e-01 -7.75431395e-01 3.91490102e-01 1.76603532e+00
-2.17638507e-01 5.47675252e-01 -1.62319779e-01 1.47761774e+00
-2.08008870e-01 1.18567556e-01 -6.56723619e-01 -5.61628819e-01
4.47236449e-01 1.57783580e+00 -5.62433422e-01 -7.45414138e-01
-4.61395353e-01 8.55798423e-01 4.89179105e-01 -7.71737844e-02
-7.00442731e-01 -1.09171104e+00 4.34710294e-01 -6.07005179e-01
-1.56927258e-01 -7.04816163e-01 -3.57802659e-01 -1.57924640e+00
-2.51926154e-01 -8.63008618e-01 2.86859274e-01 -7.68970251e-01
-1.23469150e+00 1.89023718e-01 -6.05501384e-02 -4.92743909e-01
-5.36209755e-02 -1.03275299e+00 -8.98231924e-01 1.12207055e+00
-1.09217608e+00 -7.10651755e-01 2.70795345e-01 1.87775135e-01
-1.86368600e-01 2.99892575e-01 1.43802202e+00 3.60691607e-01
-8.61569107e-01 2.35232398e-01 1.88420251e-01 4.07564312e-01
2.54354775e-01 -1.40508127e+00 1.38241816e+00 7.22842515e-01
-2.69116517e-02 1.43937063e+00 8.45675409e-01 -2.44237438e-01
-2.13306403e+00 -7.26777077e-01 1.30329967e+00 -6.38752103e-01
1.38140452e+00 -1.00334823e+00 -4.12461877e-01 7.83745348e-01
1.92604229e-01 -2.89928168e-01 6.30468369e-01 1.89958453e-01
-5.86125314e-01 1.31931394e-01 -1.00293195e+00 6.02907598e-01
9.77605104e-01 -1.50966597e+00 -7.69470394e-01 9.41316605e-01
7.41734028e-01 -6.35033965e-01 -1.10000336e+00 -1.63906261e-01
6.15985632e-01 -1.09285402e+00 8.02561164e-01 -8.44746530e-01
3.86448920e-01 -2.60367870e-01 -1.04365893e-01 -8.20671201e-01
-4.16542113e-01 -1.24746239e+00 1.49448112e-01 8.39174449e-01
5.34113646e-01 -7.05548346e-01 4.26368147e-01 7.67535329e-01
-2.22350791e-01 -4.91047055e-02 -8.76833260e-01 -2.53742784e-01
5.34242272e-01 -1.02279055e+00 4.50038493e-01 5.03481090e-01
8.34297538e-01 9.68917727e-01 2.82962918e-01 2.17279449e-01
6.05027974e-01 1.34190500e-01 8.34175944e-01 -7.77478456e-01
-3.67733061e-01 -6.46162391e-01 -6.43193185e-01 -9.10308063e-01
1.48208648e-01 -1.71542037e+00 -1.23020269e-01 -1.46485376e+00
3.44843388e-01 -4.20984887e-02 1.71179458e-01 1.47372857e-01
2.35231847e-01 1.30272314e-01 2.66088367e-01 7.23298416e-02
-6.49413705e-01 2.33395264e-01 8.57207298e-01 -2.95071930e-01
7.55068362e-02 -1.97965071e-01 -4.87604499e-01 5.16434848e-01
5.50531507e-01 -5.23708820e-01 1.67635664e-01 -3.61920059e-01
7.67822146e-01 -1.00641526e-01 4.71374035e-01 -8.04822505e-01
3.17069739e-01 1.37390718e-01 -9.61559564e-02 -6.02512538e-01
2.78147608e-01 5.52097969e-02 -1.91071585e-01 5.09300888e-01
-4.04653728e-01 8.43797624e-02 1.89513057e-01 1.50567785e-01
1.69877827e-01 -4.49714899e-01 8.41756105e-01 -5.11924207e-01
-5.41469932e-01 3.17405947e-02 -5.15169322e-01 6.81902794e-03
6.26212299e-01 2.77205616e-01 -6.30640566e-01 -6.15802184e-02
-1.03952730e+00 3.05290699e-01 5.77403009e-01 3.55190895e-02
1.11769475e-01 -7.99939692e-01 -4.18687165e-01 2.61511505e-01
2.27449566e-01 -2.17178658e-01 1.51213765e-01 8.61886382e-01
-1.39416718e+00 1.13395488e+00 -8.83937702e-02 -3.75191569e-01
-7.98003793e-01 6.39780939e-01 5.01627982e-01 -3.00376445e-01
-7.00766265e-01 5.99728346e-01 -2.49973938e-01 -1.08852959e+00
-1.96499765e-01 -9.62049484e-01 2.34456435e-01 -5.85008800e-01
6.58811629e-01 3.54183555e-01 3.96103948e-01 -5.84569693e-01
-5.16338229e-01 3.39617014e-01 1.97986975e-01 -5.81068965e-03
1.10917592e+00 1.21367030e-01 -9.79658604e-01 4.51193660e-01
1.33657205e+00 -2.37241447e-01 -3.04406732e-01 -3.02569687e-01
2.28460446e-01 4.33600336e-01 -2.69286454e-01 -4.24009889e-01
7.89996050e-03 1.26322389e+00 6.70301914e-02 3.74488592e-01
4.22757417e-01 5.82834147e-02 9.36603785e-01 1.30331004e+00
6.21618092e-01 -1.05351198e+00 -7.18555391e-01 1.18007565e+00
1.97464153e-01 -7.25585938e-01 1.40640706e-01 -5.83229125e-01
-1.51978642e-01 1.77127266e+00 -2.22179934e-01 -4.62317318e-01
4.84844953e-01 3.54726799e-02 -1.30808443e-01 -4.33583111e-01
-8.88046086e-01 -2.31312394e-01 -1.47167489e-01 2.52250969e-01
6.76453948e-01 3.42901617e-01 -2.47519940e-01 4.23079282e-01
-8.95354927e-01 -2.20978498e-01 1.06920218e+00 1.07364488e+00
-2.29985744e-01 -1.47683799e+00 -3.39215368e-01 3.09028566e-01
-5.52606523e-01 -7.07574248e-01 -5.15576720e-01 3.54219258e-01
-9.91835892e-02 9.91896749e-01 -2.26162747e-02 -5.41786909e-01
4.94320959e-01 4.35402542e-01 1.18392122e+00 -1.03125167e+00
-8.44853044e-01 -4.66536850e-01 5.49619436e-01 -7.14313865e-01
-1.62099406e-01 -4.94576037e-01 -1.67728138e+00 -1.00580561e+00
-1.39949694e-01 2.51433313e-01 8.33693981e-01 1.01078093e+00
4.63749230e-01 4.30406302e-01 3.60064328e-01 -8.88924658e-01
-8.96500587e-01 -9.72316146e-01 -5.17854512e-01 1.64526284e-01
1.82032287e-01 3.06719150e-02 -2.34535560e-01 -1.58779681e-01] | [5.578176021575928, 4.947471618652344] |
6a8e8394-2f64-455c-a6a6-afcb5ed34567 | unsupervised-abstractive-meeting | 1805.05271 | null | http://arxiv.org/abs/1805.05271v2 | http://arxiv.org/pdf/1805.05271v2.pdf | Unsupervised Abstractive Meeting Summarization with Multi-Sentence Compression and Budgeted Submodular Maximization | We introduce a novel graph-based framework for abstractive meeting speech
summarization that is fully unsupervised and does not rely on any annotations.
Our work combines the strengths of multiple recent approaches while addressing
their weaknesses. Moreover, we leverage recent advances in word embeddings and
graph degeneracy applied to NLP to take exterior semantic knowledge into
account, and to design custom diversity and informativeness measures.
Experiments on the AMI and ICSI corpus show that our system improves on the
state-of-the-art. Code and data are publicly available, and our system can be
interactively tested. | ['Jean-Pierre Lorré', 'Michalis Vazirgiannis', 'Antoine Jean-Pierre Tixier', 'Wensi Ding', 'Zekun Zhang', 'Polykarpos Meladianos', 'Guokan Shang'] | 2018-05-14 | unsupervised-abstractive-meeting-1 | https://aclanthology.org/P18-1062 | https://aclanthology.org/P18-1062.pdf | acl-2018-7 | ['dialogue-understanding', 'meeting-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 3.99267405e-01 4.69228595e-01 -3.42020422e-01 -3.46364260e-01
-9.25612032e-01 -5.81391037e-01 5.66841304e-01 5.15683293e-01
-2.18877986e-01 7.17270672e-01 1.14326417e+00 -3.52302670e-01
-2.24586517e-01 -4.46755230e-01 -1.83814421e-01 -3.33381265e-01
-2.35677660e-01 7.44213939e-01 3.72774035e-01 -5.39882421e-01
3.72682542e-01 2.85773069e-01 -1.28679192e+00 1.17536835e-01
9.70161378e-01 4.90892559e-01 -1.58180088e-01 9.54581916e-01
-2.63348788e-01 6.85597301e-01 -9.62984681e-01 -1.96372703e-01
-3.52366924e-01 -3.71495813e-01 -9.37251866e-01 1.24349281e-01
4.50052500e-01 9.42438245e-02 -6.40738964e-01 6.39105618e-01
1.06139147e+00 2.61104584e-01 7.18241870e-01 -9.72012639e-01
-8.86244655e-01 8.77946258e-01 -1.50011420e-01 4.23463881e-01
7.35935986e-01 -1.95992619e-01 1.42192817e+00 -7.11400151e-01
5.58891892e-01 1.00467575e+00 6.58422589e-01 6.42479539e-01
-1.02736008e+00 -3.33300143e-01 2.69158900e-01 2.35249907e-01
-1.45447314e+00 -9.02239799e-01 7.16845453e-01 -1.17859021e-01
1.61172426e+00 5.11957228e-01 6.57940507e-01 1.22277594e+00
-2.76986003e-01 8.76619279e-01 2.67354369e-01 -6.36373043e-01
2.66603410e-01 -2.33974352e-01 3.64473194e-01 9.65408206e-01
4.04888004e-01 -5.92742264e-01 -8.02236080e-01 -5.18878758e-01
1.66986242e-01 -3.89309168e-01 -7.85008609e-01 -2.15297431e-01
-1.20542896e+00 7.39250243e-01 -1.27117321e-01 5.62727809e-01
-2.85312951e-01 9.36009437e-02 4.45637375e-01 4.67238389e-02
8.78307641e-01 6.18337512e-01 -1.67556062e-01 -4.65887100e-01
-1.24106050e+00 1.74283355e-01 1.27115369e+00 1.19540775e+00
4.37068850e-01 3.13710153e-01 -3.77916932e-01 9.21314597e-01
1.34101406e-01 2.89052904e-01 6.04810953e-01 -9.29445565e-01
6.25036359e-01 7.18485117e-01 -2.80833811e-01 -8.55989754e-01
-5.47630012e-01 -3.06397557e-01 -5.65819561e-01 -6.09613895e-01
-1.84809625e-01 -1.84082523e-01 -1.03662157e+00 1.35993862e+00
-7.77606294e-02 3.19031954e-01 2.77577430e-01 4.90080297e-01
1.28227198e+00 6.96296275e-01 -4.09546673e-01 -4.67127293e-01
1.12487662e+00 -1.17743337e+00 -1.17727208e+00 -3.32361996e-01
6.38286471e-01 -6.48037553e-01 9.69225883e-01 2.57973492e-01
-1.27851701e+00 -4.81729992e-02 -1.31882286e+00 -7.97837302e-02
-3.10520142e-01 -9.70662758e-02 6.09500468e-01 7.98113465e-01
-1.57069981e+00 5.62298357e-01 -7.58897185e-01 -8.67991209e-01
3.54746491e-01 3.40410084e-01 -2.54499525e-01 1.26789600e-01
-1.06352043e+00 7.15517282e-01 6.04857862e-01 -3.35341692e-01
-2.75656044e-01 -5.19662857e-01 -1.04084957e+00 1.28091916e-01
6.56949580e-01 -1.06981862e+00 1.37184179e+00 -5.16378105e-01
-1.92057741e+00 4.47763234e-01 -2.97357738e-01 -6.27199173e-01
4.96420339e-02 -1.49668992e-01 -5.76831877e-01 5.79226375e-01
-1.97997987e-01 3.75035018e-01 4.51916695e-01 -1.01212132e+00
-1.40816569e-01 3.69468741e-02 9.15284380e-02 5.54386377e-01
-7.69446850e-01 4.39308286e-02 -5.68824172e-01 -7.05445766e-01
-2.94032060e-02 -7.50563979e-01 -8.48238468e-02 -6.47767186e-01
-6.40448451e-01 -3.85906577e-01 6.24019802e-01 -6.79466963e-01
1.77738142e+00 -1.89262855e+00 4.96638983e-01 -1.29815843e-02
3.48457545e-01 4.33722079e-01 -1.27764419e-01 1.27400362e+00
2.80223399e-01 5.37397683e-01 -4.51296091e-01 -4.87223148e-01
1.85457081e-01 3.70016515e-01 -4.49018478e-02 2.72456974e-01
1.14531115e-01 9.63881254e-01 -9.39787149e-01 -6.00847423e-01
1.17251553e-01 4.18806940e-01 -7.01194584e-01 1.62303239e-01
-1.49118990e-01 1.02164015e-01 -4.84283507e-01 4.43163931e-01
1.32915512e-01 -1.19305201e-01 3.76431495e-01 -5.83837591e-02
-7.22143725e-02 9.12529290e-01 -8.43804359e-01 2.07829762e+00
-3.94733906e-01 6.10791564e-01 -1.01470836e-01 -1.21787798e+00
6.93165302e-01 6.13968551e-01 3.60996366e-01 7.80755654e-02
7.08782822e-02 2.30916306e-01 -8.72265846e-02 -5.07913828e-01
9.47063625e-01 1.75532252e-01 -7.50486255e-02 3.67884696e-01
4.94895935e-01 -5.12886286e-01 3.73770148e-01 6.28646314e-01
1.38373756e+00 -3.17543149e-01 5.54107189e-01 -5.01902223e-01
4.22690809e-01 -1.30701348e-01 3.63642901e-01 8.00279915e-01
4.37467583e-02 9.00957525e-01 5.27339578e-01 1.30916461e-01
-8.65546584e-01 -1.09383273e+00 2.75520593e-01 1.04391646e+00
-2.28407085e-01 -1.00141358e+00 -7.48758972e-01 -8.56615663e-01
-1.75105527e-01 1.05238533e+00 -4.03289258e-01 -5.49595989e-02
-5.69070995e-01 -5.41079223e-01 8.43446851e-01 6.36801541e-01
6.92545995e-02 -8.46443653e-01 -6.48955815e-03 2.30233878e-01
-2.95708805e-01 -1.23922539e+00 -8.42627883e-01 -1.31170318e-01
-8.36382747e-01 -8.63769114e-01 -6.84896588e-01 -9.55503941e-01
3.53326529e-01 2.74921268e-01 1.33564699e+00 -9.23605859e-02
-9.14109200e-02 1.02443373e+00 -6.23513460e-01 -4.71082121e-01
-4.27961618e-01 6.63931906e-01 8.37259889e-02 -2.73271710e-01
1.67254299e-01 -8.10218155e-01 -3.75754774e-01 -2.59332448e-01
-9.26003993e-01 -2.09635317e-01 3.14125031e-01 6.43202782e-01
2.00790063e-01 -1.83529198e-01 1.06433153e+00 -1.10690367e+00
1.11542916e+00 -1.88262135e-01 1.31680295e-01 2.53649503e-01
-6.18443608e-01 1.89296320e-01 4.12931144e-01 -1.12135962e-01
-8.84794295e-01 -4.17887628e-01 -4.34642583e-01 -7.19349980e-02
1.59876496e-01 7.95530319e-01 -2.03073025e-01 1.95682749e-01
4.60617840e-01 2.03074515e-01 -1.86281383e-01 -3.72661889e-01
6.67076826e-01 1.04571557e+00 4.32666689e-01 -1.67074263e-01
4.97107893e-01 3.75854552e-01 -4.45397019e-01 -1.46046042e+00
-6.59454226e-01 -8.16123903e-01 -4.38023359e-01 3.38807069e-02
8.09359431e-01 -9.38365340e-01 -3.60561222e-01 -1.21166147e-01
-1.25646317e+00 -3.55643481e-02 -5.37299693e-01 4.53074664e-01
-5.00586271e-01 8.01202238e-01 -5.01304567e-01 -8.07602584e-01
-5.82513988e-01 -7.00173557e-01 1.07564020e+00 1.02377281e-01
-5.25165617e-01 -1.34819543e+00 1.75449699e-01 4.09047067e-01
4.47982192e-01 -2.21541077e-02 7.63390720e-01 -1.29909945e+00
-1.52161434e-01 -2.24850222e-01 9.74747092e-02 3.75451744e-01
2.40524679e-01 7.09298253e-02 -9.23666000e-01 -2.16505691e-01
-2.11245641e-01 -1.48173720e-01 1.14373732e+00 4.43041950e-01
8.56667519e-01 -6.68787003e-01 -3.05291653e-01 2.20031708e-01
9.57200646e-01 -3.62644672e-01 5.76858640e-01 -3.81067991e-02
8.95904303e-01 5.76228678e-01 7.02968836e-02 5.20160496e-01
6.73755288e-01 6.34310007e-01 1.21112853e-01 1.12234950e-01
-2.95337081e-01 -3.37362468e-01 3.92135561e-01 1.98980486e+00
-1.01573549e-01 -9.00337994e-01 -7.53065526e-01 9.41483140e-01
-2.08243179e+00 -1.04037023e+00 -7.49404505e-02 1.87120485e+00
7.78355479e-01 1.29942358e-01 1.94084063e-01 3.50089759e-01
5.02203286e-01 7.72819042e-01 1.96853206e-02 -8.17104399e-01
-1.96739197e-01 4.91835296e-01 6.93609774e-01 6.27985120e-01
-9.38951790e-01 1.06053817e+00 7.22958708e+00 7.82846689e-01
-4.99649435e-01 1.05439462e-01 -8.58393908e-02 -6.35860786e-02
-8.18316579e-01 -1.27889171e-01 -4.63008970e-01 -7.67838955e-02
1.27159059e+00 -6.27844155e-01 3.90129656e-01 3.11187297e-01
1.83554456e-01 2.88031995e-01 -1.03966236e+00 8.23661208e-01
6.78570092e-01 -1.64637852e+00 2.09633067e-01 -1.68714523e-01
6.02065265e-01 1.17787741e-01 -1.21298097e-01 3.14767063e-01
3.05051893e-01 -9.32055116e-01 3.09798986e-01 3.32160026e-01
5.11549056e-01 -7.29089200e-01 8.31394494e-01 7.29997754e-02
-1.18525970e+00 3.40515554e-01 -1.87403202e-01 -1.66549027e-01
4.01919931e-01 4.88680452e-01 -1.09824443e+00 1.24826360e+00
2.84931600e-01 8.97612631e-01 -4.77749139e-01 7.60914922e-01
-4.11254823e-01 1.00863087e+00 -2.35491395e-01 -3.58415067e-01
1.85108125e-01 1.16878927e-01 1.03768516e+00 1.70948756e+00
3.02713960e-01 5.56253158e-02 5.57161927e-01 3.20629239e-01
-3.17516953e-01 4.34686333e-01 -9.05593395e-01 -4.31368947e-01
6.12688184e-01 8.66885483e-01 -4.41492856e-01 -5.73650956e-01
-4.13570017e-01 1.17174554e+00 2.06683531e-01 4.84250247e-01
-5.73946118e-01 -7.14410782e-01 5.06845474e-01 3.35123427e-02
5.56518972e-01 -4.66145217e-01 -1.40728101e-01 -1.24769139e+00
8.11162964e-02 -8.26911688e-01 5.20203888e-01 -3.68832767e-01
-1.20461571e+00 6.22654855e-01 2.15299457e-01 -9.24236178e-01
-4.75277901e-01 -2.90276825e-01 -7.09952235e-01 3.88322949e-01
-1.49289358e+00 -9.41743135e-01 1.72586218e-02 1.76514149e-01
8.06596994e-01 -3.41794431e-01 1.19718969e+00 2.46741802e-01
-5.99761665e-01 6.87724888e-01 5.59714399e-02 -1.18315279e-01
5.36388040e-01 -1.29319310e+00 7.92962193e-01 8.88736427e-01
5.31131268e-01 4.06631678e-01 8.16828132e-01 -4.91156250e-01
-1.38803780e+00 -9.83995855e-01 1.38387084e+00 -3.33350927e-01
9.56533074e-01 -4.43603963e-01 -8.36359143e-01 7.72120178e-01
8.25081706e-01 -4.06513333e-01 9.64196742e-01 3.45249802e-01
-1.60645500e-01 1.67083457e-01 -7.76179314e-01 7.43499517e-01
1.43256998e+00 -3.18654835e-01 -1.12143028e+00 5.56024134e-01
1.20510757e+00 -3.43362987e-01 -6.62467957e-01 4.08872753e-01
1.73973441e-01 -7.42069423e-01 8.35188746e-01 -4.61114913e-01
-1.13579050e-01 -7.65330866e-02 -1.46879733e-01 -1.48041403e+00
-2.17434257e-01 -1.10680795e+00 -3.36014599e-01 1.51884842e+00
7.57778406e-01 -9.14976716e-01 5.03437638e-01 7.60094300e-02
-8.36192906e-01 -5.69600880e-01 -1.07589078e+00 -9.38624740e-01
-2.08084375e-01 -3.63053352e-01 4.97353286e-01 7.54261196e-01
6.42141104e-01 7.94228077e-01 -2.55953997e-01 2.11127505e-01
2.16329008e-01 -2.34279901e-01 6.76193297e-01 -1.14897633e+00
-2.64810532e-01 -5.26493192e-01 -6.03824079e-01 -1.04852200e+00
4.21766490e-01 -1.17832017e+00 -1.16014689e-01 -2.31383777e+00
-1.30882561e-01 2.23186269e-01 -2.17915297e-01 4.56900120e-01
-1.37716815e-01 2.06607476e-01 1.00195426e-02 -1.75678022e-02
-9.42751646e-01 8.80093932e-01 5.95980406e-01 -4.28498566e-01
-4.45065111e-01 -2.03480586e-01 -9.26727057e-01 6.34643018e-01
9.54135060e-01 -4.09108549e-01 -6.45907104e-01 -4.98264015e-01
-1.51143134e-01 -2.20770091e-01 -1.02519445e-01 -1.14654553e+00
3.52431446e-01 2.73819357e-01 -4.06683683e-01 -4.99470562e-01
3.71877044e-01 -2.71129996e-01 -2.93027431e-01 -1.33840293e-02
-4.30329353e-01 1.53518468e-01 3.83749545e-01 7.85516620e-01
-2.37557635e-01 -2.59312779e-01 9.31810290e-02 1.29229367e-01
-3.48509103e-01 1.28081605e-01 -5.52176774e-01 6.16251111e-01
5.92038333e-01 -4.47514690e-02 -2.59997398e-01 -8.18832159e-01
-3.65800202e-01 2.58037150e-01 3.95316929e-01 5.63945830e-01
8.31377923e-01 -1.15474856e+00 -1.06395805e+00 -1.13650858e-01
3.90360385e-01 -3.71273011e-01 3.02859731e-02 7.80566633e-01
-5.32382429e-01 3.69482368e-01 5.69300830e-01 -4.60996568e-01
-1.23156846e+00 3.52557421e-01 -1.37908012e-01 -3.26398671e-01
-9.39010739e-01 6.78892076e-01 -1.85444757e-01 -4.16907728e-01
4.18853283e-01 -4.08912867e-01 -4.16797936e-01 8.27436820e-02
4.36514616e-01 3.01408738e-01 2.58891255e-01 -4.74960238e-01
-8.23895037e-01 3.25773001e-01 -9.81939882e-02 -4.46788847e-01
1.45786226e+00 -1.78269416e-01 3.40729468e-02 8.21984589e-01
1.18157268e+00 4.99806345e-01 -4.69353706e-01 -9.46151689e-02
2.56660908e-01 -2.60969065e-02 9.51158926e-02 -5.51521301e-01
-7.27380931e-01 6.35941505e-01 -1.04221568e-01 4.82401490e-01
9.48412657e-01 6.32099882e-02 8.44198525e-01 4.51270700e-01
-1.06033899e-01 -1.21755457e+00 3.61176915e-02 6.09960437e-01
8.47209752e-01 -7.19308436e-01 3.88681531e-01 -5.05624533e-01
-9.17971730e-01 1.01907611e+00 -9.61169004e-02 -1.60946861e-01
4.83688086e-01 2.66933650e-01 6.81577772e-02 -3.22676420e-01
-9.03087497e-01 -4.41837013e-01 5.94995856e-01 7.58525372e-01
8.06495070e-01 8.52110907e-02 -6.78028166e-01 5.48458636e-01
-2.87416995e-01 -3.19686294e-01 7.83571601e-01 9.75065053e-01
-5.18406451e-01 -1.26132119e+00 1.94912374e-01 4.28661257e-01
-5.87772131e-01 -5.68102658e-01 -7.30090678e-01 7.02054262e-01
-5.15022576e-01 1.40879321e+00 -1.82098880e-01 -3.41394156e-01
6.32791162e-01 4.16232139e-01 4.82390046e-01 -1.08689344e+00
-6.57575548e-01 -3.69995758e-02 8.28134477e-01 -2.88924843e-01
-5.78465641e-01 -5.69269776e-01 -1.17996061e+00 -1.71663880e-01
-4.62347984e-01 3.72421384e-01 4.48773563e-01 7.40638435e-01
6.73972130e-01 1.04842210e+00 3.92821729e-01 -9.57108855e-01
-3.48424941e-01 -1.11851954e+00 -3.64915103e-01 7.23712817e-02
3.07477564e-01 -3.43516827e-01 -5.04590213e-01 -1.70979097e-01] | [12.52270793914795, 9.510708808898926] |
b51e9cb0-ccac-4802-a2ea-16f95420e00a | deep-cognitive-reasoning-network-for-multi | null | null | https://aclanthology.org/2021.findings-acl.19 | https://aclanthology.org/2021.findings-acl.19.pdf | Deep Cognitive Reasoning Network for Multi-hop Question Answering over Knowledge Graphs | null | ['Jie Wang', 'Feng Wu', 'Zhanqiu Zhang', 'Jianyu Cai'] | null | null | null | null | findings-acl-2021-8 | ['multi-hop-question-answering'] | ['knowledge-base'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.251688003540039, 3.7372524738311768] |
4081fa85-fe5f-4510-aa43-a164a6e966a2 | the-first-shared-task-on-discourse-1 | 2005.13399 | null | https://arxiv.org/abs/2005.13399v1 | https://arxiv.org/pdf/2005.13399v1.pdf | The First Shared Task on Discourse Representation Structure Parsing | The paper presents the IWCS 2019 shared task on semantic parsing where the goal is to produce Discourse Representation Structures (DRSs) for English sentences. DRSs originate from Discourse Representation Theory and represent scoped meaning representations that capture the semantics of negation, modals, quantification, and presupposition triggers. Additionally, concepts and event-participants in DRSs are described with WordNet synsets and the thematic roles from VerbNet. To measure similarity between two DRSs, they are represented in a clausal form, i.e. as a set of tuples. Participant systems were expected to produce DRSs in this clausal form. Taking into account the rich lexical information, explicit scope marking, a high number of shared variables among clauses, and highly-constrained format of valid DRSs, all these makes the DRS parsing a challenging NLP task. The results of the shared task displayed improvements over the existing state-of-the-art parser. | ['Rik van Noord', 'Johan Bos', 'Hessel Haagsma', 'Lasha Abzianidze'] | 2020-05-27 | the-first-shared-task-on-discourse | https://aclanthology.org/W19-1201 | https://aclanthology.org/W19-1201.pdf | ws-2019-5 | ['drs-parsing'] | ['natural-language-processing'] | [ 3.64446342e-01 9.34006631e-01 -4.31120694e-01 -6.67133212e-01
-6.99359596e-01 -7.15143025e-01 6.62751615e-01 6.76359475e-01
-1.09524682e-01 8.98638010e-01 9.69037056e-01 -3.17065358e-01
1.32547364e-01 -1.10897028e+00 -4.80713755e-01 -1.58379734e-01
-1.16431946e-02 5.57940364e-01 6.34278595e-01 -8.59556973e-01
1.84277594e-01 2.26745382e-02 -1.51933980e+00 1.11844432e+00
5.15425384e-01 7.16491997e-01 4.78873372e-01 2.07365170e-01
-9.00631309e-01 1.37096882e+00 -9.94244337e-01 -3.11927408e-01
-4.08633560e-01 -5.43941081e-01 -1.36260426e+00 -3.81322175e-01
-1.57273531e-01 1.45062283e-01 1.82807788e-01 1.01964545e+00
6.54757991e-02 2.86332428e-01 4.35136765e-01 -1.02602971e+00
-4.83833462e-01 1.28959107e+00 -1.79132745e-01 3.04710776e-01
1.11306596e+00 -4.02030110e-01 1.59707689e+00 -4.87214565e-01
1.26110971e+00 1.96839690e+00 2.05286473e-01 8.76467347e-01
-9.68035638e-01 -1.11554377e-01 2.51661807e-01 9.90962163e-02
-8.13415825e-01 -2.52417266e-01 6.08634770e-01 -3.14211488e-01
1.59193075e+00 5.48189998e-01 2.86429107e-01 9.89122629e-01
-1.63828135e-01 6.56653702e-01 7.97906339e-01 -8.17135870e-01
3.96778315e-01 1.29881680e-01 6.78318143e-01 2.38545328e-01
1.18390679e-01 -2.99883127e-01 -5.57502747e-01 -6.73972592e-02
4.66504246e-01 -7.75193930e-01 -8.64991620e-02 2.06832364e-01
-1.14215052e+00 1.12253273e+00 4.06761140e-01 7.78295934e-01
-4.10738975e-01 -1.67310461e-01 7.70287871e-01 8.36789757e-02
9.70916152e-02 2.98946828e-01 -4.09624279e-01 7.18094707e-02
-9.49466452e-02 7.01587439e-01 9.28243160e-01 8.66522372e-01
4.82444555e-01 -1.89827830e-01 -2.59741068e-01 5.54165721e-01
5.22935212e-01 1.81700304e-01 5.57334542e-01 -1.18423164e+00
8.99544597e-01 1.16170216e+00 1.94960549e-01 -9.51914847e-01
-3.79903585e-01 3.82676810e-01 -7.69020244e-02 -3.40818465e-01
2.64599532e-01 -4.79075536e-02 -3.93070310e-01 2.19320059e+00
4.91133273e-01 -2.89532810e-01 8.94225597e-01 8.64276052e-01
1.45880580e+00 9.76682842e-01 6.70045376e-01 -4.02734160e-01
2.06443691e+00 -2.77776033e-01 -1.16063392e+00 -5.40217221e-01
9.50292110e-01 -3.69483352e-01 1.15577292e+00 -3.26832235e-01
-1.43816090e+00 -1.40737310e-01 -1.01637399e+00 -4.65811938e-01
-3.12633455e-01 -4.22562212e-01 7.20208108e-01 3.06511760e-01
-6.91958904e-01 2.32293874e-01 -5.18685520e-01 -2.20227107e-01
-1.58702768e-02 -4.87625562e-02 -3.79389107e-01 2.26141229e-01
-1.76352549e+00 1.06700397e+00 1.14437127e+00 -3.27176064e-01
-2.79754132e-01 -4.96002644e-01 -1.34445465e+00 1.61501631e-01
4.82601345e-01 -4.04690057e-01 1.42174292e+00 -1.04943633e+00
-1.16545641e+00 1.44089448e+00 -4.57594872e-01 -7.05847442e-01
-2.46387661e-01 -4.05285694e-02 -4.75222230e-01 2.00463414e-01
3.84659380e-01 4.77382600e-01 3.71288173e-02 -1.08618689e+00
-6.65157914e-01 -4.83266532e-01 6.01323962e-01 5.28832018e-01
2.93098003e-01 1.02411914e+00 2.07967788e-01 -2.15959087e-01
4.46503490e-01 -3.03853214e-01 -8.53632763e-02 -6.46781385e-01
-3.53792965e-01 -7.83562124e-01 5.84154904e-01 -5.90199947e-01
1.38788128e+00 -2.27573442e+00 2.99929947e-01 -1.56564027e-01
-1.27559528e-01 1.60077348e-01 1.18377686e-01 5.29683590e-01
-3.87275338e-01 3.38291734e-01 -2.45069355e-01 1.02899462e-01
3.19145322e-01 6.03687882e-01 -5.83268821e-01 -1.60847589e-01
5.83205998e-01 7.61956871e-01 -7.83640742e-01 -5.98773718e-01
1.33388326e-01 3.71717066e-02 -2.76032358e-01 3.38574380e-01
-9.77404773e-01 1.00248568e-01 -6.59552753e-01 3.52899045e-01
3.40593189e-01 -4.21866290e-02 9.23942864e-01 -1.06100962e-01
-1.18531011e-01 1.34815097e+00 -1.43159366e+00 1.40532839e+00
-1.05128780e-01 2.15405926e-01 2.07808137e-01 -1.05302083e+00
9.61764812e-01 6.33970141e-01 -2.63837606e-01 -7.43030965e-01
1.61295503e-01 2.81562001e-01 6.82829395e-02 -4.23636705e-01
6.37840331e-01 -6.52352512e-01 -7.97683179e-01 2.03780949e-01
-9.53526497e-02 -8.91808718e-02 4.68220562e-01 3.78743708e-01
9.32908595e-01 1.49492517e-01 1.10569870e+00 -3.50030720e-01
9.12038445e-01 4.17910546e-01 9.82976317e-01 6.80457503e-02
1.38171583e-01 3.51903379e-01 1.10193241e+00 -2.81401932e-01
-5.99751413e-01 -1.19321990e+00 -1.65806413e-01 1.14399052e+00
1.81373596e-01 -7.33381331e-01 -8.01481664e-01 -4.61921662e-01
-4.33610171e-01 1.34365320e+00 -4.25558686e-01 3.70519102e-01
-1.36455584e+00 -4.76088434e-01 5.24526536e-01 7.75057256e-01
5.82331568e-02 -1.73117852e+00 -1.11800456e+00 5.61770320e-01
-5.90741694e-01 -1.35288966e+00 4.18235749e-01 2.12464660e-01
-4.98989552e-01 -1.44142449e+00 2.55596727e-01 -9.98984277e-01
1.16792910e-01 -2.73009360e-01 1.50460351e+00 1.97354734e-01
2.04399582e-02 -1.63373679e-01 -5.14235556e-01 -7.13614404e-01
-8.18198860e-01 -4.82973993e-01 -5.23365021e-01 -5.49356759e-01
6.09258890e-01 -1.58216804e-01 1.00697942e-01 -3.68796140e-01
-8.40919435e-01 5.54647632e-02 -3.61002892e-01 5.06636083e-01
5.65742791e-01 -2.85201997e-01 5.82522988e-01 -1.33810949e+00
8.12416136e-01 -6.55888557e-01 -4.95267451e-01 3.90654474e-01
2.74455130e-01 2.36577377e-01 2.82927990e-01 -1.75790880e-02
-1.65069616e+00 -4.22371238e-01 -2.49901101e-01 6.34587884e-01
-2.01524600e-01 8.07986975e-01 -6.52678788e-01 1.03506148e+00
7.58847058e-01 -4.34875637e-01 -8.03941712e-02 -1.78236112e-01
4.89877284e-01 4.29645717e-01 5.10265827e-01 -9.93938982e-01
1.15247205e-01 8.94366354e-02 2.79044230e-02 -6.86091185e-01
-1.13918531e+00 -3.44078094e-01 -3.66871655e-01 2.53916502e-01
1.24147475e+00 -1.09194732e+00 -5.19652545e-01 -1.15061976e-01
-1.72175443e+00 -1.21501898e-02 -7.44670630e-01 1.76230803e-01
-5.79018235e-01 3.30408871e-01 -7.95650482e-01 -9.71342802e-01
-3.77590001e-01 -7.79313624e-01 9.27274883e-01 2.65582025e-01
-8.94007504e-01 -1.11500704e+00 -1.11063302e-01 2.39855498e-01
1.21735886e-01 7.16189563e-01 1.61526597e+00 -1.16388345e+00
-1.91327214e-01 3.31213623e-01 -1.78437322e-01 8.48459378e-02
-2.40714345e-02 -2.40154371e-01 -6.81206763e-01 2.71188945e-01
1.86399013e-01 -4.42008227e-01 2.89182425e-01 2.04248756e-01
3.55232835e-01 -4.29937124e-01 -1.84997082e-01 -6.23190254e-02
1.39669180e+00 3.42125297e-01 8.37523520e-01 5.13888121e-01
7.49341175e-02 1.25964582e+00 9.13804829e-01 1.68163478e-01
7.14289010e-01 5.36404252e-01 4.81008053e-01 4.36463565e-01
-2.14968935e-01 -8.47911760e-02 4.07862246e-01 5.68547904e-01
2.75651664e-01 -3.12541306e-01 -1.08907747e+00 6.83062017e-01
-1.91364408e+00 -1.02025878e+00 -7.42781579e-01 1.65018392e+00
1.19639659e+00 2.64709920e-01 -1.62041813e-01 1.86321884e-01
8.56225431e-01 5.45310616e-01 1.62102535e-01 -7.85711944e-01
-6.14198327e-01 3.67404670e-01 -2.00783610e-01 7.41019845e-01
-7.86792278e-01 1.28698182e+00 5.82325220e+00 3.60760748e-01
-6.93874657e-01 1.83253199e-01 -4.75474335e-02 3.51133019e-01
-7.17243433e-01 3.84078354e-01 -8.57349515e-01 1.19862683e-01
1.27043450e+00 -3.25600803e-01 -7.12052956e-02 6.27620935e-01
-6.74324036e-02 -2.63133019e-01 -1.06684983e+00 4.34672743e-01
-1.38992250e-01 -1.61004865e+00 1.07227623e-01 -4.87510413e-01
2.92085648e-01 -2.35090405e-01 -5.42837203e-01 3.01916003e-01
3.11127424e-01 -1.03387225e+00 1.00548601e+00 8.43731612e-02
3.87559742e-01 -4.18976456e-01 9.98136818e-01 2.94665486e-01
-1.26963305e+00 -3.24644819e-02 -3.99006397e-01 -3.93240720e-01
6.94141388e-01 1.45323902e-01 -5.77765286e-01 7.60538161e-01
4.14457858e-01 2.85300046e-01 -4.63346019e-02 1.81829646e-01
-7.22367823e-01 5.76684713e-01 -2.52044737e-01 -1.73933536e-01
1.26156539e-01 -6.51641190e-02 7.57834733e-01 1.30303049e+00
6.22281246e-03 7.54899025e-01 1.84596643e-01 9.44395244e-01
1.38117909e-01 2.37129346e-01 -3.91383857e-01 7.60645941e-02
8.79087627e-01 7.53820598e-01 -6.33084536e-01 -4.78462577e-01
-3.03348511e-01 2.92691410e-01 3.14227074e-01 -1.29940957e-01
-5.10183752e-01 4.51786481e-02 6.43084943e-01 2.76750833e-01
6.62760511e-02 2.20357120e-01 -1.70834228e-01 -7.05927253e-01
2.55995125e-01 -7.18151391e-01 7.69852221e-01 -8.02199304e-01
-1.21310186e+00 6.55208349e-01 5.36260009e-01 -5.56937218e-01
-7.46069014e-01 -7.49341130e-01 -8.45934570e-01 9.76100683e-01
-1.48774827e+00 -1.01174140e+00 6.57828450e-02 5.30635238e-01
4.72037435e-01 4.00620922e-02 1.47135258e+00 -2.03884631e-01
-2.54404426e-01 -1.24782443e-01 -8.24526906e-01 1.75097525e-01
2.84799337e-01 -1.33870089e+00 3.65805894e-01 7.17926085e-01
-3.51748973e-01 7.57044077e-01 9.02058721e-01 -7.26154268e-01
-9.61071730e-01 -7.37406969e-01 1.73161113e+00 -3.18352669e-01
7.44668126e-01 -7.85540342e-02 -1.44285285e+00 1.04019821e+00
4.11628604e-01 -1.53808758e-01 7.29434013e-01 2.95718759e-01
-4.53819156e-01 3.68749648e-01 -1.31954145e+00 3.21473926e-01
9.42907214e-01 -4.81718302e-01 -1.94589376e+00 2.30120718e-01
1.14027834e+00 -9.61065710e-01 -6.39628410e-01 2.62067795e-01
-1.70111269e-01 -7.94142187e-01 8.79227638e-01 -9.13432598e-01
4.24854696e-01 -3.66829425e-01 -6.88515544e-01 -7.67090023e-01
1.29255801e-01 -3.17457199e-01 -3.37325707e-02 1.53555405e+00
5.70385695e-01 -3.35386395e-01 3.03017676e-01 8.03366959e-01
-3.34274799e-01 -5.97860217e-02 -1.20406330e+00 -3.62730265e-01
1.75160512e-01 -4.88165528e-01 8.72007668e-01 7.81099558e-01
6.53540432e-01 9.91086006e-01 3.71132940e-01 -1.48153141e-01
3.50912392e-01 2.54087150e-01 -4.49137110e-03 -1.55578768e+00
-7.78358728e-02 -1.41896889e-01 -1.78026691e-01 -3.58491778e-01
7.35325515e-01 -1.05061305e+00 2.29178229e-03 -1.86966848e+00
-1.94044426e-01 -4.57530558e-01 1.99345395e-01 5.71613073e-01
-1.55758068e-01 -4.85724837e-01 4.34528679e-01 -4.20717262e-02
-5.38318694e-01 2.96549648e-01 7.99705148e-01 1.19774461e-01
-3.66603673e-01 -4.17247027e-01 -8.75349581e-01 1.05289626e+00
8.44517946e-01 -4.59230512e-01 -4.69683677e-01 -3.26607883e-01
5.52446306e-01 6.05604827e-01 1.80643007e-01 -3.73853683e-01
-9.25910380e-03 -5.44923127e-01 -3.78734410e-01 -4.47338670e-01
5.14881611e-01 -4.62740511e-01 1.55930847e-01 2.54804492e-01
-5.46555042e-01 2.12144509e-01 2.94860989e-01 7.78318429e-03
-5.63301384e-01 -6.51970446e-01 5.39666235e-01 -5.04298031e-01
-1.11241257e+00 -7.75831223e-01 -4.46518332e-01 7.08257198e-01
1.16350865e+00 -2.10428566e-01 -6.41999781e-01 1.31758586e-01
-6.79093301e-01 7.65343085e-02 1.23673655e-01 4.28646296e-01
7.17890561e-01 -1.11815703e+00 -7.14711368e-01 -6.05182759e-02
1.22485040e-02 4.90602821e-01 1.91653669e-01 1.22919530e-01
-4.29484129e-01 3.75254035e-01 -2.99588263e-01 -9.42394584e-02
-1.40739191e+00 3.00070167e-01 9.54772383e-02 -5.81888437e-01
-7.35562384e-01 7.98698723e-01 9.31049138e-02 -5.02746761e-01
-3.53955589e-02 -4.65732068e-01 -8.75438690e-01 2.77300298e-01
8.29748452e-01 -9.82139930e-02 -3.26736346e-02 -1.06973982e+00
-8.23755085e-01 8.41984749e-02 2.62927800e-01 -2.54136771e-01
1.23001420e+00 -1.10181898e-01 -6.47380114e-01 5.49925387e-01
7.34023750e-01 8.73847902e-02 -3.89062464e-01 -2.18470663e-01
7.63145089e-01 2.79840995e-02 -3.74583215e-01 -7.20453084e-01
-3.11714441e-01 4.78398740e-01 -2.23563731e-01 4.48657840e-01
6.44004643e-01 6.13332450e-01 6.31222785e-01 8.69790316e-02
3.57614219e-01 -1.04326558e+00 -2.09855065e-01 1.08038807e+00
1.09841156e+00 -6.88203752e-01 -2.73594648e-01 -1.06418943e+00
-9.52159703e-01 1.07723570e+00 6.68826818e-01 -3.42830755e-02
-1.32625760e-03 4.28614587e-01 7.15989694e-02 -5.61126173e-01
-9.64160860e-01 -1.66559532e-01 -3.15037556e-02 8.03286254e-01
9.13685083e-01 4.71617788e-01 -9.14070964e-01 1.23581553e+00
-6.08147144e-01 -4.64287043e-01 6.73345268e-01 1.04728389e+00
-5.56193829e-01 -1.24881577e+00 -3.80270541e-01 -1.14042565e-01
-7.26869285e-01 -1.58836588e-01 -3.92485291e-01 1.11906219e+00
2.02563610e-02 1.16412032e+00 3.66353303e-01 2.39326775e-01
6.43259704e-01 2.32876584e-01 3.82120103e-01 -1.21455598e+00
-1.03989697e+00 -4.17593360e-01 9.54665661e-01 -6.68623626e-01
-8.39924455e-01 -7.97635555e-01 -2.45954180e+00 -7.42953131e-03
-1.12192482e-01 3.32674235e-01 5.10016501e-01 1.30010664e+00
-2.80611906e-02 6.23415112e-01 -6.21729083e-02 -1.50431260e-01
-6.29908204e-01 -7.59222150e-01 -3.79992753e-01 5.86755693e-01
-7.63212517e-02 -4.56863493e-01 -2.18077540e-01 -1.25582710e-01] | [10.320425987243652, 9.298236846923828] |
b5b92f11-fa06-41b1-8c46-adb3c776c744 | topic-guided-variational-autoencoders-for | 1903.07137 | null | http://arxiv.org/abs/1903.07137v1 | http://arxiv.org/pdf/1903.07137v1.pdf | Topic-Guided Variational Autoencoders for Text Generation | We propose a topic-guided variational autoencoder (TGVAE) model for text
generation. Distinct from existing variational autoencoder (VAE) based
approaches, which assume a simple Gaussian prior for the latent code, our model
specifies the prior as a Gaussian mixture model (GMM) parametrized by a neural
topic module. Each mixture component corresponds to a latent topic, which
provides guidance to generate sentences under the topic. The neural topic
module and the VAE-based neural sequence module in our model are learned
jointly. In particular, a sequence of invertible Householder transformations is
applied to endow the approximate posterior of the latent code with high
flexibility during model inference. Experimental results show that our TGVAE
outperforms alternative approaches on both unconditional and conditional text
generation, which can generate semantically-meaningful sentences with various
topics. | ['Lawrence Carin', 'Wenlin Wang', 'Changyou Chen', 'Zhe Gan', 'Hongteng Xu', 'Ruiyi Zhang', 'Guoyin Wang', 'Dinghan Shen'] | 2019-03-17 | null | null | null | null | ['conditional-text-generation'] | ['natural-language-processing'] | [-5.95331490e-02 6.53500974e-01 -7.75469020e-02 -2.66959667e-01
-9.28415656e-01 -5.12549400e-01 1.17503417e+00 -3.68105233e-01
1.64533257e-01 6.77692831e-01 6.95503771e-01 -2.29831621e-01
4.72861081e-01 -1.07979846e+00 -8.77727389e-01 -7.81889677e-01
4.57740456e-01 6.92393959e-01 -3.49335849e-01 -1.25308976e-01
-1.44002572e-01 -3.11545938e-01 -1.21064150e+00 1.63349122e-01
1.11793280e+00 4.95184690e-01 5.25719285e-01 8.05250108e-01
-3.84203851e-01 5.59318423e-01 -8.00087035e-01 -6.03910208e-01
-3.05944979e-01 -7.73012340e-01 -4.70675468e-01 3.91294479e-01
-1.44652948e-01 -5.25195420e-01 -2.75347739e-01 1.04944587e+00
2.63589859e-01 3.74769241e-01 1.28778994e+00 -1.30377102e+00
-1.09124637e+00 1.03615248e+00 -6.72594532e-02 -5.29577971e-01
1.52701378e-01 1.95873231e-02 1.27851450e+00 -9.80592132e-01
8.19848001e-01 1.54128504e+00 2.77991354e-01 8.99415553e-01
-1.52054036e+00 -3.63063842e-01 1.36099890e-01 -3.40239763e-01
-1.26618803e+00 -3.69323760e-01 9.81122911e-01 -5.65780640e-01
7.43264198e-01 -1.08457971e-02 7.34843314e-01 1.71474361e+00
2.98528880e-01 1.09845424e+00 4.09742773e-01 -3.25157136e-01
7.13136613e-01 2.99699515e-01 -3.50797474e-01 5.11525810e-01
-1.51425973e-01 -7.59713575e-02 -4.07754481e-01 -4.21498030e-01
9.50049937e-01 -1.57843933e-01 -6.06887080e-02 -4.28457260e-01
-1.11816239e+00 1.52629364e+00 -5.43765426e-02 9.93924066e-02
-6.41879559e-01 4.79140073e-01 2.09958643e-01 -2.46864602e-01
6.33060694e-01 1.29458711e-01 -2.87867188e-01 -8.77537206e-02
-1.10406029e+00 6.71336889e-01 1.00372064e+00 1.18336487e+00
6.36232197e-01 6.30854487e-01 -8.43093991e-01 8.43527555e-01
8.23638737e-01 4.76794511e-01 5.08504331e-01 -1.18661654e+00
2.85697788e-01 1.51851684e-01 1.49423197e-01 -6.31468236e-01
5.39658487e-01 -2.17650026e-01 -1.06148386e+00 -1.80060819e-01
-7.49546215e-02 -5.05743325e-01 -9.75952268e-01 1.99471617e+00
3.16880465e-01 3.85988504e-01 3.52886051e-01 4.51057196e-01
7.73229778e-01 1.30165112e+00 1.82662636e-01 -2.29722425e-01
1.35248983e+00 -8.66006136e-01 -1.11107147e+00 2.66371761e-02
1.47135466e-01 -5.11045575e-01 9.29639220e-01 1.68869291e-02
-1.22510481e+00 -4.49128777e-01 -7.06623375e-01 -1.88081548e-01
-1.06915787e-01 2.04440817e-01 4.59259510e-01 6.57389164e-01
-1.35645521e+00 1.61670968e-01 -9.72227693e-01 -3.65395285e-02
6.89814985e-02 -7.60409608e-02 1.32718444e-01 3.03816587e-01
-1.29930270e+00 4.70395327e-01 5.16123831e-01 -1.95723578e-01
-1.55360687e+00 -6.85684919e-01 -1.29508853e+00 3.45799714e-01
-5.62582426e-02 -1.22890306e+00 1.49035347e+00 -5.93621075e-01
-2.19299531e+00 3.80135238e-01 -6.79099917e-01 -8.06866229e-01
3.57126445e-01 2.23053303e-02 8.45694542e-03 1.16096780e-01
1.36137828e-01 1.29244232e+00 1.30789864e+00 -1.47306943e+00
-2.76701242e-01 4.53740209e-01 -1.02929942e-01 1.86539114e-01
-3.81254792e-01 -1.36616305e-01 -6.62480712e-01 -1.14050198e+00
-2.76449531e-01 -7.80299485e-01 -2.64522493e-01 -3.38199914e-01
-7.52075493e-01 -6.72452986e-01 9.67238069e-01 -8.25061440e-01
1.20665586e+00 -1.80966771e+00 5.91102660e-01 -1.37042522e-01
7.24755973e-02 -4.45879877e-01 -2.97693759e-02 5.35975754e-01
2.36388907e-01 3.39078784e-01 -5.91607571e-01 -9.78087425e-01
5.23612380e-01 3.64848793e-01 -9.79000509e-01 -7.29873255e-02
2.58659214e-01 1.06359017e+00 -7.06807911e-01 -6.38548195e-01
1.68871656e-01 8.28871191e-01 -1.11153150e+00 5.48423052e-01
-1.04981637e+00 4.01689023e-01 -4.83092427e-01 1.01940848e-01
4.60035741e-01 -3.29777360e-01 2.30847418e-01 2.52271205e-01
1.32762715e-01 3.34310323e-01 -8.18741500e-01 1.77974319e+00
-3.66748750e-01 6.35950327e-01 -2.11241934e-03 -8.25565636e-01
9.38422263e-01 8.27623427e-01 1.06545500e-01 3.99376482e-01
1.08651526e-01 -4.38157767e-01 -5.31129062e-01 -1.80103183e-01
9.41051066e-01 -2.19325110e-01 -2.73800701e-01 8.07200313e-01
4.86971706e-01 -5.20859301e-01 4.24947925e-02 6.36659801e-01
3.95052254e-01 4.52598572e-01 8.67459029e-02 -3.02956343e-01
1.84956491e-01 -2.77871817e-01 4.96155262e-01 8.16436887e-01
1.94354877e-01 7.31711388e-01 6.64185107e-01 6.92339465e-02
-1.41896927e+00 -1.40084386e+00 2.87418198e-02 9.48300421e-01
-3.28649849e-01 -5.79325974e-01 -1.22163534e+00 -2.85733789e-01
-3.50164086e-01 1.49345171e+00 -6.81036472e-01 -1.00702355e-02
-2.92053252e-01 -5.94631612e-01 4.35670584e-01 3.57295454e-01
3.14400971e-01 -1.12053156e+00 -1.84231013e-01 3.03059518e-01
-6.54210806e-01 -9.41838920e-01 -6.97287619e-01 -4.09162283e-01
-9.58874285e-01 -3.24220419e-01 -1.04043543e+00 -7.46854246e-01
6.20572209e-01 -3.17606688e-01 1.09374511e+00 -4.66373801e-01
2.61424541e-01 5.66081107e-01 -2.23049328e-01 -3.47521126e-01
-1.03563690e+00 -6.77634180e-02 -6.32804856e-02 9.54834893e-02
1.00759797e-01 -5.90617537e-01 -3.13389897e-01 -2.96322495e-01
-1.09671640e+00 3.89052421e-01 2.48074397e-01 9.29630101e-01
5.34482837e-01 -3.41093470e-03 3.38267475e-01 -6.27622008e-01
9.94821489e-01 -8.14650357e-01 -6.25798047e-01 1.62805259e-01
-3.04154158e-01 3.32576662e-01 3.83428752e-01 -4.76082653e-01
-1.54485226e+00 -2.10770637e-01 -1.82636395e-01 -5.88257372e-01
-3.48697126e-01 5.56242704e-01 -2.68076837e-01 1.01743329e+00
2.79199392e-01 7.35493600e-01 -8.17047618e-03 -3.27472687e-01
8.60929728e-01 6.04633152e-01 4.83083904e-01 -8.39032888e-01
7.53009379e-01 4.42619652e-01 -4.71916437e-01 -8.52874935e-01
-6.79365456e-01 2.30747581e-01 -4.40012544e-01 -1.44096285e-01
1.55970502e+00 -1.30422807e+00 -3.93596053e-01 3.38983834e-01
-1.67356455e+00 -4.79559302e-01 -3.31386268e-01 2.93005854e-01
-8.48053515e-01 3.64096880e-01 -8.19799960e-01 -1.12415576e+00
-4.72696632e-01 -1.18680632e+00 1.45704138e+00 1.34012341e-01
-3.79735827e-01 -1.39778829e+00 2.31187254e-01 1.51209727e-01
2.53735214e-01 1.46363333e-01 1.06184244e+00 -4.45991129e-01
-8.15141261e-01 1.20450765e-01 2.65516788e-01 2.93383300e-01
2.35857107e-02 2.41283342e-01 -9.66202736e-01 -1.34287074e-01
1.74916252e-01 -1.42824009e-01 8.71320128e-01 8.56352270e-01
1.03287947e+00 -7.50801504e-01 -3.05235237e-01 5.39473534e-01
1.08649099e+00 1.45018220e-01 5.98052263e-01 -2.44213670e-01
6.88055873e-01 4.24292207e-01 2.65428834e-02 6.70506835e-01
6.57917142e-01 5.49332559e-01 2.29784861e-01 3.09565634e-01
1.42748386e-01 -6.82276428e-01 7.98879623e-01 1.06339645e+00
2.29837716e-01 -7.07949758e-01 -6.24594569e-01 7.33137429e-01
-1.82822478e+00 -1.01786542e+00 2.29252920e-01 1.62286842e+00
1.10425186e+00 -2.66847938e-01 -6.44242531e-03 -3.65593106e-01
9.10446286e-01 4.30852503e-01 -4.09002662e-01 -3.40996802e-01
-2.80715376e-02 4.20064330e-02 -3.81379068e-01 7.17202187e-01
-1.04617572e+00 1.26907372e+00 7.08343077e+00 9.57987845e-01
-5.54953754e-01 2.69644141e-01 4.72404897e-01 1.20026372e-01
-1.02702749e+00 7.52544254e-02 -8.47117603e-01 6.54265583e-01
1.10290635e+00 -4.15898055e-01 3.57370108e-01 1.03487408e+00
3.99227411e-01 3.18926722e-01 -9.59772766e-01 5.26418567e-01
8.60255882e-02 -1.64449358e+00 4.82661486e-01 3.24287385e-01
1.10452533e+00 -3.35904866e-01 3.13266635e-01 4.74732101e-01
1.21619332e+00 -8.72863293e-01 9.74059880e-01 6.86275482e-01
6.36996210e-01 -8.00988674e-01 2.07613349e-01 5.37566483e-01
-9.41315055e-01 2.36685634e-01 -5.46897471e-01 2.58165479e-01
5.08278012e-01 6.08252823e-01 -8.03159356e-01 1.70552373e-01
3.39255661e-01 6.12639308e-01 1.44437239e-01 3.55003178e-01
-6.43607795e-01 1.06045568e+00 -2.79486571e-02 -1.77915052e-01
9.76897478e-02 -6.35365725e-01 9.46983099e-01 1.11353624e+00
6.38179779e-01 -5.06102256e-02 7.46323690e-02 1.80706489e+00
-2.75463045e-01 -1.36819318e-01 -6.68429434e-01 -2.41150111e-01
6.19108558e-01 1.06121564e+00 -4.05316681e-01 -7.40915716e-01
-2.32688040e-02 1.37197638e+00 8.57169852e-02 8.76581192e-01
-8.76004100e-01 -1.45886257e-01 7.76916027e-01 -4.57578331e-01
7.75084794e-01 -3.29353869e-01 -2.02072188e-01 -1.45673335e+00
-3.32265377e-01 -6.93852127e-01 1.84308276e-01 -1.01355922e+00
-1.05334187e+00 4.57450151e-01 2.41809770e-01 -1.00345707e+00
-1.08682597e+00 -1.17109917e-01 -1.10670817e+00 1.10541165e+00
-8.87953222e-01 -1.28771186e+00 3.15127999e-01 4.31849480e-01
8.66363347e-01 -4.97613400e-01 1.13023722e+00 -5.85712850e-01
-6.20982409e-01 4.32499975e-01 3.38673919e-01 1.45937979e-01
2.31214747e-01 -1.45284653e+00 6.94758475e-01 9.80157971e-01
2.16145352e-01 9.79543984e-01 1.07025635e+00 -9.06052113e-01
-1.09780431e+00 -1.38360620e+00 1.06797481e+00 -5.81812382e-01
5.08560717e-01 -7.30587423e-01 -6.14125729e-01 1.02817202e+00
8.86889398e-01 -8.34739804e-01 7.96872735e-01 -6.30612001e-02
-2.37978533e-01 6.69608176e-01 -8.29404116e-01 8.57552648e-01
4.45485175e-01 -8.09991360e-01 -6.86329484e-01 3.76112312e-01
1.50429666e+00 -3.34160507e-01 -9.21299219e-01 -1.64960220e-01
2.03213364e-01 -5.91151118e-01 9.06324625e-01 -5.73008239e-01
1.07012308e+00 -1.52895764e-01 -2.04551697e-01 -1.48922396e+00
-2.02862963e-01 -7.94973552e-01 -7.19180703e-01 1.54816258e+00
3.50174427e-01 -2.10379109e-01 7.70963609e-01 5.95329225e-01
-1.80972338e-01 -4.07195121e-01 -7.61926591e-01 -4.11259055e-01
4.60690498e-01 -7.00092435e-01 8.43693078e-01 7.76549160e-01
-1.72464833e-01 3.45124543e-01 -6.93297088e-01 1.85562208e-01
7.87883341e-01 1.03076398e-01 7.32763350e-01 -1.01848614e+00
-6.43548012e-01 -4.35818911e-01 2.97501504e-01 -1.36336589e+00
6.20644808e-01 -9.57544446e-01 4.43297923e-01 -1.76757812e+00
3.51668566e-01 3.17677230e-01 2.76418209e-01 1.36912271e-01
-3.89742613e-01 -3.12343597e-01 6.22462183e-02 1.61404446e-01
-3.18140984e-01 1.37394965e+00 1.06192696e+00 -1.71299979e-01
-2.67555654e-01 -1.79354614e-03 -8.04175138e-01 5.51370502e-01
6.26837134e-01 -4.41033632e-01 -5.30437469e-01 -4.21203434e-01
1.04678303e-01 3.50726813e-01 3.97590637e-01 -5.24351060e-01
1.96889475e-01 -1.80252925e-01 2.04792395e-01 -7.96290457e-01
4.53890085e-01 -7.89823383e-02 9.83606204e-02 1.37751311e-01
-6.48788095e-01 -4.34220761e-01 -4.69667390e-02 7.98116863e-01
-2.61097729e-01 -1.80605039e-01 4.06169683e-01 -1.37065379e-02
-1.95108548e-01 4.77250844e-01 -9.92955863e-01 -1.13178894e-01
6.27340496e-01 1.65355325e-01 5.67741506e-02 -1.09762716e+00
-7.81872392e-01 2.02939466e-01 3.19635898e-01 4.66908544e-01
7.74829805e-01 -1.58468950e+00 -8.57585490e-01 -2.08366401e-02
-1.91502973e-01 2.62763917e-01 3.84353042e-01 1.49466932e-01
3.19160111e-02 5.64956665e-01 3.03217143e-01 -5.42189717e-01
-5.84072769e-01 4.97291893e-01 1.70859724e-01 -3.17725152e-01
-5.09210527e-01 9.61030364e-01 5.93244851e-01 -6.42209947e-01
1.59166113e-01 -1.53386071e-01 -2.59071469e-01 -1.81820970e-02
3.88558149e-01 1.12160984e-02 -7.67954409e-01 -5.50338447e-01
3.05942118e-01 5.28516248e-02 2.02574536e-01 -9.01091695e-01
1.14929497e+00 -2.45999277e-01 -1.98662296e-01 5.89746416e-01
9.76964772e-01 -1.71181813e-01 -1.56018043e+00 -7.87742138e-02
-3.93129021e-01 1.94672197e-01 1.38649032e-01 -3.26204747e-01
-6.75347805e-01 1.08238661e+00 -3.32585759e-02 1.78908780e-01
6.34373009e-01 1.41870216e-01 8.36400807e-01 2.22806245e-01
-2.47920863e-02 -1.06945467e+00 2.04835668e-01 7.29423642e-01
1.10902202e+00 -7.75613248e-01 -4.78335291e-01 -1.32700756e-01
-1.00631380e+00 8.17881405e-01 3.11881840e-01 -7.10143968e-02
7.38883257e-01 1.39468655e-01 -1.37654409e-01 3.82927582e-02
-1.16370749e+00 1.64291903e-01 3.71921748e-01 7.99752474e-01
4.55374509e-01 2.12453693e-01 2.49073692e-02 9.41260695e-01
-7.25875080e-01 -2.78416604e-01 5.51469386e-01 3.98827076e-01
-5.00814736e-01 -1.18541896e+00 -3.53527993e-01 1.23762935e-01
-2.58526593e-01 -4.50078756e-01 -1.57365233e-01 3.39437932e-01
-1.39291689e-01 9.83197153e-01 5.35652697e-01 -2.41022795e-01
-4.27992642e-01 5.61726093e-01 -1.69049297e-02 -7.50568807e-01
-1.45306915e-01 6.61026657e-01 -2.22149387e-01 -2.35116318e-01
-1.00810103e-01 -8.83183837e-01 -1.13463676e+00 -1.61416441e-01
-1.21859811e-01 4.33068395e-01 6.97509766e-01 9.41967666e-01
3.20085526e-01 7.10742474e-01 4.54518229e-01 -8.38444412e-01
-5.27521551e-01 -1.26378107e+00 -5.57278335e-01 2.27902278e-01
1.72917366e-01 -2.51101583e-01 -2.18887702e-01 7.18016207e-01] | [11.850859642028809, 9.040060043334961] |
b23a2975-7b37-4cce-8475-757cba37b9a4 | kalman-based-spectro-temporal-ecg-analysis | 1812.05555 | null | http://arxiv.org/abs/1812.05555v1 | http://arxiv.org/pdf/1812.05555v1.pdf | Kalman-based Spectro-Temporal ECG Analysis using Deep Convolutional Networks for Atrial Fibrillation Detection | In this article, we propose a novel ECG classification framework for atrial
fibrillation (AF) detection using spectro-temporal representation (i.e., time
varying spectrum) and deep convolutional networks. In the first step we use a
Bayesian spectro-temporal representation based on the estimation of
time-varying coefficients of Fourier series using Kalman filter and smoother.
Next, we derive an alternative model based on a stochastic oscillator
differential equation to accelerate the estimation of the spectro-temporal
representation in lengthy signals. Finally, after comparative evaluations of
different convolutional architectures, we propose an efficient deep
convolutional neural network to classify the 2D spectro-temporal ECG data.
The ECG spectro-temporal data are classified into four different classes: AF,
non-AF normal rhythm (Normal), non-AF abnormal rhythm (Other), and noisy
segments (Noisy). The performance of the proposed methods is evaluated and
scored with the PhysioNet/Computing in Cardiology (CinC) 2017 dataset. The
experimental results show that the proposed method achieves the overall F1
score of 80.2%, which is in line with the state-of-the-art algorithms. | ['Simo Särkkä', 'Zheng Zhao', 'Ali Bahrami Rad'] | 2018-12-12 | null | null | null | null | ['ecg-classification', 'atrial-fibrillation-detection'] | ['medical', 'medical'] | [ 2.47400135e-01 -4.69905674e-01 3.22989881e-01 -2.35469714e-01
-7.76120901e-01 -4.84459102e-01 -1.62814632e-01 2.49589115e-01
-4.86391008e-01 8.86612058e-01 -3.63368839e-02 -3.91733348e-01
-5.12032330e-01 -3.24724376e-01 -2.03522563e-01 -7.12112010e-01
-7.80727565e-01 -6.91959932e-02 -4.05159473e-01 2.42130086e-01
-4.95757386e-02 5.99508047e-01 -1.12065053e+00 4.26484823e-01
9.36569393e-01 1.32288563e+00 -3.74949127e-01 1.17673695e+00
4.77858037e-01 3.36195230e-01 -9.91196811e-01 1.29469723e-01
2.31650211e-02 -9.18659270e-01 -4.35423136e-01 -3.64411741e-01
5.76535426e-02 -1.12982027e-01 -1.78158268e-01 7.01803684e-01
1.09173071e+00 -1.81165963e-01 6.52116060e-01 -5.00388503e-01
-8.54167119e-02 5.02919734e-01 -2.58377343e-01 8.51267874e-01
2.26583079e-01 2.60684453e-02 4.52217728e-01 -8.73023570e-01
3.68048817e-01 7.35784411e-01 1.41211689e+00 -2.21293103e-02
-1.14260495e+00 -6.05485618e-01 -3.06890309e-01 2.63517648e-01
-1.73740458e+00 -1.49004042e-01 8.96402478e-01 -4.44440544e-01
6.25162601e-01 2.42644832e-01 1.09003651e+00 1.04481423e+00
6.30576015e-01 4.01129067e-01 1.03440058e+00 -4.17582929e-01
1.67750329e-01 -4.85127360e-01 2.87615061e-01 6.48085594e-01
1.61591411e-01 3.48829240e-01 -4.22222227e-01 -6.66686654e-01
5.78579485e-01 7.73690268e-02 -7.34519303e-01 2.07978159e-01
-1.41688645e+00 5.57093143e-01 1.84199885e-01 5.47897577e-01
-8.49911928e-01 2.67718285e-01 7.14615524e-01 3.22872102e-01
4.43029583e-01 3.80671352e-01 -6.82791591e-01 -4.31712359e-01
-1.27584851e+00 3.44597876e-01 7.85130680e-01 2.76719064e-01
2.68662483e-01 4.26814914e-01 -6.21531487e-01 6.06577277e-01
3.83722007e-01 5.17474890e-01 6.95735455e-01 -6.87587142e-01
5.77453785e-02 2.06313968e-01 1.32089451e-01 -8.37473392e-01
-8.00788462e-01 -1.07943964e+00 -1.52391589e+00 -4.28804941e-02
4.04553741e-01 -5.82069159e-01 -8.66380632e-01 1.45997965e+00
2.20386181e-02 6.55786037e-01 2.20824823e-01 7.58747816e-01
8.94662797e-01 4.43832666e-01 -1.38619810e-01 -8.06418419e-01
1.49755406e+00 -3.15923601e-01 -1.16791987e+00 3.61001670e-01
5.85126638e-01 -6.11427486e-01 3.06200922e-01 9.10732329e-01
-9.17196512e-01 -8.94471824e-01 -1.21884882e+00 4.06573057e-01
4.42100503e-03 7.84880340e-01 2.99186885e-01 9.58549917e-01
-1.00935292e+00 1.11951029e+00 -9.26718175e-01 9.65396762e-02
5.66074848e-01 2.29228377e-01 -7.46189356e-02 4.58978593e-01
-1.55832660e+00 3.58179331e-01 3.13538939e-01 5.53448617e-01
-8.21481586e-01 -8.06237936e-01 -6.48396790e-01 3.25528495e-02
-3.96613181e-01 -7.71331668e-01 9.61369753e-01 -8.84099185e-01
-1.35641181e+00 4.58881408e-01 -1.60516068e-01 -8.16787720e-01
5.86201608e-01 -4.37150508e-01 -7.08697796e-01 3.08399022e-01
-1.30160555e-01 -2.49228671e-01 1.02129412e+00 -5.60998142e-01
-5.40297329e-01 -2.29433700e-01 -2.77461022e-01 -2.11504489e-01
1.16738230e-02 -1.07261337e-01 1.27635553e-01 -1.11277544e+00
2.58047044e-01 -9.93425846e-01 -1.73114076e-01 -1.76162481e-01
-3.32282484e-01 -9.20581594e-02 4.21405137e-01 -8.30498219e-01
1.71709776e+00 -2.37918806e+00 -4.49272357e-02 3.25632304e-01
4.55697447e-01 5.15183210e-01 3.76493484e-01 3.06039780e-01
-4.30857271e-01 1.15655571e-01 -5.75614691e-01 -8.42746422e-02
-4.87127751e-01 -9.70676169e-02 -1.89270705e-01 6.79699898e-01
2.33606875e-01 6.40388906e-01 -7.39731610e-01 -3.05853188e-01
8.06884095e-02 8.87213945e-01 -1.44529313e-01 -4.35569417e-03
4.37622726e-01 7.32014298e-01 -3.33530158e-01 3.34705532e-01
7.90320575e-01 -1.30279392e-01 3.47726464e-01 -6.00131989e-01
-8.19526315e-02 2.84820765e-01 -1.33896136e+00 1.79874527e+00
-3.52555841e-01 6.13557518e-01 -4.37530160e-01 -1.01880777e+00
1.01769757e+00 8.53035510e-01 7.39885807e-01 -4.63262349e-02
1.89937025e-01 4.35337186e-01 5.17424643e-01 -4.82656926e-01
-2.51554132e-01 6.04762742e-03 3.77814770e-01 6.28095269e-01
1.52620226e-01 2.84605205e-01 9.75180641e-02 -3.21345747e-01
1.22698057e+00 2.34950721e-01 4.60642189e-01 -6.78972960e-01
7.25189269e-01 -6.83475316e-01 7.85189390e-01 9.87695038e-01
-4.28243190e-01 8.91301453e-01 5.03327847e-01 -1.17687654e+00
-4.08886343e-01 -1.03009593e+00 -4.81480926e-01 1.78919941e-01
-3.62024665e-01 -7.69104421e-01 -7.03813970e-01 -6.05878115e-01
-2.90212154e-01 4.03531753e-02 -5.92120767e-01 -2.19814971e-01
-9.20418739e-01 -1.15893912e+00 1.14986432e+00 6.55951142e-01
5.04607797e-01 -1.07290959e+00 -1.16993511e+00 5.00531018e-01
-4.42755252e-01 -7.57728457e-01 -1.30923778e-01 3.82511079e-01
-1.04529464e+00 -1.04945886e+00 -9.53502297e-01 -5.07959008e-01
2.04679389e-02 -4.93698329e-01 1.13587260e+00 1.37276705e-02
-6.90129042e-01 6.24557473e-02 -3.34405661e-01 -7.93226063e-01
-3.17047060e-01 -7.17388466e-02 1.27632320e-01 3.05895388e-01
2.22196519e-01 -7.58149505e-01 -1.23989224e+00 -8.98484364e-02
-5.91014266e-01 -3.47314060e-01 5.69836795e-01 9.12382007e-01
6.45925045e-01 -9.96066779e-02 8.95957291e-01 -7.06528246e-01
9.29020822e-01 -2.04025373e-01 -1.96543783e-01 -3.07481270e-02
-8.04143488e-01 -2.38395363e-01 5.92556059e-01 -5.94203889e-01
-4.99059051e-01 2.64163226e-01 -3.46437812e-01 -4.21195656e-01
-1.05000265e-01 7.45916784e-01 4.66916680e-01 2.52313644e-01
1.00322056e+00 3.74573290e-01 -1.23574927e-01 -1.96270451e-01
-2.47930348e-01 7.43280530e-01 4.88147259e-01 -4.06871915e-01
3.46721083e-01 5.39178073e-01 8.54634643e-02 -7.67997086e-01
-8.18420410e-01 -3.60677749e-01 -5.75267434e-01 -1.42447889e-01
9.80033338e-01 -9.07872856e-01 -6.36690199e-01 6.12326920e-01
-1.37359917e+00 1.04275249e-01 -4.22634006e-01 9.46291208e-01
-5.16018331e-01 6.65073812e-01 -6.66751266e-01 -1.09650552e+00
-1.13439143e+00 -9.85749543e-01 1.00732684e+00 -1.37871489e-01
-4.78049695e-01 -7.60250330e-01 2.65664935e-01 -3.70336711e-01
4.25160289e-01 8.94882321e-01 5.13510048e-01 -5.38673639e-01
5.06727286e-02 -2.74418056e-01 2.51310408e-01 7.43098319e-01
2.40180552e-01 -1.39856653e-03 -1.11271691e+00 -3.75811815e-01
3.88493419e-01 3.18652481e-01 8.37875783e-01 9.13607121e-01
1.33542740e+00 1.28132656e-01 -2.74504215e-01 9.12443578e-01
1.10843909e+00 4.27002132e-01 7.67566025e-01 -1.07211046e-01
2.28908166e-01 -2.43234485e-01 4.41731840e-01 8.22828114e-01
-1.25942513e-01 3.24305028e-01 -8.27158336e-03 -2.28640005e-01
-6.12639775e-03 5.59417367e-01 1.51103586e-01 7.44055748e-01
-6.68981910e-01 -3.37302350e-02 -1.08916163e+00 5.14809787e-01
-1.86460876e+00 -8.77276361e-01 -5.67095041e-01 2.18051171e+00
7.51859784e-01 1.40408412e-01 7.19436333e-02 7.96314955e-01
6.04311466e-01 -1.49603799e-01 -3.93921018e-01 -3.73075575e-01
-1.36445925e-01 1.01374257e+00 2.52057984e-02 6.37935847e-02
-1.45053124e+00 2.20262129e-02 6.16093779e+00 5.05024612e-01
-1.33773339e+00 5.63568063e-02 6.96047843e-01 2.29378819e-01
4.26937908e-01 -2.88591683e-01 -2.21768022e-01 5.84666610e-01
1.39039171e+00 -1.31019473e-01 6.79256022e-02 3.47706586e-01
3.47522169e-01 2.68757313e-01 -1.01072085e+00 1.45258915e+00
-8.67575631e-02 -1.32411444e+00 -3.25482607e-01 -2.10988194e-01
3.74006510e-01 5.22772297e-02 -1.80629775e-01 1.91236362e-01
-7.18509734e-01 -1.06503379e+00 6.03216350e-01 1.04747975e+00
1.15419698e+00 -7.99682081e-01 1.13664913e+00 2.09446251e-01
-1.44714141e+00 -2.62012035e-01 -5.31635433e-02 1.31968735e-02
-9.09405015e-03 1.23301792e+00 -5.23633957e-01 9.86600399e-01
1.02879989e+00 1.04632640e+00 -1.78828284e-01 1.08414304e+00
-1.11560198e-02 1.04854214e+00 -3.81034315e-01 2.65934646e-01
-2.87239254e-01 -7.38906711e-02 6.44179344e-01 1.37715757e+00
5.12527406e-01 1.53658286e-01 4.48049270e-02 6.70417368e-01
2.16918930e-01 -6.59218282e-02 -3.28741908e-01 9.70493183e-02
2.00072110e-01 1.13486087e+00 -6.61030829e-01 -5.45006990e-01
-4.19006087e-02 6.60898566e-01 -3.52476239e-01 3.76039147e-01
-7.91668534e-01 -9.37450349e-01 2.82947868e-01 -3.55737358e-02
1.19500108e-01 -7.36371055e-02 -3.13184291e-01 -9.82585311e-01
3.31495464e-01 -1.00049222e+00 4.89522308e-01 -2.59200335e-01
-9.97907817e-01 1.07498205e+00 -2.64659494e-01 -1.64052439e+00
-4.61080641e-01 -4.43292171e-01 -5.50341070e-01 1.22929537e+00
-1.39103329e+00 -5.31812668e-01 -3.18146735e-01 5.96908569e-01
2.39106700e-01 -2.33040184e-01 1.43831074e+00 6.36983693e-01
-3.05929147e-02 5.61971426e-01 -1.51252180e-01 1.98905945e-01
5.63027561e-01 -1.32596385e+00 3.23042542e-01 8.07089508e-01
1.06634490e-01 7.69912839e-01 5.64746797e-01 -4.04293925e-01
-9.07937944e-01 -1.06849909e+00 9.05969501e-01 -2.03896135e-01
2.01329619e-01 9.30666551e-03 -7.62889981e-01 4.94003631e-02
1.78957134e-02 4.67433840e-01 8.56075943e-01 -5.98403029e-02
6.49730908e-03 -3.03654909e-01 -9.22649145e-01 1.49431145e-02
6.78769886e-01 -5.33745527e-01 -4.31445211e-01 2.79933631e-01
1.80952132e-01 -5.27146041e-01 -1.16930139e+00 9.64789391e-01
1.00197184e+00 -1.10673642e+00 9.11344707e-01 -4.94673759e-01
1.51968062e-01 -4.16265041e-01 2.34697714e-01 -1.21457028e+00
-3.39339823e-01 -1.15438926e+00 -4.36745226e-01 4.60353911e-01
1.71780631e-01 -6.49235904e-01 4.53925490e-01 -5.49101830e-01
-4.42568868e-01 -1.14362705e+00 -1.04502594e+00 -6.38284802e-01
-1.94599524e-01 -5.98310590e-01 1.64239720e-01 7.94651449e-01
-2.57852167e-01 1.19199201e-01 -3.48268330e-01 2.09679365e-01
4.23057348e-01 1.55361354e-01 9.74353179e-02 -1.65859318e+00
-4.10660982e-01 -9.66870710e-02 -6.53191268e-01 -4.79137361e-01
-1.81407392e-01 -8.18116426e-01 -7.65015632e-02 -1.33418751e+00
-1.12227775e-01 -1.60821810e-01 -1.01095665e+00 1.34578183e-01
-2.48065621e-01 3.59937489e-01 -1.95134535e-01 8.00589174e-02
-1.79212570e-01 5.76875769e-02 8.09565783e-01 -8.46328437e-02
-3.43067378e-01 5.45512497e-01 -2.92463094e-01 8.69417667e-01
8.82563829e-01 -4.34063405e-01 -3.06082010e-01 1.14105657e-01
2.08114028e-01 2.40275517e-01 3.10514212e-01 -1.63474691e+00
-2.10746527e-01 7.30343997e-01 7.32406437e-01 -7.39080369e-01
2.34953180e-01 -6.12989306e-01 3.31426591e-01 1.01895499e+00
-2.67313749e-01 3.09160590e-01 2.82294035e-01 6.61620498e-01
-3.41397792e-01 9.55108777e-02 8.05582404e-01 5.22525646e-02
3.07402372e-01 3.72523338e-01 -7.63368487e-01 1.05889417e-01
5.94181240e-01 -1.39676437e-01 9.38245431e-02 -1.67053640e-01
-1.32429159e+00 -3.60468775e-01 -6.08529150e-01 7.07417950e-02
9.13643003e-01 -1.30282795e+00 -1.16367137e+00 2.76486188e-01
1.15188383e-01 -1.77872494e-01 5.60482085e-01 1.32063913e+00
-8.78363907e-01 2.19588906e-01 -6.66805133e-02 -1.05835545e+00
-1.22085905e+00 -3.64306271e-02 8.62350881e-01 -4.12083238e-01
-8.51480603e-01 6.26836181e-01 -1.68886781e-01 1.39135092e-01
2.85135448e-01 -7.34111071e-01 -3.72133344e-01 -1.90830119e-02
6.60919845e-01 3.67777884e-01 5.39231360e-01 -2.20868468e-01
-5.65212607e-01 5.86761594e-01 2.09516719e-01 1.65090367e-01
1.10692632e+00 4.13811892e-01 -2.19152093e-01 6.85705364e-01
1.00092554e+00 -2.50469089e-01 -5.63305080e-01 -3.50468978e-02
-9.90075171e-02 3.72603536e-02 6.31703660e-02 -8.79892290e-01
-9.51671124e-01 1.21985233e+00 1.68328512e+00 4.73070413e-01
1.39960122e+00 -7.13100493e-01 8.37520719e-01 3.59150827e-01
-1.24664018e-02 -8.16358685e-01 -1.84156641e-01 3.14565152e-01
8.51076484e-01 -6.48540854e-01 1.33924365e-01 -1.36477798e-01
-1.50240615e-01 1.74612665e+00 -2.42490962e-01 -3.67128938e-01
1.37422717e+00 1.80818886e-01 3.57134700e-01 -2.63416529e-01
-5.28863251e-01 2.34718323e-02 4.38502938e-01 5.54486156e-01
9.57296669e-01 1.69569612e-01 -7.66760707e-01 7.50118256e-01
-7.01106638e-02 4.38281149e-01 5.26882648e-01 8.35648179e-01
5.53500233e-03 -8.54915023e-01 -1.83161572e-01 6.67186677e-01
-1.12112617e+00 -1.90670595e-01 -2.47210357e-03 4.07761216e-01
4.25986320e-01 1.01711941e+00 -1.97759062e-01 -2.24672571e-01
2.13310421e-01 3.70091766e-01 3.21931034e-01 -4.25554901e-01
-8.93901110e-01 3.93430799e-01 -5.00219204e-02 -4.28972602e-01
-4.72886115e-01 -5.53527236e-01 -1.05186462e+00 5.77644110e-01
-3.43163818e-01 1.43141508e-01 7.80287206e-01 7.46748865e-01
6.14911139e-01 1.15416133e+00 4.97101367e-01 -6.46063209e-01
-5.48017085e-01 -1.24122834e+00 -6.62134290e-01 2.14931071e-01
8.77126873e-01 -3.14990729e-01 -3.90520096e-01 2.57456452e-01] | [14.290655136108398, 3.2768101692199707] |
5a41d528-85d6-4979-8e32-7f8a9f82ddc4 | mwp-bert-a-strong-baseline-for-math-word | 2107.13435 | null | https://arxiv.org/abs/2107.13435v2 | https://arxiv.org/pdf/2107.13435v2.pdf | MWP-BERT: Numeracy-Augmented Pre-training for Math Word Problem Solving | Math word problem (MWP) solving faces a dilemma in number representation learning. In order to avoid the number representation issue and reduce the search space of feasible solutions, existing works striving for MWP solving usually replace real numbers with symbolic placeholders to focus on logic reasoning. However, different from common symbolic reasoning tasks like program synthesis and knowledge graph reasoning, MWP solving has extra requirements in numerical reasoning. In other words, instead of the number value itself, it is the reusable numerical property that matters more in numerical reasoning. Therefore, we argue that injecting numerical properties into symbolic placeholders with contextualized representation learning schema can provide a way out of the dilemma in the number representation issue here. In this work, we introduce this idea to the popular pre-training language model (PLM) techniques and build MWP-BERT, an effective contextual number representation PLM. We demonstrate the effectiveness of our MWP-BERT on MWP solving and several MWP-specific understanding tasks on both English and Chinese benchmarks. | ['Jie Shao', 'Yunshi Lan', 'Wei Qin', 'Lei Wang', 'Xiangliang Zhang', 'Jipeng Zhang', 'Zhenwen Liang'] | 2021-07-28 | null | https://aclanthology.org/2022.findings-naacl.74 | https://aclanthology.org/2022.findings-naacl.74.pdf | findings-naacl-2022-7 | ['math-word-problem-solving', 'math-word-problem-solving', 'math-word-problem-solving'] | ['knowledge-base', 'reasoning', 'time-series'] | [ 1.06010400e-01 3.03437978e-01 -5.13071954e-01 -1.96091518e-01
-4.20575917e-01 -6.53697670e-01 2.47127891e-01 3.84026140e-01
-3.28238726e-01 7.23655224e-01 2.25597173e-01 -9.45433557e-01
-1.74469918e-01 -1.58177495e+00 -9.00254250e-01 -1.01307549e-01
1.53146505e-01 3.92102122e-01 9.71403942e-02 -5.06534696e-01
5.80011368e-01 4.52351272e-01 -1.19829357e+00 5.91466606e-01
9.27383959e-01 6.40105605e-01 -1.17634840e-01 3.73980165e-01
-9.67482984e-01 1.38137567e+00 -7.03808904e-01 -6.15176022e-01
2.15318441e-01 -1.27428532e-01 -1.12480295e+00 -7.13611424e-01
1.84575081e-01 -3.57356071e-01 -1.29491270e-01 9.91482913e-01
1.07059656e-02 2.51858085e-01 6.79779351e-01 -1.63170242e+00
-7.25460529e-01 1.56878603e+00 -5.31768918e-01 -1.88983493e-02
5.69244981e-01 2.83644527e-01 1.29820859e+00 -4.49334353e-01
5.91177940e-01 1.62203610e+00 7.80214608e-01 5.70604861e-01
-1.06819844e+00 -5.83425701e-01 1.49203151e-01 4.14948106e-01
-1.33422089e+00 6.92280829e-02 8.53088140e-01 -2.01934904e-01
1.40083778e+00 4.43553954e-01 9.17802036e-01 5.05072117e-01
-5.39059117e-02 7.18167245e-01 8.47533882e-01 -8.19434643e-01
1.76259894e-02 5.68643510e-02 5.55384040e-01 7.97401607e-01
3.14732850e-01 -3.03069413e-01 -3.52104545e-01 1.89154577e-02
9.99786139e-01 -2.72477925e-01 -1.40194982e-01 -2.63371080e-01
-1.25390184e+00 1.05702341e+00 3.81287545e-01 3.21217299e-01
3.11077714e-01 8.00250292e-01 5.30950725e-01 4.58003610e-01
-2.37472206e-01 9.60310042e-01 -6.50564194e-01 -3.94550979e-01
-7.81283498e-01 4.88926649e-01 1.05389333e+00 8.83935034e-01
6.88630164e-01 1.04573570e-01 -2.00167492e-01 4.39850658e-01
2.36595318e-01 2.99399406e-01 2.74814576e-01 -1.07158029e+00
9.48464453e-01 1.00993145e+00 -5.57652153e-02 -1.23472238e+00
-4.19315845e-01 5.63264117e-02 -4.73297477e-01 5.88039458e-02
8.06884408e-01 1.40061662e-01 -3.51947248e-01 1.85229862e+00
2.32083648e-02 2.42408022e-01 2.36846671e-01 5.56001961e-01
8.86018395e-01 1.13381350e+00 1.63162962e-01 -6.04663938e-02
1.33933103e+00 -9.96307671e-01 -6.12344265e-01 -1.17248133e-01
1.32777619e+00 -1.58420950e-01 1.50478041e+00 2.94107884e-01
-1.13846469e+00 -2.39297107e-01 -1.18732166e+00 -4.70231444e-01
-7.35757291e-01 -8.61100554e-02 1.30286324e+00 7.78500021e-01
-7.71092415e-01 6.75491095e-01 -3.83799851e-01 1.67486250e-01
3.93741518e-01 3.30829680e-01 -6.32115304e-02 -1.87277630e-01
-1.56800091e+00 1.14831853e+00 7.61217058e-01 1.21242472e-03
-1.48251399e-01 -1.15765882e+00 -1.15440392e+00 4.46439952e-01
8.05394292e-01 -6.43388867e-01 1.12764621e+00 -8.13758969e-01
-1.56429446e+00 5.69146991e-01 1.30245676e-02 -7.36743331e-01
3.98068994e-01 -6.24880344e-02 -7.66111016e-02 -1.11377567e-01
-2.71259546e-01 7.27585495e-01 4.78658050e-01 -8.06325793e-01
-7.03281686e-02 6.02575168e-02 6.72288835e-01 2.90030781e-02
-3.65819745e-02 8.53480548e-02 -2.96556894e-02 -6.87573791e-01
7.58035779e-02 -4.25451338e-01 -2.42698997e-01 -1.12876808e-02
-3.13432664e-02 -7.83960283e-01 2.74273634e-01 -3.83484334e-01
1.54582012e+00 -2.00545692e+00 3.35845441e-01 5.44290960e-01
3.33492517e-01 2.34099433e-01 -1.82858050e-01 3.80993426e-01
-2.33016685e-01 5.12134075e-01 4.66630012e-02 4.18206662e-01
4.63920325e-01 3.61721307e-01 -7.04631925e-01 1.16194105e-02
4.56781268e-01 1.39216089e+00 -9.63332474e-01 -7.30713487e-01
1.73976868e-01 -1.24056116e-01 -9.51327384e-01 -2.75223583e-01
-9.52750921e-01 -2.16692701e-01 -3.08457971e-01 5.72611690e-01
8.71206284e-01 -3.02721858e-01 4.50790614e-01 -4.62145768e-02
1.12714566e-01 4.94837761e-01 -1.52993286e+00 1.71637249e+00
-6.78462386e-01 3.29434514e-01 -5.18354237e-01 -1.02718258e+00
7.12600768e-01 -1.47351965e-01 -3.81388180e-02 -6.88372076e-01
1.15183592e-01 2.69718498e-01 1.80581957e-01 -2.71089435e-01
8.27048302e-01 -1.66910455e-01 -3.44738185e-01 4.64715868e-01
-1.26624167e-01 -6.48230731e-01 3.58767778e-01 3.84048015e-01
8.65708232e-01 5.49726903e-01 6.53981566e-01 1.37357293e-02
8.21899414e-01 2.19623864e-01 5.40559173e-01 7.09686935e-01
1.88139275e-01 1.57919191e-02 1.21403635e+00 -5.44775486e-01
-7.73883641e-01 -8.85434568e-01 1.71941489e-01 9.83482480e-01
-1.21143751e-01 -9.93848562e-01 -5.47400713e-01 -4.78674382e-01
-1.30924480e-02 1.09067214e+00 -2.13554919e-01 -2.80739814e-01
-1.20248997e+00 -2.12633342e-01 9.84311461e-01 9.06334877e-01
4.57826674e-01 -1.23250175e+00 -6.68551505e-01 3.22418511e-01
-2.88840830e-01 -9.87526357e-01 2.75854040e-02 1.93601251e-01
-8.46442103e-01 -9.15642142e-01 -2.22175777e-01 -6.87564015e-01
5.25839448e-01 -1.14696741e-01 1.40430999e+00 5.02345264e-01
-1.21723503e-01 1.16670050e-01 -2.69496590e-01 -3.79044384e-01
-4.97034281e-01 -7.56448135e-02 -4.20124322e-01 -8.38926733e-01
4.99109119e-01 -4.48602229e-01 1.13582378e-02 -3.54726285e-01
-7.15124726e-01 3.60597342e-01 2.14419410e-01 5.78952312e-01
1.77442297e-01 4.28805023e-01 4.47602391e-01 -8.15516531e-01
5.50933123e-01 -1.81906477e-01 -8.73802304e-01 5.97380757e-01
-1.37414500e-01 4.57173258e-01 9.53375518e-01 -5.07804215e-01
-6.11013412e-01 -2.92828381e-01 7.56851137e-02 -1.89911515e-01
4.62084651e-01 7.52312839e-01 -3.84793907e-01 3.34125347e-02
6.11359179e-01 1.36879891e-01 -1.20526493e-01 2.49834478e-01
8.65105152e-01 -1.48336425e-01 2.62679815e-01 -1.52441692e+00
8.78189147e-01 -3.81081775e-02 3.79809946e-01 -2.53021419e-01
-6.88623667e-01 -1.71366781e-02 -3.51292014e-01 1.66528270e-01
5.46439111e-01 -6.97801888e-01 -1.18468761e+00 1.87742174e-01
-1.40228999e+00 -5.44520915e-01 -5.82120240e-01 1.50326803e-01
-7.37962365e-01 4.21577096e-01 -5.81358016e-01 -9.87633884e-01
-5.61743937e-02 -1.15364647e+00 5.58813930e-01 1.05566129e-01
-6.53556645e-01 -8.44727874e-01 -2.95498759e-01 1.61021203e-01
3.24209809e-01 1.82654738e-01 2.02596831e+00 -5.26548982e-01
-7.55111337e-01 -2.22819485e-03 -5.62394023e-01 2.67336369e-01
-1.97721303e-01 -1.27682939e-01 -5.26055157e-01 4.25786674e-01
-9.09463838e-02 -4.14106727e-01 6.71616435e-01 8.48579332e-02
1.52612662e+00 -6.89627945e-01 6.94027320e-02 5.06678402e-01
1.62951982e+00 -1.31762758e-01 1.13817084e+00 5.19591570e-01
5.51183641e-01 4.36612159e-01 4.48967397e-01 3.67688835e-01
5.36587834e-01 3.57351691e-01 1.32279009e-01 3.88967335e-01
-1.15401611e-01 -5.23868561e-01 2.30157420e-01 3.47266197e-01
-1.84462503e-01 6.96613491e-02 -1.11688340e+00 2.44928166e-01
-1.80262625e+00 -9.73338664e-01 2.45823320e-02 1.92812788e+00
1.50948620e+00 1.40593052e-01 4.20711674e-02 5.38066566e-01
1.79908499e-01 2.49451578e-01 -3.30126584e-02 -8.39809835e-01
2.24571619e-02 7.54030168e-01 3.51607621e-01 5.72464406e-01
-6.31362915e-01 1.22146142e+00 5.09902239e+00 1.14017403e+00
-8.55391324e-01 -3.55397373e-01 3.03642988e-01 1.63435161e-01
-6.24195039e-01 7.56132230e-02 -7.22781718e-01 -1.22626463e-03
7.16422021e-01 -2.76668787e-01 8.29631507e-01 1.04752469e+00
-3.76024991e-01 -4.50350270e-02 -1.53036547e+00 1.16892862e+00
-2.35710964e-01 -1.65711522e+00 4.16823596e-01 -5.55383325e-01
2.46165171e-01 -9.10541654e-01 6.89003393e-02 8.23338747e-01
3.51365358e-01 -1.72117591e+00 8.77519250e-01 4.36238050e-01
5.47391236e-01 -9.33718324e-01 5.85692823e-01 2.95945108e-01
-1.18304336e+00 -2.65373170e-01 -5.47178984e-01 -6.30449831e-01
-1.85398757e-01 2.35790357e-01 -6.55859828e-01 6.27011478e-01
1.67215481e-01 5.97821414e-01 -4.97787207e-01 6.57882631e-01
-5.91873229e-01 1.66084379e-01 -2.81037629e-01 -4.55909163e-01
3.82513732e-01 -2.57849485e-01 1.31413639e-01 1.02414382e+00
2.01459303e-01 4.02820528e-01 -1.03089467e-01 1.42694998e+00
-1.55772597e-01 1.79502159e-01 -4.86393929e-01 -4.12200481e-01
5.07905960e-01 8.19276035e-01 -8.68868947e-01 -5.15866220e-01
-5.22963226e-01 3.00567210e-01 5.38385153e-01 2.26243392e-01
-1.20210886e+00 -4.25145447e-01 2.68593401e-01 9.40291137e-02
1.98899865e-01 -5.10761380e-01 -7.03467846e-01 -1.15950298e+00
1.32876024e-01 -1.25781906e+00 3.00952941e-01 -7.62835979e-01
-8.77278388e-01 -1.38772383e-01 4.19887424e-01 -8.69469106e-01
-1.17883638e-01 -9.98317957e-01 -6.12604022e-01 9.25030649e-01
-1.64597201e+00 -1.00392735e+00 5.87456487e-02 3.77915472e-01
2.46601030e-01 8.31412449e-02 1.00501132e+00 -1.81696042e-01
-2.17697442e-01 7.08974540e-01 -6.06824458e-01 2.74401098e-01
1.40638486e-01 -1.14918065e+00 1.94715038e-01 5.84169388e-01
7.66836181e-02 1.19321597e+00 5.03961384e-01 -3.61828834e-01
-1.93379080e+00 -6.61469519e-01 9.87438023e-01 -5.25746167e-01
1.02487886e+00 -2.49033853e-01 -9.06965137e-01 6.84248030e-01
-7.07579181e-02 -2.68062800e-01 5.27464867e-01 1.63945243e-01
-1.00911748e+00 9.78168100e-02 -1.14118648e+00 9.96219814e-01
9.26728427e-01 -4.07192349e-01 -1.10437334e+00 3.06767315e-01
1.20007622e+00 -6.67870402e-01 -6.41012669e-01 3.14510137e-01
4.39389378e-01 -4.63120818e-01 1.36043608e+00 -7.94404507e-01
8.60538483e-01 -4.33009088e-01 -3.35683316e-01 -9.69176054e-01
1.23054748e-02 -5.29421449e-01 -4.73436832e-01 1.18830585e+00
3.59185010e-01 -7.14289904e-01 7.48951077e-01 8.65601897e-01
1.02210499e-01 -7.92841613e-01 -8.71290803e-01 -7.58096695e-01
8.05653274e-01 -9.09542799e-01 1.16040957e+00 9.38093543e-01
4.59455311e-01 1.57910243e-01 1.36634838e-02 -1.46611154e-01
3.92261222e-02 4.16882932e-01 7.38144577e-01 -9.51270401e-01
-5.54343760e-01 -7.33039141e-01 -2.49167874e-01 -8.39427471e-01
6.64798260e-01 -1.33187091e+00 -4.80773330e-01 -1.62113023e+00
-1.44083455e-01 -6.24419749e-01 -1.34840906e-01 8.42404246e-01
2.52553653e-02 -3.29713911e-01 4.85284358e-01 -2.91757971e-01
-3.94859254e-01 1.61148846e-01 1.35817695e+00 -3.72419119e-01
-1.75427020e-01 -3.88052970e-01 -7.40776718e-01 8.54314864e-01
8.60720098e-01 -2.88845628e-01 -5.46493888e-01 -4.05228615e-01
1.20729423e+00 2.91164637e-01 5.49161375e-01 -8.74763966e-01
2.15736628e-01 -6.77092016e-01 9.99340862e-02 -5.38069487e-01
1.01665586e-01 -8.73397410e-01 -3.49464983e-01 6.45166159e-01
-3.66405308e-01 1.36591658e-01 5.04529417e-01 4.56648245e-02
-5.20851798e-02 -7.52952039e-01 3.77136081e-01 -5.97604752e-01
-1.01286936e+00 -3.54775250e-01 -2.10596710e-01 3.45310330e-01
7.19176233e-01 -8.39027241e-02 -6.32130563e-01 -7.21465796e-02
-3.32412094e-01 2.86922038e-01 2.69906104e-01 2.37980913e-02
7.40320861e-01 -1.20994759e+00 -2.40585104e-01 -2.24786382e-02
-3.75412367e-02 2.31075197e-01 8.08902904e-02 6.38566911e-01
-9.60517049e-01 6.34437621e-01 -1.55408069e-01 9.73547101e-02
-9.30342376e-01 9.01700795e-01 3.04819167e-01 -7.76098311e-01
-7.50115156e-01 8.56982112e-01 -2.23381072e-01 -6.59867465e-01
4.23541099e-01 -1.29135776e+00 -2.63885856e-01 -4.81512956e-02
5.61304390e-01 2.99189746e-01 -1.27298310e-01 2.33329386e-01
-3.33609462e-01 5.55425644e-01 1.45262942e-01 1.04821794e-01
1.22976553e+00 4.59927827e-01 -7.28807271e-01 4.36199069e-01
7.45340705e-01 2.17032433e-01 -4.34360921e-01 -1.97622240e-01
2.97108322e-01 -3.47658753e-01 -5.73625505e-01 -7.74669886e-01
-6.26831412e-01 7.85732090e-01 -2.58243144e-01 -1.25405416e-01
8.67719233e-01 -3.31401303e-02 7.53312886e-01 9.56076860e-01
6.85604870e-01 -9.19571459e-01 1.10560179e-01 9.08136010e-01
7.87693799e-01 -8.68493378e-01 2.38627851e-01 -7.23061919e-01
-4.88788426e-01 1.51226330e+00 6.68204129e-01 -2.43485644e-01
2.17772782e-01 5.41493535e-01 -3.22530389e-01 3.95965427e-02
-6.54114008e-01 1.76965460e-01 -1.97063331e-02 5.05058348e-01
5.81126451e-01 3.94420177e-02 -3.59113902e-01 1.09764051e+00
-7.81148612e-01 3.02961469e-01 7.11342633e-01 9.51028645e-01
-3.00716698e-01 -1.29975259e+00 -4.53766763e-01 2.47259170e-01
-1.66429847e-01 -7.10028172e-01 -1.31872352e-02 1.11391366e+00
8.77914056e-02 4.18116719e-01 -8.91405642e-02 8.28172266e-02
7.73574645e-03 4.78219599e-01 1.10363114e+00 -7.29080737e-01
-6.58837020e-01 -6.83933437e-01 2.80449480e-01 -5.72388411e-01
-2.24328130e-01 -1.20226607e-01 -1.93800402e+00 -6.52487278e-01
-3.62235568e-02 2.25726187e-01 2.07566723e-01 9.43917990e-01
-2.83542514e-01 8.38342309e-01 -7.43476972e-02 -1.28553286e-01
-8.27853024e-01 -4.56880927e-01 -2.93300569e-01 1.47245759e-02
-2.23783888e-02 -4.08521295e-01 -1.92044258e-01 -3.94284070e-01] | [9.491168022155762, 7.369532585144043] |
3269aa66-e864-4a72-b25c-5e1637b0278a | oracle-efficient-smoothed-online-learning-for | 2302.05430 | null | https://arxiv.org/abs/2302.05430v1 | https://arxiv.org/pdf/2302.05430v1.pdf | Oracle-Efficient Smoothed Online Learning for Piecewise Continuous Decision Making | Smoothed online learning has emerged as a popular framework to mitigate the substantial loss in statistical and computational complexity that arises when one moves from classical to adversarial learning. Unfortunately, for some spaces, it has been shown that efficient algorithms suffer an exponentially worse regret than that which is minimax optimal, even when the learner has access to an optimization oracle over the space. To mitigate that exponential dependence, this work introduces a new notion of complexity, the generalized bracketing numbers, which marries constraints on the adversary to the size of the space, and shows that an instantiation of Follow-the-Perturbed-Leader can attain low regret with the number of calls to the optimization oracle scaling optimally with respect to average regret. We then instantiate our bounds in several problems of interest, including online prediction and planning of piecewise continuous functions, which has many applications in fields as diverse as econometrics and robotics. | ['Max Simchowitz', 'Alexander Rakhlin', 'Adam Block'] | 2023-02-10 | null | null | null | null | ['econometrics'] | ['miscellaneous'] | [ 2.51825064e-01 6.55201197e-01 -1.06619976e-01 -2.85911858e-01
-1.45885754e+00 -1.02105570e+00 2.18181744e-01 4.01690722e-01
-8.13070774e-01 1.06723738e+00 -1.86911359e-01 -5.82514465e-01
-4.21072543e-01 -6.55425310e-01 -1.34969246e+00 -9.71802533e-01
-5.49560905e-01 2.26626486e-01 -1.39562972e-02 -1.99654177e-01
3.16447496e-01 4.85797107e-01 -8.91561449e-01 -2.58378834e-01
7.35027850e-01 1.09840739e+00 -2.32809454e-01 8.62593234e-01
5.96746206e-02 4.52202111e-01 -4.72687453e-01 -6.00572944e-01
9.45464253e-01 -2.39930764e-01 -7.23786175e-01 7.79839465e-03
5.57815552e-01 -3.07913154e-01 -5.40482938e-01 1.41025341e+00
4.28607136e-01 3.25517148e-01 3.14904809e-01 -1.30382872e+00
-1.66730553e-01 8.90076041e-01 -3.86965841e-01 -8.64448026e-03
1.83536723e-01 -1.11280054e-01 1.25358772e+00 -2.43242800e-01
5.35848200e-01 1.14284348e+00 7.78587162e-01 4.61834610e-01
-1.37543273e+00 -5.02524495e-01 3.02005589e-01 1.11677364e-01
-1.16616714e+00 -2.91079104e-01 4.88916516e-01 -4.54923213e-01
3.47764820e-01 5.27203441e-01 1.74952418e-01 8.78791392e-01
2.14701742e-01 8.79490733e-01 1.04586577e+00 -4.40269887e-01
6.68793440e-01 2.13161081e-01 -2.30618849e-01 6.81990623e-01
3.91171843e-01 3.96991998e-01 -2.31251299e-01 -4.06496555e-01
4.69285607e-01 -1.18357040e-01 -2.60327190e-01 -5.73452890e-01
-6.83991730e-01 1.14897382e+00 4.23879653e-01 -7.69203976e-02
-8.57398361e-02 2.03418747e-01 2.86357850e-01 8.06267560e-01
6.11670315e-01 7.42434323e-01 -6.34375930e-01 -1.18662730e-01
-6.69349432e-01 3.82230610e-01 1.32585585e+00 8.95074904e-01
7.64086068e-01 -1.79161862e-01 6.85210526e-02 3.20986062e-01
-1.74559593e-01 6.08799756e-01 -1.60337135e-01 -1.35980117e+00
8.97718787e-01 -3.18070464e-02 7.58927584e-01 -9.69373703e-01
-2.67348617e-01 -4.99460280e-01 -5.86304426e-01 5.16725540e-01
9.21902597e-01 -6.16429031e-01 -1.99753210e-01 2.08267426e+00
5.52841425e-01 -1.23724379e-01 1.39366418e-01 9.11298275e-01
-3.87106299e-01 4.93697703e-01 -4.39617366e-01 -4.81032550e-01
6.24174118e-01 -8.87578189e-01 -5.80769897e-01 -3.55256289e-01
7.12275267e-01 -4.42789495e-01 8.42485905e-01 5.02135217e-01
-1.18827987e+00 1.39810473e-01 -1.07227182e+00 1.87550336e-02
-2.37782210e-01 -7.71753013e-01 6.59044683e-01 9.31277514e-01
-8.12419951e-01 1.15013826e+00 -1.00764561e+00 -7.54645988e-02
6.40869975e-01 3.80006254e-01 -4.70243514e-01 9.09477323e-02
-8.89050245e-01 7.81008661e-01 2.64387339e-01 6.55187247e-03
-9.78977501e-01 -8.02880943e-01 -7.34163105e-01 7.46327452e-03
9.18406427e-01 -4.60482299e-01 1.48506558e+00 -1.14114380e+00
-1.61715734e+00 4.59345311e-01 3.27955261e-02 -9.57644761e-01
1.23613632e+00 -4.14518863e-01 3.42747122e-01 -1.03831835e-01
-1.73335858e-02 -3.37542593e-01 8.34205031e-01 -8.73714805e-01
-7.82270789e-01 -6.07047081e-01 7.85447359e-01 2.23959297e-01
-1.93588346e-01 -3.40480924e-01 1.60085976e-01 -3.36294800e-01
-1.25284111e-02 -1.04114878e+00 -8.07988882e-01 3.52888376e-01
-3.51956367e-01 3.86531688e-02 3.94547015e-01 -3.83741617e-01
7.52898335e-01 -1.97789538e+00 1.98434845e-01 2.30622277e-01
-3.91377397e-02 -4.58575152e-02 -4.17924859e-03 5.51497281e-01
3.62875015e-02 2.65685469e-01 -3.32218677e-01 -4.17689174e-01
3.11921358e-01 3.91333163e-01 -5.65928280e-01 1.02939439e+00
-4.31681782e-01 5.30530095e-01 -1.09971893e+00 -1.19057320e-01
-3.91497128e-02 -2.65859336e-01 -5.88301420e-01 8.57227594e-02
-2.37896562e-01 4.38765824e-01 -6.54476285e-01 1.14549272e-01
7.78380513e-01 -1.13590926e-01 1.01389378e-01 5.66940904e-01
-4.67072278e-02 3.30872200e-02 -1.49551463e+00 1.72822583e+00
-6.26965642e-01 4.57437873e-01 7.70033956e-01 -1.32607925e+00
1.80930406e-01 1.18120044e-01 4.77386653e-01 -7.61731043e-02
1.18527919e-01 2.17271522e-01 -3.22693050e-01 -4.22142118e-01
2.36681625e-01 -4.16689485e-01 -2.65307784e-01 2.80417293e-01
-2.27003783e-01 -7.48810731e-03 -1.09067813e-01 6.71505034e-02
1.35398304e+00 -9.60867628e-02 3.01518857e-01 -1.70872256e-01
5.91517687e-01 1.67011917e-02 5.49963295e-01 1.40596652e+00
-2.11366072e-01 8.53151232e-02 7.17307925e-01 -2.59826779e-01
-1.06175923e+00 -1.07143605e+00 5.42599708e-03 1.41275382e+00
2.48165965e-01 -3.55842561e-02 -7.74033606e-01 -7.22423911e-01
4.38652545e-01 7.16922283e-01 -6.79365754e-01 -8.59068558e-02
-3.38505030e-01 -5.44169545e-01 5.02482116e-01 2.23292574e-01
2.45848030e-01 -3.30324411e-01 -2.47924909e-01 1.87973365e-01
1.80004075e-01 -1.11268806e+00 -6.35190308e-01 2.86479473e-01
-8.15300167e-01 -1.16471970e+00 -3.29207271e-01 -2.30602250e-01
6.85368896e-01 1.46990716e-01 6.58467352e-01 -3.52442533e-01
-8.32908750e-02 7.32281446e-01 -1.59402892e-01 -6.37063086e-01
-4.33627784e-01 2.44007543e-01 1.21293686e-01 9.50693190e-02
-3.48000288e-01 -5.55209875e-01 -4.08216894e-01 -6.16126657e-02
-7.30256677e-01 -3.33619446e-01 3.08222622e-01 8.77561569e-01
5.20058632e-01 9.39882249e-02 2.53287107e-01 -1.20238686e+00
3.60257238e-01 -5.66055655e-01 -1.33232784e+00 5.37618138e-02
-5.76852322e-01 4.26423281e-01 1.15822899e+00 -2.28887036e-01
-6.06007755e-01 2.48986527e-01 2.00609058e-01 -3.10518384e-01
2.58867413e-01 2.93475389e-01 -2.65784562e-01 -5.69149315e-01
5.72777212e-01 1.75375670e-01 6.35959134e-02 -3.73007774e-01
4.47943449e-01 3.38696480e-01 5.60239196e-01 -7.32240081e-01
1.24147654e+00 5.58153987e-01 6.58427835e-01 -7.93563843e-01
-1.50785673e+00 -2.61831611e-01 -3.91479194e-01 8.28460827e-02
4.52023804e-01 -5.94177306e-01 -1.01746464e+00 2.46737152e-01
-1.02906561e+00 -3.32475394e-01 -6.59017146e-01 3.56950581e-01
-9.47823405e-01 4.31065410e-01 -3.82522553e-01 -1.38677478e+00
-2.71126125e-02 -7.39140570e-01 6.04243994e-01 1.45097926e-01
4.81642902e-01 -1.21667683e+00 -4.95816953e-02 1.63249910e-01
3.49501371e-01 6.67915583e-01 4.36644137e-01 -9.68780339e-01
-8.03442478e-01 -4.94064063e-01 2.08198637e-01 4.87860590e-01
-2.13612035e-01 -4.94376540e-01 -9.20544386e-01 -6.68231130e-01
2.62264371e-01 -3.30479532e-01 5.70078731e-01 2.41504237e-01
1.33877075e+00 -1.12641215e+00 -8.42994377e-02 9.76521254e-01
1.60620356e+00 -1.35160148e-01 5.19576818e-02 5.64546108e-01
3.72579128e-01 4.49637592e-01 6.86646342e-01 6.19971752e-01
4.03561220e-02 3.47728103e-01 6.48631692e-01 2.89547682e-01
8.63205731e-01 -3.74673128e-01 4.06078339e-01 4.22463715e-01
1.30503103e-01 -5.86644188e-02 -3.96835327e-01 2.93448567e-01
-1.97651362e+00 -1.07233596e+00 3.70820880e-01 2.74890137e+00
1.08534753e+00 2.18680397e-01 -2.53228825e-02 4.69129416e-04
5.77347815e-01 2.72214204e-01 -9.25220132e-01 -5.93828797e-01
-4.08714414e-02 1.60574287e-01 1.45739365e+00 1.03167927e+00
-1.21585584e+00 7.19026268e-01 6.73820734e+00 8.97337139e-01
-9.00307655e-01 2.93581128e-01 4.50031400e-01 -1.51277527e-01
8.47106567e-04 2.39916712e-01 -5.64652264e-01 3.28530401e-01
1.03153527e+00 -7.38135874e-01 1.09228134e+00 1.27020574e+00
2.00594738e-01 9.60659161e-02 -1.50001597e+00 6.51083469e-01
-2.19028607e-01 -1.11587691e+00 -5.71265697e-01 2.57955045e-01
8.60055327e-01 -2.85964878e-03 8.56784359e-02 2.20970005e-01
6.51178062e-01 -9.31038260e-01 6.54775441e-01 5.41902483e-01
4.13725436e-01 -9.75106657e-01 6.61056161e-01 8.88767004e-01
-8.67496431e-01 -4.41882372e-01 -5.19674361e-01 -2.95441926e-01
-2.03848537e-02 2.66930819e-01 -8.75859797e-01 5.74280620e-01
2.58367509e-01 6.17189147e-02 -9.19668227e-02 1.27232623e+00
-1.77492633e-01 5.65481305e-01 -7.29931951e-01 -1.44532368e-01
2.58207798e-01 -4.32629198e-01 9.54047382e-01 9.62769926e-01
1.45867482e-01 7.08628595e-02 5.38875639e-01 5.59444547e-01
-2.09664822e-01 1.49513379e-01 -6.63786054e-01 1.38812116e-03
5.17706871e-01 9.86043692e-01 -2.04295784e-01 -6.28509521e-02
-3.17280620e-01 1.02067220e+00 3.76662940e-01 3.18719149e-01
-7.67764151e-01 -5.96339464e-01 7.61092365e-01 6.07145857e-03
4.31481123e-01 -4.25107718e-01 -2.84129292e-01 -9.35215116e-01
2.70892322e-01 -6.31725788e-01 4.54324275e-01 6.74225688e-02
-1.27630615e+00 -1.27338886e-01 -2.03747734e-01 -9.47369277e-01
-3.07349652e-01 -5.99957228e-01 -4.38378364e-01 6.89499140e-01
-1.21081567e+00 -5.34294307e-01 1.50194511e-01 4.29040164e-01
2.77525306e-01 5.73514402e-03 7.74348140e-01 3.12295649e-02
-4.62973058e-01 9.87448335e-01 9.14709568e-01 -3.96378748e-02
4.66360599e-01 -1.79291677e+00 7.29826763e-02 9.94105637e-01
8.62355977e-02 2.06979454e-01 1.21227145e+00 -1.79043025e-01
-1.85583079e+00 -1.03385329e+00 5.12070060e-01 -5.01533449e-01
1.15019774e+00 -4.09244508e-01 -5.79886556e-01 9.20061290e-01
-3.73386860e-01 2.36937881e-01 4.29609120e-01 4.19393927e-02
-2.57852882e-01 -4.02425826e-01 -1.37006390e+00 5.63964307e-01
9.98821735e-01 -5.74629188e-01 -3.41824442e-01 7.68382668e-01
8.36111963e-01 -8.16007912e-01 -8.37038875e-01 -1.31495640e-01
4.67275381e-01 -7.76550353e-01 7.91569471e-01 -9.83553588e-01
-5.56394160e-02 8.26065093e-02 -2.25805819e-01 -1.25417113e+00
6.37379661e-02 -1.54502010e+00 -2.79281139e-01 8.28079283e-01
4.03403580e-01 -9.63368475e-01 9.35852885e-01 8.45776498e-01
3.33072506e-02 -8.30142081e-01 -1.40085864e+00 -1.02135420e+00
5.22445858e-01 -4.74622905e-01 2.16348499e-01 5.91273725e-01
-2.01386824e-01 4.54144785e-03 -4.35945749e-01 6.94487929e-01
9.80182111e-01 5.74386567e-02 1.11116016e+00 -1.08044636e+00
-5.94297409e-01 -3.12930316e-01 -3.29695106e-01 -1.33466554e+00
2.15142876e-01 -9.16851580e-01 8.72551277e-02 -8.90187979e-01
-1.54734373e-01 -4.09645051e-01 -3.05277586e-01 2.25650400e-01
1.05297454e-01 -4.50511217e-01 2.89278775e-01 -2.13714078e-01
-8.69416654e-01 4.81562883e-01 1.22884190e+00 -9.18906555e-02
-1.70324951e-01 6.31231070e-01 -8.77305448e-01 9.47254300e-01
5.09268999e-01 -6.49137795e-01 -1.75408363e-01 -2.64307171e-01
2.24657267e-01 4.59006995e-01 2.52069294e-01 -7.43474483e-01
4.90169138e-01 -6.26681030e-01 -1.30194485e-01 -1.30824909e-01
2.20616505e-01 -9.67915714e-01 -1.75532177e-01 5.09808302e-01
-8.06171119e-01 -3.37288469e-01 -1.52590230e-01 1.08393335e+00
3.59827012e-01 -6.27105355e-01 9.46713209e-01 -3.22022699e-02
4.02004533e-02 5.34328699e-01 -1.09504305e-01 6.07677519e-01
1.10818863e+00 2.53443331e-01 -1.59549206e-01 -7.04009652e-01
-8.01927388e-01 4.38086510e-01 4.28081423e-01 -1.96915101e-02
-3.84983490e-03 -1.09422219e+00 -6.01189971e-01 -1.11890100e-01
-4.12383825e-01 1.42716691e-01 1.44501030e-01 9.55081582e-01
-3.50102663e-01 2.64578670e-01 3.50199372e-01 -1.12466380e-01
-9.99667883e-01 6.62772238e-01 4.09643590e-01 -2.68073171e-01
-6.64803326e-01 8.38687420e-01 -1.03366219e-01 -1.40604287e-01
6.62436664e-01 -2.47996002e-01 5.90866327e-01 -2.14914650e-01
5.36914766e-01 7.70750821e-01 -8.71165171e-02 -2.17620268e-01
-4.30802219e-02 2.09175020e-01 -1.15117066e-01 -2.15888500e-01
1.27315104e+00 -3.47932667e-01 6.22058026e-02 7.18130589e-01
1.25321972e+00 3.93589824e-01 -1.70433867e+00 -4.50306445e-01
3.67622972e-01 -6.57377720e-01 -2.23900527e-01 -4.25776780e-01
-8.14246953e-01 5.27424097e-01 4.83285457e-01 5.91063857e-01
7.48074651e-01 -2.11025715e-01 8.14245045e-01 1.10464966e+00
8.29464614e-01 -1.34875250e+00 -4.39646333e-01 7.15076745e-01
7.36001492e-01 -1.21653068e+00 2.06847787e-02 -7.85786361e-02
-2.64329970e-01 1.00380623e+00 5.19219078e-02 -6.11840308e-01
6.50499284e-01 2.67984927e-01 -2.24634573e-01 6.12960815e-01
-6.96161270e-01 7.75605487e-03 -4.31224331e-02 3.57175767e-01
-2.25339696e-01 2.16621488e-01 -3.43441099e-01 8.05074871e-01
-5.62222540e-01 -3.22140932e-01 5.98766565e-01 9.60384429e-01
-6.41897142e-01 -1.05060065e+00 -5.17883599e-01 2.54877150e-01
-9.61212099e-01 1.65382206e-01 -1.46517530e-01 7.32211351e-01
-6.09662347e-02 8.93860400e-01 -1.97862491e-01 -7.43825804e-04
2.10969627e-01 -3.27870771e-02 5.19917846e-01 -3.84726971e-01
-2.87215710e-01 -2.45532289e-01 5.49331121e-03 -8.58203769e-01
-8.73690546e-02 -6.29432261e-01 -9.51616585e-01 -4.37683105e-01
-3.64804417e-01 4.74225432e-01 6.96837187e-01 1.13798642e+00
3.26823071e-02 1.38844736e-02 1.26406050e+00 -7.78244436e-01
-1.76529837e+00 -7.03767776e-01 -8.79976571e-01 1.25776991e-01
8.62953544e-01 -3.17192584e-01 -1.09645593e+00 -3.03488046e-01] | [4.7900896072387695, 3.4681856632232666] |
e6057a66-52c8-4ad7-bab1-64182b6a9c06 | hscnet-hierarchical-scene-coordinate | 2305.03595 | null | https://arxiv.org/abs/2305.03595v1 | https://arxiv.org/pdf/2305.03595v1.pdf | HSCNet++: Hierarchical Scene Coordinate Classification and Regression for Visual Localization with Transformer | Visual localization is critical to many applications in computer vision and robotics. To address single-image RGB localization, state-of-the-art feature-based methods match local descriptors between a query image and a pre-built 3D model. Recently, deep neural networks have been exploited to regress the mapping between raw pixels and 3D coordinates in the scene, and thus the matching is implicitly performed by the forward pass through the network. However, in a large and ambiguous environment, learning such a regression task directly can be difficult for a single network. In this work, we present a new hierarchical scene coordinate network to predict pixel scene coordinates in a coarse-to-fine manner from a single RGB image. The proposed method, which is an extension of HSCNet, allows us to train compact models which scale robustly to large environments. It sets a new state-of-the-art for single-image localization on the 7-Scenes, 12 Scenes, Cambridge Landmarks datasets, and the combined indoor scenes. | ['Juho Kannala', 'Giorgos Tolias', 'Yi Zhao', 'Xiaotian Li', 'Iaroslav Melekhov', 'Zakaria Laskar', 'Shuzhe Wang'] | 2023-05-05 | null | null | null | null | ['visual-localization'] | ['computer-vision'] | [ 1.61747143e-01 -4.09038723e-01 4.71952558e-02 -7.24798143e-01
-8.97703111e-01 -3.48896801e-01 5.11875570e-01 -2.10806318e-02
-8.52114081e-01 3.91875178e-01 -2.81287909e-01 -3.50389667e-02
-1.04484474e-03 -7.08293974e-01 -1.18556452e+00 -5.35890639e-01
2.30695456e-01 4.90116507e-01 3.42228830e-01 -4.04375717e-02
2.59402305e-01 9.03672695e-01 -1.58144736e+00 -3.14179435e-03
3.55130196e-01 1.39560032e+00 5.70199132e-01 5.04180729e-01
-1.26738921e-01 7.46124029e-01 -1.68973267e-01 1.87619820e-01
4.95438904e-01 -2.17199549e-01 -5.39566934e-01 -9.89882350e-02
9.91527438e-01 -1.95183426e-01 -4.75530654e-01 1.14678681e+00
4.79922980e-01 2.89202332e-01 3.43212247e-01 -1.09188735e+00
-6.55878067e-01 -1.78259816e-02 -4.02168006e-01 -1.94272697e-01
4.78386819e-01 -6.34856001e-02 7.65175104e-01 -1.01216090e+00
5.91847479e-01 1.14823294e+00 9.72982943e-01 2.39309564e-01
-1.22698760e+00 -4.91858393e-01 1.73168495e-01 3.46213311e-01
-1.82679391e+00 -2.74253875e-01 9.18974698e-01 -3.16463530e-01
1.27880502e+00 -4.86091077e-02 8.22205782e-01 7.15898335e-01
2.47215644e-01 4.79662806e-01 1.29302979e+00 -4.80741233e-01
2.84606934e-01 -1.07024394e-01 -2.51684546e-01 1.02169383e+00
-1.25899598e-01 6.58826604e-02 -6.99026525e-01 2.99846441e-01
1.07022369e+00 4.13900554e-01 -1.43546402e-01 -9.08195138e-01
-1.22978258e+00 8.41872334e-01 1.31879830e+00 3.62018943e-01
-3.36397737e-01 5.98840535e-01 -8.41812342e-02 5.95081411e-02
3.45660418e-01 5.35682321e-01 -5.65388262e-01 6.91577233e-03
-8.99412513e-01 7.53970444e-02 7.00480998e-01 1.07736325e+00
1.34210706e+00 -2.95469582e-01 2.63905525e-01 6.45243585e-01
5.27503788e-01 7.84983218e-01 4.73639309e-01 -9.75992084e-01
2.58722544e-01 5.51440775e-01 3.91140347e-04 -1.24555516e+00
-7.40553021e-01 -2.44968936e-01 -8.99717271e-01 2.82918632e-01
3.42500269e-01 4.86338347e-01 -1.12715387e+00 1.61877334e+00
2.46067107e-01 2.30289146e-01 -2.51476884e-01 1.14699674e+00
8.12196374e-01 4.31346953e-01 -2.76764601e-01 2.90364414e-01
8.15029681e-01 -1.15045846e+00 -2.79842794e-01 -6.16134763e-01
3.90266538e-01 -6.83762848e-01 9.76607680e-01 -3.73946130e-02
-8.23577881e-01 -8.50271583e-01 -1.14428329e+00 -5.60577452e-01
-8.87254238e-01 9.65996906e-02 6.64075851e-01 1.30694464e-01
-1.44100785e+00 5.16850471e-01 -1.08285844e+00 -9.16783750e-01
2.81166553e-01 6.26286924e-01 -7.70107150e-01 -4.39174265e-01
-8.51180673e-01 1.14473140e+00 3.03487718e-01 4.00899351e-01
-8.08411300e-01 -3.88206363e-01 -1.28851485e+00 -5.17644919e-02
1.38898626e-01 -7.54917860e-01 1.02151549e+00 -6.47023797e-01
-1.52247870e+00 1.01536429e+00 -3.71559411e-01 -3.57366174e-01
4.17356253e-01 -3.99679214e-01 1.71473801e-01 -6.60950765e-02
4.52970445e-01 9.42219555e-01 6.01099372e-01 -1.31415987e+00
-6.12259388e-01 -7.16107130e-01 8.68605748e-02 3.13516080e-01
1.87684163e-01 -2.99827993e-01 -7.40980685e-01 2.70077796e-03
7.95461595e-01 -1.01666939e+00 -5.92341363e-01 3.70696276e-01
-3.05938840e-01 -7.01429546e-02 4.95869339e-01 -2.81964600e-01
3.61279339e-01 -2.15578270e+00 1.76655173e-01 2.61360943e-01
-7.10292086e-02 -2.23935157e-01 -1.57274142e-01 7.77817816e-02
1.70562446e-01 -1.81104124e-01 -1.54540747e-01 -6.44774497e-01
2.10204929e-01 3.32830995e-01 -3.85600515e-02 1.00972056e+00
-7.45425969e-02 9.80729580e-01 -8.77048612e-01 -4.58378643e-01
6.47519767e-01 5.69303572e-01 -4.81305808e-01 4.14520681e-01
6.08329549e-02 6.24500155e-01 -2.77089924e-01 7.35011339e-01
7.17073917e-01 -2.39644483e-01 -3.94402891e-01 -3.78291607e-01
-3.32861036e-01 2.88904868e-02 -1.18114388e+00 2.75060153e+00
-7.67945528e-01 6.89356148e-01 -3.43336277e-02 -1.04754448e+00
1.04412544e+00 -3.38438302e-01 5.40276706e-01 -9.46058750e-01
7.75809214e-02 4.41009969e-01 -5.10437191e-01 -2.17775121e-01
5.56389034e-01 1.22297138e-01 -2.44074255e-01 -6.83979318e-02
3.50259364e-01 -6.78580165e-01 -1.05027165e-02 -2.98411876e-01
9.89102066e-01 5.31533599e-01 3.94096106e-01 1.49637181e-02
5.44309616e-01 6.29332364e-02 3.36956531e-01 1.26139569e+00
-1.55158728e-01 8.80328834e-01 -2.32435212e-01 -7.82665610e-01
-9.90835905e-01 -1.01360357e+00 -7.27149323e-02 9.14605677e-01
6.26423419e-01 -3.04743290e-01 -3.89090776e-01 -3.97309631e-01
1.82268336e-01 1.04354672e-01 -4.78815138e-01 -1.29032299e-01
-6.09705150e-01 -1.97681561e-01 2.61307716e-01 5.99713802e-01
9.14741278e-01 -9.56686139e-01 -7.61459649e-01 1.41602397e-01
5.25991991e-02 -1.37280047e+00 -2.15416461e-01 7.15637445e-01
-6.14847898e-01 -1.05162406e+00 -4.80999738e-01 -1.10377765e+00
8.06759715e-01 4.30508584e-01 9.60374832e-01 -1.58543289e-01
-4.09020185e-01 4.87358868e-01 -1.58881143e-01 -1.31364495e-01
1.87860504e-01 1.58019394e-01 2.07259789e-01 -1.65472001e-01
4.79987055e-01 -4.56266969e-01 -4.92375106e-01 2.58721441e-01
-5.35739481e-01 5.22805564e-02 7.13519931e-01 8.95070732e-01
1.20039558e+00 -2.80492365e-01 -2.72680223e-01 -3.77111465e-01
-1.54224988e-02 4.89144400e-02 -9.58900094e-01 2.57498235e-01
-2.93488741e-01 1.10417105e-01 6.37133837e-01 -1.33292884e-01
-5.97354889e-01 9.00112152e-01 -2.89843202e-01 -5.85797071e-01
-5.33369541e-01 5.12393415e-01 -1.16904013e-01 -7.66414225e-01
5.77372313e-01 2.57832408e-01 -3.39788646e-01 -5.90243757e-01
7.38996148e-01 4.81497258e-01 8.62447679e-01 -3.89868796e-01
9.01788116e-01 5.72759032e-01 3.85151297e-01 -7.04868615e-01
-9.41428363e-01 -8.79241645e-01 -1.50376976e+00 1.17495023e-01
1.07321131e+00 -1.15021527e+00 -6.59587860e-01 5.69103539e-01
-1.38018215e+00 -5.44447601e-01 -1.86072499e-01 6.17084146e-01
-8.28355014e-01 -4.95557720e-03 -3.65196407e-01 -3.37289155e-01
2.88497098e-02 -1.23675895e+00 1.43450809e+00 4.12337065e-01
2.10019886e-01 -7.66177833e-01 1.56851083e-01 -4.59693559e-02
4.04790640e-01 2.25570634e-01 3.81752431e-01 -2.79211879e-01
-9.79435384e-01 -3.43658537e-01 -5.29104233e-01 5.83359376e-02
1.23342305e-01 -4.20969695e-01 -1.09136915e+00 -3.00354451e-01
-6.60341457e-02 -5.55814028e-01 8.87077749e-01 3.38266075e-01
1.13123667e+00 2.98434734e-01 -2.88861841e-01 1.41242957e+00
1.75664985e+00 6.76052645e-02 1.74851701e-01 6.80248857e-01
1.00553012e+00 1.22702144e-01 4.92264837e-01 1.76355168e-01
6.13444865e-01 7.94457436e-01 6.91281974e-01 -5.09530723e-01
-6.20221309e-02 -4.02940035e-01 -1.29169807e-01 8.17841351e-01
6.33094087e-02 3.02965522e-01 -9.50550854e-01 3.25396955e-01
-1.95886588e+00 -5.13312280e-01 2.43437931e-01 2.04473209e+00
5.29638886e-01 -1.86518021e-02 -5.94766200e-01 -2.98544735e-01
4.48402792e-01 2.66899258e-01 -6.93184555e-01 -7.46576535e-03
-1.99124798e-01 2.96813130e-01 8.82551968e-01 5.55019617e-01
-1.43999517e+00 1.35859931e+00 5.90971613e+00 5.09700894e-01
-1.48110652e+00 1.47713184e-01 3.40239018e-01 2.09306180e-01
2.97698349e-01 6.09263591e-03 -9.05305028e-01 1.88362075e-03
7.45131493e-01 4.33738381e-01 7.11396158e-01 1.25424206e+00
-8.60760137e-02 -3.70512724e-01 -1.33111155e+00 1.53707004e+00
4.41317379e-01 -1.23186529e+00 -2.88445503e-01 -8.28847289e-02
7.93221295e-01 6.88741207e-01 -3.51365991e-02 2.18052074e-01
1.21929780e-01 -1.07719445e+00 9.44266021e-01 6.57931745e-01
7.95661032e-01 -5.55472970e-01 8.10392320e-01 3.52869719e-01
-1.39256847e+00 -1.13346420e-01 -8.06701303e-01 -1.60978705e-01
1.54147536e-01 3.14243406e-01 -7.86598563e-01 3.32322925e-01
1.13027525e+00 1.01882482e+00 -1.02595842e+00 1.25736070e+00
-4.74995434e-01 -3.10963750e-01 -6.40842676e-01 -7.49231428e-02
4.02927518e-01 -1.98769063e-01 -6.99292943e-02 1.04757941e+00
4.85637844e-01 -2.43466049e-01 5.31901479e-01 1.02027690e+00
-8.63003656e-02 -1.22574039e-01 -9.99395490e-01 4.12444323e-01
3.86682093e-01 1.26307821e+00 -8.08348119e-01 -1.08303510e-01
-4.70748216e-01 1.20570159e+00 8.32819760e-01 4.49525505e-01
-6.38175786e-01 -5.15712321e-01 6.14867449e-01 -2.76394159e-01
4.45315570e-01 -8.21813703e-01 -2.52179742e-01 -1.18537414e+00
-1.30118076e-02 -3.36021900e-01 -1.64991349e-01 -1.10682082e+00
-1.09859431e+00 6.08564794e-01 -2.02262878e-01 -1.10895348e+00
-3.11846286e-01 -9.18385148e-01 -1.26400247e-01 9.28226113e-01
-1.98044705e+00 -1.42004335e+00 -8.29927862e-01 9.79479432e-01
3.44896048e-01 1.26786873e-01 9.99083221e-01 1.90911204e-01
-1.22087717e-01 2.75628388e-01 3.11001897e-01 2.98332602e-01
7.56325424e-01 -1.33901393e+00 3.78484815e-01 7.92482853e-01
6.00716352e-01 6.88760340e-01 2.91875243e-01 -1.99371457e-01
-1.73535943e+00 -1.04253602e+00 8.57788861e-01 -4.95764673e-01
5.37668586e-01 -7.22545505e-01 -4.98138458e-01 6.43613458e-01
-6.20213412e-02 6.24160171e-01 2.34264001e-01 1.62205584e-02
-3.41952741e-01 -5.40041387e-01 -1.03516352e+00 3.26036662e-01
1.22515976e+00 -1.12234724e+00 -3.37820947e-01 3.84551823e-01
7.66732037e-01 -9.67921674e-01 -8.70470822e-01 3.16996843e-01
5.06576359e-01 -1.03070116e+00 1.29798710e+00 -9.45661291e-02
-5.97444065e-02 -5.44091225e-01 -6.89714074e-01 -1.29083943e+00
-3.34348947e-01 -2.03002632e-01 1.91515580e-01 8.69172871e-01
1.36821121e-01 -3.90149474e-01 8.10785830e-01 5.85636199e-01
-3.26798171e-01 -4.66737181e-01 -1.25260091e+00 -6.54827714e-01
-2.68891245e-01 -5.71742177e-01 5.71732521e-01 5.43527961e-01
-5.07815242e-01 2.23360226e-01 -1.57788351e-01 3.60668570e-01
5.81796050e-01 3.28346282e-01 1.06554699e+00 -1.15950561e+00
7.01534078e-02 -2.98919052e-01 -9.35402453e-01 -1.50171018e+00
3.58788997e-01 -8.30120087e-01 7.57555425e-01 -1.70787358e+00
-8.60308483e-02 -5.72579920e-01 -4.10203904e-01 6.08089685e-01
1.92637622e-01 6.00260139e-01 3.42332125e-01 2.51857162e-01
-8.85232568e-01 5.81982255e-01 9.87735212e-01 -4.15054083e-01
-2.32645646e-01 -1.91566661e-01 -1.65819407e-01 7.13403285e-01
5.51852524e-01 -4.33022290e-01 -1.24779314e-01 -6.55147731e-01
1.26396090e-01 -3.51799279e-01 6.11313224e-01 -1.37092268e+00
8.43347847e-01 -6.28103241e-02 7.96191335e-01 -8.02003682e-01
7.63014495e-01 -1.24737108e+00 3.53814997e-02 2.98336387e-01
-1.93237677e-01 2.51325756e-01 -3.65198627e-02 5.50264001e-01
-3.97402644e-01 8.62650294e-03 6.90639794e-01 -6.23876333e-01
-1.27119362e+00 5.32280207e-01 2.50188112e-01 -2.09914595e-01
7.25979447e-01 -2.49563739e-01 -1.19040571e-01 -1.76172748e-01
-5.67234218e-01 -4.54830825e-02 7.58940160e-01 3.09730619e-01
8.30770910e-01 -1.50253487e+00 -1.17910773e-01 5.52134693e-01
2.66227305e-01 6.10221505e-01 -9.85118300e-02 8.61537933e-01
-9.58392382e-01 6.98063016e-01 -3.40003550e-01 -1.12392974e+00
-8.86733055e-01 6.63750052e-01 5.16520381e-01 4.27758768e-02
-5.86283863e-01 1.07615936e+00 3.46446000e-02 -8.53417754e-01
5.04322827e-01 -7.28464305e-01 1.19321458e-01 -2.82879800e-01
1.72961563e-01 -1.04125932e-01 1.36742875e-01 -1.11180925e+00
-6.25549555e-01 1.28566885e+00 3.57222229e-01 2.16216482e-02
1.39568055e+00 -3.98651689e-01 -3.96267146e-01 6.48455024e-01
1.69202244e+00 -4.92530048e-01 -1.43331921e+00 -5.15003502e-01
-1.96968447e-02 -7.17958212e-01 1.61356956e-01 -4.57201809e-01
-9.35126603e-01 1.02240646e+00 9.45710003e-01 -1.96180299e-01
9.33286607e-01 1.37753963e-01 4.80957270e-01 9.66642320e-01
8.91671598e-01 -9.90226865e-01 8.97290260e-02 9.29622114e-01
7.49408305e-01 -1.72563410e+00 3.17123681e-02 4.52167243e-02
-1.04005918e-01 1.20277548e+00 7.79351592e-01 -2.83920079e-01
8.00188661e-01 -2.51698047e-02 2.76394755e-01 -7.33720064e-02
6.26593307e-02 -4.20872867e-01 3.67254436e-01 7.32356846e-01
2.18997076e-01 -1.33893654e-01 5.44200063e-01 9.91671309e-02
-2.25618631e-01 -1.12624146e-01 -2.91989092e-02 1.07137489e+00
-4.03333277e-01 -7.96418607e-01 -3.73767465e-01 -8.27824920e-02
4.24629636e-02 -2.17726529e-01 -3.56350154e-01 8.99018228e-01
4.18097317e-01 7.41414547e-01 1.50338009e-01 -4.02571023e-01
4.97487724e-01 -5.88527471e-02 6.75674260e-01 -4.46604162e-01
-2.47303009e-01 3.30932401e-02 -4.78233248e-01 -1.19585621e+00
-8.06560338e-01 -4.50627834e-01 -1.17070782e+00 5.68142310e-02
-1.93237394e-01 -1.82883561e-01 1.23460627e+00 1.05143297e+00
2.69800127e-01 3.32155228e-01 5.21272242e-01 -1.72716320e+00
-2.31245935e-01 -8.23676825e-01 -6.86933875e-01 3.31229448e-01
6.11655235e-01 -7.21721947e-01 -1.37043744e-01 -5.75806908e-02] | [7.687151908874512, -2.159194231033325] |
c4c624d8-1714-4c80-915a-df978eed627f | human-de-occlusion-invisible-perception-and | 2103.11597 | null | https://arxiv.org/abs/2103.11597v1 | https://arxiv.org/pdf/2103.11597v1.pdf | Human De-occlusion: Invisible Perception and Recovery for Humans | In this paper, we tackle the problem of human de-occlusion which reasons about occluded segmentation masks and invisible appearance content of humans. In particular, a two-stage framework is proposed to estimate the invisible portions and recover the content inside. For the stage of mask completion, a stacked network structure is devised to refine inaccurate masks from a general instance segmentation model and predict integrated masks simultaneously. Additionally, the guidance from human parsing and typical pose masks are leveraged to bring prior information. For the stage of content recovery, a novel parsing guided attention module is applied to isolate body parts and capture context information across multiple scales. Besides, an Amodal Human Perception dataset (AHP) is collected to settle the task of human de-occlusion. AHP has advantages of providing annotations from real-world scenes and the number of humans is comparatively larger than other amodal perception datasets. Based on this dataset, experiments demonstrate that our method performs over the state-of-the-art techniques in both tasks of mask completion and content recovery. Our AHP dataset is available at \url{https://sydney0zq.github.io/ahp/}. | ['Xinggang Wang', 'Zilong Huang', 'Yitong Wang', 'Shiyin Wang', 'Qiang Zhou'] | 2021-03-22 | null | http://openaccess.thecvf.com//content/CVPR2021/html/Zhou_Human_De-Occlusion_Invisible_Perception_and_Recovery_for_Humans_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Zhou_Human_De-Occlusion_Invisible_Perception_and_Recovery_for_Humans_CVPR_2021_paper.pdf | cvpr-2021-1 | ['human-parsing'] | ['computer-vision'] | [ 5.41474164e-01 4.54368800e-01 -1.70214608e-01 -4.30553824e-01
-5.75859845e-01 -1.76885426e-01 1.59175903e-01 -2.20126852e-01
-3.26965392e-01 5.13541400e-01 3.44989121e-01 2.02038497e-01
3.43097419e-01 -4.04567599e-01 -7.03168452e-01 -4.14530128e-01
3.71417493e-01 3.12381625e-01 3.60635608e-01 -9.76065770e-02
-5.73260635e-02 2.80050095e-02 -1.64149237e+00 5.14236808e-01
1.00784838e+00 1.12854600e+00 4.20435548e-01 4.21256930e-01
5.75520806e-02 6.20420098e-01 -4.27731723e-01 -6.58931553e-01
4.55866724e-01 -3.25349391e-01 -7.45453179e-01 5.45401514e-01
6.51946783e-01 -4.80328113e-01 -4.50438797e-01 1.10238993e+00
5.62748849e-01 1.39298007e-01 3.45825076e-01 -1.05288208e+00
-4.93117750e-01 4.64055657e-01 -1.07442701e+00 1.84313893e-01
5.36875546e-01 3.12632293e-01 7.94103801e-01 -9.49988306e-01
6.41299367e-01 1.35590053e+00 3.10749263e-01 6.43284917e-01
-1.19632471e+00 -6.24796629e-01 6.45878136e-01 6.42741472e-02
-1.17420852e+00 -4.93743122e-01 9.28875983e-01 -4.91757125e-01
4.46883172e-01 3.26270372e-01 9.03480113e-01 1.24720418e+00
-3.05504296e-02 1.20537543e+00 1.07818091e+00 -2.99948603e-01
-5.01477607e-02 -2.44117677e-02 2.92399466e-01 7.50237584e-01
2.58635223e-01 -5.66358641e-02 -6.56162500e-01 2.46298864e-01
9.39966321e-01 1.41656846e-01 -3.75966966e-01 -2.07614213e-01
-1.18188357e+00 4.18506414e-01 3.89233470e-01 -1.94935128e-02
-4.41487014e-01 -9.04162303e-02 1.22445993e-01 -1.97065428e-01
5.05593717e-01 5.28048649e-02 -3.65333408e-01 2.38609836e-01
-8.69798958e-01 3.94181818e-01 5.77911079e-01 9.97420847e-01
6.48591518e-01 -1.57172576e-01 -4.19163913e-01 8.86074781e-01
3.17302197e-01 4.56532508e-01 2.37906396e-01 -1.11665499e+00
8.35785985e-01 7.88739145e-01 2.60866225e-01 -1.03421128e+00
-4.28622276e-01 -3.11777949e-01 -8.47563267e-01 5.48358113e-02
6.52714372e-01 1.87087862e-04 -1.31085265e+00 1.68654001e+00
6.93636596e-01 -4.60489281e-02 -4.37408715e-01 1.41955984e+00
1.09924006e+00 3.96884829e-01 1.85556233e-01 1.10301435e-01
1.87952423e+00 -1.14246356e+00 -8.08778226e-01 -7.68228889e-01
-1.77693576e-01 -5.60940385e-01 1.13917184e+00 5.02940893e-01
-1.29580176e+00 -9.29439723e-01 -9.59443212e-01 -4.72941816e-01
-1.40895814e-01 3.96057904e-01 6.13650560e-01 2.70354986e-01
-6.93948388e-01 2.22085565e-01 -7.17992604e-01 -2.01318994e-01
6.86021686e-01 1.95352063e-01 -3.40842426e-01 -2.12515622e-01
-1.02463019e+00 6.30835414e-01 4.49443698e-01 4.73404706e-01
-8.64542067e-01 -3.19870651e-01 -1.04699409e+00 -1.08002953e-01
7.22138345e-01 -8.27322483e-01 1.01516747e+00 -1.05752110e+00
-1.24259079e+00 1.14637029e+00 -1.91275224e-01 -9.96102095e-02
7.18389452e-01 -7.25352287e-01 -1.48935527e-01 3.91238719e-01
2.87643492e-01 9.21325088e-01 1.18397987e+00 -1.30493426e+00
-5.01644611e-01 -5.22382855e-01 6.12421595e-02 3.93134564e-01
1.02482662e-01 -4.19279896e-02 -1.10801876e+00 -9.25289810e-01
3.34509820e-01 -7.58782208e-01 -2.42719442e-01 8.15522894e-02
-8.68578672e-01 -4.72254343e-02 2.47958004e-01 -1.29585266e+00
1.17115104e+00 -2.10472822e+00 5.57624936e-01 8.78416747e-02
5.75787425e-01 4.59449254e-02 -2.46184766e-02 9.38401446e-02
5.75261563e-02 -1.39107630e-01 -5.85885763e-01 -7.99342990e-01
1.06636159e-01 1.37989774e-01 -3.41520756e-01 4.95090276e-01
2.58566320e-01 9.14848328e-01 -5.35504222e-01 -7.80691206e-01
2.39582121e-01 4.71970260e-01 -4.71519679e-01 3.68289113e-01
-2.75225371e-01 1.00260746e+00 -3.39874595e-01 1.09102547e+00
8.49415123e-01 -2.80243963e-01 9.49870795e-02 -3.79256934e-01
1.15049802e-01 -1.06347136e-01 -1.16704881e+00 2.10657930e+00
4.93649393e-02 1.81118146e-01 5.57942390e-01 -7.93415546e-01
4.65844184e-01 2.27286980e-01 3.62093657e-01 -7.34282792e-01
4.45959359e-01 -8.18521753e-02 -1.70475975e-01 -7.94701457e-01
3.87980253e-01 2.59220570e-01 -6.34693727e-02 1.38020486e-01
2.37697661e-02 9.09088105e-02 3.18185121e-01 1.28611073e-01
7.23362744e-01 5.85372031e-01 1.72672495e-01 -7.59088397e-02
4.99516308e-01 -1.43412173e-01 9.87841547e-01 6.47816539e-01
-5.75188577e-01 8.78817260e-01 2.56289989e-01 -5.75852215e-01
-8.03601921e-01 -1.10161030e+00 -1.70175843e-02 1.37888992e+00
6.20356917e-01 -3.49001825e-01 -1.05022979e+00 -5.64755201e-01
-8.66252929e-02 3.26671958e-01 -7.50282347e-01 2.88258255e-01
-7.57567465e-01 -6.67990446e-01 1.84287831e-01 5.91516912e-01
7.40278184e-01 -1.50117648e+00 -7.81477690e-01 -1.47324419e-02
-7.47682333e-01 -1.28890169e+00 -6.23873591e-01 -1.78172708e-01
-6.25744164e-01 -1.30297673e+00 -7.74413109e-01 -7.84732699e-01
8.95700336e-01 1.45869717e-01 1.07463980e+00 1.17562599e-01
-7.40883410e-01 2.05344081e-01 -1.64431036e-01 -3.97043526e-01
1.73234969e-01 -2.07481951e-01 -1.85819402e-01 2.19413623e-01
1.54878646e-01 -5.72240591e-01 -1.08373141e+00 3.48590314e-01
-7.87405312e-01 6.15181088e-01 8.41381669e-01 5.19840121e-01
8.24980676e-01 -1.53391525e-01 2.12807477e-01 -9.03133571e-01
2.33055890e-01 -3.01701248e-01 -4.11523134e-01 2.33343929e-01
-1.33400445e-03 -3.56085122e-01 1.67011485e-01 -5.74424863e-01
-1.40539348e+00 3.42134356e-01 -1.87233910e-02 -2.89196283e-01
-5.45721292e-01 -7.38237277e-02 -5.83077133e-01 4.01192576e-01
2.77599066e-01 2.36619562e-02 -1.46826595e-01 -7.09585130e-01
5.26783824e-01 3.90827447e-01 9.00515139e-01 -5.54044068e-01
5.45019746e-01 6.79354787e-01 -4.33435529e-01 -4.85384911e-01
-1.25343478e+00 -4.39796716e-01 -9.98989284e-01 -2.87533671e-01
1.22582185e+00 -1.08616006e+00 -5.24421513e-01 5.91156542e-01
-1.07425940e+00 -3.88770074e-01 -1.15846582e-01 2.24740565e-01
-4.37019736e-01 3.92484814e-01 -8.33500803e-01 -8.70867729e-01
-4.57444221e-01 -1.04912448e+00 1.24580145e+00 4.09390002e-01
-2.04589292e-01 -4.46871519e-01 -3.30304950e-01 8.71873379e-01
-4.30463776e-02 4.42157209e-01 4.43511724e-01 -2.36679405e-01
-6.50257170e-01 -1.10557236e-01 -3.82031471e-01 8.44453722e-02
-9.68542323e-02 -4.58706737e-01 -1.19267285e+00 -2.71483243e-01
1.73040226e-01 -3.67979348e-01 1.11004639e+00 6.01893008e-01
1.28490877e+00 -2.64279068e-01 -3.77260506e-01 6.68916643e-01
9.58185434e-01 -2.02569738e-01 4.87918258e-01 1.10194646e-01
1.03540039e+00 1.04146004e+00 6.75214469e-01 4.32321131e-01
5.35006881e-01 5.03801525e-01 3.67722154e-01 -4.04026866e-01
-5.01093149e-01 -3.85570049e-01 5.55557460e-02 5.38620234e-01
-1.23653397e-01 -1.64554402e-01 -6.59176350e-01 4.10668254e-01
-1.93909264e+00 -6.96832299e-01 6.45338595e-02 2.03091788e+00
7.84004092e-01 1.92714274e-01 2.72455513e-01 -8.90767649e-02
9.95336115e-01 1.76073492e-01 -6.84575737e-01 2.55756468e-01
-3.53595838e-02 -5.79020493e-02 1.59804359e-01 4.76104677e-01
-1.45722115e+00 9.61148620e-01 5.48782253e+00 6.23218656e-01
-6.24326468e-01 3.24665278e-01 7.70723164e-01 -1.84554249e-01
3.28017026e-02 -2.09090441e-01 -7.29886472e-01 6.70838058e-01
1.39125600e-01 5.17643750e-01 4.60919887e-01 6.28589690e-01
1.22846048e-02 -3.47069949e-01 -8.32699656e-01 1.18037701e+00
3.14292520e-01 -4.99073625e-01 -2.02700600e-01 -1.26909455e-02
4.82444704e-01 -2.20805794e-01 6.17719106e-02 1.62155122e-01
6.31092191e-02 -9.59416270e-01 1.01735413e+00 7.38386869e-01
7.98902869e-01 -4.49957013e-01 5.46678782e-01 3.81507218e-01
-1.32936883e+00 -2.63757348e-01 -3.55925351e-01 -1.91685602e-01
4.18614537e-01 6.97708905e-01 -2.03059152e-01 5.20008624e-01
8.83731186e-01 6.00330293e-01 -6.80506170e-01 1.02733195e+00
-6.73436999e-01 3.51450980e-01 -1.79632634e-01 6.24056816e-01
-3.45431000e-01 -2.39380524e-01 5.52952707e-01 1.09632218e+00
-1.71197176e-01 3.95678639e-01 4.81881231e-01 1.13546646e+00
-4.87108575e-03 5.67941926e-03 -2.09183469e-01 3.01161736e-01
2.67590106e-01 1.44679427e+00 -9.29110825e-01 -4.99768764e-01
-4.90223259e-01 1.26314688e+00 5.24008691e-01 5.84836781e-01
-1.11248016e+00 -1.13372147e-01 2.99230307e-01 2.41419151e-01
4.15409297e-01 2.76554283e-02 -3.80874753e-01 -1.36169863e+00
3.69566590e-01 -8.65363896e-01 5.17502248e-01 -7.44141281e-01
-1.36120498e+00 5.48826814e-01 7.39581510e-02 -1.05113828e+00
2.35900730e-01 -4.35811698e-01 -5.26397705e-01 7.73230493e-01
-1.26841342e+00 -1.44982445e+00 -6.70104861e-01 5.43208480e-01
8.61608565e-01 3.23574454e-01 3.43268245e-01 3.25053811e-01
-1.06542778e+00 3.59817952e-01 -7.97992885e-01 2.51226574e-01
6.78812146e-01 -1.10621548e+00 3.89215738e-01 1.16951084e+00
-1.02275953e-01 5.46182573e-01 6.06208384e-01 -1.07319725e+00
-1.04236186e+00 -8.99559438e-01 3.99813354e-01 -5.29109895e-01
2.52072960e-01 -5.93594074e-01 -8.14542294e-01 6.38779342e-01
1.74367264e-01 7.80887604e-02 4.23427284e-01 -1.71639249e-01
-2.21880734e-01 3.25424001e-02 -9.67793584e-01 5.23696065e-01
1.35478580e+00 -1.94888383e-01 -7.46686280e-01 1.43235281e-01
6.72475696e-01 -8.37907493e-01 -5.66048563e-01 5.61560154e-01
6.69128835e-01 -1.05466413e+00 1.18295801e+00 -1.81657046e-01
4.88411009e-01 -3.16706389e-01 -5.67371550e-04 -6.71682060e-01
-3.12349409e-01 -5.20976126e-01 -5.10134816e-01 1.22122645e+00
1.56318799e-01 -4.41135138e-01 6.95690632e-01 9.55044508e-01
-1.23334445e-01 -7.69019842e-01 -7.44857550e-01 -2.08382949e-01
-3.83006603e-01 -3.02346170e-01 2.86154777e-01 5.50248384e-01
-1.80908397e-01 2.99666315e-01 -7.80179322e-01 2.45658159e-01
8.89049053e-01 3.53266180e-01 7.96472013e-01 -1.07167995e+00
-4.75040376e-01 -1.75945476e-01 -1.53082348e-02 -1.30950642e+00
8.01624730e-02 -6.80062532e-01 1.99814290e-01 -1.55161297e+00
4.23150927e-01 1.12610951e-03 -2.64042497e-01 6.18461728e-01
-6.32513165e-01 4.29833055e-01 3.89053702e-01 4.22315359e-01
-8.04588437e-01 4.75357443e-01 1.59466219e+00 -3.48633975e-02
-2.64792383e-01 6.50516376e-02 -7.60744452e-01 1.01573992e+00
8.29315066e-01 -2.12621078e-01 -4.04168814e-01 -5.91185987e-01
-1.34609267e-01 5.39875627e-02 7.85436451e-01 -1.00056434e+00
7.90656805e-02 8.84108916e-02 9.06361699e-01 -8.61252010e-01
6.73145115e-01 -5.88871300e-01 -5.06541468e-02 3.45822424e-01
-2.35088989e-01 -6.40685111e-02 8.95926431e-02 7.21053302e-01
-4.55364352e-03 1.02246210e-01 5.72985232e-01 -4.43935990e-01
-7.36307025e-01 4.26259845e-01 -2.88532432e-02 3.33793372e-01
7.24626064e-01 -1.96101621e-01 -2.46917695e-01 -1.28424078e-01
-9.79560971e-01 3.96569908e-01 4.29085851e-01 4.09426928e-01
7.34986901e-01 -9.81690764e-01 -5.80372036e-01 3.59842271e-01
-1.07383570e-02 3.79368037e-01 5.50491869e-01 1.12280142e+00
-3.00058454e-01 5.85073791e-02 -1.84825212e-01 -4.55593407e-01
-1.17707515e+00 7.18458772e-01 1.49544418e-01 -5.11977002e-02
-1.12219369e+00 1.06958723e+00 7.22610176e-01 -1.68874249e-01
5.70212245e-01 -4.15708572e-01 -2.12940484e-01 2.43532620e-02
7.28569806e-01 3.37148577e-01 -5.29287100e-01 -7.20325232e-01
-2.02848732e-01 6.25811279e-01 -3.92909870e-02 -1.39399603e-01
1.05162680e+00 -6.31812334e-01 -1.64406538e-01 2.91116625e-01
5.54109216e-01 3.24017555e-02 -1.68378747e+00 -1.57583177e-01
-3.34202319e-01 -6.23303354e-01 -4.22278613e-01 -9.61230695e-01
-1.12940955e+00 8.78462255e-01 6.14113092e-01 -1.67116269e-01
1.22953665e+00 2.50498325e-01 9.05181646e-01 -1.95329025e-01
2.04641044e-01 -1.23892844e+00 1.80136457e-01 1.22858472e-02
1.07852328e+00 -1.54897738e+00 6.20771572e-02 -1.03673923e+00
-9.06367183e-01 5.97018600e-01 1.09972501e+00 3.92815936e-03
2.54051000e-01 2.57049371e-02 7.44164586e-02 -3.76728803e-01
-2.82982320e-01 -4.84101117e-01 5.92444479e-01 5.44381559e-01
2.42548019e-01 9.31985676e-02 -3.05186152e-01 9.74291801e-01
-1.95929095e-01 -2.56227225e-01 -1.51726906e-03 7.69622505e-01
-4.30722892e-01 -7.12065995e-01 -5.01976192e-01 3.15646797e-01
-5.50233126e-01 -3.59903276e-02 -4.69580442e-01 6.66838825e-01
5.71693003e-01 9.48146880e-01 -9.88919064e-02 -9.53714326e-02
2.29276076e-01 4.61871959e-02 5.09829819e-01 -5.85469842e-01
-2.59888172e-01 3.89724642e-01 6.19454719e-02 -9.98123288e-01
-4.35502142e-01 -3.89202148e-01 -1.17123652e+00 1.84490979e-01
-1.16083048e-01 -3.04727823e-01 2.72717297e-01 7.93521583e-01
2.79023767e-01 6.58378780e-01 -3.54157458e-03 -1.25473332e+00
8.46944097e-03 -1.07905996e+00 -5.53605735e-01 9.41716135e-01
2.14837462e-01 -8.21614683e-01 -1.45209119e-01 2.84893155e-01] | [8.473052024841309, -0.16174066066741943] |
38753511-06f4-447c-839a-741013e639a9 | mbrain-a-multi-channel-self-supervised | 2306.13102 | null | https://arxiv.org/abs/2306.13102v1 | https://arxiv.org/pdf/2306.13102v1.pdf | MBrain: A Multi-channel Self-Supervised Learning Framework for Brain Signals | Brain signals are important quantitative data for understanding physiological activities and diseases of human brain. Most existing studies pay attention to supervised learning methods, which, however, require high-cost clinical labels. In addition, the huge difference in the clinical patterns of brain signals measured by invasive (e.g., SEEG) and non-invasive (e.g., EEG) methods leads to the lack of a unified method. To handle the above issues, we propose to study the self-supervised learning (SSL) framework for brain signals that can be applied to pre-train either SEEG or EEG data. Intuitively, brain signals, generated by the firing of neurons, are transmitted among different connecting structures in human brain. Inspired by this, we propose MBrain to learn implicit spatial and temporal correlations between different channels (i.e., contacts of the electrode, corresponding to different brain areas) as the cornerstone for uniformly modeling different types of brain signals. Specifically, we represent the spatial correlation by a graph structure, which is built with proposed multi-channel CPC. We theoretically prove that optimizing the goal of multi-channel CPC can lead to a better predictive representation and apply the instantaneou-time-shift prediction task based on it. Then we capture the temporal correlation by designing the delayed-time-shift prediction task. Finally, replace-discriminative-learning task is proposed to preserve the characteristics of each channel. Extensive experiments of seizure detection on both EEG and SEEG large-scale real-world datasets demonstrate that our model outperforms several state-of-the-art time series SSL and unsupervised models, and has the ability to be deployed to clinical practice. | ['Yafeng Li', 'Teng Liu', 'Yang Yang', 'Junru Chen', 'Donghong Cai'] | 2023-06-15 | null | null | null | null | ['self-supervised-learning', 'seizure-detection'] | ['computer-vision', 'medical'] | [ 1.84332013e-01 -2.08052605e-01 -3.13542709e-02 -3.63242954e-01
-3.38183314e-01 -2.69726217e-01 2.00217709e-01 -7.42630139e-02
8.17383081e-03 6.59528911e-01 6.29852489e-02 -5.79635911e-02
-7.08627403e-01 -4.06525820e-01 -4.75657076e-01 -9.43612158e-01
-6.68266773e-01 7.96036050e-02 7.04522710e-03 7.49932509e-03
-4.49185483e-02 3.65469187e-01 -9.45837379e-01 1.89080745e-01
1.01566339e+00 1.15101027e+00 3.00496370e-01 7.80201778e-02
2.13615716e-01 4.73228455e-01 -5.94821453e-01 2.46918485e-01
-1.32300854e-01 -7.21094131e-01 -4.45050150e-01 -1.30268009e-02
-3.27124119e-01 2.12099791e-01 -6.36566818e-01 1.08429849e+00
6.52849376e-01 -1.60702571e-01 7.82269359e-01 -1.43882346e+00
-3.41545045e-01 6.70646966e-01 -5.72289228e-01 5.84562659e-01
1.22308113e-01 6.42824024e-02 6.01778388e-01 -5.55131435e-01
3.05980772e-01 6.68295324e-01 5.91537833e-01 4.94347781e-01
-1.32075369e+00 -8.94758165e-01 1.40468761e-01 5.89996815e-01
-1.36352611e+00 -2.21152857e-01 1.33797145e+00 -6.60322368e-01
4.27980185e-01 1.96208522e-01 9.99985635e-01 1.46600544e+00
6.46447480e-01 6.41604364e-01 1.57015419e+00 4.43189824e-03
2.58245319e-01 -1.99761063e-01 2.92667538e-01 3.67285281e-01
4.18097004e-02 1.68016255e-01 -7.04915106e-01 -1.86707422e-01
6.98725522e-01 2.96720654e-01 -8.27807963e-01 -1.86722338e-01
-1.33648145e+00 3.98131073e-01 4.01183307e-01 6.93301201e-01
-5.89305460e-01 -4.79530767e-02 3.20258945e-01 3.71710420e-01
4.71039057e-01 4.01464313e-01 -4.66904670e-01 -5.00453375e-02
-9.02808309e-01 -2.93872535e-01 5.15140831e-01 8.38711619e-01
4.84624892e-01 1.27152158e-02 -2.24629164e-01 7.53395200e-01
1.34736672e-01 4.32963163e-01 7.54065275e-01 -2.67410815e-01
2.63670146e-01 3.98676753e-01 -3.34045142e-01 -1.26335919e+00
-9.76390958e-01 -7.16438770e-01 -1.31780219e+00 -4.31110442e-01
1.52343303e-01 -3.18537652e-01 -5.75984895e-01 1.72751856e+00
-1.38632581e-01 9.57883060e-01 -2.18520537e-01 8.49871099e-01
5.11300623e-01 4.63627815e-01 1.30784512e-03 -7.51814246e-01
1.17115712e+00 -4.36698586e-01 -8.77375007e-01 -1.21053293e-01
5.05348384e-01 -1.09032713e-01 7.72139072e-01 4.01965231e-01
-5.99015296e-01 -3.04083347e-01 -8.73911619e-01 6.83083832e-01
-1.32345542e-01 -3.08766253e-02 6.27166092e-01 2.51243353e-01
-7.52269089e-01 6.26347959e-01 -1.14131916e+00 -1.53228164e-01
4.40890729e-01 3.46367508e-01 -3.29575896e-01 1.54227763e-01
-1.32686412e+00 6.70690477e-01 1.96997792e-01 4.25366759e-01
-8.12165558e-01 -7.68343925e-01 -3.96926790e-01 -1.87762678e-01
1.25241220e-01 -3.14774245e-01 4.00400907e-01 -8.63004446e-01
-1.27976429e+00 5.60450613e-01 1.21610546e-02 -2.95983851e-01
2.29441434e-01 9.33530480e-02 -7.67111659e-01 2.66842276e-01
9.79031920e-02 1.97021171e-01 8.83459926e-01 -1.04118562e+00
-6.80263713e-02 -6.17545068e-01 -4.95541841e-01 -1.03758350e-01
-5.25616705e-01 -1.37231469e-01 -1.62034914e-01 -6.78627312e-01
3.82309020e-01 -8.30654442e-01 -3.06691900e-02 -4.25392240e-01
-6.14036739e-01 -1.33243322e-01 6.54342949e-01 -8.05479705e-01
1.24363601e+00 -2.27980042e+00 2.85205960e-01 6.20845556e-01
4.40525562e-01 -1.64015159e-01 -1.81111902e-01 2.50060618e-01
-4.45978284e-01 -1.38761595e-01 -4.34462726e-01 1.45061374e-01
-3.80153239e-01 7.15180263e-02 -1.58771276e-01 7.41000891e-01
-1.71953086e-02 9.04572606e-01 -1.05592775e+00 -3.88393044e-01
1.36539917e-02 5.10349751e-01 -1.09535284e-01 1.97874323e-01
4.00241345e-01 1.14293814e+00 -7.16478348e-01 4.44576949e-01
4.72325414e-01 -3.15609396e-01 1.38384402e-01 -4.02945399e-01
6.69179186e-02 2.18862608e-01 -7.21727848e-01 1.82668829e+00
-3.19071859e-01 5.61985850e-01 -2.90821761e-01 -1.66728270e+00
9.43215132e-01 5.29419959e-01 1.06305313e+00 -8.46942604e-01
2.42011353e-01 3.82824719e-01 4.11812961e-01 -7.54437804e-01
-7.01567411e-01 -1.47001380e-02 7.68996105e-02 2.70733953e-01
1.42459735e-01 1.46610513e-01 -1.96046859e-01 -7.93745965e-02
1.30398560e+00 -2.26342887e-01 2.15341821e-01 -6.06992960e-01
5.50654352e-01 -4.35620844e-01 7.33793199e-01 4.77687895e-01
-2.35727787e-01 3.13607305e-01 6.58652544e-01 -1.54259115e-01
-4.44743246e-01 -9.18048441e-01 -4.14515138e-01 5.36283314e-01
3.35685372e-01 -1.39375716e-01 -7.57942080e-01 -3.68359625e-01
-1.16395310e-01 4.16644007e-01 -5.96463501e-01 -5.75732350e-01
-5.58700800e-01 -9.82962728e-01 3.49057436e-01 5.90937018e-01
3.63288283e-01 -1.04979694e+00 -6.37683153e-01 4.60100591e-01
-2.56427616e-01 -1.12919307e+00 -4.55282629e-01 4.78807777e-01
-1.06801653e+00 -1.00751472e+00 -7.51937807e-01 -8.26652229e-01
5.55196881e-01 -9.05770659e-02 7.57248282e-01 -1.77439779e-01
-3.71761769e-01 5.67250967e-01 -3.99260759e-01 -2.77108431e-01
1.98209643e-01 -4.65228483e-02 1.70088708e-01 5.22008717e-01
2.58368045e-01 -1.23808873e+00 -7.88865149e-01 3.78106385e-01
-6.19161367e-01 9.82082337e-02 7.03293741e-01 9.73399639e-01
7.66761422e-01 9.08287242e-02 1.10485876e+00 -6.84935331e-01
8.32101524e-01 -7.31057048e-01 -3.61139894e-01 3.81235361e-01
-6.17686093e-01 -3.88277275e-03 8.55501413e-01 -8.04966807e-01
-5.45824230e-01 -7.31702447e-02 3.27388197e-01 -5.33340156e-01
-1.10059641e-02 8.10369432e-01 -2.50297368e-01 -3.69871184e-02
3.93891096e-01 7.32154965e-01 -1.81717485e-01 -2.98065513e-01
3.51753198e-02 6.39430165e-01 5.89876235e-01 -5.95554233e-01
6.05778039e-01 3.31162244e-01 1.74920157e-01 -7.66540527e-01
-4.81300265e-01 -4.03279513e-01 -7.46288776e-01 -4.09137309e-01
8.43469083e-01 -7.04902112e-01 -6.21294022e-01 6.69169605e-01
-1.16792059e+00 -3.23310047e-01 7.85204098e-02 9.12251711e-01
-4.98680562e-01 2.00481862e-01 -5.77722669e-01 -7.47753263e-01
-3.33784521e-01 -8.73914897e-01 7.99497962e-01 -4.51901481e-02
4.82744165e-02 -1.00639915e+00 6.04109131e-02 -1.66812971e-01
2.50959635e-01 4.16118950e-01 1.05967772e+00 -5.94834507e-01
-3.08139294e-01 -8.25798362e-02 -9.27087013e-03 2.79195756e-01
2.53339082e-01 -6.11531496e-01 -8.20400298e-01 -2.88280815e-01
4.63863462e-01 3.80181484e-02 5.42324841e-01 5.71963727e-01
1.49481070e+00 -1.44525662e-01 -6.18127346e-01 7.84861147e-01
1.19278026e+00 6.60814285e-01 5.37878931e-01 -1.38521016e-01
6.25441670e-01 6.75029397e-01 5.28502017e-02 4.26536798e-01
1.82087913e-01 5.57655394e-01 7.09245130e-02 5.81519753e-02
1.95612341e-01 -5.77705279e-02 2.33895510e-01 1.58644664e+00
-3.40701640e-01 6.35732189e-02 -1.02789712e+00 4.35993373e-01
-1.86745930e+00 -7.45339811e-01 -1.88030332e-01 2.14349365e+00
7.45609760e-01 -6.95515573e-02 -7.51463771e-02 2.26770371e-01
7.77501285e-01 -2.48600058e-02 -7.45833218e-01 3.57275128e-01
-1.33560762e-01 3.75024199e-01 2.50989139e-01 -1.38090298e-01
-7.55900919e-01 3.50751460e-01 5.82874632e+00 6.14655614e-01
-1.58231628e+00 3.91101748e-01 6.20599747e-01 4.32808921e-02
-1.50499120e-01 -2.85455555e-01 -1.26645505e-01 9.57915008e-01
9.64129984e-01 -3.68847698e-01 7.56987512e-01 3.32924902e-01
5.68268359e-01 3.12912494e-01 -1.32033837e+00 1.59454417e+00
-2.74737384e-02 -8.53186548e-01 -2.96285212e-01 -4.92415726e-02
5.28310895e-01 -8.86473954e-02 -1.12365589e-01 6.18530959e-02
-5.94057322e-01 -1.04081404e+00 4.91322458e-01 9.42632318e-01
7.97405303e-01 -3.73708338e-01 7.06683040e-01 4.23603237e-01
-1.33400404e+00 -1.14611864e-01 -2.22942263e-01 1.78962275e-01
2.69612037e-02 9.14720774e-01 -1.70322239e-01 6.71519399e-01
5.97587228e-01 1.23452294e+00 -4.28355753e-01 1.22396779e+00
-2.81076193e-01 9.36295450e-01 -1.80982828e-01 -1.97473746e-02
-9.95031819e-02 -3.88480335e-01 4.42456335e-01 9.88095522e-01
5.16301513e-01 3.60835582e-01 6.32288083e-02 1.01142526e+00
1.77398294e-01 9.66931880e-02 -6.50074840e-01 -3.65959294e-02
5.96091032e-01 1.20917666e+00 -6.84867442e-01 -3.42348851e-02
-5.24170578e-01 8.83453906e-01 1.83730751e-01 6.39085472e-01
-8.62153471e-01 -4.26273495e-01 1.74205065e-01 9.15440246e-02
-3.63951057e-01 -1.78160414e-01 -3.79585892e-01 -1.35725224e+00
3.02595973e-01 -6.09306991e-01 1.28844425e-01 -7.87491739e-01
-1.63207901e+00 8.17236662e-01 1.56312183e-01 -1.54674113e+00
-4.56457548e-02 -4.71695811e-01 -9.05685961e-01 8.38969350e-01
-1.47938800e+00 -6.53886080e-01 -4.97708857e-01 1.23538876e+00
7.85437524e-02 -1.40074939e-01 6.76941335e-01 4.98333067e-01
-6.41996026e-01 3.98594767e-01 6.29563956e-03 1.80593327e-01
5.03397346e-01 -9.13457990e-01 -2.32975855e-01 6.36546373e-01
1.03239954e-01 5.75475812e-01 2.42996782e-01 -4.49748755e-01
-1.47517455e+00 -1.08874929e+00 4.12242293e-01 1.76754314e-02
9.38403845e-01 -5.48568070e-01 -1.00134695e+00 4.36119080e-01
5.35151213e-02 3.05732518e-01 5.88882744e-01 1.07581646e-03
-6.11704364e-02 -5.42922318e-01 -7.87528515e-01 2.37721384e-01
1.22242141e+00 -7.71648467e-01 -7.35224903e-01 4.36183989e-01
3.31182897e-01 8.49290192e-03 -9.64713752e-01 3.99358869e-01
4.58329588e-01 -5.39296329e-01 7.24348009e-01 -5.10014594e-01
1.84716389e-01 -5.76924644e-02 1.77688032e-01 -1.84812355e+00
-4.02919143e-01 -7.77284324e-01 -1.06731154e-01 9.45549488e-01
2.56777227e-01 -1.10006893e+00 4.14455950e-01 1.92964733e-01
-3.82433444e-01 -1.05367279e+00 -1.07348430e+00 -1.03590214e+00
6.05674349e-02 -5.56706965e-01 4.75830346e-01 1.00554919e+00
6.96879506e-01 3.85570079e-01 -4.39810932e-01 2.25505173e-01
6.66650355e-01 5.35585470e-02 -1.79106444e-02 -1.48948920e+00
-2.86966652e-01 -4.11933452e-01 -7.37169266e-01 -9.81370389e-01
3.18992019e-01 -1.08636725e+00 1.01225346e-01 -1.34634852e+00
2.66844988e-01 -4.74211305e-01 -8.94399464e-01 4.74594116e-01
-8.94380063e-02 -1.05991758e-01 -1.11380845e-01 4.29639846e-01
-2.67775416e-01 6.00244880e-01 1.36735809e+00 -3.38006139e-01
-3.39523405e-01 -1.68971360e-01 -3.51285279e-01 5.71229994e-01
7.50182331e-01 -5.00530779e-01 -7.20049977e-01 -1.79580942e-01
-6.24446087e-02 5.06374776e-01 3.91168088e-01 -1.25544095e+00
5.52816391e-01 5.07657602e-02 3.20533812e-01 -2.67831475e-01
1.09810941e-01 -8.80763412e-01 1.06668346e-01 4.00453985e-01
-2.47506887e-01 3.98260318e-02 -1.21558987e-01 7.09235489e-01
-3.80613804e-01 2.30198085e-01 6.21266246e-01 1.14328049e-01
-3.61111999e-01 6.93871021e-01 -3.30623716e-01 2.81684220e-01
1.01160347e+00 -1.61433537e-02 -2.78171360e-01 -2.28822663e-01
-8.03002119e-01 2.16978103e-01 -3.89549702e-01 2.50263810e-01
7.98093736e-01 -1.46982408e+00 -5.69303691e-01 4.13769633e-01
2.36615032e-01 -3.18662316e-01 3.75741899e-01 1.64254344e+00
1.97214782e-01 4.59106803e-01 -4.44524527e-01 -7.66911149e-01
-7.43903816e-01 4.46796060e-01 5.49614072e-01 -5.11671565e-02
-9.31825101e-01 4.65567648e-01 3.59976768e-01 6.09905757e-02
1.95117056e-01 -5.09182096e-01 -4.07559931e-01 2.02434454e-02
3.12355906e-01 2.01883569e-01 1.17659949e-01 -5.22894025e-01
-6.99526191e-01 7.10551262e-01 3.03063750e-01 6.40153065e-02
1.49043906e+00 1.12208962e-01 -2.33842939e-01 7.03825057e-01
1.34053826e+00 -3.47217739e-01 -1.06266284e+00 -1.16493158e-01
5.96106239e-02 -5.53677939e-02 1.13725059e-01 -7.69760668e-01
-1.56800640e+00 1.03307319e+00 8.84203553e-01 2.41298363e-01
1.45726514e+00 -5.70383156e-03 7.42520511e-01 3.39799315e-01
9.59546804e-01 -8.94114435e-01 7.97246620e-02 1.13232061e-01
8.64771426e-01 -7.83048749e-01 -3.36478651e-01 -3.72886240e-01
-4.84817952e-01 1.21207058e+00 3.77057254e-01 -2.28219822e-01
1.18217957e+00 3.10073048e-01 -1.59614116e-01 -4.75358665e-01
-6.88593090e-01 1.52402401e-01 4.06533539e-01 7.70208895e-01
3.70466381e-01 4.25483257e-01 -5.55670023e-01 1.09548473e+00
-1.90521702e-02 2.12264836e-01 2.25406900e-01 5.99585533e-01
8.37213621e-02 -7.82801747e-01 -1.51130244e-01 8.54862094e-01
-1.41085744e-01 -7.96047226e-02 -9.27491635e-02 5.48307836e-01
2.07159251e-01 9.68048632e-01 -1.19761817e-01 -7.83368647e-01
4.02785540e-01 9.49841887e-02 5.54154098e-01 -4.68355417e-01
-2.50136256e-01 2.35560030e-01 -4.77755815e-01 -6.34130836e-01
-6.24375463e-01 -5.91616452e-01 -1.25593662e+00 2.77304739e-01
-2.76178896e-01 1.99359268e-01 2.93396413e-01 1.10206699e+00
4.63183224e-01 7.46745825e-01 8.72975767e-01 -6.40365720e-01
-3.52519900e-01 -1.12097406e+00 -1.21112192e+00 5.89280903e-01
2.40608931e-01 -9.46127594e-01 -4.17486221e-01 6.32019192e-02] | [13.139824867248535, 3.5001022815704346] |
e3d15fc8-675b-460c-a2cc-9685dde1daf0 | graph-based-multi-robot-path-finding-and | 2206.11319 | null | https://arxiv.org/abs/2206.11319v1 | https://arxiv.org/pdf/2206.11319v1.pdf | Graph-Based Multi-Robot Path Finding and Planning | Purpose of Review Planning collision-free paths for multiple robots is important for real-world multi-robot systems and has been studied as an optimization problem on graphs, called Multi-Agent Path Finding (MAPF). This review surveys different categories of classic and state-of-the-art MAPF algorithms and different research attempts to tackle the challenges of generalizing MAPF techniques to real-world scenarios. Recent Findings Solving MAPF problems optimally is computationally challenging. Recent advances have resulted in MAPF algorithms that can compute collision-free paths for hundreds of robots and thousands of navigation tasks in seconds of runtime. Many variants of MAPF have been formalized to adapt MAPF techniques to different real-world requirements, such as considerations of robot kinematics, online optimization for real-time systems, and the integration of task assignment and path planning. Summary Algorithmic techniques for MAPF problems have addressed important aspects of several multi-robot applications, including automated warehouse fulfillment and sortation, automated train scheduling, and navigation of non-holonomic robots and quadcopters. This showcases their potential for real-world applications of large-scale multi-robot systems. | ['Hang Ma'] | 2022-06-22 | null | null | null | null | ['multi-agent-path-finding'] | ['playing-games'] | [-1.00473829e-01 9.75273773e-02 6.09445888e-05 -3.14108968e-01
-1.94307223e-01 -9.11219716e-01 2.32681017e-02 5.64595520e-01
-5.03205121e-01 9.47842598e-01 -7.64840126e-01 -4.81263191e-01
-1.07981098e+00 -8.58339131e-01 -6.79169357e-01 -4.31389719e-01
-7.86988258e-01 1.48549342e+00 3.69603544e-01 -1.09232426e+00
5.39125741e-01 9.01403606e-01 -1.19100392e+00 -3.13355684e-01
6.16975844e-01 2.96782285e-01 1.02668083e+00 1.00505817e+00
7.46170133e-02 2.48440206e-02 -6.04382455e-01 1.84483901e-01
4.87487137e-01 1.36386380e-01 -1.12818360e+00 7.30549172e-02
-4.01904523e-01 2.31668189e-01 -1.25721528e-03 6.47688031e-01
1.60133302e-01 6.31443024e-01 6.36412144e-01 -2.32762289e+00
1.64133348e-02 4.29731220e-01 -5.86287796e-01 -2.57464740e-02
3.62737954e-01 -1.88308582e-01 4.31657195e-01 -2.75625795e-01
6.05413675e-01 1.43044913e+00 5.29309809e-01 -5.97580485e-02
-7.67486155e-01 -3.22371051e-02 3.60892862e-02 4.94531184e-01
-1.32325244e+00 1.88532710e-01 -7.81210661e-02 -1.24262609e-01
1.45660937e+00 7.17814192e-02 3.06333870e-01 -1.23751044e-01
1.26555204e+00 3.04042011e-01 6.15110040e-01 -3.37095648e-01
1.46784469e-01 -1.23183332e-01 2.00875048e-02 6.86352313e-01
6.67353988e-01 -2.27587864e-01 7.96495676e-02 -3.69931087e-02
7.23373830e-01 -1.68702528e-01 9.75514725e-02 -8.02695751e-01
-1.58729100e+00 9.80812848e-01 2.82094836e-01 1.19985968e-01
-5.81025243e-01 3.15691859e-01 3.26403528e-01 4.59734887e-01
-5.69049776e-01 9.45347369e-01 -6.87608600e-01 -1.49401680e-01
2.20178992e-01 8.20727646e-01 1.21860361e+00 1.80360746e+00
1.12125158e+00 -3.45209241e-01 6.31632864e-01 4.92384672e-01
2.40143523e-01 4.68453705e-01 -1.99804038e-01 -1.13831830e+00
8.56749535e-01 3.05389941e-01 5.66893458e-01 -1.03562951e+00
-1.33706462e+00 2.76962519e-01 -4.54250902e-01 5.00344515e-01
3.87392789e-01 -3.44092160e-01 -4.51637506e-01 9.31054831e-01
6.48164213e-01 -6.85619533e-01 3.29530060e-01 9.60094094e-01
-5.15805930e-02 9.02170658e-01 -4.33690518e-01 -2.13834181e-01
1.39011669e+00 -1.54465628e+00 -8.22590113e-01 -4.39654738e-01
8.47372353e-01 -7.99535930e-01 2.12028936e-01 3.97206694e-01
-1.09309566e+00 -3.15111458e-01 -1.20098352e+00 1.34398535e-01
-7.05873489e-01 -2.88485825e-01 6.02014840e-01 3.74219835e-01
-1.22469890e+00 5.51096022e-01 -7.71642804e-01 -6.11172616e-01
-3.98448467e-01 8.28617573e-01 -5.37953973e-01 -6.64127767e-01
-8.83667529e-01 1.67652309e+00 4.94075030e-01 3.17828894e-01
-5.34742057e-01 -1.54897854e-01 -9.02796745e-01 -2.54942656e-01
9.14876521e-01 -5.28418481e-01 1.41769636e+00 1.97874382e-01
-1.46093547e+00 4.29398492e-02 9.99674201e-02 4.36789766e-02
1.47133768e-01 9.45806280e-02 -3.76963854e-01 8.32468085e-03
6.53276742e-01 2.04121217e-01 1.30321950e-01 -9.92122352e-01
-1.16042817e+00 -1.52967244e-01 1.44873947e-01 7.11527288e-01
5.26407599e-01 -2.20932662e-01 -5.12044914e-02 3.45909238e-01
1.14168078e-01 -1.45707607e+00 -1.10457289e+00 -4.29432392e-01
-3.22003096e-01 -3.56457025e-01 7.49438107e-01 1.50648639e-01
4.58810747e-01 -1.71661031e+00 4.52105492e-01 4.09665555e-01
-2.68037289e-01 -4.48434353e-01 -3.38811457e-01 1.13748145e+00
3.05882961e-01 -3.52972388e-01 9.20114145e-02 1.85168669e-01
2.82809854e-01 8.68591428e-01 2.09867910e-01 8.31653655e-01
-1.08356453e-01 5.76917052e-01 -1.17946994e+00 -3.20894659e-01
4.08242583e-01 -3.36984754e-01 -1.08843893e-01 -2.84852535e-01
-5.59505709e-02 2.27653757e-01 -6.51391983e-01 4.98986542e-01
7.76922107e-01 1.89776868e-01 4.04580444e-01 3.66593003e-01
-7.72661626e-01 -2.08892733e-01 -1.64530814e+00 1.84640515e+00
-5.64132035e-01 4.72934186e-01 4.98560518e-01 -8.94649506e-01
8.48062336e-01 -1.31151140e-01 8.26788306e-01 -5.82969010e-01
3.49627174e-02 5.14686167e-01 6.83382154e-02 -2.90989101e-01
1.48942721e+00 1.56907409e-01 -6.13635123e-01 6.54909611e-01
-2.06327662e-01 -7.82150447e-01 7.96207190e-01 -7.35832676e-02
1.31150520e+00 -6.77098557e-02 4.10899848e-01 -7.91192234e-01
2.58501500e-01 1.02108717e+00 2.96004325e-01 7.56604075e-01
-3.23635221e-01 -9.43110362e-02 7.68479481e-02 -5.44838846e-01
-7.74875104e-01 -7.19247699e-01 2.66586840e-01 1.14425683e+00
1.32686853e+00 -2.21929982e-01 -1.99806884e-01 -1.40516296e-01
3.61462295e-01 6.38561010e-01 -2.17078105e-01 3.13459665e-01
-1.16442680e+00 -9.00160074e-01 -1.10082716e-01 -1.71321502e-03
-1.16577849e-01 -8.45820189e-01 -1.33549631e+00 9.60327566e-01
-1.79756522e-01 -1.44972503e+00 -3.47308457e-01 5.94004929e-01
-4.48013246e-01 -1.61005974e+00 -4.53505158e-01 -1.21363068e+00
8.74010444e-01 1.11873686e+00 7.31270373e-01 -1.68347612e-01
-5.62622786e-01 5.55611074e-01 -4.74653631e-01 -7.07815766e-01
-5.23014486e-01 2.17254803e-01 3.54017735e-01 -8.77376795e-01
-1.88534185e-01 -2.04943806e-01 -1.71775460e-01 1.13766515e+00
-4.09106106e-01 -2.00805366e-01 8.32030058e-01 2.22642809e-01
7.46367633e-01 9.21150625e-01 7.42545545e-01 -3.52618128e-01
8.83993030e-01 -6.10542476e-01 -7.90167511e-01 4.25920516e-01
-6.73714638e-01 -1.76039666e-01 6.14764214e-01 -9.94561538e-02
-5.44563055e-01 2.02416047e-01 5.10734558e-01 2.58962125e-01
-9.39240754e-02 5.09420872e-01 2.63008684e-01 -6.14706337e-01
2.74137586e-01 -2.66585261e-01 2.99771667e-01 1.43467754e-01
4.31692332e-01 2.23122507e-01 3.71745348e-01 -5.59830308e-01
5.77107728e-01 2.02355593e-01 1.02569020e+00 -6.90105379e-01
2.11407810e-01 -7.19748199e-01 -8.32540512e-01 -3.65049064e-01
7.33540297e-01 -3.14523607e-01 -1.09899724e+00 2.50930816e-01
-1.32162011e+00 -6.58610702e-01 8.30141306e-02 3.09886456e-01
-1.26102328e+00 1.79006144e-01 -3.36639404e-01 -7.22252548e-01
5.87857813e-02 -1.45816290e+00 8.42002332e-01 2.51870722e-01
-6.35907575e-02 -9.36651587e-01 2.58352011e-01 -2.32520550e-02
4.77324724e-01 4.49684173e-01 7.56895363e-01 -3.69501919e-01
-7.60751069e-01 -1.55309081e-01 9.93770584e-02 -8.88992429e-01
3.56923282e-01 -2.79213428e-01 1.89516485e-01 -4.58802074e-01
-3.57761800e-01 1.19194381e-01 -3.09527516e-01 5.07381737e-01
9.29104909e-02 -1.93260461e-01 -1.06149161e+00 -1.49204046e-01
1.69509470e+00 6.80324674e-01 2.19680324e-01 1.08595145e+00
2.69175172e-01 1.03395915e+00 1.78081894e+00 3.20154965e-01
1.04193473e+00 9.17049766e-01 7.84350097e-01 2.96633691e-01
6.72695696e-01 4.79972482e-01 1.37111589e-01 7.94903755e-01
-9.08918455e-02 -5.53794980e-01 -1.28347301e+00 6.77609444e-01
-2.37874460e+00 -3.68403047e-01 -7.90819824e-01 1.63316154e+00
-1.54446989e-01 -2.52322257e-02 2.10229740e-01 -7.51136094e-02
9.34245050e-01 -5.87569654e-01 -3.12067777e-01 -8.44140649e-01
2.12502614e-01 -3.91662896e-01 1.35743296e+00 5.91427267e-01
-1.07840359e+00 6.03447437e-01 6.23295021e+00 2.80804336e-01
-3.73908073e-01 3.70284617e-02 -2.05430701e-01 2.17435062e-01
3.07247788e-01 -7.38212559e-03 -7.96628058e-01 -6.93793669e-02
1.07638443e+00 -5.40796101e-01 7.28514194e-01 8.85696769e-01
1.45533144e-01 -6.15122557e-01 -9.11484003e-01 7.39265919e-01
-2.66851097e-01 -9.94666815e-01 -5.32396615e-01 1.81792453e-01
7.38269866e-01 1.60064369e-01 -3.25424284e-01 2.87610859e-01
6.70331776e-01 -6.48584723e-01 8.41828167e-01 6.16468750e-02
8.47299322e-02 -1.09623301e+00 9.03439403e-01 5.69505572e-01
-1.64398193e+00 -3.82849693e-01 -7.62992203e-01 -3.57720584e-01
1.20089972e+00 3.07127327e-01 -1.51476610e+00 1.41429269e+00
7.31560707e-01 2.27291018e-01 1.88626543e-01 1.35817575e+00
3.05485427e-01 -1.02805710e+00 -5.81903577e-01 -6.51184499e-01
6.58119500e-01 -6.00250185e-01 5.39801300e-01 9.47716534e-01
4.71442342e-01 1.13449872e-01 7.42513835e-01 1.70425981e-01
9.96940255e-01 -3.24518271e-02 -8.42694104e-01 2.44785011e-01
4.32345569e-01 1.47218573e+00 -1.40830326e+00 2.66380668e-01
-1.50914624e-01 5.83483398e-01 1.95747409e-02 1.58052757e-01
-9.28975821e-01 -1.12362313e+00 7.91386962e-01 -1.18703164e-01
1.84943434e-02 -9.83419120e-01 -2.29636952e-01 6.80554658e-02
-2.17083082e-01 -1.84245095e-01 2.80344754e-01 -7.86558628e-01
-6.65305078e-01 4.26983923e-01 5.18668056e-01 -1.25676930e+00
-6.09001577e-01 -8.72055650e-01 -3.66487503e-01 5.67015111e-01
-1.83469820e+00 -7.93025136e-01 -2.65804797e-01 3.46340567e-01
8.17183137e-01 -1.27621561e-01 8.31710100e-01 6.24590330e-02
-2.83393145e-01 -2.77661532e-01 3.61781865e-01 -8.10390770e-01
5.33796191e-01 -1.21430159e+00 5.65776229e-01 6.09183311e-01
-8.29544306e-01 2.14861751e-01 1.03242505e+00 -7.86615014e-01
-2.42543364e+00 -1.02307212e+00 3.88732344e-01 -3.27689499e-01
7.98393071e-01 -2.42826641e-02 -2.48741478e-01 8.45413804e-01
1.39788449e-01 -4.67589945e-01 1.49489641e-01 -1.61954224e-01
8.26901138e-01 5.76060116e-02 -1.06154263e+00 6.03986084e-01
8.95607114e-01 5.98813355e-01 -4.54450138e-02 8.08557272e-01
6.65231824e-01 -7.95113325e-01 -8.56290936e-01 2.31389344e-01
3.27989936e-01 -2.82521427e-01 7.78590500e-01 -2.16536224e-01
-2.84481615e-01 -5.40938318e-01 -8.04729909e-02 -1.86453807e+00
-7.19447792e-01 -9.19583380e-01 5.51707625e-01 5.33372939e-01
5.36391914e-01 -1.03479743e+00 4.02193278e-01 5.02424598e-01
-9.76314843e-01 -5.67513466e-01 -1.15921307e+00 -1.04851711e+00
-1.66377708e-01 -8.58279690e-03 6.40126109e-01 7.24926889e-01
5.32550454e-01 1.80439800e-01 -4.30847444e-02 7.21492767e-01
5.74968219e-01 2.64734894e-01 1.01930726e+00 -1.11917818e+00
1.03809148e-01 -2.89569438e-01 -4.00985420e-01 -9.07684743e-01
6.10249043e-02 -5.70496202e-01 7.71014035e-01 -2.37551427e+00
-6.64233863e-01 -1.20391810e+00 4.09226775e-01 4.80377257e-01
4.24814373e-01 -4.13904101e-01 4.16279882e-01 2.14653641e-01
-8.79574060e-01 6.91018552e-02 1.60535753e+00 -8.04877728e-02
-3.62035781e-01 3.64127636e-01 -2.31707022e-01 3.32999498e-01
1.04549050e+00 -5.81532121e-01 -5.94810367e-01 -6.56203210e-01
6.38334334e-01 6.41841054e-01 -3.45743418e-01 -7.85123348e-01
8.41559231e-01 -1.03878200e+00 -5.26616454e-01 -7.52800405e-01
5.97585559e-01 -1.15823138e+00 4.44735080e-01 8.28665137e-01
5.01684308e-01 1.04347253e+00 4.10821617e-01 6.89864039e-01
4.79553528e-02 -6.75662339e-01 2.88679242e-01 -4.22256291e-01
-1.28390205e+00 -1.78845469e-02 -1.01615381e+00 -5.04182696e-01
2.02366018e+00 -3.30011934e-01 -8.88056219e-01 -1.08145230e-01
-4.51813340e-01 1.16345060e+00 1.88028246e-01 4.91179556e-01
5.66290855e-01 -8.36526573e-01 -5.62451720e-01 -1.53900892e-01
-9.15941820e-02 4.07854855e-01 2.91099727e-01 9.87076640e-01
-1.19900286e+00 8.29051971e-01 -8.48643124e-01 -5.69759727e-01
-9.55157101e-01 7.07886398e-01 -4.02156860e-02 -3.09292644e-01
-4.57353652e-01 3.58635187e-01 -8.56350660e-02 -7.03660846e-01
-2.74552673e-01 -4.53203529e-01 -7.18375519e-02 -2.73664504e-01
1.27973825e-01 1.03798831e+00 5.61451577e-02 -3.66223037e-01
-7.49830544e-01 7.41323590e-01 1.63351506e-01 -2.04191342e-01
1.39626074e+00 -6.79540992e-01 -3.06371212e-01 3.41474324e-01
5.01690865e-01 -5.29448867e-01 -5.34857690e-01 4.95989412e-01
9.55692157e-02 -2.04621226e-01 -6.91210210e-01 -5.96421778e-01
-3.74429762e-01 2.82148957e-01 9.37970206e-02 5.54669797e-01
7.07242370e-01 -1.23734675e-01 4.96665537e-01 1.02964008e+00
1.58753204e+00 -1.19274056e+00 9.84603018e-02 9.90481019e-01
8.43469024e-01 -8.46229732e-01 1.10105574e-01 -8.96868706e-01
-7.36162066e-01 1.46927953e+00 8.19764555e-01 -9.99448150e-02
4.91859168e-01 4.58366573e-01 3.28711420e-02 -2.82118499e-01
-5.04834950e-01 3.64434496e-02 -5.91102302e-01 8.41333032e-01
-5.59332788e-01 3.08281034e-01 -3.44954312e-01 9.76555124e-02
-4.08771068e-01 -5.20585716e-01 1.26004255e+00 1.72274244e+00
-8.34138393e-01 -1.05645216e+00 -6.78451777e-01 2.23446652e-01
2.74407417e-01 8.18255007e-01 2.19771370e-01 1.36343145e+00
-1.34998664e-01 1.19082606e+00 2.33991608e-01 -1.13619767e-01
8.53291035e-01 -5.74559152e-01 6.70666516e-01 -6.17145598e-01
-3.01784903e-01 -2.64591902e-01 6.03640199e-01 -4.89835918e-01
-2.86362976e-01 -7.25363851e-01 -1.96397376e+00 -2.83879369e-01
-4.72418338e-01 4.78986144e-01 1.39771748e+00 9.09654379e-01
4.21540886e-01 7.26553798e-01 4.91787910e-01 -1.38995314e+00
-3.49500120e-01 -5.10045171e-01 -7.90380180e-01 -4.44393992e-01
8.37959349e-02 -1.02072513e+00 3.07059616e-01 -5.92082918e-01] | [4.950806617736816, 1.6422604322433472] |
dd5afcb3-d8f8-45f3-891f-ec01432522ea | bi-mix-bidirectional-mixing-for-domain | 2111.10339 | null | https://arxiv.org/abs/2111.10339v1 | https://arxiv.org/pdf/2111.10339v1.pdf | Bi-Mix: Bidirectional Mixing for Domain Adaptive Nighttime Semantic Segmentation | In autonomous driving, learning a segmentation model that can adapt to various environmental conditions is crucial. In particular, copying with severe illumination changes is an impelling need, as models trained on daylight data will perform poorly at nighttime. In this paper, we study the problem of Domain Adaptive Nighttime Semantic Segmentation (DANSS), which aims to learn a discriminative nighttime model with a labeled daytime dataset and an unlabeled dataset, including coarsely aligned day-night image pairs. To this end, we propose a novel Bidirectional Mixing (Bi-Mix) framework for DANSS, which can contribute to both image translation and segmentation adaptation processes. Specifically, in the image translation stage, Bi-Mix leverages the knowledge of day-night image pairs to improve the quality of nighttime image relighting. On the other hand, in the segmentation adaptation stage, Bi-Mix effectively bridges the distribution gap between day and night domains for adapting the model to the night domain. In both processes, Bi-Mix simply operates by mixing two samples without extra hyper-parameters, thus it is easy to implement. Extensive experiments on Dark Zurich and Nighttime Driving datasets demonstrate the advantage of the proposed Bi-Mix and show that our approach obtains state-of-the-art performance in DANSS. Our code is available at https://github.com/ygjwd12345/BiMix. | ['Elisa Ricci', 'Nicu Sebe', 'Mingli Ding', 'Hao Tang', 'Zhun Zhong', 'Guanglei Yang'] | 2021-11-19 | null | null | null | null | ['image-relighting'] | ['computer-vision'] | [ 5.91168217e-02 -2.00595856e-01 -3.89149822e-02 -8.13769639e-01
-7.88620114e-01 -6.87918842e-01 5.50894976e-01 -4.42528635e-01
-3.57379943e-01 5.26217639e-01 -1.04373932e-01 -4.21881944e-01
2.68330127e-01 -6.15951538e-01 -7.92998850e-01 -9.80661273e-01
7.87786484e-01 4.68655318e-01 3.12485576e-01 -3.26384217e-01
-8.99000615e-02 9.42082405e-02 -1.43389285e+00 -1.41777605e-01
1.28209329e+00 7.85073221e-01 4.97525483e-01 5.98070383e-01
-2.86409035e-02 3.18023950e-01 -1.78394377e-01 -3.57713342e-01
6.36588752e-01 -5.42025805e-01 -4.14899737e-01 2.56417602e-01
3.70351493e-01 -1.77727014e-01 -4.22754109e-01 1.05337048e+00
5.13677835e-01 3.05514038e-01 3.31342638e-01 -1.47791994e+00
-3.18634897e-01 4.79120538e-02 -6.20229185e-01 1.32919580e-01
-4.31626499e-01 5.45616508e-01 5.19229650e-01 -6.65880919e-01
2.59478837e-01 9.63133812e-01 3.68096441e-01 5.65632463e-01
-1.18586862e+00 -9.75925982e-01 3.03155988e-01 4.27194625e-01
-1.27475154e+00 -6.43714547e-01 8.75499189e-01 -3.68013382e-01
3.35081875e-01 1.14549056e-01 6.10859454e-01 1.01320112e+00
1.43453315e-01 8.11140835e-01 1.38808596e+00 -1.54793158e-01
3.37325126e-01 1.86039209e-01 -1.30446792e-01 3.37246239e-01
5.05743362e-02 1.27640769e-01 -5.28476536e-01 4.07365799e-01
2.84385890e-01 -1.57722067e-02 -1.80245847e-01 -3.12037289e-01
-9.93824542e-01 6.98059261e-01 3.55413407e-01 -3.92484777e-02
-4.50339280e-02 -6.31634379e-03 1.94594905e-01 1.84038743e-01
7.54721045e-01 -1.92738771e-02 -5.33215582e-01 -1.18570179e-01
-9.69498634e-01 1.99649796e-01 4.45810974e-01 1.08018219e+00
1.23863363e+00 -1.19971618e-01 -1.55068651e-01 9.70072865e-01
2.71149009e-01 8.50702643e-01 3.84851813e-01 -1.12776458e+00
5.11895597e-01 2.72657007e-01 1.47728235e-01 -4.74280655e-01
-4.65488404e-01 -4.22696769e-01 -6.41120315e-01 5.82776181e-02
4.63438600e-01 -2.54867285e-01 -1.19593549e+00 1.85450518e+00
6.49317205e-01 4.87472415e-01 1.82881206e-01 1.02721035e+00
4.80534881e-01 6.72071815e-01 5.27959019e-02 7.13778734e-02
1.16324723e+00 -1.35045624e+00 -6.50565505e-01 -9.53734457e-01
4.35567260e-01 -8.72404933e-01 1.28710961e+00 -7.90747106e-02
-7.75746524e-01 -7.49386370e-01 -9.66260970e-01 -3.28708202e-01
-3.97303909e-01 5.55868298e-02 3.25589836e-01 4.44919437e-01
-8.54071438e-01 1.50853857e-01 -9.44152534e-01 -4.00537282e-01
3.24553967e-01 4.51867804e-02 -7.27323592e-02 -3.81425947e-01
-1.16673338e+00 6.73091471e-01 3.77272755e-01 2.05377668e-01
-8.83215487e-01 -6.94757402e-01 -9.00989413e-01 -2.76508480e-01
4.06758577e-01 -7.12808490e-01 1.44900692e+00 -1.02846706e+00
-1.61179435e+00 9.34497893e-01 -4.84416455e-01 -5.47219574e-01
7.44482815e-01 -8.24364573e-02 -4.75873470e-01 9.14187331e-05
3.94515008e-01 9.44310486e-01 9.66143548e-01 -1.38100457e+00
-8.23531568e-01 -4.30124551e-01 4.43940386e-02 5.33686757e-01
-2.28363276e-02 -6.53018057e-01 -9.48083103e-01 -5.78027189e-01
4.05734405e-02 -1.39823437e+00 -1.89440742e-01 -7.86435828e-02
-2.93316215e-01 1.01905815e-01 1.03963351e+00 -6.68882549e-01
8.37173343e-01 -2.55594325e+00 1.66688766e-02 -6.06628656e-02
-1.26148134e-01 2.30361775e-01 -1.67018533e-01 1.17387243e-01
2.72878170e-01 -3.98087025e-01 -8.15367639e-01 -7.46125579e-01
2.16267958e-01 6.84880555e-01 -3.79009992e-01 5.00272930e-01
1.62598416e-01 1.00304449e+00 -9.03198123e-01 -4.26714361e-01
4.34578061e-01 4.96106029e-01 -4.14468765e-01 3.42066705e-01
-2.70649374e-01 9.28312778e-01 -2.71589249e-01 5.06031871e-01
9.81611013e-01 -1.83081115e-03 -9.10043903e-03 -2.01206673e-02
-1.35748625e-01 4.68174070e-02 -7.41508543e-01 1.92631841e+00
-6.58429325e-01 9.27381277e-01 4.45394106e-02 -8.35378647e-01
8.61720324e-01 -7.57015720e-02 9.36722159e-02 -1.38116539e+00
5.48603572e-02 3.68452668e-01 -2.17628524e-01 -5.54307461e-01
5.17763078e-01 -2.82532185e-01 -8.02026615e-02 1.92415491e-01
-1.97511509e-01 -5.51778078e-01 1.66573465e-01 -7.13459700e-02
4.51314360e-01 4.17686790e-01 -1.60995439e-01 -1.64897159e-01
4.74578530e-01 9.40345228e-03 1.00887108e+00 4.86217678e-01
-5.06861269e-01 8.21316779e-01 1.37146056e-01 -2.67290026e-01
-1.04866326e+00 -1.07413936e+00 -5.76477721e-02 1.02168989e+00
6.57965958e-01 1.11966506e-02 -1.08140588e+00 -6.50045514e-01
-1.07057303e-01 1.03013277e+00 -4.37966675e-01 -4.36191320e-01
-5.87089956e-01 -6.90584004e-01 2.97002792e-01 4.17549193e-01
1.01278985e+00 -7.11185038e-01 -5.57055235e-01 3.65376733e-02
-5.41795433e-01 -1.42794144e+00 -7.55127788e-01 1.49139166e-01
-6.60528719e-01 -9.07831192e-01 -6.88055575e-01 -7.64927983e-01
6.51450038e-01 7.07288563e-01 9.33403909e-01 -2.47285768e-01
-1.70637488e-01 1.04459427e-01 -1.89718202e-01 -3.86380255e-01
-3.29169154e-01 3.35459977e-01 -1.61711991e-01 3.74362290e-01
5.18316090e-01 -5.30035436e-01 -8.83708358e-01 6.17073476e-01
-1.09216547e+00 4.52992976e-01 5.32093704e-01 5.78086615e-01
8.01660299e-01 -3.46301217e-03 3.43853742e-01 -9.30588424e-01
-3.87828648e-02 -5.15153408e-01 -8.57839584e-01 3.55064683e-02
-6.89119339e-01 -8.99530351e-02 6.44958794e-01 -2.06409141e-01
-1.35404658e+00 2.25851685e-01 -1.78851128e-01 -4.21700537e-01
-2.24419147e-01 6.73389435e-03 -3.26408535e-01 1.48080930e-03
4.25014853e-01 4.30948287e-01 -6.43613413e-02 -3.95192295e-01
4.92810190e-01 8.72392535e-01 7.61982620e-01 -5.13393283e-01
1.06464505e+00 8.98819327e-01 -1.58692226e-01 -7.53648877e-01
-1.10205698e+00 -6.46640837e-01 -5.56675196e-01 -1.43519089e-01
9.92130041e-01 -1.25008762e+00 -6.37644604e-02 7.03428864e-01
-6.60207152e-01 -1.11135542e+00 -3.51744920e-01 3.60640317e-01
-7.21516550e-01 3.43566164e-02 -1.44329801e-01 -5.59960961e-01
5.95990568e-02 -1.19905150e+00 1.19446814e+00 7.72232890e-01
2.88783252e-01 -1.04314840e+00 -7.33967945e-02 7.34591246e-01
3.14971656e-01 7.67913535e-02 5.95103979e-01 -1.71675906e-01
-5.90244174e-01 1.38823345e-01 -5.01567841e-01 5.22278130e-01
2.69819319e-01 -3.04008454e-01 -1.33577943e+00 -2.96399266e-01
6.39243200e-02 -9.57297608e-02 1.03410745e+00 4.02582943e-01
1.07313967e+00 9.61635560e-02 -1.08910307e-01 1.02953863e+00
1.17886996e+00 2.49046385e-01 6.67491317e-01 6.34981275e-01
9.09793735e-01 7.32497633e-01 1.11115575e+00 2.21995339e-01
1.03883183e+00 7.44480133e-01 3.31976920e-01 -4.09338832e-01
-4.55343485e-01 -2.05279246e-01 3.48799646e-01 6.30862594e-01
4.08228934e-01 -8.98805559e-02 -8.75651658e-01 7.21440494e-01
-1.99541676e+00 -5.48099518e-01 -2.18649115e-02 2.25733876e+00
9.61262941e-01 1.38126090e-01 2.13729143e-02 -1.98467851e-01
5.99831462e-01 3.90236467e-01 -9.07777429e-01 -2.18180954e-01
-2.53289700e-01 -9.30448472e-02 9.37086582e-01 6.48914337e-01
-1.21866298e+00 1.16693437e+00 4.87394142e+00 8.62773895e-01
-1.24077296e+00 4.55445737e-01 7.84836233e-01 -7.43638948e-02
-3.32957715e-01 2.10842058e-01 -8.09849322e-01 8.06650043e-01
1.01517272e+00 -1.52513593e-01 7.30752707e-01 5.60948431e-01
5.79102397e-01 -1.26576737e-01 -6.99336946e-01 9.26185548e-01
3.17172259e-02 -8.16534936e-01 -4.05281723e-01 1.37268044e-02
1.00230920e+00 3.17492962e-01 2.46237412e-01 2.92652905e-01
1.18520558e-01 -4.79883075e-01 9.06365275e-01 4.08611059e-01
6.57762349e-01 -6.97821200e-01 4.56527710e-01 2.76747733e-01
-1.13447404e+00 -6.63863495e-02 -2.24301159e-01 2.55126268e-01
2.36988008e-01 6.68574095e-01 -6.24694407e-01 5.60216546e-01
7.47975767e-01 8.78787518e-01 -7.20237255e-01 9.61259782e-01
-6.05109274e-01 6.82749867e-01 -2.36776695e-01 6.62754238e-01
3.90445441e-01 -6.06464803e-01 3.79759669e-01 1.25763285e+00
2.26622939e-01 -9.96128470e-03 1.92916542e-01 7.54980981e-01
-5.82226785e-03 -3.46083254e-01 -4.96570140e-01 1.42981023e-01
3.30271035e-01 1.23497820e+00 -6.53486669e-01 -3.29177231e-01
-3.56297731e-01 1.14659953e+00 -1.39344722e-01 7.42430508e-01
-1.17807186e+00 -2.66174793e-01 1.02305162e+00 7.19172284e-02
2.39585638e-01 -4.65620607e-01 -4.28184897e-01 -1.10480690e+00
1.19762257e-01 -6.91231787e-01 2.37025708e-01 -8.79110217e-01
-1.01272631e+00 4.41872895e-01 -1.80809712e-03 -1.33192503e+00
8.97164345e-02 -2.29380831e-01 -4.30620223e-01 8.33723247e-01
-2.30594683e+00 -1.37827063e+00 -7.13384569e-01 6.24132097e-01
1.04205775e+00 1.98822528e-01 2.06679225e-01 5.71435809e-01
-7.17108071e-01 3.98123384e-01 3.79552811e-01 -2.60782868e-01
1.07795274e+00 -1.19537842e+00 7.74984360e-01 1.11098599e+00
6.28077090e-02 2.08922550e-01 7.02010036e-01 -3.22782218e-01
-1.10129714e+00 -1.67289090e+00 7.01178849e-01 -1.51159987e-01
3.35545689e-01 -4.66518968e-01 -9.10835862e-01 6.48777544e-01
2.00213537e-01 9.92295817e-02 5.13226569e-01 -3.66191447e-01
-3.89335811e-01 -5.36187887e-01 -8.95773649e-01 6.62246287e-01
1.00616276e+00 -6.49002314e-01 -1.21838212e-01 5.65693498e-01
8.40822339e-01 -6.58968925e-01 -3.14665198e-01 2.64816612e-01
3.17084551e-01 -1.04713428e+00 8.22698534e-01 2.53790263e-02
1.63378939e-01 -6.83102608e-01 -2.25934401e-01 -1.43431711e+00
-6.39579892e-02 -6.42047465e-01 2.77881712e-01 1.21817493e+00
2.99255431e-01 -1.03955138e+00 5.08801043e-01 5.79249144e-01
-4.99550194e-01 -3.16847861e-01 -1.00308335e+00 -9.09532487e-01
7.70902857e-02 -4.49045658e-01 6.75864160e-01 8.33021760e-01
-8.89360905e-01 1.23681389e-01 -2.47668713e-01 4.34892356e-01
6.61498129e-01 2.89275855e-01 9.94126856e-01 -8.26181948e-01
-2.62581080e-01 -2.13369489e-01 -1.72855053e-02 -1.10603094e+00
4.35728014e-01 -6.74897432e-01 4.73804444e-01 -1.55955565e+00
-4.64049168e-02 -6.64095402e-01 -2.70796120e-01 4.58930910e-01
-3.41380119e-01 6.34447217e-01 1.93664536e-01 3.08342934e-01
-5.52721739e-01 7.63059020e-01 1.26059198e+00 -2.78639346e-01
-2.81173766e-01 7.41389543e-02 -7.47346044e-01 6.22562945e-01
1.05157828e+00 -5.37346363e-01 -6.63105547e-01 -7.01609910e-01
-9.07520428e-02 -4.89708543e-01 3.82441044e-01 -9.72675860e-01
8.11675712e-02 -3.35501224e-01 9.18882713e-02 -2.86297679e-01
4.42004263e-01 -6.89913154e-01 1.36701807e-01 2.09576249e-01
-1.11678317e-02 -3.10000539e-01 3.46036971e-01 6.52374208e-01
-2.61140347e-01 1.35844164e-02 1.09500575e+00 1.20070502e-01
-9.81107652e-01 3.39799106e-01 9.16036814e-02 3.22406948e-01
1.05576503e+00 -2.04127237e-01 -4.48149055e-01 -1.83287784e-01
-4.18474495e-01 5.49789608e-01 8.57333422e-01 7.12224007e-01
2.00954035e-01 -1.16976047e+00 -5.75345337e-01 3.42007577e-01
4.33761865e-01 4.41595078e-01 5.47933459e-01 1.11990702e+00
-4.30296659e-01 8.80988538e-02 4.98922691e-02 -8.03553104e-01
-9.18524563e-01 2.67051548e-01 3.61212611e-01 2.23088041e-01
-6.71010792e-01 8.06070209e-01 7.64850497e-01 -5.54376483e-01
-1.25022352e-01 -1.45354137e-01 7.98029155e-02 -5.33445738e-03
3.71261477e-01 1.49132326e-01 1.82869211e-01 -8.39738250e-01
-2.42596939e-01 7.58093536e-01 6.33630231e-02 -2.11172149e-01
1.10501742e+00 -6.92928970e-01 1.28185511e-01 6.63196385e-01
1.20668840e+00 -2.20614165e-01 -1.99889135e+00 -2.63315499e-01
-2.84259826e-01 -7.14207768e-01 1.68412581e-01 -6.61012650e-01
-1.29298675e+00 9.48750377e-01 8.57590139e-01 -1.06625512e-01
1.48165274e+00 -1.32053569e-01 1.24624074e+00 1.43385762e-02
5.04157357e-02 -1.38619936e+00 -1.75145298e-01 4.20886934e-01
5.03895342e-01 -1.62321627e+00 -3.99014175e-01 -2.47684777e-01
-9.67945278e-01 7.84782171e-01 5.31759322e-01 1.07072338e-01
4.28017318e-01 -2.53066644e-02 6.48955286e-01 1.04527280e-01
-4.59468871e-01 -4.88688469e-01 2.23906606e-01 5.65152645e-01
4.81807068e-03 2.21379116e-01 -5.59539646e-02 1.67160407e-01
-2.09276870e-01 -1.29675061e-01 3.54441851e-01 8.74436140e-01
-2.73042262e-01 -1.15321517e+00 -2.58965403e-01 2.57409923e-02
2.31006127e-02 4.29259203e-02 -1.74637929e-01 4.83491123e-01
4.15124655e-01 1.09271884e+00 1.12870693e-01 -2.76687950e-01
4.36074823e-01 1.00782529e-01 1.60283118e-01 -4.20396537e-01
-1.19394004e-01 1.96034178e-01 -1.75948173e-01 -6.84480369e-01
-4.22703832e-01 -7.85878718e-01 -1.30636418e+00 -2.32479721e-01
-2.18992438e-02 -5.12329722e-03 1.01362062e+00 9.80540931e-01
6.12681746e-01 5.94767988e-01 8.80405247e-01 -8.60096693e-01
-1.31017968e-01 -5.47455251e-01 -5.24468422e-01 4.25799102e-01
5.92564881e-01 -6.65882111e-01 -3.81923676e-01 3.24082017e-01] | [8.845301628112793, -1.4951142072677612] |
8c905b10-5407-48e4-8183-07d8299fa2e2 | divinet-3d-reconstruction-from-disparate | 2306.04699 | null | https://arxiv.org/abs/2306.04699v3 | https://arxiv.org/pdf/2306.04699v3.pdf | DiViNeT: 3D Reconstruction from Disparate Views via Neural Template Regularization | We present a volume rendering-based neural surface reconstruction method that takes as few as three disparate RGB images as input. Our key idea is to regularize the reconstruction, which is severely ill-posed and leaving significant gaps between the sparse views, by learning a set of neural templates that act as surface priors. Our method coined DiViNet, operates in two stages. The first stage learns the templates, in the form of 3D Gaussian functions, across different scenes, without 3D supervision. In the reconstruction stage, our predicted templates serve as anchors to help "stitch" the surfaces over sparse regions. We demonstrate that our approach is not only able to complete the surface geometry but also reconstructs surface details to a reasonable extent from few disparate input views. On the DTU and BlendedMVS datasets, our approach achieves the best reconstruction quality among existing methods in the presence of such sparse views, and performs on par, if not better, with competing methods when dense views are employed as inputs. | ['Hao Zhang', 'Akshay Gadi Patil', 'Aditya Vora'] | 2023-06-07 | null | null | null | null | ['3d-reconstruction'] | ['computer-vision'] | [ 5.06814420e-01 3.48033130e-01 3.02647322e-01 -4.19794351e-01
-9.77380753e-01 -5.15323579e-01 5.49700260e-01 -4.75597590e-01
1.57803655e-01 4.63598549e-01 2.89731413e-01 -5.17455228e-02
2.21484751e-01 -8.43168795e-01 -1.07720077e+00 -5.06820917e-01
3.87951225e-01 7.01151788e-01 2.08252892e-01 -2.39105359e-01
6.51581138e-02 7.14018285e-01 -1.78962409e+00 4.18058127e-01
8.31270695e-01 1.07193911e+00 2.80960113e-01 3.40562493e-01
-2.52231926e-01 6.11034930e-01 -9.80967879e-02 -3.53410840e-01
5.96465111e-01 -1.10967971e-01 -6.16282046e-01 5.13179779e-01
7.90933371e-01 -4.53683704e-01 -4.02068347e-01 8.14803481e-01
1.12507507e-01 8.51679668e-02 6.01430237e-01 -5.75854421e-01
-3.30530405e-01 -1.08105421e-01 -7.79073179e-01 -3.81365836e-01
5.60470104e-01 8.89668390e-02 8.70532870e-01 -1.40683484e+00
9.46317255e-01 1.19008327e+00 6.85041904e-01 5.74328840e-01
-1.59690678e+00 -4.19276863e-01 3.30791563e-01 -4.21238869e-01
-1.30873132e+00 -6.36750758e-01 1.13102996e+00 -6.04794085e-01
7.51246929e-01 3.19543093e-01 8.48056436e-01 9.55075026e-01
7.60194808e-02 5.79799473e-01 1.00984693e+00 -6.79218695e-02
2.98656017e-01 -1.40394807e-01 -2.11644962e-01 7.87152350e-01
-7.72532225e-02 1.77201018e-01 -8.06328356e-01 -4.20011103e-01
1.43069673e+00 2.44703919e-01 -7.84208238e-01 -6.46485507e-01
-1.12300491e+00 7.62182534e-01 4.32952106e-01 -1.03762940e-01
-6.76444054e-01 8.80055949e-02 -1.41163453e-01 1.16601840e-01
8.77736807e-01 2.17402756e-01 -4.15394515e-01 2.15172619e-01
-9.66010094e-01 2.34672904e-01 6.37858331e-01 8.06642592e-01
1.11167455e+00 2.04712138e-01 1.86326057e-01 7.55283356e-01
4.74143595e-01 5.17501414e-01 -1.54062465e-01 -1.35695910e+00
5.45342565e-01 4.47260201e-01 2.98836082e-01 -9.62703288e-01
-1.49096195e-02 -2.11162478e-01 -8.72646987e-01 5.72142601e-01
1.14817165e-01 1.03588946e-01 -1.24590409e+00 1.47762668e+00
6.44474745e-01 3.47803056e-01 -2.05467507e-01 9.70527709e-01
8.98896694e-01 6.55876398e-01 -5.79551160e-01 -1.35598183e-01
9.84991193e-01 -6.88833833e-01 -6.28408670e-01 -3.89008403e-01
-1.32440686e-01 -7.28754222e-01 1.11649239e+00 5.38105547e-01
-1.47099674e+00 -4.21480387e-01 -9.25320566e-01 -2.75098890e-01
1.98327556e-01 -2.51870245e-01 6.79569364e-01 1.17108710e-01
-1.37571216e+00 7.84460068e-01 -1.04598439e+00 1.19165882e-01
7.49293685e-01 1.91650197e-01 -5.91314673e-01 -5.18471777e-01
-4.05092031e-01 5.16433895e-01 -4.08508688e-01 4.54677977e-02
-9.81002986e-01 -8.50327075e-01 -1.13301849e+00 -1.14135534e-01
2.54506379e-01 -1.02034497e+00 9.16664779e-01 -9.77527022e-01
-1.49856615e+00 1.00192654e+00 -4.37192053e-01 2.24212408e-01
5.61946511e-01 -1.95425451e-01 2.82794058e-01 1.19897798e-01
-6.30475879e-02 6.53740644e-01 9.67240512e-01 -1.98672557e+00
-1.89475864e-01 -6.94845200e-01 -4.26355843e-03 3.96270633e-01
4.60197121e-01 -5.45312762e-01 -7.14499772e-01 -7.06980586e-01
7.90537596e-01 -5.69342375e-01 -4.89393592e-01 6.48352802e-01
-4.38135594e-01 2.84593284e-01 5.78707337e-01 -8.32989037e-01
3.50916207e-01 -1.96576524e+00 5.48504651e-01 3.96352082e-01
5.25221884e-01 -2.99753666e-01 -1.57998145e-01 1.69883490e-01
7.49691799e-02 -1.62591040e-01 -6.18056595e-01 -9.13089216e-01
-3.06680948e-01 3.88264507e-01 -4.19126779e-01 5.84376633e-01
7.73084238e-02 7.13928759e-01 -6.82270169e-01 -3.80288553e-03
2.85193235e-01 1.01650524e+00 -6.38468862e-01 6.69098854e-01
-3.94668251e-01 9.59476292e-01 -3.83704036e-01 7.28448629e-01
9.40883696e-01 -4.89326805e-01 1.47871926e-01 -4.32891399e-01
-6.36314601e-02 2.46008769e-01 -1.12701523e+00 2.23939180e+00
-4.09350246e-01 4.76491570e-01 5.86536527e-01 -8.09177041e-01
9.11100388e-01 3.64791781e-01 7.55676925e-01 -5.68415642e-01
2.46257707e-02 1.35998771e-01 -5.36668003e-01 -2.36881047e-01
2.03722090e-01 -3.58763307e-01 3.75142664e-01 4.07895684e-01
-2.04245709e-02 -5.36763966e-01 -5.87313473e-01 1.28798443e-03
1.02813959e+00 5.95377684e-01 7.16135949e-02 8.35285485e-02
1.77967355e-01 -1.12311862e-01 6.03179276e-01 2.44378239e-01
5.07520616e-01 1.35637891e+00 2.20616102e-01 -6.60456300e-01
-1.14378345e+00 -1.28144360e+00 -5.46198264e-02 4.18934554e-01
3.40807170e-01 -2.87722349e-01 -5.66863477e-01 -5.98453283e-01
6.37601539e-02 4.33795005e-01 -7.24431217e-01 2.43447199e-01
-5.46326637e-01 -2.70645231e-01 -1.89538211e-01 4.60744202e-01
8.22322518e-02 -9.91160631e-01 -4.39237088e-01 4.88064289e-02
-1.14571616e-01 -9.99488890e-01 -2.51394749e-01 3.03548545e-01
-1.24809980e+00 -1.09055364e+00 -7.97376275e-01 -4.04097080e-01
1.02497613e+00 5.76877177e-01 1.39890981e+00 2.34451532e-01
5.83592278e-04 3.98343831e-01 3.99017632e-02 2.33542807e-02
-1.13902472e-01 -6.32152736e-01 -3.15175891e-01 1.77647114e-01
-3.32477748e-01 -1.27782238e+00 -6.09275222e-01 2.27059871e-01
-8.94463122e-01 4.86063093e-01 2.85372853e-01 8.80179703e-01
1.12487519e+00 -5.17336190e-01 -3.22176851e-02 -1.09173644e+00
-6.48191497e-02 -4.81721699e-01 -6.65951371e-01 3.83890010e-02
-2.94320852e-01 2.36654580e-02 5.82626760e-01 -2.08843067e-01
-1.09673965e+00 3.35707635e-01 -4.37872797e-01 -9.67840195e-01
-1.76178098e-01 1.18212864e-01 -3.49814266e-01 -3.40130597e-01
5.50306201e-01 1.98843703e-01 1.55702427e-01 -8.72266471e-01
3.36871147e-01 4.32841480e-02 4.21096504e-01 -6.31667137e-01
9.61212814e-01 9.61602747e-01 -1.14694372e-01 -6.34877801e-01
-7.28801250e-01 -2.39830747e-01 -8.36042404e-01 -1.49052769e-01
6.92506671e-01 -1.03643680e+00 -3.68333489e-01 3.80065292e-01
-1.27556860e+00 -5.07600427e-01 -5.17781377e-01 1.00388698e-01
-7.89875627e-01 3.11171502e-01 -5.26504934e-01 -6.30323529e-01
-2.99514025e-01 -1.20188856e+00 1.49240518e+00 -1.11641567e-02
8.84609595e-02 -8.25357378e-01 7.59902000e-02 5.01342416e-01
2.32214898e-01 5.44140279e-01 8.32591474e-01 5.50319143e-02
-1.10977125e+00 1.74269959e-01 -1.08380117e-01 2.91448742e-01
3.03616732e-01 -1.70768693e-01 -1.24300659e+00 -2.69001126e-01
3.01602095e-01 -3.73034984e-01 7.44345129e-01 4.67739612e-01
1.26198542e+00 -1.97338164e-01 -1.94544360e-01 1.34075737e+00
1.64685345e+00 -1.41958036e-02 7.85514951e-01 -2.42860634e-02
1.15570664e+00 7.32768118e-01 2.36738831e-01 3.99440646e-01
4.66553181e-01 5.99565268e-01 7.60484695e-01 -4.41614002e-01
-1.84199303e-01 -2.87574172e-01 -3.15567255e-02 8.14008951e-01
-3.47381443e-01 -5.15729487e-02 -7.52775908e-01 2.02861160e-01
-1.69637203e+00 -5.02906978e-01 1.53075695e-01 2.28532052e+00
6.47628725e-01 -1.66491210e-01 -1.88467443e-01 -1.19060233e-01
1.92076921e-01 4.58250523e-01 -7.71047354e-01 6.16967939e-02
-3.58181708e-02 4.59894866e-01 1.13015294e-01 7.51098990e-01
-5.94272912e-01 8.23406219e-01 6.46806383e+00 2.76389599e-01
-1.20810938e+00 1.37059554e-01 7.70639598e-01 -1.90326929e-01
-1.15617478e+00 6.22120034e-03 -1.64642245e-01 1.91827595e-01
3.20106953e-01 4.50346291e-01 7.53404915e-01 5.88955939e-01
7.76412562e-02 2.14004461e-02 -1.18970585e+00 1.12627089e+00
1.10884815e-01 -1.77606320e+00 8.13991353e-02 1.45175621e-01
8.86404037e-01 2.78154224e-01 -9.03309882e-02 -2.33035803e-01
4.51161116e-01 -1.20410824e+00 8.78883839e-01 7.44232655e-01
1.05318427e+00 -5.55488825e-01 3.02069038e-01 5.06615222e-01
-1.06048727e+00 5.12782395e-01 -3.91264051e-01 1.84532210e-01
4.56817865e-01 6.02086425e-01 -3.33936393e-01 5.95237315e-01
6.57652438e-01 8.76546323e-01 -2.84950435e-02 7.53703058e-01
-2.26768658e-01 2.91566044e-01 -3.91347587e-01 8.13030541e-01
-1.06772229e-01 -5.66870272e-01 5.11954844e-01 4.31337267e-01
4.06632155e-01 4.74688441e-01 2.74437904e-01 1.24567187e+00
-1.36868864e-01 -2.78831758e-02 -9.29671586e-01 4.48954463e-01
2.78620243e-01 1.09122050e+00 -5.70119798e-01 -2.16768488e-01
-5.36669195e-01 1.19591153e+00 6.82161689e-01 6.79136515e-01
-4.57252830e-01 4.64719504e-01 7.83347905e-01 4.52465653e-01
3.24226201e-01 -3.20222080e-01 -6.61867201e-01 -1.42773831e+00
3.31310064e-01 -9.05885756e-01 -6.74323440e-02 -9.59884584e-01
-1.27830231e+00 7.50317812e-01 -3.80812168e-01 -1.18574679e+00
-1.08842164e-01 -4.24308360e-01 -6.22048795e-01 1.16526365e+00
-1.53782761e+00 -1.30485201e+00 -4.97365266e-01 7.06028819e-01
5.73161364e-01 1.83526173e-01 7.95512140e-01 1.16300680e-01
-2.36758530e-01 7.51472488e-02 -1.50327850e-02 -2.69937247e-01
2.36963540e-01 -1.04608905e+00 6.89441681e-01 6.00062966e-01
3.04044574e-01 4.82407093e-01 3.79473239e-01 -7.54210353e-01
-1.60918951e+00 -8.15382183e-01 2.35684887e-01 -3.38914096e-01
-1.12705231e-02 -5.55688918e-01 -1.28732586e+00 7.75181115e-01
-4.76409122e-03 4.71751571e-01 4.26845223e-01 6.87753409e-02
-4.71901506e-01 1.15202241e-01 -1.38797343e+00 3.86469871e-01
1.24104619e+00 -5.86014032e-01 -3.74143660e-01 2.74396360e-01
7.30293691e-01 -8.16823781e-01 -8.73181105e-01 3.92671496e-01
5.68055928e-01 -1.33325350e+00 1.32756221e+00 -3.32010686e-01
5.95103085e-01 -2.35044777e-01 -6.00285888e-01 -1.34240091e+00
-2.09152892e-01 -6.61374390e-01 -2.65828580e-01 8.36943030e-01
2.38802403e-01 -5.62317371e-01 1.10884118e+00 7.09752202e-01
-2.92894006e-01 -1.15686285e+00 -9.63314593e-01 -2.29150072e-01
-2.48038381e-01 -5.15023649e-01 6.41892850e-01 1.01946461e+00
-7.79899716e-01 1.58479586e-01 -3.09638649e-01 3.16649407e-01
9.34527338e-01 3.69391918e-01 8.92665744e-01 -1.28471756e+00
-3.36257964e-01 -9.06076953e-02 -5.80452308e-02 -1.45660782e+00
1.16782755e-01 -7.60579526e-01 2.29421914e-01 -1.68842089e+00
-6.70967028e-02 -7.25337267e-01 1.06146403e-01 4.26857144e-01
-1.47814661e-01 3.79542977e-01 5.95343933e-02 3.85662436e-01
2.28202306e-02 8.44210029e-01 1.60784614e+00 4.97970469e-02
-1.48860186e-01 2.45249197e-02 -6.57143176e-01 1.01010048e+00
1.10841893e-01 -2.20442817e-01 -4.75961596e-01 -8.17334235e-01
1.55209064e-01 5.24545133e-01 5.23236394e-01 -7.60109901e-01
-9.15762782e-02 -3.28396440e-01 6.97969079e-01 -8.85916412e-01
9.34605181e-01 -9.48080182e-01 6.52391076e-01 -1.42703235e-01
3.60516198e-02 5.03425971e-02 -6.23908155e-02 6.41298175e-01
-7.84998089e-02 2.97742665e-01 7.94158876e-01 -4.07965153e-01
-3.39588195e-01 7.94210076e-01 1.51183859e-01 -2.88020857e-02
6.09192967e-01 -4.16196465e-01 1.38077170e-01 -4.46495891e-01
-7.68345773e-01 1.93566345e-02 1.09817946e+00 1.49449199e-01
1.07184231e+00 -1.32362986e+00 -6.09807611e-01 7.09049463e-01
-1.64263681e-01 9.19810474e-01 3.06417972e-01 4.42075312e-01
-5.29255152e-01 -1.48008928e-01 -7.39046559e-02 -9.31838751e-01
-9.29845572e-01 2.18595922e-01 4.54517543e-01 -2.66065379e-03
-1.32938540e+00 1.11108291e+00 6.94963574e-01 -6.49179220e-01
3.13730776e-01 -2.02824593e-01 5.09110726e-02 -2.79199868e-01
4.19170618e-01 -1.63731556e-02 1.33371383e-01 -6.67871058e-01
-4.76784557e-02 9.63229477e-01 1.55059785e-01 -2.64559507e-01
1.83791316e+00 -3.54082771e-02 -1.11040503e-01 5.41560829e-01
1.12265813e+00 1.60161629e-01 -1.96495879e+00 -4.26924676e-01
-4.52124089e-01 -9.60788846e-01 1.58595085e-01 -5.40244997e-01
-1.50979602e+00 9.19188559e-01 2.61755705e-01 -2.29893506e-01
9.92484689e-01 2.12719321e-01 9.40657318e-01 -4.20814976e-02
7.69762635e-01 -3.92246336e-01 1.11007735e-01 3.68801326e-01
1.16203749e+00 -1.17036879e+00 8.17579627e-02 -7.93406844e-01
-4.24307793e-01 9.52689886e-01 5.95198095e-01 -4.19534326e-01
6.15891397e-01 4.53415751e-01 1.34062961e-01 -7.58502364e-01
-7.08983302e-01 1.17255405e-01 5.58868229e-01 7.85101771e-01
3.22264373e-01 -1.50578827e-01 4.35014129e-01 2.55524963e-01
-7.52114058e-02 -2.61082828e-01 2.42790714e-01 7.66139805e-01
-2.95679599e-01 -9.25240815e-01 -5.81891954e-01 5.44659793e-01
-1.50267497e-01 -6.14725146e-03 -2.73512959e-01 5.89969099e-01
1.42635955e-02 5.00331759e-01 2.38766074e-01 -4.57182527e-01
5.43503821e-01 -2.08669797e-01 5.86085320e-01 -8.45659077e-01
-3.74817163e-01 3.62128526e-01 -5.79127856e-02 -1.23067522e+00
-4.38881904e-01 -5.94858050e-01 -1.07356608e+00 -1.66853845e-01
-1.47998342e-02 -1.77570790e-01 5.71142256e-01 8.92540693e-01
2.90920764e-01 3.60534579e-01 9.02646363e-01 -1.66739571e+00
-2.76047587e-01 -5.16776741e-01 -5.13831913e-01 4.24162716e-01
5.86141646e-01 -8.36969137e-01 -4.19234484e-01 2.06633620e-02] | [9.035650253295898, -3.1954963207244873] |
c7f89996-3d9d-4a9f-a97c-6ca4578636c0 | improving-multiple-pedestrian-tracking-by | null | null | http://openaccess.thecvf.com//content/CVPR2021/html/Stadler_Improving_Multiple_Pedestrian_Tracking_by_Track_Management_and_Occlusion_Handling_CVPR_2021_paper.html | http://openaccess.thecvf.com//content/CVPR2021/papers/Stadler_Improving_Multiple_Pedestrian_Tracking_by_Track_Management_and_Occlusion_Handling_CVPR_2021_paper.pdf | Improving Multiple Pedestrian Tracking by Track Management and Occlusion Handling | Multi-pedestrian trackers perform well when targets are clearly visible making the association task quite easy. However, when heavy occlusions are present, a mechanism to reidentify persons is needed. The common approach is to extract visual features from new detections and compare them with the features of previously found tracks. Since those detections can have substantial overlaps with nearby targets - especially in crowded scenarios - the extracted features are insufficient for a reliable re-identification. In contrast, we propose a novel occlusion handling strategy that explicitly models the relation between occluding and occluded tracks outperforming the feature-based approach, while not depending on a separate re-identification network. Furthermore, we improve the track management of a regression-based method in order to bypass missing detections and to deal with tracks leaving the scene at the border of the image. Finally, we apply our tracker in both temporal directions and merge tracklets belonging to the same target, which further enhances the performance. We demonstrate the effectiveness of our tracking components with ablative experiments and surpass the state-of-the-art methods on the three popular pedestrian tracking benchmarks MOT16, MOT17, and MOT20. | ['Jurgen Beyerer', 'Daniel Stadler'] | 2021-06-19 | null | null | null | cvpr-2021-1 | ['occlusion-handling'] | ['computer-vision'] | [-2.47252271e-01 -3.76473010e-01 1.44072041e-01 -8.33306164e-02
-3.52086902e-01 -5.81195295e-01 6.39813006e-01 6.29226923e-01
-7.28556514e-01 9.51660693e-01 1.65099418e-03 1.02479704e-01
2.00813301e-02 -7.88278937e-01 -7.74150252e-01 -7.23074317e-01
1.91145279e-02 4.99427348e-01 9.97500658e-01 6.80491328e-02
-4.22235504e-02 7.06865311e-01 -1.88585687e+00 1.09551176e-01
5.32155633e-01 6.86212480e-01 9.95984077e-02 4.04225916e-01
-4.22069617e-02 3.85696739e-01 -7.55477011e-01 -4.10323709e-01
3.92569304e-01 4.00512554e-02 -1.66873366e-01 1.85675681e-01
1.04355741e+00 6.72227750e-03 -3.65941226e-01 9.11612928e-01
3.67300004e-01 8.20893720e-02 4.56032783e-01 -1.26680517e+00
1.40328407e-01 1.79103434e-01 -9.06800210e-01 5.07880092e-01
3.29146415e-01 1.25975817e-01 5.43318450e-01 -6.93176389e-01
7.04801083e-01 1.20567822e+00 9.20985758e-01 2.22225547e-01
-1.31163204e+00 -7.54454672e-01 5.68289638e-01 4.21703875e-01
-1.53436518e+00 -5.15474260e-01 2.83075213e-01 -6.07951283e-01
4.24800545e-01 3.52628320e-01 6.90999508e-01 8.34551334e-01
-1.16527461e-01 6.09420657e-01 9.23502564e-01 -3.90850127e-01
-5.31558841e-02 2.98552334e-01 5.13429046e-01 5.47683001e-01
8.44429255e-01 4.49499071e-01 -4.01200026e-01 -1.58703446e-01
3.31087440e-01 9.90149528e-02 -1.52902707e-01 -8.26374292e-01
-1.28518856e+00 4.99231011e-01 5.96377909e-01 4.66592073e-01
-5.03116369e-01 -5.40903285e-02 4.12673682e-01 4.82366346e-02
2.71729022e-01 -3.65858921e-03 -4.74112853e-02 1.43154591e-01
-1.17561996e+00 2.58216798e-01 4.58135009e-01 8.80227327e-01
8.09198439e-01 -2.49552712e-01 -5.89788437e-01 3.47726613e-01
1.80807233e-01 4.46871072e-01 -5.66019118e-02 -3.43491793e-01
2.84106225e-01 7.93403149e-01 5.20953655e-01 -9.67914522e-01
-5.77064633e-01 -8.59991491e-01 -5.65929294e-01 5.29436111e-01
9.10878003e-01 -3.13503109e-02 -7.15772212e-01 1.45438802e+00
8.65470350e-01 4.44620073e-01 -3.02351773e-01 7.20218480e-01
7.35148966e-01 1.31030053e-01 1.86597198e-01 -1.73602968e-01
1.67832160e+00 -9.17852283e-01 -6.52842283e-01 -1.80191472e-01
4.89236236e-01 -8.50393474e-01 1.36448950e-01 1.99227065e-01
-6.92559361e-01 -8.69650900e-01 -9.30885792e-01 4.67305064e-01
-5.36202133e-01 5.97456634e-01 2.40038767e-01 8.11078370e-01
-8.10420334e-01 3.58174354e-01 -7.39389777e-01 -6.86625242e-01
3.52853894e-01 4.45938230e-01 -5.20845234e-01 -6.09074254e-03
-6.45506740e-01 1.08983886e+00 6.10032856e-01 1.28922552e-01
-6.37858629e-01 -5.55445075e-01 -6.70358062e-01 -9.11204591e-02
4.78416681e-01 -5.70069790e-01 6.22513056e-01 -5.13439655e-01
-7.60080159e-01 7.51992524e-01 -5.42346179e-01 -6.67904317e-01
9.57837999e-01 -3.20344359e-01 -5.09965003e-01 -2.21725360e-01
3.18757951e-01 6.36134446e-01 7.28723466e-01 -1.34445405e+00
-1.24916172e+00 -2.00714752e-01 4.50850166e-02 -1.26515254e-01
-2.05814630e-01 1.45148307e-01 -5.73638201e-01 -4.84372407e-01
-1.24286391e-01 -9.06827867e-01 -1.67603537e-01 2.68200159e-01
-4.28868920e-01 -2.63649493e-01 9.84310687e-01 -4.48966324e-01
1.07315135e+00 -2.21674562e+00 -5.06317765e-02 3.06073010e-01
4.00204659e-01 5.93775153e-01 -5.12762815e-02 2.02812821e-01
7.03485906e-02 -5.15088499e-01 4.40468431e-01 -6.68489099e-01
-1.55300707e-01 1.58385187e-02 2.27026604e-02 8.02949369e-01
-4.93172929e-02 5.44488788e-01 -7.69791842e-01 -5.15652061e-01
6.97245300e-01 4.63630080e-01 -1.74889714e-01 -1.12059116e-01
7.63424858e-02 6.91776633e-01 -2.59295225e-01 4.40820932e-01
8.30070376e-01 -3.41547504e-02 -8.62196758e-02 -2.17143312e-01
-6.87613368e-01 -2.41895299e-02 -1.80569184e+00 1.17971170e+00
-1.23636432e-01 7.77323663e-01 -5.44222370e-02 -6.79342806e-01
8.31113875e-01 2.16171131e-01 5.34559906e-01 -3.78616452e-01
9.34596658e-02 4.15252596e-02 -1.63723454e-02 -7.69942552e-02
5.63026488e-01 4.52373177e-01 2.59201825e-01 4.14056480e-02
-2.60287523e-01 1.02686739e+00 7.44244516e-01 5.12415990e-02
1.12865233e+00 1.46164790e-01 3.00430357e-01 -4.42205220e-02
8.12462568e-01 1.14089705e-01 6.88961208e-01 1.06873226e+00
-3.11437666e-01 2.27828160e-01 -1.06415469e-02 -5.96351206e-01
-6.81533039e-01 -1.12335515e+00 -1.96909934e-01 9.10292923e-01
4.73536789e-01 -5.43481469e-01 -4.50559258e-01 -7.78890193e-01
2.43289784e-01 3.71301860e-01 -6.58458531e-01 1.54389247e-01
-6.90751493e-01 -7.41980970e-01 3.70718658e-01 5.16934991e-01
4.57327336e-01 -6.90432847e-01 -8.39596450e-01 4.35160816e-01
-1.97596297e-01 -1.35721898e+00 -3.99605155e-01 -9.63243470e-02
-4.16323423e-01 -1.43589318e+00 -6.44157827e-01 -4.01758850e-01
6.85869277e-01 6.80234671e-01 8.61341298e-01 2.87518233e-01
-4.69719976e-01 3.08016688e-01 -1.59560397e-01 -3.66529077e-01
-2.48214155e-01 -9.76738855e-02 1.53583825e-01 2.74387717e-01
4.57144916e-01 -1.03078656e-01 -4.72890973e-01 7.49782383e-01
-1.46000594e-01 -1.41207725e-01 4.23090756e-01 5.34334838e-01
4.88186747e-01 1.28854111e-01 2.52629191e-01 -3.67007136e-01
-1.92849040e-01 -1.79389745e-01 -1.05469108e+00 4.30223346e-01
-5.03268987e-02 -1.22901894e-01 1.38592556e-01 -7.05318868e-01
-9.00061905e-01 4.80422527e-01 1.89083591e-01 -3.61487061e-01
-4.65564132e-01 -3.02948982e-01 -1.20378442e-01 -3.55376810e-01
6.44774318e-01 -1.03152096e-01 -2.98950851e-01 -5.89541316e-01
3.06843847e-01 2.69685447e-01 6.48551166e-01 -2.87286222e-01
1.20848918e+00 8.49936068e-01 1.17412701e-01 -8.10104191e-01
-7.28871703e-01 -1.12748337e+00 -9.88977194e-01 -6.18122041e-01
8.06922853e-01 -9.75369394e-01 -9.93635118e-01 2.20114231e-01
-1.35408783e+00 3.63120586e-01 -2.35241726e-01 6.79618716e-01
5.04027586e-03 5.47364831e-01 -1.71141982e-01 -1.08866668e+00
-3.90244275e-02 -1.07357311e+00 9.75443661e-01 4.42679405e-01
-6.69836029e-02 -6.86810076e-01 1.84328631e-02 -1.04430644e-02
2.63967246e-01 4.81735796e-01 4.15139459e-02 -6.41143203e-01
-8.53265166e-01 -4.97932434e-01 -4.07964528e-01 -2.41231650e-01
1.43808350e-01 -1.45779196e-02 -9.86958981e-01 -5.88601351e-01
-5.18690526e-01 4.35115397e-01 1.01337218e+00 5.06065547e-01
4.03447777e-01 2.32318267e-01 -1.21434474e+00 2.61464030e-01
1.14870667e+00 -5.76476343e-02 2.83053458e-01 7.46774673e-01
6.00419879e-01 6.35983348e-01 7.40335107e-01 3.33833098e-01
3.35319519e-01 1.32160556e+00 3.52199912e-01 -2.96080232e-01
-4.75453407e-01 4.57876287e-02 3.49127144e-01 -6.00009449e-02
-1.54259145e-01 -1.56247884e-01 -5.51021099e-01 5.15960634e-01
-2.16612148e+00 -1.11333859e+00 -7.41477191e-01 2.57698750e+00
1.94418907e-01 3.69588822e-01 6.68207228e-01 1.42284587e-01
1.19432414e+00 -2.07410336e-01 -1.30820870e-01 5.10030568e-01
-1.28190115e-01 -1.25100181e-01 8.44540536e-01 3.95775884e-01
-1.59194815e+00 8.04956257e-01 5.64064837e+00 6.23554349e-01
-6.95309579e-01 1.89709276e-01 -1.82151258e-01 -2.56936580e-01
5.84053576e-01 -2.49101478e-03 -1.58839524e+00 6.24091566e-01
4.96653259e-01 2.28505597e-01 -1.64360538e-01 6.22746706e-01
1.33816227e-01 -4.45644438e-01 -1.05931199e+00 9.43827033e-01
-3.26569155e-02 -1.19801641e+00 -4.94255066e-01 3.00267488e-01
3.00233126e-01 -1.52828872e-01 -3.06692719e-01 2.06712782e-01
3.35389167e-01 -4.57135648e-01 8.97216201e-01 5.53487599e-01
1.71983853e-01 -5.36082029e-01 7.32521296e-01 2.95012712e-01
-1.86262786e+00 -1.35521561e-01 -4.27935094e-01 8.01660940e-02
4.67089146e-01 5.50486624e-01 -6.62060499e-01 8.11997712e-01
8.04030061e-01 6.14672303e-01 -1.00982201e+00 2.08700347e+00
-1.36440411e-01 8.22976232e-02 -6.03197336e-01 1.90365791e-01
4.35534818e-03 1.89033762e-01 9.68074083e-01 1.29956603e+00
2.97823161e-01 -4.46695119e-01 6.22483492e-01 6.06838107e-01
4.87893999e-01 5.77739207e-03 -4.74807054e-01 7.57993102e-01
5.03176808e-01 1.27968931e+00 -9.46559906e-01 -4.21392530e-01
-6.41259134e-01 6.61318362e-01 3.14056605e-01 2.05426261e-01
-1.06740642e+00 9.73559245e-02 7.38271117e-01 3.63184690e-01
7.24217534e-01 -2.31884301e-01 2.12836534e-01 -1.04145086e+00
2.48785019e-01 -4.87474889e-01 6.29815340e-01 -1.82129547e-01
-1.07447565e+00 5.63112676e-01 1.13787204e-01 -1.43577874e+00
1.31557006e-02 -4.72637296e-01 -5.14688253e-01 6.97751224e-01
-1.58012819e+00 -1.38033032e+00 -6.07144952e-01 5.48364222e-01
2.53000528e-01 -5.01801819e-02 4.12823439e-01 8.71117413e-01
-6.52222335e-01 5.43140888e-01 1.31901037e-02 2.75224596e-01
8.38483214e-01 -9.44802880e-01 3.48207861e-01 1.24191439e+00
3.56297076e-01 3.43025804e-01 8.76355946e-01 -8.93023670e-01
-8.30698192e-01 -1.16391623e+00 8.06727588e-01 -5.52084148e-01
4.85890985e-01 -5.94848216e-01 -7.85561502e-01 6.04551911e-01
2.13135732e-03 2.82036722e-01 3.88973594e-01 2.25074351e-01
-1.77864507e-01 -1.09426744e-01 -8.95477235e-01 3.76693457e-01
1.14289880e+00 1.00249946e-01 -4.44816262e-01 2.86400795e-01
3.80120426e-02 -3.89591098e-01 -4.98713136e-01 3.02470565e-01
5.75854182e-01 -9.48640645e-01 1.42480171e+00 -2.26942688e-01
-8.03113818e-01 -9.61972594e-01 1.49349108e-01 -9.81340945e-01
-4.15712237e-01 -3.36142421e-01 -1.59332752e-01 1.48139143e+00
-3.68578359e-02 -5.51887929e-01 8.14159155e-01 1.88389763e-01
1.08118154e-01 4.65323702e-02 -1.26930690e+00 -1.27396238e+00
-5.30023932e-01 -4.99223024e-02 4.79021937e-01 5.21510601e-01
-4.72809881e-01 -2.34458335e-02 -4.75432128e-01 5.94788730e-01
1.08762956e+00 2.13305578e-02 1.23985040e+00 -1.71299267e+00
-1.07910492e-01 -3.50018233e-01 -8.88899446e-01 -8.53978097e-01
-9.53038782e-03 -5.51887751e-01 4.69087437e-02 -1.36363506e+00
1.18916199e-01 -6.79036200e-01 -3.05157155e-01 5.10369420e-01
-2.57397711e-01 4.38968748e-01 4.99431103e-01 2.08845496e-01
-8.74412298e-01 2.21379310e-01 7.05831170e-01 -2.91681796e-01
-2.13589534e-01 3.49721938e-01 -1.20582163e-01 7.13167548e-01
3.58989477e-01 -7.36810625e-01 2.88208306e-01 -8.05223584e-02
-3.56511027e-01 -3.95175606e-01 8.48429322e-01 -1.67576253e+00
6.10315502e-01 3.03119242e-01 7.55143106e-01 -1.16571426e+00
5.11051238e-01 -1.06792295e+00 4.95620191e-01 6.82084322e-01
1.43910542e-01 6.94275647e-02 4.83059317e-01 6.89354479e-01
1.49966206e-03 -2.45244026e-01 7.37495959e-01 9.91103351e-02
-8.29845965e-01 2.39398062e-01 -3.32742602e-01 -3.99881154e-01
1.47848272e+00 -4.46375161e-01 -5.20940363e-01 7.21016973e-02
-8.66918623e-01 3.80745471e-01 5.32725275e-01 5.53815186e-01
2.24901468e-01 -1.23662972e+00 -7.92039931e-01 5.53716682e-02
2.71497458e-01 -4.40689921e-01 8.95695388e-02 1.11439645e+00
-1.18478183e-02 4.03562278e-01 -2.20087022e-01 -9.17151034e-01
-1.94834566e+00 8.19726467e-01 4.45584655e-01 -3.02328378e-01
-9.68034387e-01 3.29621136e-01 4.44117486e-01 1.60703510e-01
5.42993069e-01 -8.74599069e-02 -5.80675602e-01 2.28106692e-01
8.68340254e-01 5.23656905e-01 1.06864326e-01 -9.53310668e-01
-6.13644063e-01 6.73387647e-01 -1.63928092e-01 2.04069912e-01
9.02012527e-01 -2.23168552e-01 2.23025784e-01 1.57909822e-02
4.22129750e-01 2.51082212e-01 -1.32991660e+00 -4.73304063e-01
2.99310446e-01 -7.11451828e-01 -1.96796283e-01 -4.29410487e-01
-9.47268665e-01 5.27649581e-01 1.10231745e+00 1.32990241e-01
5.51542282e-01 -2.29864605e-02 3.72254908e-01 1.60475895e-01
4.88976538e-01 -7.94737577e-01 -1.99461713e-01 1.98069677e-01
4.02336508e-01 -1.10114896e+00 1.94630161e-01 -6.70941293e-01
-1.02583647e-01 1.00235140e+00 6.40167415e-01 8.29062685e-02
4.09194559e-01 3.60682786e-01 -8.28230605e-02 -1.14644080e-01
-2.92124391e-01 -8.87310684e-01 4.43352312e-01 8.30179393e-01
7.53432289e-02 -1.00014374e-01 -1.67059869e-01 -6.06661588e-02
1.32585317e-01 -2.53679007e-01 1.86911523e-01 8.82677674e-01
-5.77623248e-01 -1.31269705e+00 -1.01694322e+00 1.85771614e-01
-1.73112541e-01 2.31329098e-01 -1.53918579e-01 1.01879680e+00
5.26078641e-01 1.16920733e+00 2.57553279e-01 -2.16526631e-02
6.16095185e-01 -1.28024071e-01 6.33591354e-01 -4.15562868e-01
-8.26893926e-01 2.25907296e-01 1.72647119e-01 -4.56678659e-01
-6.98931217e-01 -1.06583273e+00 -7.96571374e-01 -2.01814115e-01
-7.30461240e-01 3.94224860e-02 5.50016940e-01 9.56142485e-01
1.77020371e-01 7.06813335e-01 1.14298344e-01 -9.01781082e-01
-5.32775782e-02 -7.48898566e-01 -2.08647728e-01 4.20815557e-01
3.46352339e-01 -1.42856383e+00 1.29865319e-01 -1.13971017e-01] | [6.465353965759277, -1.9659761190414429] |
65514e46-86d8-4e21-8f8a-53240e80861a | ensemble-classifier-design-tuned-to-dataset | 2205.06177 | null | https://arxiv.org/abs/2205.06177v1 | https://arxiv.org/pdf/2205.06177v1.pdf | Ensemble Classifier Design Tuned to Dataset Characteristics for Network Intrusion Detection | Machine Learning-based supervised approaches require highly customized and fine-tuned methodologies to deliver outstanding performance. This paper presents a dataset-driven design and performance evaluation of a machine learning classifier for the network intrusion dataset UNSW-NB15. Analysis of the dataset suggests that it suffers from class representation imbalance and class overlap in the feature space. We employed ensemble methods using Balanced Bagging (BB), eXtreme Gradient Boosting (XGBoost), and Random Forest empowered by Hellinger Distance Decision Tree (RF-HDDT). BB and XGBoost are tuned to handle the imbalanced data, and Random Forest (RF) classifier is supplemented by the Hellinger metric to address the imbalance issue. Two new algorithms are proposed to address the class overlap issue in the dataset. These two algorithms are leveraged to help improve the performance of the testing dataset by modifying the final classification decision made by three base classifiers as part of the ensemble classifier which employs a majority vote combiner. The proposed design is evaluated for both binary and multi-category classification. Comparing the proposed model to those reported on the same dataset in the literature demonstrate that the proposed model outperforms others by a significant margin for both binary and multi-category classification cases. | ['Gursel Serpen', 'Zeinab Zoghi'] | 2022-05-08 | null | null | null | null | ['network-intrusion-detection'] | ['miscellaneous'] | [ 1.57782048e-01 -2.58762568e-01 -2.77180552e-01 -8.06258500e-01
-2.55072594e-01 -4.09159034e-01 4.18329626e-01 5.27796924e-01
-2.19755903e-01 1.04820049e+00 3.73151600e-02 -6.99438453e-01
-7.39756048e-01 -8.39828312e-01 1.21885419e-01 -9.35357094e-01
3.69344316e-02 4.19833004e-01 8.17348994e-03 -3.85378659e-01
7.77386487e-01 7.48913348e-01 -1.78825116e+00 6.09838009e-01
8.18152487e-01 1.26675475e+00 -6.02601826e-01 6.20644689e-01
-2.23974451e-01 8.95887852e-01 -7.29170740e-01 -2.44820386e-01
4.93107766e-01 -1.82948321e-01 -5.47381818e-01 -4.73182678e-01
1.84155405e-01 5.52205928e-03 5.07445812e-01 3.13508391e-01
8.08123171e-01 -9.39497799e-02 1.03041673e+00 -1.79214990e+00
5.98597229e-02 4.05616343e-01 -7.82768488e-01 4.28371191e-01
1.02969512e-01 -1.93998516e-01 6.99027002e-01 -6.65997028e-01
2.23800853e-01 9.69338298e-01 1.03896785e+00 2.63159603e-01
-9.69754457e-01 -1.10370445e+00 -2.37256184e-01 4.52602953e-01
-1.17800188e+00 -9.08561051e-02 8.16435516e-01 -5.02045989e-01
1.08200049e+00 5.41047335e-01 5.23909569e-01 5.26474595e-01
6.44322991e-01 -5.08626997e-02 1.59263694e+00 -7.58724272e-01
4.47903186e-01 5.55418134e-01 8.29038262e-01 3.87815803e-01
6.95437610e-01 5.14865160e-01 -4.53108430e-01 -5.44024289e-01
-3.44733030e-01 2.57305294e-01 2.21057385e-01 -2.76805818e-01
-4.62949157e-01 1.04529309e+00 4.13867056e-01 3.69458199e-01
-5.95612943e-01 -4.91004795e-01 5.85949957e-01 4.50345695e-01
3.46708775e-01 3.22213560e-01 -6.22946978e-01 1.01885520e-01
-9.90257800e-01 2.35919505e-01 9.36938107e-01 4.49333489e-01
5.35137296e-01 8.29809520e-04 -7.22125620e-02 6.32606626e-01
3.63687247e-01 1.08601123e-01 6.34914696e-01 -1.03042156e-01
3.14559758e-01 1.15776193e+00 -2.06259906e-01 -1.00820041e+00
-6.67759597e-01 -1.04374146e+00 -6.98302746e-01 6.03383958e-01
1.35057062e-01 -1.83586121e-01 -1.01493919e+00 1.16342747e+00
6.95197046e-01 -2.97080040e-01 2.00678483e-01 3.13455999e-01
7.75040329e-01 2.28949189e-01 3.73913407e-01 -2.31901452e-01
9.74933624e-01 -5.96529961e-01 -5.58466554e-01 2.76206434e-01
7.90891409e-01 -6.38007939e-01 1.64086133e-01 6.56038761e-01
-4.29993063e-01 -5.43982625e-01 -1.59131312e+00 7.25273669e-01
-8.08679461e-01 -3.09750557e-01 5.15301943e-01 1.31248057e+00
-5.41959167e-01 4.21856791e-01 -1.87732413e-01 -3.66588414e-01
5.51381111e-01 6.46098256e-01 -3.11033666e-01 -2.23970804e-02
-1.02072895e+00 1.18966246e+00 4.57367003e-01 -7.53396749e-02
-1.20683439e-01 -5.90597034e-01 -1.98059738e-01 -2.18018994e-01
-3.72424632e-01 -5.68423569e-01 5.11411667e-01 -9.38670635e-01
-1.14223361e+00 7.48686612e-01 1.84826300e-01 -6.37580216e-01
5.55977821e-01 1.67002395e-01 -7.94982135e-01 -2.10455418e-01
-1.09642535e-01 1.08959615e-01 6.98064446e-01 -1.18694043e+00
-1.11985660e+00 -9.67017114e-01 -4.22811985e-01 1.54342335e-02
-2.13372424e-01 -1.47362083e-01 9.95714486e-01 -5.15265584e-01
4.09090698e-01 -6.18385077e-01 -1.11604095e-01 -8.10600221e-01
-4.58351791e-01 1.20508885e-02 1.34510720e+00 -5.52694023e-01
1.58275688e+00 -1.74470568e+00 -5.89611888e-01 8.13406587e-01
-1.02688987e-02 4.16775018e-01 4.55951840e-01 5.94916761e-01
-4.24663752e-01 -1.07894894e-02 -1.47719622e-01 1.81291610e-01
-2.71330804e-01 1.81196272e-01 -3.33906770e-01 4.29355234e-01
-9.24312323e-02 2.15651244e-01 -2.19862759e-01 -4.31222111e-01
2.24774241e-01 2.15654507e-01 -4.36775267e-01 2.79005438e-01
3.26015234e-01 2.90095031e-01 -2.09758565e-01 8.78340244e-01
9.43980336e-01 4.06028748e-01 1.30515173e-01 -3.18692714e-01
-3.67126316e-02 -1.01138353e-01 -1.43714225e+00 5.98396122e-01
-2.35560670e-01 -4.06777449e-02 -3.69649291e-01 -1.40491748e+00
1.62934816e+00 2.02395737e-01 4.31839734e-01 -4.78543460e-01
3.38843763e-01 4.40629810e-01 4.35783207e-01 -3.90440762e-01
1.82580188e-01 -3.31268549e-01 4.37923893e-02 5.27235866e-01
-4.54733931e-02 2.05317378e-01 -5.59574217e-02 -2.52077371e-01
1.16013062e+00 -1.07901163e-01 9.00743484e-01 -4.48018909e-01
8.15078497e-01 3.87408048e-01 5.97286820e-01 7.89833665e-01
-4.03068393e-01 -1.46844601e-02 1.50202647e-01 -8.50823104e-01
-7.28271604e-01 -7.56031990e-01 -5.61565816e-01 1.22326291e+00
-2.44031511e-02 8.26265141e-02 -3.77704293e-01 -9.38001513e-01
3.51424396e-01 8.55467081e-01 -6.01128280e-01 -2.95641422e-01
-2.65331775e-01 -1.50898564e+00 5.20288229e-01 3.66637617e-01
7.06155360e-01 -6.76541924e-01 -8.28120530e-01 3.74829531e-01
4.24810618e-01 -4.32046294e-01 6.28209591e-01 8.72494876e-01
-1.07577920e+00 -1.26599514e+00 1.64400131e-01 -3.83682519e-01
3.46550167e-01 -6.23550825e-02 9.55158889e-01 2.54985511e-01
-5.60615897e-01 -1.20085910e-01 -7.18913078e-01 -9.19158161e-01
-3.48487884e-01 5.74138947e-02 -2.08134204e-02 1.14574917e-01
7.96158493e-01 -8.13078344e-01 -7.43223906e-01 4.68785882e-01
-4.75471973e-01 -4.84115720e-01 6.41158760e-01 1.07838476e+00
-1.82958711e-02 4.02384400e-01 1.25989366e+00 -1.07856369e+00
5.75101018e-01 -1.03718162e+00 -1.04315050e-01 1.53964430e-01
-1.22606468e+00 -1.89292535e-01 4.70570952e-01 -2.25185543e-01
-9.49003875e-01 -2.81176805e-01 -4.15361933e-02 4.34702963e-01
-2.81744957e-01 2.37668842e-01 -1.30327553e-01 -2.91146070e-01
9.47065353e-01 -1.72694892e-01 -4.68966030e-02 -3.10973167e-01
-1.77047357e-01 1.44273674e+00 1.08146789e-02 -2.97359705e-01
3.90624195e-01 3.25651139e-01 3.69879931e-01 -2.46701673e-01
-6.74915254e-01 -6.51373565e-01 -7.03370094e-01 -2.54354089e-01
4.24844503e-01 -5.77192903e-01 -5.67404091e-01 4.64103848e-01
-5.85987568e-01 2.94624358e-01 8.73696655e-02 4.36340630e-01
-1.23012587e-01 -3.64150614e-01 1.03113934e-01 -1.10520613e+00
-9.02986228e-01 -7.58146048e-01 3.95302773e-01 3.19325596e-01
-4.63542283e-01 -8.25666070e-01 9.67396945e-02 6.22165680e-01
7.13782012e-01 8.62972438e-01 1.24636531e+00 -1.63767958e+00
1.84754074e-01 -8.97741795e-01 -8.16384032e-02 3.27762902e-01
1.08097434e-01 6.63008168e-02 -1.18340147e+00 -2.62662560e-01
3.88237499e-02 4.31716479e-02 6.66637063e-01 1.95245333e-02
8.70221198e-01 -3.89339924e-02 -5.14786005e-01 3.31825852e-01
1.63313591e+00 7.89621830e-01 2.69923359e-01 8.21291208e-01
3.13867629e-01 5.57926893e-01 7.49223113e-01 6.05722427e-01
2.97256559e-01 5.11031926e-01 5.39084196e-01 1.07139342e-01
2.67770495e-02 1.50748238e-01 -2.75443703e-01 3.45536828e-01
1.08174644e-01 -5.66404825e-03 -1.15946364e+00 1.65872633e-01
-1.64016759e+00 -9.66088772e-01 -2.82646209e-01 2.20374346e+00
4.86070395e-01 4.50927645e-01 3.00736338e-01 1.19185889e+00
8.41912150e-01 -3.16569477e-01 -4.79591578e-01 -1.07479024e+00
-2.45139729e-02 5.61711729e-01 5.44119298e-01 2.42505893e-01
-1.04126036e+00 1.15819983e-01 5.68815947e+00 5.88952184e-01
-1.25927389e+00 2.59822346e-02 8.03372979e-01 5.53333648e-02
2.86488533e-01 7.41883442e-02 -1.04231143e+00 5.88446319e-01
9.60992634e-01 7.35466406e-02 -2.59642035e-01 8.83113623e-01
-1.16648898e-01 -4.09570932e-01 -7.43124068e-01 6.97282434e-01
1.17141113e-01 -9.96161759e-01 -1.82311743e-01 1.32173553e-01
6.48204803e-01 -3.31640512e-01 -1.41785994e-01 3.54731321e-01
2.39138335e-01 -8.68353724e-01 1.83623925e-01 4.95928675e-01
3.10873181e-01 -1.12791348e+00 1.24534070e+00 2.83153802e-01
-6.31052315e-01 -9.30833817e-01 5.78114949e-02 -3.80787939e-01
-4.56513047e-01 1.01858604e+00 -1.17261195e+00 6.94517553e-01
8.30350518e-01 1.78967312e-01 -7.16480672e-01 1.10157144e+00
3.29419881e-01 7.59872019e-01 -2.88458318e-01 -8.41703340e-02
-3.15504968e-02 7.52799511e-02 5.51889837e-01 1.00441945e+00
1.83018997e-01 -8.62072259e-02 1.41416201e-02 -3.12570930e-02
4.64996368e-01 4.88404542e-01 -3.27550173e-01 6.76454425e-01
7.65992582e-01 1.34119296e+00 -6.74780846e-01 -3.17880541e-01
3.74475420e-02 1.45436183e-01 1.20125629e-01 -1.73881412e-01
-5.24902225e-01 -5.58923960e-01 3.23976994e-01 2.53819525e-01
2.22878903e-01 3.88218880e-01 -1.08277035e+00 -4.06453729e-01
-2.72188216e-01 -7.56536603e-01 9.17567372e-01 -2.84790128e-01
-1.36402988e+00 7.87227929e-01 -9.72054899e-03 -1.24699914e+00
-3.98075014e-01 -5.01150310e-01 -7.97549903e-01 9.47704196e-01
-1.25191438e+00 -1.29094744e+00 -6.28295422e-01 3.91173124e-01
-4.78734784e-02 -7.15312243e-01 8.94267619e-01 2.53511190e-01
-5.90815783e-01 7.58867085e-01 1.95612371e-01 -3.52028400e-01
7.26929188e-01 -1.12169600e+00 -4.19409066e-01 4.37815994e-01
-5.35516798e-01 3.30466896e-01 7.96379983e-01 -7.05284774e-01
-8.66087973e-01 -1.13028777e+00 7.03277826e-01 -2.68746674e-01
1.08379818e-01 2.65171868e-03 -5.83104074e-01 2.07255676e-01
-6.66802824e-02 1.14341311e-01 1.44941902e+00 1.48808792e-01
-4.63682413e-01 -6.11779869e-01 -2.16388941e+00 -1.65079966e-01
4.53996599e-01 1.34211540e-01 -6.28072202e-01 -7.53631582e-03
-1.83095455e-01 -2.81972643e-02 -1.11368787e+00 9.41731274e-01
9.98257518e-01 -1.40723681e+00 6.08713090e-01 -8.24651837e-01
2.79473335e-01 -3.52182359e-01 -5.06801188e-01 -1.25716639e+00
-2.03082487e-01 -1.24605764e-02 -4.63617109e-02 1.43926156e+00
3.92928421e-01 -9.61334646e-01 7.89794803e-01 3.47478181e-01
2.14026839e-01 -1.16809452e+00 -8.66683424e-01 -4.13893729e-01
6.52080253e-02 -1.86313510e-01 9.21069384e-01 9.63675976e-01
-1.98917948e-02 2.23658130e-01 -4.59002741e-02 -2.36294214e-02
7.96271741e-01 3.44293028e-01 8.09922934e-01 -1.55119908e+00
2.32292898e-02 -2.43193671e-01 -9.65142965e-01 5.79086006e-01
-2.50567764e-01 -8.58528972e-01 -5.15748143e-01 -1.04176927e+00
4.05327231e-02 -1.04859126e+00 -5.67382336e-01 3.46831858e-01
-2.18408421e-01 5.53672135e-01 2.10422208e-03 1.50404632e-01
-2.81684013e-04 -1.16107628e-01 4.52385932e-01 1.23342022e-01
-2.41137266e-01 3.69399577e-01 -9.55289125e-01 5.32397628e-01
1.14139581e+00 -6.57718956e-01 -1.50785729e-01 4.76429641e-01
-1.60620093e-01 -8.05115551e-02 1.13977499e-01 -1.32200229e+00
1.77676201e-01 -4.72507179e-02 1.02310562e+00 -8.55255961e-01
2.00021490e-02 -1.07662702e+00 3.91986758e-01 8.14211845e-01
-2.23267213e-01 1.72115520e-01 7.27131218e-02 3.77516687e-01
-2.02833652e-01 -1.68775886e-01 1.13153362e+00 3.11433554e-01
-4.98716116e-01 -5.31346463e-02 -4.97569181e-02 -4.05250520e-01
1.60919619e+00 -7.32349634e-01 -5.45936167e-01 2.60311402e-02
-7.11268961e-01 2.20023058e-02 3.12038604e-02 2.60406166e-01
3.07821542e-01 -1.22645891e+00 -7.83297241e-01 4.78332281e-01
2.79568136e-01 -6.44261241e-01 1.40552104e-01 8.25898468e-01
-3.57625723e-01 3.13760817e-01 -8.28797281e-01 -5.11853516e-01
-1.62371731e+00 2.60550499e-01 3.94420415e-01 -5.69097161e-01
5.64563945e-02 6.91606581e-01 -7.43005812e-01 -7.22763479e-01
1.70951009e-01 4.89674062e-01 -6.31367564e-01 3.65306377e-01
4.88245577e-01 1.00969148e+00 7.79705822e-01 -5.07613122e-01
-5.92965364e-01 3.73847693e-01 -3.81072313e-01 3.11573744e-01
1.47430325e+00 4.64183949e-02 -2.29229122e-01 1.97891861e-01
1.10218525e+00 -2.36914992e-01 -2.64247686e-01 4.82050106e-02
2.95124114e-01 -3.61590326e-01 2.24821121e-01 -1.49559569e+00
-6.44496918e-01 5.75414121e-01 1.25592756e+00 3.23092222e-01
1.40449190e+00 -8.25677097e-01 2.90404350e-01 3.86980735e-02
1.78362131e-01 -1.10850954e+00 -4.19566453e-01 2.73783147e-01
5.41533530e-01 -1.27900040e+00 4.04187620e-01 -1.26409441e-01
-4.64384854e-01 1.35953438e+00 6.30008459e-01 -1.87045470e-01
1.11311364e+00 4.07796741e-01 7.70240948e-02 -4.17511612e-02
-9.28372443e-01 1.28613532e-01 2.81200670e-02 7.92133451e-01
3.26426446e-01 8.62138793e-02 -7.92546093e-01 7.37200201e-01
-4.06852692e-01 1.95356473e-01 1.78835630e-01 1.39025152e+00
-5.70474863e-01 -1.08192801e+00 -7.24016309e-01 1.13618636e+00
-6.12689972e-01 2.12068066e-01 -3.58080447e-01 7.16434836e-01
7.58666515e-01 1.31966126e+00 8.21454599e-02 -7.81091750e-01
2.42914781e-01 3.97772163e-01 2.05799490e-01 -7.90522546e-02
-1.21408248e+00 -7.02439904e-01 2.00062096e-01 -1.77708566e-01
-3.98642957e-01 -3.59556168e-01 -8.21184278e-01 -3.60852748e-01
-5.50575495e-01 4.20658231e-01 1.28900242e+00 1.05954611e+00
3.08986008e-01 5.04669607e-01 1.16148961e+00 -5.41785121e-01
-7.83546567e-01 -1.15449989e+00 -5.38166463e-01 1.61316469e-01
3.41507077e-01 -9.39317167e-01 -6.47021651e-01 -3.59111369e-01] | [8.371797561645508, 4.572478771209717] |
11c0d1e2-5095-4d4d-a03f-65ba5316edc4 | spelling-correction-for-russian-a-comparative | null | null | https://aclanthology.org/2021.ranlp-main.136 | https://aclanthology.org/2021.ranlp-main.136.pdf | Spelling Correction for Russian: A Comparative Study of Datasets and Methods | We develop a minimally-supervised model for spelling correction and evaluate its performance on three datasets annotated for spelling errors in Russian. The first corpus is a dataset of Russian social media data that was recently used in a shared task on Russian spelling correction. The other two corpora contain texts produced by learners of Russian as a foreign language. Evaluating on three diverse datasets allows for a cross-corpus comparison. We compare the performance of the minimally-supervised model to two baseline models that do not use context for candidate re-ranking, as well as to a character-level statistical machine translation system with context-based re-ranking. We show that the minimally-supervised model outperforms all of the other models. We also present an analysis of the spelling errors and discuss the difficulty of the task compared to the spelling correction problem in English. | ['Alla Rozovskaya'] | null | null | https://aclanthology.org/2021.ranlp-1.136 | https://aclanthology.org/2021.ranlp-1.136.pdf | ranlp-2021-9 | ['cross-corpus', 'spelling-correction'] | ['computer-vision', 'natural-language-processing'] | [ 5.57460725e-01 -9.61100161e-02 -1.80686966e-01 -2.46423453e-01
-1.28493369e+00 -6.26273334e-01 7.05033779e-01 7.99090326e-01
-9.38742340e-01 1.01540029e+00 5.69655240e-01 -6.65733159e-01
2.27230862e-01 -4.09772933e-01 -6.89620197e-01 -1.42818108e-01
7.42156267e-01 7.67044783e-01 3.96445870e-01 -6.30130053e-01
7.88992405e-01 2.24471435e-01 -1.15474558e+00 5.21104753e-01
1.37162745e+00 -1.60707787e-01 4.55469072e-01 7.25772798e-01
-2.96343148e-01 6.34326160e-01 -7.89970160e-01 -7.01004028e-01
-2.74717994e-02 -3.89021248e-01 -1.15059495e+00 -4.49773490e-01
8.47257674e-01 4.27016467e-01 -1.32827722e-02 1.33720648e+00
2.74462521e-01 1.22330129e-01 6.94961488e-01 -4.64352667e-01
-9.75507975e-01 8.13746333e-01 -9.88672376e-02 5.50137341e-01
6.15782082e-01 -4.00783271e-01 1.16089833e+00 -1.10120738e+00
1.01036370e+00 9.32290196e-01 8.08420300e-01 1.00353158e+00
-9.93375123e-01 -3.52553368e-01 -5.31451963e-02 1.37370870e-01
-1.27468050e+00 -3.87017429e-01 1.64755378e-02 -4.78991419e-01
1.25852656e+00 3.83327514e-01 1.56358585e-01 8.27234566e-01
4.78938855e-02 6.42470241e-01 1.20801532e+00 -1.19670260e+00
-1.18864834e-01 2.37209976e-01 4.12689209e-01 4.82950270e-01
3.91394168e-01 2.90182326e-02 -4.88011450e-01 -3.09155788e-02
1.79679796e-01 -3.96730840e-01 -2.19881907e-01 1.74890339e-01
-1.46458817e+00 6.05545998e-01 -3.10494244e-01 5.84326625e-01
2.34484181e-01 -1.24827228e-01 5.37717283e-01 6.56985402e-01
8.36101413e-01 8.12149286e-01 -1.03148878e+00 -2.92297482e-01
-1.05698812e+00 3.56089681e-01 7.20334649e-01 9.06472385e-01
6.43338799e-01 -8.54263231e-02 -2.78307050e-01 1.18642747e+00
1.33998170e-01 5.84664464e-01 7.96590269e-01 -3.22036654e-01
1.05244160e+00 4.43660080e-01 3.82004410e-01 -4.39318597e-01
-3.28864694e-01 -2.30292842e-01 -2.76131779e-01 -8.28680694e-02
7.69607484e-01 -7.01175481e-02 -7.94225395e-01 1.46348274e+00
-9.59023535e-02 2.38993857e-02 2.37093493e-01 5.24197042e-01
7.91000962e-01 6.56000853e-01 8.81676227e-02 -5.42928934e-01
9.58551526e-01 -1.40769565e+00 -6.76833570e-01 -1.78352997e-01
1.59147453e+00 -1.08639038e+00 1.32175386e+00 3.77529681e-01
-1.10516965e+00 -4.28315908e-01 -8.33044112e-01 -3.09544623e-01
-4.31123435e-01 4.27797288e-01 1.28144667e-01 8.08623672e-01
-1.07846117e+00 1.03412271e+00 -3.58231813e-01 -5.64754069e-01
-3.09345275e-01 5.37470281e-02 -4.33595240e-01 1.83043748e-01
-1.29928887e+00 1.37285304e+00 3.49977791e-01 -4.94720727e-01
-2.22837642e-01 -7.10377812e-01 -7.68871188e-01 -2.66345352e-01
-2.19523549e-01 -2.16370463e-01 1.70990336e+00 -1.20920539e+00
-1.51593268e+00 1.39818823e+00 -3.79840970e-01 -4.60001022e-01
4.25962597e-01 -3.67747515e-01 -8.55027497e-01 -3.62115324e-01
3.24156433e-01 -1.94483027e-02 3.44759107e-01 -7.54107058e-01
-9.93813515e-01 -2.43475214e-02 -5.24127543e-01 4.27156299e-01
-1.47031516e-01 8.48074973e-01 -1.00890994e-01 -1.10184407e+00
-1.16424255e-01 -1.22624195e+00 -3.16182673e-01 -9.93660688e-01
-5.13445102e-02 -3.65187466e-01 1.64256081e-01 -1.03334653e+00
1.74755371e+00 -1.76371324e+00 1.86870724e-01 5.53074218e-02
-3.72111917e-01 7.85063267e-01 -3.11937273e-01 6.28260136e-01
-6.15768254e-01 4.75483805e-01 -3.30714464e-01 -5.46992362e-01
-3.43659759e-01 -3.38357277e-02 -4.79359448e-01 3.75937462e-01
4.59901877e-02 8.18532109e-01 -1.31993210e+00 -2.07142711e-01
-2.03659847e-01 -1.54666349e-01 -4.56433177e-01 -2.46583655e-01
-1.00897349e-01 5.31824052e-01 1.09728433e-01 3.37039977e-01
3.62292528e-01 1.04967669e-01 1.26028568e-01 5.51531553e-01
-3.56672883e-01 1.17136264e+00 -1.03528559e+00 1.63039148e+00
-6.30786777e-01 5.53213298e-01 -6.23772919e-01 -4.10107613e-01
9.35370266e-01 2.83870429e-01 -2.21376836e-01 -6.91419721e-01
-1.85054168e-01 8.08515966e-01 -1.01069562e-01 -1.83852047e-01
1.23752391e+00 -9.45491269e-02 -3.11465949e-01 6.50548756e-01
9.22767371e-02 -2.51402646e-01 4.96203274e-01 3.59228402e-01
1.09478712e+00 2.10965648e-01 6.18986130e-01 -4.62100059e-01
8.40499699e-01 2.51223981e-01 6.04072332e-01 9.09142792e-01
1.87217537e-02 8.04572463e-01 2.34380867e-02 -4.69271779e-01
-1.03668261e+00 -7.59401500e-01 -1.19947344e-01 1.34029865e+00
-8.67095031e-03 -9.22363639e-01 -8.25319469e-01 -9.51395690e-01
-2.10533738e-01 1.08929634e+00 -4.23034668e-01 1.04696326e-01
-8.34512889e-01 -5.76415420e-01 6.65887296e-01 3.07947725e-01
-2.72791445e-01 -1.10436940e+00 2.97886550e-01 8.50798190e-02
-4.06482786e-01 -8.91622066e-01 -7.89022624e-01 1.46219820e-01
-8.24531317e-01 -7.12058067e-01 -4.89831865e-01 -1.06535923e+00
6.02408111e-01 1.21331148e-01 1.28116953e+00 5.96992671e-01
3.75259757e-01 9.31513309e-02 -8.22036207e-01 -6.88584149e-01
-1.18203771e+00 3.66319865e-01 4.03541476e-01 -5.86202025e-01
7.75332510e-01 -5.46322204e-02 2.97064692e-01 6.57641292e-02
-5.42054474e-01 -4.47661802e-02 1.76938117e-01 7.89576113e-01
5.11476159e-01 -7.39268601e-01 5.17103314e-01 -1.41825259e+00
6.35462642e-01 -2.95311630e-01 -5.55769205e-01 3.75298083e-01
-7.91240215e-01 1.26359880e-01 6.36496186e-01 -1.06905699e-01
-1.05398417e+00 1.44885793e-01 -4.49500173e-01 2.66666412e-01
-1.32957697e-01 4.79875773e-01 7.33981058e-02 -2.03318536e-01
1.04062152e+00 1.37180895e-01 -6.29322886e-01 -8.83897901e-01
2.50836045e-01 8.99410844e-01 6.23343945e-01 -4.13545638e-01
1.06392431e+00 -3.60243529e-01 -3.55048478e-01 -8.30806255e-01
-1.07466388e+00 -8.14623058e-01 -8.85130703e-01 1.70109808e-01
5.30801117e-01 -9.59880888e-01 1.93323866e-01 4.85497147e-01
-1.41903234e+00 -3.76474142e-01 -5.13149858e-01 5.73779285e-01
-4.81252700e-01 8.15825820e-01 -9.75337207e-01 -4.04855043e-01
-2.54386425e-01 -7.75689065e-01 7.84028649e-01 7.11192787e-02
-5.52669108e-01 -1.16222429e+00 7.75758684e-01 4.29115862e-01
-2.38293447e-02 -4.30980861e-01 8.09988618e-01 -1.22175014e+00
8.73817429e-02 3.76068391e-02 8.44435543e-02 2.36017615e-01
-4.68871519e-02 -2.66020820e-02 -6.30432487e-01 -1.53014034e-01
-3.81339937e-01 -1.60162941e-01 1.10655892e+00 -1.53224662e-01
6.71292841e-01 -1.54173002e-01 -1.82932422e-01 4.84330714e-01
1.07308948e+00 -1.65270701e-01 7.79389322e-01 8.25410008e-01
9.63626742e-01 4.77950782e-01 1.05219769e+00 1.61153916e-02
4.66627002e-01 7.68800735e-01 -7.02285171e-02 1.56397715e-01
-1.79048404e-01 -6.59048676e-01 5.01814365e-01 1.65611970e+00
-2.38169879e-01 -2.66279161e-01 -1.11815441e+00 7.98788071e-01
-1.88091707e+00 -9.30077910e-01 -8.52887571e-01 2.54590607e+00
1.24965513e+00 -6.36307225e-02 4.92510945e-02 1.25887677e-01
9.00611639e-01 -1.10487990e-01 3.80689234e-01 -1.24668360e+00
-3.00955951e-01 4.86883372e-01 8.29195917e-01 9.60215092e-01
-1.17131197e+00 1.67138851e+00 6.94333410e+00 9.41417038e-01
-7.94807494e-01 2.88274974e-01 1.24416322e-01 3.51265401e-01
-6.22655213e-01 2.85555333e-01 -1.17979681e+00 4.18269813e-01
1.39138436e+00 -1.25712916e-01 2.01355770e-01 6.24132335e-01
1.72026768e-01 -1.73364282e-02 -9.15434897e-01 6.99619055e-01
3.43207151e-01 -1.28007281e+00 8.56257007e-02 -3.81736845e-01
1.36165380e+00 3.71160209e-01 -3.28093708e-01 4.08722699e-01
5.99826753e-01 -8.96006465e-01 9.53300893e-01 3.91527623e-01
9.88184988e-01 -7.07340121e-01 6.31302953e-01 5.46299815e-01
-6.77440345e-01 2.31163993e-01 -4.95572984e-01 -4.35110509e-01
6.12566955e-02 3.92932057e-01 -8.90391886e-01 4.26096916e-01
4.24697250e-01 8.19762826e-01 -9.74992394e-01 1.12444150e+00
-8.38006556e-01 8.67816806e-01 2.11803332e-01 -1.82070017e-01
-5.41110300e-02 -1.94958001e-01 8.92581284e-01 1.92856681e+00
3.72266054e-01 -1.14103273e-01 7.98569352e-04 2.57883668e-02
-2.35802904e-01 8.30535114e-01 -3.11235219e-01 -1.85632054e-02
4.26873565e-01 9.39136744e-01 -5.05781412e-01 -5.63272893e-01
-4.45828319e-01 1.36246312e+00 7.37554073e-01 6.68415651e-02
-4.22276258e-01 -5.64810753e-01 4.87254173e-01 2.06261426e-01
1.31146535e-01 -3.03331882e-01 -4.55989003e-01 -1.58760357e+00
-1.80219561e-01 -1.04882431e+00 3.49325925e-01 -6.72681510e-01
-1.24235189e+00 5.30129910e-01 -3.07777941e-01 -1.33210087e+00
-2.12204486e-01 -7.10949838e-01 -4.90724236e-01 1.17619205e+00
-1.72295666e+00 -9.08630967e-01 3.62684101e-01 3.39003980e-01
7.92866170e-01 -4.49962080e-01 9.76150334e-01 4.17664796e-01
-4.03411955e-01 7.72743642e-01 5.34188032e-01 1.25712171e-01
1.47974348e+00 -1.64434993e+00 9.88517165e-01 1.44312739e+00
6.23309970e-01 7.69960582e-01 7.99946070e-01 -1.05181909e+00
-5.45073569e-01 -1.38461721e+00 2.32586741e+00 -1.00631297e+00
7.59263992e-01 -1.67532012e-01 -1.04727268e+00 8.13663006e-01
1.45140484e-01 -1.48934990e-01 6.94357872e-01 2.73107380e-01
-3.40205133e-01 4.05279964e-01 -8.66311789e-01 6.02171540e-01
1.19209445e+00 -5.65389693e-01 -1.26357102e+00 6.43686950e-01
6.85412526e-01 -6.59561515e-01 -3.20448697e-01 3.55197005e-02
7.23031461e-02 -4.94359195e-01 4.66796160e-01 -1.03847516e+00
6.17695630e-01 -4.58408713e-01 1.29227474e-01 -1.94295633e+00
-6.24520898e-01 -8.49449098e-01 4.59915370e-01 1.33386779e+00
1.00223672e+00 -2.37554178e-01 3.73165011e-01 1.35254920e-01
-5.48730433e-01 1.34973064e-01 -8.16094875e-01 -8.65919590e-01
6.44353390e-01 -3.79322886e-01 3.10646892e-01 1.19595480e+00
4.71442968e-01 4.16409194e-01 -4.55205590e-01 1.87645443e-02
6.50299862e-02 -2.16088608e-01 6.40906870e-01 -1.43210912e+00
-1.68552160e-01 -2.40280718e-01 -2.25072846e-01 -9.27491784e-01
4.43315744e-01 -1.42749786e+00 2.24255264e-01 -1.37893689e+00
6.09042794e-02 -4.48230535e-01 -2.33616054e-01 4.64754522e-01
-5.51218390e-01 5.72924376e-01 2.66941726e-01 3.93560022e-01
-7.15070903e-01 1.40439654e-02 6.75715327e-01 1.72917709e-01
-4.46330756e-01 6.80783838e-02 -6.63062990e-01 9.83013570e-01
8.31071496e-01 -1.05003476e+00 2.08419666e-01 -6.24808431e-01
8.20003629e-01 -1.65860981e-01 -3.02015841e-01 -8.38305652e-01
2.77917475e-01 -3.42470318e-01 -8.87967870e-02 -3.50873888e-01
-4.16951627e-01 -1.55916765e-01 -3.00338775e-01 3.31822902e-01
-6.12560213e-01 5.81126869e-01 6.58927709e-02 3.98783445e-01
-6.48319274e-02 -9.16025162e-01 9.11354780e-01 -1.62592798e-01
-7.95846879e-01 -1.15172826e-01 -8.41215134e-01 5.04376948e-01
6.95298374e-01 -1.49884015e-01 -3.07773501e-01 -3.82564985e-03
-5.43536723e-01 -3.59697670e-01 8.18520010e-01 6.61007583e-01
1.40859962e-01 -1.11411226e+00 -1.14212942e+00 8.43518227e-02
5.55775642e-01 -6.47766352e-01 -2.53039926e-01 6.03830695e-01
-9.15345490e-01 4.23878461e-01 2.15074681e-02 -6.20228276e-02
-1.54582167e+00 4.38469976e-01 1.22270405e-01 -6.95113719e-01
-2.90096939e-01 8.61257732e-01 -4.21474069e-01 -1.10096872e+00
-1.17694348e-01 -2.88890660e-01 -5.74875295e-01 -9.25319493e-02
9.35174286e-01 4.34152991e-01 6.33999169e-01 -1.04515457e+00
-4.58924741e-01 4.57619488e-01 -3.59504730e-01 -2.16898665e-01
1.05083883e+00 -5.20140886e-01 -1.96562037e-01 7.17427731e-01
6.57798171e-01 1.07675683e+00 -2.31809855e-01 -5.43905258e-01
5.89849412e-01 -3.88166398e-01 -2.64012665e-01 -1.08578944e+00
-2.40724653e-01 6.76643968e-01 9.31210816e-02 -2.99706429e-01
5.47475278e-01 -4.82245564e-01 6.61428809e-01 6.34177208e-01
4.13309038e-01 -1.60313368e+00 -3.81750524e-01 1.59627211e+00
4.78821158e-01 -1.15848029e+00 -1.42376482e-01 -3.95578414e-01
-8.83452475e-01 1.18921590e+00 4.12888616e-01 -1.29824206e-01
2.29719892e-01 5.74092306e-02 3.51128012e-01 3.92801613e-01
-5.95223188e-01 -1.50649324e-01 4.29888278e-01 6.69676304e-01
9.30081785e-01 1.10679790e-01 -1.22400475e+00 5.92027485e-01
-5.03808320e-01 -8.39598924e-02 1.05675483e+00 6.81986868e-01
-4.62924093e-01 -1.77687633e+00 -2.95522988e-01 2.55629778e-01
-7.56516755e-01 -8.04322541e-01 -7.13309467e-01 3.00752342e-01
1.08214751e-01 9.87425804e-01 -1.24037296e-01 -3.36656541e-01
3.40408951e-01 2.23883614e-01 4.34457004e-01 -1.35043848e+00
-1.29139829e+00 -4.02901411e-01 5.17263889e-01 -1.32466629e-01
-2.15798169e-01 -1.09619617e+00 -1.00869143e+00 -2.11031243e-01
-5.00270903e-01 5.48275232e-01 6.37109995e-01 1.30606091e+00
-1.07324161e-02 5.98333851e-02 3.65435451e-01 -4.63879645e-01
-6.72007084e-01 -1.02821982e+00 -4.43555981e-01 4.68834579e-01
1.91687793e-01 -1.20778549e-02 -3.85673821e-01 8.87778029e-02] | [11.079379081726074, 10.68565559387207] |
645ca4f1-948a-4d44-8860-e903702662e4 | chinese-zero-pronoun-resolution-some-recent | null | null | https://aclanthology.org/D13-1135 | https://aclanthology.org/D13-1135.pdf | Chinese Zero Pronoun Resolution: Some Recent Advances | null | ['Vincent Ng', 'Chen Chen'] | 2013-10-01 | null | null | null | emnlp-2013-10 | ['chinese-zero-pronoun-resolution'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.387261390686035, 3.557269811630249] |
63bb773d-5402-49db-a5ec-9f0b72191e3d | attention-based-memory-video-portrait-matting | 2203.06890 | null | https://arxiv.org/abs/2203.06890v2 | https://arxiv.org/pdf/2203.06890v2.pdf | Attention based Memory video portrait matting | We proposed a novel trimap free video matting method based on the attention mechanism. By the nature of the problem, most existing approaches use either multiple computational expansive modules or complex algorithms to exploit temporal information fully. We designed a temporal aggregation module to compute the temporal coherence between the current frame and its two previous frames. | ['Shufeng Song'] | 2022-03-14 | null | null | null | null | ['image-matting', 'video-matting'] | ['computer-vision', 'computer-vision'] | [-4.55010869e-02 -1.56315431e-01 -1.45655394e-01 -2.16728389e-01
-2.34219730e-01 -8.45779330e-02 4.60910797e-01 -3.31033051e-01
-3.88233602e-01 7.18315363e-01 2.45540828e-01 -6.05324768e-02
2.59727966e-02 -7.93018401e-01 -5.54006636e-01 -5.47892869e-01
-2.14615613e-01 1.02662869e-01 5.22670746e-01 -1.39506608e-01
2.82776088e-01 2.37828225e-01 -1.53017449e+00 4.87360746e-01
8.98975313e-01 1.15298581e+00 5.44967294e-01 8.46918881e-01
-3.22813869e-01 1.00370276e+00 -4.50724959e-01 -3.49721551e-01
3.65923226e-01 -5.84776878e-01 -7.06238925e-01 4.29545701e-01
3.03406388e-01 -3.49312812e-01 -8.20139110e-01 7.92046010e-01
2.24522725e-01 4.70502019e-01 6.80823773e-02 -1.07444096e+00
-6.89516246e-01 6.53222620e-01 -6.30815566e-01 7.98940957e-01
5.73752701e-01 2.96299726e-01 7.37231553e-01 -1.05739403e+00
8.13809276e-01 1.05560911e+00 5.22843242e-01 3.75504613e-01
-9.88644004e-01 -3.00897032e-01 6.19306266e-01 8.14451396e-01
-1.22563148e+00 -5.00939250e-01 8.02817822e-01 -3.62793893e-01
9.96091902e-01 3.21743309e-01 1.10010362e+00 6.49765015e-01
3.32035750e-01 7.84795463e-01 1.03766334e+00 -2.08693624e-01
4.64573167e-02 -3.94970000e-01 4.94429395e-02 6.71463490e-01
-1.76655963e-01 1.85322631e-02 -7.08548784e-01 1.32513523e-01
9.21123862e-01 1.39075741e-01 -2.92170644e-01 1.94596837e-03
-1.33365536e+00 6.24576092e-01 3.33960235e-01 3.59246761e-01
-5.37912369e-01 3.61665130e-01 3.06216002e-01 4.08915907e-01
5.05922556e-01 9.73472297e-02 -1.11887105e-01 -2.94424921e-01
-1.28113866e+00 9.84023511e-02 4.95618582e-01 8.59159291e-01
8.60318720e-01 1.73578575e-01 -2.27071509e-01 1.84710562e-01
8.27929452e-02 2.34297607e-02 5.86277246e-01 -1.17273223e+00
4.05151516e-01 6.18337691e-01 2.63524428e-03 -1.17249513e+00
1.10542230e-01 3.38114388e-02 -8.14749062e-01 1.50596499e-01
-1.98922120e-02 -5.26047722e-02 -1.02000809e+00 1.52534175e+00
2.84994274e-01 5.91652870e-01 -4.00680959e-01 1.04870796e+00
6.97775543e-01 7.85582781e-01 -3.67275365e-02 -7.24888980e-01
7.74143696e-01 -1.24972785e+00 -1.13626587e+00 4.09646658e-03
1.08954020e-01 -7.53332913e-01 7.94356287e-01 2.28074685e-01
-1.70890832e+00 -6.38625324e-01 -1.29767084e+00 -1.18649587e-01
-2.34262571e-01 -1.84678540e-01 7.99145401e-01 3.27806979e-01
-1.20964170e+00 8.36618423e-01 -9.44125772e-01 -6.90658987e-02
2.29833812e-01 3.81702304e-01 -2.85089016e-01 -2.13641836e-03
-1.01211262e+00 7.98862100e-01 6.03018761e-01 2.13321462e-01
-7.94037759e-01 -3.63001823e-01 -7.44803369e-01 9.47710946e-02
5.09047031e-01 -8.70525599e-01 9.82227564e-01 -1.56232524e+00
-1.53946877e+00 4.95283872e-01 -4.05661106e-01 -4.97186035e-01
6.05979621e-01 -5.66589534e-01 -1.61873788e-01 3.31063479e-01
-1.49888009e-01 4.96185660e-01 1.28263187e+00 -7.94472396e-01
-3.95268500e-01 -6.93815947e-02 3.46756995e-01 4.88285005e-01
-2.72288442e-01 6.79264143e-02 -8.43277276e-01 -1.10933936e+00
2.58750737e-01 -6.30700111e-01 -4.62805957e-01 -5.16183600e-02
-3.39883864e-02 1.25196755e-01 1.09785736e+00 -8.30619931e-01
1.78635597e+00 -2.14907765e+00 7.44585991e-01 -1.99900299e-01
4.28482533e-01 2.80643970e-01 -1.43785384e-02 3.13758522e-01
-1.98530361e-01 5.98293580e-02 -1.35002509e-01 -5.13406515e-01
-3.95494431e-01 2.89361238e-01 -3.47660750e-01 3.03366005e-01
7.60155171e-02 9.16655898e-01 -9.45180953e-01 -8.31305325e-01
2.67619520e-01 3.36740345e-01 -5.91878474e-01 4.95722920e-01
-3.42708260e-01 3.36309344e-01 -1.25566259e-01 5.80964267e-01
6.09458923e-01 -1.18173517e-01 1.85299516e-01 -2.78159976e-01
-2.53525645e-01 2.16701627e-01 -1.06404448e+00 1.99017775e+00
-1.71932235e-01 6.17733836e-01 1.45952374e-01 -6.32866919e-01
4.74143386e-01 5.48848927e-01 8.40945363e-01 -3.56613874e-01
8.99834782e-02 -1.01096235e-01 -1.63916975e-01 -6.90461695e-01
8.65928054e-01 -4.98849340e-03 3.70602936e-01 4.64715213e-01
5.47242872e-02 1.11712337e-01 3.71586919e-01 4.30766761e-01
1.15017951e+00 4.47859824e-01 3.94347757e-01 -1.35736227e-01
5.32863498e-01 3.42216566e-02 9.01136339e-01 4.99083489e-01
-3.70078057e-01 8.23916435e-01 3.07226956e-01 -8.59086215e-01
-1.14908266e+00 -9.46259260e-01 4.07775700e-01 1.10455656e+00
3.35499734e-01 -8.54812264e-01 -6.71859205e-01 -6.10807717e-01
-3.92230600e-01 5.99774159e-02 -8.05471599e-01 2.33586272e-03
-7.83132255e-01 -6.26656950e-01 2.10724492e-02 6.09159768e-01
8.25776935e-01 -1.00852752e+00 -7.33851790e-01 3.86044711e-01
-4.79035348e-01 -8.43218744e-01 -1.06711435e+00 -2.17912540e-01
-1.03782034e+00 -9.10437167e-01 -6.69612765e-01 -5.59286892e-01
4.91413921e-01 5.81576288e-01 1.13769913e+00 4.10631865e-01
-1.35266811e-01 8.72649774e-02 -4.84703183e-01 -1.81124068e-03
7.08322674e-02 -2.32824713e-01 -6.98364247e-03 2.20429450e-01
1.05899684e-01 -8.11491430e-01 -6.37414396e-01 2.82556891e-01
-8.11928153e-01 5.93455911e-01 4.57809567e-01 7.10321009e-01
4.21292335e-01 2.16428395e-02 1.92440897e-01 -4.43214774e-01
4.35895979e-01 -4.18710381e-01 -3.26484323e-01 2.63100028e-01
-1.90716714e-01 5.91916516e-02 2.94152111e-01 -6.97744548e-01
-1.08172953e+00 9.67665911e-02 1.39398500e-01 -9.13910031e-01
2.48910099e-01 5.55881023e-01 -1.53987020e-01 -2.98342551e-03
1.55429661e-01 3.84691417e-01 -7.57898837e-02 -2.68636405e-01
2.68992037e-01 1.01593137e-01 6.38074934e-01 -5.84852040e-01
6.48816884e-01 6.44689500e-01 -2.54679203e-01 -5.87220132e-01
-7.06927299e-01 -2.56466776e-01 -8.40391338e-01 -5.60820937e-01
1.10661435e+00 -8.57427418e-01 -6.00228667e-01 4.89058167e-01
-1.40253913e+00 -2.60548443e-01 -2.67256856e-01 4.63982463e-01
-5.43657959e-01 7.88362086e-01 -7.58004367e-01 -6.86192572e-01
-3.90996844e-01 -1.05907810e+00 8.23195577e-01 1.20006867e-01
-1.63321495e-01 -7.95882404e-01 3.59265417e-01 9.54603851e-02
5.79751670e-01 2.42744073e-01 5.07436216e-01 2.65997976e-01
-1.05880594e+00 2.72103678e-02 -2.11767897e-01 1.69613142e-03
7.89037123e-02 2.75275141e-01 -6.91187561e-01 -2.11810887e-01
3.73005807e-01 7.84370229e-02 1.09946311e+00 3.83761257e-01
1.04045475e+00 -6.48251057e-01 -1.59210250e-01 8.35776567e-01
1.44815922e+00 2.70178199e-01 1.01572990e+00 2.15125412e-01
8.20215046e-01 1.61406860e-01 5.85873663e-01 5.39008379e-01
1.57535285e-01 8.35507393e-01 3.67964625e-01 4.67075370e-02
-6.57962412e-02 -5.30614657e-03 5.98712444e-01 1.16183114e+00
-6.01463735e-01 -2.25813031e-01 -4.71162677e-01 6.73104584e-01
-2.34570670e+00 -1.50555539e+00 1.07503913e-01 1.97290242e+00
6.88516676e-01 2.28881851e-01 1.18532039e-01 -2.80632302e-02
7.89074421e-01 3.38017315e-01 -4.07649249e-01 -5.10066152e-01
-1.84175923e-01 6.96863085e-02 2.47528851e-01 3.70397270e-01
-1.04285181e+00 8.15595627e-01 8.13743401e+00 8.39203537e-01
-9.34134662e-01 1.23779021e-01 4.86155242e-01 -3.48997921e-01
-2.93424517e-01 1.81465819e-01 -3.93098533e-01 6.67236447e-01
6.90659046e-01 -4.00662184e-01 3.97154272e-01 5.81002653e-01
3.52336526e-01 -3.42529088e-01 -9.13253903e-01 1.01951778e+00
2.01339573e-01 -1.43712533e+00 2.66200274e-01 -1.18030444e-01
7.99494088e-01 -3.09593618e-01 1.10254012e-01 -2.59431824e-02
1.24323182e-01 -9.78770375e-01 7.94130504e-01 7.98305511e-01
5.90284646e-01 -5.15784383e-01 5.23371756e-01 6.22344650e-02
-1.64598894e+00 2.18208302e-02 -2.10732937e-01 -4.65833306e-01
4.85016346e-01 4.74829435e-01 -2.12708712e-01 6.92243338e-01
7.57960856e-01 8.28156352e-01 -5.40946543e-01 1.22904205e+00
-3.82789448e-02 3.20074141e-01 -2.82678992e-01 4.39102173e-01
3.75466406e-01 -3.22854906e-01 9.33502078e-01 1.03385758e+00
4.13173944e-01 2.54577577e-01 3.56552213e-01 5.82216382e-01
1.34190410e-01 -5.89842722e-02 -6.12294972e-01 -6.06573410e-02
1.04732409e-01 1.19709909e+00 -6.67445898e-01 -7.37128854e-01
-7.10013211e-01 9.49250340e-01 2.93462217e-01 1.95070475e-01
-1.24940050e+00 -7.40337893e-02 6.74473941e-01 1.71114326e-01
4.99656320e-01 -6.36606038e-01 -3.92632335e-01 -1.48145056e+00
1.71303213e-01 -7.04380035e-01 6.73309684e-01 -9.20097828e-01
-9.01081979e-01 7.12583363e-01 -8.27836096e-02 -1.31105614e+00
5.78619279e-02 -2.07672417e-02 -1.01484084e+00 6.72663391e-01
-1.06209421e+00 -9.45641339e-01 -4.40359056e-01 7.69162476e-01
9.63405132e-01 -8.39185715e-02 4.26104695e-01 4.08021629e-01
-5.23003340e-01 2.32086495e-01 -3.64322752e-01 -5.20234331e-02
5.46696067e-01 -1.04606616e+00 2.93466896e-01 1.30355501e+00
2.62373835e-02 6.53080523e-01 7.42570758e-01 -7.48624325e-01
-1.24655044e+00 -8.01255226e-01 7.96511829e-01 -1.69621855e-01
5.97357452e-01 -3.27520631e-02 -1.05019236e+00 8.30404639e-01
7.99926758e-01 -2.30565324e-01 2.65358567e-01 -2.18291864e-01
-1.51362732e-01 9.99596249e-03 -8.10309350e-01 7.05507100e-01
1.07898605e+00 -3.91585320e-01 -8.75819504e-01 -1.02937505e-01
1.07511508e+00 -3.79199505e-01 -6.19544029e-01 3.75212938e-01
4.52327728e-01 -1.13521922e+00 7.49196589e-01 -4.50278312e-01
5.14073193e-01 -5.05607665e-01 -1.09692089e-01 -8.46492171e-01
-5.78814983e-01 -1.03651249e+00 -6.17181480e-01 9.41943169e-01
4.59038280e-02 -5.23958467e-02 8.55448961e-01 7.10519314e-01
-1.59756497e-01 -5.77006459e-01 -9.39621449e-01 -6.53101087e-01
-5.41665733e-01 -3.06635201e-01 3.86243343e-01 8.22509229e-01
2.50656635e-01 1.10081203e-01 -7.98324645e-01 -1.72346979e-02
4.02196020e-01 4.13331300e-01 4.64461803e-01 -6.54615223e-01
-4.95544672e-01 -4.20490086e-01 -4.69944566e-01 -9.59278107e-01
-2.26711124e-01 -4.69590753e-01 -9.30666178e-02 -1.21528900e+00
5.43527842e-01 4.89342436e-02 -3.66015971e-01 3.82483095e-01
-4.30172265e-01 2.48972028e-01 3.78164560e-01 2.57875085e-01
-9.42003071e-01 6.28032207e-01 1.13520241e+00 -1.78838611e-01
-2.90983379e-01 -3.44850242e-01 -2.55825400e-01 6.38889670e-01
7.07895219e-01 -3.16697270e-01 -3.87338161e-01 -6.08321011e-01
1.63993433e-01 2.06709936e-01 2.46573359e-01 -1.22962773e+00
2.93403774e-01 -5.60268223e-01 1.84360057e-01 -7.85864115e-01
5.52269816e-01 -7.27223754e-01 4.92045432e-01 4.32982922e-01
-1.74323335e-01 7.86950111e-01 2.11526334e-01 6.40861154e-01
-4.95991111e-01 3.57080400e-02 8.29380989e-01 -4.64245468e-01
-1.10849345e+00 6.52788937e-01 -5.96055567e-01 -1.59157515e-01
1.23961961e+00 -2.17527151e-01 -7.17791766e-02 -4.91593093e-01
-1.04029977e+00 1.25159189e-01 6.72749162e-01 3.09527993e-01
9.59598243e-01 -1.50734782e+00 -5.27830780e-01 8.64056498e-02
-4.26041156e-01 -1.18208900e-01 3.68238777e-01 1.03217125e+00
-7.56327450e-01 1.04742005e-01 -6.56488478e-01 -3.44796211e-01
-1.48721266e+00 1.15015435e+00 2.32630908e-01 -3.67801964e-01
-6.82932556e-01 7.29540944e-01 3.33484173e-01 5.73744833e-01
1.28849745e-01 -6.65747449e-02 -1.23072885e-01 1.10468015e-01
6.98557854e-01 3.94904971e-01 -2.61659056e-01 -5.23370445e-01
-9.57135558e-02 5.10095298e-01 -1.44471943e-01 -3.84575069e-01
1.18787396e+00 -4.95224297e-01 -3.22527111e-01 5.06565630e-01
8.55708957e-01 -3.45082939e-01 -1.45938408e+00 -1.68025538e-01
-1.18893780e-01 -9.18152988e-01 -7.42111206e-02 -1.76448226e-01
-1.19609213e+00 6.22774422e-01 1.87449202e-01 1.71379849e-01
1.43179274e+00 -3.30134362e-01 9.67130363e-01 -1.03021845e-01
2.06040934e-01 -1.28186417e+00 3.50317806e-01 5.39004862e-01
9.17097986e-01 -9.61122096e-01 3.01993281e-01 -4.12934273e-01
-5.63569725e-01 9.79210019e-01 8.55100870e-01 -2.20825866e-01
5.85298717e-01 3.75423968e-01 -1.75306603e-01 -1.57823488e-01
-1.32065332e+00 -2.78115809e-01 2.97338665e-01 4.14255410e-01
4.04131830e-01 -4.46939707e-01 -5.59117317e-01 1.82512313e-01
2.83049732e-01 1.47078857e-01 4.92325395e-01 1.07003367e+00
-5.28878093e-01 -9.82554436e-01 -3.32639158e-01 1.31709039e-01
-4.76832330e-01 -8.85188878e-02 -2.51949668e-01 4.44848508e-01
1.91575050e-01 7.39281833e-01 1.62077308e-01 -6.04600370e-01
-1.81236729e-01 2.25386117e-03 5.14757574e-01 -4.01336163e-01
-5.78716636e-01 2.54236817e-01 -6.27506822e-02 -8.79372120e-01
-8.62569153e-01 -7.16570079e-01 -8.37434292e-01 -6.43978953e-01
-2.20685035e-01 -1.12886585e-01 5.13181835e-02 8.28684330e-01
3.19980353e-01 5.13173103e-01 5.72662592e-01 -1.21564901e+00
1.76387399e-01 -7.74950981e-01 -3.12068939e-01 4.91405249e-01
1.91560268e-01 -6.30859673e-01 -8.75700936e-02 3.76283348e-01] | [10.554170608520508, -0.9342142343521118] |
24424c0b-0bc4-4627-a6da-489a27be45c4 | ca-net-comprehensive-attention-convolutional | 2009.10549 | null | https://arxiv.org/abs/2009.10549v2 | https://arxiv.org/pdf/2009.10549v2.pdf | CA-Net: Comprehensive Attention Convolutional Neural Networks for Explainable Medical Image Segmentation | Accurate medical image segmentation is essential for diagnosis and treatment planning of diseases. Convolutional Neural Networks (CNNs) have achieved state-of-the-art performance for automatic medical image segmentation. However, they are still challenged by complicated conditions where the segmentation target has large variations of position, shape and scale, and existing CNNs have a poor explainability that limits their application to clinical decisions. In this work, we make extensive use of multiple attentions in a CNN architecture and propose a comprehensive attention-based CNN (CA-Net) for more accurate and explainable medical image segmentation that is aware of the most important spatial positions, channels and scales at the same time. In particular, we first propose a joint spatial attention module to make the network focus more on the foreground region. Then, a novel channel attention module is proposed to adaptively recalibrate channel-wise feature responses and highlight the most relevant feature channels. Also, we propose a scale attention module implicitly emphasizing the most salient feature maps among multiple scales so that the CNN is adaptive to the size of an object. Extensive experiments on skin lesion segmentation from ISIC 2018 and multi-class segmentation of fetal MRI found that our proposed CA-Net significantly improved the average segmentation Dice score from 87.77% to 92.08% for skin lesion, 84.79% to 87.08% for the placenta and 93.20% to 95.88% for the fetal brain respectively compared with U-Net. It reduced the model size to around 15 times smaller with close or even better accuracy compared with state-of-the-art DeepLabv3+. In addition, it has a much higher explainability than existing networks by visualizing the attention weight maps. Our code is available at https://github.com/HiLab-git/CA-Net | ['Sébastien Ourselin', 'Tao Song', 'Ran Gu', 'Tom Vercauteren', 'Rui Huang', 'Michael Aertsen', 'Guotai Wang', 'Shaoting Zhang', 'Jan Deprest'] | 2020-09-22 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 9.20023769e-02 2.64430493e-01 2.56724600e-02 -4.94146585e-01
-5.02346396e-01 -3.37518394e-01 -2.50503211e-03 1.84654564e-01
-3.32550794e-01 4.02799368e-01 6.20018877e-02 -3.30773950e-01
-7.87462071e-02 -6.06142521e-01 -6.84453189e-01 -6.92622006e-01
8.47731680e-02 3.51007760e-01 4.67461795e-01 -6.26450852e-02
1.03826910e-01 5.04086077e-01 -8.04282129e-01 3.61375511e-01
1.26592982e+00 9.90082681e-01 2.92519957e-01 5.29626012e-01
-2.36699089e-01 3.61047953e-01 -3.54995370e-01 -1.99125558e-01
6.06043525e-02 -5.57494104e-01 -8.60038877e-01 -4.94740009e-02
3.22171718e-01 -2.19550207e-01 -2.29290739e-01 1.23359239e+00
6.21317744e-01 -8.53610188e-02 6.67963326e-01 -9.51899707e-01
-8.17402005e-01 6.86743438e-01 -1.05941117e+00 4.83108640e-01
-3.81431222e-01 2.15964019e-02 5.35748482e-01 -4.91808534e-01
3.54436427e-01 1.02259672e+00 5.68019807e-01 7.04541981e-01
-8.25779676e-01 -8.94982159e-01 2.00436786e-01 3.02101612e-01
-1.29586411e+00 1.30838290e-01 6.66254699e-01 -3.02037686e-01
6.88211858e-01 3.80345434e-01 7.36342430e-01 5.03204525e-01
3.12941760e-01 7.53441811e-01 8.12240243e-01 -7.34538436e-02
-6.83183372e-02 -3.04160386e-01 8.23117420e-02 9.47292447e-01
3.17743301e-01 -4.28374797e-01 1.32815555e-01 2.39334702e-01
1.17247427e+00 2.72951365e-01 -3.33702385e-01 -1.06141724e-01
-1.27452230e+00 7.04031646e-01 1.13565254e+00 6.62397087e-01
-4.78643328e-01 2.64945358e-01 3.84024799e-01 -2.88003564e-01
4.36000764e-01 4.57866818e-01 -5.95614672e-01 1.06458969e-01
-7.83486903e-01 -7.87688792e-02 2.64239758e-01 6.29148066e-01
5.55524409e-01 -5.46580628e-02 -4.35936898e-01 8.40440631e-01
2.52900064e-01 2.56496161e-01 6.40961409e-01 -8.56468201e-01
3.81539345e-01 7.61844873e-01 -2.89949387e-01 -1.01844656e+00
-9.43611681e-01 -7.41539657e-01 -1.21079397e+00 -9.54883918e-02
2.65841097e-01 -2.03644171e-01 -1.45004499e+00 1.60768545e+00
5.09383023e-01 3.75917256e-01 -2.07499415e-01 1.12740386e+00
1.05287695e+00 4.95839000e-01 2.10972250e-01 -8.81542414e-02
1.50670207e+00 -1.17105889e+00 -8.19792151e-01 -1.38723701e-01
4.97556627e-01 -6.41880929e-01 9.19520676e-01 9.54177827e-02
-1.18457150e+00 -4.78595793e-01 -9.24942851e-01 -8.97415429e-02
-2.47127652e-01 9.36140269e-02 7.47421861e-01 5.35523474e-01
-9.60743368e-01 5.30655503e-01 -1.10968745e+00 -2.64868617e-01
9.40740287e-01 5.85835278e-01 -1.11501575e-01 -2.60680057e-02
-1.04995227e+00 6.68046057e-01 2.64059275e-01 3.00744921e-01
-6.22868240e-01 -9.38797355e-01 -7.26545215e-01 2.30284244e-01
1.18947051e-01 -6.01071775e-01 1.02803588e+00 -1.13118064e+00
-1.26847017e+00 6.79275990e-01 -3.32086869e-02 -2.79023528e-01
5.07486761e-01 5.25188744e-02 -2.03626931e-01 4.30666804e-01
1.64284363e-01 1.01794446e+00 4.20106620e-01 -1.03554809e+00
-5.19500136e-01 -4.58070606e-01 -6.08107746e-02 2.02759176e-01
-3.14590037e-01 -3.13879810e-02 -8.65397394e-01 -7.22194135e-01
4.83330280e-01 -8.85999501e-01 -4.11304414e-01 1.61810115e-01
-4.78557318e-01 -4.20763083e-02 6.83779061e-01 -8.31592500e-01
1.22807634e+00 -2.02827597e+00 3.90172340e-02 9.20941159e-02
4.76862580e-01 4.93965745e-01 -8.55527371e-02 -3.08843225e-01
-2.26949960e-01 4.78098571e-01 -4.86430526e-01 -3.29049341e-02
-4.62976784e-01 2.70268116e-02 4.32525456e-01 5.65087080e-01
1.92996830e-01 1.07536566e+00 -7.76477814e-01 -6.76397860e-01
3.61251414e-01 7.42506981e-01 -7.23725379e-01 8.92377198e-02
1.62697703e-01 7.79507458e-01 -5.98732531e-01 7.08271623e-01
8.37823927e-01 -4.75881159e-01 -1.52133824e-02 -4.11371320e-01
-2.25445628e-02 -2.18554035e-01 -8.52038801e-01 1.79653144e+00
-3.58168393e-01 5.89091599e-01 8.66606906e-02 -1.10160506e+00
6.85063303e-01 3.47560585e-01 7.73543835e-01 -6.64539039e-01
5.06672084e-01 2.60122269e-01 6.26230836e-01 -6.33830249e-01
-4.80592996e-02 -6.57605454e-02 2.05235362e-01 4.53748479e-02
-1.05365910e-01 1.73028618e-01 -1.80961993e-02 2.11852118e-01
8.64462137e-01 -3.02350134e-01 2.07801431e-01 -4.93482530e-01
5.58784425e-01 -2.89480358e-01 7.39655852e-01 5.65900922e-01
-4.20534521e-01 1.05515492e+00 4.93029088e-01 -5.53164065e-01
-8.08976591e-01 -7.01913118e-01 -3.42550486e-01 7.77204633e-01
4.59164888e-01 4.49573398e-02 -1.10818946e+00 -7.61663496e-01
-1.11220755e-01 3.10066104e-01 -1.13475597e+00 -5.65162152e-02
-6.37887120e-01 -8.31835628e-01 4.45173085e-01 8.25452268e-01
7.47848451e-01 -1.25391400e+00 -6.23942316e-01 1.78006589e-01
-1.67967036e-01 -8.19488287e-01 -7.04245031e-01 -8.46675113e-02
-8.53895247e-01 -1.08251905e+00 -1.15769422e+00 -8.29797626e-01
1.11034787e+00 7.22885802e-02 7.43567109e-01 4.95289803e-01
-5.45118332e-01 -1.65644959e-01 -3.35347801e-01 -5.20301580e-01
-4.67356294e-02 3.33626300e-01 -4.70536023e-01 3.91481966e-02
-1.44094914e-01 -3.71572733e-01 -1.11689532e+00 1.78735495e-01
-9.93661940e-01 4.26689118e-01 6.91598058e-01 8.51584733e-01
6.92205310e-01 -2.39796266e-01 5.41923106e-01 -1.11073256e+00
3.26914012e-01 -5.70319772e-01 -2.43014842e-01 3.00974458e-01
-5.57375491e-01 -2.41532013e-01 4.41437870e-01 -3.90422225e-01
-9.18220222e-01 1.21901603e-02 -2.72528023e-01 -4.15637702e-01
-1.24008551e-01 2.69159645e-01 -3.89280021e-02 -1.32599756e-01
3.44818801e-01 1.01224454e-02 -7.02751055e-02 -3.16560864e-01
6.76107928e-02 5.23179173e-01 3.36130142e-01 -2.68735647e-01
3.11385691e-01 5.10215104e-01 -4.47537191e-02 -4.18365568e-01
-7.02407360e-01 -2.87770063e-01 -7.14118302e-01 -1.64008602e-01
1.25273633e+00 -5.11394739e-01 -6.93031967e-01 5.25946498e-01
-1.14940715e+00 -3.90042365e-01 1.06841907e-01 4.00683880e-01
-1.24768652e-01 3.10447186e-01 -7.77924836e-01 -2.32141763e-01
-7.86316335e-01 -1.61880267e+00 8.63286734e-01 7.42704511e-01
6.61556562e-03 -1.01098549e+00 -3.85329574e-01 3.89412344e-01
7.10367143e-01 5.01796782e-01 9.03579235e-01 -6.15016997e-01
-3.24189186e-01 -2.15246738e-03 -8.12363386e-01 5.05887270e-02
4.12941813e-01 7.53933266e-02 -7.10439503e-01 -1.82690933e-01
-4.22216952e-01 1.68795839e-01 9.15223658e-01 1.03088284e+00
1.82642674e+00 -7.00009391e-02 -4.53990191e-01 1.09910429e+00
1.35796952e+00 4.75331813e-01 5.43052018e-01 2.31684521e-01
1.00146282e+00 3.72974664e-01 4.09711868e-01 2.00454339e-01
4.06130493e-01 4.29103643e-01 7.75896490e-01 -7.96824336e-01
-3.37445825e-01 2.29294419e-01 -4.53810960e-01 7.24387765e-01
-9.90510434e-02 -1.41343847e-01 -1.04245424e+00 7.81061232e-01
-1.78092694e+00 -4.30749208e-01 -2.93739051e-01 1.81728196e+00
9.12116110e-01 -2.16668304e-02 -1.31860211e-01 -2.45672524e-01
9.19659138e-01 -7.77031183e-02 -7.58243203e-01 -4.46376681e-01
2.73733884e-01 2.90375113e-01 6.93654835e-01 4.67705458e-01
-1.15883172e+00 8.57210159e-01 5.20730543e+00 8.52078438e-01
-1.42435324e+00 3.31481606e-01 1.17574668e+00 2.78898124e-02
-2.02117652e-01 -3.25185061e-01 -3.64661366e-01 6.37186468e-01
3.82105947e-01 1.18978381e-01 2.34871984e-01 6.11358583e-01
2.17757642e-01 9.18037668e-02 -5.78353286e-01 8.24012756e-01
-1.12904653e-01 -1.43521833e+00 5.16079180e-03 -1.06383435e-01
9.57663178e-01 1.51238009e-01 9.79987681e-02 3.06555834e-02
-1.81076586e-01 -1.24721456e+00 4.76988047e-01 3.12336087e-01
8.78990471e-01 -8.15544844e-01 9.53829944e-01 1.00229196e-01
-1.15136075e+00 1.06235497e-01 -3.41420233e-01 2.94573456e-01
1.08677670e-02 4.14076716e-01 -6.87289596e-01 3.70531350e-01
9.71036375e-01 6.28916502e-01 -6.02845788e-01 1.24816597e+00
-9.34867859e-02 5.60146868e-01 -1.32816359e-01 1.38528692e-02
5.10521173e-01 6.89223930e-02 1.25577062e-01 1.43984795e+00
2.69798100e-01 4.75440592e-01 8.46726075e-03 8.14916968e-01
-1.99676514e-01 3.23166788e-01 -1.82425193e-02 4.17056143e-01
3.14072490e-01 1.55766535e+00 -1.26554275e+00 -5.31397343e-01
-3.00965011e-01 9.67595100e-01 2.43178621e-01 3.39691043e-01
-1.14569020e+00 -5.87123632e-01 3.81959885e-01 6.71042968e-03
2.98014253e-01 2.19330251e-01 -6.28837168e-01 -9.22514737e-01
-3.39599639e-01 -5.67846477e-01 3.66685778e-01 -6.99961722e-01
-9.59431350e-01 8.44859660e-01 -2.23411515e-01 -8.95413220e-01
4.68250811e-01 -4.62805927e-01 -8.44118536e-01 9.23018813e-01
-1.58100498e+00 -1.11005604e+00 -6.66746795e-01 5.22253633e-01
6.35780573e-01 1.38677895e-01 5.13110638e-01 4.92356181e-01
-7.73878872e-01 7.22154379e-01 -6.78513944e-02 3.47165823e-01
5.92808008e-01 -1.34325910e+00 2.37700552e-01 6.98606908e-01
-2.18525887e-01 6.96936905e-01 3.56366277e-01 -6.09016418e-01
-8.27653408e-01 -1.15419185e+00 3.52611750e-01 6.54772148e-02
3.97143960e-01 2.31100786e-02 -1.07689989e+00 5.20048559e-01
3.69083285e-01 5.35849333e-01 4.36457217e-01 -2.16026127e-01
8.97176415e-02 -2.06143752e-01 -1.31932116e+00 4.04544950e-01
6.76017880e-01 -8.85822531e-03 -3.78037989e-02 3.04864317e-01
7.74541616e-01 -9.16564465e-01 -9.46866333e-01 7.08185971e-01
6.35624170e-01 -8.06167781e-01 8.17372859e-01 -3.94857883e-01
6.24783397e-01 -3.02141428e-01 3.52265358e-01 -1.14323473e+00
-5.56362808e-01 -2.01414019e-01 1.94969654e-01 9.31497335e-01
5.58697164e-01 -7.27819085e-01 6.67034984e-01 6.52592003e-01
-4.80210632e-01 -1.35107028e+00 -8.43078554e-01 -8.06112587e-02
3.16607714e-01 -1.92078829e-01 6.22276187e-01 1.10301685e+00
-3.85420710e-01 -4.47188057e-02 -7.96269178e-02 2.43655428e-01
4.41818208e-01 2.90880408e-02 2.30730653e-01 -8.62085283e-01
9.83292237e-03 -6.79481864e-01 -2.83221394e-01 -7.60520756e-01
-2.44888350e-01 -9.08366799e-01 -7.29145631e-02 -1.96861744e+00
5.50678372e-01 -4.62492734e-01 -7.37941206e-01 7.72327006e-01
-5.31126499e-01 5.27441144e-01 1.97652355e-01 6.05862848e-02
-4.86935586e-01 1.95054710e-01 1.85105705e+00 -2.29600891e-01
-1.48445308e-01 -1.01583920e-01 -9.86170053e-01 7.03521729e-01
1.00349891e+00 -3.76997620e-01 -2.25379944e-01 -7.44544685e-01
-2.65110552e-01 3.89239825e-02 2.25357533e-01 -8.61381590e-01
1.66683763e-01 -1.13456748e-01 7.04784393e-01 -4.16838288e-01
-1.34519458e-01 -6.98974788e-01 -1.05846576e-01 7.87084520e-01
-4.33761626e-01 -6.17139190e-02 4.24976945e-01 2.21291825e-01
-8.77326354e-02 -5.77376299e-02 1.02130544e+00 -2.02948079e-01
-4.98417556e-01 6.72213972e-01 -1.13635853e-01 -2.13124510e-02
1.11479723e+00 -5.88079020e-02 -2.11244091e-01 -1.56818375e-01
-7.61588693e-01 3.70678604e-01 1.65780708e-01 3.37542951e-01
5.57184696e-01 -1.15814674e+00 -7.08510756e-01 8.21794719e-02
-1.39190301e-01 3.08807820e-01 7.46429265e-01 1.15950167e+00
-9.44988072e-01 3.75302941e-01 -3.02946299e-01 -7.03711331e-01
-1.25688410e+00 2.42028132e-01 6.29642606e-01 -1.65911764e-01
-6.45832181e-01 1.09062076e+00 6.98617041e-01 -2.69636780e-01
3.90325822e-02 -7.77979374e-01 -4.14101064e-01 -3.00782084e-01
3.74316275e-01 2.26089105e-01 2.19802428e-02 -7.08204567e-01
-5.88671744e-01 8.80658567e-01 -1.56653643e-01 4.61217463e-01
1.32391667e+00 -5.31843565e-02 -1.87664047e-01 -1.33435443e-01
1.21211565e+00 -2.57771909e-01 -1.42761433e+00 -2.90730987e-02
-4.87808585e-01 -3.51822883e-01 2.30980501e-01 -1.12973714e+00
-1.87186038e+00 1.13117826e+00 9.25226986e-01 1.05929025e-01
1.33386004e+00 -1.50405159e-02 9.53027606e-01 -4.07509446e-01
-1.32190958e-01 -7.53229201e-01 -1.92773144e-03 1.49554163e-01
9.09515977e-01 -1.34221423e+00 7.55236968e-02 -5.28183579e-01
-7.13974953e-01 1.14066601e+00 9.79074299e-01 -1.33713037e-01
6.60546005e-01 1.59398884e-01 1.95346087e-01 -2.79405624e-01
-2.26750731e-01 -7.90999550e-03 6.27770543e-01 2.95097262e-01
7.04983532e-01 2.33348593e-01 -4.90816087e-01 8.49666119e-01
1.48478672e-01 -2.91035742e-01 3.08953822e-01 4.61926490e-01
-4.17026997e-01 -6.17777467e-01 -1.63408905e-01 5.09449482e-01
-9.26927269e-01 -1.83535323e-01 2.50503849e-02 7.70047188e-01
4.41709548e-01 5.73621333e-01 1.02078207e-01 -1.44516975e-01
1.63049638e-01 -4.69149351e-01 2.83439070e-01 -5.45584798e-01
-7.64856219e-01 3.27367127e-01 -5.13659060e-01 -5.86535096e-01
-3.65113527e-01 -4.96084034e-01 -1.93314886e+00 -4.35689948e-02
-3.16345006e-01 2.29758001e-03 7.58075476e-01 9.01257634e-01
3.84849787e-01 1.18765938e+00 4.32232380e-01 -7.67394781e-01
9.68503803e-02 -1.06163490e+00 -4.26658720e-01 3.45794797e-01
3.01921129e-01 -4.87129718e-01 -1.09777581e-02 -9.66135859e-02] | [14.582379341125488, -2.5332541465759277] |
f63e9de8-83e3-4ca2-b917-c7c7295fe340 | kit-multi-a-translation-oriented-multilingual | null | null | https://aclanthology.org/L18-1616 | https://aclanthology.org/L18-1616.pdf | KIT-Multi: A Translation-Oriented Multilingual Embedding Corpus | null | ['er', 'Thanh-Le Ha', 'Ngoc Quan Pham', 'Alex Waibel', 'Matthias Sperber', 'Jan Niehues'] | 2018-05-01 | kit-multi-a-translation-oriented-multilingual-1 | https://aclanthology.org/L18-1616 | https://aclanthology.org/L18-1616.pdf | lrec-2018-5 | ['multilingual-word-embeddings', 'cross-lingual-document-classification'] | ['methodology', 'natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.105790615081787, 3.5335988998413086] |
308cd02e-06af-4032-a770-baecb61b64c8 | removing-objects-from-neural-radiance-fields | 2212.11966 | null | https://arxiv.org/abs/2212.11966v1 | https://arxiv.org/pdf/2212.11966v1.pdf | Removing Objects From Neural Radiance Fields | Neural Radiance Fields (NeRFs) are emerging as a ubiquitous scene representation that allows for novel view synthesis. Increasingly, NeRFs will be shareable with other people. Before sharing a NeRF, though, it might be desirable to remove personal information or unsightly objects. Such removal is not easily achieved with the current NeRF editing frameworks. We propose a framework to remove objects from a NeRF representation created from an RGB-D sequence. Our NeRF inpainting method leverages recent work in 2D image inpainting and is guided by a user-provided mask. Our algorithm is underpinned by a confidence based view selection procedure. It chooses which of the individual 2D inpainted images to use in the creation of the NeRF, so that the resulting inpainted NeRF is 3D consistent. We show that our method for NeRF editing is effective for synthesizing plausible inpaintings in a multi-view coherent manner. We validate our approach using a new and still-challenging dataset for the task of NeRF inpainting. | ['Sara Vicente', 'Michael Firman', 'Gabriel Brostow', 'Marc Pollefeys', 'Aron Monszpart', 'Guillermo Garcia-Hernando', 'Silvan Weder'] | 2022-12-22 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Weder_Removing_Objects_From_Neural_Radiance_Fields_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Weder_Removing_Objects_From_Neural_Radiance_Fields_CVPR_2023_paper.pdf | cvpr-2023-1 | ['image-inpainting'] | ['computer-vision'] | [ 6.88505948e-01 1.17049403e-01 2.36614466e-01 -4.86376762e-01
-8.15933287e-01 -7.31230319e-01 6.05739772e-01 -2.77890623e-01
-1.61242634e-01 7.90035129e-01 6.52172446e-01 8.58465806e-02
-8.42771158e-02 -7.72852778e-01 -9.27790761e-01 -4.20549870e-01
7.23373532e-01 5.40976748e-02 -8.69130269e-02 -3.25117618e-01
2.44148925e-01 8.37305069e-01 -1.69025457e+00 3.72437060e-01
7.47159064e-01 8.96726668e-01 4.07339156e-01 7.05118239e-01
4.80942130e-02 8.06965947e-01 -5.39776683e-01 -5.24349809e-01
8.02608252e-01 -5.49103856e-01 -5.75851798e-01 2.85032272e-01
8.89088631e-01 -6.23440087e-01 -2.10952178e-01 7.97990441e-01
3.18787247e-01 3.94025505e-01 5.67163467e-01 -9.33476031e-01
-4.51905906e-01 1.13998421e-01 -6.03758574e-01 -2.16243714e-01
6.69803560e-01 3.95975523e-02 6.81093156e-01 -9.37466681e-01
1.14698815e+00 1.11957800e+00 6.38942420e-01 4.96308833e-01
-1.34198821e+00 -3.14705163e-01 1.30195275e-01 -3.81809115e-01
-1.22425020e+00 -6.73074961e-01 1.03042114e+00 -2.07293168e-01
7.92107165e-01 7.75801659e-01 9.97282565e-01 1.15522492e+00
3.96568745e-01 4.57804441e-01 1.24274695e+00 -5.07112980e-01
3.19884479e-01 5.62998652e-02 -5.87980330e-01 4.52594727e-01
-3.97648960e-02 8.25923532e-02 -9.20312583e-01 -1.82382986e-01
1.16535056e+00 1.78299174e-01 -4.99518871e-01 -4.39992011e-01
-1.21683955e+00 6.33255720e-01 4.41376507e-01 -1.68900222e-01
-6.72539115e-01 2.70392030e-01 -1.42816365e-01 2.86820143e-01
7.94305444e-01 6.16607606e-01 -3.57472420e-01 -1.14296526e-02
-1.23423815e+00 4.83775944e-01 4.06307369e-01 7.85536587e-01
6.31367564e-01 -2.20408157e-01 -8.33917558e-02 9.16087151e-01
5.58378398e-02 2.94189185e-01 -9.38525051e-02 -1.61489224e+00
3.88455689e-02 3.15402657e-01 4.12872434e-01 -9.17638779e-01
-2.34599225e-02 -6.45555779e-02 -6.23386681e-01 7.77120888e-01
-2.38685962e-02 7.64938518e-02 -9.99166489e-01 1.54762530e+00
5.24178386e-01 5.90266241e-03 -1.09871127e-01 9.15240943e-01
6.23975396e-01 6.13438010e-01 -3.48938316e-01 -2.38244124e-02
1.09399581e+00 -7.76053309e-01 -6.53592706e-01 -2.69629270e-01
4.03074967e-03 -1.00581014e+00 8.77113044e-01 7.46044040e-01
-1.25075567e+00 -2.04910174e-01 -9.17672038e-01 -6.44910336e-01
-9.63735394e-03 -2.49290064e-01 5.95247865e-01 3.82219940e-01
-1.04469764e+00 6.94812179e-01 -4.35531646e-01 -3.84239018e-01
4.44352239e-01 -4.52052653e-02 -7.49478519e-01 -3.61701429e-01
-8.81998718e-01 9.87276196e-01 1.77694842e-01 -5.35844825e-02
-6.74775660e-01 -8.33236396e-01 -8.12680840e-01 -2.86472261e-01
3.44322473e-01 -1.25871408e+00 1.23308623e+00 -1.07072103e+00
-1.42711830e+00 8.95382226e-01 -1.20343357e-01 -2.31300473e-01
8.82711709e-01 -4.63665336e-01 -1.55873835e-01 2.14214832e-01
1.74026459e-01 1.11119270e+00 1.36566436e+00 -1.68121231e+00
-4.27943021e-01 -1.02785848e-01 2.97854573e-01 7.81200767e-01
1.19266957e-01 -2.56188571e-01 -4.35252130e-01 -1.12182379e+00
2.26586431e-01 -7.03450322e-01 -2.39955127e-01 7.01729894e-01
-4.26849157e-01 5.07250071e-01 8.38753581e-01 -8.83690834e-01
7.42548645e-01 -2.21173620e+00 1.71316355e-01 2.05212504e-01
3.28135490e-01 -2.16267675e-01 6.21507019e-02 3.98566991e-01
-9.38307568e-02 -5.60172275e-02 -5.55770040e-01 -6.99355960e-01
-1.37713671e-01 2.35681161e-01 -4.89062607e-01 3.60812545e-01
8.76829848e-02 4.88942474e-01 -9.59803164e-01 -2.04826996e-01
5.40376902e-01 8.84567618e-01 -7.72963822e-01 3.90229195e-01
-3.57314289e-01 6.88932955e-01 -1.11628495e-01 5.19058108e-01
8.75363648e-01 3.35519835e-02 -2.64926374e-01 -5.12458205e-01
-2.42115617e-01 1.28151283e-01 -1.24264526e+00 2.22603941e+00
-6.43980980e-01 6.31770134e-01 2.18848273e-01 -1.64907724e-01
8.35773826e-01 3.21621560e-02 5.03714025e-01 -4.21286672e-01
-3.69178876e-02 7.28653073e-02 -6.27043903e-01 -1.13915786e-01
9.77046907e-01 -3.73999715e-01 6.85191080e-02 6.40205324e-01
-2.47609634e-02 -8.47527206e-01 -1.24272488e-01 3.50004315e-01
1.05109298e+00 6.73602879e-01 4.01400685e-01 -3.87409627e-02
-4.43528108e-02 -1.64521098e-01 3.41209173e-01 8.50447476e-01
1.01804785e-01 1.47505248e+00 -8.81674215e-02 -5.03670633e-01
-1.16333413e+00 -1.25226986e+00 -9.84148756e-02 5.90689301e-01
5.73522002e-02 -4.57308322e-01 -8.06513608e-01 -6.47442758e-01
-1.33712977e-01 9.70990181e-01 -6.09604001e-01 3.61690857e-02
-3.35434943e-01 -1.79418474e-01 6.10039793e-02 1.77778333e-01
4.99975175e-01 -9.35475528e-01 -1.04938638e+00 1.73479944e-01
-3.25894415e-01 -1.04101753e+00 -7.09781885e-01 2.56049305e-01
-8.19132090e-01 -7.85339653e-01 -8.90976191e-01 -3.47760230e-01
8.08463097e-01 6.51521385e-01 1.28377318e+00 -2.99998689e-02
-2.40536749e-01 7.08646834e-01 -3.07134449e-01 -3.43752325e-01
-4.91500735e-01 -3.17762643e-01 -2.10629001e-01 1.63749475e-02
-2.02537715e-01 -8.44830573e-01 -9.28878725e-01 -4.17444557e-02
-1.23715854e+00 7.60468364e-01 3.21097642e-01 4.91009206e-01
7.79824138e-01 -6.84123859e-02 3.68849039e-02 -9.65297639e-01
4.87631530e-01 -2.52425402e-01 -4.07393575e-01 8.28296468e-02
-3.26896042e-01 1.33294195e-01 4.87733752e-01 -1.65871441e-01
-1.44718754e+00 2.56050766e-01 -2.01146677e-01 -6.50227427e-01
-3.04866463e-01 1.06234387e-01 -3.56316380e-02 -2.08131388e-01
6.61257446e-01 -3.31677571e-02 -2.77938813e-01 -4.88561571e-01
7.48273194e-01 3.37392032e-01 6.31908596e-01 -4.01047349e-01
8.14175546e-01 8.20099175e-01 -2.46565361e-02 -7.51191854e-01
-8.33898306e-01 -3.26596238e-02 -5.70641220e-01 -4.32170898e-01
9.02149796e-01 -1.04426670e+00 -1.08112507e-01 3.41147542e-01
-1.30475402e+00 -2.68231720e-01 -6.59034073e-01 2.56742179e-01
-6.35115802e-01 2.24709854e-01 -2.45579883e-01 -6.80724561e-01
-1.92429051e-01 -1.04792237e+00 1.44608378e+00 2.17416182e-01
-4.62489277e-01 -6.72679842e-01 1.62840679e-01 5.23041904e-01
3.43209177e-01 7.22815394e-01 6.87239230e-01 2.73557931e-01
-8.25683296e-01 1.23533741e-01 -8.59618858e-02 3.75666171e-01
2.71899223e-01 2.19120737e-02 -1.21343279e+00 -9.50845852e-02
2.25467339e-01 -2.09115818e-01 8.51053059e-01 3.49343032e-01
1.06386733e+00 -3.12944233e-01 -9.29721370e-02 9.09134448e-01
1.49014592e+00 3.30541171e-02 9.11471605e-01 4.27858680e-01
9.69443917e-01 5.81641853e-01 3.98343861e-01 5.72930753e-01
3.28642070e-01 7.05761909e-01 6.25060499e-01 -1.66874915e-01
-2.42729589e-01 -3.56107891e-01 3.03868592e-01 4.65722114e-01
-1.52265966e-01 -3.35485697e-01 -4.93862957e-01 4.99268204e-01
-1.65772045e+00 -9.75252628e-01 1.04236796e-01 2.24463606e+00
8.15453291e-01 -4.53348607e-02 -3.33624303e-01 -1.26591325e-01
4.38385993e-01 4.07350034e-01 -6.06818199e-01 -5.53848326e-01
-1.86522663e-01 4.30690378e-01 5.63047588e-01 6.27902269e-01
-9.21699286e-01 6.10420585e-01 5.58836794e+00 3.74990761e-01
-8.76342237e-01 9.41053331e-02 7.31577277e-01 -4.51173723e-01
-9.25758123e-01 1.72688574e-01 -3.15055639e-01 8.09245706e-02
4.25456464e-01 1.45192280e-01 7.13814855e-01 4.59262103e-01
2.60817111e-01 -5.67130923e-01 -9.55068409e-01 1.37498069e+00
3.68378282e-01 -1.55362177e+00 1.52244151e-01 -1.01183675e-01
1.06767809e+00 -1.64408207e-01 -1.79010332e-02 -3.47344488e-01
3.49272132e-01 -1.01515019e+00 1.13704717e+00 9.01667535e-01
9.19921875e-01 -9.08806503e-01 -2.85577290e-02 1.24698199e-01
-8.20100963e-01 2.11983323e-01 -1.37171730e-01 1.08734265e-01
4.46344286e-01 9.02125299e-01 -5.46697378e-01 8.26148212e-01
8.33412290e-01 6.47763193e-01 -4.93383557e-01 8.02965760e-01
-1.53700560e-01 -7.58829117e-02 -5.00490725e-01 6.40222669e-01
-9.59531963e-02 -3.28938752e-01 6.58961654e-01 7.66800642e-01
6.41800046e-01 2.18484476e-01 -2.74562091e-01 9.81486440e-01
-2.68093139e-01 -1.39356345e-01 -8.75000536e-01 2.70271748e-01
4.29414064e-01 1.34917474e+00 -8.87901366e-01 -1.88480001e-02
-2.04165950e-01 1.66392565e+00 1.87804192e-01 3.04337472e-01
-6.89585626e-01 -1.46278694e-01 7.12337255e-01 2.37552181e-01
3.68478656e-01 -6.95294365e-02 -2.93910950e-01 -1.34217978e+00
2.15826035e-01 -9.19809282e-01 2.81891748e-02 -1.45423627e+00
-1.29851198e+00 7.29452014e-01 1.73695877e-01 -1.30387139e+00
-2.74409533e-01 -2.15174749e-01 -4.43332136e-01 9.57014501e-01
-1.29341471e+00 -1.20339906e+00 -5.38561463e-01 5.80299199e-01
5.08070588e-01 3.35093826e-01 8.10931802e-01 1.39988780e-01
3.34965475e-02 6.49200380e-02 -3.60557809e-02 -5.63958883e-01
9.40884292e-01 -1.11981285e+00 6.13611460e-01 9.99909937e-01
2.91490674e-01 5.98250806e-01 9.11397696e-01 -6.90322340e-01
-1.30098283e+00 -1.14061546e+00 6.08267665e-01 -5.58999419e-01
-1.74112529e-01 -4.90737081e-01 -5.43792784e-01 7.11532056e-01
3.74072373e-01 1.29325658e-01 4.13077861e-01 -4.63064134e-01
-2.74069488e-01 -3.81168090e-02 -1.49276149e+00 8.78151178e-01
1.31754351e+00 -6.34008050e-01 -4.42799538e-01 3.03191930e-01
5.29995680e-01 -7.30085373e-01 -7.21821070e-01 2.83203721e-01
6.41839027e-01 -1.51202500e+00 1.14067614e+00 -8.93304124e-02
6.05305970e-01 -5.88677227e-01 -3.92307431e-01 -1.51736224e+00
-1.11649945e-01 -9.26596940e-01 -6.14335947e-02 9.95025754e-01
-1.17021069e-01 -3.05478871e-01 6.71893179e-01 7.55314887e-01
-1.44760296e-01 -4.66043890e-01 -7.55300522e-01 -2.92836249e-01
-3.77099067e-01 -5.21280468e-01 5.29022694e-01 8.59499872e-01
-9.00582671e-01 5.99469356e-02 -5.85435867e-01 3.04142050e-02
7.85049915e-01 2.09996596e-01 9.69435394e-01 -1.14336026e+00
-3.99777412e-01 -7.25964084e-02 2.09448822e-02 -8.85829449e-01
-1.77720010e-01 -6.68386638e-01 1.58741593e-01 -1.88009834e+00
-2.66758762e-02 -3.02041709e-01 8.34905803e-02 3.47903967e-01
4.10951022e-03 5.18257797e-01 4.66702551e-01 2.07342967e-01
-1.38943955e-01 5.89001358e-01 1.36865151e+00 1.42044082e-01
-1.93788007e-01 -4.80567694e-01 -9.02914286e-01 8.87420177e-01
6.55494452e-01 -5.81323683e-01 -6.41214609e-01 -4.51754510e-01
4.37059253e-01 3.71153429e-02 6.04610503e-01 -1.03684521e+00
-2.02662796e-01 -2.16037780e-01 8.18557799e-01 -6.91837251e-01
7.12249875e-01 -9.32391942e-01 8.71265650e-01 -5.72056510e-02
-6.65387437e-02 1.70282960e-01 9.83686745e-02 5.22245824e-01
1.01254605e-01 -2.02167079e-01 7.08661556e-01 -5.77217877e-01
-7.23193049e-01 2.39638060e-01 -8.37774798e-02 -1.44615144e-01
7.64200747e-01 -4.90048826e-01 -8.85916129e-03 -7.90216088e-01
-5.37009835e-01 -3.56911272e-01 1.05435359e+00 4.30751324e-01
9.21875119e-01 -1.36652100e+00 -5.49034774e-01 3.65784228e-01
2.78478563e-02 3.58214706e-01 4.09882724e-01 4.18551594e-01
-7.40037918e-01 -2.11151198e-01 -3.83857071e-01 -3.22353423e-01
-1.16950083e+00 2.17594251e-01 1.93810254e-01 -2.49068625e-02
-8.98606598e-01 8.81131768e-01 4.24218625e-01 -3.90470952e-01
-4.40194532e-02 -2.99136281e-01 2.00773582e-01 -3.69956605e-02
4.64865834e-01 9.34746787e-02 3.01411003e-01 -6.33355260e-01
-1.90611303e-01 4.22322810e-01 -1.17058434e-01 -6.41426325e-01
1.74100065e+00 -4.38701451e-01 -1.01148620e-01 4.36593503e-01
1.01252007e+00 4.49099064e-01 -1.89033747e+00 2.29988322e-01
-6.14113331e-01 -8.63823354e-01 1.89595357e-01 -8.79004538e-01
-9.83741522e-01 3.31866592e-01 4.48510438e-01 -4.55381460e-02
1.39694965e+00 -2.75506407e-01 9.84222293e-01 7.90671855e-02
5.95595896e-01 -9.38039660e-01 -9.67190564e-02 2.53410965e-01
1.28416872e+00 -9.57111597e-01 3.65458906e-01 -3.84217054e-01
-6.11822307e-01 1.17227507e+00 4.10002679e-01 -1.83621436e-01
4.23190475e-01 3.10411870e-01 1.57041937e-01 -1.83941275e-01
-5.86277902e-01 1.54545873e-01 2.78948247e-01 8.17940831e-01
3.28549087e-01 -1.99717849e-01 -1.77693944e-02 3.63337919e-02
-3.88914853e-01 -2.06977222e-02 7.89906085e-01 1.19984329e+00
5.72342379e-03 -1.16344082e+00 -6.31568611e-01 4.54203367e-01
-2.14391306e-01 -2.18719959e-01 -4.48974669e-01 4.37044382e-01
2.90455341e-01 6.32195830e-01 1.65170468e-02 -1.20612904e-01
3.12007070e-01 -1.44336969e-02 9.85268891e-01 -6.67662442e-01
-5.68155408e-01 5.71555607e-02 1.64694756e-01 -8.71044278e-01
-6.79871976e-01 -6.40079856e-01 -8.50715697e-01 -2.52569109e-01
-2.51043458e-02 -4.21475619e-01 6.33811474e-01 6.72728837e-01
5.38405597e-01 3.36035073e-01 5.80826998e-01 -1.31361818e+00
1.13594159e-01 -3.39693904e-01 -3.57598543e-01 5.38225591e-01
5.12049675e-01 -7.03906178e-01 -1.81034759e-01 3.31032425e-01] | [9.287914276123047, -3.0858051776885986] |
ca4c3ad7-b12e-4b13-9a5b-0c22152bfacd | 191013408 | 1910.13408 | null | https://arxiv.org/abs/1910.13408v2 | https://arxiv.org/pdf/1910.13408v2.pdf | A framework for deep learning emulation of numerical models with a case study in satellite remote sensing | Numerical models based on physics represent the state-of-the-art in earth system modeling and comprise our best tools for generating insights and predictions. Despite rapid growth in computational power, the perceived need for higher model resolutions overwhelms the latest-generation computers, reducing the ability of modelers to generate simulations for understanding parameter sensitivities and characterizing variability and uncertainty. Thus, surrogate models are often developed to capture the essential attributes of the full-blown numerical models. Recent successes of machine learning methods, especially deep learning, across many disciplines offer the possibility that complex nonlinear connectionist representations may be able to capture the underlying complex structures and nonlinear processes in earth systems. A difficult test for deep learning-based emulation, which refers to function approximation of numerical models, is to understand whether they can be comparable to traditional forms of surrogate models in terms of computational efficiency while simultaneously reproducing model results in a credible manner. A deep learning emulation that passes this test may be expected to perform even better than simple models with respect to capturing complex processes and spatiotemporal dependencies. Here we examine, with a case study in satellite-based remote sensing, the hypothesis that deep learning approaches can credibly represent the simulations from a surrogate model with comparable computational efficiency. Our results are encouraging in that the deep learning emulation reproduces the results with acceptable accuracy and often even faster performance. We discuss the broader implications of our results in light of the pace of improvements in high-performance implementations of deep learning as well as the growing desire for higher-resolution simulations in the earth sciences. | ['Weile Wang', 'Auroop R. Ganguly', 'Thomas Vandal', 'Ramakrishna Nemani', 'Kate Duffy'] | 2019-10-29 | null | null | null | null | ['cloud-detection'] | ['computer-vision'] | [-2.71515667e-01 -1.70562446e-01 1.95549369e-01 -1.24106392e-01
-4.96332675e-01 -6.70620024e-01 1.08773768e+00 2.47454390e-01
-4.36547101e-02 9.42487776e-01 2.66668946e-01 -1.11398566e+00
-4.50417697e-01 -1.06720424e+00 -6.49557769e-01 -6.46472573e-01
-5.13674915e-01 7.20891476e-01 -2.06851199e-01 -7.00708151e-01
-1.55428961e-01 9.87508774e-01 -1.37070358e+00 -6.80640265e-02
1.01065886e+00 7.07810342e-01 2.49257265e-03 6.48259401e-01
-3.34557854e-02 5.93875349e-01 -2.49761224e-01 2.08665177e-01
6.26416877e-02 -3.94204110e-01 -5.35632908e-01 -3.81205738e-01
1.03354722e-01 -2.59134859e-01 -5.33879042e-01 6.96977615e-01
5.07012308e-01 3.19071323e-01 8.25260699e-01 -8.71415854e-01
-3.34832221e-01 2.79751182e-01 -2.18149886e-01 3.50993752e-01
-5.55240810e-02 4.83797669e-01 6.55659735e-01 -5.29860020e-01
1.38161451e-01 1.31281614e+00 1.10528076e+00 2.57272758e-02
-1.79648638e+00 -5.00013471e-01 4.83625643e-02 -4.22744602e-01
-1.37309062e+00 -4.62770194e-01 2.71042138e-01 -8.11663330e-01
1.13076246e+00 2.37804785e-01 8.37799251e-01 6.87573850e-01
5.32353044e-01 2.98875906e-02 1.17649817e+00 -7.65715837e-02
2.07082093e-01 -4.41456819e-03 -2.68179715e-01 2.69293100e-01
6.02215409e-01 7.61416972e-01 1.15760788e-02 -4.93084222e-01
9.26745534e-01 -2.57377386e-01 -3.42877001e-01 1.32003352e-01
-9.41263318e-01 9.88538980e-01 6.30712807e-01 2.73141414e-01
-5.97571611e-01 5.59008420e-01 1.84245050e-01 6.01449460e-02
7.00544298e-01 8.53570282e-01 -5.90370417e-01 -4.73836362e-02
-1.41880035e+00 7.02981114e-01 9.24523354e-01 3.26321006e-01
5.59825480e-01 8.32422376e-01 1.95135489e-01 4.33817118e-01
4.49271172e-01 8.67573380e-01 1.98987037e-01 -9.04099524e-01
-3.11035253e-02 3.58917475e-01 5.12904584e-01 -1.03194582e+00
-5.92566729e-01 -7.85389841e-01 -1.08143294e+00 5.36409557e-01
2.39795864e-01 -4.35088664e-01 -8.54508936e-01 1.66398299e+00
6.28745258e-02 2.35113800e-01 1.97000518e-01 6.98692381e-01
4.83233750e-01 1.38446045e+00 3.36662918e-01 1.24079697e-01
8.71113837e-01 -2.17124149e-01 -1.89003840e-01 -4.78732049e-01
5.89946449e-01 -5.56945205e-01 8.85339737e-01 -2.97070980e-01
-1.03773475e+00 -5.47987938e-01 -9.22035456e-01 4.65147644e-01
-4.85911667e-01 -2.46473059e-01 1.03461766e+00 3.61679465e-01
-1.18518972e+00 1.16965997e+00 -1.21987987e+00 -4.07505453e-01
1.30782053e-01 1.21925637e-01 1.36779556e-02 3.55103821e-01
-1.48155010e+00 1.30405688e+00 2.65045017e-01 6.04582965e-01
-1.00220037e+00 -1.13348186e+00 -7.10255802e-01 5.82688749e-01
-3.25046659e-01 -8.50423336e-01 1.20497012e+00 -9.46202993e-01
-9.94415879e-01 3.91591042e-01 3.01959384e-02 -5.57254016e-01
6.07062161e-01 2.61680394e-01 -4.87845451e-01 -2.15004534e-01
-1.38374388e-01 5.04398882e-01 3.49727243e-01 -1.38546276e+00
-2.89557457e-01 -1.80042442e-02 1.46822836e-02 1.90386206e-01
2.87620902e-01 -2.35141382e-01 4.48998570e-01 -4.47223783e-01
1.82112783e-01 -9.30199087e-01 -5.24907291e-01 1.84518114e-01
2.34420627e-01 1.82276025e-01 5.04190862e-01 -6.60141468e-01
9.04800534e-01 -1.80099702e+00 -2.70398051e-01 2.38752156e-01
9.18250531e-02 4.40308511e-01 -4.29515280e-02 8.66522431e-01
-2.30538353e-01 6.18165851e-01 -5.52553356e-01 -1.40283331e-01
9.55085233e-02 3.84977669e-01 -8.09353173e-01 4.86281753e-01
6.22429848e-01 8.44152033e-01 -8.64945531e-01 2.97893491e-02
4.68783766e-01 6.24202907e-01 -2.32685804e-01 1.48366243e-01
-3.82410169e-01 6.23298824e-01 -3.54675740e-01 2.08274797e-01
7.08591104e-01 -3.06531101e-01 2.25124329e-01 2.30969220e-01
-3.45140487e-01 5.11012435e-01 -1.17031157e+00 8.06358874e-01
-7.40138769e-01 9.23468590e-01 2.66195834e-01 -1.07138038e+00
8.83512020e-01 3.97829056e-01 2.13168725e-01 -6.16816342e-01
-1.98990628e-01 4.20719475e-01 5.91230810e-01 -1.56725883e-01
4.01545405e-01 -7.41693676e-01 2.27109820e-01 5.45237064e-01
-3.83219361e-01 -8.01085353e-01 -2.41562635e-01 -6.09242842e-02
3.99618566e-01 4.50397819e-01 8.89723226e-02 -8.42276096e-01
1.54626235e-01 3.95213366e-01 1.67715028e-01 8.30833495e-01
1.69541776e-01 3.61437917e-01 3.51195395e-01 -7.53911674e-01
-1.36734223e+00 -9.84616756e-01 -5.78990102e-01 5.96494794e-01
-4.59306151e-01 1.31673902e-01 -2.36614272e-01 4.37910229e-01
4.75198597e-01 7.68447399e-01 -6.47708893e-01 -2.98640877e-01
-3.18811268e-01 -1.50076723e+00 9.29164767e-01 5.58576465e-01
3.10556620e-01 -9.08799112e-01 -5.49785495e-01 5.81245542e-01
1.45763665e-01 -7.63352931e-01 6.88300967e-01 3.06932837e-01
-1.26844907e+00 -7.21739471e-01 -4.64122444e-01 -2.14611739e-01
2.24451914e-01 -4.03645113e-02 1.33004999e+00 2.08801568e-01
8.74137692e-03 -4.36884239e-02 1.77649081e-01 -3.77709538e-01
-9.96628106e-01 9.77805033e-02 2.32516289e-01 -5.73188663e-01
-4.05627908e-03 -1.04819167e+00 -6.12721801e-01 -5.72341233e-02
-1.11443102e+00 3.77446041e-02 3.82684290e-01 6.75749660e-01
1.77426338e-01 4.30448540e-02 7.59859622e-01 -4.09930706e-01
6.80337369e-01 -8.89040768e-01 -8.41567516e-01 9.98674780e-02
-8.22731376e-01 2.82326132e-01 8.03994179e-01 -1.60712600e-01
-1.15745747e+00 -5.00011086e-01 -2.25118458e-01 3.56217287e-02
-2.53631920e-01 1.27302623e+00 3.83795112e-01 -1.93275675e-01
9.44669485e-01 3.37651759e-01 7.87281394e-02 -3.84719163e-01
1.39426395e-01 4.42122877e-01 3.21352363e-01 -1.02703059e+00
8.87546778e-01 5.86308300e-01 3.59326601e-01 -1.10035443e+00
-3.59228849e-01 4.83144820e-02 -2.39557028e-01 -1.28324166e-01
3.06656092e-01 -1.21428049e+00 -2.95096606e-01 4.92616087e-01
-9.49632823e-01 -7.60768116e-01 -2.64840126e-01 6.79111362e-01
-3.98634344e-01 -1.38827041e-01 -4.53494996e-01 -9.91301894e-01
-1.60405010e-01 -9.15219128e-01 7.53834188e-01 3.74365002e-01
-2.52406716e-01 -1.63019264e+00 4.42562282e-01 -4.48894680e-01
1.07194543e+00 6.43141210e-01 1.29813313e+00 -3.48723620e-01
-3.78530204e-01 -2.18797550e-01 -2.94203192e-01 1.74995258e-01
8.73818323e-02 4.18225497e-01 -1.28216088e+00 -2.18539327e-01
-6.47829697e-02 -1.02993146e-01 9.35937405e-01 7.24999011e-01
7.72272229e-01 -1.35686293e-01 -3.08905631e-01 8.36554766e-01
1.64984572e+00 2.23209541e-02 6.99180663e-01 2.85608590e-01
3.00317198e-01 5.66378891e-01 -1.33786500e-01 3.84714872e-01
1.31572917e-01 1.67095378e-01 3.37904364e-01 -5.16687334e-01
8.75404254e-02 -2.64281541e-01 -8.89562368e-02 5.51563084e-01
-2.61075824e-01 -1.01186454e-01 -1.64709055e+00 5.42137623e-01
-1.67543674e+00 -1.24460781e+00 -3.46201211e-01 2.22932243e+00
6.27097726e-01 2.30940953e-01 -2.11042151e-01 -1.69202715e-01
3.45657974e-01 4.72433776e-01 -6.76863551e-01 -6.32659197e-01
-3.06930602e-01 3.47984701e-01 5.61624765e-01 5.76199114e-01
-8.99728477e-01 7.70655930e-01 7.29232597e+00 2.13765785e-01
-1.56689453e+00 -1.18052609e-01 8.70179236e-01 1.73363388e-01
-4.18909073e-01 2.21315846e-01 -6.97255492e-01 2.31370836e-01
1.78577924e+00 -3.62375379e-01 5.27641416e-01 5.79359114e-01
9.28668618e-01 -2.57605582e-01 -9.92386460e-01 3.48668903e-01
-5.21808743e-01 -1.91568327e+00 2.16057613e-01 9.42048579e-02
9.97211695e-01 4.65578228e-01 7.11377263e-02 4.70610946e-01
6.16754174e-01 -1.55783510e+00 6.69230342e-01 7.87770510e-01
6.45780802e-01 -5.49833953e-01 7.84882247e-01 4.35421050e-01
-1.16032040e+00 1.42765000e-01 -5.59495687e-01 -7.59243846e-01
6.43639490e-02 4.78810161e-01 -5.66713810e-01 4.30280834e-01
6.14079475e-01 4.85931754e-01 -2.48757616e-01 1.15984678e+00
2.42957115e-01 1.07785904e+00 -7.07318425e-01 2.01890200e-01
7.30064869e-01 -3.21594536e-01 3.47844660e-01 1.25135124e+00
4.02923644e-01 1.58942953e-01 -7.08520189e-02 1.25404847e+00
3.72751564e-01 -2.57873327e-01 -7.79728532e-01 -4.06306654e-01
4.66814607e-01 9.07539427e-01 -5.32898068e-01 -3.29114050e-01
-2.98221916e-01 6.12193123e-02 7.78959394e-02 6.82040989e-01
-7.32185245e-01 7.43691847e-02 1.04683173e+00 5.02632856e-01
8.15646490e-04 -5.12959719e-01 -3.39037955e-01 -8.17447662e-01
-3.91010344e-01 -9.72033918e-01 -5.76788746e-02 -1.22759116e+00
-1.21577680e+00 5.50668180e-01 1.62411392e-01 -1.01185822e+00
-3.79605800e-01 -7.22857893e-01 -9.94050086e-01 1.60762000e+00
-1.74999571e+00 -9.62552845e-01 -3.16852599e-01 -4.47850535e-03
6.84710313e-03 2.55045503e-01 1.14050877e+00 -1.41953960e-01
-1.62806332e-01 -2.65436232e-01 9.51805949e-01 -1.26791403e-01
3.39244783e-01 -9.84778702e-01 8.62182796e-01 6.49975657e-01
-2.53475904e-01 6.37229204e-01 9.51197863e-01 -6.47331893e-01
-1.24961960e+00 -1.14439082e+00 5.39019108e-01 -2.92273849e-01
1.00978911e+00 1.60904881e-02 -1.29819262e+00 6.48823023e-01
-1.63710147e-01 8.26361030e-02 4.96063441e-01 7.43184388e-02
-6.98531643e-02 -1.15176566e-01 -8.39344561e-01 5.51851690e-01
3.58905852e-01 -6.35495245e-01 -6.70840085e-01 2.74811298e-01
2.27911025e-01 -2.64540583e-01 -9.41874325e-01 5.20940781e-01
6.45992458e-01 -5.51428437e-01 7.83536136e-01 -1.08504152e+00
7.43661165e-01 -2.90300369e-01 -2.95935601e-01 -1.72065747e+00
-5.80306530e-01 -4.47082758e-01 5.66945858e-02 8.96572888e-01
5.65079451e-01 -7.66917348e-01 4.42699641e-01 8.31791639e-01
3.78512852e-02 -5.19917011e-01 -7.65300572e-01 -8.55276763e-01
1.06957090e+00 -3.88272703e-01 8.85196686e-01 1.07010221e+00
-3.86587411e-01 1.77520346e-02 3.96975055e-02 6.25746429e-01
4.59871560e-01 2.21809134e-01 3.69146794e-01 -1.77365077e+00
-1.91290498e-01 -7.44959831e-01 -2.65286505e-01 -4.99331772e-01
1.32329747e-01 -6.35884106e-01 -1.50629580e-01 -1.75903177e+00
-1.56307116e-01 -7.99215853e-01 -1.09963499e-01 1.16924644e-01
-8.08772892e-02 5.98822758e-02 3.66437361e-02 4.07826781e-01
7.16734588e-01 6.56863034e-01 9.33664739e-01 -5.68861328e-02
-1.38926178e-01 -5.96456788e-02 -5.18979669e-01 8.09414804e-01
9.74317372e-01 -4.14815187e-01 -2.86196142e-01 -5.19559979e-01
5.85695505e-01 2.60316849e-01 7.31269717e-01 -1.17814267e+00
-2.90558264e-02 -4.11780804e-01 8.31081748e-01 -2.89966255e-01
1.60885513e-01 -6.45009220e-01 6.01971626e-01 4.71931517e-01
-1.47932515e-01 1.79330006e-01 1.10628152e+00 2.33159423e-01
-1.06194131e-01 -2.51609962e-02 8.36614788e-01 -3.66200179e-01
-4.92059320e-01 3.36380750e-01 -8.68824840e-01 2.74072271e-02
4.34287399e-01 -1.02334850e-01 -4.21426356e-01 -7.41647542e-01
-7.89920270e-01 2.61033684e-01 5.93482971e-01 6.93434430e-03
4.62461114e-02 -9.54944134e-01 -1.07584047e+00 2.06386074e-01
-1.95914298e-01 -7.21560419e-02 1.52262360e-01 4.53582674e-01
-9.15480494e-01 5.78658402e-01 -1.56225190e-01 -4.29424465e-01
-2.10981384e-01 1.08708784e-01 1.22649789e+00 -2.41396755e-01
-4.93127823e-01 4.27950650e-01 3.10926467e-01 -6.93726003e-01
-4.28991199e-01 -5.03832042e-01 2.76512980e-01 -1.14459895e-01
2.68418133e-01 1.83429271e-01 1.10365957e-01 -5.08888185e-01
-2.09865198e-01 1.59535214e-01 5.74585378e-01 -2.35415533e-01
1.64048386e+00 1.42854780e-01 2.66647674e-02 6.65880203e-01
6.05677068e-01 -4.22696173e-01 -1.57159054e+00 -4.61463025e-03
-1.81699291e-01 -9.74250659e-02 5.43885052e-01 -9.61474538e-01
-8.22721303e-01 1.20556188e+00 5.37167430e-01 3.46214622e-01
6.19853437e-01 -4.04669791e-01 2.60903448e-01 6.53805554e-01
-6.36779517e-02 -7.60449946e-01 -6.25048161e-01 5.27191281e-01
1.02886462e+00 -1.15014434e+00 1.79022059e-01 3.13152671e-01
-1.60080701e-01 1.10543585e+00 4.13481683e-01 -3.34771603e-01
1.00471723e+00 4.62481886e-01 -5.42778000e-02 -9.61206034e-02
-7.16663957e-01 7.40616396e-02 -3.68772186e-02 4.95818883e-01
5.59349835e-01 3.12908947e-01 1.03867106e-01 1.67955175e-01
-2.07414150e-01 -7.35359341e-02 6.47752345e-01 6.78011715e-01
-6.59802437e-01 -6.17168069e-01 -6.76300228e-01 4.87547994e-01
-2.55856425e-01 -5.27708113e-01 1.02975540e-01 1.04369581e+00
-3.82167310e-01 7.12021172e-01 5.11169553e-01 2.26287395e-01
9.74602327e-02 2.34581307e-01 8.79912451e-02 -5.74836135e-01
-4.79270577e-01 -1.14071727e-01 9.30037946e-02 -1.28547698e-01
-2.20913544e-01 -6.33595705e-01 -1.01098812e+00 -9.69702542e-01
-3.72406431e-02 3.91955167e-01 7.33684242e-01 1.02365887e+00
3.81984413e-01 6.18449748e-01 3.84797513e-01 -1.35261893e+00
-8.63055825e-01 -1.15077746e+00 -6.76991403e-01 -1.57318577e-01
5.29397309e-01 -6.67006671e-01 -5.40926814e-01 -2.33594567e-01] | [6.584314346313477, 3.1027779579162598] |
1d94d4f7-791f-4db7-a0a1-75eaf4e8503a | explaining-large-language-model-based-neural | 2301.13820 | null | https://arxiv.org/abs/2301.13820v1 | https://arxiv.org/pdf/2301.13820v1.pdf | Explaining Large Language Model-Based Neural Semantic Parsers (Student Abstract) | While large language models (LLMs) have demonstrated strong capability in structured prediction tasks such as semantic parsing, few amounts of research have explored the underlying mechanisms of their success. Our work studies different methods for explaining an LLM-based semantic parser and qualitatively discusses the explained model behaviors, hoping to inspire future research toward better understanding them. | ['Ziyu Yao', 'Bailin Wang', 'Yilun Zhou', 'Daking Rai'] | 2023-01-25 | null | null | null | null | ['semantic-parsing'] | ['natural-language-processing'] | [ 2.32086107e-01 1.30004013e+00 -8.49509776e-01 -9.66341496e-01
-6.83934033e-01 -3.10368389e-01 6.08686447e-01 2.66973734e-01
-4.16421294e-02 3.38984758e-01 5.93197167e-01 -1.05907643e+00
2.22677693e-01 -6.48601174e-01 -6.42104149e-01 1.09681070e-01
1.45441353e-01 8.10665846e-01 4.57070112e-01 4.06722389e-02
6.08377159e-01 1.63084611e-01 -1.18088114e+00 6.18180275e-01
7.02357888e-01 3.44959706e-01 5.63486516e-01 4.65262592e-01
-1.03394425e+00 1.49496245e+00 -5.15532672e-01 -6.83014810e-01
-1.34956136e-01 -6.19063556e-01 -1.28041422e+00 -1.24660917e-01
1.70330241e-01 1.45899355e-01 -2.46640995e-01 9.71278071e-01
-1.86948687e-01 -8.65165815e-02 2.83714473e-01 -9.62989390e-01
-8.09908271e-01 1.40682054e+00 -2.60074213e-02 2.19351918e-01
4.83540565e-01 -1.44203424e-01 1.42248166e+00 -5.07661879e-01
8.89853716e-01 1.85245323e+00 7.50154972e-01 1.02457464e+00
-1.05956483e+00 -5.57709515e-01 2.49787256e-01 1.53292581e-01
-8.39707077e-01 -4.37693834e-01 6.94807410e-01 2.35714186e-02
1.72198272e+00 -1.69110641e-01 3.00668627e-01 8.81413221e-01
4.84867871e-01 9.42028999e-01 1.28487754e+00 -8.62163842e-01
2.02414110e-01 1.86643928e-01 7.72245944e-01 6.84759736e-01
5.98120749e-01 8.62527415e-02 -9.69688177e-01 -1.60208330e-01
5.83210945e-01 -6.49570763e-01 4.62619990e-01 -1.38379887e-01
-6.80134714e-01 1.35312688e+00 1.56255692e-01 5.13014019e-01
-1.55435294e-01 1.96608305e-01 3.30038905e-01 2.56468028e-01
5.80998898e-01 7.51774788e-01 -8.11117887e-01 -4.29888070e-01
-7.82818794e-01 2.93134719e-01 1.07136190e+00 1.12729645e+00
5.84954202e-01 1.70645908e-01 7.01700270e-01 8.43014777e-01
5.16981065e-01 1.40349612e-01 6.60731733e-01 -1.13634431e+00
6.94646537e-01 7.53958225e-01 -6.18138462e-02 -7.77033806e-01
-6.69349253e-01 1.97299823e-01 5.18391654e-02 -2.82638341e-01
3.65605295e-01 7.75358379e-02 -6.80998683e-01 1.86085331e+00
-3.54726046e-01 -1.01079501e-01 3.31241608e-01 6.60951853e-01
8.72822821e-01 5.74770033e-01 1.16744936e+00 -6.71066716e-02
1.24308383e+00 -1.04352748e+00 -4.55927253e-01 -1.13800967e+00
1.21261358e+00 -6.21196210e-01 1.11050665e+00 -1.74178451e-01
-1.34497380e+00 -4.18350339e-01 -6.42638147e-01 -3.95597339e-01
-1.40122741e-01 -3.42573434e-01 1.32078457e+00 6.06677473e-01
-1.11640561e+00 5.99755228e-01 -1.41734791e+00 -9.26092446e-01
2.36494094e-01 2.56397307e-01 -1.12276174e-01 1.26808032e-01
-1.16376686e+00 1.24290967e+00 5.80486000e-01 -3.57835501e-01
-5.45007586e-01 -3.31147730e-01 -1.10825980e+00 4.22130227e-02
2.34955654e-01 -7.02922523e-01 1.61827707e+00 -9.76788759e-01
-1.10610878e+00 1.21168339e+00 -8.64981651e-01 -8.66377831e-01
-1.58562869e-01 -1.18685804e-01 -2.79193014e-01 4.09984402e-03
2.02880532e-01 9.24298227e-01 4.05602038e-01 -1.09766662e+00
-4.97514337e-01 -5.95228076e-01 2.02854082e-01 1.45551816e-01
-2.96636790e-01 6.44946516e-01 -3.12104505e-02 -5.71088314e-01
7.15881765e-01 -8.76868904e-01 -5.03117800e-01 -6.59832895e-01
-3.81861657e-01 -5.30174792e-01 2.78669953e-01 -6.41618252e-01
1.36821353e+00 -1.71727061e+00 -1.87175646e-01 -3.51960599e-01
1.14567146e-01 1.58885002e-01 -4.08265293e-01 7.99757123e-01
-4.33470048e-02 6.48418546e-01 -2.58809328e-02 -6.18295670e-01
-3.87201309e-02 4.86459672e-01 -6.06121361e-01 -3.02374661e-01
2.33940646e-01 1.18272984e+00 -7.50615358e-01 -6.89112425e-01
2.38303974e-01 1.77023962e-01 -4.97533530e-01 2.91723371e-01
-5.51230073e-01 5.06902561e-02 -8.97578835e-01 6.19729102e-01
1.08067527e-01 -5.05003333e-01 5.71519434e-01 3.37395757e-01
-5.65488338e-02 1.22022140e+00 -5.94889462e-01 1.51798201e+00
-5.30660331e-01 6.48786485e-01 1.05057256e-02 -1.15742123e+00
8.74235809e-01 2.92503864e-01 3.73369306e-02 -7.35681772e-01
-1.58036858e-01 2.96439081e-01 1.65508360e-01 -3.87881309e-01
4.60298210e-01 -5.33602059e-01 -1.28679216e-01 5.19501925e-01
-1.00800611e-01 -1.68568328e-01 -1.11274578e-01 4.26420361e-01
1.22115719e+00 2.05495521e-01 3.66829455e-01 -3.18709880e-01
2.72664368e-01 6.18885159e-01 5.70088327e-01 1.02848768e+00
-2.00632170e-01 1.79657146e-01 4.59777832e-01 -7.87112772e-01
-1.07883775e+00 -6.60774469e-01 -2.58812476e-02 1.22758186e+00
2.91931689e-01 -8.81458163e-01 -9.71764684e-01 -6.73511267e-01
-1.15696497e-01 1.31579459e+00 -3.56446445e-01 4.81590517e-02
-8.18413019e-01 -6.83762252e-01 7.14394867e-01 9.94442940e-01
2.60720938e-01 -1.56532502e+00 -5.86127520e-01 3.79985094e-01
-1.92649975e-01 -1.36976695e+00 2.84897953e-01 4.19336706e-01
-1.62748170e+00 -1.06057501e+00 4.52667952e-01 -1.09183788e+00
6.54675484e-01 2.97126889e-01 1.55222464e+00 3.06125492e-01
6.66365996e-02 2.63873547e-01 -3.93915802e-01 -3.81520122e-01
-1.09598172e+00 2.65891969e-01 -1.58816501e-01 -1.06065714e+00
1.07725203e+00 -2.58313268e-01 -8.24399143e-02 1.48885116e-01
-5.36236346e-01 3.60804796e-01 4.42883849e-01 6.72615886e-01
1.73588261e-01 -2.91180909e-01 7.02395201e-01 -1.35804355e+00
9.13812518e-01 -3.95737022e-01 -4.17671323e-01 2.42983714e-01
-1.00637770e+00 2.54754096e-01 6.49271309e-01 6.90110773e-02
-1.42478621e+00 -2.59385705e-01 -4.04796749e-01 4.08479840e-01
-4.07316804e-01 4.58193749e-01 -1.15493955e-02 -9.29204375e-02
3.13874602e-01 6.51810914e-02 -2.38813534e-01 -6.87668562e-01
2.66076982e-01 3.82294565e-01 1.37615874e-01 -7.50727534e-01
6.78413749e-01 1.35337999e-02 -2.35024720e-01 -6.91955328e-01
-1.24800813e+00 -2.06078380e-01 -5.09189188e-01 4.51036513e-01
1.04325080e+00 -7.62523234e-01 -2.45634407e-01 -1.11449167e-01
-1.31002676e+00 -4.69801426e-01 3.18226144e-02 3.17790955e-01
-7.98596203e-01 4.17059749e-01 -1.16415226e+00 -9.58574116e-01
-1.84673443e-01 -1.08552313e+00 9.55919862e-01 2.38724738e-01
-9.35445309e-01 -1.57417846e+00 -9.25465524e-02 9.00181890e-01
3.59749287e-01 -7.45399594e-01 1.52039182e+00 -1.03719652e+00
-6.86214864e-01 6.73780069e-02 -1.37388691e-01 -4.32796776e-02
-2.13167354e-01 -4.92335737e-01 -7.92601943e-01 1.64746255e-01
1.44648522e-01 -4.53707904e-01 6.23147607e-01 5.82886279e-01
1.13861167e+00 -4.93814126e-02 -7.29529679e-01 1.85973257e-01
1.35486853e+00 2.51678735e-01 3.18139762e-01 6.85484231e-01
4.76747036e-01 8.86831462e-01 7.13296771e-01 -1.05871044e-01
7.06719160e-01 4.51341093e-01 1.31579384e-01 1.76175550e-01
-3.09844375e-01 -8.33950818e-01 4.16616946e-01 8.12976837e-01
1.58175573e-01 -1.69680431e-01 -1.04297686e+00 5.20211041e-01
-1.78896022e+00 -7.00133860e-01 -9.79377851e-02 1.65369129e+00
4.48877066e-01 4.70605910e-01 -3.98504257e-01 -2.96382636e-01
5.27666748e-01 5.24841189e-01 -4.79351252e-01 -1.03303337e+00
-2.55508959e-01 3.49445015e-01 4.44977015e-01 8.26563895e-01
-7.38562524e-01 1.96334517e+00 8.24749565e+00 2.40542009e-01
-5.49405456e-01 1.67471647e-01 6.07346833e-01 5.14779210e-01
-6.21521235e-01 6.75241232e-01 -9.94949341e-01 2.10334267e-02
1.47727525e+00 -1.14320345e-01 3.38428289e-01 1.17872727e+00
5.55940688e-01 -9.65082273e-02 -1.11326158e+00 3.74176711e-01
-1.68194667e-01 -1.32494342e+00 4.06746328e-01 -1.49494246e-01
2.99852371e-01 2.01693043e-01 -5.12352109e-01 5.57177544e-01
5.93384445e-01 -9.95455384e-01 5.27781904e-01 -1.23492897e-01
-3.25474329e-02 -2.38188148e-01 6.20882511e-01 5.91287017e-01
-1.13776124e+00 -2.04786077e-01 -6.57918692e-01 -6.28684878e-01
4.43312943e-01 -6.43277019e-02 -7.80086279e-01 6.69775382e-02
3.03508431e-01 5.85627317e-01 -7.30997801e-01 1.96995884e-01
-5.58099031e-01 1.38750577e+00 8.61189812e-02 -6.32590726e-02
4.44070816e-01 3.29122879e-02 3.51208210e-01 1.15765774e+00
-1.25885144e-01 2.80742109e-01 2.76616067e-01 9.83644187e-01
1.20014489e-01 3.41900527e-01 -5.99287033e-01 -7.53990471e-01
7.19873071e-01 7.38730431e-01 -9.91690934e-01 -3.19299489e-01
-9.23544943e-01 8.03393662e-01 5.78590035e-01 4.11133990e-02
-4.95094508e-01 2.83935428e-01 7.16938317e-01 3.03757161e-01
-1.66955397e-01 -3.90311748e-01 -9.51923072e-01 -1.28700578e+00
-7.23641366e-02 -8.10485125e-01 5.34013033e-01 -1.02431333e+00
-1.16652834e+00 5.48893273e-01 -7.68219531e-02 -1.64910719e-01
-3.93805116e-01 -7.55926073e-01 -7.13482082e-01 7.05569565e-01
-1.48790181e+00 -1.43367851e+00 1.99166745e-01 1.23282202e-01
1.14920056e+00 -2.58170456e-01 1.28123426e+00 -4.05720025e-01
-3.89632165e-01 2.75716573e-01 -5.75838804e-01 8.15957487e-02
3.92869800e-01 -1.07180572e+00 1.12605250e+00 7.05243886e-01
4.44480121e-01 9.86957133e-01 8.18035364e-01 -9.12942529e-01
-1.40242589e+00 -6.46811783e-01 1.63853872e+00 -7.29939461e-01
7.66752601e-01 -1.15526006e-01 -1.04399323e+00 1.25152898e+00
1.32528748e-02 -5.08560479e-01 5.51594019e-01 4.18010682e-01
-3.13453525e-01 5.32895744e-01 -8.34894657e-01 2.80199200e-01
1.18697274e+00 -5.24536788e-01 -1.09717119e+00 3.73201847e-01
8.29676747e-01 -2.14840263e-01 -7.19699442e-01 3.08013111e-01
2.20464736e-01 -1.04780710e+00 8.39331746e-01 -1.24215627e+00
7.73299277e-01 4.00540084e-01 -2.20931500e-01 -8.75929713e-01
-1.72830194e-01 -2.75743484e-01 -9.03371572e-02 9.56423283e-01
6.41425967e-01 -7.93422580e-01 1.31482947e+00 1.08738947e+00
-4.39820826e-01 -6.21212900e-01 -6.63830996e-01 -7.73469090e-01
4.73144740e-01 -8.08820367e-01 3.77873272e-01 8.27508211e-01
4.29300964e-01 4.25759166e-01 -5.80960289e-02 3.03010345e-01
6.47045553e-01 2.21995115e-01 3.65324676e-01 -1.37734699e+00
-2.31947169e-01 -4.66762811e-01 -3.12134653e-01 -1.16293120e+00
1.01517045e+00 -1.03240943e+00 -1.48508146e-01 -1.78907883e+00
5.73900461e-01 -5.24091423e-01 -1.13347121e-01 5.91326714e-01
-1.92635760e-01 -1.65086091e-01 2.72272170e-01 2.79982090e-01
-6.26383483e-01 1.01805456e-01 7.47546375e-01 1.40120104e-01
2.93545797e-02 -1.90797418e-01 -1.19171119e+00 1.15589130e+00
7.41438925e-01 -7.22213507e-01 -5.12152314e-01 -8.19768846e-01
8.53098929e-02 3.82608861e-01 -3.47281471e-02 -6.69150352e-01
-4.99617457e-02 -5.23751676e-01 -3.91171053e-02 -1.43115833e-01
1.08882584e-01 -4.92119163e-01 -3.08880329e-01 4.75677013e-01
-1.01102901e+00 2.31704250e-01 2.49073297e-01 3.13325256e-01
-4.55928564e-01 -7.36896098e-01 5.12063026e-01 -5.42927325e-01
-1.26577818e+00 -3.99571657e-02 -6.18660510e-01 3.12390000e-01
7.12847590e-01 -1.95581988e-01 -6.71584979e-02 -4.03425813e-01
-8.77891302e-01 5.28808646e-02 7.91084886e-01 7.99570739e-01
4.60921794e-01 -7.75498390e-01 -2.19760969e-01 -4.37005721e-02
6.80126026e-02 -3.35051894e-01 -2.30694905e-01 2.61294872e-01
-4.55206066e-01 1.13709080e+00 2.82772541e-01 -2.13248149e-01
-1.18543029e+00 3.22002858e-01 1.64816648e-01 -3.01106572e-01
-8.93446445e-01 9.40169752e-01 3.50545913e-01 -5.89877844e-01
1.38666809e-01 -1.34900361e-01 -7.98083991e-02 -6.02500439e-01
3.40815514e-01 1.09676398e-01 -2.40256116e-01 -4.45418924e-01
-5.66729367e-01 4.10245419e-01 -7.32259452e-02 8.73943716e-02
1.39825130e+00 -3.67554098e-01 -2.94350147e-01 4.91634160e-01
7.34168351e-01 -1.66358322e-01 -7.16565609e-01 -1.13959543e-01
9.89694417e-01 -1.05539203e-01 -1.44180372e-01 -6.57966256e-01
-5.21173477e-01 1.06781483e+00 3.89435664e-02 2.38975823e-01
5.89742839e-01 6.95222497e-01 9.96794105e-01 5.36737263e-01
7.96973228e-01 -1.06743300e+00 -9.79838818e-02 7.23789215e-01
2.31340960e-01 -1.56481647e+00 -3.07427257e-01 -5.82769096e-01
-8.38712871e-01 1.34946156e+00 7.43647993e-01 1.39880583e-01
4.18138772e-01 6.29145384e-01 4.37756836e-01 -3.35418344e-01
-1.09540653e+00 1.07177205e-01 -2.88532466e-01 5.21337211e-01
1.12378204e+00 2.78093725e-01 -6.44948244e-01 8.34370017e-01
-4.82195944e-01 -1.64612651e-01 4.56744730e-01 9.21007633e-01
-7.85161614e-01 -1.63437760e+00 -1.32699773e-01 3.26403469e-01
-6.13681853e-01 -5.24838388e-01 -7.19760358e-01 7.91840196e-01
-3.57424766e-01 1.27115011e+00 5.74586578e-02 -8.15059170e-02
-3.87227349e-02 7.57170320e-01 3.50133240e-01 -1.20426369e+00
-4.88693684e-01 -3.27646673e-01 7.32978821e-01 -9.38372374e-01
-4.06700611e-01 -7.08484411e-01 -1.91961193e+00 -8.97085518e-02
-1.52482018e-01 4.20922995e-01 6.95037007e-01 1.12346458e+00
1.88391432e-01 -5.54450899e-02 1.40030226e-02 -1.85653791e-01
-5.18214226e-01 -1.02087796e+00 -4.10987139e-01 5.32666862e-01
-4.35885370e-01 -4.76900160e-01 -3.19129378e-01 -1.26096327e-02] | [10.496962547302246, 9.333677291870117] |
68370834-6251-4c11-9f7b-791a4a41ff0d | ntu-rgbd-120-a-large-scale-benchmark-for-3d | 1905.04757 | null | https://arxiv.org/abs/1905.04757v2 | https://arxiv.org/pdf/1905.04757v2.pdf | NTU RGB+D 120: A Large-Scale Benchmark for 3D Human Activity Understanding | Research on depth-based human activity analysis achieved outstanding performance and demonstrated the effectiveness of 3D representation for action recognition. The existing depth-based and RGB+D-based action recognition benchmarks have a number of limitations, including the lack of large-scale training samples, realistic number of distinct class categories, diversity in camera views, varied environmental conditions, and variety of human subjects. In this work, we introduce a large-scale dataset for RGB+D human action recognition, which is collected from 106 distinct subjects and contains more than 114 thousand video samples and 8 million frames. This dataset contains 120 different action classes including daily, mutual, and health-related activities. We evaluate the performance of a series of existing 3D activity analysis methods on this dataset, and show the advantage of applying deep learning methods for 3D-based human action recognition. Furthermore, we investigate a novel one-shot 3D activity recognition problem on our dataset, and a simple yet effective Action-Part Semantic Relevance-aware (APSR) framework is proposed for this task, which yields promising results for recognition of the novel action classes. We believe the introduction of this large-scale dataset will enable the community to apply, adapt, and develop various data-hungry learning techniques for depth-based and RGB+D-based human activity understanding. [The dataset is available at: http://rose1.ntu.edu.sg/Datasets/actionRecognition.asp] | ['Ling-Yu Duan', 'Gang Wang', 'Mauricio Perez', 'Alex C. Kot', 'Amir Shahroudy', 'Jun Liu'] | 2019-05-12 | null | null | null | null | ['one-shot-3d-action-recognition'] | ['computer-vision'] | [ 3.95591736e-01 -3.77199173e-01 -4.60529834e-01 -3.26470613e-01
-6.34295464e-01 -2.79727042e-01 5.05864263e-01 -3.59698534e-01
-1.42747134e-01 4.88056302e-01 7.26596832e-01 2.35228449e-01
-8.21994171e-02 -6.42113388e-01 -3.45522881e-01 -6.87505186e-01
-1.42461970e-01 2.38881692e-01 4.85715389e-01 -1.36459097e-01
2.57256866e-01 8.65164936e-01 -1.78529155e+00 5.36046088e-01
1.52231097e-01 1.27436590e+00 -1.52951270e-01 7.41095841e-01
5.59471287e-02 1.19198716e+00 -7.68069327e-01 1.10398635e-01
4.38887715e-01 -7.23366857e-01 -9.15800035e-01 6.43705010e-01
3.65902960e-01 -5.62052608e-01 -8.22858930e-01 3.92317802e-01
7.75843740e-01 4.11737561e-01 5.45837581e-01 -1.57512188e+00
-4.70240176e-01 -2.76859701e-01 -3.96738440e-01 6.45056427e-01
1.09167397e+00 3.38778734e-01 3.97885323e-01 -5.68465590e-01
5.95104992e-01 1.07946980e+00 5.17048359e-01 8.28075111e-01
-3.85576129e-01 -3.52764100e-01 5.52005246e-02 6.16107523e-01
-1.14203930e+00 -4.79012877e-01 9.85768437e-01 -4.18964475e-01
1.48212123e+00 1.77630782e-01 1.30731404e+00 1.50410855e+00
9.43207443e-02 1.59351397e+00 9.20776546e-01 -3.20263088e-01
4.33338702e-01 -5.76317489e-01 -1.31170183e-01 7.18585551e-01
6.55446574e-02 3.11955959e-02 -9.05132711e-01 5.27824126e-02
9.83944952e-01 4.29625630e-01 -9.96157005e-02 -7.12590039e-01
-1.48106515e+00 4.18595880e-01 1.60291225e-01 3.39987636e-01
-4.29351777e-01 2.73913264e-01 6.20365977e-01 5.87706175e-03
3.80970746e-01 5.73082231e-02 -5.27315080e-01 -9.51662242e-01
-4.60986555e-01 2.56247848e-01 4.69802350e-01 8.82834613e-01
4.02195692e-01 2.64516063e-02 -2.66170055e-01 7.30913222e-01
1.47418275e-01 6.31942332e-01 7.84527361e-01 -1.39060509e+00
4.98184144e-01 9.98232484e-01 3.09131555e-02 -9.45658147e-01
-5.44972062e-01 2.97034800e-01 -6.60257757e-01 1.27128676e-01
4.22843605e-01 2.39499018e-01 -9.00801897e-01 1.19379389e+00
6.15239978e-01 1.32908702e-01 1.72243729e-01 1.02644956e+00
8.98856461e-01 3.53230596e-01 6.80747554e-02 8.86091555e-04
1.06552422e+00 -1.14692152e+00 -6.57450020e-01 -1.81749761e-01
9.95619178e-01 -1.42192110e-01 1.05210555e+00 4.09800470e-01
-9.30200100e-01 -7.39464700e-01 -9.28059459e-01 -1.62006900e-01
-5.32989144e-01 1.09695226e-01 1.05853641e+00 8.30068827e-01
-7.09725320e-01 3.15884084e-01 -1.09926915e+00 -7.87647665e-01
9.02195096e-01 5.72072379e-02 -8.24180424e-01 -2.63939738e-01
-9.64412332e-01 6.72096014e-01 3.70958537e-01 -9.94051397e-02
-1.06108308e+00 -2.57552385e-01 -9.88821030e-01 -6.33289993e-01
4.41297144e-01 -4.29835260e-01 1.17865992e+00 -9.28579092e-01
-1.55111349e+00 1.17877877e+00 6.71518072e-02 -2.92000175e-01
5.36268353e-01 -3.09633523e-01 -5.41603148e-01 5.03323615e-01
1.16178200e-01 5.83885074e-01 5.19705296e-01 -4.68026400e-01
-6.34702146e-01 -8.18944991e-01 2.51636803e-01 4.26265568e-01
-1.92234010e-01 4.19675605e-03 -5.74585080e-01 -7.12487400e-01
1.96413681e-01 -8.56322825e-01 -6.06816225e-02 3.38856220e-01
9.03443247e-02 -2.88864374e-01 6.87852144e-01 -5.28405249e-01
8.79721522e-01 -2.08023882e+00 3.90146598e-02 -3.62918198e-01
-3.16262878e-02 2.84105927e-01 -6.34799898e-02 3.91004354e-01
6.37854189e-02 -2.28600815e-01 -9.76311713e-02 1.08681001e-01
-8.21116865e-02 5.18701434e-01 2.88280785e-01 6.29149795e-01
2.63666138e-02 1.09367764e+00 -9.69628811e-01 -5.98732650e-01
7.51769066e-01 4.01832283e-01 -4.02691513e-01 1.93938062e-01
6.21786229e-02 6.06443703e-01 -7.68210113e-01 1.35085523e+00
1.82038859e-01 -1.85352191e-01 -8.05894136e-02 -1.38023019e-01
3.02276820e-01 3.89437117e-02 -1.30023539e+00 2.25743103e+00
2.89524719e-02 4.40245539e-01 -6.38186634e-01 -1.31098461e+00
7.53739119e-01 2.93219805e-01 1.34068990e+00 -1.07433188e+00
1.81577846e-01 -1.05780527e-01 -4.73886549e-01 -8.41655910e-01
1.40859336e-01 2.79137611e-01 -2.35735297e-01 5.45949936e-01
6.46691844e-02 1.75061598e-01 2.58672088e-01 -1.07480071e-01
1.61893797e+00 6.34813786e-01 7.48085380e-01 2.31213287e-01
5.94894469e-01 2.35197827e-01 6.16962194e-01 6.04754508e-01
-1.14182830e+00 5.70964277e-01 2.66415000e-01 -9.04147089e-01
-6.43096149e-01 -9.91557837e-01 1.83839783e-01 1.12065554e+00
3.59138191e-01 -3.26357663e-01 -5.06145895e-01 -9.28040922e-01
-6.38540462e-02 1.29763901e-01 -7.47937441e-01 -3.34949195e-01
-7.21511722e-01 -6.70141399e-01 7.18332529e-01 1.14300096e+00
1.19653678e+00 -1.23568046e+00 -1.01045024e+00 1.27962515e-01
-3.37616533e-01 -1.19261074e+00 -1.93406761e-01 -5.66631258e-02
-1.21603680e+00 -1.63721704e+00 -7.98173368e-01 -5.29374838e-01
4.30755347e-01 5.91687322e-01 9.19639111e-01 -2.57169783e-01
-5.85683286e-01 1.17667663e+00 -7.59311318e-01 -3.13098639e-01
-1.21869974e-01 -3.46453965e-01 2.47169137e-01 -9.46978256e-02
8.48553836e-01 -3.31893653e-01 -8.30620825e-01 6.05613828e-01
-7.41907418e-01 -1.37349144e-01 5.50717413e-01 3.78588527e-01
8.69502366e-01 4.91680689e-02 2.51789689e-01 -1.38449863e-01
1.12111028e-02 -2.28051737e-01 1.65215675e-02 2.63789415e-01
-9.57425684e-02 -3.44550401e-01 -1.34581357e-01 -4.97299105e-01
-9.85063493e-01 4.72612023e-01 -1.18372753e-01 -3.92805040e-01
-6.30454719e-01 -1.20837785e-01 -5.66860437e-01 1.72646139e-02
7.41376638e-01 4.94561553e-01 -3.90779646e-03 -4.98166502e-01
2.47381821e-01 6.46862924e-01 4.95237172e-01 -3.95283133e-01
2.75646120e-01 7.34358430e-01 1.25564158e-01 -9.51552749e-01
-7.84269452e-01 -7.67034173e-01 -1.19110072e+00 -6.45372748e-01
1.12822092e+00 -1.21301150e+00 -5.38542032e-01 1.16246831e+00
-7.06802070e-01 -5.46204627e-01 -5.28648615e-01 6.18202448e-01
-1.07890761e+00 5.88320673e-01 -4.26469654e-01 -7.14949310e-01
-7.94409402e-03 -8.20139170e-01 1.41672993e+00 1.11698940e-01
-2.21157983e-01 -8.02986503e-01 2.49508396e-01 1.12773013e+00
4.26312461e-02 6.96402371e-01 2.94845939e-01 -4.37783062e-01
-4.87143636e-01 -4.18564141e-01 1.95527881e-01 4.41082954e-01
4.62811619e-01 -4.20046270e-01 -7.11933970e-01 4.29304950e-02
-3.03052664e-02 -6.47436440e-01 5.30034602e-01 3.89439583e-01
1.34085655e+00 9.02258679e-02 -4.10702765e-01 3.98675710e-01
9.20617759e-01 4.78774309e-01 9.79640603e-01 4.09201533e-01
8.77943158e-01 2.70083070e-01 9.31524694e-01 7.17600584e-01
2.09712759e-01 8.21219504e-01 3.20301145e-01 5.65825626e-02
-1.82818234e-01 -1.10466391e-01 5.02182186e-01 5.32314599e-01
-6.00506246e-01 -1.78949103e-01 -9.07517612e-01 3.52983028e-01
-1.97181177e+00 -1.39278722e+00 7.90503770e-02 1.89871693e+00
4.99017715e-01 -6.81720441e-03 6.14797890e-01 5.80482304e-01
3.16365242e-01 3.59012246e-01 -8.18675160e-01 7.10101277e-02
-8.07161853e-02 1.13784000e-01 3.67490381e-01 -1.73574984e-01
-1.53276730e+00 7.34215558e-01 6.55644035e+00 6.01835728e-01
-6.15897715e-01 1.18122518e-01 3.38371575e-01 -4.49089468e-01
5.15467644e-01 -5.28004944e-01 -6.61502838e-01 2.47537628e-01
6.82953060e-01 7.18703270e-02 -7.25998655e-02 1.23289192e+00
3.44805181e-01 -4.11411524e-01 -1.07539833e+00 1.51360464e+00
5.00193179e-01 -1.11756051e+00 5.70650725e-03 5.73998280e-02
5.65946937e-01 -7.51487315e-02 -6.17930949e-01 3.67689639e-01
-1.37954757e-01 -7.63473213e-01 4.46783572e-01 5.96051037e-01
4.41457838e-01 -5.00153959e-01 4.71648872e-01 2.68676192e-01
-1.41789567e+00 -3.00196320e-01 -2.06112668e-01 -4.13937330e-01
2.47301966e-01 2.31478736e-01 -2.29896858e-01 4.29517657e-01
1.12489522e+00 1.54888058e+00 -7.18201220e-01 8.03606629e-01
-1.44669339e-01 1.95241719e-01 -8.17318857e-02 -7.53077567e-02
9.96958315e-02 6.12487681e-02 2.81632602e-01 8.53450954e-01
2.16976777e-01 5.42157829e-01 3.19075644e-01 1.96115538e-01
2.28453085e-01 -2.19854102e-01 -8.36943209e-01 -2.98318654e-01
-5.67614511e-02 7.47483611e-01 -7.71854401e-01 -3.95930529e-01
-6.09345198e-01 1.36884809e+00 -7.58001208e-02 9.76172909e-02
-1.03047526e+00 -8.58857408e-02 9.68505025e-01 1.19456559e-01
2.08332419e-01 -4.34653252e-01 1.30345151e-01 -1.23874998e+00
2.09090173e-01 -1.00994432e+00 7.58446753e-01 -9.38845038e-01
-1.04229963e+00 4.19666432e-02 3.26796830e-01 -1.58936787e+00
-1.83781818e-01 -8.66296768e-01 -1.77592397e-01 4.03584093e-02
-9.45064723e-01 -1.07992828e+00 -7.88461506e-01 1.08013046e+00
9.15688336e-01 -4.46665496e-01 7.17056870e-01 3.59390199e-01
-5.49172640e-01 1.84842318e-01 -1.73135892e-01 4.61263806e-01
3.47664386e-01 -9.60303903e-01 3.06042790e-01 6.50061548e-01
2.93266445e-01 -1.39107630e-01 8.26386455e-03 -6.43085718e-01
-1.84144521e+00 -1.06950486e+00 6.49624348e-01 -1.12632883e+00
2.09279746e-01 -1.85537070e-01 -3.88055861e-01 7.31880009e-01
-4.72433478e-01 2.76315093e-01 9.40022588e-01 -3.67471218e-01
-2.05097552e-02 -1.35617122e-01 -1.17345667e+00 3.64887357e-01
1.99652457e+00 -4.22076792e-01 -8.15990090e-01 5.47274113e-01
3.24561030e-01 -4.05980855e-01 -1.10858631e+00 6.40287161e-01
7.30386019e-01 -1.19376373e+00 1.33173954e+00 -7.33141959e-01
3.16324443e-01 -2.75999278e-01 -5.60966611e-01 -6.67443335e-01
-2.72222400e-01 -6.34546280e-02 -7.43846357e-01 6.74718142e-01
-3.83399069e-01 -2.53087252e-01 1.20646548e+00 4.10694718e-01
-2.52704531e-01 -7.79298663e-01 -1.04729772e+00 -9.66202378e-01
-4.31610823e-01 -7.70599306e-01 4.33247447e-01 7.03259230e-01
2.60668248e-02 -1.86097503e-01 -3.91742796e-01 -2.75525510e-01
4.96305913e-01 3.55448686e-02 1.12490201e+00 -9.36948419e-01
-2.05423415e-01 -2.08643451e-01 -1.21137798e+00 -1.28681242e+00
-1.16664536e-01 -4.72191125e-01 -2.17181221e-01 -1.83855391e+00
2.94171702e-02 -1.82993565e-04 -3.12078208e-01 7.76545227e-01
2.38113970e-01 4.25978690e-01 -1.81602359e-01 3.11893463e-01
-1.01986670e+00 7.17665672e-01 1.47124696e+00 -3.98785859e-01
-1.18286133e-01 1.79960936e-01 -2.39214137e-01 7.28306651e-01
7.32715249e-01 -1.28767714e-01 -6.92674994e-01 -2.66601086e-01
-2.67308384e-01 -4.64521088e-02 5.91650546e-01 -1.51093221e+00
-9.20565724e-02 -4.86319602e-01 8.95996928e-01 -7.05981314e-01
6.71868265e-01 -6.42797947e-01 1.25990108e-01 6.40721977e-01
-1.49649099e-01 -2.46016726e-01 -4.08851020e-02 7.03942776e-01
-2.01166362e-01 3.82485181e-01 5.21420896e-01 -6.95624828e-01
-1.63766515e+00 5.35071552e-01 -5.10401666e-01 1.87599435e-01
1.45700300e+00 -1.14247012e+00 -8.66038434e-04 -4.88065779e-02
-7.42128730e-01 -1.68650225e-01 3.42857927e-01 8.85852039e-01
6.74633026e-01 -1.80655265e+00 -3.14079344e-01 3.05636853e-01
6.93997145e-01 -2.03361467e-01 4.91983324e-01 8.70731294e-01
-6.06098890e-01 4.70263481e-01 -8.29736412e-01 -6.93789482e-01
-1.28946543e+00 4.21341628e-01 5.26065111e-01 8.97575915e-02
-9.04216766e-01 5.66552043e-01 -1.44739196e-01 -2.97922820e-01
5.22625148e-01 -4.35063511e-01 -1.64593697e-01 -8.11398029e-02
7.39185274e-01 8.99702728e-01 -1.13959178e-01 -8.54074180e-01
-8.48009050e-01 6.85971737e-01 5.21524251e-01 2.39360943e-01
1.18713343e+00 -4.88562547e-02 4.58723933e-01 4.92785752e-01
1.14651120e+00 -7.65612781e-01 -1.42924941e+00 -2.60963701e-02
-8.88622329e-02 -8.52947116e-01 -2.09579244e-01 -6.52941942e-01
-1.17580354e+00 7.13305414e-01 9.74623263e-01 -2.71501541e-01
1.31870747e+00 1.92785114e-01 9.16812479e-01 6.65432870e-01
7.42166400e-01 -1.42258203e+00 8.04313302e-01 3.49010170e-01
8.23310971e-01 -1.55584741e+00 2.90165812e-01 -6.13204427e-02
-7.95886755e-01 8.70824516e-01 8.15546632e-01 1.30621389e-01
6.11322165e-01 -1.78020433e-01 1.27637878e-01 -5.26578307e-01
-3.16146582e-01 -3.75262260e-01 8.66269767e-02 1.08987844e+00
2.21647620e-01 -1.67247560e-02 -5.61828353e-02 2.91565537e-01
2.88823038e-01 4.62999403e-01 2.87180454e-01 1.60611689e+00
-4.99491066e-01 -9.39120233e-01 -1.95740059e-01 2.50741363e-01
-1.44776329e-01 7.39233732e-01 -5.84661782e-01 9.39953566e-01
2.29884163e-01 8.51553440e-01 -6.78559095e-02 -3.96423757e-01
7.07948208e-01 2.74631739e-01 9.37063396e-01 -3.80161256e-01
-1.25824496e-01 -4.07410532e-01 4.95586991e-02 -1.28204203e+00
-1.23675561e+00 -8.68139744e-01 -1.20658278e+00 -2.24064991e-01
3.29865843e-01 -5.24879932e-01 3.80156964e-01 1.05865765e+00
5.32548130e-01 2.87174046e-01 4.28855687e-01 -1.02502298e+00
-1.20930821e-01 -7.46653736e-01 -7.93631077e-01 8.04771602e-01
-1.54169768e-01 -1.05797029e+00 -1.47571400e-01 2.72071689e-01] | [7.871738433837891, 0.4557730257511139] |
56edc23c-e546-4832-9718-66436db94792 | cosmix-compositional-semantic-mix-for-domain | 2207.09778 | null | https://arxiv.org/abs/2207.09778v1 | https://arxiv.org/pdf/2207.09778v1.pdf | CoSMix: Compositional Semantic Mix for Domain Adaptation in 3D LiDAR Segmentation | 3D LiDAR semantic segmentation is fundamental for autonomous driving. Several Unsupervised Domain Adaptation (UDA) methods for point cloud data have been recently proposed to improve model generalization for different sensors and environments. Researchers working on UDA problems in the image domain have shown that sample mixing can mitigate domain shift. We propose a new approach of sample mixing for point cloud UDA, namely Compositional Semantic Mix (CoSMix), the first UDA approach for point cloud segmentation based on sample mixing. CoSMix consists of a two-branch symmetric network that can process labelled synthetic data (source) and real-world unlabelled point clouds (target) concurrently. Each branch operates on one domain by mixing selected pieces of data from the other one, and by using the semantic information derived from source labels and target pseudo-labels. We evaluate CoSMix on two large-scale datasets, showing that it outperforms state-of-the-art methods by a large margin. Our code is available at https://github.com/saltoricristiano/cosmix-uda. | ['Fabio Poiesi', 'Elisa Ricci', 'Nicu Sebe', 'Giuseppe Fiameni', 'Fabio Galasso', 'Cristiano Saltori'] | 2022-07-20 | null | null | null | null | ['lidar-semantic-segmentation', 'point-cloud-segmentation'] | ['computer-vision', 'computer-vision'] | [ 3.31079125e-01 8.78079981e-02 -4.39266562e-01 -6.10410333e-01
-9.01647806e-01 -6.15119874e-01 7.93105602e-01 -2.94250362e-02
-4.89726305e-01 5.27144551e-01 -4.02584523e-01 -3.23149443e-01
9.97771323e-02 -1.01043665e+00 -1.06334007e+00 -5.79790354e-01
3.29763174e-01 1.30503368e+00 7.30509043e-01 -8.55062976e-02
1.32942989e-01 5.92799664e-01 -1.88410234e+00 8.08428451e-02
1.16859984e+00 7.41752625e-01 4.23845708e-01 4.76906449e-01
-5.70826471e-01 7.09661171e-02 -2.94448704e-01 -1.17063589e-01
7.50295162e-01 -1.70395821e-01 -7.85939276e-01 9.68203545e-02
6.89897418e-01 1.65188149e-01 5.82451075e-02 1.24537826e+00
3.50828439e-01 8.89943466e-02 8.82604539e-01 -1.85462523e+00
-1.14049293e-01 5.30865371e-01 -5.53399682e-01 -7.08305389e-02
-4.06644821e-01 2.14386001e-01 7.25544751e-01 -8.95347893e-01
7.67281592e-01 1.55042160e+00 6.68173671e-01 6.81130886e-01
-1.49988878e+00 -1.12944269e+00 2.54623353e-01 1.19726673e-01
-1.34501696e+00 -1.24313861e-01 9.01933134e-01 -6.76181078e-01
4.85207021e-01 -6.53537782e-03 5.91792166e-01 1.08949673e+00
-1.89036638e-01 9.42568183e-01 1.37004685e+00 -1.58888504e-01
6.24576509e-01 1.97948053e-01 3.26584637e-01 2.46642411e-01
4.78380054e-01 2.92147577e-01 -3.86312336e-01 -9.36169475e-02
2.75525212e-01 -1.78197727e-01 1.96537688e-01 -1.00462127e+00
-1.02420855e+00 9.75725114e-01 4.60306197e-01 -3.39418203e-01
-1.28594920e-01 2.78176777e-02 3.39533687e-01 1.73849791e-01
6.76138699e-01 1.67942151e-01 -5.65239787e-01 2.32446924e-01
-9.00659800e-01 5.90350389e-01 5.56344211e-01 1.32063603e+00
1.25021827e+00 -1.27832070e-01 2.63672501e-01 9.07841265e-01
5.36881387e-01 9.80279148e-01 2.22380668e-01 -1.32926381e+00
3.67002338e-01 4.41673905e-01 7.17084408e-02 -1.56075701e-01
-3.56616139e-01 -1.90444529e-01 -4.66304928e-01 6.71390951e-01
4.96982068e-01 -3.70938480e-02 -1.50310731e+00 1.58538461e+00
6.88934743e-01 5.29365063e-01 3.22800279e-01 6.54934585e-01
9.07439291e-01 5.13183951e-01 1.68940842e-01 3.79522324e-01
1.08977938e+00 -9.86278117e-01 -6.95461780e-02 -6.19933069e-01
4.27618951e-01 -4.22659487e-01 1.16134965e+00 4.15183067e-01
-5.38938046e-01 -8.11759651e-01 -1.23228598e+00 2.45699212e-02
-6.97358668e-01 -4.10663575e-01 2.32464090e-01 7.48533010e-01
-8.79017532e-01 5.55530787e-01 -7.52913654e-01 -6.53763652e-01
7.83459723e-01 2.46750370e-01 2.36561410e-02 -2.72848070e-01
-8.39914799e-01 6.51460230e-01 5.69456577e-01 -4.41369832e-01
-1.08052492e+00 -7.41893232e-01 -8.20996046e-01 -5.99685133e-01
3.83755952e-01 -7.65064120e-01 1.46506357e+00 -6.29195631e-01
-1.35515213e+00 1.16635728e+00 -2.94744670e-01 -8.50523293e-01
6.16500497e-01 -2.06713647e-01 -1.83246359e-01 -5.92738204e-02
6.48014486e-01 1.53225374e+00 9.31180239e-01 -1.90678942e+00
-1.09994233e+00 -5.96590281e-01 -3.03711206e-01 1.93882331e-01
4.75741714e-01 -5.23936152e-01 -3.25837880e-01 -2.52102315e-01
3.57392192e-01 -1.22870862e+00 -3.91713172e-01 -3.38467397e-03
-3.45231444e-01 -2.98319638e-01 9.97607470e-01 -1.23947449e-01
2.41517842e-01 -2.17291021e+00 -2.13715229e-02 2.43193924e-01
1.37714699e-01 2.09385380e-01 -2.40338922e-01 9.51191038e-02
-6.20180694e-03 3.57237868e-02 -9.70751703e-01 -5.75279355e-01
5.80619015e-02 7.08538055e-01 -4.13906276e-01 3.56907517e-01
1.68489277e-01 7.66888201e-01 -1.11727226e+00 -5.43225408e-01
4.20875043e-01 1.08312927e-01 -4.14650112e-01 -7.11445436e-02
-8.49189341e-01 8.90715182e-01 -2.83760875e-01 5.92266440e-01
1.45830965e+00 1.19140185e-01 -1.77248776e-01 2.98203498e-01
-5.33215478e-02 6.46563172e-02 -1.28428674e+00 1.85191286e+00
-1.98375672e-01 5.40783703e-01 -2.39298940e-02 -8.98112118e-01
1.16022444e+00 -2.29133472e-01 6.04060590e-01 -7.08755255e-01
1.35182753e-01 5.71506143e-01 -2.45690316e-01 -1.98322549e-01
6.68684363e-01 -2.56869555e-01 -9.44944844e-02 1.36691943e-01
8.72068256e-02 -1.01730645e+00 2.03476697e-01 6.44099265e-02
6.29168212e-01 4.76321340e-01 -1.31516486e-01 -3.74358714e-01
3.43506217e-01 6.13456488e-01 8.03131640e-01 8.88989747e-01
-4.90649790e-01 8.44789743e-01 1.08715340e-01 -1.81050375e-01
-1.24611914e+00 -1.45016873e+00 -3.01034570e-01 7.89060235e-01
6.97845936e-01 -5.31418733e-02 -8.49743903e-01 -6.97803795e-01
5.47038198e-01 1.18429935e+00 -3.61363798e-01 -5.49345315e-02
-3.75620842e-01 -5.45800984e-01 4.51762259e-01 5.18940210e-01
7.49598324e-01 -9.77806985e-01 -5.55249751e-01 -8.90251026e-02
-1.65215850e-01 -1.31864321e+00 6.05357736e-02 3.99062902e-01
-1.06186926e+00 -9.32971120e-01 -4.50124979e-01 -5.57370603e-01
3.12180370e-01 5.13137877e-01 1.19424188e+00 -5.32291651e-01
1.32861912e-01 2.71744609e-01 -2.89100260e-01 -9.53012407e-01
-6.59371912e-01 2.77256697e-01 5.12477458e-02 -5.69936372e-02
7.41026580e-01 -6.36541665e-01 -2.23466516e-01 5.18647373e-01
-1.02939785e+00 1.30460754e-01 3.66589546e-01 2.98521966e-01
8.87678623e-01 -1.83889210e-01 3.63970309e-01 -1.21535885e+00
2.84716040e-01 -5.85283995e-01 -8.33290935e-01 -3.34960997e-01
-7.34113336e-01 1.11406125e-01 3.03746723e-02 -1.98655680e-01
-9.40879703e-01 4.27489072e-01 -2.08098590e-01 -6.96027696e-01
-7.68547177e-01 -7.94381499e-02 -2.32922241e-01 1.59551874e-02
1.08804703e+00 5.44822365e-02 1.39463067e-01 -4.88196760e-01
7.92210162e-01 8.07548106e-01 8.27606678e-01 -8.09575975e-01
1.09766865e+00 1.04567897e+00 1.26731932e-01 -8.88722122e-01
-5.99198878e-01 -9.32668209e-01 -9.45123434e-01 -1.66876018e-01
1.03582811e+00 -1.09389627e+00 3.20019498e-02 7.22819865e-01
-1.06975317e+00 -6.50218129e-01 -8.24964464e-01 3.76333028e-01
-7.34266341e-01 2.17397615e-01 3.81744541e-02 -6.12828910e-01
1.20411523e-01 -1.22285283e+00 1.36470902e+00 1.72408685e-01
9.21755880e-02 -5.71715891e-01 2.54681945e-01 3.73605877e-01
-1.10141188e-01 3.22184235e-01 6.91010416e-01 -7.15657532e-01
-7.54422724e-01 2.48277739e-01 -3.18439126e-01 5.82234800e-01
-8.67113546e-02 8.10543634e-03 -1.32909060e+00 -1.09933726e-01
-6.01357855e-02 -6.40410557e-02 1.20227909e+00 5.22830129e-01
9.28849339e-01 3.61865371e-01 -5.18389821e-01 6.76678419e-01
1.21369278e+00 2.05436245e-01 4.40803528e-01 5.47104299e-01
7.32289016e-01 5.87393224e-01 9.38954651e-01 -1.24734258e-02
7.38597572e-01 3.76628131e-01 8.24921906e-01 -1.32602766e-01
-2.53334552e-01 -3.55484426e-01 9.07236710e-02 4.18974638e-01
2.34111920e-01 -9.05928612e-02 -1.26990724e+00 9.33902621e-01
-1.99355638e+00 -5.08017063e-01 -4.95897561e-01 2.19671535e+00
4.60391223e-01 5.11992931e-01 5.32925189e-01 -7.69551145e-03
7.92584836e-01 7.01550543e-02 -9.34254825e-01 -1.92747563e-01
-2.47623652e-01 3.41374278e-01 1.08990383e+00 5.88555098e-01
-1.20195627e+00 1.26572478e+00 5.18099928e+00 8.96054327e-01
-8.74273181e-01 4.75901008e-01 2.52807975e-01 2.40977541e-01
-5.21523714e-01 2.12475836e-01 -8.44550014e-01 4.36204225e-01
8.82543564e-01 8.03032611e-03 1.33391961e-01 1.13276720e+00
-9.87584144e-02 -2.63800919e-01 -9.73396838e-01 9.15080905e-01
-2.63444752e-01 -1.25548410e+00 -3.73930181e-03 1.21187143e-01
9.36437190e-01 8.94502163e-01 -4.94197905e-02 3.76815051e-01
7.44470537e-01 -4.85214591e-01 9.61718082e-01 2.40380257e-01
6.60907686e-01 -6.53261423e-01 4.54556525e-01 6.04242682e-01
-1.16216648e+00 -1.27944546e-02 -5.24496019e-01 2.23600045e-01
5.60117029e-02 5.42128325e-01 -9.51323926e-01 7.13936865e-01
1.00515997e+00 8.37924540e-01 -5.69860697e-01 1.06745970e+00
-2.95841902e-01 5.89117527e-01 -6.36332393e-01 4.86079425e-01
3.15234363e-01 -6.71844363e-01 9.32742953e-01 1.13118505e+00
1.50333688e-01 -3.81920189e-01 2.34874263e-01 9.69176888e-01
-1.19065475e-02 -1.79262757e-01 -9.60186243e-01 2.98706293e-01
6.29402161e-01 9.56071794e-01 -9.93045092e-01 -5.08687019e-01
-1.65574059e-01 6.73521399e-01 -1.26228243e-01 3.06460768e-01
-8.06548774e-01 -1.64053917e-01 1.00088704e+00 2.42495999e-01
3.68634552e-01 -4.54736978e-01 -8.02970529e-01 -7.91421473e-01
-1.28485724e-01 -5.19302130e-01 3.60233754e-01 -9.31656301e-01
-1.24254429e+00 4.35183704e-01 5.80501914e-01 -1.54695809e+00
-2.87626721e-02 -6.58913910e-01 -3.12867761e-01 7.38746643e-01
-1.99110425e+00 -1.19093752e+00 -7.25017846e-01 4.31386888e-01
8.46555352e-01 -3.81334312e-02 3.84331286e-01 1.31338775e-01
-2.00348884e-01 3.32393385e-02 2.05991670e-01 -2.10559100e-01
7.12949395e-01 -1.39334011e+00 1.05399525e+00 8.89779687e-01
2.01500788e-01 -7.32049765e-03 7.05381811e-01 -7.73387671e-01
-6.58753276e-01 -1.59750128e+00 4.61875975e-01 -7.22838700e-01
3.34932059e-01 -6.31138086e-01 -9.55374181e-01 7.28870273e-01
1.21130005e-01 -7.53185824e-02 1.68847024e-01 -2.37049893e-01
-4.69756573e-01 -3.04749042e-01 -1.43276834e+00 5.24475157e-01
1.30974364e+00 -3.29665035e-01 -5.94454944e-01 4.02873218e-01
1.03835917e+00 -4.75956857e-01 -3.64521742e-01 8.45242679e-01
4.66915667e-02 -1.13940263e+00 9.61669445e-01 -2.48053759e-01
2.18232244e-01 -7.51599491e-01 -3.45746487e-01 -1.32107830e+00
-4.97008748e-02 -3.06961745e-01 2.70085454e-01 9.47926760e-01
5.04290164e-01 -1.01449358e+00 1.06201029e+00 1.53830424e-01
-4.79594201e-01 -1.38113603e-01 -1.03140116e+00 -1.24345446e+00
5.98943472e-01 -9.26974058e-01 9.47369039e-01 8.00477445e-01
-8.12722087e-01 3.84619832e-01 2.94181824e-01 4.39089000e-01
9.85514462e-01 2.91674942e-01 1.32881165e+00 -1.81805015e+00
1.63321495e-01 -4.36286539e-01 -4.05469179e-01 -1.08953595e+00
2.16220081e-01 -1.20499337e+00 2.95654416e-01 -1.60101783e+00
-1.73262760e-01 -8.54256749e-01 6.20300733e-02 2.61233538e-01
1.68798342e-01 3.05939108e-01 2.66942143e-01 4.74000990e-01
-2.43859962e-01 6.04960561e-01 1.00343156e+00 -4.46412235e-01
-2.88092792e-01 2.59688020e-01 -4.00806010e-01 8.60545397e-01
1.16442263e+00 -8.67642701e-01 -5.45616984e-01 -3.49319309e-01
-1.96155146e-01 -5.61093271e-01 4.44507092e-01 -1.49227619e+00
1.64516289e-02 -3.17814589e-01 -4.91173528e-02 -1.06510174e+00
4.95943606e-01 -7.66638041e-01 2.62082160e-01 3.56854379e-01
1.15057386e-01 -3.21741462e-01 3.84372503e-01 5.96521318e-01
-1.38645485e-01 -2.30634898e-01 1.24059010e+00 -1.54345796e-01
-1.15483367e+00 3.49093765e-01 -5.01493551e-02 2.94083804e-01
1.15453589e+00 -5.73680997e-01 -2.31599301e-01 -4.54298481e-02
-6.86795950e-01 4.19092506e-01 8.18284929e-01 5.75137436e-01
4.21393275e-01 -1.22742939e+00 -7.57937372e-01 4.01647985e-01
6.09711707e-01 9.73286867e-01 -1.32191069e-02 4.61486459e-01
-4.66189831e-01 1.33208334e-01 -2.01626003e-01 -1.34711146e+00
-1.04752910e+00 5.29216826e-01 4.67382699e-01 1.99327022e-01
-6.57791257e-01 9.64577377e-01 3.69521946e-01 -1.26763403e+00
-6.38237819e-02 -6.11559570e-01 1.17540710e-01 3.67262997e-02
-1.94521844e-01 3.51924747e-01 1.04148351e-01 -7.77262807e-01
-3.30734134e-01 8.09472978e-01 2.26698592e-01 -4.06773388e-01
1.06845200e+00 2.90340674e-03 1.02149718e-01 7.19437540e-01
8.52746606e-01 -2.66053379e-01 -1.50396967e+00 -5.57665646e-01
1.75752833e-01 -5.50692320e-01 -2.80153513e-01 -5.95785797e-01
-9.30651128e-01 9.93773699e-01 8.69500637e-01 9.98358615e-03
8.52415323e-01 2.06184953e-01 6.07611358e-01 2.31036440e-01
5.64723432e-01 -1.21159923e+00 -2.28813529e-01 5.99052966e-01
5.99514067e-01 -1.44031131e+00 -3.19355190e-01 -5.77237666e-01
-8.50843191e-01 5.48447251e-01 5.92343748e-01 -2.80973464e-01
7.99344361e-01 9.19663236e-02 4.35533792e-01 -2.01434702e-01
-1.71125308e-01 -7.01472461e-01 -1.00999601e-01 1.07988691e+00
-4.96097147e-01 3.33042413e-01 1.14715733e-01 1.80557564e-01
-2.95772880e-01 -2.68203057e-02 3.47405821e-01 9.59226131e-01
-6.37309909e-01 -1.33805788e+00 -6.21458590e-01 3.16451877e-01
3.61987442e-01 2.24969611e-01 -4.99752671e-01 9.28090990e-01
6.29675388e-01 8.80246401e-01 3.29039395e-01 -4.94129986e-01
5.33521950e-01 2.74481893e-01 2.69279838e-01 -7.69050241e-01
6.08322099e-02 -5.11126779e-02 -8.72586742e-02 -4.89936501e-01
-6.48314297e-01 -1.13075304e+00 -1.45286500e+00 -5.70928566e-02
-1.06930472e-02 1.21289857e-01 1.05513263e+00 6.23952389e-01
4.85231131e-01 3.50493401e-01 4.94911611e-01 -1.15056503e+00
-2.69486248e-01 -9.50938404e-01 -5.75410187e-01 3.34676325e-01
4.54885364e-01 -9.92627203e-01 -2.92208076e-01 9.71701555e-03] | [8.149197578430176, -2.6085610389709473] |
2fd5e3a6-50c4-4d36-aca5-caa811a8354a | efficient-algorithms-for-exact-graph-matching | 2305.19666 | null | https://arxiv.org/abs/2305.19666v2 | https://arxiv.org/pdf/2305.19666v2.pdf | Efficient Algorithms for Exact Graph Matching on Correlated Stochastic Block Models with Constant Correlation | We consider the problem of graph matching, or learning vertex correspondence, between two correlated stochastic block models (SBMs). The graph matching problem arises in various fields, including computer vision, natural language processing and bioinformatics, and in particular, matching graphs with inherent community structure has significance related to de-anonymization of correlated social networks. Compared to the correlated Erdos-Renyi (ER) model, where various efficient algorithms have been developed, among which a few algorithms have been proven to achieve the exact matching with constant edge correlation, no low-order polynomial algorithm has been known to achieve exact matching for the correlated SBMs with constant correlation. In this work, we propose an efficient algorithm for matching graphs with community structure, based on the comparison between partition trees rooted from each vertex, by extending the idea of Mao et al. (2021) to graphs with communities. The partition tree divides the large neighborhoods of each vertex into disjoint subsets using their edge statistics to different communities. Our algorithm is the first low-order polynomial-time algorithm achieving exact matching between two correlated SBMs with high probability in dense graphs. | ['Hye Won Chung', 'Dongpil Shin', 'Joonhyuk Yang'] | 2023-05-31 | null | null | null | null | ['graph-matching'] | ['graphs'] | [ 0.41280547 0.29342628 -0.2730768 0.04222735 -0.3929605 -0.6257282
0.38333195 0.7622422 -0.0447631 0.47117275 -0.26477045 -0.34141544
-0.61133116 -1.2922541 -0.50561047 -0.9157914 -0.5094566 1.0153807
0.49358824 -0.04409095 0.02488483 0.4463059 -1.1577901 -0.09344656
0.8983442 0.33292815 -0.13380142 0.74845463 -0.07078172 0.5261127
-0.02065226 -0.76828545 0.5420987 -0.68147725 -0.90058196 0.03690619
0.2624778 0.552589 -0.9079038 1.5595106 0.30706713 -0.2838963
0.64630276 -1.9037145 -0.5177119 0.98162043 -1.0570112 0.09917287
0.4727962 -0.5795341 1.2265003 -0.02152834 0.78814 1.253948
0.82159245 0.17991678 -1.6185493 -0.8620984 -0.4059028 0.39423543
-1.6603945 0.08459254 0.8075884 -0.5474995 0.24555871 0.40497765
0.7410329 0.38901848 -0.0147411 0.4287899 1.1090083 -0.489525
0.177574 -0.08031772 0.24403058 0.6361419 1.0087858 0.11774593
-0.16725548 -0.7303043 0.39455518 0.12873606 -0.27170268 -1.0178713
-0.95898736 0.8731006 0.47741953 0.7230619 0.08800749 0.22000954
0.41580638 0.7203658 0.45626804 0.11221402 -0.11076808 0.59445286
-0.75881153 0.14999986 1.1056334 1.2189287 0.9852926 -0.60448676
0.14166671 0.28156707 0.03349779 0.60410553 -0.24940099 -0.469712
0.54076993 0.72477925 -0.39139554 -1.8278292 -0.4120234 -0.43075413
-1.5896555 -0.40462488 0.6120121 0.43112102 -0.20349722 1.7453305
0.44129923 0.42029327 0.08494242 0.4329808 0.46917155 0.23060837
-0.3003213 -0.41765308 1.3012497 -0.48610553 -0.45346978 0.18987282
0.83011115 -0.6800204 0.20769416 -0.02750893 -0.68771416 0.03718745
-0.83735657 0.3216969 -0.08560134 -0.72278625 0.68026525 0.90269333
-1.2281933 0.8845649 -0.5261211 -0.7530895 0.58694905 0.38740873
-0.68078613 -0.38929248 -0.99862766 0.402369 0.16505578 -0.39568517
-0.20369424 -0.4549474 -0.7910057 0.00868972 0.3623752 -1.0112432
0.3881171 -1.0585868 -0.6696162 1.2189912 -0.39175695 -0.7523066
0.42399904 0.82061833 -0.36389333 0.36632726 0.22387126 -0.16057357
0.46530983 -0.99707645 -0.40307775 -0.7455 -0.0639225 -0.3150706
-0.37165633 0.12952502 -0.05579519 -0.42245045 0.58127075 -1.2195376
-0.7359296 -0.0327512 -0.4319475 -0.08811806 0.5325737 -0.31978017
1.3195331 -1.9046819 0.3492598 1.0113486 0.8329287 -0.157461
-0.11319675 0.7268284 -0.23421502 0.18946001 -0.60069984 -0.01470804
-0.23953934 0.08236657 0.16039398 1.109048 -0.37819037 0.73386246
-1.1845725 -0.64405984 -0.252752 -0.02993146 -0.60797065 -0.16431256
0.23663524 0.27352783 -0.2944048 0.40202945 1.2244298 -0.5968413
0.8217655 0.1229552 0.3597205 -0.30784336 -1.4166824 1.1989077
-0.09599953 0.7410946 0.23625551 -1.2702614 1.1427683 0.468597
0.7974271 -0.24633396 0.04568747 0.40768453 0.01349979 -0.15245117
0.39551586 -0.09475096 -0.02734309 0.5491925 -0.4903794 0.05574089
0.42008522 0.68745714 1.7983679 -0.81798214 0.43661863 -0.59078294
0.71201265 -0.09766686 0.46575084 0.67720497 -0.11554044 0.548639
0.7399606 -0.20099545 -1.2774508 -0.95829475 0.03436432 0.07795287
0.40895182 -0.6460643 -0.7578565 -0.17661527 0.17961486 -0.01215169
-0.5183094 -0.35176092 -0.28883323 -0.63271606 0.3126126 0.0661356
0.41535664 -0.5037465 0.20597005 0.13291106 -0.2485231 -1.1791848
-0.70039696 -0.14830004 -1.1400288 -1.7191173 -0.5373343 -1.1394876
1.0637366 0.64862686 1.2559652 0.44997072 -0.27344635 0.21069582
-0.23533201 0.08723874 -0.7379033 0.2789314 0.02082562 0.13977239
0.46768808 -0.93576777 -0.27677613 0.49545997 -0.94278485 -0.10110389
0.17110424 0.5940928 0.6412036 0.39558104 0.0911301 -1.2091032
0.33782515 -0.76729804 -0.94100684 0.33135876 -0.6832475 0.23081526
0.63063973 -0.1466407 -0.22185062 0.3891628 0.36211988 -0.33093506
0.31030205 0.31772804 -0.27013126 -0.4887797 0.46183047 0.19996198
0.17628358 -0.26131698 0.33379087 0.6987049 0.354842 -0.34279162
1.3071346 0.80149496 0.9301673 -0.5668775 -0.32219553 -0.94165796
-0.7309971 -0.04025678 0.6025187 -0.69680655 -0.8587443 0.49622503
-1.1177715 0.35490212 -0.03149363 0.21834567 -0.6767371 0.9982808
-0.47652793 -0.7338818 -0.31878933 -0.60530555 0.6239815 -0.03041966
-0.1951963 -1.0760193 0.56263626 0.13457881 0.09455751 0.48842636
1.3008194 -0.69588274 -0.77600986 -0.49193063 -0.23824218 -0.27275935
0.1683383 -0.11002333 -0.38192567 -0.621489 -0.40325433 0.33080095
0.5350009 0.2556366 0.76634055 -0.29667887 -0.94887775 0.6818576
1.4928246 -0.02976191 0.7147397 0.18806314 0.6753418 0.63339657
0.09711548 0.05986617 0.22566812 0.571014 0.18390235 0.03300423
0.303639 -0.5225329 -0.16169177 1.026848 0.04492365 -0.2773007
-0.6457936 0.6447357 -1.9744128 -1.2602934 -1.0312258 2.564309
0.46582744 -0.18107848 0.22008349 0.38604274 1.5344249 -0.04546394
-0.15065333 -0.24293406 -0.45985028 0.10632193 0.95086163 0.500776
-0.7476444 0.49942473 5.2955446 0.871309 -0.15604709 0.09630925
0.5363764 0.5151118 -0.8563573 0.6690701 -0.27807853 0.37398663
0.72963285 -0.94040745 0.51326126 0.6881952 -0.24574319 0.13089772
-0.94451183 1.1107706 -0.23387408 -1.202079 -0.20406784 0.4451072
1.3076233 -0.3943719 -0.25513038 -0.41429165 0.50119704 -0.6898965
0.13508089 0.40622744 0.5893056 -0.9984971 0.59170276 0.39362928
-1.7846557 0.17985328 -0.6969428 0.23158956 -0.15689835 0.85831493
-0.5492502 1.0048587 0.4961962 0.77258384 -0.33601055 1.4142324
0.16357009 0.5110091 -0.5077293 -0.10825451 -0.25913695 -0.89242166
0.99741125 0.62891644 0.4386683 0.11103093 -0.01034024 0.54318243
-0.13901608 0.47719046 -0.9465427 -0.14200646 0.6049423 1.1324362
-1.1883005 -0.155492 -0.25012216 0.9942822 0.28806925 -0.06343311
-0.3690576 -0.42062718 0.69792664 0.61680233 0.28144255 -0.18422954
-0.25727296 -1.0021143 -0.12269868 -0.8027972 0.7065801 -0.2702413
-1.5549394 0.7000473 -0.11040009 -1.3419088 0.01597335 -0.07088225
-0.4793299 0.7588574 -0.8625048 -0.7564807 -0.11477765 0.7409815
-0.45895284 -0.14341648 0.83986956 0.3994616 -0.0956208 0.5105625
0.60347027 0.02703438 0.308424 -1.1727911 0.58740854 0.8763571
0.35211152 0.38622728 0.8896197 -0.8027019 -1.3299067 -0.9545678
1.5630538 -0.06726225 0.80688584 -0.47337952 -0.8268752 0.47292155
-0.04503325 0.18714999 0.48252267 0.09043623 -0.49818942 -0.3721719
-1.3602511 0.6417223 1.5435305 -0.4947736 0.06923522 0.598938
0.6243235 -0.05963166 -0.879242 0.14983848 0.23449197 -0.79338276
0.789404 -0.6521345 0.13364585 -0.5581646 0.05894732 -1.1086563
-0.59464484 -1.0071166 0.22791573 1.2207845 0.1363907 -0.8556758
1.2680283 0.15204199 0.6783472 -0.42542124 -1.2048534 -0.98881274
-0.06697514 -0.17787267 0.8770259 1.073675 0.01122694 0.31548208
-0.39104933 0.34928942 1.2460643 0.5456438 0.9406597 -1.8928697
-0.28449082 -0.4936462 -1.1483947 -0.62082016 0.29605293 -1.4292622
-0.46842772 -1.5549685 0.79725164 -0.69995254 0.19637598 -0.03984374
0.0170553 0.16782513 0.04405718 0.3145139 -0.49024135 0.25289354
1.0626366 -0.33505318 -0.04974769 0.607187 -0.6158187 0.30100802
0.6217414 -1.1657522 -0.19666073 0.34497553 0.5156842 0.32159904
0.29014915 -0.841035 0.5062462 0.04341591 -0.42515978 -0.5095019
-0.32747492 -1.2319524 1.0919583 1.0008534 -0.19369173 0.4014298
-0.31851673 1.142748 -0.13697636 -0.3920374 0.85060936 0.05205359
-0.20054875 0.80595714 -0.3262469 0.18656743 1.3452691 -0.34932375
-0.17437291 -0.73310035 -0.5432172 0.2839798 0.90441513 0.12403423
0.29249355 -1.444052 -0.8460642 -0.07399951 0.30949587 -0.08918341
0.26078412 0.8802022 -0.508319 0.24101306 0.0338439 -0.6533335
-1.8314836 0.9541368 0.32209983 -0.5574314 -0.4059252 0.56733054
0.00769141 -0.27661642 -0.15991166 0.28205094 0.1034137 -0.11824004
0.14301631 0.7100754 -0.07256968 -0.8948299 -0.35283533 0.79284793
0.06831717 0.26628074 1.1568056 -0.19320537 -0.8293999 0.07360747
1.5367879 0.23171948 -0.28452736 -0.7589974 0.3538251 -0.6883146
-0.45753738 0.3464148 -1.4243492 0.34905878 0.25369647 0.5947402
1.0057304 0.23207717 0.66239846 0.21224806 0.866971 -0.50761586
-0.34543854 0.2481745 0.32453775 -0.67509604 0.12457838 -0.9469672
-0.08757564 1.0623934 0.18038347 -0.3331766 0.9544483 0.31395054
-0.5521108 -0.18168633 -0.45931458 -0.32297868 0.04606012 0.7998927
-0.08510584 0.26563278 -0.672501 0.24281004 -0.2494594 -0.16164152
0.4807997 0.34940067 -0.20919813 -1.4746298 -0.37712908 0.6245406
-0.20651332 -0.13636336 -0.61392593 0.5969189 -0.14085649 0.92111635
0.2287465 -0.42361498 0.11097754 -0.5016646 0.49626255 -0.3701809
-0.38088533 -0.5217768 -0.14341256 -0.1936141 -0.56384367 -0.82167447
-0.861163 -0.9908832 -0.53028077 0.5117372 0.13983113 0.69530064
0.14095327 -0.13374193 1.1108325 -0.08062945 -0.3784612 -0.40929577
-1.2153372 0.47800964 -0.04724306 -0.25448826 -0.7166813 -0.41939265] | [6.918465614318848, 5.194827079772949] |
73e39d1e-1e45-4435-9823-4e25879fa15a | pie-a-parameter-and-inference-efficient | 2204.13957 | null | https://arxiv.org/abs/2204.13957v2 | https://arxiv.org/pdf/2204.13957v2.pdf | PIE: a Parameter and Inference Efficient Solution for Large Scale Knowledge Graph Embedding Reasoning | Knowledge graph (KG) embedding methods which map entities and relations to unique embeddings in the KG have shown promising results on many reasoning tasks. However, the same embedding dimension for both dense entities and sparse entities will cause either over parameterization (sparse entities) or under fitting (dense entities). Normally, a large dimension is set to get better performance. Meanwhile, the inference time grows log-linearly with the number of entities for all entities are traversed and compared. Both the parameter and inference become challenges when working with huge amounts of entities. Thus, we propose PIE, a \textbf{p}arameter and \textbf{i}nference \textbf{e}fficient solution. Inspired from tensor decomposition methods, we find that decompose entity embedding matrix into low rank matrices can reduce more than half of the parameters while maintaining comparable performance. To accelerate model inference, we propose a self-supervised auxiliary task, which can be seen as fine-grained entity typing. By randomly masking and recovering entities' connected relations, the task learns the co-occurrence of entity and relations. Utilizing the fine grained typing, we can filter unrelated entities during inference and get targets with possibly sub-linear time requirement. Experiments on link prediction benchmarks demonstrate the proposed key capabilities. Moreover, we prove effectiveness of the proposed solution on the Open Graph Benchmark large scale challenge dataset WikiKG90Mv2 and achieve the state of the art performance. | ['Wei Chu', 'Xiexiong Lin', 'Taifeng Wang', 'Linlin Chao'] | 2022-04-29 | null | null | null | null | ['entity-typing'] | ['natural-language-processing'] | [-4.06685650e-01 4.28834587e-01 -3.95707816e-01 -2.22903237e-01
-1.78311363e-01 -5.30131638e-01 2.58057386e-01 3.82730603e-01
-4.25878793e-01 6.80525482e-01 3.34284276e-01 -3.16432387e-01
-6.02386475e-01 -1.27966070e+00 -8.43415797e-01 -4.55461383e-01
-4.20893013e-01 8.94762635e-01 1.97708070e-01 -7.21050277e-02
-2.49683172e-01 4.02237996e-02 -1.21472585e+00 1.43178895e-01
1.12734020e+00 9.15941000e-01 -1.77178428e-01 3.65296572e-01
-3.77281904e-01 7.04933345e-01 -2.26712286e-01 -1.04970276e+00
1.80958137e-01 4.08592075e-01 -1.10474300e+00 -4.22381401e-01
4.46715683e-01 -2.71555901e-01 -8.88437748e-01 1.09490466e+00
4.94277477e-01 1.23791158e-01 5.86818635e-01 -1.45615768e+00
-9.58357275e-01 1.07958114e+00 -6.25057697e-01 3.37297976e-01
3.95884305e-01 -4.06961590e-01 1.54235613e+00 -1.11801958e+00
7.50399530e-01 1.24827242e+00 6.74888253e-01 1.69628307e-01
-1.38808095e+00 -6.26483858e-01 3.88354391e-01 5.84803045e-01
-1.69951224e+00 -1.36191487e-01 3.97331089e-01 -4.11253601e-01
1.16789639e+00 4.37805742e-01 3.22653919e-01 7.80788660e-01
-2.57062703e-01 5.94675124e-01 4.61472720e-01 1.50375990e-02
-2.40462467e-01 3.46077591e-01 5.97535789e-01 1.00254488e+00
1.00492311e+00 -4.01080132e-01 -4.62121844e-01 -3.13337654e-01
3.24308068e-01 1.81222782e-01 -5.09992301e-01 -3.42453539e-01
-1.24091101e+00 6.21017277e-01 8.52272391e-01 4.70471494e-02
-3.43818635e-01 2.42883757e-01 5.35297632e-01 3.27581286e-01
2.43249759e-01 3.64142209e-01 -6.88447118e-01 1.81578070e-01
-3.02244782e-01 1.55388251e-01 1.05918574e+00 1.16597378e+00
8.82844090e-01 -2.84375697e-01 -1.81468278e-01 8.52757573e-01
3.06325465e-01 4.21169072e-01 1.11548461e-01 -4.36004966e-01
1.02094471e+00 1.06819737e+00 -2.17317231e-02 -1.79527199e+00
-3.46221387e-01 -3.44216079e-01 -1.08557594e+00 -6.46351457e-01
1.81875989e-01 -1.49955258e-01 -5.94503522e-01 1.68621242e+00
7.22738087e-01 3.36313367e-01 4.16269414e-02 7.20398188e-01
1.07691646e+00 6.06593311e-01 8.10444206e-02 1.25035718e-01
1.77189827e+00 -9.92400706e-01 -7.35959530e-01 -2.86158510e-02
9.00729954e-01 -3.62004906e-01 6.64337993e-01 -9.44125280e-02
-7.25402117e-01 -2.10387170e-01 -8.80605817e-01 -3.25769782e-01
-7.58038878e-01 1.72449593e-02 1.20787346e+00 5.08379042e-01
-5.86909533e-01 6.21054828e-01 -7.09433794e-01 -1.03670493e-01
3.00204307e-01 5.64087629e-01 -8.26195002e-01 -3.14902365e-01
-1.58075106e+00 7.17900753e-01 7.96384037e-01 4.29183275e-01
-3.63468498e-01 -9.94997740e-01 -1.01973295e+00 3.95999491e-01
6.58236027e-01 -9.34965968e-01 3.08507413e-01 1.62343085e-01
-7.69577444e-01 4.87342626e-01 -1.19081892e-01 -3.62401247e-01
1.03951193e-01 -3.99195313e-01 -7.44022012e-01 -9.81694367e-03
1.32908881e-01 1.56162426e-01 4.78893340e-01 -8.99799526e-01
-5.83703578e-01 -5.13291955e-01 5.78237057e-01 1.93731084e-01
-9.05257940e-01 -4.27271426e-01 -8.39104831e-01 -4.66453969e-01
3.18534493e-01 -8.72237504e-01 5.38888648e-02 -3.16072494e-01
-7.92537034e-01 -4.75005478e-01 6.43134773e-01 -7.44567275e-01
1.66504788e+00 -1.90851521e+00 4.42711622e-01 4.36365485e-01
8.49712074e-01 9.50446427e-02 5.48724197e-02 5.22340894e-01
-1.89817458e-01 2.42371961e-01 1.41982198e-01 -5.23939803e-02
2.50394791e-01 4.74977791e-01 -2.31985018e-01 2.47425005e-01
7.13235438e-02 1.13391972e+00 -9.21297669e-01 -5.81581056e-01
-2.09413037e-01 3.46971184e-01 -7.78924286e-01 4.63678949e-02
7.96167552e-02 -2.35560060e-01 -6.34812713e-01 5.76794684e-01
7.37680852e-01 -7.50528216e-01 5.18874168e-01 -9.98414636e-01
4.72971618e-01 3.42898667e-01 -1.58329225e+00 1.30647182e+00
-4.83354241e-01 1.68559819e-01 -1.32676497e-01 -1.12899852e+00
5.04133999e-01 2.09236309e-01 4.13508117e-01 -1.98999569e-01
-6.40597939e-02 8.64398032e-02 -2.79887449e-02 -6.59069240e-01
4.54602689e-01 3.52210462e-01 3.68194692e-02 2.36187041e-01
3.17393869e-01 6.72111332e-01 5.30180097e-01 8.02057266e-01
1.41539299e+00 -2.15462267e-01 -1.35225961e-02 -5.96815720e-02
5.44868052e-01 -1.65445805e-01 6.83320165e-01 4.33821172e-01
3.99461806e-01 -2.02666476e-01 8.03559065e-01 -5.08538365e-01
-6.12038910e-01 -1.06440282e+00 -2.78706513e-02 1.11290026e+00
4.20695603e-01 -9.80855465e-01 -2.02136859e-01 -9.44264233e-01
4.92772520e-01 2.80264437e-01 -7.87941277e-01 -1.59025565e-01
-4.66072649e-01 -1.02783680e+00 5.70620179e-01 5.84957182e-01
3.03070396e-01 -4.82883453e-01 2.53241777e-01 1.53186575e-01
-3.28692436e-01 -1.48797226e+00 -3.57843935e-01 -8.47612042e-03
-6.70122147e-01 -1.22805548e+00 -2.44689047e-01 -8.78240705e-01
9.92145300e-01 6.93369880e-02 1.17935884e+00 1.90058038e-01
-2.05142975e-01 1.20327741e-01 -3.02009225e-01 1.21733479e-01
2.87246644e-01 2.30747506e-01 3.53026181e-01 -2.67640278e-02
3.94400299e-01 -7.75520980e-01 -6.75437272e-01 2.10950330e-01
-7.72122383e-01 -4.01382297e-02 6.84180200e-01 1.03906226e+00
5.04860997e-01 4.02087778e-01 3.56689662e-01 -1.37350023e+00
5.61364710e-01 -7.52579153e-01 -3.81393820e-01 5.56609571e-01
-9.91622269e-01 4.94699299e-01 6.65772021e-01 -3.82944912e-01
-7.49979615e-01 -4.69128132e-01 2.63070673e-01 -5.05303204e-01
4.76958394e-01 9.74187672e-01 -2.10024998e-01 -4.36936356e-02
4.15500283e-01 -4.75656055e-02 -5.31930327e-01 -4.94327664e-01
6.58387244e-01 4.15666461e-01 3.63760978e-01 -8.15979064e-01
1.25760901e+00 3.74898940e-01 8.55604485e-02 -3.87575597e-01
-8.35221469e-01 -5.86063981e-01 -2.99834013e-01 3.38692099e-01
6.41908705e-01 -1.13164306e+00 -1.11975121e+00 -1.64525002e-01
-1.04289699e+00 3.10426384e-01 2.59903166e-02 5.38027108e-01
2.71309316e-01 4.86797690e-01 -7.55423844e-01 -3.54118407e-01
-4.78074908e-01 -7.40894794e-01 8.00695300e-01 1.12018637e-01
1.00434735e-01 -1.03706348e+00 1.49685010e-01 5.00115097e-01
1.04131736e-01 1.09722503e-01 1.33586645e+00 -7.18497694e-01
-8.94481301e-01 -3.22217762e-01 -7.00773060e-01 -1.03917252e-02
1.09704725e-01 -3.99183512e-01 -6.36582673e-01 -1.69037476e-01
-8.18134964e-01 -1.37976900e-01 9.07418847e-01 -3.17276925e-01
1.04166460e+00 -7.46928394e-01 -7.85647988e-01 8.30476344e-01
1.42905056e+00 -5.50131023e-01 4.55069810e-01 1.24400727e-01
1.43741810e+00 4.25868243e-01 5.31582534e-01 4.22122627e-01
9.43412364e-01 6.00579143e-01 1.98402166e-01 1.66341588e-01
1.29582837e-01 -5.59573293e-01 1.15248309e-02 1.09532785e+00
-3.94871444e-01 -2.15941742e-01 -8.69372725e-01 7.26105571e-01
-1.91578448e+00 -8.87178063e-01 -2.46501759e-01 2.01810956e+00
9.68690634e-01 5.46276011e-02 -1.50408968e-01 5.11973761e-02
6.33457601e-01 1.89901665e-02 -4.16342854e-01 -5.71655296e-02
4.32389304e-02 1.38725996e-01 7.05986857e-01 4.48363304e-01
-8.68459344e-01 8.88332546e-01 4.60679293e+00 7.75908291e-01
-6.42542601e-01 1.91896990e-01 1.56997025e-01 4.77847010e-02
-6.85185909e-01 1.67249754e-01 -9.77965474e-01 4.24082816e-01
7.93862045e-01 -5.42075694e-01 4.73037779e-01 8.15003574e-01
-4.33332950e-01 2.49612153e-01 -1.13605988e+00 1.08768225e+00
-1.22246012e-01 -1.44384205e+00 1.51216954e-01 1.03508532e-01
6.18906736e-01 -3.66085675e-03 -2.96557695e-01 8.63208473e-01
4.55388516e-01 -8.90723646e-01 1.05039790e-01 3.29979658e-01
7.36901283e-01 -5.53747594e-01 8.55690241e-01 4.56290096e-02
-1.69297659e+00 -6.17829412e-02 -6.88323438e-01 2.52785772e-01
1.62989601e-01 8.78686249e-01 -7.91582942e-01 1.16233242e+00
9.32323635e-01 6.91836476e-01 -5.72542191e-01 6.70671523e-01
-2.20823944e-01 3.79587501e-01 -5.96139669e-01 2.93356776e-02
-1.63177088e-01 -3.37967068e-01 3.97634208e-01 1.10975230e+00
5.45571446e-02 2.60180116e-01 1.84062168e-01 6.90809608e-01
-5.77298582e-01 1.88894510e-01 -4.60538268e-01 -3.69046837e-01
7.89004385e-01 1.45398510e+00 -3.39563251e-01 -3.30214500e-01
-5.77368617e-01 1.05227029e+00 9.48175192e-01 5.13879776e-01
-1.04480398e+00 -6.53536916e-01 8.74945402e-01 2.26055505e-03
5.18304467e-01 -1.84126850e-02 5.99992238e-02 -1.52013350e+00
4.29677248e-01 -4.72500205e-01 7.63267279e-01 -2.54118621e-01
-1.37423730e+00 4.69224900e-01 -1.13920579e-02 -8.24234128e-01
1.36989713e-01 -6.34283185e-01 -2.00538799e-01 7.43686080e-01
-1.57541895e+00 -1.14883590e+00 -4.58043873e-01 6.18835568e-01
-2.73545444e-01 8.68498757e-02 9.28743541e-01 8.78111720e-01
-1.08044231e+00 1.03535593e+00 9.96023715e-02 4.72630531e-01
5.31149447e-01 -1.50595617e+00 3.16845477e-01 5.44176459e-01
2.48097271e-01 1.06724870e+00 5.07065237e-01 -7.31909335e-01
-1.88045323e+00 -1.23722005e+00 9.79207098e-01 -4.73754674e-01
9.63714600e-01 -3.87569755e-01 -9.22340930e-01 9.41669285e-01
-2.12938234e-01 4.71127123e-01 7.19124496e-01 9.37869370e-01
-8.51015389e-01 -3.10976118e-01 -9.56482172e-01 6.16248131e-01
1.44281220e+00 -5.80441475e-01 -5.32256007e-01 5.01533926e-01
9.87805843e-01 -3.90081644e-01 -1.62886262e+00 5.44488847e-01
5.07391930e-01 -4.69779640e-01 1.26418519e+00 -1.00496161e+00
2.58412212e-01 -4.44673777e-01 -1.43830031e-01 -1.08750415e+00
-5.33801138e-01 -3.95366341e-01 -1.02824271e+00 1.38507795e+00
7.42485702e-01 -8.22025299e-01 8.47079337e-01 7.34232247e-01
3.04452419e-01 -1.04188848e+00 -6.45866215e-01 -8.11116278e-01
-4.05586362e-01 -4.44131350e-04 6.66344404e-01 1.32193160e+00
2.03800470e-01 6.78241134e-01 -2.40103945e-01 7.82646298e-01
6.28617585e-01 3.88802260e-01 8.41664135e-01 -1.38339972e+00
-4.59043801e-01 2.26230174e-02 -8.10016692e-01 -1.01772714e+00
2.43938237e-01 -1.20521188e+00 -5.47035933e-01 -1.67756259e+00
4.10698682e-01 -8.64124000e-01 -4.61423427e-01 6.77752554e-01
-6.33617043e-01 -1.27964646e-01 -1.49924085e-01 6.09635375e-02
-6.57642424e-01 5.66907585e-01 8.77449691e-01 -3.52978587e-01
4.99650389e-02 -4.48436379e-01 -7.53764033e-01 4.36302423e-01
2.91742235e-01 -5.10168076e-01 -5.53517580e-01 -7.35170424e-01
8.74542892e-01 -8.47269073e-02 1.88942224e-01 -5.64267457e-01
4.63516653e-01 2.13543713e-01 1.79777980e-01 -4.30002660e-01
4.57012057e-01 -8.89439821e-01 3.66999924e-01 7.61666521e-02
-4.48402297e-03 3.56542803e-02 3.40821557e-02 1.00197053e+00
-2.19036549e-01 4.00614663e-05 6.14287741e-02 1.56687081e-01
-7.77488828e-01 7.42620289e-01 5.31017423e-01 2.53705442e-01
9.18973744e-01 -2.41155177e-03 -6.40038788e-01 8.02234709e-02
-8.01325262e-01 6.59459710e-01 3.89438495e-02 4.49105918e-01
5.22564590e-01 -1.54665303e+00 -5.51480532e-01 -1.16971120e-01
2.98783600e-01 3.06364298e-01 4.82947946e-01 1.14613783e+00
-3.27488065e-01 4.15828705e-01 2.95011401e-01 -2.42554963e-01
-1.17249203e+00 7.18772233e-01 3.42398770e-02 -6.98241293e-01
-6.16609991e-01 1.12200880e+00 1.39268935e-01 -6.16757095e-01
1.76292866e-01 -1.07998982e-01 -2.81819910e-01 2.76360452e-01
3.93797874e-01 5.31243861e-01 8.55035037e-02 -3.76667857e-01
-4.81161773e-01 4.21746492e-01 -5.19930363e-01 4.92224902e-01
1.49161243e+00 -1.12892702e-01 -5.02476394e-01 4.34342362e-02
1.25705457e+00 3.04145008e-01 -4.48087096e-01 -5.27435362e-01
1.15211673e-01 -5.35903752e-01 -3.91264148e-02 -3.96015644e-01
-1.16266644e+00 4.15237397e-01 1.62016198e-01 3.17125857e-01
7.69284189e-01 1.34041384e-01 8.93833518e-01 8.26159418e-01
4.99044657e-01 -8.35552990e-01 -3.82361174e-01 3.22426766e-01
5.63181758e-01 -1.15533519e+00 2.34903470e-01 -9.89564478e-01
-5.20566642e-01 7.69994020e-01 7.59276688e-01 -9.54556651e-03
8.73910785e-01 -1.01692873e-04 -6.01909280e-01 -6.04550838e-01
-8.56531978e-01 -2.18469694e-01 5.20912826e-01 3.51416618e-01
2.22109124e-01 2.70980924e-01 -2.53430098e-01 7.49711156e-01
-1.35652229e-01 -2.55292386e-01 1.08416490e-01 5.06360233e-01
-9.14933607e-02 -1.10316348e+00 1.26035642e-02 8.91271532e-01
-3.52641821e-01 -4.15349215e-01 -1.16821034e-02 6.93080068e-01
1.89604640e-01 6.40761137e-01 -1.94458067e-01 -6.12120509e-01
4.96712208e-01 3.69255766e-02 3.79062444e-01 -6.24062300e-01
-1.68506816e-01 -6.62551999e-01 4.73952979e-01 -6.56154275e-01
-1.79108977e-02 -2.42038563e-01 -1.29006338e+00 -7.43793309e-01
-7.18305409e-01 3.53857905e-01 2.00671166e-01 7.00460494e-01
7.27469981e-01 6.37981772e-01 4.65321273e-01 -8.87858495e-02
-4.84619558e-01 -9.11386132e-01 -7.04954684e-01 6.02037489e-01
-8.08764994e-02 -9.34323013e-01 -3.26174647e-01 -2.76631534e-01] | [8.733111381530762, 7.877845764160156] |
c4fff2cc-4a48-4a03-87f7-b2abf0e20da1 | scaling-open-vocabulary-object-detection | 2306.09683 | null | https://arxiv.org/abs/2306.09683v1 | https://arxiv.org/pdf/2306.09683v1.pdf | Scaling Open-Vocabulary Object Detection | Open-vocabulary object detection has benefited greatly from pretrained vision-language models, but is still limited by the amount of available detection training data. While detection training data can be expanded by using Web image-text pairs as weak supervision, this has not been done at scales comparable to image-level pretraining. Here, we scale up detection data with self-training, which uses an existing detector to generate pseudo-box annotations on image-text pairs. Major challenges in scaling self-training are the choice of label space, pseudo-annotation filtering, and training efficiency. We present the OWLv2 model and OWL-ST self-training recipe, which address these challenges. OWLv2 surpasses the performance of previous state-of-the-art open-vocabulary detectors already at comparable training scales (~10M examples). However, with OWL-ST, we can scale to over 1B examples, yielding further large improvement: With an L/14 architecture, OWL-ST improves AP on LVIS rare classes, for which the model has seen no human box annotations, from 31.2% to 44.6% (43% relative improvement). OWL-ST unlocks Web-scale training for open-world localization, similar to what has been seen for image classification and language modelling. | ['Neil Houlsby', 'Alexey Gritsenko', 'Matthias Minderer'] | 2023-06-16 | null | null | null | null | ['open-vocabulary-object-detection'] | ['computer-vision'] | [ 2.11337879e-01 3.32545996e-01 -2.46733993e-01 -3.12822104e-01
-1.14566386e+00 -7.06398666e-01 6.25735939e-01 1.67330474e-01
-9.47466195e-01 3.71611357e-01 -2.44829580e-02 -3.16931456e-01
6.81883991e-01 -3.69628906e-01 -9.83823955e-01 -1.56769276e-01
7.80341849e-02 6.67429328e-01 9.97231424e-01 -1.88503832e-01
-2.95488387e-01 6.25174716e-02 -1.69201469e+00 6.00338042e-01
2.23354176e-01 9.94629264e-01 2.68133104e-01 8.20441544e-01
-2.12686941e-01 6.01910710e-01 -4.18572217e-01 -2.52205074e-01
4.81252760e-01 1.18837126e-01 -6.27007842e-01 -5.87883778e-02
1.38753736e+00 -4.88554835e-01 -3.14154714e-01 9.84433651e-01
5.22513747e-01 -3.44264477e-01 5.41658401e-01 -1.13742995e+00
-7.43787825e-01 5.13570666e-01 -8.23901296e-01 3.93361628e-01
2.28023067e-01 3.54059458e-01 1.12681878e+00 -1.19660556e+00
8.96041632e-01 1.18074656e+00 1.04052162e+00 7.45037436e-01
-1.37496352e+00 -9.98170197e-01 1.51494835e-02 -3.07213087e-02
-1.67059803e+00 -4.43197161e-01 -1.42913133e-01 -6.14515007e-01
1.43832850e+00 -7.49004260e-02 5.44771194e-01 1.19498324e+00
-2.13298336e-01 1.02323163e+00 1.02149642e+00 -6.46828949e-01
-8.71464312e-02 4.77669775e-01 1.30081072e-01 9.31791186e-01
3.83776993e-01 1.82028264e-01 -4.72545266e-01 1.05248153e-01
6.07500196e-01 -2.94239670e-01 1.13994747e-01 -6.08318746e-01
-1.24889612e+00 8.57785881e-01 6.81311607e-01 1.12567335e-01
3.34879011e-01 3.13453943e-01 5.94512761e-01 2.78480560e-01
5.38975537e-01 3.46942425e-01 -6.58236027e-01 2.82458931e-01
-1.10739553e+00 3.68465222e-02 6.87127531e-01 1.34207714e+00
7.95257211e-01 -9.39663276e-02 -1.11574873e-01 7.57352173e-01
3.93147945e-01 8.79738986e-01 3.30697030e-01 -6.92187309e-01
4.62984413e-01 4.94666040e-01 4.77015525e-02 -2.15027839e-01
-5.58644712e-01 -3.65279198e-01 -2.24713922e-01 5.36259592e-01
6.72995746e-01 -1.97946772e-01 -1.26778972e+00 1.54202151e+00
1.04748815e-01 1.00977212e-01 1.10109160e-02 7.07009912e-01
9.34081376e-01 3.44570845e-01 3.22545111e-01 1.86160296e-01
1.68652904e+00 -1.21606505e+00 -2.40787670e-01 -8.08750868e-01
9.01357830e-01 -7.52185106e-01 1.20294046e+00 1.60025001e-01
-6.04390085e-01 -6.05032146e-01 -1.40351045e+00 -2.65239716e-01
-6.29792690e-01 2.36876696e-01 5.15422285e-01 6.94073796e-01
-1.17353892e+00 -2.54823454e-02 -6.15339279e-01 -1.08062744e+00
7.06097841e-01 2.69296914e-01 -6.48419619e-01 -4.40116003e-02
-1.00902271e+00 1.16022027e+00 6.19688272e-01 -5.53737342e-01
-1.05630088e+00 -7.08296657e-01 -8.70778441e-01 -2.32029751e-01
6.37603819e-01 -4.61729616e-01 1.45865643e+00 -8.30642581e-01
-7.70313442e-01 1.36501932e+00 1.95226043e-01 -9.08109546e-01
6.09406650e-01 -2.03074977e-01 -4.87816989e-01 3.40942629e-02
3.91972691e-01 1.56768298e+00 1.00556040e+00 -1.03426802e+00
-1.15032148e+00 -2.01652363e-01 6.92331716e-02 1.67807966e-01
-3.85219902e-01 2.90929466e-01 -7.67577708e-01 -3.56030315e-01
-2.58828789e-01 -9.03872967e-01 -1.72252461e-01 5.75500071e-01
1.57012895e-01 -3.50563049e-01 1.01646900e+00 -4.18791085e-01
6.79897964e-01 -2.20341396e+00 -6.37663305e-01 -2.62186348e-01
2.71027356e-01 5.14266312e-01 -4.13856745e-01 1.76974535e-01
1.61983922e-01 -7.87530392e-02 4.53229658e-02 -4.64000344e-01
-2.09130645e-02 3.38855952e-01 -3.26484919e-01 3.78426105e-01
3.07137102e-01 1.00744176e+00 -7.97489345e-01 -7.78839469e-01
2.33005896e-01 2.12117583e-01 -6.66372061e-01 -1.75665796e-01
-6.04286790e-01 -1.21853597e-01 -1.40610728e-02 6.35689676e-01
4.86809701e-01 -4.28868592e-01 -6.56641275e-02 -2.17812642e-01
-2.43359402e-01 -5.36620878e-02 -1.18279159e+00 1.82689345e+00
-3.93210739e-01 1.11422861e+00 1.30294457e-01 -6.75422311e-01
6.70644939e-01 5.44668324e-02 1.39057145e-01 -6.27741992e-01
7.09091946e-02 2.04269886e-01 -5.52957766e-02 -3.60884249e-01
3.62981647e-01 1.35026842e-01 -2.95889750e-02 1.88827544e-01
6.40565813e-01 -1.81837574e-01 3.92010689e-01 4.33517396e-01
1.23579228e+00 1.96171939e-01 3.39098901e-01 -1.79831415e-01
2.51986325e-01 3.86882782e-01 1.21940956e-01 1.17177999e+00
-4.48759794e-01 5.41012526e-01 7.51065239e-02 -4.48961377e-01
-1.40994251e+00 -9.90923524e-01 -4.81152207e-01 1.51090884e+00
9.77280140e-02 -7.01664925e-01 -7.57944703e-01 -9.87473190e-01
3.01429629e-01 4.57022309e-01 -4.74966735e-01 1.35478720e-01
-2.73364693e-01 -5.89320064e-01 1.20394874e+00 7.79952586e-01
5.36375463e-01 -9.10489082e-01 -6.49424434e-01 6.87951921e-03
5.75991385e-02 -1.66322970e+00 -4.27740455e-01 5.76560080e-01
-4.32607621e-01 -1.05581415e+00 -4.69386727e-01 -1.02037275e+00
3.85243088e-01 3.84086132e-01 1.07732725e+00 -1.67162701e-01
-9.34467018e-01 4.19012189e-01 -1.46091059e-01 -7.66350806e-01
-4.57716346e-01 -2.87947375e-02 3.60768557e-01 -4.52168465e-01
7.41286695e-01 1.15911260e-01 -4.88482118e-01 3.04154932e-01
-5.46750963e-01 -3.43471766e-02 8.22620571e-01 9.00240183e-01
5.45060754e-01 -3.74711424e-01 1.99784383e-01 -6.52724385e-01
4.44400348e-02 -8.98241624e-02 -8.15690458e-01 2.10438088e-01
-7.12525666e-01 -2.12734751e-03 1.42321140e-01 -6.08931899e-01
-8.48359466e-01 4.98206466e-01 -2.38784194e-01 -5.35756767e-01
-3.53556067e-01 -1.40323520e-01 2.04018310e-01 -4.78508383e-01
1.29360235e+00 -1.19385934e-02 8.54129996e-03 -4.02266264e-01
8.48247707e-01 8.91275287e-01 6.71171844e-01 -1.13920212e-01
8.88206363e-01 6.65852129e-01 -4.78676647e-01 -8.47296834e-01
-1.26496589e+00 -1.02453136e+00 -7.53224254e-01 2.12212890e-01
1.10097241e+00 -1.51029491e+00 -4.61681098e-01 9.94167700e-02
-9.31726038e-01 -4.34097618e-01 -4.92005080e-01 1.89681709e-01
-4.05843854e-01 3.06825697e-01 -5.89991689e-01 -6.74354255e-01
-2.58393258e-01 -9.14985180e-01 1.47752297e+00 -1.97824001e-01
-1.52598796e-02 -5.46835124e-01 -1.77718803e-01 5.11824906e-01
2.26581365e-01 -3.25510353e-01 2.10419551e-01 -8.40739250e-01
-6.39367282e-01 -5.05820215e-01 -7.58295417e-01 4.86241400e-01
-4.50427055e-01 -4.19133931e-01 -1.21230733e+00 -4.55635458e-01
-4.45219606e-01 -9.14326906e-01 1.13087630e+00 3.89279574e-02
5.88837922e-01 1.40692905e-01 -7.81661153e-01 5.41329026e-01
1.30374420e+00 -3.72630060e-01 1.85099900e-01 5.14412403e-01
6.26864314e-01 1.15042761e-01 6.99705422e-01 1.14171669e-01
2.14240670e-01 8.41767609e-01 3.18320692e-01 -4.38935399e-01
-6.44912660e-01 -5.06834567e-01 4.10477519e-01 -6.98632281e-03
4.32885528e-01 -4.78735380e-02 -1.17704415e+00 4.28490847e-01
-1.88052666e+00 -9.66534615e-01 -2.27021262e-01 1.86967945e+00
6.89144790e-01 6.34316266e-01 1.33376077e-01 -2.63311446e-01
4.72463280e-01 5.99887483e-02 -6.57472014e-01 9.47839245e-02
-2.60768920e-01 2.12643705e-02 1.20750880e+00 5.01255870e-01
-1.62350965e+00 1.46628118e+00 7.34504175e+00 1.09513819e+00
-9.04666603e-01 6.83976591e-01 1.12254597e-01 -2.25308225e-01
3.99580926e-01 4.11047377e-02 -1.62800503e+00 5.99299334e-02
8.55223954e-01 3.04750919e-01 1.23760365e-01 1.28653371e+00
-3.07061136e-01 -1.68011218e-01 -1.23654711e+00 1.24656081e+00
4.70983177e-01 -1.41442549e+00 1.26245935e-02 -5.17603680e-02
6.17128491e-01 1.05239844e+00 -2.79017657e-01 7.70071805e-01
5.51653266e-01 -7.65892446e-01 8.41449976e-01 -1.65005207e-01
1.31122863e+00 -1.34508952e-01 6.64035082e-01 6.31639421e-01
-1.29311132e+00 -2.74473041e-01 -6.33752823e-01 -1.76385436e-02
8.13974161e-03 1.00356467e-01 -1.22248840e+00 -1.23340860e-01
1.07754838e+00 4.45354670e-01 -1.03316522e+00 8.75170708e-01
-3.73329550e-01 6.10802948e-01 -8.16443801e-01 -4.09700535e-02
3.31079781e-01 4.29864973e-01 4.17167664e-01 1.64090121e+00
-1.98424518e-01 -2.62882739e-01 6.22595131e-01 5.91406882e-01
-1.92409873e-01 1.37415513e-01 -7.08205044e-01 2.48640314e-01
3.75241905e-01 1.37212396e+00 -6.32845044e-01 -5.92749417e-01
-9.83394802e-01 1.05183625e+00 6.57730579e-01 4.83495742e-02
-1.00122988e+00 -3.93560082e-02 5.54202497e-01 3.27628285e-01
5.17834723e-01 -2.02118516e-01 3.44543643e-02 -1.20272124e+00
-1.54617444e-01 -7.00622201e-01 5.61560333e-01 -7.83230364e-01
-1.21535349e+00 3.52909595e-01 -1.11752205e-01 -1.01897812e+00
1.03489704e-01 -1.03943682e+00 -7.85444155e-02 5.02063453e-01
-1.35888147e+00 -1.73274577e+00 -2.74348080e-01 4.68191415e-01
8.25057328e-01 -2.97063351e-01 8.51538897e-01 4.09264177e-01
-1.43945038e-01 8.37618649e-01 1.06935218e-01 3.31932515e-01
1.13536692e+00 -1.34241498e+00 9.13068771e-01 9.17630196e-01
8.71048987e-01 3.02738577e-01 5.45320570e-01 -6.14781439e-01
-1.12389958e+00 -1.22165942e+00 6.21114612e-01 -1.13872755e+00
1.11743104e+00 -1.09172654e+00 -5.38409531e-01 1.02242661e+00
9.25288722e-02 7.03758121e-01 3.37951034e-01 2.90999979e-01
-9.21238899e-01 -1.08207136e-01 -8.66058230e-01 5.47338486e-01
1.45501924e+00 -7.00207591e-01 -7.09640980e-01 7.36644804e-01
9.47693944e-01 -4.76706117e-01 -4.66678023e-01 4.90238577e-01
6.15288019e-01 -6.44943237e-01 1.16907370e+00 -6.13129675e-01
-1.68284774e-01 -4.81440067e-01 -2.23199829e-01 -7.51446247e-01
-4.14645106e-01 -1.44547299e-01 -1.43591268e-02 9.74027693e-01
5.68153858e-01 -4.93086755e-01 9.56403792e-01 1.92527518e-01
-9.54729412e-03 -1.64510608e-01 -8.16519320e-01 -1.11754990e+00
-2.88155209e-02 -6.79866195e-01 -1.17976755e-01 6.64687216e-01
3.71720158e-02 7.82893836e-01 -2.46717960e-01 2.63277382e-01
7.14146018e-01 -2.73786545e-01 1.08354843e+00 -1.11751342e+00
-4.37838584e-01 -3.35113674e-01 -6.87377453e-01 -1.39344609e+00
3.78458574e-02 -1.08437061e+00 1.98117226e-01 -1.39882147e+00
4.49840158e-01 -3.97907138e-01 -1.28306329e-01 8.53932619e-01
-3.21665220e-02 1.05185223e+00 3.88757229e-01 2.76706040e-01
-1.08306766e+00 -1.11172693e-02 7.03687251e-01 -4.09704983e-01
2.35968128e-01 -5.05681574e-01 -5.02776682e-01 9.72327530e-01
3.86824697e-01 -3.76528054e-01 -2.56070584e-01 -4.49766934e-01
2.46503428e-02 -4.82160777e-01 5.12314737e-01 -1.39611888e+00
3.78236800e-01 4.00099754e-01 6.22541547e-01 -7.07244575e-01
2.89133191e-01 -7.96290457e-01 -3.60417783e-01 4.41014349e-01
-2.92609364e-01 -4.32493210e-01 5.47212660e-01 7.85122752e-01
5.13990857e-02 -1.70838937e-01 8.90852988e-01 -1.63024709e-01
-1.35018146e+00 2.12189153e-01 -3.92724425e-01 2.54165083e-01
1.12210214e+00 -1.31171703e-01 -6.12084270e-01 -2.95981895e-02
-8.62342477e-01 4.53845114e-01 4.91641819e-01 7.83633113e-01
2.94798315e-01 -9.29136753e-01 -5.80548882e-01 2.54396349e-01
8.04380536e-01 -4.17581536e-02 -2.44511336e-01 5.94515681e-01
-4.60113406e-01 6.19685292e-01 6.12524860e-02 -1.10385835e+00
-1.46392679e+00 8.68476272e-01 3.22390318e-01 -1.87540993e-01
-8.68453920e-01 1.13738739e+00 5.31189620e-01 -4.47625101e-01
4.94465262e-01 -3.23716551e-01 5.77042699e-02 3.19592692e-02
6.71493590e-01 -1.72186583e-01 1.80085972e-01 -4.55037028e-01
-4.91430491e-01 5.69130421e-01 -3.34319472e-01 -1.14707910e-01
8.19964111e-01 -2.00199615e-02 4.85985041e-01 2.34116629e-01
9.21001434e-01 -5.78788854e-02 -1.42809272e+00 -3.67124379e-01
1.65618166e-01 -6.17115311e-02 8.51940587e-02 -9.27463949e-01
-3.88523549e-01 8.65299523e-01 1.04766798e+00 -1.87752396e-01
4.50725615e-01 5.42187035e-01 6.33307636e-01 7.56764293e-01
5.48718870e-01 -1.18438649e+00 2.63187021e-01 7.19400406e-01
6.70537472e-01 -1.81471801e+00 -4.50906670e-03 -3.76732796e-01
-4.09782231e-01 7.16008425e-01 1.04193926e+00 -1.17906801e-01
5.33702612e-01 7.68426538e-01 3.49700540e-01 -2.32566297e-01
-7.32692063e-01 -7.74100125e-01 1.96447313e-01 7.73726821e-01
3.07934612e-01 -4.52596042e-03 1.19971149e-01 1.13241270e-01
3.72405797e-02 -8.38845968e-02 6.82070181e-02 7.68310666e-01
-8.43301594e-01 -8.37524891e-01 -4.76602226e-01 5.09473026e-01
-1.69673264e-01 -4.67062354e-01 -3.12964201e-01 1.00378096e+00
5.29475212e-01 7.44878232e-01 1.99371800e-01 -9.97579917e-02
4.32681113e-01 2.90893793e-01 5.39177299e-01 -1.03211319e+00
-2.38662481e-01 7.99822882e-02 3.07747334e-01 -6.17720664e-01
-2.34650627e-01 -6.62492990e-01 -1.30320859e+00 2.40809888e-01
-7.74535894e-01 -3.09742600e-01 6.14090204e-01 7.99877644e-01
2.43056118e-01 2.31482670e-01 -2.41634712e-01 -8.63849580e-01
-7.26903319e-01 -1.05897546e+00 -2.80823976e-01 3.55785698e-01
3.48489255e-01 -7.40120590e-01 -1.56763047e-01 1.39690697e-01] | [9.515212059020996, 1.4129977226257324] |
878da011-d608-48c5-a0b9-323cb323b482 | bitmix-data-augmentation-for-image | 2006.16625 | null | https://arxiv.org/abs/2006.16625v1 | https://arxiv.org/pdf/2006.16625v1.pdf | BitMix: Data Augmentation for Image Steganalysis | Convolutional neural networks (CNN) for image steganalysis demonstrate better performances with employing concepts from high-level vision tasks. The major employed concept is to use data augmentation to avoid overfitting due to limited data. To augment data without damaging the message embedding, only rotating multiples of 90 degrees or horizontally flipping are used in steganalysis, which generates eight fixed results from one sample. To overcome this limitation, we propose BitMix, a data augmentation method for spatial image steganalysis. BitMix mixes a cover and stego image pair by swapping the random patch and generates an embedding adaptive label with the ratio of the number of pixels modified in the swapped patch to those in the cover-stego pair. We explore optimal hyperparameters, the ratio of applying BitMix in the mini-batch, and the size of the bounding box for swapping patch. The results reveal that using BitMix improves the performance of spatial image steganalysis and better than other data augmentation methods. | ['Heung-Kyu Lee', 'Seung-Hun Nam', 'In-Jae Yu', 'Wonhyuk Ahn'] | 2020-06-30 | null | null | null | null | ['steganalysis'] | ['computer-vision'] | [ 9.29417133e-01 2.47348368e-01 -1.29384562e-01 1.77969709e-01
-2.73902595e-01 -1.47585437e-01 4.11521912e-01 -3.74185473e-01
-4.79492843e-01 3.13407153e-01 7.79945701e-02 -5.03997743e-01
5.15853941e-01 -7.49598145e-01 -8.34728897e-01 -1.17717528e+00
3.24435905e-02 -2.94101954e-01 2.37505451e-01 -1.74369290e-01
3.65317315e-01 2.06680745e-01 -1.34315562e+00 3.31200957e-01
5.80744684e-01 8.20520520e-01 1.68421030e-01 5.89798629e-01
1.96116835e-01 5.87583423e-01 -6.21865213e-01 -7.60041699e-02
6.58014596e-01 -9.13954854e-01 -4.23625708e-01 1.48670480e-01
2.25596890e-01 -4.30963755e-01 -5.35948396e-01 1.06653380e+00
3.58255833e-01 -2.19263166e-01 4.24797595e-01 -1.15084219e+00
-6.36490703e-01 6.53728962e-01 -8.30545425e-01 1.67552605e-01
-3.07683080e-01 3.68628651e-01 3.52833271e-01 -3.67331713e-01
4.83940363e-01 8.58773708e-01 5.06657302e-01 5.30143976e-01
-1.11025846e+00 -8.61782253e-01 -5.11944175e-01 1.01809673e-01
-1.05215549e+00 -4.54896003e-01 9.27429020e-01 -1.06626786e-01
6.13087416e-01 3.20538521e-01 7.01535761e-01 9.56819952e-01
2.47790933e-01 4.32041198e-01 1.02283406e+00 -7.03401744e-01
2.57465653e-02 7.41480514e-02 -4.84606981e-01 7.45587349e-01
3.68258119e-01 3.12356055e-01 -3.37776721e-01 6.18700646e-02
1.04789948e+00 -3.41981719e-03 -3.29802245e-01 -3.83615524e-01
-1.50201428e+00 9.50330913e-01 7.52562642e-01 3.79608601e-01
-2.26914316e-01 6.50722027e-01 2.17723921e-01 4.58846599e-01
1.88832760e-01 6.64461792e-01 -3.47289667e-02 3.09239119e-01
-1.03766227e+00 -1.70582965e-01 3.44202489e-01 4.31599230e-01
7.63587296e-01 3.73492241e-01 -6.66849464e-02 4.08753276e-01
8.55360031e-02 5.33984065e-01 7.90820122e-01 -7.25237250e-01
5.84833562e-01 5.52066863e-01 -3.11550856e-01 -1.25536644e+00
-5.00916876e-02 -3.58745843e-01 -1.19122386e+00 3.18548143e-01
2.20310450e-01 -1.29477441e-01 -1.23542333e+00 1.46143377e+00
6.77112192e-02 4.47681159e-01 4.66183499e-02 6.81860089e-01
5.12545228e-01 6.69184387e-01 -6.84097558e-02 4.85741580e-03
1.20940137e+00 -9.19830680e-01 -5.41812718e-01 -4.74939793e-01
9.29283679e-01 -6.66407287e-01 7.30873644e-01 1.12568177e-01
-6.73601508e-01 -5.97363472e-01 -1.37699509e+00 1.32993892e-01
-4.23375040e-01 6.59990609e-02 2.46366665e-01 8.86842489e-01
-6.59576416e-01 5.04301071e-01 -5.80751836e-01 2.44970679e-01
3.45453084e-01 4.19463634e-01 -5.82737148e-01 -1.20889097e-01
-1.25133991e+00 6.38573468e-01 8.09742928e-01 -4.15223278e-02
-8.48082364e-01 -3.19599122e-01 -1.06281936e+00 1.29005283e-01
6.71099648e-02 -2.91146755e-01 3.76201749e-01 -1.20357478e+00
-1.17523956e+00 6.93519235e-01 3.42965841e-01 -7.67986059e-01
3.07630539e-01 3.28903347e-01 -3.22462320e-01 8.16047788e-02
-1.56938881e-01 1.13353968e+00 1.60928726e+00 -1.05922008e+00
-2.09212318e-01 -8.67938250e-02 -6.00899518e-01 -6.25581443e-02
-5.13185740e-01 -3.44353586e-01 -4.91922200e-01 -8.82123768e-01
3.94070089e-01 -9.98134196e-01 -2.99808830e-01 -2.45371833e-01
-4.40184265e-01 6.26777112e-01 1.45776653e+00 -8.59014809e-01
1.06240380e+00 -2.35420036e+00 -6.81295022e-02 3.68977249e-01
2.21914217e-01 8.02463710e-01 -5.48085868e-01 1.56686127e-01
-4.99491900e-01 2.05089718e-01 -4.79894996e-01 -2.36700669e-01
-4.27519500e-01 -3.76349427e-02 -2.78873146e-01 5.75338304e-01
1.66459635e-01 1.08609021e+00 -6.00689054e-01 -2.91627795e-01
4.73558187e-01 6.85221374e-01 -4.67260569e-01 6.87573431e-03
-1.30463764e-01 5.08953035e-01 2.17750645e-03 3.14545780e-01
8.64465833e-01 -4.56213325e-01 3.01581949e-01 -2.91638583e-01
1.56090513e-01 -3.81939188e-02 -9.82031286e-01 1.14855635e+00
-1.87164932e-01 9.91648316e-01 -3.96944404e-01 -8.37865710e-01
1.01840568e+00 2.78879195e-01 3.94460499e-01 -7.00975001e-01
4.35679555e-01 1.21731617e-01 9.23817158e-02 -6.74026251e-01
5.08467972e-01 7.17858002e-02 1.09744139e-01 5.73303401e-01
-2.71674514e-01 -8.13549832e-02 -2.07311407e-01 -1.48939490e-01
1.04973423e+00 -2.01726139e-01 1.71658218e-01 3.22396368e-01
3.11514884e-01 -2.18635127e-01 2.13995457e-01 6.31174207e-01
-6.87203556e-02 7.47219563e-01 7.00895250e-01 -4.35952276e-01
-1.59672427e+00 -1.35676965e-01 3.10225189e-01 8.61621737e-01
1.81492910e-01 -2.39712503e-02 -9.46894288e-01 -7.21086860e-01
-1.89469039e-01 2.57815987e-01 -8.15568388e-01 -4.36224788e-01
-8.36641669e-01 -8.17729294e-01 7.81776667e-01 1.42056614e-01
1.31323373e+00 -1.28885341e+00 -7.53570914e-01 -1.14274947e-02
-3.79938483e-01 -1.04615474e+00 -5.08707643e-01 1.10842526e-01
-7.92781591e-01 -1.00360608e+00 -8.75848174e-01 -9.18702245e-01
8.72575164e-01 4.63514715e-01 3.47420305e-01 3.12934279e-01
-2.33726427e-01 -5.06433189e-01 -4.68994439e-01 6.23135269e-03
-6.60854518e-01 1.91000253e-01 -4.61156517e-01 8.22606310e-02
2.85878256e-02 -2.61959225e-01 -7.46614337e-01 3.48466218e-01
-1.41480434e+00 4.17876214e-01 9.00223970e-01 7.58038342e-01
1.36381760e-01 2.41924644e-01 -2.35218909e-02 -7.92686164e-01
2.88046420e-01 -4.60363060e-01 -3.77084047e-01 2.28899568e-02
-5.47964573e-01 3.35521996e-01 2.13353217e-01 -7.30717659e-01
-2.02033058e-01 -1.07603602e-01 9.45169553e-02 -5.39145350e-01
1.18647367e-01 3.03675979e-01 -1.53638609e-02 -4.51096714e-01
5.80969453e-01 5.38515627e-01 5.44558227e-01 -1.14904054e-01
2.42363840e-01 5.86451292e-01 3.88298124e-01 4.53508437e-01
8.68459821e-01 5.81253827e-01 1.15127094e-01 -8.13536704e-01
8.50764569e-03 -1.18761085e-01 -3.26829553e-01 9.04475078e-02
8.18298519e-01 -5.27899921e-01 -4.85278428e-01 8.27738702e-01
-9.42123652e-01 -4.85837132e-01 -1.48266315e-01 2.53220499e-01
-1.96265236e-01 4.95957732e-01 -1.97202548e-01 -4.68531132e-01
-3.49806458e-01 -1.15082669e+00 7.40728080e-01 -7.59354606e-02
1.13070957e-01 -6.26321197e-01 -1.09822981e-01 2.60012269e-01
5.93319833e-01 6.25525355e-01 7.70219088e-01 -5.42518735e-01
-6.90410852e-01 -4.52756643e-01 -4.35828209e-01 6.46820724e-01
1.57947376e-01 -3.94223809e-01 -6.48154378e-01 -5.59403360e-01
3.26199755e-02 -8.52478817e-02 1.25107360e+00 3.11924487e-01
1.10419810e+00 -6.33536339e-01 -4.26810175e-01 8.94472778e-01
1.35346150e+00 1.84937924e-01 1.30927372e+00 5.79770088e-01
7.73542821e-01 4.45018500e-01 1.35976836e-01 1.31424949e-01
-1.33288771e-01 4.61352825e-01 8.43882620e-01 -3.57866585e-01
-3.11280876e-01 -1.63907275e-01 3.10247570e-01 1.72274500e-01
2.51357585e-01 -6.67458117e-01 -7.08334267e-01 5.09594917e-01
-1.25180757e+00 -9.19499755e-01 1.17606565e-01 2.07318783e+00
5.90186834e-01 8.76013339e-02 -9.35861692e-02 5.92975140e-01
9.01317239e-01 6.46473646e-01 -4.05224562e-01 6.84706075e-03
-2.22123310e-01 1.29156947e-01 1.14710820e+00 4.94824588e-01
-1.08701169e+00 9.26221013e-01 6.05446863e+00 8.33961725e-01
-1.57223868e+00 8.16685259e-02 7.32374072e-01 1.02403715e-01
-2.59541988e-01 -2.22420357e-02 -4.05906498e-01 8.28473091e-01
6.71919584e-01 8.85607421e-01 5.05422175e-01 3.84977639e-01
-5.33413291e-02 -7.41863102e-02 -3.97362649e-01 8.93458307e-01
3.74885738e-01 -1.46277821e+00 6.18958250e-02 4.67079163e-01
9.61736023e-01 -7.20460936e-02 6.96024358e-01 -2.44905874e-01
-5.10957539e-02 -1.04131854e+00 3.22349548e-01 1.39502689e-01
1.12012160e+00 -4.20343161e-01 9.18583572e-01 7.21829161e-02
-7.43214965e-01 -2.29561299e-01 4.84769838e-03 3.27620178e-01
-6.58256859e-02 2.94296652e-01 -1.18046296e+00 -5.45373373e-02
3.48924696e-01 3.64473462e-01 -5.60645223e-01 6.21267915e-01
-2.60427296e-01 6.43071532e-01 -3.26902777e-01 2.50830024e-01
3.53507847e-01 5.35281710e-02 4.98400688e-01 9.11185741e-01
5.09513795e-01 -1.18311718e-01 -4.87690449e-01 6.12868428e-01
-1.91967577e-01 -2.39161447e-01 -7.88684249e-01 -1.82193115e-01
5.91571748e-01 8.32637966e-01 -6.28135860e-01 -2.68946528e-01
7.31684640e-02 1.13736916e+00 -2.94400126e-01 4.26021457e-01
-6.69586778e-01 -5.61905682e-01 3.03773701e-01 4.24746811e-01
8.23773026e-01 -2.81103224e-01 -4.04620081e-01 -6.43351912e-01
-1.54988348e-01 -1.05790246e+00 7.43681937e-02 -6.43236220e-01
-2.32094720e-01 4.12178427e-01 -1.03134885e-01 -1.23967922e+00
-1.21623121e-01 -4.19809461e-01 -5.16892731e-01 5.97841263e-01
-1.54023576e+00 -1.30024445e+00 -5.88546813e-01 3.00345153e-01
1.45353153e-01 -4.10741746e-01 4.32962209e-01 2.34252848e-02
-4.91320610e-01 6.75861001e-01 1.56899989e-01 4.65347290e-01
4.30967510e-01 -5.48340380e-01 7.71812618e-01 9.68699813e-01
-1.34348616e-01 4.01865810e-01 7.40133882e-01 -7.47587621e-01
-8.81108046e-01 -9.72094417e-01 6.49372816e-01 6.98373616e-02
1.39341623e-01 -1.79933980e-01 -8.83465052e-01 6.15132272e-01
3.01337332e-01 9.43832397e-02 3.32484931e-01 -9.38128173e-01
-5.74328601e-01 6.06585220e-02 -1.42062724e+00 5.74501395e-01
4.56465274e-01 -2.41445899e-01 5.86750098e-02 -1.16645075e-01
6.67444110e-01 -3.68570834e-01 -2.54564553e-01 4.51089084e-01
4.93635863e-01 -9.43669856e-01 9.29568172e-01 -1.24560557e-01
5.05347252e-01 -3.86029750e-01 -1.94689371e-02 -1.02075946e+00
-1.33783609e-01 -7.32593119e-01 -1.77511692e-01 4.62751776e-01
2.27987051e-01 -6.90978229e-01 1.11040342e+00 -2.36064065e-02
2.33048469e-01 -4.57742512e-01 -8.33618760e-01 -3.32685262e-01
-2.62130857e-01 2.23276913e-02 7.28162587e-01 1.05418515e+00
-4.06478137e-01 -2.47156695e-01 -8.93428981e-01 2.22279772e-01
6.48309112e-01 -4.79494333e-01 7.79125690e-01 -6.01185083e-01
-1.47360176e-01 -1.72686204e-01 -5.29844582e-01 -8.48638833e-01
-3.11442882e-01 -5.23445427e-01 -4.97964285e-02 -9.30625319e-01
-1.21802300e-01 -5.54258525e-01 -4.42482121e-02 7.95582771e-01
-6.56945631e-02 1.02945280e+00 1.62014276e-01 3.50357443e-01
-2.16125976e-02 2.55832791e-01 1.36874354e+00 -1.80994883e-01
-2.89085090e-01 -3.16785485e-01 -4.59121972e-01 4.14755732e-01
1.00702894e+00 -6.71198308e-01 -1.61742836e-01 -4.47867751e-01
3.35722268e-01 -3.27998027e-02 6.29636765e-01 -1.18725693e+00
3.50778699e-02 -1.06869088e-02 5.26483893e-01 -3.40402246e-01
3.34426194e-01 -8.80415916e-01 3.85534108e-01 9.32668686e-01
-3.43783379e-01 -1.52120650e-01 7.48609677e-02 4.82659340e-01
7.23335147e-02 -3.61895472e-01 1.03674757e+00 -1.70970201e-01
-5.46902716e-01 7.81512484e-02 -3.19354296e-01 -5.93145072e-01
1.12648237e+00 -6.73103273e-01 -3.27347487e-01 -3.51454973e-01
-2.98169583e-01 -4.21376675e-01 5.03073573e-01 3.84298563e-01
8.59586358e-01 -1.22206461e+00 -5.28602004e-01 9.35826421e-01
-1.50393859e-01 -2.56212622e-01 1.74077749e-01 7.32068181e-01
-9.21941340e-01 3.11778992e-01 -4.24310654e-01 -3.13407481e-01
-1.43428779e+00 5.43489456e-01 4.25604731e-01 -1.90381452e-01
-5.33501625e-01 7.60859489e-01 -1.45081535e-01 3.57404836e-02
1.83521137e-01 -1.60516590e-01 -1.86577037e-01 -1.94693089e-01
6.04840577e-01 3.94550860e-01 -6.80735335e-02 -5.18410563e-01
1.59062281e-01 4.57748502e-01 -2.66412377e-01 -1.89505398e-01
1.13162732e+00 -1.43556803e-01 -1.94745645e-01 -2.77516425e-01
1.46157026e+00 -2.71186024e-01 -1.35997963e+00 -2.37120539e-01
-5.19143760e-01 -5.90303838e-01 2.37071648e-01 -3.50121975e-01
-1.30960405e+00 6.23116851e-01 9.83205497e-01 3.62994194e-01
1.10631633e+00 -2.43099540e-01 1.01641047e+00 1.68999642e-01
-1.65057436e-01 -7.69572258e-01 4.92973655e-01 2.48043478e-01
5.76752067e-01 -1.28439593e+00 -1.26622647e-01 -9.95517969e-02
-6.87055767e-01 7.99509048e-01 4.78468955e-01 -1.86692789e-01
4.64239746e-01 5.78879192e-02 2.82027181e-02 -2.68997222e-01
-7.51730725e-02 6.93202913e-02 1.39997348e-01 6.82027519e-01
-1.09924883e-01 -2.28290781e-01 -7.19180852e-02 -5.30887425e-01
-1.15947790e-01 -2.52718210e-01 3.42388600e-01 9.68431056e-01
-5.45618713e-01 -9.92767692e-01 -6.81561828e-01 1.81343913e-01
-3.91356826e-01 -2.68104702e-01 -1.57296643e-01 6.46534026e-01
5.14290750e-01 6.08023524e-01 3.44815582e-01 -5.88427246e-01
-3.59167278e-01 -1.73584029e-01 2.67998397e-01 -2.38723114e-01
-3.37555230e-01 -5.66004589e-02 -4.93707985e-01 -1.01096801e-01
-3.18999410e-01 -1.48118004e-01 -8.83905947e-01 -5.15828431e-01
-5.41867077e-01 -5.19343726e-02 1.04273260e+00 8.79664421e-01
3.28919113e-01 4.82129306e-01 7.82699585e-01 -7.67576218e-01
-2.55951583e-01 -1.27501237e+00 -2.73666680e-01 2.86540121e-01
9.15987551e-01 -7.66257569e-02 -5.64995766e-01 2.52303958e-01] | [4.2979278564453125, 8.05726432800293] |
b10f7d8c-204d-4a84-b101-51401f6cbaf5 | siamese-nestedunet-networks-for-change | null | null | https://dl.acm.org/doi/abs/10.1145/3437802.3437810 | https://dl.acm.org/doi/abs/10.1145/3437802.3437810 | Siamese NestedUNet Networks for Change Detection of High Resolution Satellite Image | Change detection is an important task in remote sensing (RS) image analysis. With the development of deep learning and the increase of RS data, there are more and more change detection methods based on supervised learning. In this paper, we improve the semantic segmentation network UNet++ and propose a fully convolutional siamese network (Siam-NestedUNet) for change detection. We combine three types of siamese structures with UNet++ respectively to explore the impact of siamese structures on the change detection task under the condition of a backbone network with strong feature extraction capabilities. In addition, for the characteristics of multiple outputs in Siam-NestedUNet, we design a set of experiments to explore the importance level of the output at different semantic levels. According to the experimental results, our method improves greatly on a number of indicators, including precision, recall, F1-Score and overall accuracy, and has better performance than other SOTA change detection methods. Our implementation will be released at https://github.com/likyoo/Siam-NestedUNet. | ['Sheng Fang', 'Zhe Li', 'Kaiyu Li'] | 2020-10-27 | null | null | null | null | ['change-detection-for-remote-sensing-images'] | ['miscellaneous'] | [ 9.83396918e-02 -4.45847362e-01 1.15401074e-01 -4.88811225e-01
-2.08647639e-01 -4.98761714e-01 5.60833514e-01 -4.80639264e-02
-4.15258318e-01 3.83550942e-01 4.68289927e-02 -1.87508747e-01
-8.81235376e-02 -1.15541661e+00 -5.99892557e-01 -5.75297832e-01
-2.33595759e-01 -5.60733639e-02 6.14423573e-01 -3.79614443e-01
1.03697978e-01 4.62942749e-01 -1.27529657e+00 -5.19474521e-02
1.11650789e+00 1.02271414e+00 3.14777046e-01 3.23543519e-01
-2.41353467e-01 3.10963064e-01 -3.32017034e-01 -2.14022845e-02
4.17284995e-01 -4.02250737e-01 -6.05938673e-01 -1.96267754e-01
3.37736428e-01 -2.45791614e-01 -2.04883680e-01 1.39916408e+00
5.72629571e-01 2.73964275e-02 3.92945409e-01 -1.03953624e+00
-4.43297952e-01 8.37355316e-01 -8.69553983e-01 5.67542672e-01
-2.26480111e-01 2.32476681e-01 9.73710358e-01 -8.46372545e-01
6.13813043e-01 1.18317914e+00 9.10356462e-01 -1.60764500e-01
-7.79457152e-01 -9.56304073e-01 5.20714879e-01 2.70352542e-01
-1.38180995e+00 -1.79161802e-01 7.44745910e-01 -3.51101309e-01
7.05474973e-01 1.54347166e-01 7.60891676e-01 5.66660821e-01
-8.91926512e-02 8.43604624e-01 9.36908245e-01 -2.06799414e-02
1.99520633e-01 -1.89784735e-01 2.14352489e-01 7.37293422e-01
2.98717231e-01 -1.05999686e-01 1.49215036e-03 2.48801172e-01
8.02337945e-01 3.12737942e-01 -3.82063478e-01 -1.29762724e-01
-1.20071971e+00 6.88072979e-01 1.06017172e+00 5.41585147e-01
-2.01673985e-01 1.72896296e-01 3.43670130e-01 1.75714314e-01
6.85318947e-01 4.16835040e-01 -6.12630606e-01 5.11380024e-02
-9.92307305e-01 -1.44318994e-02 3.63541484e-01 6.60257459e-01
8.72842491e-01 -1.53304249e-01 -1.06573336e-01 9.05473769e-01
2.11187780e-01 6.59511030e-01 3.62228423e-01 -6.60821319e-01
5.81333697e-01 8.93205106e-01 1.59311984e-02 -1.12078083e+00
-5.00137806e-01 -7.72807419e-01 -9.78322327e-01 -1.06480867e-01
-7.25812651e-03 -2.00434864e-01 -1.08686399e+00 1.48857117e+00
3.76519293e-01 2.05575123e-01 -2.02397123e-01 8.63578796e-01
8.48998725e-01 8.31789076e-01 -9.90101099e-02 9.39790010e-02
1.09417295e+00 -1.09041345e+00 -5.80070198e-01 -1.42481640e-01
5.62009573e-01 -3.68418425e-01 1.33853066e+00 3.82824279e-02
-3.71638387e-01 -5.31685650e-01 -9.47625399e-01 1.55488670e-01
-6.03505611e-01 3.57524753e-01 5.64401388e-01 2.31630132e-01
-8.60318959e-01 5.94474852e-01 -1.02544093e+00 -6.65656984e-01
7.27115750e-01 5.04024886e-02 -3.77340615e-02 1.70690343e-02
-1.29108977e+00 5.65918744e-01 3.41135681e-01 4.53392893e-01
-7.59246767e-01 -6.48394644e-01 -6.02989674e-01 1.41677260e-01
4.88930076e-01 -3.53772432e-01 1.04305542e+00 -1.05926752e+00
-1.12089586e+00 5.35368264e-01 3.71562876e-02 -1.46672249e-01
6.10327423e-01 -3.00578415e-01 -4.66168582e-01 1.48949385e-01
1.89054102e-01 8.07566762e-01 5.05814612e-01 -1.08202004e+00
-8.96178663e-01 -4.73237276e-01 1.08354375e-01 1.55504361e-01
-3.29092592e-01 -2.16115676e-02 -3.87642086e-01 -7.46416867e-01
3.75693530e-01 -8.04552078e-01 -3.32978040e-01 2.56990254e-01
-3.10204297e-01 -2.98838019e-01 8.96810353e-01 -6.90277159e-01
1.30768526e+00 -2.36582994e+00 -1.75640732e-01 2.35633820e-01
1.26244664e-01 6.01064146e-01 -3.69992524e-01 2.26421595e-01
4.84879650e-02 6.06913984e-01 -8.11542571e-01 1.63766459e-01
-2.11802110e-01 1.51615039e-01 -3.23515274e-02 1.97741747e-01
3.66261274e-01 8.99768114e-01 -9.43182409e-01 -3.51451904e-01
7.51591101e-02 7.40630552e-02 -3.48944217e-01 1.32086491e-02
-3.26519608e-01 3.60773593e-01 -6.59255028e-01 7.81326532e-01
7.90266812e-01 -2.19318524e-01 -1.50616318e-01 -1.23294167e-01
-4.66615319e-01 7.76260272e-02 -1.19416916e+00 1.52488053e+00
-2.38443494e-01 5.13335347e-01 -6.65660622e-03 -8.22671831e-01
9.91756201e-01 -1.78114817e-01 4.41639513e-01 -7.31968820e-01
-8.62308964e-02 3.60585123e-01 8.58294219e-02 -6.54083192e-01
4.11822826e-01 1.22326352e-01 7.38321021e-02 1.75818190e-01
-3.63629609e-01 -2.91876611e-03 2.35107005e-01 1.02263585e-01
9.91728902e-01 1.96906254e-01 2.07845807e-01 -3.64820957e-01
3.35501760e-01 2.60925204e-01 6.95523620e-01 6.46545649e-01
-5.02687633e-01 4.34729934e-01 3.16181004e-01 -3.51544648e-01
-6.50921881e-01 -7.25263000e-01 -2.61886209e-01 8.61840189e-01
4.68039811e-01 -6.61281869e-02 -4.61196899e-01 -6.89540148e-01
9.93386507e-02 3.62562686e-01 -5.31549990e-01 -1.14183813e-01
-5.00107527e-01 -1.13170481e+00 6.57785833e-01 5.49878061e-01
1.31098390e+00 -1.15684342e+00 -4.60712999e-01 1.47355497e-01
-1.37876183e-01 -8.77758741e-01 -2.94177681e-01 -2.58193891e-02
-1.04623103e+00 -9.73167479e-01 -6.04564428e-01 -8.10372174e-01
4.94348168e-01 5.74071109e-01 6.90552235e-01 -2.22825389e-02
-7.50814676e-02 -1.57800734e-01 -6.11446202e-01 -3.02306205e-01
2.95994915e-02 5.47860682e-01 -4.84249502e-01 -7.51350448e-02
1.55524313e-01 -6.10654235e-01 -8.49008381e-01 2.95984805e-01
-1.25174403e+00 -2.22630873e-02 6.78500950e-01 3.19116563e-01
4.29174542e-01 4.78670150e-01 4.32924598e-01 -7.24754691e-01
3.13710332e-01 -5.37771583e-01 -4.73202556e-01 1.93363026e-01
-7.23816335e-01 -1.32229462e-01 3.18294764e-01 -2.59208649e-01
-1.06290424e+00 -1.27010986e-01 -2.32346490e-01 -3.95498574e-02
-2.36261055e-01 8.81510317e-01 -3.34448010e-01 1.38576448e-01
5.22740364e-01 2.06690237e-01 -1.65661260e-01 -7.40551293e-01
2.19240561e-01 8.23567688e-01 2.85441101e-01 2.21959464e-02
8.29163074e-01 5.60339391e-01 -5.00099480e-01 -8.21430922e-01
-9.17301059e-01 -5.33247232e-01 -5.18465281e-01 -1.50758430e-01
7.25063860e-01 -1.07692444e+00 6.55554533e-02 1.04039109e+00
-7.39000082e-01 -5.38712621e-01 -1.09118022e-01 3.55572104e-01
1.67659655e-01 3.42573851e-01 -3.90020311e-01 -4.20185685e-01
-5.87665260e-01 -9.02904153e-01 9.53083396e-01 5.35112262e-01
2.02929869e-01 -8.06189179e-01 8.14122185e-02 1.09225832e-01
6.26749635e-01 3.77450496e-01 7.18310595e-01 -2.70620823e-01
-5.02580464e-01 1.47721335e-01 -5.81599116e-01 2.70799220e-01
2.93570489e-01 2.55970031e-01 -6.94348097e-01 -2.60402948e-01
-2.45543107e-01 2.13999167e-01 1.38898337e+00 4.00195211e-01
1.15811634e+00 -2.54821420e-01 -4.43797141e-01 7.34580517e-01
1.66568446e+00 2.45273784e-01 6.82246864e-01 6.82160497e-01
8.15868258e-01 3.77288967e-01 5.56832910e-01 2.59343714e-01
5.96649051e-01 4.02946949e-01 7.51412511e-01 -3.59688222e-01
-2.38223508e-01 -1.52676344e-01 4.94937509e-01 7.44720340e-01
6.92486465e-02 -5.44733778e-02 -1.09583795e+00 8.41531873e-01
-1.76432574e+00 -9.26346779e-01 -4.62983161e-01 1.74032581e+00
8.17328334e-01 5.63838594e-02 -6.98097572e-02 -1.39799723e-02
9.29379225e-01 3.97688836e-01 -9.02646780e-01 1.39370412e-01
-1.76036224e-01 3.79042476e-02 4.70702559e-01 2.72093564e-01
-1.34728944e+00 1.15317702e+00 5.19092655e+00 7.13752031e-01
-1.33203125e+00 1.39405176e-01 3.88732702e-01 1.43332496e-01
-3.09192032e-01 -8.34997520e-02 -6.95461929e-01 6.09443605e-01
3.56794089e-01 2.81865448e-01 1.02422714e-01 4.79566216e-01
5.49313486e-01 -4.01009947e-01 -4.74501222e-01 5.87387502e-01
-2.13992760e-01 -1.00761521e+00 1.63779184e-01 -2.16931090e-01
9.49525297e-01 6.40410841e-01 -1.62663192e-01 3.23287189e-01
2.45420575e-01 -6.29259646e-01 6.77957237e-01 3.84025007e-01
6.24650955e-01 -4.73239869e-01 9.09985304e-01 3.00124705e-01
-1.29519117e+00 -4.91166413e-02 -1.89546883e-01 -2.92531792e-02
-1.24192908e-01 9.66342449e-01 -3.95645767e-01 6.46548092e-01
9.76145029e-01 1.29246998e+00 -8.71119380e-01 1.23906338e+00
-5.19085765e-01 9.25352335e-01 -5.60967505e-01 3.71507779e-02
4.15382743e-01 -3.42207253e-01 5.35556912e-01 1.22698081e+00
3.29314053e-01 1.11318082e-01 2.48160511e-01 8.32463324e-01
-6.45693317e-02 4.19450514e-02 -3.76110911e-01 -2.08949465e-02
6.10825121e-01 1.22802031e+00 -9.72858787e-01 -3.28248531e-01
-8.46908391e-02 7.70306647e-01 1.10888109e-01 3.43536377e-01
-8.77379954e-01 -6.29489779e-01 6.55219257e-01 -4.32541408e-03
4.84053433e-01 -3.43058378e-01 -2.85280079e-01 -1.25936306e+00
2.07415625e-01 -6.03138506e-01 3.15124273e-01 -7.43343532e-01
-1.01280880e+00 4.45340574e-01 -2.53792554e-02 -1.07699275e+00
6.00324690e-01 -2.41669908e-01 -8.51507604e-01 5.66802561e-01
-2.00023341e+00 -9.92672682e-01 -7.58140326e-01 3.36444587e-01
5.28435409e-01 2.17316449e-01 3.39173973e-01 3.40628803e-01
-1.05437660e+00 2.13164017e-01 2.67683178e-01 4.45936739e-01
4.41649199e-01 -1.06448019e+00 5.86472392e-01 1.33813465e+00
-9.96814147e-02 2.57993042e-01 3.34315658e-01 -6.33548379e-01
-7.39722490e-01 -1.55738342e+00 5.55916548e-01 6.35249019e-02
7.39766538e-01 -5.71047850e-02 -9.37127233e-01 4.79571968e-01
-1.04983933e-01 -1.19818293e-01 2.62758136e-01 -5.86912315e-03
-2.66116351e-01 -4.78140980e-01 -1.11808324e+00 4.27687764e-01
1.33449662e+00 -2.66002774e-01 -4.10889804e-01 5.55663109e-02
8.72847140e-01 -1.35237038e-01 -7.81071663e-01 6.90665126e-01
3.10004801e-01 -7.50654995e-01 5.83446622e-01 -1.64063066e-01
5.80381453e-01 -7.03351855e-01 -9.17696282e-02 -1.33849537e+00
-4.07763839e-01 9.28983614e-02 3.02793026e-01 1.21210408e+00
4.37691361e-01 -8.25498402e-01 3.97369951e-01 -4.23173681e-02
-1.76291123e-01 -4.69871491e-01 -6.62567496e-01 -7.61918604e-01
6.17060624e-02 -1.92260981e-01 8.44307065e-01 1.19616771e+00
-6.02422655e-01 3.58554065e-01 1.15219779e-01 4.64093029e-01
2.72419810e-01 4.45006847e-01 5.40553749e-01 -1.32603765e+00
1.20326951e-01 -7.00431168e-01 -2.16076642e-01 -1.07159400e+00
-3.56575340e-01 -1.01050234e+00 3.41473147e-02 -1.86654997e+00
2.64114320e-01 -6.27649426e-01 -3.30728799e-01 1.00038958e+00
-4.59199429e-01 2.17042282e-01 2.13329867e-02 5.88096261e-01
-5.37937105e-01 8.51856768e-01 1.40448225e+00 -2.91621596e-01
-3.93868059e-01 -2.04255208e-02 -5.58033466e-01 5.99070430e-01
1.27403915e+00 -5.18763006e-01 -6.55986145e-02 -8.30045342e-01
2.79985368e-01 -4.92680043e-01 4.18629289e-01 -1.06411517e+00
-9.48374718e-02 -1.85096249e-01 1.77586779e-01 -5.97020149e-01
-4.25863415e-01 -6.78904533e-01 1.84510008e-01 6.87484741e-01
-2.02482387e-01 -6.49774596e-02 9.61730927e-02 4.09068465e-01
-3.41988742e-01 -6.37907013e-02 9.11783636e-01 -1.82098150e-01
-1.12605977e+00 4.83729661e-01 -2.00130567e-01 -5.15736677e-02
8.26078475e-01 -1.55323029e-01 -5.41699886e-01 -6.50575850e-03
-5.57814538e-01 5.99091828e-01 3.48486274e-01 5.34728408e-01
3.28875005e-01 -1.08746552e+00 -6.76691473e-01 5.73681742e-02
2.85512179e-01 3.88662994e-01 5.22811450e-02 9.96178865e-01
-7.40713537e-01 9.04718190e-02 -1.53753400e-01 -7.07026362e-01
-8.81417811e-01 1.14913005e-02 6.06148958e-01 -2.94140726e-01
-6.74278498e-01 7.28575528e-01 2.62695290e-02 -8.32681477e-01
2.05551484e-03 -7.84394503e-01 -2.48417184e-01 1.94736212e-01
2.40701780e-01 4.87785488e-01 -1.45741813e-02 -2.96508968e-01
-3.72407764e-01 5.04970729e-01 1.12259157e-01 1.28581464e-01
1.65869713e+00 -2.47386366e-01 -3.27349603e-01 3.54211897e-01
1.04923081e+00 -2.77220130e-01 -1.29384923e+00 -4.01701063e-01
9.03078467e-02 -2.24625766e-01 2.90242493e-01 -1.10754120e+00
-1.47164917e+00 7.16829956e-01 1.00383627e+00 1.25652328e-01
1.15857804e+00 -6.77516162e-02 8.23008776e-01 4.12503570e-01
1.53573126e-01 -9.85095441e-01 -1.02036513e-01 6.31735563e-01
8.43262255e-01 -1.42315423e+00 -1.64855972e-01 -3.64036441e-01
-2.64810711e-01 7.72800386e-01 7.30065525e-01 7.57749146e-03
8.62685204e-01 3.36584798e-03 2.44418055e-01 -3.92344892e-01
-1.01665057e-01 -5.99366903e-01 -7.12250173e-02 2.38761306e-01
1.44067228e-01 2.27760613e-01 -4.07170147e-01 3.08611035e-01
-1.01760812e-01 -9.36225150e-03 3.72021258e-01 7.83360243e-01
-8.74092460e-01 -8.10711861e-01 -1.86133251e-01 5.84634960e-01
-1.38495535e-01 -2.24110737e-01 -4.34778094e-01 8.13078165e-01
3.98211837e-01 9.43428934e-01 7.06662983e-02 -2.62909085e-01
4.81898308e-01 -2.76752889e-01 -7.03774765e-02 -5.43179870e-01
-6.65124059e-01 -1.46254050e-02 -1.65467113e-01 -3.95777851e-01
-8.23420644e-01 -7.49056458e-01 -1.41728199e+00 -1.41954601e-01
-3.75704587e-01 4.70337681e-02 6.04758978e-01 8.85492980e-01
4.36328918e-01 4.72899348e-01 7.97435522e-01 -4.28429991e-01
-4.44676101e-01 -1.27634966e+00 -6.10736251e-01 1.30608633e-01
1.96339488e-01 -4.55975950e-01 -2.47275472e-01 -1.88100278e-01] | [9.668091773986816, -1.2758382558822632] |
772c98f4-a24f-41f1-a666-e04701b15d4d | protecting-individual-interests-across | 2105.03714 | null | https://arxiv.org/abs/2105.03714v2 | https://arxiv.org/pdf/2105.03714v2.pdf | Consistency of Constrained Spectral Clustering under Graph Induced Fair Planted Partitions | Spectral clustering is popular among practitioners and theoreticians alike. While performance guarantees for spectral clustering are well understood, recent studies have focused on enforcing ``fairness'' in clusters, requiring them to be ``balanced'' with respect to a categorical sensitive node attribute (e.g. the race distribution in clusters must match the race distribution in the population). In this paper, we consider a setting where sensitive attributes indirectly manifest in an auxiliary \textit{representation graph} rather than being directly observed. This graph specifies node pairs that can represent each other with respect to sensitive attributes and is observed in addition to the usual \textit{similarity graph}. Our goal is to find clusters in the similarity graph while respecting a new individual-level fairness constraint encoded by the representation graph. We develop variants of unnormalized and normalized spectral clustering for this task and analyze their performance under a \emph{fair} planted partition model induced by the representation graph. This model uses both the cluster membership of the nodes and the structure of the representation graph to generate random similarity graphs. To the best of our knowledge, these are the first consistency results for constrained spectral clustering under an individual-level fairness constraint. Numerical results corroborate our theoretical findings. | ['Ambedkar Dukkipati', 'Shubham Gupta'] | 2021-05-08 | null | null | null | null | ['stochastic-block-model'] | ['graphs'] | [ 2.71834344e-01 4.07362968e-01 -4.55834955e-01 -6.40958369e-01
-3.23293567e-01 -7.22032428e-01 4.50938880e-01 5.23795605e-01
-4.04299408e-01 5.20945191e-01 1.10598050e-01 -9.17633846e-02
-5.45655370e-01 -9.21536744e-01 -4.36600417e-01 -8.41719389e-01
-2.23210260e-01 8.95893216e-01 -1.08697847e-01 -1.42214447e-01
1.39906798e-02 2.21201390e-01 -1.43597388e+00 -1.67524174e-01
1.04220533e+00 4.98171896e-01 -2.61524349e-01 2.72386342e-01
2.22096682e-01 7.09981203e-01 -3.77786964e-01 -5.91205418e-01
6.44414783e-01 -7.00199246e-01 -8.50405276e-01 2.54054457e-01
2.34423861e-01 5.21646976e-01 -3.00325871e-01 1.50926220e+00
3.45875323e-01 3.26409310e-01 1.00830388e+00 -1.97756004e+00
-7.80866146e-01 1.18369913e+00 -9.99283552e-01 -1.53360711e-02
2.68538833e-01 -2.99436301e-01 1.52532661e+00 -1.05960660e-01
5.15140593e-01 1.09783292e+00 3.64727736e-01 6.01487160e-01
-1.84930372e+00 -8.25299978e-01 4.79081944e-02 -6.52266070e-02
-1.84255004e+00 -5.47974288e-01 8.16210330e-01 -6.10385001e-01
-2.31228191e-02 7.93841183e-01 2.02757418e-01 6.55199349e-01
-5.29685259e-01 1.71298340e-01 1.15018284e+00 -4.80423927e-01
6.06798112e-01 5.43270297e-02 4.16031569e-01 3.92579138e-01
6.66776597e-01 1.21537019e-02 -3.78620088e-01 -4.84709620e-01
3.51661354e-01 -2.00602174e-01 -1.97615027e-01 -9.11332190e-01
-8.33723128e-01 1.03434646e+00 4.59564567e-01 1.29307434e-01
-1.75339609e-01 2.58844703e-01 2.93080926e-01 2.91509658e-01
4.16143835e-01 2.50874490e-01 8.88064429e-02 4.72566724e-01
-1.04011405e+00 -8.66043717e-02 7.64063716e-01 1.10360479e+00
9.53789413e-01 -2.29428023e-01 -1.92778468e-01 6.22824728e-01
2.52031147e-01 5.48996925e-01 -1.10954501e-01 -1.39083803e+00
2.52436012e-01 5.94740272e-01 -3.70670892e-02 -1.18787336e+00
-4.44361627e-01 -4.67127144e-01 -1.37274361e+00 -8.42610225e-02
7.84447730e-01 6.71823323e-02 -5.72925627e-01 2.57529259e+00
2.74284303e-01 2.39460424e-01 -1.88377649e-01 1.00170994e+00
1.89327806e-01 2.16675475e-01 1.78300455e-01 -5.33706486e-01
1.20023394e+00 -1.46544293e-01 -5.02064288e-01 1.77800745e-01
5.36979258e-01 -4.07089591e-01 9.34832275e-01 -2.07381845e-02
-9.65579331e-01 -1.86917130e-02 -7.68092155e-01 3.51302683e-01
7.31951892e-02 -4.07298923e-01 3.22725266e-01 1.34763110e+00
-1.31119800e+00 3.58142704e-01 -4.27708685e-01 -5.63994825e-01
4.69134837e-01 3.67537618e-01 -3.58184785e-01 -4.44720387e-02
-1.24016798e+00 3.27189952e-01 3.43166888e-01 -4.91720885e-01
-3.37651372e-01 -5.60461283e-01 -7.87113726e-01 2.12147996e-01
6.15812600e-01 -7.50908971e-01 6.78608418e-01 -1.38012648e+00
-8.42115998e-01 1.31775165e+00 8.00863504e-02 -2.28135034e-01
4.50416982e-01 6.14534378e-01 -3.11615527e-01 2.20654696e-01
5.04204988e-01 3.13914865e-01 5.91232002e-01 -1.58428228e+00
-4.23355043e-01 -5.52690089e-01 -7.45560527e-02 1.65966898e-01
-3.32707286e-01 2.92232305e-01 -2.42166162e-01 -6.19713545e-01
1.26562849e-01 -1.08972263e+00 -4.53900099e-01 -1.20131806e-01
-9.53066647e-01 -1.07916743e-01 3.19367588e-01 -8.65238195e-04
1.48524439e+00 -2.25124574e+00 6.49165884e-02 1.19182563e+00
5.77092171e-01 -3.48331541e-01 -1.36736020e-01 2.80794263e-01
-4.54975337e-01 4.77821022e-01 -6.36388540e-01 -2.77465671e-01
4.09917414e-01 2.44179547e-01 -8.68630596e-03 1.25797212e+00
-4.65528190e-01 3.96572143e-01 -9.10904467e-01 -6.03757977e-01
-3.49171571e-02 -1.81247339e-01 -7.57774651e-01 -4.69514616e-02
1.28705531e-01 1.92839831e-01 -2.52407342e-01 2.73608148e-01
7.27862358e-01 -3.34315479e-01 7.25178778e-01 1.01188533e-01
1.65790439e-01 -2.52330601e-01 -1.57988501e+00 1.19128835e+00
2.69664019e-01 2.56937407e-02 3.91594917e-01 -1.40391278e+00
8.23201120e-01 1.44242987e-01 8.72362018e-01 -3.03191632e-01
1.84967533e-01 1.82779916e-02 3.03404331e-01 1.45821109e-01
4.01035160e-01 -5.60949504e-01 -5.23620486e-01 9.46857691e-01
-3.59356910e-01 2.66602010e-01 1.31749555e-01 8.86596024e-01
8.56542945e-01 -6.01759553e-01 4.44262952e-01 -1.02129555e+00
3.49630058e-01 -2.78203547e-01 8.06456923e-01 8.78525496e-01
-5.18175066e-01 5.99241316e-01 9.33204651e-01 2.28925049e-01
-1.25668037e+00 -1.31073701e+00 -1.48432374e-01 1.32122540e+00
2.98493981e-01 -4.26279932e-01 -9.10399079e-01 -4.27760363e-01
1.65316314e-02 5.94165623e-01 -9.95038807e-01 -3.90213490e-01
-1.22237590e-03 -9.00960207e-01 7.00354993e-01 2.44427592e-01
2.66069084e-01 -2.68567741e-01 -3.55440199e-01 -2.67260551e-01
-3.73311818e-01 -8.10357690e-01 -8.90947104e-01 1.31079915e-03
-3.35091949e-01 -1.37751615e+00 -3.23998034e-01 -5.45304537e-01
1.02719045e+00 2.02544481e-01 1.13340235e+00 1.08588494e-01
-1.33113429e-01 5.59853196e-01 -3.23461980e-01 6.72247484e-02
-2.44232193e-01 2.62210472e-03 3.44364077e-01 4.53910828e-01
2.83272266e-01 -7.65706658e-01 -5.79695106e-01 5.03117442e-01
-1.11344302e+00 -2.23559231e-01 -6.92316219e-02 5.69858670e-01
4.06007916e-01 3.23784232e-01 4.67432797e-01 -1.22215116e+00
5.90379417e-01 -7.85273612e-01 -3.53065163e-01 3.01200569e-01
-7.04286754e-01 1.18834168e-01 4.68793303e-01 -9.76089463e-02
-8.31085682e-01 6.50442541e-02 5.59274852e-01 -2.81924546e-01
2.63277199e-02 3.30682307e-01 -5.86857378e-01 3.55094492e-01
8.21893692e-01 -3.55820686e-01 5.88480905e-02 -6.49466887e-02
7.17064917e-01 5.47740877e-01 6.47735119e-01 -1.14230239e+00
9.72705781e-01 7.08167732e-01 3.67424965e-01 -3.36210668e-01
-5.95794380e-01 -4.48934168e-01 -6.17849827e-01 -1.96548104e-01
6.97802126e-01 -6.09359860e-01 -1.07396972e+00 7.69567713e-02
-3.66754502e-01 -1.52224183e-01 -6.62634015e-01 2.53706098e-01
-6.07859612e-01 6.43979728e-01 -3.26895416e-01 -1.09086168e+00
6.31774813e-02 -8.05483401e-01 6.04286432e-01 -1.05520815e-01
-3.98496062e-01 -8.62231672e-01 9.94540751e-02 1.69435605e-01
1.19842775e-01 4.39469665e-01 9.91432607e-01 -8.47775042e-01
-2.47671202e-01 -1.88870002e-02 -3.50113034e-01 -5.65617979e-02
3.19429576e-01 1.42215729e-01 -7.98227191e-01 -4.84675705e-01
-2.91787058e-01 -1.60888553e-01 8.19666266e-01 5.51124930e-01
1.19713676e+00 -5.17054796e-01 -4.40406293e-01 4.21991765e-01
1.34296787e+00 -1.63276121e-01 3.28875154e-01 -9.04371142e-02
6.96402073e-01 1.01073003e+00 6.18488155e-02 7.17087448e-01
5.29544652e-01 6.80096924e-01 3.28810900e-01 -3.78353894e-02
3.26388776e-01 -1.00612439e-01 -2.28226632e-02 5.17181635e-01
-1.62977055e-01 -3.04179758e-01 -9.45839703e-01 4.83261287e-01
-2.00086522e+00 -1.20975244e+00 -4.40018743e-01 2.62840915e+00
7.95380414e-01 -2.03840271e-01 5.84335327e-01 2.19871685e-01
1.49713945e+00 2.77610924e-02 -3.57551396e-01 -2.65433967e-01
-4.62523133e-01 1.94507122e-01 6.63436711e-01 7.91766942e-01
-1.01357877e+00 5.33289671e-01 5.72119713e+00 9.54534650e-01
-4.46649343e-01 6.81396797e-02 1.01164329e+00 -1.67033657e-01
-6.69286668e-01 2.77324796e-01 8.47344846e-02 5.97439468e-01
9.16591704e-01 -7.64899194e-01 4.71403629e-01 4.41982836e-01
1.89308718e-01 -1.05234958e-01 -1.15163505e+00 7.59014606e-01
-9.52493846e-02 -7.19639778e-01 -7.77650923e-02 4.38607663e-01
8.45047116e-01 -4.76168215e-01 2.87565619e-01 -1.72751680e-01
1.08924937e+00 -1.24739921e+00 7.13011861e-01 4.28662390e-01
1.23899269e+00 -1.28375649e+00 3.09295595e-01 2.25916103e-01
-1.38084352e+00 4.48813029e-02 -1.97917640e-01 3.38177010e-03
-1.98438793e-01 8.41555059e-01 -1.32112607e-01 7.50303805e-01
6.55850351e-01 4.09845382e-01 -4.21664596e-01 7.64546752e-01
1.79051965e-01 7.32978046e-01 -2.44466677e-01 5.39049387e-01
-6.72127679e-02 -5.96741021e-01 5.47858179e-01 9.01327491e-01
1.67895690e-01 4.41510439e-01 4.90801215e-01 1.07355082e+00
-3.92226845e-01 3.78845572e-01 -5.30391455e-01 9.29803923e-02
8.43555927e-01 1.11185110e+00 -1.21061480e+00 -3.17919433e-01
-2.22238284e-02 7.52419353e-01 2.83808708e-01 4.06698674e-01
-6.62729502e-01 -2.62854606e-01 1.01315510e+00 3.32731158e-01
-2.46725500e-01 1.86091796e-01 -4.31585342e-01 -1.01662707e+00
-4.43576932e-01 -6.65965199e-01 1.07466233e+00 -3.98221850e-01
-1.78488696e+00 1.57260656e-01 1.25075042e-01 -8.16247821e-01
1.47296712e-01 -9.33821872e-03 -4.71786678e-01 9.39718366e-01
-8.46710443e-01 -7.37099409e-01 4.46251631e-02 1.08494854e+00
-4.90223229e-01 7.06100315e-02 7.50847399e-01 1.28926113e-01
-7.41618276e-01 9.00110602e-01 2.85079926e-01 1.49194211e-01
7.23689318e-01 -1.51763940e+00 -1.79594651e-01 9.89359975e-01
7.40973055e-02 7.98359513e-01 9.17101681e-01 -6.50940478e-01
-7.39335716e-01 -1.14707613e+00 7.83600330e-01 -4.11313355e-01
6.19263709e-01 -5.39270341e-01 -7.21073627e-01 6.97608948e-01
1.59591585e-01 6.14195913e-02 1.28392339e+00 2.19578773e-01
-8.53648901e-01 -2.38704473e-01 -1.53470981e+00 6.16734624e-01
1.30358863e+00 -6.24099851e-01 -1.30132958e-01 6.42768815e-02
4.79671836e-01 1.97810292e-01 -8.58150065e-01 3.01444501e-01
1.13917086e-02 -1.01950979e+00 8.29104960e-01 -6.71971560e-01
1.50090352e-01 -4.68062878e-01 -5.35330713e-01 -1.37597883e+00
-7.86477029e-01 -7.77508974e-01 4.49652076e-01 1.43657374e+00
2.20488250e-01 -6.61209106e-01 9.15797412e-01 1.03491795e+00
2.76298523e-01 5.18039390e-02 -1.27403784e+00 -7.78441191e-01
4.11179334e-01 -2.29834422e-01 7.87253320e-01 1.56987333e+00
4.67532843e-01 4.07533556e-01 -4.43378478e-01 2.38593847e-01
1.28893471e+00 2.11923927e-01 3.86895448e-01 -1.72689915e+00
-1.80161849e-01 -8.42429101e-01 -4.28300649e-01 -1.18406124e-01
5.58787942e-01 -1.51222348e+00 -1.49738982e-01 -1.07860780e+00
9.44339871e-01 -8.60653579e-01 -2.55998343e-01 3.79085064e-01
-3.07820201e-01 3.59548032e-01 2.50716597e-01 2.43729949e-01
-6.98021948e-01 3.33620787e-01 6.77976072e-01 -9.30952374e-03
-5.91948964e-02 -5.54804578e-02 -1.34601939e+00 3.12684864e-01
1.01552176e+00 -4.83655661e-01 -6.37934923e-01 3.06144834e-01
3.61849785e-01 2.58068275e-02 3.80338132e-01 -4.85947400e-01
3.93337101e-01 -5.17788231e-01 -3.11638452e-02 5.38336858e-02
-9.04274583e-02 -9.55437839e-01 6.30785525e-01 3.73288423e-01
-7.08109319e-01 3.15253399e-02 -3.12465489e-01 7.53684282e-01
2.14590833e-01 1.35929823e-01 1.05755663e+00 7.45453611e-02
-1.31906956e-01 2.92659938e-01 -2.59655505e-01 6.24161065e-01
1.39197874e+00 -2.79186696e-01 -4.56319779e-01 -7.41376698e-01
-7.19228864e-01 3.99409860e-01 1.01438153e+00 -1.04303226e-01
2.64142931e-01 -1.49217999e+00 -1.03058839e+00 -1.19382881e-01
3.45503896e-01 -3.35697085e-01 1.86599582e-01 9.00683820e-01
5.10063507e-02 -4.32376005e-02 -2.11437568e-01 -4.57963288e-01
-1.19578683e+00 1.02575076e+00 4.45850044e-01 1.76406860e-01
-2.49178097e-01 7.14170635e-01 6.00482523e-01 -4.19906735e-01
-1.23386998e-02 3.51875454e-01 1.16671450e-01 2.01911747e-01
-5.27064577e-02 6.19666398e-01 -4.76950884e-01 -1.16277909e+00
-5.58453202e-01 3.30026120e-01 2.55200624e-01 -3.55966002e-01
8.64039063e-01 -4.05067593e-01 -5.10243356e-01 3.29115540e-01
8.81367803e-01 2.85356581e-01 -7.69954205e-01 -3.80686402e-01
2.53159136e-01 -7.14726686e-01 -3.87764394e-01 -4.23796266e-01
-1.46241748e+00 3.86398405e-01 1.14206858e-01 5.98979831e-01
1.08260310e+00 3.72728139e-01 -1.05228126e-01 -1.57624185e-01
2.42432684e-01 -1.26750743e+00 -1.45413741e-01 -1.50440171e-01
3.04216176e-01 -8.53068411e-01 -3.25540304e-02 -7.45565057e-01
-6.70855463e-01 3.92027646e-01 3.57608289e-01 -1.22393467e-01
7.78206766e-01 6.12963922e-02 -2.36097813e-01 -1.96704596e-01
-3.44703257e-01 -4.70207900e-01 2.30084255e-01 8.93472731e-01
3.93345267e-01 7.23180592e-01 -3.02275747e-01 6.29114091e-01
-3.08379650e-01 -6.04407728e-01 7.50984371e-01 3.72817248e-01
-4.50007915e-01 -1.16443956e+00 -6.62837267e-01 6.78406119e-01
-4.22972113e-01 -1.30436629e-01 -8.52407336e-01 4.12156105e-01
1.86524510e-01 1.39333177e+00 3.50317508e-01 -2.08190665e-01
1.97047204e-01 -1.63203001e-01 3.28074634e-01 -6.37197316e-01
-4.42824274e-01 -4.35180739e-02 7.84434602e-02 -4.71014708e-01
-7.10906684e-01 -8.82382333e-01 -1.14115334e+00 -9.95153666e-01
-1.98175684e-01 6.86986446e-01 -1.52032867e-01 4.17325258e-01
5.43426760e-02 1.96268037e-01 9.28336501e-01 -1.79042697e-01
-4.16634887e-01 -4.86499935e-01 -1.33814836e+00 8.99314225e-01
-3.71633493e-03 -5.46737790e-01 -4.84934419e-01 -4.54866415e-04] | [7.1525983810424805, 5.106605529785156] |
5a716cac-57df-4757-a44a-8eb670d7684e | structuring-user-generated-content-on-social | 2210.15377 | null | https://arxiv.org/abs/2210.15377v2 | https://arxiv.org/pdf/2210.15377v2.pdf | Retrieving Users' Opinions on Social Media with Multimodal Aspect-Based Sentiment Analysis | People post their opinions and experiences on social media, yielding rich databases of end-users' sentiments. This paper shows to what extent machine learning can analyze and structure these databases. An automated data analysis pipeline is deployed to provide insights into user-generated content for researchers in other domains. First, the domain expert can select an image and a term of interest. Then, the pipeline uses image retrieval to find all images showing similar content and applies aspect-based sentiment analysis to outline users' opinions about the selected term. As part of an interdisciplinary project between architecture and computer science researchers, an empirical study of Hamburg's Elbphilharmonie was conveyed. Therefore, we selected 300 thousand posts with the hashtag \enquote{\texttt{hamburg}} from the platform Flickr. Image retrieval methods generated a subset of slightly more than 1.5 thousand images displaying the Elbphilharmonie. We found that these posts mainly convey a neutral or positive sentiment towards it. With this pipeline, we suggest a new semantic computing method that offers novel insights into end-users opinions, e.g., for architecture domain experts. | ['Georg Groh', 'Tobias Eder', 'Miriam Anschütz'] | 2022-10-27 | null | null | null | null | ['aspect-based-sentiment-analysis'] | ['natural-language-processing'] | [-4.33314532e-01 3.94513682e-02 3.28478105e-02 -4.48759496e-01
-8.16186249e-01 -1.03644824e+00 8.10834885e-01 6.84399962e-01
-6.08979762e-01 4.78731096e-03 5.62525868e-01 1.04837641e-02
2.12352127e-01 -7.83167005e-01 -2.39865318e-01 -5.67462027e-01
2.12606028e-01 4.07441169e-01 -1.33656813e-02 -5.62269568e-01
8.99477959e-01 2.75097132e-01 -1.77324796e+00 8.75176370e-01
1.49261996e-01 1.35359764e+00 2.69219071e-01 4.99968976e-01
-5.19521952e-01 9.67501104e-01 -8.15042198e-01 -8.82732391e-01
1.14625275e-01 5.27461916e-02 -8.68713260e-01 2.46050403e-01
6.09581351e-01 -1.00892879e-01 1.13167427e-01 1.14077258e+00
6.05619431e-01 -1.53830394e-01 5.03275812e-01 -1.24750566e+00
-7.77340531e-01 6.40685380e-01 -3.28644633e-01 2.59385794e-01
6.36078656e-01 -1.57387108e-01 1.41498554e+00 -1.10313404e+00
1.10375607e+00 1.18264484e+00 4.50822115e-01 -1.78011626e-01
-6.23935521e-01 -3.19183677e-01 8.67406558e-03 5.06744869e-02
-1.55870366e+00 -2.03228503e-01 7.65447080e-01 -6.81571782e-01
5.79281926e-01 4.06927288e-01 9.90605652e-01 7.76217461e-01
1.24273058e-02 9.32582915e-01 1.21681917e+00 -4.93334949e-01
4.88650590e-01 9.14695203e-01 5.57898208e-02 4.81456488e-01
-1.45015538e-01 -7.96507359e-01 -6.49666369e-01 -3.63225520e-01
1.44745067e-01 3.39755505e-01 1.36440203e-01 -1.97754845e-01
-1.35155070e+00 9.69935775e-01 6.13234639e-01 5.32589197e-01
-5.20332575e-01 -7.31740743e-02 4.21061188e-01 5.49687803e-01
9.40134287e-01 6.81581795e-01 -4.08038348e-01 -6.51914254e-02
-9.89969313e-01 2.70327508e-01 9.51717019e-01 1.05801141e+00
1.19392455e+00 -6.46696687e-01 4.72707860e-02 1.00235403e+00
4.54203635e-01 5.91723621e-01 3.43472689e-01 -1.01757562e+00
6.08852543e-02 8.85531008e-01 2.44845122e-01 -1.70964432e+00
-2.54287738e-02 -3.92104745e-01 -1.79039910e-01 7.95036703e-02
-2.06735656e-01 -1.64578408e-01 -6.00723982e-01 8.52280855e-01
3.14949334e-01 -6.22546017e-01 -6.33107498e-02 1.14455128e+00
1.05122113e+00 5.72907329e-01 2.52563283e-02 2.73591131e-01
1.90099061e+00 -6.16706729e-01 -3.50608885e-01 -1.78516984e-01
6.57097101e-01 -1.40265191e+00 1.16583931e+00 4.45618927e-01
-8.85258913e-01 -2.80862153e-01 -7.04343319e-01 7.25854486e-02
-8.76529396e-01 -1.95900835e-02 5.46493649e-01 5.77374101e-01
-1.33249176e+00 3.48249018e-01 -7.10503161e-02 -9.03755486e-01
5.41547537e-01 -2.98482254e-02 -1.34410392e-02 -8.72184932e-02
-8.83526325e-01 5.26116729e-01 -2.81920999e-01 -4.92919415e-01
-5.81480682e-01 -7.57980108e-01 -6.05961144e-01 -3.70286793e-01
2.31897160e-01 -4.97469813e-01 1.23915446e+00 -1.57974243e+00
-9.37627673e-01 1.55627418e+00 1.31172519e-02 -9.58715752e-02
2.31526867e-01 -8.65295380e-02 -6.94346070e-01 6.55284464e-01
6.71104193e-01 7.84665763e-01 9.30449307e-01 -1.10533023e+00
-9.17722225e-01 -4.99299169e-01 4.08640504e-01 3.06871921e-01
-8.81324291e-01 5.04421413e-01 -7.73799062e-01 -6.94338977e-01
-1.38493285e-01 -1.10939860e+00 -1.89093769e-01 -1.69122711e-01
-5.09491041e-02 -2.72273034e-01 7.02385962e-01 -3.10578644e-01
1.27286518e+00 -2.21775222e+00 -1.94827124e-01 6.79054916e-01
3.70513231e-01 -4.74174529e-01 -1.28859049e-03 9.67536151e-01
2.73732960e-01 3.75879616e-01 3.21611971e-01 -1.10031225e-01
9.97736007e-02 -2.21221820e-01 -3.33640307e-01 3.90325129e-01
-2.29525015e-01 7.43509352e-01 -1.08184183e+00 -7.61454344e-01
4.78056446e-03 4.11786139e-01 -5.68309009e-01 -2.22731411e-01
-3.02849382e-01 1.55255329e-02 -8.75886321e-01 9.32121754e-01
3.63834411e-01 -4.93808627e-01 -3.83272171e-02 -5.54595768e-01
-5.18833637e-01 -1.60204813e-01 -5.62707603e-01 1.63078344e+00
-6.87010467e-01 9.84240770e-01 1.16764426e-01 -5.19907236e-01
9.17273998e-01 9.10072476e-02 8.39679718e-01 -7.14118838e-01
3.52642447e-01 1.78317040e-01 -7.90759563e-01 -5.52988887e-01
9.07199740e-01 -1.01654880e-01 -6.36338532e-01 6.19484782e-01
-3.36434215e-01 -5.19571841e-01 1.27315983e-01 5.30675530e-01
1.06857908e+00 -3.26122522e-01 -1.42137498e-01 -6.46966159e-01
3.41020614e-01 8.06225538e-01 -2.94842273e-01 4.36860830e-01
-6.71104342e-02 8.15590024e-01 2.69731045e-01 -8.13847959e-01
-1.19020092e+00 -3.97328079e-01 1.50337573e-02 1.43344998e+00
1.84864357e-01 -9.18210030e-01 -6.93533242e-01 -4.68187004e-01
-8.78515765e-02 5.26445091e-01 -4.40289885e-01 4.22785163e-01
-3.53160203e-02 -4.83851373e-01 -6.21976405e-02 -4.13547158e-02
5.30107141e-01 -1.13515031e+00 -6.53152943e-01 -1.14425741e-01
-4.02659237e-01 -1.08980298e+00 -3.99548262e-01 -4.21164513e-01
-2.59227723e-01 -7.73843586e-01 -1.07048631e+00 -9.48923945e-01
9.20989454e-01 4.72956359e-01 1.67219555e+00 3.19197685e-01
-3.56685728e-01 1.10412395e+00 -7.98412383e-01 -4.86729205e-01
-3.29187326e-03 1.01022631e-01 -3.24530423e-01 2.26345971e-01
8.42325330e-01 -6.59716949e-02 -1.07603133e+00 4.39943850e-01
-9.96574521e-01 -4.59012210e-01 4.22861814e-01 1.11474041e-02
4.06393021e-01 2.41099089e-01 1.43853799e-01 -9.10986185e-01
8.51782203e-01 -8.00251842e-01 -2.94339389e-01 6.67060763e-02
-5.43054819e-01 -3.78990501e-01 2.24282742e-01 -2.56368145e-03
-8.09889019e-01 7.55416648e-03 6.29789755e-02 -2.08280608e-01
-9.47338045e-02 7.88162887e-01 3.28244656e-01 1.49477571e-01
7.25576580e-01 -2.83804327e-01 -1.84676751e-01 -1.58612967e-01
3.97080123e-01 1.18080688e+00 6.28459156e-02 -2.97694832e-01
5.12317777e-01 1.01006305e+00 -6.14833236e-01 -1.36726737e+00
-1.10096896e+00 -1.06030893e+00 -2.12069348e-01 -7.69395232e-01
1.04806709e+00 -1.34467340e+00 -5.81985414e-01 2.83783734e-01
-7.54999936e-01 -2.34143361e-02 -1.81323528e-01 1.21628098e-01
-2.52780825e-01 2.00859621e-01 -2.75869817e-01 -6.59499645e-01
-3.89793634e-01 -8.80140960e-01 1.30372739e+00 -7.67540336e-02
-7.90638804e-01 -9.32761669e-01 -5.23448102e-02 7.68845558e-01
4.82213378e-01 2.62110531e-02 4.30689067e-01 -6.93961561e-01
-3.88456732e-01 -5.51905334e-01 -3.47911686e-01 2.25504085e-01
-2.52190441e-01 2.32929617e-01 -8.11924815e-01 -2.11770892e-01
-1.11932054e-01 -3.22225243e-01 4.73089218e-01 2.05191553e-01
9.88196194e-01 -4.10557777e-01 -2.73959160e-01 -1.56973988e-01
1.57306278e+00 -1.07127853e-01 6.77906036e-01 9.39406335e-01
5.28447390e-01 8.95748556e-01 9.22739685e-01 8.56998205e-01
6.28791571e-01 3.75011027e-01 4.18034256e-01 -6.58723488e-02
2.34990194e-01 1.28219053e-01 4.94265020e-01 7.61599898e-01
1.90077871e-02 -1.40513286e-01 -9.41690147e-01 9.04344559e-01
-1.60362792e+00 -8.30711722e-01 -2.16263607e-01 1.64143848e+00
3.84048074e-01 2.20844403e-01 1.61089540e-01 -3.46537501e-01
7.27178931e-01 3.52224112e-01 -1.19928874e-01 -3.56770784e-01
-1.61075741e-01 1.23197466e-01 7.39965677e-01 1.69230998e-01
-9.13117170e-01 9.61561978e-01 5.82051802e+00 7.72306085e-01
-9.53060448e-01 -4.37022224e-02 9.18770134e-01 -1.23628274e-01
-7.96950340e-01 1.66735813e-01 -7.95289457e-01 3.60900760e-01
9.00860131e-01 4.24224511e-02 3.10343117e-01 1.12381649e+00
1.90056011e-01 -2.59070545e-01 -6.80764556e-01 1.02780747e+00
5.11635363e-01 -1.49915957e+00 -5.65629750e-02 1.06707983e-01
9.08255756e-01 1.45667687e-01 2.81103253e-01 6.20238595e-02
2.72265464e-01 -5.56532681e-01 1.04755747e+00 5.89131355e-01
5.67099035e-01 -8.33899379e-01 7.51846313e-01 -1.43102214e-01
-9.99539614e-01 -1.68205753e-01 -3.05705041e-01 -1.18009008e-01
-1.25191063e-01 1.01871622e+00 -8.47360432e-01 -2.69121546e-02
1.22248507e+00 9.97118473e-01 -7.99907684e-01 5.11829019e-01
-6.14491329e-02 2.10537970e-01 -1.77792624e-01 -5.04602253e-01
4.35001075e-01 -2.14311287e-01 4.28791642e-01 1.34176481e+00
6.34825230e-01 -1.37838110e-01 1.04164906e-01 4.14140642e-01
-1.22769192e-01 4.46910024e-01 -6.49106264e-01 -4.80870038e-01
4.18279171e-01 1.90059710e+00 -1.25657165e+00 -5.20057201e-01
-4.32535112e-01 1.09065008e+00 -1.62357271e-01 2.21411034e-01
-4.82069165e-01 -3.44903558e-01 4.86152738e-01 6.86069310e-01
4.44911301e-01 -7.50181973e-02 -5.57206757e-02 -9.81660485e-01
-9.88135338e-02 -1.05621088e+00 3.96063417e-01 -1.27812350e+00
-1.26021183e+00 7.23783195e-01 -1.85560063e-01 -1.19020057e+00
9.24916565e-02 -5.36613703e-01 -4.49013382e-01 3.27731609e-01
-1.19213247e+00 -1.22581398e+00 -4.46211368e-01 6.05592668e-01
6.88280642e-01 -4.24575686e-01 5.82916141e-01 3.49259257e-01
-8.80601779e-02 -2.04778701e-01 -1.32307082e-01 -3.00947465e-02
8.90707433e-01 -1.23425937e+00 1.30307272e-01 6.87726960e-02
2.05368444e-01 5.97358346e-01 9.00889635e-01 -6.24289513e-01
-1.42404580e+00 -8.37244749e-01 1.07159102e+00 -7.17081249e-01
1.08016419e+00 -1.19042285e-01 -6.70912266e-02 4.17389691e-01
8.88027251e-01 -4.44225043e-01 1.08822560e+00 2.87002437e-02
-7.91613460e-02 -1.04809953e-02 -9.46592808e-01 8.96103740e-01
5.48287809e-01 -8.15403819e-01 -5.10098040e-01 6.23545945e-01
4.21448976e-01 3.19010377e-01 -9.62006152e-01 -2.40310490e-01
5.32684207e-01 -1.05766547e+00 1.04914868e+00 -7.33107030e-02
8.06928635e-01 -2.25321412e-01 -3.96887660e-01 -1.31566370e+00
6.80634333e-03 -4.93111044e-01 7.01294363e-01 1.27897477e+00
5.67006707e-01 -2.23331571e-01 8.03042352e-01 8.10021460e-01
-1.06303193e-01 -2.17562512e-01 -9.64234173e-02 9.91694108e-02
-1.72064930e-01 -6.24266207e-01 4.36444730e-01 9.71557617e-01
1.39178202e-01 4.92141008e-01 1.46290019e-01 -1.25927031e-01
3.80786598e-01 3.50482106e-01 7.91831613e-01 -1.23660266e+00
2.05452248e-01 -4.39968884e-01 -4.08984482e-01 -7.32904553e-01
-8.83985087e-02 -9.51525688e-01 -3.67951572e-01 -1.57233810e+00
1.27667859e-01 -3.26918155e-01 -2.48909041e-01 1.28145516e-01
5.55071652e-01 7.85553098e-01 -2.79475581e-02 6.14693582e-01
-1.16126132e+00 -1.41870439e-01 1.05277789e+00 -2.49649510e-01
1.61630604e-02 -1.84385598e-01 -1.24624908e+00 1.06637239e+00
6.98827624e-01 -3.36600065e-01 -1.75983906e-01 -2.67022312e-01
1.41995120e+00 -5.72432756e-01 4.95593131e-01 -8.65339637e-01
3.14840376e-01 1.08310297e-01 3.23349804e-01 -9.44273233e-01
1.99165508e-01 -9.85089540e-01 2.72866748e-02 4.14440148e-02
-3.11611146e-01 3.20388049e-01 -1.34317666e-01 2.80620009e-01
-5.34821749e-01 -3.47209334e-01 3.19607735e-01 -8.04983199e-01
-1.11443937e+00 2.71639810e-03 -9.27719295e-01 -6.14100471e-02
8.89224946e-01 -2.78793741e-02 -2.00820088e-01 -8.17526400e-01
-8.74185920e-01 1.25903144e-01 8.19568396e-01 2.97534853e-01
6.22568846e-01 -1.28594923e+00 -6.38948858e-01 -8.57393965e-02
5.89048088e-01 -3.64152938e-01 2.30525851e-01 7.13700235e-01
-7.13400066e-01 9.03919488e-02 5.59230754e-03 -5.76917171e-01
-1.32438028e+00 4.20307666e-01 -2.74111450e-01 1.60150275e-01
-3.53923976e-01 8.23879421e-01 -5.70911728e-02 -2.71492213e-01
-1.07846864e-01 7.56910443e-02 -4.67847288e-01 1.10160089e+00
5.49530685e-01 1.52356058e-01 1.69837788e-01 -1.01999104e+00
-4.29665536e-01 7.21773922e-01 -7.71142170e-02 -3.19111407e-01
1.62550020e+00 -6.03559911e-01 -4.68307763e-01 3.18288207e-01
1.65028048e+00 2.78244317e-01 -5.92920363e-01 -6.70017675e-02
2.35031977e-01 -5.72177768e-01 4.14200783e-01 -5.98698676e-01
-1.16719425e+00 4.45477098e-01 4.70289558e-01 9.21277702e-01
1.04346681e+00 4.97462034e-01 6.82463646e-01 4.58378047e-01
3.05603683e-01 -1.69186366e+00 5.46872556e-01 3.64916474e-01
1.04954767e+00 -1.37605989e+00 3.27459186e-01 -7.39387646e-02
-1.16577101e+00 1.11434412e+00 -6.58661351e-02 -1.87952191e-01
1.04558396e+00 -1.26717597e-01 5.12868464e-01 -1.11867738e+00
-5.46081126e-01 -3.27416092e-01 2.03928038e-01 9.94446948e-02
5.47107697e-01 7.93584213e-02 -1.18570700e-01 2.16171518e-01
-7.64229953e-01 -1.92980856e-01 5.48981249e-01 1.19510901e+00
-7.18881547e-01 -9.09902513e-01 -5.21289825e-01 3.85397613e-01
-8.57045710e-01 -3.14389288e-01 -8.72887552e-01 4.33455497e-01
-5.33870161e-02 1.14177179e+00 2.82453954e-01 -3.68905365e-01
3.37218255e-01 -1.73462078e-01 -1.34573206e-01 -7.13235021e-01
-9.29223239e-01 3.44359964e-01 3.20270091e-01 -5.08550942e-01
-9.25594807e-01 -7.33676791e-01 -1.03824770e+00 -2.22725406e-01
4.24194336e-02 2.94517249e-01 1.08997154e+00 5.75485289e-01
6.16305351e-01 1.88978806e-01 7.92773724e-01 -6.95898712e-01
2.30581030e-01 -5.12569010e-01 -4.41254050e-01 7.39520133e-01
3.16067487e-02 -5.87957986e-02 -5.89996457e-01 3.20488274e-01] | [12.862785339355469, 5.305628299713135] |
a4a2e1a7-0bfb-41b0-bfaa-c37975f44014 | integrated-sensing-and-communication-for-6g | 2208.02157 | null | https://arxiv.org/abs/2208.02157v4 | https://arxiv.org/pdf/2208.02157v4.pdf | Integrated Sensing and Communication for 6G: Ten Key Machine Learning Roles | Integrating sensing and communication is a defining theme for future wireless systems. This is motivated by the promising performance gains, especially as they assist each other, and by the better utilization of the wireless and hardware resources. Realizing these gains in practice, however, is subject to several challenges where leveraging machine learning can provide a potential solution. This article focuses on ten key machine learning roles for joint sensing and communication, sensing-aided communication, and communication-aided sensing systems, explains why and how machine learning can be utilized, and highlights important directions for future research. The article also presents real-world results for some of these machine learning roles based on the large-scale real-world dataset DeepSense 6G, which could be adopted in investigating a wide range of integrated sensing and communication problems. | ['Ahmed Alkhateeb', 'Umut Demirhan'] | 2022-08-03 | null | null | null | null | ['machine-learning', 'machine-learning'] | ['methodology', 'miscellaneous'] | [ 5.46056807e-01 2.60659784e-01 -7.97353745e-01 -2.07344592e-01
-9.52214479e-01 -1.86831281e-01 2.06988100e-02 -2.84961879e-01
-8.86502117e-03 8.94156337e-01 2.35896245e-01 -7.27224231e-01
-3.32109481e-01 -7.20234573e-01 -3.09947789e-01 -9.89239097e-01
-6.78655922e-01 -1.82333514e-02 -3.52200419e-01 1.02895916e-01
-3.65821898e-01 2.74118990e-01 -6.82780325e-01 -1.54390618e-01
5.45637369e-01 1.63808703e+00 8.58604461e-02 8.27504218e-01
2.07485005e-01 6.05053604e-01 -6.71243846e-01 -1.38683572e-01
2.85430133e-01 -2.84735709e-01 -1.83135197e-01 -1.35765687e-01
-1.93251550e-01 -4.64387834e-01 -2.36279771e-01 5.64054370e-01
9.53232765e-01 -3.83079708e-01 4.00244981e-01 -1.24737120e+00
-3.42028350e-01 1.11023784e+00 -8.91513705e-01 1.41090482e-01
3.50343496e-01 -4.88262117e-01 1.12546837e+00 -2.62906939e-01
-3.54045749e-01 9.23071384e-01 7.94658482e-01 3.03261161e-01
-8.91833544e-01 -1.25433969e+00 -1.98600486e-01 -2.85299458e-02
-1.18060803e+00 -7.81815350e-01 8.48635137e-01 -4.25283983e-02
6.96060479e-01 3.64018768e-01 8.00121546e-01 8.83580446e-01
1.48022622e-01 8.34956229e-01 7.84091830e-01 -7.99753904e-01
3.37938070e-01 2.41883799e-01 -1.42481953e-01 5.93910992e-01
2.91432559e-01 6.87422156e-01 -4.94850010e-01 -4.97279406e-01
6.00623429e-01 9.08963084e-02 -2.64362156e-01 -1.27209365e-01
-1.16122925e+00 5.99952340e-01 5.77206314e-01 2.08462000e-01
-4.11992043e-01 6.35818124e-01 -1.58806130e-01 3.96817356e-01
3.99272829e-01 2.95778185e-01 -4.47976410e-01 -2.85268068e-01
-1.29370606e+00 -6.54830754e-01 8.01767230e-01 1.04457831e+00
5.48511982e-01 4.06921834e-01 7.83509314e-02 8.41088712e-01
8.49834204e-01 1.06767404e+00 -1.90034434e-01 -1.09154201e+00
4.77295011e-01 -1.84441045e-01 2.11653501e-01 -8.18965852e-01
-7.63606608e-01 -9.87568140e-01 -1.24137568e+00 -1.14221491e-01
5.07419743e-02 -1.02618098e+00 -3.37663919e-01 1.72701526e+00
-1.31155588e-02 5.86205244e-01 2.43555218e-01 5.70343733e-01
6.63641810e-01 3.64982069e-01 -7.18833357e-02 -5.13556004e-01
1.18485391e+00 -8.40436339e-01 -7.28101313e-01 -4.60096538e-01
5.02126157e-01 -6.72504544e-01 2.40446448e-01 4.29348439e-01
-9.89096165e-01 -5.33905268e-01 -1.52137804e+00 6.12571478e-01
9.27253589e-02 4.55004066e-01 8.77083719e-01 1.47426784e+00
-1.08898401e+00 -2.01882780e-01 -1.10614729e+00 -5.31737626e-01
7.02386498e-01 6.83687925e-01 5.83315909e-01 5.22804540e-03
-1.09605801e+00 1.93151101e-01 -2.07415760e-01 -6.06207773e-02
-4.42691743e-01 -7.46227443e-01 -4.05414969e-01 1.82575956e-01
3.38828206e-01 -6.45797133e-01 1.30861199e+00 -3.22909296e-01
-1.45131588e+00 3.26598644e-01 -1.82741508e-01 -6.67347848e-01
2.32052311e-01 -8.23853016e-02 -1.02948880e+00 -4.90253642e-02
-1.22319691e-01 2.88972795e-01 6.87198699e-01 -1.10453773e+00
-8.60752344e-01 -3.53215516e-01 2.07719374e-02 -2.08253786e-01
-7.63011277e-01 -2.87804365e-01 -2.13504329e-01 -5.96080959e-01
4.29096848e-01 -1.04760575e+00 -5.51594079e-01 1.31784275e-01
-7.45597303e-01 4.00727749e-01 8.71769369e-01 -3.16223472e-01
1.12044251e+00 -2.19281864e+00 -4.41269875e-01 6.55431390e-01
4.85578865e-01 1.10518686e-01 -9.56618488e-02 4.51892465e-01
3.83099169e-01 3.36951375e-01 2.61647135e-01 -3.55277270e-01
-2.96121150e-01 6.56100139e-02 -8.53189304e-02 3.05331886e-01
-5.28799057e-01 9.21421170e-01 -7.81234622e-01 -1.99853569e-01
3.41392398e-01 3.83634239e-01 -2.21964419e-01 -4.21523154e-02
1.32955581e-01 7.43702888e-01 -9.55308139e-01 9.57583368e-01
7.55948484e-01 -6.15374625e-01 3.62668663e-01 -4.24833268e-01
3.06770682e-01 -5.69748804e-02 -9.16567743e-01 1.34217095e+00
-9.42956686e-01 5.29005706e-01 5.08009613e-01 -1.33469248e+00
8.63637984e-01 5.10663390e-01 1.21966064e+00 -9.73373413e-01
1.29114434e-01 3.31531256e-01 1.36471549e-02 -5.11461139e-01
-8.52051452e-02 -5.53884991e-02 -2.73221228e-02 8.61930549e-01
-4.35707778e-01 2.07283065e-01 -3.97821605e-01 9.83647183e-02
1.36511946e+00 -5.79929531e-01 6.94667995e-01 -1.26138821e-01
1.25678584e-01 -4.31044072e-01 3.90238017e-01 1.15616024e+00
-4.92375195e-01 -1.17902346e-01 -2.23980740e-01 1.45286560e-01
-3.84944826e-01 -9.68284309e-01 -8.50809962e-02 1.24878705e+00
4.58770454e-01 1.06128277e-02 -2.60829497e-02 -1.57464862e-01
3.50053370e-01 1.23377755e-01 -2.72991598e-01 -1.40392795e-01
-1.78733449e-02 -1.04420030e+00 8.46011221e-01 7.58016348e-01
7.58491635e-01 -3.26344371e-01 -5.89382052e-01 3.64088893e-01
-1.98385209e-01 -1.54755545e+00 -1.24476217e-02 4.85417992e-01
-7.57550895e-01 -8.61010790e-01 -6.36040390e-01 -3.06660533e-01
-1.19648799e-02 9.47835803e-01 9.60891724e-01 2.94469327e-01
7.88360015e-02 7.22325981e-01 -5.28844297e-01 -8.47991586e-01
-1.57740802e-01 3.39531809e-01 1.13036275e-01 -2.33390573e-02
2.50399590e-01 -9.53108609e-01 -6.41381383e-01 3.83562952e-01
5.84828518e-02 -4.07451779e-01 1.01950455e+00 4.92838234e-01
2.31072649e-01 3.12974244e-01 9.29606557e-01 -6.89093590e-01
3.78273934e-01 -9.60521340e-01 -4.19782132e-01 1.18538700e-01
-6.47905052e-01 -3.49629760e-01 1.85472637e-01 -1.04186311e-01
-9.19251621e-01 -4.95051444e-02 -4.29738253e-01 1.92254066e-01
2.96614528e-01 5.29996157e-01 -3.37214231e-01 -6.19180918e-01
6.88162744e-01 -7.45863616e-02 -6.24669157e-02 -1.96927443e-01
2.72167653e-01 1.26194477e+00 8.12243298e-02 -6.00424290e-01
1.02471447e+00 8.21846783e-01 2.31843218e-01 -1.26228201e+00
-1.07203877e+00 -6.21173978e-01 -2.14854375e-01 -1.19960770e-01
4.67247099e-01 -1.54511273e+00 -5.78504801e-01 3.51458006e-02
-7.30195343e-01 -4.85429734e-01 -4.22737077e-02 7.89925396e-01
-2.86401689e-01 1.03030652e-01 -3.17271322e-01 -1.31725025e+00
-3.81203473e-01 -9.23937917e-01 1.05461061e+00 2.51034468e-01
-2.66310692e-01 -1.20092320e+00 -2.99079329e-01 7.80099988e-01
8.65253866e-01 2.69474462e-02 6.26350105e-01 -3.78935635e-01
-7.03501165e-01 -4.06855345e-02 -4.25008893e-01 -7.39839077e-02
3.95548701e-01 -3.31346720e-01 -1.15256095e+00 -5.54694235e-01
-2.78901190e-01 -2.18457788e-01 8.15149844e-01 9.14272189e-01
1.26336360e+00 -5.95571063e-02 -1.19527853e+00 8.07985723e-01
1.26272845e+00 4.47446376e-01 4.65085179e-01 -3.12311143e-01
3.65253508e-01 -2.04750642e-01 5.97165287e-01 7.62221813e-01
3.67962569e-01 6.10390067e-01 6.60955429e-01 -6.34903312e-01
-1.20354131e-01 4.04150560e-02 6.58795536e-02 5.39138496e-01
9.25383996e-03 -5.06015122e-01 -5.72879851e-01 1.63439840e-01
-1.67680025e+00 -9.43053246e-01 -5.24095818e-02 1.75737834e+00
2.52586573e-01 1.43153474e-01 3.68752867e-01 6.58001840e-01
3.04180950e-01 3.45282972e-01 -7.25756109e-01 1.13248631e-01
-5.49391918e-02 2.49056607e-01 8.48217726e-01 4.25918430e-01
-9.75638747e-01 3.16525578e-01 7.72224998e+00 8.13639402e-01
-1.33757150e+00 5.15282214e-01 7.18106210e-01 -6.66827941e-03
-4.87155560e-03 -3.84821117e-01 -5.51590741e-01 3.98681462e-01
8.73453677e-01 1.60034955e-01 3.11776936e-01 7.66029596e-01
6.86190844e-01 -2.77148247e-01 -1.02970397e+00 1.21365857e+00
-3.55323464e-01 -1.56071568e+00 -5.73957980e-01 5.05001724e-01
6.75055325e-01 4.11804765e-01 1.54467493e-01 2.02649087e-01
2.60721356e-01 -7.31543601e-01 3.62357408e-01 2.31089771e-01
1.03182113e+00 -2.54358232e-01 6.97156072e-01 2.84496397e-01
-1.50599849e+00 -6.84613645e-01 -5.77602312e-02 -4.77172077e-01
1.57308713e-01 1.34490061e+00 -4.24461186e-01 5.44802129e-01
7.48640478e-01 9.78538871e-01 6.12250483e-03 1.06678963e+00
-9.20491368e-02 1.01526976e+00 -5.17984331e-01 -3.68283629e-01
-2.03314438e-01 6.36765659e-02 3.38101834e-01 1.12149584e+00
6.75094426e-01 2.22465470e-01 6.21336818e-01 4.49252516e-01
-1.15739025e-01 -3.16085637e-01 -4.35728580e-01 1.13853589e-01
1.23275113e+00 1.31816351e+00 -4.08830047e-01 -1.05636388e-01
-8.16427708e-01 2.79791385e-01 -5.65346897e-01 7.36089945e-01
-7.94011056e-01 -5.46336584e-02 7.71751761e-01 1.34781465e-01
4.90914248e-02 -3.69890630e-01 -8.48696649e-01 -6.73555613e-01
-9.51727182e-02 -4.82914895e-01 2.30356574e-01 -4.70826924e-01
-1.01687491e+00 -7.41624758e-02 -3.71508062e-01 -1.30978954e+00
1.79253593e-02 -2.79845744e-01 -6.09947860e-01 4.71747339e-01
-1.91539025e+00 -1.22227693e+00 -5.95246494e-01 5.67591548e-01
1.48727551e-01 -4.46075141e-01 9.13913012e-01 8.53073895e-01
-3.17499489e-01 8.69493306e-01 3.37800145e-01 2.20933169e-01
3.00816983e-01 -8.37057948e-01 -8.27659145e-02 5.97135723e-01
2.38886863e-01 1.81449935e-01 3.96088928e-01 -8.17418918e-02
-1.55326891e+00 -1.06448829e+00 2.51868010e-01 1.07318200e-01
5.32488108e-01 -2.57742882e-01 1.12439968e-01 4.08851534e-01
1.36785731e-01 2.37216711e-01 1.42678022e+00 4.65411127e-01
-9.17216316e-02 -7.30319321e-01 -1.19921196e+00 3.17837715e-01
1.22519577e+00 -4.14098799e-01 5.61003327e-01 3.27812374e-01
4.23767716e-01 -1.90755904e-01 -9.70529497e-01 4.45991993e-01
1.05107999e+00 -8.25455666e-01 1.07732558e+00 6.24731146e-02
-2.35400364e-01 1.21904492e-01 -6.82176292e-01 -1.32616568e+00
-2.66584456e-01 -9.88081634e-01 -7.03606784e-01 1.10008538e+00
6.53232872e-01 -6.81010067e-01 1.28118074e+00 5.41357026e-02
1.80644110e-01 -9.99437749e-01 -8.22293818e-01 -8.66603971e-01
-2.14422539e-01 -7.62935221e-01 8.21661949e-01 6.50906980e-01
1.86580509e-01 4.28827316e-01 -7.07065940e-01 3.55706811e-01
8.01070392e-01 -5.51761910e-02 8.12183440e-01 -1.39998078e+00
-5.61539054e-01 -2.77550429e-01 -2.29371637e-01 -1.97922385e+00
-2.34531745e-01 -4.46525872e-01 -2.36928552e-01 -1.35295737e+00
6.74721897e-02 -1.26029611e+00 -6.03848815e-01 5.51859021e-01
2.29535311e-01 6.19436741e-01 3.49549688e-02 1.00162491e-01
-6.62134767e-01 3.23152751e-01 1.11110580e+00 -4.97178435e-01
-1.76181018e-01 1.03083992e+00 -1.47319639e+00 5.20822942e-01
1.13575137e+00 8.31272453e-02 -6.39990509e-01 -4.62320447e-01
2.03657612e-01 4.90394115e-01 1.07602395e-01 -1.47591186e+00
3.05373847e-01 -2.44124949e-01 3.48702759e-01 -2.29263380e-01
7.80200303e-01 -1.57480395e+00 -5.15710190e-02 5.63202798e-01
-3.53257470e-02 -6.51782334e-01 -7.65285864e-02 8.54091525e-01
1.29768297e-01 6.73034191e-01 9.00662005e-01 2.33238950e-01
-4.65803504e-01 3.32429439e-01 -4.26596969e-01 -1.96304366e-01
1.08198357e+00 -5.21400809e-01 -1.81764215e-01 -1.02451181e+00
-8.23117316e-01 3.73200804e-01 -4.84876513e-01 5.39052896e-02
4.52556759e-01 -1.24876177e+00 -5.27651548e-01 2.19955280e-01
6.58340082e-02 -7.93124795e-01 -1.06717855e-01 1.10174525e+00
2.57732272e-01 3.59061122e-01 2.36435473e-01 -9.40901160e-01
-1.19251788e+00 -8.27503577e-02 6.46050572e-01 -1.17096767e-01
-7.84371570e-02 7.74481177e-01 -5.56740642e-01 -1.91060230e-02
6.45154715e-01 -2.48781905e-01 5.96791506e-02 -3.11684132e-01
4.05251235e-01 6.14186168e-01 2.75393445e-02 -2.71827877e-01
-4.11388010e-01 7.61341751e-01 4.58283424e-01 1.26020893e-01
1.03619111e+00 -7.49320626e-01 4.28810418e-01 4.55107130e-02
8.54999602e-01 4.00256693e-01 -9.30438280e-01 -6.94112539e-01
-1.96550444e-01 -2.49004200e-01 2.83404827e-01 -1.03325152e+00
-1.64282966e+00 8.02092373e-01 9.91136789e-01 5.69222331e-01
1.41798067e+00 1.88168362e-01 9.70284283e-01 5.45226216e-01
1.01138341e+00 -8.53711069e-01 2.44527623e-01 1.28886001e-02
1.69954434e-01 -1.58909380e+00 1.92573458e-01 -7.41833866e-01
2.94453353e-01 7.87229121e-01 2.38772526e-01 4.04485315e-01
1.67643213e+00 7.63526976e-01 2.57718712e-01 5.21262698e-02
-3.87477249e-01 -3.48323017e-01 -1.10145330e-01 1.18702185e+00
6.06530428e-01 5.40007353e-01 1.09034464e-01 7.75900364e-01
-2.81082481e-01 -1.94377564e-02 1.95598722e-01 8.75911653e-01
-7.25070715e-01 -1.31088400e+00 -4.27522302e-01 1.15231335e+00
-4.79668736e-01 5.99799827e-02 -1.59466594e-01 4.53335434e-01
3.37094218e-01 1.92649591e+00 -3.72503102e-01 -8.48170519e-01
-1.33097515e-01 -8.03733170e-01 2.39661545e-01 -4.28237438e-01
2.14562476e-01 2.88503051e-01 2.24988714e-01 -6.71471119e-01
-8.47809434e-01 -5.17675638e-01 -1.12314606e+00 -4.17101324e-01
-4.66467351e-01 9.78146717e-02 5.85440993e-01 1.38886368e+00
3.76313627e-01 8.98137152e-01 9.52932239e-01 -6.47061765e-01
-5.07959008e-01 -5.91754019e-01 -9.25932288e-01 -4.24996763e-01
7.41802871e-01 -5.80313325e-01 -1.66616321e-01 -2.64586717e-01] | [6.29931640625, 1.3016233444213867] |
f066476a-68f3-4646-a32a-7736ffee87e1 | learning-to-learn-end-to-end-goal-oriented | 2110.15724 | null | https://arxiv.org/abs/2110.15724v1 | https://arxiv.org/pdf/2110.15724v1.pdf | Learning to Learn End-to-End Goal-Oriented Dialog From Related Dialog Tasks | For each goal-oriented dialog task of interest, large amounts of data need to be collected for end-to-end learning of a neural dialog system. Collecting that data is a costly and time-consuming process. Instead, we show that we can use only a small amount of data, supplemented with data from a related dialog task. Naively learning from related data fails to improve performance as the related data can be inconsistent with the target task. We describe a meta-learning based method that selectively learns from the related dialog task data. Our approach leads to significant accuracy improvements in an example dialog task. | ['Satinder Singh', 'Jonathan K. Kummerfeld', 'Janarthanan Rajendran'] | 2021-10-10 | null | https://aclanthology.org/2021.nlp4convai-1.16 | https://aclanthology.org/2021.nlp4convai-1.16.pdf | emnlp-nlp4convai-2021-11 | ['goal-oriented-dialog'] | ['natural-language-processing'] | [ 6.08941391e-02 3.86230230e-01 -3.34065482e-02 -9.85974729e-01
-1.01459324e+00 -7.05272913e-01 5.96864343e-01 2.60919631e-01
-6.56753540e-01 1.00164151e+00 3.65496576e-01 -2.36688375e-01
2.02687964e-01 -4.22821522e-01 -4.37034726e-01 -1.96618259e-01
8.72166753e-02 1.23880804e+00 4.71338123e-01 -7.61015713e-01
2.49730751e-01 -5.10686971e-02 -9.07243669e-01 5.82952201e-01
6.67240143e-01 6.83242440e-01 5.31895697e-01 8.15725088e-01
-5.28318584e-01 1.08989811e+00 -8.58878314e-01 -2.44137660e-01
1.49395198e-01 -6.88233674e-01 -1.42865193e+00 6.74318671e-02
3.96951258e-01 -6.85179174e-01 -7.03962967e-02 8.70084286e-01
2.44339645e-01 4.54120755e-01 5.01033723e-01 -1.31382763e+00
-2.36898884e-01 7.16487408e-01 1.11385718e-01 2.17674956e-01
1.56797066e-01 1.13860235e-01 1.06411695e+00 -8.87912154e-01
5.38092852e-01 1.48430490e+00 4.18957740e-01 9.81059253e-01
-1.42115986e+00 -2.92271286e-01 2.85456479e-01 -7.60718361e-02
-5.89058459e-01 -6.59491062e-01 7.01454639e-01 -4.29460615e-01
1.10433376e+00 -1.70883402e-01 3.31816614e-01 1.13341975e+00
1.86810922e-02 8.13831329e-01 9.98898566e-01 -3.46621960e-01
4.46955711e-01 3.70355397e-01 6.98620617e-01 6.89879417e-01
-3.74828219e-01 3.90746035e-02 -5.07892132e-01 -2.96996355e-01
2.06745088e-01 -2.04053983e-01 -5.10972291e-02 -2.73304105e-01
-8.42893362e-01 1.07425046e+00 4.24237102e-01 3.20885926e-01
-1.65672898e-01 -1.36972353e-01 3.53664100e-01 8.75499547e-01
6.07176065e-01 8.28775287e-01 -1.03760564e+00 -2.13083178e-01
-4.98762995e-01 5.41053653e-01 1.35676324e+00 1.00545835e+00
9.21354771e-01 -2.25790769e-01 -2.87337974e-02 1.01812792e+00
2.55474031e-01 -3.22501883e-02 5.83491564e-01 -1.20159411e+00
7.26166487e-01 7.80835629e-01 2.93721586e-01 -2.00687453e-01
-7.27210224e-01 4.43380773e-01 -1.89498842e-01 2.14202642e-01
1.21404028e+00 -7.95701563e-01 -4.41799462e-01 1.97238898e+00
1.88550666e-01 -4.67036039e-01 4.58567083e-01 6.51251853e-01
8.36183429e-01 6.47905827e-01 1.87602654e-01 -2.19644144e-01
1.26895607e+00 -1.22691822e+00 -7.27909267e-01 -7.29671776e-01
7.95692325e-01 -5.57774007e-01 1.19055724e+00 2.23167047e-01
-1.05202270e+00 -2.97819912e-01 -8.43760133e-01 -3.11488360e-01
-2.74344623e-01 -3.43382269e-01 6.51031017e-01 1.78514838e-01
-1.02048790e+00 3.71506959e-01 -4.86459136e-01 -5.64881265e-01
1.11205272e-01 5.29229105e-01 -2.68611372e-01 1.20048381e-01
-9.57897902e-01 1.44012856e+00 5.83299041e-01 -3.86399597e-01
-8.54283094e-01 -5.11642575e-01 -7.16705203e-01 2.01916829e-01
6.97863519e-01 -4.07207996e-01 2.01058745e+00 -8.33939433e-01
-1.56342351e+00 7.28359103e-01 -3.04507524e-01 -5.99716783e-01
2.73161978e-01 -2.77932703e-01 2.27558002e-01 -1.15583479e-01
-4.01864082e-01 8.38432908e-01 4.03706521e-01 -1.23960817e+00
-8.84103477e-01 -4.29577947e-01 4.06109661e-01 2.96782315e-01
-3.50671947e-01 6.36958480e-02 -5.33219837e-02 -3.14379275e-01
-7.01102242e-02 -1.11959994e+00 -2.61490226e-01 -2.23117456e-01
-2.71167845e-01 -8.00196528e-01 7.94499874e-01 -5.39940596e-01
5.84603429e-01 -1.77497041e+00 2.12746799e-01 -4.07114327e-01
2.93443054e-01 2.38294661e-01 -1.16548337e-01 4.05382454e-01
3.06953192e-01 -1.40921205e-01 -1.82912230e-01 -3.51960659e-01
-1.29442886e-01 2.06639156e-01 -3.21389407e-01 -1.54247969e-01
1.47505492e-01 8.21062326e-01 -9.57320392e-01 -3.68944585e-01
2.09886193e-01 -1.35864407e-01 -7.33906806e-01 8.59269321e-01
-8.89977217e-01 6.20133340e-01 -4.80801374e-01 1.41228572e-01
1.53593495e-01 -1.77828491e-01 3.03830117e-01 -1.20653482e-02
1.72925010e-01 8.60230863e-01 -4.53225791e-01 1.81064081e+00
-8.04663181e-01 7.23805189e-01 3.79798084e-01 -1.00509596e+00
1.04133189e+00 3.95751268e-01 2.46360496e-01 -4.17800516e-01
2.23184630e-01 7.67613426e-02 4.41127479e-01 -5.32936871e-01
4.71399128e-01 -4.41676795e-01 -9.03522000e-02 8.33173037e-01
5.15156806e-01 -2.45607063e-01 -5.90280257e-03 1.52333662e-01
1.03791726e+00 -3.04523587e-01 2.58522123e-01 -3.24974149e-01
3.87211680e-01 5.51380873e-01 2.97486901e-01 5.43802738e-01
-2.65224725e-01 1.20426156e-01 5.79934120e-01 -4.58871365e-01
-9.22991276e-01 -7.25014746e-01 1.99171573e-01 1.75762761e+00
-1.65297929e-02 -3.28404129e-01 -6.94033146e-01 -1.07054186e+00
-2.08621651e-01 8.59783351e-01 -3.16496879e-01 -1.36451840e-01
-6.77093148e-01 -6.83859646e-01 2.15961546e-01 5.23054421e-01
5.71950555e-01 -1.15104079e+00 -3.66729110e-01 2.60151982e-01
-1.43981293e-01 -1.03945184e+00 -3.45395923e-01 6.31937265e-01
-1.01471782e+00 -1.05948734e+00 -3.93851191e-01 -9.95021105e-01
5.57972729e-01 2.20448300e-01 1.28435469e+00 1.15296640e-01
2.40806580e-01 1.31755874e-01 -1.32963121e-01 -3.67888957e-01
-9.13057566e-01 1.86209410e-01 -1.30341873e-01 -5.33374250e-01
7.58692086e-01 -2.86102176e-01 2.19636634e-02 4.42620337e-01
-2.97648489e-01 -3.99278663e-02 2.05447599e-01 1.29256380e+00
-1.48244143e-01 -2.62503147e-01 1.02229965e+00 -1.03221858e+00
1.18053532e+00 -5.03606856e-01 -7.26287246e-01 3.31422031e-01
-4.71387267e-01 4.10287142e-01 7.40124941e-01 -5.20978391e-01
-1.35258842e+00 3.64253759e-01 -1.67645589e-02 -7.42199197e-02
-4.47692424e-01 3.93350333e-01 -3.03662747e-01 3.57735336e-01
7.78825521e-01 -1.69460550e-01 3.01020265e-01 -8.02353859e-01
3.43163818e-01 9.57584798e-01 3.92050892e-01 -7.03072071e-01
2.79432923e-01 -8.04079995e-02 -4.13152575e-01 -7.49903083e-01
-1.08560479e+00 -2.62514085e-01 -7.26880848e-01 -4.85597961e-02
7.22245216e-01 -7.52731085e-01 -8.70825410e-01 1.85888484e-01
-1.30011177e+00 -1.02547550e+00 -9.97313187e-02 3.48305047e-01
-4.80430663e-01 -6.21547140e-02 -6.14529729e-01 -7.31735766e-01
-2.61572510e-01 -1.17890871e+00 5.61440349e-01 9.38166678e-02
-6.10097647e-01 -1.20342994e+00 2.13588670e-01 5.64637661e-01
4.59137022e-01 -5.42300463e-01 1.05928791e+00 -1.76669717e+00
-2.45380208e-01 -1.41451567e-01 -2.48948764e-02 2.68510073e-01
4.77897912e-01 -6.06663823e-01 -1.07547033e+00 -2.76798636e-01
3.91075820e-01 -1.19548631e+00 5.76048195e-01 1.05404809e-01
7.12692857e-01 -5.55323958e-01 -1.78113982e-01 -2.15002269e-01
8.07379186e-01 5.65569282e-01 -2.82838523e-01 -4.55794819e-02
4.33523268e-01 1.34076464e+00 7.11830616e-01 1.49279460e-01
5.91933846e-01 8.49130988e-01 5.88839836e-02 2.38500103e-01
1.67620685e-02 -1.44656286e-01 3.74715924e-01 2.40766436e-01
5.67097247e-01 -3.04184586e-01 -1.08560407e+00 5.68868458e-01
-2.16754103e+00 -7.48904228e-01 1.14852063e-01 2.01197863e+00
1.24144673e+00 1.91599950e-01 4.93361115e-01 -4.64706361e-01
6.96735561e-01 -1.14286892e-01 -8.39894831e-01 -2.50482380e-01
3.65750372e-01 -1.06474645e-01 4.83679883e-02 1.12739110e+00
-9.78983164e-01 1.37851298e+00 7.35766792e+00 2.53217697e-01
-9.43352342e-01 1.82285845e-01 5.11301935e-01 -7.07940955e-04
-4.43572849e-02 1.69466928e-01 -8.82912278e-01 3.07717234e-01
1.18763649e+00 -2.79557407e-01 5.00451148e-01 1.17902637e+00
2.06220895e-02 -2.75594443e-01 -1.59178960e+00 5.39513230e-01
-1.14626981e-01 -1.15155566e+00 -2.85927474e-01 8.82906467e-03
2.40483299e-01 9.65643898e-02 -4.81241256e-01 6.65674508e-01
1.09905958e+00 -9.96057451e-01 1.47087470e-01 -1.77700948e-02
-8.25557392e-03 -5.46072781e-01 6.97134554e-01 8.87148619e-01
-6.52132273e-01 -6.52980208e-02 -5.92004359e-01 -2.53234982e-01
1.67632684e-01 1.00628927e-01 -1.49879229e+00 -1.12201922e-01
3.06069940e-01 4.81177002e-01 -4.92884904e-01 4.99485046e-01
-2.74229228e-01 4.80902076e-01 -2.43765116e-01 -3.54674548e-01
1.65046066e-01 7.15726614e-02 3.60964030e-01 9.46849108e-01
-2.03339130e-01 2.81352639e-01 5.93726575e-01 8.77657652e-01
-2.74643064e-01 -1.59687161e-01 -9.21264768e-01 7.64784515e-02
6.45986199e-01 1.27890944e+00 -2.67625391e-01 -5.83704948e-01
-4.76011336e-01 6.34358287e-01 7.66073644e-01 3.57256792e-02
-2.44795755e-01 -9.08151865e-02 3.31380755e-01 -4.46314961e-01
1.05388127e-01 -2.30036423e-01 -1.84171215e-01 -1.17197227e+00
-2.04502568e-01 -9.81459320e-01 5.51281214e-01 -5.16889930e-01
-1.57360017e+00 7.60050774e-01 1.45159706e-01 -7.79292881e-01
-1.00779140e+00 -8.62351954e-01 -7.38664210e-01 9.99230266e-01
-9.89050686e-01 -8.41580629e-01 -2.74170607e-01 7.04858363e-01
1.20363712e+00 -6.45928919e-01 1.11512983e+00 -2.60562956e-01
-3.44599605e-01 3.75059336e-01 -3.25659573e-01 3.79571795e-01
1.04254031e+00 -1.48858190e+00 3.49924862e-01 2.96410531e-01
5.98760583e-02 7.58003592e-01 8.55167329e-01 -5.32596171e-01
-1.33363223e+00 -7.13392317e-01 1.00007915e+00 -7.03235686e-01
7.23675728e-01 -4.75135505e-01 -1.10699761e+00 1.08231854e+00
6.64408922e-01 -2.95672119e-01 7.74529815e-01 6.37682199e-01
-4.04185414e-01 3.01048309e-02 -1.21078563e+00 6.57851934e-01
7.58592725e-01 -4.99644667e-01 -1.33588183e+00 5.87968230e-01
6.26314700e-01 -4.32133585e-01 -7.23028004e-01 3.80926132e-02
1.55805796e-01 -5.68079889e-01 7.57080853e-01 -1.14414310e+00
3.69587600e-01 1.85935721e-01 -3.13622653e-01 -1.58073974e+00
2.48591661e-01 -5.41084886e-01 -9.62288007e-02 1.24989128e+00
8.34953904e-01 -3.25041085e-01 8.30330193e-01 1.40700042e+00
-9.13706198e-02 -4.33521420e-01 -6.94950283e-01 -7.28421032e-01
5.13250053e-01 -2.47554220e-02 2.26190120e-01 9.78776276e-01
7.96422064e-01 1.20098674e+00 -3.20535868e-01 -2.11542979e-01
3.16914529e-01 3.47916782e-01 1.05682468e+00 -1.41020751e+00
-1.83218241e-01 -2.03391999e-01 3.05448264e-01 -1.43028831e+00
6.23429954e-01 -8.26372504e-01 4.85456824e-01 -1.30708039e+00
1.01398997e-01 -5.56886494e-01 2.26167083e-01 8.45880985e-01
-3.66035461e-01 -2.39373252e-01 2.64178872e-01 3.10849905e-01
-5.98572910e-01 5.17054319e-01 7.53220379e-01 -2.90463984e-01
-5.78279972e-01 2.50558257e-01 -6.98814511e-01 8.67444277e-01
1.05084503e+00 -4.19198185e-01 -7.84029782e-01 -3.53908330e-01
-2.91235030e-01 5.58686674e-01 2.56187558e-01 -6.21784747e-01
4.47602540e-01 -2.74928957e-01 6.65172338e-02 -3.88212264e-01
8.25080633e-01 -8.47063303e-01 -6.25275671e-01 4.57775742e-01
-1.07414675e+00 6.19489886e-02 2.48248801e-01 3.62994999e-01
-1.82763517e-01 -7.25398004e-01 8.08363199e-01 -3.70044112e-01
-6.56086564e-01 -1.51166037e-01 -5.97908497e-01 4.12577450e-01
8.06426287e-01 4.10227567e-01 -6.04241967e-01 -7.07047284e-01
-9.01563406e-01 5.40684998e-01 4.28557217e-01 5.15289545e-01
3.49829227e-01 -7.79388428e-01 -6.09816015e-01 -2.39276782e-01
-4.94388072e-03 4.25568484e-02 -1.82534412e-01 3.41253489e-01
5.16485609e-02 4.22599792e-01 -3.48505557e-01 -5.29262006e-01
-1.47263098e+00 5.95490932e-01 4.53280956e-01 -3.72645557e-01
-6.18808210e-01 8.04593623e-01 1.71515241e-01 -8.53073478e-01
4.74643260e-01 -2.42936954e-01 -2.74837792e-01 1.30728468e-01
6.06312394e-01 -1.50666252e-01 5.26599400e-02 -7.86567107e-02
-1.35078609e-01 -1.96249828e-01 -5.81558049e-01 -6.72609746e-01
1.30402005e+00 -1.50087044e-01 1.00885384e-01 7.49573886e-01
9.76690173e-01 -3.48856688e-01 -1.47539544e+00 -5.27780652e-01
3.90320271e-01 -2.01119155e-01 -1.59877151e-01 -9.91702974e-01
-8.05853128e-01 1.00115323e+00 2.18221545e-01 4.35270071e-01
6.51121795e-01 2.19655216e-01 5.95328689e-01 1.32404816e+00
4.76482749e-01 -1.04058754e+00 2.93392718e-01 9.23371077e-01
9.20109034e-01 -1.88251662e+00 -1.70675561e-01 -1.18952967e-01
-1.04610169e+00 1.11448300e+00 1.14531958e+00 4.80980165e-02
6.23793662e-01 2.97605872e-01 3.69511098e-01 -3.07034612e-01
-1.31964409e+00 -5.29593490e-02 -5.27748838e-02 5.61138690e-01
6.36301279e-01 -5.14905214e-01 -3.83182094e-02 5.95694125e-01
-3.85882527e-01 -2.82239199e-01 4.94136840e-01 1.13340783e+00
-8.25453460e-01 -1.37277901e+00 -1.91788405e-01 3.67704540e-01
-5.91845512e-02 -9.35582966e-02 -1.04191303e+00 7.91212440e-01
-5.44858992e-01 1.43136692e+00 1.27894953e-01 -3.50653350e-01
1.43343404e-01 8.33278358e-01 2.87516207e-01 -9.16896343e-01
-8.98457646e-01 -1.23213409e-02 5.29493868e-01 -3.46447676e-01
-4.76660252e-01 -3.95986021e-01 -1.34774387e+00 -2.83538252e-01
-2.79938102e-01 4.23205584e-01 4.43036884e-01 1.26264381e+00
2.99750715e-01 2.74596453e-01 6.57585621e-01 -6.65167153e-01
-6.23288631e-01 -1.37915564e+00 -7.85546675e-02 4.20450002e-01
4.70749527e-01 -5.62450826e-01 -2.41308838e-01 1.48570249e-02] | [12.865556716918945, 8.038320541381836] |
d98ef86c-0564-43d1-aaea-7098a0f7cb5d | a-dataset-and-preliminary-results-for-umpire | 1809.06217 | null | http://arxiv.org/abs/1809.06217v1 | http://arxiv.org/pdf/1809.06217v1.pdf | A Dataset and Preliminary Results for Umpire Pose Detection Using SVM Classification of Deep Features | In recent years, there has been increased interest in video summarization and
automatic sports highlights generation. In this work, we introduce a new
dataset, called SNOW, for umpire pose detection in the game of cricket. The
proposed dataset is evaluated as a preliminary aid for developing systems to
automatically generate cricket highlights. In cricket, the umpire has the
authority to make important decisions about events on the field. The umpire
signals important events using unique hand signals and gestures. We identify
four such events for classification namely SIX, NO BALL, OUT and WIDE based on
detecting the pose of the umpire from the frames of a cricket video.
Pre-trained convolutional neural networks such as Inception V3 and VGG19
net-works are selected as primary candidates for feature extraction. The
results are obtained using a linear SVM classifier. The highest classification
performance was achieved for the SVM trained on features extracted from the
VGG19 network. The preliminary results suggest that the proposed system is an
effective solution for the application of cricket highlights generation. | ['Tizhoosh Hamid R.', 'Paul Sruthy', 'Venugopal Harshwin', 'Ravi Aravind'] | 2018-09-11 | null | null | null | null | ['game-of-cricket'] | ['playing-games'] | [ 9.76962373e-02 -3.21911252e-03 2.73244902e-02 6.93033338e-02
-4.42922980e-01 -3.63180757e-01 4.78546709e-01 3.34271282e-01
-5.82359970e-01 5.49046457e-01 3.53873849e-01 2.82787412e-01
-2.84874644e-02 -6.96395457e-01 -5.74100316e-01 -6.19430900e-01
-3.29398721e-01 2.44956866e-01 5.45833528e-01 -6.27459645e-01
5.84619880e-01 6.98897302e-01 -2.00543809e+00 7.09163666e-01
1.49361283e-01 9.87176955e-01 1.52231425e-01 1.12803471e+00
3.61761391e-01 9.86902595e-01 -1.26106369e+00 -2.09409878e-01
3.13185930e-01 -4.63684291e-01 -3.67124289e-01 -1.15913808e-01
5.57396352e-01 -3.31810445e-01 -5.66264927e-01 5.09684563e-01
8.21491063e-01 3.62005532e-01 6.00992978e-01 -1.32687938e+00
4.96010244e-01 7.16652036e-01 -3.75572264e-01 7.26099491e-01
7.10447907e-01 9.06724557e-02 8.75881970e-01 -7.30090797e-01
8.73202503e-01 7.54780352e-01 6.62689209e-01 3.22006732e-01
-4.71397907e-01 -7.36049891e-01 -1.87800393e-01 5.95997393e-01
-1.27897954e+00 -9.44129080e-02 9.53678548e-01 -7.30000019e-01
6.77233994e-01 6.66313529e-01 1.16000807e+00 8.82300377e-01
9.27813500e-02 1.18747663e+00 8.98994565e-01 -6.79950833e-01
8.15763623e-02 -4.93744053e-02 3.14743310e-01 3.43782723e-01
2.47810796e-01 3.76378484e-02 -1.12633789e+00 1.51274297e-02
6.88004971e-01 -4.25992876e-01 -1.26408130e-01 1.53049767e-01
-1.06977856e+00 7.83354461e-01 3.52647007e-01 2.45034784e-01
-5.62393308e-01 1.28223464e-01 8.56727839e-01 -3.57853435e-02
2.24097580e-01 4.58750933e-01 1.64108813e-01 -5.95628202e-01
-1.22836411e+00 7.58957326e-01 5.92960000e-01 6.87219858e-01
-8.87120068e-02 2.13031992e-01 -3.64775509e-01 5.41895092e-01
-9.11779255e-02 2.12473556e-01 6.42373934e-02 -5.39573312e-01
6.52465284e-01 6.14506781e-01 1.34204835e-01 -1.18996286e+00
-6.80174112e-01 -2.20594466e-01 -1.55016840e-01 4.58362877e-01
5.54008603e-01 -4.50921506e-01 -6.53810918e-01 8.30524385e-01
3.07332367e-01 -1.09535396e-01 -4.14980650e-02 1.38709521e+00
1.49328542e+00 7.96579242e-01 -1.28062278e-01 2.11055994e-01
1.44848168e+00 -6.18893743e-01 -7.37625718e-01 2.29566861e-02
7.94115216e-02 -8.33579242e-01 7.27794409e-01 7.75732934e-01
-9.36034441e-01 -7.02765226e-01 -1.35708964e+00 3.18620563e-01
-4.77811620e-02 6.90965593e-01 5.02340019e-01 5.80254674e-01
-3.64206433e-01 5.11467040e-01 -7.38489032e-01 -5.08841038e-01
6.95644468e-02 2.95688540e-01 -2.40669325e-01 5.92346728e-01
-1.09292519e+00 1.03580928e+00 7.11287320e-01 3.98673773e-01
-6.52824283e-01 -8.61090124e-02 -7.64466047e-01 -1.20041549e-01
3.40339661e-01 2.09415749e-01 9.19539928e-01 -9.17834520e-01
-1.34481251e+00 1.04722929e+00 4.45132405e-01 -5.91454387e-01
9.02618527e-01 -3.57748896e-01 -3.26633155e-01 5.74855983e-01
1.24029316e-01 3.74178231e-01 5.90782523e-01 -9.23590958e-01
-1.10503471e+00 -2.65282899e-01 6.17712364e-02 4.08438295e-01
-8.22750479e-02 6.00612581e-01 -3.48272920e-01 -8.16396058e-01
-3.01768165e-02 -9.74595547e-01 1.80558294e-01 -5.50361395e-01
-4.82834727e-01 -1.51022553e-01 7.99133420e-01 -1.14372265e+00
1.40720892e+00 -1.87511086e+00 1.64911732e-01 4.78704840e-01
1.57268062e-01 5.24209559e-01 4.09499168e-01 5.22123873e-01
6.83265179e-02 -6.25282049e-01 3.10102791e-01 5.24341643e-01
-2.89892644e-01 -1.56400889e-01 -3.91928852e-03 2.88258374e-01
6.44648150e-02 4.64087784e-01 -7.74967432e-01 -5.86350083e-01
4.99386638e-01 1.78580269e-01 -9.79285985e-02 2.26846442e-01
1.25147834e-01 3.19730282e-01 -3.59516054e-01 6.57929718e-01
3.18066895e-01 6.57604218e-01 -3.59630808e-02 -2.80116230e-01
-3.73695076e-01 -8.52904320e-02 -1.36358094e+00 1.24252558e+00
1.53702959e-01 1.18963802e+00 -1.31960988e-01 -9.76313233e-01
1.20105207e+00 4.09483343e-01 5.35673559e-01 -1.41578108e-01
4.52677965e-01 9.26796570e-02 9.40070897e-02 -1.02089059e+00
1.11200297e+00 2.42268413e-01 -3.44504297e-01 1.67271923e-02
2.27005687e-02 1.23375356e-01 6.38526857e-01 1.83166683e-01
8.58643115e-01 3.76736879e-01 2.78863102e-01 4.41059731e-02
2.45333925e-01 4.79663432e-01 2.50023246e-01 6.84239030e-01
-1.95051640e-01 7.55865335e-01 5.50077319e-01 -6.06669903e-01
-9.17848051e-01 -6.91779494e-01 2.53958791e-01 1.07536566e+00
1.25642687e-01 -4.19007927e-01 -1.02781546e+00 -3.32411349e-01
-2.92972028e-01 4.72415984e-01 -6.14440322e-01 -2.38927584e-02
-8.94865334e-01 -5.77612340e-01 7.30374634e-01 6.84128761e-01
4.96533900e-01 -1.55544269e+00 -1.38056087e+00 3.06311071e-01
-3.57966483e-01 -1.00560057e+00 -2.13263288e-01 1.07713915e-01
-5.86786866e-01 -1.28565502e+00 -9.33027923e-01 -7.69603252e-01
2.25380048e-01 6.39863908e-02 6.46960914e-01 -1.67272091e-01
-6.72691405e-01 2.83803552e-01 -8.39627087e-01 -8.27755868e-01
-3.68884444e-01 2.13295534e-01 -1.17311686e-01 5.18227071e-02
5.52487493e-01 -1.31693035e-01 -3.11773628e-01 2.57402062e-01
-3.34446043e-01 4.05231975e-02 2.02384979e-01 5.75970411e-01
1.13251261e-01 -1.64137393e-01 3.28252316e-01 -4.07008827e-01
7.07947373e-01 -1.64843887e-01 -3.58372241e-01 -2.29161367e-01
3.08182031e-01 -4.34404641e-01 3.34367156e-02 -4.18232143e-01
-8.10683608e-01 2.14196354e-01 -5.59440367e-02 -2.02811643e-01
-2.81694829e-01 4.35443580e-01 -1.30342826e-01 -1.11229688e-01
1.13039756e+00 3.68973352e-02 -9.55693945e-02 -2.21369922e-01
-1.32408395e-01 1.00713527e+00 8.29012215e-01 -4.42407519e-01
4.11924601e-01 2.07746103e-01 -4.04604599e-02 -1.49098670e+00
-4.61452901e-01 -7.24921584e-01 -7.34697342e-01 -1.38172519e+00
9.08294976e-01 -7.45958686e-01 -9.07784462e-01 7.48862207e-01
-1.17296839e+00 1.30927682e-01 -1.21039808e-01 7.22788870e-01
-4.01279092e-01 1.34652257e-01 -4.73142177e-01 -1.06504691e+00
-5.43958664e-01 -7.95809031e-01 8.96114409e-01 6.26752913e-01
-6.41728401e-01 -2.05619231e-01 -1.30580693e-01 7.52232969e-01
-7.49509335e-02 8.72474849e-01 1.83861163e-02 -5.67670107e-01
-3.00538689e-01 -8.06719601e-01 1.90561697e-01 1.48766950e-01
-1.08515859e-01 2.42239684e-01 -8.87692153e-01 5.78995515e-03
-5.70180714e-01 -1.19478561e-01 7.60450661e-01 6.68716609e-01
6.39038384e-01 7.93780684e-02 -1.04816213e-01 2.91795135e-01
8.23783934e-01 5.27882993e-01 7.22864091e-01 7.48528779e-01
6.49524391e-01 7.82926977e-01 1.19833255e+00 7.24605739e-01
1.16710179e-01 1.01649976e+00 3.91558826e-01 -2.20132902e-01
-2.50083566e-01 -2.10609674e-01 5.73062837e-01 2.80060738e-01
-1.25354624e+00 2.40779296e-02 -7.60410428e-01 5.73040009e-01
-1.74602735e+00 -1.32768226e+00 -6.44222081e-01 2.08088589e+00
4.31211025e-01 3.19871932e-01 8.62174451e-01 7.20377386e-01
9.74981666e-01 1.56446323e-01 -7.21124187e-02 -6.44003034e-01
-9.20850560e-02 2.76366353e-01 5.57113826e-01 4.34011929e-02
-1.41103971e+00 9.93581355e-01 5.56938887e+00 7.59767950e-01
-1.33946538e+00 -1.49640590e-01 5.86519875e-02 -4.99312758e-01
6.63128197e-01 -4.07676041e-01 -9.68864679e-01 4.36227679e-01
5.37567139e-01 7.49853626e-02 -1.51610434e-01 8.56270492e-01
6.04269683e-01 -4.57645446e-01 -6.92569256e-01 9.34609234e-01
4.02771503e-01 -1.40322864e+00 -3.29392165e-01 -2.60896474e-01
4.17924076e-01 -4.49544579e-01 -4.21705067e-01 4.66536619e-02
-2.40163609e-01 -6.41504228e-01 1.10925627e+00 4.90450472e-01
6.12262547e-01 -9.85160589e-01 9.15203273e-01 1.66889071e-01
-1.19602692e+00 6.20741695e-02 -8.47276822e-02 -4.42771435e-01
8.86093453e-02 -2.92446353e-02 -1.20337546e+00 3.75319868e-01
7.33588040e-01 4.62577611e-01 -3.75404507e-01 1.43716073e+00
-3.84613931e-01 7.93939054e-01 -3.30372036e-01 -5.86137056e-01
6.83465824e-02 1.40861630e-01 1.03368342e+00 1.54003024e+00
2.70058393e-01 1.28011033e-01 2.10683078e-01 1.72352955e-01
5.45238853e-01 3.92858863e-01 -3.84551764e-01 7.75710121e-02
8.08102414e-02 1.18215048e+00 -1.25862181e+00 -3.95284832e-01
-3.99099663e-03 7.50046670e-01 -3.47190470e-01 -2.51684543e-02
-9.00050402e-01 -8.72434914e-01 1.92796871e-01 6.01080596e-01
2.12546825e-01 3.27284308e-03 -2.10934386e-01 -9.54972327e-01
1.57450452e-01 -9.69427586e-01 5.15801907e-01 -8.52815390e-01
-4.84430373e-01 7.33382285e-01 2.92553127e-01 -1.49647319e+00
-2.93918431e-01 -6.56557679e-01 -9.36176300e-01 5.08433878e-01
-3.55267555e-01 -1.23719013e+00 -7.13718355e-01 3.72154027e-01
7.71399438e-01 -4.79028970e-01 4.50187385e-01 2.03757867e-01
-4.58036393e-01 4.18603867e-01 -3.64372075e-01 5.40320337e-01
3.87293965e-01 -1.02572215e+00 1.12570770e-01 9.34117794e-01
2.31730729e-01 -8.67768843e-03 1.12606800e+00 -8.41268837e-01
-9.82632041e-01 -7.10666358e-01 7.39857793e-01 -2.47464612e-01
2.59323508e-01 -4.07536805e-01 -2.81669617e-01 5.48368931e-01
-1.17894076e-01 -5.92146337e-01 5.00827134e-01 -1.42582074e-01
4.16671008e-01 -1.58177376e-01 -8.79721642e-01 3.59905601e-01
6.25686049e-01 -5.61461784e-02 -8.97647917e-01 2.28717923e-01
-2.00899258e-01 -9.10739720e-01 -6.49538338e-01 1.05259255e-01
9.24186409e-01 -8.39244187e-01 8.39901149e-01 -5.73598206e-01
7.07469821e-01 -2.38763854e-01 8.62482488e-02 -1.12171841e+00
2.50564426e-01 -5.75948775e-01 2.57288754e-01 9.95838642e-01
1.31027073e-01 7.92210847e-02 1.33497071e+00 2.48989865e-01
-1.73125938e-01 -2.18115017e-01 -8.60590816e-01 -3.56049538e-01
-6.77632928e-01 -4.53292340e-01 2.76023448e-02 6.26420081e-01
3.82833421e-01 1.27220780e-01 -7.78015316e-01 5.87394973e-03
6.19595945e-01 -4.25229929e-02 1.18525136e+00 -1.25538504e+00
-1.39474735e-01 -3.08105230e-01 -1.21692967e+00 -5.11967301e-01
-4.69806075e-01 -6.16848648e-01 -2.43369676e-02 -1.41971111e+00
-7.35816360e-02 2.13203244e-02 -3.77685577e-02 2.72013992e-01
-1.17276877e-01 4.95901257e-01 6.18059099e-01 -1.10006511e-01
-2.13161558e-01 -6.20934414e-03 1.11501694e+00 -4.75386642e-02
-5.89560807e-01 6.01996303e-01 -1.07674494e-01 8.14204693e-01
6.98680341e-01 -3.08186293e-01 8.29724297e-02 2.52835751e-01
-5.34959286e-02 4.09116566e-01 4.42896634e-01 -1.44121063e+00
1.01491034e-01 1.24422228e-02 4.71537620e-01 -9.30030107e-01
6.17494881e-01 -3.89396012e-01 1.28586233e-01 5.66450775e-01
-4.01740819e-01 -2.75674522e-01 3.09146643e-01 4.66451161e-02
-3.27159733e-01 -4.06370282e-01 4.30109590e-01 -1.69615105e-01
-1.03062975e+00 -2.97587931e-01 -8.52476120e-01 -1.62513539e-01
1.23427498e+00 -7.67695367e-01 1.44051202e-02 -4.67437267e-01
-1.20355523e+00 -5.99316321e-04 -1.83333695e-01 4.56339121e-01
5.99080384e-01 -1.07910359e+00 -9.68337595e-01 -3.57217304e-02
1.36702642e-01 -3.34008753e-01 2.68162638e-01 5.92279971e-01
-1.08907104e+00 6.55206442e-02 -9.40845907e-01 -5.02601445e-01
-1.99190247e+00 -1.61652416e-01 8.71217847e-02 7.12685511e-02
-7.69091487e-01 8.84209692e-01 -5.21632493e-01 2.23408937e-01
3.47730458e-01 -3.32568616e-01 -9.97535646e-01 7.01655447e-01
5.87458313e-01 8.37382734e-01 1.82825670e-01 -9.85814512e-01
-2.73718983e-01 4.83798951e-01 1.76957414e-01 -9.58737507e-02
1.35354078e+00 4.64806408e-01 2.65125304e-01 4.67152208e-01
3.83527696e-01 1.91590965e-01 -1.07800925e+00 3.42094183e-01
-1.65662616e-01 -5.19735277e-01 -2.14052033e-02 -6.34277344e-01
-1.02993536e+00 9.03425097e-01 6.14558995e-01 2.12992996e-01
9.80590403e-01 -1.98925123e-01 5.90094030e-01 1.83647722e-01
5.99795520e-01 -1.52679205e+00 -5.73304966e-02 2.35792786e-01
1.22614908e+00 -9.42124188e-01 1.09057985e-01 -3.55338544e-01
-1.11242259e+00 1.46543753e+00 7.49108493e-01 -5.06216168e-01
1.20797403e-01 2.27215484e-01 2.49785811e-01 -3.20236266e-01
-8.01905468e-02 -2.73637384e-01 4.67133731e-01 7.50533462e-01
3.95392358e-01 4.46746171e-01 -8.79017949e-01 7.84878552e-01
-7.44478583e-01 1.37489960e-01 1.01233101e+00 1.06425178e+00
-6.65879011e-01 -6.64540827e-01 -9.80702341e-01 6.93219483e-01
-5.41371703e-01 3.23438823e-01 -5.88136911e-01 1.12658405e+00
3.43326628e-01 9.89985108e-01 -8.03518668e-02 -6.82976842e-01
7.52892673e-01 -8.81124735e-02 5.99790692e-01 -4.53510463e-01
-1.16128385e+00 1.80967197e-01 6.25803113e-01 -3.50021392e-01
-3.96912187e-01 -7.93837190e-01 -1.13066328e+00 -9.83791426e-02
-2.86411017e-01 3.53642464e-01 7.42277563e-01 7.37863421e-01
-1.99063286e-01 6.06727779e-01 2.21979067e-01 -1.49140561e+00
3.18366028e-02 -1.20237935e+00 -8.52028191e-01 3.64685029e-01
-2.01702371e-01 -8.68555665e-01 -3.66785228e-02 3.40983182e-01] | [7.695512771606445, 0.153154194355011] |
4afe073b-d6d7-4c4a-bc81-c244fd30a497 | pai-gcn-permutable-anisotropic-graph | 2004.09995 | null | https://arxiv.org/abs/2004.09995v3 | https://arxiv.org/pdf/2004.09995v3.pdf | Learning Local Neighboring Structure for Robust 3D Shape Representation | Mesh is a powerful data structure for 3D shapes. Representation learning for 3D meshes is important in many computer vision and graphics applications. The recent success of convolutional neural networks (CNNs) for structured data (e.g., images) suggests the value of adapting insight from CNN for 3D shapes. However, 3D shape data are irregular since each node's neighbors are unordered. Various graph neural networks for 3D shapes have been developed with isotropic filters or predefined local coordinate systems to overcome the node inconsistency on graphs. However, isotropic filters or predefined local coordinate systems limit the representation power. In this paper, we propose a local structure-aware anisotropic convolutional operation (LSA-Conv) that learns adaptive weighting matrices for each node according to the local neighboring structure and performs shared anisotropic filters. In fact, the learnable weighting matrix is similar to the attention matrix in the random synthesizer -- a new Transformer model for natural language processing (NLP). Comprehensive experiments demonstrate that our model produces significant improvement in 3D shape reconstruction compared to state-of-the-art methods. | ['Juyong Zhang', 'Guangtao Zhai', 'Yiyan Yang', 'Junchi Yan', 'Xiaokang Yang', 'Zhongpai Gao'] | 2020-04-21 | null | null | null | null | ['3d-shape-representation'] | ['computer-vision'] | [-1.11708596e-01 2.84725577e-01 3.91995460e-02 -3.91596615e-01
5.14072459e-03 -2.83166468e-01 3.61254245e-01 -4.29585874e-02
1.52811348e-01 1.67808607e-02 5.98169804e-01 -2.56239414e-01
-2.07201634e-02 -1.32005751e+00 -8.73710155e-01 -5.47448456e-01
-3.12064635e-03 5.78756034e-01 1.83437526e-01 -3.01097423e-01
2.74956256e-01 1.13951135e+00 -1.04770195e+00 4.97956961e-01
7.28154361e-01 8.97266150e-01 3.72045696e-01 5.24461210e-01
-8.03643882e-01 4.50916111e-01 -2.38934413e-01 -2.25221261e-01
3.42790633e-01 -2.98963878e-02 -7.07481027e-01 1.35902718e-01
7.80931354e-01 -2.87040979e-01 -6.17243648e-01 1.21332538e+00
6.52495801e-01 -4.12251763e-02 9.08675969e-01 -9.62954521e-01
-1.60712409e+00 5.63778400e-01 -6.06476545e-01 -3.90642434e-02
2.06980109e-01 3.17913340e-03 1.12631512e+00 -1.22996700e+00
8.20636213e-01 1.81357551e+00 8.22897494e-01 4.57738340e-01
-1.02796185e+00 -3.79863828e-01 4.48958009e-01 -9.46974382e-02
-1.38768327e+00 -8.12932253e-02 1.35488999e+00 -3.56586546e-01
1.24938083e+00 2.68521216e-02 8.61009061e-01 9.79786932e-01
3.12238544e-01 8.23350787e-01 4.13961112e-01 -1.04233116e-01
-5.69147393e-02 -4.21761036e-01 -2.85426408e-01 1.12255287e+00
2.59406656e-01 -2.34396473e-01 -2.56293148e-01 -3.02517358e-02
1.65152967e+00 -4.26702760e-02 -1.11826733e-01 -7.27458894e-01
-1.32919395e+00 5.78698397e-01 1.03500175e+00 4.17478412e-01
-4.37758356e-01 4.38334882e-01 2.18877435e-01 4.23872948e-01
8.28255653e-01 4.41397458e-01 -3.47738802e-01 4.71977234e-01
-4.10973758e-01 1.87378585e-01 6.10942245e-01 1.32921433e+00
6.75552905e-01 7.18136132e-01 -8.67975652e-02 9.41439927e-01
5.20601928e-01 5.74935794e-01 -3.11564421e-03 -8.35576653e-01
4.75047261e-01 1.15536523e+00 -3.94088686e-01 -1.72400367e+00
-6.58653796e-01 -6.34929359e-01 -1.50950193e+00 2.05152884e-01
8.60588625e-02 3.57194066e-01 -1.01782417e+00 1.47769201e+00
3.45607728e-01 3.32667857e-01 -2.42067784e-01 1.00961816e+00
1.40355015e+00 7.04826415e-01 -2.45387062e-01 4.33057368e-01
1.05479550e+00 -7.38424897e-01 -5.59959412e-01 8.17311108e-02
4.29110408e-01 -7.66060889e-01 1.13005042e+00 -3.68390456e-02
-1.17383289e+00 -7.80398786e-01 -9.53601480e-01 -4.66733158e-01
-3.36954892e-01 -1.02011837e-01 5.62846065e-01 3.24824631e-01
-1.25703406e+00 6.20616794e-01 -6.84440792e-01 -3.03404480e-01
6.76451445e-01 2.71935433e-01 -3.35814118e-01 -8.53827521e-02
-8.23391080e-01 5.38490772e-01 -8.16792175e-02 7.49721751e-02
-5.47215939e-01 -8.34886909e-01 -1.08429837e+00 7.51395002e-02
5.89695983e-02 -1.09229684e+00 7.09415019e-01 -5.69504142e-01
-1.52023935e+00 9.85008240e-01 -6.57612234e-02 -8.36257190e-02
1.14443883e-01 1.19028673e-01 -2.04322517e-01 2.51149107e-02
-5.95063437e-03 6.89958513e-01 1.12971354e+00 -1.38073051e+00
1.03191324e-02 -3.81184787e-01 2.55664140e-01 2.36512080e-01
-1.85642272e-01 -3.49282414e-01 -4.81544644e-01 -1.15655410e+00
7.85662532e-01 -5.60545444e-01 -4.47854608e-01 4.42559600e-01
-3.08317572e-01 -5.08943856e-01 9.10620034e-01 -3.47594291e-01
8.98011267e-01 -2.17593002e+00 3.93477559e-01 4.99903083e-01
6.31941915e-01 -7.99680427e-02 -5.79582036e-01 1.94547847e-01
-6.64186850e-02 3.95098835e-01 -1.77872226e-01 -1.87040031e-01
-1.71652791e-04 3.34075332e-01 -1.92160293e-01 2.72998542e-01
5.31660974e-01 1.17980623e+00 -9.33908403e-01 -2.66692996e-01
3.27598363e-01 8.01351488e-01 -9.00324166e-01 1.46724179e-01
-2.42717877e-01 3.58267993e-01 -6.95072472e-01 6.99932396e-01
1.02260125e+00 -4.91291136e-01 -1.51479676e-01 -8.45087469e-01
8.33118334e-02 2.48526365e-01 -1.31694508e+00 2.08628798e+00
-3.84717524e-01 1.76113054e-01 2.52714574e-01 -1.01016605e+00
1.35126114e+00 1.67911708e-01 5.02367079e-01 -5.26369333e-01
1.36728778e-01 2.72783995e-01 -2.87967250e-02 -3.78629059e-01
4.16872352e-01 2.30024531e-01 1.85481593e-01 3.41532171e-01
1.61974624e-01 -7.39089429e-01 -3.06006640e-01 2.05824763e-01
9.88169074e-01 -1.79502685e-02 3.24083716e-02 -5.88868856e-01
4.84709799e-01 -4.27597642e-01 4.64971125e-01 5.50119698e-01
2.22573802e-01 1.07309735e+00 2.62666225e-01 -1.05465281e+00
-1.19845152e+00 -1.11905921e+00 1.24294519e-01 5.77559710e-01
2.05007464e-01 -4.39732254e-01 -4.66552913e-01 -4.85668123e-01
2.25217178e-01 3.23621213e-01 -6.59990072e-01 -1.71858266e-01
-8.36115658e-01 -1.35239765e-01 3.49223584e-01 7.13789642e-01
6.79338872e-01 -1.15690684e+00 2.03979481e-02 2.35558972e-01
2.94176131e-01 -1.02150679e+00 -8.76676559e-01 -2.29923695e-01
-1.14992571e+00 -8.75015259e-01 -7.59943247e-01 -1.18935132e+00
1.18491137e+00 5.20751417e-01 1.35681403e+00 4.56823856e-01
2.55290885e-02 5.61207652e-01 -2.59608090e-01 -1.94080174e-01
-2.10679412e-01 3.03540617e-01 -6.79916963e-02 8.97528157e-02
7.29130134e-02 -9.52574432e-01 -5.75955033e-01 1.61254257e-01
-8.58965874e-01 2.22703919e-01 5.11257231e-01 6.96462870e-01
9.41847324e-01 -1.69135049e-01 3.22509825e-01 -1.07850111e+00
8.48238945e-01 -2.34170154e-01 -2.95647830e-01 1.60695881e-01
-8.42315927e-02 3.66956562e-01 5.95894039e-01 -4.01578724e-01
-8.01991820e-01 -5.49248978e-02 -4.52052176e-01 -6.80763423e-01
-1.51918426e-01 5.76060951e-01 -3.91120166e-01 -3.64640653e-01
5.36263347e-01 2.23623607e-02 7.69491261e-03 -6.61351442e-01
4.28384870e-01 1.39717177e-01 2.64900416e-01 -8.46544504e-01
1.00245154e+00 3.78662080e-01 3.10940027e-01 -1.12557185e+00
-5.80504954e-01 4.28515822e-02 -7.60402977e-01 -1.13164462e-01
8.65313470e-01 -8.58757794e-01 -4.82263207e-01 5.88321328e-01
-1.50396693e+00 -2.68845409e-01 -4.22345549e-01 1.77845672e-01
-5.20745158e-01 2.57267386e-01 -7.35787034e-01 -2.36770183e-01
-4.13016438e-01 -1.18274462e+00 1.24331808e+00 -2.63896603e-02
-1.68184027e-01 -1.14205313e+00 -2.90447235e-01 -1.12409182e-01
7.45320499e-01 2.39370361e-01 1.41672170e+00 -1.42327994e-01
-8.25592875e-01 3.36262845e-02 -5.40827930e-01 9.15791988e-02
2.75040388e-01 3.15457582e-02 -5.93213439e-01 -1.23613596e-01
-1.85367554e-01 -2.97386926e-02 6.88778877e-01 6.25216246e-01
1.69135952e+00 -2.46995464e-01 -5.88516593e-02 9.56462681e-01
1.04813397e+00 -8.51050690e-02 4.57094491e-01 -1.82823077e-01
1.25547588e+00 3.20105463e-01 -2.65618622e-01 2.16113165e-01
5.47875345e-01 3.14828902e-01 6.90020800e-01 -2.38152340e-01
-5.89532912e-01 -5.56202471e-01 -7.02933520e-02 1.42081070e+00
-4.40154821e-01 -9.68911946e-02 -7.62778878e-01 4.40131545e-01
-1.56346691e+00 -5.75595200e-01 -2.43059903e-01 1.76694477e+00
4.23073649e-01 1.74232781e-01 -3.79186481e-01 -1.29843399e-01
6.40860200e-01 6.83382213e-01 -6.00063860e-01 -4.63298708e-01
-4.58147794e-01 2.87830651e-01 2.76001483e-01 5.87038219e-01
-7.93793201e-01 1.01507235e+00 5.88286543e+00 7.06122935e-01
-1.09744787e+00 -1.92624569e-01 5.03750026e-01 4.33807403e-01
-1.06027341e+00 -4.19678152e-01 -3.19766015e-01 -1.07719406e-01
9.24299806e-02 5.79871275e-02 4.13994372e-01 7.36968577e-01
-3.28528620e-02 7.54263222e-01 -9.35831547e-01 1.20591044e+00
1.16461188e-01 -1.77651417e+00 8.31092954e-01 -1.86957009e-02
9.13039446e-01 8.71763453e-02 1.81868255e-01 -6.13128766e-02
2.79107481e-01 -1.05034959e+00 8.50175917e-01 5.54339349e-01
9.51971591e-01 -6.58720672e-01 4.55585182e-01 9.61347148e-02
-1.60829163e+00 4.70550060e-01 -9.11256790e-01 7.04647154e-02
2.73311082e-02 7.51451075e-01 -4.96096700e-01 4.92607713e-01
6.54085219e-01 1.19976652e+00 -2.87308544e-01 8.14443648e-01
-2.48113647e-01 2.71949381e-01 -4.42600697e-01 -1.08692154e-01
3.98422509e-01 -4.16361481e-01 8.30571890e-01 9.01473641e-01
6.17091179e-01 2.58530021e-01 1.74663737e-01 1.32058275e+00
-5.65837979e-01 1.97811007e-01 -1.15940928e+00 -1.32762119e-01
3.88469219e-01 9.93425131e-01 -8.16277921e-01 -1.16026372e-01
-5.03714919e-01 9.11738217e-01 5.11423409e-01 5.12381554e-01
-3.25604826e-01 -2.66510606e-01 6.85048580e-01 3.53110880e-01
2.68038422e-01 -8.06952178e-01 -5.59057474e-01 -1.04082692e+00
-1.29751101e-01 -6.28698528e-01 6.17685989e-02 -8.71687114e-01
-1.65570438e+00 6.39632642e-01 -3.42854470e-01 -1.26712060e+00
4.57816362e-01 -9.27908361e-01 -6.51210070e-01 7.20136464e-01
-1.28735137e+00 -1.35414839e+00 -2.84216970e-01 6.64664805e-01
6.04919255e-01 -3.14688206e-01 8.12031448e-01 1.76971138e-01
3.20982275e-04 4.80632871e-01 -3.98465931e-01 4.54413414e-01
2.25161627e-01 -1.17233527e+00 1.03027332e+00 4.43621188e-01
4.04420078e-01 6.54438436e-01 1.59735814e-01 -7.67282724e-01
-1.91445398e+00 -1.15134907e+00 7.30815113e-01 -1.57762721e-01
4.53931302e-01 -4.41383004e-01 -1.11005819e+00 5.40284395e-01
1.93382502e-01 4.45323765e-01 2.43948728e-01 -1.39085473e-02
-6.73310280e-01 -4.24115285e-02 -1.08400929e+00 8.43199611e-01
1.79934347e+00 -6.00529730e-01 -3.89269978e-01 1.56772017e-01
1.01051354e+00 -8.16784263e-01 -9.64626729e-01 5.85737228e-01
4.01868403e-01 -6.70503974e-01 1.25165784e+00 -7.38822460e-01
4.48526412e-01 -4.08379644e-01 -3.73146445e-01 -1.57269859e+00
-8.36476624e-01 -4.50876534e-01 -1.50071472e-01 7.33824432e-01
4.36135948e-01 -4.88299161e-01 8.18415761e-01 3.64018172e-01
-5.92237473e-01 -9.01223719e-01 -8.61161113e-01 -3.78534168e-01
1.09782286e-01 -5.51924646e-01 1.05700314e+00 1.05452299e+00
-6.53083324e-01 4.13962483e-01 -2.01452747e-01 3.51816982e-01
5.92921078e-01 3.78418833e-01 7.23831356e-01 -1.32497251e+00
1.44210845e-01 -7.22731650e-01 -5.89890897e-01 -1.63270986e+00
1.99921444e-01 -1.36540020e+00 -4.65321392e-01 -1.86367154e+00
-3.09384257e-01 -6.28779233e-01 -1.44234225e-01 3.28020185e-01
1.40662014e-01 1.56867236e-01 1.01607226e-01 -2.01277748e-01
-3.01021576e-01 8.73089492e-01 2.19287586e+00 -7.07162142e-01
-1.35716975e-01 -1.73470244e-01 -5.92860103e-01 7.33562052e-01
7.52695739e-01 -1.93887696e-01 -4.87094969e-01 -1.27733243e+00
6.75536811e-01 -3.50190997e-01 1.77062646e-01 -6.78623259e-01
3.47975582e-01 -6.90018088e-02 4.66398269e-01 -8.33604872e-01
1.44578472e-01 -1.08692956e+00 1.05109826e-01 1.51819661e-01
-2.15477675e-01 4.12067711e-01 5.63244708e-02 6.10708058e-01
-1.55403480e-01 -4.56548762e-03 5.52157640e-01 -5.93211174e-01
-5.09511352e-01 9.71575379e-01 -9.07929018e-02 2.54723877e-02
3.44800681e-01 -3.32817227e-01 -6.41170740e-02 -4.26482946e-01
-6.96618438e-01 -1.40272640e-02 4.63072836e-01 6.35688424e-01
1.26614761e+00 -2.02532339e+00 -8.41371238e-01 6.97971761e-01
-1.32879047e-02 7.07154155e-01 2.74571389e-01 3.93631816e-01
-8.07322145e-01 2.32140161e-02 -3.11899751e-01 -6.75458252e-01
-7.90319204e-01 2.99919873e-01 4.32521850e-01 -3.72344851e-02
-9.14733171e-01 1.08929074e+00 2.75406539e-01 -9.77257133e-01
2.14136153e-01 -9.10501957e-01 -5.98550290e-02 -3.00616562e-01
2.32983634e-01 7.44803324e-02 7.65416324e-02 -6.42982006e-01
-1.45406976e-01 1.12189198e+00 9.13467556e-02 5.80623209e-01
1.79731870e+00 9.63153616e-02 -5.02935529e-01 3.50243658e-01
1.20495868e+00 1.57247819e-02 -9.85691667e-01 -5.71236134e-01
-2.68059641e-01 -5.50800383e-01 8.08961019e-02 -7.57453442e-02
-1.67145014e+00 1.10066772e+00 1.42877027e-01 2.73501486e-01
7.36542284e-01 2.67494731e-02 7.26968706e-01 5.80033123e-01
4.27923679e-01 -7.98079669e-01 2.78847069e-01 1.02786982e+00
1.63255656e+00 -1.00858974e+00 -1.14739556e-02 -7.42826164e-01
-2.30124116e-01 1.35514486e+00 6.78706884e-01 -6.93157911e-01
1.23689449e+00 3.49914372e-01 -5.38842604e-02 -6.37636065e-01
-3.58050555e-01 2.18545701e-02 7.01962233e-01 8.13204288e-01
5.11200786e-01 1.37327969e-01 1.74069270e-01 4.29849118e-01
-1.74880937e-01 -5.67617118e-01 1.02629669e-01 5.16266525e-01
-2.39273310e-01 -9.59070861e-01 -1.30183056e-01 5.98734260e-01
7.57047459e-02 -1.63800102e-02 -4.91682231e-01 5.47167063e-01
-3.99577357e-02 4.44303721e-01 3.08268517e-01 -4.45882499e-01
5.39431512e-01 -1.34589404e-01 6.87177479e-01 -6.67895019e-01
-3.58161628e-01 7.46333525e-02 -1.77123129e-01 -6.15420222e-01
-6.50618136e-01 -2.67622858e-01 -1.37291932e+00 -3.95904779e-01
-5.49290590e-02 -3.46523106e-01 2.75833756e-01 4.81447250e-01
6.88366294e-01 6.82753265e-01 5.95870197e-01 -1.19027770e+00
-7.99302980e-02 -8.10050249e-01 -6.82678103e-01 6.16634429e-01
1.69482067e-01 -7.65515089e-01 -4.61521447e-02 -1.67875916e-01] | [8.197859764099121, -3.7360832691192627] |
3d54d9e4-3502-4d0d-b98e-839c7548e924 | hierarchical-semantic-tree-concept-whitening | 2307.04343 | null | https://arxiv.org/abs/2307.04343v1 | https://arxiv.org/pdf/2307.04343v1.pdf | Hierarchical Semantic Tree Concept Whitening for Interpretable Image Classification | With the popularity of deep neural networks (DNNs), model interpretability is becoming a critical concern. Many approaches have been developed to tackle the problem through post-hoc analysis, such as explaining how predictions are made or understanding the meaning of neurons in middle layers. Nevertheless, these methods can only discover the patterns or rules that naturally exist in models. In this work, rather than relying on post-hoc schemes, we proactively instill knowledge to alter the representation of human-understandable concepts in hidden layers. Specifically, we use a hierarchical tree of semantic concepts to store the knowledge, which is leveraged to regularize the representations of image data instances while training deep models. The axes of the latent space are aligned with the semantic concepts, where the hierarchical relations between concepts are also preserved. Experiments on real-world image datasets show that our method improves model interpretability, showing better disentanglement of semantic concepts, without negatively affecting model classification performance. | ['Dajiang Zhu', 'Tianming Liu', 'Ninghao Liu', 'Changying Li', 'Yanjun Lyu', 'Xiaowei Yu', 'David Liu', 'Zhengliang Liu', 'Zihao Wu', 'Lin Zhao', 'Lu Zhang', 'Haixing Dai'] | 2023-07-10 | null | null | null | null | ['image-classification', 'disentanglement'] | ['computer-vision', 'methodology'] | [ 2.13894457e-01 5.37796199e-01 -1.89075470e-01 -6.42731428e-01
4.53396082e-01 -4.34101224e-01 6.30709827e-01 3.21218103e-01
-1.30799532e-01 3.23348433e-01 4.31465119e-01 -4.00272012e-01
-1.33679315e-01 -7.37695932e-01 -7.35613167e-01 -5.21914542e-01
1.01235241e-01 2.05096453e-01 1.42033547e-02 7.84875825e-02
2.77545154e-01 1.08465083e-01 -1.65855658e+00 7.45022118e-01
7.58490324e-01 1.06727290e+00 2.91604191e-01 1.04801513e-01
-3.73746216e-01 1.09056580e+00 -4.26826954e-01 -4.01682347e-01
-1.11349963e-01 -3.35442483e-01 -9.02910531e-01 1.84166715e-01
2.03360036e-01 -1.29950076e-01 -8.96043852e-02 1.07057941e+00
-4.30249780e-01 -1.67033285e-01 5.90846419e-01 -1.36714697e+00
-9.49298203e-01 8.37790668e-01 -2.63147205e-01 -2.41153724e-02
-5.61334901e-02 -4.85083386e-02 1.40808690e+00 -7.15629637e-01
3.02783251e-01 1.47911167e+00 4.69941139e-01 6.95539713e-01
-1.53350770e+00 -7.96612620e-01 6.99369371e-01 5.81056237e-01
-1.23608649e+00 -2.33913288e-01 8.46123755e-01 -7.19422102e-01
7.47277558e-01 2.68079907e-01 9.30038273e-01 1.14023352e+00
1.81918070e-01 6.68315053e-01 1.08841670e+00 -4.79440361e-01
4.51703429e-01 3.41100246e-01 3.93170357e-01 6.60816491e-01
4.97333795e-01 -5.49501814e-02 -7.56709933e-01 -8.66592154e-02
7.18041122e-01 3.31351876e-01 -1.33710429e-01 -6.02249444e-01
-1.00763273e+00 1.00485516e+00 6.59405887e-01 3.68566781e-01
-4.40298796e-01 8.00099894e-02 6.79620653e-02 5.70305362e-02
3.88155162e-01 6.64298952e-01 -7.34434009e-01 3.35567266e-01
-6.08590722e-01 4.74460796e-02 5.58906078e-01 7.63072193e-01
1.01620519e+00 -1.60925299e-01 7.70430192e-02 5.92511177e-01
6.84571803e-01 -2.92769764e-02 7.69202888e-01 -8.11666608e-01
1.78659707e-01 1.16339350e+00 -3.42743427e-01 -1.32802510e+00
-2.38355815e-01 -5.22797585e-01 -9.09059823e-01 -1.30495131e-02
8.77245795e-03 2.91866213e-01 -1.14516246e+00 1.88829374e+00
-1.03155002e-01 1.33482471e-01 2.07406119e-01 6.79135442e-01
4.15827572e-01 4.51213062e-01 5.01280963e-01 1.87199190e-01
1.51555514e+00 -8.17709565e-01 -7.49569476e-01 -6.50644541e-01
7.42317855e-01 -2.39252269e-01 1.14894927e+00 2.48636514e-01
-5.44937193e-01 -6.20530605e-01 -1.04459262e+00 -1.71957627e-01
-5.64331353e-01 -9.59685221e-02 8.05139184e-01 2.85529613e-01
-1.00473797e+00 5.61038315e-01 -9.20111954e-01 -3.61230314e-01
7.55682886e-01 3.57688576e-01 -2.95876741e-01 5.25172241e-02
-1.11112523e+00 9.70371664e-01 9.25058186e-01 1.86070114e-01
-7.49599040e-01 -6.84117317e-01 -8.09771657e-01 4.85103846e-01
2.89284348e-01 -7.78749764e-01 9.16342676e-01 -1.36383712e+00
-1.11543059e+00 8.79330635e-01 -3.25512022e-01 -4.90417600e-01
-6.87066466e-02 -3.26572686e-01 -2.47673839e-01 -9.37606245e-02
-6.84645921e-02 1.08539271e+00 8.27658117e-01 -1.70183349e+00
-5.87365091e-01 -3.98898542e-01 3.42261642e-01 -1.19577805e-02
-7.66831934e-01 -3.00694436e-01 -9.08559933e-02 -4.64187145e-01
7.23601222e-01 -1.05021524e+00 -2.73652434e-01 1.49618328e-01
-7.03579724e-01 -1.49520427e-01 8.84996235e-01 -4.78759289e-01
1.30856502e+00 -2.13682866e+00 1.62608549e-01 3.64568502e-01
7.72698462e-01 2.68518943e-02 -3.49017084e-02 1.64396882e-01
-2.71749973e-01 5.91403186e-01 -2.39625975e-01 -2.57999390e-01
8.39721709e-02 7.77127385e-01 -6.34456754e-01 -5.13553806e-02
3.74797314e-01 8.99762154e-01 -7.36077070e-01 -2.67783821e-01
7.90498182e-02 3.17750603e-01 -7.68549263e-01 1.63713887e-01
-4.51092839e-01 3.44594061e-01 -6.96310222e-01 2.25900114e-01
3.63398880e-01 -7.35580683e-01 4.95611042e-01 -2.38801628e-01
2.70014465e-01 6.23365760e-01 -8.46318901e-01 1.37406433e+00
-3.97564679e-01 7.40482628e-01 -4.96995836e-01 -1.07564354e+00
7.96007991e-01 8.16064849e-02 -1.46706598e-02 -5.33734262e-01
-1.77945733e-01 -1.41820461e-01 3.21180761e-01 -3.98418933e-01
3.26479852e-01 -2.15590402e-01 2.48053417e-01 3.55319947e-01
-4.57797237e-02 3.87867868e-01 -2.30976731e-01 1.82868153e-01
6.87117577e-01 -1.24847125e-02 4.43267107e-01 -5.17067552e-01
3.83496165e-01 1.07619790e-02 5.89607894e-01 5.22586048e-01
1.57045096e-01 3.50950897e-01 7.87974536e-01 -8.24100912e-01
-7.84722090e-01 -1.01514888e+00 -3.00739128e-02 9.96029794e-01
6.43438920e-02 -8.58911395e-01 -7.97069252e-01 -6.28776729e-01
-9.71578211e-02 1.02223420e+00 -1.00644803e+00 -6.29794300e-01
-2.87807852e-01 -5.94241381e-01 2.00276151e-01 6.57282829e-01
4.48617041e-01 -9.06142294e-01 -8.67579520e-01 3.59006897e-02
-1.80816263e-01 -1.06415498e+00 2.99065113e-01 4.37337101e-01
-1.23809230e+00 -9.90045607e-01 1.24427661e-01 -5.36311269e-01
1.21430612e+00 2.37060666e-01 1.21435595e+00 6.53524339e-01
7.29239732e-02 2.24322248e-02 -4.02618676e-01 -4.50059474e-01
-3.47831964e-01 1.36004806e-01 -1.17137529e-01 2.26880774e-01
6.88762009e-01 -6.82084680e-01 -5.74374318e-01 2.70852387e-01
-1.15785313e+00 6.96637154e-01 6.57424927e-01 7.42180288e-01
3.87882680e-01 1.80833802e-01 3.11152756e-01 -1.09741592e+00
4.07340914e-01 -5.47680676e-01 -2.61080623e-01 3.54074806e-01
-9.20881987e-01 5.80998421e-01 6.39633894e-01 -4.62227851e-01
-9.27507281e-01 5.17549329e-02 3.21776897e-01 -4.00913715e-01
-3.39371741e-01 7.34005570e-01 -3.73330057e-01 4.54177082e-01
5.03927171e-01 2.55181849e-01 -2.03592286e-01 -5.53911924e-01
4.25524563e-01 3.42602521e-01 1.49615154e-01 -5.65612972e-01
8.47387493e-01 5.48998833e-01 -1.97803915e-01 -5.63995898e-01
-1.38725734e+00 -2.89657922e-03 -8.44733715e-01 1.49777442e-01
1.04309189e+00 -8.41464460e-01 -4.23537761e-01 1.10805102e-01
-1.28926730e+00 -1.06178068e-01 -5.89493513e-02 2.99594015e-01
-2.83115476e-01 8.79905000e-03 -3.86459589e-01 -4.75175411e-01
2.53926396e-01 -1.07144678e+00 7.39779890e-01 1.40283823e-01
-7.47727931e-01 -1.34911227e+00 -3.51558805e-01 3.84487420e-01
3.14888537e-01 9.45033655e-02 1.84661734e+00 -8.84405077e-01
-7.61494756e-01 8.54387581e-02 -3.12599391e-01 2.19242111e-01
2.80280709e-01 -2.28066549e-01 -1.35697448e+00 -8.24288726e-02
7.60202929e-02 -1.92409500e-01 8.57610643e-01 1.06192648e-01
1.81439459e+00 -8.51485550e-01 -4.85142738e-01 4.31523025e-01
1.07493734e+00 1.03381403e-01 6.23421967e-01 5.64336658e-01
8.60494375e-01 9.97269452e-01 1.08573578e-01 1.70428246e-01
5.13368487e-01 4.46108669e-01 6.35197580e-01 -2.03768000e-01
2.92464018e-01 -6.29339218e-01 1.20101571e-01 6.34941041e-01
4.07925621e-02 -1.38756707e-01 -1.02802730e+00 2.74905294e-01
-2.01674318e+00 -6.91195250e-01 -5.39094396e-02 1.88119459e+00
8.99858534e-01 2.28674129e-01 -5.68894386e-01 4.50797416e-02
5.17881930e-01 2.05114678e-01 -6.42822742e-01 -2.43798524e-01
-7.55397417e-03 -1.61104426e-01 -1.32643618e-03 4.37644869e-01
-7.82732487e-01 1.15070057e+00 6.04688835e+00 3.07117701e-01
-1.10533881e+00 -1.47270322e-01 8.41719687e-01 1.07588895e-01
-5.94419003e-01 4.39010650e-01 -6.19636238e-01 2.07662851e-01
7.84371972e-01 -1.07366309e-01 1.43798873e-01 8.61061037e-01
1.60638168e-01 6.71094880e-02 -1.62120199e+00 7.10022688e-01
-5.95429505e-04 -1.36866558e+00 8.00990880e-01 2.68834174e-01
5.89967966e-01 -5.19446015e-01 1.28932714e-01 2.48720035e-01
3.55993062e-01 -1.32985365e+00 8.58442008e-01 6.54218435e-01
1.40916660e-01 -4.69200730e-01 5.50785184e-01 4.05930281e-01
-7.69413769e-01 -9.39297453e-02 -5.33020556e-01 -4.33282197e-01
-3.39780658e-01 4.99417245e-01 -8.66978109e-01 1.47965461e-01
7.53568351e-01 9.69450831e-01 -9.65016782e-01 4.94798571e-01
-6.98486626e-01 7.20594883e-01 7.86485299e-02 2.43866071e-01
2.16312155e-01 6.98373318e-02 -2.51215268e-02 8.50430846e-01
9.69597548e-02 1.01028502e-01 3.79083566e-02 1.32070386e+00
-7.54865110e-02 -2.79334128e-01 -5.20666242e-01 -1.91612080e-01
3.08361679e-01 1.05579734e+00 -7.58113921e-01 -5.14374554e-01
-2.44589299e-01 8.41930926e-01 4.06073898e-01 5.56049883e-01
-6.73716068e-01 2.21897438e-01 1.04848909e+00 2.06597835e-01
1.33806020e-01 -3.33035052e-01 -7.10245669e-01 -1.18227577e+00
1.16062285e-02 -7.85313308e-01 2.62840062e-01 -9.42419052e-01
-1.09351122e+00 6.29749000e-01 1.75362110e-01 -9.93451953e-01
-1.98401093e-01 -8.61801684e-01 -5.74536741e-01 7.34080911e-01
-1.47228825e+00 -1.09545743e+00 -2.08703607e-01 4.15922582e-01
4.25352365e-01 -1.26912985e-02 1.08679104e+00 -2.72909254e-01
-4.72291797e-01 2.83529162e-01 -8.87058005e-02 1.15971513e-01
1.20332718e-01 -1.13064837e+00 2.50418246e-01 7.15975523e-01
4.75552052e-01 1.26958323e+00 7.86811292e-01 -4.63072956e-01
-8.84753346e-01 -9.57629621e-01 1.07934821e+00 -5.14249682e-01
7.63820589e-01 -7.02231944e-01 -1.41349685e+00 9.28379476e-01
8.90540257e-02 -1.97987422e-01 1.20024908e+00 3.86580229e-01
-7.40832329e-01 -1.23174533e-01 -6.28568411e-01 7.60126948e-01
1.05526769e+00 -7.05654681e-01 -9.16923463e-01 9.44228396e-02
8.61160100e-01 9.10681859e-02 -4.75240737e-01 2.80792326e-01
5.57094455e-01 -8.85338485e-01 8.54843438e-01 -1.10804617e+00
8.70435357e-01 -3.40926528e-01 -1.25927761e-01 -1.43867755e+00
-5.55233121e-01 1.12725899e-01 -1.90767482e-01 1.11583233e+00
6.47695303e-01 -5.65219104e-01 9.97874618e-01 1.10421920e+00
4.51935604e-02 -7.00285137e-01 -6.19257987e-01 -4.13357615e-01
-1.43658176e-01 -3.29008788e-01 6.52287602e-01 1.18287575e+00
-1.04784265e-01 4.98731703e-01 -1.19237676e-01 4.25607711e-01
3.83908391e-01 1.36891529e-01 4.09215987e-01 -1.66352844e+00
-2.32011572e-01 -3.67910415e-01 -5.83407044e-01 -1.00101793e+00
4.69627887e-01 -9.36456025e-01 -1.67089283e-01 -1.57758093e+00
4.17277068e-01 -3.87076855e-01 -7.11483538e-01 1.13019097e+00
-2.42623672e-01 -8.16221610e-02 3.51852506e-01 5.45514107e-01
-6.15795374e-01 6.30606711e-01 1.05348587e+00 -1.56693295e-01
1.80750310e-01 -3.13389599e-01 -1.15673888e+00 1.18435407e+00
8.73033643e-01 -7.51984954e-01 -8.08192313e-01 -8.97743762e-01
4.21545446e-01 -6.11868739e-01 8.00928473e-01 -8.95042837e-01
1.81011930e-01 -4.26439345e-01 6.89301550e-01 -2.32378021e-01
3.11835527e-01 -1.16212654e+00 1.95003062e-01 5.92973053e-01
-7.82384992e-01 1.70339569e-01 1.43116727e-01 5.90598643e-01
-4.33999062e-01 -1.75929457e-01 5.75400829e-01 -1.35840282e-01
-9.02418852e-01 2.58819107e-02 -4.51603085e-01 -2.52398014e-01
8.44223201e-01 -2.04826891e-01 -2.70593733e-01 -2.75932252e-01
-7.23196924e-01 2.07201108e-01 3.60569984e-01 8.02818954e-01
7.31551051e-01 -1.14188707e+00 -1.92426056e-01 3.65054637e-01
4.60907876e-01 1.35286644e-01 2.54936367e-02 2.98247963e-01
-1.33000657e-01 5.86903691e-01 -2.74553090e-01 -6.76956415e-01
-1.12781394e+00 4.23911661e-01 3.59286785e-01 -1.71233818e-01
-4.80225861e-01 6.47866488e-01 8.66133213e-01 -3.08804095e-01
1.23634964e-01 -7.39128530e-01 -3.81929934e-01 7.74869770e-02
4.36768532e-01 -2.77110010e-01 -1.90783516e-01 -4.42101300e-01
-2.22360358e-01 2.80495286e-01 -5.04582763e-01 2.40715459e-01
1.41963685e+00 -2.02220410e-01 -3.10751230e-01 6.95067286e-01
1.00981867e+00 -5.33326507e-01 -1.36506474e+00 -3.05751771e-01
6.93191588e-01 -3.91848683e-01 1.26555607e-01 -9.07494605e-01
-1.02556968e+00 1.17963147e+00 4.09538597e-01 2.69442409e-01
1.05698681e+00 2.18122527e-01 3.35625529e-01 6.12888157e-01
5.21825403e-02 -6.64939821e-01 2.65597165e-01 5.85516870e-01
8.11582208e-01 -1.18238008e+00 -2.47810990e-01 -4.92626160e-01
-5.80763221e-01 1.30846381e+00 9.02397096e-01 8.95976648e-02
5.13044477e-01 -1.01732276e-01 1.34010926e-01 -2.74439305e-01
-1.00976992e+00 1.06523573e-01 3.25482190e-01 3.87277633e-01
4.82728302e-01 1.19783245e-01 8.23563263e-02 8.87018919e-01
-4.33418542e-01 -1.62492275e-01 3.73642921e-01 6.83905661e-01
-5.05121052e-01 -1.09160924e+00 -1.17136627e-01 3.63909572e-01
-1.81522533e-01 -4.44094300e-01 -5.98831415e-01 6.65803671e-01
3.15897316e-01 8.06864023e-01 1.79437190e-01 -5.63794494e-01
1.42686009e-01 3.87850344e-01 -7.65316039e-02 -7.80199111e-01
-2.50346035e-01 -3.67597044e-01 -1.52586326e-01 -5.88032722e-01
-4.52501774e-01 -3.08526218e-01 -1.53229976e+00 -1.16440311e-01
-1.68114305e-02 1.90226763e-01 5.58583617e-01 1.26997948e+00
5.79704344e-01 5.91885686e-01 3.36283326e-01 -3.67935956e-01
-3.18695098e-01 -7.17280507e-01 -3.12659204e-01 5.78311861e-01
2.87004977e-01 -7.50063717e-01 -3.46100748e-01 5.14396250e-01] | [8.985823631286621, 5.788405418395996] |
b0007640-1306-479b-9b13-e341372b6a9c | efficient-object-level-visual-context | 2101.05208 | null | https://arxiv.org/abs/2101.05208v1 | https://arxiv.org/pdf/2101.05208v1.pdf | Efficient Object-Level Visual Context Modeling for Multimodal Machine Translation: Masking Irrelevant Objects Helps Grounding | Visual context provides grounding information for multimodal machine translation (MMT). However, previous MMT models and probing studies on visual features suggest that visual information is less explored in MMT as it is often redundant to textual information. In this paper, we propose an object-level visual context modeling framework (OVC) to efficiently capture and explore visual information for multimodal machine translation. With detected objects, the proposed OVC encourages MMT to ground translation on desirable visual objects by masking irrelevant objects in the visual modality. We equip the proposed with an additional object-masking loss to achieve this goal. The object-masking loss is estimated according to the similarity between masked objects and the source texts so as to encourage masking source-irrelevant objects. Additionally, in order to generate vision-consistent target words, we further propose a vision-weighted translation loss for OVC. Experiments on MMT datasets demonstrate that the proposed OVC model outperforms state-of-the-art MMT models and analyses show that masking irrelevant objects helps grounding in MMT. | ['Deyi Xiong', 'Dexin Wang'] | 2020-12-18 | null | null | null | null | ['multimodal-machine-translation'] | ['natural-language-processing'] | [ 4.66485620e-01 1.10279601e-02 -3.21653038e-01 -4.63032760e-02
-6.58667803e-01 -4.69989538e-01 7.89573789e-01 1.06261708e-01
-3.52841914e-02 4.17716086e-01 3.86207759e-01 -5.10079563e-01
2.79181898e-01 -2.73174673e-01 -8.02224517e-01 -6.52729988e-01
5.82568467e-01 1.93827406e-01 -1.23570576e-01 -1.26940191e-01
5.01588583e-01 2.62173086e-01 -1.35909617e+00 9.40092862e-01
1.16339970e+00 7.98998713e-01 9.28321779e-01 3.58000755e-01
-5.82008600e-01 6.02978408e-01 -5.36802530e-01 -6.68677986e-01
2.81151116e-01 -7.51977205e-01 -6.22659624e-01 4.84232545e-01
7.08171368e-01 7.37404600e-02 -2.57737152e-02 1.19899225e+00
4.95766371e-01 3.68015990e-02 8.52698147e-01 -1.44936717e+00
-1.37930417e+00 4.87126499e-01 -9.47931170e-01 1.08118556e-01
2.55270541e-01 4.34619695e-01 1.00814867e+00 -1.62211156e+00
8.19035113e-01 1.62450635e+00 7.17753470e-02 5.60077667e-01
-1.25848103e+00 -4.11953121e-01 5.06821215e-01 4.37405527e-01
-1.08209455e+00 -4.23610806e-01 9.75933135e-01 -2.35566139e-01
7.70288110e-01 6.27435744e-01 6.95036948e-01 1.21628821e+00
3.09227675e-01 1.12773788e+00 1.40127790e+00 -8.45016837e-01
7.05789076e-04 6.12361968e-01 -2.13320911e-01 7.15585589e-01
1.64512232e-01 3.45988385e-02 -8.58982265e-01 1.70174241e-01
6.08612537e-01 -5.72164170e-02 -2.36520886e-01 -4.79730099e-01
-1.51604557e+00 6.28863275e-01 5.41982055e-01 2.98241287e-01
-3.61683041e-01 5.17295189e-02 3.29687186e-02 1.47871315e-01
4.73918170e-01 1.30158886e-01 6.45801798e-02 2.53504634e-01
-7.45408893e-01 -1.47654802e-01 -7.06358925e-02 1.36527729e+00
7.03829825e-01 1.02121323e-01 -9.93247211e-01 8.45021605e-01
7.07672834e-01 8.12750220e-01 1.70411274e-01 -7.48292804e-01
1.15022790e+00 9.93132412e-01 9.47899092e-03 -1.07791018e+00
4.52924892e-02 -4.92429316e-01 -7.49833822e-01 2.31000632e-01
4.44619022e-02 2.64950722e-01 -1.05173254e+00 1.54476750e+00
3.26599270e-01 -3.99178535e-01 -1.08565921e-02 1.32429016e+00
8.89194191e-01 7.61642098e-01 2.37572089e-01 -5.42226434e-01
1.48902810e+00 -1.05129671e+00 -1.17781425e+00 -5.62621832e-01
4.16866153e-01 -1.31210542e+00 1.63073647e+00 -1.19054280e-02
-1.23042285e+00 -5.64062953e-01 -8.53053451e-01 -2.94516116e-01
-2.49271229e-01 5.15960336e-01 4.28484797e-01 6.63671255e-01
-9.12790239e-01 -9.33641195e-02 -3.31291020e-01 -4.97606099e-01
5.93618393e-01 -1.08017825e-01 -9.14713591e-02 -1.30400330e-01
-8.51097524e-01 1.11184573e+00 3.27025414e-01 3.10415864e-01
-9.44836080e-01 -3.43975335e-01 -8.80010188e-01 -2.95843892e-02
2.82142371e-01 -8.57267559e-01 8.27689052e-01 -1.41079772e+00
-1.06249642e+00 8.39846373e-01 -5.95127344e-01 -3.32742110e-02
4.94726896e-01 8.43819231e-02 -1.91845536e-01 3.52429867e-01
1.72138020e-01 1.04446208e+00 1.24192691e+00 -2.01028800e+00
-5.82797110e-01 -3.84559780e-01 -1.48402035e-01 6.69099033e-01
-4.92044359e-01 4.10326570e-02 -6.90267563e-01 -1.07142496e+00
2.65474945e-01 -7.48117328e-01 1.06882334e-01 2.91853905e-01
-5.35649717e-01 7.36715877e-03 1.12624753e+00 -6.94995582e-01
1.10618782e+00 -1.96548414e+00 4.11805153e-01 -8.30267221e-02
2.09308252e-01 -7.75962248e-02 -4.06535059e-01 3.55957747e-01
2.72947460e-01 -3.89329195e-02 -9.42233130e-02 -6.70193136e-01
6.21972270e-02 2.32132107e-01 -2.55013198e-01 2.75858164e-01
4.55861092e-01 1.34732592e+00 -6.44622803e-01 -1.00698090e+00
3.79632890e-01 4.59139287e-01 -2.02440932e-01 -4.42231074e-03
-4.37286943e-01 5.57230532e-01 -2.12407842e-01 1.11843348e+00
8.67095530e-01 -2.40319669e-01 1.46108747e-01 -3.77768964e-01
1.33705931e-03 -2.20564887e-01 -6.68233693e-01 1.71148968e+00
-3.22052240e-01 9.49415743e-01 1.73560381e-02 -7.37689614e-01
8.31434906e-01 1.03513174e-01 -1.00498445e-01 -1.09952998e+00
2.49588713e-01 -7.41383247e-03 -1.02656081e-01 -6.16952598e-01
7.00324059e-01 -7.11714942e-03 2.79719830e-01 2.27101877e-01
-2.47092068e-01 7.26724118e-02 -1.57636821e-01 3.63722771e-01
1.95093706e-01 2.38262340e-01 1.00633251e-02 -8.88572410e-02
3.82757038e-01 3.05239081e-01 4.61661220e-01 5.06275415e-01
-3.61327618e-01 5.70597708e-01 1.36571884e-01 7.47065851e-03
-9.47348475e-01 -9.19346869e-01 1.98797673e-01 1.15966916e+00
6.70888186e-01 -1.12590738e-01 -8.40885878e-01 -6.96608424e-01
-2.56745189e-01 1.01296127e+00 -7.35438585e-01 -7.50146285e-02
-5.24051785e-01 -4.61612582e-01 -1.24881919e-02 5.05215526e-01
5.10977983e-01 -1.12308812e+00 -4.37698871e-01 -1.89594716e-01
-6.98695898e-01 -1.07419717e+00 -1.07616496e+00 -2.07404107e-01
-9.77884471e-01 -6.42472446e-01 -9.70531583e-01 -1.07179844e+00
1.03177524e+00 8.68808270e-01 9.65618432e-01 -5.95862493e-02
-2.28598222e-01 5.21155536e-01 -4.31234896e-01 -5.02574861e-01
-4.78879184e-01 -4.45670456e-01 -1.09776013e-01 3.67765307e-01
3.61088037e-01 2.69602668e-02 -7.03264117e-01 3.32776904e-01
-8.33713055e-01 6.47063613e-01 8.97677541e-01 7.84857571e-01
7.99791336e-01 -4.91108090e-01 1.25358567e-01 -1.07066602e-01
6.08093560e-01 -1.49221301e-01 -3.35786730e-01 7.71627128e-01
-6.36644840e-01 1.49087161e-02 6.81674853e-02 -9.86149788e-01
-1.25808907e+00 -1.93813562e-01 7.46138632e-01 -8.64908516e-01
2.37550780e-01 2.30172664e-01 -4.41983849e-01 -1.87983185e-01
3.38006526e-01 6.02964938e-01 -1.15607329e-01 -2.90127158e-01
6.92826390e-01 4.92206097e-01 1.61976125e-02 -5.79094708e-01
7.21740127e-01 3.61712992e-01 1.29998505e-01 -6.31481707e-01
-4.48746026e-01 -1.45653695e-01 -2.63694376e-01 -5.32177329e-01
9.55172718e-01 -7.27792442e-01 -5.54446161e-01 -1.13271281e-01
-1.33985114e+00 -1.29339388e-02 9.92996429e-05 4.31205422e-01
-5.08797586e-01 5.94804764e-01 -1.49568662e-01 -1.25830126e+00
-4.91063714e-01 -1.35815609e+00 1.39260817e+00 1.38924584e-01
5.19877709e-02 -7.21713245e-01 -3.34811062e-01 7.47105896e-01
2.71103770e-01 -1.15385517e-01 1.15742946e+00 -7.78982788e-02
-1.00968516e+00 4.04431462e-01 -6.76136017e-01 3.39188464e-02
3.90987769e-02 -2.12987632e-01 -8.61586809e-01 -2.06498861e-01
-5.03149396e-03 -7.68966526e-02 9.18276846e-01 5.46452701e-01
8.35617661e-01 -5.24254680e-01 -3.74675155e-01 3.56210113e-01
1.23476338e+00 1.78316176e-01 6.48877680e-01 3.37774694e-01
8.58879805e-01 9.24282730e-01 1.09109282e+00 4.57271114e-02
3.14757288e-01 8.87673140e-01 6.89576328e-01 -4.70076293e-01
-5.75681984e-01 -3.15883845e-01 4.92550164e-01 7.42588162e-01
-7.81191885e-02 -3.74404341e-01 -6.35147154e-01 6.15643263e-01
-2.04005647e+00 -7.22158015e-01 -3.18359345e-01 1.85826850e+00
7.39213169e-01 -1.52819559e-01 -1.67299241e-01 -2.90976763e-01
9.51395452e-01 -2.07806490e-02 -2.20715314e-01 -4.34867412e-01
-5.58268070e-01 -4.16318059e-01 1.03132166e-01 4.68506873e-01
-7.42025018e-01 1.09205687e+00 5.48726320e+00 1.08569860e+00
-8.77667010e-01 4.32739198e-01 6.35838926e-01 -2.13411286e-01
-7.84910202e-01 1.09638341e-01 -5.54409921e-01 3.70833337e-01
1.17895119e-01 1.08564995e-01 4.39830393e-01 5.47744036e-01
5.74790657e-01 -5.61806448e-02 -9.40522671e-01 1.27815771e+00
3.60048115e-01 -1.27975714e+00 7.36239552e-01 2.29553029e-01
8.31962764e-01 -6.12547100e-01 6.65396750e-01 1.85854331e-01
-3.35726798e-01 -8.65760088e-01 1.18183792e+00 4.00159538e-01
7.90238798e-01 -6.39818788e-01 3.97514462e-01 1.21859133e-01
-1.20403457e+00 -9.82640907e-02 -4.81498778e-01 2.77737021e-01
-5.38567565e-02 2.37024203e-01 -8.55112493e-01 5.89562595e-01
4.40401345e-01 3.90394121e-01 -8.56333792e-01 6.62289739e-01
-1.27409428e-01 2.34809980e-01 3.00642937e-01 -2.27548614e-01
3.44778061e-01 -1.97232574e-01 8.79773200e-01 1.19102776e+00
3.71716082e-01 -2.17672378e-01 1.00792959e-01 1.24595046e+00
-5.71036227e-02 4.86493796e-01 -4.76555526e-01 -7.13272244e-02
4.92695600e-01 1.09876919e+00 -6.49126649e-01 -3.22191685e-01
-4.35854375e-01 1.28274488e+00 1.99765921e-01 7.70225585e-01
-8.01809311e-01 1.90764621e-01 4.03817534e-01 7.46963639e-03
2.91679740e-01 -4.95549776e-02 -8.01016867e-01 -1.24162328e+00
3.61861378e-01 -9.89379525e-01 1.29152730e-01 -1.23346329e+00
-1.09846139e+00 4.13746685e-01 -5.75348511e-02 -1.68627381e+00
3.58843535e-01 -6.63743854e-01 -3.10105562e-01 1.08972287e+00
-1.45830917e+00 -1.83861828e+00 -2.81180680e-01 6.69500291e-01
1.08798289e+00 -1.86640367e-01 2.61307627e-01 -1.47366896e-01
-3.08317363e-01 7.58208930e-01 -2.02047899e-01 -2.38046974e-01
8.06522191e-01 -9.27358031e-01 5.35624735e-02 1.07500076e+00
3.68819058e-01 7.70470083e-01 8.28314483e-01 -9.45232749e-01
-1.92745316e+00 -1.22137368e+00 9.22501087e-01 -6.25217021e-01
2.18292847e-01 -3.40091556e-01 -7.84465432e-01 5.04080117e-01
7.60534227e-01 -3.51490557e-01 4.17454511e-01 -2.14987874e-01
-4.21778560e-01 1.00409687e-01 -1.13146687e+00 1.09253228e+00
1.00228691e+00 -4.04309243e-01 -7.09191203e-01 2.00527817e-01
8.10113549e-01 -1.63897350e-01 -1.94618836e-01 2.41759822e-01
4.90275860e-01 -4.99091417e-01 1.02799022e+00 -4.23374265e-01
5.53695261e-01 -6.01108730e-01 -4.99591112e-01 -1.16404307e+00
-2.50901312e-01 -5.61285675e-01 -3.77034128e-01 1.23891783e+00
4.87683415e-01 -2.33934984e-01 4.03945476e-01 1.89338624e-01
-2.55124927e-01 -4.58086818e-01 -1.04568624e+00 -7.27099836e-01
-1.32029220e-01 -3.46939653e-01 3.80393296e-01 1.06634259e+00
-7.76990782e-03 3.71391535e-01 -6.13663018e-01 2.66458452e-01
9.37895238e-01 3.77425700e-01 5.27701735e-01 -6.50000930e-01
-6.10011332e-02 -6.93943977e-01 -8.35332647e-02 -9.06963766e-01
-2.70118386e-01 -8.62363756e-01 4.21103947e-02 -1.85664451e+00
7.81730473e-01 1.13921938e-03 -2.29152486e-01 4.03611273e-01
-6.23578668e-01 3.86496961e-01 4.79016781e-01 3.17720324e-01
-6.17758214e-01 8.87316644e-01 1.99305558e+00 -6.27929270e-01
-1.54663250e-01 -4.97901797e-01 -7.53104389e-01 2.95663953e-01
5.86378992e-01 -2.73032755e-01 -5.33553958e-01 -7.00866699e-01
1.41331494e-01 -5.49328327e-02 6.63301289e-01 -2.71767139e-01
-8.84949565e-02 -5.81143737e-01 7.58296609e-01 -1.06344473e+00
4.63632017e-01 -9.69568670e-01 -1.16132654e-01 3.05753082e-01
-4.26881760e-01 6.23284886e-03 2.81483710e-01 7.68751383e-01
2.10933089e-02 4.58645727e-03 5.36916375e-01 -1.31723154e-02
-5.76455116e-01 -5.34762777e-02 -3.72720212e-01 -2.39639029e-01
9.41957831e-01 -5.26671648e-01 -5.90621173e-01 -2.58580744e-01
-4.41671103e-01 2.55483180e-01 5.80203474e-01 7.62154460e-01
1.12102783e+00 -1.79594409e+00 -8.81444573e-01 -8.56149122e-02
6.54538393e-01 -6.52769387e-01 2.64170438e-01 1.07182038e+00
-1.18599586e-01 4.97464895e-01 -2.03665525e-01 -9.11188066e-01
-1.71149218e+00 1.05969453e+00 -6.45497665e-02 2.32959121e-01
-5.37579834e-01 6.79949641e-01 7.10523546e-01 -4.62435260e-02
2.99876422e-01 -3.88923764e-01 -2.15090603e-01 -6.38462901e-02
5.58959901e-01 2.08033979e-01 -1.80492997e-01 -7.62011647e-01
-4.14480269e-01 6.46657944e-01 -7.35817328e-02 -4.05817181e-01
5.48095584e-01 -7.19300568e-01 -8.45180079e-02 2.55928516e-01
7.50973821e-01 -1.09796477e-02 -1.11692560e+00 -3.62157196e-01
-2.56825745e-01 -7.56254792e-01 5.44736795e-02 -1.09135187e+00
-7.61549950e-01 1.17199802e+00 9.86049950e-01 -2.89079159e-01
1.32506371e+00 1.27024978e-01 3.73006642e-01 5.70449196e-02
2.63885260e-01 -1.07489955e+00 5.85076392e-01 1.23657897e-01
1.33000004e+00 -1.43071210e+00 -1.11669257e-01 -5.34101129e-01
-1.14395082e+00 7.68647015e-01 7.11015463e-01 6.29221022e-01
-2.59349495e-01 -2.91969806e-01 2.52020895e-01 -8.13960955e-02
-7.61799216e-01 -3.75226498e-01 8.85701239e-01 8.21600974e-01
2.37713054e-01 1.13609061e-01 -4.31771040e-01 4.35559452e-01
3.98754895e-01 -5.34272969e-01 1.34461239e-01 8.44650209e-01
-6.86574996e-01 -8.27847898e-01 -9.35482800e-01 -4.93599586e-02
-1.28318086e-01 -5.21760046e-01 -7.82805264e-01 5.63440502e-01
1.29845336e-01 1.31911647e+00 -2.08107516e-01 -2.72614002e-01
6.30863607e-02 4.49199192e-02 8.55911314e-01 -4.05226171e-01
-4.95925456e-01 5.61253548e-01 -8.20544362e-02 -3.80348206e-01
-4.47908819e-01 -4.34781104e-01 -8.55897069e-01 -2.59089749e-02
-4.24368054e-01 -1.84639230e-01 9.15584445e-01 8.48458946e-01
3.98815781e-01 5.19898474e-01 4.74908829e-01 -8.13965559e-01
-3.61444317e-02 -1.09592497e+00 -1.20202221e-01 4.86670613e-01
2.70739704e-01 -6.84338510e-01 -2.04745814e-01 2.23038137e-01] | [11.463756561279297, 1.484451413154602] |
c1803c7c-6612-4a16-965d-5cc66e4f5e0c | exploring-diffusion-models-for-unsupervised | 2304.05841 | null | https://arxiv.org/abs/2304.05841v2 | https://arxiv.org/pdf/2304.05841v2.pdf | Exploring Diffusion Models for Unsupervised Video Anomaly Detection | This paper investigates the performance of diffusion models for video anomaly detection (VAD) within the most challenging but also the most operational scenario in which the data annotations are not used. As being sparse, diverse, contextual, and often ambiguous, detecting abnormal events precisely is a very ambitious task. To this end, we rely only on the information-rich spatio-temporal data, and the reconstruction power of the diffusion models such that a high reconstruction error is utilized to decide the abnormality. Experiments performed on two large-scale video anomaly detection datasets demonstrate the consistent improvement of the proposed method over the state-of-the-art generative models while in some cases our method achieves better scores than the more complex models. This is the first study using a diffusion model and examining its parameters' influence to present guidance for VAD in surveillance scenarios. | ['Elisa Ricci', 'Cigdem Beyan', "Nicola Dall'Asen", 'Anil Osman Tur'] | 2023-04-12 | null | null | null | null | ['video-anomaly-detection'] | ['computer-vision'] | [ 1.53064564e-01 -3.32748234e-01 1.38159111e-01 -1.81309015e-01
-2.39296883e-01 -4.95396882e-01 9.94448543e-01 3.80094051e-01
-2.90973306e-01 3.42681706e-01 2.06446648e-01 -5.00612736e-01
-2.06784070e-01 -6.69926763e-01 -5.32272696e-01 -6.52939558e-01
-5.16848207e-01 5.60393512e-01 7.93486357e-01 -1.34342179e-01
1.58667415e-01 5.69891691e-01 -1.57116044e+00 5.22433817e-02
8.27850699e-01 1.08550584e+00 -1.78591028e-01 9.64286089e-01
-2.04037964e-01 9.67091322e-01 -6.99707150e-01 -3.25986505e-01
2.38517240e-01 -3.86756539e-01 -2.75913388e-01 2.17715994e-01
4.29582864e-01 -6.14706039e-01 -3.95316452e-01 9.54598010e-01
4.21910703e-01 -1.05151376e-02 8.35520804e-01 -1.30034971e+00
-3.39255840e-01 1.30132660e-01 -5.61602354e-01 1.16315591e+00
5.31093121e-01 3.56535226e-01 5.31664193e-01 -5.13791561e-01
6.41732514e-01 1.07362342e+00 6.66423142e-01 4.54478592e-01
-7.66678929e-01 -4.82545674e-01 5.85136294e-01 3.05933982e-01
-1.15114999e+00 -3.64937156e-01 7.99208522e-01 -6.87787175e-01
1.09722686e+00 2.19383508e-01 7.23399341e-01 1.78140855e+00
2.50477850e-01 6.36975229e-01 7.72469461e-01 -7.23321885e-02
3.88491035e-01 -6.15822412e-02 9.16895419e-02 8.72590125e-01
4.31466609e-01 -5.69047146e-02 -3.62990290e-01 -5.51857471e-01
6.35544658e-01 2.87376851e-01 -1.70361266e-01 -1.70489147e-01
-8.24301243e-01 7.98764169e-01 -7.56231546e-02 5.02964377e-01
-8.69618833e-01 -1.89033881e-01 4.98784125e-01 3.22887748e-01
8.95890892e-01 1.05174974e-01 -8.30440745e-02 -7.19387710e-01
-1.05423462e+00 2.87813038e-01 7.04263449e-01 6.99212253e-01
5.05002104e-02 3.72852653e-01 -6.63918614e-01 4.66741234e-01
3.41692328e-01 4.90869462e-01 1.42746702e-01 -4.39123690e-01
5.43427527e-01 7.10645795e-01 1.47636622e-01 -1.28934312e+00
-3.87433112e-01 -5.63686490e-01 -8.43815148e-01 8.94499496e-02
5.75752437e-01 -3.11478376e-01 -1.34723711e+00 1.50273931e+00
3.21798295e-01 8.07883322e-01 -1.09267578e-01 7.45942533e-01
6.05654836e-01 7.10107982e-01 3.06011885e-01 -3.52734923e-01
9.95697916e-01 -5.48836648e-01 -9.28328574e-01 -2.32269973e-01
5.80185413e-01 -7.77298987e-01 4.18086410e-01 4.61004585e-01
-9.14536417e-01 -3.40016395e-01 -8.16258729e-01 5.96129298e-01
-2.78765500e-01 -1.34021059e-01 4.71284270e-01 6.76130712e-01
-9.91857648e-01 1.58657119e-01 -1.21829581e+00 -8.26398849e-01
6.86794579e-01 -1.26793772e-01 -1.27746880e-01 -2.65691698e-01
-9.71653819e-01 8.22980762e-01 7.93130845e-02 -2.00211722e-02
-1.20140445e+00 -5.25643289e-01 -6.65953815e-01 -2.31140584e-01
6.39412403e-01 -5.68172634e-01 8.61562431e-01 -7.57954538e-01
-8.75955880e-01 5.17532825e-01 -3.29128683e-01 -8.92601728e-01
9.61379528e-01 -4.11009938e-01 -7.77142286e-01 2.28870869e-01
3.29360664e-02 6.12688176e-02 1.08191764e+00 -1.13060999e+00
-7.61144280e-01 -4.46931481e-01 2.90521346e-02 -4.55833487e-02
-2.07230747e-01 3.71775739e-02 -5.61611831e-01 -9.18000877e-01
1.19035810e-01 -8.31077754e-01 -2.43635491e-01 -1.16017446e-01
-2.58497864e-01 -3.65943253e-01 1.36129701e+00 -9.22565460e-01
1.61537969e+00 -2.18239045e+00 2.77035981e-02 4.84059930e-01
2.62669206e-01 5.03243089e-01 1.34122089e-01 4.50824499e-01
3.14581424e-01 1.44160658e-01 -3.72912794e-01 -3.07777256e-01
-2.46095791e-01 5.14286041e-01 -3.19998831e-01 4.94163334e-01
4.92644250e-01 5.91966689e-01 -9.33859527e-01 -4.65213716e-01
1.23689823e-01 3.53826761e-01 -5.01744926e-01 4.74703014e-01
-1.89565003e-01 6.47852242e-01 -7.57441938e-01 9.17297006e-01
3.88040274e-01 -5.06493039e-02 -3.47214162e-01 1.44636929e-01
4.21710350e-02 -2.53751218e-01 -1.17974305e+00 1.48083091e+00
6.10824898e-02 7.64140368e-01 -1.42282695e-01 -9.29381251e-01
6.78390801e-01 4.46065575e-01 8.16550136e-01 -5.64601600e-01
-1.42359465e-01 -1.53897544e-02 9.67577994e-02 -1.00249720e+00
4.10710961e-01 3.91438067e-01 3.63369524e-01 2.88869202e-01
2.97709219e-02 3.71385396e-01 3.16223472e-01 5.02655149e-01
1.61185932e+00 1.89436153e-02 1.27265915e-01 2.85618678e-02
3.52879435e-01 8.29878151e-02 7.23821461e-01 1.29123735e+00
-4.39008117e-01 5.78996122e-01 7.68772125e-01 -4.41448927e-01
-1.11322045e+00 -1.03441203e+00 1.05187646e-03 6.60384953e-01
-2.53798850e-02 -4.35516000e-01 -6.71336234e-01 -1.16767788e+00
-2.27896139e-01 8.47071707e-01 -8.75625134e-01 -9.77553427e-02
-3.80842835e-01 -1.09896767e+00 8.05928886e-01 4.54630226e-01
5.10937572e-01 -7.92016804e-01 -6.66662335e-01 2.12870196e-01
-9.93210822e-02 -1.19663405e+00 -4.64345142e-02 -3.04526389e-01
-8.71591091e-01 -1.28340399e+00 -7.44035840e-01 -1.33327723e-01
6.93205059e-01 1.80840120e-02 1.18500376e+00 1.95664048e-01
-1.56509265e-01 9.06437278e-01 -7.08501637e-01 -8.18197250e-01
-4.73776311e-01 -2.27514774e-01 4.33556959e-02 3.40154499e-01
6.52015865e-01 -3.63675594e-01 -5.85698545e-01 2.36729845e-01
-1.23593771e+00 -4.95840967e-01 5.67970276e-01 4.84262735e-01
3.28850925e-01 3.46032709e-01 5.02195239e-01 -1.01195884e+00
7.54917502e-01 -9.97387826e-01 -3.72471690e-01 -2.84506730e-03
-6.20818973e-01 -2.81722218e-01 1.78614393e-01 -5.09837866e-01
-1.10651410e+00 -2.28003591e-01 -2.86041588e-01 -6.44788384e-01
-6.27131641e-01 2.77257174e-01 2.73182362e-01 2.44721293e-01
6.71704233e-01 2.88459957e-01 -1.83604330e-01 -3.08933526e-01
-2.70399332e-01 3.85627151e-01 1.96248934e-01 -2.36691520e-01
8.15338790e-01 6.87930763e-01 3.69530544e-02 -1.21371615e+00
-5.47765553e-01 -6.30410075e-01 -5.83477736e-01 -4.22726303e-01
1.03674042e+00 -6.80694044e-01 1.66014761e-01 6.31472766e-01
-1.18139589e+00 -1.31306592e-02 -1.73883036e-01 5.03188908e-01
-5.38449958e-02 6.66428149e-01 -2.53453791e-01 -1.23651290e+00
-1.08985938e-02 -1.07927489e+00 9.99184906e-01 -1.48507850e-02
-1.40291855e-01 -1.12847245e+00 2.39179522e-01 1.72333419e-01
7.15173066e-01 5.25867999e-01 7.73601234e-01 -1.11173463e+00
-6.10486686e-01 -3.22189927e-01 -4.65871021e-02 1.60617739e-01
4.14722264e-02 3.12954009e-01 -9.26264644e-01 -2.87701432e-02
1.05006117e-02 2.33893514e-01 9.39776540e-01 6.13591433e-01
1.15056670e+00 -2.45286256e-01 -4.12850380e-01 3.66050780e-01
1.11776102e+00 3.26806217e-01 8.06025743e-01 3.67336124e-01
6.79644287e-01 4.24615502e-01 7.78582513e-01 7.15949714e-01
2.84574926e-01 4.86233890e-01 7.25462317e-01 2.26611067e-02
1.69324651e-01 4.02899022e-05 4.21457142e-01 2.99606562e-01
-4.84259158e-01 -6.72984779e-01 -1.10883737e+00 7.42688179e-01
-2.13225365e+00 -1.41708314e+00 -2.79981077e-01 2.05449748e+00
1.58493534e-01 3.87486011e-01 3.46574008e-01 1.72970265e-01
4.47482258e-01 5.47943652e-01 -3.28998774e-01 -3.80820900e-01
-1.97335839e-01 -2.41688326e-01 3.64570588e-01 7.23224804e-02
-1.33528972e+00 6.48349702e-01 6.53084183e+00 6.38059080e-01
-8.56514990e-01 5.18534817e-02 5.17146051e-01 -3.95340621e-02
-8.51613879e-02 -4.09323692e-01 -4.57976252e-01 8.14665735e-01
9.36881840e-01 4.03990120e-01 1.00422725e-01 6.91092789e-01
5.07643402e-01 -2.97867149e-01 -7.88032889e-01 9.14592206e-01
1.78002089e-01 -1.05672944e+00 2.74128467e-01 -1.11054806e-02
6.85137451e-01 9.09838453e-02 1.35208294e-01 9.47887227e-02
5.55018112e-02 -7.77300358e-01 4.52618450e-01 9.55022693e-01
6.38749376e-02 -6.70571744e-01 1.07548153e+00 3.47538590e-01
-8.94977927e-01 -2.83543527e-01 -1.46188721e-01 2.81711906e-01
3.19010496e-01 7.18081236e-01 -8.58248949e-01 4.18402344e-01
9.71854985e-01 9.12254333e-01 -9.23076451e-01 1.31215370e+00
4.08896469e-02 1.30809546e+00 -3.00323933e-01 1.78102821e-01
6.35984540e-01 -1.77256033e-01 1.24779165e+00 1.59068739e+00
6.85656667e-01 1.65823057e-01 1.80331692e-01 5.48703671e-01
3.87543589e-01 -2.44656168e-02 -1.15893042e+00 -8.04781467e-02
-2.42216093e-03 9.79087949e-01 -7.19092488e-01 -4.45849597e-01
-6.23831928e-01 1.01019526e+00 -4.68270220e-02 5.10366678e-01
-1.06388068e+00 2.90593326e-01 6.41537547e-01 3.53702039e-01
3.92738104e-01 -3.76874685e-01 2.87583750e-03 -1.11045241e+00
3.32655758e-01 -8.12020719e-01 7.51837969e-01 -4.56761777e-01
-1.54086936e+00 8.26704919e-01 1.73927784e-01 -1.36386406e+00
-5.52481413e-01 -3.78133535e-01 -8.18337917e-01 5.44416189e-01
-1.54088211e+00 -1.10702944e+00 -5.19773364e-01 8.64788771e-01
8.18018615e-01 -4.62059200e-01 5.33097863e-01 4.29264873e-01
-6.78789794e-01 1.86395481e-01 -1.51690662e-01 2.06262767e-01
4.26505774e-01 -1.30416334e+00 2.52214462e-01 1.43845713e+00
2.49973416e-01 1.42250925e-01 9.73632276e-01 -1.12029552e+00
-1.12025130e+00 -1.17639863e+00 4.66557324e-01 -7.71231711e-01
6.23219252e-01 -8.88976157e-02 -1.23822129e+00 6.09056294e-01
1.24659494e-01 -6.47261515e-02 5.61145246e-01 -3.02233864e-02
4.86447923e-02 2.81025946e-01 -1.21629620e+00 3.64488423e-01
1.30715668e+00 -2.50019342e-01 -6.31669164e-01 3.08519721e-01
2.20580101e-01 -4.36992109e-01 -5.86493731e-01 4.16583002e-01
1.58418909e-01 -1.10900056e+00 7.88671792e-01 -7.96360016e-01
2.05426350e-01 -3.35195392e-01 2.47434042e-02 -1.23792088e+00
-3.11241686e-01 -3.11376363e-01 -6.34033561e-01 1.31651855e+00
3.86685044e-01 -5.92704058e-01 5.63618243e-01 5.65852284e-01
-5.32559045e-02 -5.18310726e-01 -1.01193750e+00 -6.33635044e-01
-7.16962516e-01 -8.29433918e-01 3.45332950e-01 1.05435622e+00
-8.18590343e-01 -1.37738198e-01 -6.19955719e-01 7.82062352e-01
5.87287784e-01 -4.96846616e-01 7.42948174e-01 -1.38580310e+00
2.38779224e-02 -2.81149864e-01 -1.05885851e+00 -6.80080473e-01
-1.63935751e-01 -2.67842144e-01 -4.03978258e-01 -1.55484414e+00
-1.67640954e-01 -3.87531370e-01 -3.92969042e-01 7.21975043e-02
-3.19569141e-01 4.00050022e-02 -2.77018636e-01 1.92691281e-01
-7.89058030e-01 3.31712425e-01 6.96409643e-01 -1.21046156e-01
-3.84700954e-01 3.44125032e-01 -2.06226781e-01 8.50621521e-01
6.60299361e-01 -4.98863935e-01 -5.29643536e-01 -4.35439020e-01
2.52304878e-02 -2.31393725e-01 6.12914145e-01 -1.07689703e+00
3.93895000e-01 -1.38387367e-01 4.69967306e-01 -7.22826838e-01
1.83633551e-01 -1.13118231e+00 -9.20232832e-02 2.99193263e-01
-3.54597233e-02 5.73956788e-01 3.07412595e-01 1.24735284e+00
-2.10591882e-01 2.58445479e-02 3.13897610e-01 -2.49401666e-02
-1.14949870e+00 6.76231146e-01 -6.52285695e-01 1.15584649e-01
1.34693575e+00 -3.36621135e-01 -1.47195026e-01 -6.99837625e-01
-7.61290312e-01 9.01986957e-02 3.08703154e-01 7.74067104e-01
7.44741321e-01 -1.05009794e+00 -8.86716008e-01 3.07042360e-01
2.64392763e-01 -4.07631919e-02 4.21627194e-01 1.03952360e+00
-4.99614120e-01 2.35290140e-01 -1.11338593e-01 -1.06903279e+00
-1.25831842e+00 5.59470475e-01 2.97429264e-01 -2.98478067e-01
-7.97572970e-01 3.98670822e-01 -7.09871575e-02 2.66906559e-01
3.51574123e-01 -2.54844010e-01 -5.79327881e-01 9.81716216e-02
5.62942684e-01 6.40154779e-01 -2.60480354e-03 -6.38033271e-01
-2.98157543e-01 2.62722880e-01 -2.96653118e-02 6.71609119e-02
1.29407752e+00 -1.42993346e-01 2.19300926e-01 3.25710267e-01
4.22079623e-01 -1.67136341e-02 -1.28179646e+00 -2.66585797e-01
7.21863583e-02 -7.47871518e-01 1.25979990e-01 -6.06023490e-01
-9.87199664e-01 7.08301723e-01 9.71157789e-01 6.72869623e-01
1.13386428e+00 -7.46614262e-02 4.48036909e-01 1.70234561e-01
1.56984348e-02 -1.11625123e+00 4.63437848e-02 2.37860247e-01
7.81983674e-01 -1.47309387e+00 -1.62936300e-01 -2.41976380e-01
-8.98925066e-01 8.19496870e-01 7.75593162e-01 -5.10454252e-02
7.65229166e-01 -7.97383301e-03 1.06632322e-01 -4.74240154e-01
-7.27617383e-01 -2.69239336e-01 6.00371540e-01 7.77836144e-01
4.99130562e-02 -3.45505655e-01 -1.03406934e-02 1.36881143e-01
4.22160655e-01 -1.69621408e-01 4.00484711e-01 1.04647541e+00
-2.81438321e-01 -6.28597140e-01 -4.01163995e-01 9.12320197e-01
-8.74771476e-01 1.51436895e-01 -4.19576466e-01 7.67334342e-01
1.83002725e-01 1.12315500e+00 2.78458834e-01 -1.30421743e-01
5.17652392e-01 2.34212041e-01 4.89829062e-03 -2.83481449e-01
-3.89329910e-01 -3.27278562e-02 1.71537250e-01 -7.98579991e-01
-7.12967813e-01 -8.39636505e-01 -8.18515599e-01 -1.53197289e-01
-1.98170971e-02 -2.92185266e-02 5.66992521e-01 1.07749701e+00
5.14437437e-01 7.26071954e-01 3.91594350e-01 -3.43159109e-01
-2.61356831e-01 -9.16155875e-01 -3.96939009e-01 7.02569842e-01
5.57940364e-01 -8.55207562e-01 -3.49737763e-01 -1.59515947e-01] | [7.858697414398193, 1.5687147378921509] |
20888f9e-0eab-42a2-a263-41a9c9c6bbba | quantitative-manipulation-of-custom | null | null | http://openaccess.thecvf.com//content/CVPR2023/html/Do_Quantitative_Manipulation_of_Custom_Attributes_on_3D-Aware_Image_Synthesis_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Do_Quantitative_Manipulation_of_Custom_Attributes_on_3D-Aware_Image_Synthesis_CVPR_2023_paper.pdf | Quantitative Manipulation of Custom Attributes on 3D-Aware Image Synthesis | While 3D-based GAN techniques have been successfully applied to render photo-realistic 3D images with a variety of attributes while preserving view consistency, there has been little research on how to fine-control 3D images without limiting to a specific category of objects of their properties. To fill such research gap, we propose a novel image manipulation model of 3D-based GAN representations for a fine-grained control of specific custom attributes. By extending the latest 3D-based GAN models (e.g., EG3D), our user-friendly quantitative manipulation model enables a fine yet normalized control of 3D manipulation of multi-attribute quantities while achieving view consistency. We validate the effectiveness of our proposed technique both qualitatively and quantitatively through various experiments. | ['Jin Young Choi', 'Chul Lee', 'Taehyeong Kim', 'EunKyung Yoo', 'Hoseok Do'] | 2023-01-01 | null | null | null | cvpr-2023-1 | ['image-manipulation', '3d-aware-image-synthesis'] | ['computer-vision', 'computer-vision'] | [ 2.18433499e-01 2.37613857e-01 -1.50088057e-01 -3.70562881e-01
-2.71720827e-01 -6.81775689e-01 6.77702844e-01 -4.40932184e-01
3.41151506e-01 6.40116096e-01 3.19767386e-01 -3.11755203e-02
-7.03457892e-02 -1.10186028e+00 -6.82703257e-01 -5.81750691e-01
4.70199138e-01 3.54516000e-01 8.46206918e-02 -2.46539146e-01
7.04539344e-02 8.49470556e-01 -1.42387760e+00 -1.88710973e-01
9.40789104e-01 1.10124838e+00 -3.22908461e-02 4.20209706e-01
8.58131945e-02 4.52075690e-01 -4.55075085e-01 -2.32363656e-01
6.77651227e-01 -7.55781829e-01 -3.72647792e-01 5.40382445e-01
6.46579623e-01 -4.71693754e-01 -7.61887133e-02 8.78102601e-01
4.70126480e-01 -1.16500802e-01 6.81912065e-01 -1.37082911e+00
-1.19680357e+00 1.34074828e-02 -8.16837072e-01 -2.48163104e-01
3.57662857e-01 2.66040891e-01 7.04291940e-01 -3.56003702e-01
6.60399914e-01 1.48374116e+00 3.85100603e-01 7.72441387e-01
-1.30616915e+00 -5.34631550e-01 2.16977239e-01 -1.80108294e-01
-1.20771205e+00 -1.50565058e-01 1.17407870e+00 -3.76141578e-01
6.57693088e-01 4.60370719e-01 9.95046794e-01 1.13141859e+00
4.51464742e-01 3.56620252e-01 1.85009968e+00 -3.81407350e-01
3.01386863e-01 -4.38777246e-02 -4.15998340e-01 6.84425652e-01
1.00623660e-01 8.14646855e-02 -5.15326023e-01 1.60468891e-01
1.36624408e+00 -4.39147949e-02 -2.18753144e-01 -6.92280352e-01
-1.22830236e+00 6.26496077e-01 4.77499604e-01 -4.80833985e-02
-4.46564406e-01 4.68541026e-01 1.68613344e-02 2.02142462e-01
6.56399965e-01 3.41059715e-01 -3.13221663e-01 -9.88541543e-02
-6.48463309e-01 2.83031374e-01 2.90999979e-01 1.58084118e+00
8.75776529e-01 3.82713109e-01 -4.05690551e-01 6.79019213e-01
1.56425208e-01 6.53630257e-01 -9.39189177e-03 -1.25485456e+00
-3.55996229e-02 7.61987567e-01 1.39507353e-01 -8.38061452e-01
-9.36914086e-02 -3.09684217e-01 -1.25802636e+00 7.88219988e-01
5.35281980e-03 1.27407953e-01 -1.11591983e+00 1.93686152e+00
5.41799426e-01 -2.85659343e-01 -3.58333051e-01 8.93007636e-01
5.16345203e-01 4.65828300e-01 3.97303849e-02 -1.36444211e-01
1.20082223e+00 -7.11125135e-01 -6.94890916e-01 2.28077337e-01
-4.78081452e-03 -5.84164858e-01 1.48067796e+00 1.68225512e-01
-1.45514286e+00 -6.35108054e-01 -9.83649254e-01 -3.05008173e-01
-1.29773930e-01 -2.86696911e-01 7.36580074e-01 1.03403449e+00
-1.42110419e+00 4.80899774e-02 -6.33695304e-01 -5.24925888e-01
8.24708819e-01 2.78371602e-01 -3.61978740e-01 -4.39820774e-02
-7.65098035e-01 8.84037077e-01 8.10215808e-03 -2.04041928e-01
-1.00832558e+00 -8.84969950e-01 -6.10795438e-01 -8.50686952e-02
3.29532087e-01 -1.58338797e+00 7.60805845e-01 -6.15134478e-01
-1.86620736e+00 1.21198034e+00 1.27963901e-01 -5.14111519e-02
5.81119001e-01 -1.00103788e-01 -4.07485254e-02 -1.07095696e-01
5.54546341e-03 1.03776169e+00 9.95299578e-01 -1.59322965e+00
-2.38392949e-01 -5.50747454e-01 7.22665429e-01 4.13693309e-01
-3.98712754e-01 -2.17373788e-01 -3.64309669e-01 -8.90186787e-01
5.98687194e-02 -9.54338312e-01 -1.12451538e-01 5.37844062e-01
-5.03777087e-01 1.39244214e-01 1.09850657e+00 -1.61517680e-01
6.54696763e-01 -1.86932135e+00 3.78501058e-01 -1.05917498e-01
5.05844831e-01 -9.05436203e-02 -4.68640812e-02 3.37689131e-01
2.59966195e-01 4.68452096e-01 -2.75366127e-01 -4.01695341e-01
1.32268593e-01 3.12432289e-01 -1.34714484e-01 3.64083141e-01
2.17124820e-01 1.07261348e+00 -5.42668581e-01 -4.85848159e-01
6.02883875e-01 7.58820713e-01 -8.18035185e-01 3.24422061e-01
-4.19041246e-01 9.05711949e-01 -8.11801910e-01 8.58939946e-01
9.34323490e-01 -9.52512845e-02 -2.47336373e-01 -4.28362280e-01
1.24761220e-02 -1.82363987e-01 -8.78417373e-01 1.83028197e+00
-6.96664095e-01 2.70400554e-01 8.46739411e-02 -5.43641567e-01
1.03274632e+00 2.68756021e-02 5.27835727e-01 -6.99414194e-01
1.76308021e-01 -1.41432837e-01 -3.97213757e-01 -1.40351161e-01
5.57115316e-01 -3.48927677e-01 -2.90327877e-01 4.00986224e-01
-1.32082447e-01 -8.93365681e-01 -2.81272382e-01 9.18458253e-02
7.97316909e-01 5.61158657e-01 4.04378742e-01 -4.52662498e-01
4.16032225e-01 -1.62039071e-01 3.63748103e-01 6.92968726e-01
-1.51573392e-02 9.60579872e-01 2.18840003e-01 -2.19741076e-01
-1.43384981e+00 -1.12952602e+00 -1.44335292e-02 5.69246829e-01
3.98357719e-01 -2.16656536e-01 -8.61162186e-01 -5.63946903e-01
4.26370325e-03 6.64452732e-01 -9.15366054e-01 -6.96511269e-02
-3.60741824e-01 -4.56325144e-01 2.86869854e-01 4.91773218e-01
9.37941790e-01 -7.09168553e-01 -7.06522346e-01 -1.75722465e-01
2.62387604e-01 -1.12139416e+00 -5.20197034e-01 -2.12278828e-01
-9.79604065e-01 -6.24171257e-01 -6.62564635e-01 -4.27197814e-01
7.45705426e-01 2.37279907e-01 1.30492997e+00 7.62967467e-02
3.39212343e-02 7.36657143e-01 -4.60796684e-01 -4.06320572e-01
-5.84167957e-01 5.35811186e-02 6.89345002e-02 -3.28069814e-02
-2.97206312e-01 -1.13311827e+00 -8.19546819e-01 5.69709599e-01
-1.02787840e+00 5.03239870e-01 4.96494323e-01 6.43715084e-01
9.45925236e-01 -6.38376400e-02 4.64264601e-01 -8.99995089e-01
5.55946708e-01 7.16907009e-02 -5.81654251e-01 1.93175778e-01
-6.37132287e-01 -3.36274467e-02 6.61780655e-01 -3.19191486e-01
-1.16975629e+00 -1.53715298e-01 -8.92596170e-02 -6.20786190e-01
-2.75124967e-01 -1.40368134e-01 -7.62433827e-01 -3.70671332e-01
3.75494123e-01 1.62511438e-01 -6.19642087e-04 -3.63730848e-01
8.13095748e-01 4.16111618e-01 4.73809630e-01 -7.87426829e-01
1.14383757e+00 7.78576910e-01 4.60677356e-01 -4.41447437e-01
-6.86579585e-01 2.39965305e-01 -6.68684423e-01 -4.08953846e-01
1.01376843e+00 -7.99094319e-01 -7.18912780e-01 6.90027952e-01
-7.68526256e-01 -3.57020706e-01 -7.40930557e-01 9.25337821e-02
-1.12465131e+00 1.85074300e-01 -3.58814359e-01 -4.74103451e-01
-4.32985991e-01 -1.24537361e+00 1.40140355e+00 1.91955149e-01
-1.80031672e-01 -9.11193430e-01 -3.47437002e-02 3.75507295e-01
7.05913067e-01 7.51056075e-01 1.24963033e+00 3.80148590e-01
-9.89953339e-01 5.94279207e-02 -2.08422527e-01 3.68638635e-01
5.78677475e-01 -6.51437119e-02 -1.00811648e+00 -2.07852200e-01
2.20240295e-01 -2.97167718e-01 4.06774759e-01 4.70137864e-01
1.40822840e+00 -1.35628611e-01 -1.96226195e-01 9.67589915e-01
1.47967136e+00 1.26003340e-01 8.33725691e-01 1.75880805e-01
9.85077679e-01 4.40240391e-02 6.60301924e-01 5.13546705e-01
3.04583192e-01 9.60076809e-01 8.15957367e-01 -1.34864584e-01
-5.97448885e-01 -4.76011693e-01 -9.23622847e-02 7.43525565e-01
-5.02836585e-01 -6.07098758e-01 -2.97433943e-01 2.38795415e-01
-1.44006968e+00 -6.07025683e-01 2.14949608e-01 1.98181188e+00
6.06504500e-01 -5.78478270e-04 -4.31418940e-02 -5.76768145e-02
5.77800214e-01 4.60339814e-01 -7.66255081e-01 -4.14532274e-01
-3.76840562e-01 2.24674031e-01 4.08486873e-01 2.71274596e-01
-7.99088597e-01 8.89480472e-01 6.36293697e+00 9.23933446e-01
-1.00329351e+00 5.91526292e-02 8.42905283e-01 -9.08157229e-02
-1.01138484e+00 8.60168189e-02 -6.16828740e-01 2.76677549e-01
6.50333762e-02 -1.34558901e-01 3.27894568e-01 7.38472223e-01
4.33856189e-01 -5.71855605e-02 -1.07448280e+00 1.14933014e+00
3.05638965e-02 -1.59095991e+00 6.52793884e-01 3.38634729e-01
1.10206461e+00 -7.67538726e-01 2.63867110e-01 -3.91806550e-02
2.38857448e-01 -1.06754935e+00 1.02372599e+00 6.60710037e-01
1.27731180e+00 -8.03158820e-01 1.35590330e-01 9.41190496e-02
-7.94233918e-01 1.55454844e-01 -2.54419476e-01 1.00683212e-01
3.56847465e-01 7.08036780e-01 -3.59298527e-01 6.94067180e-01
7.81091213e-01 7.47284889e-01 -4.47929531e-01 5.51561534e-01
-1.96407869e-01 2.84112662e-01 -1.67296439e-01 1.37533501e-01
9.89526138e-02 -5.17384470e-01 6.68958068e-01 5.72259665e-01
6.10082507e-01 2.48262942e-01 -1.87626466e-01 1.25198472e+00
-3.11446309e-01 -2.36264542e-01 -8.09961140e-01 2.67458707e-01
4.74458486e-01 1.16526759e+00 -6.40050650e-01 -5.41028148e-03
-1.06401891e-01 1.27371359e+00 2.69455880e-01 1.58706263e-01
-9.07783568e-01 1.08658060e-01 8.92903030e-01 4.78975296e-01
2.88890153e-01 -5.81860900e-01 -5.98035455e-01 -9.63191330e-01
1.45138398e-01 -8.62844110e-01 -3.05721313e-02 -1.27346039e+00
-1.39009011e+00 4.58375812e-01 2.26443201e-01 -1.26997221e+00
-1.66047037e-01 -3.06029528e-01 -4.59856719e-01 9.17126417e-01
-1.27260172e+00 -1.67902410e+00 -6.75877213e-01 7.49984384e-01
4.17127818e-01 -4.37770113e-02 8.92094254e-01 2.49485765e-02
-1.54285237e-01 4.74485457e-01 -1.96145877e-01 -5.04191756e-01
5.32908797e-01 -1.14676178e+00 5.56964934e-01 5.91287732e-01
-2.23824367e-01 3.32822174e-01 5.26879549e-01 -5.56069374e-01
-1.79762542e+00 -9.98667896e-01 -1.43454075e-02 -6.57462597e-01
4.50569130e-02 -4.76167560e-01 -3.95724565e-01 6.11923397e-01
5.93953431e-01 3.69576998e-02 2.01436698e-01 -4.00103867e-01
-2.17774585e-01 -3.96981597e-01 -1.67374599e+00 7.01657951e-01
1.72760010e+00 -4.40641522e-01 -9.18735564e-02 6.26017377e-02
8.65243137e-01 -5.87215066e-01 -1.15693200e+00 6.63299680e-01
4.73348022e-01 -1.21708393e+00 1.11267471e+00 -2.87083864e-01
4.82020766e-01 -4.76500362e-01 -3.72026086e-01 -1.56784117e+00
-4.01595503e-01 -6.93103731e-01 8.53919685e-02 1.52550936e+00
-3.27849030e-01 -4.95443761e-01 8.18399251e-01 5.02793133e-01
-3.90642792e-01 -8.55736017e-01 -8.42122972e-01 -8.20047319e-01
2.90294904e-02 -2.88118362e-01 9.75681663e-01 7.95915008e-01
-7.92005479e-01 3.91548052e-02 -6.56611562e-01 3.54541764e-02
7.62994111e-01 4.09747064e-01 1.08366227e+00 -8.82210314e-01
-1.30084947e-01 -4.35203910e-01 -7.81663299e-01 -1.26058292e+00
-1.78681239e-02 -5.80394745e-01 -3.67798537e-01 -1.56288362e+00
2.91409016e-01 -9.40060258e-01 7.44526833e-02 2.78043598e-01
5.12021296e-02 6.35030150e-01 1.54344112e-01 1.32430583e-01
-1.57925457e-01 9.09591079e-01 2.00004864e+00 -1.27377898e-01
1.10731637e-02 -3.34291905e-01 -9.39913511e-01 4.48994011e-01
8.25451016e-01 -2.76277095e-01 -8.58781576e-01 -6.53882921e-01
1.81593075e-01 -1.20414877e-02 5.23753643e-01 -9.52265441e-01
-3.31490397e-01 -5.31625688e-01 5.02293527e-01 -5.07721066e-01
5.98600626e-01 -9.83568907e-01 5.84348142e-01 -2.00062916e-02
-8.78523886e-02 8.35997388e-02 1.52921528e-01 5.55458844e-01
5.38027193e-03 3.35866630e-01 9.71086979e-01 -2.11824551e-01
-5.48640907e-01 6.69260383e-01 -1.11197278e-01 1.45350605e-01
1.23730612e+00 -5.33270478e-01 -2.11340263e-01 -7.11012900e-01
-3.15500408e-01 -2.53436416e-01 1.27254939e+00 3.72883022e-01
6.29122257e-01 -1.65838552e+00 -5.80773056e-01 3.93619001e-01
1.09147899e-01 2.06009611e-01 4.46001232e-01 2.84937888e-01
-5.48470795e-01 1.36494681e-01 -8.53802025e-01 -8.52186799e-01
-9.79515612e-01 5.59532344e-01 3.25097710e-01 -2.66845766e-02
-7.33831584e-01 4.90563333e-01 5.51588953e-01 -3.16801518e-01
-1.48601502e-01 -4.44936216e-01 2.23466009e-01 -4.47936624e-01
-1.22183403e-02 1.94369271e-01 2.85669719e-03 -4.16877449e-01
-2.25985751e-01 1.13277495e+00 2.54859388e-01 -1.02673387e-02
1.29818559e+00 -5.27442396e-01 -1.58676393e-02 2.09043160e-01
8.52412820e-01 1.96059033e-01 -1.63450718e+00 3.25154334e-01
-1.00580192e+00 -8.49506259e-01 4.93917577e-02 -7.68130898e-01
-1.40691483e+00 6.72792792e-01 6.09071732e-01 1.51840761e-01
1.54587770e+00 1.11433394e-01 6.62465036e-01 -2.09579170e-01
8.30184042e-01 -6.18413568e-01 1.14401475e-01 9.31759849e-02
1.07117343e+00 -9.92860258e-01 1.24194205e-01 -7.37440646e-01
-5.63934147e-01 6.70351088e-01 6.89245999e-01 -1.86695516e-01
5.45968831e-01 3.16015482e-01 -4.83539850e-02 -2.25636795e-01
-4.13608462e-01 1.46218643e-01 1.40533641e-01 1.12624860e+00
2.07324252e-01 1.03920154e-01 -1.01048231e-01 1.66027881e-02
-2.39664063e-01 5.95907122e-02 5.72225332e-01 7.15620399e-01
1.34477749e-01 -1.30664361e+00 -1.96384698e-01 2.91425437e-01
-1.38930038e-01 2.92508304e-02 -2.12163985e-01 1.00938809e+00
1.56998500e-01 6.25934601e-01 5.13257869e-02 -4.00995791e-01
4.09170359e-01 -3.04967791e-01 1.00531280e+00 -4.79212195e-01
-3.48592073e-01 -2.30182722e-01 -3.43528949e-02 -5.44726193e-01
-9.17566240e-01 -3.42782736e-01 -6.92768395e-01 -4.87479657e-01
4.46406715e-02 -5.24585426e-01 5.30500412e-01 6.04623318e-01
6.23269796e-01 4.22087789e-01 8.42010021e-01 -9.04270172e-01
-2.24604145e-01 -6.86741173e-01 -7.51077771e-01 6.38085067e-01
3.29056159e-02 -9.55573559e-01 -9.08959955e-02 -2.99593825e-02] | [12.270013809204102, -0.5684561133384705] |
ff43f3bd-b74f-4fbd-a19b-9b02d249071c | adversarial-amendment-is-the-only-force | 2305.10766 | null | https://arxiv.org/abs/2305.10766v1 | https://arxiv.org/pdf/2305.10766v1.pdf | Adversarial Amendment is the Only Force Capable of Transforming an Enemy into a Friend | Adversarial attack is commonly regarded as a huge threat to neural networks because of misleading behavior. This paper presents an opposite perspective: adversarial attacks can be harnessed to improve neural models if amended correctly. Unlike traditional adversarial defense or adversarial training schemes that aim to improve the adversarial robustness, the proposed adversarial amendment (AdvAmd) method aims to improve the original accuracy level of neural models on benign samples. We thoroughly analyze the distribution mismatch between the benign and adversarial samples. This distribution mismatch and the mutual learning mechanism with the same learning ratio applied in prior art defense strategies is the main cause leading the accuracy degradation for benign samples. The proposed AdvAmd is demonstrated to steadily heal the accuracy degradation and even leads to a certain accuracy boost of common neural models on benign classification, object detection, and segmentation tasks. The efficacy of the AdvAmd is contributed by three key components: mediate samples (to reduce the influence of distribution mismatch with a fine-grained amendment), auxiliary batch norm (to solve the mutual learning mechanism and the smoother judgment surface), and AdvAmd loss (to adjust the learning ratios according to different attack vulnerabilities) through quantitative and ablation experiments. | ['Zhongxue Gan', 'Tao Chen', 'Chong Yu'] | 2023-05-18 | null | null | null | null | ['adversarial-defense'] | ['adversarial'] | [ 2.31593624e-01 2.40293071e-01 2.15672374e-01 -1.19386919e-01
-5.42391002e-01 -5.85703313e-01 7.43416488e-01 -1.91380709e-01
-4.14614916e-01 6.34063005e-01 -5.72758764e-02 -2.54945755e-01
-1.71661992e-02 -8.54637325e-01 -8.13121498e-01 -9.44433868e-01
1.73282832e-01 -7.78893456e-02 3.41322303e-01 -5.44783115e-01
1.42218262e-01 4.80133235e-01 -9.87287700e-01 1.77533284e-01
1.22707915e+00 9.98650849e-01 -3.59151155e-01 4.42437381e-01
-1.67928517e-01 7.58614719e-01 -9.70403492e-01 -8.37029994e-01
6.17976248e-01 -2.40618691e-01 -3.50803614e-01 -4.78289366e-01
3.83611888e-01 -4.14090365e-01 -5.49482882e-01 1.56525207e+00
7.65694797e-01 -1.41334116e-01 8.79208267e-01 -1.39085686e+00
-9.35726166e-01 7.36808300e-01 -6.23821974e-01 2.13507056e-01
-1.43985063e-01 4.42184269e-01 3.64552528e-01 -5.43364167e-01
1.19773678e-01 1.38158321e+00 7.95071542e-01 9.50674832e-01
-6.96387053e-01 -1.02970552e+00 3.46443117e-01 1.65501967e-01
-1.10915089e+00 -1.81802049e-01 9.48734283e-01 -4.26418006e-01
2.77852863e-01 3.82823735e-01 2.58198142e-01 1.51097345e+00
6.21423602e-01 5.21977067e-01 1.01349938e+00 -2.39724189e-01
9.68217552e-02 3.17684889e-01 5.01622632e-02 3.60611200e-01
1.94141164e-01 4.65296239e-01 4.90429578e-03 -1.40330613e-01
6.98183060e-01 -1.95786819e-01 -2.57130712e-01 1.00643881e-01
-5.15259564e-01 7.98564672e-01 7.47807801e-01 2.24067658e-01
-2.71085292e-01 4.88505550e-02 6.41144276e-01 3.85821074e-01
5.31116605e-01 7.79930532e-01 -3.76174599e-01 2.15369195e-01
-3.45026672e-01 3.01058084e-01 5.31771719e-01 5.09846151e-01
2.61503104e-02 7.67869115e-01 -4.63899910e-01 8.24288368e-01
2.23128170e-01 6.67519987e-01 5.84683299e-01 -4.47886229e-01
4.22058553e-01 5.46724558e-01 -2.94780821e-01 -1.23178792e+00
-1.26371309e-01 -9.36327219e-01 -1.08130896e+00 6.32342219e-01
5.42272270e-01 -3.97926629e-01 -9.67052281e-01 1.91374946e+00
4.17627424e-01 1.99029505e-01 2.21978918e-01 7.34018981e-01
7.52310872e-01 5.01960456e-01 2.04214871e-01 5.73442690e-02
9.79058444e-01 -8.07654738e-01 -8.66683781e-01 -3.16746503e-01
2.37383336e-01 -7.64481544e-01 9.83594060e-01 3.05295944e-01
-9.41625714e-01 -5.06031871e-01 -1.23844767e+00 5.07768631e-01
-4.91683632e-01 -2.57792026e-01 2.93379009e-01 1.07909000e+00
-5.54418802e-01 6.70123041e-01 -4.53398675e-01 3.24336022e-01
7.45123684e-01 2.80015856e-01 -2.61975467e-01 1.22029729e-01
-1.72381926e+00 9.54220533e-01 2.70206541e-01 3.20251554e-01
-9.61271524e-01 -1.04911864e+00 -6.20365024e-01 -9.31878611e-02
1.23199150e-01 -3.68099332e-01 6.78410590e-01 -1.33107769e+00
-1.50018215e+00 5.80906391e-01 7.38336444e-01 -7.58429527e-01
1.03633940e+00 -3.87229711e-01 -4.93902922e-01 -2.02128559e-01
-3.38932574e-01 3.68950248e-01 1.31560576e+00 -1.27956629e+00
-1.85470693e-02 -4.46143061e-01 -4.96743321e-02 1.53655916e-01
-8.33999157e-01 -1.11789785e-01 3.76023129e-02 -1.23624766e+00
-1.06275931e-01 -8.64919305e-01 -2.67706305e-01 -1.01409256e-01
-3.89146358e-01 2.33139828e-01 9.85646844e-01 -7.79333234e-01
9.77824867e-01 -2.22490907e+00 2.97039226e-02 2.73829430e-01
1.73083156e-01 9.45644498e-01 -2.02695787e-01 -1.09993167e-01
-2.85915971e-01 1.84231102e-01 -3.17046702e-01 1.33717060e-01
-1.16496958e-01 1.10875368e-01 -6.00748539e-01 5.45240581e-01
4.11919564e-01 8.09633434e-01 -6.82934165e-01 -7.94546753e-02
1.08524337e-01 6.08043611e-01 -4.13225234e-01 2.70298630e-01
-1.08229026e-01 5.21822274e-01 -4.41919178e-01 4.90499228e-01
1.03808284e+00 4.96772736e-01 -3.63373935e-01 -4.00986969e-01
4.87821072e-01 -4.98204231e-01 -1.04511809e+00 8.46274257e-01
-1.19982123e-01 2.72281855e-01 1.01156771e-01 -8.09358716e-01
1.21112025e+00 1.74537763e-01 6.80558309e-02 -7.60720432e-01
5.31928957e-01 1.84604749e-01 2.81168967e-01 -3.54034513e-01
8.32969546e-02 -2.04543725e-01 -6.66087195e-02 -1.32076263e-01
-2.52843738e-01 -1.83808461e-01 -7.18872130e-01 6.56684935e-02
1.08535147e+00 -2.21278951e-01 -8.49428475e-02 -3.11768264e-01
7.61659682e-01 -2.95740217e-01 7.26871371e-01 6.53346121e-01
-5.49628496e-01 3.21539909e-01 5.17775536e-01 -3.24446678e-01
-1.05099010e+00 -1.04340065e+00 -1.58825248e-01 8.83393764e-01
2.59528071e-01 3.78866553e-01 -9.83381450e-01 -9.76795793e-01
7.30852857e-02 6.75632358e-01 -9.17546332e-01 -9.96702611e-01
-4.37981635e-01 -9.11696732e-01 1.14439833e+00 6.15810037e-01
9.08605874e-01 -9.45398152e-01 9.07380655e-02 -2.73683965e-01
3.74582112e-01 -9.23243463e-01 -4.18944567e-01 1.44280940e-01
-5.71318507e-01 -9.43443000e-01 -7.54715919e-01 -5.77315331e-01
7.23144054e-01 -3.17668229e-01 4.99693811e-01 4.91532423e-02
-1.99310377e-01 -7.64623582e-02 -3.96630198e-01 -7.67628372e-01
-8.70390952e-01 -1.34511501e-01 1.66696474e-01 1.68237135e-01
-1.60525352e-01 -5.84527731e-01 -4.01204467e-01 3.69348168e-01
-1.01974404e+00 -4.52903122e-01 5.27194560e-01 1.02457738e+00
2.43017793e-01 2.58296251e-01 8.99155974e-01 -1.13123322e+00
8.84782195e-01 -6.68461084e-01 -4.77768779e-01 7.70241022e-02
-8.79784048e-01 -2.72119185e-04 9.39627111e-01 -9.16605353e-01
-1.12438810e+00 -4.03822660e-01 -6.89311087e-01 -9.18412685e-01
6.98341010e-03 2.34781634e-02 -4.45631236e-01 -5.50685525e-01
9.48702216e-01 1.30809531e-01 1.74839064e-01 -1.82593867e-01
4.30812031e-01 2.64551580e-01 6.31199837e-01 -4.19216335e-01
1.09972489e+00 1.39012456e-01 4.48679477e-02 -3.78277928e-01
-7.27820218e-01 1.97917104e-01 -2.93572426e-01 -2.91206002e-01
7.12599516e-01 -5.73525608e-01 -4.51184392e-01 1.14299715e+00
-8.98964405e-01 -2.83852816e-01 -2.60517359e-01 2.27280632e-01
-3.87613218e-05 3.51895928e-01 -7.58632421e-01 -8.16819370e-01
-6.69030786e-01 -1.17212093e+00 3.92180383e-01 3.19764942e-01
1.32665604e-01 -1.03778887e+00 -2.09821895e-01 3.61665338e-01
7.45164454e-01 8.35873604e-01 9.36065257e-01 -1.00708914e+00
7.67689645e-02 -5.13840795e-01 -1.69968665e-01 9.13987398e-01
3.26914564e-02 1.65262967e-01 -1.12971818e+00 -3.36039305e-01
4.32148665e-01 -2.60986179e-01 7.85001457e-01 2.37263098e-01
1.04044735e+00 -6.44057870e-01 1.35554736e-02 7.25597799e-01
1.17745197e+00 5.36176383e-01 9.10731375e-01 3.82087797e-01
9.71524417e-01 5.39480627e-01 4.77956176e-01 7.02336580e-02
-3.54713291e-01 3.95188570e-01 1.05409157e+00 -2.41297841e-01
-2.23230898e-01 -2.61690523e-02 3.50125015e-01 5.76009393e-01
2.69948840e-01 -2.38211542e-01 -8.19645286e-01 1.81656510e-01
-1.50924957e+00 -8.61245096e-01 -1.80889398e-01 2.01421738e+00
7.84246802e-01 6.27373278e-01 -1.98608518e-01 2.80398667e-01
9.09285426e-01 2.26744324e-01 -9.30410922e-01 -5.31045437e-01
-3.59899759e-01 7.23211095e-02 6.81043506e-01 4.68383193e-01
-1.13339829e+00 9.39213991e-01 6.17013264e+00 1.37561250e+00
-1.22403848e+00 1.82107449e-01 8.78415048e-01 6.64019436e-02
-2.50844121e-01 -4.87487674e-01 -6.14162743e-01 5.81194997e-01
7.21149087e-01 1.35579214e-01 2.44336516e-01 9.61594701e-01
-1.51621446e-01 5.68547189e-01 -5.47859550e-01 5.84322572e-01
-1.25743272e-02 -1.22453868e+00 2.92155415e-01 4.07209322e-02
8.99211586e-01 -2.60295719e-01 5.83771706e-01 4.65209126e-01
2.73521006e-01 -1.05963469e+00 7.40808308e-01 5.67338109e-01
6.33725405e-01 -1.06125283e+00 1.05639708e+00 2.79593080e-01
-7.47730196e-01 -3.16927761e-01 -2.03156784e-01 1.85078651e-01
-2.62959093e-01 5.02390802e-01 -5.48037827e-01 5.06663382e-01
5.28886735e-01 3.66544247e-01 -6.95082903e-01 5.70654690e-01
-2.42008418e-01 8.00241947e-01 1.84540208e-02 1.42983750e-01
3.02366525e-01 -2.02848696e-05 9.55033720e-01 8.46444130e-01
-2.28982508e-01 -1.24813065e-01 -1.69610173e-01 9.00194764e-01
-1.14450417e-01 -3.55583020e-02 -5.87700427e-01 5.26223704e-02
6.46722198e-01 1.12254965e+00 -3.91362101e-01 3.98382694e-02
1.88352346e-01 9.23443019e-01 1.86290368e-01 2.53199130e-01
-1.19227552e+00 -4.79355335e-01 6.63383961e-01 3.92295755e-02
-3.61093283e-02 2.95456111e-01 -6.58477068e-01 -6.38960481e-01
-2.07525529e-02 -1.19687736e+00 2.44312212e-01 -3.22128087e-01
-1.70554924e+00 8.95060956e-01 -3.40780854e-01 -1.20565927e+00
7.97709227e-02 -5.29023886e-01 -1.07841182e+00 1.05316174e+00
-1.17029297e+00 -1.10177803e+00 -2.34507516e-01 7.54247010e-01
3.57243776e-01 -7.12428808e-01 5.81335902e-01 5.42020261e-01
-1.00027585e+00 1.29816353e+00 4.85612936e-02 3.46044332e-01
7.42620289e-01 -1.03765035e+00 4.01044071e-01 1.16197085e+00
-5.17923236e-01 5.17754197e-01 8.13751996e-01 -7.69289851e-01
-1.06375849e+00 -1.41849804e+00 -9.56344754e-02 -4.57888663e-01
6.67041719e-01 -2.01450109e-01 -1.26190531e+00 2.90103465e-01
-2.64853556e-02 1.41639680e-01 3.90717566e-01 -3.94467056e-01
-6.37966812e-01 -3.42121184e-01 -1.64542115e+00 7.36339629e-01
6.47911131e-01 -3.87538612e-01 -4.60678756e-01 2.04867199e-01
1.18992734e+00 -4.67840105e-01 -8.82149339e-01 7.70090938e-01
4.99547869e-01 -7.01993346e-01 1.05850339e+00 -6.83741212e-01
5.37207246e-01 -8.33287165e-02 -1.24494918e-01 -1.29276562e+00
-2.40060955e-01 -5.06667733e-01 -8.18009228e-02 1.60418999e+00
2.65893966e-01 -9.33445275e-01 6.59693599e-01 5.08667171e-01
-2.27556005e-01 -1.09297073e+00 -9.53364193e-01 -8.39895308e-01
6.15234971e-01 -3.00067633e-01 6.45252764e-01 1.22149217e+00
-5.36792457e-01 -5.12458421e-02 -4.69336033e-01 3.08222443e-01
6.89719200e-01 -7.43600309e-01 6.91566885e-01 -8.32944095e-01
-2.21547142e-01 -7.10688949e-01 -4.35825974e-01 -2.74945527e-01
1.49342328e-01 -6.87248409e-01 -7.07286447e-02 -8.46882284e-01
-3.27401698e-01 -7.26525724e-01 -4.13350970e-01 3.75856251e-01
-6.97005093e-01 3.40719193e-01 2.00414896e-01 2.64899254e-01
5.43707609e-02 7.30226278e-01 1.31443942e+00 -3.69657904e-01
-6.45140558e-02 2.56991595e-01 -9.07357752e-01 8.86877298e-01
7.98731744e-01 -5.62988758e-01 -4.83346015e-01 -2.58859336e-01
-1.83796436e-02 -1.72743723e-01 3.85818928e-01 -1.02437532e+00
-7.49412505e-03 -5.63038960e-02 5.50769329e-01 -4.56991196e-02
1.52376890e-01 -8.44643354e-01 1.79236010e-02 8.92201424e-01
-4.40362632e-01 -1.43806010e-01 4.42414224e-01 7.53598690e-01
-5.11997938e-02 -7.62403160e-02 1.31797910e+00 9.58917066e-02
-3.88994068e-01 3.02915573e-01 -1.13818064e-01 1.07592352e-01
1.26914263e+00 -1.49994582e-01 -5.42764544e-01 -1.53419793e-01
-5.16500115e-01 1.50467083e-01 9.50838625e-02 6.20873034e-01
5.52897096e-01 -1.38720953e+00 -7.96354890e-01 3.46567363e-01
-2.00203538e-01 -2.01150090e-01 5.33365071e-01 5.79544008e-01
-3.89577031e-01 -1.05504334e-01 -5.46653628e-01 -2.56458640e-01
-1.24664640e+00 6.85945749e-01 7.09202051e-01 -4.04395849e-01
-3.45313162e-01 1.26080298e+00 3.44790965e-01 -5.35583198e-01
4.87731576e-01 2.45974898e-01 -4.39101547e-01 -1.46051079e-01
4.88293737e-01 5.68419218e-01 9.20011625e-02 -4.31497544e-01
-2.65503824e-01 2.47154146e-01 -3.75599831e-01 3.13610315e-01
9.87053692e-01 1.65457815e-01 4.06820998e-02 1.21223023e-02
8.95287693e-01 2.14764252e-01 -1.43933737e+00 -7.84185678e-02
-5.75464547e-01 -3.67107779e-01 3.18472460e-02 -1.04784203e+00
-1.42604971e+00 6.76164150e-01 9.51501489e-01 3.09683353e-01
1.14401317e+00 -4.65405464e-01 8.53221357e-01 -1.08191751e-01
-7.52121359e-02 -6.46984041e-01 3.89925122e-01 4.34449077e-01
1.03879070e+00 -1.11304557e+00 -1.77631617e-01 -3.99466395e-01
-7.93904781e-01 8.42649877e-01 9.88030255e-01 -4.63880748e-01
5.97318292e-01 4.89303768e-01 2.45013595e-01 8.34209174e-02
-2.09011868e-01 5.16720772e-01 4.64271218e-01 7.31642365e-01
-5.69860972e-02 -1.25886977e-01 -3.57535392e-01 1.00218105e+00
-1.55788660e-01 -6.35253012e-01 1.37072742e-01 5.97454131e-01
-2.04372063e-01 -9.59762990e-01 -5.75988650e-01 4.14440244e-01
-7.02188969e-01 -7.23794997e-02 -5.03110588e-01 5.66778898e-01
4.53945309e-01 7.76451409e-01 -2.99336106e-01 -8.99296463e-01
4.63757724e-01 -1.43230706e-01 3.14748555e-01 -2.11335823e-01
-1.05387521e+00 -2.83971131e-01 -3.38559419e-01 -5.01144469e-01
1.94580957e-01 -2.22639188e-01 -1.05772018e+00 -3.35261732e-01
-5.06407380e-01 -2.43568560e-03 6.01095796e-01 1.02778697e+00
9.17147398e-02 1.16781807e+00 9.64667141e-01 -5.13794899e-01
-1.12590694e+00 -9.85544682e-01 -3.83075893e-01 5.63821733e-01
1.95620149e-01 -6.15336537e-01 -7.12586343e-01 -2.76618958e-01] | [5.560213565826416, 7.9413065910339355] |
571279a1-d76b-4587-9a12-7772705f1bab | bayesian-renormalization | 2305.10491 | null | https://arxiv.org/abs/2305.10491v2 | https://arxiv.org/pdf/2305.10491v2.pdf | Bayesian Renormalization | In this note we present a fully information theoretic approach to renormalization inspired by Bayesian statistical inference, which we refer to as Bayesian Renormalization. The main insight of Bayesian Renormalization is that the Fisher metric defines a correlation length that plays the role of an emergent RG scale quantifying the distinguishability between nearby points in the space of probability distributions. This RG scale can be interpreted as a proxy for the maximum number of unique observations that can be made about a given system during a statistical inference experiment. The role of the Bayesian Renormalization scheme is subsequently to prepare an effective model for a given system up to a precision which is bounded by the aforementioned scale. In applications of Bayesian Renormalization to physical systems, the emergent information theoretic scale is naturally identified with the maximum energy that can be probed by current experimental apparatus, and thus Bayesian Renormalization coincides with ordinary renormalization. However, Bayesian Renormalization is sufficiently general to apply even in circumstances in which an immediate physical scale is absent, and thus provides an ideal approach to renormalization in data science contexts. To this end, we provide insight into how the Bayesian Renormalization scheme relates to existing methods for data compression and data generation such as the information bottleneck and the diffusion learning paradigm. | ['Alexander G. Stapleton', 'Marc S. Klinger', 'David S. Berman'] | 2023-05-17 | null | null | null | null | ['data-compression'] | ['time-series'] | [ 5.00093400e-01 3.14852834e-01 4.35828418e-02 -2.75835693e-01
-1.36907488e-01 -2.99616307e-01 9.46600258e-01 4.02400851e-01
-4.20196235e-01 6.24729276e-01 3.72056395e-01 -2.62976319e-01
-5.14104426e-01 -1.18821621e+00 -5.09293735e-01 -1.34996748e+00
-9.57856551e-02 4.84383762e-01 1.60920039e-01 -5.20923793e-01
7.32442558e-01 4.75948215e-01 -1.79750967e+00 -1.18426114e-01
6.82828546e-01 7.96082616e-01 1.88595816e-01 8.02711964e-01
-3.74276843e-03 1.30511969e-01 -2.78809309e-01 6.87480867e-02
3.64895344e-01 -7.89073944e-01 -9.02952373e-01 -2.66583681e-01
-7.90650398e-02 -3.15756142e-01 -4.95317906e-01 1.35192490e+00
3.14532816e-01 3.76783043e-01 9.65595841e-01 -4.50050205e-01
-6.33931577e-01 8.00252616e-01 -1.55944392e-01 4.66095567e-01
2.10346013e-01 -6.20822832e-02 1.06162035e+00 -2.59762019e-01
5.88895679e-01 1.40937614e+00 2.69016027e-01 4.60400790e-01
-1.58565462e+00 -1.32575825e-01 -3.71500492e-01 -1.44912124e-01
-1.14661050e+00 -4.75116342e-01 7.33530283e-01 -5.46734154e-01
3.69610429e-01 2.61089839e-02 6.82367146e-01 5.37882507e-01
9.02047873e-01 -1.36864260e-01 1.20418632e+00 -1.03580189e+00
9.08302784e-01 -2.41428420e-01 5.71601510e-01 4.62833107e-01
4.53229845e-01 3.70646954e-01 -7.07085133e-01 -2.39209026e-01
7.76553154e-01 -2.00430870e-01 -2.57906675e-01 -3.44549358e-01
-1.35157967e+00 1.01341152e+00 3.40764850e-01 3.53287756e-01
-2.42271140e-01 3.22767168e-01 2.36061916e-01 5.22383392e-01
4.87944037e-01 4.74256665e-01 -3.29204768e-01 7.52007961e-02
-4.69774455e-01 4.95698750e-01 8.40449631e-01 4.99731600e-01
8.47715199e-01 -3.78976911e-01 6.32348955e-02 1.95111334e-01
4.22086507e-01 6.41866386e-01 2.86132812e-01 -1.21304834e+00
-2.24330887e-01 3.14424008e-01 1.95767596e-01 -4.49046522e-01
-4.38438267e-01 -2.15812996e-01 -9.61423039e-01 3.29453707e-01
7.76600003e-01 5.43597415e-02 -5.16389310e-01 1.98804915e+00
3.20113599e-01 -2.91943282e-01 2.33928308e-01 7.38054931e-01
5.39901137e-01 6.06614769e-01 -2.72303909e-01 -4.93359357e-01
1.50641179e+00 6.73388690e-02 -5.89574158e-01 2.59272069e-01
5.40327251e-01 -4.42145497e-01 8.53542984e-01 2.04832375e-01
-1.06229889e+00 -2.45663449e-01 -1.15487421e+00 -1.32969871e-01
-4.42819804e-01 -6.51249051e-01 8.58866334e-01 8.09495807e-01
-1.19629037e+00 1.29651928e+00 -7.16344118e-01 -6.42043114e-01
-1.96207121e-01 2.66667187e-01 2.45881383e-03 2.59847581e-01
-1.28539622e+00 8.63750994e-01 6.12619698e-01 -2.26655062e-02
-3.35359752e-01 -5.14621973e-01 -3.29345971e-01 6.86761364e-02
6.86598336e-03 -8.38481963e-01 1.20937335e+00 -4.91760373e-01
-1.61104596e+00 7.90642500e-01 -2.78037816e-01 -4.70024914e-01
1.34545982e-01 3.39643627e-01 3.73700224e-02 3.77686352e-01
-2.39971310e-01 5.58125794e-01 9.05748367e-01 -8.03574860e-01
3.83553728e-02 -4.63210791e-01 4.23249692e-01 -2.75996160e-02
1.50842845e-01 -4.34550904e-02 3.64963263e-01 -2.46486709e-01
9.06302810e-01 -1.06361818e+00 -2.89215803e-01 -2.64722168e-01
3.93690483e-04 -2.28996053e-01 2.06020862e-01 -2.15371028e-01
7.26163685e-01 -1.93948972e+00 3.49048138e-01 3.47892016e-01
5.84383130e-01 -3.36624444e-01 3.10509741e-01 8.16382408e-01
-2.07530171e-01 1.37434468e-01 -3.62576902e-01 2.93218851e-01
1.61267385e-01 -1.08634289e-02 -4.12853807e-01 6.39706194e-01
-7.75103271e-02 8.31375599e-01 -9.34696078e-01 -2.94528101e-02
2.27539092e-01 3.90495598e-01 -6.85734034e-01 -1.06507398e-01
-1.61719441e-01 8.32449317e-01 -6.19611084e-01 1.09809088e-02
8.82190645e-01 -2.95454860e-01 1.53956309e-01 -3.45120654e-02
-3.88247311e-01 7.16240823e-01 -1.04803395e+00 1.68943536e+00
3.21960971e-02 4.93815362e-01 -8.15643668e-02 -1.04641545e+00
7.78295696e-01 -3.46193612e-02 4.09884363e-01 -3.96041423e-01
3.12372565e-01 -6.87433630e-02 4.66411531e-01 -6.86949492e-02
6.96267188e-01 -6.00012779e-01 -2.88945258e-01 1.02169836e+00
-7.01716989e-02 -3.20314080e-01 2.64290094e-01 4.22119111e-01
1.24995506e+00 -2.00716004e-01 5.12003958e-01 -1.20807123e+00
4.33113784e-01 -5.08365333e-01 1.99945316e-01 1.27036679e+00
-1.95612073e-01 3.15139711e-01 6.88355207e-01 -3.54065478e-01
-1.36230779e+00 -1.40579486e+00 -8.30991685e-01 1.01376724e+00
3.23570848e-01 -5.73485911e-01 -1.13331807e+00 4.62192222e-02
-2.71482449e-02 4.62878585e-01 -7.80829132e-01 -4.05177772e-01
-1.99863061e-01 -1.08106208e+00 4.68869172e-02 -1.84014812e-01
4.06141400e-01 -1.06496847e+00 -6.47590220e-01 6.44051433e-02
-2.86544440e-03 -6.83445990e-01 -1.49797453e-02 2.79178262e-01
-1.22445619e+00 -8.77197921e-01 -2.54545003e-01 3.34304273e-02
5.80571532e-01 2.38249332e-01 9.67763245e-01 1.05843786e-02
-4.06228453e-01 2.93280274e-01 -4.06980723e-01 -2.11009547e-01
-1.11962116e+00 6.94609247e-03 3.90797168e-01 -2.53819913e-01
4.88608301e-01 -8.64143312e-01 -8.82437289e-01 -1.40386134e-01
-1.13670051e+00 1.55420899e-01 2.79453725e-01 4.09560800e-01
-2.86752265e-02 1.05651952e-01 6.00563347e-01 -5.48620403e-01
6.36197984e-01 -3.25987786e-01 -7.81180084e-01 7.56136999e-02
-4.97231930e-01 8.27125013e-01 3.84115070e-01 9.12753716e-02
-9.57957566e-01 -3.13162833e-01 1.09026313e-01 7.51800418e-01
-2.14986712e-01 4.00618970e-01 8.94735083e-02 -1.07643805e-01
7.01758862e-01 5.12844771e-02 5.60759082e-02 -4.40708876e-01
6.19665265e-01 6.44723356e-01 2.12249801e-01 -7.43385434e-01
4.90140587e-01 7.35418379e-01 7.19736516e-01 -1.05784357e+00
-9.83693659e-01 -2.63883531e-01 -1.17578125e+00 -1.89019650e-01
8.69543791e-01 -4.49966133e-01 -1.17802835e+00 4.57334757e-01
-1.04220605e+00 9.49010104e-02 -7.72730350e-01 5.33917487e-01
-8.26281607e-01 3.60114962e-01 -6.88357353e-01 -1.18692362e+00
-1.42017692e-01 -9.45095181e-01 1.00866044e+00 1.43332377e-01
1.17191393e-02 -1.12199247e+00 1.18469141e-01 -7.21307993e-02
5.34767449e-01 1.37512043e-01 1.25770378e+00 -1.73067287e-01
-6.16199911e-01 -1.13204271e-01 -2.40510374e-01 -1.34260319e-02
-2.09002960e-02 -5.34899198e-02 -9.91171837e-01 -3.06817442e-01
6.19946241e-01 3.40075046e-02 1.14473796e+00 6.30735338e-01
7.41956115e-01 2.06261516e-01 -1.53023720e-01 1.29218176e-01
1.24917197e+00 -1.27543777e-01 5.71798027e-01 -2.85980225e-01
3.69870305e-01 8.19787443e-01 -4.79373261e-02 4.85821545e-01
9.10537615e-02 5.06079376e-01 -8.72195810e-02 6.33827806e-01
4.50942889e-02 -5.84892221e-02 1.41347617e-01 1.32832837e+00
-2.94125050e-01 4.77275513e-02 -6.95539236e-01 -1.49092466e-01
-1.71872044e+00 -1.07732761e+00 -3.33882511e-01 2.81890178e+00
8.17311347e-01 2.66845673e-01 -8.35461095e-02 1.62100345e-01
1.06508482e+00 6.81700483e-02 -5.68046510e-01 -4.26533222e-01
-6.46446869e-02 1.86405241e-01 7.00621247e-01 6.85145617e-01
-8.40609491e-01 4.91560847e-01 7.49165058e+00 7.01741934e-01
-7.16337562e-01 1.79962650e-01 2.36432612e-01 1.83085188e-01
-2.67088443e-01 4.61811572e-01 -7.41629303e-01 4.89575684e-01
1.49843061e+00 -5.04252791e-01 8.18336964e-01 2.59771526e-01
4.33007956e-01 -6.47837639e-01 -1.22362518e+00 6.71484828e-01
-4.12837386e-01 -1.25745130e+00 -1.06390435e-02 6.87208295e-01
5.57099700e-01 1.54968321e-01 -3.45527798e-01 -2.20880821e-01
3.69364589e-01 -6.86395943e-01 4.18442219e-01 8.00671756e-01
5.27139306e-01 -6.09512806e-01 4.05894339e-01 5.81834793e-01
-8.67596686e-01 1.16693802e-01 -8.51977229e-01 -5.58430016e-01
2.22570434e-01 1.18958926e+00 -2.42783800e-01 3.05298805e-01
2.41259396e-01 5.07772505e-01 -3.16403955e-01 7.21844614e-01
-5.55359712e-03 5.17707944e-01 -5.91181457e-01 -6.01762757e-02
-1.04904816e-01 -8.17437232e-01 6.79101527e-01 7.03236997e-01
3.20836544e-01 2.95388788e-01 -2.35206321e-01 9.69479263e-01
-5.14326543e-02 -1.39506206e-01 -6.46382034e-01 -1.52863577e-01
2.82603592e-01 1.03138161e+00 -1.24620521e+00 -2.02687085e-01
-2.92847957e-02 5.91433406e-01 2.31585160e-01 7.04495609e-02
-2.77771115e-01 -3.57481867e-01 3.23670805e-01 1.68532506e-01
9.84479412e-02 -5.45381725e-01 -3.44367214e-02 -1.34484553e+00
-1.11000501e-01 -4.22985405e-02 -6.07843569e-04 -6.14856362e-01
-1.27587759e+00 -1.46508038e-01 3.42871606e-01 -7.80845463e-01
-1.85514554e-01 -6.95088327e-01 -4.17258352e-01 8.87682736e-01
-1.13303185e+00 -2.73116708e-01 8.66707414e-03 2.88565531e-02
-2.37501055e-01 2.59993076e-01 1.06022131e+00 -2.99614638e-01
-1.64492041e-01 -1.16830610e-01 9.44493115e-01 -4.96427029e-01
4.05414104e-01 -1.54912329e+00 8.32998514e-01 5.75082123e-01
-2.05166370e-01 9.40725744e-01 1.13903010e+00 -7.17068851e-01
-1.70071554e+00 -4.50991869e-01 7.67502308e-01 -3.75771463e-01
8.79929423e-01 -7.92275250e-01 -9.79783237e-01 9.12507027e-02
-5.72761595e-02 -1.80324912e-01 4.96091843e-01 2.14372829e-01
-2.64183104e-01 9.25169662e-02 -1.22996664e+00 5.40577352e-01
1.22385705e+00 -7.24711537e-01 -6.63892865e-01 6.31071031e-01
7.65794337e-01 1.52813084e-02 -1.02450800e+00 1.47113323e-01
6.63405061e-01 -1.09307313e+00 6.86334431e-01 -4.80129600e-01
2.90248305e-01 -8.25623348e-02 -1.60364702e-01 -1.07517898e+00
-5.04628599e-01 -1.04791033e+00 1.78760126e-01 8.75589788e-01
-4.21676934e-02 -1.01337767e+00 3.67433190e-01 8.23176146e-01
4.24517572e-01 -9.39870328e-02 -1.13141632e+00 -7.21040428e-01
6.83105588e-01 -4.09465939e-01 4.59187686e-01 5.91073334e-01
3.43604982e-01 3.09436649e-01 1.55457094e-01 -6.59641623e-02
1.08181477e+00 8.97010863e-02 2.29347527e-01 -1.82533193e+00
-3.58121067e-01 -4.70191211e-01 -5.81461251e-01 -1.25579441e+00
3.93528901e-02 -1.10450995e+00 1.89149946e-01 -9.12114739e-01
5.03363132e-01 -2.27939397e-01 -3.13108385e-01 -7.03420281e-01
2.26462007e-01 6.71257898e-02 -1.11358479e-01 5.42590499e-01
-2.75290966e-01 5.18348932e-01 1.03503704e+00 5.10353923e-01
-8.02103728e-02 -1.14955343e-01 -6.77033007e-01 7.30609894e-01
8.76162589e-01 -5.55208743e-01 -1.05359808e-01 3.27912897e-01
7.14639664e-01 1.16751067e-01 3.96047473e-01 -5.96784353e-01
1.35398820e-01 -5.36156967e-02 3.04814894e-02 -3.79709363e-01
1.24892451e-01 -3.06828111e-01 -1.03726402e-01 3.76631737e-01
-7.40207076e-01 -4.08152580e-01 -3.01175326e-01 7.49627829e-01
4.22696412e-01 -8.85350049e-01 9.51831460e-01 -3.14471394e-01
-6.50163367e-02 -7.76788592e-02 -7.76133835e-01 6.96972460e-02
4.99981344e-01 2.60777771e-01 -4.94780213e-01 -4.24390674e-01
-5.99454820e-01 -3.80962551e-01 8.07637632e-01 -3.88921890e-03
2.09737226e-01 -1.18954897e+00 -6.52430952e-01 2.16178626e-01
-2.41218903e-03 -3.74091029e-01 1.67318389e-01 7.03148842e-01
-3.97264600e-01 5.22007167e-01 8.12326595e-02 -5.93402445e-01
-6.50104225e-01 5.79995573e-01 1.90633267e-01 3.59094217e-02
-8.19852352e-01 1.78720981e-01 4.97802436e-01 -3.34033608e-01
-5.01005232e-01 -5.68940580e-01 2.56648004e-01 -2.08346829e-01
8.63869071e-01 4.52215105e-01 2.77232211e-02 -6.51825726e-01
-1.00031227e-01 7.20006287e-01 1.85860228e-02 -4.50660855e-01
1.00458086e+00 -5.00330329e-01 -6.28178000e-01 9.52014863e-01
9.47683752e-01 -5.69853410e-02 -1.14089692e+00 -2.82749295e-01
-1.24824099e-01 -2.06315801e-01 2.80924201e-01 -2.66126812e-01
-3.05471599e-01 9.12575483e-01 5.55067182e-01 1.16664422e+00
7.07536161e-01 4.90376621e-01 4.63345110e-01 6.97864294e-01
4.53704447e-01 -1.04517162e+00 -5.88785052e-01 5.39076149e-01
7.88565159e-01 -1.16669464e+00 1.89600408e-01 -2.74452776e-01
2.92706400e-01 1.42741680e+00 -3.91527027e-01 -4.35637951e-01
1.07199872e+00 6.60202727e-02 -6.34706795e-01 -3.16020727e-01
-6.19047880e-01 -2.74519503e-01 2.79592454e-01 1.80728987e-01
4.85768795e-01 2.77399123e-01 -6.22495115e-01 2.88511571e-02
-4.99472499e-01 -1.33808434e-01 9.08132553e-01 8.73950899e-01
-1.05383348e+00 -1.05409276e+00 -2.96230137e-01 4.27618206e-01
-3.35863829e-01 -2.38385946e-01 -2.11674601e-01 3.31690878e-01
-2.04345703e-01 1.06573868e+00 1.51884124e-01 5.44677638e-02
-2.54167050e-01 3.19958180e-01 9.04234231e-01 -5.25498331e-01
-4.28432561e-02 -1.79730266e-01 -2.99265712e-01 -2.49520451e-01
-7.18541086e-01 -9.13502455e-01 -1.35523558e+00 -8.15031409e-01
-4.90969032e-01 5.03605485e-01 9.22370374e-01 1.37744796e+00
2.25660175e-01 3.60413045e-01 4.51919883e-01 -1.15664661e+00
-7.28726089e-01 -9.42894101e-01 -9.47144747e-01 1.93998083e-01
4.97684240e-01 -6.26358986e-01 -9.00088549e-01 -9.78634208e-02] | [5.721335411071777, 4.771697998046875] |
af00bcaf-51d7-4f65-954e-38378334e845 | subspace-regularizers-for-few-shot-class-1 | 2110.07059 | null | https://arxiv.org/abs/2110.07059v2 | https://arxiv.org/pdf/2110.07059v2.pdf | Subspace Regularizers for Few-Shot Class Incremental Learning | Few-shot class incremental learning -- the problem of updating a trained classifier to discriminate among an expanded set of classes with limited labeled data -- is a key challenge for machine learning systems deployed in non-stationary environments. Existing approaches to the problem rely on complex model architectures and training procedures that are difficult to tune and re-use. In this paper, we present an extremely simple approach that enables the use of ordinary logistic regression classifiers for few-shot incremental learning. The key to this approach is a new family of subspace regularization schemes that encourage weight vectors for new classes to lie close to the subspace spanned by the weights of existing classes. When combined with pretrained convolutional feature extractors, logistic regression models trained with subspace regularization outperform specialized, state-of-the-art approaches to few-shot incremental image classification by up to 22% on the miniImageNet dataset. Because of its simplicity, subspace regularization can be straightforwardly extended to incorporate additional background information about the new classes (including class names and descriptions specified in natural language); these further improve accuracy by up to 2%. Our results show that simple geometric regularization of class representations offers an effective tool for continual learning. | ['Derry Tanti Wijaya', 'Jacob Andreas', 'Ekin Akyürek', 'Afra Feyza Akyürek'] | 2021-10-13 | subspace-regularizers-for-few-shot-class | https://openreview.net/forum?id=boJy41J-tnQ | https://openreview.net/pdf?id=boJy41J-tnQ | iclr-2022-4 | ['few-shot-class-incremental-learning'] | ['methodology'] | [ 4.33091432e-01 -1.21658668e-01 -2.95030236e-01 -6.36307001e-01
-7.42293537e-01 -4.51159149e-01 6.59913003e-01 1.02814078e-01
-6.08462632e-01 6.22130334e-01 1.82825346e-02 -1.39446519e-02
2.30458882e-02 -5.00210464e-01 -4.87376869e-01 -5.76780379e-01
-1.79450288e-01 4.32258159e-01 5.99384546e-01 -3.53681028e-01
1.91008046e-01 5.77188075e-01 -2.04410005e+00 4.09145325e-01
4.36976343e-01 6.68089509e-01 -1.13121338e-01 7.14564681e-01
-3.48384559e-01 8.84374738e-01 -2.23193929e-01 -5.93105219e-02
2.48312041e-01 -4.61464107e-01 -6.86119378e-01 2.74870485e-01
6.15177691e-01 -4.30251956e-01 -4.16620582e-01 8.48637283e-01
4.94413286e-01 6.38621330e-01 7.83170223e-01 -1.21507776e+00
-4.70413119e-01 3.81656796e-01 -3.52602035e-01 4.68870133e-01
2.30649441e-01 -2.96552274e-02 5.70168912e-01 -1.29957438e+00
8.76865685e-01 8.23715091e-01 1.04279935e+00 8.20891023e-01
-1.45347154e+00 -5.65851808e-01 3.53006899e-01 6.25598371e-01
-1.36740887e+00 -8.50825608e-01 7.98779666e-01 -4.68225032e-01
1.16628397e+00 1.77615538e-01 5.29500782e-01 9.02875900e-01
-2.62227952e-01 6.30097866e-01 9.12085354e-01 -7.78804243e-01
5.28537512e-01 5.93906462e-01 6.44188523e-01 5.79537213e-01
1.08395703e-03 -1.25462443e-01 -5.07441700e-01 -1.63196489e-01
2.75014997e-01 2.92244792e-01 -4.94968668e-02 -1.13784146e+00
-7.54898965e-01 9.27401364e-01 1.93744868e-01 4.85660315e-01
-5.84511086e-02 -9.27990377e-02 6.45041347e-01 3.32838684e-01
7.22986162e-01 4.03253496e-01 -5.46903849e-01 -9.94113162e-02
-1.03272736e+00 -6.41098842e-02 7.27398157e-01 1.05908048e+00
9.72563028e-01 2.34481871e-01 -9.82769802e-02 1.14563739e+00
-1.71160206e-01 6.94545805e-02 7.70361185e-01 -9.07852292e-01
-1.32221103e-01 4.66212362e-01 -1.53380156e-01 -5.13866484e-01
-3.97018105e-01 -6.07302725e-01 -3.35007578e-01 3.00210416e-01
2.07075238e-01 -8.79486278e-02 -1.01964283e+00 1.53983200e+00
3.64910305e-01 5.46310544e-01 6.61396831e-02 2.38486275e-01
7.69714475e-01 4.58040863e-01 2.09401935e-01 -5.87216556e-01
8.80057633e-01 -1.02091944e+00 -4.49020058e-01 -2.90675312e-01
6.69828296e-01 -3.90535086e-01 1.01293623e+00 8.99618417e-02
-5.78067422e-01 -5.01862705e-01 -1.15086818e+00 2.61529207e-01
-6.95164680e-01 -2.07767636e-01 7.24590242e-01 7.02541769e-01
-9.14855421e-01 7.60943651e-01 -9.08226967e-01 -7.16690719e-01
6.59670293e-01 3.76826048e-01 -4.67267841e-01 -3.84518802e-01
-7.13973463e-01 8.96274328e-01 3.88106078e-01 -3.37124854e-01
-6.30356967e-01 -6.81219816e-01 -1.03425360e+00 2.32628863e-02
4.38189715e-01 -2.74727374e-01 1.40048444e+00 -1.01177597e+00
-1.40575266e+00 8.18030417e-01 -3.87611121e-01 -4.80403721e-01
2.81302691e-01 -1.13514014e-01 -2.92765141e-01 2.00890169e-01
1.55283853e-01 5.47939479e-01 1.09876227e+00 -9.38373327e-01
-4.87455547e-01 -3.30916971e-01 -2.15140149e-01 4.67988610e-01
-8.03941429e-01 2.33032946e-02 -2.85173893e-01 -4.76647079e-01
1.02428958e-01 -9.78715599e-01 -3.64500672e-01 -6.94133341e-02
1.10035025e-01 -2.55613297e-01 1.01968479e+00 -2.52166092e-01
1.03628862e+00 -2.16100931e+00 5.88786677e-02 -8.57419819e-02
3.14636901e-03 3.57014418e-01 -7.03147650e-02 6.62708059e-02
-4.19959068e-01 -3.49972308e-01 -3.50395948e-01 -2.04408154e-01
-3.18777382e-01 1.70050189e-01 -3.37552160e-01 4.94093716e-01
1.46712810e-01 8.49505186e-01 -1.00315523e+00 -2.15920910e-01
3.68122578e-01 4.85591412e-01 -4.90978241e-01 -5.28205745e-03
5.31297475e-02 1.73590839e-01 1.64926440e-01 5.55208206e-01
4.10895407e-01 -1.64758563e-01 5.30809921e-04 -1.13051897e-02
-2.82033049e-02 -2.28642315e-01 -1.37874389e+00 1.68512833e+00
-1.56909347e-01 8.21705163e-01 -4.53783393e-01 -1.32075179e+00
7.22941577e-01 1.16204068e-01 6.65004134e-01 -1.47801340e-01
-5.37662096e-02 7.69882873e-02 -3.39557707e-01 -3.61917317e-01
2.54221261e-01 -4.46174979e-01 -4.90571447e-02 3.31495076e-01
7.29707479e-01 -1.42168894e-01 3.03408802e-01 1.28766209e-01
1.02140558e+00 1.88478991e-01 6.12528026e-01 5.34812100e-02
4.02213544e-01 2.33579203e-01 5.66275835e-01 1.02352154e+00
-4.30700868e-01 5.20170033e-01 2.84581333e-02 -7.80424416e-01
-1.09172809e+00 -9.12900269e-01 -3.49879622e-01 1.68333101e+00
-1.95023790e-01 -3.92274112e-01 -5.34565687e-01 -7.79827058e-01
-2.31266603e-01 1.04996502e+00 -6.47717476e-01 -5.81453025e-01
-4.86784011e-01 -9.84916747e-01 6.63606972e-02 6.57112479e-01
2.47748598e-01 -7.78401852e-01 -6.54268026e-01 4.86486465e-01
9.24237669e-02 -1.10122120e+00 -1.59849867e-01 6.58332646e-01
-1.18717504e+00 -1.17673194e+00 -9.48510706e-01 -9.10796404e-01
8.13452184e-01 4.69691694e-01 6.51761234e-01 -2.70723611e-01
-7.47098446e-01 9.98157442e-01 -3.54882896e-01 -4.17489201e-01
-2.57217377e-01 9.02790576e-02 2.69920766e-01 1.07978888e-01
7.90275931e-01 -4.52308714e-01 -3.51993263e-01 1.73889041e-01
-6.07842207e-01 -2.82911748e-01 2.63505489e-01 9.70736265e-01
5.50645590e-01 -1.53081134e-01 5.33912838e-01 -1.38728940e+00
3.70137185e-01 -4.81105059e-01 -1.92835242e-01 4.16234046e-01
-7.21767366e-01 1.91909410e-02 4.33108926e-01 -1.02320278e+00
-1.08443904e+00 6.40946269e-01 1.20904811e-01 -5.71025193e-01
-2.66254067e-01 2.32394740e-01 4.18104261e-01 -4.39410418e-01
1.25496566e+00 4.75942522e-01 -4.13654149e-02 -4.12870705e-01
8.38367581e-01 6.29015326e-01 4.87656027e-01 -8.42123404e-02
7.39133179e-01 5.45203686e-01 -1.56753480e-01 -1.22301447e+00
-9.83945489e-01 -9.98930335e-01 -1.28938496e+00 -3.02728146e-01
5.60307503e-01 -1.01775885e+00 2.23355636e-01 4.91422623e-01
-7.17952013e-01 -3.13215792e-01 -7.77522802e-01 5.08526862e-01
-7.24754155e-01 4.62013304e-01 -3.75383586e-01 -7.99042940e-01
-1.34471059e-01 -6.68979824e-01 6.62703454e-01 2.24654317e-01
-2.74193078e-01 -8.02948892e-01 2.22517222e-01 4.89128269e-02
7.38321304e-01 -6.27797544e-02 8.89714003e-01 -1.02437341e+00
-4.18032594e-02 -6.15141630e-01 1.57939523e-01 3.38865101e-01
1.48466155e-01 -2.83571512e-01 -1.14834726e+00 -5.68788171e-01
1.41906232e-01 -5.47867537e-01 1.14622664e+00 2.67435938e-01
9.52922165e-01 6.94142878e-02 -3.40124756e-01 6.42536819e-01
1.29717183e+00 2.71958619e-01 4.62681174e-01 4.06139851e-01
4.93985265e-01 4.23164278e-01 3.82281780e-01 3.89511257e-01
1.06671505e-01 4.99585897e-01 -1.17350273e-01 2.95917094e-01
-3.16989630e-01 1.22715637e-01 1.20958045e-01 6.39602661e-01
-1.27955273e-01 6.14933670e-01 -7.88164258e-01 3.70464414e-01
-1.79605508e+00 -1.36625278e+00 3.52751404e-01 2.34816504e+00
6.82697833e-01 1.28295437e-01 -5.32985404e-02 5.51099479e-02
8.02170038e-01 -1.09535223e-02 -9.22934532e-01 -1.78896189e-01
5.34727536e-02 2.14915201e-01 3.02782446e-01 3.17961097e-01
-1.58088982e+00 1.19556355e+00 7.40293598e+00 6.78286076e-01
-1.08200407e+00 1.67521372e-01 5.37089825e-01 -3.28881711e-01
2.76220143e-01 3.06778196e-02 -1.03841031e+00 5.81545942e-02
1.16633940e+00 -2.40734369e-01 5.27540982e-01 1.37079465e+00
-4.21662301e-01 -1.30449025e-05 -1.12599504e+00 1.20784962e+00
7.23490775e-01 -1.35024357e+00 -1.88369527e-01 -4.17006552e-01
8.78345668e-01 2.62085021e-01 7.54884928e-02 1.02914834e+00
1.85774103e-01 -4.84253496e-01 3.03917736e-01 5.05313277e-01
1.12451124e+00 -5.04234850e-01 4.66830850e-01 4.68225777e-01
-1.14057577e+00 -5.61916709e-01 -6.83727384e-01 -5.51196039e-02
-8.85817558e-02 1.08458335e-02 -8.95683587e-01 -2.01027721e-01
6.91831589e-01 7.96831250e-01 -8.80655885e-01 1.26871133e+00
1.90214124e-02 4.75042909e-01 -2.46779919e-01 -5.72671229e-03
3.18989567e-02 2.63487548e-01 5.47074497e-01 1.18191886e+00
1.75381854e-01 2.03505293e-01 1.46951407e-01 2.44892970e-01
1.50973359e-02 2.57541478e-01 -9.64428961e-01 1.54409781e-01
3.38782907e-01 1.12355888e+00 -9.28743422e-01 -6.22796476e-01
-6.96637332e-01 1.23784542e+00 7.04606116e-01 4.34908837e-01
-4.25607830e-01 -5.51607788e-01 3.27734619e-01 1.39175087e-01
5.66838920e-01 -2.98045665e-01 -5.13021015e-02 -1.45347774e+00
-4.02466148e-01 -5.78801751e-01 5.88838577e-01 -5.63541293e-01
-1.20238316e+00 7.30168164e-01 2.82772303e-01 -1.36924458e+00
-5.57929516e-01 -6.36314213e-01 -6.06818676e-01 3.09299707e-01
-1.31206429e+00 -1.06753755e+00 -4.26691622e-01 7.33202159e-01
9.67405796e-01 -7.46583939e-01 1.33361745e+00 2.04073220e-01
-3.49961042e-01 5.34623921e-01 5.12034178e-01 -4.34617605e-03
8.41910481e-01 -1.00323296e+00 1.07582383e-01 7.57833540e-01
5.56160212e-01 6.42111897e-01 6.13932014e-01 -5.17311692e-01
-9.94326293e-01 -1.08054435e+00 7.09899247e-01 -5.38311362e-01
6.15552366e-01 -4.09838080e-01 -1.10639453e+00 7.88074255e-01
-1.33153051e-01 4.77097392e-01 1.21341658e+00 3.11268985e-01
-5.55649996e-01 -1.28385186e-01 -9.52111244e-01 5.09346783e-01
9.64809000e-01 -6.42803311e-01 -8.32962811e-01 4.22841311e-01
6.32873178e-01 -3.64581347e-02 -2.42877021e-01 3.49073797e-01
4.94850487e-01 -6.18938267e-01 9.74030912e-01 -1.14373899e+00
-1.65206134e-01 8.01408961e-02 -3.20471108e-01 -1.26720715e+00
-5.67775965e-01 -3.95131588e-01 -3.21886718e-01 8.42806041e-01
2.55658597e-01 -3.22628289e-01 9.24757421e-01 8.31846058e-01
3.58383432e-02 -5.27978003e-01 -1.03673863e+00 -8.27455938e-01
-2.14069728e-02 -5.97503185e-01 -3.08963899e-02 1.05955851e+00
1.91841811e-01 6.66314185e-01 -3.06028396e-01 -3.60773951e-01
7.47934043e-01 -5.41110672e-02 6.26439989e-01 -1.50081241e+00
-2.02614427e-01 -2.49980152e-01 -7.20273793e-01 -5.21168709e-01
2.64562845e-01 -1.06990612e+00 -3.54437120e-02 -1.37615943e+00
4.38411564e-01 -3.49588662e-01 -5.98089814e-01 7.15966105e-01
-8.47591385e-02 4.97551084e-01 1.86680287e-01 4.52709615e-01
-9.57055867e-01 4.81781602e-01 3.23740482e-01 -2.94653684e-01
-5.99244475e-01 1.80982515e-01 -6.54327393e-01 9.54493523e-01
5.82354367e-01 -5.88415325e-01 -4.94833350e-01 -6.18904270e-02
-2.32778415e-01 -4.51176643e-01 1.03667989e-01 -1.20089018e+00
5.86737096e-01 1.11014485e-01 6.10082865e-01 -4.37181681e-01
4.74407732e-01 -8.11791539e-01 -1.51829466e-01 5.22963881e-01
-3.63376349e-01 -5.63236952e-01 2.93443620e-01 9.12594199e-01
5.63248768e-02 -6.38538420e-01 9.40576315e-01 -4.95904714e-01
-1.31244433e+00 2.99140424e-01 -6.31041825e-01 1.57870967e-02
1.49879420e+00 -4.09745067e-01 -8.49590153e-02 -3.19715291e-01
-1.25015879e+00 -1.55423135e-01 2.95523256e-01 4.18539673e-01
7.12306678e-01 -1.32805824e+00 -5.13122737e-01 3.54879200e-01
5.06937504e-01 -4.18459326e-01 2.55038887e-01 5.30948818e-01
-2.76048750e-01 1.26258045e-01 -3.33133399e-01 -6.59283578e-01
-1.46002662e+00 7.18512535e-01 3.12817335e-01 5.60205802e-02
-7.88479328e-01 1.04561734e+00 -2.58565605e-01 -6.09494507e-01
3.63920689e-01 3.40923816e-01 -4.94245052e-01 5.40681303e-01
8.00003409e-01 4.66906577e-01 6.22439496e-02 -7.40780115e-01
-3.72169226e-01 4.56112713e-01 -4.38309431e-01 -1.23990122e-02
1.86887658e+00 -3.44352424e-02 1.42252624e-01 9.64058518e-01
1.32813787e+00 -2.81821430e-01 -1.38853574e+00 -7.58574426e-01
1.75915167e-01 -3.41473997e-01 3.56661119e-02 -6.54587150e-01
-4.06806618e-01 8.13833356e-01 1.02226651e+00 -2.38900557e-01
8.30320716e-01 1.41907543e-01 2.26205409e-01 8.12978566e-01
4.99579996e-01 -1.43493581e+00 3.04328024e-01 6.39647543e-01
6.13521278e-01 -1.50845337e+00 1.36894509e-01 -2.25384966e-01
-6.56329691e-01 1.47271490e+00 4.57060605e-01 -9.40813497e-02
9.44056153e-01 -3.53073739e-02 -9.67484415e-02 -3.36038992e-02
-6.27317369e-01 -2.83199221e-01 4.45254952e-01 7.74093747e-01
3.26893777e-01 -1.66764215e-01 9.87954661e-02 4.91858304e-01
2.59884059e-01 5.67096807e-02 5.53284764e-01 1.13408804e+00
-7.65734315e-01 -8.34579706e-01 -1.49270877e-01 7.44271934e-01
-1.08925350e-01 -1.15448691e-01 -7.45938867e-02 5.06560326e-01
5.12115732e-02 7.31043994e-01 1.42546207e-01 -2.80950934e-01
2.67411798e-01 7.75501430e-01 5.14262199e-01 -1.19462478e+00
-3.17820581e-03 7.25940540e-02 -1.02950715e-01 -3.76693577e-01
-3.75905871e-01 -1.00460255e+00 -9.38416064e-01 2.76156336e-01
-5.33686996e-01 -4.11918983e-02 6.80695772e-01 9.02279913e-01
1.44787803e-01 4.36010569e-01 6.62950695e-01 -1.06833434e+00
-8.21614206e-01 -8.91092896e-01 -5.95575511e-01 3.76064271e-01
8.14202577e-02 -7.52584100e-01 -4.71763283e-01 3.20380509e-01] | [9.93230152130127, 3.020188570022583] |
a2f09d0e-8e0e-4b38-9a2f-d48ba138ba14 | pifpaf-composite-fields-for-human-pose | 1903.06593 | null | http://arxiv.org/abs/1903.06593v2 | http://arxiv.org/pdf/1903.06593v2.pdf | PifPaf: Composite Fields for Human Pose Estimation | We propose a new bottom-up method for multi-person 2D human pose estimation
that is particularly well suited for urban mobility such as self-driving cars
and delivery robots. The new method, PifPaf, uses a Part Intensity Field (PIF)
to localize body parts and a Part Association Field (PAF) to associate body
parts with each other to form full human poses. Our method outperforms previous
methods at low resolution and in crowded, cluttered and occluded scenes thanks
to (i) our new composite field PAF encoding fine-grained information and (ii)
the choice of Laplace loss for regressions which incorporates a notion of
uncertainty. Our architecture is based on a fully convolutional, single-shot,
box-free design. We perform on par with the existing state-of-the-art bottom-up
method on the standard COCO keypoint task and produce state-of-the-art results
on a modified COCO keypoint task for the transportation domain. | ['Alexandre Alahi', 'Sven Kreiss', 'Lorenzo Bertoni'] | 2019-03-15 | pifpaf-composite-fields-for-human-pose-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Kreiss_PifPaf_Composite_Fields_for_Human_Pose_Estimation_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Kreiss_PifPaf_Composite_Fields_for_Human_Pose_Estimation_CVPR_2019_paper.pdf | cvpr-2019-6 | ['2d-human-pose-estimation'] | ['computer-vision'] | [-1.89831406e-01 2.65644789e-01 1.06067330e-01 -3.64729017e-01
-6.30235076e-01 -3.18908319e-02 7.04134524e-01 -7.73363784e-02
-6.60473049e-01 5.85991681e-01 3.75401378e-01 4.81844097e-01
-2.34317058e-03 -6.58544362e-01 -8.75465333e-01 -2.27695808e-01
-2.24350363e-01 1.30015445e+00 8.13888550e-01 -6.51852489e-01
-3.05494905e-01 6.22432768e-01 -1.82140899e+00 -5.17325737e-02
5.89747608e-01 9.00388360e-01 5.33911139e-02 6.65686548e-01
3.66386801e-01 4.24725503e-01 -1.30104870e-01 -6.64127052e-01
4.14825350e-01 9.52037796e-02 -7.09406614e-01 1.42866254e-01
9.01404500e-01 -3.26409996e-01 -5.04886270e-01 5.38644969e-01
6.75432563e-01 4.43596125e-01 7.03623950e-01 -1.61501622e+00
1.48945063e-01 -5.05906343e-02 -6.81124747e-01 -3.59579772e-02
6.99923575e-01 1.33856252e-01 6.58067405e-01 -1.08204496e+00
9.58200514e-01 1.85494649e+00 1.20066512e+00 4.53039736e-01
-1.02215183e+00 -3.21870953e-01 4.47082788e-01 2.38557160e-01
-1.43426585e+00 -3.81020397e-01 3.95446241e-01 -5.02343595e-01
1.07155621e+00 -8.63991901e-02 9.48011935e-01 9.66028512e-01
2.96639949e-01 1.21259212e+00 7.92745352e-01 -2.83775091e-01
-4.87093814e-02 -2.68794507e-01 -2.15949625e-01 1.10005999e+00
3.17487270e-01 -2.09800433e-02 -6.12529278e-01 -1.15303151e-01
5.83993435e-01 5.09083122e-02 2.61953712e-01 -1.07080662e+00
-1.18012822e+00 7.42835820e-01 5.52250326e-01 -1.32268846e-01
-6.16926908e-01 4.63656515e-01 2.12366238e-01 -3.06369662e-01
5.50963581e-01 -1.31016850e-01 -5.55520296e-01 2.01078840e-02
-1.01107383e+00 1.23570848e+00 6.67468727e-01 1.04261029e+00
8.14147413e-01 -4.71718371e-01 -5.68951428e-01 6.77899301e-01
5.53546607e-01 8.14226866e-01 -4.51020636e-02 -1.10236084e+00
4.81622398e-01 3.60321164e-01 5.08714855e-01 -7.70862341e-01
-9.05721605e-01 -2.90398508e-01 -3.98325562e-01 3.89483005e-01
5.72400033e-01 -1.05961196e-01 -1.22329187e+00 1.66032672e+00
9.08101678e-01 -8.19234923e-02 -5.25291800e-01 1.10685635e+00
8.50077569e-01 2.08277136e-01 1.50229499e-01 4.12318230e-01
1.77963471e+00 -1.40918493e+00 -5.29058099e-01 -6.34502470e-01
2.48429716e-01 -2.15031847e-01 5.24209738e-01 2.94474423e-01
-1.21582019e+00 -7.81948984e-01 -7.89013922e-01 -3.78211021e-01
-5.38748026e-01 4.89447787e-02 6.86711967e-01 6.09533250e-01
-8.56518686e-01 5.12170613e-01 -9.70172822e-01 -7.13248014e-01
6.07718587e-01 2.86239743e-01 -8.21391344e-01 -1.90898702e-01
-1.31388736e+00 1.27561033e+00 2.63321754e-02 2.77683251e-02
-7.57463515e-01 -6.92059338e-01 -1.34133291e+00 -2.09067836e-01
5.57430029e-01 -1.24691594e+00 1.30340755e+00 -2.83783823e-01
-1.35571289e+00 9.44085836e-01 -1.81636423e-01 -6.87592924e-01
1.04953504e+00 -7.90761411e-01 -3.41521204e-02 3.11867833e-01
4.80749249e-01 1.20796514e+00 8.69634569e-01 -1.11827374e+00
-8.39290559e-01 -6.25181377e-01 -3.12431902e-01 4.81035411e-01
4.84199703e-01 -1.30355582e-01 -9.79326427e-01 -5.09960115e-01
-5.19604497e-02 -1.18014371e+00 -4.97819602e-01 3.46781015e-01
-3.75645339e-01 -4.07121271e-01 6.86959863e-01 -5.54474473e-01
7.08335102e-01 -1.62283993e+00 3.31263572e-01 1.82781026e-01
3.88184428e-01 6.29921407e-02 9.16369557e-02 3.02202046e-01
3.59248132e-01 -4.86301720e-01 -1.87903002e-01 -9.17713702e-01
4.50242400e-01 1.20439626e-01 4.42055583e-01 8.54189634e-01
1.85344011e-01 1.16098523e+00 -9.16161656e-01 -8.04717481e-01
5.03995895e-01 6.13379717e-01 -4.30974126e-01 1.23560086e-01
-4.17454801e-02 2.74785966e-01 -3.25240493e-01 7.14711666e-01
8.44237506e-01 1.18997678e-01 -5.33868253e-01 -9.80750918e-02
-2.15105936e-01 -2.51054794e-01 -1.46143866e+00 2.09555173e+00
-6.61578402e-03 2.92313278e-01 3.38976502e-01 -5.90763688e-01
6.42853141e-01 1.12660058e-01 6.98933542e-01 -4.71059501e-01
1.83134228e-02 -6.83804005e-02 -5.88827610e-01 -3.05885285e-01
6.95582449e-01 -7.48436153e-02 -3.99764717e-01 -3.01171821e-02
4.30638343e-01 -1.78527504e-01 2.58804172e-01 2.96206981e-01
1.01498508e+00 7.82556474e-01 3.36483806e-01 -3.47813725e-01
3.91986191e-01 -1.97158698e-02 5.01331151e-01 9.17482615e-01
-5.92707574e-01 9.00211155e-01 2.28301540e-01 -6.25792384e-01
-8.94456625e-01 -1.20069945e+00 -1.05219230e-01 1.31930757e+00
2.42959216e-01 -3.87156665e-01 -7.09121883e-01 -7.58307815e-01
5.08042574e-01 2.63932258e-01 -8.26829195e-01 2.82630622e-01
-6.19525492e-01 -3.11926454e-01 4.42687124e-01 7.52491891e-01
5.51425695e-01 -8.43797326e-01 -9.04778302e-01 3.29076916e-01
-3.60311866e-01 -1.33882678e+00 -4.99566108e-01 8.68124366e-02
-2.94178516e-01 -1.06785917e+00 -1.29544163e+00 -3.50688517e-01
2.26026967e-01 9.70477834e-02 1.35650301e+00 -3.82836461e-01
-4.08454299e-01 7.10800707e-01 -1.70650691e-01 -7.22356737e-01
2.27025762e-01 9.56094824e-03 1.78822473e-01 -8.16853717e-02
3.54691058e-01 -1.56339377e-01 -8.46740186e-01 4.08354610e-01
-1.81109279e-01 -2.47967727e-02 6.43971264e-01 5.29723406e-01
5.36379933e-01 -2.62177169e-01 1.60520330e-01 -6.59300685e-01
1.64210740e-02 -2.78189629e-01 -5.06752014e-01 1.47848176e-02
-1.05395868e-01 1.84140019e-02 -6.15081601e-02 -1.85652435e-01
-1.12728834e+00 5.56175828e-01 -2.67665774e-01 -4.35699075e-01
-4.21443164e-01 -1.51736856e-01 -2.54630536e-01 -1.51309162e-01
6.58872247e-01 -3.96298915e-01 1.42039046e-01 -4.99754786e-01
6.44800544e-01 2.26216123e-01 8.89110267e-01 -5.08047700e-01
7.22815692e-01 9.04581606e-01 4.38464820e-01 -8.86680782e-01
-4.95759279e-01 -8.90612781e-01 -1.16891944e+00 -4.86886382e-01
1.23833895e+00 -1.31609356e+00 -8.54336262e-01 5.92030942e-01
-9.84236121e-01 -3.56920213e-01 -1.90498918e-01 5.01419008e-01
-7.94514477e-01 2.69481838e-01 -4.83176887e-01 -9.62001979e-01
-2.45373905e-01 -9.42307532e-01 1.96523190e+00 1.77668363e-01
-2.64919549e-01 -6.29311383e-01 2.31493935e-01 4.45743114e-01
2.00976327e-01 6.46330714e-01 -1.59774702e-02 -2.56952852e-01
-4.35349435e-01 -5.02191305e-01 -3.72754671e-02 -2.44043618e-01
-5.44533908e-01 -3.96488637e-01 -8.58018100e-01 -3.06248158e-01
-7.18125820e-01 -4.63058650e-01 1.17071187e+00 7.21608937e-01
3.99153143e-01 5.71532249e-02 -9.12002802e-01 3.84929508e-01
8.68933439e-01 -5.77154100e-01 6.08179450e-01 3.27855229e-01
8.27380359e-01 8.69925022e-01 7.04298794e-01 5.77803016e-01
9.78955209e-01 1.15465665e+00 3.32421005e-01 -3.08662832e-01
-2.52014965e-01 -3.84754092e-01 6.82533858e-03 -2.32728004e-01
-3.26102972e-01 -4.78236228e-02 -8.77823472e-01 7.10203111e-01
-2.29255700e+00 -9.63573754e-01 -2.07697034e-01 2.02507210e+00
2.15020254e-01 2.61237055e-01 8.08928967e-01 -1.72194093e-01
4.49774623e-01 4.53207195e-02 -3.18531245e-01 1.21702120e-01
7.60131478e-02 8.20831284e-02 6.38179302e-01 5.04867733e-01
-1.69149995e+00 1.04819643e+00 6.45819283e+00 7.01804578e-01
-3.38402390e-01 2.81404674e-01 2.78104424e-01 4.35127615e-04
2.39016041e-01 -3.36547077e-01 -1.18556011e+00 1.32507235e-01
4.59977508e-01 5.81895649e-01 -1.15098894e-01 9.85732436e-01
6.87025487e-02 -5.09683967e-01 -9.00492072e-01 9.76437688e-01
1.87539831e-01 -7.95290351e-01 -6.39034033e-01 3.04010008e-02
4.92342919e-01 1.98335111e-01 -3.29236388e-01 4.54019845e-01
3.20156544e-01 -1.04525423e+00 1.24904013e+00 8.46190274e-01
4.94589925e-01 -8.20283651e-01 6.30822659e-01 4.18576926e-01
-1.66956437e+00 5.74808642e-02 -1.77234605e-01 -2.22396944e-02
5.83770573e-01 3.39363396e-01 -6.57508075e-01 5.35494864e-01
1.01671457e+00 5.39957225e-01 -6.53195858e-01 1.13499272e+00
-2.05398828e-01 -2.06495300e-01 -7.31544316e-01 1.15161613e-01
1.56594485e-01 3.09073985e-01 7.27816761e-01 1.53373992e+00
5.60495369e-02 4.37782370e-02 5.95910251e-01 5.63748121e-01
2.23113343e-01 -2.27639198e-01 -4.53646481e-01 7.89943397e-01
1.65881991e-01 1.39125609e+00 -8.18358898e-01 -4.51308161e-01
-3.68791580e-01 1.07788026e+00 3.70532423e-01 1.10941045e-01
-7.78706789e-01 -2.53319800e-01 9.13176060e-01 5.91856003e-01
7.50642240e-01 -4.07551050e-01 1.42622992e-01 -1.07984233e+00
-3.53534557e-02 -3.49759549e-01 4.51881319e-01 -8.90066504e-01
-1.15168703e+00 3.45337689e-01 6.98452532e-01 -1.01045787e+00
-3.38265240e-01 -6.63264155e-01 -2.17570737e-01 8.49704266e-01
-1.27760363e+00 -1.79122961e+00 -4.96630579e-01 7.42115200e-01
4.56343621e-01 1.20402262e-01 5.75292349e-01 2.62428313e-01
-7.68142715e-02 4.53120798e-01 -5.91704249e-01 1.39765441e-02
7.69455552e-01 -1.39459491e+00 8.33220541e-01 8.71063590e-01
-1.19748808e-01 2.54724354e-01 9.11676466e-01 -9.96939540e-01
-1.26677883e+00 -1.10202801e+00 9.89646316e-01 -1.02207661e+00
-5.69938459e-02 -7.45318472e-01 -4.05756831e-01 6.57904863e-01
-1.49500340e-01 4.79823768e-01 9.58524123e-02 2.04334557e-01
-7.90207759e-02 -3.84725705e-02 -1.41605532e+00 3.82288963e-01
1.36619663e+00 4.91610579e-02 -7.27251589e-01 3.64299268e-01
5.58695555e-01 -8.53398263e-01 -5.87808251e-01 5.81767917e-01
9.70248401e-01 -9.76162255e-01 1.44221544e+00 -4.71186817e-01
-9.94874761e-02 -4.61950541e-01 -3.11379414e-02 -1.12016630e+00
-8.06358516e-01 -6.16901994e-01 -3.85233611e-01 6.50985658e-01
1.83801204e-02 -1.66898534e-01 9.16204929e-01 6.10518217e-01
-1.53869897e-01 -7.27524102e-01 -1.07265997e+00 -6.79142773e-01
-1.64738387e-01 -5.21615922e-01 4.23927158e-01 8.19744989e-02
-3.52998227e-01 3.39265734e-01 -6.59889698e-01 1.43805712e-01
1.03325558e+00 -2.87274390e-01 1.26485443e+00 -1.56953430e+00
-1.77964941e-01 -1.75907552e-01 -8.26174796e-01 -1.03329289e+00
6.89146519e-02 -3.31340015e-01 4.19070512e-01 -1.88197911e+00
9.26736146e-02 -1.53615713e-01 2.92263687e-01 4.65734214e-01
-2.63206780e-01 5.36893010e-01 3.02175671e-01 -1.05673544e-01
-1.11001551e+00 5.84414005e-01 1.10674143e+00 6.26760647e-02
-3.35459113e-02 1.19572408e-01 -3.36183101e-01 8.79053354e-01
8.43633115e-02 -3.10657412e-01 5.83890453e-02 -5.80607988e-02
1.38704762e-01 -6.32805331e-03 7.15481341e-01 -1.38622952e+00
3.55280012e-01 8.34876075e-02 8.18061233e-01 -1.10130036e+00
9.41327810e-01 -7.11526513e-01 9.98449922e-02 5.60981274e-01
1.82485104e-01 -4.17773984e-03 8.94918442e-02 7.85215020e-01
2.53502399e-01 2.13143274e-01 8.10710847e-01 -4.38278794e-01
-1.07540905e+00 6.21784449e-01 -2.51868129e-01 1.37421861e-01
9.41789269e-01 -2.75408089e-01 6.04994074e-02 -4.09361243e-01
-7.69897223e-01 5.87300897e-01 3.98111880e-01 6.67831123e-01
5.06822467e-01 -1.42886674e+00 -9.28868890e-01 3.14552963e-01
3.06392312e-01 1.73017249e-01 3.86769265e-01 8.82930934e-01
-5.04874706e-01 5.05382478e-01 -2.06937447e-01 -8.05195332e-01
-1.24177051e+00 4.64474410e-01 3.17370296e-01 -4.22093272e-01
-9.50358272e-01 9.47548807e-01 1.35448292e-01 -6.96282864e-01
2.03802302e-01 -2.27883339e-01 -2.49219716e-01 1.08614109e-01
5.74640632e-01 7.36718953e-01 1.22036092e-01 -1.23507154e+00
-8.56020033e-01 7.63238370e-01 2.96456516e-01 -4.41101223e-01
1.15213823e+00 -3.48231047e-01 3.80260050e-01 1.82744369e-01
9.23305809e-01 -4.42955159e-02 -1.77006412e+00 -2.25441322e-01
-2.84982324e-01 -4.58831400e-01 -2.23306537e-01 -7.21265316e-01
-5.77027082e-01 6.18103981e-01 6.44394994e-01 -3.04750144e-01
5.82249403e-01 3.43889624e-01 7.90499091e-01 2.75584102e-01
6.55453622e-01 -1.31293166e+00 8.27233493e-02 7.57140279e-01
8.21672380e-01 -1.26102674e+00 3.96473587e-01 -3.02973449e-01
-8.69185627e-01 8.66509259e-01 6.03465319e-01 -2.30452895e-01
7.29083121e-01 2.77419180e-01 -1.23463891e-01 -4.18291509e-01
-3.79735857e-01 -8.58362734e-01 7.47398078e-01 1.00491762e+00
4.33775559e-02 1.38879627e-01 -9.93078668e-03 6.50208354e-01
-1.72146067e-01 -2.35364623e-02 -1.56864300e-01 1.05352998e+00
-6.01660728e-01 -7.92451501e-01 -6.74173295e-01 2.09091544e-01
-2.47240975e-01 4.50967103e-01 -2.41593093e-01 1.09287179e+00
6.45273745e-01 8.30535352e-01 1.27308175e-01 -7.14334995e-02
8.07577431e-01 1.69182032e-01 8.32235277e-01 -5.16227603e-01
-2.86617190e-01 1.73812747e-01 2.61102676e-01 -1.16226935e+00
-6.07005239e-01 -9.84166265e-01 -1.06038046e+00 -9.08814967e-02
-1.20493107e-01 -2.53966212e-01 5.53416967e-01 1.11893690e+00
3.16792279e-01 3.87875825e-01 -9.03171077e-02 -1.85355496e+00
-9.93780345e-02 -1.02846313e+00 -5.98930597e-01 5.27075827e-01
3.66920471e-01 -1.39695084e+00 2.79104501e-01 -2.58586735e-01] | [7.070920467376709, -0.856907069683075] |
9f5703d8-dd22-4522-8675-fff43507b483 | kg-fid-infusing-knowledge-graph-in-fusion-in-1 | 2110.04330 | null | https://arxiv.org/abs/2110.04330v2 | https://arxiv.org/pdf/2110.04330v2.pdf | KG-FiD: Infusing Knowledge Graph in Fusion-in-Decoder for Open-Domain Question Answering | Current Open-Domain Question Answering (ODQA) model paradigm often contains a retrieving module and a reading module. Given an input question, the reading module predicts the answer from the relevant passages which are retrieved by the retriever. The recent proposed Fusion-in-Decoder (FiD), which is built on top of the pretrained generative model T5, achieves the state-of-the-art performance in the reading module. Although being effective, it remains constrained by inefficient attention on all retrieved passages which contain a lot of noise. In this work, we propose a novel method KG-FiD, which filters noisy passages by leveraging the structural relationship among the retrieved passages with a knowledge graph. We initiate the passage node embedding from the FiD encoder and then use graph neural network (GNN) to update the representation for reranking. To improve the efficiency, we build the GNN on top of the intermediate layer output of the FiD encoder and only pass a few top reranked passages into the higher layers of encoder and decoder for answer generation. We also apply the proposed GNN based reranking method to enhance the passage retrieval results in the retrieving module. Extensive experiments on common ODQA benchmark datasets (Natural Question and TriviaQA) demonstrate that KG-FiD can improve vanilla FiD by up to 1.5% on answer exact match score and achieve comparable performance with FiD with only 40% of computation cost. | ['Michael Zeng', 'Yiming Yang', 'Xiang Ren', 'Yichong Xu', 'Shuohang Wang', 'Wenhao Yu', 'Yuwei Fang', 'Chenguang Zhu', 'Donghan Yu'] | 2021-10-08 | kg-fid-infusing-knowledge-graph-in-fusion-in | https://aclanthology.org/2022.acl-long.340 | https://aclanthology.org/2022.acl-long.340.pdf | acl-2022-5 | ['triviaqa'] | ['miscellaneous'] | [ 1.51388124e-01 1.91754878e-01 2.45833337e-01 -1.45921111e-01
-1.33770597e+00 -4.62677002e-01 5.31336367e-01 2.97082841e-01
-4.68587756e-01 5.94379544e-01 5.81016839e-01 -3.08446974e-01
-1.59806117e-01 -1.19051850e+00 -9.91807103e-01 -1.98405966e-01
3.56402665e-01 6.34744167e-01 6.92360640e-01 -6.24097824e-01
2.43940562e-01 -2.27033615e-01 -1.34167743e+00 6.65781260e-01
1.32440686e+00 1.09175789e+00 2.29023010e-01 9.63671267e-01
-5.41827619e-01 1.24316847e+00 -6.47437453e-01 -8.10709953e-01
1.90796610e-02 -6.90367818e-01 -1.23336375e+00 -5.11940360e-01
4.48879302e-01 -6.48347437e-01 -8.81752551e-01 7.16133654e-01
7.60925770e-01 5.63295484e-01 5.05374193e-01 -6.28128052e-01
-1.39515615e+00 7.98391223e-01 -8.82352963e-02 5.22671461e-01
8.27060342e-01 2.94801258e-02 1.43432331e+00 -1.08971334e+00
4.77819234e-01 1.26659918e+00 2.81217635e-01 5.09311497e-01
-8.36349249e-01 -1.84089258e-01 1.33792669e-01 5.95832169e-01
-1.22272670e+00 -1.46336958e-01 6.35251284e-01 1.70232609e-01
1.14007747e+00 2.13443145e-01 2.44032696e-01 6.74729764e-01
1.31223708e-01 9.51087952e-01 4.23816085e-01 -2.43442714e-01
-5.25281057e-02 -2.22717598e-01 4.98666435e-01 8.31407547e-01
-1.52977973e-01 -3.08516204e-01 -4.39449936e-01 -2.66210079e-01
2.63895303e-01 -5.03972955e-02 -5.59799731e-01 1.21231794e-01
-8.66025686e-01 9.90908861e-01 1.03637862e+00 2.33856160e-02
-5.61225653e-01 6.48977235e-02 1.66732997e-01 8.09240639e-01
2.97542095e-01 4.72470433e-01 -2.54499793e-01 2.15933502e-01
-7.76775777e-01 3.90314519e-01 1.06716943e+00 8.23098898e-01
8.69879961e-01 -5.04801869e-01 -9.31429446e-01 9.96407628e-01
5.06486773e-01 4.81845319e-01 4.77399707e-01 -6.42977536e-01
8.72415066e-01 9.89131093e-01 -7.53433332e-02 -1.14364576e+00
-4.66777906e-02 -6.52002573e-01 -5.22325635e-01 -8.23431671e-01
3.57745811e-02 -4.10211422e-02 -1.03656721e+00 1.44680166e+00
3.99339080e-01 2.23301247e-01 2.46042356e-01 1.00613701e+00
1.45734298e+00 1.18247616e+00 5.44698304e-03 1.96490929e-01
1.27182639e+00 -1.47204828e+00 -5.37275553e-01 -2.07371548e-01
6.85347259e-01 -8.62027347e-01 1.12893021e+00 -1.98928908e-01
-1.16148651e+00 -6.66158319e-01 -9.01425660e-01 -7.49451756e-01
-3.05698574e-01 5.17699905e-02 3.86346988e-02 3.21473368e-02
-1.22152519e+00 4.06080872e-01 -2.42490798e-01 -2.50843167e-01
1.44123450e-01 2.53719330e-01 6.71627522e-02 -6.89781427e-01
-1.75774992e+00 9.22591627e-01 3.13584387e-01 8.55620801e-02
-9.40372229e-01 -8.34916770e-01 -9.46112514e-01 2.76171625e-01
3.31713349e-01 -1.43025124e+00 1.27942538e+00 -2.94287235e-01
-1.59404361e+00 3.85803699e-01 -3.90199602e-01 -5.70277870e-01
1.13183074e-01 -3.29977870e-01 -5.51789701e-01 5.26553631e-01
1.00519709e-01 7.11438596e-01 6.95839524e-01 -9.59794164e-01
-4.77517873e-01 -1.28442392e-01 6.07019007e-01 4.18391347e-01
-1.39788687e-01 -2.40618363e-01 -8.13277304e-01 -3.27438980e-01
8.98955613e-02 -6.18800402e-01 -2.02806499e-02 -5.07653773e-01
-3.28025401e-01 -7.03087091e-01 5.39176643e-01 -1.13035989e+00
1.72439432e+00 -1.86126876e+00 1.71116993e-01 1.11507043e-01
2.28394091e-01 4.74040627e-01 -7.03858852e-01 8.42457473e-01
3.16259921e-01 -2.01135561e-01 -1.90275729e-01 -9.73754302e-02
1.01657629e-01 1.66317776e-01 -6.68446302e-01 -5.78315929e-02
3.97361934e-01 1.50920355e+00 -1.03574920e+00 -3.32472682e-01
-2.23965675e-01 3.33524108e-01 -8.28037500e-01 5.31203926e-01
-6.82569742e-01 1.84039935e-01 -6.17548704e-01 3.56183052e-01
5.33823192e-01 -6.33016229e-01 -3.37952018e-01 -1.17883839e-01
4.79138374e-01 1.00394189e+00 -8.11415553e-01 1.83135557e+00
-5.37683070e-01 1.74464568e-01 -3.06385696e-01 -5.42826533e-01
9.96429026e-01 2.14142978e-01 -9.18803886e-02 -1.05901718e+00
6.62932843e-02 3.40369701e-01 -6.51773363e-02 -7.52368212e-01
8.54096115e-01 1.87970817e-01 -2.96142958e-02 4.56735462e-01
3.73085290e-01 -4.86301966e-02 3.45357984e-01 8.54342759e-01
1.49497855e+00 -8.15209299e-02 -2.20854253e-01 1.30661443e-01
1.07561243e+00 -4.88957651e-02 6.86668232e-02 9.00278091e-01
1.81422845e-01 6.77224815e-01 2.86677420e-01 -4.77888919e-02
-7.34304786e-01 -1.09776366e+00 4.01824266e-01 1.26381016e+00
2.69204468e-01 -4.14436072e-01 -6.47365689e-01 -8.63125443e-01
8.63811076e-02 8.10245991e-01 -4.95057106e-01 -6.98797464e-01
-8.21614683e-01 -3.45908552e-01 5.18348396e-01 4.22823548e-01
7.01869011e-01 -1.08943939e+00 1.31142542e-01 4.02666926e-01
-6.17663145e-01 -1.08433962e+00 -7.26494730e-01 -3.31534177e-01
-6.77870691e-01 -9.73160744e-01 -8.92350435e-01 -9.45832789e-01
6.10377312e-01 3.38470250e-01 1.48391581e+00 4.01363492e-01
3.41704100e-01 4.90149230e-01 -7.65878439e-01 8.00953060e-02
-3.61679405e-01 4.53961372e-01 -6.12328649e-01 9.54185352e-02
3.83439124e-01 -3.11438888e-01 -9.95411217e-01 1.80348605e-01
-1.11878955e+00 -2.46783793e-01 5.07781446e-01 9.42194402e-01
6.20741785e-01 -3.34464192e-01 1.12590265e+00 -7.97011256e-01
1.13131666e+00 -8.83219957e-01 -2.63287902e-01 4.97328222e-01
-3.70607793e-01 3.80652696e-01 7.38896608e-01 -1.40543222e-01
-1.16567051e+00 -6.46682024e-01 -5.76505780e-01 -2.19814956e-01
4.07032847e-01 8.08293819e-01 8.08513537e-02 1.05329482e-02
6.06311142e-01 3.24601471e-01 -3.47730279e-01 -6.09005630e-01
7.48199582e-01 7.29002476e-01 4.95463252e-01 -3.58808249e-01
8.97824407e-01 7.66758025e-02 -5.01214862e-01 -2.65342057e-01
-1.28974688e+00 -6.78023636e-01 -1.51192024e-01 -4.09908369e-02
7.51610458e-01 -9.61096764e-01 -4.98986661e-01 1.10565096e-01
-1.26417172e+00 5.03784269e-02 -3.01125199e-01 2.45804340e-01
-2.10843071e-01 3.72913212e-01 -7.97913313e-01 -4.11864638e-01
-8.86059523e-01 -7.90601075e-01 1.04020309e+00 5.09427786e-01
1.28281370e-01 -8.51807833e-01 2.97488689e-01 9.12945032e-01
6.28108799e-01 -3.67051929e-01 1.29429054e+00 -9.01007354e-01
-8.70340347e-01 -3.30634713e-01 -3.65980297e-01 3.77195567e-01
-2.39913046e-01 -5.66999435e-01 -7.79340088e-01 -1.87946290e-01
-1.21215872e-01 -4.61121976e-01 1.46613491e+00 -1.78016052e-01
8.47128212e-01 -3.72205257e-01 4.48424742e-02 1.33012861e-01
1.50802147e+00 -2.17410177e-02 7.39359617e-01 1.45965099e-01
6.84635580e-01 4.24050122e-01 3.70884150e-01 5.06040603e-02
9.26091731e-01 3.06343079e-01 4.39448565e-01 3.96216273e-01
-5.84001124e-01 -7.57224619e-01 5.17637312e-01 1.57728434e+00
4.84590113e-01 -6.39692307e-01 -5.98793507e-01 8.06157947e-01
-1.80428267e+00 -6.96592093e-01 -1.22601457e-01 2.02833629e+00
9.40925837e-01 -7.11491033e-02 -2.15568960e-01 -1.76721320e-01
4.82059538e-01 2.68624276e-01 -5.18905282e-01 -4.21451837e-01
2.54640523e-02 5.54321051e-01 -2.14829128e-02 8.11607182e-01
-6.55304134e-01 9.57960665e-01 5.10276508e+00 8.16870570e-01
-4.56928104e-01 5.63265868e-02 2.62337804e-01 4.10298109e-02
-7.27948189e-01 1.09915577e-01 -9.04511988e-01 3.29485625e-01
1.18440652e+00 -2.48691767e-01 4.63133931e-01 4.56274152e-01
-2.09714830e-01 -2.86578033e-02 -9.12595987e-01 5.63548565e-01
4.51080918e-01 -1.33461142e+00 7.40852237e-01 -5.87441683e-01
7.54774749e-01 1.77951992e-01 -9.28112715e-02 9.85909760e-01
2.01442644e-01 -8.18467796e-01 1.47537678e-01 8.63004923e-01
2.33507633e-01 -9.15503800e-01 9.51639414e-01 4.34792697e-01
-1.12417078e+00 -1.80972040e-01 -8.12192678e-01 1.13076858e-01
2.81227112e-01 4.84838128e-01 -7.10169077e-01 1.00796342e+00
5.04149199e-01 5.04939914e-01 -6.82204723e-01 1.23188162e+00
-5.41757226e-01 7.18868852e-01 -2.33930528e-01 -2.35469550e-01
6.00891769e-01 2.28699110e-02 5.25264740e-01 8.72110844e-01
1.93987548e-01 3.23489934e-01 -1.20005302e-01 7.85537064e-01
-7.01540053e-01 2.42653981e-01 -2.79139012e-01 3.47667523e-02
4.06968057e-01 9.85277712e-01 7.30946511e-02 -4.45723683e-01
-6.12742066e-01 1.16149163e+00 6.56583071e-01 4.56473827e-01
-6.92328036e-01 -6.55776978e-01 1.63750470e-01 -1.09847344e-01
5.96851647e-01 2.24463359e-01 2.91415274e-01 -1.31852853e+00
3.08258682e-01 -1.04307067e+00 8.18930387e-01 -8.20325017e-01
-1.43375647e+00 7.80261397e-01 -4.28832293e-01 -9.07288134e-01
-2.89147317e-01 -2.63473392e-01 -5.13072431e-01 1.16259587e+00
-2.06446743e+00 -8.28604579e-01 -1.88378111e-01 7.65412569e-01
5.38683832e-01 7.33933896e-02 7.69346535e-01 6.42892778e-01
-5.24134219e-01 6.43495917e-01 1.53353497e-01 2.30966434e-01
5.28508723e-01 -1.18388486e+00 5.25699854e-01 7.43850291e-01
2.64582157e-01 6.87249482e-01 2.15030730e-01 -5.77175260e-01
-1.51830494e+00 -1.21565163e+00 1.35765910e+00 -4.02457774e-01
6.97956920e-01 -3.29501480e-02 -1.31739497e+00 3.75733078e-01
6.37536407e-01 -1.63007349e-01 4.70094055e-01 -1.74398080e-01
-4.74756241e-01 -8.13862532e-02 -8.44982326e-01 5.66473365e-01
7.17145681e-01 -8.26480567e-01 -1.18536484e+00 3.58602196e-01
1.34416544e+00 -4.43369865e-01 -9.25725281e-01 4.90162969e-01
5.58953620e-02 -5.67116618e-01 1.09012008e+00 -7.33448684e-01
5.20411909e-01 -4.20141160e-01 -1.01693600e-01 -1.25653005e+00
-3.29949915e-01 -5.03148794e-01 -6.13096237e-01 1.26983154e+00
6.88868940e-01 -4.63776171e-01 4.91677284e-01 3.60029161e-01
-2.57012516e-01 -1.13328552e+00 -8.30552638e-01 -4.50506955e-01
1.95672318e-01 -9.46165472e-02 6.37574136e-01 3.55006427e-01
-1.83157951e-01 9.85527456e-01 -4.77680229e-02 2.50169456e-01
2.69696593e-01 4.09940004e-01 5.34507215e-01 -7.66346693e-01
-3.34955156e-01 -3.40301618e-02 -3.17323878e-02 -1.88902342e+00
3.22492997e-04 -1.27699828e+00 4.84147072e-02 -2.18569851e+00
1.96655005e-01 -9.55080427e-03 -4.66642529e-01 5.50799891e-02
-9.03829455e-01 2.66558095e-03 2.06268385e-01 1.13452166e-01
-1.13227189e+00 1.07231760e+00 1.54533577e+00 -2.80036211e-01
-4.61428314e-02 -1.91844091e-01 -8.19720149e-01 3.88259739e-01
6.13726497e-01 -5.39733171e-01 -6.71401978e-01 -1.02503002e+00
5.57227314e-01 2.35403806e-01 3.76367092e-01 -9.09401596e-01
6.07361078e-01 5.49933195e-01 1.39604703e-01 -8.14396620e-01
2.56740391e-01 -5.13910532e-01 -4.62060452e-01 1.80008829e-01
-6.92694664e-01 1.73867062e-01 7.42227435e-02 8.46135259e-01
-5.96632242e-01 -3.91925097e-01 4.19641346e-01 -4.91018295e-02
-6.27548456e-01 4.09487396e-01 -6.01664837e-03 6.45594239e-01
3.83305758e-01 2.57595658e-01 -8.08612823e-01 -6.88259482e-01
-4.42823678e-01 8.41506958e-01 -2.26957589e-01 5.70039034e-01
9.23658133e-01 -1.33675623e+00 -1.04716706e+00 -4.94483393e-03
2.14171588e-01 -2.12332755e-02 6.53304815e-01 7.57998109e-01
-4.55359936e-01 7.40414739e-01 5.58514178e-01 -1.14825405e-01
-7.34263182e-01 4.97073770e-01 4.15968686e-01 -8.67499650e-01
-5.10206342e-01 1.26778436e+00 -2.45145559e-02 -7.17869222e-01
5.60854822e-02 -2.71852165e-01 -5.94841897e-01 -1.71243586e-02
7.34840691e-01 4.69931483e-01 2.70363510e-01 -4.73595828e-01
-1.98518395e-01 3.29471707e-01 -4.58056778e-01 4.77547757e-03
1.06346858e+00 -3.02534580e-01 -3.66825253e-01 -9.91398096e-02
1.48737144e+00 -1.24477677e-01 -6.82377517e-01 -8.10010195e-01
-3.02133095e-02 -1.55815229e-01 -4.56806757e-02 -9.78786707e-01
-8.53297532e-01 8.05847764e-01 1.64388448e-01 1.04572184e-01
1.22301614e+00 2.12295711e-01 1.60560822e+00 7.27309167e-01
1.61897972e-01 -7.69588590e-01 2.72172093e-01 9.95348215e-01
1.10981297e+00 -9.51239467e-01 -3.88604581e-01 -2.21733779e-01
-4.61132795e-01 8.54881525e-01 6.72163725e-01 -3.61671478e-01
5.07156372e-01 -4.03770745e-01 4.13848162e-02 -3.84099185e-01
-1.12397790e+00 -4.72332686e-01 7.27532148e-01 1.57790542e-01
2.21292019e-01 -3.56998265e-01 -4.87561494e-01 6.83234751e-01
-2.01584071e-01 -3.37293483e-02 1.41045883e-01 7.09233522e-01
-7.52222478e-01 -9.76945639e-01 -3.48278172e-02 7.32490659e-01
-4.97601509e-01 -6.61413670e-01 -3.27109098e-01 2.25934163e-01
-2.29988575e-01 1.40245295e+00 -1.53632745e-01 -5.57426035e-01
5.87782621e-01 3.68255705e-01 2.21638203e-01 -6.59523308e-01
-9.74936068e-01 -4.96219963e-01 2.92865157e-01 -5.25727510e-01
-1.60542727e-01 -2.00117931e-01 -1.27298498e+00 -2.03234509e-01
-5.81850111e-01 5.82507551e-01 1.15260319e-03 1.01757646e+00
7.76847839e-01 7.36413002e-01 5.44320345e-01 2.17883795e-01
-7.80596495e-01 -1.12070632e+00 -2.41259888e-01 2.91200757e-01
4.24966604e-01 -1.51983902e-01 -2.88813949e-01 -2.94927388e-01] | [11.172150611877441, 8.05397891998291] |
c69736fb-85eb-4088-af74-64ea487177a8 | metrics-matter-in-surgical-phase-recognition | 2305.13961 | null | https://arxiv.org/abs/2305.13961v1 | https://arxiv.org/pdf/2305.13961v1.pdf | Metrics Matter in Surgical Phase Recognition | Surgical phase recognition is a basic component for different context-aware applications in computer- and robot-assisted surgery. In recent years, several methods for automatic surgical phase recognition have been proposed, showing promising results. However, a meaningful comparison of these methods is difficult due to differences in the evaluation process and incomplete reporting of evaluation details. In particular, the details of metric computation can vary widely between different studies. To raise awareness of potential inconsistencies, this paper summarizes common deviations in the evaluation of phase recognition algorithms on the Cholec80 benchmark. In addition, a structured overview of previously reported evaluation results on Cholec80 is provided, taking known differences in evaluation protocols into account. Greater attention to evaluation details could help achieve more consistent and comparable results on the surgical phase recognition task, leading to more reliable conclusions about advancements in the field and, finally, translation into clinical practice. | ['Stefanie Speidel', 'Dominik Rivoir', 'Isabel Funke'] | 2023-05-23 | null | null | null | null | ['surgical-phase-recognition'] | ['computer-vision'] | [ 2.50695735e-01 6.41630143e-02 -7.91255713e-01 -3.80597144e-01
-1.00850403e+00 -6.75031722e-01 3.58126074e-01 8.46770167e-01
-6.96576655e-01 5.95599771e-01 6.35373056e-01 -4.09569412e-01
-5.47241807e-01 -3.32141250e-01 -1.74838468e-01 -7.58429825e-01
-4.74946171e-01 4.82944965e-01 9.73544046e-02 -1.26890391e-02
4.63101238e-01 5.87571859e-01 -1.32361937e+00 2.87420422e-01
4.03967172e-01 7.45758414e-01 3.63119900e-01 5.56234717e-01
5.75059056e-02 4.23312634e-01 -4.95248467e-01 1.68857891e-02
1.61459550e-01 -3.75224590e-01 -9.11191404e-01 -1.86205804e-01
2.99175624e-02 9.07562152e-02 -7.47215897e-02 7.39395261e-01
8.27440560e-01 -2.94913322e-01 6.78689241e-01 -6.52490258e-01
2.37165704e-01 5.27515233e-01 -2.10725777e-02 2.15377569e-01
5.38127780e-01 5.40639609e-02 3.63576770e-01 -5.09597003e-01
9.21298504e-01 4.78537530e-01 7.19522834e-01 5.22484422e-01
-1.04078341e+00 -4.96912688e-01 -1.07161395e-01 9.90351737e-02
-1.25977874e+00 -2.57169306e-01 5.31297922e-01 -5.48257828e-01
7.23092556e-01 5.29192209e-01 1.26050580e+00 7.20220685e-01
7.64878988e-01 6.53295279e-01 1.31572759e+00 -4.36664909e-01
1.37750044e-01 1.91191763e-01 -2.41263509e-02 7.45525539e-01
8.03580046e-01 3.77868116e-01 -3.25299799e-01 -1.35469168e-01
4.25413102e-01 -1.81385398e-01 -4.89940584e-01 -9.92801189e-01
-1.56542003e+00 4.85396802e-01 5.46764195e-01 6.39286280e-01
-3.34339589e-01 -7.97402337e-02 1.04293025e+00 -2.74256226e-02
-1.16469897e-01 5.65372467e-01 1.35672530e-02 -5.00765264e-01
-8.50559473e-01 -8.08868110e-02 9.21648860e-01 8.00876856e-01
1.17966339e-01 -4.62848812e-01 -1.92519903e-01 6.05331779e-01
3.48104149e-01 3.23216319e-02 7.26895392e-01 -8.13630939e-01
1.46839647e-02 4.44139689e-01 -1.34319946e-01 -8.35552216e-01
-9.18457806e-01 -4.96579260e-01 -7.62049317e-01 2.61640579e-01
3.24229956e-01 7.90166035e-02 -8.91690552e-01 1.23997653e+00
-9.40602794e-02 -3.42394292e-01 5.48169613e-02 1.01512206e+00
1.09429884e+00 -3.27869475e-01 3.65112960e-01 -4.40506965e-01
1.78706312e+00 -6.88467681e-01 -8.08158398e-01 -3.84010017e-01
1.10918379e+00 -9.95754421e-01 5.74759364e-01 5.23278356e-01
-7.41702020e-01 -3.55711393e-02 -1.24584496e+00 2.51983523e-01
-2.60653645e-01 4.14558321e-01 1.01633763e+00 8.18222463e-01
-8.49121153e-01 3.86018485e-01 -1.27882004e+00 -8.17588210e-01
1.31154716e-01 7.39618003e-01 -5.44697583e-01 -8.44629928e-02
-8.64982843e-01 1.33069980e+00 4.45471793e-01 3.11828047e-01
-7.90001988e-01 -4.93376374e-01 -9.33740079e-01 -6.89252198e-01
2.36138806e-01 -7.14512885e-01 1.43301535e+00 -3.34261775e-01
-1.30363655e+00 1.44822896e+00 -5.58796264e-02 -3.11781257e-01
3.17454129e-01 9.85164940e-02 -3.30121517e-01 9.35872793e-02
-1.02114737e-01 4.27679867e-01 -2.05014925e-02 -9.33567822e-01
-6.79100513e-01 -4.83755171e-01 -1.09815076e-01 6.36208892e-01
2.48689726e-01 -4.99731302e-02 -5.74148238e-01 -6.18732810e-01
3.14946890e-01 -1.37217426e+00 -7.23819435e-01 2.35727102e-01
1.24847688e-01 1.06934555e-01 1.11540824e-01 -3.20378333e-01
1.30246663e+00 -2.04534841e+00 1.53263226e-01 1.91013869e-02
1.02723427e-01 -2.39113159e-03 2.96737194e-01 4.85108346e-01
-2.73138881e-01 -7.26109520e-02 -8.83383006e-02 -1.08185552e-01
-5.32870770e-01 6.03732526e-01 3.47574353e-01 1.02151537e+00
-2.39850938e-01 6.60980225e-01 -1.04525542e+00 -6.64113224e-01
5.42729735e-01 1.69950470e-01 -4.99506772e-01 -4.15832587e-02
5.31933665e-01 4.98259425e-01 -3.46973836e-01 8.19697797e-01
1.74256369e-01 8.24024677e-02 4.52690929e-01 -3.98946732e-01
-2.40434542e-01 5.71362019e-01 -9.11342740e-01 2.37850499e+00
-5.44502497e-01 6.39853597e-01 4.29078117e-02 -7.64922380e-01
5.78552425e-01 4.61059660e-01 9.59625959e-01 -4.69485760e-01
2.65695065e-01 6.35269940e-01 5.04401445e-01 -2.65578240e-01
5.83031476e-01 -2.99127400e-01 -1.33252397e-01 -9.97799784e-02
-1.25722870e-01 -3.67157936e-01 4.01131511e-01 -9.97509360e-02
9.11399364e-01 -1.46477044e-01 8.61203790e-01 -6.10513866e-01
3.40782493e-01 4.46777552e-01 4.01914388e-01 4.32627499e-01
-6.56183481e-01 7.75864899e-01 4.66778785e-01 -3.29902917e-01
-2.32072771e-01 -1.00756693e+00 -5.89428008e-01 1.37453929e-01
4.59863037e-01 -6.63145304e-01 -1.77194163e-01 -8.62392724e-01
-1.52341694e-01 2.72099853e-01 -1.03488588e+00 -3.43573809e-01
-6.15356445e-01 -1.04681659e+00 2.96419799e-01 4.77405608e-01
-3.14760327e-01 -8.32801223e-01 -9.53040004e-01 3.75678629e-01
-4.11898762e-01 -1.10779369e+00 -5.45868203e-02 6.13766015e-01
-1.44292259e+00 -1.48754191e+00 -9.15201008e-01 -1.16406059e+00
8.53503883e-01 4.33377892e-01 1.00344217e+00 -7.90955406e-03
-5.90427756e-01 6.26582563e-01 -3.62439424e-01 -6.36566460e-01
-6.39063478e-01 1.56450421e-01 -6.85699359e-02 -8.50986183e-01
2.94622093e-01 5.30179702e-02 -1.08396387e+00 3.70813549e-01
-6.71772599e-01 -9.96774659e-02 1.12850904e+00 7.88280070e-01
6.99849188e-01 -5.35045385e-01 -2.83418149e-02 -9.03177440e-01
3.75266194e-01 -2.29460701e-01 -3.72818381e-01 2.40059644e-02
-1.25783408e+00 7.80584887e-02 -8.67966563e-02 -1.12351038e-01
-5.07588446e-01 2.20427420e-02 -1.73414703e-02 2.08861768e-01
-2.46857002e-01 7.83946037e-01 5.42072475e-01 -4.64289635e-01
9.53793585e-01 -1.07378354e-02 5.02279937e-01 -7.00819194e-02
-2.30589852e-01 4.85910743e-01 3.63633871e-01 -3.44205588e-01
3.63434702e-01 5.31508386e-01 1.19706303e-01 -7.45874107e-01
-4.53533113e-01 -8.86632562e-01 -3.12455952e-01 -2.35784769e-01
6.43238604e-01 -8.30126166e-01 -4.13289160e-01 2.48300850e-01
-6.52514398e-01 -3.84620786e-01 -3.69694203e-01 9.80226040e-01
-8.25024009e-01 4.17401582e-01 -6.70509577e-01 -1.70186430e-01
-5.41680634e-01 -1.94613206e+00 1.04575610e+00 2.06881419e-01
-7.25110054e-01 -1.13848698e+00 5.02850115e-01 1.80981338e-01
4.35215950e-01 3.58274162e-01 5.68191290e-01 -5.28362870e-01
-3.78342956e-01 -7.09974170e-01 2.49587834e-01 -9.85657126e-02
5.72782218e-01 -5.88043518e-02 -5.41185617e-01 -3.81978303e-01
-1.53372809e-01 2.16186345e-01 2.09675118e-01 6.60700977e-01
8.86226952e-01 2.93176025e-01 -8.35261166e-01 3.95195931e-01
1.53626037e+00 2.25444227e-01 7.87615299e-01 7.74345458e-01
-6.82955980e-02 5.39251983e-01 1.18515027e+00 1.80829108e-01
3.67380559e-01 8.75476360e-01 4.80796248e-01 -1.62177086e-01
-4.43043530e-01 2.20216691e-01 -7.48464912e-02 7.68600047e-01
-1.82256281e-01 2.75081605e-01 -1.15056121e+00 5.96180320e-01
-1.18911600e+00 -5.08468270e-01 -1.45781219e-01 2.46042418e+00
6.47147238e-01 2.47610062e-01 -8.59777182e-02 2.03457162e-01
3.33675832e-01 2.12286741e-01 6.21014088e-02 -2.83634394e-01
2.60328531e-01 6.32178709e-02 8.65730584e-01 2.76458263e-01
-9.39741790e-01 3.31984073e-01 7.72470045e+00 3.55443120e-01
-1.41904926e+00 -2.03168243e-02 1.35315448e-01 -8.57958049e-02
2.16379911e-01 1.96038127e-01 -4.05263036e-01 7.92467520e-02
6.08388484e-01 -2.64336199e-01 -1.69137761e-01 6.51169300e-01
2.50551701e-01 -6.84483707e-01 -1.12705600e+00 1.08057773e+00
2.28819415e-01 -1.36553061e+00 -5.35086632e-01 1.38362035e-01
2.31405020e-01 2.74454534e-01 -7.21918270e-02 1.92128703e-01
-3.29935968e-01 -6.72819257e-01 2.41587639e-01 3.34594846e-01
8.41684043e-01 -3.06487739e-01 1.09770799e+00 -1.90547302e-01
-1.07337391e+00 1.93855196e-01 8.17549527e-02 5.29474542e-02
3.68786231e-02 3.28845352e-01 -1.29665887e+00 8.05654824e-01
5.58684886e-01 6.87446594e-01 -4.23719376e-01 1.54835117e+00
-1.23564072e-01 2.85161018e-01 -1.93176627e-01 -1.28577650e-01
-3.25249545e-02 4.81317677e-02 5.57115138e-01 1.14753008e+00
1.35119762e-02 -4.42413241e-02 -7.20728636e-02 -1.31657898e-01
5.65485716e-01 3.82254094e-01 -5.25563717e-01 -7.19984248e-02
-3.37377936e-02 1.39724994e+00 -1.11769748e+00 -2.67284699e-02
-5.64225078e-01 5.37183821e-01 -2.27419776e-03 -9.70168114e-02
-5.38931549e-01 -2.37912610e-01 6.69331372e-01 4.18977737e-01
-3.47060174e-01 -2.91459680e-01 -5.26209116e-01 -8.09243083e-01
-2.92616598e-02 -8.05747330e-01 8.64131689e-01 -3.49138111e-01
-3.65199089e-01 4.54135180e-01 2.07354069e-01 -1.79906166e+00
-2.28598356e-01 -8.50399017e-01 -1.84431180e-01 5.15200794e-01
-1.28531373e+00 -5.99130988e-01 -5.68044603e-01 -6.43211231e-02
5.28676569e-01 2.29230627e-01 1.27987266e+00 6.77978098e-02
-2.38062292e-01 5.40624082e-01 -1.36220232e-01 3.06715798e-02
1.31880426e+00 -1.19210756e+00 -3.94865125e-01 3.83303642e-01
-3.06089640e-01 6.46316469e-01 8.60691130e-01 -5.48301756e-01
-1.55756819e+00 -5.34759283e-01 5.69170117e-01 -6.07835352e-01
4.64356571e-01 2.23663285e-01 -3.76918107e-01 5.64312577e-01
2.72450298e-01 -2.76617199e-01 1.23147142e+00 2.41605952e-01
3.82248193e-01 -1.98756397e-01 -9.85551775e-01 8.40735555e-01
7.80631423e-01 -7.08306860e-03 -7.10549414e-01 2.43674263e-01
-5.68142049e-02 -1.08006680e+00 -1.51624238e+00 9.95138168e-01
7.73570836e-01 -9.26742494e-01 8.52823853e-01 -7.47367293e-02
7.07188174e-02 -1.82498917e-01 2.30041355e-01 -1.05831516e+00
-2.38190755e-01 -2.04788670e-01 3.71944666e-01 3.27743709e-01
4.06423569e-01 -6.53964460e-01 1.02624166e+00 6.00346208e-01
-5.98760784e-01 -1.07101917e+00 -7.51178205e-01 -4.24791366e-01
1.66305806e-02 -3.78594637e-01 -1.80268452e-01 5.93368471e-01
7.41033852e-01 -2.77209014e-01 1.23883419e-01 4.59222682e-02
3.99552524e-01 2.20333889e-01 5.82737982e-01 -1.02973580e+00
-5.45667820e-02 -8.47584009e-01 -9.08491552e-01 -4.97170597e-01
-4.20794278e-01 -1.06895506e+00 1.61598206e-01 -2.03273559e+00
3.81926000e-01 -4.88017231e-01 -3.13613087e-01 1.25339404e-01
-2.39446405e-02 4.50261384e-01 2.25678291e-02 3.55642557e-01
-3.42450082e-01 1.09503761e-01 1.36504972e+00 -1.59873571e-02
-2.18272015e-01 2.75143653e-01 -7.84356594e-01 6.25954628e-01
7.80873835e-01 -7.29042172e-01 -3.14460337e-01 -8.67599621e-02
9.15760919e-02 1.81179002e-01 -2.29022533e-01 -1.20061874e+00
4.66684669e-01 2.31850874e-02 1.39192045e-01 -6.64556623e-01
1.46635458e-01 -9.50919628e-01 3.85215551e-01 1.44087541e+00
-5.58905005e-02 1.03686728e-01 4.96132791e-01 3.41902375e-01
-5.81535757e-01 -2.47699693e-01 7.33467638e-01 -2.70326287e-01
-8.50950003e-01 1.93358451e-01 -8.05689633e-01 -1.77154496e-01
1.29175985e+00 -6.91145182e-01 3.91890891e-02 3.14905085e-02
-9.63100612e-01 1.39074996e-01 7.08641291e-01 5.90786874e-01
4.35833603e-01 -1.22918820e+00 -6.25387132e-01 -8.78459886e-02
6.75913930e-01 2.46443331e-01 4.69754428e-01 1.80035031e+00
-1.07906318e+00 8.74164820e-01 -1.25433460e-01 -7.70212531e-01
-1.37787187e+00 5.56547344e-01 5.80923855e-01 -5.33228159e-01
-6.11551821e-01 1.25522405e-01 9.94857475e-02 -3.94376099e-01
5.23389816e-01 -4.77131844e-01 -3.24874312e-01 7.72408471e-02
2.01263383e-01 9.90223438e-02 4.47414786e-01 -3.60590696e-01
-9.63672996e-01 6.72793150e-01 -1.31294027e-01 3.85107845e-02
8.14304709e-01 -1.29660800e-01 8.61995444e-02 5.01058817e-01
9.16795790e-01 1.52765244e-01 -2.95311600e-01 1.10147886e-01
1.83210477e-01 -1.57162756e-01 -3.05070486e-02 -9.59751487e-01
-8.37548852e-01 5.41902304e-01 1.06657898e+00 -5.05741000e-01
1.39796686e+00 7.89147057e-03 9.25505459e-02 -1.47446617e-02
8.11164796e-01 -7.83522606e-01 -2.25733981e-01 1.74121588e-01
9.35333252e-01 -1.33557415e+00 3.49593461e-01 -7.32643306e-01
-6.55753791e-01 1.17524087e+00 2.44914412e-01 -2.76361201e-02
7.21040428e-01 5.02956569e-01 5.56796134e-01 -2.12924093e-01
-3.87471139e-01 -2.07115822e-02 4.54750538e-01 5.81964195e-01
1.09107363e+00 2.38796130e-01 -1.39549506e+00 1.44235030e-01
-1.21764831e-01 2.53370792e-01 6.18907571e-01 1.21827841e+00
-6.30540103e-02 -1.31005299e+00 -8.83584246e-02 5.13415277e-01
-6.06568635e-01 1.91822007e-01 2.10427828e-02 1.20749938e+00
-5.22697210e-01 6.28487051e-01 -3.01889151e-01 -1.56535581e-03
6.90470576e-01 -2.12526858e-01 7.04706550e-01 -7.56118417e-01
-7.04417169e-01 7.62798861e-02 4.53888029e-01 -8.85020196e-01
-4.88987029e-01 -9.06031549e-01 -1.18646979e+00 2.89802581e-01
-2.06984326e-01 4.51422364e-01 1.11914217e+00 5.08573472e-01
9.93396416e-02 8.66105080e-01 1.08211167e-01 -6.43974423e-01
-3.61135066e-01 -8.30241382e-01 -4.37442899e-01 3.69712077e-02
5.64935524e-03 -8.48960459e-01 -3.43031734e-01 -4.56926107e-01] | [14.082347869873047, -3.3373119831085205] |
3429158f-5757-4562-a39d-391ca1fddc56 | deep-rigid-instance-scene-flow | 1904.08913 | null | http://arxiv.org/abs/1904.08913v1 | http://arxiv.org/pdf/1904.08913v1.pdf | Deep Rigid Instance Scene Flow | In this paper we tackle the problem of scene flow estimation in the context
of self-driving. We leverage deep learning techniques as well as strong priors
as in our application domain the motion of the scene can be composed by the
motion of the robot and the 3D motion of the actors in the scene. We formulate
the problem as energy minimization in a deep structured model, which can be
solved efficiently in the GPU by unrolling a Gaussian-Newton solver. Our
experiments in the challenging KITTI scene flow dataset show that we outperform
the state-of-the-art by a very large margin, while being 800 times faster. | ['Raquel Urtasun', 'Shenlong Wang', 'Wei-Chiu Ma', 'Yuwen Xiong', 'Rui Hu'] | 2019-04-18 | deep-rigid-instance-scene-flow-1 | http://openaccess.thecvf.com/content_CVPR_2019/html/Ma_Deep_Rigid_Instance_Scene_Flow_CVPR_2019_paper.html | http://openaccess.thecvf.com/content_CVPR_2019/papers/Ma_Deep_Rigid_Instance_Scene_Flow_CVPR_2019_paper.pdf | cvpr-2019-6 | ['scene-flow-estimation'] | ['computer-vision'] | [-1.12160191e-01 -7.62243792e-02 5.92762604e-03 -2.18900777e-02
-2.23378494e-01 -5.00081301e-01 5.70516646e-01 -1.35460541e-01
-7.98570514e-01 4.40401286e-01 1.71977773e-01 -3.08481246e-01
2.80184895e-01 -5.33088148e-01 -8.62848520e-01 -5.89370012e-01
-3.61422747e-02 6.14062488e-01 4.70073819e-01 -1.30888239e-01
2.67603874e-01 4.33208793e-01 -1.24982941e+00 -4.46647517e-02
5.55651069e-01 8.18670034e-01 1.67277902e-01 1.15561295e+00
8.98449495e-02 1.46461201e+00 -1.47454143e-01 -1.51375830e-01
4.53858376e-01 -2.77885288e-01 -1.05029321e+00 3.90506893e-01
8.10668468e-01 -5.44941962e-01 -9.19025064e-01 6.56994224e-01
2.50204682e-01 5.90579510e-01 4.00654644e-01 -1.17442787e+00
4.06813383e-01 -8.31597820e-02 -7.90319622e-01 3.42329562e-01
3.58632654e-01 4.07820076e-01 7.70831525e-01 -8.03564072e-01
9.63803709e-01 1.23567545e+00 3.64403665e-01 2.16308028e-01
-1.17833173e+00 -1.81244463e-01 1.87762529e-01 3.55479866e-01
-1.04736876e+00 -4.81992930e-01 7.57283866e-01 -8.05072069e-01
1.08655024e+00 -1.72699690e-01 7.16727555e-01 8.44522059e-01
3.34262878e-01 8.77660453e-01 5.78543723e-01 -1.48628205e-01
3.98476303e-01 -2.96308666e-01 4.83349487e-02 1.10145652e+00
1.31924152e-01 5.61629161e-02 -5.05506396e-01 4.29689325e-02
9.04897988e-01 -2.27259904e-01 -3.32641155e-02 -5.88726997e-01
-1.29032171e+00 1.09999359e+00 5.82328379e-01 -1.54899105e-01
-2.60117382e-01 8.06604683e-01 4.75608468e-01 -1.14843532e-01
4.02373165e-01 1.44487424e-02 -2.98741490e-01 -3.04950207e-01
-8.31390500e-01 5.87327540e-01 1.11503625e+00 7.51238227e-01
1.01442873e+00 1.54238731e-01 1.41814157e-01 9.19901207e-02
3.34290564e-01 3.57552201e-01 7.62135834e-02 -1.56292963e+00
5.81432641e-01 1.95518494e-01 2.15740874e-01 -1.00751138e+00
-4.63849634e-01 -2.46790126e-01 -5.84095538e-01 4.66715306e-01
7.12930858e-01 -6.23743117e-01 -7.86975563e-01 1.42628181e+00
7.75682211e-01 7.98459888e-01 -1.90789133e-01 1.03741741e+00
4.96902943e-01 9.43846703e-01 -1.16557956e-01 3.85389514e-02
1.13907993e+00 -1.57681382e+00 -5.68487227e-01 -8.06153059e-01
9.15046513e-01 -6.31983876e-01 5.51125348e-01 2.91058272e-01
-1.12771559e+00 -6.78085446e-01 -9.69391167e-01 -4.15431768e-01
4.74737026e-02 -7.51588047e-02 7.84367919e-01 1.79182023e-01
-9.68917012e-01 5.88648617e-01 -1.32333708e+00 -2.44873196e-01
5.35314322e-01 1.96979746e-01 -5.08922696e-01 -2.01966584e-01
-5.03624141e-01 9.38095689e-01 4.00811106e-01 8.94458815e-02
-1.31547344e+00 -9.25292611e-01 -1.30508780e+00 -9.64423493e-02
5.44896662e-01 -1.10826683e+00 1.26276124e+00 -6.18344367e-01
-1.72159302e+00 8.72126997e-01 -3.44275802e-01 -4.40359831e-01
9.59710896e-01 -5.68513155e-01 3.98622096e-01 2.32108042e-01
8.34442228e-02 7.05291748e-01 7.25186169e-01 -8.05278897e-01
-8.57910872e-01 -1.84521437e-01 4.71372664e-01 2.53492266e-01
2.85677582e-01 -1.54756933e-01 -5.58249295e-01 -1.98675111e-01
-3.00181419e-01 -1.34904134e+00 -7.52258062e-01 2.35019624e-01
-2.26318300e-01 1.12698562e-01 1.08703613e+00 -5.12401283e-01
6.86995804e-01 -2.06085086e+00 4.14246321e-01 -2.39059001e-01
3.08386296e-01 8.68917070e-03 1.40952766e-01 1.41496584e-01
1.46632880e-01 -5.95998168e-01 -4.15289342e-01 -7.22484410e-01
-8.94575641e-02 3.32104981e-01 -2.40150005e-01 1.01521170e+00
6.97680861e-02 9.77048993e-01 -1.19462478e+00 -5.75890541e-01
4.80720937e-01 4.44476128e-01 -8.46698046e-01 5.16555190e-01
-8.48270580e-02 6.35002673e-01 -4.35837865e-01 -1.26734763e-01
6.05867386e-01 -3.99411052e-01 -4.38865162e-02 3.20607834e-02
-2.75967449e-01 2.44298384e-01 -1.34535325e+00 2.31487942e+00
-6.18855536e-01 1.03468978e+00 2.98384130e-01 -1.06449389e+00
4.22840029e-01 -1.09009512e-01 6.02127433e-01 -2.20551163e-01
2.05086350e-01 -5.54288737e-02 -2.33230796e-02 -6.91078961e-01
5.41983247e-01 -1.43747348e-02 8.22723191e-03 3.65342796e-01
8.35555047e-03 -4.27809417e-01 4.62905467e-01 3.17344695e-01
1.36689198e+00 2.91623980e-01 3.82231995e-02 -3.60008866e-01
5.11060953e-01 3.95413369e-01 4.11641628e-01 6.31720006e-01
-2.21317455e-01 2.98141152e-01 5.23756862e-01 -7.20970809e-01
-1.01765907e+00 -7.75887728e-01 2.00640634e-01 9.16049898e-01
3.81239951e-01 -3.93165410e-01 -7.84061074e-01 -6.74094677e-01
-1.21410303e-02 4.88716394e-01 -5.72172165e-01 1.68783516e-01
-1.08621573e+00 -5.44779480e-01 1.28233865e-01 6.40468121e-01
5.12650132e-01 -8.21704626e-01 -1.37349474e+00 3.83897811e-01
-1.48609251e-01 -1.87880945e+00 -5.98576844e-01 -1.35780508e-02
-7.12996483e-01 -1.24776375e+00 -3.64701509e-01 -7.21643448e-01
4.98391569e-01 3.31237853e-01 1.09550130e+00 -7.13297650e-02
-7.11805582e-01 3.22731614e-01 1.68849289e-01 -8.32024515e-02
-1.51237160e-01 8.15952197e-02 -3.27994943e-01 1.01220094e-01
-3.20039600e-01 -4.22789842e-01 -8.40970039e-01 -5.85856959e-02
-6.91854417e-01 3.06143641e-01 1.46274015e-01 5.02456605e-01
3.59119743e-01 9.71549898e-02 -1.99739322e-01 -8.60603333e-01
-3.38632539e-02 -4.94526535e-01 -9.19734776e-01 -3.34706873e-01
2.42323771e-01 1.71715587e-01 4.98423278e-01 -2.39026502e-01
-1.11020863e+00 7.76614904e-01 -6.07204661e-02 -6.31860077e-01
-2.05093607e-01 1.73949927e-01 1.23065494e-01 -3.16977382e-01
4.32459086e-01 -1.50293633e-01 -2.47119427e-01 -1.18059404e-01
5.56002855e-01 -4.18708138e-02 7.88404584e-01 -5.68870246e-01
7.33345509e-01 1.15233040e+00 6.31074429e-01 -9.36129093e-01
-9.09451723e-01 -6.55844390e-01 -8.65829408e-01 -4.92781639e-01
1.30809510e+00 -9.41322684e-01 -8.76738608e-01 4.71264362e-01
-1.47767997e+00 -8.49897385e-01 -3.87242407e-01 5.04829764e-01
-9.28859651e-01 4.23960149e-01 -6.72549844e-01 -4.68748391e-01
-3.20576765e-02 -1.36006153e+00 1.18297374e+00 1.27830297e-01
-2.00408027e-02 -1.36999583e+00 5.55069745e-01 3.99552405e-01
2.76312888e-01 7.23857999e-01 3.31733108e-01 -5.30064432e-03
-8.48635077e-01 -2.50159167e-02 -1.51583850e-01 -5.88091388e-02
-3.22942019e-01 9.72156227e-02 -8.63610029e-01 -2.12759301e-01
1.59647956e-01 -1.73431307e-01 1.00452459e+00 5.37172437e-01
8.91444147e-01 -9.63267982e-02 -2.40086079e-01 1.09651458e+00
1.52927959e+00 -2.14638859e-01 3.59794676e-01 3.05646509e-01
1.15312147e+00 7.94656873e-01 5.38374424e-01 5.87118566e-01
5.97310483e-01 6.81680620e-01 7.09192097e-01 -1.18746974e-01
-1.34591401e-01 -2.27757007e-01 2.30554014e-01 3.42116356e-01
8.48504305e-02 -1.73116401e-01 -1.12693739e+00 6.14479065e-01
-2.39108062e+00 -8.05360675e-01 -5.97824395e-01 1.75560176e+00
1.61397949e-01 9.16623026e-02 1.00378290e-01 -5.31400479e-02
3.18562627e-01 5.01683474e-01 -6.03147984e-01 -3.45462471e-01
2.58707076e-01 4.50929403e-02 6.54580235e-01 1.06351542e+00
-1.41788113e+00 1.16497755e+00 6.65232134e+00 3.55012000e-01
-1.10311508e+00 1.24100871e-01 6.46918237e-01 -2.57369697e-01
2.86230564e-01 2.91462213e-01 -6.98796213e-01 2.82271862e-01
8.37945521e-01 -2.27233365e-01 3.17166418e-01 8.48292530e-01
4.07530099e-01 -4.29250866e-01 -1.27066064e+00 1.16621459e+00
8.97093117e-02 -1.42602408e+00 -5.17087579e-01 2.29355872e-01
9.78374124e-01 4.97606516e-01 -1.98652729e-01 3.33556032e-04
4.48107272e-01 -7.90557861e-01 8.44161272e-01 2.45506167e-01
2.85657376e-01 -5.26632011e-01 4.27776992e-01 5.71084082e-01
-1.13391316e+00 -8.50132853e-03 -2.25266188e-01 -4.94981110e-01
6.93812191e-01 6.53000176e-01 -4.29521710e-01 3.86974066e-01
5.36199510e-01 1.08779585e+00 -1.66316152e-01 1.06094909e+00
-3.36868346e-01 4.19314355e-01 -4.36647445e-01 4.79706824e-01
6.96877480e-01 -4.24947977e-01 6.74179554e-01 1.43578792e+00
5.90290837e-02 1.07656315e-01 5.38021982e-01 7.36631691e-01
-2.08813455e-02 -1.48058489e-01 -5.71090221e-01 5.71961999e-01
-3.57650042e-01 1.38332891e+00 -7.03593433e-01 -4.51165766e-01
-4.43852216e-01 1.04008114e+00 5.50504208e-01 3.69241804e-01
-9.22327459e-01 -9.03608948e-02 8.32231998e-01 1.20353699e-01
5.20189404e-01 -6.19435370e-01 -1.06248222e-01 -1.36811173e+00
-3.48321311e-02 -1.85909048e-01 2.47174844e-01 -7.13567317e-01
-7.78994322e-01 3.82228047e-01 -6.82407022e-02 -8.33045542e-01
-3.27252626e-01 -7.02922046e-01 -8.50325823e-01 5.30614674e-01
-1.72704268e+00 -6.93308592e-01 -6.88670933e-01 6.88014925e-01
6.82139993e-01 2.76733160e-01 2.13615358e-01 1.54192001e-01
-7.72782922e-01 -2.74235338e-01 -1.55251548e-01 1.23820059e-01
3.26050133e-01 -1.19188619e+00 6.84190392e-01 1.17152333e+00
5.39733768e-02 -1.04218863e-01 7.45749295e-01 -3.85251254e-01
-1.63976765e+00 -1.13805914e+00 7.34431803e-01 -3.55582386e-01
1.02127552e+00 -5.36496758e-01 -6.04202986e-01 8.49927545e-01
2.44776189e-01 6.91058576e-01 1.74527302e-01 -5.33966362e-01
3.88207124e-03 1.83093145e-01 -7.48334050e-01 3.50986689e-01
1.19417131e+00 -3.66719872e-01 -2.67477721e-01 5.11057079e-01
5.00904441e-01 -9.73152220e-01 -4.35118675e-01 1.40790537e-01
3.15213650e-01 -8.79910588e-01 9.62845445e-01 -5.84592223e-01
6.70724332e-01 -3.22653294e-01 9.22576711e-02 -1.07910383e+00
-2.66691506e-01 -9.58625376e-01 -5.01108587e-01 5.05235672e-01
-8.46649557e-02 -1.37377858e-01 1.07685721e+00 7.17351854e-01
-8.30934197e-02 -5.32365203e-01 -1.04652750e+00 -4.44997758e-01
-1.07309900e-01 -5.64007282e-01 -1.43130878e-02 5.64164281e-01
-1.25762179e-01 5.30355275e-01 -4.36882585e-01 1.36186063e-01
8.83928061e-01 -7.16591813e-03 1.18440688e+00 -8.61328840e-01
-4.46166933e-01 -4.03022736e-01 -5.82067251e-01 -1.56283665e+00
7.26866424e-01 -7.01267481e-01 3.44020456e-01 -1.51968777e+00
9.81599391e-02 -8.32619146e-02 3.59097749e-01 1.60026491e-01
-3.86909068e-01 -2.12588198e-02 3.88498783e-01 -5.63327074e-02
-9.10845518e-01 5.22217572e-01 1.21192575e+00 -1.60819575e-01
-8.86928365e-02 -1.71221063e-01 -7.40384907e-02 9.33588445e-01
4.36745405e-01 -6.13560498e-01 -3.54609936e-01 -8.93333614e-01
1.91503406e-01 2.33184978e-01 4.59602058e-01 -1.10014093e+00
6.59205794e-01 -1.34553984e-01 1.53516568e-02 -4.79452401e-01
5.55010200e-01 -9.72426414e-01 -7.45160282e-02 7.32564270e-01
-1.55740768e-01 1.58629403e-01 3.37091357e-01 6.73040450e-01
-3.78172360e-02 -9.14497375e-02 8.28113616e-01 -1.78752825e-01
-7.72090793e-01 6.12399936e-01 -4.36989307e-01 4.59846914e-01
1.26044858e+00 3.97173166e-02 -2.64386743e-01 -3.73299241e-01
-4.81149971e-01 3.77374619e-01 4.85424101e-01 2.97940671e-01
3.87375653e-01 -9.52468514e-01 -8.83994281e-01 5.19604981e-02
-3.24810922e-01 4.35244590e-01 3.17334682e-01 8.57571721e-01
-1.16170180e+00 4.09313828e-01 -1.17923647e-01 -8.83532286e-01
-9.35682893e-01 5.57978332e-01 5.73903620e-01 -3.86317700e-01
-8.37358236e-01 7.24422038e-01 7.19782770e-01 -1.05921634e-01
1.76732078e-01 -1.88619390e-01 -5.93079440e-03 -2.74231464e-01
4.74227309e-01 8.93065751e-01 -8.95109400e-02 -9.40217376e-01
-5.46817839e-01 7.79813409e-01 2.21535742e-01 -2.98698723e-01
1.34485412e+00 -9.56680104e-02 -9.46224555e-02 2.11772084e-01
1.59031081e+00 -1.72548786e-01 -1.85436237e+00 -3.69231068e-02
-2.11537465e-01 -6.62505746e-01 3.80079389e-01 -1.94509383e-02
-1.29566252e+00 8.82798553e-01 1.41431376e-01 -2.03628555e-01
7.39019454e-01 -6.14611581e-02 9.04277265e-01 3.13816100e-01
2.30021983e-01 -8.26253295e-01 8.60555544e-02 6.84851229e-01
4.78645772e-01 -1.29395223e+00 7.16795474e-02 -4.29213613e-01
-5.18924475e-01 1.12163770e+00 5.04056454e-01 -8.48027885e-01
8.08240473e-01 5.37810087e-01 6.24019234e-03 -1.13444947e-01
-9.55412805e-01 -2.08814070e-01 1.81432590e-01 2.34651059e-01
6.99192435e-02 -2.82123715e-01 2.37830997e-01 -2.83705354e-01
-1.87079996e-01 -7.37082120e-03 7.36998975e-01 1.04488087e+00
-3.15657377e-01 -7.93386519e-01 -1.07457377e-01 -4.09793332e-02
-3.07217479e-01 3.34222376e-01 -5.89703768e-02 6.11148894e-01
9.60061401e-02 1.04373264e+00 3.48999351e-01 7.02604279e-02
5.07130504e-01 -4.57586527e-01 6.14765346e-01 -4.55469906e-01
-4.83618051e-01 -9.84653831e-02 -7.72442743e-02 -1.19428694e+00
-5.44070721e-01 -8.12423229e-01 -1.25323057e+00 -4.18743402e-01
8.94539729e-02 -1.32280022e-01 6.60865009e-01 1.10174918e+00
1.80335954e-01 5.33987343e-01 6.53844178e-01 -1.41368449e+00
1.10075984e-03 -5.36948621e-01 -2.48746023e-01 4.19547260e-01
7.60284424e-01 -4.36360031e-01 -4.96166468e-01 2.58583277e-01] | [8.554159164428711, -1.8919517993927002] |
242a6df8-e131-42dc-b48c-c81a56a9399d | transfer-learning-for-scene-text-recognition | 2201.03180 | null | https://arxiv.org/abs/2201.03180v1 | https://arxiv.org/pdf/2201.03180v1.pdf | Transfer Learning for Scene Text Recognition in Indian Languages | Scene text recognition in low-resource Indian languages is challenging because of complexities like multiple scripts, fonts, text size, and orientations. In this work, we investigate the power of transfer learning for all the layers of deep scene text recognition networks from English to two common Indian languages. We perform experiments on the conventional CRNN model and STAR-Net to ensure generalisability. To study the effect of change in different scripts, we initially run our experiments on synthetic word images rendered using Unicode fonts. We show that the transfer of English models to simple synthetic datasets of Indian languages is not practical. Instead, we propose to apply transfer learning techniques among Indian languages due to similarity in their n-gram distributions and visual features like the vowels and conjunct characters. We then study the transfer learning among six Indian languages with varying complexities in fonts and word length statistics. We also demonstrate that the learned features of the models transferred from other Indian languages are visually closer (and sometimes even better) to the individual model features than those transferred from English. We finally set new benchmarks for scene-text recognition on Hindi, Telugu, and Malayalam datasets from IIIT-ILST and Bangla dataset from MLT-17 by achieving 6%, 5%, 2%, and 23% gains in Word Recognition Rates (WRRs) compared to previous works. We further improve the MLT-17 Bangla results by plugging in a novel correction BiLSTM into our model. We additionally release a dataset of around 440 scene images containing 500 Gujarati and 2535 Tamil words. WRRs improve over the baselines by 8%, 4%, 5%, and 3% on the MLT-19 Hindi and Bangla datasets and the Gujarati and Tamil datasets. | ['C. V. Jawahar', 'Rohit Saluja', 'Sanjana Gunna'] | 2022-01-10 | null | null | null | null | ['scene-text-recognition'] | ['computer-vision'] | [ 3.84849310e-01 -6.07985914e-01 2.74917006e-01 -5.68230271e-01
-8.41141820e-01 -7.86591828e-01 8.92462015e-01 -2.69345760e-01
-8.07570100e-01 5.66565156e-01 2.80597478e-01 -6.93141699e-01
3.67446095e-01 -6.57704353e-01 -9.78938401e-01 -5.72388172e-01
1.71354473e-01 3.79821569e-01 2.31845781e-01 -2.26434439e-01
1.90973669e-01 4.78296995e-01 -1.00303650e+00 6.71422839e-01
8.76731157e-01 5.14999747e-01 2.85565078e-01 1.05680048e+00
-6.64542615e-02 9.40538466e-01 -7.21518099e-01 -5.09603500e-01
4.46493596e-01 -5.23832381e-01 -5.97837746e-01 3.76528352e-01
9.84552145e-01 -5.43958902e-01 -6.82119787e-01 7.74355888e-01
6.25068963e-01 -5.39673269e-02 9.32619452e-01 -7.00152755e-01
-1.21113098e+00 8.39038193e-01 -9.15611088e-01 2.19281930e-02
4.60357927e-02 1.48339048e-01 6.92913055e-01 -1.18620253e+00
5.37633419e-01 1.40640819e+00 7.32620955e-01 4.51657176e-01
-1.05800176e+00 -6.42110586e-01 1.44512758e-01 3.71492743e-01
-1.26645565e+00 -5.53065658e-01 3.89881730e-01 -2.63823688e-01
1.22247696e+00 3.40451330e-01 5.06677963e-02 1.41326737e+00
-8.72551277e-02 9.24968779e-01 1.34149325e+00 -8.16347718e-01
-1.07531510e-01 3.39533448e-01 -5.02197966e-02 5.82299948e-01
-1.36906113e-02 -4.23278183e-01 -3.44505042e-01 4.26266015e-01
6.35447383e-01 -2.85097152e-01 -6.34509325e-02 -3.91642824e-02
-1.42591310e+00 8.76875877e-01 1.88457474e-01 2.79431701e-01
1.08994402e-01 -1.74410604e-02 5.29740930e-01 4.52195674e-01
3.73148948e-01 1.40686661e-05 -6.15040004e-01 -1.81353018e-01
-7.20896423e-01 -1.45295039e-01 5.95482171e-01 1.12592649e+00
5.46508431e-01 6.27229989e-01 -8.15535372e-04 1.49807179e+00
6.20533861e-02 1.05801618e+00 7.31457055e-01 -3.94582540e-01
9.68872011e-01 7.90056586e-02 -1.76789403e-01 -7.78512001e-01
-8.46921504e-02 -9.88190398e-02 -8.08067799e-01 3.91085744e-02
5.21478176e-01 -2.92247474e-01 -1.53126228e+00 1.42283189e+00
-1.25192672e-01 -4.60152030e-01 3.76972914e-01 5.70886016e-01
7.86076665e-01 1.00023699e+00 -6.38324916e-02 1.72967881e-01
1.11991942e+00 -1.26499915e+00 -4.38155383e-01 -3.72677565e-01
8.65793943e-01 -1.06633043e+00 1.61165941e+00 2.49429494e-01
-7.77721226e-01 -6.69466972e-01 -9.21823740e-01 -1.82027057e-01
-5.14639378e-01 5.27070403e-01 1.94091350e-01 8.72317970e-01
-9.53448355e-01 2.40860030e-01 -5.93429208e-01 -9.24306393e-01
3.32015246e-01 1.30030483e-01 -3.56434137e-01 -3.93570453e-01
-9.56214488e-01 1.02149487e+00 4.46877986e-01 -1.89205091e-02
-7.99576521e-01 -4.10683244e-01 -9.31180060e-01 -2.59357303e-01
7.94560015e-02 1.14093773e-01 1.01327002e+00 -1.11438453e+00
-1.61451328e+00 9.11252856e-01 -1.30355926e-02 -3.78656238e-01
7.60017514e-01 -1.76428169e-01 -5.99183917e-01 -3.73354144e-02
-3.18630897e-02 6.33004606e-01 7.58114815e-01 -1.18971407e+00
-6.61263704e-01 -2.35171393e-01 -3.14836830e-01 3.73572707e-01
-4.65132773e-01 3.10505271e-01 -6.78563058e-01 -1.05527198e+00
-1.90354824e-01 -9.21832800e-01 1.12910584e-01 -2.17170045e-01
-4.18599784e-01 2.84480453e-01 9.97129560e-01 -1.13396502e+00
5.54876328e-01 -2.24891353e+00 -7.42760301e-02 -7.75570050e-02
-3.84277105e-01 2.96812296e-01 -6.29970133e-01 3.88536930e-01
-4.70176013e-03 7.26546049e-02 -2.23144874e-01 -1.99533209e-01
3.10965115e-03 4.31122124e-01 -3.94620836e-01 4.55663830e-01
1.82184964e-01 8.76142859e-01 -3.11168492e-01 -3.91109824e-01
4.10897046e-01 3.94182771e-01 -2.96092689e-01 -1.44584715e-01
1.12484936e-02 2.57111281e-01 -5.48458099e-03 5.50691068e-01
6.82839096e-01 1.52762458e-01 1.28832042e-01 3.46302427e-02
-2.60472782e-02 3.49406213e-01 -9.56105649e-01 1.35961497e+00
-6.16906047e-01 9.32408869e-01 -2.15580150e-01 -9.57640946e-01
1.05980754e+00 -7.01651424e-02 -1.18102849e-01 -1.13790965e+00
5.85482381e-02 2.46119097e-01 3.12067628e-01 -3.19607019e-01
6.38047755e-01 -3.86825926e-03 -2.27918223e-01 4.96662289e-01
1.24576434e-01 -3.55724573e-01 3.31237227e-01 5.08179106e-02
6.70054436e-01 4.44052741e-02 -1.81727409e-01 -1.84884742e-01
2.84705788e-01 -9.67618749e-02 4.21245135e-02 1.02957118e+00
4.25912850e-02 9.26188648e-01 2.28665397e-01 -4.06094640e-01
-1.43902600e+00 -1.12655592e+00 -1.99546725e-01 1.58833587e+00
-4.34453696e-01 -1.23417102e-01 -7.92387545e-01 -5.74955046e-01
-1.72780022e-01 9.61323559e-01 -6.25010133e-01 3.73319983e-02
-7.92264819e-01 -8.91963542e-01 8.90999198e-01 7.46833801e-01
9.62046742e-01 -1.20841956e+00 -2.04317659e-01 -5.26660159e-02
1.16528030e-02 -1.63295555e+00 -6.37850702e-01 3.62772167e-01
-6.33151233e-01 -4.12970603e-01 -1.05312216e+00 -1.04498804e+00
5.29258311e-01 2.66604513e-01 6.47185862e-01 -3.95794839e-01
-1.93978220e-01 3.10781389e-01 -5.48301756e-01 -4.15801615e-01
-6.54778242e-01 1.91545054e-01 2.91654412e-02 1.54255331e-01
2.01255813e-01 -8.08622539e-02 -6.05745614e-02 2.82083631e-01
-8.92393529e-01 2.48110473e-01 4.44603354e-01 7.34112680e-01
7.44947419e-02 -2.14780733e-01 1.42698437e-01 -9.47584271e-01
4.27010536e-01 1.15546184e-02 -5.33392847e-01 4.36974347e-01
-2.98033804e-01 -2.07270935e-01 7.96137035e-01 -7.09376752e-01
-1.16322470e+00 5.52631946e-06 -6.62040934e-02 -1.53899208e-01
-2.51296163e-01 3.81702811e-01 -1.80144548e-01 1.73628777e-02
7.40250707e-01 6.55510187e-01 -4.48441029e-01 -3.94285709e-01
6.46632195e-01 1.09735262e+00 6.70980990e-01 -6.19565010e-01
1.02988493e+00 3.14325809e-01 -7.11095631e-01 -1.38894093e+00
-4.30305153e-01 5.46984710e-02 -8.02963972e-01 1.91808000e-01
8.44778001e-01 -1.03113067e+00 -1.61440820e-01 1.20290589e+00
-1.04308677e+00 -8.17420900e-01 -1.58203617e-01 4.15052861e-01
-3.34073246e-01 6.47925615e-01 -1.04287016e+00 -5.52537441e-01
-1.86808690e-01 -1.14414084e+00 9.76421773e-01 2.53056027e-02
4.79327701e-02 -1.04555333e+00 -1.67960495e-01 2.69117802e-01
5.25084674e-01 -1.16156682e-01 1.16017842e+00 -8.09621453e-01
-5.39067537e-02 -6.23708367e-02 -5.05515695e-01 7.24043906e-01
3.98980290e-01 8.65030661e-02 -9.28233922e-01 -5.11129081e-01
-1.18188821e-01 -5.45571983e-01 9.36405420e-01 7.49430284e-02
9.25082147e-01 -2.24769756e-01 2.73469120e-01 7.63791740e-01
1.28255272e+00 4.12015945e-01 8.49284768e-01 6.60827458e-01
1.14151907e+00 6.56428993e-01 1.73496425e-01 2.04231918e-01
4.69823658e-01 5.05451798e-01 -1.34918705e-01 -3.56574684e-01
-3.24478358e-01 -2.44205669e-01 8.21693301e-01 1.22024584e+00
8.33764076e-02 -3.79720569e-01 -1.20485806e+00 7.19096899e-01
-1.45995998e+00 -5.14349759e-01 -2.75931776e-01 2.13394403e+00
9.48194087e-01 1.01342350e-01 9.27065164e-02 2.65167374e-02
6.91948533e-01 1.16601929e-01 -4.14259553e-01 -7.68499315e-01
-6.47881806e-01 2.18353420e-01 7.79943943e-01 4.65430230e-01
-1.13949645e+00 1.52888227e+00 6.28140116e+00 8.84915590e-01
-1.45029891e+00 -3.77664552e-03 8.79576623e-01 2.36061458e-02
8.35803570e-04 -3.39860767e-01 -7.47804582e-01 2.69102067e-01
1.09629798e+00 1.77471325e-01 7.00167000e-01 5.24670959e-01
-5.86317899e-03 9.43469927e-02 -1.02818346e+00 1.02577031e+00
4.71400499e-01 -1.00453103e+00 3.62277657e-01 -1.33839592e-01
9.94163990e-01 5.67464948e-01 3.81368756e-01 5.13407290e-01
5.15314698e-01 -1.25444043e+00 9.13310409e-01 -9.00030658e-02
1.16179478e+00 -6.42825961e-01 4.38236892e-01 9.89965126e-02
-7.62859643e-01 1.68966070e-01 -6.79964840e-01 3.81095037e-02
-3.16693932e-01 -6.41931295e-02 -9.49656069e-01 2.32361570e-01
8.47386360e-01 7.00239241e-01 -1.04475999e+00 5.03674090e-01
-2.63477117e-01 9.80187178e-01 -3.81393969e-01 -2.67371356e-01
4.98811275e-01 -1.32155180e-01 2.97959801e-02 1.77610862e+00
2.37649292e-01 -3.71696711e-01 -6.94519579e-02 5.26901603e-01
-2.59081930e-01 4.28225189e-01 -6.74498379e-01 7.93131590e-02
2.02008169e-02 1.04018652e+00 -7.81710029e-01 -4.66716290e-01
-6.47793531e-01 1.35222399e+00 1.97680667e-01 7.73443639e-01
-9.85768616e-01 -6.53783381e-01 1.88323498e-01 -2.19778493e-01
7.95713842e-01 -4.47955519e-01 -3.80817413e-01 -1.33514762e+00
1.67572901e-01 -1.48178601e+00 9.73404944e-02 -8.72170687e-01
-1.25706172e+00 7.96854138e-01 -1.38234958e-01 -8.29359293e-01
-1.03128262e-01 -1.10956216e+00 -4.87948328e-01 1.07430828e+00
-1.30693471e+00 -1.54600906e+00 5.31763956e-02 7.94135213e-01
1.03229368e+00 -4.88323629e-01 7.45573342e-01 2.01137871e-01
-5.57004988e-01 1.00633955e+00 7.67285705e-01 7.13832319e-01
9.85241354e-01 -1.16733003e+00 1.07669520e+00 9.86559093e-01
4.98367071e-01 3.14713776e-01 3.97463381e-01 -3.83309692e-01
-1.43558288e+00 -1.36864376e+00 7.08351016e-01 -5.88627398e-01
8.84222209e-01 -9.75862205e-01 -1.08592284e+00 1.03977227e+00
5.32675862e-01 -1.79178804e-01 3.52691144e-01 -1.02565698e-02
-7.93914855e-01 -4.39438149e-02 -8.59009027e-01 1.00495315e+00
6.67815804e-01 -6.83122933e-01 -4.86516953e-01 3.37370455e-01
6.66073740e-01 -2.94705003e-01 -5.58389306e-01 9.79849473e-02
5.60608089e-01 -7.02094138e-01 7.31329799e-01 -6.38104975e-01
5.06005943e-01 -5.59517406e-02 -7.11206377e-01 -1.23855805e+00
-1.66563928e-01 -3.03951740e-01 6.06715739e-01 1.41824114e+00
6.89657569e-01 -7.00220525e-01 4.86639768e-01 1.56588420e-01
9.70171914e-02 -4.26777117e-02 -8.36599648e-01 -8.31667840e-01
7.88284004e-01 -7.31912673e-01 9.32275951e-02 1.20028353e+00
-3.89517576e-01 6.73065424e-01 -6.49640560e-01 -1.11914389e-01
3.63764077e-01 -1.60144582e-01 8.62486959e-01 -4.76108849e-01
-4.13497180e-01 -2.73776084e-01 -1.41764790e-01 -1.02656722e+00
4.81314436e-02 -9.06612873e-01 1.05252944e-01 -1.38131583e+00
4.09695894e-01 -1.25337273e-01 -9.15232226e-02 7.46667445e-01
2.34813318e-02 6.15755618e-01 5.82499146e-01 1.11035034e-01
-3.94823134e-01 3.62505108e-01 1.08755660e+00 -4.97187972e-01
-2.05869935e-02 -4.14504826e-01 -4.09020841e-01 6.65104806e-01
8.87693226e-01 -3.00594479e-01 -1.81879178e-01 -1.16129947e+00
-1.75968587e-01 -3.50706607e-01 8.30489472e-02 -8.39337468e-01
3.22853625e-02 1.97608601e-02 7.26166964e-01 -5.37494361e-01
1.25512078e-01 -5.64096570e-01 -4.04576153e-01 3.67292076e-01
-3.81698370e-01 6.64263815e-02 6.95657670e-01 1.28213242e-01
1.23986699e-01 -1.12319864e-01 8.62780213e-01 -4.76520173e-02
-7.52944827e-01 -2.17196062e-01 -8.11364710e-01 3.64345275e-02
6.72825456e-01 -5.19248843e-01 -4.31364715e-01 -5.37422001e-01
-4.25405711e-01 5.50689176e-02 4.90217179e-01 8.30084205e-01
6.24014020e-01 -1.14867425e+00 -1.27293372e+00 3.19206566e-01
2.08124623e-01 -3.91679972e-01 2.71424651e-02 5.27636647e-01
-8.41644943e-01 4.77525353e-01 -3.13424259e-01 -7.16262519e-01
-1.31542552e+00 3.73256981e-01 2.47288272e-01 -2.05803793e-02
-3.83278996e-01 4.36000615e-01 3.86803925e-01 -1.01667142e+00
3.13966274e-01 -3.96116316e-01 8.49429965e-02 -2.67222881e-01
3.20269138e-01 1.67489901e-01 1.12758100e-01 -8.59848499e-01
-2.49552876e-01 8.83453548e-01 -5.50245106e-01 -3.39805067e-01
1.37358582e+00 -1.41502962e-01 4.87942323e-02 5.91625988e-01
1.21157110e+00 2.67821014e-01 -1.09988046e+00 -3.88750046e-01
-1.58067308e-02 -2.25728735e-01 -2.89627761e-01 -1.10256219e+00
-8.79735589e-01 1.13015640e+00 7.54619002e-01 -1.21112049e-01
1.07660985e+00 -1.70823410e-01 5.43203592e-01 6.65301621e-01
1.44545332e-01 -1.06700301e+00 1.52341276e-02 1.06553698e+00
9.46823061e-01 -1.19698656e+00 -2.89233357e-01 2.12363899e-02
-1.04612935e+00 1.15794945e+00 5.66715300e-01 6.26418144e-02
5.09193130e-02 3.62397492e-01 5.34526467e-01 3.95138562e-01
-4.50599194e-01 -1.01946173e-02 1.54548451e-01 8.20338130e-01
5.69800317e-01 6.12271428e-02 1.33893276e-02 1.61033586e-01
-3.59204769e-01 -5.41347980e-01 7.72566378e-01 8.32481384e-01
-2.57536829e-01 -1.01147926e+00 -5.70348501e-01 3.17174643e-01
-5.82911849e-01 -4.88711774e-01 -5.59515178e-01 1.13460827e+00
-2.56711692e-01 8.91654015e-01 9.16763991e-02 -4.31851804e-01
3.92096072e-01 4.73244525e-02 4.56008732e-01 -5.66926301e-01
-4.61420894e-01 4.38072443e-01 5.37522472e-02 1.27624005e-01
-1.47706091e-01 -6.35281980e-01 -1.01893198e+00 -3.23178649e-01
-1.97074994e-01 -2.95324773e-01 7.82993436e-01 7.56139159e-01
-6.33156449e-02 3.98741245e-01 6.03179276e-01 -6.27916396e-01
-5.55613875e-01 -1.39762843e+00 -5.02947748e-01 3.02865773e-01
1.00967936e-01 1.19055554e-01 -2.56623626e-01 4.17705327e-01] | [11.870315551757812, 2.352259635925293] |
bbdfa774-dec7-4bbe-86e3-a15c5671d1c9 | natlogattack-a-framework-for-attacking | 2307.02849 | null | https://arxiv.org/abs/2307.02849v1 | https://arxiv.org/pdf/2307.02849v1.pdf | NatLogAttack: A Framework for Attacking Natural Language Inference Models with Natural Logic | Reasoning has been a central topic in artificial intelligence from the beginning. The recent progress made on distributed representation and neural networks continues to improve the state-of-the-art performance of natural language inference. However, it remains an open question whether the models perform real reasoning to reach their conclusions or rely on spurious correlations. Adversarial attacks have proven to be an important tool to help evaluate the Achilles' heel of the victim models. In this study, we explore the fundamental problem of developing attack models based on logic formalism. We propose NatLogAttack to perform systematic attacks centring around natural logic, a classical logic formalism that is traceable back to Aristotle's syllogism and has been closely developed for natural language inference. The proposed framework renders both label-preserving and label-flipping attacks. We show that compared to the existing attack models, NatLogAttack generates better adversarial examples with fewer visits to the victim models. The victim models are found to be more vulnerable under the label-flipping setting. NatLogAttack provides a tool to probe the existing and future NLI models' capacity from a key viewpoint and we hope more logic-based attacks will be further explored for understanding the desired property of reasoning. | ['Xiaodan Zhu', "Zi'ou Zheng"] | 2023-07-06 | null | null | null | null | ['natural-language-inference'] | ['natural-language-processing'] | [ 5.51237762e-02 4.98961806e-01 -1.18237145e-01 -1.38346776e-01
-4.41067159e-01 -8.90333295e-01 8.70279610e-01 2.06198916e-01
-7.01711923e-02 7.69437373e-01 4.71115112e-02 -8.37452888e-01
-3.50126743e-01 -1.42257762e+00 -8.35326314e-01 -5.62958717e-01
-1.47321418e-01 6.95301712e-01 4.26900625e-01 -5.99953949e-01
2.72342622e-01 6.35508955e-01 -1.24261057e+00 5.29574871e-01
7.61440456e-01 6.66357517e-01 -6.42353952e-01 6.64573669e-01
-6.80180863e-02 1.69734478e+00 -7.63503790e-01 -9.31263745e-01
2.55552769e-01 -3.94611418e-01 -1.08578980e+00 -9.58213270e-01
4.75444794e-01 -7.38459826e-02 -4.39189255e-01 1.69666421e+00
2.36253589e-01 -1.81751937e-01 4.77963030e-01 -1.67738616e+00
-7.83384025e-01 1.31581140e+00 -7.95745254e-02 1.42250702e-01
4.89371151e-01 3.46932560e-01 1.00273263e+00 -8.45765099e-02
5.15004933e-01 1.75318110e+00 5.47038674e-01 7.03528881e-01
-9.43330288e-01 -8.94854546e-01 -6.87545761e-02 6.37613952e-01
-1.29171920e+00 -7.05101341e-02 8.77843142e-01 -1.36161536e-01
7.85378516e-01 4.20087993e-01 2.04279974e-01 1.36495996e+00
4.48713660e-01 6.00431800e-01 1.38206506e+00 -6.98815227e-01
4.26441282e-01 2.85005659e-01 3.16645712e-01 8.24314773e-01
4.77048695e-01 4.51623112e-01 -3.39702010e-01 -5.06972730e-01
3.00366551e-01 -4.24052179e-01 -5.17771319e-02 -1.73040852e-01
-7.34857976e-01 1.22503853e+00 5.75156987e-01 3.40463817e-01
-1.00608774e-01 4.78222072e-01 6.13021374e-01 5.15867829e-01
-9.46015865e-02 7.44959295e-01 -2.70473659e-01 7.33744502e-02
-5.71251154e-01 6.07096076e-01 1.22513676e+00 6.29193485e-01
3.07036102e-01 -3.67227430e-03 -1.68991983e-01 -5.92494756e-02
3.70517135e-01 6.95993900e-01 1.32224232e-01 -1.14080310e+00
3.49429667e-01 7.54315078e-01 -7.55442306e-03 -1.32060921e+00
-2.29893669e-01 -2.34760016e-01 -6.25395060e-01 4.81540442e-01
7.68892407e-01 -6.55397177e-02 -1.72509134e-01 1.91155028e+00
2.39363223e-01 2.12133959e-01 3.67616534e-01 5.03238142e-01
3.54926020e-01 5.44509232e-01 7.61628570e-03 1.29274577e-01
1.38259327e+00 -6.33551359e-01 -6.12092733e-01 1.96797326e-01
7.18370557e-01 -3.04122746e-01 8.72197807e-01 6.42332971e-01
-1.03782141e+00 -1.35672018e-01 -1.08006334e+00 3.15372676e-01
-7.20365107e-01 -7.26962209e-01 9.23665941e-01 1.13566363e+00
-7.99699068e-01 3.32077026e-01 -7.02408195e-01 -2.46673170e-03
5.82990885e-01 2.20905274e-01 -1.05546087e-01 -1.64320573e-01
-2.01607728e+00 1.33967376e+00 6.04408383e-01 7.85467699e-02
-1.23950768e+00 -3.81900251e-01 -6.54480577e-01 -1.47411851e-02
8.53162348e-01 -5.03786147e-01 1.03534007e+00 -7.65739560e-01
-1.60834098e+00 7.08495200e-01 3.31478029e-01 -1.14345396e+00
6.42788470e-01 3.79886553e-02 -5.70427537e-01 2.56520480e-01
-2.85253406e-01 2.22148985e-01 4.62386757e-01 -1.14993203e+00
-1.86867863e-01 -1.95164546e-01 8.04064393e-01 -3.76525611e-01
-1.84835881e-01 2.86357284e-01 4.32164788e-01 -3.63399327e-01
-3.21978360e-01 -9.19213951e-01 -3.12989615e-02 5.22467680e-02
-5.89926481e-01 -4.38185185e-01 8.19278777e-01 -5.51468469e-02
1.16868532e+00 -1.55359304e+00 -4.58281301e-02 6.39670134e-01
1.99655116e-01 7.18642473e-01 6.27587885e-02 7.44497657e-01
-1.13684274e-01 3.09825063e-01 -1.36533797e-01 4.37125921e-01
4.01642948e-01 2.52464563e-01 -9.19657886e-01 4.96193409e-01
4.43546139e-02 1.04647660e+00 -1.01983821e+00 -5.01161456e-01
1.28318369e-01 1.11964919e-01 -6.84524894e-01 1.61813006e-01
-7.62361169e-01 2.35103369e-01 -4.24716771e-01 5.79207718e-01
5.53887725e-01 -1.42083734e-01 3.46497118e-01 -1.02734357e-01
3.08050126e-01 3.43144506e-01 -1.16202605e+00 1.36091590e+00
-2.77219117e-01 4.05989379e-01 -2.14698464e-01 -1.14367211e+00
8.23853195e-01 2.78500497e-01 -2.52689362e-01 -5.69091678e-01
3.21435690e-01 7.13069066e-02 4.86985505e-01 -5.09464502e-01
4.50037904e-02 -3.59626353e-01 -4.61077124e-01 7.02689171e-01
-3.11881959e-01 -2.29907930e-01 6.40311465e-02 4.99409348e-01
1.37716985e+00 -1.69542506e-02 4.27483886e-01 -4.12527233e-01
9.94366586e-01 1.69932507e-02 3.04717839e-01 1.20529056e+00
-4.23143864e-01 -1.58260137e-01 8.95642042e-01 -6.20467186e-01
-6.15623236e-01 -1.35678709e+00 3.35254520e-02 6.90636277e-01
7.05827996e-02 -4.11487341e-01 -8.12393427e-01 -1.01162589e+00
-1.07385352e-01 1.19279659e+00 -6.29523277e-01 -8.05819929e-01
-4.95076805e-01 -3.16564173e-01 1.74047637e+00 4.33492064e-01
8.68445218e-01 -1.49496865e+00 -6.31609440e-01 -1.05658166e-01
-9.33417082e-02 -8.50773931e-01 4.15829211e-01 1.05488442e-01
-2.67564088e-01 -1.26554787e+00 1.35582566e-01 -3.42926770e-01
5.07460594e-01 -5.05476832e-01 1.13679373e+00 3.20529163e-01
-1.74771808e-02 1.47994295e-01 -2.85195917e-01 -6.25439942e-01
-1.13256764e+00 -3.62943597e-02 7.81062469e-02 -3.16658795e-01
6.35087132e-01 -5.16655684e-01 -2.04376519e-01 2.19768554e-01
-1.15958476e+00 -4.33092356e-01 1.38068944e-01 6.23652875e-01
-2.42251262e-01 5.81320763e-01 6.68975949e-01 -1.34243274e+00
8.14575315e-01 -5.03182173e-01 -8.54386926e-01 5.95004618e-01
-4.72747445e-01 3.93307358e-01 1.20269799e+00 -3.28590691e-01
-9.62158918e-01 -5.46804368e-01 -1.13038562e-01 -3.81605297e-01
-1.80112466e-01 5.77268422e-01 -3.17069471e-01 -1.24669164e-01
1.11370647e+00 -1.06132254e-01 -1.48024842e-01 2.42194548e-01
6.01179421e-01 4.85082477e-01 5.63919842e-01 -1.36104250e+00
1.11646843e+00 4.85308319e-01 4.29077208e-01 -1.89923793e-01
-9.17529702e-01 4.39914316e-01 -1.35758847e-01 -7.65530318e-02
6.02498829e-01 -4.00586486e-01 -1.39137983e+00 3.84733409e-01
-1.38667142e+00 -4.23820257e-01 -2.15136290e-01 2.36215852e-02
-4.75321114e-01 4.87781942e-01 -6.39647782e-01 -1.11337543e+00
-2.91341513e-01 -1.17695761e+00 3.71564716e-01 9.73349530e-03
-4.82219458e-01 -1.18348956e+00 3.24049622e-01 3.42500478e-01
4.66106862e-01 3.05407703e-01 1.37320173e+00 -1.13112414e+00
-5.40926039e-01 -3.69155854e-01 -3.91650423e-02 3.19349021e-01
-9.93478522e-02 7.65332133e-02 -1.16043115e+00 -9.46197361e-02
2.12064222e-01 -6.51673198e-01 4.44382429e-01 -1.76694199e-01
1.08603871e+00 -5.61261117e-01 3.65360826e-02 2.62535155e-01
1.44141233e+00 3.19579035e-01 1.11740255e+00 3.67429107e-01
4.10019964e-01 6.21898115e-01 3.61102581e-01 6.16673343e-02
1.64677203e-01 3.94547105e-01 7.15442836e-01 5.30057728e-01
1.48577228e-01 -5.01571417e-01 5.01703620e-01 2.87235737e-01
1.20330274e-01 -3.12314898e-01 -1.23628676e+00 1.25968114e-01
-1.68653429e+00 -1.40519381e+00 -1.40357588e-03 1.93421841e+00
9.76609647e-01 5.88451326e-01 -1.08415440e-01 4.72954363e-01
6.52926505e-01 3.45261171e-02 -3.33505899e-01 -8.43973339e-01
-1.15998991e-01 6.02044344e-01 3.22520167e-01 6.80827260e-01
-8.47068131e-01 1.16162980e+00 6.26709557e+00 9.42589641e-01
-9.79972124e-01 -3.22850607e-02 1.98807567e-01 2.19023898e-01
-5.40304661e-01 3.74083132e-01 -6.19780898e-01 4.32900518e-01
1.19838846e+00 -3.18686068e-01 6.81387246e-01 8.68821383e-01
-2.04383940e-01 -6.20020926e-02 -1.29740739e+00 3.11215371e-01
1.33840367e-01 -1.41036987e+00 3.26775014e-01 -1.50315419e-01
6.28970921e-01 -4.52561438e-01 7.30283111e-02 5.69947124e-01
9.71812308e-01 -1.36220646e+00 6.82967544e-01 4.84126568e-01
3.21142942e-01 -1.04123402e+00 9.29970801e-01 4.74220097e-01
-7.16110468e-01 -2.66968787e-01 -2.06234589e-01 -3.60225797e-01
-2.16534510e-01 3.26142341e-01 -7.34375477e-01 5.32936871e-01
3.35860670e-01 2.71485075e-02 -5.82313597e-01 3.73111606e-01
-7.38722086e-01 6.96648717e-01 -2.42561519e-01 -3.08494866e-01
3.10191870e-01 4.12917584e-02 4.31606472e-01 9.09779131e-01
-2.25832477e-01 -2.11208254e-01 4.25639451e-02 1.31743741e+00
-1.00732140e-01 -4.87092316e-01 -1.01556289e+00 -9.11493823e-02
7.85217285e-01 8.75431895e-01 -6.76701486e-01 -4.04627711e-01
-7.05173314e-02 4.72392976e-01 4.59827185e-01 4.01598774e-02
-1.23269308e+00 -3.96998882e-01 2.04170346e-01 -1.39977753e-01
-6.05563559e-02 1.21433526e-01 -2.22824678e-01 -1.02010548e+00
-2.41722047e-01 -1.44998097e+00 4.76312608e-01 -6.78973436e-01
-1.46515763e+00 5.12706339e-01 2.65189886e-01 -7.04757154e-01
-2.20871076e-01 -7.60453641e-01 -9.55775142e-01 6.67104304e-01
-1.29696369e+00 -1.13641191e+00 2.32040733e-02 8.72923374e-01
-6.45521730e-02 -3.54308426e-01 1.29586399e+00 -2.52738073e-02
-5.02344251e-01 7.86189854e-01 -3.72153252e-01 3.65276426e-01
5.12842417e-01 -1.05689740e+00 1.93395652e-02 1.11212265e+00
1.87743753e-01 1.07011819e+00 9.66477275e-01 -6.28326535e-01
-1.47060382e+00 -9.09585714e-01 7.14737356e-01 -7.42996812e-01
1.04428005e+00 -2.36405492e-01 -8.28898668e-01 8.52443755e-01
3.23892683e-01 -3.39720473e-02 3.79933566e-01 1.16386130e-01
-9.17823792e-01 -2.35193074e-01 -1.45916188e+00 9.09553766e-01
7.74246454e-01 -9.14752066e-01 -1.03772748e+00 3.55076134e-01
6.25581205e-01 -1.65579598e-02 -5.40971398e-01 1.66024610e-01
5.13850570e-01 -1.10942006e+00 9.22514439e-01 -8.38913858e-01
6.95815086e-01 -3.31386149e-01 -3.34275395e-01 -9.81342375e-01
7.67689198e-02 -7.78038740e-01 -2.52174228e-01 1.22260404e+00
3.45393956e-01 -9.46936011e-01 7.41212010e-01 7.49658763e-01
3.98045689e-01 -4.77815360e-01 -9.99060333e-01 -8.95296752e-01
7.34931350e-01 -6.57094419e-01 5.61109602e-01 1.01361370e+00
3.79628807e-01 2.11208120e-01 -2.50840127e-01 3.47850323e-01
6.55825078e-01 -1.34946057e-03 6.94985151e-01 -1.20800638e+00
-3.35790157e-01 -4.50471014e-01 -5.26131809e-01 -2.76514947e-01
7.77886927e-01 -1.24199593e+00 -2.16530889e-01 -1.08061171e+00
-8.12123418e-02 -3.10737818e-01 -2.87451357e-01 5.58091521e-01
9.89899114e-02 4.10207249e-02 1.95333093e-01 -1.64248988e-01
-6.26815975e-01 1.75048128e-01 9.55139041e-01 -3.65935445e-01
3.02890539e-01 -1.71261370e-01 -8.72004926e-01 9.80964005e-01
9.59472418e-01 -6.52045488e-01 -6.77114725e-01 1.50949426e-03
8.77568662e-01 -8.44405964e-02 6.91385627e-01 -1.08540332e+00
5.97033918e-01 -3.66903484e-01 -3.21256906e-01 -1.14710234e-01
-1.44488811e-02 -1.01006460e+00 5.62701859e-02 8.98593128e-01
-8.34742010e-01 7.35412017e-02 6.49426803e-02 3.66941005e-01
-2.74160244e-02 -4.64945287e-01 9.04275119e-01 -3.47464859e-01
-3.61562937e-01 -1.77816212e-01 -5.72284818e-01 3.04358542e-01
1.15972030e+00 2.90657878e-01 -7.26708889e-01 -4.45765406e-01
-3.77065569e-01 2.49867551e-02 2.31197879e-01 3.61147197e-03
4.79079843e-01 -1.09125090e+00 -6.24167800e-01 1.11157343e-01
-7.65244886e-02 -2.33313575e-01 2.66219191e-02 5.67344606e-01
-8.74982357e-01 4.91632044e-01 -2.19885319e-01 -1.56483084e-01
-9.20460105e-01 1.03635895e+00 4.39272672e-01 -7.06402719e-01
-4.35771614e-01 6.57659590e-01 -1.08834282e-02 -6.14306390e-01
2.89730936e-01 -3.19585830e-01 1.63081307e-02 -3.82753551e-01
5.31013846e-01 4.42884415e-01 -1.24747641e-01 -8.63636807e-02
-4.59684461e-01 1.58742726e-01 -9.74494517e-02 -4.81493920e-02
8.76240969e-01 4.14307356e-01 -4.77295160e-01 3.54322731e-01
8.39132786e-01 1.91319197e-01 -3.77659410e-01 -1.19073957e-01
1.24789193e-01 -1.63619637e-01 -2.85057127e-01 -1.21550333e+00
-7.98191369e-01 8.70560348e-01 2.28575021e-01 5.56176901e-01
6.66119516e-01 -1.58433065e-01 5.37983119e-01 8.93911302e-01
7.39770412e-01 -7.15711296e-01 3.04414728e-03 6.14369512e-01
7.85492301e-01 -9.34108913e-01 -7.76291452e-03 -3.20423096e-01
-4.48650926e-01 1.02597487e+00 5.60452402e-01 -5.68960130e-01
4.97922570e-01 6.76660419e-01 9.14337337e-02 -3.45707119e-01
-8.05897057e-01 2.90017515e-01 -2.78355926e-01 6.40344203e-01
-1.84534434e-02 8.70255679e-02 -6.87083602e-02 4.88992721e-01
-3.99524033e-01 1.04710288e-01 5.39147973e-01 8.54466736e-01
-1.69459775e-01 -1.31003916e+00 -6.99020624e-01 3.21546756e-03
-6.08401358e-01 -2.12965921e-01 -6.09965205e-01 8.02155495e-01
1.53333798e-01 1.18600559e+00 -4.25343961e-01 -2.55703658e-01
4.83658947e-02 2.56975114e-01 5.86136580e-01 -3.26672494e-01
-8.54964375e-01 -7.74082720e-01 2.68320054e-01 -4.22403604e-01
-4.13860023e-01 -1.25281781e-01 -1.27400923e+00 -8.58219206e-01
-2.48275772e-01 4.11174506e-01 1.77153379e-01 9.69279110e-01
1.09977927e-02 5.08230269e-01 4.01207268e-01 -2.14295790e-01
-1.03807092e+00 -6.08073533e-01 -3.31171721e-01 3.37990403e-01
-8.10138434e-02 -5.56749403e-01 -6.44471943e-01 -3.02106887e-01] | [8.933648109436035, 7.156285762786865] |
e5e9b04d-8498-4199-b6d7-ba8362286cd1 | evaluating-factual-consistency-of-texts-with | 2305.13309 | null | https://arxiv.org/abs/2305.13309v1 | https://arxiv.org/pdf/2305.13309v1.pdf | Evaluating Factual Consistency of Texts with Semantic Role Labeling | Automated evaluation of text generation systems has recently seen increasing attention, particularly checking whether generated text stays truthful to input sources. Existing methods frequently rely on an evaluation using task-specific language models, which in turn allows for little interpretability of generated scores. We introduce SRLScore, a reference-free evaluation metric designed with text summarization in mind. Our approach generates fact tuples constructed from Semantic Role Labels, applied to both input and summary texts. A final factuality score is computed by an adjustable scoring mechanism, which allows for easy adaption of the method across domains. Correlation with human judgments on English summarization datasets shows that SRLScore is competitive with state-of-the-art methods and exhibits stable generalization across datasets without requiring further training or hyperparameter tuning. We experiment with an optional co-reference resolution step, but find that the performance boost is mostly outweighed by the additional compute required. Our metric is available online at https://github.com/heyjing/SRLScore. | ['Michael Gertz', 'Dennis Aumiller', 'Jing Fan'] | 2023-05-22 | null | null | null | null | ['semantic-role-labeling', 'text-summarization'] | ['natural-language-processing', 'natural-language-processing'] | [ 4.67512369e-01 5.29407322e-01 -4.13531005e-01 -4.71050173e-01
-1.38019538e+00 -1.01307607e+00 1.11980951e+00 5.34398079e-01
-3.49085331e-01 1.12661171e+00 8.34574342e-01 -4.01538648e-02
-1.01899013e-01 -6.27282619e-01 -3.35523993e-01 -2.32906282e-01
4.27803576e-01 7.56373525e-01 2.27606595e-01 -3.58331114e-01
5.60694396e-01 -2.76795268e-01 -1.50239015e+00 6.78427100e-01
1.16053545e+00 8.58957946e-01 -2.17208248e-02 8.65352333e-01
-3.25973958e-01 9.57410693e-01 -1.11132836e+00 -7.36469150e-01
-3.34551111e-02 -6.93140447e-01 -1.10496306e+00 -8.96964073e-02
4.05997366e-01 3.78077812e-02 7.15982169e-03 8.71873498e-01
5.20206809e-01 1.27188355e-01 7.89619863e-01 -1.07227027e+00
-6.34739876e-01 1.09413493e+00 -1.10827051e-01 1.99478939e-01
8.12717199e-01 1.03727341e-01 1.53853333e+00 -7.81945646e-01
8.55879307e-01 1.17601299e+00 4.66201484e-01 5.31966746e-01
-1.33903801e+00 -3.29738408e-01 -3.05298120e-02 3.72644067e-02
-9.94704664e-01 -6.55318022e-01 4.77872580e-01 -3.25076103e-01
1.08044028e+00 5.49596429e-01 2.26622105e-01 1.14631891e+00
-6.61145002e-02 8.35474908e-01 8.95177245e-01 -4.36588794e-01
3.94754976e-01 3.24269831e-01 2.26111904e-01 3.81985307e-01
6.65565014e-01 -5.22082329e-01 -6.18138611e-01 -4.14099336e-01
1.56699076e-01 -5.81022859e-01 -4.54739630e-01 1.71698127e-02
-1.28867209e+00 7.62152433e-01 1.08017094e-01 1.84331715e-01
-3.32203269e-01 -7.51881450e-02 6.80782199e-01 3.64333153e-01
5.96545339e-01 9.80260313e-01 -5.93788207e-01 -3.71493042e-01
-1.24843848e+00 5.39745271e-01 1.08678937e+00 1.01012635e+00
5.38614213e-01 -2.88218498e-01 -7.01246262e-01 1.00795841e+00
-2.61911422e-01 4.58747804e-01 8.99503589e-01 -1.18200433e+00
7.70836771e-01 8.31394494e-01 3.07828963e-01 -6.87382817e-01
-3.17101717e-01 -2.23533124e-01 -6.10289574e-01 -2.49299705e-01
4.00269270e-01 -2.98082262e-01 -4.76101667e-01 1.66737938e+00
4.54240553e-02 -3.99026960e-01 3.64107519e-01 5.59555173e-01
9.39059019e-01 4.34426188e-01 -5.65032512e-02 -3.21981221e-01
1.33004379e+00 -8.03820729e-01 -6.46380484e-01 -7.97435567e-02
7.62260675e-01 -7.56591797e-01 1.39632666e+00 4.01087999e-01
-1.23781836e+00 -1.87563151e-01 -1.09168828e+00 -2.97672719e-01
-2.29158118e-01 2.24991173e-01 4.28623766e-01 6.23348534e-01
-1.01905489e+00 6.75115287e-01 -4.72798437e-01 -4.08951908e-01
2.75856525e-01 2.08512247e-01 -2.45404556e-01 2.26770371e-01
-1.20331597e+00 8.94511342e-01 6.39325023e-01 -5.16569495e-01
-3.17657828e-01 -5.08368969e-01 -7.13019252e-01 9.42378789e-02
6.11535609e-01 -1.02773166e+00 1.72502315e+00 -8.86481702e-01
-1.53311419e+00 7.46519387e-01 -4.08740073e-01 -5.93885303e-01
7.18718708e-01 -1.84446752e-01 -2.64166594e-01 2.59310633e-01
4.79651690e-01 4.25961375e-01 4.62194949e-01 -1.05315971e+00
-5.71171761e-01 -7.32969418e-02 2.13327616e-01 5.04022181e-01
-2.98162222e-01 1.23834841e-01 -1.57884046e-01 -7.07463801e-01
-1.39948592e-01 -6.07823431e-01 -2.50679790e-03 -5.76134086e-01
-6.14698648e-01 -5.95199943e-01 2.27750316e-01 -5.82350791e-01
1.59595525e+00 -1.41170502e+00 -2.82577798e-02 2.15321276e-02
5.16217668e-03 1.38852254e-01 -1.93138018e-01 6.09153867e-01
1.31771147e-01 3.59126776e-01 -4.93383378e-01 -1.67417705e-01
1.38052538e-01 -2.21243590e-01 -5.43638110e-01 -1.54013216e-01
3.20432246e-01 9.04589593e-01 -1.12684977e+00 -5.39208233e-01
-2.09097207e-01 -1.22741632e-01 -4.78535056e-01 2.88928282e-02
-5.55838645e-01 -4.19158861e-02 -3.52943212e-01 3.13204318e-01
1.61524579e-01 -5.02716720e-01 2.14290947e-01 1.69732660e-01
1.32117271e-01 9.90984738e-01 -1.08655047e+00 1.73987281e+00
-3.97536218e-01 4.45602566e-01 -3.47984701e-01 -6.99335992e-01
8.50574374e-01 3.47412467e-01 1.01407386e-01 -5.56173444e-01
-1.11068245e-02 3.25374633e-01 -1.52319402e-01 -4.08601105e-01
9.87895668e-01 -4.97848839e-02 -3.65150779e-01 8.73332024e-01
6.44672215e-02 -3.82411659e-01 7.32742965e-01 6.97416008e-01
1.29984081e+00 -1.90257300e-02 6.16442502e-01 -2.42808804e-01
5.64681709e-01 3.48605365e-01 3.78937542e-01 8.89442563e-01
2.83695519e-01 6.73967004e-01 7.40908802e-01 7.87827186e-03
-1.00969803e+00 -9.05976057e-01 -6.05108738e-02 1.08863449e+00
-2.37685591e-02 -8.45470667e-01 -9.32796419e-01 -9.10102785e-01
3.66054266e-03 1.36001146e+00 -5.24650335e-01 -5.72029687e-02
-5.24384141e-01 -7.16026664e-01 7.33968198e-01 4.48587626e-01
2.51978874e-01 -1.13773179e+00 -6.28638864e-01 2.82851726e-01
-6.02251530e-01 -8.61957729e-01 -4.89447773e-01 -2.70773321e-02
-8.92630279e-01 -1.05740345e+00 -4.58814383e-01 -3.47433925e-01
4.01836962e-01 1.12308182e-01 1.49660528e+00 -7.15162158e-02
6.10074326e-02 3.13243747e-01 -4.76466715e-01 -4.61824834e-01
-7.74890900e-01 5.18117011e-01 -8.73558670e-02 -4.22535717e-01
4.41093594e-01 -2.94784039e-01 -5.09511948e-01 1.88252792e-01
-9.43019986e-01 -1.06826704e-02 4.61078107e-01 9.29151237e-01
2.67340750e-01 -2.20518693e-01 1.06990230e+00 -1.30054307e+00
1.23401070e+00 -5.24626911e-01 -2.94788718e-01 4.29051310e-01
-8.40041518e-01 3.91222268e-01 7.38466322e-01 -7.74533823e-02
-1.23505652e+00 -4.77290809e-01 1.32601932e-01 1.93084776e-01
-4.38137613e-02 5.89018345e-01 -1.33063585e-01 7.43551612e-01
1.14408171e+00 7.68958926e-02 -1.70756698e-01 -2.34612539e-01
5.47437191e-01 7.85952032e-01 5.38873136e-01 -6.67265892e-01
4.89014447e-01 2.13933319e-01 -4.07532513e-01 -5.98583877e-01
-1.23785114e+00 -4.70962346e-01 -3.52588356e-01 2.32043974e-02
3.39731157e-01 -8.12794149e-01 -3.33186090e-01 -7.17742294e-02
-1.27971113e+00 -3.11090410e-01 -5.88742316e-01 1.44576684e-01
-7.13364840e-01 2.96775848e-01 -4.10309672e-01 -6.75992906e-01
-7.07334101e-01 -6.18455648e-01 1.10105288e+00 -6.08698139e-03
-9.90433812e-01 -9.48453188e-01 1.18925512e-01 6.14120424e-01
2.55356550e-01 1.33747637e-01 8.30107033e-01 -1.06235218e+00
-2.87465245e-01 -3.59532923e-01 -1.25192598e-01 3.86178732e-01
1.33606985e-01 -9.98881236e-02 -9.18595433e-01 4.24713455e-02
-8.18281397e-02 -5.90987563e-01 1.08053780e+00 2.90739864e-01
1.08685935e+00 -8.63956869e-01 -7.00287372e-02 9.38602462e-02
1.10739827e+00 -2.05482781e-01 4.51957762e-01 3.60824436e-01
3.41854542e-01 7.51380026e-01 7.71148920e-01 5.40117204e-01
5.61653912e-01 5.76546311e-01 -1.27752289e-01 3.38100195e-01
-1.97634980e-01 -4.37964916e-01 4.57483917e-01 6.35224581e-01
4.48076129e-02 -6.69425428e-01 -7.77494133e-01 4.76708233e-01
-1.92489338e+00 -1.33025992e+00 -1.86708514e-02 2.29992437e+00
1.24384475e+00 3.33857059e-01 1.93002984e-01 2.19367668e-01
6.47431016e-01 1.46035805e-01 -5.34365773e-01 -5.32213211e-01
-3.27021927e-01 3.06688219e-01 3.64698559e-01 6.03237033e-01
-7.98693180e-01 8.91876519e-01 6.43166161e+00 9.30615008e-01
-8.71541560e-01 -4.36570216e-03 5.53516090e-01 -3.61981392e-01
-7.26098478e-01 6.61862493e-02 -6.89686120e-01 5.53869247e-01
1.02522779e+00 -9.63236809e-01 1.65238723e-01 6.69346035e-01
2.04144150e-01 -2.40864635e-01 -1.20225453e+00 6.48934126e-01
2.42209926e-01 -1.41123796e+00 2.10787177e-01 -3.12473714e-01
9.13053155e-01 1.77509133e-02 -2.62667686e-01 2.49011248e-01
6.05732441e-01 -8.98125768e-01 7.79647768e-01 3.21619511e-01
8.71697247e-01 -5.56069255e-01 8.42555046e-01 4.17730510e-01
-6.14551544e-01 -3.31042744e-02 -2.57261246e-01 -1.19361028e-01
9.26797315e-02 5.82938552e-01 -1.23191226e+00 6.16261184e-01
1.88696370e-01 5.78459084e-01 -7.44281411e-01 6.47716880e-01
-6.34469926e-01 7.57335782e-01 -1.19575791e-01 -3.44171703e-01
-2.93151401e-02 6.92791939e-02 7.78320909e-01 1.46753716e+00
2.68353373e-01 1.32436529e-02 9.69837680e-02 7.66546369e-01
-4.82227445e-01 2.95017630e-01 -5.80719411e-01 -4.87031452e-02
6.25357449e-01 1.16917527e+00 -5.61917067e-01 -8.10904205e-01
2.99479850e-02 9.52742815e-01 2.94214636e-01 2.03652412e-01
-5.23442447e-01 -6.11842453e-01 3.17536056e-01 2.81654298e-01
5.80112711e-02 1.97179988e-01 -7.38254428e-01 -1.39530575e+00
3.89419079e-01 -8.72492135e-01 6.63144350e-01 -7.23649144e-01
-1.18316126e+00 5.25681734e-01 1.72339603e-01 -1.08095396e+00
-8.10638845e-01 -2.29251161e-01 -7.45352924e-01 7.00871944e-01
-1.26518953e+00 -6.10565960e-01 -2.76821911e-01 1.53671905e-01
7.79317319e-01 -1.30135566e-01 8.57862711e-01 -2.46597677e-01
-4.14989710e-01 7.67550170e-01 -5.07436953e-02 -7.88501650e-02
1.03541386e+00 -1.56039822e+00 4.74783897e-01 7.94354141e-01
-2.27473099e-02 6.53807580e-01 1.02118480e+00 -6.66521430e-01
-8.52422833e-01 -1.11107683e+00 1.44338524e+00 -9.96094167e-01
7.03901827e-01 -1.54585019e-01 -9.14277375e-01 4.81552273e-01
5.68342149e-01 -7.30912209e-01 8.95999372e-01 3.39668453e-01
-5.91648817e-01 6.98672459e-02 -1.07020569e+00 5.33311903e-01
1.13461483e+00 -2.46793434e-01 -1.02279031e+00 5.05993128e-01
6.98133171e-01 -2.50129938e-01 -6.30062222e-01 2.60716975e-01
3.85937423e-01 -8.02130818e-01 5.37828267e-01 -6.87624693e-01
9.20772731e-01 -3.12368691e-01 -7.60315359e-02 -1.43524408e+00
-2.58280426e-01 -7.22763658e-01 -1.04122691e-01 1.35726345e+00
8.58806133e-01 -6.15342140e-01 5.82988322e-01 7.51731038e-01
-1.65919855e-01 -5.73244035e-01 -6.02657318e-01 -8.93956244e-01
7.31292963e-02 -2.17403769e-01 6.92317665e-01 8.89804959e-01
5.73257148e-01 9.56640601e-01 1.00034267e-01 -2.93016970e-01
4.69334215e-01 1.78652719e-01 8.33781838e-01 -1.21883237e+00
-2.64353603e-01 -8.43532443e-01 -7.31633008e-02 -7.11210132e-01
2.20553890e-01 -1.20589471e+00 -1.90933738e-02 -1.88866341e+00
3.33033144e-01 -1.23374723e-01 -1.04778036e-01 5.94149172e-01
-4.11576599e-01 1.47318169e-01 1.22672124e-02 2.27025554e-01
-9.92254019e-01 6.02598608e-01 9.03826237e-01 -6.70389831e-02
-4.19352055e-01 -3.00797913e-02 -1.38106644e+00 6.56983793e-01
1.17497635e+00 -4.30830091e-01 -6.10933721e-01 -3.84890467e-01
2.95616776e-01 2.51681674e-02 2.26079434e-01 -8.78677487e-01
2.68076986e-01 -1.83002278e-01 2.00106755e-01 -3.34984213e-01
-8.93025380e-03 -4.87096794e-02 -9.11314562e-02 1.16980426e-01
-9.27698970e-01 2.11497992e-01 -3.77532318e-02 4.55970645e-01
-2.98718184e-01 -4.89609718e-01 5.67258835e-01 -1.97921172e-01
-2.93947607e-01 -2.27381110e-01 -1.86078921e-01 6.64265931e-01
5.48568606e-01 -2.39833981e-01 -6.76401675e-01 -4.61904764e-01
-2.54035443e-01 3.10008079e-01 5.93702793e-01 4.53696221e-01
3.43732685e-01 -1.05154288e+00 -1.09978426e+00 -2.02019364e-01
3.66015941e-01 -1.77790970e-02 -2.60590669e-02 4.44031060e-01
-2.32395649e-01 5.91271162e-01 1.91371843e-01 -2.21744955e-01
-1.13194907e+00 1.27440050e-01 -8.68620500e-02 -5.84045827e-01
-4.06879991e-01 6.52512908e-01 -1.10392541e-01 -3.13908368e-01
1.49533553e-02 -3.06275219e-01 -2.37289414e-01 4.25151438e-01
6.96631491e-01 6.10519826e-01 3.32427531e-01 -3.43238384e-01
-2.30129957e-01 5.95138259e-02 -2.60341465e-01 -5.01711845e-01
1.10038602e+00 8.93314928e-03 -4.47647013e-02 4.72258538e-01
8.58011186e-01 1.43463165e-01 -7.52704918e-01 -2.52936214e-01
4.40623969e-01 -2.52981246e-01 -1.99814379e-01 -1.29643369e+00
-3.77302825e-01 4.59820479e-01 -2.69275755e-01 3.50038022e-01
9.73894477e-01 7.62507394e-02 6.62611842e-01 5.91730416e-01
2.59994835e-01 -1.33167553e+00 8.86062086e-02 6.43980086e-01
1.20756233e+00 -1.15410900e+00 1.79634362e-01 -2.17639014e-01
-1.01910508e+00 8.33317101e-01 4.47522551e-01 -1.80633441e-02
7.02478662e-02 -5.69018684e-02 -1.90663654e-02 -3.95888612e-02
-1.27142143e+00 2.08535809e-02 3.39721620e-01 2.34298721e-01
7.68570900e-01 2.42042482e-01 -6.42032146e-01 7.33131886e-01
-8.17602575e-01 -7.15877563e-02 8.04192603e-01 6.47554159e-01
-5.83988011e-01 -1.24743021e+00 -8.74154940e-02 9.54491258e-01
-5.11575222e-01 -2.19629765e-01 -7.65787303e-01 4.83751476e-01
-3.44987631e-01 1.25805521e+00 -6.26459420e-02 -8.61331373e-02
4.56810534e-01 3.40201944e-01 3.87475133e-01 -9.49774027e-01
-8.45946074e-01 -2.31330946e-01 6.08114481e-01 -3.62035424e-01
-2.71539211e-01 -9.96443152e-01 -1.23170805e+00 -2.42165565e-01
-2.38623664e-01 5.76862991e-01 4.21458811e-01 5.99062741e-01
6.26158297e-01 2.67658591e-01 4.44375873e-01 -3.92276078e-01
-9.18460965e-01 -1.19198358e+00 -1.75798446e-01 5.91299593e-01
2.08278149e-01 -3.43261570e-01 -4.23160464e-01 1.61834314e-01] | [11.974128723144531, 9.226944923400879] |
680852d8-5801-4752-ba18-1a3891b2da87 | a-scalable-architecture-for-web-deployment-of | null | null | https://aclanthology.org/L12-1226 | https://aclanthology.org/L12-1226.pdf | A Scalable Architecture For Web Deployment of Spoken Dialogue Systems | We describe a scalable architecture, particularly well-suited to cloud-based computing, which can be used for Web-deployment of spoken dialogue systems. In common with similar platforms, like WAMI and the Nuance Mobile Developer Platform, we use a client/server approach in which speech recognition is carried out on the server side; our architecture, however, differs from these systems in offering considerably more elaborate server-side functionality, based on large-scale grammar-based language processing and generic dialogue management. We describe two substantial applications, built using our framework, which we argue would have been hard to construct in WAMI or NMDP. Finally, we present a series of evaluations carried out using CALL-SLT, a speech translation game, where we contrast performance in Web and desktop versions. Task Error Rate in the Web version is only slightly inferior that in the desktop one, and the average additional latency is under half a second. The software is generally available for research purposes. | ['Manny Rayner', 'Matthew Fuchs', 'Nikos Tsourakis'] | 2012-05-01 | null | null | null | lrec-2012-5 | ['dialogue-management'] | ['natural-language-processing'] | [-1.02125414e-01 2.36896083e-01 6.95662200e-01 -2.10429519e-01
-1.05264270e+00 -8.74108970e-01 6.80145860e-01 -3.12133372e-01
-6.10974014e-01 6.25772953e-01 1.90602630e-01 -8.80880415e-01
1.56577229e-01 -2.90872157e-01 1.02964595e-01 -6.64046764e-01
-9.95754972e-02 9.54742491e-01 6.00401044e-01 -7.88101375e-01
1.63426474e-01 2.03846082e-01 -1.59477675e+00 3.11772108e-01
6.87590599e-01 6.87937975e-01 5.63942015e-01 1.35900915e+00
-5.39149761e-01 7.78899372e-01 -9.09285963e-01 -2.50269085e-01
-1.57638505e-01 -3.21506709e-01 -1.28692293e+00 1.02103680e-01
6.35121912e-02 -4.08101022e-01 3.16398311e-03 5.35220742e-01
9.64847863e-01 2.13200897e-01 -1.03464849e-01 -1.11688697e+00
2.45300487e-01 3.46076846e-01 1.25327408e-01 1.70748651e-01
1.11017299e+00 1.57977879e-01 7.85528600e-01 -6.53693378e-01
6.02943718e-01 1.10646272e+00 4.69313711e-01 7.07024872e-01
-8.58852565e-01 4.31417450e-02 -1.35961652e-01 -2.17771024e-01
-1.31744039e+00 -1.04855895e+00 1.05152220e-01 6.78362325e-02
1.58486092e+00 1.01468241e+00 2.82911390e-01 1.07749271e+00
-8.18988457e-02 7.36188173e-01 1.22591603e+00 -8.59738708e-01
3.56005758e-01 3.80003899e-01 -7.24677220e-02 4.46599483e-01
-4.28325385e-01 -6.41945124e-01 -6.52420223e-01 -5.28721869e-01
6.87003493e-01 -6.35187089e-01 -1.93498611e-01 2.19866540e-02
-1.23390901e+00 4.80445385e-01 -6.37695372e-01 6.42504156e-01
-3.57698232e-01 -3.61025900e-01 7.21268952e-01 7.28953600e-01
6.80463076e-01 7.28130117e-02 -7.75203645e-01 -1.27392900e+00
-8.10962677e-01 3.02880973e-01 1.83789575e+00 1.11262321e+00
2.98594564e-01 -1.59046218e-01 6.26423508e-02 1.02058613e+00
2.27050036e-01 2.81364858e-01 7.90187836e-01 -1.08361804e+00
5.68222582e-01 9.20492262e-02 1.77039474e-01 -2.22015649e-01
-5.87861478e-01 1.79201394e-01 -3.16132396e-01 6.19448349e-02
6.39684737e-01 -6.48124754e-01 -1.46501213e-01 1.00099325e+00
4.09334838e-01 -7.76282623e-02 2.32199565e-01 9.10189331e-01
7.19632804e-01 5.29890299e-01 -2.33659089e-01 -3.24258476e-01
1.48959100e+00 -1.08930516e+00 -7.47762620e-01 -7.32682180e-03
9.98476148e-01 -1.15763700e+00 1.66669023e+00 6.01497531e-01
-1.49883723e+00 1.83264092e-01 -6.80451870e-01 1.62383504e-02
-3.76447648e-01 -1.00755699e-01 7.49549806e-01 1.14844906e+00
-1.86406314e+00 2.31605530e-01 -9.03450608e-01 -1.05202317e+00
-7.10767627e-01 4.05116498e-01 -2.42166013e-01 4.83447254e-01
-8.50078821e-01 9.94710624e-01 8.83221552e-02 -1.80900261e-01
-3.89376074e-01 1.12861104e-01 -4.61617291e-01 1.98631227e-01
4.55422938e-01 -2.77094752e-01 2.14394236e+00 -7.07841039e-01
-2.47269559e+00 9.68099713e-01 -8.02918002e-02 -1.11507319e-01
4.81763065e-01 1.34112567e-01 -4.17018801e-01 2.12953556e-02
-1.42199606e-01 4.14258353e-02 4.27794695e-01 -7.92605281e-01
-8.95515740e-01 -3.79010499e-01 2.87275523e-01 6.78647339e-01
-2.15055197e-01 8.25965285e-01 -7.09247768e-01 5.73691055e-02
-1.28440395e-01 -8.48491132e-01 -6.72915578e-02 -6.84314966e-01
-9.98327658e-02 -3.67358059e-01 6.48281813e-01 -8.77428770e-01
1.38568473e+00 -1.84705412e+00 -6.03463054e-02 1.07471697e-01
-8.02105591e-02 3.98504049e-01 3.07567865e-02 1.10801065e+00
3.40270519e-01 8.02152529e-02 1.47136655e-02 -4.20767814e-01
2.89879978e-01 3.68898898e-01 6.80657178e-02 -1.01941429e-01
-2.74510235e-01 3.66612881e-01 -6.68486118e-01 -2.77698487e-01
-8.14785250e-03 -4.50199656e-02 -3.46333325e-01 3.24693769e-01
-1.40487522e-01 1.79460540e-01 -2.36251116e-01 4.22138214e-01
5.00780702e-01 -3.12166009e-03 6.21811092e-01 8.45839381e-01
-5.95912933e-01 9.44835484e-01 -1.08771455e+00 2.12178063e+00
-8.57582867e-01 5.15342295e-01 9.61168170e-01 -5.43581665e-01
5.97954512e-01 9.90303040e-01 1.15535133e-01 -4.33816522e-01
-2.02849492e-01 3.89562875e-01 -8.05406552e-03 -6.97361529e-01
8.10915589e-01 6.20835423e-01 5.58598079e-02 9.12052333e-01
8.98610353e-02 -2.23365098e-01 1.01293288e-01 2.66292363e-01
1.23059809e+00 7.41461888e-02 1.63712114e-01 -2.33751550e-01
6.45769179e-01 8.58236849e-02 5.34315072e-02 7.24379063e-01
-2.20398769e-01 3.71435791e-01 6.69134796e-01 -3.82862911e-02
-9.82659936e-01 -5.85612833e-01 7.21725151e-02 1.91680706e+00
-4.79835212e-01 -7.73658037e-01 -1.30452967e+00 -4.80252087e-01
-8.17616999e-01 6.89400792e-01 4.88130361e-01 6.61297083e-01
-4.46109474e-01 -5.54929614e-01 9.39095140e-01 1.32521912e-02
5.74859560e-01 -1.13387406e+00 -6.49240911e-01 4.01354641e-01
-3.01764101e-01 -1.07761896e+00 -5.56200385e-01 -6.77114278e-02
-6.66861713e-01 -6.03573442e-01 -6.07006133e-01 -7.43489027e-01
-4.65032905e-02 3.92739326e-01 1.33344841e+00 3.50625277e-01
-7.74826333e-02 8.31704915e-01 -5.57957470e-01 -2.90659875e-01
-6.70071661e-01 5.19866347e-01 6.22501001e-02 -3.57641965e-01
3.68733197e-01 -5.39498210e-01 -2.02737182e-01 1.76994205e-01
-7.72348702e-01 1.03335977e-01 2.36994237e-01 6.15070939e-01
-3.83959860e-01 -3.10920238e-01 4.96909708e-01 -9.43721592e-01
1.26672721e+00 -3.83433282e-01 -4.26674128e-01 2.78118342e-01
-5.46057343e-01 -5.28506279e-01 3.10946614e-01 6.75759539e-02
-1.25900340e+00 -8.61431211e-02 -9.74112391e-01 4.02723283e-01
-4.66367513e-01 4.28373605e-01 -1.74175501e-01 -1.63755361e-02
4.90389287e-01 5.05435765e-01 4.81224477e-01 -7.51465023e-01
1.52123138e-01 1.55203283e+00 2.10850120e-01 -5.07912099e-01
-8.39120243e-03 -1.02411039e-01 -7.78183818e-01 -1.26035762e+00
3.24070245e-01 -6.94457710e-01 -3.69177014e-01 -3.47231627e-02
5.26940882e-01 -7.42834985e-01 -1.15556657e+00 5.96128285e-01
-1.26767230e+00 -7.57611334e-01 1.83098644e-01 1.77377105e-01
-8.39373052e-01 4.58939761e-01 -8.54060709e-01 -1.20201838e+00
-6.92219317e-01 -9.54231262e-01 1.09228957e+00 2.12410867e-01
-3.62048090e-01 -1.14288449e+00 8.25679824e-02 5.32106876e-01
8.63413751e-01 -7.22699821e-01 7.41505086e-01 -1.29298544e+00
-9.57218483e-02 -2.20662415e-01 1.66422334e-02 7.39094568e-03
-2.39522085e-01 1.41893700e-01 -1.25643504e+00 -3.82637799e-01
9.44657326e-02 -4.03517514e-01 -2.16665462e-01 -1.14948370e-01
6.55524373e-01 -5.06101549e-01 1.35264210e-02 -1.62910819e-02
8.32438231e-01 2.25897148e-01 6.55873954e-01 4.79990751e-01
1.55121028e-01 8.03991437e-01 5.18257558e-01 6.76912725e-01
8.03660631e-01 1.03680706e+00 -8.41625556e-02 -1.83700994e-02
2.82423675e-01 1.92017660e-01 6.56213343e-01 1.41737509e+00
-2.34149694e-01 -2.28044152e-01 -1.19078374e+00 1.14688799e-01
-2.15428877e+00 -6.88250244e-01 -1.75418392e-01 2.30226207e+00
9.15468335e-01 -2.38151148e-01 7.99374819e-01 -3.63854989e-02
3.05938065e-01 -1.89707637e-01 8.94581527e-02 -1.05394995e+00
1.52452990e-01 3.19164574e-01 1.17346264e-01 8.70913982e-01
-6.08199477e-01 1.20535183e+00 7.30893946e+00 7.58430421e-01
-1.11872447e+00 6.10961735e-01 2.60985941e-01 -7.51381963e-02
-6.50605485e-02 -9.33966339e-02 -5.17504036e-01 4.37662750e-01
1.81285143e+00 -2.04693154e-01 8.57622743e-01 8.69170904e-01
4.72068310e-01 -4.86876398e-01 -7.51375318e-01 8.21388841e-01
-4.82200161e-02 -1.22524178e+00 -6.06435895e-01 3.19880128e-01
-1.09181471e-01 4.68813926e-01 -4.04516250e-01 3.42943430e-01
3.92760724e-01 -6.21765614e-01 3.59201789e-01 1.80937663e-01
8.52946520e-01 -6.33558154e-01 7.65305161e-01 8.45568538e-01
-7.21592665e-01 3.99328142e-01 -1.53767765e-01 -4.06978220e-01
-1.51815876e-01 -2.12107137e-01 -1.14914799e+00 4.13934350e-01
8.51490319e-01 -4.49993253e-01 -3.51516306e-01 6.99123621e-01
3.26672107e-01 6.53029859e-01 -4.04386282e-01 -3.63405973e-01
1.89595520e-01 -4.85643864e-01 5.28465867e-01 1.48629260e+00
2.14491278e-01 8.44834372e-02 3.04054052e-01 3.15869674e-02
3.85584533e-01 5.59490025e-01 -7.48514235e-01 1.43772975e-01
4.93910700e-01 1.72763991e+00 -6.16331279e-01 -4.30989176e-01
-5.91704845e-01 1.43520045e+00 1.20009452e-01 2.54310519e-01
-3.89241308e-01 -7.41552413e-01 7.11372375e-01 -1.59426793e-01
-6.53819293e-02 -6.79665387e-01 -9.98405516e-02 -1.35074222e+00
1.60488188e-01 -1.38420451e+00 7.64353052e-02 -7.14599431e-01
-7.29507923e-01 1.05878365e+00 -2.18493864e-01 -6.68072045e-01
-1.19663906e+00 -5.98236322e-01 -7.36990392e-01 1.22929585e+00
-8.36734354e-01 -9.55047131e-01 -1.92955323e-02 6.85913384e-01
7.94725716e-01 -3.19280654e-01 1.63030756e+00 2.92153329e-01
-6.36121273e-01 5.14573216e-01 3.42975616e-01 -1.47840649e-01
6.23644173e-01 -1.45677149e+00 6.65676177e-01 4.41968709e-01
1.24972366e-01 6.38743877e-01 7.01677799e-01 -2.09251761e-01
-1.75619161e+00 -2.59614795e-01 1.40400994e+00 -5.01502395e-01
8.32376838e-01 -9.16606069e-01 -7.81540334e-01 5.33331811e-01
9.21333313e-01 -6.11639678e-01 8.25588524e-01 4.53918070e-01
2.99369097e-01 1.49454832e-01 -1.16302299e+00 7.15911090e-01
8.93657446e-01 -6.92238331e-01 -1.72267094e-01 8.84046614e-01
5.23930132e-01 -6.92382336e-01 -8.04979146e-01 -5.97288311e-01
4.35104221e-01 -1.11539924e+00 3.27517152e-01 -5.22328138e-01
-1.27513200e-01 1.06335998e-01 -6.13022931e-02 -1.33269608e+00
7.00442344e-02 -1.54025733e+00 1.99836075e-01 1.39754653e+00
4.94683921e-01 -1.08538389e+00 6.57167256e-01 1.09302998e+00
-1.91396162e-01 -2.82565415e-01 -1.12009323e+00 -6.30816936e-01
7.29330117e-03 -4.44062263e-01 6.28945172e-01 4.78201330e-01
8.24029624e-01 6.35883152e-01 -4.13085036e-02 -1.98912919e-02
-1.38508439e-01 -3.81274939e-01 9.03648794e-01 -8.15130293e-01
-5.84645987e-01 -3.53677005e-01 4.03688885e-02 -1.09168971e+00
-9.68952551e-02 -5.45737386e-01 1.70231864e-01 -1.21153367e+00
-4.07708347e-01 -3.52344364e-01 4.81244415e-01 4.62700516e-01
2.56822824e-01 -4.01997231e-02 3.10139477e-01 3.36630464e-01
-6.62258208e-01 7.96252862e-02 6.52076244e-01 3.80129665e-01
-2.98267603e-01 3.55516464e-01 -4.78975147e-01 6.37054503e-01
6.70428514e-01 9.12677795e-02 -3.94616038e-01 -4.10777837e-01
3.51432036e-03 5.08168101e-01 -2.41936982e-01 -9.23387825e-01
6.20627403e-01 3.03730983e-02 -5.64667583e-01 -1.21372633e-01
3.74691308e-01 -6.98568702e-01 -2.02436715e-01 -1.29864188e-02
-1.01189241e-01 4.55638885e-01 9.98500884e-02 -7.45065734e-02
-1.08839661e-01 -4.38387811e-01 -4.63627558e-03 -3.40687543e-01
-6.02566302e-01 -2.11321712e-01 -1.40713966e+00 -1.02491595e-01
7.67845929e-01 -1.45331442e-01 -3.01103920e-01 -9.15163934e-01
-7.29345500e-01 7.40874037e-02 4.47866678e-01 1.96813032e-01
3.16988915e-01 -5.39061546e-01 -5.37125945e-01 1.47369117e-01
-1.03225624e-02 -3.92652512e-01 2.96465727e-03 1.08821976e+00
-1.02939129e+00 6.61428511e-01 4.41444814e-02 -4.05715317e-01
-1.54761744e+00 4.69419323e-02 2.37025976e-01 -1.72577530e-01
-5.92045784e-01 5.94480097e-01 -4.20029461e-01 -8.57675076e-01
3.59056592e-01 -1.24989236e-02 -3.38926278e-02 -1.89719781e-01
8.14926147e-01 4.55420345e-01 7.70498157e-01 -3.25613141e-01
-3.97282779e-01 -3.62320155e-01 -9.35163870e-02 -9.67805028e-01
9.91503060e-01 -7.01440215e-01 -5.05527079e-01 6.87167883e-01
6.20108366e-01 1.27077952e-01 -7.38757133e-01 -1.36308149e-01
4.55017000e-01 -1.20766282e-01 -4.68069762e-02 -9.59170282e-01
-3.50175589e-01 7.16741621e-01 5.19683838e-01 7.96657026e-01
9.79768157e-01 -3.05396914e-01 7.82804132e-01 7.98498094e-01
7.56270289e-01 -1.36437905e+00 -5.68089008e-01 9.54974651e-01
9.10403490e-01 -1.03726399e+00 -5.90744972e-01 -3.90445828e-01
-9.68149900e-01 1.20848036e+00 4.24913645e-01 3.41142446e-01
3.64097506e-01 6.58926606e-01 5.46079814e-01 -7.08504319e-02
-1.35116863e+00 -3.37633550e-01 -4.85684842e-01 6.00043774e-01
1.01349485e+00 2.02535868e-01 -4.03784931e-01 4.49340582e-01
-3.54270518e-01 -8.17818046e-02 7.43796051e-01 1.22435701e+00
-3.31835151e-01 -1.42420447e+00 -4.37879205e-01 1.56648859e-01
-6.41010880e-01 -4.53486890e-01 -6.70111120e-01 6.78679347e-01
-5.75104713e-01 1.52375638e+00 2.10554361e-01 -3.34538668e-01
3.48748356e-01 6.87786281e-01 1.87462613e-01 -8.69449794e-01
-1.26182163e+00 2.37005308e-01 9.58595574e-01 -6.15400374e-01
8.49176124e-02 -6.55439496e-01 -1.16962016e+00 -7.81519532e-01
-8.67849588e-02 4.13824677e-01 1.24948311e+00 8.07222784e-01
6.66862965e-01 3.47213447e-02 4.58307147e-01 -8.40087295e-01
-8.07622850e-01 -1.32461357e+00 -7.94164836e-01 -1.80218592e-01
-1.50621772e-01 2.64729798e-01 -6.15066402e-02 -1.79925367e-01] | [13.005367279052734, 7.871457576751709] |
f22c4635-551d-4dde-ae1c-9d812fb6545f | robust-statistical-ranking-theory-and | 1408.3467 | null | https://arxiv.org/abs/1408.3467v2 | https://arxiv.org/pdf/1408.3467v2.pdf | Evaluating Visual Properties via Robust HodgeRank | Nowadays, how to effectively evaluate visual properties has become a popular topic for fine-grained visual comprehension. In this paper we study the problem of how to estimate such visual properties from a ranking perspective with the help of the annotators from online crowdsourcing platforms. The main challenges of our task are two-fold. On one hand, the annotations often contain contaminated information, where a small fraction of label flips might ruin the global ranking of the whole dataset. On the other hand, considering the large data capacity, the annotations are often far from being complete. What is worse, there might even exist imbalanced annotations where a small subset of samples are frequently annotated. Facing such challenges, we propose a robust ranking framework based on the principle of Hodge decomposition of imbalanced and incomplete ranking data. According to the HodgeRank theory, we find that the major source of the contamination comes from the cyclic ranking component of the Hodge decomposition. This leads us to an outlier detection formulation as sparse approximations of the cyclic ranking projection. Taking a step further, it facilitates a novel outlier detection model as Huber's LASSO in robust statistics. Moreover, simple yet scalable algorithms are developed based on Linearized Bregman Iteration to achieve an even less biased estimator. Statistical consistency of outlier detection is established in both cases under nearly the same conditions. Our studies are supported by experiments with both simulated examples and real-world data. The proposed framework provides us a promising tool for robust ranking with large scale crowdsourcing data arising from computer vision. | ['Xiaochun Cao', 'Yuan YAO', 'Jiechao Xiong', 'Qianqian Xu', 'Qingming Huang'] | 2014-08-15 | null | null | null | null | ['graph-sampling'] | ['graphs'] | [-3.09701283e-02 -3.59333083e-02 6.63120672e-02 -3.99248376e-02
-8.28300416e-01 -5.48367500e-01 2.58092821e-01 3.30670834e-01
-1.86774984e-01 7.70437360e-01 2.12783620e-01 3.03181142e-01
-8.33428055e-02 -3.83340359e-01 -8.76927555e-01 -9.40338314e-01
1.80097520e-01 4.08163220e-01 1.25253081e-01 -2.25958601e-01
2.93105930e-01 1.97357699e-01 -1.81425548e+00 1.97603360e-01
1.13503098e+00 1.04931843e+00 -2.40532495e-02 2.65900314e-01
-4.53976840e-02 6.93741322e-01 -5.48828900e-01 -5.49329102e-01
5.08921862e-01 1.52556216e-02 -4.22742635e-01 3.22339177e-01
5.52388072e-01 -3.40186268e-01 -5.11321612e-02 1.52493203e+00
6.73939645e-01 1.31151661e-01 5.41933835e-01 -1.57109308e+00
-4.98862445e-01 1.79817155e-01 -9.97131348e-01 -1.39759749e-01
4.65568483e-01 -5.62132858e-02 9.93371189e-01 -1.24508560e+00
6.84584916e-01 1.05919778e+00 6.69457436e-01 1.46230400e-01
-9.38940048e-01 -3.20229024e-01 1.22541919e-01 2.35614017e-01
-1.59932840e+00 -2.11735979e-01 9.01143134e-01 -7.72304535e-01
3.07898992e-03 5.54299831e-01 6.00560308e-01 8.79730523e-01
-2.00174212e-01 6.81290090e-01 1.44755697e+00 -2.46661752e-01
2.52336651e-01 2.33296886e-01 1.37975395e-01 7.16826439e-01
6.36471927e-01 -3.64509642e-01 -7.46519923e-01 -5.57191789e-01
2.66994268e-01 2.17317715e-01 -3.99405003e-01 -5.51722765e-01
-1.23547506e+00 5.18592477e-01 2.42043823e-01 -4.35705334e-02
-1.56168714e-01 -1.21258989e-01 5.27958751e-01 1.35243401e-01
6.18464053e-01 6.73730448e-02 -2.36739703e-02 2.69465476e-01
-8.74602854e-01 2.74472475e-01 6.48853481e-01 9.89718735e-01
8.08352828e-01 -1.79243818e-01 -2.67065316e-01 7.32841372e-01
2.80300558e-01 5.46982825e-01 3.46399039e-01 -6.79635525e-01
6.86564386e-01 7.36780405e-01 7.06926823e-01 -1.63550436e+00
-2.12063864e-01 -3.88773322e-01 -1.01898551e+00 3.12022805e-01
9.34253216e-01 2.62971014e-01 -4.16936904e-01 1.61249828e+00
5.84559977e-01 3.35536562e-02 -3.04030836e-01 1.25890064e+00
4.76823479e-01 3.41983855e-01 -1.71814829e-01 -5.18217802e-01
1.49689078e+00 -7.02531278e-01 -7.36208618e-01 1.48904309e-01
4.63067681e-01 -7.18964398e-01 1.24783015e+00 6.16143107e-01
-7.43194997e-01 -4.30880427e-01 -1.00509381e+00 -1.67030081e-01
-2.92242974e-01 5.16932309e-01 4.24415529e-01 5.37917793e-01
-7.57750392e-01 3.70037436e-01 -3.86283934e-01 -2.62507915e-01
2.40622103e-01 1.24610350e-01 -5.61944783e-01 -2.23337904e-01
-8.97639990e-01 5.73487282e-01 1.96749404e-01 5.55206776e-01
-6.52723491e-01 -2.36787215e-01 -4.43312913e-01 -2.30105340e-01
6.91054702e-01 -4.24478143e-01 7.01621592e-01 -1.10045338e+00
-1.01306808e+00 9.05808687e-01 -3.79113525e-01 -7.55367130e-02
1.16719413e+00 -2.74243504e-01 -1.23590454e-02 -5.72791323e-02
4.00438786e-01 -1.54954419e-01 9.80442166e-01 -1.36329710e+00
-3.05641711e-01 -7.19422877e-01 -2.53735304e-01 4.11899179e-01
-4.15233791e-01 -6.79328144e-02 -3.89751285e-01 -7.08962500e-01
3.98564458e-01 -9.12346125e-01 -2.06846595e-01 1.11780822e-01
-6.40437365e-01 -1.70149669e-01 3.82853687e-01 -6.23527169e-01
1.19619608e+00 -2.20394683e+00 1.56042114e-01 4.14359003e-01
6.21166527e-01 -1.02847464e-01 1.21517219e-01 3.90031934e-01
-7.13308528e-02 1.09953769e-02 -1.06511272e-01 -1.88879564e-01
2.19146349e-03 -2.30965525e-01 -6.03534698e-01 9.03777182e-01
-1.87848732e-01 5.39097130e-01 -1.06632352e+00 -5.10604978e-01
-1.12124845e-01 5.25570288e-02 -1.62483469e-01 2.16584563e-01
-2.74743382e-02 6.47036076e-01 -3.59873831e-01 7.19725788e-01
9.34292138e-01 -2.18645409e-01 1.07221074e-01 -4.27618802e-01
-8.22679847e-02 -4.10660386e-01 -1.78118992e+00 1.42263091e+00
1.17230080e-01 3.62097204e-01 -1.90711357e-02 -9.34929252e-01
8.12286675e-01 1.72462061e-01 4.64094549e-01 -4.04068232e-01
2.37249695e-02 5.36727786e-01 -4.14324015e-01 -7.11735904e-01
6.55878186e-01 2.66215727e-02 8.65274851e-05 3.08342844e-01
-3.99664462e-01 3.14045578e-01 3.24225038e-01 3.11871588e-01
1.00202358e+00 -6.88998178e-02 4.67836112e-01 -1.58603162e-01
5.83028436e-01 1.08768709e-01 8.14738035e-01 7.84131944e-01
-2.56916672e-01 7.47402847e-01 7.91693628e-01 -6.17773235e-01
-1.14455485e+00 -8.29058468e-01 1.09677864e-02 1.01204062e+00
3.55803549e-01 -3.31669867e-01 -7.33921647e-01 -6.32264495e-01
-3.82165089e-02 -1.41067788e-01 -5.46970904e-01 2.44174272e-01
-1.88861951e-01 -1.02272403e+00 3.95941019e-01 1.61116734e-01
3.43019933e-01 -6.40138149e-01 -2.43083924e-01 -1.38427585e-01
-4.07005757e-01 -1.01092565e+00 -3.56476724e-01 -3.10436994e-01
-8.05012465e-01 -1.36971235e+00 -9.94332969e-01 -3.33934754e-01
1.06711030e+00 5.42289078e-01 9.45286691e-01 2.55818903e-01
-1.82226568e-01 1.51492655e-01 -4.02419657e-01 -3.06639999e-01
5.82587197e-02 -4.08983350e-01 4.99556452e-01 4.63205457e-01
3.19673449e-01 -2.82870471e-01 -7.43611693e-01 5.39210856e-01
-9.30058956e-01 -1.23433657e-01 3.72486174e-01 8.11966360e-01
8.71309757e-01 1.34766951e-01 5.10218441e-01 -9.01086450e-01
5.38132787e-01 -5.58504820e-01 -8.12584221e-01 3.02985996e-01
-4.12709862e-01 -1.33036092e-01 7.54156709e-01 -4.42948729e-01
-7.25422144e-01 2.55709797e-01 3.80388975e-01 -3.95006478e-01
1.85396541e-02 2.27183551e-01 -3.32996905e-01 -1.08813129e-01
6.89624429e-01 2.21412517e-02 -1.36062101e-01 -4.54494983e-01
4.25215483e-01 7.26107895e-01 3.50884765e-01 -6.39319301e-01
1.00715780e+00 1.02961206e+00 6.94228113e-02 -8.61513317e-01
-8.87541950e-01 -7.26627767e-01 -5.15049279e-01 -3.39469820e-01
6.15282953e-01 -1.08309877e+00 -8.23264003e-01 5.55125892e-01
-1.10289323e+00 1.90693364e-01 -1.71949252e-01 2.89086968e-01
-3.72197419e-01 8.56194496e-01 -3.02444518e-01 -1.18275261e+00
-3.83562557e-02 -1.07858074e+00 1.14520156e+00 7.03175813e-02
1.07196808e-01 -4.46430415e-01 2.19217926e-01 5.64878583e-01
-3.26388143e-02 4.61394876e-01 5.67247570e-01 -6.47235215e-01
-8.08027387e-01 -3.75861883e-01 -4.03809816e-01 2.68738180e-01
-3.21134120e-01 -2.17724778e-02 -1.05039442e+00 -3.39475095e-01
1.16796188e-01 -1.88523889e-01 6.32423341e-01 1.00644059e-01
1.02930295e+00 -3.75327110e-01 -4.59842943e-02 4.04842705e-01
1.51771224e+00 -4.30616379e-01 5.10041773e-01 2.90733695e-01
1.01202989e+00 8.38568032e-01 1.00818396e+00 7.60677218e-01
3.92762899e-01 7.20914662e-01 4.91222411e-01 9.09481645e-02
2.28486165e-01 -3.03808004e-01 3.83538127e-01 1.06620646e+00
-3.75673264e-01 -4.98636737e-02 -8.55780423e-01 5.09297013e-01
-2.23524880e+00 -8.55043769e-01 -5.80911934e-01 2.79346085e+00
5.67873240e-01 -2.76729435e-01 1.94045380e-01 3.00278991e-01
1.04634261e+00 1.17586292e-01 -3.23039353e-01 1.53000742e-01
-3.01214218e-01 -6.38297915e-01 5.12617171e-01 2.76001334e-01
-9.20161605e-01 5.88605940e-01 5.11246920e+00 1.00464416e+00
-6.86649859e-01 3.68837059e-01 6.40052736e-01 -3.61457691e-02
-2.18425602e-01 6.88795745e-02 -6.15792453e-01 8.31509292e-01
1.51483253e-01 3.56906429e-02 3.22476864e-01 8.17969382e-01
3.69537443e-01 -4.09599990e-01 -7.90017605e-01 1.25030720e+00
3.01000029e-01 -6.73507094e-01 -1.17129199e-01 1.04737654e-01
8.96962225e-01 -2.26333916e-01 -7.96899274e-02 -2.47755289e-01
-2.60485616e-02 -8.35447252e-01 8.72388899e-01 7.75188327e-01
7.79613972e-01 -6.19180799e-01 7.49908209e-01 5.15772343e-01
-1.07951832e+00 -1.12239063e-01 -8.64395440e-01 -2.18677908e-01
-4.62696925e-02 1.35352862e+00 -3.84843975e-01 5.27605295e-01
6.49487317e-01 4.95399624e-01 -6.42344236e-01 1.22765040e+00
-2.87000209e-01 1.59057334e-01 -2.77829230e-01 6.46684989e-02
-3.19830805e-01 -5.25561631e-01 7.68345773e-01 6.37950897e-01
4.86762643e-01 1.27174277e-02 1.50874630e-01 5.15005112e-01
-1.66258737e-01 5.93700409e-01 -9.40390289e-01 3.10973316e-01
3.33868623e-01 1.39457774e+00 -7.79260337e-01 -2.69198030e-01
-3.82709682e-01 9.85669136e-01 5.53191245e-01 4.84164864e-01
-7.39785671e-01 -6.20120876e-02 2.05447927e-01 1.75709680e-01
-1.78607792e-01 -8.21049511e-02 -3.50734770e-01 -1.63773847e+00
6.30526543e-01 -1.05827785e+00 1.94985271e-01 -7.56010115e-01
-1.52009010e+00 2.68551588e-01 -2.77102023e-01 -1.67253196e+00
3.19114476e-01 -5.80607057e-01 -3.85153830e-01 5.83019316e-01
-1.33177924e+00 -9.18185115e-01 -6.73877358e-01 5.95323205e-01
1.31868213e-01 -6.66076392e-02 4.30367738e-01 4.20006633e-01
-6.62377775e-01 3.88315052e-01 4.04197603e-01 -4.15151427e-03
9.77322161e-01 -1.26467347e+00 -3.02361637e-01 1.00661874e+00
2.12673679e-01 5.65963089e-01 7.79688478e-01 -8.03565621e-01
-1.34965825e+00 -8.11809361e-01 7.03841746e-01 -6.75245047e-01
6.67418242e-01 -4.63019729e-01 -9.39691544e-01 2.65335590e-01
-2.85528749e-01 2.93881625e-01 5.81913233e-01 -1.70358084e-02
-4.35262680e-01 -2.31396601e-01 -9.81844485e-01 4.86679167e-01
1.02565336e+00 -6.24631524e-01 -3.45473021e-01 7.17218518e-01
2.05820784e-01 -4.02757615e-01 -5.90342879e-01 2.46518731e-01
5.68767309e-01 -1.22243834e+00 7.55341291e-01 -3.89969707e-01
3.03261727e-01 -8.40752602e-01 -2.14232519e-01 -1.04813921e+00
1.60180584e-01 -6.79396808e-01 -1.54732496e-01 1.30543494e+00
-1.02454245e-01 -4.59223747e-01 6.93833113e-01 6.09038591e-01
3.31432343e-01 -5.79881310e-01 -9.47682261e-01 -6.86201990e-01
-4.89167482e-01 -3.22619170e-01 2.68188477e-01 8.67636085e-01
-7.51238763e-02 2.61585452e-02 -7.82841027e-01 3.17704976e-01
9.69516098e-01 9.89485439e-03 9.81440365e-01 -1.44857991e+00
-1.65049300e-01 7.11143538e-02 -5.10167837e-01 -7.37679362e-01
1.46695664e-02 -6.43374860e-01 1.66676432e-01 -1.19379568e+00
6.18374884e-01 -4.95461732e-01 -2.17006460e-01 1.40637159e-01
-4.46026921e-01 4.92766261e-01 2.15230018e-01 6.78040862e-01
-9.26778853e-01 5.17607331e-01 9.81482327e-01 1.29409283e-02
-2.18459032e-03 4.29560356e-02 -5.81094205e-01 1.05755818e+00
4.54442620e-01 -6.17017984e-01 -3.25684965e-01 -3.76959950e-01
1.03804576e+00 -1.66512594e-01 5.30479848e-01 -8.41167033e-01
4.21358019e-01 2.59802081e-02 2.29065806e-01 -4.07148778e-01
8.78970325e-02 -1.00789499e+00 6.87039047e-02 6.53892159e-02
-1.47637045e-02 2.19053254e-01 -4.07481521e-01 1.13648748e+00
-3.03706110e-01 -3.12595099e-01 5.59255779e-01 -1.35326207e-01
-4.56943154e-01 2.47825935e-01 1.67650077e-02 2.59441048e-01
1.00206590e+00 -1.24737576e-01 -5.31243980e-01 -3.72195423e-01
-6.35820866e-01 1.21581711e-01 8.85549843e-01 1.62842777e-02
4.50087219e-01 -1.53389239e+00 -6.18992090e-01 4.27760109e-02
3.31496567e-01 2.03969926e-02 4.67564732e-01 1.28880942e+00
-3.80742311e-01 -1.91664079e-03 -5.41178361e-02 -5.94298065e-01
-1.20469081e+00 5.68809927e-01 -8.73488113e-02 -2.56654412e-01
-2.81437844e-01 5.56412816e-01 1.68544382e-01 -1.92311212e-01
2.39817694e-01 -8.27197079e-03 -1.61624417e-01 6.00838482e-01
5.88211775e-01 8.91896486e-01 8.33018869e-02 -9.36370254e-01
-4.04509455e-01 7.48967707e-01 2.98576444e-01 -1.00854911e-01
1.02693856e+00 -4.21638846e-01 -5.33411264e-01 5.66667140e-01
9.17510867e-01 6.94124043e-01 -1.08751810e+00 -1.35120839e-01
1.11530453e-01 -7.86170006e-01 -4.62623656e-01 -2.68309355e-01
-7.92358696e-01 7.95459867e-01 5.12897789e-01 3.59732687e-01
1.00068939e+00 -2.72463828e-01 3.98158461e-01 3.82442266e-01
5.66466510e-01 -1.28980470e+00 1.33720785e-01 1.17534980e-01
8.91380429e-01 -1.57046938e+00 2.39866719e-01 -5.05690157e-01
-7.96241343e-01 8.77436697e-01 4.02534693e-01 -2.04704493e-01
1.57057911e-01 -2.72088289e-01 5.29052876e-02 -2.33612210e-01
-2.87330478e-01 -3.45573664e-01 4.30074722e-01 3.65082473e-01
2.26893440e-01 7.44992644e-02 -5.93745351e-01 6.32085025e-01
1.58432767e-01 -1.46676853e-01 6.86700344e-01 3.65472078e-01
-4.50153142e-01 -7.95273900e-01 -7.30875492e-01 3.70187968e-01
-5.17918169e-01 -3.06092259e-02 -2.59140134e-01 4.85734522e-01
3.03714126e-01 8.40781808e-01 -4.35381740e-01 -1.23636283e-01
3.21460754e-01 9.05243009e-02 1.10332884e-01 -5.01572430e-01
-1.07445613e-01 9.35407057e-02 -1.38661265e-01 -8.33865583e-01
-4.56968546e-01 -5.74603736e-01 -8.89796615e-01 -6.36595413e-02
-4.21429336e-01 1.55123383e-01 6.30326390e-01 8.51249278e-01
2.44444430e-01 3.29386592e-02 6.14706397e-01 -7.91867256e-01
-6.59379125e-01 -7.50427067e-01 -9.16037083e-01 7.89984107e-01
2.39759773e-01 -8.00677836e-01 -6.34923220e-01 -6.58647940e-02] | [7.767514228820801, 4.44369649887085] |
e1bc71c1-3ce2-4e6d-90bf-bb15f22ebb61 | a-trillion-genetic-programming-instructions | 2205.03251 | null | https://arxiv.org/abs/2205.03251v1 | https://arxiv.org/pdf/2205.03251v1.pdf | A Trillion Genetic Programming Instructions per Second | We summarise how a 3.0 GHz 16 core AVX512 computer can interpret the equivalent of up to on average 1103370000000 GPop/s. Citations to existing publications are given. Implementation stress is placed on both parallel computing, bandwidth limits and avoiding repeated calculation. Information theory suggests in digital computing, failed disruption propagation gives huge speed ups as FDP and incremental evaluation can be used to reduce fitness evaluation time in phenotypically converged populations. Conversely FDP may be responsible for evolution stagnation. So the wider Evolutionary Computing, Artificial Life, Unconventional Computing and Software Engineering community may need to avoid deep nesting. | ['W. B. Langdon'] | 2022-05-06 | null | null | null | null | ['artificial-life'] | ['miscellaneous'] | [-9.72483307e-02 -1.20954834e-01 3.32134098e-01 1.84778154e-01
2.62702644e-01 -4.58732754e-01 1.25209779e-01 1.66076511e-01
-4.09819216e-01 9.91022646e-01 -3.60533893e-01 -6.15336537e-01
-5.13475657e-01 -8.14379632e-01 -6.31357282e-02 -7.37422347e-01
-5.67608953e-01 3.36747944e-01 1.94581732e-01 -5.18413007e-01
8.11932921e-01 4.87912655e-01 -1.89336050e+00 -4.67504352e-01
1.00792146e+00 5.66058993e-01 -1.72640830e-01 1.15555418e+00
2.80101657e-01 2.48703122e-01 -1.11347234e+00 -2.29542419e-01
3.21213663e-01 -2.99716115e-01 -6.62701190e-01 -3.18501860e-01
-6.19953692e-01 7.83981085e-02 -4.35390174e-02 9.99456108e-01
9.58953083e-01 1.98873401e-01 3.33781600e-01 -1.36629140e+00
-4.93923903e-01 5.18375218e-01 -8.78574491e-01 9.33837831e-01
2.38816544e-01 6.04507685e-01 3.43173742e-01 -5.42032361e-01
3.75865966e-01 9.02022481e-01 9.01865304e-01 6.87496811e-02
-1.00926185e+00 -3.11928362e-01 -5.69290400e-01 -1.98575705e-01
-1.90424180e+00 -3.17280918e-01 8.33185613e-02 5.85172065e-02
1.73025513e+00 8.77850771e-01 1.20152295e+00 4.82027620e-01
9.94711995e-01 -1.10194415e-01 5.83492577e-01 -3.57918322e-01
4.68321800e-01 -1.09991238e-01 -9.21430066e-02 5.45126438e-01
1.06902277e+00 5.07098019e-01 -5.19836307e-01 -6.83281660e-01
6.30210519e-01 -6.08291149e-01 -2.75694966e-01 3.20555180e-01
-6.24828279e-01 5.66228092e-01 5.98564260e-02 1.23822957e-01
-5.59774220e-01 5.11431038e-01 5.19061744e-01 5.08650482e-01
6.66215867e-02 1.09129965e+00 -6.12268388e-01 -1.05324399e+00
-6.62400842e-01 1.83392778e-01 7.89271057e-01 6.79736674e-01
3.33364099e-01 3.81570846e-01 5.64346492e-01 3.35068047e-01
3.29702288e-01 1.95683718e-01 6.61749840e-01 -1.22452652e+00
-2.51604527e-01 6.74480796e-01 -6.57234862e-02 -1.12269270e+00
-7.39031136e-01 -7.04991281e-01 -7.41405427e-01 4.35467601e-01
-9.40047801e-02 -5.99724293e-01 -5.44491112e-01 9.32652473e-01
4.86521095e-01 -2.79843621e-02 2.11381063e-01 6.66867077e-01
1.76810279e-01 7.97122836e-01 -1.71079859e-01 -3.33901107e-01
1.19354856e+00 -5.27647257e-01 -3.81406993e-01 -6.18756898e-02
5.67342997e-01 -9.74411011e-01 7.39569962e-01 6.43603921e-01
-1.22957087e+00 -6.21494949e-02 -1.36021745e+00 5.40875554e-01
-3.36404115e-01 -4.70811427e-01 1.14200532e+00 1.38209116e+00
-1.24044442e+00 1.00386286e+00 -1.31365430e+00 -4.35199767e-01
3.90721470e-01 7.16554463e-01 4.54853147e-01 2.20114678e-01
-9.18393672e-01 8.98969054e-01 5.83562851e-01 7.60961547e-02
-2.21477136e-01 -9.23883617e-01 -2.10800320e-01 3.22038591e-01
-8.06702999e-04 -8.89424503e-01 3.97759140e-01 -7.85293460e-01
-1.48885393e+00 3.67081016e-01 5.45928895e-01 -2.62111783e-01
2.75889695e-01 4.08445209e-01 -5.41891932e-01 7.23721460e-02
-4.03232336e-01 3.22226405e-01 2.44166657e-01 -6.95824862e-01
-4.75421786e-01 -2.91029483e-01 -4.96705264e-01 3.50163162e-01
-3.28081161e-01 2.03327596e-01 2.17432246e-01 -4.04497176e-01
1.87762380e-02 -1.04983139e+00 -5.65747380e-01 -4.58568990e-01
1.48094162e-01 -5.43670021e-02 5.19180715e-01 -3.88086945e-01
1.46327698e+00 -1.79285777e+00 -4.37056459e-02 3.88451934e-01
3.43944356e-02 1.47002503e-01 2.09007151e-02 6.76786423e-01
1.78384423e-01 5.53579807e-01 4.41294499e-02 5.64874113e-01
-4.31004316e-02 1.97189033e-01 1.04600735e-01 6.76736057e-01
1.67535618e-01 4.16059941e-01 -8.79759431e-01 -8.60565007e-02
-9.77186784e-02 6.14549398e-01 -7.18838930e-01 -4.75822628e-01
4.73479182e-01 -5.39375022e-02 -2.89220065e-01 1.10250366e+00
8.16751778e-01 -1.65075481e-01 3.16265941e-01 6.89267576e-01
-4.64734077e-01 -1.38381958e-01 -9.70815659e-01 1.51246560e+00
-2.38564953e-01 7.51300514e-01 -3.61126885e-02 -5.14409840e-01
7.41596520e-01 1.37079030e-01 4.02309120e-01 -4.58925992e-01
5.61349392e-01 3.52235168e-01 8.26942384e-01 -1.61148801e-01
1.03116524e+00 1.43155873e-01 7.01268390e-02 4.89385158e-01
-2.93903291e-01 -4.26311553e-01 2.60283053e-01 1.60816565e-01
1.29897344e+00 1.10621415e-01 1.13487773e-01 -9.87890601e-01
1.13450460e-01 4.77383494e-01 6.67124450e-01 5.97236514e-01
-3.89352262e-01 2.30745256e-01 6.03451908e-01 -6.94520235e-01
-1.23313081e+00 -7.34127343e-01 -5.51823437e-01 8.41684043e-01
3.82773936e-01 -7.21960664e-01 -6.10604763e-01 2.73731768e-01
-1.49164021e-01 9.34152007e-01 -1.61935598e-01 -6.44788027e-01
-3.19265336e-01 -1.82691407e+00 8.09246182e-01 1.29404277e-01
3.38923842e-01 -7.98973918e-01 -1.60382962e+00 3.62625420e-01
6.07175708e-01 -4.61078100e-02 8.58201310e-02 3.47100616e-01
-1.17712951e+00 -5.51459610e-01 -5.46866693e-02 -1.43139660e-01
4.47098076e-01 -6.69521571e-04 1.40672374e+00 9.31529284e-01
-1.16176808e+00 4.08196673e-02 -2.32999325e-01 -3.61225575e-01
-3.98508251e-01 -4.15422842e-02 2.78749615e-01 -1.07381833e+00
3.74684423e-01 -9.46015239e-01 -6.64613724e-01 8.69332105e-02
-3.03627282e-01 -2.63630092e-01 2.52748042e-01 1.03917420e+00
2.01823428e-01 8.42334926e-01 3.32660854e-01 -1.19616874e-01
8.47156942e-01 -7.14587748e-01 -7.05683291e-01 1.97498903e-01
-1.15413189e+00 2.36742776e-02 5.35202146e-01 -3.14000934e-01
-8.77014160e-01 -3.96608204e-01 4.34384635e-03 1.23289088e-02
3.82391930e-01 3.88227314e-01 4.23852533e-01 -5.35024226e-01
8.41434777e-01 7.88772292e-03 8.03333148e-02 1.06696054e-01
-2.38589421e-01 5.48377156e-01 1.27874285e-01 -6.02319598e-01
3.15875441e-01 -7.22575560e-03 3.10251445e-01 -1.03636718e+00
5.44762254e-01 4.04821932e-01 -3.43256369e-02 -3.52543294e-01
3.15868556e-01 -4.28914040e-01 -1.22196960e+00 1.90021098e-01
-8.29810381e-01 -2.84566320e-02 -3.90997678e-01 2.03069866e-01
-9.35463831e-02 4.13174778e-02 -5.08351862e-01 -1.07955706e+00
-4.62379873e-01 -8.41277122e-01 5.50992548e-01 9.25707400e-01
-4.70772743e-01 -9.02174234e-01 2.60571122e-01 -1.50165513e-01
7.13333845e-01 3.62249941e-01 5.73372483e-01 -1.84007436e-01
-1.15984477e-01 -1.76439777e-01 3.13454092e-01 -4.21866536e-01
-2.10139528e-01 8.13043892e-01 -4.62800473e-01 -4.79748100e-01
1.53623689e-02 7.89010599e-02 2.12600127e-01 4.06815916e-01
8.53969634e-01 -1.80371866e-01 -5.26296675e-01 8.39790106e-01
1.69724727e+00 7.18627214e-01 7.63846278e-01 4.95535165e-01
1.63501874e-01 4.89139467e-01 4.31021541e-01 1.01340377e+00
-1.11808181e-01 2.22254202e-01 3.05860549e-01 3.25251371e-01
2.70041794e-01 1.92279950e-01 1.09991543e-01 9.52795506e-01
-3.73928905e-01 -7.44100034e-01 -1.44411016e+00 1.67909175e-01
-1.44937468e+00 -8.29612017e-01 -4.16882694e-01 2.05192113e+00
5.66502512e-01 2.20320448e-01 -3.03053595e-02 1.41419917e-01
6.81173325e-01 -4.86365527e-01 -7.71627486e-01 -1.15507746e+00
-3.76240280e-03 2.96134293e-01 8.18560183e-01 3.26613426e-01
-2.14471325e-01 5.24390280e-01 7.35851288e+00 8.28397512e-01
-9.70140874e-01 -1.98813658e-02 8.50586295e-01 -6.29753113e-01
-1.96754083e-01 6.28898218e-02 -3.08899403e-01 8.10653687e-01
1.64459288e+00 -1.00875747e+00 6.65472388e-01 6.39974713e-01
-1.28348973e-02 -6.87806487e-01 -3.04024726e-01 1.07548416e+00
-3.11232477e-01 -1.18988252e+00 -4.14078414e-01 5.42022943e-01
6.78516805e-01 -2.42718793e-02 6.42229542e-02 2.62465961e-02
5.91993272e-01 -1.30901945e+00 3.86848122e-01 3.37415487e-01
3.96248072e-01 -1.66540730e+00 8.38272691e-01 1.51327804e-01
-8.40976179e-01 -2.21100301e-01 -7.60472596e-01 -6.20753348e-01
8.88920724e-02 2.54501700e-01 -4.42553580e-01 5.77577889e-01
9.41898823e-01 -1.27991736e-01 -6.87626660e-01 1.10344756e+00
2.67556816e-01 3.28253716e-01 -8.73496532e-01 -2.99897999e-01
8.32559168e-02 -4.85422373e-01 7.44889140e-01 6.68603063e-01
7.23877549e-01 3.25359643e-01 -4.14950788e-01 5.26191950e-01
4.85884696e-01 -1.23803854e-01 -2.81783760e-01 -3.83682936e-01
9.81239021e-01 1.14954805e+00 -1.42345011e+00 -1.54061913e-01
2.22506240e-01 9.73294437e-01 -3.03141564e-01 -2.23647133e-01
-9.58146393e-01 -9.26957011e-01 1.06295061e+00 -2.63589174e-01
-6.13974556e-02 -2.72944063e-01 -7.58027315e-01 -7.85634995e-01
-6.91923082e-01 -7.64120519e-01 3.41180295e-01 -5.58863580e-01
-7.94043481e-01 7.82885432e-01 -5.09634733e-01 -6.28015101e-01
-2.42879719e-01 -4.97021705e-01 -8.52791309e-01 7.59847879e-01
-7.54090309e-01 -1.50540536e-02 -9.26705375e-02 -8.25601891e-02
2.37304524e-01 -3.44098479e-01 7.87890315e-01 1.83034375e-01
-1.00340652e+00 5.60443282e-01 4.29591447e-01 -9.14146304e-01
-3.43563631e-02 -1.13019502e+00 4.78366047e-01 1.05693638e+00
-4.94126529e-01 7.28466332e-01 1.06375098e+00 -7.69207418e-01
-2.19185567e+00 -3.69278044e-01 3.82834435e-01 -3.14015776e-01
5.90951324e-01 3.02748382e-02 -7.52029061e-01 -1.08305234e-02
6.44210815e-01 -3.16812724e-01 7.04348326e-01 5.69987409e-02
5.95472157e-01 2.93995380e-01 -1.35658991e+00 8.05135667e-01
1.09719217e+00 3.63331139e-01 -7.84052387e-02 2.30708405e-01
4.73570943e-01 -4.19320196e-01 -1.10208273e+00 -1.09345585e-01
5.54812729e-01 -9.83138204e-01 8.96595776e-01 -3.09411675e-01
5.27654700e-02 -2.62721419e-01 1.47205949e-01 -1.09762383e+00
-3.95156205e-01 -1.25787258e+00 2.71194309e-01 1.01180649e+00
2.12043166e-01 -1.03397477e+00 6.79490089e-01 9.73133802e-01
-3.56591940e-01 -8.14657211e-01 -1.08172119e+00 -9.00053680e-01
1.25819638e-01 -4.10395786e-02 8.05055082e-01 9.71360743e-01
3.46794546e-01 8.68808553e-02 -4.25908975e-02 2.69309938e-01
6.18193507e-01 -1.89908952e-01 2.45432958e-01 -1.17173791e+00
-4.78011966e-01 -7.26853967e-01 -6.13968015e-01 6.93180561e-02
-6.37392938e-01 -5.51575303e-01 -2.84863681e-01 -9.04816866e-01
-8.79850090e-02 -5.61911821e-01 7.08888145e-03 2.16789454e-01
-5.22026569e-02 4.20600832e-01 -4.96310741e-02 -8.88023451e-02
-4.88159247e-02 3.80284816e-01 7.95499206e-01 4.76072341e-01
-4.04102445e-01 -5.65218687e-01 -7.43564665e-01 6.49776340e-01
1.01626825e+00 -7.57697165e-01 -5.07292986e-01 -2.98204154e-01
9.88798738e-01 3.32190096e-02 -1.03920259e-01 -1.06682086e+00
4.17204052e-01 -2.87959039e-01 4.24344748e-01 -1.48374796e-01
1.10395253e-01 -6.37051702e-01 1.02933979e+00 1.31970453e+00
3.82922024e-01 7.68985927e-01 5.85887730e-01 1.69774950e-01
1.96352452e-01 -5.23365140e-01 7.43392467e-01 -3.40202935e-02
-3.15240532e-01 -1.88855156e-01 -9.16726053e-01 -1.27002016e-01
1.36197329e+00 -6.07424080e-01 -5.81851304e-01 1.84465453e-01
-2.90425599e-01 4.46755111e-01 1.00057566e+00 8.25014263e-02
3.93624634e-01 -9.17814612e-01 -5.73218703e-01 1.39681801e-01
-6.29565895e-01 -5.48081994e-01 3.55453521e-01 6.82504535e-01
-1.45920718e+00 2.62774955e-02 -7.61776865e-01 -2.98781514e-01
-1.33240306e+00 4.49593157e-01 3.80190998e-01 4.56402779e-01
-5.91668904e-01 1.32122612e+00 -8.16010058e-01 9.39091593e-02
-2.54884481e-01 4.19710040e-01 3.53127301e-01 2.29266230e-02
5.64467251e-01 1.02099872e+00 7.00650066e-02 1.18229557e-02
-8.27029526e-01 2.59542525e-01 2.60994822e-01 -1.37969762e-01
1.51226246e+00 -1.98946089e-01 -4.81714100e-01 -7.55037069e-02
7.67591298e-01 -2.31254920e-01 -8.76493156e-01 9.90188003e-01
1.09934896e-01 -7.86860108e-01 4.45745260e-01 -7.63450325e-01
-7.94509888e-01 5.49027085e-01 8.04610133e-01 5.54477811e-01
1.49569774e+00 -6.75996900e-01 3.00587416e-01 3.13924968e-01
5.07391036e-01 -1.38661730e+00 -3.27800214e-01 2.97257006e-01
4.47588980e-01 -6.26566291e-01 5.12636840e-01 -8.23691413e-02
-4.38187420e-01 1.53425908e+00 5.50104439e-01 -3.22500110e-01
4.25093144e-01 9.43139255e-01 -6.21203005e-01 -3.29967916e-01
-1.31984556e+00 4.58985806e-01 -6.21927321e-01 5.97764492e-01
4.70367014e-01 -1.03969291e-01 -8.58003974e-01 3.92470419e-01
-6.40887797e-01 -2.63032764e-01 1.01232362e+00 1.32875526e+00
-4.94693220e-01 -1.02574718e+00 -5.81195295e-01 3.57100040e-01
-5.36630452e-01 -4.08629328e-01 -9.01526883e-02 7.43019938e-01
1.23804912e-01 8.68808746e-01 6.42206371e-01 -3.03565115e-01
-3.11500341e-01 -2.56729454e-01 6.23700321e-01 -9.92160290e-02
-1.24797428e+00 -4.11221832e-02 2.58476377e-01 -5.42556405e-01
2.39347309e-01 -9.41628337e-01 -1.41854203e+00 -1.23300076e+00
-4.03037339e-01 5.11371076e-01 1.01813889e+00 8.43539685e-02
9.49614704e-01 8.01514626e-01 3.33620578e-01 -7.25674450e-01
-2.52885312e-01 -5.36216974e-01 -8.38203192e-01 -6.42547131e-01
-4.18194473e-01 -5.94380319e-01 -5.23495257e-01 -3.72959405e-01] | [5.655646324157715, 3.9695873260498047] |
9951ca0e-2af8-488b-8903-7be1bfcae872 | direct-dense-pose-estimation | 2204.01263 | null | https://arxiv.org/abs/2204.01263v1 | https://arxiv.org/pdf/2204.01263v1.pdf | Direct Dense Pose Estimation | Dense human pose estimation is the problem of learning dense correspondences between RGB images and the surfaces of human bodies, which finds various applications, such as human body reconstruction, human pose transfer, and human action recognition. Prior dense pose estimation methods are all based on Mask R-CNN framework and operate in a top-down manner of first attempting to identify a bounding box for each person and matching dense correspondences in each bounding box. Consequently, these methods lack robustness due to their critical dependence on the Mask R-CNN detection, and the runtime increases drastically as the number of persons in the image increases. We therefore propose a novel alternative method for solving the dense pose estimation problem, called Direct Dense Pose (DDP). DDP first predicts the instance mask and global IUV representation separately and then combines them together. We also propose a simple yet effective 2D temporal-smoothing scheme to alleviate the temporal jitters when dealing with video data. Experiments demonstrate that DDP overcomes the limitations of previous top-down baseline methods and achieves competitive accuracy. In addition, DDP is computationally more efficient than previous dense pose estimation methods, and it reduces jitters when applied to a video sequence, which is a problem plaguing the previous methods. | ['Luc van Gool', 'Christian Theobalt', 'Lingjie Liu', 'Liqian Ma'] | 2022-04-04 | null | null | null | null | ['pose-transfer'] | ['computer-vision'] | [ 1.80431664e-01 -1.02316372e-01 1.11539394e-01 -5.21280318e-02
-4.68312591e-01 -1.07747652e-01 2.82032818e-01 -3.55179727e-01
-5.11195481e-01 6.46788180e-01 2.64516920e-01 4.90121573e-01
1.88043207e-01 -6.20926678e-01 -7.34820843e-01 -5.72784185e-01
9.11555588e-02 7.05530882e-01 6.08052790e-01 -8.97871405e-02
-2.02314183e-01 4.91628796e-01 -1.49232936e+00 -1.30171597e-01
5.57342052e-01 1.15895796e+00 3.70645244e-03 3.19388092e-01
2.73340076e-01 4.44371283e-01 -5.63396454e-01 -4.63217258e-01
5.24289310e-01 -4.40262347e-01 -6.41065240e-01 4.28821266e-01
4.98779476e-01 -4.79727954e-01 -5.83154261e-01 9.07384932e-01
5.53712010e-01 3.05349499e-01 3.50459695e-01 -1.14712906e+00
4.58669439e-02 -1.36056170e-01 -9.37945962e-01 -1.61161885e-01
6.91123009e-01 6.40239567e-02 5.74331820e-01 -9.92046654e-01
6.74292326e-01 1.40038514e+00 1.05799401e+00 7.77455091e-01
-1.05506575e+00 -6.47297025e-01 1.10762455e-01 1.20874673e-01
-1.56380236e+00 -3.61932456e-01 6.58019543e-01 -4.58276600e-01
6.23731732e-01 2.70711333e-01 1.10804260e+00 1.00917280e+00
-1.84692293e-01 8.32263112e-01 8.94663692e-01 -2.98327625e-01
7.11468374e-03 -4.27187890e-01 -2.61844456e-01 9.28069770e-01
2.89846748e-01 2.06159297e-02 -6.52196229e-01 -6.85011744e-02
1.40346336e+00 1.80499494e-01 -3.49898607e-01 -6.69968963e-01
-1.39024889e+00 5.36938787e-01 5.73351026e-01 -5.78648336e-02
-4.64065909e-01 2.98396558e-01 2.96955317e-01 -1.92625597e-01
3.80920172e-01 1.66355163e-01 -3.27500254e-01 -3.85264261e-03
-9.52487350e-01 5.57201803e-01 6.54760718e-01 9.63107467e-01
7.46349990e-01 -2.66647577e-01 -2.13467404e-01 5.34413457e-01
3.44517082e-01 4.44738954e-01 2.23085687e-01 -9.02806580e-01
4.69135374e-01 6.18578792e-01 3.33243668e-01 -1.34069550e+00
-4.26079869e-01 -5.56476451e-02 -9.38763797e-01 -8.42764750e-02
5.78410447e-01 1.92677081e-02 -1.05627513e+00 1.52346909e+00
8.36352170e-01 2.88685411e-01 -4.64532703e-01 1.23763704e+00
7.49854028e-01 4.58438247e-01 -6.60248920e-02 -9.38544571e-02
1.32793629e+00 -1.00623846e+00 -7.47817755e-01 -4.99882877e-01
1.52093619e-01 -5.80385625e-01 8.19167972e-01 2.65736699e-01
-1.13452423e+00 -5.31279266e-01 -8.65798593e-01 -2.79608458e-01
5.14733940e-02 2.66702741e-01 5.58518052e-01 4.53909993e-01
-7.42068470e-01 6.45864248e-01 -1.13188529e+00 -3.68685871e-01
4.13956195e-01 6.80308461e-01 -4.93858874e-01 5.22621907e-03
-9.43467379e-01 8.34527493e-01 2.62180746e-01 5.46878815e-01
-6.04137361e-01 -3.28589737e-01 -1.17793202e+00 -1.75873995e-01
6.49460435e-01 -9.15914178e-01 1.11199510e+00 -8.06719601e-01
-1.68526828e+00 9.32552695e-01 -2.52051622e-01 -2.73837984e-01
8.60175371e-01 -7.21753180e-01 7.16032013e-02 1.56816676e-01
-7.71651492e-02 7.19902456e-01 1.15816700e+00 -9.94753599e-01
-3.81414384e-01 -6.04808211e-01 -1.00324929e-01 3.39703649e-01
-1.23579882e-01 -3.08198351e-02 -1.08117449e+00 -6.86476052e-01
4.47470367e-01 -1.09386861e+00 -5.46840072e-01 3.39144558e-01
-3.73277992e-01 -1.74350683e-02 8.32278728e-01 -8.07302237e-01
1.12928379e+00 -1.96583068e+00 6.89734042e-01 3.54053706e-01
3.77154112e-01 3.31337631e-01 1.35769680e-01 2.35913359e-02
2.70155847e-01 -4.32707369e-01 -2.24271834e-01 -6.36094570e-01
-8.57904702e-02 3.00160587e-01 1.40356034e-01 8.58019829e-01
1.80562690e-01 9.22274649e-01 -6.82857811e-01 -7.56267607e-01
3.67793530e-01 6.33631051e-01 -5.55504560e-01 4.02608156e-01
-1.19691424e-01 8.29564512e-01 -2.54133761e-01 6.96298718e-01
6.38248086e-01 -4.28044021e-01 1.65469989e-01 -5.74936986e-01
5.46478331e-02 7.56910294e-02 -1.38949990e+00 1.99221051e+00
-1.35537520e-01 2.72008866e-01 2.18934461e-01 -8.39708567e-01
7.42300451e-01 2.44097158e-01 6.90497518e-01 -4.11723047e-01
3.82940650e-01 3.36388126e-02 -3.28312486e-01 -2.85123944e-01
3.65301013e-01 -2.73473144e-01 -8.05122480e-02 1.20554917e-01
-8.53048339e-02 5.60179017e-02 -8.84511098e-02 -1.86048791e-01
9.28522766e-01 5.29749334e-01 4.43446130e-01 3.62318084e-02
4.66218770e-01 -2.02860564e-01 8.13809156e-01 2.22047716e-01
-3.67240578e-01 9.37684059e-01 1.59731314e-01 -6.03375256e-01
-8.92351508e-01 -1.00607061e+00 1.53232962e-01 6.20774090e-01
4.98199701e-01 -5.23322940e-01 -8.61586094e-01 -5.50607860e-01
8.81447420e-02 -2.66998231e-01 -6.37152255e-01 9.80544612e-02
-1.03395653e+00 -3.29830974e-01 2.40307227e-01 8.00793469e-01
7.45049953e-01 -9.20115888e-01 -9.71659839e-01 1.49696633e-01
-4.90454584e-01 -1.36374104e+00 -6.43858910e-01 -1.23411201e-01
-8.68289113e-01 -1.10409677e+00 -1.01193178e+00 -6.41468704e-01
8.47498477e-01 3.41868065e-02 9.65936899e-01 1.15588717e-01
-5.72489977e-01 3.23987007e-01 -1.87910721e-01 -5.90616874e-02
2.10549101e-01 -6.80766180e-02 2.76597738e-01 1.53010666e-01
3.51492822e-01 -3.52363437e-01 -8.13864529e-01 5.62506199e-01
-4.66739386e-01 2.21571058e-01 5.27353168e-01 7.79194891e-01
7.09973276e-01 -2.03529477e-01 1.34526137e-02 -4.70174164e-01
-7.57167190e-02 2.19503660e-02 -5.89942098e-01 5.67109659e-02
2.61409413e-02 -2.22803801e-01 8.36524665e-02 -4.73721236e-01
-8.93663168e-01 7.55683303e-01 -2.64635891e-01 -8.70236516e-01
-9.82655957e-02 3.95473577e-02 -3.86968315e-01 -1.21522456e-01
4.01320577e-01 -3.32241431e-02 3.19556296e-01 -5.11762440e-01
1.59908667e-01 1.85425684e-01 7.64510155e-01 -4.00072694e-01
1.06452477e+00 6.60737574e-01 8.05864856e-02 -8.47702205e-01
-8.82859528e-01 -5.59553742e-01 -1.05680835e+00 -3.96225095e-01
1.13314092e+00 -9.56378162e-01 -9.78771329e-01 4.95038629e-01
-1.33454156e+00 -2.35005587e-01 -2.34861732e-01 6.11025810e-01
-6.61954999e-01 5.21625698e-01 -7.09886670e-01 -6.30298495e-01
-2.15515107e-01 -9.75568533e-01 1.40378559e+00 6.99084923e-02
-4.78512526e-01 -6.13515019e-01 -4.78438707e-03 4.00907516e-01
6.08668998e-02 5.98415554e-01 1.80857271e-01 4.95480150e-02
-6.71983480e-01 -3.50486577e-01 -2.00752914e-01 1.11792706e-01
1.55184101e-02 -4.90155727e-01 -7.66178787e-01 -4.48281050e-01
-4.84020896e-02 -2.66679734e-01 5.47959566e-01 5.09235561e-01
1.08327889e+00 -3.12494367e-01 -4.18970764e-01 7.57885277e-01
1.14278889e+00 -2.21420556e-01 8.03878129e-01 2.92706519e-01
1.26376081e+00 5.98025918e-01 7.55116761e-01 6.01298094e-01
2.17004016e-01 1.13532186e+00 3.63244414e-01 -3.64105552e-01
-1.96767122e-01 -3.66842031e-01 3.19528610e-01 6.26410902e-01
-5.02431512e-01 2.42927194e-01 -7.12105632e-01 2.57239521e-01
-2.10237336e+00 -6.62982881e-01 6.24847487e-02 2.41925073e+00
7.87354112e-01 -8.42795521e-02 5.24279237e-01 1.23787791e-01
8.07712257e-01 -1.06836949e-02 -4.34147060e-01 3.30923110e-01
2.23779544e-01 3.31094027e-01 4.02867258e-01 1.80079043e-01
-1.32473910e+00 1.02406955e+00 6.29799080e+00 4.61300462e-01
-8.14532220e-01 1.53285697e-01 1.78189009e-01 -3.33938688e-01
2.97850728e-01 -2.93034345e-01 -9.23379362e-01 2.23605528e-01
3.62441331e-01 2.06612706e-01 1.29530504e-01 7.17782080e-01
1.43746287e-02 -1.86782911e-01 -1.35125804e+00 1.57260811e+00
2.14808047e-01 -1.01150167e+00 -1.76634714e-01 1.24900721e-01
5.98593533e-01 -3.80651236e-01 -1.52202532e-01 6.74338862e-02
-1.44154921e-01 -1.01895237e+00 9.16511297e-01 3.58527988e-01
7.32353568e-01 -6.97697043e-01 6.41322970e-01 3.47956985e-01
-1.49484682e+00 1.63478494e-01 -4.18220758e-01 -2.75027722e-01
3.65558594e-01 4.31899786e-01 -4.39172596e-01 3.81426811e-01
9.08645213e-01 7.56687045e-01 -3.59988779e-01 1.01258886e+00
-2.36588180e-01 4.48527709e-02 -4.25453901e-01 1.76589653e-01
-1.58018380e-01 -1.86636522e-01 5.04706323e-01 9.43396628e-01
3.11978310e-01 2.15423435e-01 5.30544281e-01 7.83266664e-01
9.35109984e-03 8.31713527e-02 -4.65323061e-01 2.70409703e-01
1.36996910e-01 1.03928733e+00 -9.62025523e-01 -3.08743268e-01
-4.75903481e-01 1.40186357e+00 2.73872972e-01 1.41641170e-01
-8.90197933e-01 -4.54465598e-02 6.38767123e-01 5.46003222e-01
5.14703572e-01 -4.11858201e-01 -7.34171122e-02 -1.31682265e+00
2.64586031e-01 -8.85856330e-01 2.44209841e-01 -4.50366616e-01
-1.06639886e+00 3.62712801e-01 4.09179851e-02 -1.57300913e+00
-1.67959601e-01 -4.94462222e-01 -5.70649132e-02 5.82353175e-01
-9.32050467e-01 -1.10925865e+00 -6.52304828e-01 9.59822953e-01
6.25606477e-01 3.74304175e-01 6.48138762e-01 4.47683066e-01
-5.59453547e-01 5.63331485e-01 -5.38591802e-01 3.34602177e-01
5.99985182e-01 -9.16681468e-01 4.51070249e-01 9.00977194e-01
-4.71320823e-02 5.63409626e-01 5.46838880e-01 -8.62200201e-01
-1.54085600e+00 -1.10869908e+00 6.08525693e-01 -4.72537071e-01
2.28755131e-01 -5.20032525e-01 -7.44033396e-01 7.00145185e-01
-2.82912225e-01 2.28647679e-01 3.98390621e-01 8.89808908e-02
-1.06394455e-01 -6.28626272e-02 -1.01848912e+00 5.43656826e-01
1.32311094e+00 -2.97863364e-01 -6.14365935e-01 2.96516597e-01
4.96782571e-01 -1.00180185e+00 -7.89905787e-01 5.63788176e-01
6.55580819e-01 -9.42555130e-01 1.20634913e+00 -1.93757847e-01
3.16174716e-01 -5.61168730e-01 3.92177850e-02 -8.50723445e-01
-3.05646837e-01 -6.21308982e-01 -4.48362082e-01 7.30763197e-01
-2.11388007e-01 -2.87092596e-01 1.09622300e+00 6.64140761e-01
1.15034971e-02 -6.98128045e-01 -1.03677130e+00 -8.06426823e-01
-5.08632004e-01 -2.05096230e-01 3.98639768e-01 6.78632021e-01
-2.43343383e-01 1.87326148e-01 -8.59013855e-01 2.40561813e-01
9.10053134e-01 3.44777294e-02 1.15536797e+00 -1.19481015e+00
-4.36520904e-01 -9.71529186e-02 -7.45636106e-01 -1.50944698e+00
-6.43626750e-02 -3.31137896e-01 5.39743900e-01 -1.33146894e+00
1.15065835e-01 -1.69858590e-01 1.81409508e-01 4.76546854e-01
-2.09249467e-01 6.26123905e-01 2.92909920e-01 3.83058697e-01
-6.37575567e-01 6.66381180e-01 1.32283926e+00 6.14835061e-02
-2.37186074e-01 1.46546498e-01 -3.41921188e-02 1.03493571e+00
3.13816696e-01 -2.41782814e-01 -1.17512763e-01 -2.44946197e-01
-4.62538078e-02 1.18591852e-01 7.10941136e-01 -1.38761342e+00
2.16542065e-01 3.13305110e-02 7.52929568e-01 -9.11363840e-01
6.26093090e-01 -8.66015851e-01 4.28148270e-01 7.16347158e-01
1.06867872e-01 -2.89978348e-02 1.02707215e-01 5.95578074e-01
-1.48068503e-01 2.10739613e-01 9.82283413e-01 -4.00790215e-01
-7.63254225e-01 7.24735439e-01 7.64276162e-02 -2.43243594e-02
1.10460496e+00 -4.67557579e-01 2.94454843e-01 -1.91459402e-01
-9.58418608e-01 6.40452653e-02 6.68614507e-01 3.80554497e-01
8.56500983e-01 -1.46981061e+00 -4.32412654e-01 3.34267616e-01
-8.47665295e-02 4.22084510e-01 1.96762577e-01 1.11152852e+00
-7.00543344e-01 2.16038331e-01 -2.21373424e-01 -8.64896953e-01
-1.51316738e+00 5.07081866e-01 3.14353287e-01 -2.03164458e-01
-1.20095026e+00 1.02913272e+00 4.60310489e-01 -1.79780409e-01
6.27439320e-01 -2.26073518e-01 4.66820970e-02 -1.83564782e-01
5.53377092e-01 4.69740182e-01 -5.27187400e-02 -9.82996345e-01
-5.56792855e-01 1.06515419e+00 1.04685977e-01 1.16905905e-02
1.23263264e+00 -1.25223994e-01 -1.39218360e-01 2.32329696e-01
1.07976604e+00 -1.96204245e-01 -1.48876488e+00 -3.01578045e-01
-2.45178163e-01 -8.83905470e-01 -2.98583984e-01 -2.37802595e-01
-1.16734898e+00 8.12815189e-01 5.73863804e-01 -3.43410343e-01
1.11736095e+00 1.66972578e-01 1.15126264e+00 9.79090407e-02
6.12212479e-01 -1.17489171e+00 4.07813996e-01 3.77595931e-01
9.85346496e-01 -1.03736317e+00 2.55764186e-01 -8.11132193e-01
-4.86982644e-01 8.80013287e-01 9.47659194e-01 -1.85302779e-01
5.06796300e-01 1.49861455e-01 -2.10306048e-01 -3.11696082e-01
-1.27269134e-01 -3.50044191e-01 5.44489920e-01 6.28939867e-01
2.19124988e-01 5.45585109e-03 -3.09989333e-01 3.50051194e-01
-9.61061716e-02 1.08655982e-01 -9.12320837e-02 1.05763519e+00
-3.19225609e-01 -9.97645497e-01 -7.63472140e-01 1.63006589e-01
-2.13293448e-01 3.15399170e-01 -3.42311323e-01 7.90682256e-01
2.82972455e-01 3.96323562e-01 1.02827772e-01 -4.70911592e-01
6.14514112e-01 -1.59670681e-01 8.43064964e-01 -7.15427697e-01
-3.82953435e-01 4.54363376e-01 -1.39646620e-01 -1.06056726e+00
-6.60301805e-01 -6.78473175e-01 -1.20267701e+00 -2.13059217e-01
-3.45679879e-01 -2.91835457e-01 3.18311036e-01 1.09860706e+00
9.49895829e-02 2.93582201e-01 2.05220431e-01 -1.56707072e+00
-4.97987449e-01 -8.04337800e-01 -5.26151478e-01 6.64163053e-01
2.33929634e-01 -1.17937768e+00 -1.02198482e-01 9.47259441e-02] | [7.089582443237305, -0.9200857877731323] |
a858234b-876b-423b-9b2d-0a1fc152e261 | ccg-supertagging-with-a-recurrent-neural | null | null | https://aclanthology.org/P15-2041 | https://aclanthology.org/P15-2041.pdf | CCG Supertagging with a Recurrent Neural Network | null | ['Stephen Clark', 'Michael Auli', 'Wenduan Xu'] | 2015-07-01 | ccg-supertagging-with-a-recurrent-neural-1 | https://aclanthology.org/P15-2041 | https://aclanthology.org/P15-2041.pdf | ijcnlp-2015-7 | ['ccg-supertagging'] | ['natural-language-processing'] | [-8.63703638e-02 1.71006292e-01 -6.22772932e-01 -4.08054382e-01
-8.41685571e-03 -9.08429027e-01 6.55310392e-01 -6.53472245e-01
-2.85945535e-01 1.06888819e+00 -4.63127941e-02 -1.01159286e+00
-3.91567826e-01 -9.63214397e-01 -4.95059669e-01 -6.31337762e-01
-9.79754329e-01 7.25764990e-01 3.30370307e-01 -6.93831444e-01
7.03166842e-01 7.88774848e-01 -1.68942046e+00 7.18545914e-01
7.04417467e-01 8.52217197e-01 2.49141872e-01 1.14950800e+00
-1.95044339e-01 1.55633950e+00 -7.48382092e-01 -5.46825826e-01
3.13719302e-01 -1.23176083e-01 -7.22945035e-01 -1.01074085e-01
9.28529128e-02 -8.59008506e-02 -2.09758401e-01 9.22211111e-01
5.37373662e-01 4.49454933e-02 1.08379531e+00 -1.42548037e+00
-5.91619551e-01 6.10313773e-01 -4.01565880e-02 1.21627934e-01
1.03678203e+00 -5.39447069e-01 1.19919395e+00 -1.13026452e+00
7.20913768e-01 1.26888943e+00 8.66221786e-01 5.44149756e-01
-1.22286928e+00 -1.94712028e-01 -3.26822817e-01 -9.51717794e-02
-1.46558487e+00 -3.25250506e-01 4.25783843e-02 -2.08119690e-01
1.66093647e+00 1.26596653e+00 1.20609856e+00 1.01401424e+00
1.26658809e+00 8.34431887e-01 1.04267764e+00 -5.13792276e-01
3.35295945e-01 3.66983831e-01 1.54683650e-01 6.33519173e-01
8.40953708e-01 5.26628852e-01 -7.06372619e-01 -9.13127720e-01
9.33553874e-01 -2.94925272e-01 1.71355158e-01 -5.05680561e-01
-9.05919552e-01 6.91228509e-01 1.78732842e-01 3.83959889e-01
-1.39880210e-01 9.89067405e-02 1.26390755e-01 5.30987144e-01
-2.58292928e-02 6.47037446e-01 -9.11868811e-01 -1.33165747e-01
-8.71728659e-01 5.10332465e-01 1.25398111e+00 1.52653182e+00
1.24482810e-01 2.94908643e-01 -9.34252143e-02 3.17179203e-01
8.92314315e-01 1.01808000e+00 4.28362608e-01 -1.36146402e+00
-6.87414408e-02 1.72361732e-01 5.01781464e-01 -8.52631688e-01
-6.33224547e-01 -9.64177120e-03 -8.93263519e-01 4.49267089e-01
3.49161088e-01 4.57367361e-01 -8.02827001e-01 5.07305264e-01
4.33481112e-02 -2.34125629e-01 4.53833073e-01 5.55570945e-02
4.99930978e-01 3.76208365e-01 -1.34477139e-01 -5.73289394e-01
1.06082785e+00 -1.36716676e+00 -1.35299087e+00 2.33215362e-01
9.05734658e-01 -1.07320261e+00 4.35900748e-01 5.33875942e-01
-1.55548143e+00 -1.37560293e-01 -1.08699942e+00 1.78573877e-01
-7.27255583e-01 -3.14239264e-01 8.57801437e-01 1.43120694e+00
-1.60129595e+00 9.73287821e-01 -4.91727620e-01 6.59165755e-02
1.20568443e-02 8.24621081e-01 -2.64718989e-03 4.62812334e-01
-1.33193445e+00 1.08501506e+00 2.22979754e-01 -1.21242590e-01
-1.65216476e-01 -2.13068098e-01 -8.23704481e-01 -5.63443303e-01
-4.78693932e-01 -5.29636025e-01 1.44139910e+00 -2.59346128e-01
-1.65295815e+00 9.71794367e-01 -1.42069459e-01 -1.97814897e-01
6.14786744e-01 -1.28011424e-02 -8.31891418e-01 2.42498964e-01
-1.89849049e-01 5.76383233e-01 9.28263724e-01 -1.35132408e+00
-7.59897232e-01 -1.67359829e-01 -1.23336017e-01 2.66287565e-01
-1.25510961e-01 1.89734384e-01 2.11616129e-01 -1.12999000e-01
3.27147305e-01 -7.20919967e-01 -2.53068686e-01 -5.32041907e-01
-1.46512717e-01 -7.10518599e-01 7.70373225e-01 -4.51523662e-01
1.83705616e+00 -1.67618537e+00 -2.06720144e-01 3.98590982e-01
3.57815564e-01 -1.24705513e-03 2.40583986e-01 1.08380008e+00
-2.76906848e-01 7.32199550e-01 4.11965609e-01 -1.10722095e-01
2.24991128e-01 4.85861301e-01 -4.16602850e-01 3.05609167e-01
-1.29282743e-01 1.15307164e+00 -1.16605783e+00 -5.23096442e-01
4.11106765e-01 9.42391157e-02 -4.58732933e-01 4.79237735e-01
2.99364805e-01 1.47170946e-01 -3.56553018e-01 1.39399457e+00
1.15709066e+00 -1.31984919e-01 1.45911396e-01 5.30878425e-01
-3.79135728e-01 3.55090618e-01 -6.96863770e-01 1.05554795e+00
7.08333924e-02 5.00173986e-01 1.02364108e-01 -7.94621468e-01
3.33247900e-01 8.61540735e-01 4.37155962e-01 -9.67555881e-01
-2.26398129e-02 6.92409754e-01 1.12803578e-01 -6.07703328e-01
7.58228302e-01 3.94563079e-02 -3.64872098e-01 6.40070081e-01
-2.37588286e-01 -6.59476995e-01 9.64643434e-02 3.08100313e-01
6.22585893e-01 -5.18246442e-02 5.62923312e-01 -1.05613089e+00
7.86340594e-01 -1.82965681e-01 -1.81299388e-01 1.03415680e+00
-3.09923887e-01 3.19085121e-01 2.99841821e-01 -6.65102363e-01
-6.45341039e-01 -1.12307119e+00 -4.89381433e-01 1.30636716e+00
3.24267983e-01 -4.39044595e-01 -9.54439282e-01 -2.49762803e-01
1.77620783e-01 6.89606130e-01 -5.90509653e-01 3.84124845e-01
-5.03739953e-01 -8.32535863e-01 7.39044368e-01 3.45434904e-01
-5.07752821e-02 -1.33414865e+00 -6.58416986e-01 1.25490099e-01
-2.20292807e-01 -6.63697243e-01 -6.23428151e-02 4.48765576e-01
-1.35989368e+00 -5.18594682e-01 -6.66252747e-02 -8.20914626e-01
5.87345481e-01 2.46782884e-01 1.27047324e+00 5.39230824e-01
-2.31483161e-01 4.26904231e-01 -1.21292919e-01 -4.95818377e-01
-4.59671497e-01 -8.00336525e-02 5.28869390e-01 -5.87835789e-01
5.19427478e-01 -2.50617653e-01 -7.29350567e-01 5.37953973e-01
-6.88540697e-01 1.62748516e-01 1.79803044e-01 1.04410267e+00
1.35816500e-01 -9.34035778e-02 1.22507080e-01 -6.38007045e-01
8.72274399e-01 -1.69219792e-01 -3.78732830e-01 5.77745810e-02
-6.77108407e-01 -3.74140263e-01 3.21430594e-01 -3.25342178e-01
-1.01981449e+00 -4.87835288e-01 -9.82677937e-02 2.45538145e-01
1.11353043e-02 -1.46784872e-01 6.47139177e-02 -5.24923325e-01
8.02199244e-01 9.25758183e-02 1.99174434e-02 -6.80815242e-03
3.01039815e-01 7.09525108e-01 -6.82967342e-03 -6.68678164e-01
8.44880998e-01 4.91470337e-01 7.98524171e-02 -9.57177758e-01
-1.52186140e-01 -2.60129690e-01 -9.51962709e-01 -6.54426932e-01
6.56643391e-01 -6.78531289e-01 -9.10833478e-01 3.91110867e-01
-9.38691139e-01 -3.38627815e-01 -3.91645581e-01 4.25431967e-01
-1.01278400e+00 2.75717527e-02 -3.90154392e-01 -1.27895141e+00
-5.10977268e-01 -1.02017939e+00 9.43384409e-01 5.30070923e-02
-5.10597289e-01 -1.26927447e+00 5.87685481e-02 2.71537274e-01
1.81734428e-01 -1.73075795e-01 6.90226793e-01 -2.38256708e-01
-4.24233019e-01 -1.53791070e-01 2.34436691e-02 -1.39755070e-01
1.70832314e-02 4.95917559e-01 -9.81751978e-01 -5.31145096e-01
6.65065646e-02 -1.92070693e-01 -1.08835101e-01 6.52520418e-01
5.91872573e-01 -2.29931593e-01 -8.56000841e-01 5.40386558e-01
1.38545322e+00 3.85070026e-01 5.32770038e-01 7.28214979e-01
1.41836226e-01 5.53460240e-01 9.17806149e-01 4.63203549e-01
1.30579369e-02 3.28798652e-01 2.40537539e-01 1.49327129e-01
1.11720070e-01 -1.54819340e-01 3.77893507e-01 1.16112018e+00
-8.18235934e-01 -2.69281328e-01 -5.07867396e-01 4.42987174e-01
-1.72482407e+00 -1.40330648e+00 -4.32368398e-01 6.90478683e-01
6.25676990e-01 1.56016424e-01 -1.48347050e-01 3.35214496e-01
4.99015123e-01 -2.03574806e-01 -1.19133167e-01 -1.06291151e+00
-1.43546045e-01 3.15233678e-01 7.37729073e-01 1.00061214e+00
-7.20721722e-01 1.03317809e+00 1.29781246e+01 1.02230716e+00
2.21112028e-01 1.03134915e-01 5.16071796e-01 3.48020852e-01
-4.36954498e-01 -4.56139445e-02 -1.04416132e+00 2.72933897e-02
1.38140702e+00 -4.30666685e-01 6.85999811e-01 5.44219851e-01
3.44648361e-01 -4.23268199e-01 -1.26188684e+00 5.26221812e-01
9.73738134e-02 -1.40886843e+00 -2.83300440e-04 6.85225725e-01
7.73699820e-01 -5.08050561e-01 6.22419357e-01 3.24184299e-01
6.09259963e-01 -1.14389277e+00 8.60300779e-01 2.53660440e-01
1.03040910e+00 -6.05088234e-01 5.67372203e-01 1.68872893e-01
-1.14389896e+00 -2.20873043e-01 -8.77727985e-01 -1.00755692e+00
3.93533185e-02 -1.81779593e-01 -4.29956943e-01 3.48861217e-01
9.58353162e-01 2.99398601e-01 -3.93658698e-01 9.95779395e-01
-4.78694476e-02 1.04875881e-02 -2.71853864e-01 -4.48467314e-01
4.83122796e-01 -3.54241252e-01 4.66730654e-01 1.00164843e+00
2.48499006e-01 3.51035744e-01 -9.84472036e-02 4.01770771e-01
5.45058846e-01 3.29446048e-02 -1.19659424e+00 -1.78908288e-01
2.83276141e-01 9.16795909e-01 -4.83487815e-01 -4.22520459e-01
-2.00212970e-01 8.62069130e-01 -3.55488248e-02 5.01107454e-01
-6.11489356e-01 -4.35615242e-01 9.72222984e-01 -1.27327025e-01
-1.14700586e-01 -3.48497719e-01 -6.23769283e-01 -7.30352640e-01
-5.89872956e-01 -4.54965204e-01 5.93606755e-02 -5.53365827e-01
-1.39813089e+00 5.79277515e-01 -2.27688253e-02 -1.40553558e+00
-6.99901402e-01 -1.27676582e+00 -4.76714373e-01 4.92853165e-01
-1.11898029e+00 -1.10984349e+00 2.50124663e-01 4.52870727e-01
1.64141744e-01 -5.34416080e-01 1.39563632e+00 3.57715860e-02
1.00637585e-01 9.24474537e-01 6.69434488e-01 -7.37814724e-01
5.56605101e-01 -1.27867436e+00 5.68737745e-01 -1.37897313e-01
-4.31265175e-01 9.05828118e-01 6.28349900e-01 -5.39804697e-01
-1.41196322e+00 -3.66917729e-01 1.08350635e+00 -9.83769417e-01
6.55218959e-01 -3.86345625e-01 4.23767231e-02 7.88592756e-01
7.15902448e-01 -6.18741751e-01 8.21781039e-01 -1.83753878e-01
1.80774391e-01 5.75296998e-01 -1.39248300e+00 6.12354755e-01
1.66275799e+00 -4.63594139e-01 -6.25784039e-01 7.60327101e-01
8.13696027e-01 -6.94087505e-01 -1.30082703e+00 3.34633321e-01
8.65424156e-01 -8.75409484e-01 1.61978090e+00 -1.32660246e+00
-4.43697497e-02 2.81152606e-01 -2.61993498e-01 -9.32519078e-01
-5.96193194e-01 -1.23518765e+00 -5.33532679e-01 -5.83747849e-02
5.96577883e-01 -1.13057327e+00 3.42365682e-01 8.74560475e-01
-2.82833427e-01 -6.42737269e-01 -1.06996536e+00 -1.32016802e+00
-3.35779637e-02 -1.45572275e-01 4.90409225e-01 7.63798356e-01
6.83744550e-01 1.09839931e-01 -6.36873543e-02 -1.00294888e-01
5.33176839e-01 9.55312885e-03 4.41501856e-01 -1.34294486e+00
3.86843324e-01 -5.75816095e-01 -3.07655483e-01 -9.45992947e-01
-8.85957032e-02 -8.12076271e-01 -6.53862000e-01 -1.28511906e+00
-8.32044985e-03 -1.91056758e-01 -1.11109078e-01 -1.65725678e-01
3.67937148e-01 2.13746816e-01 1.20859891e-02 1.03788137e-01
-3.71160030e-01 6.18435517e-02 1.29639816e+00 6.91750320e-05
-1.62315920e-01 4.85058486e-01 -4.81304944e-01 7.84440815e-01
8.58408585e-02 -2.99253196e-01 -6.78878546e-01 6.11881316e-02
6.69384480e-01 4.61409837e-02 3.24159935e-02 -7.42885649e-01
5.37211418e-01 -3.75702560e-01 4.78586555e-01 -1.32223868e+00
1.30741090e-01 -9.61415648e-01 6.95283338e-02 9.40189242e-01
2.74610907e-01 1.20424610e-02 8.66204947e-02 5.39500564e-02
-1.44871444e-01 -5.70943117e-01 9.21121240e-01 -4.07591403e-01
-4.92852688e-01 -5.20386267e-03 -1.03226590e+00 8.97834301e-02
9.92593169e-01 -7.84614205e-01 -3.59281451e-01 -4.20183957e-01
-8.29068601e-01 -1.95836127e-02 6.50830388e-01 3.03609259e-02
7.30431557e-01 -1.51530886e+00 -2.30721906e-01 7.18729138e-01
-3.22939813e-01 -3.74400020e-01 -1.70157343e-01 6.58265352e-01
-1.32361674e+00 1.02442718e+00 -5.41665435e-01 -4.55340147e-01
-1.14228773e+00 4.78126436e-01 4.28307921e-01 -2.41845414e-01
-2.15481281e-01 1.11954463e+00 2.71224789e-02 -8.23025763e-01
1.85185194e-01 -9.21545625e-02 -7.54407048e-01 3.55081353e-03
6.88606799e-01 1.05194807e+00 -2.91290224e-01 -6.04341030e-01
-4.56784427e-01 6.65885091e-01 2.32151806e-01 -2.87484169e-01
9.17833567e-01 -2.19243199e-01 -9.89108324e-01 4.28274393e-01
8.38715494e-01 -1.36269778e-01 -3.69319022e-02 4.16855574e-01
1.25943512e-01 -8.02164078e-01 -4.32406247e-01 -3.55811834e-01
-1.85641110e-01 5.54822803e-01 5.34874737e-01 8.99602413e-01
8.57008278e-01 -3.02566767e-01 8.18335712e-01 9.66778398e-01
5.72402716e-01 -1.68019545e+00 -2.13140488e-01 6.89524531e-01
9.12339568e-01 -9.28350806e-01 5.44190466e-01 -7.27165341e-01
-4.14997995e-01 1.32979155e+00 4.68304873e-01 -1.55325383e-01
1.27306652e+00 5.74917436e-01 1.14069022e-02 -3.70670199e-01
-9.44949508e-01 1.12705544e-01 3.75366658e-01 1.11147714e+00
5.06513238e-01 5.07374525e-01 -9.66500878e-01 3.21953118e-01
-7.28706717e-01 -2.34555230e-01 4.90474731e-01 1.41972518e+00
-6.43810987e-01 -1.20391917e+00 -7.23931909e-01 4.61561680e-01
-5.49773455e-01 -1.16372630e-01 -5.28106689e-01 8.46754074e-01
-5.76629937e-02 1.49448860e+00 -1.97535474e-03 -5.03491640e-01
4.11356747e-01 1.41089618e-01 7.21762300e-01 -1.23501487e-01
-9.04846430e-01 4.15413082e-01 3.84890139e-01 -1.23056793e+00
-8.58632207e-01 -1.05834293e+00 -1.40667629e+00 -1.19437599e+00
-5.12782812e-01 1.89310342e-01 3.83317530e-01 3.90289724e-01
-2.06836104e-01 2.85260603e-02 9.80917513e-01 -1.07949340e+00
-3.90341938e-01 -9.39418614e-01 -1.00026262e+00 -7.84516707e-02
2.89751232e-01 -8.00943017e-01 -7.83523321e-01 2.83909619e-01] | [-7.261399269104004, 3.6752259731292725] |
fbcf7743-aa38-4101-ad3c-96da02daa283 | adversarial-generation-of-time-frequency | 1902.04072 | null | https://arxiv.org/abs/1902.04072v2 | https://arxiv.org/pdf/1902.04072v2.pdf | Adversarial Generation of Time-Frequency Features with application in audio synthesis | Time-frequency (TF) representations provide powerful and intuitive features for the analysis of time series such as audio. But still, generative modeling of audio in the TF domain is a subtle matter. Consequently, neural audio synthesis widely relies on directly modeling the waveform and previous attempts at unconditionally synthesizing audio from neurally generated invertible TF features still struggle to produce audio at satisfying quality. In this article, focusing on the short-time Fourier transform, we discuss the challenges that arise in audio synthesis based on generated invertible TF features and how to overcome them. We demonstrate the potential of deliberate generative TF modeling by training a generative adversarial network (GAN) on short-time Fourier features. We show that by applying our guidelines, our TF-based network was able to outperform a state-of-the-art GAN generating waveforms directly, despite the similar architecture in the two networks. | ['Nathanaël Perraudin', 'Andrés Marafioti', 'Piotr Majdak', 'Nicki Holighaus'] | 2019-02-11 | adversarial-generation-of-time-frequency-1 | https://tifgan.github.io/ | http://proceedings.mlr.press/v97/marafioti19a/marafioti19a.pdf | 36th-international-conference-on-machine | ['audio-generation'] | ['audio'] | [ 6.39929116e-01 3.89886469e-01 4.05632526e-01 -1.50223345e-01
-1.28254747e+00 -7.44969428e-01 6.59367263e-01 -5.91420770e-01
3.20558757e-01 7.91341126e-01 2.42184266e-01 -2.28462398e-01
5.66030070e-02 -8.59035313e-01 -9.24328744e-01 -6.79485142e-01
-1.70792565e-01 1.33788347e-01 -3.10285866e-01 -2.77485728e-01
-1.73758700e-01 2.48185590e-01 -1.76057947e+00 4.72382456e-01
6.19554102e-01 1.15211654e+00 -2.10684732e-01 1.04342806e+00
3.37253441e-03 6.50283992e-01 -1.20003724e+00 -4.19875830e-01
1.98649198e-01 -9.83147740e-01 -5.03127337e-01 -2.44904816e-01
3.60606045e-01 -2.01280132e-01 -3.58058333e-01 5.08554816e-01
6.64784670e-01 9.85845476e-02 7.88565099e-01 -1.44347548e+00
-7.43725002e-01 8.92547965e-01 1.44354537e-01 -1.32562630e-02
5.54180205e-01 2.25441381e-01 1.12406301e+00 -7.24856973e-01
5.18774509e-01 1.04702961e+00 9.44018543e-01 7.03227460e-01
-1.48890865e+00 -7.94875145e-01 -3.77690285e-01 -5.77785783e-02
-1.21430886e+00 -5.90319037e-01 1.11946464e+00 -3.56325358e-01
9.33083236e-01 3.64566565e-01 8.34735096e-01 1.52964389e+00
2.65087098e-01 5.34450054e-01 9.74319696e-01 -4.90065664e-01
3.13478224e-02 -3.02416474e-01 -6.39096677e-01 1.74629927e-01
-6.01715505e-01 4.49569345e-01 -8.01858842e-01 -8.17812085e-02
8.21222007e-01 -5.47951758e-01 -3.00812215e-01 1.70695305e-01
-1.16142988e+00 8.75939846e-01 2.19963893e-01 4.80510086e-01
-3.80862594e-01 8.23564410e-01 2.80277133e-01 5.74038625e-01
4.14968640e-01 8.26513767e-01 -2.51861274e-01 -5.34353495e-01
-1.33333707e+00 4.71245974e-01 6.23991191e-01 7.58279145e-01
3.34709138e-01 1.13400722e+00 -2.78952152e-01 4.98692095e-01
3.31937000e-02 6.00821018e-01 5.79203129e-01 -1.02000403e+00
6.79634884e-02 -2.42728159e-01 -2.67164588e-01 -7.67255545e-01
-1.51132420e-02 -6.21196687e-01 -6.58860683e-01 1.75065428e-01
4.29098696e-01 -4.62273300e-01 -1.14507341e+00 2.07448769e+00
-7.39863962e-02 4.96743500e-01 3.38677224e-03 5.68030298e-01
8.80956769e-01 9.22098339e-01 -1.71590120e-01 -6.75321966e-02
9.11710262e-01 -5.83162427e-01 -9.54601586e-01 8.62345919e-02
-5.53263314e-02 -9.72507179e-01 1.12416255e+00 4.85180110e-01
-1.45106459e+00 -8.22581291e-01 -1.10184872e+00 -6.70298040e-02
-2.77136862e-01 -1.96429974e-04 5.34572423e-01 9.10751581e-01
-1.19694805e+00 8.09306324e-01 -6.14567399e-01 2.29261771e-01
1.13336667e-01 3.02934498e-01 -2.65424043e-01 4.94134843e-01
-1.46449757e+00 5.45064092e-01 -2.11962517e-02 -1.25927880e-01
-1.27689779e+00 -1.23661697e+00 -7.99474180e-01 2.98764855e-01
-1.38708189e-01 -8.86598408e-01 1.60663807e+00 -9.49036479e-01
-1.99838710e+00 3.41107905e-01 -1.40103206e-01 -8.00431252e-01
4.55541760e-01 -2.11305648e-01 -7.82513797e-01 2.55125821e-01
-1.38241038e-01 7.79951215e-01 1.47078145e+00 -1.02787328e+00
-1.98544040e-01 4.08752948e-01 -1.13395192e-01 -4.07139450e-01
-1.11687593e-01 -2.09104270e-01 4.05052155e-01 -1.18397200e+00
-2.84473479e-01 -8.46432269e-01 1.13160580e-01 -3.56419265e-01
-4.01181936e-01 6.51626810e-02 9.00508702e-01 -6.55722857e-01
1.12419105e+00 -2.06865644e+00 -1.09712351e-02 1.23016283e-01
4.53109518e-02 2.11690262e-01 -1.92821890e-01 7.98936069e-01
-4.84583110e-01 3.49822819e-01 -6.85631037e-02 -4.01069075e-01
3.57771397e-01 -1.20530985e-01 -1.13844502e+00 5.46503775e-02
7.19996810e-01 1.08176327e+00 -8.22236836e-01 3.67245302e-02
9.76419598e-02 1.00908220e+00 -7.85919964e-01 3.47588390e-01
-4.14989233e-01 8.63647401e-01 -3.42331156e-02 3.48024756e-01
1.74956605e-01 1.24346800e-01 -2.15645701e-01 -2.72786081e-01
-9.75975208e-03 7.24823236e-01 -6.34109318e-01 1.77402735e+00
-7.54889727e-01 9.11620855e-01 -1.30561337e-01 -8.47801685e-01
9.46339488e-01 8.31578135e-01 5.19784331e-01 -4.67301786e-01
1.55699611e-01 4.01341796e-01 1.54141679e-01 -1.01045556e-01
7.89084211e-02 -6.27325177e-01 -1.41413927e-01 5.98627210e-01
5.88663757e-01 -8.37148726e-01 -6.84475601e-02 -2.06626847e-01
1.07602012e+00 3.25121939e-01 -7.45486170e-02 1.89464897e-01
2.84055620e-01 -4.26024973e-01 7.86850527e-02 5.27624190e-01
4.17984188e-01 1.09986365e+00 4.15452570e-01 -2.08016083e-01
-1.28454733e+00 -1.30776620e+00 9.52704698e-02 9.11197960e-01
-8.34305227e-01 -6.74812198e-01 -8.95959020e-01 -2.30339393e-01
-1.32087082e-01 1.00815177e+00 -7.60379791e-01 -4.69939113e-01
-6.56166136e-01 -1.58881739e-01 1.16591382e+00 4.84215349e-01
-2.67392397e-02 -1.17981207e+00 -5.37477016e-01 6.21906638e-01
-1.86950549e-01 -8.77542615e-01 -4.05368447e-01 4.45360869e-01
-5.84241509e-01 -2.87498891e-01 -9.55532253e-01 -3.75117332e-01
-1.38812177e-02 -3.21786016e-01 1.26227832e+00 -3.04108322e-01
-3.97667319e-01 4.24693733e-01 -3.31500679e-01 -8.16770077e-01
-7.88923860e-01 2.07623690e-01 -3.72793600e-02 2.55161915e-02
-2.75307864e-01 -1.28267777e+00 -4.41416860e-01 4.22968566e-02
-1.10127783e+00 -6.13063462e-02 2.77992755e-01 9.24840808e-01
3.72115403e-01 6.95620924e-02 1.18867636e+00 -5.57591021e-01
8.60064507e-01 -3.65551502e-01 -1.18432350e-01 -2.34852284e-01
-3.38096201e-01 5.54359257e-02 9.34127271e-01 -7.55620003e-01
-8.06849778e-01 -1.86428607e-01 -7.24002302e-01 -6.48155153e-01
5.02924249e-02 3.98066878e-01 4.74614743e-03 -9.99145769e-03
7.70006418e-01 3.59770685e-01 -9.28695686e-03 -3.08220983e-01
6.67676568e-01 4.04942006e-01 7.88276970e-01 -6.86195612e-01
1.12162900e+00 2.26777762e-01 6.19819127e-02 -5.93701184e-01
-6.61962628e-01 3.27050328e-01 -1.67481571e-01 -3.11575830e-01
6.66546583e-01 -7.77748048e-01 -4.93835807e-01 3.51824731e-01
-9.75711286e-01 -3.56526732e-01 -9.06298280e-01 3.66480291e-01
-1.20547462e+00 -2.48343199e-01 -6.45446062e-01 -8.80182147e-01
-3.59761328e-01 -8.17581236e-01 1.18994248e+00 -7.96090215e-02
-5.86709678e-01 -8.67770970e-01 1.35819957e-01 -6.14517815e-02
9.68150437e-01 7.06653297e-01 9.90167201e-01 -3.47776890e-01
-4.42481488e-01 -1.57096326e-01 4.45296347e-01 4.08936977e-01
2.51380056e-01 3.10874611e-01 -1.54222190e+00 -5.48521802e-02
1.99584961e-01 -3.98076147e-01 6.20817482e-01 4.57919151e-01
1.11601830e+00 -3.53346527e-01 2.64966816e-01 8.84289086e-01
9.74153399e-01 3.40703934e-01 8.48027706e-01 -1.17884822e-01
4.03470129e-01 2.72206545e-01 1.78844079e-01 4.47863132e-01
-2.08039105e-01 5.52168787e-01 2.60514230e-01 -2.34292626e-01
-5.75725198e-01 -7.34176457e-01 5.04900157e-01 8.09028387e-01
-1.89980656e-01 -3.65851790e-01 -5.12306929e-01 5.58480382e-01
-1.20354331e+00 -1.25688434e+00 1.88835055e-01 1.86162162e+00
1.14560318e+00 1.65643215e-01 2.75331289e-01 8.37389410e-01
4.04813439e-01 9.97526422e-02 -3.88564974e-01 -5.91712177e-01
-1.30011842e-01 1.09544063e+00 -9.49595422e-02 1.70078963e-01
-9.06888485e-01 6.10767543e-01 7.56467533e+00 9.11721289e-01
-1.68040514e+00 1.64760754e-01 3.90405029e-01 -3.01565886e-01
-7.06863403e-01 -3.14849168e-01 -3.15324068e-01 4.11009878e-01
1.57242131e+00 -5.79494476e-01 7.61616528e-01 5.36793768e-01
1.10715970e-01 7.22826123e-01 -1.28348172e+00 7.40976989e-01
-1.69180006e-01 -1.42515051e+00 3.04175079e-01 -7.47665018e-03
6.98391199e-01 -4.39497709e-01 8.14190447e-01 4.50169742e-01
5.33333421e-02 -1.73386502e+00 1.18136406e+00 4.68141556e-01
1.30346847e+00 -8.64550591e-01 2.39477038e-01 2.21875519e-03
-1.25056362e+00 6.41062334e-02 1.75331786e-01 9.80979204e-03
4.67450798e-01 7.81564474e-01 -1.07166755e+00 4.72247183e-01
3.84150535e-01 4.13363606e-01 -7.39666894e-02 8.03001940e-01
-2.05144003e-01 1.03695333e+00 -2.59011120e-01 3.88248026e-01
2.51461774e-01 1.52116984e-01 5.09099305e-01 1.13383436e+00
9.93633568e-01 -1.41494527e-01 -4.22110707e-01 1.24985421e+00
-1.07588097e-01 -3.29095840e-01 -7.84242809e-01 -6.40342295e-01
2.76876420e-01 1.01102161e+00 -3.07152003e-01 -3.66896391e-02
-7.93471709e-02 6.43310249e-01 -1.17886089e-01 3.10686529e-01
-1.27179527e+00 -5.06914616e-01 6.04404271e-01 3.43947381e-01
5.84494591e-01 -3.22400242e-01 -8.32451954e-02 -7.81179488e-01
-4.91821915e-02 -1.11956334e+00 1.13834832e-02 -1.16396332e+00
-1.33992290e+00 1.03112721e+00 -2.36601710e-01 -1.48860955e+00
-1.08773756e+00 -2.62444735e-01 -7.34700441e-01 1.00139248e+00
-1.21173513e+00 -1.37569308e+00 2.01474857e-02 5.87843895e-01
4.35879856e-01 -1.00604944e-01 1.18589830e+00 3.42170626e-01
2.55917519e-01 8.38585138e-01 -9.65607539e-02 -1.67562231e-01
6.86801374e-01 -1.27860236e+00 7.92381763e-01 5.88380873e-01
4.91030335e-01 6.84871674e-01 1.03532672e+00 -2.33464152e-01
-1.34859812e+00 -1.09729302e+00 6.50742292e-01 -2.67803997e-01
6.91262722e-01 -5.73079586e-01 -7.06583679e-01 6.71496034e-01
4.56598490e-01 -1.47258565e-01 8.18429887e-01 -2.74611086e-01
-3.93596679e-01 -2.24232860e-02 -1.13104951e+00 4.58493263e-01
1.01916111e+00 -9.84724820e-01 -6.13847911e-01 1.08687252e-01
9.82410669e-01 -3.82741451e-01 -1.11494040e+00 3.75956774e-01
5.90123832e-01 -9.45045412e-01 1.24421322e+00 -5.64186811e-01
7.27621019e-01 -2.77637690e-01 -5.86687736e-02 -1.71994662e+00
-2.24654973e-01 -1.40093017e+00 -2.09057093e-01 1.47361577e+00
3.67997855e-01 -5.61877549e-01 5.52798748e-01 9.41503700e-03
-4.07851577e-01 -3.60010922e-01 -1.00238144e+00 -8.75071526e-01
2.92457104e-01 -6.72286630e-01 8.63989353e-01 6.48387730e-01
-6.35444224e-02 2.51047641e-01 -6.08362079e-01 -1.70671359e-01
1.47940740e-01 1.58626124e-01 7.89035261e-01 -1.09537244e+00
-5.52180469e-01 -4.91750240e-01 -4.00307268e-01 -6.85503185e-01
2.44783342e-01 -6.87097609e-01 1.35549724e-01 -1.07569492e+00
-5.78023851e-01 -3.46942276e-01 -1.00725323e-01 3.03283662e-01
1.57338917e-01 6.50031030e-01 3.56856883e-01 -3.31140101e-01
3.24096441e-01 7.22724557e-01 1.36497855e+00 -1.09961644e-01
2.09251195e-01 1.15553990e-01 -6.08851135e-01 3.99702221e-01
7.37130225e-01 -4.66257125e-01 -7.05354571e-01 -1.43858284e-01
2.19274312e-01 3.37548286e-01 4.75376457e-01 -1.35206401e+00
-1.34277586e-02 2.26348251e-01 4.94907707e-01 -3.55050892e-01
9.12116289e-01 -5.75750470e-01 8.13331366e-01 1.62641391e-01
-5.08953154e-01 9.75517333e-02 3.95410836e-01 3.49968642e-01
-6.13077283e-01 -5.21189421e-02 4.94511366e-01 -2.73496397e-02
3.09390463e-02 5.70289269e-02 -5.78580976e-01 2.57983506e-01
5.89089572e-01 1.26228493e-03 -2.43477985e-01 -1.03824329e+00
-9.50225115e-01 -7.65501738e-01 2.05901384e-01 3.31813455e-01
5.37671208e-01 -1.52561188e+00 -7.20284224e-01 5.61535060e-01
-3.71461511e-01 -3.45716000e-01 2.20003501e-01 5.29493630e-01
-1.05487928e-01 5.19275129e-01 -3.99598986e-01 -4.85523432e-01
-9.40081060e-01 3.60083938e-01 3.79593015e-01 -1.75992712e-01
-5.30391514e-01 9.12983716e-01 1.71334282e-01 -1.45818874e-01
7.70889297e-02 -4.62539166e-01 8.63005444e-02 -3.71285677e-02
3.22122216e-01 -1.17580099e-02 2.98456967e-01 -4.11599576e-01
-8.58902559e-02 5.91259599e-01 6.40017390e-01 -6.99175119e-01
1.46220231e+00 3.88228387e-01 3.13135594e-01 5.96122324e-01
1.32560337e+00 4.07980114e-01 -1.25331032e+00 2.91395515e-01
-5.66278636e-01 -3.59250963e-01 -1.05723634e-01 -7.28640616e-01
-1.14887094e+00 1.22910702e+00 2.28023037e-01 7.34471560e-01
1.34030497e+00 -2.69900680e-01 1.03798032e+00 -2.73786969e-02
4.24188912e-01 -5.17391145e-01 1.86103821e-01 4.64662969e-01
1.11996722e+00 -3.17306817e-01 -4.07844841e-01 -4.13099140e-01
-4.06779885e-01 1.28833652e+00 1.93318340e-03 -4.33563650e-01
5.99017620e-01 6.70192480e-01 1.41314298e-01 1.09982342e-01
-9.37020540e-01 2.23737713e-02 6.01641834e-01 8.02175641e-01
7.33792305e-01 -2.02622578e-01 8.36870298e-02 6.33943021e-01
-1.20091546e+00 5.19672446e-02 4.68695432e-01 7.39120901e-01
7.87194222e-02 -1.31722999e+00 -3.97850424e-01 3.57394814e-01
-9.62379396e-01 -2.50364810e-01 -2.96490818e-01 6.93555653e-01
9.19416696e-02 1.01534164e+00 4.49146703e-02 -6.11452937e-01
9.56384018e-02 4.05388564e-01 7.04514325e-01 -3.25031251e-01
-9.15913045e-01 2.43521914e-01 2.11754799e-01 -4.05647695e-01
-2.66387731e-01 -5.98304331e-01 -1.09482884e+00 -1.68482468e-01
-1.76349461e-01 1.64552808e-01 7.81605244e-01 5.86596191e-01
3.92327487e-01 1.16000819e+00 7.61049271e-01 -1.09111321e+00
-5.91864228e-01 -9.65005159e-01 -4.47220534e-01 4.10865396e-01
6.35663509e-01 -5.47924876e-01 -6.63333893e-01 5.40899992e-01] | [15.6019287109375, 5.932265758514404] |
f4f9c528-1c11-4fcb-a243-cff68ed2df93 | relationship-extraction-for-knowledge-graph | 2201.01647 | null | https://arxiv.org/abs/2201.01647v4 | https://arxiv.org/pdf/2201.01647v4.pdf | Comparison of biomedical relationship extraction methods and models for knowledge graph creation | Biomedical research is growing at such an exponential pace that scientists, researchers, and practitioners are no more able to cope with the amount of published literature in the domain. The knowledge presented in the literature needs to be systematized in such a way that claims and hypotheses can be easily found, accessed, and validated. Knowledge graphs can provide such a framework for semantic knowledge representation from literature. However, in order to build a knowledge graph, it is necessary to extract knowledge as relationships between biomedical entities and normalize both entities and relationship types. In this paper, we present and compare few rule-based and machine learning-based (Naive Bayes, Random Forests as examples of traditional machine learning methods and DistilBERT, PubMedBERT, T5 and SciFive-based models as examples of modern deep learning transformers) methods for scalable relationship extraction from biomedical literature, and for the integration into the knowledge graphs. We examine how resilient are these various methods to unbalanced and fairly small datasets. Our experiments show that transformer-based models handle well both small (due to pre-training on a large dataset) and unbalanced datasets. The best performing model was the PubMedBERT-based model fine-tuned on balanced data, with a reported F1-score of 0.92. DistilBERT-based model followed with F1-score of 0.89, performing faster and with lower resource requirements. BERT-based models performed better then T5-based generative models. | ['Wolfgang Thielemann', 'Nikola Milosevic'] | 2022-01-05 | null | null | null | null | ['key-information-extraction'] | ['natural-language-processing'] | [-1.11820288e-01 4.48325098e-01 -4.69792217e-01 -7.58942962e-02
-4.32180673e-01 -4.87456173e-01 4.67754334e-01 7.84118831e-01
-3.24511915e-01 1.20928693e+00 1.25403315e-01 -6.42651916e-01
-6.62197173e-01 -1.13882697e+00 -7.08872199e-01 -2.89260745e-01
-3.97960236e-03 8.83700788e-01 1.76519901e-01 -1.77408949e-01
1.17648870e-01 5.93254209e-01 -1.10128117e+00 2.25096598e-01
8.56705964e-01 7.93617308e-01 -2.08154880e-03 3.63556474e-01
-3.66464138e-01 9.33882952e-01 -6.74377561e-01 -7.33197451e-01
-2.34861657e-01 -2.59533167e-01 -1.12884390e+00 -5.12281418e-01
-2.57280674e-02 2.34871864e-01 -1.93267137e-01 7.60043323e-01
5.15435696e-01 -2.45972201e-01 9.01931345e-01 -1.16997552e+00
-6.30215168e-01 9.80085909e-01 -3.42444897e-01 2.50272781e-01
3.50164533e-01 -2.79022634e-01 6.52483702e-01 -6.60527110e-01
1.02313447e+00 1.12226057e+00 1.02827692e+00 3.10158312e-01
-1.05415499e+00 -8.41409922e-01 -3.29651654e-01 4.73056674e-01
-1.35411119e+00 -2.62826234e-01 3.88970673e-01 -5.57809472e-01
1.17219007e+00 1.59989014e-01 7.41898715e-01 1.13996434e+00
2.93179214e-01 8.80911499e-02 1.00378013e+00 -5.49984634e-01
2.52256840e-01 3.73572826e-01 2.10689038e-01 7.74780154e-01
9.24293876e-01 -2.44990006e-01 -5.73109090e-01 -2.05571666e-01
6.24180377e-01 -1.07260354e-01 -6.58168271e-02 -1.20540947e-01
-1.14132631e+00 6.45857513e-01 5.91073751e-01 7.81346738e-01
-6.42338037e-01 -3.76616940e-02 4.23284411e-01 1.20225288e-01
3.19031298e-01 7.27946341e-01 -7.82210410e-01 -9.43241045e-02
-1.18213224e+00 2.51406848e-01 1.12642717e+00 9.67364192e-01
4.76764649e-01 -7.77377188e-02 -2.45636165e-01 6.58420026e-01
1.86317474e-01 2.47602686e-01 5.30446351e-01 -6.38852239e-01
4.76463027e-02 8.62671971e-01 -2.73578733e-01 -1.23853743e+00
-6.03348911e-01 -7.92448282e-01 -9.45778251e-01 -4.21076715e-01
4.48906720e-01 -1.01091433e-02 -1.11182225e+00 1.58749270e+00
2.53543109e-01 -8.63943398e-02 1.37659922e-01 2.60632545e-01
1.29513288e+00 1.73042089e-01 4.07785147e-01 -2.94564337e-01
1.54520655e+00 -3.65115702e-01 -9.23444211e-01 1.78523064e-01
7.59640336e-01 -7.08326697e-01 4.11217034e-01 3.28937352e-01
-9.58352149e-01 -1.51210815e-01 -9.26590681e-01 -2.05354840e-01
-9.85286534e-01 -9.27823260e-02 7.10996985e-01 6.20992243e-01
-1.00825238e+00 6.53178990e-01 -6.47340775e-01 -5.11196136e-01
9.52525616e-01 4.39001054e-01 -4.79956329e-01 -1.32830277e-01
-1.41487348e+00 1.43456531e+00 6.64414108e-01 -1.59065038e-01
-5.33310175e-01 -9.58807349e-01 -5.83009362e-01 1.74064636e-01
3.67660761e-01 -1.09381080e+00 8.14429045e-01 -2.69094825e-01
-8.18288743e-01 8.00918400e-01 -9.43882540e-02 -6.28330529e-01
4.08441722e-01 1.46708488e-01 -3.77538234e-01 1.13644779e-01
2.10359082e-01 4.52152580e-01 1.63055196e-01 -1.23332882e+00
-3.12872797e-01 -5.02640009e-01 -5.87603524e-02 -2.46902466e-01
-3.47443670e-01 -7.68657327e-02 -1.06608137e-01 -2.40651682e-01
-3.10483538e-02 -5.92574835e-01 -8.25049877e-02 -2.37191826e-01
-6.04455471e-01 -3.26683372e-01 6.32641733e-01 -8.30590546e-01
1.26390779e+00 -1.48809636e+00 -1.13875814e-01 3.04513395e-01
5.40434003e-01 5.04606068e-01 3.45878959e-01 5.04065156e-01
-1.80859745e-01 4.47323084e-01 -2.13238131e-02 3.02736133e-01
-1.92287028e-01 4.40862656e-01 9.55059081e-02 2.90309414e-02
1.31225929e-01 1.03104353e+00 -1.07929325e+00 -8.33654761e-01
-9.86696221e-03 5.79006016e-01 -3.81687731e-01 -2.37197593e-01
-1.95515513e-01 3.51919830e-02 -5.42701721e-01 7.37816274e-01
3.26358348e-01 -4.74586219e-01 4.94933605e-01 -4.75584984e-01
1.29874453e-01 2.88205296e-01 -6.67950511e-01 1.36203384e+00
-4.87497091e-01 4.46058869e-01 -3.45812738e-01 -1.31197298e+00
1.08257508e+00 6.05860591e-01 5.63668787e-01 -5.67100704e-01
2.02782184e-01 4.88231152e-01 1.84247911e-01 -6.39340162e-01
1.30741909e-01 -4.24221754e-01 2.15954661e-01 5.32779731e-02
3.48574817e-01 4.58553769e-02 4.84836340e-01 4.04792219e-01
1.13252354e+00 1.67438954e-01 6.14136994e-01 -2.95231581e-01
3.32525700e-01 3.49969119e-01 3.93888474e-01 5.20258069e-01
2.84385622e-01 1.59327284e-01 7.42759109e-01 -4.06179518e-01
-9.21933293e-01 -6.80082023e-01 -3.80415350e-01 4.58189338e-01
-4.30962026e-01 -5.04235506e-01 -6.43102109e-01 -6.52257562e-01
1.61834583e-01 8.39170992e-01 -6.38300598e-01 -1.32848412e-01
-2.84972966e-01 -9.93398190e-01 8.18840444e-01 4.15057987e-01
3.88624877e-01 -9.00478482e-01 -4.60567176e-01 2.56676972e-01
-2.56724387e-01 -1.01699018e+00 4.41426665e-01 4.17392612e-01
-9.37698126e-01 -1.44165015e+00 -5.87730348e-01 -5.82443118e-01
4.72340137e-01 -4.54735577e-01 1.27388859e+00 -6.95850477e-02
-4.30351377e-01 1.63438395e-02 -2.80161738e-01 -8.43499541e-01
-5.46323359e-01 1.99864060e-01 -2.05034181e-01 -6.03006840e-01
6.22876287e-01 -5.15511274e-01 -3.97546947e-01 -5.90384118e-02
-8.68907630e-01 -8.44334513e-02 7.81920135e-01 9.40772355e-01
6.03810728e-01 1.98922038e-01 1.03396678e+00 -9.82574046e-01
6.11717582e-01 -8.10482383e-01 -7.32478797e-02 4.40742344e-01
-1.15832138e+00 1.40593514e-01 4.04657185e-01 -2.63322383e-01
-7.49602795e-01 -2.26125121e-01 -2.77111530e-01 -2.47868717e-01
-6.36155754e-02 1.00523138e+00 -1.14543654e-01 1.35053173e-01
9.73118424e-01 -8.74123275e-02 3.61033082e-02 -4.49259043e-01
3.67100894e-01 6.34556592e-01 3.52142692e-01 -5.20079553e-01
4.56324190e-01 1.82072163e-01 4.06829923e-01 -6.69820189e-01
-8.62712026e-01 -2.72434741e-01 -6.26349747e-01 3.35603300e-03
8.19609940e-01 -6.15907788e-01 -6.46968424e-01 -4.41771857e-02
-1.07901227e+00 4.37006317e-02 -3.68186414e-01 5.02969384e-01
-2.21297503e-01 1.01486258e-01 -2.72767782e-01 -6.54913008e-01
-5.52562952e-01 -8.45465302e-01 6.66216910e-01 1.26942664e-01
-6.45274401e-01 -1.24112117e+00 1.86580848e-02 4.78484750e-01
4.90300983e-01 5.89410841e-01 1.44710124e+00 -1.22124577e+00
-2.45553061e-01 -2.69398719e-01 -4.09512043e-01 2.88444348e-02
3.16954494e-01 -5.87666221e-02 -7.99464762e-01 2.05667272e-01
-2.82131761e-01 -9.10983905e-02 7.72586167e-01 4.06538248e-01
9.88764048e-01 -4.94903713e-01 -7.49561846e-01 2.67930746e-01
1.43929684e+00 3.18507344e-01 7.08940268e-01 5.30582905e-01
7.86414444e-01 5.83632350e-01 7.99340904e-02 -4.63546440e-02
5.47462285e-01 4.37600851e-01 1.03274867e-01 -1.02026984e-01
-3.14735949e-01 -2.07569197e-01 -3.25089812e-01 8.26453030e-01
-3.23739320e-01 -1.05114177e-01 -1.27423155e+00 6.66881800e-01
-1.76452947e+00 -9.56559598e-01 -4.06927556e-01 1.95004761e+00
1.39207709e+00 2.17416301e-01 9.00533050e-02 3.01464289e-01
4.70202535e-01 -5.54527879e-01 -3.47736239e-01 -4.25962240e-01
-1.00710690e-01 6.44563079e-01 5.47104359e-01 1.09090447e-01
-7.34199643e-01 6.96277797e-01 6.65911722e+00 8.55933487e-01
-8.49474669e-01 9.16153118e-02 6.58895433e-01 1.22908019e-01
-3.00710648e-01 -4.61462699e-03 -6.50818586e-01 3.26135695e-01
1.22611296e+00 -5.75442076e-01 2.66359653e-02 5.99469483e-01
9.69029311e-03 -5.70983030e-02 -1.05041587e+00 8.08316171e-01
-1.94674969e-01 -1.73630273e+00 1.51914313e-01 1.96110487e-01
6.41618192e-01 -9.37782824e-02 -3.95238966e-01 3.36035669e-01
4.79456037e-01 -1.25709343e+00 2.97721028e-01 8.52120638e-01
6.59254074e-01 -5.54025412e-01 1.10278487e+00 1.62034482e-01
-7.68580496e-01 6.58503100e-02 -3.34785134e-01 8.90220255e-02
-2.01773360e-01 1.08031738e+00 -1.24669468e+00 1.15948534e+00
7.59853601e-01 5.99549174e-01 -5.09473026e-01 1.04560649e+00
5.31782443e-03 6.11566484e-01 -1.53891519e-01 -2.27720857e-01
-3.48626375e-02 3.13006580e-01 2.09966049e-01 1.49608934e+00
4.15639579e-01 -5.45537211e-02 -9.98535752e-02 8.53604138e-01
-2.15207219e-01 2.71804482e-01 -5.64467847e-01 -2.68041670e-01
4.34482425e-01 1.28956664e+00 -9.21013236e-01 -5.39443076e-01
-1.83091164e-01 1.47218063e-01 1.27718285e-01 1.76249951e-01
-6.27086461e-01 -3.82293820e-01 1.01096161e-01 3.87815058e-01
1.89585477e-01 2.01567218e-01 -4.11741406e-01 -8.18690300e-01
-3.00630242e-01 -9.41201270e-01 4.76319373e-01 -8.23281348e-01
-1.46367764e+00 5.75727820e-01 2.95163810e-01 -6.31925762e-01
-2.56047994e-01 -7.38869309e-01 6.08340930e-03 9.84509051e-01
-1.37577260e+00 -1.28195333e+00 -1.22893363e-01 3.63052249e-01
-1.18705511e-01 -1.77000180e-01 1.10921049e+00 4.41169173e-01
-3.55183870e-01 3.02469283e-01 2.13666391e-02 2.10340321e-01
5.47112226e-01 -1.33273923e+00 -8.72310996e-02 2.59503752e-01
1.30318969e-01 9.75259006e-01 6.43063307e-01 -9.09537375e-01
-1.12822127e+00 -9.41228807e-01 1.30308163e+00 -4.96881723e-01
7.46987939e-01 1.81733593e-02 -1.09463346e+00 4.33733195e-01
8.98955092e-02 -2.44832665e-01 9.29414928e-01 4.08536226e-01
-3.53766292e-01 -1.45707920e-01 -1.24543524e+00 3.07442546e-01
9.46485400e-01 -3.19139481e-01 -8.62437427e-01 5.39052546e-01
3.96384865e-01 -3.15910220e-01 -1.31609356e+00 5.20703256e-01
7.17121780e-01 -6.33608758e-01 8.95709515e-01 -7.54051387e-01
5.06484985e-01 -9.10919458e-02 9.87067670e-02 -1.18916547e+00
-2.98932225e-01 -2.80683517e-01 -1.98340118e-01 1.08962715e+00
8.28009248e-01 -5.67332804e-01 5.85376263e-01 5.04178941e-01
2.52159797e-02 -1.08336663e+00 -7.90049136e-01 -7.18675017e-01
3.18018794e-01 -2.61677712e-01 6.32515550e-01 1.30986977e+00
2.35595763e-01 5.76110721e-01 1.47654697e-01 -3.27736109e-01
3.12672406e-01 -3.96638811e-02 4.37083632e-01 -1.74422991e+00
3.16409878e-02 -6.61825359e-01 -6.19557559e-01 1.85772460e-02
-1.46932140e-01 -1.16207588e+00 -4.95909005e-01 -2.36208534e+00
4.50681180e-01 -5.73921382e-01 -4.45727289e-01 9.28601682e-01
-1.32271554e-02 -5.12169953e-03 -2.90878177e-01 7.13662803e-02
-1.41796589e-01 -4.99303192e-02 1.10805237e+00 -1.77041709e-01
-4.11396809e-02 -3.16305429e-01 -1.00239432e+00 5.40458322e-01
8.34164083e-01 -6.72689795e-01 -4.66752768e-01 8.82466231e-03
7.39494324e-01 -8.30086172e-02 4.00231183e-01 -7.64370918e-01
3.65734220e-01 5.52936085e-02 7.14681864e-01 -5.65100014e-01
-4.11513150e-02 -5.69937468e-01 5.30097723e-01 5.94133496e-01
-3.14083964e-01 7.90497959e-02 3.73235375e-01 3.20630521e-01
-1.63887188e-01 -2.37835005e-01 3.60206246e-01 -3.25812787e-01
-2.64928669e-01 5.12928292e-02 -1.43567950e-01 8.11951011e-02
7.85677791e-01 -2.15348735e-01 -6.89410090e-01 -1.57352924e-01
-9.82029259e-01 5.98463118e-02 2.58533880e-02 2.10439846e-01
5.04794180e-01 -9.26449597e-01 -7.31283545e-01 -2.50809044e-01
-3.16647887e-02 -1.18419386e-01 1.13577656e-02 1.02099669e+00
-5.74143291e-01 7.38829792e-01 -3.48101497e-01 -2.59748906e-01
-1.05815077e+00 7.08648503e-01 2.27263883e-01 -5.95289290e-01
-3.19297731e-01 6.50245309e-01 -1.92166999e-01 -3.10161918e-01
6.22615814e-02 -3.01090747e-01 -5.61677933e-01 3.58625054e-01
8.02332535e-02 5.55512726e-01 4.22222227e-01 -4.50011671e-01
-6.56544864e-01 3.33439410e-01 -7.53961205e-02 1.10579796e-01
1.48898697e+00 4.79961634e-01 -4.12314326e-01 3.33303422e-01
9.20456886e-01 -4.43074517e-02 -6.32009581e-02 8.79513845e-02
2.58474648e-01 1.77237704e-01 1.43614456e-01 -1.32950497e+00
-9.67227936e-01 6.17975473e-01 1.96718886e-01 4.88426477e-01
9.16249216e-01 8.01801383e-02 4.56802517e-01 3.91819119e-01
2.47500136e-01 -8.75016987e-01 -2.86879182e-01 1.65140748e-01
7.32418716e-01 -8.37649763e-01 5.12483537e-01 -4.80892420e-01
-3.74036878e-01 1.17938089e+00 1.63676143e-01 3.12672049e-01
9.02644932e-01 4.29177344e-01 -1.20514616e-01 -6.64951324e-01
-6.76763415e-01 -1.54031843e-01 4.80764300e-01 7.42717147e-01
7.02907920e-01 6.47998974e-02 -5.02206445e-01 7.22956896e-01
-4.04541582e-01 5.92518568e-01 2.14637786e-01 8.66662264e-01
-1.98521301e-01 -1.15737224e+00 -2.02596292e-01 8.75655711e-01
-8.49492252e-01 -2.43236721e-01 -6.04805529e-01 9.95889485e-01
3.86003256e-01 9.02731240e-01 -2.49980822e-01 -2.36312181e-01
3.58895093e-01 3.75565827e-01 5.21219313e-01 -5.34033298e-01
-6.80488646e-01 -3.74237001e-02 5.07448375e-01 -2.28430182e-01
-6.98024750e-01 -3.28112215e-01 -1.30686307e+00 -2.39613235e-01
-4.26946104e-01 2.29008824e-01 7.14291930e-01 1.10797882e+00
5.32756925e-01 9.15834427e-01 -1.33717373e-01 -1.03238679e-01
-1.51731551e-01 -1.08132303e+00 -3.37519825e-01 1.14749648e-01
-1.05947733e-01 -8.07718635e-01 -1.63128097e-02 2.43416205e-01] | [8.487245559692383, 8.553481101989746] |
8e5978f6-e61a-4e64-8773-55decc631ce9 | watclaimcheck-a-new-dataset-for-claim | null | null | https://aclanthology.org/2022.acl-long.92 | https://aclanthology.org/2022.acl-long.92.pdf | WatClaimCheck: A new Dataset for Claim Entailment and Inference | We contribute a new dataset for the task of automated fact checking and an evaluation of state of the art algorithms. The dataset includes claims (from speeches, interviews, social media and news articles), review articles published by professional fact checkers and premise articles used by those professional fact checkers to support their review and verify the veracity of the claims. An important challenge in the use of premise articles is the identification of relevant passages that will help to infer the veracity of a claim. We show that transferring a dense passage retrieval model trained with review articles improves the retrieval quality of passages in premise articles. We report results for the prediction of claim veracity by inference from premise articles. | ['Pascal Poupart', 'Ruizhe Wang', 'Kashif Khan'] | null | null | null | null | acl-2022-5 | ['passage-retrieval'] | ['natural-language-processing'] | [-2.40730554e-01 5.04795253e-01 -6.20826364e-01 3.19321334e-01
-1.74719167e+00 -7.26943433e-01 8.06576431e-01 8.23116601e-01
-1.75405934e-01 1.12848735e+00 9.10754621e-01 -5.57061315e-01
-2.30950996e-01 -8.80377352e-01 -8.51420939e-01 1.57296047e-01
4.89608258e-01 6.22725964e-01 4.15918142e-01 -5.31232297e-01
1.08240235e+00 1.24377899e-01 -9.57561374e-01 9.78739917e-01
6.25617921e-01 1.12236011e+00 -4.70667452e-01 7.15716660e-01
-3.40711474e-01 2.09627604e+00 -1.06149936e+00 -1.34187436e+00
8.38987231e-02 -3.55544269e-01 -1.05386567e+00 -2.14538604e-01
5.68033695e-01 -3.45577717e-01 -4.75430042e-01 9.54683125e-01
2.95197129e-01 -3.72297972e-01 7.58824110e-01 -1.01425290e+00
-1.26217806e+00 1.17995393e+00 -4.96978760e-01 8.60248983e-01
8.20529699e-01 -2.16898978e-01 1.18784106e+00 -7.56221592e-01
1.22810459e+00 6.34669840e-01 1.09486294e+00 -1.25192478e-02
-6.05730653e-01 -1.97853342e-01 -6.29318595e-01 5.13720632e-01
-1.10121226e+00 -9.29502428e-01 7.63988554e-01 -7.88739741e-01
8.54997396e-01 1.80155560e-01 4.99735653e-01 1.25480962e+00
2.22823843e-01 7.74767637e-01 9.55724835e-01 -4.48015720e-01
1.54977068e-01 3.40366274e-01 5.02274990e-01 6.35918617e-01
3.52476209e-01 -5.58014691e-01 -7.36716926e-01 -9.54618156e-01
7.72557827e-03 -3.09949994e-01 -3.20152760e-01 6.09438419e-01
-7.63182819e-01 1.37795937e+00 1.50679117e-02 3.03730160e-01
-9.60591495e-01 -1.41215220e-01 9.00326550e-01 6.63959861e-01
9.85633910e-01 4.48203653e-01 -2.67255425e-01 -1.93812609e-01
-1.37913263e+00 5.96305013e-01 1.58018303e+00 5.03335297e-01
2.06696585e-01 -6.03688300e-01 -6.49468064e-01 8.71398449e-01
6.57429323e-02 7.34650195e-01 4.37070467e-02 -1.08318758e+00
9.72885549e-01 5.56714714e-01 6.00681305e-01 -1.72819591e+00
-6.41890764e-02 -5.80642343e-01 -2.65644908e-01 -4.02939856e-01
1.16029203e-01 -5.12203872e-02 -2.53709644e-01 7.23292410e-01
7.81709030e-02 1.06042549e-02 2.48710826e-01 4.78377521e-01
8.97698164e-01 6.27099872e-01 -2.44012237e-01 -5.94808578e-01
1.34278858e+00 -7.36609399e-01 -1.05668437e+00 1.78350538e-01
5.42065203e-01 -1.20571196e+00 4.22356039e-01 2.04197913e-01
-1.29724991e+00 -1.54184131e-02 -9.01531279e-01 -2.98198283e-01
-7.52968118e-02 2.92442709e-01 3.60013336e-01 2.36877963e-01
-6.16929293e-01 4.08187151e-01 -4.45186272e-02 1.13274669e-02
9.52888131e-01 -5.66717982e-01 -2.56227851e-01 -1.60819381e-01
-1.39366531e+00 9.43452001e-01 -1.03175044e-01 -1.55657843e-01
-5.62272549e-01 -8.40994418e-01 -5.24119198e-01 1.23985140e-02
8.06616724e-01 -5.08003950e-01 1.29753745e+00 -4.16143507e-01
-9.48787093e-01 1.29025185e+00 3.29746790e-02 -1.03359544e+00
7.83874929e-01 -1.32899910e-01 -8.16264331e-01 7.98470378e-01
9.05330956e-01 -6.18489683e-01 7.29546607e-01 -8.28901112e-01
-4.03696656e-01 -9.13549121e-03 -1.33081721e-02 -5.22408187e-01
9.73820314e-02 5.04117727e-01 -1.37727587e-02 -5.75282693e-01
-1.90469474e-01 -7.86487281e-01 4.53825325e-01 -5.85752487e-01
-1.06634557e+00 -4.85508859e-01 5.24330199e-01 -1.47338939e+00
1.44595993e+00 -1.94866550e+00 -8.43094945e-01 2.03539565e-01
4.98486757e-01 1.22496085e-02 -2.25540195e-02 1.11535299e+00
6.14248812e-01 5.88407218e-01 9.86525863e-02 8.35018903e-02
6.93042129e-02 -3.33262175e-01 -1.22328341e+00 5.66700339e-01
-2.76595280e-02 1.07033563e+00 -6.07154012e-01 -6.48627400e-01
-6.75342917e-01 1.94975380e-02 -2.46199086e-01 -2.40609869e-01
-3.87669563e-01 -1.24799855e-01 -6.95303500e-01 6.43895984e-01
3.89633238e-01 -8.38851154e-01 -1.98859051e-01 -9.50780064e-02
8.54898468e-02 1.18199372e+00 -3.47303689e-01 8.63601983e-01
-6.96784258e-02 9.94460285e-01 -1.64730951e-01 -5.91175079e-01
8.22773337e-01 6.09132290e-01 4.51668829e-01 -6.47337854e-01
1.16373241e-01 3.27461064e-01 -5.81768274e-01 -9.48440433e-01
7.82332718e-01 -1.71886384e-01 -1.19809337e-01 1.07253253e+00
-5.36070228e-01 -1.02908881e-02 4.35455829e-01 9.01323855e-01
1.47468090e+00 -6.72184408e-01 3.57464790e-01 2.97004789e-01
4.35648680e-01 8.55345488e-01 2.78717875e-01 1.19391084e+00
-7.72145856e-03 4.09730703e-01 8.85588288e-01 -5.78748763e-01
-1.43286908e+00 -4.52566087e-01 -7.02757686e-02 5.58371544e-01
-2.82930046e-01 -4.43259865e-01 -2.92274505e-01 -8.03225815e-01
3.11414003e-01 9.06477511e-01 -9.67318118e-01 2.15801746e-01
-3.04489076e-01 -1.35433331e-01 7.96854377e-01 5.19138388e-02
5.18657267e-01 -1.00038373e+00 -2.83471793e-01 4.25044864e-01
-9.11794603e-01 -1.48895156e+00 -8.15896541e-02 -5.80211699e-01
-2.08120227e-01 -1.72156620e+00 -3.64869386e-01 -2.46706620e-01
-3.35439518e-02 1.38475671e-01 1.26224351e+00 8.13201666e-01
2.97696441e-01 2.87076384e-01 -7.03936756e-01 -3.85610402e-01
-1.16336191e+00 7.42067099e-02 -3.84187877e-01 -1.81215003e-01
4.05770451e-01 -2.03950763e-01 -1.88864425e-01 -1.46627784e-01
-8.85101199e-01 -3.62298876e-01 2.29063928e-01 6.16353035e-01
3.66686881e-01 -3.44428346e-02 9.81418848e-01 -1.30377686e+00
1.31948066e+00 -1.00921595e+00 -2.37964928e-01 4.08756375e-01
-4.89436209e-01 -2.84855753e-01 1.92475900e-01 -1.91743553e-01
-7.49199092e-01 -1.10075700e+00 -2.62324333e-01 -1.42310709e-01
4.96352375e-01 1.19707203e+00 9.47414458e-01 2.80866623e-01
1.14058959e+00 -1.08676702e-01 4.37399857e-02 -3.78665060e-01
1.10426731e-01 8.92321408e-01 5.44731200e-01 -1.40994936e-01
5.42951286e-01 5.89183092e-01 -3.64444852e-01 -4.66178626e-01
-1.94953144e+00 -5.03172517e-01 -3.64831500e-02 -2.72405267e-01
3.39809448e-01 -8.87269080e-01 -6.61987543e-01 -6.51925206e-02
-1.52212119e+00 2.18023941e-01 -4.39089864e-01 2.36652732e-01
-3.74435157e-01 3.84351999e-01 -1.20584106e+00 -8.90743554e-01
-4.80467647e-01 -3.52216214e-01 7.21223652e-01 -3.83881956e-01
-3.33820254e-01 -6.86144054e-01 5.02593458e-01 1.23977077e+00
2.96960503e-01 3.17379475e-01 9.01194572e-01 -1.20430756e+00
-2.20567778e-01 -1.01845241e+00 -2.27991506e-01 1.33008510e-01
-4.05501783e-01 2.59859264e-01 -6.59617126e-01 3.94276023e-01
2.71143645e-01 -7.02560186e-01 1.03843009e+00 3.69616985e-01
4.35354233e-01 -1.40697396e+00 -2.38660052e-01 -4.25765485e-01
1.36917984e+00 -3.21443826e-01 9.45593655e-01 7.00130105e-01
8.85703266e-02 3.73014510e-01 2.12433755e-01 6.40715361e-01
4.62138891e-01 2.75779426e-01 -1.11354716e-01 7.27049470e-01
-1.27248704e-01 -5.12898982e-01 2.39768196e-02 9.20464516e-01
-1.28988614e-02 -5.00767171e-01 -7.88498342e-01 7.95256674e-01
-1.78504765e+00 -1.93681109e+00 -5.46286047e-01 1.50889909e+00
9.39720511e-01 2.95989186e-01 1.96405753e-01 3.31822455e-01
8.70484591e-01 1.03961229e-01 -2.70903781e-02 -2.77956724e-01
-4.58889663e-01 -2.79670209e-01 3.72673839e-01 5.97401917e-01
-7.70623267e-01 4.80123520e-01 7.06792641e+00 8.31351638e-01
-4.54652369e-01 6.48559093e-01 5.87803662e-01 -1.03948124e-01
-6.33329034e-01 -1.23707717e-02 -6.01096451e-01 6.73044682e-01
9.33039606e-01 -2.85562515e-01 1.46311224e-02 6.98039412e-01
1.65941611e-01 -3.86160314e-01 -5.58518589e-01 4.81453329e-01
6.38871908e-01 -2.53718448e+00 -1.60384715e-01 5.21575361e-02
1.26475453e+00 3.24597716e-01 -1.60282075e-01 4.69182342e-01
3.43078047e-01 -3.79244447e-01 1.17905021e+00 9.97198105e-01
3.64518732e-01 -1.81297913e-01 1.21134830e+00 4.09177810e-01
1.05683722e-01 -1.45338595e-01 -4.68933433e-01 -1.11727133e-01
5.15928626e-01 1.39117360e+00 -9.72294629e-01 3.06187540e-01
2.80677617e-01 6.01174116e-01 -6.06546640e-01 1.00543976e+00
-5.95865130e-01 9.67298746e-01 -9.72023159e-02 -2.37787455e-01
2.88311124e-01 2.82632560e-01 6.54250562e-01 1.01034486e+00
1.73719734e-01 2.35273704e-01 -9.19869840e-02 9.21781361e-01
-8.62296045e-01 2.23012388e-01 -5.95361054e-01 -5.45163989e-01
6.34157777e-01 7.58020401e-01 -5.06125569e-01 -8.12356114e-01
-2.10937560e-01 5.69393158e-01 5.17387390e-01 2.71838427e-01
-5.98228633e-01 6.47347271e-02 -3.01295608e-01 4.79780674e-01
7.28499889e-01 1.62942737e-01 -3.10652047e-01 -1.12443590e+00
2.79030323e-01 -1.01911628e+00 6.17122710e-01 -1.11795700e+00
-1.79041851e+00 6.61212385e-01 -4.85304564e-01 -8.99680078e-01
-2.90835530e-01 -9.77330208e-02 -5.00448585e-01 6.56184912e-01
-1.66118002e+00 -7.71807432e-01 2.40898356e-01 3.74872953e-01
8.81861299e-02 -4.29348856e-01 2.59918481e-01 6.08337335e-02
-9.98647138e-02 3.54985613e-03 1.44737540e-02 6.01415098e-01
5.04635155e-01 -5.70164442e-01 1.13067962e-01 6.49624586e-01
4.05699641e-01 5.02283275e-01 8.88540149e-01 -1.27348113e+00
-1.01369190e+00 -5.53919971e-01 1.52692604e+00 -1.10322499e+00
1.38680482e+00 4.24068272e-01 -9.33850944e-01 4.58634853e-01
2.17141524e-01 -2.50924259e-01 9.53899801e-01 3.67399305e-01
-8.11078191e-01 3.32304090e-01 -1.04759240e+00 7.82053471e-02
3.79187167e-01 -1.33162773e+00 -1.20469701e+00 1.03490305e+00
6.92901850e-01 -8.90704468e-02 -7.02467859e-01 -9.34505165e-02
3.55257899e-01 -7.46051729e-01 5.29933751e-01 -9.31445658e-01
1.17428279e+00 -3.42311151e-02 7.31022134e-02 -7.39679217e-01
-2.58003056e-01 -4.39668864e-01 -6.13924205e-01 1.08503759e+00
9.28880870e-01 -4.27100658e-01 4.59430844e-01 6.64888799e-01
1.34643167e-01 -4.23499614e-01 -9.54175115e-01 -3.74723107e-01
-3.77000272e-02 -5.94098866e-01 3.04507315e-01 1.14515615e+00
4.25446481e-01 4.58846599e-01 -4.17130291e-01 1.08178481e-02
3.23932260e-01 5.13472319e-01 5.64144194e-01 -1.29736567e+00
-1.91110238e-01 -3.43641251e-01 -2.95737922e-01 -4.82813656e-01
1.84426263e-01 -8.41903806e-01 -2.37189099e-01 -2.01316476e+00
5.63626826e-01 -2.76009053e-01 2.84713149e-01 2.50960410e-01
-9.90871191e-02 4.03930247e-01 -2.25123484e-02 9.72054005e-01
-8.56046915e-01 4.49857339e-02 1.04902017e+00 -4.69969720e-01
1.92653686e-01 1.37593001e-01 -1.10411131e+00 4.34356153e-01
6.78241849e-01 -9.11640167e-01 2.56263107e-01 -2.18536675e-01
1.54076803e+00 4.61767286e-01 7.43369758e-01 -7.20610857e-01
6.68534458e-01 -3.99371348e-02 -5.37047237e-02 -8.83930981e-01
-1.21540621e-01 -2.47356847e-01 -8.06961507e-02 9.64376181e-02
-8.33935440e-01 6.89839423e-02 -2.33433187e-01 9.18008327e-01
-3.86845231e-01 -6.67380512e-01 1.49350673e-01 -2.21008256e-01
-7.56072551e-02 -1.25948675e-02 -5.40789366e-01 5.80107212e-01
2.89255738e-01 1.35785580e-01 -1.21149981e+00 -9.56504107e-01
-4.67602909e-01 -2.79514104e-01 2.50647366e-01 -2.49904357e-02
5.14608562e-01 -1.11521482e+00 -1.55959845e+00 -5.12374282e-01
1.85048595e-01 -1.01766503e+00 2.06090033e-01 1.36184120e+00
-4.89697784e-01 3.93837005e-01 2.79240489e-01 2.99017102e-01
-6.64368689e-01 9.11762893e-01 -2.66845286e-01 -4.04386520e-01
-8.81561518e-01 6.32637501e-01 -1.06416726e+00 2.67419279e-01
-3.10144246e-01 3.10221817e-02 -3.82886529e-01 3.07797134e-01
1.18074226e+00 6.20715380e-01 3.03895652e-01 -6.24579489e-01
-6.57390729e-02 -1.98662966e-01 -8.20169225e-02 -3.54720682e-01
1.47553110e+00 -1.96550384e-01 -3.77401859e-01 4.72917080e-01
1.05633104e+00 9.66613650e-01 -5.53630829e-01 -4.49117720e-01
1.79860920e-01 -4.34690684e-01 2.03401968e-01 -1.23720276e+00
-7.63544023e-01 9.49229375e-02 -6.78482354e-01 1.15908432e+00
1.63625106e-01 4.28773522e-01 1.10844970e+00 7.07597554e-01
1.36363164e-01 -1.44865799e+00 -6.68552099e-03 6.01736248e-01
1.34970856e+00 -1.27853465e+00 2.98757941e-01 -2.41346836e-01
-8.52465749e-01 9.88610327e-01 -4.51012045e-01 -2.41791815e-01
6.78547621e-01 2.20735312e-01 1.30147502e-01 -1.16338694e+00
-8.34855199e-01 7.79569596e-02 4.98665780e-01 2.05885112e-01
1.50977194e-01 -2.00258330e-01 -6.12270117e-01 7.14863420e-01
-2.41657883e-01 2.20219806e-01 9.44536746e-01 5.91427982e-01
-5.98915815e-01 -1.90444425e-01 -4.09461647e-01 1.05241680e+00
-1.24514234e+00 -4.65126634e-01 -7.12088585e-01 4.10259187e-01
-2.61955738e-01 1.44440699e+00 -3.12409818e-01 -4.18596491e-02
8.24309066e-02 9.30333212e-02 1.21086039e-01 -4.53602970e-01
-8.43468547e-01 -5.30617595e-01 1.08082438e+00 -1.39793456e-01
-7.50558913e-01 -7.38720894e-01 -6.64780676e-01 -9.65307415e-01
-4.79640812e-01 9.07069325e-01 6.60814166e-01 1.40257311e+00
4.23599899e-01 -3.15069817e-02 4.18720067e-01 4.57125336e-01
-4.79020983e-01 -1.07857311e+00 -4.70409214e-01 5.94105244e-01
4.77130592e-01 -3.07715207e-01 -6.03180289e-01 2.03469619e-01] | [8.58165168762207, 9.858927726745605] |
992c0a15-f886-4376-b50b-ec62cabbadcb | synergy-with-translation-artifacts-for | 2210.09588 | null | https://arxiv.org/abs/2210.09588v1 | https://arxiv.org/pdf/2210.09588v1.pdf | Synergy with Translation Artifacts for Training and Inference in Multilingual Tasks | Translation has played a crucial role in improving the performance on multilingual tasks: (1) to generate the target language data from the source language data for training and (2) to generate the source language data from the target language data for inference. However, prior works have not considered the use of both translations simultaneously. This paper shows that combining them can synergize the results on various multilingual sentence classification tasks. We empirically find that translation artifacts stylized by translators are the main factor of the performance gain. Based on this analysis, we adopt two training methods, SupCon and MixUp, considering translation artifacts. Furthermore, we propose a cross-lingual fine-tuning algorithm called MUSC, which uses SupCon and MixUp jointly and improves the performance. Our code is available at https://github.com/jongwooko/MUSC. | ['Se-Young Yun', 'Jongwoo Ko', 'Jaehoon Oh'] | 2022-10-18 | null | null | null | null | ['sentence-classification'] | ['natural-language-processing'] | [-2.68715054e-01 -2.82045513e-01 -6.02357268e-01 -3.73250186e-01
-1.23848104e+00 -6.93392098e-01 7.55791306e-01 -2.75528640e-01
-4.97981995e-01 1.12334967e+00 3.86848658e-01 -7.79438376e-01
2.99227148e-01 -4.31540042e-01 -8.27573776e-01 -4.62395877e-01
5.77231348e-01 3.53552610e-01 -2.54956245e-01 -2.87392557e-01
-6.93809688e-02 2.68113595e-02 -9.08023655e-01 3.20251673e-01
1.52656555e+00 2.84394562e-01 3.45054835e-01 3.02855909e-01
-2.04893768e-01 5.93312323e-01 -6.07032597e-01 -6.26212358e-01
3.80293280e-01 -7.62264669e-01 -7.61396766e-01 -7.14229867e-02
3.00504565e-01 -7.34324753e-02 5.03927916e-02 1.01516426e+00
6.31304383e-01 -1.48985326e-01 5.19304335e-01 -9.99640703e-01
-6.74380898e-01 1.34345245e+00 -6.12521887e-01 2.05242828e-01
2.28526294e-01 3.44453491e-02 9.59569931e-01 -9.94786799e-01
7.74013698e-01 1.21389723e+00 5.01129746e-01 3.98553818e-01
-1.20697546e+00 -8.43929827e-01 8.26224014e-02 -1.15948368e-03
-1.30720115e+00 -6.46529794e-01 7.91704118e-01 -3.61005574e-01
9.39552546e-01 1.07604871e-02 3.00692111e-01 1.41049862e+00
4.15490419e-01 9.40616786e-01 1.44104552e+00 -8.70305300e-01
-2.24006847e-01 5.02129495e-01 -3.99407893e-02 5.32550037e-01
2.33256757e-01 -8.85545462e-02 -4.95283633e-01 -6.81090653e-02
4.10248756e-01 -3.61725897e-01 -2.78458476e-01 1.60520926e-01
-1.21309853e+00 9.62637365e-01 6.49583340e-02 5.99568367e-01
-2.33243614e-01 -3.32940184e-02 4.08727676e-01 6.48792684e-01
7.58678794e-01 6.16493285e-01 -5.87732077e-01 -1.21566989e-01
-9.58435774e-01 -4.14446555e-02 6.24931753e-01 9.46180940e-01
8.03449988e-01 2.46172622e-02 -1.03541918e-01 1.11966336e+00
1.25107825e-01 8.26646447e-01 8.00946593e-01 -6.14754319e-01
1.01106489e+00 6.56767130e-01 -8.91981050e-02 -4.05759126e-01
-1.07006423e-01 -5.54396391e-01 -4.23497468e-01 -3.67076874e-01
4.81373429e-01 -5.22924125e-01 -6.88700020e-01 1.82954025e+00
1.52777001e-01 -2.80375212e-01 2.13073432e-01 7.26534426e-01
5.13823569e-01 5.72806716e-01 -4.31257812e-03 -1.30270094e-01
1.41829181e+00 -1.21601725e+00 -8.14706802e-01 -3.74984920e-01
1.13552082e+00 -1.29669917e+00 1.56403244e+00 2.10898265e-01
-8.06787431e-01 -4.87896085e-01 -1.02479827e+00 -1.97519436e-01
-2.99660951e-01 7.96463728e-01 6.07036054e-01 5.86887181e-01
-7.61673510e-01 3.87260884e-01 -7.97628224e-01 -3.82562429e-01
-2.30751932e-03 7.00786859e-02 -3.76669616e-01 -1.05371572e-01
-1.44017529e+00 1.03156137e+00 3.02392185e-01 -7.02296272e-02
-4.74166363e-01 -4.35091615e-01 -7.64733493e-01 -4.52412874e-01
2.48978913e-01 -6.94319129e-01 1.27189624e+00 -1.26775706e+00
-1.62829709e+00 7.67298281e-01 -4.32803363e-01 -2.99131423e-01
8.16656888e-01 -5.74980795e-01 -3.05134505e-01 -2.69935042e-01
4.22191620e-01 4.56824422e-01 6.18957281e-01 -1.15056729e+00
-6.85433149e-01 -1.64069876e-01 -1.26809299e-01 3.29442620e-01
-4.11281228e-01 2.52331227e-01 -7.77568638e-01 -9.49225187e-01
-1.88821256e-01 -1.09096074e+00 1.44711593e-02 -8.95130754e-01
-5.21345854e-01 -2.50194758e-01 3.03448796e-01 -1.06830204e+00
1.37632191e+00 -1.91117334e+00 9.24702957e-02 2.67723072e-02
-2.34011903e-01 2.75956839e-01 -2.28974998e-01 5.60917318e-01
2.15439964e-02 3.16832870e-01 -5.47757037e-02 -5.97600698e-01
-1.47554338e-01 1.38874725e-01 -3.71839345e-01 1.52548850e-01
3.17376256e-02 1.02499914e+00 -8.37890089e-01 -4.80767697e-01
-1.09261945e-02 3.05294096e-01 -2.23867461e-01 -5.01320623e-02
-2.37252358e-02 5.86154699e-01 -3.29791576e-01 3.98592263e-01
5.82476854e-01 -8.20785388e-02 4.54871535e-01 1.19635705e-02
-1.32158369e-01 8.40769947e-01 -8.56488824e-01 1.79672766e+00
-8.63215208e-01 5.54705441e-01 -2.27322102e-01 -3.71052474e-01
8.78289998e-01 4.91397649e-01 1.59781143e-01 -8.29612553e-01
3.04139346e-01 6.66865051e-01 1.64428890e-01 -4.30743486e-01
5.54263294e-01 -2.08027408e-01 -1.33883208e-01 8.55018079e-01
-5.35595194e-02 -5.49115278e-02 5.47022104e-01 1.82779357e-01
6.30906045e-01 3.25302571e-01 3.68472010e-01 -4.06816125e-01
4.76675540e-01 1.43431127e-01 7.97885358e-01 4.71764654e-01
1.03877999e-01 3.02964002e-01 3.54157150e-01 -9.92758051e-02
-1.06418884e+00 -1.00303686e+00 -1.06157228e-01 1.14692104e+00
-1.91378459e-01 -5.06733835e-01 -9.40515757e-01 -8.66030931e-01
-1.36687279e-01 9.37874079e-01 -3.06028724e-01 4.61937068e-03
-7.14784086e-01 -8.75870764e-01 7.16507971e-01 3.87591004e-01
3.81197095e-01 -8.36492956e-01 -1.03747956e-01 1.37296453e-01
-8.47550452e-01 -1.12844920e+00 -8.73593748e-01 1.93222150e-01
-8.15482676e-01 -7.85751939e-01 -4.45094585e-01 -7.86905825e-01
6.66767299e-01 2.85932511e-01 1.18614590e+00 9.78462920e-02
2.59687394e-01 -1.76708370e-01 -4.51913327e-01 -3.73903751e-01
-8.76937687e-01 7.26742148e-01 2.50214368e-01 -1.25307038e-01
5.01648188e-01 -4.16888744e-01 -4.35343422e-02 4.40355778e-01
-7.16237962e-01 4.13635910e-01 6.74449563e-01 7.26575851e-01
4.41199809e-01 -7.00680092e-02 6.24867022e-01 -1.23215115e+00
9.84668374e-01 -3.35154086e-01 -5.06915331e-01 3.33078265e-01
-7.43915975e-01 2.70716965e-01 8.70966971e-01 -5.24402976e-01
-9.99023974e-01 -1.05322510e-01 -1.79122031e-01 -1.81236923e-01
-1.95378195e-02 6.98183060e-01 -2.12402299e-01 2.82963246e-01
6.21025145e-01 2.09159940e-01 -2.00860247e-01 -6.89698458e-01
2.93978870e-01 1.00792587e+00 6.09236322e-02 -8.74730051e-01
8.20970476e-01 7.68670291e-02 -5.61179101e-01 -5.41858077e-01
-6.87695563e-01 -1.97325423e-01 -8.45696807e-01 9.91487280e-02
5.50617814e-01 -1.23742175e+00 2.16407537e-01 3.24447632e-01
-1.11798096e+00 -5.24868727e-01 1.53332338e-01 7.82545626e-01
-2.10258469e-01 2.49983922e-01 -7.80251265e-01 -4.53838736e-01
-4.46494967e-01 -1.45000434e+00 8.23120058e-01 -9.62472484e-02
-3.58847201e-01 -1.22580349e+00 1.95449606e-01 6.46607816e-01
1.72692969e-01 -1.27682522e-01 9.45665359e-01 -6.76594496e-01
-2.95165509e-01 -1.66579150e-02 3.23833078e-02 4.82025117e-01
4.57176775e-01 1.13380261e-01 -8.59468639e-01 -3.25717926e-01
-5.19333966e-02 -1.65863276e-01 5.32262087e-01 3.13673615e-01
3.57946604e-01 -3.26816767e-01 -1.21730044e-01 5.66073060e-01
1.23881292e+00 1.03123831e-02 4.13031131e-01 4.10372913e-01
7.48691618e-01 5.96262217e-01 6.09789133e-01 2.14146916e-02
6.65453970e-01 8.98598075e-01 -3.32151949e-01 -1.61329687e-01
-3.28237295e-01 -4.84559327e-01 8.26576412e-01 1.42047870e+00
-7.50293210e-02 -1.46466345e-01 -1.09696114e+00 5.59487879e-01
-1.88230896e+00 -6.25079393e-01 -2.32838884e-01 2.11851382e+00
1.20489359e+00 1.43553272e-01 7.68197849e-02 -1.51089981e-01
6.60346746e-01 -1.84431225e-01 -2.27480546e-01 -3.34838599e-01
-2.34548658e-01 1.14154689e-01 5.83375990e-01 7.93409288e-01
-8.81338477e-01 1.48080683e+00 5.74071503e+00 1.04731894e+00
-1.49022400e+00 4.36241835e-01 4.62585092e-01 -1.05688594e-01
-6.05390429e-01 1.11503728e-01 -9.56309736e-01 5.57876289e-01
9.43291903e-01 -3.79664421e-01 4.67126667e-01 5.92226803e-01
6.16795599e-01 -4.52614166e-02 -1.05881989e+00 6.74329400e-01
7.74539709e-02 -8.48681688e-01 2.90142953e-01 1.15968592e-01
9.57309186e-01 2.64198512e-01 -1.12727903e-01 3.56480241e-01
4.72183585e-01 -7.22510636e-01 9.43146110e-01 5.32081500e-02
6.92808568e-01 -6.91262305e-01 9.10146296e-01 4.62139726e-01
-1.00487030e+00 3.01251650e-01 -1.76789403e-01 1.07690236e-02
1.55854836e-01 6.83672011e-01 -9.87877488e-01 9.72738206e-01
3.27140957e-01 7.03064680e-01 -8.05434227e-01 4.96840030e-01
-9.11317229e-01 9.91963983e-01 -1.04459979e-01 2.23185062e-01
1.42127141e-01 -3.91511470e-01 4.62011516e-01 1.22680366e+00
3.80914956e-01 -7.18600333e-01 2.02361330e-01 7.48171389e-01
-2.65451789e-01 5.21888196e-01 -5.64760566e-01 -1.59731165e-01
6.15649283e-01 1.05162239e+00 -5.25458694e-01 -2.25061163e-01
-4.52456981e-01 9.70377028e-01 4.66930419e-01 4.47200269e-01
-8.92441690e-01 -3.51618379e-01 5.82300007e-01 9.45337489e-02
3.55233885e-02 -3.53619933e-01 -5.40485561e-01 -1.52708173e+00
2.93550104e-01 -1.33053672e+00 2.42668405e-01 -4.09751326e-01
-1.08082283e+00 8.56206894e-01 -1.28370956e-01 -1.33918393e+00
-4.08111215e-01 -3.80329311e-01 -4.73895520e-01 1.28774047e+00
-1.58051419e+00 -1.29521000e+00 2.02296555e-01 4.63278383e-01
6.95371091e-01 -2.44438604e-01 6.96536601e-01 5.95143855e-01
-9.05821562e-01 8.73904228e-01 1.89516827e-01 3.42745513e-01
1.14288211e+00 -1.11692822e+00 5.34070373e-01 1.34380567e+00
3.29590499e-01 9.69191313e-01 5.68465889e-01 -7.16183245e-01
-1.16527915e+00 -1.03975880e+00 1.56550670e+00 -6.65842593e-01
7.14044988e-01 -4.82969552e-01 -7.35378981e-01 8.01563799e-01
5.17129719e-01 -6.91921473e-01 7.89391339e-01 3.64790559e-01
-4.57802325e-01 -5.96982017e-02 -6.96745813e-01 8.67235541e-01
8.24611247e-01 -6.90744698e-01 -5.48061907e-01 2.06183732e-01
7.29960859e-01 -3.38015646e-01 -7.94492185e-01 1.36587232e-01
4.17082787e-01 -6.57675087e-01 4.95761424e-01 -4.27874118e-01
7.33805060e-01 -2.86944121e-01 -9.36819389e-02 -1.85407460e+00
-1.40013322e-01 -5.79608738e-01 3.57202828e-01 1.37609828e+00
1.04415524e+00 -9.41828012e-01 2.95627803e-01 2.60506928e-01
-1.41007602e-01 -7.19027460e-01 -7.20900178e-01 -1.00283861e+00
5.35785437e-01 -3.23473245e-01 6.65469527e-01 1.08808315e+00
-4.11185250e-02 7.94230402e-01 -4.89317387e-01 -9.16060954e-02
3.62823457e-01 2.31083766e-01 1.01362455e+00 -7.04656959e-01
-4.06007856e-01 -5.04005671e-01 2.96900958e-01 -1.02853918e+00
2.63944417e-01 -1.26410246e+00 1.17175067e-02 -1.62130082e+00
1.84130520e-01 -5.40657938e-01 -9.15486068e-02 6.79098725e-01
-6.39790595e-01 2.05894426e-01 3.02429229e-01 5.17983258e-01
-1.02222875e-01 3.63034397e-01 1.33343267e+00 1.23501465e-01
-3.97825748e-01 3.10391723e-03 -1.03832448e+00 4.32645589e-01
1.08427215e+00 -6.78942740e-01 -3.65280539e-01 -1.01191175e+00
1.99505195e-01 -2.36082986e-01 -3.29717517e-01 -6.71540260e-01
-2.19042644e-01 -2.38794625e-01 -3.64956260e-03 -1.42740786e-01
1.26944399e-02 -5.56027174e-01 1.48387760e-01 4.82088953e-01
-2.63823956e-01 3.76751214e-01 2.53177881e-01 -8.80862027e-02
-2.75920570e-01 -2.34267667e-01 6.47731125e-01 -2.58721989e-02
-3.27288628e-01 -8.88397396e-02 -3.32959741e-01 2.83409089e-01
6.05159819e-01 1.29395112e-01 -3.98216456e-01 -2.54775941e-01
-3.27988744e-01 2.51737207e-01 5.26105821e-01 7.10079193e-01
-5.96392527e-02 -1.44996428e+00 -1.06500614e+00 2.32301012e-01
7.08245710e-02 -4.52659309e-01 -2.56863385e-01 1.12081873e+00
-4.92623031e-01 6.30804837e-01 -2.31736619e-03 -3.83993328e-01
-1.30651951e+00 6.64542690e-02 2.31090263e-01 -5.62792182e-01
-2.56461114e-01 6.53210998e-01 -1.31188817e-02 -8.46786857e-01
-6.66986406e-02 -3.80065322e-01 1.37911499e-01 6.03960864e-02
3.28024745e-01 1.54803649e-01 2.68068075e-01 -6.96037769e-01
-2.31474444e-01 5.09027064e-01 -2.74990648e-01 -3.73847306e-01
1.03394175e+00 -2.65589833e-01 -2.12811217e-01 6.53050840e-01
1.01914525e+00 6.42916858e-01 -7.50280023e-01 -4.30159777e-01
1.42267182e-01 -3.98034096e-01 -1.12270862e-01 -9.38848853e-01
-8.35674644e-01 7.15932667e-01 2.70598177e-02 -2.41373539e-01
1.03487122e+00 -6.26099929e-02 8.82702768e-01 2.02590287e-01
5.66143513e-01 -1.20946729e+00 -2.31554732e-01 7.81861246e-01
8.25391769e-01 -1.38151884e+00 -5.28561510e-02 -5.51662087e-01
-9.23946440e-01 7.90332854e-01 4.78302985e-01 1.01486057e-01
3.59991699e-01 3.29798490e-01 7.16489255e-01 3.62355024e-01
-7.75944471e-01 -1.51549438e-02 3.75051856e-01 1.11369543e-01
9.84260738e-01 5.17034531e-01 -7.46118963e-01 6.99455142e-01
-6.99244320e-01 -5.33594750e-02 3.20950836e-01 7.26817548e-01
-8.22928250e-02 -1.78206432e+00 -4.41254705e-01 3.08321923e-01
-4.72243041e-01 -4.62592959e-01 -6.51756644e-01 7.76757300e-01
1.96744397e-01 1.04459643e+00 -2.98036546e-01 -5.60773790e-01
1.47902429e-01 4.30147946e-01 5.03330350e-01 -6.82664931e-01
-8.36087644e-01 2.82479048e-01 3.40050310e-01 -1.85223579e-01
-1.76083431e-01 -6.75685525e-01 -1.00820315e+00 -4.31266278e-01
-3.43622208e-01 4.59412962e-01 7.10016310e-01 1.06562388e+00
4.38542932e-01 4.32929486e-01 6.91993415e-01 -3.36832076e-01
-6.47747755e-01 -1.43851960e+00 -2.14420229e-01 4.50914085e-01
-6.92660511e-02 -4.46437150e-01 -3.75912845e-01 2.34911427e-01] | [11.428635597229004, 10.242596626281738] |
cb5716ec-445a-467f-9a06-36f6e49f21af | geolocation-estimation-of-photos-using-a | null | null | http://openaccess.thecvf.com/content_ECCV_2018/html/Eric_Muller-Budack_Geolocation_Estimation_of_ECCV_2018_paper.html | http://openaccess.thecvf.com/content_ECCV_2018/papers/Eric_Muller-Budack_Geolocation_Estimation_of_ECCV_2018_paper.pdf | Geolocation Estimation of Photos using a Hierarchical Model and Scene Classification | While the successful estimation of a photo's geolocation enables a number of interesting applications, it is also a very challenging task. Due to the complexity of the problem, most existing approaches are restricted to specific areas, imagery, or worldwide landmarks. Only a few proposals predict GPS coordinates without any limitations. In this paper, we introduce several deep learning methods, which pursue the latter approach and treat geolocalization as a classification problem where the earth is subdivided into geographical cells. We propose to exploit hierarchical knowledge of multiple partitionings and additionally extract and take the photo's scene content into account, i.e., indoor, natural, or urban setting etc. As a result, contextual information at different spatial resolutions as well as more specific features for various environmental settings are incorporated in the learning process of the convolutional neural network. Experimental results on two benchmarks demonstrate the effectiveness of our approach outperforming the state of the art while using a significant lower number of training images and without relying on retrieval methods that require an appropriate reference dataset. | ['Kader Pustu-Iren', 'Eric Muller-Budack', 'Ralph Ewerth'] | 2018-09-01 | null | null | null | eccv-2018-9 | ['photo-geolocation-estimation'] | ['computer-vision'] | [ 7.79559789e-03 -2.02775896e-01 -1.38961032e-01 -4.62483048e-01
-7.16106057e-01 -6.15999699e-01 8.84029508e-01 3.77673179e-01
-5.81926763e-01 7.67751098e-01 1.91113263e-01 5.70903346e-02
-2.23063350e-01 -1.03809655e+00 -8.18106174e-01 -7.68286526e-01
-2.59354915e-02 2.95313239e-01 9.26129520e-02 9.75653082e-02
4.10296738e-01 5.98510325e-01 -1.54430866e+00 -2.38954902e-01
9.27959859e-01 1.25544322e+00 2.29597539e-01 3.02211255e-01
1.80352442e-02 5.04923463e-01 -4.16980416e-01 -5.54136820e-02
3.61286879e-01 3.71891037e-02 -7.29793370e-01 7.84920231e-02
8.02350342e-01 -1.89780802e-01 -2.46815249e-01 1.02909112e+00
5.72627068e-01 3.43396246e-01 5.69685936e-01 -9.48201180e-01
-6.30685687e-01 2.97894150e-01 -5.59504032e-01 6.21697679e-02
2.07553774e-01 -8.64626020e-02 7.66995609e-01 -8.12037766e-01
3.92823607e-01 6.53156102e-01 9.02476192e-01 -3.13496649e-01
-9.73653972e-01 -4.49665397e-01 2.75791645e-01 2.50007421e-01
-1.97642922e+00 -5.73910415e-01 6.79116428e-01 -3.43607336e-01
7.50195086e-01 1.55766860e-01 3.69803041e-01 5.12292087e-01
4.81215455e-02 5.43378830e-01 1.01967919e+00 -3.00721884e-01
3.72560918e-01 1.62069406e-02 -1.50163591e-01 5.29388011e-01
9.37847942e-02 -5.09669721e-01 -1.72507867e-01 3.94099392e-02
6.27234876e-01 1.33132711e-01 -4.71404076e-01 -5.66932261e-01
-1.26225603e+00 4.73337263e-01 9.44377840e-01 5.05626619e-01
-6.83755517e-01 2.00740069e-01 8.21512714e-02 -1.84833378e-01
6.40446961e-01 2.96171576e-01 -2.94774264e-01 2.66568899e-01
-1.34107125e+00 -2.32003443e-03 6.26717806e-01 9.24129963e-01
1.32540727e+00 -1.20833084e-01 1.31408006e-01 6.80114508e-01
1.04595885e-01 4.75215286e-01 4.95173842e-01 -5.43805420e-01
5.38728595e-01 7.08187759e-01 3.86192530e-01 -1.43139172e+00
-6.13411188e-01 -6.21522307e-01 -1.16747570e+00 -2.74748325e-01
2.51826167e-01 -9.87302065e-02 -1.03622663e+00 1.49198663e+00
5.08252144e-01 7.26065695e-01 -1.32343382e-01 8.00298452e-01
8.45216632e-01 6.29772425e-01 1.60998821e-01 3.05997849e-01
1.15642655e+00 -9.56438243e-01 -2.39347741e-01 -5.08287728e-01
6.97954595e-01 -4.86762851e-01 5.84769070e-01 4.54405732e-02
-6.62364483e-01 -6.48756027e-01 -9.73918438e-01 -1.94106236e-01
-1.01979196e+00 7.60967195e-01 7.57095039e-01 5.86673975e-01
-1.58777869e+00 3.86228651e-01 -6.04030430e-01 -8.02881896e-01
2.30930924e-01 5.64451694e-01 -6.49329901e-01 -1.52634993e-01
-1.22788084e+00 6.46844089e-01 4.95561332e-01 3.26813847e-01
-2.93144554e-01 -4.30915713e-01 -1.04254425e+00 3.67635936e-01
5.11283539e-02 -8.33216906e-01 5.61650395e-01 -9.60369945e-01
-1.11476457e+00 8.01555455e-01 -1.12919528e-02 -5.16699553e-01
6.35294080e-01 5.23687601e-02 -2.43183583e-01 -1.16322341e-03
4.20221388e-01 7.48649716e-01 5.22945940e-01 -1.00377858e+00
-9.90399241e-01 -4.66006607e-01 4.33109254e-01 4.86371785e-01
-2.53288567e-01 -2.71851569e-01 -8.67067516e-01 -3.54178309e-01
3.67124140e-01 -1.04706168e+00 -3.38126242e-01 -2.02392191e-01
-4.75511730e-01 -1.87698588e-01 3.49564105e-01 -5.98479033e-01
1.20141768e+00 -2.19354892e+00 -9.56982076e-02 2.56385475e-01
-3.59580964e-02 4.47731800e-02 -7.92847350e-02 6.69595957e-01
7.24920928e-02 2.94711173e-01 -1.00218780e-01 -4.40417975e-01
1.66988894e-01 -1.64399251e-01 -2.08238274e-01 8.94721031e-01
7.43554607e-02 6.92478836e-01 -8.48674417e-01 -4.23264831e-01
6.13105178e-01 6.75585568e-01 -3.34531993e-01 -2.72999734e-01
1.15304209e-01 6.15818560e-01 -6.30780935e-01 6.41451061e-01
1.00422847e+00 -3.15118104e-01 1.25225291e-01 -2.00699925e-01
-4.56769228e-01 2.21543893e-01 -1.42840350e+00 1.72977376e+00
-6.70043647e-01 6.27233863e-01 -1.91729814e-01 -1.07433581e+00
7.20849693e-01 3.26183364e-02 5.85625708e-01 -7.36486495e-01
-7.61742443e-02 2.48743668e-01 -5.36572337e-01 -2.98646063e-01
9.67625439e-01 3.47446769e-01 -1.87321380e-01 8.68546590e-02
-1.75696850e-01 1.97061285e-01 -8.60027000e-02 -2.33228430e-01
6.34020627e-01 2.35470742e-01 6.35453284e-01 -2.97485054e-01
6.73008323e-01 1.86961237e-02 5.52271485e-01 9.40032840e-01
-1.11356825e-01 7.96929061e-01 -4.42660004e-02 -6.82748139e-01
-7.71067202e-01 -4.83692944e-01 -2.45445684e-01 9.55595136e-01
6.21889949e-01 -1.39998227e-01 -6.90179765e-01 -3.76270175e-01
-1.35654267e-02 3.75279337e-01 -7.94878304e-01 2.35866919e-01
-5.11935055e-01 -1.02051187e+00 5.04523277e-01 4.96876925e-01
1.02826679e+00 -6.51446164e-01 -4.57465380e-01 4.51043509e-02
-3.99942011e-01 -1.18322253e+00 -8.95541534e-02 -2.17976123e-01
-5.06355643e-01 -1.01865804e+00 -8.09988737e-01 -7.82828569e-01
5.60185611e-01 6.94600880e-01 1.07851136e+00 1.38397911e-03
2.94545963e-02 3.76266122e-01 -2.38938928e-01 -1.26754135e-01
5.06545484e-01 6.74874544e-01 -1.07053757e-01 4.16068733e-01
4.25980389e-01 -6.22232974e-01 -8.88869941e-01 3.22854519e-01
-7.22631335e-01 -6.63243830e-02 7.58445501e-01 6.96985543e-01
6.62521362e-01 3.89979243e-01 3.45803499e-01 -7.38884270e-01
3.93678807e-02 -8.95328104e-01 -6.57004952e-01 4.37812239e-01
-2.33198345e-01 -1.98440149e-01 3.57720315e-01 -2.14903001e-02
-9.24536943e-01 5.12358129e-01 3.18132043e-02 5.18559031e-02
-5.49623072e-01 8.76366913e-01 -3.70052397e-01 -3.81638467e-01
4.27772939e-01 5.00353217e-01 -8.34729970e-01 -3.64089519e-01
3.54378700e-01 6.98791623e-01 4.25846696e-01 -2.49543160e-01
8.54379058e-01 6.96491063e-01 7.41735613e-03 -9.97039855e-01
-5.82715571e-01 -9.07010376e-01 -9.41102564e-01 -6.62047043e-02
7.24268615e-01 -1.32058668e+00 -4.60603774e-01 3.63913000e-01
-8.65867376e-01 -1.44447610e-01 2.44349048e-01 5.44323504e-01
-3.68433386e-01 4.47209775e-01 -5.15992716e-02 -7.61972249e-01
-1.43099412e-01 -9.55289066e-01 1.44416296e+00 4.83899891e-01
2.77037829e-01 -1.09673572e+00 -1.40603594e-02 2.15055466e-01
4.43015754e-01 4.50921208e-01 5.96938670e-01 -5.28977692e-01
-1.09229493e+00 -5.94496250e-01 -5.72870255e-01 -2.05747887e-01
2.39043295e-01 -1.95652634e-01 -1.22368336e+00 -2.18951583e-01
-2.83728331e-01 -7.30676055e-02 1.00806713e+00 5.24436176e-01
1.06213796e+00 -2.82704264e-01 -6.41185701e-01 9.46456015e-01
1.90713513e+00 -1.51276946e-01 7.31085837e-01 5.83737612e-01
8.26505005e-01 6.21725857e-01 7.96536505e-01 4.92046952e-01
8.36383760e-01 6.62855625e-01 7.60954022e-01 -2.82284409e-01
2.12616324e-01 -5.21786846e-02 -2.07874849e-01 2.78996140e-01
-3.10362756e-01 -4.85348821e-01 -1.16490066e+00 9.34063196e-01
-2.00487399e+00 -8.85367513e-01 -9.92077291e-02 2.32092690e+00
2.88382530e-01 -3.16148549e-01 -2.86155999e-01 -9.25705731e-02
7.27073312e-01 5.58215439e-01 -4.41687912e-01 3.20731223e-01
-2.72986710e-01 -9.51161981e-02 1.14851546e+00 3.86456311e-01
-1.77873743e+00 1.06883848e+00 5.38569355e+00 7.08558023e-01
-1.42738950e+00 -1.97295606e-01 7.04285264e-01 3.45108241e-01
-3.06167305e-02 -9.36556160e-02 -7.89305031e-01 4.49140906e-01
5.69801450e-01 2.16015637e-01 2.97703683e-01 8.67564023e-01
9.96058658e-02 -4.43950236e-01 -7.88271129e-01 1.10479259e+00
8.98952484e-02 -1.27983701e+00 -3.12261999e-01 1.51543066e-01
9.42481339e-01 4.94016618e-01 2.16601923e-01 2.46955469e-01
3.05308606e-02 -9.35554266e-01 7.19927788e-01 7.93479323e-01
7.41262555e-01 -6.16500616e-01 9.33541596e-01 4.36216861e-01
-1.52234542e+00 -5.52305952e-02 -5.63529313e-01 -1.07369818e-01
-2.10492998e-01 4.04410422e-01 -8.44128966e-01 1.05200386e+00
7.90921271e-01 8.08208287e-01 -9.30660725e-01 1.51472580e+00
-9.47968215e-02 3.10697228e-01 -5.32902420e-01 1.58685163e-01
6.32913411e-01 -1.61677226e-01 -6.95594877e-04 1.24577689e+00
7.35356688e-01 -1.22768812e-01 4.22004908e-01 3.82121742e-01
-1.17977090e-01 3.53769779e-01 -7.88685620e-01 2.53470331e-01
6.03339136e-01 1.36591554e+00 -8.00119996e-01 -2.14678019e-01
-3.98513317e-01 9.02555466e-01 3.50276917e-01 6.13445818e-01
-7.64885604e-01 -4.13813740e-01 4.30271029e-01 1.73924729e-01
2.95487434e-01 -3.73608112e-01 -2.04514325e-01 -1.26674438e+00
-6.09231927e-03 -2.87014872e-01 1.92459136e-01 -8.22821617e-01
-9.48325574e-01 4.80245113e-01 -2.11183101e-01 -1.41289222e+00
-2.64788359e-01 -3.49174500e-01 -4.32315558e-01 1.06054008e+00
-2.04679704e+00 -1.63079154e+00 -7.78290987e-01 4.37839806e-01
3.61242265e-01 1.24125756e-01 7.60121524e-01 5.52911043e-01
-5.08027017e-01 2.44463369e-01 6.97721601e-01 2.12398455e-01
8.71264756e-01 -1.32414353e+00 4.45718408e-01 8.73495877e-01
1.02138169e-01 5.76798379e-01 3.37985426e-01 -3.13957244e-01
-1.04795206e+00 -1.35374558e+00 1.23703814e+00 -2.08354205e-01
5.47830999e-01 -1.73767537e-01 -5.68299174e-01 7.88472474e-01
2.12293342e-01 2.23970324e-01 5.64686954e-01 1.33462027e-01
-5.57043217e-02 -3.70906115e-01 -1.06529891e+00 4.60121751e-01
8.68900955e-01 -5.60998619e-01 -9.80872214e-02 4.08749342e-01
2.33806700e-01 -5.13704121e-01 -7.06301212e-01 3.16581935e-01
3.41161400e-01 -1.05072403e+00 1.14606404e+00 6.66764658e-03
6.57503009e-02 -5.53407669e-01 -2.29030982e-01 -1.21159744e+00
-5.96177459e-01 -1.10768164e-02 5.32683969e-01 1.35609841e+00
5.70798144e-02 -8.24526131e-01 6.64316714e-01 6.10928595e-01
-1.20753743e-01 -3.67535740e-01 -8.56362939e-01 -3.51346731e-01
-2.10593477e-01 -1.69071168e-01 9.82944369e-01 1.36711192e+00
-4.61191356e-01 3.90145332e-02 -5.14796555e-01 8.64200056e-01
2.95394868e-01 4.28938717e-01 8.96420836e-01 -1.35613358e+00
3.28736424e-01 -3.11590463e-01 -7.51823962e-01 -1.03215861e+00
1.45664826e-01 -5.55402577e-01 1.22057870e-01 -1.84056592e+00
-1.94879889e-01 -8.02368760e-01 -4.54198152e-01 4.85455841e-01
-1.16233818e-01 5.83600879e-01 -1.80787072e-01 4.60189551e-01
-8.09904397e-01 4.07612056e-01 4.43341345e-01 -5.19726098e-01
-1.56775713e-01 1.37448683e-01 -4.18285668e-01 6.41946495e-01
9.28746581e-01 -2.55907089e-01 -4.30025995e-01 -8.94905627e-01
4.78772998e-01 -3.82764935e-01 6.04775190e-01 -1.45907724e+00
6.64361179e-01 -1.92066774e-01 5.38033187e-01 -7.67657816e-01
3.79534394e-01 -7.82983601e-01 3.43323350e-01 -3.83545570e-02
7.62205124e-02 -1.30744502e-01 1.03006065e-01 8.41761529e-01
-4.88874257e-01 -6.26670569e-02 5.19980252e-01 -1.80796459e-01
-1.21663439e+00 4.61451560e-01 -7.25642443e-02 -3.29419583e-01
8.12602341e-01 -4.44133013e-01 -7.39995688e-02 -3.53810996e-01
-4.42994386e-01 2.55917460e-01 9.08415616e-01 1.27427459e-01
1.27184972e-01 -1.28972065e+00 -3.91877174e-01 -5.81520684e-02
3.39400500e-01 2.20245078e-01 6.07134998e-01 9.23047185e-01
-8.15861523e-01 7.63853967e-01 -1.11831665e-01 -6.42029703e-01
-8.83667350e-01 5.37829041e-01 4.91906583e-01 -1.49473593e-01
-3.36771339e-01 5.79905212e-01 4.19323266e-01 -5.23890197e-01
2.14059070e-01 -3.90973210e-01 -5.84873915e-01 3.73715341e-01
2.97012299e-01 1.37493804e-01 1.63215429e-01 -1.14386952e+00
-4.36452061e-01 8.33758891e-01 4.06098962e-01 2.98658490e-01
1.38936651e+00 -5.67213714e-01 -9.83306020e-02 1.61683649e-01
9.33621883e-01 4.11713608e-02 -1.12382543e+00 -5.69919109e-01
3.03479105e-01 -5.66458285e-01 1.59965113e-01 -5.09548724e-01
-1.12978172e+00 8.26897502e-01 6.93665862e-01 2.77056277e-01
1.24759746e+00 -2.56545871e-01 4.45045620e-01 5.34077406e-01
7.13229597e-01 -1.27034783e+00 -5.56253016e-01 4.19195086e-01
5.54707885e-01 -1.51190269e+00 2.14057893e-01 -2.81660557e-01
-2.76245713e-01 9.42618787e-01 2.82936066e-01 -1.83642492e-01
7.71646023e-01 -4.88714635e-01 -1.28932800e-02 -3.82006983e-03
-5.14347591e-02 -5.58489084e-01 4.32709485e-01 5.36067307e-01
3.78473222e-01 2.72906963e-02 -8.29008147e-02 2.54307270e-01
-5.73630743e-02 -5.65844700e-02 3.07369709e-01 7.13391960e-01
-4.43197519e-01 -6.61170423e-01 -3.73653680e-01 1.23604521e-01
-2.88371831e-01 -2.74936348e-01 -3.45977157e-01 9.32017505e-01
5.37886083e-01 8.18005443e-01 1.68380097e-01 -1.91648498e-01
2.00760290e-02 -2.41895244e-01 -1.77114215e-02 -5.15563667e-01
-5.10856688e-01 -6.23938553e-02 5.43464087e-02 -5.49891949e-01
-8.29864562e-01 -6.15831196e-01 -7.75542676e-01 -2.54883707e-01
-2.53265291e-01 -6.34131208e-02 8.87813807e-01 9.29236412e-01
6.11607254e-01 1.93624631e-01 6.65876389e-01 -1.37049007e+00
-3.15017775e-02 -9.67998862e-01 -6.29167855e-01 7.66966399e-03
3.78239006e-01 -6.43883884e-01 3.33121084e-02 -1.59749761e-02] | [7.682529926300049, -1.8209214210510254] |
196161f5-f6e6-4d37-b955-42b12f20e2f1 | surrogate-assisted-semi-supervised-inference | 2105.01264 | null | https://arxiv.org/abs/2105.01264v1 | https://arxiv.org/pdf/2105.01264v1.pdf | Surrogate Assisted Semi-supervised Inference for High Dimensional Risk Prediction | Risk modeling with EHR data is challenging due to a lack of direct observations on the disease outcome, and the high dimensionality of the candidate predictors. In this paper, we develop a surrogate assisted semi-supervised-learning (SAS) approach to risk modeling with high dimensional predictors, leveraging a large unlabeled data on candidate predictors and surrogates of outcome, as well as a small labeled data with annotated outcomes. The SAS procedure borrows information from surrogates along with candidate predictors to impute the unobserved outcomes via a sparse working imputation model with moment conditions to achieve robustness against mis-specification in the imputation model and a one-step bias correction to enable interval estimation for the predicted risk. We demonstrate that the SAS procedure provides valid inference for the predicted risk derived from a high dimensional working model, even when the underlying risk prediction model is dense and the risk model is mis-specified. We present an extensive simulation study to demonstrate the superiority of our SSL approach compared to existing supervised methods. We apply the method to derive genetic risk prediction of type-2 diabetes mellitus using a EHR biobank cohort. | ['Tianxi Cai', 'Zijian Guo', 'Jue Hou'] | 2021-05-04 | null | null | null | null | ['genetic-risk-prediction'] | ['medical'] | [ 3.12161297e-01 4.34952319e-01 -7.25892425e-01 -9.31959033e-01
-1.14900684e+00 -1.30186096e-01 7.86286294e-02 2.37427101e-01
7.91282207e-02 1.08753622e+00 8.27239275e-01 -3.31962109e-01
-3.41868997e-01 -7.32231259e-01 -8.04521799e-01 -4.96951461e-01
-4.19875413e-01 7.24324942e-01 -8.10530961e-01 4.78302360e-01
-2.29157105e-01 -3.16355899e-02 -9.09896076e-01 2.38570824e-01
1.35845840e+00 7.07759798e-01 -4.30487424e-01 2.99096197e-01
1.35043830e-01 8.67717862e-01 -8.89648572e-02 -6.58446312e-01
3.50918949e-01 -4.60217685e-01 -3.57364476e-01 -2.97421008e-01
4.83313054e-02 -5.09474337e-01 3.82686197e-03 6.40392423e-01
3.67597759e-01 -4.90679175e-01 1.21430516e+00 -1.35367835e+00
-5.55695415e-01 7.40562141e-01 -3.83444339e-01 -6.33561015e-01
2.51445949e-01 -2.61913300e-01 6.91971064e-01 -7.84118593e-01
5.58240056e-01 9.91189599e-01 1.30314124e+00 2.90891767e-01
-1.72533202e+00 -6.75014436e-01 -2.96691090e-01 -2.47959256e-01
-1.36880767e+00 -6.03279173e-01 2.80639917e-01 -6.92408442e-01
3.36768031e-01 3.61011140e-02 5.09313643e-01 1.17253232e+00
3.20760787e-01 2.90678978e-01 1.14247572e+00 -3.76425326e-01
6.40167356e-01 5.72707653e-02 4.03856248e-01 5.05101085e-01
3.30449551e-01 6.95276260e-01 -2.51564831e-01 -1.25769556e+00
7.22351015e-01 4.59550470e-01 2.37062916e-01 -5.14622867e-01
-1.18564582e+00 8.62498641e-01 -1.24369286e-01 -6.78381026e-01
-6.13579094e-01 -5.66951372e-02 2.35675007e-01 3.14602911e-01
6.37226343e-01 -1.45325005e-01 -6.90476656e-01 6.45629540e-02
-1.08574045e+00 1.81002673e-02 1.05926895e+00 1.10754442e+00
4.02238250e-01 -9.62434039e-02 -5.26458442e-01 8.42589259e-01
4.48415458e-01 6.79878533e-01 6.31731004e-02 -1.21475816e+00
5.30089796e-01 7.07527637e-01 4.41402823e-01 -4.55747813e-01
-4.89572406e-01 -4.62008417e-01 -1.39486551e+00 4.01066728e-02
8.10902357e-01 -5.91341496e-01 -9.50435817e-01 2.04285908e+00
3.52482766e-01 5.61257720e-01 3.98899168e-01 4.79028642e-01
4.66086030e-01 2.81991005e-01 3.62713814e-01 -6.03587627e-01
1.15396631e+00 -3.27240855e-01 -6.09572887e-01 1.29613474e-01
1.07344484e+00 -1.70164898e-01 4.60864067e-01 2.49829516e-01
-1.03677881e+00 -2.04410434e-01 -4.25982982e-01 1.00071043e-01
1.84396967e-01 3.24720681e-01 6.08155131e-01 7.45574415e-01
-3.34514320e-01 4.05047357e-01 -9.08922195e-01 -1.57115564e-01
7.10476518e-01 4.63101953e-01 -3.28619480e-01 -3.55414808e-01
-1.16534114e+00 7.05720901e-01 -1.55028095e-02 6.35079741e-02
-7.26708114e-01 -1.39661288e+00 -8.08806062e-01 8.15167204e-02
2.23265626e-02 -1.04779923e+00 7.47903764e-01 -8.47809076e-01
-9.42075253e-01 6.17952824e-01 -5.21258056e-01 -5.27416885e-01
7.90744126e-01 -1.72152072e-02 -4.39982444e-01 -3.22074354e-01
2.24305749e-01 6.86699972e-02 5.33297956e-01 -9.40637290e-01
-2.73348153e-01 -7.20492959e-01 -9.55815256e-01 -6.86403289e-02
2.11566955e-01 1.99911790e-03 1.29310623e-01 -9.12178755e-01
1.26756623e-01 -9.10385907e-01 -6.03221476e-01 -3.93795371e-02
-6.19591951e-01 2.93257833e-01 -1.46927819e-01 -8.59004557e-01
1.24029517e+00 -1.91846967e+00 -5.76215126e-02 6.27960265e-01
1.71985254e-01 -3.05066675e-01 -7.86017179e-02 4.16248500e-01
-2.37928450e-01 -1.35182798e-01 -5.12931466e-01 -2.89959908e-01
-1.12460263e-01 2.61389315e-01 -3.20360959e-01 5.44673622e-01
8.25050250e-02 9.01842296e-01 -6.11791313e-01 -4.71935838e-01
-1.43702865e-01 4.85673994e-01 -6.24731362e-01 4.15656209e-01
1.33048430e-01 6.87003613e-01 -3.70937824e-01 7.93631911e-01
8.22870255e-01 -4.36923712e-01 5.52579820e-01 -1.26617383e-02
8.42201263e-02 -1.03972079e-02 -1.35191059e+00 1.38615596e+00
-2.00427055e-01 -4.27405089e-01 -3.03078024e-03 -9.66645122e-01
1.05176127e+00 4.82660294e-01 7.44820356e-01 -7.22868517e-02
-2.18351692e-01 2.42691860e-01 -2.06893578e-01 -4.29456830e-01
-5.89910090e-01 -3.67069066e-01 -2.39466578e-01 5.22468090e-01
-9.73161608e-02 7.42688596e-01 -3.74888986e-01 3.69366258e-02
1.25447524e+00 1.80331856e-01 5.31714141e-01 -5.95321730e-02
1.51717529e-01 9.38128084e-02 1.22909200e+00 1.13351929e+00
7.77495652e-02 6.45812929e-01 7.14548945e-01 -6.52927041e-01
-1.04298043e+00 -1.45974433e+00 -9.60288048e-01 6.13083780e-01
-5.11557639e-01 -2.38379180e-01 -2.40010858e-01 -6.50069714e-01
5.65060735e-01 6.10385716e-01 -6.92233026e-01 -1.22115321e-01
-2.21982926e-01 -1.61886728e+00 6.82487726e-01 7.42422402e-01
1.99030042e-02 -4.04695839e-01 1.37469545e-01 5.25179029e-01
-8.47723559e-02 -5.26173770e-01 -1.01367466e-01 3.02144647e-01
-1.15408504e+00 -1.14098978e+00 -7.96055973e-01 -6.69649839e-01
9.14756298e-01 -6.09558225e-01 1.12240326e+00 -1.76638424e-01
1.11676857e-01 1.21014593e-02 9.63053666e-03 -4.63055670e-01
-6.93190873e-01 -2.99169242e-01 2.71044314e-01 1.58902958e-01
6.04896009e-01 -6.18071258e-01 -7.22898006e-01 2.79687017e-01
-2.90898532e-01 1.63699940e-01 6.61332965e-01 1.15915751e+00
7.46465385e-01 -4.64767188e-01 1.29654932e+00 -1.33289719e+00
-1.96748059e-02 -1.11442268e+00 -6.82134509e-01 5.17573953e-01
-1.18900526e+00 1.40383765e-01 3.85318071e-01 -3.50717843e-01
-1.21900392e+00 6.21858776e-01 1.80027336e-02 -6.09197794e-03
-3.46598804e-01 7.58096397e-01 -1.90530643e-01 3.89756143e-01
4.70667660e-01 -4.72885147e-02 4.11668450e-01 -8.03568661e-01
1.70126393e-01 8.39325368e-01 3.10622394e-01 -6.16345704e-01
5.00535965e-01 4.10847396e-01 4.45040047e-01 -8.91641602e-02
-9.39859033e-01 -3.02039504e-01 -7.65806019e-01 3.86409193e-01
6.31896198e-01 -1.49763691e+00 -7.34935522e-01 2.59872884e-01
-6.73647225e-01 -3.24839145e-01 -2.92535067e-01 1.04760861e+00
-8.40243101e-01 9.65988934e-02 -7.15417981e-01 -1.02139187e+00
-3.61105770e-01 -7.47103155e-01 9.95848656e-01 -3.50060165e-01
-3.84804130e-01 -1.04906368e+00 4.76925492e-01 3.06124777e-01
6.75027743e-02 4.16779965e-01 1.38430321e+00 -8.46073627e-01
-3.58653337e-01 -4.44383323e-01 -2.79682487e-01 -3.24234590e-02
1.21224545e-01 -5.04490197e-01 -5.79250276e-01 -1.05568074e-01
-1.39669850e-01 -2.59787947e-01 7.66860068e-01 8.43341827e-01
1.08623505e+00 -5.49150467e-01 -3.43768328e-01 9.40564692e-01
1.28369367e+00 -2.47473627e-01 5.84492862e-01 -1.83858693e-01
4.60853457e-01 6.57586873e-01 3.73678684e-01 8.74099791e-01
6.78603768e-01 6.28651023e-01 5.30141499e-03 -3.36552799e-01
4.15559679e-01 -5.21509171e-01 2.15217769e-01 4.16306853e-01
3.35722119e-02 2.81212598e-01 -1.06311977e+00 4.01943266e-01
-2.03577924e+00 -9.80408192e-01 -6.65855467e-01 2.68630195e+00
1.15666544e+00 -4.49426830e-01 3.41644138e-01 -1.93529770e-01
7.13231027e-01 -7.27737188e-01 -8.33714187e-01 -1.98207960e-01
-2.03177750e-01 -1.74517110e-02 5.57764351e-01 3.03513825e-01
-9.12283957e-01 7.50137074e-03 7.44166279e+00 2.20909953e-01
-2.05307320e-01 1.87677458e-01 9.60515976e-01 -1.03446230e-01
-1.40774176e-01 2.15675190e-01 -8.01053345e-01 6.93291366e-01
1.39460540e+00 -1.12615541e-01 8.97959098e-02 6.23796582e-01
5.87440252e-01 1.09789722e-01 -1.37024534e+00 6.69809699e-01
-2.35549629e-01 -1.33330631e+00 -3.07551146e-01 3.67230475e-01
8.61201346e-01 -1.96726806e-02 -2.44609773e-01 2.18258023e-01
7.31148064e-01 -1.08022511e+00 1.34778678e-01 1.04038119e+00
1.22779715e+00 -6.09055459e-01 9.29836750e-01 4.81514722e-01
-5.81710696e-01 -3.76202375e-01 -2.99167037e-01 -9.51328576e-02
3.41877520e-01 1.06701255e+00 -7.37069368e-01 4.02182460e-01
2.79058039e-01 8.14989567e-01 -2.56248206e-01 1.03096807e+00
1.99034303e-01 1.14829063e+00 -2.82384574e-01 8.49310756e-01
-6.29357874e-01 -4.72067714e-01 1.01497240e-01 8.68768990e-01
6.62592292e-01 3.53396624e-01 1.95660487e-01 1.05006087e+00
-1.35902595e-02 1.97578564e-01 -5.08913696e-01 3.89045954e-01
5.08950830e-01 7.11069226e-01 2.11200073e-01 -3.98014843e-01
-7.53868639e-01 3.43446493e-01 2.74302691e-01 5.67518115e-01
-5.62623024e-01 1.54050052e-01 5.36819100e-01 1.94070026e-01
-3.85526419e-02 1.07193403e-01 -7.22455323e-01 -1.45292819e+00
-2.17531011e-01 -7.28339851e-01 8.89789999e-01 -6.26018941e-01
-1.98206770e+00 -1.28385842e-01 -9.55725312e-02 -1.28666377e+00
-4.78381753e-01 -2.15591360e-02 -5.52207589e-01 1.38243639e+00
-1.39600170e+00 -1.25556517e+00 1.09375149e-01 5.00526726e-01
-2.77663738e-01 -2.80163527e-01 1.20454800e+00 1.19358711e-01
-7.63364255e-01 6.56732380e-01 7.05828786e-01 2.60864735e-01
9.83850181e-01 -1.20132208e+00 -8.41792524e-02 2.80036002e-01
-4.40639257e-01 7.70654798e-01 2.68162578e-01 -1.34423053e+00
-1.43340921e+00 -1.44470155e+00 1.11048663e+00 -6.68093443e-01
3.68820012e-01 -3.08511943e-01 -9.88795578e-01 9.90423620e-01
-5.48159957e-01 2.77384400e-01 1.49450743e+00 6.42047644e-01
-3.62361461e-01 -2.41880581e-01 -1.60103130e+00 1.81496695e-01
9.07414138e-01 -1.58546343e-01 -6.38344944e-01 2.88251370e-01
2.60298073e-01 -1.32208988e-01 -1.57211876e+00 6.09536171e-01
1.01334858e+00 -6.21604860e-01 1.15519369e+00 -1.11698079e+00
4.92139220e-01 -1.09975874e-01 -2.61397332e-01 -9.24983382e-01
-3.75229657e-01 -3.60862464e-01 -4.01203245e-01 1.29464400e+00
6.00449681e-01 -6.40263677e-01 9.15691733e-01 1.32412589e+00
2.82990426e-01 -4.17744488e-01 -1.18982255e+00 -6.58520162e-01
3.21017146e-01 -3.80058825e-01 8.40323925e-01 9.70415652e-01
1.49015099e-01 -6.51662648e-02 -9.50040340e-01 4.63449687e-01
1.37144125e+00 4.24464256e-01 5.85913658e-01 -1.64786100e+00
-3.69826883e-01 6.01804316e-01 -2.10529655e-01 -3.72940630e-01
4.09799784e-01 -1.01736593e+00 -3.33643883e-01 -1.15919423e+00
7.32806921e-01 -9.63172793e-01 -4.82471704e-01 6.84698522e-01
-2.05542266e-01 2.19914149e-02 -3.10163796e-01 3.84766400e-01
-2.11331338e-01 4.61067528e-01 7.20115900e-01 1.24042459e-01
-5.17447293e-01 5.98315835e-01 -7.52458990e-01 6.57244265e-01
6.40408754e-01 -9.26472425e-01 -3.75972718e-01 -9.26264375e-02
2.81352818e-01 9.16064441e-01 7.42007852e-01 -4.93929237e-01
-9.40356925e-02 -3.65236104e-01 1.03591526e+00 -5.06024003e-01
8.68902430e-02 -9.66918230e-01 7.27596045e-01 5.88991225e-01
-9.16567147e-01 -1.94849744e-01 -4.03898418e-01 1.07663727e+00
1.97222501e-01 1.13282919e-01 4.48569089e-01 1.57205194e-01
2.09372967e-01 3.50367159e-01 -1.78404540e-01 1.53675050e-01
6.94081128e-01 -3.63468565e-02 -2.19034314e-01 -3.15915167e-01
-1.34760106e+00 3.52802008e-01 4.55677956e-01 -1.02500394e-01
5.20789742e-01 -1.56575441e+00 -1.07651412e+00 7.57711470e-01
3.00541162e-01 -3.62480342e-01 1.48696601e-01 1.14473033e+00
2.26720944e-01 1.47522867e-01 -3.83415483e-02 -4.27262098e-01
-8.25082600e-01 6.78032279e-01 1.49134323e-01 -6.96031675e-02
-8.27147484e-01 -7.11693689e-02 1.09078452e-01 -5.31690598e-01
2.82405198e-01 -1.36983871e-01 6.87767863e-02 -1.08854569e-01
5.64398527e-01 5.82642436e-01 -1.80499598e-01 -4.32162106e-01
-2.34007284e-01 4.17524844e-01 2.59833276e-01 1.18320979e-01
1.48115087e+00 -3.78725350e-01 -4.38161902e-02 5.16335070e-01
9.83292699e-01 -2.95958053e-02 -1.24138081e+00 -3.92141610e-01
1.90996319e-01 -3.33261997e-01 -1.99933201e-01 -1.22232473e+00
-5.18836617e-01 3.76639515e-01 6.62547648e-01 -3.30785871e-01
9.03582633e-01 -1.32588327e-01 4.17427599e-01 4.82698642e-02
3.33918184e-01 -6.59804761e-01 -9.71927583e-01 1.03388265e-01
4.88284945e-01 -1.53699398e+00 5.58970384e-02 -2.92672843e-01
-6.47914827e-01 7.63745844e-01 7.44622722e-02 6.59513324e-02
9.62837279e-01 3.90337586e-01 1.16230249e-01 2.21914381e-01
-1.14907992e+00 1.76117152e-01 3.05695623e-01 9.79437590e-01
3.98350835e-01 2.56855786e-01 -2.81597495e-01 1.23544216e+00
1.73362613e-01 7.84369051e-01 4.94233251e-01 5.44891238e-01
1.39982402e-01 -1.33310282e+00 -4.96376216e-01 1.35088468e+00
-6.37689829e-01 -3.25584084e-01 3.21594290e-02 1.57605633e-01
1.60664245e-02 8.64806831e-01 3.06164604e-02 2.40636602e-01
3.28743428e-01 5.88775098e-01 -3.63682322e-02 -6.41397119e-01
-3.64120543e-01 8.89165923e-02 3.07281464e-01 -4.64846700e-01
-3.42954099e-01 -9.74280953e-01 -1.09952962e+00 -1.20966889e-01
7.76039343e-03 2.40927890e-01 2.44681060e-01 9.01796401e-01
8.17400455e-01 -1.34507105e-01 7.74597228e-01 -1.05588071e-01
-1.15680516e+00 -5.95397294e-01 -8.62232864e-01 5.42066038e-01
4.34467226e-01 -4.76214975e-01 -3.21586937e-01 3.05674940e-01] | [7.701855659484863, 4.938989162445068] |
8bdf924e-c9fa-4704-9771-d5aeb35c3d3c | a-neural-transition-based-model-for-nested | 1810.01808 | null | http://arxiv.org/abs/1810.01808v1 | http://arxiv.org/pdf/1810.01808v1.pdf | A Neural Transition-based Model for Nested Mention Recognition | It is common that entity mentions can contain other mentions recursively.
This paper introduces a scalable transition-based method to model the nested
structure of mentions. We first map a sentence with nested mentions to a
designated forest where each mention corresponds to a constituent of the
forest. Our shift-reduce based system then learns to construct the forest
structure in a bottom-up manner through an action sequence whose maximal length
is guaranteed to be three times of the sentence length. Based on Stack-LSTM
which is employed to efficiently and effectively represent the states of the
system in a continuous space, our system is further incorporated with a
character-based component to capture letter-level patterns. Our model achieves
the state-of-the-art results on ACE datasets, showing its effectiveness in
detecting nested mentions. | ['Bailin Wang', 'Yu Wang', 'Wei Lu', 'Hongxia Jin'] | 2018-10-03 | a-neural-transition-based-model-for-nested-1 | https://aclanthology.org/D18-1124 | https://aclanthology.org/D18-1124.pdf | emnlp-2018-10 | ['nested-named-entity-recognition', 'nested-mention-recognition'] | ['natural-language-processing', 'natural-language-processing'] | [ 2.81012297e-01 3.40630293e-01 -2.62963384e-01 -6.49798274e-01
-1.01888525e+00 -7.32632160e-01 4.57841158e-01 3.24301839e-01
-2.63366461e-01 7.20663667e-01 4.13948953e-01 -8.00252855e-01
5.13605356e-01 -1.06932032e+00 -1.02918875e+00 -2.87792295e-01
-4.52699453e-01 3.91233951e-01 5.01792789e-01 -6.54235333e-02
2.60774434e-01 4.16825533e-01 -1.20047247e+00 7.19821453e-01
8.94773304e-01 3.96029353e-01 2.50618756e-01 7.96000898e-01
-8.27530265e-01 1.00543869e+00 -6.05170667e-01 -4.94654000e-01
1.34769842e-01 -1.76255405e-01 -1.41585970e+00 1.66821361e-01
6.53308272e-01 -6.59719408e-02 -3.11061829e-01 9.03460920e-01
-7.12327287e-02 1.47966012e-01 5.02790689e-01 -6.41324759e-01
-6.21195078e-01 1.25041449e+00 -5.37293971e-01 3.33995402e-01
3.76142472e-01 -9.99029651e-02 1.24820960e+00 -7.64819741e-01
7.71060526e-01 1.17918336e+00 5.58414102e-01 3.10535461e-01
-1.17098260e+00 -3.24955732e-01 7.29392231e-01 1.41141579e-01
-1.03419566e+00 -3.92983198e-01 2.75016040e-01 -1.92385212e-01
1.70906579e+00 2.12255970e-01 5.67155659e-01 3.92751783e-01
3.12723964e-01 8.30527008e-01 8.10448468e-01 -7.11021423e-01
2.38524094e-01 -3.27179432e-01 6.65732861e-01 1.03492451e+00
1.41436290e-02 -6.19888842e-01 -4.67609644e-01 -1.07704774e-01
4.88360405e-01 -9.59219262e-02 1.46861270e-01 -7.05295056e-02
-8.82191956e-01 7.65156269e-01 6.72472894e-01 3.26659977e-01
-5.30996978e-01 3.08323234e-01 4.87106234e-01 3.43768328e-01
3.45340788e-01 3.46830785e-01 -7.32432067e-01 -2.05048639e-02
-1.16156518e+00 1.20654315e-01 1.08365488e+00 1.15940547e+00
8.92137706e-01 -1.57648996e-01 -4.01031584e-01 5.02625048e-01
3.15701574e-01 1.27462611e-01 8.21735784e-02 -7.86551118e-01
7.28465140e-01 9.16691363e-01 -3.38518284e-02 -3.37853312e-01
-2.98015088e-01 -4.01814550e-01 -6.00907683e-01 -2.09906965e-01
2.63433844e-01 -9.82375890e-02 -1.21980190e+00 1.69029260e+00
3.99152249e-01 4.23271596e-01 7.76849268e-03 2.54188001e-01
6.62474811e-01 8.58303130e-01 2.57812232e-01 -2.64540553e-01
1.59717715e+00 -1.15205789e+00 -4.76231426e-01 -5.12104928e-01
8.81366968e-01 -2.62539923e-01 7.96629190e-01 -1.97005808e-01
-1.12598372e+00 -1.68069258e-01 -7.99924135e-01 -2.04532176e-01
-2.67508328e-01 2.03561530e-01 6.27016902e-01 5.68713546e-01
-1.19125795e+00 5.32591879e-01 -1.12797499e+00 -1.03697494e-01
1.42333508e-01 4.20829356e-01 -3.80253911e-01 2.72472855e-02
-1.39538050e+00 1.04564631e+00 6.00432217e-01 2.40808845e-01
-5.94481289e-01 -4.04856920e-01 -1.13272583e+00 2.44760558e-01
4.94203538e-01 -7.14433610e-01 1.59579790e+00 -4.43068802e-01
-1.42101622e+00 8.07500005e-01 -1.02126205e+00 -8.99202108e-01
-1.16939478e-01 -1.80071652e-01 -1.73976377e-01 6.33096248e-02
1.97078466e-01 4.18216377e-01 4.08905864e-01 -9.72308874e-01
-8.52530777e-01 -2.41443157e-01 4.29460585e-01 1.00991547e-01
-7.80410171e-02 4.15359586e-01 -5.46182871e-01 -7.64369518e-02
4.04318392e-01 -9.09819424e-01 -5.79062700e-01 -6.82763994e-01
-8.26579332e-01 -3.93034577e-01 2.17083260e-01 -8.40620935e-01
1.67701340e+00 -1.51413822e+00 -8.18775222e-02 4.81658429e-02
1.40182078e-01 3.21402669e-01 -2.24092826e-02 7.48547971e-01
-5.20693650e-03 1.69242993e-01 -5.05312204e-01 -5.17387688e-01
-2.82719452e-02 2.13805452e-01 -3.90877217e-01 2.08953217e-01
5.07209122e-01 1.16071367e+00 -9.75542963e-01 -5.73329389e-01
-1.45730317e-01 9.92709473e-02 -4.20569181e-01 1.39326170e-01
-4.26645070e-01 -1.75478488e-01 -3.76685143e-01 3.22622329e-01
7.36800492e-01 -3.29879284e-01 5.59428394e-01 1.49847478e-01
-4.50458139e-01 1.35180283e+00 -1.13318789e+00 1.40311766e+00
-4.49450970e-01 2.64315128e-01 -4.61545475e-02 -8.21377039e-01
9.59147692e-01 2.60183156e-01 -1.86530009e-01 -7.39909112e-02
-3.30133498e-01 1.42050147e-01 6.49804482e-03 -2.52599359e-01
7.39484310e-01 -4.43593413e-02 -3.06472629e-01 4.92129147e-01
-3.66348736e-02 2.35306606e-01 5.48919737e-01 5.13173163e-01
1.41539490e+00 2.25543126e-01 4.18806374e-01 -1.38869241e-01
5.41996539e-01 5.86013906e-02 6.68016791e-01 8.97467911e-01
2.80099898e-03 1.27591491e-01 6.27956152e-01 -4.20244753e-01
-8.47804129e-01 -1.03264427e+00 8.26222077e-02 1.37838483e+00
-3.59840006e-01 -6.48733556e-01 -9.96975482e-01 -8.66172671e-01
-3.10467482e-01 9.88219082e-01 -5.68973303e-01 -1.63670480e-02
-1.08070743e+00 -4.60922301e-01 6.32096350e-01 7.88460135e-01
4.73770767e-01 -1.33516121e+00 -6.71100914e-01 5.73435068e-01
-2.72908211e-01 -1.01924276e+00 -5.16113639e-01 7.81080484e-01
-9.81820941e-01 -6.96485519e-01 -2.86237597e-01 -1.22748315e+00
7.33699560e-01 3.34965698e-02 1.27183056e+00 6.94074184e-02
2.10016623e-01 -3.92355919e-01 -2.76868224e-01 -6.64729550e-02
-5.16426563e-01 7.01708734e-01 -2.71383911e-01 -3.36161613e-01
6.07275486e-01 -3.69693100e-01 -1.39091402e-01 -4.98898953e-01
-5.86770654e-01 1.28727794e-01 5.65589130e-01 6.33037925e-01
4.15818483e-01 -3.34989816e-01 4.03128266e-01 -1.26022327e+00
3.88091445e-01 -2.97265321e-01 -4.73151267e-01 4.47899252e-01
-2.67850965e-01 4.54013556e-01 6.16245806e-01 -1.41409516e-01
-9.80597913e-01 6.20300472e-01 -2.36996904e-01 6.85030743e-02
-3.68107319e-01 8.75256777e-01 -1.56616077e-01 4.84985292e-01
2.74383694e-01 4.62110400e-01 -4.60632980e-01 -5.76203525e-01
6.03114903e-01 7.35624075e-01 6.96425080e-01 -5.85144222e-01
6.50746286e-01 -5.67389727e-02 -1.25052363e-01 -5.85586190e-01
-9.84728396e-01 -4.69704241e-01 -1.08381724e+00 8.52700472e-02
6.02919817e-01 -8.19214225e-01 -6.93350732e-01 4.28669214e-01
-1.57241797e+00 -2.89327532e-01 -7.60633573e-02 -6.34962693e-02
-1.49185151e-01 4.37408835e-01 -1.44250607e+00 -1.00501382e+00
-6.60099506e-01 -8.66531909e-01 9.30174649e-01 1.96837857e-01
-1.87453017e-01 -5.93961060e-01 1.29250988e-01 2.34609805e-02
4.80129421e-02 -6.80194423e-02 1.11619675e+00 -9.16093171e-01
-8.33508909e-01 -1.08447991e-01 -7.97801372e-03 -6.26174659e-02
-2.05414854e-02 1.73488602e-01 -7.42744148e-01 -7.56994560e-02
-3.04226428e-01 1.34875299e-02 1.06577313e+00 2.82783180e-01
5.72921455e-01 -5.57135582e-01 -5.13386369e-01 2.33281806e-01
1.20807934e+00 1.49433896e-01 5.80066264e-01 7.08208084e-01
5.69374800e-01 5.60502768e-01 2.81799376e-01 6.49199560e-02
8.54991019e-01 3.69243622e-01 3.53319883e-01 1.84846729e-01
1.60432622e-01 -4.56687301e-01 5.25883138e-01 1.00504637e+00
2.10954711e-01 -1.07062809e-01 -1.18761241e+00 7.37489581e-01
-1.75094116e+00 -9.68854249e-01 -7.08141208e-01 2.02245307e+00
1.19877470e+00 4.49035257e-01 8.81205425e-02 -1.08523807e-02
1.07660472e+00 9.74306688e-02 -3.50390375e-01 -9.33972180e-01
1.66383713e-01 1.75484836e-01 6.84114456e-01 6.22492313e-01
-1.32599843e+00 1.56572795e+00 7.29662800e+00 3.04499030e-01
-9.28961754e-01 -1.82686821e-01 4.95448440e-01 1.46238700e-01
-2.69873291e-01 2.02957436e-01 -1.19391000e+00 3.14959913e-01
1.49693334e+00 -1.67816982e-01 1.50464281e-01 7.58224189e-01
6.33505508e-02 -1.02777526e-01 -1.00343549e+00 -2.12502945e-02
-3.25536847e-01 -1.48178029e+00 -8.35134089e-02 -1.28050357e-01
7.18792975e-01 2.60078430e-01 -2.98613697e-01 4.75808382e-01
7.87766635e-01 -7.74826586e-01 7.06899643e-01 3.08013737e-01
6.70934379e-01 -8.09097111e-01 5.72735846e-01 6.57283425e-01
-1.54469502e+00 -2.53623072e-03 -4.57961202e-01 -3.89928907e-01
5.96648633e-01 6.16043150e-01 -1.16182435e+00 5.32029688e-01
3.73718947e-01 3.05037409e-01 -3.56741160e-01 1.02971339e+00
-6.35283887e-01 9.68033195e-01 -2.84980923e-01 -2.84539312e-01
5.82725585e-01 -1.14739120e-01 1.94194257e-01 1.81892264e+00
7.42867403e-03 2.07811996e-01 3.99280787e-01 6.74578607e-01
-2.37246051e-01 -5.44587485e-02 -3.14745814e-01 7.03674480e-02
8.72280896e-01 1.13234437e+00 -8.75285745e-01 -7.62312353e-01
-5.39798617e-01 8.71045887e-01 8.77742052e-01 1.40865281e-01
-6.61090076e-01 -6.48164928e-01 4.55711871e-01 -4.45546204e-04
8.83537412e-01 -3.77738565e-01 -3.50258082e-01 -1.14370120e+00
2.13526096e-02 -7.33253956e-01 3.03001493e-01 -5.25811315e-01
-9.35770214e-01 8.89870644e-01 -1.63406804e-01 -5.82627118e-01
-4.24157083e-01 -3.63270193e-01 -8.24796140e-01 1.20631266e+00
-1.30937171e+00 -1.08922100e+00 3.94051731e-01 2.05462933e-01
5.41442335e-01 2.30154932e-01 1.09310293e+00 -7.40543529e-02
-7.17150986e-01 3.68425131e-01 -8.33761469e-02 5.81217349e-01
2.76781917e-01 -1.55438685e+00 1.46444678e+00 1.35807323e+00
4.22170669e-01 1.26441169e+00 7.10319579e-01 -8.98397744e-01
-1.23367548e+00 -1.23640919e+00 1.76524842e+00 -3.36226493e-01
7.70382464e-01 -5.53496003e-01 -1.24361920e+00 1.23001885e+00
2.86619365e-01 -2.38321662e-01 4.27720577e-01 5.75856566e-01
-5.52426636e-01 2.48595580e-01 -7.73721457e-01 5.38021803e-01
1.10513151e+00 -8.12596917e-01 -1.28169990e+00 1.89606816e-01
1.09392035e+00 -7.24542677e-01 -5.50213933e-01 8.93016383e-02
1.72588572e-01 -5.70612311e-01 5.54500937e-01 -9.95178223e-01
6.64994359e-01 -3.30335170e-01 6.55275062e-02 -1.29472423e+00
-7.51867592e-01 -4.69010234e-01 -6.72276974e-01 1.34090340e+00
8.62151921e-01 -3.46034735e-01 1.00719070e+00 7.55440772e-01
-2.98796713e-01 -9.26498532e-01 -8.91204357e-01 -5.84378064e-01
1.68260321e-01 -3.43853384e-01 7.54951119e-01 5.23892879e-01
4.65151370e-01 7.97631621e-01 -5.71349002e-02 4.54041183e-01
2.18512967e-01 4.94559288e-01 2.97359377e-01 -1.00470841e+00
-1.80689573e-01 -7.20003992e-02 -1.40754908e-01 -1.50569189e+00
4.97116804e-01 -1.10477388e+00 5.20320475e-01 -2.07038927e+00
2.85913050e-01 -2.01975852e-01 -3.41193080e-01 8.50262284e-01
-4.14474547e-01 -1.97004646e-01 2.22901404e-01 -3.31921503e-02
-6.92297757e-01 9.78857055e-02 7.05620706e-01 -1.50905401e-01
-2.28096381e-01 1.06764711e-01 -6.32713616e-01 6.09745860e-01
6.47700846e-01 -6.57839477e-01 -2.59505343e-02 -4.78104264e-01
1.26224563e-01 2.78649330e-01 -2.77959317e-01 -6.21880054e-01
5.45180261e-01 -1.24625959e-01 7.76550323e-02 -9.46133375e-01
1.34297118e-01 -3.35823536e-01 -2.10175097e-01 6.55601144e-01
-7.38152683e-01 2.55320787e-01 6.91660270e-02 3.61942261e-01
-1.26258001e-01 -5.33554196e-01 3.74932319e-01 -3.57922256e-01
-6.46756649e-01 1.51920453e-01 -7.58607864e-01 3.54858786e-02
6.70960486e-01 5.92522211e-02 -2.67933607e-01 -1.80991799e-01
-5.76139152e-01 3.26292306e-01 2.03291103e-01 2.12028742e-01
2.88535237e-01 -9.07656848e-01 -7.31706977e-01 1.36462906e-02
-2.36714587e-01 1.54868200e-01 6.63208887e-02 3.94648999e-01
-4.48996365e-01 6.61607742e-01 2.60812581e-01 -2.42144123e-01
-1.41653144e+00 6.44808471e-01 2.11673170e-01 -7.29177594e-01
-8.39463592e-01 1.08691037e+00 -1.74476728e-01 -7.13670135e-01
9.14479718e-02 -6.17220163e-01 -2.06543133e-01 -9.95685458e-02
7.57719457e-01 9.44623947e-02 2.08359823e-01 -6.19280040e-01
-7.12687016e-01 8.92657265e-02 -3.68479937e-01 -1.30547941e-01
1.41014659e+00 -4.11414132e-02 -4.75300461e-01 5.98733425e-01
9.72120106e-01 -1.23962853e-02 -1.16682565e+00 -3.46041322e-01
7.04057395e-01 1.00833140e-01 -2.64406085e-01 -6.99920952e-01
-6.51808023e-01 6.96057260e-01 -2.44707033e-01 2.39993408e-01
7.26741493e-01 1.75921872e-01 9.22394574e-01 7.59659708e-01
5.95770657e-01 -8.23585689e-01 -5.21027625e-01 1.27208877e+00
3.65670711e-01 -6.85738385e-01 -2.48875037e-01 -6.86582327e-01
-5.96490443e-01 1.09371078e+00 5.56920469e-01 -2.67765373e-01
2.25335956e-01 7.90540695e-01 -3.30907814e-02 1.90395027e-01
-1.18125975e+00 -1.93056837e-01 7.68646672e-02 3.18964481e-01
7.57670343e-01 1.96338519e-01 -2.67415285e-01 6.10771179e-01
-3.06160063e-01 2.91983522e-02 7.61010706e-01 1.05531049e+00
-1.05185986e+00 -1.29128528e+00 -2.08075553e-01 6.44113541e-01
-6.25604868e-01 -6.69544637e-01 -1.96925387e-01 4.16096449e-01
-2.41849571e-01 1.03441310e+00 2.80612439e-01 -2.44154662e-01
3.93808901e-01 6.58753395e-01 3.09172183e-01 -1.14688635e+00
-9.85210657e-01 -1.04596741e-01 4.54637319e-01 -3.17735404e-01
-1.73257709e-01 -8.94281626e-01 -1.75342739e+00 -3.32779169e-01
-3.48550677e-01 3.38373244e-01 3.06851327e-01 1.14344752e+00
3.66243869e-01 7.19197690e-01 4.38737571e-01 -4.49518085e-01
-6.88401520e-01 -1.11965430e+00 -5.26503026e-01 -7.79336989e-02
2.74741501e-01 -1.29245460e-01 4.81289178e-02 1.11366153e-01] | [10.339324951171875, 9.536539077758789] |
e4158ea5-787b-47dc-ae1c-45015579ae26 | online-hybrid-lightweight-representations | 2205.11179 | null | https://arxiv.org/abs/2205.11179v1 | https://arxiv.org/pdf/2205.11179v1.pdf | Online Hybrid Lightweight Representations Learning: Its Application to Visual Tracking | This paper presents a novel hybrid representation learning framework for streaming data, where an image frame in a video is modeled by an ensemble of two distinct deep neural networks; one is a low-bit quantized network and the other is a lightweight full-precision network. The former learns coarse primary information with low cost while the latter conveys residual information for high fidelity to original representations. The proposed parallel architecture is effective to maintain complementary information since fixed-point arithmetic can be utilized in the quantized network and the lightweight model provides precise representations given by a compact channel-pruned network. We incorporate the hybrid representation technique into an online visual tracking task, where deep neural networks need to handle temporal variations of target appearances in real-time. Compared to the state-of-the-art real-time trackers based on conventional deep neural networks, our tracking algorithm demonstrates competitive accuracy on the standard benchmarks with a small fraction of computational cost and memory footprint. | ['Bohyung Han', 'Eunhyeok Park', 'Minji Kim', 'Ilchae Jung'] | 2022-05-23 | null | null | null | null | ['visual-tracking'] | ['computer-vision'] | [ 1.70934007e-01 -1.31141409e-01 -5.97504675e-01 -3.26911844e-02
-6.77428365e-01 -2.82978892e-01 5.83964586e-01 -2.08030343e-01
-4.15542394e-01 5.19288957e-01 -3.79120819e-02 -2.61079371e-01
2.38183647e-01 -5.50134420e-01 -1.12382388e+00 -8.12169015e-01
-5.35238624e-01 1.52738705e-01 5.51941752e-01 2.30278866e-03
-1.22688174e-01 6.78786695e-01 -1.74425685e+00 2.40408137e-01
2.13451833e-01 1.72590566e+00 4.29210365e-02 7.76673257e-01
1.32180220e-02 1.14354002e+00 -7.44369030e-01 -9.44274738e-02
7.79561520e-01 1.30072445e-01 5.34495413e-02 -4.71149176e-01
1.18808448e+00 -6.53504491e-01 -1.11525774e+00 1.32887602e+00
5.74141502e-01 -1.03065953e-01 2.66356289e-01 -1.27833366e+00
-5.82544327e-01 4.51071769e-01 -5.84001720e-01 3.81083220e-01
-8.47207606e-02 9.25045609e-02 4.39012527e-01 -4.56031203e-01
4.55119461e-01 1.40099704e+00 1.10446954e+00 5.08132875e-01
-9.66324329e-01 -1.26801372e+00 1.90140903e-01 2.40519673e-01
-1.36748302e+00 -6.65011585e-01 3.94551188e-01 -2.65582561e-01
9.48032022e-01 -2.34267220e-01 6.77109838e-01 8.99675667e-01
5.55598617e-01 6.17984951e-01 4.88314688e-01 -1.40272170e-01
1.78795546e-01 -2.43416443e-01 1.78455293e-01 8.73280764e-01
5.05294740e-01 8.77934694e-01 -8.00507426e-01 -3.86208922e-01
1.15252852e+00 4.57947880e-01 -4.78868544e-01 -3.78531933e-01
-1.15131199e+00 6.85686707e-01 7.28404701e-01 -5.74918054e-02
-3.79106879e-01 1.17651963e+00 7.39900947e-01 4.77490872e-01
2.71998137e-01 -4.16274786e-01 -3.66570562e-01 -2.06959978e-01
-1.31988990e+00 2.36896530e-01 6.83793426e-01 1.06624234e+00
5.57247102e-01 6.05025589e-01 -5.17692447e-01 2.26619676e-01
5.13548434e-01 7.97205031e-01 5.74249923e-01 -1.07999802e+00
2.58969545e-01 1.53366193e-01 1.08679242e-01 -9.99427259e-01
-1.02127701e-01 -6.44916058e-01 -1.06947529e+00 6.43526852e-01
2.98372984e-01 -1.11240596e-01 -9.39317346e-01 1.81417179e+00
3.01910996e-01 8.55534077e-01 -6.87157586e-02 9.25514162e-01
6.78803205e-01 7.77383864e-01 -2.31929347e-02 -1.38737872e-01
1.39964497e+00 -1.11009979e+00 -7.97568321e-01 1.41181529e-01
9.69946310e-02 -5.25496960e-01 -1.55940533e-01 4.40163255e-01
-1.07256794e+00 -1.03087115e+00 -1.36752713e+00 -7.26719992e-03
-3.01750302e-01 4.22278434e-01 8.41394901e-01 8.80202949e-01
-1.56575441e+00 8.11381698e-01 -1.08937335e+00 1.07497424e-01
5.47818303e-01 6.94357038e-01 -7.70254806e-02 2.24278197e-01
-1.13915133e+00 4.99660969e-01 4.12920028e-01 2.38017634e-01
-1.18962812e+00 -7.16504037e-01 -8.90596747e-01 1.80726454e-01
7.76403472e-02 -4.94957387e-01 1.38053811e+00 -9.97313797e-01
-1.63518894e+00 3.21873307e-01 -8.91836286e-02 -1.15672326e+00
4.51066554e-01 -3.47294807e-01 -4.92920250e-01 2.74023980e-01
-2.53287166e-01 8.17001581e-01 1.39962494e+00 -8.07848573e-01
-8.54997396e-01 -2.23722354e-01 -1.23015426e-01 -2.86748618e-01
-3.14646453e-01 2.02002060e-02 -6.45439565e-01 -9.63397145e-01
-1.14532158e-01 -9.03772652e-01 -9.37738940e-02 1.03044271e+00
1.63002431e-01 -2.32679978e-01 1.42257607e+00 -4.35255915e-01
1.11809576e+00 -2.14426351e+00 -4.01626110e-01 8.59945547e-03
5.05248308e-01 7.87935793e-01 -1.72660947e-01 -7.04331696e-02
-3.52351135e-03 -5.55056930e-01 4.67498422e-01 -3.81865084e-01
9.49068591e-02 1.43107042e-01 -6.23959482e-01 9.15685356e-01
-1.94357753e-01 8.92610908e-01 -9.57577646e-01 -3.53182346e-01
2.02872649e-01 9.16399300e-01 -2.62758791e-01 1.31346300e-01
-1.49558157e-01 9.76703912e-02 -2.56941676e-01 6.38392389e-01
9.10588562e-01 -3.98442715e-01 3.92974205e-02 -4.81417865e-01
-2.90342301e-01 1.41779780e-01 -1.22104669e+00 1.88584316e+00
-1.38518317e-02 1.01008129e+00 3.77819270e-01 -6.72899127e-01
9.56003726e-01 4.17947203e-01 4.85265672e-01 -7.30290949e-01
2.01345801e-01 -3.30582485e-02 -4.27368194e-01 2.33489409e-01
9.00239468e-01 3.60769510e-01 2.00118452e-01 3.04575622e-01
1.59727737e-01 7.48093724e-01 -1.60885245e-01 1.85549214e-01
1.31221902e+00 5.06300986e-01 3.92452180e-02 -6.34774566e-02
2.43341178e-01 -1.03796378e-01 8.36815596e-01 1.14923513e+00
-5.63329399e-01 2.54920691e-01 3.48216146e-02 -8.91892314e-01
-8.68541718e-01 -1.07421291e+00 9.96525493e-03 1.20468378e+00
3.01685393e-01 -4.86243218e-01 -3.56419384e-01 -2.53703654e-01
2.15322927e-01 -2.82849409e-02 -4.79763687e-01 -2.29705706e-01
-5.98470628e-01 -1.34981632e-01 9.16149497e-01 8.85717154e-01
7.58049726e-01 -7.87561953e-01 -1.14509690e+00 2.90143102e-01
4.32954103e-01 -9.93342638e-01 -6.02100492e-01 3.60755891e-01
-9.05260205e-01 -1.06711459e+00 -7.46061265e-01 -7.53427982e-01
2.82241046e-01 5.10083377e-01 1.07760251e+00 9.35115442e-02
-2.37126261e-01 3.18325698e-01 -2.39282519e-01 -3.06712002e-01
-1.05634555e-01 -4.21221286e-01 2.17141449e-01 -1.10478051e-01
4.12843525e-01 -2.76885211e-01 -6.71052217e-01 -1.08466059e-01
-6.66900098e-01 -2.30864644e-01 5.30372262e-01 1.03580165e+00
5.76121032e-01 -7.71085620e-02 5.04423305e-03 -3.45501602e-01
1.24303751e-01 -2.67060548e-01 -1.21554971e+00 1.44565254e-01
-2.77032733e-01 1.56930998e-01 6.21317923e-01 -8.38452697e-01
-8.41851234e-01 5.74421883e-01 2.08545968e-01 -1.24362397e+00
2.82962531e-01 -1.11542575e-01 5.42330801e-01 -8.18354547e-01
5.19426525e-01 3.64469409e-01 2.57934749e-01 -3.99154514e-01
4.61857200e-01 4.77014929e-01 1.07127571e+00 -6.00236177e-01
8.10684264e-01 5.80434620e-01 3.67225669e-02 -5.98849654e-01
-3.54816645e-01 -4.94844884e-01 -4.22516078e-01 -2.71371186e-01
4.04244810e-01 -1.41206706e+00 -1.06867623e+00 5.34955442e-01
-1.43394768e+00 -2.93611199e-01 -3.97406608e-01 5.72610736e-01
-4.60350871e-01 4.32245314e-01 -9.92189765e-01 -7.68683076e-01
-6.15212798e-01 -1.04949713e+00 1.45090413e+00 3.07867974e-01
5.14834464e-01 -6.19079649e-01 2.46692881e-01 -4.73022014e-01
7.02660024e-01 4.13443059e-01 1.98586822e-01 -2.47819766e-01
-1.06088722e+00 -2.05595538e-01 -5.44709086e-01 2.16681346e-01
-1.39410853e-01 -8.84954631e-02 -1.07039046e+00 -8.81956577e-01
-1.65097401e-01 -4.54407513e-01 1.17602360e+00 6.81272805e-01
1.17876828e+00 -3.04084778e-01 -5.05243540e-01 9.87054586e-01
1.54162657e+00 1.28193513e-01 5.01104772e-01 1.60170168e-01
5.28327644e-01 -3.84247422e-01 4.40347910e-01 6.24140799e-01
1.14903226e-01 7.37425327e-01 5.65625966e-01 2.94121895e-02
-4.52556461e-01 -1.28388032e-01 6.98019505e-01 5.83893478e-01
-2.28049979e-02 -3.38141024e-02 -3.76365125e-01 3.14181894e-01
-2.10069227e+00 -1.13951623e+00 2.09849209e-01 2.37877965e+00
5.55655599e-01 7.61981457e-02 -3.30870994e-03 -2.69938827e-01
9.58431900e-01 4.94630188e-01 -6.54620230e-01 -6.61162138e-02
1.13447018e-01 2.41032481e-01 1.08920455e+00 -6.95241988e-02
-1.29517436e+00 7.31586516e-01 6.97964096e+00 9.25877988e-01
-1.41932416e+00 3.44759673e-01 1.74007148e-01 -3.18612367e-01
4.26459104e-01 -3.40352148e-01 -1.08879280e+00 4.75339532e-01
1.44751465e+00 9.74101722e-02 2.54054993e-01 1.23789847e+00
-2.51176775e-01 3.40361565e-01 -1.11179256e+00 1.43528867e+00
-2.97576040e-02 -1.73611736e+00 1.01212263e-01 2.34501250e-02
5.25990427e-01 5.07199824e-01 2.62858629e-01 4.85598683e-01
4.70461696e-01 -9.05476391e-01 9.00230110e-01 6.74104989e-01
1.07521152e+00 -6.27234876e-01 5.43689311e-01 1.79752171e-01
-1.99810803e+00 -2.11628661e-01 -8.80250454e-01 -2.08638571e-02
-7.39221647e-02 2.37876162e-01 -2.61579275e-01 3.38218629e-01
9.96230841e-01 7.94925749e-01 -3.92411560e-01 1.32374501e+00
3.01981270e-01 4.51262742e-01 -4.94875073e-01 1.20248690e-01
4.05108362e-01 3.42285365e-01 4.23161983e-01 1.30568755e+00
4.13276762e-01 -1.00059286e-01 4.13230926e-01 5.70760906e-01
-2.72036165e-01 -5.61372161e-01 -7.59657919e-01 1.47069395e-01
7.19125926e-01 1.20938647e+00 -4.01131481e-01 -6.94188774e-01
-5.69885314e-01 7.88338006e-01 3.76290381e-01 2.44856581e-01
-1.06629837e+00 -2.73891062e-01 8.81828904e-01 -2.05736071e-01
7.92054832e-01 -2.26114869e-01 3.25593829e-01 -1.01703274e+00
-1.65303558e-01 -7.95730650e-01 3.21036011e-01 -5.92528939e-01
-1.07031250e+00 6.81851923e-01 -3.09058160e-01 -1.67414558e+00
-4.00451809e-01 -9.76558208e-01 -3.12085062e-01 7.67928004e-01
-1.79987442e+00 -1.16718388e+00 -4.94797170e-01 8.43352318e-01
3.03725898e-01 -4.75567460e-01 7.34279096e-01 5.22075593e-01
-2.27481902e-01 9.30099070e-01 4.15948063e-01 4.04202610e-01
6.96567297e-01 -8.74096751e-01 5.42112470e-01 8.60321522e-01
1.44563928e-01 6.60264373e-01 3.44605029e-01 -6.89826012e-01
-1.94510710e+00 -1.61638665e+00 3.73576403e-01 9.05685499e-02
5.35652518e-01 -1.72364607e-01 -7.89537549e-01 8.81590247e-01
2.53918201e-01 8.85217249e-01 2.57081807e-01 -3.56031537e-01
-7.69218147e-01 -3.50653559e-01 -1.05075955e+00 1.23192251e-01
8.20065379e-01 -5.55383563e-01 -5.10681748e-01 3.09123486e-01
8.62843335e-01 -8.78121972e-01 -7.99297810e-01 9.86116827e-02
8.27984691e-01 -7.26074040e-01 1.12839425e+00 -3.28769326e-01
-3.16789299e-01 -6.18545473e-01 -6.93906471e-02 -7.99689233e-01
-6.17272735e-01 -9.12379622e-01 -9.93238568e-01 6.05643988e-01
-3.42154443e-01 -5.09723186e-01 1.05232847e+00 2.63227344e-01
-8.49514529e-02 -4.32822913e-01 -1.24632001e+00 -9.51379657e-01
-4.24081385e-01 -2.81349719e-01 5.31224906e-01 5.22821963e-01
-5.02355337e-01 -2.69810379e-01 -8.21126044e-01 4.74310249e-01
1.19648814e+00 4.41620052e-02 6.31119907e-01 -1.30801582e+00
-3.22824180e-01 -2.28480950e-01 -9.54340160e-01 -1.68728077e+00
1.11452386e-01 -5.81132412e-01 8.16304460e-02 -7.55052686e-01
1.92196313e-02 -3.38699907e-01 -6.51333749e-01 5.35536528e-01
4.16689306e-01 4.61689889e-01 3.41117889e-01 2.34087601e-01
-1.03552794e+00 6.31634533e-01 6.95105314e-01 -4.99542803e-01
1.23743825e-01 -1.85102314e-01 -2.19458729e-01 5.56192160e-01
2.63209701e-01 -6.32018685e-01 -2.51080632e-01 -5.60807884e-01
-3.33397180e-01 3.68144512e-01 6.78824782e-01 -1.48931563e+00
1.13069069e+00 4.15436268e-01 8.61038625e-01 -1.00818312e+00
5.61675608e-01 -1.07473588e+00 2.38462344e-01 8.98925602e-01
-2.38006726e-01 2.24848941e-01 4.06359226e-01 9.81603622e-01
-1.75904796e-01 2.01761067e-01 7.82617629e-01 1.35256991e-01
-1.05373919e+00 7.12707758e-01 -9.91798490e-02 -3.43316972e-01
8.65201890e-01 -3.99092734e-01 -5.95278025e-01 -3.46027672e-01
-4.20163602e-01 1.79507528e-02 1.65337786e-01 5.06246924e-01
6.22310460e-01 -1.76735210e+00 -4.16449040e-01 2.67112464e-01
-3.05250853e-01 -3.34736854e-01 9.84420106e-02 3.80381495e-01
-5.67699730e-01 7.20734596e-01 -5.10162055e-01 -9.12688076e-01
-1.15898156e+00 6.46636426e-01 5.13012171e-01 -1.91313282e-01
-8.56962144e-01 7.28573740e-01 -3.78513671e-02 1.37860864e-01
9.51567173e-01 -4.73932326e-01 1.77105553e-02 -3.45860392e-01
1.16788816e+00 3.85200620e-01 -9.91176292e-02 -8.21688235e-01
-2.93173850e-01 5.51094353e-01 -8.62045884e-02 3.56521457e-01
1.07960308e+00 1.70127854e-01 6.39150068e-02 1.22652911e-01
1.16719031e+00 -5.67866027e-01 -1.90982842e+00 -5.59235334e-01
-3.45808476e-01 -5.06350338e-01 5.86805165e-01 -3.29787940e-01
-1.59534049e+00 6.58668101e-01 1.37783086e+00 -1.33506715e-01
1.12702858e+00 -4.86859798e-01 9.14566457e-01 7.52669871e-01
8.06393564e-01 -8.66769373e-01 -9.07349735e-02 5.88386297e-01
4.62141961e-01 -1.00256968e+00 1.08459577e-01 -8.31528753e-02
-6.35812730e-02 1.45616484e+00 5.35396338e-01 -5.23273289e-01
7.95184731e-01 6.65109456e-01 -6.83570653e-02 -9.39554796e-02
-9.77480352e-01 -1.33466339e-02 2.23806337e-01 8.74822795e-01
2.17184961e-01 -2.41873771e-01 3.18894267e-01 2.26097435e-01
2.14422658e-01 3.76765996e-01 -4.70595099e-02 9.56773818e-01
-5.47282457e-01 -6.94478095e-01 -6.01254880e-01 3.55269492e-01
-4.23332095e-01 -1.18648894e-02 3.67508471e-01 6.74489319e-01
1.41655192e-01 5.10669112e-01 4.85486954e-01 -3.06886047e-01
-8.49008113e-02 -3.79564315e-01 4.63682950e-01 -2.30883792e-01
-8.50588977e-01 1.31666943e-01 -3.24578017e-01 -1.18028700e+00
-5.61163545e-01 -2.53842473e-01 -1.14907050e+00 -5.02496004e-01
-1.27070367e-01 4.88567166e-02 7.17527032e-01 5.93901634e-01
5.66901624e-01 7.01812088e-01 2.54045874e-01 -1.54532468e+00
-1.03287446e+00 -6.45819366e-01 -6.50338411e-01 -1.62419975e-01
9.90919173e-01 -8.67043555e-01 2.29111370e-02 1.08450592e-01] | [6.3238396644592285, -2.122180938720703] |
e17c86ed-8b55-45b6-a85f-6d379e818615 | deepirisnet2-learning-deep-iriscodes-from | 1902.05390 | null | http://arxiv.org/abs/1902.05390v1 | http://arxiv.org/pdf/1902.05390v1.pdf | DeepIrisNet2: Learning Deep-IrisCodes from Scratch for Segmentation-Robust Visible Wavelength and Near Infrared Iris Recognition | We first, introduce a deep learning based framework named as DeepIrisNet2 for
visible spectrum and NIR Iris representation. The framework can work without
classical iris normalization step or very accurate iris segmentation; allowing
to work under non-ideal situation. The framework contains spatial transformer
layers to handle deformation and supervision branches after certain
intermediate layers to mitigate overfitting. In addition, we present a dual CNN
iris segmentation pipeline comprising of a iris/pupil bounding boxes detection
network and a semantic pixel-wise segmentation network. Furthermore, to get
compact templates, we present a strategy to generate binary iris codes using
DeepIrisNet2. Since, no ground truth dataset are available for CNN training for
iris segmentation, We build large scale hand labeled datasets and make them
public; i) iris, pupil bounding boxes, ii) labeled iris texture. The networks
are evaluated on challenging ND-IRIS-0405, UBIRIS.v2, MICHE-I, and CASIA v4
Interval datasets. Proposed approach significantly improves the
state-of-the-art and achieve outstanding performance surpassing all previous
methods. | ['Akanksha Joshi', 'R. Raghavendra', 'Padmaja Joshi', 'Abhishek Gangwar'] | 2019-02-06 | null | null | null | null | ['iris-segmentation'] | ['medical'] | [ 3.89861643e-01 1.91067174e-01 -3.51737887e-01 -4.56117451e-01
-6.35499001e-01 -5.17557979e-01 3.42908412e-01 -2.90335268e-01
-1.44529447e-01 3.85125697e-01 7.44979456e-02 -2.97819763e-01
-2.18290046e-01 -6.76579833e-01 -5.38598120e-01 -7.21330881e-01
3.47410023e-01 6.26429200e-01 -1.01463862e-01 -7.81717361e-04
3.09730262e-01 5.80558181e-01 -1.49223864e+00 1.53556958e-01
1.11668670e+00 1.01024556e+00 -6.86296642e-01 7.30879545e-01
2.24461764e-01 7.25578368e-01 -4.27624792e-01 -4.49903309e-01
7.93431163e-01 -4.57545847e-01 -1.05762041e+00 3.34364235e-01
1.01812148e+00 -3.98292542e-01 -2.25753188e-02 1.13637948e+00
7.87676275e-01 3.45587521e-03 6.00732982e-01 -6.82223558e-01
-8.69225264e-01 5.32502651e-01 -1.21504366e+00 1.32385969e-01
2.81641856e-02 7.44961977e-01 3.74588370e-01 -1.20716982e-01
6.55183554e-01 1.02857029e+00 8.72202933e-01 6.52271748e-01
-9.89961743e-01 -4.94009227e-01 -4.54055846e-01 -3.96626920e-01
-1.35628390e+00 -4.68145043e-01 2.99393088e-01 -6.79529130e-01
6.41415834e-01 4.33307379e-01 3.91484320e-01 7.05404639e-01
-3.71755630e-01 6.82657182e-01 1.65536511e+00 -2.82872707e-01
-3.74916881e-01 3.18883695e-02 2.62360871e-01 8.69697094e-01
-8.25386569e-02 6.15996540e-01 5.05085438e-02 1.67858765e-01
1.04032207e+00 -8.25401917e-02 -2.35251449e-02 2.21202821e-02
-9.75456834e-01 3.15388650e-01 8.83750677e-01 -7.86788203e-03
-3.29868019e-01 -1.77796215e-01 2.99344659e-01 1.71081111e-01
3.37120086e-01 2.23042309e-01 -3.72365952e-01 2.42612123e-01
-1.21549273e+00 7.24025071e-02 8.76770169e-02 7.54212797e-01
5.05032361e-01 -1.45865381e-01 -7.23292589e-01 8.75192583e-01
3.36696148e-01 3.28712970e-01 2.42713779e-01 -3.31154048e-01
1.94329590e-01 9.39933836e-01 2.23429389e-02 -5.65005720e-01
-7.20165849e-01 -8.93836677e-01 -9.50779557e-01 3.69138420e-01
6.20601475e-01 -2.08864197e-01 -1.86123753e+00 1.05159020e+00
5.69967270e-01 5.04786193e-01 1.50839880e-01 1.24295723e+00
1.31033516e+00 2.07266301e-01 -7.91502073e-02 3.20983917e-01
1.36620343e+00 -1.21359956e+00 -2.59829402e-01 1.91489622e-01
3.50234836e-01 -1.24852526e+00 8.93633783e-01 4.32825565e-01
-1.18681145e+00 -7.28632092e-01 -6.17637873e-01 -4.27370399e-01
-3.06558460e-01 9.43855405e-01 6.59354806e-01 8.61551285e-01
-1.11312246e+00 4.02103275e-01 -7.37106264e-01 -4.85110253e-01
7.40543842e-01 7.48663187e-01 -1.16394304e-01 2.26573378e-01
-5.00569403e-01 5.23845077e-01 5.24984360e-01 3.53711158e-01
-5.36870778e-01 -6.08375907e-01 -8.24829459e-01 -2.74477333e-01
-4.82252194e-03 -7.44988322e-01 1.03869832e+00 -1.08203971e+00
-1.77081001e+00 1.63968492e+00 5.86923957e-03 -6.83642983e-01
4.32465822e-01 2.05545798e-01 -3.35119665e-01 -4.41749208e-02
-3.73470813e-01 7.84809530e-01 7.71860659e-01 -9.35819745e-01
-6.14427507e-01 -5.76226711e-01 3.09387743e-02 1.67448118e-01
2.34987602e-01 4.48593497e-01 -6.34844005e-01 -5.19009709e-01
8.02625865e-02 -1.02211940e+00 -1.56973720e-01 -2.34022081e-01
-1.02939081e+00 -2.00618222e-01 4.10966277e-01 -8.98823202e-01
9.17636037e-01 -2.03284597e+00 -2.68277258e-01 5.24360180e-01
1.80114269e-01 8.38613391e-01 -2.71256655e-01 -3.22173715e-01
-3.64032030e-01 -2.86472421e-02 -2.15283811e-01 -3.58074665e-01
-2.21404508e-01 -1.08898729e-01 7.11162984e-02 8.08084905e-01
9.21561494e-02 8.78439903e-01 -3.80628794e-01 -4.73806292e-01
5.91317892e-01 5.76253593e-01 -5.48379779e-01 8.39587115e-03
-2.32834652e-01 8.20291936e-01 -1.91507265e-01 1.35566020e+00
1.13877392e+00 -3.89521390e-01 -5.78587532e-01 -3.03584486e-01
-3.14682186e-01 -5.86792082e-02 -1.27109885e+00 1.73088276e+00
-1.55882925e-01 2.47414738e-01 1.26543596e-01 -7.81931102e-01
9.93670344e-01 1.97913021e-01 3.42788279e-01 -4.92428333e-01
5.54826081e-01 1.24982618e-01 7.58424848e-02 -5.21612942e-01
2.82503217e-01 1.58397570e-01 4.87725168e-01 2.20737651e-01
1.56186208e-01 1.66162357e-01 8.32556188e-02 -4.21924859e-01
4.42247152e-01 3.84773493e-01 5.96607812e-02 -5.16310818e-02
7.57940233e-01 2.49387726e-01 5.64148962e-01 5.19492149e-01
-3.93031478e-01 8.23058486e-01 3.81854087e-01 -8.93379033e-01
-9.05331552e-01 -8.14957023e-01 -7.33464599e-01 7.10414052e-01
1.44213840e-01 1.62193954e-01 -1.02728367e+00 -5.36538243e-01
-1.04389578e-01 -8.36371630e-02 -8.17137182e-01 5.48820853e-01
-1.21498920e-01 -1.33230090e+00 7.95773745e-01 1.78077802e-01
8.61552775e-01 -1.01314962e+00 -1.91552505e-01 -2.15631887e-01
1.85983062e-01 -9.01320159e-01 -5.52876413e-01 -4.20790464e-01
-7.08890021e-01 -1.56134760e+00 -7.09651053e-01 -8.73877823e-01
8.56645107e-01 -3.55195671e-01 1.00131643e+00 2.28381887e-01
-9.54640150e-01 -2.95711040e-01 -4.02021632e-02 -5.34826517e-01
-9.16259587e-02 1.78141013e-01 -2.31833175e-01 2.58057564e-01
7.69153714e-01 -1.10153452e-01 -9.37510073e-01 2.33357534e-01
-6.65120065e-01 1.52899787e-01 6.73527241e-01 8.86287153e-01
9.26895320e-01 -7.68702403e-02 -8.36354941e-02 -1.07139921e+00
7.17308186e-03 -3.97215262e-02 -1.11979640e+00 2.15165898e-01
-5.75811267e-01 -2.38133803e-01 3.02030802e-01 6.53991429e-03
-1.13059831e+00 3.75437975e-01 -3.11345935e-01 -2.49464557e-01
-7.40633309e-01 1.32879719e-01 4.44623888e-01 -5.02662599e-01
9.77001369e-01 -9.34316684e-03 4.04066555e-02 -7.18482137e-01
3.69032979e-01 1.23346484e+00 1.15012658e+00 -4.48990911e-01
7.66079605e-01 5.61628163e-01 1.19667485e-01 -4.79163826e-01
-1.14641023e+00 -6.99827492e-01 -9.03057754e-01 1.66995853e-01
1.07682967e+00 -1.00053132e+00 -1.05042410e+00 7.84723163e-01
-8.28002214e-01 -3.05475593e-01 -1.86304927e-01 3.27200323e-01
-4.59356248e-01 1.62540618e-02 -8.28976572e-01 -4.72781628e-01
-7.38198400e-01 -1.44949281e+00 1.33507431e+00 1.13801897e+00
2.68746763e-01 -7.09989309e-01 2.31863722e-01 1.11949766e+00
3.29774767e-01 7.83733130e-01 5.04254878e-01 -4.40337777e-01
-6.73138857e-01 -1.22034073e-01 -8.82310271e-01 2.84404814e-01
9.80786681e-02 4.89959151e-01 -1.33963132e+00 -3.13485801e-01
-6.97577834e-01 -4.43937272e-01 9.95942414e-01 8.04440022e-01
1.47302341e+00 -2.35517859e-01 -2.73434669e-01 1.61973941e+00
1.60998785e+00 1.34165064e-02 8.95156384e-01 2.92438895e-01
7.13806748e-01 5.44718683e-01 2.26873100e-01 1.04421489e-01
2.17255905e-01 5.87705851e-01 3.42322230e-01 -1.11418140e+00
-5.62990606e-01 1.58717837e-02 -3.87155533e-01 1.46847546e-01
-5.62053621e-01 6.75322488e-02 -1.06278396e+00 7.95942783e-01
-1.52977097e+00 -8.19912195e-01 -3.45279336e-01 2.17831779e+00
1.19512880e+00 -4.00573879e-01 3.34579527e-01 -2.36578196e-01
6.26427233e-01 -2.13831663e-01 -6.66531086e-01 -4.52061534e-01
-2.11776629e-01 8.76248658e-01 8.01689923e-01 4.48435187e-01
-1.58533847e+00 1.30107176e+00 5.84599638e+00 6.91320360e-01
-1.37621570e+00 3.48837376e-02 1.00552785e+00 -6.05770610e-02
3.81399930e-01 -2.80457795e-01 -8.83146584e-01 1.85425654e-01
8.04987133e-01 6.21799111e-01 4.74071860e-01 4.69410062e-01
5.66644296e-02 7.72215128e-02 -4.49957401e-01 1.10138953e+00
-1.07469819e-01 -1.39624751e+00 -3.54766607e-01 5.18035963e-02
1.13334894e+00 4.51475352e-01 4.18480635e-01 -7.98508748e-02
4.56726313e-01 -1.62551582e+00 -2.40838513e-01 7.48603940e-01
1.38899195e+00 -8.50790858e-01 9.56848323e-01 -1.23539522e-01
-8.21189940e-01 8.60912427e-02 -4.52675581e-01 3.60789686e-01
-4.19132262e-01 4.07994598e-01 -7.94873953e-01 5.82782805e-01
6.88607037e-01 8.14808309e-01 -8.37996006e-01 1.52064705e+00
-7.43315322e-03 5.65458894e-01 -3.42729658e-01 7.20556557e-01
4.10090029e-01 -4.46382284e-01 2.95671135e-01 9.97819424e-01
5.79764806e-02 1.45591691e-01 1.05365135e-01 1.04139268e+00
-1.85520872e-01 3.38807166e-01 -2.65291840e-01 4.37617227e-02
-9.16420296e-02 1.39208817e+00 -6.76451921e-01 -1.62833065e-01
-3.89317900e-01 8.06665659e-01 -8.65485072e-02 3.59602720e-01
-6.11492753e-01 -4.26856160e-01 8.85613203e-01 9.12308469e-02
-1.84152834e-02 5.61877728e-01 -4.88551468e-01 -1.03202665e+00
-3.51711005e-01 -1.05775476e+00 3.34590644e-01 -6.99311435e-01
-1.10162115e+00 7.77527392e-01 -5.08650780e-01 -1.05716324e+00
2.25758851e-02 -6.97103739e-01 -6.18324816e-01 1.40935671e+00
-1.91222179e+00 -1.92386568e+00 -5.82812428e-01 7.91410565e-01
1.79727629e-01 -5.22992015e-01 6.80639625e-01 3.91190439e-01
-1.09924197e+00 9.44604039e-01 7.40921944e-02 6.09994948e-01
9.39408123e-01 -1.39348471e+00 4.36353683e-01 1.04392779e+00
-4.73762006e-02 7.04413652e-01 2.59039223e-01 -5.88616371e-01
-9.42999303e-01 -1.35427701e+00 3.77917200e-01 -4.08658117e-01
2.18590990e-01 1.69505969e-01 -4.23880696e-01 7.96504021e-01
3.48351419e-01 3.72111201e-01 6.01372242e-01 1.58861727e-01
-8.41332674e-02 -7.67845586e-02 -1.52126324e+00 3.47779542e-01
7.52241135e-01 -2.96693772e-01 -1.90008000e-01 9.36915040e-01
3.01385731e-01 -1.43710709e+00 -1.00561130e+00 6.04316473e-01
3.73610705e-01 -1.09171855e+00 1.16927469e+00 -6.16747439e-01
3.55059355e-01 -6.60228074e-01 4.70399499e-01 -8.56792450e-01
6.45635277e-02 -9.55767035e-01 2.37529784e-01 1.17123830e+00
4.20229137e-01 -4.98000532e-01 1.17109454e+00 3.17831248e-01
-1.44264787e-01 -7.07200527e-01 -5.07703424e-01 -4.68705922e-01
-5.80382813e-03 1.62691802e-01 9.75392044e-01 1.03942275e+00
-6.75438881e-01 -2.14499384e-01 -3.47313106e-01 5.09784102e-01
8.85136485e-01 3.64494324e-01 1.00775218e+00 -1.35253024e+00
-2.21295968e-01 -5.57072163e-01 -5.69792032e-01 -7.72016346e-01
-2.64426112e-01 -8.76262963e-01 -2.65866309e-01 -1.19881105e+00
1.26809239e-01 -7.22202361e-01 -1.80744156e-01 1.10483825e+00
1.14335241e-02 9.86434162e-01 -3.16231221e-01 1.89518765e-01
-2.06259564e-02 -2.23713458e-01 1.56848836e+00 -1.95988819e-01
-5.84857643e-01 3.49116683e-01 -6.47667229e-01 5.91960788e-01
8.03306520e-01 -5.96196204e-02 -2.53568679e-01 -3.56513470e-01
-1.26760930e-01 -7.84038287e-03 4.32607025e-01 -9.98265445e-01
2.85507172e-01 -8.39300379e-02 5.69180906e-01 -7.15369880e-01
-1.77753329e-01 -5.06263971e-01 5.61405793e-02 7.13501126e-02
-2.70413935e-01 -5.13179302e-01 4.83249843e-01 -6.72859251e-02
-2.38730967e-01 -3.96841094e-02 1.38184786e+00 8.81377682e-02
-4.65553343e-01 6.49630427e-01 5.25785148e-01 -1.75137445e-01
9.77844775e-01 -3.65065277e-01 -6.87204242e-01 3.98700088e-01
-8.78628850e-01 3.45422834e-01 7.54106760e-01 2.83553839e-01
2.54310817e-01 -8.59234035e-01 -8.70644331e-01 7.28914440e-01
3.54876339e-01 2.56513268e-01 4.45843726e-01 1.09586239e+00
-1.11394227e+00 5.02546072e-01 -4.15058315e-01 -8.92215967e-01
-1.52914703e+00 4.93601859e-01 9.52096283e-01 -1.57462396e-02
-7.33102262e-01 1.24348438e+00 -4.31607366e-02 -9.27703500e-01
3.99084479e-01 -6.00866437e-01 -4.57409620e-01 -3.29614639e-01
7.65113533e-01 2.63108667e-02 2.52231836e-01 -7.92163551e-01
-2.89516728e-02 1.11680591e+00 -8.14425126e-02 5.11097670e-01
1.05505192e+00 -4.10983562e-02 -4.92179692e-01 -5.55349231e-01
7.00185001e-01 -1.74484044e-01 -1.01485646e+00 -4.50675726e-01
-4.20692742e-01 -5.54056346e-01 4.82210249e-01 -1.40773654e+00
-1.46225345e+00 7.57190526e-01 1.36587977e+00 -3.88663173e-01
1.52424347e+00 -3.81584913e-01 8.85249734e-01 -1.92326933e-01
-1.87678337e-01 -8.69070113e-01 -7.44608462e-01 1.96997762e-01
3.75440568e-01 -1.61340082e+00 -1.05521828e-01 -3.89071584e-01
-4.36396778e-01 1.14001894e+00 6.92305088e-01 5.75916357e-02
4.87576306e-01 5.20573556e-02 5.96976101e-01 -4.45286810e-01
-3.71277821e-03 -7.52956331e-01 1.00591266e+00 7.88575232e-01
6.92310154e-01 1.19788505e-01 5.76629639e-02 9.74702239e-02
-4.17186767e-01 3.46714258e-01 4.63696063e-01 1.76076725e-01
-5.22584468e-02 -1.15209842e+00 -5.80895960e-01 6.97219491e-01
-7.24267840e-01 -2.01243922e-01 -3.23061347e-01 6.64987147e-01
5.74346006e-01 7.26803601e-01 2.42495552e-01 -1.61838517e-01
7.16250092e-02 -1.49132997e-01 4.21392381e-01 -6.13342583e-01
-1.06375957e+00 2.95801699e-01 -2.62392610e-01 -5.58604360e-01
-6.21635139e-01 -4.01455969e-01 -9.04914200e-01 -5.48084497e-01
-1.97440952e-01 -3.54330271e-01 6.17087543e-01 9.16068792e-01
2.46422336e-01 4.78433460e-01 3.72870922e-01 -3.80125672e-01
-2.62304604e-01 -1.00251985e+00 -6.57798231e-01 3.19429636e-01
7.37348437e-01 -2.98853189e-01 6.75514787e-02 2.44378805e-01] | [3.747234344482422, -3.629610300064087] |
d55b5cdb-d029-47be-a033-873d3b66fadf | navgpt-explicit-reasoning-in-vision-and | 2305.16986 | null | https://arxiv.org/abs/2305.16986v2 | https://arxiv.org/pdf/2305.16986v2.pdf | NavGPT: Explicit Reasoning in Vision-and-Language Navigation with Large Language Models | Trained with an unprecedented scale of data, large language models (LLMs) like ChatGPT and GPT-4 exhibit the emergence of significant reasoning abilities from model scaling. Such a trend underscored the potential of training LLMs with unlimited language data, advancing the development of a universal embodied agent. In this work, we introduce the NavGPT, a purely LLM-based instruction-following navigation agent, to reveal the reasoning capability of GPT models in complex embodied scenes by performing zero-shot sequential action prediction for vision-and-language navigation (VLN). At each step, NavGPT takes the textual descriptions of visual observations, navigation history, and future explorable directions as inputs to reason the agent's current status, and makes the decision to approach the target. Through comprehensive experiments, we demonstrate NavGPT can explicitly perform high-level planning for navigation, including decomposing instruction into sub-goal, integrating commonsense knowledge relevant to navigation task resolution, identifying landmarks from observed scenes, tracking navigation progress, and adapting to exceptions with plan adjustment. Furthermore, we show that LLMs is capable of generating high-quality navigational instructions from observations and actions along a path, as well as drawing accurate top-down metric trajectory given the agent's navigation history. Despite the performance of using NavGPT to zero-shot R2R tasks still falling short of trained models, we suggest adapting multi-modality inputs for LLMs to use as visual navigation agents and applying the explicit reasoning of LLMs to benefit learning-based models. | ['Qi Wu', 'Yicong Hong', 'Gengze Zhou'] | 2023-05-26 | null | null | null | null | ['instruction-following', 'vision-and-language-navigation', 'visual-navigation'] | ['natural-language-processing', 'robots', 'robots'] | [ 4.67731357e-02 4.03010398e-01 2.11555269e-02 -1.71239913e-01
-6.97470248e-01 -4.41652715e-01 8.66035581e-01 6.49496391e-02
-5.31831920e-01 2.35258356e-01 7.63369143e-01 -6.64864898e-01
-4.59503904e-02 -9.12266135e-01 -7.47783065e-01 -3.18392426e-01
-2.45932981e-01 8.26453984e-01 1.40355870e-01 -6.70613110e-01
5.61743081e-01 3.71908695e-01 -1.48921049e+00 3.43035460e-01
5.98133028e-01 6.75254822e-01 7.19067574e-01 8.21246266e-01
-2.21117318e-01 1.57331812e+00 -1.02399580e-01 4.83038202e-02
1.26422241e-01 -4.31108713e-01 -9.09549952e-01 -1.54983476e-01
1.69256344e-01 -6.85022295e-01 -6.38781667e-01 6.48662448e-01
2.60732293e-01 8.95573914e-01 7.38400042e-01 -1.14317322e+00
-1.12325871e+00 6.35743380e-01 1.42729163e-01 1.05864778e-01
8.70485365e-01 7.56413460e-01 9.05593216e-01 -8.84133279e-01
9.31697488e-01 1.55282772e+00 3.79200250e-01 9.87888038e-01
-9.70684409e-01 -1.58576742e-01 5.24759173e-01 4.13266182e-01
-8.99117529e-01 -5.55580378e-01 4.03219968e-01 -3.90574962e-01
1.65814924e+00 -1.28759503e-01 7.87198901e-01 1.40332103e+00
4.24340129e-01 1.02130163e+00 9.88351226e-01 -3.53478402e-01
6.05040610e-01 -5.32543361e-01 -1.30861282e-01 1.30844760e+00
-4.16926980e-01 5.27576387e-01 -8.55514884e-01 1.74665496e-01
8.25506866e-01 -1.33558750e-01 -6.45838752e-02 -4.94428605e-01
-1.52567208e+00 6.40172541e-01 8.83682489e-01 1.58395395e-02
-5.36108315e-01 5.10537744e-01 3.18384618e-01 2.07194641e-01
-2.14788333e-01 8.32906067e-01 -2.47133479e-01 -4.06341344e-01
-3.88571262e-01 2.77014345e-01 5.81810474e-01 1.09946597e+00
6.54835165e-01 3.26488674e-01 -3.14439356e-01 3.55544448e-01
3.84884089e-01 5.61879635e-01 7.49538720e-01 -1.54397619e+00
4.32996899e-01 5.65222085e-01 8.41994956e-02 -7.78665245e-01
-8.17991436e-01 4.56597507e-02 -2.03862458e-01 5.65962136e-01
1.52718559e-01 2.21731782e-01 -1.31143379e+00 1.94198275e+00
2.89382450e-02 -2.89344698e-01 6.23980880e-01 7.89698243e-01
8.07465374e-01 8.26023459e-01 4.58558679e-01 2.77308315e-01
1.24774921e+00 -1.32683396e+00 -3.94133538e-01 -7.97178149e-01
1.09134877e+00 4.21624742e-02 1.55611217e+00 3.31614524e-01
-1.06557786e+00 -5.71332216e-01 -9.30656314e-01 -4.84450042e-01
-6.12045944e-01 -2.79287219e-01 7.66955495e-01 -1.64012924e-01
-1.44340658e+00 3.81987691e-01 -1.05714715e+00 -8.35355341e-01
4.34405506e-01 1.13842964e-01 -5.86375058e-01 -2.95615435e-01
-8.53214204e-01 1.46745753e+00 4.45068449e-01 4.88411747e-02
-1.59561849e+00 -2.97530740e-01 -1.37076759e+00 1.16100311e-02
4.62280869e-01 -8.89270782e-01 1.47954404e+00 -3.71270716e-01
-1.69600129e+00 6.94853365e-01 -2.81642437e-01 -6.14887893e-01
1.49792910e-01 -9.92374569e-02 -2.03658953e-01 4.24021011e-04
3.69138151e-01 1.21057105e+00 3.98625880e-01 -1.12219059e+00
-5.86263239e-01 -3.45024168e-01 3.69763851e-01 7.46711552e-01
3.06531787e-01 -4.69474375e-01 -3.99020731e-01 2.63017090e-03
3.90327275e-01 -8.63105595e-01 -5.55822730e-01 1.03148958e-02
-1.97118536e-01 -1.97532415e-01 3.72006714e-01 -6.62137091e-01
5.98953664e-01 -2.11789203e+00 5.61286688e-01 -2.08339587e-01
7.09261075e-02 -2.47379214e-01 -6.05066717e-01 5.79644024e-01
4.60664213e-01 -4.33358923e-02 4.60739024e-02 -4.83997941e-01
2.02902421e-01 5.32486796e-01 -6.25844538e-01 -6.96856678e-02
-6.61727935e-02 1.39017951e+00 -1.28958154e+00 -3.41655999e-01
4.66204047e-01 2.26756305e-01 -6.62139297e-01 3.52454841e-01
-7.00926304e-01 5.71250498e-01 -4.58933026e-01 5.64165592e-01
-2.14012414e-01 -2.05958411e-01 1.42673613e-03 1.17828444e-01
-1.62851632e-01 4.89006490e-01 -5.16816080e-01 2.19496226e+00
-7.77069211e-01 7.17942178e-01 -2.73980528e-01 -3.71494353e-01
6.18447125e-01 2.27272194e-02 -1.20995887e-01 -1.48196352e+00
-6.89401403e-02 4.83348072e-02 -5.55237308e-02 -7.44742036e-01
6.57998800e-01 -1.22519940e-01 -3.07994336e-01 5.08156598e-01
2.37802312e-01 -3.40819716e-01 1.36537656e-01 4.15681571e-01
1.31768179e+00 7.81738162e-01 5.65889478e-01 1.68936089e-01
1.30647957e-01 5.36457300e-01 3.88576202e-02 1.05718398e+00
-2.80090928e-01 2.65851647e-01 1.27596051e-01 -5.98661602e-01
-8.90387833e-01 -1.22022271e+00 5.43543279e-01 1.59685957e+00
1.72713384e-01 -4.25622255e-01 -5.63729346e-01 -3.14355522e-01
-3.35142434e-01 1.50165701e+00 -8.92281294e-01 -2.79513925e-01
-6.60776675e-01 -2.03878105e-01 4.61746216e-01 7.24456787e-01
5.93497515e-01 -1.65640461e+00 -1.28332782e+00 2.57133782e-01
-2.65107363e-01 -1.05572307e+00 -2.21991420e-01 5.65936923e-01
-7.04813957e-01 -8.84603322e-01 -2.25071043e-01 -8.91589940e-01
7.83419192e-01 2.20466077e-01 1.11533666e+00 1.18353732e-01
5.66907674e-02 8.48022282e-01 -2.73262560e-01 -6.62846789e-02
-7.11061716e-01 -1.96100399e-01 -2.62596807e-03 -8.99799049e-01
1.55287638e-01 -4.53637838e-01 -4.93183643e-01 9.95829478e-02
-4.13873881e-01 6.80641413e-01 6.43125236e-01 6.76466465e-01
5.09882748e-01 -3.36147577e-01 2.50391096e-01 -4.25178856e-01
7.33023822e-01 -4.16333497e-01 -4.43590909e-01 1.63911477e-01
-5.86095452e-01 4.67327058e-01 5.31364322e-01 -3.43363702e-01
-1.25555098e+00 -2.50224799e-01 -1.41556248e-01 -1.21950254e-01
-2.88106292e-01 6.76978648e-01 -3.95504991e-03 4.94518951e-02
9.17621553e-01 7.43116975e-01 -1.01013876e-01 3.26820835e-02
8.62119555e-01 -1.68656353e-02 8.96232665e-01 -7.35855877e-01
3.75429451e-01 4.29122925e-01 3.93217020e-02 -3.80209208e-01
-7.87719727e-01 -1.91463336e-01 -4.60619152e-01 -1.22630961e-01
1.09338796e+00 -9.28740561e-01 -1.00214362e+00 1.55352369e-01
-9.83964801e-01 -1.22768223e+00 -4.04549092e-01 2.43244305e-01
-1.28861165e+00 -3.56094167e-02 -7.18113542e-01 -6.96991622e-01
-2.88040321e-02 -1.35844564e+00 9.05804515e-01 1.97547585e-01
-5.59703529e-01 -1.06438565e+00 2.71016330e-01 3.73126745e-01
4.38603967e-01 -1.28474221e-01 1.25918913e+00 -6.49741054e-01
-8.98442924e-01 7.57718319e-03 3.36088799e-02 -4.54897285e-01
-1.65190518e-01 -4.85972613e-01 -6.51986837e-01 1.61716447e-03
-1.09074108e-01 -7.47363985e-01 6.68169022e-01 2.52695054e-01
8.61078382e-01 -2.21068799e-01 -3.79422247e-01 7.74645627e-01
1.29438615e+00 5.65317452e-01 6.52553320e-01 8.05534244e-01
6.70066595e-01 4.58986759e-01 7.36012459e-01 2.24517044e-02
1.01816857e+00 4.04817790e-01 8.51307690e-01 3.88739944e-01
-2.32143775e-01 -9.01661873e-01 7.20900953e-01 5.08309901e-01
-6.75879568e-02 -3.67604822e-01 -1.10062289e+00 3.81351173e-01
-1.92935002e+00 -1.11919606e+00 5.65850675e-01 1.84417748e+00
4.89179045e-01 2.21865028e-01 -3.52847546e-01 -6.75178885e-01
-2.53214955e-01 3.75298351e-01 -9.44769025e-01 -6.55791461e-01
3.08509618e-02 -1.95394203e-01 2.79663742e-01 7.73884773e-01
-6.08445525e-01 1.52315772e+00 6.55850363e+00 3.95932525e-01
-7.25461662e-01 3.04683391e-02 2.61654973e-01 1.62350405e-02
-3.76028985e-01 1.31882623e-01 -6.13661706e-01 -1.30284891e-01
1.03501892e+00 6.54650480e-02 9.71363008e-01 8.44078302e-01
1.49778575e-01 -3.41820985e-01 -1.35753894e+00 8.27642143e-01
2.43861079e-01 -1.46811247e+00 4.30374473e-01 -4.97749262e-02
4.23663825e-01 3.47472042e-01 1.66522384e-01 1.06144845e+00
1.00861561e+00 -1.36390769e+00 1.01396120e+00 5.56564510e-01
4.10152227e-01 -2.88313121e-01 1.93509191e-01 7.00262904e-01
-1.06389868e+00 -4.64587003e-01 -1.88746035e-01 -4.04700071e-01
3.71228904e-01 -6.10940754e-01 -1.10297585e+00 1.94267794e-01
3.77313912e-01 7.39551008e-01 -4.89142776e-01 5.50101399e-01
-5.14161527e-01 -1.54500594e-02 -2.33160090e-02 -1.44903198e-01
6.15707338e-01 -1.49939224e-01 4.47924018e-01 7.55069613e-01
5.38242042e-01 3.80139172e-01 3.20219576e-01 9.33750033e-01
2.62174636e-01 -2.83952892e-01 -8.79700482e-01 -6.74119964e-02
4.31546926e-01 8.61741602e-01 -6.69377267e-01 -3.81485462e-01
-2.82634616e-01 1.11242366e+00 7.66670585e-01 7.11652219e-01
-6.14526153e-01 3.68979245e-01 6.28187001e-01 -2.39130557e-01
6.78986311e-02 -6.58675790e-01 -1.69701308e-01 -9.83212233e-01
-4.44079876e-01 -9.10497546e-01 2.45516092e-01 -1.64663064e+00
-5.57049632e-01 7.32757926e-01 -1.26630798e-01 -9.68635023e-01
-6.47464633e-01 -9.31519926e-01 -6.28337204e-01 6.62254632e-01
-1.23759687e+00 -1.35828090e+00 -3.97378981e-01 5.36010087e-01
1.04837608e+00 -1.96754277e-01 1.20362449e+00 -6.70426309e-01
-2.54844874e-01 -2.71233451e-02 -4.23228234e-01 -3.57438400e-02
2.77924627e-01 -1.30861163e+00 8.93994987e-01 6.82713866e-01
2.18033522e-01 7.86942840e-01 9.68071997e-01 -7.82569170e-01
-1.62579286e+00 -6.19176984e-01 4.50667322e-01 -9.95388746e-01
7.83076704e-01 -1.92759603e-01 -7.29817688e-01 1.27519143e+00
2.06638709e-01 -3.17378312e-01 4.69169021e-01 7.58062229e-02
-4.67831284e-01 6.19675338e-01 -9.42760050e-01 1.50919485e+00
1.60424781e+00 -7.83398330e-01 -1.10327947e+00 2.92670071e-01
8.88675869e-01 -7.58008897e-01 -3.80423784e-01 -2.14490276e-02
4.47202176e-01 -8.74804854e-01 1.08017755e+00 -8.47426474e-01
5.46984136e-01 -2.73455024e-01 -4.72055674e-01 -1.53019679e+00
-5.73079348e-01 -3.29803854e-01 -2.10114136e-01 4.46232915e-01
3.94972712e-01 -3.89808118e-01 5.71805000e-01 7.90734172e-01
-5.22166610e-01 -6.42775834e-01 -8.13453496e-01 -3.84727925e-01
-1.68386504e-01 -7.36087382e-01 4.10212576e-01 3.57385844e-01
2.77591318e-01 3.18078071e-01 -8.84990245e-02 2.90773273e-01
4.03973013e-01 -4.20704558e-02 8.54080081e-01 -7.77037680e-01
-1.71376720e-01 -5.22120476e-01 -2.34259278e-01 -1.42372429e+00
4.29993749e-01 -1.01659131e+00 6.07829452e-01 -2.24854350e+00
4.94642071e-02 -1.72977656e-01 -2.80797213e-01 8.16219091e-01
1.77064836e-01 -4.28659886e-01 3.45140696e-01 2.09426537e-01
-8.76904786e-01 6.11979961e-01 1.51671469e+00 -2.29335755e-01
-2.93402165e-01 -4.98453617e-01 -6.49316430e-01 8.99132431e-01
5.67517698e-01 -4.79576960e-02 -8.15396667e-01 -7.49034703e-01
5.46992481e-01 2.98065186e-01 6.33999586e-01 -1.13542056e+00
5.67846298e-01 -4.86543834e-01 4.13638353e-01 -5.26309013e-01
5.77176929e-01 -6.34715915e-01 -1.06957935e-01 5.09206474e-01
-6.99368298e-01 3.37411284e-01 4.14352655e-01 7.22100496e-01
1.39823750e-01 -5.06376885e-02 2.30611935e-01 -6.75448477e-01
-1.82616842e+00 1.66253790e-01 -8.32187235e-01 -4.96302098e-02
7.65005946e-01 -4.70787108e-01 -7.45076835e-01 -4.90496635e-01
-8.85916889e-01 6.73460662e-01 8.20274711e-01 5.00166416e-01
1.01852810e+00 -1.14929187e+00 -1.12886518e-01 1.82783946e-01
4.14842486e-01 3.12309302e-02 4.83891666e-01 5.85269332e-01
-4.53433275e-01 5.37303030e-01 -5.36625803e-01 -3.39984357e-01
-6.61442399e-01 8.49650264e-01 4.80253190e-01 -1.38113260e-01
-1.01768553e+00 9.29465234e-01 5.35434067e-01 -8.37905765e-01
1.80515811e-01 -6.09458804e-01 -4.37458575e-01 -3.41476679e-01
5.49199283e-01 -5.38265891e-02 -3.92751843e-01 -5.33503950e-01
-2.89818048e-01 4.17652249e-01 5.84817678e-02 -5.18362880e-01
1.13104844e+00 -3.45082104e-01 1.32107586e-01 7.21303463e-01
7.09073961e-01 -2.61538088e-01 -1.78957212e+00 -5.93883954e-02
-5.93651198e-02 6.01848550e-02 -1.47382915e-01 -1.11891270e+00
-3.58064771e-01 9.12763238e-01 2.66201258e-01 -3.54983062e-01
7.03548312e-01 4.63763401e-02 5.66396832e-01 1.02802527e+00
9.27944005e-01 -8.93147469e-01 6.59199655e-01 1.01715302e+00
1.03672636e+00 -1.44236362e+00 -3.47887844e-01 4.65410113e-01
-9.04731870e-01 1.06759226e+00 1.00610268e+00 1.33592263e-01
-5.34490198e-02 -2.27371510e-02 1.75974593e-01 -4.29709882e-01
-1.21599030e+00 -2.78983355e-01 1.95706099e-01 9.47760105e-01
-6.66350573e-02 1.09280974e-01 6.76267207e-01 2.19482258e-01
-4.01571333e-01 -1.14357341e-02 5.56464612e-01 1.12263501e+00
-8.12968612e-01 -3.95496100e-01 -2.46377196e-02 2.83973038e-01
3.65567535e-01 -3.76807898e-01 -2.41129640e-02 7.92601466e-01
-1.63665086e-01 7.56560385e-01 9.15013254e-02 -3.90155405e-01
3.34180743e-01 2.23186135e-01 5.67408085e-01 -8.77501667e-01
-2.58401334e-01 -5.44692397e-01 1.87172681e-01 -1.08142614e+00
-7.17037097e-02 -5.19030750e-01 -1.98948181e+00 -1.10994287e-01
4.71374661e-01 -2.65199929e-01 5.09461105e-01 1.20303428e+00
3.74518156e-01 7.30558157e-01 -2.89743245e-01 -1.13489497e+00
-4.12086815e-01 -7.17323661e-01 4.59148884e-02 3.48912656e-01
4.62937176e-01 -7.68894911e-01 -3.03249031e-01 -1.48478433e-01] | [4.412583351135254, 0.6926828026771545] |
103ab0db-01ae-4ee8-9a54-b4ff461109ba | evolutionary-preference-learning-via-graph | 2206.12779 | null | https://arxiv.org/abs/2206.12779v2 | https://arxiv.org/pdf/2206.12779v2.pdf | Evolutionary Preference Learning via Graph Nested GRU ODE for Session-based Recommendation | Session-based recommendation (SBR) aims to predict the user next action based on the ongoing sessions. Recently, there has been an increasing interest in modeling the user preference evolution to capture the fine-grained user interests. While latent user preferences behind the sessions drift continuously over time, most existing approaches still model the temporal session data in discrete state spaces, which are incapable of capturing the fine-grained preference evolution and result in sub-optimal solutions. To this end, we propose Graph Nested GRU ordinary differential equation (ODE), namely GNG-ODE, a novel continuum model that extends the idea of neural ODEs to continuous-time temporal session graphs. The proposed model preserves the continuous nature of dynamic user preferences, encoding both temporal and structural patterns of item transitions into continuous-time dynamic embeddings. As the existing ODE solvers do not consider graph structure change and thus cannot be directly applied to the dynamic graph, we propose a time alignment technique, called t-Alignment, to align the updating time steps of the temporal session graphs within a batch. Empirical results on three benchmark datasets show that GNG-ODE significantly outperforms other baselines. | ['Sunghun Kim', 'Yan Zhang', 'Xing Xie', 'Chaozhuo Li', 'Peiyan Zhang', 'Jiayan Guo'] | 2022-06-26 | null | null | null | null | ['session-based-recommendations'] | ['miscellaneous'] | [-1.95252284e-01 -4.55283642e-01 -4.73497003e-01 -4.67136294e-01
1.09584503e-01 -7.47612178e-01 2.76706725e-01 3.74040276e-01
-3.87562513e-02 3.34130347e-01 5.42185247e-01 -4.18275774e-01
-6.46889985e-01 -6.50708199e-01 -4.31474686e-01 -5.84349155e-01
-4.62978452e-01 4.33545351e-01 1.44677669e-01 -4.08204347e-01
1.28543407e-01 1.80357084e-01 -1.03346050e+00 3.80613133e-02
1.22901094e+00 8.35741043e-01 1.02919340e-01 6.01184249e-01
-1.71606466e-01 3.64355057e-01 2.06018850e-01 -2.36789659e-01
2.56212711e-01 -4.82197315e-01 -5.36472440e-01 2.10621491e-01
2.00762361e-01 -2.26587996e-01 -9.43480194e-01 7.13634193e-01
2.86871850e-01 8.43546093e-01 5.73735058e-01 -1.42616773e+00
-1.21865547e+00 5.56525826e-01 -5.04878521e-01 3.57482135e-01
3.77457350e-01 -1.31566495e-01 1.41635740e+00 -4.16266054e-01
4.87280667e-01 1.15682209e+00 5.61744809e-01 7.27790594e-01
-1.46620464e+00 -8.76652151e-02 1.07937968e+00 4.44804430e-01
-9.92035031e-01 2.69642055e-01 1.12469900e+00 -4.62554604e-01
8.58108997e-01 4.01544034e-01 9.36267614e-01 1.32831955e+00
2.45319977e-01 8.39093029e-01 8.46136391e-01 3.43497485e-01
3.09252590e-01 -1.20548591e-01 7.21899569e-01 5.08646667e-01
7.67208487e-02 -7.81524330e-02 -5.87325454e-01 -3.99994552e-01
8.45885456e-01 6.25383973e-01 -4.35231864e-01 -8.54556561e-01
-9.45935011e-01 7.48820186e-01 5.76609194e-01 2.43971452e-01
-3.87598068e-01 -8.54323059e-02 3.10902417e-01 6.89948678e-01
5.98759413e-01 2.74177998e-01 -6.65874302e-01 -9.74747762e-02
-6.68912053e-01 3.11198086e-01 9.22213554e-01 8.23602796e-01
7.16659367e-01 1.34667633e-02 -2.74164706e-01 6.81588769e-01
5.89005232e-01 -4.56073731e-02 8.08779597e-01 -4.60017145e-01
2.62222439e-01 9.63055193e-01 1.24367252e-01 -1.24621999e+00
-3.45371604e-01 -6.53215647e-01 -9.11681294e-01 -5.89700699e-01
3.51097047e-01 4.40790243e-02 -9.68683183e-01 1.79002619e+00
3.44834596e-01 7.12729335e-01 -2.01682225e-01 9.59737360e-01
5.37455797e-01 9.11402285e-01 -5.44876941e-02 -3.73438030e-01
7.93882072e-01 -1.06455112e+00 -7.06263185e-01 -3.84563595e-01
5.41702688e-01 -1.87461257e-01 1.29665399e+00 2.95961294e-02
-8.89768660e-01 -3.94355476e-01 -7.89359570e-01 7.52399638e-02
-2.92689174e-01 -4.68216747e-01 6.90309823e-01 2.92529464e-01
-1.16920626e+00 9.38979268e-01 -9.13642466e-01 -4.91654098e-01
-7.88927544e-03 5.11812806e-01 -4.80204001e-02 -1.41150624e-01
-1.40911448e+00 4.05283809e-01 4.66550067e-02 3.50566387e-01
-4.43228006e-01 -1.01494217e+00 -7.84369469e-01 2.17122123e-01
5.28396130e-01 -5.31377912e-01 1.08774638e+00 -7.41212785e-01
-1.55290842e+00 3.44885260e-01 -3.95902157e-01 -3.18427145e-01
4.81378257e-01 1.37260705e-01 -7.23934591e-01 -6.39050245e-01
-4.80807841e-01 -1.92661300e-01 6.89840078e-01 -1.01420522e+00
-4.75364834e-01 -5.76687396e-01 2.23058715e-01 4.18267757e-01
-7.97441363e-01 -4.74850237e-01 -7.04956710e-01 -6.66781008e-01
7.27066025e-02 -1.26440752e+00 -4.45005447e-01 -3.05485815e-01
4.89184354e-03 -4.94488239e-01 8.76180410e-01 -6.30729556e-01
1.97887337e+00 -1.97580838e+00 7.92481661e-01 2.68028527e-01
2.81724691e-01 1.41728833e-01 -2.91130871e-01 6.79325342e-01
7.11729527e-02 2.28870083e-02 -1.09618440e-01 -3.65856737e-01
3.45490754e-01 6.19673073e-01 -3.07751447e-01 3.59940171e-01
-4.43424612e-01 1.17577863e+00 -1.15014303e+00 2.86927372e-02
-1.63990986e-02 4.31570828e-01 -8.77801836e-01 3.93572360e-01
-2.97127575e-01 5.82521617e-01 -7.47719586e-01 1.45607024e-01
6.06905282e-01 -5.28820693e-01 5.76069951e-01 -1.36395469e-01
5.07365949e-02 3.43210071e-01 -1.18585193e+00 1.77922297e+00
-3.71289492e-01 2.88576841e-01 -1.44713208e-01 -9.81977284e-01
7.79272377e-01 1.45320967e-01 8.40380251e-01 -8.71911466e-01
-9.55699310e-02 4.68026698e-02 9.01073888e-02 -4.55897033e-01
6.68729246e-01 3.58979791e-01 -2.62949497e-01 5.29096723e-01
-1.10521175e-01 5.44508040e-01 2.16814399e-01 4.81353492e-01
1.04124331e+00 1.40678108e-01 -2.41926615e-03 -2.83590555e-01
4.25247401e-01 -3.58064175e-01 8.26388538e-01 4.46075916e-01
-2.72240609e-01 4.16875631e-01 5.70650041e-01 -4.70316470e-01
-8.21319163e-01 -1.05682957e+00 3.46271753e-01 1.34890628e+00
4.34164882e-01 -6.01463854e-01 -3.88183028e-01 -7.01154470e-01
3.04741681e-01 5.59330583e-01 -9.59415197e-01 -5.20484507e-01
-6.39790952e-01 -8.53745222e-01 -3.65383834e-01 5.18655419e-01
1.19040616e-01 -8.98746490e-01 6.71659317e-03 5.64096451e-01
-2.21015930e-01 -7.54729807e-01 -1.46956778e+00 -1.90927267e-01
-1.00020289e+00 -8.01832676e-01 -7.09790409e-01 -6.86443686e-01
6.06190979e-01 3.78770232e-01 8.55571985e-01 -1.06325969e-01
-1.04363389e-01 8.26597691e-01 -4.74010348e-01 2.73170680e-01
4.13528420e-02 8.46271366e-02 2.52448231e-01 6.99176967e-01
3.04326594e-01 -8.94647539e-01 -1.03338432e+00 4.42594409e-01
-8.75767827e-01 -9.20808315e-02 1.53913349e-01 6.80994153e-01
6.19883299e-01 -1.45230725e-01 4.86433327e-01 -1.22068560e+00
1.03899503e+00 -8.35240185e-01 -2.23142371e-01 5.00484943e-01
-1.14740384e+00 1.87498435e-01 6.65205538e-01 -9.46176410e-01
-1.00779092e+00 -2.95751870e-01 1.02732256e-01 -5.78835726e-01
2.04330638e-01 9.41982150e-01 -1.39831096e-01 2.07646221e-01
2.71017015e-01 3.16064239e-01 -2.79782265e-01 -6.94272101e-01
3.89500588e-01 4.73379582e-01 2.52236873e-01 -4.96826828e-01
6.01088703e-01 2.02004820e-01 -2.59499043e-01 -4.88437384e-01
-7.80859768e-01 -8.61453950e-01 -5.65376401e-01 -1.28769428e-01
5.48968375e-01 -4.80098754e-01 -4.92041469e-01 5.40326893e-01
-5.91974974e-01 -8.26921165e-01 -1.96640313e-01 2.74148941e-01
-4.25762922e-01 5.75544477e-01 -7.32808769e-01 -7.08188772e-01
-9.34508666e-02 -7.56769776e-01 5.76027930e-01 4.28617805e-01
-1.22096442e-01 -1.68785429e+00 3.98279458e-01 -7.08072409e-02
4.94359702e-01 2.26339087e-01 9.04838741e-01 -5.27271271e-01
-3.77139270e-01 -2.53859043e-01 1.00249805e-01 -4.84845825e-02
3.51549000e-01 -1.44284904e-01 -3.14818293e-01 -6.47899568e-01
-2.13408172e-01 2.22858250e-01 6.86471701e-01 3.93802136e-01
1.28361392e+00 -4.78400081e-01 -3.50195378e-01 6.01132870e-01
1.36875713e+00 3.15941811e-01 4.57499564e-01 -1.64105911e-02
1.06215668e+00 5.17488718e-01 5.30991793e-01 4.28494751e-01
8.31551135e-01 8.88642490e-01 3.02638054e-01 2.23384395e-01
2.56707162e-01 -6.32289767e-01 5.41043341e-01 1.21601713e+00
-2.07523063e-01 -5.70890248e-01 -4.96331275e-01 6.99173570e-01
-2.59410930e+00 -7.95293868e-01 -4.56064939e-01 2.34031844e+00
4.27863717e-01 7.99089205e-03 4.51613784e-01 -2.63841718e-01
7.15614438e-01 3.85699958e-01 -1.10494506e+00 -5.83362699e-01
2.25872979e-01 -3.01676542e-01 4.73249823e-01 5.57942510e-01
-7.05488205e-01 6.76763654e-01 5.36150122e+00 2.84122169e-01
-1.06526721e+00 4.00831960e-02 2.71335453e-01 -1.33462757e-01
-6.24836564e-01 -4.31105457e-02 -5.20857334e-01 6.51325107e-01
9.93001521e-01 -6.63372159e-01 1.00434065e+00 4.55796808e-01
5.00273108e-01 7.19856143e-01 -1.30474591e+00 9.77107704e-01
2.64300760e-02 -9.37718332e-01 -7.80940875e-02 3.18281204e-01
9.03377891e-01 -1.50059298e-01 3.91084671e-01 4.95252311e-01
4.99824673e-01 -6.05949938e-01 1.93595991e-01 7.08449841e-01
3.98276657e-01 -4.36127752e-01 1.28545389e-01 3.18575114e-01
-1.62018561e+00 -4.21741396e-01 -3.96760970e-01 -3.40462513e-02
3.95900935e-01 2.90011793e-01 -3.06091338e-01 8.43216121e-01
6.62032366e-01 1.33716583e+00 -5.13367593e-01 1.17372024e+00
2.28462219e-01 8.20812821e-01 -5.44511080e-02 -1.37182912e-02
3.18785548e-01 -6.55044258e-01 6.55974448e-01 8.37842226e-01
5.31607151e-01 6.30717650e-02 3.08425158e-01 6.62464499e-01
1.41891912e-01 5.47071919e-02 -3.34244549e-01 -5.27645648e-01
8.14735517e-02 1.15023768e+00 -4.14452672e-01 9.36652794e-02
-6.58677578e-01 1.42520201e+00 3.94719541e-01 7.15529680e-01
-9.25771296e-01 4.54542451e-02 1.03050220e+00 2.42803887e-01
3.20542336e-01 -4.84842837e-01 1.28864273e-01 -1.56093478e+00
-1.12021184e-02 -6.25807345e-01 8.64340007e-01 -1.73506007e-01
-1.70036900e+00 4.31660473e-01 -1.92806721e-01 -1.24966860e+00
-1.36999831e-01 -4.84795034e-01 -8.76902461e-01 7.59963930e-01
-1.32332444e+00 -1.03393245e+00 -2.68341422e-01 8.51522446e-01
6.43000007e-01 2.92915374e-01 7.08881795e-01 4.41052258e-01
-7.85897374e-01 7.12098420e-01 4.43011373e-01 -3.16517115e-01
6.36822641e-01 -1.56807601e+00 6.52604342e-01 6.10253215e-01
9.88493860e-02 8.72932374e-01 8.50822866e-01 -6.98704422e-01
-1.68984950e+00 -1.29346478e+00 8.13469887e-01 -4.30530846e-01
9.92915332e-01 -5.32182157e-01 -1.41874278e+00 8.28185380e-01
-2.16975901e-02 6.51572421e-02 6.67306840e-01 4.14562285e-01
-2.81559110e-01 -1.30991980e-01 -7.17801929e-01 6.60605550e-01
1.58887458e+00 -5.29351354e-01 -4.23962235e-01 7.55274445e-02
8.53714287e-01 -3.49571854e-01 -9.01173651e-01 2.11704701e-01
5.24841368e-01 -4.81370062e-01 8.68021131e-01 -1.03096879e+00
-3.28695513e-02 -4.69223075e-02 6.12890348e-02 -1.76326406e+00
-8.47136557e-01 -9.19050336e-01 -9.22336698e-01 1.08258975e+00
2.61037946e-01 -7.78632820e-01 7.94357121e-01 1.02147949e+00
-1.31552964e-01 -9.22337294e-01 -5.38418412e-01 -6.55930161e-01
-1.35128036e-01 6.95101023e-02 7.91722655e-01 1.23941064e+00
2.21451417e-01 2.18842313e-01 -7.44268298e-01 2.15454757e-01
5.50827503e-01 5.24523973e-01 3.66425902e-01 -1.59930444e+00
-6.81027353e-01 -4.32443708e-01 -1.87998295e-01 -1.40111935e+00
4.49653603e-02 -9.30389464e-01 -2.19297230e-01 -1.96846843e+00
1.13495350e-01 -3.26961160e-01 -9.92272139e-01 1.38169855e-01
-3.24070275e-01 -3.34719926e-01 -1.29895896e-01 6.85491040e-02
-7.67603815e-01 6.79929674e-01 1.42049325e+00 -6.18312731e-02
-7.72253096e-01 2.30803773e-01 -8.71875823e-01 3.07614893e-01
6.61989927e-01 -1.61834747e-01 -9.93573427e-01 -3.60798895e-01
4.10688519e-01 9.67135280e-02 1.93619691e-02 -4.90594804e-01
3.92436057e-01 -6.07658565e-01 -2.30594099e-01 -3.54828089e-01
1.44374311e-01 -8.82449985e-01 4.02234167e-01 1.66824862e-01
-5.14857709e-01 1.34899005e-01 -1.38641940e-02 1.33209801e+00
8.32674280e-02 3.64668757e-01 3.15472066e-01 1.71913251e-01
-9.94406283e-01 1.22969675e+00 -1.89356923e-01 -1.60743579e-01
8.88388395e-01 -2.50745118e-01 6.29685149e-02 -4.26039457e-01
-1.22952652e+00 7.49912679e-01 4.29507256e-01 1.04686117e+00
3.56293947e-01 -1.55659699e+00 -2.71192759e-01 -1.29888104e-02
7.42601976e-02 -2.73041338e-01 8.09257686e-01 7.17884719e-01
2.43820742e-01 2.32325241e-01 3.15550878e-03 -2.20887080e-01
-1.19591975e+00 8.24785113e-01 3.87915879e-01 -6.75643325e-01
-6.22943401e-01 6.36434257e-01 3.58669549e-01 -5.82025528e-01
3.34992439e-01 -2.84247994e-01 -5.94294667e-01 2.15975165e-01
1.70412838e-01 4.60085630e-01 -2.11387977e-01 -5.76988399e-01
-1.41814485e-01 5.20185828e-01 -2.77196586e-01 2.44823605e-01
1.59828210e+00 -6.19954228e-01 3.83706056e-02 9.60819900e-01
1.30644798e+00 -2.22837165e-01 -1.54908347e+00 -5.20947814e-01
5.72722172e-03 -6.59242332e-01 1.56647280e-01 -5.08027136e-01
-1.09352767e+00 7.85005808e-01 5.62637269e-01 8.50846291e-01
1.03407967e+00 -3.12579632e-01 1.13319194e+00 7.86109120e-02
2.51355827e-01 -1.08170056e+00 2.22559929e-01 5.17683566e-01
7.04105735e-01 -7.80409157e-01 -4.51625109e-01 -3.49455088e-01
-9.12163496e-01 8.35620642e-01 8.28872621e-01 -2.83528328e-01
9.60167289e-01 -4.05195892e-01 -3.26516330e-01 -2.70766877e-02
-9.01993632e-01 -5.48944212e-02 7.88411677e-01 3.93147230e-01
3.98428798e-01 2.17257276e-01 -4.62381780e-01 8.40299428e-01
2.46831521e-01 -5.51553145e-02 3.85630995e-01 6.33938611e-01
2.63152309e-02 -1.56624913e+00 4.57742900e-01 7.22161889e-01
-1.01225093e-01 2.85729855e-01 -3.70807618e-01 2.23735660e-01
-3.78217518e-01 8.20711017e-01 1.13499232e-01 -6.13502741e-01
6.44703150e-01 7.99102634e-02 3.29511791e-01 -5.83176315e-01
-2.71400392e-01 4.12910394e-02 -3.26060742e-01 -7.50548840e-01
-1.02149047e-01 -8.79567027e-01 -1.18968272e+00 -4.72175956e-01
-4.35995087e-02 3.06201905e-01 2.20228329e-01 6.32757187e-01
6.97057486e-01 7.51827002e-01 9.06389892e-01 -7.00613558e-01
-6.21701181e-01 -6.26919329e-01 -9.09143448e-01 7.71349192e-01
4.89656389e-01 -5.73947489e-01 -4.66300756e-01 -2.21309677e-01] | [10.190589904785156, 5.589132785797119] |
4e74967e-0e0a-4bea-9d53-e27db7dfbdeb | efficient-gesture-recognition-for-the | 2205.06980 | null | https://arxiv.org/abs/2205.06980v1 | https://arxiv.org/pdf/2205.06980v1.pdf | Efficient Gesture Recognition for the Assistance of Visually Impaired People using Multi-Head Neural Networks | This paper proposes an interactive system for mobile devices controlled by hand gestures aimed at helping people with visual impairments. This system allows the user to interact with the device by making simple static and dynamic hand gestures. Each gesture triggers a different action in the system, such as object recognition, scene description or image scaling (e.g., pointing a finger at an object will show a description of it). The system is based on a multi-head neural network architecture, which initially detects and classifies the gestures, and subsequently, depending on the gesture detected, performs a second stage that carries out the corresponding action. This multi-head architecture optimizes the resources required to perform different tasks simultaneously, and takes advantage of the information obtained from an initial backbone to perform different processes in a second stage. To train and evaluate the system, a dataset with about 40k images was manually compiled and labeled including different types of hand gestures, backgrounds (indoors and outdoors), lighting conditions, etc. This dataset contains synthetic gestures (whose objective is to pre-train the system in order to improve the results) and real images captured using different mobile phones. The results obtained and the comparison made with the state of the art show competitive results as regards the different actions performed by the system, such as the accuracy of classification and localization of gestures, or the generation of descriptions for objects and scenes. | ['Miguel Ángel Lozano', 'Antonio Javier Gallego', 'Samer Alashhab'] | 2022-05-14 | null | null | null | null | ['gesture-recognition'] | ['computer-vision'] | [ 2.33868539e-01 -3.58425945e-01 3.67208496e-02 -4.41207618e-01
-2.07570463e-01 -3.96796286e-01 6.17341697e-01 -3.92189980e-01
-7.31862843e-01 3.63745660e-01 1.49941579e-01 -1.23155110e-01
3.09864990e-02 -5.18305898e-01 -2.36811861e-01 -7.80358374e-01
1.19438826e-03 8.62554371e-01 5.78775406e-01 -7.93349147e-02
2.81299025e-01 9.21838880e-01 -1.98880720e+00 5.58034539e-01
2.22592309e-01 6.23853266e-01 6.52990341e-01 1.15071309e+00
-2.51553893e-01 5.15848219e-01 -8.19582582e-01 1.28555194e-01
-6.59206929e-03 -2.30885118e-01 -6.29900396e-01 2.74775028e-01
3.66031557e-01 -8.06245446e-01 -2.54862309e-01 7.83953071e-01
1.08616924e+00 3.08341295e-01 6.43764257e-01 -1.10956335e+00
-5.81059456e-02 2.15217173e-01 1.10797383e-01 -1.47988275e-01
7.86972344e-01 4.49846655e-01 4.53416228e-01 -7.32555330e-01
6.23310804e-01 1.44742274e+00 2.06416711e-01 1.05024862e+00
-8.40255439e-01 -5.75996637e-01 1.41092330e-01 4.48060781e-01
-1.39786851e+00 -4.94656593e-01 4.17177081e-01 -5.70793986e-01
8.82798195e-01 5.79530954e-01 6.90927565e-01 1.11960161e+00
-4.36684549e-01 8.81323278e-01 9.32214200e-01 -8.62702727e-01
2.70058304e-01 2.40526069e-02 2.66350091e-01 5.86744428e-01
-1.49797410e-01 -1.07917756e-01 -4.42446738e-01 2.08878249e-01
8.48005295e-01 8.96955281e-02 -4.63481903e-01 -2.11081058e-01
-1.10137212e+00 1.76389456e-01 3.57585818e-01 6.63146734e-01
-5.27418017e-01 -4.86788489e-02 1.53948680e-01 2.34415904e-01
-2.73407847e-01 -3.51551510e-02 -3.96849930e-01 -1.55146778e-01
-8.32530260e-01 1.96521074e-01 7.34947741e-01 9.13585603e-01
1.66219383e-01 -5.56697309e-01 -5.73542237e-01 6.80931389e-01
4.84051108e-01 6.34341240e-01 7.62777925e-01 -6.56458676e-01
6.24667048e-01 7.42103517e-01 2.99766481e-01 -4.38969165e-01
-7.13506460e-01 5.10527790e-01 -3.59259427e-01 7.84201264e-01
9.30539370e-01 -2.09732652e-01 -1.36009645e+00 1.23609841e+00
3.53360325e-01 -1.59165353e-01 -2.80941635e-01 1.01680660e+00
1.14723599e+00 2.98006237e-01 3.17536771e-01 -5.57082668e-02
1.35945570e+00 -8.02097023e-01 -8.05026054e-01 9.60762128e-02
3.14123929e-01 -7.01044679e-01 1.50826263e+00 4.32841271e-01
-8.54642868e-01 -7.55552411e-01 -5.30430555e-01 5.04622720e-02
-5.35004377e-01 7.22336471e-01 2.38832638e-01 7.97775328e-01
-1.16977286e+00 2.34271720e-01 -8.32629800e-01 -6.11716568e-01
1.52162731e-01 8.99400413e-01 -2.68670440e-01 4.37980220e-02
-5.98684311e-01 6.96381629e-01 3.95276248e-01 1.83636144e-01
-6.00843906e-01 8.26726779e-02 -4.88634557e-01 6.40296489e-02
1.13562584e-01 -3.90204936e-01 1.12250257e+00 -9.02215660e-01
-1.93079138e+00 8.80151927e-01 -3.24367106e-01 1.11790458e-02
9.14276779e-01 -2.93339372e-01 -1.69384673e-01 2.62715280e-01
-1.09330840e-01 7.45603800e-01 1.01469767e+00 -9.94567990e-01
-9.02968645e-01 -6.12746000e-01 1.47620723e-01 3.57521445e-01
-3.89281869e-01 5.22090971e-01 -1.13958454e+00 -3.50412607e-01
9.86468270e-02 -1.12984204e+00 1.24365106e-01 -1.65518045e-01
-6.30416572e-01 -2.21984968e-01 1.01778650e+00 -8.55184674e-01
1.12128067e+00 -2.02563906e+00 3.53744298e-01 4.17946547e-01
-1.10875286e-01 5.33495545e-01 1.11895073e-02 1.80362687e-01
7.54888356e-02 -1.88659996e-01 8.50061402e-02 -4.56323773e-01
-1.75024867e-01 2.56988227e-01 2.27084637e-01 1.28721073e-01
-2.61640549e-01 5.23823202e-01 -5.66886485e-01 -5.24074852e-01
8.16578865e-01 7.39872456e-01 -3.11387092e-01 5.68770885e-01
-2.27870986e-01 9.36510324e-01 -3.04215640e-01 5.70266664e-01
2.21340612e-01 3.12053822e-02 3.58870476e-01 7.48992488e-02
-1.46944210e-01 7.94169493e-03 -1.58515501e+00 1.27208209e+00
-5.36277711e-01 8.12792897e-01 -9.04768407e-02 -5.17999113e-01
4.18530375e-01 5.95317125e-01 3.19025457e-01 -4.05977756e-01
3.57278347e-01 2.32148781e-01 8.05843398e-02 -1.32489848e+00
5.05204685e-02 6.60765648e-01 4.14354146e-01 4.34522510e-01
-2.70006895e-01 1.00470103e-01 1.28065363e-01 -3.11895937e-01
1.03629613e+00 2.97968119e-01 1.45859629e-01 5.14952600e-01
8.25233042e-01 -9.47707668e-02 -2.17240006e-01 7.00704038e-01
2.00562719e-02 5.79388499e-01 2.64091883e-02 -5.64379156e-01
-7.97175765e-01 -6.95171237e-01 2.10140228e-01 1.42469299e+00
-9.75110196e-03 6.93165660e-02 -1.03343105e+00 -6.56531215e-01
-2.34444708e-01 4.39380527e-01 -4.03328985e-01 4.10683364e-01
-9.42659676e-01 -5.37450314e-01 4.26275462e-01 6.99192524e-01
8.09511006e-01 -1.77609527e+00 -1.30395162e+00 -8.49804655e-02
7.04476088e-02 -9.41216350e-01 -1.41642287e-01 -9.77329537e-02
-6.82290137e-01 -1.23993945e+00 -1.18678772e+00 -1.00444424e+00
9.05583084e-01 1.96262244e-02 5.81346273e-01 1.80698201e-01
-2.04812780e-01 4.64882970e-01 -6.03714764e-01 -2.67437428e-01
-3.74249041e-01 1.14830725e-01 -2.94188596e-02 1.80125296e-01
3.85935992e-01 -2.83804685e-01 -3.91535044e-01 3.72501016e-01
-7.55926311e-01 -1.44442156e-01 6.60002410e-01 7.35769749e-01
3.05216581e-01 -7.92752877e-02 -2.16974810e-01 -4.10367638e-01
7.02400208e-01 2.87828714e-01 -5.13409674e-01 4.79517937e-01
-3.91853321e-03 1.12655155e-01 1.91851020e-01 -9.14936781e-01
-1.05122769e+00 7.31103957e-01 -2.51303136e-01 -2.90202528e-01
-7.79553592e-01 -2.32401147e-01 -6.81783497e-01 -3.09205592e-01
4.96830374e-01 2.61791438e-01 -2.67738491e-01 -7.69149959e-01
4.71291989e-01 1.39966583e+00 5.31139791e-01 5.53674884e-02
4.60871696e-01 3.36398423e-01 -2.35240951e-01 -1.01920307e+00
1.10957354e-01 -6.66880727e-01 -1.25283217e+00 -7.04185426e-01
9.65579450e-01 -3.04084927e-01 -9.95946288e-01 1.01298237e+00
-1.44434190e+00 -7.44970977e-01 7.02150464e-02 6.63437843e-01
-3.91447842e-01 -1.57234743e-01 -2.28434905e-01 -1.06407118e+00
-1.94239676e-01 -1.42312670e+00 1.16957247e+00 3.55860531e-01
-2.46887475e-01 -5.12826979e-01 -2.63857007e-01 2.22475961e-01
1.97947115e-01 8.74735136e-03 9.19564188e-01 -6.71984196e-01
-5.46833992e-01 -5.08854806e-01 -1.39615178e-01 2.43557587e-01
2.27594018e-01 -7.27931112e-02 -1.08187246e+00 3.38422805e-02
-3.49823117e-01 -6.80902824e-02 5.31497657e-01 5.11935651e-01
1.22888196e+00 -2.87628651e-01 -5.46702266e-01 2.62327433e-01
1.06015968e+00 7.88496256e-01 7.47589231e-01 3.95259023e-01
9.18945551e-01 7.44956315e-01 5.06256223e-01 2.82406777e-01
2.01934606e-01 1.33437073e+00 2.94612855e-01 3.01482854e-03
-5.65871358e-01 2.59885222e-01 2.46783867e-01 6.40862882e-02
-6.07023418e-01 -2.64779299e-01 -9.87917423e-01 2.19290152e-01
-1.70299852e+00 -9.00857270e-01 -1.85423851e-01 2.35444355e+00
3.80020648e-01 -1.71295375e-01 5.38090587e-01 6.70925915e-01
9.63270307e-01 -2.56712228e-01 -5.12710869e-01 -1.14849024e-01
1.99717775e-01 4.82588530e-01 3.15221667e-01 6.45267069e-01
-1.12944829e+00 1.15425396e+00 5.77959824e+00 3.20380002e-01
-1.53221023e+00 -5.46859875e-02 8.84612724e-02 -2.46848181e-01
4.46287870e-01 -7.95310974e-01 -7.04008281e-01 4.98947680e-01
4.30905759e-01 8.58392060e-01 7.66086578e-01 7.70975769e-01
3.95324379e-01 -2.04712823e-01 -1.19662201e+00 1.17582273e+00
1.99270546e-01 -7.15332806e-01 1.01595916e-01 -1.32002041e-01
2.00620994e-01 1.44901825e-02 -1.77040681e-01 5.19091338e-02
-1.78810850e-01 -8.35128188e-01 8.32210302e-01 7.93014228e-01
1.02593863e+00 -2.46067852e-01 6.84215546e-01 4.95713323e-01
-1.13614666e+00 -4.04188365e-01 3.00502449e-01 -1.79246366e-01
-5.04174829e-02 -3.27381343e-01 -1.31942511e+00 -1.86382383e-01
7.42756546e-01 -1.87843051e-02 -6.44975662e-01 1.17437589e+00
-7.15876997e-01 2.84687489e-01 -2.98513234e-01 -6.29949450e-01
-1.52167156e-01 1.40940174e-01 6.06422424e-01 1.33827281e+00
1.82105690e-01 2.12995127e-01 -7.84399211e-02 3.74894381e-01
2.15204194e-01 2.64358580e-01 -4.23520148e-01 3.81671995e-01
4.26479101e-01 8.35284352e-01 -9.10511494e-01 -6.01548851e-01
-2.65679836e-01 1.46137202e+00 -1.66716129e-01 6.07258499e-01
-4.31280553e-01 -6.91773593e-01 3.50262016e-01 1.97027966e-01
1.46153793e-01 -2.74199516e-01 -3.54075342e-01 -8.85836244e-01
3.30151230e-01 -8.93208146e-01 1.29550159e-01 -1.02219892e+00
-2.43007317e-01 6.45606577e-01 -7.90079162e-02 -1.17869210e+00
-7.13458419e-01 -1.05121613e+00 -4.74005580e-01 1.06755364e+00
-8.91820848e-01 -1.08687818e+00 -8.05199564e-01 9.04854059e-01
8.35329711e-01 -2.36669973e-01 1.00034964e+00 4.26782042e-01
-5.77278495e-01 4.14595902e-01 -1.43738687e-01 3.11314851e-01
5.03080606e-01 -1.18222344e+00 2.77777791e-01 6.27027094e-01
2.45771348e-01 4.40909535e-01 3.58838975e-01 -6.07355356e-01
-9.82315898e-01 -6.10598445e-01 9.26385045e-01 -4.87459779e-01
2.78503280e-02 -1.92073151e-01 -4.61924464e-01 5.43086827e-01
-2.68369943e-01 -2.73621202e-01 2.29167432e-01 -2.43406460e-01
2.01625153e-01 4.08035815e-02 -1.32458758e+00 7.16235936e-01
1.28793228e+00 -4.54476893e-01 -6.07871830e-01 5.04457653e-01
2.17346936e-01 -6.48300350e-01 -1.22520171e-01 1.26633897e-01
1.17968822e+00 -8.40913773e-01 9.95159864e-01 -8.20889950e-01
-8.04797038e-02 -2.81521112e-01 -1.32780939e-01 -9.53805983e-01
9.44628865e-02 -3.09814662e-01 -4.19733189e-02 8.38778853e-01
2.32499421e-01 -2.86480963e-01 8.37930262e-01 8.94628584e-01
2.48655662e-01 -4.73697782e-01 -8.45648408e-01 -5.05429506e-01
-8.23479831e-01 -6.24363422e-01 7.89192319e-01 3.27355564e-01
-6.13867640e-02 -3.17909033e-03 -1.39805943e-01 4.20201778e-01
4.54339199e-02 -1.76745608e-01 1.13697207e+00 -1.21361125e+00
-1.31261766e-01 -5.43028653e-01 -6.69763267e-01 -9.97793019e-01
-1.48768902e-01 -4.34168518e-01 1.17442608e-01 -1.64211023e+00
-3.03368896e-01 -4.74817365e-01 3.29576403e-01 7.67653167e-01
1.24007445e-02 2.54399180e-01 5.08464873e-01 5.53204060e-01
-2.96285659e-01 -1.33600801e-01 9.89511549e-01 -1.16395928e-01
-9.46345091e-01 6.13318861e-01 2.44837940e-01 1.03158259e+00
6.29406691e-01 -1.11409351e-01 -6.51559010e-02 -4.74203616e-01
-4.32046652e-01 5.46817929e-02 6.71043038e-01 -1.39223373e+00
2.54945666e-01 1.10681638e-01 6.66044831e-01 -5.69819093e-01
5.03199220e-01 -1.00659204e+00 -7.14899972e-02 9.53363478e-01
-3.47604305e-01 -1.35477915e-01 -2.07468018e-01 -3.41837220e-02
1.51012525e-01 -4.11138833e-01 4.85990137e-01 -2.07251757e-01
-9.15061057e-01 -1.70188084e-01 -2.77701318e-01 -6.43898726e-01
9.56490040e-01 -4.52710062e-01 2.75721997e-01 -3.73500109e-01
-1.41999030e+00 -1.10161297e-01 1.50687054e-01 4.20914948e-01
5.91367900e-01 -1.02313828e+00 -3.00221205e-01 5.01058459e-01
-6.57625049e-02 -7.81080425e-02 -1.05722226e-01 5.07814169e-01
-7.74748683e-01 5.59681296e-01 -3.92925262e-01 -6.77117169e-01
-1.89935911e+00 2.37079695e-01 4.28839743e-01 2.35308763e-02
-5.93565702e-01 8.23618650e-01 -4.36259061e-01 -2.15313539e-01
9.13716137e-01 -6.46441400e-01 -8.28720987e-01 1.55297965e-01
8.09592426e-01 7.59535730e-01 1.30667463e-01 -7.78864563e-01
-4.15232003e-01 8.61631334e-01 3.85258347e-01 -5.88009596e-01
1.11412168e+00 9.02287439e-02 2.07417637e-01 3.26717526e-01
7.01025426e-01 -6.37076423e-02 -1.07490396e+00 4.70179431e-02
-5.42795174e-02 -4.94617909e-01 -1.93086341e-01 -1.17522585e+00
-9.44723785e-01 9.00897503e-01 1.38194525e+00 1.15661442e-01
1.33293033e+00 -9.96720195e-02 3.90399843e-01 7.06714571e-01
6.67678952e-01 -1.03015661e+00 5.07653914e-02 3.33346814e-01
1.06022072e+00 -1.07966411e+00 -3.71982276e-01 -3.33710521e-01
-5.18093526e-01 1.26491416e+00 3.96586031e-01 1.68400541e-01
3.93564910e-01 8.55285227e-02 2.86782205e-01 -1.11079521e-01
2.35498846e-01 -5.85462809e-01 4.47190523e-01 7.64752269e-01
1.30944833e-01 3.55671376e-01 -4.63961571e-01 2.64661491e-01
-1.55203298e-01 3.53584886e-01 -1.22748187e-03 7.19185889e-01
-3.84819984e-01 -1.25382292e+00 -8.36548626e-01 3.20553422e-01
-9.52728540e-02 2.03951851e-01 -7.05546260e-01 9.50515032e-01
6.44419909e-01 9.47852135e-01 -1.38826206e-01 -5.82055449e-01
6.06861055e-01 2.67571509e-01 4.69622493e-01 -7.43534744e-01
-7.66745389e-01 1.98058952e-02 -1.21419214e-01 -6.35880470e-01
-4.10626471e-01 -9.17622685e-01 -1.40047729e+00 3.42460304e-01
-8.40032175e-02 -4.14506972e-01 1.13861072e+00 1.06730032e+00
-1.07911132e-01 6.48754060e-01 2.36531049e-01 -1.46271133e+00
-2.50453621e-01 -1.34074330e+00 -1.89649627e-01 4.79227871e-01
2.59052157e-01 -6.80747688e-01 -1.20935723e-01 2.93569326e-01] | [6.51102876663208, -0.28187716007232666] |
a6b37fd3-0ab0-4df1-b4e1-7b81736c5cfb | stable-remaster-bridging-the-gap-between-old | 2306.06803 | null | https://arxiv.org/abs/2306.06803v1 | https://arxiv.org/pdf/2306.06803v1.pdf | Stable Remaster: Bridging the Gap Between Old Content and New Displays | The invention of modern displays has enhanced the viewer experience for any kind of content: ranging from sports to movies in 8K high-definition resolution. However, older content developed for CRT or early Plasma screen TVs has become outdated quickly and no longer meets current aspect ratio and resolution standards. In this paper, we explore whether we can solve this problem with the use of diffusion models to adapt old content to meet contemporary expectations. We explore the ability to combine multiple independent computer vision tasks to attempt to solve the problem of expanding aspect ratios of old animated content such that the new content would be indistinguishable from the source material to a brand-new viewer. These existing capabilities include Stable Diffusion, Content-Aware Scene Detection, Object Detection, and Key Point Matching. We were able to successfully chain these tasks together in a way that generated reasonable outputs, however, future work needs to be done to improve and expand the application to non-animated content as well. | ['Yian Wong', 'Shuvam Keshari', 'Nathan Paull'] | 2023-06-11 | null | null | null | null | ['key-point-matching'] | ['natural-language-processing'] | [ 4.07860309e-01 -1.72876582e-01 5.25880337e-01 -2.10266322e-01
-6.00533426e-01 -7.77995408e-01 6.42291903e-01 -1.18946442e-02
-4.21655208e-01 4.01208192e-01 2.60273647e-02 -2.13584095e-01
-2.56547090e-02 -6.84954703e-01 -2.09747672e-01 -3.46618056e-01
8.47757980e-03 5.11423945e-01 1.11491752e+00 -3.39757830e-01
4.22883362e-01 6.20857060e-01 -1.78423011e+00 2.37347141e-01
5.84554791e-01 8.92421782e-01 5.99349320e-01 1.04701316e+00
-8.82657990e-02 5.03581762e-01 -6.44518733e-01 -1.76814899e-01
3.75421405e-01 -4.38407004e-01 -5.71100712e-01 4.45349626e-02
7.60515571e-01 -2.97531277e-01 -2.52347708e-01 9.04521942e-01
6.56385660e-01 1.20253734e-01 4.34693903e-01 -7.78201818e-01
-5.76388299e-01 9.65973809e-02 -8.92363012e-01 6.16638720e-01
7.14622617e-01 1.05004981e-01 3.73958796e-01 -6.11775458e-01
8.93203676e-01 1.03907180e+00 6.72996163e-01 2.76758015e-01
-1.18153262e+00 -5.27684271e-01 -2.73367837e-02 -3.56486775e-02
-1.25728285e+00 -6.05501294e-01 7.59272158e-01 -3.89234871e-01
8.85823548e-01 5.43822825e-01 8.89307797e-01 7.32997775e-01
2.57165164e-01 6.94683492e-02 1.19783306e+00 -4.30950761e-01
2.33187936e-02 4.55125064e-01 -2.90784806e-01 4.92316693e-01
1.94471702e-01 -1.95805449e-02 -5.19537807e-01 4.83023226e-02
1.10520267e+00 -7.36991346e-01 -2.63329595e-01 -4.39703405e-01
-1.11407232e+00 6.16641521e-01 1.99524507e-01 4.48307127e-01
3.41667309e-02 -2.20197573e-01 2.51472831e-01 3.58430684e-01
5.56235433e-01 9.03255701e-01 -2.39252448e-01 -4.69681114e-01
-9.51925099e-01 2.80701548e-01 8.00735056e-01 8.27512145e-01
3.70319784e-01 2.34651104e-01 1.80022150e-01 6.80351853e-01
3.97639349e-02 4.81074005e-01 2.65958667e-01 -1.45340109e+00
1.35356262e-01 1.31698757e-01 1.08656868e-01 -1.16432381e+00
-5.66889048e-01 -2.15172231e-01 -2.44617939e-01 6.92966580e-01
5.56942284e-01 -2.45614931e-01 -8.37694347e-01 1.19361544e+00
4.36056137e-01 -2.59192195e-02 -2.55932450e-01 1.00905836e+00
8.31644773e-01 7.41846800e-01 -2.52300978e-01 -8.63444805e-02
1.20794749e+00 -6.07249320e-01 -6.30845845e-01 -2.27177024e-01
1.64804935e-01 -1.29099917e+00 1.07436228e+00 6.90343618e-01
-1.39618361e+00 -8.06626558e-01 -1.24872541e+00 2.36557443e-02
-9.53651220e-03 -4.66302127e-01 4.88967061e-01 1.03266776e+00
-1.16217875e+00 4.89090621e-01 -4.85622823e-01 -5.47948301e-01
1.19565442e-01 2.28121594e-01 -1.99067414e-01 5.00992015e-02
-9.15301800e-01 1.16968489e+00 1.16706140e-01 -2.82105595e-01
-2.57787943e-01 -7.37681568e-01 -5.58480024e-01 -2.07007945e-01
4.60145056e-01 -7.51385629e-01 1.36459684e+00 -1.31104851e+00
-1.65494287e+00 8.86231422e-01 2.15649173e-01 -7.10289702e-02
4.99255300e-01 -1.24835163e-01 -6.33278549e-01 2.31680259e-01
9.35550593e-03 4.53932226e-01 9.43128824e-01 -1.44296074e+00
-1.01021147e+00 -1.99172541e-01 -1.21358275e-01 4.20370251e-01
-1.46877170e-01 1.78958997e-01 -8.81606400e-01 -5.92260063e-01
2.32201800e-01 -9.20664549e-01 6.66865101e-03 2.36187309e-01
2.22856686e-01 3.46398056e-01 1.13336706e+00 -6.56237483e-01
1.00653958e+00 -2.18168950e+00 1.36743441e-01 2.09760085e-01
3.00105810e-01 2.68131226e-01 8.98015276e-02 2.63558567e-01
1.82887971e-01 -1.36019975e-01 1.99990422e-01 -3.60306390e-02
-4.94133115e-01 -2.33462676e-01 -1.61897633e-02 4.25692737e-01
-2.56618232e-01 4.15318221e-01 -6.98495626e-01 -4.08630610e-01
2.62075067e-01 4.49358672e-01 -3.56875837e-01 -1.79144755e-01
-1.76324382e-01 4.47316557e-01 -2.13684902e-01 5.06016970e-01
7.18094409e-01 -2.17984132e-02 4.94247861e-02 -8.66264701e-02
-4.86114234e-01 -2.19998330e-01 -1.54808724e+00 1.65923631e+00
-2.76983500e-01 1.16362298e+00 4.38015282e-01 -3.87255073e-01
9.19665992e-01 -1.93026394e-01 5.95209777e-01 -1.17665327e+00
2.70625856e-02 7.11884573e-02 -5.81769012e-02 -7.06101656e-01
1.37165594e+00 -4.07062322e-01 6.06635995e-02 2.43840367e-01
-4.70389903e-01 -4.85111505e-01 -7.37552568e-02 2.61676252e-01
7.97153234e-01 2.43150324e-01 -1.48753852e-01 -1.26285449e-01
1.18079081e-01 3.01076174e-01 1.08889654e-01 9.39931691e-01
2.49281023e-02 1.04717422e+00 -2.98962817e-02 -1.81666330e-01
-1.37249517e+00 -1.38244390e+00 -2.88286299e-01 9.22648191e-01
5.56963980e-01 -4.43758845e-01 -5.67288280e-01 -1.30879626e-01
-2.03116044e-01 5.90628505e-01 -4.76009846e-02 9.92487594e-02
-6.23085380e-01 -6.85184479e-01 2.67162889e-01 3.25423151e-01
5.24818718e-01 -7.24501789e-01 -9.55893099e-01 4.40771997e-01
4.94007692e-02 -1.06444573e+00 -1.75642163e-01 -1.17267057e-01
-7.29196727e-01 -6.50361300e-01 -9.64458287e-01 -5.02232552e-01
2.23043859e-01 4.40631509e-01 9.38192487e-01 -2.14964807e-01
-4.82610732e-01 7.31193721e-01 -3.41192752e-01 -3.09709579e-01
-5.52461088e-01 -2.43504882e-01 8.07500780e-02 -2.32830733e-01
-5.47422245e-02 -3.85420889e-01 -7.08075345e-01 3.96722227e-01
-8.75524044e-01 1.94972634e-01 3.57931018e-01 1.15760721e-01
4.18860734e-01 2.61711627e-01 4.24376309e-01 -8.16178083e-01
6.07498229e-01 -5.63593730e-02 -5.79111040e-01 -9.64336395e-02
-6.68770492e-01 -7.54240602e-02 2.89980531e-01 -7.92087793e-01
-1.46130681e+00 2.12813653e-02 -2.26410970e-01 2.01209518e-03
1.10781312e-01 2.21814558e-01 6.73405081e-02 -3.33262682e-01
8.97316337e-01 5.61199198e-03 -1.52048603e-01 -4.99934047e-01
5.07550359e-01 7.14042962e-01 7.71031439e-01 -3.43540460e-01
7.26506531e-01 6.80789292e-01 -2.06934214e-01 -1.15988362e+00
-2.36987025e-01 -3.42677802e-01 -3.38348538e-01 -5.74031055e-01
8.15058649e-01 -7.95764565e-01 -2.48796478e-01 5.84930599e-01
-8.32995772e-01 -3.01172137e-01 -4.81239438e-01 4.96952862e-01
-5.63841105e-01 5.39351344e-01 -4.42149311e-01 -6.72555208e-01
9.89183486e-02 -8.09725404e-01 8.53742719e-01 6.31677151e-01
-3.38487595e-01 -7.24137187e-01 3.67612004e-01 3.55492771e-01
6.96235716e-01 1.34641036e-01 6.85813546e-01 1.37513671e-02
-6.80979311e-01 -2.38367215e-01 -3.19687426e-01 -3.92695470e-03
8.94960091e-02 1.83215961e-01 -9.30979013e-01 -2.93827653e-01
2.10177451e-01 1.39205948e-01 6.83950543e-01 4.70407486e-01
6.07626617e-01 2.68630385e-01 -4.28143084e-01 6.03232920e-01
1.44083905e+00 5.99825323e-01 8.98927808e-01 7.71822274e-01
4.68110472e-01 5.06726861e-01 5.20346403e-01 2.55844891e-01
2.07282782e-01 1.11227036e+00 -4.15620767e-02 -3.06396514e-01
-4.48920220e-01 -4.03605253e-02 6.52342886e-02 5.53418100e-01
-2.91782022e-01 -4.23360586e-01 -8.43104362e-01 1.63925737e-01
-1.57903540e+00 -1.18566048e+00 -4.00584370e-01 2.17422152e+00
4.33732808e-01 5.76495409e-01 2.26473182e-01 -1.46418184e-01
5.93535960e-01 -6.96922839e-02 -5.35433352e-01 -6.82083189e-01
-1.75457373e-01 6.02244437e-02 5.52956820e-01 3.36458892e-01
-8.30997825e-01 6.34644330e-01 7.32933426e+00 6.85252666e-01
-1.17309272e+00 3.86807919e-02 4.30500060e-01 -2.18463644e-01
-4.00499821e-01 2.11110339e-02 -6.83566272e-01 3.84122044e-01
8.61816585e-01 -1.68591663e-01 2.84638673e-01 6.64368987e-01
2.27098674e-01 -7.63078690e-01 -6.00183368e-01 1.09131944e+00
3.37959945e-01 -1.31980419e+00 -3.89823258e-01 -8.98943990e-02
6.63940609e-01 -6.43574148e-02 3.30017924e-01 -3.20457444e-02
-7.59593993e-02 -6.80096507e-01 8.71680319e-01 8.86078358e-01
9.71949279e-01 -4.13505465e-01 2.74636447e-01 1.81937620e-01
-1.20175254e+00 3.72256301e-02 -1.96546659e-01 -2.56689519e-01
4.47108567e-01 1.62071988e-01 -5.64610958e-01 3.65372688e-01
1.06689751e+00 2.36285865e-01 -8.36794198e-01 1.43204486e+00
4.72800255e-01 1.04130670e-01 -5.13880551e-01 1.78902894e-02
-1.13059118e-01 -3.71944085e-02 7.70883441e-01 1.08716679e+00
5.77348769e-01 1.02063283e-01 -1.06210984e-01 4.62605953e-01
4.33994561e-01 -3.34800258e-02 -6.83229804e-01 3.75842512e-01
1.06610991e-01 1.15084076e+00 -1.20147693e+00 -2.29657546e-01
-6.22744918e-01 1.27492440e+00 5.30886911e-02 3.00513864e-01
-8.26803982e-01 -3.51039410e-01 2.57070035e-01 9.11058426e-01
3.77979994e-01 -5.08790493e-01 -2.37946242e-01 -1.03419983e+00
-1.02497354e-01 -8.86169016e-01 2.13339642e-01 -1.22650158e+00
-1.21886301e+00 6.68712378e-01 2.37870440e-01 -1.00182617e+00
-1.12330414e-01 -3.56109232e-01 -4.34179425e-01 9.40131545e-01
-9.45847273e-01 -7.64873981e-01 -2.64573812e-01 3.74678552e-01
6.49917424e-01 -1.08806215e-01 5.36203563e-01 3.90309274e-01
-2.18296200e-02 2.58985698e-01 5.49549341e-01 -4.78990495e-01
8.70416105e-01 -9.86762404e-01 3.65958422e-01 8.26065123e-01
1.93393275e-01 1.00396514e-01 1.13335264e+00 -7.48203695e-01
-1.18371081e+00 -3.26262891e-01 3.50611240e-01 -7.09953606e-01
4.08599377e-01 -3.36754173e-01 -8.63127470e-01 2.29209855e-01
1.59443796e-01 -5.58238983e-01 3.24565291e-01 1.14829957e-01
-8.14437345e-02 -1.50224984e-01 -1.03950644e+00 7.29057908e-01
9.21366215e-01 -3.33955824e-01 -4.25806671e-01 -1.09290712e-01
4.77849662e-01 -7.15671480e-01 -6.00901425e-01 4.13066685e-01
7.89027154e-01 -1.19892967e+00 9.48222101e-01 1.33180665e-02
4.83527146e-02 -4.49095100e-01 2.43051490e-03 -9.55102861e-01
-4.73586828e-01 -6.83823645e-01 9.03610215e-02 1.18008232e+00
4.21381354e-01 -3.74693781e-01 7.51649737e-01 9.24224675e-01
-1.92344844e-01 -2.90344030e-01 -1.03407061e+00 -6.65669441e-01
-2.20280170e-01 -5.14417350e-01 1.26750588e-01 6.20374620e-01
-1.92594081e-01 5.91199994e-01 -4.37169552e-01 3.39752734e-02
3.92108381e-01 1.92991644e-01 6.93801820e-01 -1.23085082e+00
-5.08918464e-01 -4.75842565e-01 -5.04025877e-01 -1.05144525e+00
-6.43724084e-01 -4.09335405e-01 -1.71460375e-01 -1.59991646e+00
1.47216812e-01 -7.44105279e-01 3.40421379e-01 -4.24928486e-01
-1.55071467e-02 6.09555542e-01 4.81224835e-01 1.33150578e-01
-4.05921429e-01 2.92375144e-02 1.39641094e+00 -2.83976328e-02
-5.09358466e-01 8.42822529e-03 -8.88015032e-01 7.64423609e-01
7.68033564e-01 -3.18678439e-01 -5.31233072e-01 -4.02475536e-01
5.15268326e-01 1.71189293e-01 1.04260840e-01 -1.46556318e+00
2.56588370e-01 1.38634026e-01 8.47979724e-01 -3.67305338e-01
5.68414986e-01 -8.64780605e-01 6.22155726e-01 2.17129979e-02
1.11261364e-02 2.74580330e-01 6.63353920e-01 5.94881415e-01
2.28070736e-01 -3.41295600e-01 7.41799355e-01 -1.32523283e-01
-1.10276020e+00 -6.67420775e-02 -8.96672487e-01 -3.08760218e-02
1.22787118e+00 -8.07965875e-01 -3.55183005e-01 -6.13360703e-01
-8.59133422e-01 -2.13156596e-01 9.31433916e-01 5.36326885e-01
6.99017465e-01 -1.00683665e+00 -5.42940676e-01 1.90414041e-01
-1.56988829e-01 -3.52819890e-01 5.47815442e-01 4.42516148e-01
-9.69524503e-01 8.67361426e-02 -4.17796522e-01 -6.04183257e-01
-1.55131400e+00 7.09698081e-01 2.66328305e-01 1.98010683e-01
-1.12098360e+00 6.75484598e-01 3.20515595e-02 1.85173541e-01
5.54298609e-02 3.55266854e-02 -1.68280944e-01 4.27287072e-02
5.97828567e-01 3.24999928e-01 -1.24887750e-01 -6.65610969e-01
-1.04559347e-01 8.51095140e-01 -9.13887918e-02 -5.55534482e-01
1.08790004e+00 -6.38324738e-01 4.18013126e-01 5.17157972e-01
7.76557088e-01 3.51857275e-01 -1.48102653e+00 1.31508917e-01
-3.02444309e-01 -7.58570135e-01 1.36287481e-01 -7.46242762e-01
-7.94486940e-01 5.12049735e-01 1.11413407e+00 4.05303836e-01
1.28496301e+00 1.67299822e-01 7.60414183e-01 4.38380428e-02
2.74536192e-01 -1.38232958e+00 3.61595750e-01 2.06974223e-01
5.80496848e-01 -9.78367567e-01 3.01336348e-01 -3.35168064e-01
-7.83186913e-01 1.21896720e+00 6.40438318e-01 2.35567763e-01
4.79619473e-01 6.86747551e-01 1.58202618e-01 -4.33102101e-01
-3.93033236e-01 -1.99700102e-01 2.50321805e-01 9.06310737e-01
3.48592997e-01 -2.69902170e-01 -1.21364318e-01 -3.66298966e-02
-1.20996371e-01 -1.49302989e-01 8.35428476e-01 1.03967094e+00
-8.46385837e-01 -8.48607063e-01 -6.43659830e-01 5.76859713e-01
-3.07563692e-01 9.14122537e-02 -2.38173455e-01 8.13631892e-01
2.17180967e-01 8.96737516e-01 3.40506613e-01 -3.69514585e-01
5.25424898e-01 -2.32163265e-01 8.08589339e-01 -4.65519607e-01
-4.94022787e-01 4.08307284e-01 3.05897027e-01 -2.97719479e-01
-5.29314101e-01 -6.94925845e-01 -1.02313972e+00 -4.99624640e-01
-4.20564026e-01 -1.29723489e-01 7.86314666e-01 4.63215828e-01
2.02900037e-01 5.37770987e-01 2.19057247e-01 -9.54984486e-01
3.90171036e-02 -6.10834897e-01 -7.07463324e-01 1.34768516e-01
1.22395873e-01 -4.16173249e-01 -4.11884598e-02 1.78987175e-01] | [11.024794578552246, -2.116525888442993] |
a6d87bac-2024-466a-9382-07fbe99fb13e | environmental-noise-embeddings-for-robust | 1601.02553 | null | http://arxiv.org/abs/1601.02553v2 | http://arxiv.org/pdf/1601.02553v2.pdf | Environmental Noise Embeddings for Robust Speech Recognition | We propose a novel deep neural network architecture for speech recognition
that explicitly employs knowledge of the background environmental noise within
a deep neural network acoustic model. A deep neural network is used to predict
the acoustic environment in which the system in being used. The discriminative
embedding generated at the bottleneck layer of this network is then
concatenated with traditional acoustic features as input to a deep neural
network acoustic model. Through a series of experiments on Resource Management,
CHiME-3 task, and Aurora4, we show that the proposed approach significantly
improves speech recognition accuracy in noisy and highly reverberant
environments, outperforming multi-condition training, noise-aware training,
i-vector framework, and multi-task learning on both in-domain noise and unseen
noise. | ['Bhiksha Raj', 'Suyoun Kim', 'Ian Lane'] | 2016-01-11 | null | null | null | null | ['robust-speech-recognition'] | ['speech'] | [ 1.27578259e-01 -5.48367739e-01 6.85192764e-01 -5.79986036e-01
-1.03565180e+00 -2.49450073e-01 3.11060846e-01 -2.06731215e-01
-6.13494217e-01 2.09743872e-01 4.64708954e-01 -6.96534336e-01
-1.30665138e-01 -5.91206312e-01 -5.90591431e-01 -7.09248066e-01
-9.03761312e-02 1.75886184e-01 -8.30052942e-02 -1.66845188e-01
-2.97899872e-01 6.09506547e-01 -1.47508895e+00 2.13815644e-01
-4.90363501e-02 1.40034032e+00 5.03319681e-01 1.46606612e+00
7.41466656e-02 8.30500662e-01 -1.22967124e+00 4.49140184e-02
3.68592054e-01 2.64117010e-02 -4.34931546e-01 -2.42320970e-01
2.87727922e-01 -4.11786139e-01 -9.41363633e-01 6.80264831e-01
1.43215585e+00 7.43085146e-01 2.84559160e-01 -5.11796117e-01
-3.09156686e-01 5.77116430e-01 1.43733695e-01 7.49965906e-01
-5.81793971e-02 3.38236630e-01 8.35651457e-01 -1.08418381e+00
-2.65515387e-01 1.32484519e+00 6.84439898e-01 2.55044848e-01
-9.90076303e-01 -6.31471336e-01 1.16395719e-01 3.90995413e-01
-1.24796605e+00 -1.04766762e+00 6.92171454e-01 -2.45372266e-01
1.56213486e+00 3.73878568e-01 1.66774362e-01 1.53014982e+00
-1.15593120e-01 6.69128835e-01 6.65756583e-01 -4.72203374e-01
3.50634575e-01 -1.64178461e-01 1.06668547e-01 2.05791399e-01
-5.28068066e-01 5.47631621e-01 -6.13587439e-01 -3.17105502e-01
2.15547308e-01 -4.56532955e-01 -1.46231890e-01 4.68935758e-01
-6.54545248e-01 4.46604371e-01 1.42042890e-01 4.08133000e-01
-5.92345178e-01 3.54480445e-01 5.76928735e-01 5.34673870e-01
4.71243501e-01 3.05603296e-01 -7.77681053e-01 -4.68266249e-01
-7.36339808e-01 -7.24439323e-02 1.00491023e+00 4.56178039e-01
7.41812110e-01 1.07582819e+00 -9.16649029e-02 1.57317853e+00
4.07837123e-01 8.41503084e-01 6.99320078e-01 -6.35898650e-01
4.52482909e-01 -4.40443128e-01 -2.05845237e-01 -6.86858296e-01
-3.99552822e-01 -1.02092874e+00 -6.88853264e-01 3.03702988e-02
-2.34141469e-01 -5.61957300e-01 -1.21502864e+00 1.50601113e+00
1.85608156e-02 7.51986921e-01 4.51874197e-01 6.64480448e-01
7.95442522e-01 1.01821327e+00 -4.51065637e-02 1.67058975e-01
8.79171550e-01 -1.28204632e+00 -8.26089263e-01 -6.21686816e-01
4.12860572e-01 -9.90565777e-01 6.98492765e-01 5.89985013e-01
-7.67461419e-01 -8.96562517e-01 -1.18256688e+00 2.87712306e-01
-4.70108837e-01 3.55837382e-02 5.93070164e-02 1.03088152e+00
-1.22282231e+00 2.55559325e-01 -5.86021245e-01 1.90833613e-01
1.67892322e-01 4.24077123e-01 -8.93518552e-02 -3.92550230e-02
-1.39071107e+00 9.90487039e-01 2.76978344e-01 5.59071720e-01
-1.64279342e+00 -5.61591268e-01 -1.00979543e+00 2.96097010e-01
2.25595295e-01 -2.93042779e-01 1.60898554e+00 -6.70965850e-01
-1.99521279e+00 4.50450256e-02 -1.05715342e-01 -6.73086524e-01
-1.00991298e-02 -6.56972528e-01 -1.23957253e+00 -2.15036467e-01
-4.61621791e-01 -1.76891699e-01 9.85475063e-01 -1.20641100e+00
-5.42439759e-01 1.45381643e-02 -2.11433187e-01 3.05844873e-01
-3.93389940e-01 3.35612118e-01 -3.50949287e-01 -6.54049218e-01
-2.12373316e-01 -6.30829990e-01 -7.31534243e-01 -7.57587790e-01
-5.01986563e-01 9.88877863e-02 1.00131595e+00 -1.07461524e+00
1.26759541e+00 -2.46998739e+00 -1.84145436e-01 6.25727117e-01
-2.53909379e-01 8.11754107e-01 -5.61251044e-01 2.63612509e-01
-1.23539492e-01 -1.56848595e-01 2.09839836e-01 -6.44383371e-01
6.98447675e-02 3.45263541e-01 -3.53754818e-01 1.98367715e-01
7.05673322e-02 2.64689296e-01 -6.12541199e-01 1.76843956e-01
5.16179681e-01 7.07524061e-01 -3.50148469e-01 6.30582988e-01
1.12786002e-01 2.05109283e-01 -7.79481828e-02 3.72888297e-01
5.51347196e-01 4.54542249e-01 -1.19287059e-01 3.66941355e-02
8.69382918e-02 7.65567839e-01 -1.27565289e+00 1.50176418e+00
-1.30512857e+00 9.34023440e-01 5.03784180e-01 -9.00185227e-01
1.08508444e+00 7.36664772e-01 3.68211828e-02 -9.11710739e-01
4.09144759e-02 2.80752093e-01 2.18232855e-01 -5.57939887e-01
3.97059828e-01 -1.31073788e-01 7.64901713e-02 3.61583948e-01
2.86379516e-01 -4.63057198e-02 -6.28303111e-01 -2.77318418e-01
1.42932105e+00 -4.62222129e-01 -3.00493054e-02 2.14203641e-01
6.09130383e-01 -8.51733685e-01 5.17705917e-01 1.11173797e+00
-2.95045733e-01 5.54468215e-01 -1.74130484e-01 -4.20746982e-01
-9.90801394e-01 -1.02184510e+00 -2.47944206e-01 1.59165013e+00
-4.70369607e-01 -2.53915995e-01 -4.75754440e-01 -2.50789225e-01
-9.89397615e-02 8.39177489e-01 -4.06969458e-01 -2.82231122e-01
-8.16854954e-01 -7.79277980e-01 1.03229904e+00 6.04623973e-01
1.39311165e-01 -1.14972234e+00 -3.36290747e-02 6.68958724e-01
-1.96407791e-02 -1.40296483e+00 -2.56476790e-01 9.39499021e-01
-1.85613930e-01 -3.12522531e-01 -2.87771463e-01 -8.22787642e-01
-1.96516424e-01 1.28034547e-01 9.15971458e-01 -2.02050254e-01
-2.66587019e-01 3.79936486e-01 -2.27985948e-01 -4.90518600e-01
-8.96722794e-01 -1.29910588e-01 5.26427925e-01 2.07968235e-01
2.26527378e-01 -8.56579781e-01 -4.73911792e-01 2.75508314e-01
-6.36843085e-01 -5.25897741e-01 5.51337242e-01 9.86779273e-01
2.65202343e-01 2.87979007e-01 8.45111907e-01 -2.50375539e-01
6.30969286e-01 -6.08642697e-01 -4.30229098e-01 4.87995781e-02
-3.97787467e-02 9.05315101e-04 6.29802167e-01 -4.38273609e-01
-1.12435591e+00 -1.11412115e-01 -1.02141750e+00 -5.16864538e-01
-4.94118333e-01 5.82016051e-01 -4.29279119e-01 8.14898536e-02
6.25933349e-01 4.06631291e-01 -4.02043402e-01 -8.67257476e-01
1.75251842e-01 1.32476640e+00 7.52249122e-01 -3.96097064e-01
7.12689102e-01 -1.06194727e-01 -3.44815582e-01 -1.24101627e+00
-7.35097528e-01 -6.87028527e-01 -3.65079701e-01 -1.50457174e-01
5.92088223e-01 -1.08172321e+00 -3.70469928e-01 7.82588303e-01
-1.16578043e+00 -6.22381508e-01 1.30229630e-03 6.40201628e-01
-3.23666573e-01 1.97469100e-01 -6.17957771e-01 -1.30605555e+00
-3.86599839e-01 -1.36178470e+00 8.71252239e-01 -1.37363940e-01
2.05842257e-01 -9.75343943e-01 2.43741840e-01 4.08387810e-01
9.10689175e-01 -4.22914594e-01 7.83525944e-01 -1.23360980e+00
-2.23949105e-01 -2.29976833e-01 1.77217051e-01 1.28864670e+00
2.08336785e-01 -2.30901584e-01 -1.85521042e+00 -2.90743589e-01
2.26464748e-01 -4.15614218e-01 9.25059736e-01 3.63979846e-01
1.52358198e+00 -2.26498544e-01 2.19870016e-01 7.60203183e-01
1.23742604e+00 5.45144081e-01 4.90106553e-01 3.26536536e-01
6.90576375e-01 2.00125530e-01 -1.59675367e-02 4.26269114e-01
-3.98447625e-02 7.28073776e-01 5.12881994e-01 -1.84724852e-01
-2.46175051e-01 1.71094760e-01 5.65412164e-01 1.38377106e+00
4.88556862e-01 -3.37188870e-01 -9.99164224e-01 6.60805821e-01
-1.24530065e+00 -8.36944699e-01 4.36344713e-01 2.01420212e+00
5.79866111e-01 1.51752442e-01 -2.36505330e-01 3.30238909e-01
4.11918700e-01 4.00279880e-01 -4.51129317e-01 -7.83677161e-01
-3.64615172e-01 7.31018186e-01 2.61115253e-01 8.08423102e-01
-1.28804171e+00 8.10164154e-01 7.22728348e+00 9.41525698e-01
-1.23250401e+00 3.96276385e-01 6.91369295e-01 -2.17691720e-01
-8.43150094e-02 -6.26885056e-01 -6.07005894e-01 8.24328363e-02
1.89306641e+00 2.49358624e-01 5.19046843e-01 9.18874502e-01
4.48367000e-01 2.86055505e-01 -9.34160948e-01 9.21050429e-01
1.04764864e-01 -1.03169417e+00 -2.28249490e-01 -4.10142839e-02
2.64837056e-01 7.27677941e-01 5.34907401e-01 6.11712873e-01
4.89860833e-01 -1.29558742e+00 6.27321422e-01 1.82431608e-01
6.39796972e-01 -9.19638455e-01 8.34936261e-01 2.48203024e-01
-9.34876323e-01 -4.28577423e-01 -3.64913493e-01 6.96331933e-02
1.25275314e-01 5.62386096e-01 -1.37669539e+00 4.52024549e-01
7.77745366e-01 1.73077844e-02 -2.11446896e-01 1.04320061e+00
7.69764930e-02 1.27653170e+00 -4.50097710e-01 1.36890069e-01
5.43564022e-01 2.74069786e-01 7.15258360e-01 1.68694854e+00
3.47811937e-01 -1.89991906e-01 4.05269004e-02 2.65160531e-01
-4.27327603e-02 -2.96836440e-02 -5.32577932e-01 1.39813438e-01
5.57220221e-01 1.04041040e+00 2.96909541e-01 -3.04407328e-01
-4.59669441e-01 6.88638330e-01 2.26131439e-01 8.01541686e-01
-5.36015332e-01 -3.60522300e-01 1.15200138e+00 -6.24066174e-01
6.52822495e-01 -3.92575264e-01 -1.46376923e-01 -5.75705469e-01
-1.17121249e-01 -9.68697906e-01 -5.35271205e-02 -6.41385913e-01
-1.05209601e+00 1.21526289e+00 -4.90884900e-01 -7.95389175e-01
-5.27554512e-01 -7.88562477e-01 -8.16900730e-01 1.50057256e+00
-1.50760901e+00 -6.46197021e-01 1.26320392e-01 4.09924120e-01
8.08894932e-01 -8.38989437e-01 1.31654930e+00 7.41797209e-01
-9.20231104e-01 8.50162268e-01 6.11922204e-01 3.13865751e-01
4.71936345e-01 -1.31095767e+00 9.57870662e-01 9.67288613e-01
4.91219789e-01 2.36209661e-01 7.31228232e-01 -1.42679974e-01
-1.27671289e+00 -1.38427770e+00 4.88105476e-01 -1.74523309e-01
8.57401729e-01 -6.48647845e-01 -9.17119145e-01 4.05872405e-01
2.51879275e-01 1.26708314e-01 1.16590393e+00 4.17714983e-01
-4.93158162e-01 -3.14728737e-01 -7.28619993e-01 2.66107112e-01
6.54548407e-01 -9.89362001e-01 -4.27301228e-01 3.31450015e-01
1.08222902e+00 -4.75228071e-01 -7.60452271e-01 1.62583172e-01
2.80987740e-01 -5.72747827e-01 1.27169514e+00 -7.18997359e-01
-2.16404036e-01 2.86253300e-02 -8.48063171e-01 -1.76377237e+00
-2.19068483e-01 -8.67012143e-01 -2.25952253e-01 9.22307014e-01
8.32490921e-01 -4.24621552e-01 4.57385808e-01 1.06839165e-01
-8.73633802e-01 -5.79217792e-01 -1.29413164e+00 -8.56793344e-01
-1.10992454e-01 -1.26240253e+00 7.65077233e-01 4.55187976e-01
-8.86452854e-01 2.16568351e-01 -7.14010298e-01 8.63276243e-01
2.87229180e-01 -6.28144503e-01 7.08030581e-01 -6.51329279e-01
-7.40863323e-01 -1.72613874e-01 -4.67856139e-01 -1.18162835e+00
2.98939735e-01 -5.48593700e-01 5.41611671e-01 -1.21642256e+00
-6.05107248e-01 -5.41935146e-01 -1.00415933e+00 3.53457093e-01
-1.78430080e-01 -4.23583426e-02 -2.27693394e-02 -2.85128087e-01
-3.89945239e-01 8.02100599e-01 5.70135474e-01 -2.99261153e-01
-2.18058884e-01 5.27340710e-01 -4.04450178e-01 5.59099913e-01
7.68040180e-01 -3.91406745e-01 -1.17507398e-01 -8.53518963e-01
-3.07400435e-01 3.34247127e-02 1.21270038e-01 -1.43389738e+00
2.88280219e-01 1.97035640e-01 2.09421292e-01 -5.18130064e-01
8.74743223e-01 -8.67132545e-01 -1.27952427e-01 1.33268222e-01
-4.26230729e-01 -2.05135316e-01 5.91467619e-01 7.23620176e-01
-4.29776907e-01 -2.52497196e-01 5.20585001e-01 3.16160828e-01
-8.44223619e-01 2.90341884e-01 -7.76772559e-01 -2.43042707e-01
1.57762811e-01 8.70169625e-02 -1.06551342e-01 -5.77496409e-01
-1.13902140e+00 2.59109810e-02 -6.42249227e-01 6.53532207e-01
8.02002788e-01 -1.35544527e+00 -8.31820726e-01 4.87402707e-01
-9.12178308e-02 -1.29041255e-01 5.31902969e-01 2.14997664e-01
-2.51136571e-01 2.08058298e-01 4.09378260e-01 -6.27021849e-01
-1.11363328e+00 1.21921577e-01 9.58588243e-01 -3.82747725e-02
-4.80374962e-01 1.20355618e+00 7.49666393e-02 -7.23549306e-01
8.81278574e-01 -2.82125175e-01 6.55277967e-02 -4.13659662e-01
8.27059031e-01 2.47065276e-01 6.98318362e-01 -8.02476406e-01
-2.57771134e-01 -1.77957177e-01 -2.31173150e-02 -6.58670962e-01
1.48356342e+00 -9.35745463e-02 3.04213285e-01 6.49969399e-01
1.50372517e+00 -4.57705483e-02 -1.07283449e+00 -8.88085008e-01
-8.49755574e-03 -2.05549210e-01 7.24522412e-01 -1.10436487e+00
-9.22936797e-01 1.11577678e+00 9.58355725e-01 1.00978740e-01
1.28451300e+00 -3.56599897e-01 8.62941504e-01 8.87414932e-01
-6.59811422e-02 -1.34175730e+00 1.82107449e-01 1.06612885e+00
9.87186193e-01 -1.09310806e+00 -6.60420239e-01 2.22854152e-01
-3.46202999e-01 1.01189375e+00 6.03815913e-01 2.78189152e-01
1.29729009e+00 8.79573345e-01 7.07401514e-01 1.59177005e-01
-9.72582519e-01 -1.87961116e-01 1.89945862e-01 7.81011939e-01
9.17437002e-02 1.48582846e-01 1.07154334e+00 6.58560276e-01
-4.10853565e-01 -6.84901059e-01 1.86433077e-01 7.28981555e-01
-6.62120998e-01 -8.73992741e-01 -5.45498729e-01 3.92195195e-01
-5.10605693e-01 -5.10968328e-01 -8.83470196e-03 6.56459853e-02
4.58515733e-02 1.37064266e+00 9.02445242e-02 -8.57792139e-01
6.38475716e-01 3.11122596e-01 -1.49766639e-01 -7.04219460e-01
-1.02407229e+00 4.16168123e-01 5.91862917e-01 -4.68085945e-01
4.03014570e-02 -4.62952465e-01 -7.49468386e-01 1.45972282e-01
-4.14499938e-01 2.04521477e-01 1.16962802e+00 1.07362497e+00
1.01990841e-01 1.16183233e+00 1.09721696e+00 -8.26439261e-01
-8.78915012e-01 -1.35610390e+00 -5.22790492e-01 -2.52428651e-02
9.28749263e-01 -2.48293608e-01 -6.32518947e-01 -4.28319484e-01] | [14.850733757019043, 6.045862197875977] |
ab9bc604-51ec-40cd-8201-7f046dc15af0 | anchor-transform-learning-sparse-1 | 2003.08197 | null | https://arxiv.org/abs/2003.08197v4 | https://arxiv.org/pdf/2003.08197v4.pdf | Anchor & Transform: Learning Sparse Embeddings for Large Vocabularies | Learning continuous representations of discrete objects such as text, users, movies, and URLs lies at the heart of many applications including language and user modeling. When using discrete objects as input to neural networks, we often ignore the underlying structures (e.g., natural groupings and similarities) and embed the objects independently into individual vectors. As a result, existing methods do not scale to large vocabulary sizes. In this paper, we design a simple and efficient embedding algorithm that learns a small set of anchor embeddings and a sparse transformation matrix. We call our method Anchor & Transform (ANT) as the embeddings of discrete objects are a sparse linear combination of the anchors, weighted according to the transformation matrix. ANT is scalable, flexible, and end-to-end trainable. We further provide a statistical interpretation of our algorithm as a Bayesian nonparametric prior for embeddings that encourages sparsity and leverages natural groupings among objects. By deriving an approximate inference algorithm based on Small Variance Asymptotics, we obtain a natural extension that automatically learns the optimal number of anchors instead of having to tune it as a hyperparameter. On text classification, language modeling, and movie recommendation benchmarks, we show that ANT is particularly suitable for large vocabulary sizes and demonstrates stronger performance with fewer parameters (up to 40x compression) as compared to existing compression baselines. | ['Yu-An Wang', 'Paul Pu Liang', 'Manzil Zaheer', 'Amr Ahmed'] | 2020-03-18 | anchor-transform-learning-sparse-embeddings | https://openreview.net/forum?id=Vd7lCMvtLqg | https://openreview.net/pdf?id=Vd7lCMvtLqg | iclr-2021-1 | ['movie-recommendation'] | ['miscellaneous'] | [ 3.13683897e-02 -1.25169471e-01 -5.58441162e-01 -5.11671066e-01
-5.44811904e-01 -7.34957874e-01 6.93814814e-01 5.36865830e-01
-4.78996158e-01 2.00498998e-01 5.69642842e-01 -3.23045284e-01
-3.61495584e-01 -7.68374026e-01 -9.00751591e-01 -6.04341209e-01
-1.97080344e-01 7.34154999e-01 4.43414859e-02 7.46345222e-02
1.85020566e-01 2.77895510e-01 -1.40325999e+00 3.89339253e-02
4.38472122e-01 9.68108475e-01 8.36827233e-02 6.26144826e-01
-2.54564136e-01 2.40462005e-01 -2.49023899e-01 -3.97038251e-01
3.05786818e-01 6.72329143e-02 -5.47229946e-01 1.12539038e-01
7.20864594e-01 -8.06555688e-01 -5.33933938e-01 9.25967991e-01
2.52727807e-01 4.32904363e-01 1.08078170e+00 -1.16292584e+00
-1.13276708e+00 7.60674953e-01 -5.61023057e-01 3.01296711e-02
-7.24510327e-02 -3.29018325e-01 1.56259167e+00 -1.14209795e+00
3.31116468e-01 1.16625619e+00 7.41442978e-01 2.40931928e-01
-1.59255159e+00 -3.90970558e-01 2.20049068e-01 -9.46587771e-02
-1.48816442e+00 -4.76913303e-01 5.06021500e-01 -4.80789989e-01
8.83405209e-01 1.86753914e-01 2.38988101e-01 1.04482353e+00
-1.79877639e-01 7.67308652e-01 2.49433890e-01 -5.73187649e-01
6.37590289e-01 2.80048043e-01 3.46238434e-01 6.82077944e-01
6.09824359e-01 -4.24577653e-01 -5.13654768e-01 -6.94601178e-01
7.65631676e-01 6.34188950e-01 -9.46527347e-02 -6.52772725e-01
-1.09806192e+00 1.37332213e+00 1.96271047e-01 -1.18007334e-02
-2.99557984e-01 4.35336053e-01 3.66306990e-01 6.18127696e-02
4.43373710e-01 1.18426837e-01 -4.74216223e-01 -1.22258402e-01
-8.89424264e-01 3.80964100e-01 9.16966856e-01 1.00715649e+00
7.07975924e-01 -1.67592227e-01 -4.69873697e-02 1.08639729e+00
4.58029360e-01 4.39186782e-01 6.10177279e-01 -1.00907946e+00
4.49456692e-01 2.52001762e-01 1.74354851e-01 -1.22406662e+00
-4.66592982e-02 -2.07541138e-01 -7.91696131e-01 -3.18390757e-01
3.51077378e-01 1.09802499e-01 -5.52703500e-01 1.88717675e+00
2.72595495e-01 2.42311284e-01 -2.51687527e-01 5.62261641e-01
3.01754683e-01 8.08880150e-01 -9.96661335e-02 8.89413729e-02
1.34893131e+00 -8.69906843e-01 -5.69189548e-01 -2.00258225e-01
7.69240558e-01 -4.10244703e-01 1.23567784e+00 3.59311640e-01
-9.74632144e-01 -3.89063179e-01 -8.22306395e-01 -4.32114750e-01
-3.51357728e-01 2.40472164e-02 7.60864377e-01 5.66308677e-01
-9.23567772e-01 7.05847263e-01 -9.70008254e-01 -2.36809060e-01
2.70343781e-01 2.92462260e-01 -2.69097328e-01 -1.53863221e-01
-8.83609295e-01 2.87373036e-01 2.61451066e-01 -3.93309295e-01
-3.57515931e-01 -8.30613792e-01 -8.68811548e-01 3.45770389e-01
2.09806010e-01 -7.21518040e-01 1.14229727e+00 -5.88855207e-01
-1.25234616e+00 3.93623412e-01 -3.33113700e-01 -7.27113903e-01
-6.26491681e-02 -3.52239877e-01 -8.71000662e-02 2.01275438e-01
-2.63754209e-03 7.05672145e-01 1.14184213e+00 -8.35803151e-01
-5.24412811e-01 -1.55943140e-01 5.44555746e-02 4.38197665e-02
-1.32643771e+00 -1.32185608e-01 -5.45823872e-01 -9.10793304e-01
2.71012709e-02 -8.38077962e-01 -3.09345406e-02 5.22677183e-01
-7.26247206e-02 -4.85411406e-01 6.60134673e-01 -5.25225282e-01
1.42397261e+00 -2.47498870e+00 3.25491190e-01 3.70623410e-01
5.91221750e-01 -8.43521729e-02 -1.94957420e-01 5.66993475e-01
1.34097397e-01 2.15572730e-01 -1.97354957e-01 -7.33803570e-01
5.53447187e-01 5.02971232e-01 -7.21113086e-01 5.53096592e-01
-1.90867662e-01 7.33171701e-01 -8.00091445e-01 -3.57578158e-01
-2.69596241e-02 6.55461431e-01 -1.10605502e+00 1.69860542e-01
-3.21097940e-01 -4.84938085e-01 -4.43135619e-01 2.90982336e-01
3.86472166e-01 -6.87069774e-01 2.30520904e-01 -2.81932771e-01
3.77736509e-01 4.04829413e-01 -1.42670333e+00 1.64374697e+00
-4.53394383e-01 5.70622563e-01 -2.85273660e-02 -1.04120302e+00
7.68791854e-01 2.54966438e-01 5.02096295e-01 1.25863403e-01
-3.99746634e-02 1.40347794e-01 -4.15334135e-01 -3.04528594e-01
6.65286660e-01 -2.14536674e-02 -8.92952383e-02 7.34029293e-01
2.08175689e-01 1.99893638e-01 1.18439436e-01 5.53408265e-01
1.13270438e+00 -3.54175389e-01 3.40455532e-01 8.45106132e-03
-5.34341298e-02 -5.26831865e-01 2.29725078e-01 9.39429462e-01
2.41435379e-01 6.30645037e-01 5.13971627e-01 -5.42061746e-01
-1.31147742e+00 -1.19502008e+00 -2.85559207e-01 1.51986969e+00
-2.39556834e-01 -8.77000868e-01 -4.21170413e-01 -5.20466506e-01
5.13569593e-01 6.34033203e-01 -6.95903182e-01 -2.56201833e-01
-2.82253802e-01 -5.79222202e-01 1.88056663e-01 7.95297742e-01
-2.53860354e-01 -4.53109235e-01 -5.27341813e-02 1.64477617e-01
-5.56188030e-03 -1.17302549e+00 -1.06275511e+00 2.81673074e-01
-1.07494700e+00 -4.86654311e-01 -4.71057802e-01 -8.11974943e-01
7.13402271e-01 5.32926977e-01 9.56126094e-01 -6.64157271e-02
3.36367004e-02 5.13392985e-01 -3.81705016e-01 -1.22912206e-01
-1.72623470e-01 5.26673906e-02 3.80562752e-01 2.37445176e-01
5.72250724e-01 -8.76344800e-01 -5.23149967e-01 8.45105276e-02
-1.36182415e+00 -1.94666281e-01 5.17679036e-01 9.08534229e-01
5.96705914e-01 -1.81381442e-02 1.99426636e-01 -9.35415447e-01
6.36962414e-01 -8.33500624e-01 -4.53529984e-01 9.27131549e-02
-4.99713391e-01 3.64500165e-01 8.84061456e-01 -8.75477493e-01
-2.37056851e-01 -1.90934211e-01 1.49996519e-01 -6.54220343e-01
-3.46864127e-02 5.95308542e-01 1.45552188e-01 2.13775307e-01
6.51890337e-01 2.23666951e-01 1.32714137e-01 -7.05287635e-01
7.63485253e-01 9.44615245e-01 3.54871154e-01 -8.60331059e-01
9.93663490e-01 5.44228673e-01 -3.02854687e-01 -9.41841602e-01
-8.62590790e-01 -6.65077209e-01 -5.42790174e-01 5.48423111e-01
4.16448414e-01 -9.05212879e-01 -4.04732317e-01 -2.01975346e-01
-9.01617706e-01 -2.67817616e-01 -5.42897701e-01 6.23703480e-01
-4.60378408e-01 6.26427889e-01 -6.44272864e-01 -4.61366475e-01
-2.65994340e-01 -6.71839237e-01 1.19016898e+00 -2.75423199e-01
-3.94994140e-01 -1.15360630e+00 1.48113966e-01 7.75394514e-02
3.76131892e-01 -1.82471916e-01 1.18767107e+00 -9.89227772e-01
-4.23588753e-01 -3.75527054e-01 -1.62063539e-01 4.92612481e-01
1.50540471e-01 -5.87057248e-02 -5.08000553e-01 -5.10026634e-01
-1.36412844e-01 -4.09761786e-01 7.92455256e-01 3.87713790e-01
1.57722747e+00 -8.17747653e-01 -2.53450185e-01 7.85414815e-01
1.27371180e+00 -4.18872088e-01 1.95015490e-01 1.30538195e-01
6.95613086e-01 3.71587545e-01 -2.43880600e-02 8.17024231e-01
3.66468161e-01 6.44462466e-01 1.56209707e-01 3.71576071e-01
2.22542211e-01 -4.74225581e-01 4.90989804e-01 1.09412324e+00
4.54470843e-01 -1.91129357e-01 -5.92634201e-01 6.32273972e-01
-1.71976912e+00 -8.73838723e-01 4.52102631e-01 2.39576244e+00
1.10393393e+00 1.47089317e-01 2.27951676e-01 7.50274807e-02
4.69237566e-01 1.84483781e-01 -5.84006608e-01 -2.08227500e-01
1.55058488e-01 2.59654075e-01 6.60311639e-01 5.69481909e-01
-9.36431050e-01 5.86680472e-01 6.05478859e+00 9.83061850e-01
-7.36289382e-01 1.50651693e-01 4.51783478e-01 -4.65985268e-01
-6.74222708e-01 -4.04087529e-02 -1.15473664e+00 5.50595164e-01
1.15610576e+00 -1.11216992e-01 7.05128133e-01 9.51759696e-01
7.80604705e-02 4.76741552e-01 -1.43403924e+00 9.33872640e-01
1.57725349e-01 -1.48129272e+00 4.34235483e-01 2.91828394e-01
6.57389641e-01 4.43720892e-02 3.43869537e-01 2.38694951e-01
5.40986300e-01 -9.85504806e-01 6.43999040e-01 1.82000160e-01
7.51102149e-01 -4.94574130e-01 2.94062376e-01 3.30919653e-01
-9.71942604e-01 -3.36177230e-01 -8.94042373e-01 5.01688831e-02
2.83342823e-02 6.11747742e-01 -7.11561382e-01 -4.72415425e-02
4.64191824e-01 8.32832694e-01 -3.52763742e-01 6.71530664e-01
1.82911724e-01 9.48215425e-01 -8.85743916e-01 -2.39015102e-01
2.47550115e-01 -2.32247740e-01 2.81340718e-01 1.18412590e+00
3.74048650e-01 2.85029504e-03 2.48426750e-01 6.48994088e-01
-4.57983375e-01 2.57663220e-01 -5.50290644e-01 -3.13237250e-01
9.18031931e-01 1.12325704e+00 -5.86412191e-01 -4.09358561e-01
-6.17519140e-01 6.82950795e-01 5.96213996e-01 4.95959491e-01
-7.09558070e-01 -4.02440280e-01 8.56481254e-01 2.61467159e-01
9.90162849e-01 -6.58260763e-01 -1.49176091e-01 -1.19617105e+00
1.24710254e-01 -8.36379230e-01 3.15025806e-01 -4.14387077e-01
-1.47832322e+00 1.19764060e-01 1.93458989e-01 -1.08688056e+00
-2.87830651e-01 -7.72678673e-01 -3.40014458e-01 5.15149832e-01
-1.18028390e+00 -8.41381669e-01 1.06491074e-01 4.58039463e-01
4.26214010e-01 -2.80943334e-01 9.90783215e-01 1.95769861e-01
-3.32926065e-01 8.60895395e-01 9.32190716e-01 2.69141227e-01
5.77274442e-01 -1.28229165e+00 4.03868437e-01 4.71132636e-01
6.82149887e-01 1.07607126e+00 6.65808380e-01 -2.47813642e-01
-1.57961380e+00 -9.77860093e-01 8.27353776e-01 -4.67063159e-01
1.13870680e+00 -7.27391362e-01 -8.99019659e-01 9.45737481e-01
-1.37073368e-01 3.47423494e-01 1.08757150e+00 4.60010231e-01
-8.24660659e-01 -2.61920810e-01 -8.88920486e-01 6.01920426e-01
7.96979785e-01 -7.30649710e-01 -5.83796263e-01 6.16286695e-01
1.14940608e+00 9.63326544e-03 -1.05194592e+00 -1.12627245e-01
6.67058825e-01 -4.42904383e-01 1.21449769e+00 -9.71510589e-01
3.67368370e-01 -1.01424314e-01 -7.26504862e-01 -1.03717959e+00
-5.67122817e-01 -5.44372559e-01 -8.63849699e-01 1.06969118e+00
1.73774317e-01 -5.63016772e-01 7.19389617e-01 7.53854215e-01
3.09899062e-01 -7.42805243e-01 -8.32616866e-01 -7.92629302e-01
1.39613017e-01 -3.29522043e-01 5.53289592e-01 9.46828127e-01
-8.12402368e-02 2.67226607e-01 -4.13210481e-01 3.19822788e-01
7.92290747e-01 1.15808196e-01 7.64736712e-01 -1.23913729e+00
-8.48256528e-01 -3.82208079e-01 -3.75465333e-01 -1.68350601e+00
1.38290569e-01 -1.02240491e+00 -2.86087781e-01 -1.18717039e+00
2.14433596e-01 -5.81476569e-01 -3.36823702e-01 4.89263862e-01
7.34446049e-02 9.95329991e-02 -8.26564729e-02 4.31919724e-01
-6.26013100e-01 6.78887784e-01 5.31019032e-01 -1.22961298e-01
-7.75992572e-02 -9.28635448e-02 -9.38880444e-01 6.84720039e-01
6.29130483e-01 -6.92898870e-01 -4.53404099e-01 -6.69063866e-01
4.31699336e-01 -3.01410466e-01 2.25070179e-01 -5.63968003e-01
4.29226428e-01 -5.48446290e-02 1.81551978e-01 -3.12216997e-01
3.80442441e-01 -8.77909243e-01 -2.17466205e-01 -5.75881004e-02
-7.95230150e-01 -8.53631496e-02 -1.14964008e-01 9.32021976e-01
6.86966404e-02 -5.17060399e-01 4.23036903e-01 2.45120436e-01
-7.91644230e-02 5.70560932e-01 -2.19408721e-01 2.79855698e-01
5.32125235e-01 -3.68914679e-02 3.97404619e-02 -4.86784875e-01
-5.53291976e-01 -2.32175663e-02 3.83664191e-01 4.20939833e-01
6.20804489e-01 -1.56826115e+00 -5.45767009e-01 2.78359681e-01
1.10459112e-01 1.69501677e-01 -5.43566942e-02 4.67267573e-01
-2.23353997e-01 3.08892936e-01 3.07751209e-01 -5.13501406e-01
-1.04164708e+00 5.43524981e-01 -2.13103835e-02 3.74939777e-02
-6.60109341e-01 8.61475706e-01 2.21963748e-01 -3.49226147e-01
7.35701621e-01 -5.25535405e-01 4.61980514e-02 -2.41017248e-03
7.32679605e-01 1.02947310e-01 -1.66357324e-01 -2.45771408e-01
4.92383465e-02 4.31338519e-01 -3.57600600e-01 -2.87370943e-03
1.67617822e+00 -1.69856280e-01 -1.12009957e-01 5.91334999e-01
1.72685575e+00 1.78059474e-01 -1.33462548e+00 -8.91863823e-01
-1.16182566e-01 -7.30552077e-01 9.22633186e-02 3.28271203e-02
-8.52648497e-01 7.58476377e-01 1.93167493e-01 4.24422354e-01
6.87696755e-01 2.00208187e-01 9.38484788e-01 8.68283689e-01
7.46938512e-02 -8.95440400e-01 2.10083008e-01 4.27837133e-01
4.92583930e-01 -9.24285054e-01 9.85603482e-02 2.14078017e-02
-2.62612760e-01 1.27655673e+00 -9.44391079e-03 -2.46940449e-01
1.04407990e+00 2.68930495e-01 -4.75397259e-01 -1.27512170e-02
-9.49666858e-01 3.67852122e-01 3.78237247e-01 4.25833076e-01
2.89644450e-01 -1.93446632e-02 4.88672890e-02 6.62599921e-01
-2.33536184e-01 -9.63365510e-02 3.72685105e-01 8.24144304e-01
-8.21397364e-01 -1.03718483e+00 -2.17826411e-01 8.83400857e-01
-4.45100009e-01 -3.93050581e-01 1.73040733e-01 3.86831552e-01
-2.22822651e-01 5.29016018e-01 3.49720180e-01 -2.64132142e-01
-7.60940015e-02 1.08149134e-01 1.37984201e-01 -7.59223700e-01
1.16689034e-01 -5.99053577e-02 -3.13166916e-01 -2.68759370e-01
-1.44728392e-01 -7.05924630e-01 -1.00828052e+00 -5.83239079e-01
-2.28492394e-01 2.39558324e-01 6.08280361e-01 8.87906849e-01
4.92146850e-01 -7.49567598e-02 6.97990775e-01 -7.64375627e-01
-1.22380912e+00 -8.27558935e-01 -6.65677190e-01 4.92408633e-01
3.79243821e-01 -5.76522291e-01 -6.57573044e-01 2.05858469e-01] | [8.92679214477539, 4.57017707824707] |
ffd54eb5-9da6-44cc-95d2-86a2a8156b88 | generating-highly-realistic-images-of-skin | 1809.01410 | null | http://arxiv.org/abs/1809.01410v2 | http://arxiv.org/pdf/1809.01410v2.pdf | Generating Highly Realistic Images of Skin Lesions with GANs | As many other machine learning driven medical image analysis tasks, skin
image analysis suffers from a chronic lack of labeled data and skewed class
distributions, which poses problems for the training of robust and
well-generalizing models. The ability to synthesize realistic looking images of
skin lesions could act as a reliever for the aforementioned problems.
Generative Adversarial Networks (GANs) have been successfully used to
synthesize realistically looking medical images, however limited to low
resolution, whereas machine learning models for challenging tasks such as skin
lesion segmentation or classification benefit from much higher resolution data.
In this work, we successfully synthesize realistically looking images of skin
lesions with GANs at such high resolution. Therefore, we utilize the concept of
progressive growing, which we both quantitatively and qualitatively compare to
other GAN architectures such as the DCGAN and the LAPGAN. Our results show that
with the help of progressive growing, we can synthesize highly realistic
dermoscopic images of skin lesions that even expert dermatologists find hard to
distinguish from real ones. | ['Nassir Navab', 'Shadi Albarqouni', 'Christoph Baur'] | 2018-09-05 | null | null | null | null | ['skin-lesion-segmentation'] | ['medical'] | [ 8.22648168e-01 5.82526565e-01 -1.95608940e-02 -1.53961241e-01
-9.68339980e-01 -5.71788788e-01 5.46099722e-01 -2.47362956e-01
-1.73016310e-01 9.12089646e-01 -1.35319993e-01 -2.11002409e-01
1.23563848e-01 -8.55823398e-01 -5.86223960e-01 -1.03768086e+00
4.27810967e-01 3.60720575e-01 4.86777499e-02 -2.63333887e-01
-2.10194796e-01 5.46774328e-01 -1.21866691e+00 2.84377754e-01
1.08075047e+00 5.99270880e-01 -2.08395943e-02 8.13866556e-01
-7.35458955e-02 8.07359993e-01 -5.31530678e-01 -6.17877662e-01
2.20761448e-01 -7.85601735e-01 -6.36472583e-01 2.81965733e-01
4.57693160e-01 -4.11864549e-01 -6.59161359e-02 1.04344916e+00
4.22874302e-01 -3.95732701e-01 8.05554152e-01 -1.10430562e+00
-7.03857481e-01 2.65136689e-01 -6.96718574e-01 -2.16178626e-01
2.17360750e-01 3.89949262e-01 3.64435971e-01 -3.48960012e-01
9.28822398e-01 8.12334955e-01 4.87653017e-01 1.30186915e+00
-1.23005021e+00 -2.07833633e-01 -2.51086950e-01 -1.73675343e-01
-9.20056343e-01 -1.97668523e-01 7.86496222e-01 -3.54555041e-01
1.04756504e-01 5.33465803e-01 6.88957155e-01 1.51995516e+00
1.45290896e-01 6.05781317e-01 1.71849537e+00 -4.57003772e-01
2.75071323e-01 3.66739899e-01 -7.12356031e-01 6.55999422e-01
1.79491505e-01 3.32151279e-02 2.38480568e-02 1.55700386e-01
1.13719058e+00 3.95070314e-02 -3.87890369e-01 -2.13794291e-01
-8.51783872e-01 7.14823008e-01 5.92955828e-01 4.55905855e-01
-5.48835695e-01 2.30997831e-01 -2.61827875e-02 2.44101938e-02
4.01874900e-01 6.59993172e-01 1.07222363e-01 3.71634513e-02
-1.02374899e+00 -5.24668433e-02 4.20021594e-01 5.03978908e-01
3.20075423e-01 3.44359517e-01 -1.04432076e-01 7.43228912e-01
-2.88414788e-02 5.56221306e-01 4.14536506e-01 -8.86515200e-01
-6.24270663e-02 5.92711329e-01 -2.80127078e-02 -6.51128352e-01
-1.26491874e-01 -3.19238305e-01 -1.06639802e+00 5.30839324e-01
5.94769061e-01 -1.50299966e-01 -1.42098737e+00 1.71808255e+00
4.84385252e-01 -1.77832365e-01 1.05409905e-01 1.03445446e+00
7.35554576e-01 4.10512179e-01 3.38867724e-01 -3.50193568e-02
1.14069510e+00 -7.65362799e-01 -7.02440381e-01 -2.91021794e-01
2.54132599e-01 -7.00729847e-01 1.11406946e+00 4.85281229e-01
-1.28578269e+00 -2.76158661e-01 -8.75162780e-01 6.59110025e-02
-1.61276102e-01 -1.36971533e-01 7.62355089e-01 9.38410223e-01
-1.02899992e+00 4.03766692e-01 -9.36669409e-01 -4.28247184e-01
8.47751200e-01 -3.54264118e-02 -6.21806324e-01 -3.74972224e-01
-8.17360818e-01 8.72366667e-01 5.02276346e-02 1.51510283e-01
-9.72057581e-01 -8.26069236e-01 -7.59549022e-01 -1.85389295e-01
2.28454024e-01 -8.48026514e-01 8.85199547e-01 -1.24829137e+00
-1.40082884e+00 1.24856138e+00 1.13503218e-01 -5.53973317e-01
8.98387372e-01 2.38685310e-01 -3.28622997e-01 4.29937571e-01
-2.71662235e-01 9.25799012e-01 1.11784744e+00 -1.59348536e+00
-1.74606740e-01 -3.74405980e-01 4.34303749e-03 -8.65567289e-03
-1.92144513e-01 -2.16374293e-01 -9.48569253e-02 -5.83227515e-01
-3.01508367e-01 -9.48549330e-01 -6.58299565e-01 3.31508279e-01
-6.54027402e-01 6.35344625e-01 8.17607641e-01 -9.58771527e-01
6.65634692e-01 -1.88417959e+00 1.13631956e-01 4.10527624e-02
3.44774187e-01 6.05022609e-01 -3.01746935e-01 3.15634996e-01
-1.33920252e-01 3.66765797e-01 -5.10334313e-01 -6.17361106e-02
-3.93823266e-01 2.69693166e-01 -4.95937392e-02 3.98832709e-01
5.00217021e-01 1.27909100e+00 -8.52293253e-01 -5.78523874e-01
4.53799993e-01 9.31301415e-01 -2.45215759e-01 2.02167362e-01
-3.87756348e-01 9.50020075e-01 -3.21192354e-01 6.85234547e-01
5.88849604e-01 -3.65008861e-01 1.56468973e-01 -5.08518741e-02
4.92311835e-01 -2.50755340e-01 -4.28864747e-01 1.56746268e+00
-6.86660588e-01 5.66079080e-01 1.09211035e-01 -6.47103131e-01
6.54357374e-01 3.81508738e-01 4.87137079e-01 -8.05070698e-01
1.59019321e-01 8.00284967e-02 2.31116414e-01 -5.30805349e-01
1.01375170e-01 -6.92495406e-01 1.88028663e-01 3.72143328e-01
-2.04049796e-01 -5.94300270e-01 1.34839248e-02 9.68493372e-02
9.88484144e-01 5.53491786e-02 2.76221279e-02 2.23703608e-01
4.60704058e-01 7.12703615e-02 1.88010082e-01 2.92418331e-01
-1.37697710e-02 1.10721278e+00 5.68018496e-01 -3.00429940e-01
-1.33335114e+00 -1.17474294e+00 -1.40211403e-01 2.67048001e-01
4.06233706e-02 1.73739910e-01 -1.39345479e+00 -7.22226441e-01
-3.43533784e-01 5.67063391e-01 -8.60156834e-01 -1.95426598e-01
-4.25865263e-01 -8.52248132e-01 5.60493112e-01 5.50971091e-01
4.03341025e-01 -1.19999552e+00 -5.22893131e-01 9.64045152e-02
-4.04536985e-02 -1.12952209e+00 1.69087283e-03 -1.25475079e-01
-5.96531987e-01 -1.17498541e+00 -1.28067744e+00 -6.54537618e-01
1.11022258e+00 -1.47711262e-01 1.06445444e+00 9.77787822e-02
-9.38411117e-01 2.86895633e-01 -3.08160663e-01 -4.63462114e-01
-1.06695986e+00 -1.05858773e-01 -3.32504660e-01 8.14751983e-02
-1.54812038e-01 -5.36597610e-01 -9.28589404e-01 2.16841131e-01
-1.36425626e+00 2.50827342e-01 8.20305288e-01 1.00593412e+00
8.65709662e-01 2.61375271e-02 6.30658805e-01 -1.48020184e+00
6.02180421e-01 -1.70615524e-01 -3.52614105e-01 3.43444228e-01
-4.56284165e-01 -1.03233226e-01 8.09461594e-01 -4.41670090e-01
-1.25384355e+00 -8.65594000e-02 -4.48190808e-01 -2.57677436e-01
-4.76304233e-01 1.99490204e-01 3.68252732e-02 -4.92632091e-01
1.06017661e+00 2.74850070e-01 3.38761359e-01 -1.46663949e-01
4.93941516e-01 4.54509765e-01 5.83198965e-01 -2.52229452e-01
1.03436899e+00 7.87385762e-01 1.08353175e-01 -7.44092584e-01
-6.81324422e-01 1.13426439e-01 -3.39517772e-01 -2.10712969e-01
9.14767444e-01 -5.22307575e-01 -1.78782314e-01 5.51293790e-01
-7.22878814e-01 -5.99320769e-01 -7.78246701e-01 -4.59084399e-02
-6.82839036e-01 2.46303305e-01 -8.29351842e-01 -6.61202431e-01
-3.91401142e-01 -9.91863370e-01 1.15352130e+00 5.98109305e-01
-1.57916933e-01 -1.33958316e+00 -2.89736502e-02 6.63432419e-01
6.47386551e-01 1.10700226e+00 1.08196640e+00 -1.17345713e-01
-5.01292944e-01 -3.71777922e-01 -6.25202581e-02 5.92382252e-01
3.71681720e-01 1.94538891e-01 -1.06898987e+00 -3.67593497e-01
-3.39552946e-02 -4.59196657e-01 8.08433414e-01 5.48539877e-01
1.22415924e+00 -2.32216835e-01 -1.44939259e-01 7.94406593e-01
1.63839650e+00 1.27532989e-01 1.01893258e+00 -2.09175482e-01
6.49269700e-01 8.67249966e-01 5.45584500e-01 -1.90522835e-01
-9.88500044e-02 2.70695537e-01 6.73999846e-01 -1.00101948e+00
-6.04362667e-01 -4.37423706e-01 -1.36602268e-01 4.05226409e-01
-3.50253284e-01 -3.21950883e-01 -6.56674802e-01 6.25150681e-01
-1.19720566e+00 -7.58119941e-01 -5.47728948e-02 1.99607098e+00
8.11124444e-01 -5.14649525e-02 9.02420729e-02 1.53686866e-01
7.73494720e-01 3.77415344e-02 -7.84129977e-01 -4.35270160e-01
-1.14102945e-01 6.49787784e-01 4.89566028e-01 2.05654725e-01
-8.70552421e-01 7.80586362e-01 6.50021315e+00 8.39945674e-01
-1.50220490e+00 5.63427731e-02 1.05739355e+00 3.65700461e-02
-5.65328062e-01 -3.12852800e-01 -1.10554881e-01 5.09898841e-01
6.97337449e-01 6.66349009e-02 2.99002558e-01 7.28432953e-01
3.41233946e-02 -9.05279294e-02 -8.98752093e-01 8.50228667e-01
2.68357550e-03 -1.25988674e+00 2.63453007e-01 3.11731011e-01
1.14488924e+00 -4.48336244e-01 6.94037795e-01 -3.56479406e-01
2.81626314e-01 -1.62697554e+00 1.48057371e-01 5.35845220e-01
1.46872234e+00 -7.47212589e-01 6.99992478e-01 2.32248262e-01
-4.73155946e-01 2.75697172e-01 -1.52424291e-01 6.08230531e-01
2.49581367e-01 6.81726158e-01 -1.14239252e+00 3.84756148e-01
1.51935518e-01 2.47270197e-01 -7.11448908e-01 7.79778779e-01
-4.83308107e-01 4.66877580e-01 -7.65826087e-03 2.61057168e-01
1.50632799e-01 -1.86884671e-01 2.03326255e-01 7.71315873e-01
3.01392198e-01 -1.84675574e-01 -5.31879187e-01 1.13504207e+00
2.98086517e-02 2.70535657e-03 -7.66712070e-01 -3.45799923e-01
-2.67702416e-02 1.56954670e+00 -7.18036413e-01 7.05228522e-02
7.66947633e-03 1.07905173e+00 -3.21235433e-02 3.85003626e-01
-8.04690063e-01 4.36787978e-02 2.70717412e-01 5.49745500e-01
-3.53412814e-02 1.29848123e-01 -2.25915328e-01 -1.05893481e+00
-5.07204719e-02 -1.03981221e+00 6.21692166e-02 -8.79529953e-01
-1.29843366e+00 9.06229198e-01 -4.62142438e-01 -1.13829792e+00
-6.39017582e-01 -6.15340054e-01 -6.55250967e-01 7.14216113e-01
-1.69256854e+00 -1.60769606e+00 -5.62644839e-01 5.77198267e-01
1.65370584e-01 1.10033214e-01 9.83032882e-01 3.39594111e-02
-2.98185885e-01 5.64888418e-01 -9.49844271e-02 9.13152620e-02
7.28681147e-01 -1.48389435e+00 3.30199033e-01 7.43161738e-01
1.60330430e-01 3.13740283e-01 6.42486751e-01 -4.24860299e-01
-1.11442471e+00 -1.06171978e+00 1.95531964e-01 -2.72999138e-01
1.38622224e-01 -1.83851048e-01 -8.61496270e-01 3.63847673e-01
3.28483909e-01 2.79868841e-01 7.01193988e-01 -4.31547254e-01
-1.93361431e-01 -3.87908295e-02 -1.74344981e+00 6.95863307e-01
6.53484285e-01 -4.50942010e-01 2.89865509e-02 5.65293849e-01
2.22937018e-01 -3.76170069e-01 -1.09136331e+00 3.75996172e-01
3.81235301e-01 -1.16167641e+00 1.02445173e+00 -5.81018269e-01
9.57072318e-01 2.56678499e-02 3.63300532e-01 -1.58383775e+00
2.76967824e-01 -4.42332566e-01 3.28163385e-01 1.20174730e+00
3.63062024e-01 -6.01597548e-01 1.15993118e+00 6.52660310e-01
3.18358958e-01 -8.06550384e-01 -6.16991580e-01 -6.41048610e-01
3.77319723e-01 5.96485771e-02 2.95980573e-01 9.69876289e-01
-3.15753281e-01 -1.15360603e-01 -3.54667097e-01 -2.54284233e-01
7.50279546e-01 6.57825693e-02 6.00751042e-01 -8.47383261e-01
-2.93424666e-01 -3.01037431e-01 -3.84059787e-01 -2.42673352e-01
-1.25948742e-01 -6.85622811e-01 -7.19168484e-02 -1.57093465e+00
2.85454243e-01 -6.04015112e-01 5.68742529e-02 3.20484251e-01
-1.63481295e-01 7.45803297e-01 2.71602105e-02 -7.46813938e-02
-8.61721113e-02 1.34792581e-01 2.06836915e+00 -2.02023253e-01
2.15059683e-01 -1.83100075e-01 -8.76029849e-01 6.89068556e-01
7.87490845e-01 -3.20967793e-01 -5.49143791e-01 -1.48968756e-01
3.40917446e-02 3.05534691e-01 5.37985206e-01 -9.35961843e-01
-2.81945486e-02 -1.41855463e-01 5.22994459e-01 1.13799252e-01
4.61031228e-01 -6.70664549e-01 4.89148468e-01 6.58814311e-01
-3.22131157e-01 -5.78595281e-01 1.96989197e-02 4.24607605e-01
-4.99688417e-01 -1.26950145e-01 1.26504111e+00 -5.81761003e-01
-3.38268638e-01 2.93456107e-01 -1.85614064e-01 4.24459279e-02
1.29268420e+00 -2.39636421e-01 -5.44033885e-01 -6.78311169e-01
-8.38827550e-01 -2.23916501e-01 1.04556143e+00 1.28829643e-01
7.28593946e-01 -1.21769726e+00 -8.44447792e-01 1.65877193e-01
-4.88527417e-02 3.75223517e-01 6.33343995e-01 8.84618521e-01
-9.82572198e-01 1.92709342e-01 -6.16174221e-01 -5.76949894e-01
-1.12371778e+00 5.73344588e-01 4.75813955e-01 -4.99400586e-01
-4.13782716e-01 7.27742195e-01 3.37059289e-01 -2.24483892e-01
-2.57735878e-01 -1.00250887e-02 1.13716640e-01 -3.27015430e-01
4.53784406e-01 -1.32255539e-01 6.40635639e-02 -3.82523090e-01
-5.06487116e-02 6.17492080e-01 -1.50630131e-01 4.62031960e-02
1.20310402e+00 1.05693586e-01 8.93646330e-02 -1.76139131e-01
8.37057412e-01 1.25691175e-01 -1.40396678e+00 1.17026024e-01
-4.62668538e-01 -5.53795278e-01 -2.06699923e-01 -1.07928681e+00
-1.44632459e+00 1.11604762e+00 5.68496943e-01 3.85488778e-01
1.43330872e+00 -7.12670311e-02 8.79325271e-01 -3.99573952e-01
4.73714650e-01 -7.41352916e-01 1.60934940e-01 -4.58309412e-01
7.22678125e-01 -1.27638781e+00 -2.95278639e-01 -8.14377606e-01
-8.37221682e-01 8.49286079e-01 5.66684186e-01 -2.71303684e-01
-1.39143132e-02 6.69618309e-01 4.41749543e-01 -4.55674902e-02
-5.76516211e-01 -2.55438507e-01 1.94289550e-01 1.17309761e+00
2.78182477e-01 6.51778877e-02 -2.84129202e-01 2.17176117e-02
-6.39383271e-02 -8.51721764e-02 7.20059812e-01 6.50264740e-01
1.25939816e-01 -1.38895917e+00 -3.45645040e-01 3.03962886e-01
-7.74604619e-01 7.11762533e-02 -6.61426067e-01 1.08636308e+00
1.89366881e-02 6.52124345e-01 -8.52551982e-02 -1.72772989e-01
-1.38347538e-03 -1.57641485e-01 9.86650288e-01 -5.66290617e-01
-4.79120523e-01 -5.52253518e-03 -5.22757648e-03 -3.82360101e-01
-4.97738808e-01 -3.57038349e-01 -9.42358971e-01 -2.13340878e-01
2.45255437e-02 -2.39416689e-01 9.02949870e-01 6.65924549e-01
-5.16363792e-02 7.38623023e-01 5.38297176e-01 -3.48011822e-01
-2.79753655e-01 -7.16252744e-01 -7.75956929e-01 8.16606224e-01
2.55745888e-01 -1.54256329e-01 -2.61296183e-01 2.55658031e-01] | [14.19669246673584, -1.9991936683654785] |
8198dd11-3aeb-4364-b67f-0454f4fd8032 | subspace-sparse-representation | 1507.01307 | null | http://arxiv.org/abs/1507.01307v1 | http://arxiv.org/pdf/1507.01307v1.pdf | Subspace-Sparse Representation | Given an overcomplete dictionary $A$ and a signal $b$ that is a linear
combination of a few linearly independent columns of $A$, classical sparse
recovery theory deals with the problem of recovering the unique sparse
representation $x$ such that $b = A x$. It is known that under certain
conditions on $A$, $x$ can be recovered by the Basis Pursuit (BP) and the
Orthogonal Matching Pursuit (OMP) algorithms. In this work, we consider the
more general case where $b$ lies in a low-dimensional subspace spanned by some
columns of $A$, which are possibly linearly dependent. In this case, the
sparsest solution $x$ is generally not unique, and we study the problem that
the representation $x$ identifies the subspace, i.e. the nonzero entries of $x$
correspond to dictionary atoms that are in the subspace. Such a representation
$x$ is called subspace-sparse. We present sufficient conditions for
guaranteeing subspace-sparse recovery, which have clear geometric
interpretations and explain properties of subspace-sparse recovery. We also
show that the sufficient conditions can be satisfied under a randomized model.
Our results are applicable to the traditional sparse recovery problem and we
get conditions for sparse recovery that are less restrictive than the canonical
mutual coherent condition. We also use the results to analyze the sparse
representation based classification (SRC) method, for which we get conditions
to show its correctness. | ['R. Vidal', 'C. You'] | 2015-07-06 | null | null | null | null | ['sparse-representation-based-classification'] | ['computer-vision'] | [ 3.15271318e-01 -1.87916085e-01 -2.70693719e-01 2.29357220e-02
-6.11581624e-01 -4.98165935e-01 -1.18071057e-01 -4.55239058e-01
2.23905072e-01 6.26650274e-01 2.68827885e-01 -9.67907682e-02
-4.16652381e-01 -7.58331895e-01 -7.45454252e-01 -1.03158724e+00
-2.89012313e-01 2.25996822e-01 -5.72187901e-01 -4.72241253e-01
2.64127374e-01 4.73798066e-01 -1.41582441e+00 7.48105496e-02
6.99364603e-01 8.47716570e-01 2.70890653e-01 3.02875221e-01
-3.24118957e-02 5.41005850e-01 -1.90249145e-01 2.28898332e-01
6.66741133e-01 -7.43307710e-01 -6.59719706e-01 3.56185585e-01
3.10182720e-01 3.89237739e-02 -7.11853683e-01 1.35265434e+00
1.00625500e-01 1.85446322e-01 5.04417598e-01 -1.11110938e+00
-5.52343369e-01 6.86298549e-01 -9.44253266e-01 1.13192365e-01
6.02390230e-01 -4.82531071e-01 7.83617675e-01 -1.13445544e+00
7.70102203e-01 9.23169315e-01 6.98102772e-01 2.30449900e-01
-1.22937441e+00 -7.55546272e-01 -5.63748740e-03 3.05369198e-01
-2.05634785e+00 -8.46821189e-01 1.01250100e+00 -3.90028417e-01
6.34665489e-01 4.99574870e-01 3.85018677e-01 5.97104907e-01
-5.61269335e-02 5.09216964e-01 1.13390779e+00 -7.73436666e-01
3.45576882e-01 -6.40535504e-02 4.86327291e-01 8.05476308e-01
7.30226874e-01 2.20974043e-01 -5.96678495e-01 -3.44712675e-01
8.05158317e-01 1.84256688e-01 -7.88402677e-01 -7.60375738e-01
-1.28036308e+00 1.08964288e+00 3.61507505e-01 7.84880996e-01
-6.08550906e-01 -1.03024721e-01 -5.17503098e-02 3.85131955e-01
-1.80487260e-01 1.63530931e-01 1.67618722e-01 4.06565785e-01
-1.08583677e+00 9.92994308e-02 9.58864212e-01 1.27548480e+00
1.15512872e+00 6.03267372e-01 4.53861177e-01 7.19110847e-01
2.35569566e-01 1.06179953e+00 4.42518860e-01 -1.12193859e+00
3.72801512e-01 2.51484275e-01 1.31813154e-01 -1.51896942e+00
1.22690098e-02 -6.38078868e-01 -1.09263051e+00 -1.71996593e-01
2.17435613e-01 3.06297272e-01 -5.48079729e-01 1.86410153e+00
-6.71210140e-02 6.26467586e-01 3.86227041e-01 1.01435959e+00
5.95615983e-01 7.16091990e-01 -4.09426898e-01 -8.06932926e-01
8.95721018e-01 -4.79703695e-01 -7.53733099e-01 -2.52579004e-01
4.70888972e-01 -8.60353410e-01 4.69785005e-01 3.27057689e-01
-1.07634747e+00 -3.40827912e-01 -1.00425351e+00 5.09782672e-01
1.75782353e-01 2.10670546e-01 5.52293062e-01 4.37224448e-01
-1.05025887e+00 1.68645546e-01 -5.05519986e-01 -4.20619726e-01
-2.47382030e-01 4.07125831e-01 -7.48987854e-01 -7.86858261e-01
-8.24254572e-01 5.41787863e-01 1.21100657e-01 3.18083495e-01
-8.88239861e-01 -1.28614143e-01 -1.05847335e+00 8.94857198e-02
2.35995546e-01 -6.42473280e-01 5.62191427e-01 -1.10113800e+00
-7.35017538e-01 8.56712162e-01 -8.25239003e-01 -2.98484683e-01
-2.20159203e-01 1.88517556e-01 -5.73543191e-01 4.91235763e-01
4.57428098e-01 -2.05496311e-01 1.18256104e+00 -1.50811839e+00
-1.67691931e-01 -6.29873216e-01 -3.54471266e-01 3.78540643e-02
-2.11658478e-02 -5.83265834e-02 -3.56498837e-01 -6.18862271e-01
1.14320838e+00 -1.04814279e+00 -5.04874051e-01 -6.77073002e-01
-5.82305491e-01 3.93281758e-01 9.52486157e-01 -5.92122912e-01
1.39459705e+00 -2.50841689e+00 5.32771409e-01 8.74707162e-01
1.97366998e-01 -1.38620049e-01 -3.24739635e-01 7.40672588e-01
-7.78925121e-01 -1.99617386e-01 -7.26670623e-01 -6.26305863e-02
-3.40497077e-01 1.32878259e-01 -6.62406921e-01 9.76302922e-01
-4.51940209e-01 3.53540719e-01 -5.18388450e-01 -2.38455161e-01
1.15904272e-01 3.67287934e-01 -5.40222287e-01 5.88705726e-02
2.79951304e-01 4.93623406e-01 -5.45545816e-01 1.04546964e+00
8.18112433e-01 -2.93316901e-01 3.36857319e-01 -2.75315434e-01
-2.41082221e-01 -3.65391344e-01 -1.79676092e+00 1.72630882e+00
-6.85629398e-02 4.01851565e-01 7.77669728e-01 -1.76129568e+00
1.04879367e+00 4.12299305e-01 7.55163312e-01 -3.73367310e-01
5.40010929e-02 6.38540804e-01 -1.54089388e-02 -3.83009136e-01
3.95001262e-01 -5.61611176e-01 -9.81605351e-02 3.71444553e-01
-2.42658466e-01 9.88700464e-02 3.84105630e-02 4.06635970e-01
1.19745886e+00 -4.92434710e-01 5.92719197e-01 -5.53609014e-01
7.50061035e-01 1.24120533e-01 5.84852278e-01 8.27636838e-01
-7.89599270e-02 5.68203032e-01 -7.32151121e-02 -4.00860608e-02
-7.98035502e-01 -8.44778597e-01 -2.21331775e-01 4.51886028e-01
4.48567837e-01 -3.32521379e-01 -4.01835769e-01 1.24730682e-02
-3.59349549e-02 4.60620612e-01 -3.11343908e-01 2.34703869e-02
-8.62932146e-01 -5.30130863e-01 -3.03758532e-02 3.08363885e-02
4.36042398e-01 -7.51338124e-01 -1.16690397e-02 2.62135714e-01
-4.52866167e-01 -7.85088778e-01 -3.80860686e-01 9.56435800e-02
-1.08298600e+00 -1.33449113e+00 -8.47994030e-01 -9.50615585e-01
1.05744922e+00 1.11930537e+00 8.43476295e-01 9.31322575e-02
-1.36783570e-01 7.76639283e-01 -3.60728979e-01 3.25727344e-01
-3.61274630e-02 -3.94442230e-01 4.59691256e-01 2.97714353e-01
3.57797295e-01 -8.25444400e-01 -4.43493754e-01 3.92817289e-01
-9.09272611e-01 -3.95188957e-01 5.40038824e-01 7.79019773e-01
1.02458322e+00 4.76485848e-01 4.28670794e-01 -7.35250711e-01
2.90786713e-01 -8.91656458e-01 -3.25209141e-01 6.83005005e-02
-4.15606439e-01 -5.60812885e-03 6.73873663e-01 -4.85500097e-02
-5.80036759e-01 3.83103579e-01 7.68333972e-02 -1.02746093e+00
1.55791581e-01 9.57641363e-01 -3.63281518e-01 -4.25053328e-01
8.32239687e-01 6.52568877e-01 7.38590807e-02 -6.15274727e-01
3.74391615e-01 4.96833444e-01 6.09058201e-01 -6.99642956e-01
1.07203114e+00 7.56557286e-01 2.42130473e-01 -1.23783994e+00
-4.38228607e-01 -9.95157957e-01 -3.58073622e-01 1.35666206e-01
1.00028269e-01 -1.16389561e+00 -2.50377506e-01 -2.13927463e-01
-7.40734577e-01 2.78604627e-01 -4.52354342e-01 7.20567942e-01
-9.64724720e-01 7.43557334e-01 -2.92035699e-01 -8.00627530e-01
-1.38727054e-01 -1.07665300e+00 6.76547170e-01 -2.14230925e-01
-2.02656731e-01 -6.47679746e-01 3.41482371e-01 2.13401571e-01
3.45971704e-01 1.73384264e-01 8.01681221e-01 -3.18004847e-01
-6.92103088e-01 -2.61926353e-01 -4.34347056e-02 1.42039359e-01
-4.86789495e-02 -4.76985931e-01 -3.97497892e-01 -7.62886107e-01
5.81970930e-01 2.04814970e-01 9.14961457e-01 8.08973849e-01
8.64900410e-01 -6.15457833e-01 -6.57264709e-01 8.45677376e-01
1.97631037e+00 3.35093349e-01 7.51974940e-01 8.73986110e-02
6.48895025e-01 5.17087221e-01 3.12255710e-01 4.15990233e-01
-7.25369677e-02 5.53580761e-01 2.85968989e-01 1.47928998e-01
2.12956488e-01 5.21212630e-02 3.41270298e-01 1.16343689e+00
-2.30263174e-01 1.52918652e-01 -7.42251158e-01 8.30052733e-01
-1.85529387e+00 -1.44462502e+00 -2.78124541e-01 2.38963318e+00
4.84834760e-01 -6.80901110e-01 6.45660460e-02 4.71871227e-01
1.05487990e+00 2.33941704e-01 -3.11745286e-01 -6.81445077e-02
-3.55193824e-01 6.13680363e-01 6.16442502e-01 8.01830351e-01
-8.09967339e-01 5.88301063e-01 7.35366344e+00 6.01418495e-01
-9.99979198e-01 1.98964942e-02 -2.69672368e-02 6.95308298e-02
-6.99819088e-01 2.12548435e-01 -7.37109423e-01 2.03663066e-01
7.70010889e-01 -2.32920930e-01 6.48306847e-01 1.13183415e+00
1.40897885e-01 2.07917422e-01 -9.66726959e-01 1.51562691e+00
4.30827409e-01 -1.35853946e+00 5.81599101e-02 1.80463091e-01
8.62258554e-01 -2.80982763e-01 1.12666607e-01 8.70364755e-02
2.29271650e-01 -1.16109240e+00 3.29499453e-01 2.70591706e-01
9.95769560e-01 -6.05836511e-01 2.89651990e-01 5.52013755e-01
-1.20899856e+00 -1.55405164e-01 -6.45660937e-01 -5.07289991e-02
1.77228913e-01 7.63495445e-01 -2.81216055e-01 6.27201080e-01
4.93900537e-01 1.16876090e+00 2.05738246e-01 9.08679485e-01
1.78521663e-01 4.93045807e-01 -2.80830175e-01 5.97424686e-01
5.42856678e-02 -7.52825558e-01 9.53496039e-01 1.05090928e+00
1.16889489e+00 8.98528337e-01 8.66533965e-02 7.01447666e-01
1.90321431e-01 2.74140120e-01 -1.19982302e+00 -3.84339318e-02
6.68327272e-01 8.67728472e-01 -3.27154100e-01 -2.89472103e-01
-4.12225842e-01 9.18633997e-01 -1.11604474e-01 5.87100923e-01
-1.94326490e-01 -1.49602473e-01 7.61793911e-01 2.74620980e-01
5.80033183e-01 -3.20374697e-01 -2.10907862e-01 -1.68986177e+00
-1.45760790e-01 -1.17722964e+00 4.56746817e-01 -5.49449027e-01
-1.32770371e+00 6.02409244e-01 -9.33959186e-02 -1.67043436e+00
-3.10894549e-01 -3.22345495e-01 -3.57935488e-01 1.17387795e+00
-1.39728999e+00 -6.45760834e-01 -3.98886353e-02 1.38823378e+00
1.14035860e-01 -4.51616108e-01 1.07972562e+00 1.19798891e-01
-6.74671650e-01 1.05502568e-01 4.79737073e-01 -1.89300016e-01
2.41543785e-01 -7.01514482e-01 -7.57377982e-01 1.28170025e+00
2.16818884e-01 1.23172116e+00 8.80582213e-01 -5.57015836e-01
-1.94862914e+00 -7.96646833e-01 9.80777383e-01 1.07634246e-01
4.68549579e-01 3.44595820e-01 -8.21000993e-01 1.06943655e+00
6.52416572e-02 -5.33437403e-03 1.13815498e+00 1.40545219e-01
-6.55315101e-01 -1.74435601e-01 -1.12944734e+00 1.93671837e-01
1.03341830e+00 -5.96994221e-01 -7.02199280e-01 3.10259342e-01
1.51175186e-01 -1.84886366e-01 -6.57320976e-01 4.36802775e-01
2.98151731e-01 -8.78078043e-01 1.39057028e+00 -4.09459680e-01
1.61485791e-01 -7.09325075e-01 -9.76246119e-01 -1.08687699e+00
-5.85486114e-01 -5.67073107e-01 -9.26013067e-02 3.92799228e-01
-1.58833981e-01 -4.75076646e-01 8.21612656e-01 2.71285325e-01
-1.97769225e-01 -8.90323445e-02 -1.34395254e+00 -8.52050304e-01
-2.88318396e-01 -2.39750609e-01 4.49512988e-01 1.43881655e+00
3.76950115e-01 2.00886637e-01 -6.51183248e-01 3.87997836e-01
9.71346200e-01 7.82724321e-01 5.10283351e-01 -1.05608535e+00
-5.38461149e-01 -1.36729240e-01 -2.45996758e-01 -1.11668575e+00
4.70954448e-01 -9.47882950e-01 8.24555457e-02 -1.14934671e+00
3.38351727e-01 -9.50245798e-01 -3.13580483e-01 1.63235858e-01
2.74932057e-01 1.93853542e-01 2.81185120e-01 1.02739275e+00
-2.01665282e-01 2.47661814e-01 1.00204039e+00 -2.09576458e-01
-2.64844060e-01 -1.01782501e-01 -1.10889685e+00 6.33204043e-01
4.79530364e-01 -3.22735995e-01 -3.17926377e-01 -3.57075840e-01
-9.05407444e-02 8.17074716e-01 1.11776188e-01 -9.01482105e-01
1.44131988e-01 -1.55259281e-01 -9.31483507e-02 -4.98829991e-01
3.66491556e-01 -1.16072643e+00 8.38871598e-01 5.31631768e-01
4.72463714e-03 -1.79583803e-01 -2.02428579e-01 7.03461826e-01
-5.35689056e-01 -5.70646942e-01 7.31912494e-01 -2.97713429e-01
-7.85257399e-01 4.08630848e-01 -2.07269117e-01 -4.60588932e-01
8.96870255e-01 -4.65649933e-01 -9.33829322e-03 -5.59229732e-01
-1.01630604e+00 -3.47599268e-01 5.92871964e-01 -3.29441607e-01
7.76954889e-01 -1.57888949e+00 -8.25031519e-01 5.87289870e-01
1.66289330e-01 -2.73786485e-01 5.14270484e-01 8.38357449e-01
-4.29173917e-01 6.26077533e-01 -9.90754738e-02 -6.35155022e-01
-9.74646986e-01 8.43032360e-01 1.62058905e-01 2.45709307e-02
-5.73939860e-01 6.38203979e-01 2.10996866e-01 1.13813832e-01
-5.77752590e-02 1.08699694e-01 -1.81685194e-01 -3.34671915e-01
6.91866517e-01 2.55626947e-01 -2.42611423e-01 -1.39678240e+00
-4.65077341e-01 8.87959957e-01 2.74202853e-01 -1.95617333e-01
1.25792849e+00 -2.32656091e-01 -6.45769060e-01 2.44269725e-02
1.47813058e+00 7.04993129e-01 -5.13701856e-01 -5.84315538e-01
-5.16358495e-01 -7.28721023e-01 -3.96121144e-02 -5.17213270e-02
-1.38089931e+00 4.40731049e-01 2.67422974e-01 2.33063489e-01
1.39515805e+00 1.49720430e-01 5.66947937e-01 3.75729233e-01
8.22552443e-01 -4.80993360e-01 -4.11666393e-01 3.69083613e-01
1.02896273e+00 -9.41554010e-01 -4.25046012e-02 -8.36005867e-01
-1.99222788e-01 1.00176144e+00 4.07839157e-02 -5.32785833e-01
7.58888662e-01 -5.92327826e-02 -2.77611166e-01 -3.06984931e-01
-1.03846200e-01 -3.04945558e-01 5.83565310e-02 6.53977871e-01
3.62291038e-01 3.11083317e-01 -6.99501455e-01 6.50717258e-01
-2.80443311e-01 -7.80524611e-02 3.92228752e-01 7.61426270e-01
-6.60935879e-01 -1.20320630e+00 -1.12939239e+00 1.90212071e-01
-7.12481365e-02 -1.72089756e-01 3.35058533e-02 4.29393649e-01
-3.45122516e-02 1.01400995e+00 -2.53022522e-01 -3.22506040e-01
1.27939329e-01 -2.88920477e-02 4.70099419e-01 -7.19345450e-01
8.51366222e-02 4.67470795e-01 -2.17047542e-01 -7.26376534e-01
-7.00261533e-01 -8.10685456e-01 -1.15567398e+00 -6.80132687e-01
-1.26855269e-01 6.28641963e-01 3.20447654e-01 6.89341843e-01
1.87992692e-01 -1.93626955e-01 8.68770897e-01 -4.61979121e-01
-5.00039518e-01 -5.65142334e-01 -1.44186032e+00 4.32988495e-01
4.02529597e-01 -4.30521637e-01 -7.61039019e-01 3.01380634e-01] | [7.3438496589660645, 4.3992133140563965] |
b3ae61c5-da89-473e-9084-a96395d5b155 | semantic-head-enhanced-pedestrian-detection | 1911.11985 | null | https://arxiv.org/abs/1911.11985v1 | https://arxiv.org/pdf/1911.11985v1.pdf | Semantic Head Enhanced Pedestrian Detection in a Crowd | Pedestrian detection in the crowd is a challenging task because of intra-class occlusion. More prior information is needed for the detector to be robust against it. Human head area is naturally a strong cue because of its stable appearance, visibility and relative location to body. Inspired by it, we adopt an extra branch to conduct semantic head detection in parallel with traditional body branch. Instead of manually labeling the head regions, we use weak annotations inferred directly from body boxes, which is named as `semantic head'. In this way, the head detection is formulated into using a special part of labeled box to detect the corresponding part of human body, which surprisingly improves the performance and robustness to occlusion. Moreover, the head-body alignment structure is explicitly explored by introducing Alignment Loss, which functions in a self-supervised manner. Based on these, we propose the head-body alignment net (HBAN) in this work, which aims to enhance pedestrian detection by fully utilizing the human head prior. Comprehensive evaluations are conducted to demonstrate the effectiveness of HBAN on CityPersons dataset. | ['Huimin Ma', 'Ruiqi Lu'] | 2019-11-27 | null | null | null | null | ['head-detection'] | ['computer-vision'] | [-2.81473517e-01 2.08883330e-01 4.30775136e-02 -4.78939205e-01
-9.26356763e-02 -5.35776876e-02 4.76579756e-01 -2.63965223e-02
-5.83117664e-01 4.53205884e-01 3.35410535e-01 1.82461441e-01
7.08004355e-01 -6.89245403e-01 -6.32589877e-01 -8.12332571e-01
4.11430120e-01 7.66041651e-02 9.62165952e-01 -1.63587049e-01
-1.48302823e-01 2.22907096e-01 -1.40176892e+00 -2.34668463e-01
7.57404387e-01 9.18752193e-01 1.87773168e-01 1.07154995e-01
1.89457387e-01 6.18593812e-01 -4.41370517e-01 -6.47195280e-01
2.57281244e-01 -3.61695558e-01 -3.57757360e-01 5.68734109e-01
3.89369130e-01 -5.99768817e-01 -1.87138215e-01 1.07883906e+00
7.66983509e-01 1.79177582e-01 4.66580480e-01 -1.22746801e+00
1.04074821e-01 2.20862731e-01 -1.20195472e+00 1.92913756e-01
4.17849839e-01 4.34901446e-01 8.17110598e-01 -7.56671309e-01
2.04971656e-01 1.47711241e+00 5.51520228e-01 5.21878302e-01
-7.64802575e-01 -7.94485748e-01 6.31443918e-01 1.37862444e-01
-1.42462826e+00 -5.48468411e-01 8.58111441e-01 -3.54982138e-01
1.71529740e-01 -6.64805099e-02 7.57180810e-01 7.78144360e-01
-2.81433254e-01 1.21334815e+00 8.37091565e-01 -3.83891851e-01
2.05050372e-02 1.23083852e-01 3.39225024e-01 9.62243855e-01
4.94197011e-01 1.67035293e-02 -2.18645766e-01 2.51993418e-01
5.21959126e-01 -7.21955970e-02 -2.44380504e-01 -6.91830635e-01
-7.61591375e-01 5.54159343e-01 7.66496539e-01 -2.28738293e-01
-1.95198506e-01 -8.51452053e-02 5.50002575e-01 -6.08341753e-01
2.01442480e-01 -5.90414643e-01 -1.62623242e-01 4.22730863e-01
-8.47742796e-01 2.03813270e-01 3.46042931e-01 1.07440567e+00
6.65372193e-01 -1.43413663e-01 -5.14200211e-01 7.36789763e-01
7.92495966e-01 5.17404854e-01 2.36116275e-01 -5.92374086e-01
5.74710846e-01 8.42674553e-01 2.99407542e-01 -9.57249701e-01
-6.05462432e-01 -7.11464584e-01 -8.45401525e-01 8.68307501e-02
6.73270226e-01 -2.31378511e-01 -8.81000519e-01 1.69881070e+00
1.06721079e+00 -1.11686885e-01 -2.94157833e-01 1.30260825e+00
1.02107465e+00 3.30298841e-01 4.90702778e-01 -1.62711546e-01
1.86973321e+00 -1.23951626e+00 -6.00683272e-01 -6.03671491e-01
3.64688575e-01 -5.42590082e-01 8.79491866e-01 -6.63510384e-03
-8.00728738e-01 -6.50366545e-01 -8.63897204e-01 -1.49010733e-01
-4.94587086e-02 3.98407400e-01 5.42116523e-01 7.83504307e-01
-5.07796347e-01 -1.92998186e-01 -7.57145941e-01 -4.90746647e-01
5.70207357e-01 1.62827894e-01 -2.55979717e-01 -1.82605833e-01
-9.91078436e-01 7.53286719e-01 5.57624757e-01 3.91484797e-01
-7.00329840e-01 4.89485078e-02 -1.17619133e+00 -1.21443957e-01
5.55199087e-01 -8.48734975e-01 1.09570801e+00 -5.58849275e-01
-1.39194369e+00 9.46406841e-01 -4.30214852e-01 -3.90384585e-01
9.32905614e-01 -2.87725657e-01 -1.53328598e-01 2.21299864e-02
3.74437004e-01 8.56752515e-01 9.27794993e-01 -1.29111218e+00
-9.56561804e-01 -6.60150170e-01 -1.37582749e-01 2.02182695e-01
-2.07560793e-01 2.89612502e-01 -1.11659491e+00 -5.60556233e-01
2.50529319e-01 -8.88772190e-01 -2.50040948e-01 8.88097882e-02
-7.59831965e-01 -4.03063416e-01 7.23511755e-01 -9.54055011e-01
1.20508158e+00 -2.02511191e+00 -1.78984880e-01 1.74006179e-01
3.45526904e-01 3.51079702e-01 2.86658138e-01 -2.36676931e-01
2.87441581e-01 -4.50046480e-01 -3.05732667e-01 -5.82495451e-01
-7.46454075e-02 1.29451707e-01 1.51961118e-01 7.69108713e-01
1.09793156e-01 7.49641657e-01 -7.99888194e-01 -1.11159551e+00
2.89292932e-01 3.67706090e-01 -3.75830710e-01 2.36442894e-01
3.26076150e-02 5.53628922e-01 -6.38548434e-01 8.20000291e-01
1.03481770e+00 -1.23386025e-01 -1.71380445e-01 -3.61701578e-01
-2.51331359e-01 -1.21511519e-01 -1.24827409e+00 1.22106445e+00
-9.64130759e-02 9.62103307e-02 3.97327513e-01 -6.37417495e-01
8.65447998e-01 7.48311654e-02 4.34474787e-03 -6.36805952e-01
3.40617925e-01 -5.78268431e-02 1.90967575e-01 -4.56832379e-01
6.95477203e-02 1.66602209e-01 1.93192393e-01 -8.16179737e-02
-3.17428708e-01 4.47637916e-01 3.26642066e-01 1.51637578e-04
5.88796675e-01 2.87219524e-01 3.98079962e-01 -1.60083324e-01
9.80976939e-01 -4.41220999e-01 1.03210711e+00 5.10060787e-01
-7.24468529e-01 5.91584086e-01 1.71658814e-01 -3.16965431e-01
-7.39653766e-01 -7.52187967e-01 -1.23051122e-01 1.21052897e+00
5.03743589e-01 -3.33025664e-01 -1.08612299e+00 -7.79847383e-01
-1.63693562e-01 2.62627929e-01 -5.86468160e-01 -1.19331656e-02
-8.55549574e-01 -8.04511428e-01 2.38113031e-01 8.66463661e-01
1.03120792e+00 -9.46812868e-01 -7.99410164e-01 1.42873973e-01
-2.93097675e-01 -1.25259352e+00 -6.30630016e-01 -1.05515651e-01
-4.45007294e-01 -1.19856048e+00 -1.17907965e+00 -6.79671884e-01
7.10871398e-01 4.17168856e-01 7.01499760e-01 1.59948900e-01
-2.53730655e-01 -7.97877237e-02 -1.59264207e-01 -4.58550185e-01
1.32395491e-01 -1.15330301e-01 7.26694614e-02 3.56432766e-01
2.70556450e-01 -3.48531663e-01 -1.14827132e+00 6.59859538e-01
-4.11078811e-01 3.18531431e-02 7.05886900e-01 6.43092692e-01
3.69210184e-01 1.73697636e-01 1.27697721e-01 -5.59932172e-01
-4.93245982e-02 -8.29601884e-02 -6.73104584e-01 1.55455723e-01
-8.49912991e-04 -2.76506662e-01 3.53383601e-01 -1.86201915e-01
-1.22998309e+00 5.46719790e-01 -2.43100598e-01 -9.11950916e-02
-4.20872480e-01 -2.76914656e-01 -1.08417952e+00 -5.61646484e-02
2.41009817e-01 2.25101799e-01 -7.48656690e-02 -5.74406207e-01
3.72427344e-01 5.94528675e-01 7.75987804e-01 -5.20199537e-01
8.00175786e-01 7.00772226e-01 1.19043209e-01 -7.97149301e-01
-1.15268350e+00 -8.21297228e-01 -8.42589021e-01 -3.73257279e-01
1.02623200e+00 -1.02473998e+00 -8.56659651e-01 6.41623914e-01
-1.28431141e+00 7.01862723e-02 2.37415820e-01 3.38366061e-01
-9.44233872e-03 7.92498350e-01 -5.75759411e-01 -1.12227976e+00
-3.86982083e-01 -1.07986939e+00 1.38043559e+00 4.33603615e-01
1.40054196e-01 -6.20627880e-01 -4.35405225e-01 4.34001476e-01
9.66444891e-03 2.90108740e-01 2.65535802e-01 -3.26792359e-01
-3.81481975e-01 -4.83397782e-01 -6.13037109e-01 3.84269863e-01
-1.33106768e-01 -1.98740497e-01 -1.17325175e+00 -2.69772232e-01
-2.35135004e-01 -2.07074266e-03 1.01813340e+00 5.55969656e-01
7.64320791e-01 -2.41007626e-01 -6.61937892e-01 5.42506814e-01
6.92590356e-01 -1.94105402e-01 4.25249040e-01 4.79240358e-01
9.24526453e-01 9.96391833e-01 6.13846123e-01 5.82379282e-01
6.08631134e-01 8.17028224e-01 4.75565314e-01 -3.88847172e-01
-3.67675930e-01 -4.24127221e-01 4.12851006e-01 3.61183763e-01
-5.34671731e-02 -9.83091816e-02 -7.52573669e-01 1.77499622e-01
-1.78761268e+00 -5.52916765e-01 -3.74920756e-01 2.19702387e+00
5.54269075e-01 4.44039077e-01 8.35698068e-01 1.29625976e-01
1.21150017e+00 2.13637985e-02 -4.54971105e-01 5.79639733e-01
-1.55115902e-01 -5.14602005e-01 5.16707420e-01 3.08390111e-01
-1.52078176e+00 9.57394183e-01 4.93200731e+00 8.27799499e-01
-7.78994918e-01 1.55360445e-01 5.43818355e-01 3.17214906e-01
4.05459940e-01 3.25553827e-02 -1.50615323e+00 8.28090250e-01
3.89411561e-02 3.53503078e-01 -3.10874552e-01 9.65857983e-01
3.38356346e-01 -1.81690663e-01 -7.94210434e-01 9.38523889e-01
7.59576866e-03 -4.16618407e-01 -1.71859011e-01 9.07804370e-02
3.15171391e-01 -4.62364376e-01 -2.96832800e-01 2.81815827e-01
6.45431653e-02 -5.28658032e-01 1.03343141e+00 2.11833879e-01
3.36344898e-01 -7.26514637e-01 8.60284090e-01 6.00482643e-01
-1.71481323e+00 -6.52784035e-02 -5.08929312e-01 -4.40321825e-02
4.69604850e-01 6.99703872e-01 -4.66268718e-01 3.97772431e-01
6.12790704e-01 6.47694528e-01 -9.49199677e-01 1.49848545e+00
-7.18081176e-01 5.19887567e-01 -4.84894931e-01 2.89177269e-01
3.71810747e-03 -5.47416434e-02 5.54601848e-01 1.32106221e+00
-1.57238901e-01 1.63694322e-01 7.94901013e-01 6.10297561e-01
1.55484125e-01 2.26885125e-01 -2.44424924e-01 8.41797352e-01
3.56430709e-01 1.36377788e+00 -8.11409533e-01 -3.46200645e-01
-4.83068913e-01 9.78553832e-01 1.77703351e-01 2.46969178e-01
-8.83733690e-01 -2.43664697e-01 1.93408072e-01 4.20153558e-01
2.93622643e-01 1.00669689e-01 8.58353898e-02 -1.08327806e+00
4.74112034e-01 -4.45343077e-01 4.64935780e-01 -5.60097337e-01
-1.13971913e+00 4.48021621e-01 -6.75070733e-02 -1.26332402e+00
2.82534331e-01 -3.88583839e-01 -7.30637372e-01 7.31119931e-01
-1.54212511e+00 -1.52173257e+00 -7.43575037e-01 6.02496505e-01
3.64191771e-01 2.84597069e-01 1.02075174e-01 4.86897260e-01
-1.19832790e+00 9.30241108e-01 -6.91010237e-01 5.07159352e-01
7.42514908e-01 -1.00499260e+00 2.63806224e-01 1.11063623e+00
-6.20165169e-01 5.56448817e-01 6.77385807e-01 -7.79161930e-01
-7.54824519e-01 -1.34764850e+00 6.20564163e-01 -4.00219619e-01
2.61173397e-01 -3.52116197e-01 -8.02552700e-01 4.38337445e-01
-2.40332279e-02 3.56700748e-01 2.66161919e-01 -1.47371083e-01
-2.95023710e-01 -2.16526300e-01 -8.69971812e-01 5.46264946e-01
1.12151837e+00 5.19447550e-02 -5.93315721e-01 3.00135583e-01
6.45540476e-01 -3.40199023e-01 -2.91376442e-01 5.50788581e-01
3.33282292e-01 -9.97764647e-01 1.18275595e+00 -1.95690766e-01
9.50438157e-02 -8.05959761e-01 7.79415220e-02 -6.37536824e-01
-2.98975408e-01 -2.86511749e-01 -2.97603577e-01 1.64174855e+00
-6.49203509e-02 -3.91418785e-01 9.19146478e-01 6.12038255e-01
-1.16074644e-01 -5.07699430e-01 -7.40403414e-01 -7.55060494e-01
-3.46535981e-01 -7.59332776e-02 4.25821215e-01 4.87044871e-01
-3.06041509e-01 6.06737435e-01 -5.35425842e-01 4.75556105e-01
1.07720828e+00 -4.85924296e-02 1.12572694e+00 -1.21733224e+00
-1.72659427e-01 -4.30271000e-01 -6.18106544e-01 -1.51688421e+00
8.07846114e-02 -4.02130187e-01 3.39279473e-01 -1.39951646e+00
4.41374063e-01 -2.54764825e-01 -1.17171682e-01 4.28653389e-01
-6.83029652e-01 2.74538934e-01 2.06285417e-01 1.84018850e-01
-9.54617262e-01 7.97132850e-01 1.34188056e+00 -3.04125324e-02
-6.73446292e-03 3.79540443e-01 -4.38969672e-01 1.24233890e+00
4.54141021e-01 -1.76519796e-01 -5.55266626e-02 -1.18737206e-01
-4.54585016e-01 -1.55853825e-02 6.26944542e-01 -1.16561031e+00
4.23586696e-01 1.41828194e-01 6.41504109e-01 -8.40274453e-01
3.92227232e-01 -5.37014127e-01 -4.39488232e-01 4.52804625e-01
1.41244799e-01 -2.20937788e-01 -2.49087773e-02 7.22040594e-01
-1.64029315e-01 -7.09147155e-02 1.14556897e+00 -2.24769339e-01
-7.76161730e-01 3.94528389e-01 1.22815870e-01 1.11632906e-01
1.05149925e+00 -1.98552564e-01 -5.84624037e-02 -6.32797480e-02
-5.52759409e-01 6.68107986e-01 5.41538358e-01 1.27027541e-01
3.94601822e-01 -1.10844386e+00 -6.69717968e-01 1.85532242e-01
3.01430404e-01 3.17543417e-01 2.67870933e-01 1.16185665e+00
-2.18030617e-01 2.79199660e-01 1.01115398e-01 -6.79424882e-01
-1.32327032e+00 7.38795102e-01 4.02638644e-01 7.25451112e-02
-8.88072252e-01 9.32951510e-01 8.08810532e-01 -1.86659008e-01
6.26035333e-01 -1.08870156e-01 -4.94229138e-01 2.23558489e-02
7.45933414e-01 3.21390361e-01 -2.64186025e-01 -1.06018627e+00
-6.65345967e-01 7.17537403e-01 -9.78455245e-02 2.16735810e-01
6.53613210e-01 -4.57366168e-01 -3.55968438e-02 -1.95382059e-01
8.83321106e-01 1.25140771e-01 -1.51642931e+00 -3.93472821e-01
-1.04637809e-01 -4.43700373e-01 -2.39745644e-03 -4.07387704e-01
-1.18695104e+00 1.03614938e+00 6.95213974e-01 -3.92308801e-01
9.01988566e-01 -2.39345785e-02 9.05939817e-01 4.64470275e-02
3.18913013e-01 -9.91376877e-01 1.55158909e-02 2.79780924e-01
5.04707456e-01 -1.47400486e+00 5.66575304e-02 -8.09753776e-01
-6.23024285e-01 7.43345380e-01 1.03998911e+00 1.81762964e-01
4.79206502e-01 -4.73223813e-02 -8.09284300e-02 -3.94795723e-02
7.29808286e-02 -7.50226676e-01 3.49502623e-01 6.41571403e-01
2.76817411e-01 1.97192542e-02 -3.42542320e-01 7.60086775e-01
-5.60700148e-02 -2.43191659e-01 6.64532138e-03 8.53317380e-01
-7.55987167e-01 -8.69620383e-01 -6.76423132e-01 7.10049719e-02
-2.86011666e-01 1.22869983e-01 -3.05234879e-01 6.57913446e-01
3.98196399e-01 1.06164598e+00 -2.64921665e-01 -9.73744094e-02
5.00542343e-01 -1.78086549e-01 1.57607213e-01 -5.50971568e-01
-1.61783606e-01 4.37181473e-01 -6.56120479e-02 -4.53282982e-01
-3.72302353e-01 -7.22741723e-01 -1.26297522e+00 -7.63902292e-02
-6.24282897e-01 5.47395721e-02 2.91907877e-01 1.01557219e+00
-8.00656602e-02 3.88346404e-01 4.17255312e-01 -1.05110109e+00
-4.23197716e-01 -9.77685571e-01 -4.93929505e-01 5.07721722e-01
2.69365907e-01 -1.03807449e+00 -1.96974725e-01 -3.06560080e-02] | [7.973312854766846, -0.5777560472488403] |
e803c8cb-b92b-4885-8f25-20b02d4f6cdd | patch-autoaugment | 2103.11099 | null | https://arxiv.org/abs/2103.11099v2 | https://arxiv.org/pdf/2103.11099v2.pdf | Local Patch AutoAugment with Multi-Agent Collaboration | Data augmentation (DA) plays a critical role in improving the generalization of deep learning models. Recent works on automatically searching for DA policies from data have achieved great success. However, existing automated DA methods generally perform the search at the image level, which limits the exploration of diversity in local regions. In this paper, we propose a more fine-grained automated DA approach, dubbed Patch AutoAugment, to divide an image into a grid of patches and search for the joint optimal augmentation policies for the patches. We formulate it as a multi-agent reinforcement learning (MARL) problem, where each agent learns an augmentation policy for each patch based on its content together with the semantics of the whole image. The agents cooperate with each other to achieve the optimal augmentation effect of the entire image by sharing a team reward. We show the effectiveness of our method on multiple benchmark datasets of image classification and fine-grained image recognition (e.g., CIFAR-10, CIFAR-100, ImageNet, CUB-200-2011, Stanford Cars and FGVC-Aircraft). Extensive experiments demonstrate that our method outperforms the state-of-the-art DA methods while requiring fewer computational resources. | ['Zhibo Chen', 'Xin Jin', 'Xin Li', 'Ruoyu Feng', 'Tao Yu', 'Shiqi Lin'] | 2021-03-20 | local-patch-autoaugment-with-multi-agent | https://openreview.net/forum?id=RuC5ilX2m6O | https://openreview.net/pdf?id=RuC5ilX2m6O | null | ['fine-grained-image-recognition'] | ['computer-vision'] | [-2.23591849e-01 4.76350123e-03 -1.97138026e-01 -1.43794492e-01
-8.20521116e-01 -3.48562270e-01 4.05785263e-01 8.91199335e-02
-5.54557741e-01 8.09425890e-01 3.79831307e-02 5.79267517e-02
9.14349034e-02 -7.94149518e-01 -1.04961205e+00 -1.00010192e+00
2.80689895e-02 7.25553572e-01 7.47996643e-02 -2.52266347e-01
-6.25160988e-03 3.66749018e-01 -1.39733815e+00 2.56213784e-01
9.11697686e-01 1.25072277e+00 3.85657996e-01 2.86270320e-01
3.27099697e-03 1.06822097e+00 -6.35379910e-01 -1.27674237e-01
4.45697486e-01 -3.20501477e-01 -8.14941823e-01 5.46228707e-01
3.72249573e-01 -3.70647997e-01 -2.74066389e-01 1.18464124e+00
2.72479117e-01 2.12553054e-01 4.90614414e-01 -1.62599003e+00
-8.45637918e-01 6.30843818e-01 -9.63903964e-01 2.08254650e-01
-4.51298028e-01 1.84542209e-01 9.72316146e-01 -6.92932606e-01
2.72777677e-01 1.33351386e+00 2.91617036e-01 6.94556594e-01
-1.33397412e+00 -6.61650538e-01 7.61785567e-01 1.11162573e-01
-1.23464906e+00 8.29124600e-02 8.70924830e-01 -4.49828416e-01
7.13680208e-01 -1.57290891e-01 5.44518054e-01 8.83054495e-01
-1.26021817e-01 1.16631579e+00 1.34080529e+00 -4.16935503e-01
4.10781711e-01 4.13198508e-02 -3.53078300e-04 9.22165036e-01
-7.32125193e-02 1.30432814e-01 -3.48250180e-01 -2.92996198e-01
1.08721483e+00 7.41867051e-02 1.14177831e-01 -5.54814994e-01
-1.25289381e+00 1.11959159e+00 9.05043304e-01 1.06641501e-01
-8.62690091e-01 2.43750900e-01 3.11949879e-01 3.24105442e-01
5.77626169e-01 6.21249080e-01 -5.10401964e-01 3.33464265e-01
-2.11924493e-01 5.20823717e-01 2.71214366e-01 7.17369318e-01
1.13301945e+00 2.27616847e-01 -3.45238566e-01 9.96500134e-01
2.28646517e-01 2.71798164e-01 5.95790029e-01 -9.22301233e-01
4.38938409e-01 6.81480706e-01 3.23384553e-01 -8.42696071e-01
-2.35070512e-01 -5.32228589e-01 -1.12036860e+00 4.08034086e-01
2.67916411e-01 -3.98371786e-01 -9.18322921e-01 1.93466747e+00
4.70576346e-01 4.29689080e-01 1.60946503e-01 9.69420433e-01
4.96122271e-01 7.38408983e-01 3.76992315e-01 3.31533365e-02
1.18582523e+00 -1.71073830e+00 -4.63704050e-01 -5.77126026e-01
4.16976154e-01 -2.62207806e-01 1.09799325e+00 3.69203657e-01
-8.54489326e-01 -7.14671731e-01 -9.70553398e-01 5.46509445e-01
-2.30088785e-01 3.63769859e-01 5.83820581e-01 5.42161725e-02
-9.67553139e-01 2.51364917e-01 -7.12570727e-01 1.59900397e-01
8.59564066e-01 2.88539380e-01 -1.30210400e-01 -5.41263781e-02
-9.15248811e-01 5.73709309e-01 3.35961372e-01 -2.39422426e-01
-1.54575324e+00 -7.72796392e-01 -6.35717332e-01 3.19642313e-02
5.61919928e-01 -4.44456816e-01 1.36515176e+00 -1.52689433e+00
-1.25708032e+00 6.55984640e-01 3.10599267e-01 -9.11108494e-01
2.78717428e-01 -2.37445503e-01 -6.72034621e-02 1.36738960e-02
3.39773476e-01 1.30311763e+00 1.13628674e+00 -1.50626397e+00
-9.04995501e-01 -4.10235435e-01 3.66597533e-01 4.50229436e-01
-5.08582115e-01 -2.44101360e-01 -1.73373103e-01 -1.09571517e+00
-4.77725744e-01 -9.81093168e-01 -8.12810421e-01 -2.66200334e-01
-1.16252758e-01 -5.34876347e-01 7.26668179e-01 -3.24205637e-01
9.79502439e-01 -2.20659375e+00 4.31612492e-01 5.35021499e-02
2.76476979e-01 2.32814372e-01 -5.70240915e-01 4.88729924e-02
1.69193223e-01 -9.34348479e-02 -1.12862326e-01 -3.22339326e-01
-1.32446930e-01 4.42520648e-01 -4.78080124e-01 3.14781964e-01
2.95882732e-01 9.75497007e-01 -8.20992827e-01 -3.08011293e-01
1.25645390e-02 1.13532431e-01 -7.11384475e-01 3.67156535e-01
-7.47632682e-01 6.06799960e-01 -7.45298147e-01 5.38420081e-01
4.69853908e-01 -5.48441589e-01 -1.83489323e-01 -3.12629156e-02
4.75094952e-02 -3.67840141e-01 -1.03343141e+00 1.62775362e+00
-3.10591668e-01 3.08044571e-02 1.07469030e-01 -1.22506988e+00
1.06028306e+00 1.15035243e-01 5.40345371e-01 -7.80331254e-01
2.77890712e-02 7.87011255e-03 2.63591334e-02 -2.07899407e-01
2.86282778e-01 1.02961875e-01 -6.35050386e-02 6.62334859e-01
1.33168682e-01 1.67514175e-01 8.62004980e-02 3.66227329e-02
8.28747153e-01 -8.43773186e-02 2.24235594e-01 -3.24046314e-01
4.88665164e-01 3.65927607e-01 5.79374135e-01 1.03488326e+00
-4.17013288e-01 1.82974890e-01 1.46273419e-01 -7.87697315e-01
-1.28418624e+00 -7.16912448e-01 3.80030304e-01 1.45731938e+00
2.12805957e-01 -5.62870204e-02 -1.04985189e+00 -1.09456313e+00
3.31648998e-02 1.95646554e-01 -1.03656995e+00 -5.98355234e-02
-5.66850960e-01 -8.09406698e-01 3.85535657e-01 6.41116440e-01
1.11149383e+00 -1.48336089e+00 -4.64467764e-01 3.57522726e-01
-1.34983107e-01 -1.15900934e+00 -5.33206403e-01 2.65865959e-02
-6.90470397e-01 -8.25155139e-01 -7.81218827e-01 -1.15004623e+00
8.65807831e-01 4.29850161e-01 1.08856177e+00 1.61280587e-01
-1.97212934e-01 4.52056080e-01 -4.70244855e-01 -4.86853510e-01
-3.31647158e-01 1.15306295e-01 -1.46192694e-02 4.38634962e-01
1.03613488e-01 -2.75146991e-01 -6.15244687e-01 6.82058811e-01
-8.92456949e-01 1.03690036e-01 9.00730073e-01 1.25756085e+00
1.08359301e+00 3.28491293e-02 1.10742247e+00 -7.26131618e-01
6.59111440e-01 -5.95804870e-01 -7.71306515e-01 3.03479135e-01
-4.65018272e-01 7.04343170e-02 7.32677579e-01 -7.80853271e-01
-9.49002624e-01 2.23343268e-01 1.60362184e-01 -8.96682322e-01
-3.72681379e-01 4.62561816e-01 -6.99224845e-02 -1.65175021e-01
8.15892875e-01 4.32269067e-01 2.28982940e-01 -4.19561744e-01
3.88899446e-01 3.58093530e-01 5.59646547e-01 -8.65778744e-01
5.47489643e-01 4.27064240e-01 -2.09448889e-01 -4.82771814e-01
-1.20456421e+00 -2.99264729e-01 -3.93199205e-01 -1.30448088e-01
9.04550493e-01 -1.20930517e+00 -5.18611670e-01 6.12097383e-01
-8.14766467e-01 -7.42772996e-01 -4.35775757e-01 2.62179434e-01
-7.55580306e-01 -1.10442080e-01 -3.66478384e-01 -4.95804578e-01
-3.84050459e-01 -1.36647105e+00 9.22229230e-01 4.47438836e-01
4.19933200e-01 -7.87541449e-01 2.22248212e-01 4.33373421e-01
3.36571068e-01 3.39084901e-02 9.47953343e-01 -5.87542474e-01
-6.64753914e-01 1.29831925e-01 -2.01784372e-01 5.65380573e-01
9.34883207e-02 -5.32655478e-01 -7.82555103e-01 -4.48531926e-01
-2.82466680e-01 -9.25816953e-01 8.94191742e-01 5.02179205e-01
1.72868323e+00 -7.18484759e-01 -1.52433231e-01 2.96387196e-01
1.34440994e+00 2.74474889e-01 4.66888100e-01 7.16928124e-01
6.78373575e-01 3.32566410e-01 1.15032530e+00 5.81701577e-01
3.65837634e-01 5.73390365e-01 8.82130504e-01 -1.59986228e-01
-6.32507354e-02 -1.26476198e-01 1.31363586e-01 2.09174454e-01
1.08282916e-01 -1.62298486e-01 -7.25647449e-01 6.65038824e-01
-2.26409340e+00 -7.00216055e-01 6.30810916e-01 1.85276592e+00
7.21060574e-01 -1.47863477e-01 5.35157144e-01 -2.49449179e-01
6.95146918e-01 1.57623544e-01 -8.21986318e-01 -7.87610486e-02
-2.37434000e-01 -9.43333656e-02 3.19293678e-01 2.85188287e-01
-1.53596866e+00 1.27293646e+00 5.51590538e+00 1.04886723e+00
-9.11499143e-01 1.24966338e-01 1.19512212e+00 1.50222301e-01
6.27600327e-02 -4.91974205e-01 -8.54258716e-01 2.18933940e-01
3.72718602e-01 1.44389004e-01 8.98563981e-01 1.16567218e+00
-1.84941575e-01 1.57278284e-01 -8.17698479e-01 8.90167236e-01
-4.96693961e-02 -1.67129910e+00 2.94050992e-01 7.50005990e-02
1.29116988e+00 4.62164581e-01 2.95984864e-01 5.07986248e-01
7.59877026e-01 -1.02352202e+00 5.76364815e-01 2.69482404e-01
3.32078844e-01 -9.62661862e-01 4.96061593e-01 5.54189563e-01
-9.18591261e-01 -4.51151818e-01 -4.70878601e-01 8.81123245e-02
-4.17821229e-01 1.20353408e-01 -7.02984989e-01 1.80564910e-01
9.31444705e-01 6.83999538e-01 -5.59012711e-01 8.91743422e-01
-1.39462009e-01 5.35681903e-01 -5.39659336e-02 -8.82169157e-02
8.66836131e-01 -2.24077478e-01 4.50252205e-01 6.91777468e-01
-2.88712215e-02 1.04608729e-01 8.58079314e-01 7.49432743e-01
-4.74159658e-01 6.09673113e-02 -5.07113039e-01 -7.27899298e-02
4.77793425e-01 1.44090962e+00 -5.01738966e-01 -4.31595236e-01
-4.93456751e-01 7.19552398e-01 8.27686548e-01 3.48318219e-01
-8.08381617e-01 1.01420783e-01 8.22207868e-01 -2.92623430e-01
5.82010031e-01 6.32191892e-04 -2.78497804e-02 -9.27105725e-01
-2.16981247e-01 -1.44355726e+00 6.49132729e-01 -5.28526008e-01
-1.51755488e+00 9.30770993e-01 -2.26026431e-01 -1.20590913e+00
-1.02450483e-01 -4.12665009e-01 -4.78554606e-01 6.07795238e-01
-1.51355708e+00 -1.26839554e+00 -3.76559854e-01 9.21062291e-01
8.56179237e-01 -7.39099324e-01 7.97421575e-01 1.26797363e-01
-5.80931365e-01 5.27761579e-01 5.57340682e-02 3.45132321e-01
4.39848512e-01 -1.32820761e+00 2.19531298e-01 5.05543113e-01
2.95870930e-01 -1.01196036e-01 3.39585841e-01 -5.53817809e-01
-1.06144929e+00 -1.58932912e+00 -3.46760713e-02 4.73497510e-02
7.38395095e-01 -4.50223498e-02 -9.01157618e-01 7.57915258e-01
4.24742669e-01 4.26412523e-01 3.15120757e-01 -1.38396397e-01
-2.99650967e-01 -2.86745489e-01 -1.18248427e+00 6.16963565e-01
7.10291266e-01 -1.03428317e-02 -2.30780914e-01 4.22408730e-01
8.76758754e-01 -2.91966796e-01 -7.68788397e-01 3.93770307e-01
2.19286978e-02 -5.39196491e-01 1.02315176e+00 -9.84926045e-01
4.18497711e-01 -3.11218977e-01 -3.09618413e-01 -1.88661313e+00
-5.41696489e-01 -2.87585646e-01 -1.33909628e-01 1.06759298e+00
2.61695653e-01 -5.98432302e-01 9.25537765e-01 1.35759249e-01
-1.73340216e-01 -9.89613712e-01 -8.14294040e-01 -7.13630021e-01
1.34459794e-01 7.93609843e-02 8.43806028e-01 8.78526449e-01
-5.24790525e-01 3.17073941e-01 -5.12880385e-01 3.91360134e-01
8.33674252e-01 3.15752059e-01 9.37503457e-01 -1.03016973e+00
-6.45741880e-01 -4.99748170e-01 -1.55821308e-01 -1.01766264e+00
3.57812852e-01 -6.59375012e-01 1.33830667e-01 -1.13280487e+00
3.60585958e-01 -9.33284104e-01 -4.62284178e-01 8.82991850e-01
-1.35531768e-01 2.12291956e-01 2.15824455e-01 3.55442971e-01
-9.53035831e-01 8.61655295e-01 1.51378238e+00 -4.48184997e-01
-1.27435505e-01 -2.15009805e-02 -8.33889663e-01 7.71369457e-01
1.06335545e+00 -3.86822194e-01 -4.12353575e-01 -6.33717775e-01
-1.10461168e-01 -8.28450844e-02 5.46923876e-01 -8.04204643e-01
8.08770359e-02 -4.03582424e-01 3.72617900e-01 -3.88589680e-01
2.12377235e-01 -7.17363894e-01 -3.26073706e-01 5.12779415e-01
-5.46127856e-01 2.44162291e-01 3.42483997e-01 7.11413860e-01
-4.38780487e-01 -1.11104481e-01 1.01227832e+00 -3.26476485e-01
-8.86647463e-01 8.14025521e-01 -2.23571345e-01 1.32200614e-01
1.29556286e+00 4.91177917e-01 -6.32568717e-01 -2.47377276e-01
-7.75129020e-01 5.82431555e-01 1.00796610e-01 5.78864396e-01
5.68155289e-01 -1.56657732e+00 -8.93182814e-01 1.65444851e-01
1.70804933e-01 2.68944502e-01 3.69826823e-01 5.98427594e-01
-1.69885248e-01 1.31010622e-01 -6.25482857e-01 -7.04223871e-01
-1.07067168e+00 8.21531475e-01 5.82776964e-01 -7.14648962e-01
-4.68202829e-01 9.04116273e-01 8.13300490e-01 -3.77400279e-01
4.05417711e-01 5.02627203e-03 -3.67107451e-01 -2.33831167e-01
6.38846576e-01 1.88777223e-01 -9.95594263e-02 -3.91565531e-01
-4.55494113e-02 2.43968889e-01 -5.22106349e-01 3.32837040e-03
1.50455153e+00 -3.39580737e-02 3.05609871e-02 1.47966202e-02
8.30411553e-01 -4.96347070e-01 -1.67729151e+00 -6.98811829e-01
-3.43360931e-01 -3.81853104e-01 1.82129472e-01 -9.72670496e-01
-1.47137976e+00 5.98319590e-01 5.73754013e-01 2.17776358e-01
1.24193645e+00 2.96148837e-01 4.39953148e-01 4.92738128e-01
4.62679625e-01 -9.68607724e-01 8.22598040e-01 3.43843818e-01
1.22752655e+00 -1.48756218e+00 -3.69231492e-01 4.04220670e-02
-1.12179780e+00 6.66949987e-01 1.31874239e+00 -5.29372990e-01
5.45404434e-01 2.98041720e-02 -1.68250792e-03 -8.08167309e-02
-8.19529712e-01 -1.82572007e-01 1.39962479e-01 6.47625983e-01
-2.64100611e-01 1.67237327e-01 2.57679701e-01 6.46791399e-01
2.08471790e-01 -3.01463991e-01 1.67504743e-01 8.23047400e-01
-6.19234145e-01 -1.07443643e+00 -4.04682994e-01 5.21683395e-01
-3.94017369e-01 1.58705875e-01 -2.99798727e-01 7.21286356e-01
5.74308336e-02 7.34377682e-01 2.33971208e-01 -2.66361773e-01
4.60086428e-02 -3.34502846e-01 5.09059131e-01 -4.63309884e-01
-6.01496875e-01 2.62774497e-01 -2.52195746e-01 -4.45672750e-01
-4.19659913e-01 -6.49274170e-01 -1.23899364e+00 4.42253314e-02
-4.19623367e-02 2.99910277e-01 4.24394011e-01 9.45461214e-01
5.78856707e-01 6.92894995e-01 9.56576705e-01 -8.23212445e-01
-8.29091489e-01 -1.00623643e+00 -4.60260004e-01 2.57118374e-01
4.50942039e-01 -9.17864203e-01 1.32725850e-01 -1.41817346e-01] | [9.778618812561035, 2.3983757495880127] |
0e7b8e5d-f087-47ee-bbac-c6af4908b071 | revisiting-stereo-depth-estimation-from-a | 2011.02910 | null | https://arxiv.org/abs/2011.02910v4 | https://arxiv.org/pdf/2011.02910v4.pdf | Revisiting Stereo Depth Estimation From a Sequence-to-Sequence Perspective with Transformers | Stereo depth estimation relies on optimal correspondence matching between pixels on epipolar lines in the left and right images to infer depth. In this work, we revisit the problem from a sequence-to-sequence correspondence perspective to replace cost volume construction with dense pixel matching using position information and attention. This approach, named STereo TRansformer (STTR), has several advantages: It 1) relaxes the limitation of a fixed disparity range, 2) identifies occluded regions and provides confidence estimates, and 3) imposes uniqueness constraints during the matching process. We report promising results on both synthetic and real-world datasets and demonstrate that STTR generalizes across different domains, even without fine-tuning. | ['Francis X. Creighton', 'Andy Ding', 'Nathan Drenkow', 'Mathias Unberath', 'Russell H. Taylor', 'Xingtong Liu', 'Zhaoshuo Li'] | 2020-11-05 | null | http://openaccess.thecvf.com//content/ICCV2021/html/Li_Revisiting_Stereo_Depth_Estimation_From_a_Sequence-to-Sequence_Perspective_With_Transformers_ICCV_2021_paper.html | http://openaccess.thecvf.com//content/ICCV2021/papers/Li_Revisiting_Stereo_Depth_Estimation_From_a_Sequence-to-Sequence_Perspective_With_Transformers_ICCV_2021_paper.pdf | iccv-2021-1 | ['stereo-depth-estimation'] | ['computer-vision'] | [ 4.47921544e-01 -1.38241768e-01 -3.21774296e-02 -3.38897079e-01
-6.30811930e-01 -5.11537373e-01 4.79338199e-01 -1.82467267e-01
-4.02949601e-01 9.21067834e-01 2.09803239e-01 -1.79558173e-02
2.51885623e-01 -7.47938275e-01 -6.32411003e-01 -3.61547291e-01
3.98560971e-01 2.95665354e-01 5.88344157e-01 6.27056509e-03
7.84094393e-01 5.04786015e-01 -1.56831706e+00 1.43262982e-01
1.04542351e+00 9.81559455e-01 3.54172260e-01 3.67958337e-01
1.66265294e-02 6.47027850e-01 -1.59430116e-01 -4.40002024e-01
6.70680821e-01 -3.72580737e-01 -6.33736610e-01 4.22141403e-01
9.24750149e-01 -5.62082648e-01 -5.10365725e-01 1.03214800e+00
3.36588830e-01 6.33989200e-02 3.34057540e-01 -9.55253422e-01
-9.48971510e-02 -2.10512340e-01 -1.03948927e+00 2.07574084e-01
7.90764391e-01 2.29293779e-01 9.89225745e-01 -9.59457755e-01
9.33507502e-01 1.12676060e+00 8.48212898e-01 3.43928456e-01
-1.34153545e+00 -4.08338457e-01 4.73533943e-02 1.20245971e-01
-1.54556584e+00 -5.26750922e-01 8.16152692e-01 -7.22479045e-01
1.00557375e+00 2.86448095e-03 8.70216310e-01 7.71232665e-01
2.72825472e-02 7.81795979e-01 1.19387889e+00 -4.09760684e-01
2.07014292e-01 -1.44591093e-01 -1.34948716e-01 5.15828729e-01
1.53121203e-01 5.17808259e-01 -6.77815914e-01 -2.71774009e-02
1.28929722e+00 -2.75372177e-01 -7.99863696e-01 -7.89753377e-01
-1.29083765e+00 8.41606975e-01 1.78737581e-01 6.89912513e-02
-1.49472117e-01 8.23592842e-02 2.99695551e-01 1.18815288e-01
5.89615524e-01 3.17165077e-01 -2.31385335e-01 -1.53351411e-01
-8.51543486e-01 1.12451404e-01 4.79420841e-01 1.15013647e+00
1.17590833e+00 -1.17235005e-01 2.26707280e-01 7.14599609e-01
1.87261123e-02 2.61873186e-01 2.48672768e-01 -1.53021908e+00
8.30643952e-01 3.04219991e-01 2.21545085e-01 -1.06348753e+00
-2.67233312e-01 -1.43082827e-01 -5.28690338e-01 4.49775040e-01
4.89553511e-01 3.36654969e-02 -6.70868158e-01 1.67583334e+00
2.69514024e-01 4.00891930e-01 -2.16059864e-01 1.02672100e+00
5.20125031e-01 3.31734180e-01 -4.92376953e-01 -2.01830089e-01
7.33233809e-01 -7.83042490e-01 -5.66499770e-01 -6.12578511e-01
4.21051562e-01 -6.71348095e-01 9.43496585e-01 3.45125914e-01
-1.35371649e+00 -5.12583196e-01 -1.09833419e+00 -4.53455955e-01
1.09815672e-01 -2.72282273e-01 4.29437488e-01 3.84454668e-01
-1.29405642e+00 5.69673419e-01 -6.10179126e-01 -3.12662005e-01
1.80878147e-01 2.47681886e-01 -4.04317945e-01 -1.75932333e-01
-9.63445961e-01 8.21895599e-01 2.22413182e-01 -7.42803365e-02
-3.81995440e-01 -7.41040289e-01 -1.29534459e+00 -1.69437036e-01
2.47771442e-01 -1.12545490e+00 9.95039344e-01 -8.26998174e-01
-1.54093564e+00 1.13830888e+00 -6.23992085e-01 -1.80298910e-01
1.00141156e+00 -2.65884668e-01 2.80542374e-01 3.08299214e-01
2.78647751e-01 9.97570157e-01 5.37507355e-01 -1.16560543e+00
-9.32362258e-01 -4.37019557e-01 1.41197592e-01 3.86296451e-01
1.75895661e-01 -4.17865485e-01 -7.88959384e-01 -6.35263801e-01
6.46591723e-01 -7.68806517e-01 -2.10750267e-01 4.29124594e-01
-1.92453742e-01 2.93276101e-01 5.49775302e-01 -4.44195330e-01
9.05608714e-01 -1.99060595e+00 2.91703492e-01 2.19396606e-01
2.90590733e-01 -1.77739277e-01 6.08779602e-02 3.43424022e-01
2.05466896e-02 -4.73620683e-01 -4.66818810e-01 -4.94711339e-01
-4.23912257e-01 1.52415693e-01 -2.29775801e-01 6.95780575e-01
-1.12366162e-01 8.12864900e-01 -8.98757815e-01 -6.35761917e-01
7.58498728e-01 4.01555985e-01 -7.84026682e-01 7.72589445e-02
-9.54812169e-02 8.20774734e-01 -1.25645414e-01 4.55515176e-01
9.56595957e-01 -1.69429317e-01 -3.66163161e-03 -1.91581085e-01
-4.12134796e-01 2.36768603e-01 -1.26898193e+00 2.03714156e+00
-4.88487303e-01 1.07169497e+00 -6.40016794e-02 -6.36587501e-01
9.11065996e-01 -1.92481130e-02 5.96760571e-01 -8.09749842e-01
-2.58996487e-02 3.05824071e-01 -5.07416785e-01 -4.71377790e-01
5.75229168e-01 -1.78391725e-01 2.95060366e-01 9.37970132e-02
-3.61782968e-01 -4.62653577e-01 1.80638507e-01 -7.76275545e-02
8.13467085e-01 3.82853270e-01 4.98794734e-01 -3.18004906e-01
7.52174079e-01 -8.84780362e-02 8.50854397e-01 3.24622542e-01
-1.53792560e-01 1.17832518e+00 3.64966631e-01 -4.92833048e-01
-1.29987752e+00 -1.16451716e+00 -3.73636067e-01 1.97756320e-01
8.37745130e-01 -9.91240144e-02 -4.70958799e-01 -2.47997507e-01
2.15460166e-01 4.74949747e-01 -5.79632998e-01 3.43227953e-01
-7.14548707e-01 3.45638134e-02 6.98552504e-02 8.59944642e-01
6.31971478e-01 -5.21018565e-01 -8.36630464e-01 6.40538484e-02
-6.11419618e-01 -1.47804403e+00 -8.90338063e-01 -2.12190583e-01
-1.11694801e+00 -1.13871634e+00 -7.89770663e-01 -6.48444653e-01
7.23023832e-01 6.30563915e-01 1.14048731e+00 -8.43050852e-02
-2.54192561e-01 3.61140072e-01 1.35531602e-02 1.72303140e-01
4.44532372e-02 -2.72514731e-01 -2.50955582e-01 -1.71477735e-01
2.13715881e-01 -7.81872272e-01 -7.65234411e-01 7.01605260e-01
-4.26861554e-01 3.42512220e-01 1.98497415e-01 8.72226417e-01
7.61334479e-01 -2.94024169e-01 3.45603526e-02 -5.87242186e-01
-2.02072840e-02 4.88686748e-02 -1.19538999e+00 -5.04486822e-03
-4.87767905e-01 3.42066810e-02 1.76716521e-01 -3.27243865e-03
-1.03361571e+00 2.22655877e-01 -1.05764069e-01 -7.06041515e-01
1.47437692e-01 -5.94593808e-02 -1.32854939e-01 -2.76126742e-01
4.80737060e-01 2.68413335e-01 -1.17135327e-02 -2.71343797e-01
1.09157726e-01 2.72367835e-01 7.48476565e-01 -3.73129249e-01
7.04478264e-01 1.12788963e+00 2.17161581e-01 -7.96281815e-01
-5.27081132e-01 -7.62892783e-01 -1.06177580e+00 -2.41157815e-01
8.31167161e-01 -1.02160573e+00 -8.18224132e-01 3.65353137e-01
-1.37197244e+00 -3.11541706e-01 -2.42442504e-01 7.44132698e-01
-9.95218873e-01 8.80570590e-01 -5.66498578e-01 -6.54901981e-01
5.01924418e-02 -1.17730367e+00 1.26010215e+00 -1.59269068e-02
-2.76716858e-01 -1.22828305e+00 9.29081142e-02 3.70568156e-01
6.67169690e-03 2.84264892e-01 4.49250728e-01 4.32055801e-01
-1.16534162e+00 -2.67967414e-02 -4.28341448e-01 2.49945596e-01
5.00238240e-02 -1.04169384e-01 -1.06841505e+00 -3.23724359e-01
1.27640098e-01 -8.11739266e-02 7.99261212e-01 7.40674138e-01
9.64030564e-01 1.72296211e-01 -3.36610377e-01 1.12688148e+00
1.64607668e+00 5.17296493e-02 1.02371287e+00 5.06322145e-01
7.20662475e-01 9.53225434e-01 8.31817985e-01 4.54223931e-01
5.15373707e-01 9.74087000e-01 3.56860012e-01 -1.26381382e-01
-2.27072105e-01 -4.88073081e-01 -1.83357298e-02 4.48811024e-01
3.44625264e-02 -1.21470653e-01 -6.99337125e-01 6.00336552e-01
-1.84305406e+00 -9.60655034e-01 -1.49003103e-01 2.37777805e+00
6.16761804e-01 3.03620566e-02 -4.60051224e-02 8.43788311e-02
9.00665164e-01 1.85831368e-01 -5.71641266e-01 -9.82244760e-02
-3.15709651e-01 -4.76148650e-02 5.61266243e-01 1.14285195e+00
-7.54813790e-01 8.05268347e-01 6.69572926e+00 5.99475443e-01
-1.02594912e+00 -1.80714317e-02 5.33958077e-01 5.68977743e-02
-5.42469680e-01 9.98713300e-02 -7.33754039e-01 4.09205198e-01
-3.98608595e-02 -1.98784471e-01 2.90645629e-01 5.85165203e-01
2.86101371e-01 -4.59584773e-01 -1.51391065e+00 1.47486043e+00
2.05761075e-01 -1.46095932e+00 -1.85120091e-01 2.18047097e-01
9.34738696e-01 -3.52561362e-02 -5.01232184e-02 -4.62164879e-01
-1.71998818e-03 -7.22274363e-01 8.06173921e-01 2.82856941e-01
8.68002594e-01 -5.90720892e-01 5.27121305e-01 2.44083822e-01
-1.29804027e+00 1.68313876e-01 -3.66076052e-01 -2.00334683e-01
5.14935791e-01 5.97922206e-01 -5.86399138e-01 4.59193438e-01
5.26459217e-01 1.09815526e+00 -9.16926488e-02 1.23809969e+00
-3.49092931e-01 -4.00095016e-01 -2.11737305e-01 5.94726205e-01
6.42540082e-02 -6.23037100e-01 5.24434805e-01 8.89475048e-01
3.43780398e-01 5.69828302e-02 -9.54667702e-02 9.04380918e-01
2.94001043e-01 -1.77241504e-01 -7.04521537e-01 8.22827637e-01
5.08806765e-01 6.23650730e-01 -5.38627148e-01 -2.49495625e-01
-6.05906606e-01 1.15306926e+00 2.49538392e-01 3.61329854e-01
-6.44936800e-01 -9.69916284e-02 8.74137282e-01 3.69797379e-01
3.87317270e-01 -3.49329948e-01 -7.08034933e-01 -1.38449216e+00
3.26528430e-01 -4.19156045e-01 2.92044938e-01 -1.04686570e+00
-1.03437579e+00 4.68799680e-01 -9.15208608e-02 -1.72037983e+00
-4.55604851e-01 -4.05764639e-01 -3.17249984e-01 1.02947533e+00
-1.96025693e+00 -6.54275060e-01 -7.91321218e-01 6.59589231e-01
7.80844092e-01 4.25751537e-01 3.80532086e-01 4.01506424e-01
-1.86263874e-01 4.07193094e-01 5.38802445e-02 -1.22348048e-01
5.31045854e-01 -1.00871480e+00 6.36812627e-01 9.01500642e-01
-2.67658174e-01 2.62296051e-01 6.62487745e-01 -5.26139975e-01
-1.02128172e+00 -6.83673322e-01 9.01228607e-01 -4.99418557e-01
3.52115482e-01 -1.84685230e-01 -7.98928261e-01 6.34766400e-01
3.53965499e-02 4.29329500e-02 2.08074600e-01 -2.40715444e-01
-3.33943039e-01 3.56481224e-02 -1.24071193e+00 5.45800686e-01
1.44998932e+00 -6.37190402e-01 -4.13932830e-01 9.85376388e-02
4.49935287e-01 -8.84231865e-01 -6.81446731e-01 4.98700619e-01
6.65259242e-01 -1.60624397e+00 1.10674739e+00 1.35524139e-01
6.21679962e-01 -2.95169324e-01 -3.32166106e-01 -9.76368487e-01
-1.32748187e-01 -6.47947729e-01 1.57484770e-01 7.43366063e-01
2.64893055e-01 -7.07805574e-01 1.04059041e+00 8.26665640e-01
-2.35487863e-01 -7.69059062e-01 -1.05459154e+00 -8.46999109e-01
-1.81039602e-01 -3.93003374e-01 3.78550231e-01 9.03976619e-01
7.39441216e-02 1.21616751e-01 -3.32047939e-01 1.29525825e-01
8.41377676e-01 3.45096707e-01 8.62215221e-01 -1.12467206e+00
-4.67730790e-01 -4.54130113e-01 -7.69476593e-01 -1.81548965e+00
1.04750566e-01 -3.22180837e-01 6.43341197e-03 -1.40827000e+00
-1.03745842e-03 -4.99173462e-01 4.10658300e-01 -9.86150056e-02
-3.93831842e-02 3.17375571e-01 -3.36975381e-02 3.77672970e-01
-2.03819469e-01 6.12176597e-01 1.35043228e+00 1.68233335e-01
-1.15255997e-01 -3.45302634e-02 -2.75782466e-01 8.46160471e-01
4.00507838e-01 -4.78472970e-02 -6.18023694e-01 -7.09035158e-01
1.14720114e-01 3.76242906e-01 2.52601445e-01 -8.96784782e-01
7.65841529e-02 -2.66286433e-01 2.52552837e-01 -8.40409815e-01
7.48749435e-01 -7.96377063e-01 1.09187044e-01 4.32652742e-01
-1.10911198e-01 1.34745821e-01 9.64793190e-02 4.68308896e-01
-3.24883580e-01 -2.49918580e-01 1.13777351e+00 -1.46841526e-01
-1.10199082e+00 2.91980594e-01 2.10429251e-01 2.70789385e-01
1.10050654e+00 -1.09106255e+00 -1.31878376e-01 -4.91627365e-01
-4.27340418e-01 3.35036963e-01 1.25191963e+00 2.04642296e-01
8.98932993e-01 -1.18941653e+00 -5.81607223e-01 3.84948224e-01
2.84324974e-01 1.62962973e-01 2.50228047e-01 8.32086623e-01
-9.69828129e-01 5.95841885e-01 -1.31570920e-01 -1.15176928e+00
-1.25720835e+00 5.22464395e-01 5.01871943e-01 1.06159076e-01
-1.13164389e+00 8.72087896e-01 7.50398338e-01 -4.64486778e-02
2.72825658e-01 -1.53987780e-01 7.37516508e-02 -2.57716179e-01
4.30608809e-01 4.18802202e-01 -6.66176318e-04 -7.38944769e-01
-2.31799066e-01 1.33669865e+00 2.40024880e-01 -4.54669595e-01
1.04791236e+00 -6.83875859e-01 8.03932399e-02 3.31732810e-01
1.36338973e+00 1.61520094e-02 -1.68682575e+00 -3.12899023e-01
-2.05680937e-01 -1.48918819e+00 9.85116437e-02 -1.86055571e-01
-1.18711650e+00 9.82801139e-01 2.96381086e-01 -3.02184969e-01
9.68341708e-01 -1.14776514e-01 9.09903228e-01 -2.42582578e-02
5.63434303e-01 -9.73358393e-01 1.60171106e-01 4.31296110e-01
8.34047198e-01 -1.33349800e+00 4.62784283e-02 -8.15614104e-01
-5.94969451e-01 1.09233022e+00 8.03596973e-01 -3.24109420e-02
4.19251800e-01 2.07085818e-01 -1.68787166e-02 -7.14414418e-02
-5.20054817e-01 -2.25114644e-01 1.69825964e-02 7.99365461e-01
4.38759565e-01 -4.81338382e-01 -2.21748799e-01 -4.82015908e-01
-8.31010044e-02 1.29894599e-01 5.22767782e-01 8.32341015e-01
-3.97347778e-01 -1.03774834e+00 -3.80733788e-01 -1.05449790e-02
1.54964358e-01 -1.56484887e-01 -7.83945769e-02 8.00208628e-01
6.68747872e-02 7.69005954e-01 4.28654522e-01 -2.42364094e-01
3.77492428e-01 -5.04903615e-01 8.38573396e-01 -5.04092336e-01
6.40894547e-02 1.67993531e-01 1.13400564e-01 -6.76939011e-01
-5.57305634e-01 -8.25460613e-01 -1.12647235e+00 -4.75684226e-01
-3.74351561e-01 -1.32027760e-01 4.68296677e-01 7.06796587e-01
3.92908394e-01 5.83821386e-02 8.57754350e-01 -1.08762646e+00
-1.89860627e-01 -6.09470308e-01 -5.29232979e-01 4.24494267e-01
4.69178706e-01 -8.06532502e-01 -5.29759347e-01 -2.82108840e-02] | [8.896645545959473, -2.466895341873169] |
4ee99238-d19d-4c81-8203-8b04f4ef6436 | keypose-multi-view-3d-labeling-and-keypoint | 1912.02805 | null | https://arxiv.org/abs/1912.02805v2 | https://arxiv.org/pdf/1912.02805v2.pdf | KeyPose: Multi-View 3D Labeling and Keypoint Estimation for Transparent Objects | Estimating the 3D pose of desktop objects is crucial for applications such as robotic manipulation. Many existing approaches to this problem require a depth map of the object for both training and prediction, which restricts them to opaque, lambertian objects that produce good returns in an RGBD sensor. In this paper we forgo using a depth sensor in favor of raw stereo input. We address two problems: first, we establish an easy method for capturing and labeling 3D keypoints on desktop objects with an RGB camera; and second, we develop a deep neural network, called $KeyPose$, that learns to accurately predict object poses using 3D keypoints, from stereo input, and works even for transparent objects. To evaluate the performance of our method, we create a dataset of 15 clear objects in five classes, with 48K 3D-keypoint labeled images. We train both instance and category models, and show generalization to new textures, poses, and objects. KeyPose surpasses state-of-the-art performance in 3D pose estimation on this dataset by factors of 1.5 to 3.5, even in cases where the competing method is provided with ground-truth depth. Stereo input is essential for this performance as it improves results compared to using monocular input by a factor of 2. We will release a public version of the data capture and labeling pipeline, the transparent object database, and the KeyPose models and evaluation code. Project website: https://sites.google.com/corp/view/keypose. | ['Anelia Angelova', 'Xingyu Liu', 'Rico Jonschkowski', 'Kurt Konolige'] | 2019-12-05 | keypose-multi-view-3d-labeling-and-keypoint-1 | http://openaccess.thecvf.com/content_CVPR_2020/html/Liu_KeyPose_Multi-View_3D_Labeling_and_Keypoint_Estimation_for_Transparent_Objects_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Liu_KeyPose_Multi-View_3D_Labeling_and_Keypoint_Estimation_for_Transparent_Objects_CVPR_2020_paper.pdf | cvpr-2020-6 | ['transparent-objects'] | ['computer-vision'] | [ 7.93351978e-02 -5.88257983e-02 9.94145796e-02 -4.98684317e-01
-7.18272746e-01 -9.12550509e-01 4.66759205e-01 -1.79626063e-01
-2.13889003e-01 2.08473995e-01 -7.78666735e-02 -2.20427006e-01
8.67483690e-02 -6.73009634e-01 -1.25579870e+00 -3.32093358e-01
5.23495413e-02 8.58321369e-01 5.80336928e-01 1.38931572e-01
3.03884596e-01 9.93454576e-01 -1.74675500e+00 4.00042444e-01
2.22287342e-01 1.38783932e+00 4.07220930e-01 1.00910485e+00
-8.30531642e-02 4.81700271e-01 -3.05895329e-01 -2.47005433e-01
7.60133088e-01 2.08275869e-01 -8.93108487e-01 1.23092914e-02
9.10369456e-01 -9.68703747e-01 -2.65871435e-01 6.98874116e-01
5.24290383e-01 -1.87029615e-01 4.22422260e-01 -1.31798804e+00
-3.73052537e-01 7.32044503e-02 -4.08952564e-01 -3.33058387e-01
6.25717044e-01 1.85452774e-01 6.93853855e-01 -1.00065613e+00
7.36385643e-01 1.18477392e+00 7.16612995e-01 5.40063441e-01
-1.09360671e+00 -5.52639604e-01 7.46737868e-02 -1.10315911e-01
-1.32586336e+00 -4.35311496e-01 5.14955163e-01 -6.21785164e-01
1.13311684e+00 2.59562373e-01 6.52564049e-01 1.15960908e+00
1.36963706e-02 7.15408981e-01 1.19914973e+00 -3.50307077e-01
1.26483083e-01 1.01266809e-01 -8.26695263e-02 6.09156966e-01
-1.74012743e-02 1.01378687e-01 -5.36722422e-01 -1.03810713e-01
1.25626731e+00 1.95712626e-01 -1.77005053e-01 -1.06384277e+00
-1.40890718e+00 9.96844918e-02 5.17526805e-01 -2.44255796e-01
-2.24475279e-01 4.93117005e-01 -2.40511931e-02 2.38201678e-01
3.38176399e-01 4.72961903e-01 -8.24956656e-01 -4.73353744e-01
-4.02957976e-01 3.64428580e-01 8.34244967e-01 1.50101960e+00
8.25139582e-01 -5.55395603e-01 3.15374196e-01 4.45264757e-01
2.23488003e-01 6.73128247e-01 7.35864555e-03 -1.53641474e+00
3.59240651e-01 5.62150240e-01 4.87797946e-01 -6.40316665e-01
-4.24715281e-01 -4.77321744e-02 -1.78811178e-01 6.66881382e-01
6.66319013e-01 1.82146654e-01 -1.05333161e+00 1.24602628e+00
5.92342556e-01 -1.75714239e-01 -2.93830097e-01 9.50846851e-01
8.13400865e-01 3.56623828e-01 -6.57528102e-01 5.05029023e-01
1.09768164e+00 -7.48065472e-01 -1.66357115e-01 -2.77067333e-01
4.30980980e-01 -9.07279789e-01 1.42816591e+00 7.41091430e-01
-1.00021410e+00 -5.02041638e-01 -8.67922485e-01 -4.30552810e-01
-4.58345175e-01 1.74976196e-02 8.36491287e-01 4.89685953e-01
-1.14748025e+00 7.43722975e-01 -1.12746334e+00 -4.88434047e-01
3.74946207e-01 6.57709599e-01 -4.77874160e-01 -3.27399582e-01
-4.86079037e-01 9.35856164e-01 1.84623614e-01 1.32634975e-02
-9.82402146e-01 -6.96459591e-01 -6.04827642e-01 -3.27960759e-01
4.68007416e-01 -5.43412149e-01 1.52505207e+00 -5.53942621e-01
-1.53572297e+00 1.20686102e+00 8.41663452e-04 6.76209778e-02
7.36765862e-01 -7.97984600e-01 2.89251566e-01 2.00771451e-01
-3.13399099e-02 9.18618441e-01 7.83235312e-01 -1.59546614e+00
-4.81584102e-01 -6.77172005e-01 5.93250811e-01 2.45832905e-01
5.74361952e-03 -2.18462184e-01 -8.98733020e-01 -7.33126402e-02
4.53277797e-01 -1.10191858e+00 5.18795587e-02 6.17020905e-01
-5.03201187e-01 7.14039877e-02 7.53848493e-01 -5.27370453e-01
8.54311213e-02 -2.03226972e+00 3.21910456e-02 2.34065741e-01
1.31241649e-01 -1.70686498e-01 3.84722352e-02 1.75176308e-01
1.45591885e-01 -1.84581071e-01 2.23374337e-01 -5.85765719e-01
1.31237283e-01 1.79128110e-01 -3.33676994e-01 5.48294783e-01
-5.91720939e-02 6.49746358e-01 -6.74398482e-01 -1.49031147e-01
4.37771231e-01 6.27973676e-01 -6.90555930e-01 2.76730418e-01
-4.36464310e-01 3.22915286e-01 -3.21608782e-01 1.02132440e+00
7.77516067e-01 -3.83197576e-01 -1.31856114e-01 -3.35013419e-01
-1.78963840e-01 4.54773635e-01 -1.38968790e+00 2.15527892e+00
-4.41409975e-01 4.95768726e-01 1.47232682e-01 -3.31890345e-01
7.77960241e-01 6.84117973e-02 5.92488825e-01 -3.67820412e-01
1.52163310e-02 3.93644303e-01 -4.26138520e-01 -3.09045583e-01
5.46760738e-01 2.84252405e-01 9.99513790e-02 4.86001432e-01
-5.46412207e-02 -7.63114512e-01 -3.58006299e-01 7.65523687e-02
1.11456418e+00 8.84990573e-01 -2.35160142e-01 -3.26161943e-02
-1.82874337e-01 1.70851171e-01 6.99118376e-02 7.46804953e-01
1.83661878e-01 1.03653014e+00 4.08773929e-01 -5.91296494e-01
-1.24043787e+00 -1.09790170e+00 -2.16324970e-01 8.88533950e-01
4.40139830e-01 -2.88397193e-01 -6.23324692e-01 -4.21639174e-01
3.24023724e-01 3.06950480e-01 -4.26772863e-01 1.84412345e-01
-5.32160461e-01 -1.81859061e-01 1.99911818e-01 7.92850912e-01
4.05879140e-01 -7.39643574e-01 -1.00716925e+00 -1.70302138e-01
7.73450583e-02 -1.29033911e+00 -1.10625684e-01 4.67933863e-01
-8.25501263e-01 -1.21664441e+00 -6.04381561e-01 -3.88702095e-01
6.39617503e-01 4.93197709e-01 1.17007113e+00 -2.39655562e-02
-2.42231175e-01 9.42846835e-01 -4.29521948e-01 -5.21256983e-01
-5.14337905e-02 1.81770250e-01 2.23062664e-01 -6.15795553e-01
3.24863493e-01 -5.15697658e-01 -7.41053939e-01 4.27016139e-01
-6.58863366e-01 1.81155413e-01 4.14946437e-01 3.00337374e-01
7.34299600e-01 -2.99328238e-01 -4.07690108e-01 -4.82111216e-01
-1.16324008e-01 -1.40408734e-02 -9.44679081e-01 4.31447802e-03
-2.60054380e-01 -4.04355377e-02 1.47272334e-01 -2.32212141e-01
-7.31376469e-01 5.02010524e-01 -8.31675008e-02 -7.56718159e-01
-3.65938395e-01 -9.22435597e-02 -2.05053106e-01 -3.99693549e-01
7.17741847e-01 -1.81037992e-01 -4.18016501e-02 -7.32908309e-01
3.78234178e-01 7.33641624e-01 4.93878573e-01 -8.98061037e-01
7.58996010e-01 8.40396464e-01 1.67589299e-02 -5.18836677e-01
-7.48081565e-01 -4.46770489e-01 -1.01065862e+00 -1.76674500e-01
6.78785384e-01 -9.68449593e-01 -1.09968686e+00 6.56572938e-01
-1.32946789e+00 -7.58965015e-01 -2.67470121e-01 4.92224485e-01
-9.12842453e-01 -3.06818914e-02 -5.87398708e-01 -6.25654161e-01
9.62326676e-02 -1.31796598e+00 1.61876130e+00 -7.71131292e-02
-8.22513029e-02 -4.35694635e-01 -2.93911994e-01 2.70290434e-01
1.16439156e-01 3.35304350e-01 5.52553594e-01 -2.77944088e-01
-1.20600140e+00 -2.51843333e-01 -2.73710251e-01 2.72507727e-01
1.75179735e-01 -3.31132337e-02 -1.33325231e+00 -1.98470190e-01
-5.37999831e-02 -6.37347579e-01 4.47999209e-01 2.35505491e-01
1.49753690e+00 7.69313425e-02 -4.29748476e-01 8.48157048e-01
1.40485930e+00 1.42811999e-01 4.51079935e-01 5.31678379e-01
1.04940903e+00 6.37532473e-01 6.91717505e-01 3.55933636e-01
4.96008247e-01 8.49146307e-01 9.65712845e-01 2.49867979e-02
6.35696352e-02 -1.59608155e-01 1.59715548e-01 4.14748937e-01
-3.35659385e-01 -2.81062797e-02 -1.22655714e+00 1.61075905e-01
-1.48198664e+00 -4.58393186e-01 -7.99736455e-02 2.32885575e+00
8.66505921e-01 3.27105910e-01 6.30983189e-02 5.12544326e-02
2.95171291e-01 -2.57190049e-01 -8.52828562e-01 -1.65349036e-01
2.42066354e-01 2.51496941e-01 7.96597123e-01 5.27861893e-01
-9.61209595e-01 9.02732670e-01 5.95644236e+00 1.80884898e-01
-1.19685173e+00 -8.86552483e-02 4.10010695e-01 -3.22182775e-01
-2.47330830e-01 -1.33224592e-01 -9.12182748e-01 9.99696925e-02
5.82178652e-01 4.52124566e-01 6.65616155e-01 1.22929931e+00
-1.05364680e-01 -2.42820740e-01 -1.55653226e+00 1.20690048e+00
-1.71429161e-02 -1.05419874e+00 -2.99892187e-01 2.83237636e-01
5.02617657e-01 3.99484992e-01 -5.20226397e-02 -1.40602281e-02
4.40847337e-01 -8.94137084e-01 1.07115328e+00 5.00838935e-01
9.30933058e-01 -4.27573472e-01 4.69316542e-01 4.32572871e-01
-7.85363913e-01 2.05194399e-01 -3.54196876e-01 -1.64161369e-01
-1.28932640e-01 4.33127761e-01 -9.62171137e-01 1.83900803e-01
1.22930956e+00 5.81016660e-01 -5.21176100e-01 9.31404889e-01
-1.47860572e-01 2.87417788e-02 -8.17221642e-01 1.54300243e-01
-7.75174350e-02 1.73037782e-01 2.26838455e-01 7.34931588e-01
4.14629817e-01 1.26458228e-01 1.20356508e-01 8.27972531e-01
-9.93514806e-02 -4.59531039e-01 -6.23591959e-01 1.94241673e-01
5.99195600e-01 9.55209792e-01 -8.87852013e-01 -1.37846157e-01
-3.11814904e-01 1.18393791e+00 2.41477922e-01 2.30317414e-01
-5.74233174e-01 -3.31544787e-01 6.24063253e-01 3.53721470e-01
4.44439143e-01 -6.81154490e-01 -3.39607626e-01 -1.09414268e+00
4.73133028e-01 -7.58561194e-01 -1.96942165e-01 -1.26525962e+00
-9.53495979e-01 3.79171491e-01 3.02716970e-01 -1.27644169e+00
-2.24180266e-01 -1.32516205e+00 1.49609834e-01 8.77607286e-01
-1.24109399e+00 -1.19724429e+00 -7.90289342e-01 5.87294042e-01
3.31882715e-01 3.86505514e-01 9.33801889e-01 1.29437551e-01
1.87789783e-01 1.35119274e-01 1.02124214e-01 -8.64636600e-02
6.93370700e-01 -1.42996395e+00 7.66481996e-01 3.21439117e-01
1.11253887e-01 7.37367570e-01 5.71854472e-01 -4.07366574e-01
-1.92952859e+00 -6.71683669e-01 2.54353613e-01 -1.21291161e+00
3.24550718e-01 -9.04935777e-01 -5.12800336e-01 1.02258170e+00
-2.86697358e-01 2.70022452e-01 2.81172603e-01 1.58760533e-01
-4.05036062e-01 -1.02364145e-01 -1.19278538e+00 4.69240725e-01
1.52243721e+00 -7.31318474e-01 -3.09569895e-01 6.07288659e-01
8.94972682e-01 -1.20738125e+00 -9.39611197e-01 3.95621747e-01
1.03306663e+00 -1.19593942e+00 1.21751535e+00 -3.00044000e-01
5.10160506e-01 -3.52522075e-01 -5.46469510e-01 -1.02171302e+00
-7.43852044e-03 -3.59326154e-01 -2.67275065e-01 8.66905868e-01
1.17868006e-01 -4.21605140e-01 1.13898599e+00 9.40264761e-01
-1.83740065e-01 -7.28551626e-01 -7.56606400e-01 -7.97706187e-01
-1.28166065e-01 -7.86158323e-01 7.70679593e-01 7.28938282e-01
-4.78955060e-01 -1.44428402e-01 -6.57442212e-02 3.44201952e-01
5.41923344e-01 4.14725929e-01 1.24461114e+00 -1.11027539e+00
-3.47100377e-01 -6.35347366e-02 -5.33623219e-01 -1.49234092e+00
-1.00602694e-01 -5.14712274e-01 1.41421705e-01 -1.56907976e+00
-6.99786916e-02 -8.13499570e-01 2.41131678e-01 5.93361378e-01
3.31758916e-01 3.87151927e-01 1.37428164e-01 3.48378867e-01
-4.36456352e-01 2.11014792e-01 1.17406011e+00 5.78183867e-02
-1.66484132e-01 -6.75114840e-02 -3.93791407e-01 8.83736610e-01
7.20345736e-01 -2.92583406e-01 -2.28471026e-01 -9.18588519e-01
3.29278529e-01 -1.82764530e-01 7.98083782e-01 -1.12421048e+00
-8.92649293e-02 -1.21265464e-01 7.47535110e-01 -7.83311188e-01
9.49820817e-01 -1.22219086e+00 2.11134598e-01 2.99458712e-01
-1.70982867e-01 3.13835405e-02 2.47196496e-01 1.67886540e-01
4.47258443e-01 -1.46951124e-01 3.97708684e-01 -5.25214672e-01
-7.69905388e-01 4.12445396e-01 1.28305778e-01 -3.00956339e-01
8.08934152e-01 -5.54200947e-01 -3.65092099e-01 -2.87277907e-01
-6.10887229e-01 -4.91885096e-03 1.05434871e+00 5.71180999e-01
6.02980971e-01 -1.30948544e+00 -1.84332058e-01 2.63283759e-01
1.59666732e-01 6.65959537e-01 -1.79490596e-01 4.52482373e-01
-1.03769958e+00 3.27320486e-01 -9.71417055e-02 -1.11548340e+00
-1.13185942e+00 3.84748340e-01 3.90154183e-01 4.72925633e-01
-6.24115407e-01 9.98528600e-01 1.05832212e-01 -8.14155579e-01
4.59705740e-01 -8.65697443e-01 5.31296372e-01 -4.33665603e-01
3.48206043e-01 2.49012217e-01 3.71545553e-01 -3.40639412e-01
-4.53469008e-01 7.82500803e-01 4.63758856e-02 -1.35926515e-01
1.49169862e+00 2.85406783e-02 -8.31009075e-02 6.45399690e-01
1.27951574e+00 2.89926901e-02 -1.68981302e+00 -1.65362194e-01
-2.62797415e-01 -8.91689956e-01 -9.33974162e-02 -7.06185043e-01
-7.40963280e-01 9.36055005e-01 7.41938293e-01 -3.73448730e-02
7.75395811e-01 3.58635008e-01 4.57853764e-01 7.60214806e-01
8.43826473e-01 -7.65790522e-01 1.99165285e-01 5.95879853e-01
9.24994349e-01 -1.28448296e+00 9.09163281e-02 -5.10844409e-01
-1.84362605e-01 1.09764993e+00 8.17411482e-01 -8.67612101e-03
4.61558610e-01 5.58687806e-01 7.77206793e-02 -4.36059177e-01
-3.95586848e-01 1.13307193e-01 8.31365287e-02 7.02758789e-01
2.18660802e-01 -2.38592736e-02 7.18127012e-01 1.35069475e-01
-5.20702958e-01 8.91409069e-02 3.25531602e-01 1.44848228e+00
-2.93156475e-01 -1.05439508e+00 -6.03159249e-01 3.37558597e-01
-2.11999103e-01 6.69726804e-02 -4.83964384e-01 7.88466692e-01
1.51684299e-01 4.59954292e-01 2.06879586e-01 -4.86704707e-01
5.60801268e-01 -1.82648405e-01 9.73340034e-01 -7.50891149e-01
-2.51034319e-01 -8.43549967e-02 -2.40833312e-02 -1.04181361e+00
-5.77802896e-01 -6.69598818e-01 -1.10847747e+00 -3.60675037e-01
-2.68573850e-01 -4.13281649e-01 1.02648318e+00 6.33930624e-01
4.08748657e-01 1.40635774e-01 3.75401706e-01 -1.64910805e+00
-5.09401858e-01 -6.90972507e-01 -4.76056993e-01 2.58720279e-01
3.93899530e-01 -7.56338537e-01 -2.32139856e-01 2.17460901e-01] | [7.012991905212402, -2.1491899490356445] |
b844a6b7-e260-44ca-a9ff-a69ea5f57576 | dmm-net-differentiable-mask-matching-network | 1909.12471 | null | https://arxiv.org/abs/1909.12471v1 | https://arxiv.org/pdf/1909.12471v1.pdf | DMM-Net: Differentiable Mask-Matching Network for Video Object Segmentation | In this paper, we propose the differentiable mask-matching network (DMM-Net) for solving the video object segmentation problem where the initial object masks are provided. Relying on the Mask R-CNN backbone, we extract mask proposals per frame and formulate the matching between object templates and proposals at one time step as a linear assignment problem where the cost matrix is predicted by a CNN. We propose a differentiable matching layer by unrolling a projected gradient descent algorithm in which the projection exploits the Dykstra's algorithm. We prove that under mild conditions, the matching is guaranteed to converge to the optimum. In practice, it performs similarly to the Hungarian algorithm during inference. Meanwhile, we can back-propagate through it to learn the cost matrix. After matching, a refinement head is leveraged to improve the quality of the matched mask. Our DMM-Net achieves competitive results on the largest video object segmentation dataset YouTube-VOS. On DAVIS 2017, DMM-Net achieves the best performance without online learning on the first frames. Without any fine-tuning, DMM-Net performs comparably to state-of-the-art methods on SegTrack v2 dataset. At last, our matching layer is very simple to implement; we attach the PyTorch code ($<50$ lines) in the supplementary material. Our code is released at https://github.com/ZENGXH/DMM_Net. | ['Sanja Fidler', 'Raquel Urtasun', 'Li Gu', 'Yuwen Xiong', 'Xiaohui Zeng', 'Renjie Liao'] | 2019-09-27 | dmm-net-differentiable-mask-matching-network-1 | http://openaccess.thecvf.com/content_ICCV_2019/html/Zeng_DMM-Net_Differentiable_Mask-Matching_Network_for_Video_Object_Segmentation_ICCV_2019_paper.html | http://openaccess.thecvf.com/content_ICCV_2019/papers/Zeng_DMM-Net_Differentiable_Mask-Matching_Network_for_Video_Object_Segmentation_ICCV_2019_paper.pdf | iccv-2019-10 | ['one-shot-visual-object-segmentation'] | ['computer-vision'] | [ 1.07805863e-01 2.75477588e-01 -3.93837273e-01 -3.52448761e-01
-8.91350389e-01 -4.97647703e-01 8.47086534e-02 -4.35030192e-01
-5.34724057e-01 2.12290853e-01 -4.01304215e-02 -3.60146880e-01
2.62974441e-01 -5.13997853e-01 -1.28945434e+00 -3.80339891e-01
1.04567535e-01 4.39906478e-01 4.49934930e-01 2.29968831e-01
1.52223617e-01 2.85977691e-01 -1.10811555e+00 6.79554716e-02
6.90992236e-01 1.26378012e+00 4.48839605e-01 6.92768276e-01
6.89769536e-02 4.78652179e-01 2.74030888e-03 -6.23514593e-01
9.09239471e-01 -1.89326689e-01 -9.52694833e-01 4.10130322e-01
8.60461473e-01 -7.09874988e-01 -7.64558077e-01 1.21870577e+00
2.78891921e-01 1.51384875e-01 3.93136352e-01 -1.20164323e+00
-3.99225533e-01 5.83584785e-01 -6.56452954e-01 5.23778349e-02
6.81335153e-03 4.45055425e-01 1.03519225e+00 -1.28933239e+00
7.35418379e-01 1.00652719e+00 6.76150799e-01 6.43407881e-01
-1.16743636e+00 -6.10288739e-01 5.23471653e-01 1.82802290e-01
-1.33168364e+00 -4.72713917e-01 5.51690459e-01 -5.19632220e-01
7.07754672e-01 1.27720200e-02 9.08888996e-01 7.45363712e-01
-1.45765200e-01 1.23946965e+00 3.94474566e-01 4.83887121e-02
1.27300262e-01 -3.43861252e-01 1.40651301e-01 1.14443672e+00
1.01056561e-01 -5.99502362e-02 -4.70738232e-01 2.10407123e-01
9.79129672e-01 1.88160732e-01 -3.40306908e-01 -5.15807986e-01
-1.03665817e+00 6.42817914e-01 5.70429921e-01 -1.38567656e-01
-2.60767847e-01 5.75316787e-01 1.99118152e-01 1.25901267e-01
4.80725497e-01 9.85764116e-02 -6.55192435e-01 -7.09874332e-02
-1.29967260e+00 4.41132933e-01 6.49719536e-01 1.34166574e+00
9.29758906e-01 -1.13705561e-01 -2.41347745e-01 5.75974762e-01
4.64410305e-01 3.59306782e-01 4.81435508e-02 -1.51177180e+00
6.93073630e-01 3.57552201e-01 1.20110651e-02 -8.91868353e-01
-3.14478219e-01 -4.88416404e-01 -6.91077352e-01 -3.80692147e-02
5.17869294e-01 -3.22225928e-01 -1.18641686e+00 1.66230142e+00
5.40621758e-01 6.93715215e-01 -4.18732852e-01 1.21155095e+00
7.62298346e-01 7.60039091e-01 -3.60037029e-01 8.24374333e-02
1.17262506e+00 -1.49030876e+00 -4.26920831e-01 -3.68910432e-01
4.88229990e-01 -5.82165658e-01 8.42236519e-01 1.76882774e-01
-1.45417106e+00 -5.03033578e-01 -8.49384606e-01 -2.76357740e-01
9.46566835e-02 1.98419556e-01 4.60128576e-01 3.20286930e-01
-1.12330782e+00 8.10901403e-01 -1.14396393e+00 -1.57319102e-02
9.90722239e-01 5.86908698e-01 -1.24404676e-01 -9.93428901e-02
-5.40833652e-01 3.67828608e-01 2.95143038e-01 3.73478711e-01
-1.27851522e+00 -9.36321914e-01 -1.01105833e+00 1.65227596e-02
8.53068054e-01 -8.57538283e-01 1.54801190e+00 -8.67833853e-01
-1.56802356e+00 9.27563250e-01 -4.85810995e-01 -5.79130292e-01
9.06045675e-01 -3.72933030e-01 2.34701380e-01 2.03620180e-01
2.46387526e-01 1.16203845e+00 8.99553716e-01 -1.01651156e+00
-8.24626207e-01 -1.18894622e-01 1.20534532e-01 -3.46923135e-02
-7.83132389e-03 -9.28769633e-02 -1.43205619e+00 -6.11470282e-01
2.99010396e-01 -9.93770719e-01 -4.78020400e-01 4.62866843e-01
-5.15324712e-01 -7.02113807e-02 6.69884861e-01 -7.40401328e-01
1.11802983e+00 -2.25914907e+00 2.56697804e-01 2.83494979e-01
5.47406375e-01 2.87878662e-01 -3.52955014e-01 -8.25769678e-02
2.39371583e-01 1.47358850e-01 -6.60593450e-01 -8.12188268e-01
1.53485239e-01 1.46378189e-01 -1.54565170e-01 7.42859066e-01
1.65855527e-01 1.31201637e+00 -6.74963176e-01 -4.81587201e-01
1.09496981e-01 2.33912736e-01 -9.39621150e-01 3.69749486e-01
-4.89422143e-01 2.96707988e-01 -2.15551168e-01 6.45033836e-01
8.51324260e-01 -5.61896563e-01 7.29494393e-02 -2.87653506e-01
-7.01868609e-02 2.26661086e-01 -1.27115536e+00 2.16911507e+00
-3.45089585e-02 7.07928360e-01 3.90683234e-01 -9.95854497e-01
5.17776310e-01 4.29248884e-02 5.78608811e-01 -4.47378963e-01
3.09207916e-01 2.26729900e-01 -1.40249297e-01 -3.98466647e-01
3.13528538e-01 4.74221736e-01 3.09166759e-01 1.59017250e-01
2.71358460e-01 3.45116891e-02 3.39518666e-01 2.72530854e-01
9.70096886e-01 3.41641068e-01 -1.39646128e-01 -1.40352249e-01
2.31425375e-01 -6.67797104e-02 1.04365194e+00 8.35371733e-01
-2.08389774e-01 8.29115689e-01 5.34535885e-01 -4.39714432e-01
-9.82475698e-01 -1.04515910e+00 3.30549255e-02 9.40573871e-01
3.45597386e-01 -4.24141705e-01 -1.10698426e+00 -8.35994482e-01
1.11474656e-02 1.04329921e-01 -5.08668780e-01 2.25261256e-01
-8.16563249e-01 -2.90962130e-01 2.33830437e-01 6.53451204e-01
6.22183323e-01 -9.61611271e-01 -4.32902396e-01 2.63167948e-01
-4.76646014e-02 -1.26944554e+00 -1.25556791e+00 -5.82573339e-02
-9.99204278e-01 -1.11285722e+00 -8.92558038e-01 -1.03060794e+00
8.86189938e-01 2.99845845e-01 1.07750988e+00 4.31742013e-01
-2.07113549e-01 2.31049567e-01 -7.85211846e-02 -8.07254091e-02
9.73283276e-02 4.88660693e-01 -3.82787883e-01 1.41606346e-01
7.17989132e-02 -5.05730510e-01 -9.28298473e-01 4.18943435e-01
-6.72124028e-01 4.31157738e-01 3.96427363e-01 6.05451226e-01
9.12925899e-01 -4.48301762e-01 4.15705703e-02 -8.40576708e-01
-1.49056613e-01 -3.90176922e-01 -1.12104893e+00 -2.23431792e-02
-4.05294180e-01 -7.40707517e-02 2.72575468e-01 -3.86271954e-01
-4.53091949e-01 4.37263399e-01 -3.44429046e-01 -7.61752665e-01
8.22286680e-02 1.77066475e-01 -3.19503635e-01 -1.75759792e-01
-1.20071247e-02 -1.07818395e-02 -5.43553308e-02 -5.22259712e-01
4.59075719e-01 2.34063074e-01 8.60843182e-01 -6.58596575e-01
1.00903881e+00 5.53809702e-01 -2.21054435e-01 -2.99476564e-01
-1.05159974e+00 -5.89725792e-01 -5.79516113e-01 -3.08634847e-01
1.01391375e+00 -9.03566301e-01 -9.63599026e-01 5.05105257e-01
-1.29142606e+00 -9.50189531e-01 -2.35921934e-01 3.15562159e-01
-6.96334779e-01 2.89484292e-01 -8.30004156e-01 -4.77844238e-01
-3.04354250e-01 -1.29959691e+00 1.04244959e+00 2.03371152e-01
9.99138430e-02 -6.86819971e-01 -2.49949932e-01 4.32482630e-01
1.90243110e-01 -2.59892410e-03 2.79094398e-01 -3.70496541e-01
-1.20353544e+00 2.21709400e-01 -3.45029116e-01 2.25620598e-01
-1.11968257e-01 -4.72404622e-02 -7.29265630e-01 -2.58766562e-01
-1.90120772e-01 -4.64758314e-02 1.25502539e+00 7.55106032e-01
1.52695918e+00 -5.46347797e-01 -2.50757843e-01 1.39003539e+00
1.31151938e+00 4.87626530e-03 4.76337284e-01 3.43591154e-01
1.01487648e+00 2.26012930e-01 5.43868721e-01 3.98081690e-01
5.60484290e-01 6.48864865e-01 7.12318063e-01 -2.32380748e-01
-1.69090912e-01 -3.56536925e-01 3.09254080e-01 6.78846776e-01
1.37494639e-01 -1.18738569e-01 -7.03189135e-01 5.38621724e-01
-2.20424509e+00 -6.84314907e-01 -1.36375446e-02 2.05074811e+00
7.40496755e-01 3.01998824e-01 2.67601609e-01 -4.06354636e-01
7.78410733e-01 2.40712211e-01 -8.51619422e-01 6.46669790e-02
-7.03817829e-02 1.77703589e-01 8.65597248e-01 8.96733403e-01
-1.30656862e+00 1.28837132e+00 5.66932487e+00 7.09421158e-01
-9.64694142e-01 2.79276341e-01 7.08205581e-01 -2.37006232e-01
-2.22288251e-01 8.46501514e-02 -9.21303153e-01 5.64493477e-01
5.41490078e-01 1.18608154e-01 7.51176476e-01 7.29673862e-01
2.91704863e-01 1.06068268e-01 -1.17506742e+00 1.04985940e+00
-1.69020265e-01 -1.83286321e+00 -2.11992726e-01 1.95081998e-02
7.83478320e-01 4.53365475e-01 8.20465609e-02 8.86448920e-02
1.64832503e-01 -7.96813369e-01 1.09481931e+00 3.54214311e-01
6.58614516e-01 -5.70731163e-01 4.76021111e-01 2.05701053e-01
-1.41520858e+00 -4.29661758e-02 -4.24524367e-01 9.67989489e-02
5.28563857e-01 4.96202886e-01 -4.69923228e-01 3.24839115e-01
6.98542178e-01 9.61552203e-01 -2.94133812e-01 1.37029493e+00
-2.65171736e-01 6.70860171e-01 -4.17102784e-01 4.40748662e-01
3.10898900e-01 -4.52772856e-01 4.52730536e-01 1.35957992e+00
1.79737076e-01 9.16273370e-02 4.16845173e-01 1.09914982e+00
-5.74868858e-01 -4.98309247e-02 -1.01255737e-01 1.55389354e-01
3.52610618e-01 1.27964187e+00 -8.76270354e-01 -5.32440305e-01
-5.19936681e-01 1.12174010e+00 3.76580387e-01 4.77288663e-01
-1.00608897e+00 9.02726650e-02 7.82096863e-01 1.75941274e-01
8.40287864e-01 -2.99074322e-01 -4.04941410e-01 -1.25016880e+00
2.58217216e-01 -8.84121537e-01 2.24527359e-01 -4.76368636e-01
-1.15343809e+00 3.60185295e-01 -2.06170782e-01 -1.03273535e+00
1.13898635e-01 -6.75083399e-01 -5.71345568e-01 4.82405782e-01
-1.52594578e+00 -8.75941992e-01 -2.21641645e-01 4.74261880e-01
7.35210538e-01 5.59916683e-02 9.45568830e-02 5.59677303e-01
-9.20380592e-01 7.29454935e-01 -2.48768359e-01 6.17444217e-01
3.51515889e-01 -1.14885879e+00 8.82135749e-01 1.12195015e+00
2.50962704e-01 3.53860974e-01 3.17503333e-01 -6.24224842e-01
-1.47011113e+00 -1.36103952e+00 3.99784356e-01 -6.14202209e-02
5.63456416e-01 -5.42683303e-01 -9.37685072e-01 9.17686760e-01
1.12466611e-01 4.89869982e-01 2.47368604e-01 -3.35288256e-01
-3.06485116e-01 -1.47605479e-01 -8.32662106e-01 5.60821950e-01
1.38485050e+00 -3.08501989e-01 -2.00611353e-01 3.87846231e-01
1.13892949e+00 -1.02240181e+00 -6.46606863e-01 3.14802349e-01
4.58478808e-01 -7.02175558e-01 7.94611275e-01 -6.38045609e-01
4.48831648e-01 -5.73783815e-01 -1.88448012e-01 -7.03539312e-01
-2.92369425e-01 -1.04446816e+00 -4.72003728e-01 9.92679834e-01
7.51335442e-01 -4.97938573e-01 1.09285438e+00 7.38612056e-01
-5.20549536e-01 -1.24265254e+00 -9.09597218e-01 -7.23136067e-01
-1.22388795e-01 -7.37186134e-01 5.80965281e-01 5.03474534e-01
-5.93153238e-01 -7.43126571e-02 -3.72675717e-01 4.17430907e-01
7.45973110e-01 1.61773950e-01 9.28651035e-01 -7.26785004e-01
-7.06346273e-01 -5.85797429e-01 -1.92790061e-01 -1.90371192e+00
3.54541540e-01 -1.06764460e+00 1.93906486e-01 -1.64123738e+00
3.26695085e-01 -3.06596309e-01 -1.21278509e-01 5.33187509e-01
-2.24319115e-01 3.38002235e-01 6.91484749e-01 1.22461148e-01
-8.32829714e-01 4.73667413e-01 1.32517958e+00 -3.12176824e-01
-2.53409982e-01 1.98910162e-01 -5.48485279e-01 8.59292626e-01
8.02852929e-01 -6.82947099e-01 -5.50100803e-02 -7.75257409e-01
1.30183876e-01 1.01789441e-02 4.53048646e-01 -7.99890995e-01
5.08778691e-01 -1.03486322e-01 1.80847630e-01 -6.61437809e-01
2.89049149e-01 -6.99798346e-01 -1.65430065e-02 5.21896899e-01
-3.01221490e-01 1.33133933e-01 5.17108738e-02 3.45558852e-01
4.40707132e-02 -3.19212198e-01 6.83452964e-01 6.62467927e-02
-6.63337171e-01 1.15475976e+00 -9.07879844e-02 3.57522011e-01
7.72393286e-01 -2.01600030e-01 1.64098181e-02 -2.19094843e-01
-7.03579068e-01 5.94327271e-01 5.76503634e-01 2.07290053e-01
6.82685614e-01 -1.14382052e+00 -6.21852994e-01 -3.79696526e-02
-3.60754073e-01 6.41669810e-01 2.02262595e-01 1.08253086e+00
-5.66456735e-01 1.03377402e-01 2.97549754e-01 -7.58491099e-01
-1.05459619e+00 4.90214974e-01 5.69341242e-01 1.28493933e-02
-8.92988384e-01 1.06665182e+00 3.63100171e-01 -2.83364892e-01
6.40926957e-01 -5.16679227e-01 3.24552178e-01 -3.04105014e-01
4.72454071e-01 4.10671234e-01 -2.14482442e-01 -3.62345695e-01
-3.33165199e-01 7.64386415e-01 -9.31749344e-02 -1.93734705e-01
1.30635560e+00 -5.15795387e-02 -1.19084321e-01 8.71151015e-02
1.45326781e+00 -1.86924353e-01 -1.96777797e+00 -2.90178210e-01
-6.72618002e-02 -5.08444190e-01 -1.40069664e-01 -2.62562931e-01
-1.75600445e+00 6.82635069e-01 3.59819263e-01 -3.10903281e-01
1.08069146e+00 1.93225697e-01 1.09228373e+00 3.54518443e-01
1.36499122e-01 -1.04550350e+00 -7.45016560e-02 5.47937930e-01
6.14664674e-01 -1.24977660e+00 -7.49766901e-02 -6.48522556e-01
-3.58533919e-01 9.74141359e-01 7.09417999e-01 -3.84675562e-01
7.24830449e-01 3.73340368e-01 -2.06438899e-02 -1.23964690e-01
-5.80264330e-01 -2.02406406e-01 4.53763217e-01 1.96110696e-01
1.28704295e-01 -3.65915708e-02 -1.57781303e-01 4.85118598e-01
-1.72386333e-01 3.66281793e-02 3.59536022e-01 6.78785563e-01
-3.68398070e-01 -9.23094332e-01 -9.00006741e-02 4.29439574e-01
-5.24790406e-01 -1.35754749e-01 -2.82546096e-02 5.04700840e-01
1.69051543e-01 7.16565669e-01 2.25388408e-01 -2.09890872e-01
2.81001568e-01 -2.65433937e-01 4.39556628e-01 -5.58519483e-01
-4.48463619e-01 2.15619624e-01 -2.64189333e-01 -1.18513668e+00
-3.49581331e-01 -6.35222673e-01 -1.51945388e+00 -3.19434047e-01
-1.73520848e-01 -1.15877829e-01 4.78017002e-01 1.02026284e+00
5.27703524e-01 4.44617927e-01 4.76861358e-01 -1.18540955e+00
-3.02327335e-01 -5.45597374e-01 -2.06416443e-01 7.32934251e-02
4.27910328e-01 -3.83470327e-01 -1.61542684e-01 2.16177747e-01] | [9.201544761657715, -0.10571889579296112] |
e7de6d47-6836-4aea-b771-f0e480ad19a9 | dreeam-guiding-attention-with-evidence-for | 2302.08675 | null | https://arxiv.org/abs/2302.08675v1 | https://arxiv.org/pdf/2302.08675v1.pdf | DREEAM: Guiding Attention with Evidence for Improving Document-Level Relation Extraction | Document-level relation extraction (DocRE) is the task of identifying all relations between each entity pair in a document. Evidence, defined as sentences containing clues for the relationship between an entity pair, has been shown to help DocRE systems focus on relevant texts, thus improving relation extraction. However, evidence retrieval (ER) in DocRE faces two major issues: high memory consumption and limited availability of annotations. This work aims at addressing these issues to improve the usage of ER in DocRE. First, we propose DREEAM, a memory-efficient approach that adopts evidence information as the supervisory signal, thereby guiding the attention modules of the DocRE system to assign high weights to evidence. Second, we propose a self-training strategy for DREEAM to learn ER from automatically-generated evidence on massive data without evidence annotations. Experimental results reveal that our approach exhibits state-of-the-art performance on the DocRED benchmark for both DocRE and ER. To the best of our knowledge, DREEAM is the first approach to employ ER self-training. | ['Naoaki Okazaki', 'An Wang', 'Youmi Ma'] | 2023-02-17 | null | null | null | null | ['document-level-relation-extraction'] | ['natural-language-processing'] | [ 9.29583609e-02 5.58068037e-01 -5.15079439e-01 4.35494818e-03
-9.00762796e-01 -2.14467421e-01 7.49051571e-01 8.09899092e-01
-6.17237866e-01 9.34004188e-01 1.27509013e-01 -3.52299869e-01
-4.16146010e-01 -7.83545315e-01 -7.94034600e-01 -1.67878151e-01
-1.09979838e-01 5.83143175e-01 4.35036451e-01 -3.13925058e-01
2.38031074e-01 4.77790266e-01 -1.33443451e+00 4.81690466e-01
7.53386796e-01 1.15797579e+00 9.63259116e-03 6.50638998e-01
-9.24774632e-02 1.23863745e+00 -7.46924222e-01 -7.89400458e-01
-1.52452961e-01 -2.91763507e-02 -1.35861373e+00 -1.76414419e-02
3.76508445e-01 -1.85202733e-01 -6.51983023e-01 8.62980723e-01
3.31437558e-01 -1.03054531e-02 6.35050833e-01 -9.07893419e-01
-7.88922071e-01 1.22250819e+00 -5.82423329e-01 9.29345429e-01
2.79872179e-01 -4.28516388e-01 1.40239346e+00 -1.30086970e+00
9.23069119e-01 9.17134702e-01 3.30379307e-01 2.43808240e-01
-9.19209778e-01 -4.81181890e-01 2.05623314e-01 5.05670488e-01
-1.49167097e+00 -6.25080585e-01 6.48273766e-01 -2.27976099e-01
1.61428380e+00 3.06867152e-01 3.63262683e-01 8.40567172e-01
4.70816903e-02 1.07869613e+00 7.43801475e-01 -7.98495829e-01
4.70875204e-02 1.31744638e-01 8.03114653e-01 7.23518133e-01
5.29226363e-01 -2.35823601e-01 -8.54448080e-01 -1.51243940e-01
3.53856206e-01 -3.62792909e-01 -3.84222865e-01 1.75516307e-01
-7.82237887e-01 5.53647757e-01 4.16309923e-01 5.82324684e-01
-6.16763234e-01 -9.92283970e-02 5.77047825e-01 2.66482681e-01
5.26963174e-01 6.10799432e-01 -7.75624514e-01 1.15005739e-01
-5.01446784e-01 5.23884483e-02 8.30047488e-01 9.20849144e-01
3.36341113e-01 -4.74916399e-01 -4.51299310e-01 9.77861762e-01
1.70436963e-01 1.31478667e-01 5.90811491e-01 -4.32720900e-01
1.07935846e+00 8.65501523e-01 5.97826988e-02 -1.04721749e+00
-3.17609221e-01 -7.34053731e-01 -8.28898013e-01 -4.06706929e-01
-2.25528982e-02 -1.29428461e-01 -6.78579271e-01 1.34835732e+00
2.57692158e-01 2.43588313e-01 4.69828725e-01 6.65947378e-01
1.38328469e+00 5.81089675e-01 1.58504948e-01 -3.48874897e-01
1.51361465e+00 -1.13870835e+00 -9.53456223e-01 -5.31018913e-01
7.68161058e-01 -4.60442483e-01 6.69004738e-01 3.91376495e-01
-1.12923241e+00 -2.66484261e-01 -1.33852947e+00 -2.33959809e-01
-5.94246328e-01 5.63821852e-01 6.39871716e-01 7.01910853e-02
-4.74870473e-01 6.04979217e-01 -5.59865654e-01 -8.40257034e-02
4.37533736e-01 3.03945571e-01 -4.60999697e-01 2.41256151e-02
-1.52322721e+00 1.17749870e+00 7.40485013e-01 3.95841628e-01
-3.40728253e-01 -3.10418904e-01 -9.82701361e-01 4.91286784e-01
1.00245357e+00 -4.16054606e-01 1.37706900e+00 -2.04639718e-01
-1.10229862e+00 7.65532255e-01 -1.26467332e-01 -9.00985301e-01
-1.40491903e-01 -6.54952705e-01 -6.89818799e-01 1.90520301e-01
2.80631389e-02 2.56169766e-01 3.84247065e-01 -1.14548612e+00
-6.74642861e-01 -2.12334529e-01 -1.40913785e-01 5.42851277e-02
-6.12178445e-01 2.08456978e-01 -6.98909879e-01 -5.39300323e-01
3.88728715e-02 -5.73496521e-01 5.98788895e-02 -9.24648404e-01
-6.32820666e-01 -9.08913434e-01 6.02405608e-01 -6.08894289e-01
1.82990754e+00 -1.88180006e+00 -6.81589991e-02 2.69327700e-01
5.49032390e-01 6.87961340e-01 -3.59546468e-02 4.69770402e-01
-9.77786407e-02 2.39643663e-01 4.88318428e-02 -2.51002073e-01
-1.39377505e-01 1.72272041e-01 -4.94341105e-01 -7.12349564e-02
7.32168257e-01 1.10802424e+00 -1.03801179e+00 -8.96226227e-01
-2.95765102e-01 2.53968030e-01 -9.25535394e-04 2.23215029e-01
-2.49222964e-01 -1.94624722e-01 -5.22958517e-01 4.55725223e-01
3.27403814e-01 -6.53356671e-01 4.04827744e-01 -1.70049489e-01
2.04014376e-01 9.22206163e-01 -1.07691705e+00 1.04493940e+00
-4.44850206e-01 5.58296263e-01 -4.23734248e-01 -8.95170450e-01
1.02125919e+00 5.75186789e-01 2.28572205e-01 -6.95423186e-01
1.22928552e-01 4.05192047e-01 1.06525011e-01 -5.51577628e-01
9.44922686e-01 3.98903102e-01 -7.16238143e-03 2.07439631e-01
2.25537941e-01 3.12988997e-01 7.44470835e-01 5.90349972e-01
1.55404902e+00 -2.67976254e-01 7.44810879e-01 2.43484706e-01
6.76812410e-01 -1.11747116e-01 5.25169134e-01 6.40184343e-01
4.87326294e-01 -3.90650257e-02 4.32361454e-01 -2.63660073e-01
-4.96019334e-01 -5.35830081e-01 -1.61896959e-01 8.71256590e-01
4.79785278e-02 -9.59909379e-01 -4.09732670e-01 -1.13928592e+00
-5.98892011e-03 7.29066074e-01 -5.30833483e-01 -1.34861410e-01
-7.71866560e-01 -7.55491257e-01 4.35999334e-01 8.14097643e-01
5.32634556e-01 -1.31283939e+00 -5.72444975e-01 6.29438758e-01
-3.49626809e-01 -1.49722719e+00 -2.25679696e-01 6.69273019e-01
-7.19887495e-01 -1.25793087e+00 -2.01015756e-01 -5.15502393e-01
5.19552052e-01 3.27435374e-01 1.63501298e+00 5.12616873e-01
-8.67381133e-03 1.42427713e-01 -5.70670009e-01 -4.57169861e-01
-2.55892813e-01 3.98186177e-01 -3.08430880e-01 -2.85492957e-01
8.09727967e-01 -3.36845756e-01 -3.22336346e-01 3.15655112e-01
-6.96484268e-01 -1.07653610e-01 1.07146752e+00 9.46124732e-01
7.03196049e-01 3.14345539e-01 8.04165006e-01 -1.30830324e+00
1.01140332e+00 -5.90838075e-01 -4.57836658e-01 6.00619555e-01
-1.04859459e+00 2.82566786e-01 4.12800342e-01 -2.82722205e-01
-9.35257673e-01 -2.92466670e-01 -2.07015611e-02 -1.26247695e-02
5.38826250e-02 1.03918386e+00 -4.69446406e-02 4.28582162e-01
6.90062225e-01 1.08084744e-02 -8.65434885e-01 -4.18000400e-01
2.26657957e-01 9.04558539e-01 7.54008710e-01 -7.25219846e-01
4.22433019e-01 -1.27090856e-01 1.27818942e-01 -4.34410721e-01
-1.53972864e+00 -8.41561615e-01 -5.84274113e-01 1.25421226e-01
4.82286721e-01 -7.98806906e-01 -5.25475919e-01 -1.84116721e-01
-1.33872736e+00 -4.67204526e-02 -3.39576662e-01 2.08451450e-01
-3.26572396e-02 2.31525600e-01 -8.40963423e-01 -7.80162275e-01
-7.38666236e-01 -6.59831464e-01 1.09229302e+00 1.14566185e-01
-3.09374988e-01 -5.77641428e-01 1.60040602e-01 4.73672599e-01
-2.23856702e-01 -2.19389886e-01 5.97151518e-01 -1.06902099e+00
-5.58838010e-01 -5.13454974e-01 -4.76344883e-01 3.77576165e-02
-5.22343330e-02 -1.88445836e-01 -8.65583837e-01 1.23371884e-01
-3.83945674e-01 -5.30644119e-01 1.03641677e+00 -2.28128463e-01
1.28738630e+00 -5.84681809e-01 -7.22445905e-01 -1.59851447e-01
1.09104598e+00 1.84947908e-01 5.59150636e-01 5.67698598e-01
4.34693009e-01 4.73332644e-01 1.15460074e+00 2.43892595e-01
5.91547668e-01 8.57006907e-01 1.21964328e-01 -4.63812351e-02
-3.09226632e-01 -2.78886706e-01 1.54301956e-01 8.73913348e-01
-1.69079602e-01 -7.71835446e-01 -9.39997852e-01 6.38626516e-01
-2.16522837e+00 -8.65733802e-01 -3.82344186e-01 1.68948901e+00
1.19645929e+00 5.80384493e-01 -2.10268766e-01 5.52632570e-01
4.65254605e-01 -1.04841881e-01 -2.14854941e-01 -3.14792842e-01
-1.22648031e-01 3.21726680e-01 3.58967602e-01 2.69501805e-01
-1.34358108e+00 9.73006308e-01 5.31342554e+00 9.04790580e-01
-6.27985120e-01 1.51033819e-01 5.78462541e-01 3.36123079e-01
1.30537495e-01 -1.72965705e-01 -1.38741088e+00 2.22616270e-01
1.08948147e+00 -2.80600637e-02 -1.51821166e-01 8.62395704e-01
-2.89297879e-01 -9.09207314e-02 -1.32427347e+00 6.45548761e-01
9.48266461e-02 -1.45188165e+00 -2.65690595e-01 2.73308232e-02
5.33126175e-01 -4.59912568e-02 -3.78835231e-01 4.62345660e-01
5.02718031e-01 -8.27224791e-01 4.44190025e-01 3.32578599e-01
3.77684206e-01 -6.73018157e-01 1.28768051e+00 4.03539717e-01
-1.29710078e+00 -7.73411691e-02 -3.45113724e-01 2.68185169e-01
1.06887467e-01 8.54825437e-01 -1.26573288e+00 6.84923232e-01
4.68209684e-01 5.56162119e-01 -6.60099745e-01 7.50259936e-01
-7.19272614e-01 6.46292567e-01 -1.92392185e-01 -1.07177414e-01
-4.17743474e-02 3.98676336e-01 4.17918354e-01 1.71428454e+00
6.06215969e-02 3.18213075e-01 1.76872700e-01 4.72800851e-01
-6.04301989e-01 2.04099953e-01 -4.54755604e-01 -1.47538871e-01
7.29003191e-01 1.36329544e+00 -6.10162973e-01 -5.69566607e-01
-3.00456792e-01 6.72936738e-01 8.94802570e-01 -1.02786273e-01
-4.33655977e-01 -4.24629837e-01 -1.88046634e-01 -7.55164633e-03
5.90025425e-01 -4.05823439e-02 -2.23813981e-01 -1.02349722e+00
2.51709431e-01 -7.13820577e-01 8.33244324e-01 -7.97887027e-01
-1.16947222e+00 8.66054237e-01 -1.26752099e-02 -9.02515829e-01
-3.90522897e-01 -5.04694223e-01 -2.80925632e-01 5.26465654e-01
-1.80653560e+00 -8.51919413e-01 -6.95672929e-02 4.91024479e-02
6.28520966e-01 -2.28567302e-01 7.98678339e-01 5.16442776e-01
-8.46928895e-01 6.75123334e-01 -3.44224870e-01 4.90564346e-01
4.31573540e-01 -1.52589309e+00 3.80359262e-01 1.04596376e+00
7.23277926e-01 7.69683838e-01 4.39809173e-01 -8.40218484e-01
-1.22204518e+00 -9.62167263e-01 1.49959004e+00 -5.24188757e-01
8.30882490e-01 -7.41579980e-02 -1.09428680e+00 6.69235647e-01
2.27444932e-01 2.38216683e-01 6.91194415e-01 6.27650082e-01
-4.13353324e-01 -9.50823724e-02 -7.70208836e-01 4.88698810e-01
1.03452671e+00 -4.90912974e-01 -1.05334961e+00 2.35699087e-01
7.27160692e-01 -4.95747775e-01 -1.19653916e+00 5.79565883e-01
9.70305577e-02 -3.83995801e-01 9.63606179e-01 -5.53408682e-01
7.62069583e-01 -2.52251297e-01 8.41283649e-02 -1.21315753e+00
-2.16186747e-01 -3.71834576e-01 -1.25067246e+00 1.59741271e+00
8.06504548e-01 -3.14556748e-01 5.01640737e-01 3.64749551e-01
-1.60083041e-01 -1.07212591e+00 -6.85544491e-01 -7.50428379e-01
-3.51212651e-01 -2.56204754e-01 6.31262898e-01 7.91836679e-01
3.87689769e-01 1.00903749e+00 -1.45014599e-01 2.88461417e-01
2.41025433e-01 2.12618738e-01 6.24365509e-01 -1.35708618e+00
-4.57932651e-01 -2.54604340e-01 -4.50316817e-02 -1.17288518e+00
2.90416330e-01 -1.08264947e+00 2.90385574e-01 -1.64434361e+00
3.48690480e-01 -5.65867245e-01 -5.79501927e-01 6.48215830e-01
-5.65357268e-01 7.06766546e-02 -1.56709850e-01 1.96438372e-01
-1.03597784e+00 4.16542411e-01 9.45954919e-01 -2.15487182e-01
-1.07752889e-01 -7.30270445e-02 -8.70250344e-01 7.11440742e-01
7.78348207e-01 -7.40099430e-01 -2.89909720e-01 -1.48478106e-01
6.11697733e-01 2.14873031e-02 1.56315938e-01 -5.42974949e-01
4.18397129e-01 5.93814924e-02 2.51322508e-01 -9.63217795e-01
1.06043942e-01 -7.81262577e-01 -4.22045469e-01 8.04499090e-02
-3.98688048e-01 9.63461846e-02 1.98195711e-01 5.37273526e-01
-4.37946945e-01 -5.28801262e-01 2.09627151e-01 8.92713442e-02
-6.89392924e-01 5.11865951e-02 -2.96870917e-01 1.76764950e-01
4.83992815e-01 2.41190150e-01 -7.64446795e-01 1.72209159e-01
-4.54339266e-01 3.41390342e-01 -4.85846549e-01 2.43990019e-01
8.70585322e-01 -1.16551971e+00 -7.56729841e-01 -2.24597931e-01
3.64134640e-01 4.32411700e-01 -2.14733377e-01 6.30353808e-01
1.62663624e-01 7.19622314e-01 4.38692540e-01 -1.67437658e-01
-1.69971466e+00 6.55789316e-01 -4.28744219e-02 -1.23673081e+00
-9.22418177e-01 9.11200583e-01 -4.18924123e-01 1.94886960e-02
2.10240662e-01 -3.56937945e-01 -7.88357317e-01 7.25860298e-02
6.51186883e-01 2.35867605e-01 6.82190776e-01 -2.56998152e-01
-2.48409927e-01 9.16749164e-02 -6.34573340e-01 6.58187270e-02
1.61464322e+00 -6.74209893e-02 -2.95347631e-01 2.17791066e-01
6.14327729e-01 3.15938175e-01 -7.31059313e-01 -5.59912622e-01
8.69073391e-01 -2.94742137e-01 3.81176502e-01 -1.11736226e+00
-8.56220961e-01 4.89096671e-01 1.00283772e-01 4.28302914e-01
9.80876565e-01 4.15834576e-01 9.68630612e-01 9.56626415e-01
2.09040746e-01 -1.11688018e+00 1.93595320e-01 7.22507775e-01
1.08512306e+00 -1.31180394e+00 3.17554057e-01 -8.47076952e-01
-4.71239090e-01 1.03655481e+00 7.88402498e-01 2.08535850e-01
5.42571366e-01 5.28345883e-01 -3.59580159e-01 -6.27906501e-01
-1.17311847e+00 -3.77447486e-01 8.09638977e-01 1.72414318e-01
6.98471963e-01 -1.49563119e-01 -7.32890368e-01 9.76459563e-01
-1.38380239e-02 1.57787383e-01 3.25342208e-01 1.04749990e+00
-3.52940053e-01 -1.29889572e+00 -2.65722536e-02 5.21558225e-01
-7.32226670e-01 -4.63305354e-01 -6.22232795e-01 5.61760843e-01
9.77534950e-02 1.00282192e+00 -1.63084343e-01 -2.01498285e-01
4.16261613e-01 1.83595642e-01 4.79350835e-01 -8.61494780e-01
-7.16814756e-01 -9.83050168e-02 8.74759674e-01 -1.82262316e-01
-4.44233656e-01 -4.24218416e-01 -1.33869791e+00 1.22672275e-01
-7.85507500e-01 5.02043128e-01 4.06560183e-01 1.19744086e+00
5.24723709e-01 9.38189209e-01 3.59193355e-01 -7.53099620e-02
-3.60379100e-01 -1.18375146e+00 -2.32379243e-01 2.09500521e-01
9.30830836e-02 -8.62565994e-01 -1.41046168e-02 -1.85891941e-01] | [9.3016996383667, 8.690985679626465] |
d696b102-97ac-4117-97ee-9d1a04b99c04 | lidar-gait-benchmarking-3d-gait-recognition | 2211.10598 | null | https://arxiv.org/abs/2211.10598v2 | https://arxiv.org/pdf/2211.10598v2.pdf | LidarGait: Benchmarking 3D Gait Recognition with Point Clouds | Video-based gait recognition has achieved impressive results in constrained scenarios. However, visual cameras neglect human 3D structure information, which limits the feasibility of gait recognition in the 3D wild world. Instead of extracting gait features from images, this work explores precise 3D gait features from point clouds and proposes a simple yet efficient 3D gait recognition framework, termed LidarGait. Our proposed approach projects sparse point clouds into depth maps to learn the representations with 3D geometry information, which outperforms existing point-wise and camera-based methods by a significant margin. Due to the lack of point cloud datasets, we built the first large-scale LiDAR-based gait recognition dataset, SUSTech1K, collected by a LiDAR sensor and an RGB camera. The dataset contains 25,239 sequences from 1,050 subjects and covers many variations, including visibility, views, occlusions, clothing, carrying, and scenes. Extensive experiments show that (1) 3D structure information serves as a significant feature for gait recognition. (2) LidarGait outperforms existing point-based and silhouette-based methods by a significant margin, while it also offers stable cross-view results. (3) The LiDAR sensor is superior to the RGB camera for gait recognition in the outdoor environment. The source code and dataset have been made available at https://lidargait.github.io. | ['Shiqi Yu', 'George Q. Huang', 'Rui Wang', 'Wei Wu', 'Chao Fan', 'Chuanfu Shen'] | 2022-11-19 | null | http://openaccess.thecvf.com//content/CVPR2023/html/Shen_LidarGait_Benchmarking_3D_Gait_Recognition_With_Point_Clouds_CVPR_2023_paper.html | http://openaccess.thecvf.com//content/CVPR2023/papers/Shen_LidarGait_Benchmarking_3D_Gait_Recognition_With_Point_Clouds_CVPR_2023_paper.pdf | cvpr-2023-1 | ['gait-recognition-in-the-wild', 'gait-recognition'] | ['computer-vision', 'computer-vision'] | [-3.83988559e-01 -8.57075036e-01 -6.57755136e-02 -2.82063723e-01
-3.80072474e-01 -3.04942280e-01 1.51381224e-01 -2.34232828e-01
-3.16763580e-01 4.33113039e-01 -6.06472325e-03 3.41906250e-01
1.52024657e-01 -8.19549978e-01 -5.40627718e-01 -7.22543716e-01
-2.86884785e-01 4.35868949e-01 5.25364935e-01 -2.11273655e-01
1.36562020e-01 6.75744712e-01 -1.77873576e+00 -2.38030493e-01
5.53311288e-01 1.04406416e+00 3.03647071e-01 4.88190740e-01
5.04734330e-02 1.04987741e-01 -4.46787685e-01 -3.35964680e-01
3.28031689e-01 -3.41829583e-02 -1.13868944e-01 4.81596380e-01
5.94847441e-01 -5.30842662e-01 -6.73100650e-01 6.92393899e-01
5.49731970e-01 -8.49655867e-02 4.35028523e-01 -1.56520605e+00
-5.29285550e-01 -5.92579246e-01 -8.18948388e-01 5.49993105e-02
7.99614251e-01 3.95720780e-01 6.09726846e-01 -8.43353808e-01
5.43069124e-01 1.13600945e+00 8.46561730e-01 3.28655392e-01
-7.15525389e-01 -5.07696033e-01 -1.67914584e-01 5.17727733e-01
-1.61700308e+00 -1.95532963e-01 1.02738893e+00 -3.08658838e-01
1.13876665e+00 1.98769003e-01 1.18114722e+00 1.20449924e+00
1.09069496e-01 7.43518174e-01 1.02825987e+00 -1.83184147e-01
2.12578252e-01 -4.16892052e-01 1.13532590e-02 9.60066020e-01
7.69767344e-01 1.04782864e-01 -7.25872576e-01 -1.83764681e-01
1.02455080e+00 4.34963793e-01 -2.43112952e-01 -8.50122511e-01
-1.18808806e+00 4.59701866e-01 5.70712805e-01 -4.94759008e-02
-3.65197748e-01 3.15845370e-01 3.39563906e-01 2.26195514e-01
1.60042852e-01 -4.17798877e-01 -1.67236924e-01 -4.69578594e-01
-7.98544288e-01 1.90306589e-01 4.06997919e-01 1.01704645e+00
8.13548088e-01 2.77678609e-01 5.94591737e-01 7.63595521e-01
5.35149395e-01 1.31951022e+00 5.09254456e-01 -8.82915914e-01
6.93252742e-01 7.79272377e-01 8.58045593e-02 -1.43372023e+00
-3.58608693e-01 1.17920220e-01 -6.94855571e-01 1.44973695e-01
2.54079372e-01 3.22354168e-01 -9.11026120e-01 9.87613261e-01
3.27652097e-01 1.49042502e-01 -2.81994760e-01 1.29766273e+00
8.54076028e-01 3.51397216e-01 -2.89574593e-01 2.49261223e-02
1.34351504e+00 -5.02812445e-01 -3.84050250e-01 -3.38991940e-01
4.41352874e-01 -5.29695868e-01 1.03899074e+00 4.55180168e-01
-6.67925596e-01 -5.00566185e-01 -1.06272268e+00 6.58717901e-02
-2.64555573e-01 2.65510887e-01 4.48132932e-01 9.19352710e-01
-1.01324654e+00 2.58148223e-01 -1.19921827e+00 -8.80714118e-01
3.07837486e-01 2.52308130e-01 -8.07291627e-01 -5.23674428e-01
-7.27272689e-01 7.54234731e-01 1.73721924e-01 1.20491996e-01
-3.63798946e-01 -6.76630810e-02 -1.16272795e+00 -3.60737175e-01
3.82182330e-01 -6.62635624e-01 4.97007370e-01 -8.20908695e-02
-1.12896228e+00 1.00089729e+00 -3.29497576e-01 -1.17386855e-01
6.21732533e-01 -4.01393503e-01 -3.10905933e-01 5.26811242e-01
2.54845709e-01 3.45955253e-01 8.27468097e-01 -1.14443839e+00
-2.65460551e-01 -8.16443861e-01 -2.87068456e-01 1.69456303e-01
-2.21938267e-01 -1.99513361e-01 -7.59966493e-01 -4.31984425e-01
5.13110816e-01 -9.95725155e-01 3.56304995e-03 3.22605044e-01
-2.68282950e-01 6.04684204e-02 1.10781205e+00 -6.74988389e-01
6.86026990e-01 -2.03368664e+00 -1.21052936e-01 6.50114864e-02
6.10421551e-03 2.11685285e-01 1.46259934e-01 5.72823882e-01
2.71286070e-01 -1.05790138e-01 -2.08463758e-01 -5.16055882e-01
-1.14351427e-02 6.89316571e-01 9.57751572e-02 7.43929088e-01
8.64135027e-02 8.19985867e-01 -7.53329396e-01 -8.22308123e-01
8.06370080e-01 5.53987443e-01 -4.37458038e-01 -2.30378695e-02
3.63757849e-01 1.81067631e-01 -6.18298411e-01 1.41917288e+00
9.14724946e-01 4.44275737e-02 -1.20627917e-01 -2.85431027e-01
-7.61419535e-02 -3.64135861e-01 -1.31312215e+00 1.88212240e+00
3.12308073e-02 3.16575289e-01 -1.31658509e-01 -1.08409035e+00
1.27272642e+00 1.83563799e-01 9.11369622e-01 -7.80844688e-01
-5.45691215e-02 2.32652992e-01 -7.28551269e-01 -6.59785926e-01
4.37558502e-01 1.04794435e-01 -4.06686217e-01 3.93202081e-02
3.00256871e-02 -6.46397993e-02 1.72422007e-01 -1.41308382e-01
1.20741618e+00 4.46209729e-01 4.02681231e-01 1.57208994e-01
3.45513433e-01 1.83782429e-01 8.42755914e-01 3.59184325e-01
-6.53001070e-01 9.44293559e-01 -2.01042462e-02 -6.42339110e-01
-9.50672150e-01 -1.49557543e+00 9.43821892e-02 1.85205817e-01
4.72337991e-01 -5.94766021e-01 -2.56044865e-01 -3.12854588e-01
3.48951459e-01 -1.09437332e-01 -2.69690007e-01 2.95647290e-02
-7.84702837e-01 -4.61643338e-01 5.29414237e-01 7.71199048e-01
1.18504441e+00 -5.09983778e-01 -1.02921283e+00 -1.51048079e-02
-5.40982366e-01 -1.37101841e+00 -4.74164486e-01 -2.93121487e-01
-1.15905023e+00 -1.19120312e+00 -8.42512429e-01 -4.52166706e-01
4.91080254e-01 6.86450660e-01 8.35408092e-01 1.77277699e-01
-5.70450366e-01 8.12902689e-01 -5.29617488e-01 3.99718769e-02
4.93623257e-01 -4.25684303e-01 6.10589683e-01 -1.82445824e-01
6.65113151e-01 -7.72006094e-01 -6.84694350e-01 6.03804886e-01
-1.30562708e-01 -2.83001095e-01 4.41516846e-01 6.07355535e-01
7.50093281e-01 5.35177141e-02 -6.17382713e-02 1.37030676e-01
1.47564663e-03 -2.35729113e-01 -3.72049689e-01 -7.31793195e-02
-2.11367518e-01 -4.83738661e-01 2.21837237e-01 -1.23257555e-01
-5.69717586e-01 3.55926663e-01 1.60261467e-02 -9.12425041e-01
-4.81224447e-01 1.90712497e-01 -3.54305536e-01 -1.94373131e-01
3.34003359e-01 4.81190056e-01 2.45847270e-01 -5.80481648e-01
-4.42906804e-02 5.55694103e-01 6.60340488e-01 -5.61331749e-01
1.05087745e+00 7.98811913e-01 1.87647894e-01 -1.33726478e+00
2.24360563e-02 -7.81636178e-01 -1.18884695e+00 -5.67157269e-01
7.20285714e-01 -1.10920823e+00 -7.77281821e-01 7.27726817e-01
-7.63093650e-01 1.55726984e-01 -5.52882180e-02 6.75239921e-01
-8.48547757e-01 8.52766335e-01 -6.22420728e-01 -9.01137471e-01
-2.24403918e-01 -9.45579290e-01 1.39976656e+00 1.33474231e-01
-1.51563674e-01 -6.47030652e-01 -1.69407442e-01 5.92794180e-01
-8.90131146e-02 7.49533057e-01 3.75385702e-01 8.39222595e-02
-6.83865726e-01 -4.54448313e-01 -5.97185493e-02 2.71531101e-02
2.81582206e-01 9.09425169e-02 -7.40981936e-01 -2.60955662e-01
-1.25907868e-01 -2.23153174e-01 5.69043815e-01 4.14076954e-01
5.39194822e-01 1.62305370e-01 -4.22571659e-01 6.89485252e-01
1.46772385e+00 1.82153478e-01 7.48331368e-01 5.82661986e-01
9.91122246e-01 3.60114127e-01 8.23393881e-01 4.97595102e-01
7.14611650e-01 9.46290374e-01 4.07725096e-01 8.98925662e-02
-1.00500584e-01 -4.11199242e-01 5.90106845e-01 1.08861387e+00
-5.05334675e-01 1.97044760e-01 -1.08361006e+00 4.04008657e-01
-1.80560958e+00 -1.00011194e+00 -2.63932586e-01 2.22165680e+00
7.78112048e-03 4.07148972e-02 5.26019752e-01 4.95844990e-01
6.40011966e-01 6.99743852e-02 -5.26111424e-01 -1.84983108e-02
-1.84043199e-01 -4.26580496e-02 5.44624388e-01 -6.60227761e-02
-1.13013136e+00 6.85806274e-01 5.76914167e+00 4.18996006e-01
-9.21724141e-01 -1.48111582e-03 -2.42447421e-01 -2.81606078e-01
4.41665173e-01 -3.07975888e-01 -6.52939081e-01 6.48156345e-01
5.29006064e-01 -6.09114356e-02 -1.39932200e-01 1.08363593e+00
3.74888062e-01 -2.15059206e-01 -7.39093482e-01 1.55815697e+00
2.84552008e-01 -9.00559902e-01 -4.38731313e-02 4.99004304e-01
9.32501107e-02 1.17368102e-01 -1.39553607e-01 2.45262310e-02
-1.69310421e-02 -6.15467727e-01 7.09578395e-01 4.50505406e-01
8.74944866e-01 -7.02837825e-01 6.06673241e-01 3.50283712e-01
-1.79173315e+00 -3.72151285e-02 -5.00858963e-01 -2.08551273e-01
3.08730543e-01 5.02385974e-01 -3.91964436e-01 7.46668398e-01
1.18889511e+00 1.31832564e+00 -6.24872386e-01 1.30914176e+00
-1.28570795e-01 4.68503684e-01 -5.80565155e-01 9.22881514e-02
-1.29095151e-03 -4.56699669e-01 5.83227217e-01 1.05104804e+00
7.05271661e-01 1.48876920e-01 4.99252528e-01 5.52304208e-01
3.72465640e-01 -2.03056887e-01 -9.84746873e-01 1.55729264e-01
4.90061522e-01 8.85247350e-01 -8.60484838e-01 -8.93149748e-02
-4.92949337e-01 1.19119513e+00 -1.25367731e-01 2.12464273e-01
-6.84488654e-01 -2.01012999e-01 9.54570234e-01 2.34462842e-01
5.38686037e-01 -9.88880932e-01 -1.78551108e-01 -1.51949763e+00
5.24413228e-01 -4.18328255e-01 3.52220774e-01 -1.02891171e+00
-1.17943752e+00 1.80220664e-01 2.00100899e-01 -1.80932963e+00
-1.29211590e-01 -8.09743583e-01 -3.66737843e-01 6.14816308e-01
-1.22325814e+00 -1.30763781e+00 -7.06923902e-01 9.17689323e-01
4.77384090e-01 -8.44896436e-02 7.17009068e-01 4.05547321e-01
-4.55578655e-01 2.38918662e-01 -3.26436921e-03 4.47072566e-01
5.04724681e-01 -9.23440337e-01 6.25715375e-01 8.69341195e-01
4.32569720e-02 4.72101480e-01 4.49781030e-01 -9.99139190e-01
-2.13431048e+00 -8.60163093e-01 5.93850374e-01 -6.39763534e-01
3.05933028e-01 -3.60621631e-01 -6.79764211e-01 5.30534446e-01
-5.18334985e-01 2.03941047e-01 5.63024163e-01 -1.71386063e-01
-3.10644209e-01 -1.89359143e-01 -1.34563291e+00 3.09428871e-01
1.52362585e+00 -3.66601884e-01 -6.98025823e-01 -9.39527585e-04
1.99441969e-01 -4.19344127e-01 -9.49902534e-01 2.69768894e-01
7.40835488e-01 -9.08113658e-01 1.38914251e+00 4.41550650e-02
1.11976184e-01 -5.58547437e-01 -6.46457970e-01 -1.17231858e+00
-2.03655586e-01 1.09568141e-01 -3.46286058e-01 8.93537641e-01
-6.09714329e-01 -6.49763405e-01 1.21349049e+00 3.86477530e-01
-5.14921322e-02 -3.09545457e-01 -1.28787279e+00 -1.23166370e+00
-2.73445129e-01 -6.03185475e-01 4.79630113e-01 6.15238309e-01
-9.53502879e-02 -3.43523562e-01 -6.07187986e-01 2.81298399e-01
1.24051797e+00 2.41733223e-01 9.66574848e-01 -1.38235044e+00
1.23160876e-01 4.11842279e-02 -1.50820243e+00 -1.05958080e+00
-1.70632213e-01 -5.32229483e-01 -1.91876680e-01 -1.64296532e+00
-8.88004601e-02 -2.70409167e-01 1.89166099e-01 5.52393913e-01
1.81428179e-01 6.48233414e-01 3.36647570e-01 5.49989641e-01
-4.61043328e-01 7.03376114e-01 1.01178837e+00 -1.43211678e-01
1.95575088e-01 -2.07627788e-01 -1.28561631e-01 7.59985268e-01
6.99496269e-01 -1.41988263e-01 -2.66218573e-01 -5.34963131e-01
-4.04703617e-01 8.01432207e-02 7.79735506e-01 -1.44643891e+00
9.74860191e-02 -1.55350164e-01 7.45824516e-01 -1.04291213e+00
8.94527256e-01 -1.01958060e+00 2.67766416e-01 7.60259390e-01
6.99678481e-01 2.51892030e-01 7.93777183e-02 6.82042480e-01
-1.06924139e-01 2.31310412e-01 3.09750199e-01 -4.89896446e-01
-1.32977569e+00 5.66601574e-01 -3.71499687e-01 -1.73997000e-01
1.00594926e+00 -1.25722957e+00 -1.96971267e-01 -1.42000064e-01
-6.18101597e-01 1.92794368e-01 1.01075041e+00 5.65540552e-01
1.19558597e+00 -1.62889338e+00 -5.10270596e-01 5.77731669e-01
3.29294860e-01 -2.16250092e-01 3.14376175e-01 6.97494924e-01
-7.65948653e-01 4.57426786e-01 -6.48208201e-01 -1.22515047e+00
-1.59424448e+00 2.47068703e-01 2.34483093e-01 3.28757882e-01
-1.11371326e+00 3.07641298e-01 -2.81144679e-01 -4.73851025e-01
-4.54274826e-02 -2.73512810e-01 1.12403445e-01 -1.79655775e-01
4.67655033e-01 6.00014567e-01 1.87021360e-01 -1.22657788e+00
-7.69726694e-01 1.34451210e+00 3.83615226e-01 1.27251610e-01
1.19340241e+00 -4.17170882e-01 2.81432480e-01 5.16516328e-01
1.18192363e+00 -5.18146813e-01 -1.25893712e+00 -1.74521044e-01
-1.23473279e-01 -1.14045513e+00 -3.06748390e-01 -1.52923360e-01
-1.14930642e+00 9.69939768e-01 8.41781855e-01 -2.54393965e-01
1.10429955e+00 -2.32649609e-01 9.87127662e-01 4.02570188e-01
1.37747431e+00 -9.89499032e-01 2.57774293e-01 3.89675587e-01
6.86604798e-01 -1.21469021e+00 1.28864989e-01 -6.68419600e-01
-5.76065361e-01 1.07838273e+00 6.04582965e-01 -3.40223581e-01
4.82171416e-01 2.89842963e-01 -1.63921099e-02 -2.81221956e-01
-3.54854465e-01 -2.78184742e-01 -6.08794130e-02 1.31320822e+00
5.29676378e-02 2.46335760e-01 1.03043795e-01 1.33859888e-01
-3.60870153e-01 2.60899752e-01 3.10283601e-01 1.42756879e+00
-5.12091160e-01 -7.89560795e-01 -8.04638028e-01 3.99240166e-01
1.22415490e-01 4.57845837e-01 -1.99554533e-01 9.19222593e-01
6.92379922e-02 9.58512008e-01 6.85460642e-02 -7.17348158e-01
7.02926457e-01 -1.58043541e-02 6.48224950e-01 -2.53817707e-01
1.12693705e-01 -1.08894847e-01 2.36312337e-02 -9.93840873e-01
-4.41491514e-01 -1.03489661e+00 -1.16736794e+00 -5.85113883e-01
7.67765567e-02 -1.78124905e-01 6.48593903e-01 6.54021680e-01
3.18235308e-01 -7.27491006e-02 4.27484006e-01 -1.29194522e+00
-1.87120244e-01 -4.54900712e-01 -9.40387487e-01 5.34318924e-01
1.53089061e-01 -1.29896402e+00 -1.55665785e-01 2.82705009e-01] | [14.241941452026367, 1.419130563735962] |
c17b62fb-a139-4595-8b13-311d4d3fb07c | revisit-knowledge-distillation-a-teacher-free | 1909.11723 | null | https://arxiv.org/abs/1909.11723v3 | https://arxiv.org/pdf/1909.11723v3.pdf | Revisiting Knowledge Distillation via Label Smoothing Regularization | Knowledge Distillation (KD) aims to distill the knowledge of a cumbersome teacher model into a lightweight student model. Its success is generally attributed to the privileged information on similarities among categories provided by the teacher model, and in this sense, only strong teacher models are deployed to teach weaker students in practice. In this work, we challenge this common belief by following experimental observations: 1) beyond the acknowledgment that the teacher can improve the student, the student can also enhance the teacher significantly by reversing the KD procedure; 2) a poorly-trained teacher with much lower accuracy than the student can still improve the latter significantly. To explain these observations, we provide a theoretical analysis of the relationships between KD and label smoothing regularization. We prove that 1) KD is a type of learned label smoothing regularization and 2) label smoothing regularization provides a virtual teacher model for KD. From these results, we argue that the success of KD is not fully due to the similarity information between categories from teachers, but also to the regularization of soft targets, which is equally or even more important. Based on these analyses, we further propose a novel Teacher-free Knowledge Distillation (Tf-KD) framework, where a student model learns from itself or manuallydesigned regularization distribution. The Tf-KD achieves comparable performance with normal KD from a superior teacher, which is well applied when a stronger teacher model is unavailable. Meanwhile, Tf-KD is generic and can be directly deployed for training deep neural networks. Without any extra computation cost, Tf-KD achieves up to 0.65\% improvement on ImageNet over well-established baseline models, which is superior to label smoothing regularization. | ['Tao Wang', 'Li Yuan', 'Guilin Li', 'Francis E. H. Tay', 'Jiashi Feng'] | 2019-09-25 | revisiting-knowledge-distillation-via-label | http://openaccess.thecvf.com/content_CVPR_2020/html/Yuan_Revisiting_Knowledge_Distillation_via_Label_Smoothing_Regularization_CVPR_2020_paper.html | http://openaccess.thecvf.com/content_CVPR_2020/papers/Yuan_Revisiting_Knowledge_Distillation_via_Label_Smoothing_Regularization_CVPR_2020_paper.pdf | cvpr-2020-6 | ['self-knowledge-distillation'] | ['computer-vision'] | [ 1.07077420e-01 5.83123684e-01 -3.60224634e-01 -3.35414618e-01
-4.31621641e-01 -6.27629459e-01 4.89480436e-01 2.07956150e-01
-4.09711123e-01 6.57469511e-01 6.40953481e-02 -3.14665616e-01
-1.86450988e-01 -7.30561137e-01 -8.78665924e-01 -8.99478555e-01
3.82389337e-01 2.51068681e-01 4.63651747e-01 -2.95030177e-01
-1.36518925e-01 3.94903243e-01 -1.60048127e+00 -1.38232365e-01
1.36691952e+00 9.70758975e-01 2.02514619e-01 8.60154480e-02
-1.52430221e-01 9.77223098e-01 -3.87718052e-01 -4.49011028e-01
1.26429901e-01 -2.46486261e-01 -7.50047386e-01 4.42908481e-02
7.96592355e-01 -3.13971490e-01 -2.80103534e-01 1.30079663e+00
3.42663884e-01 2.38047570e-01 7.10616887e-01 -1.00308645e+00
-7.99869001e-01 9.31271851e-01 -5.80188453e-01 -1.89612374e-01
-1.77411586e-01 -4.23400104e-02 8.68870437e-01 -7.96063483e-01
1.70040682e-01 1.04870224e+00 5.35933971e-01 6.60180867e-01
-1.21499705e+00 -7.79444754e-01 4.82554674e-01 1.47601422e-02
-1.29838383e+00 -7.37908706e-02 6.94094479e-01 -5.73000789e-01
2.44415402e-01 1.22579180e-01 4.18100178e-01 9.17586207e-01
-5.27544022e-01 1.06097162e+00 1.36505663e+00 -4.89998281e-01
7.45706260e-02 8.07444990e-01 4.90899682e-01 8.12553585e-01
1.47940472e-01 1.81394652e-01 -3.37473333e-01 1.48462459e-01
5.30745625e-01 2.60825008e-02 -5.07284045e-01 -5.30912042e-01
-7.64272869e-01 8.07264209e-01 7.13271677e-01 2.75277913e-01
3.93082090e-02 5.17699085e-02 1.65359527e-01 4.77479309e-01
5.80163240e-01 2.97985762e-01 -6.04877353e-01 9.16219223e-03
-8.69889021e-01 -3.97507586e-02 7.09991097e-01 9.02310073e-01
1.09531057e+00 3.02843511e-01 -2.04879165e-01 7.38179624e-01
3.20054621e-01 5.49226642e-01 4.03213233e-01 -6.35173142e-01
-3.26988101e-02 7.13100135e-01 -1.77960575e-01 -7.81565905e-01
-8.00484717e-02 -7.28966415e-01 -7.59533584e-01 2.71395683e-01
4.72780675e-01 -6.51707873e-02 -9.28684950e-01 2.16003966e+00
4.85000253e-01 7.92971849e-01 6.80177584e-02 6.87160969e-01
1.05449796e+00 3.45297635e-01 1.88385710e-01 -3.04877371e-01
1.19895899e+00 -1.02499032e+00 -5.68127692e-01 -7.78313959e-03
8.55497539e-01 -5.47098279e-01 1.22695708e+00 5.71471810e-01
-9.53030467e-01 -6.69184446e-01 -8.44959140e-01 -2.53844950e-02
-2.88318306e-01 1.40517846e-01 6.89428270e-01 7.20059931e-01
-1.02557588e+00 5.49580514e-01 -5.52924216e-01 -8.28070641e-02
5.49541354e-01 3.89731318e-01 -1.24178372e-01 -4.88112755e-02
-1.14087009e+00 9.21072841e-01 3.35472554e-01 -1.50383607e-01
-1.11829662e+00 -1.11122918e+00 -6.55094326e-01 2.41294593e-01
6.77789152e-01 -3.73510808e-01 1.25378501e+00 -1.13710570e+00
-1.79907191e+00 8.23592484e-01 1.09013647e-01 -2.95408547e-01
5.82960963e-01 -4.09016103e-01 -6.45778282e-03 5.97808063e-02
-4.26411554e-02 5.28074265e-01 8.49710464e-01 -1.49536049e+00
-7.19047666e-01 -1.68737575e-01 2.50286013e-01 1.00854523e-01
-6.68818593e-01 -4.49569523e-01 -2.28027165e-01 -6.42238736e-01
-8.32712874e-02 -9.00340497e-01 -1.37226880e-01 -2.19977781e-01
-3.66368920e-01 -7.12331414e-01 8.90111387e-01 -2.33047664e-01
1.28740776e+00 -2.26725268e+00 5.27949892e-02 3.30544353e-01
5.46951771e-01 7.71494269e-01 -1.64005920e-01 -2.31977329e-02
-2.30523318e-01 1.25892803e-01 -8.83641616e-02 -1.90881461e-01
1.87026829e-01 3.92440885e-01 -7.06248164e-01 3.18268269e-01
8.43475014e-02 8.55583727e-01 -1.01527107e+00 -3.00232917e-01
1.43005550e-01 3.99820238e-01 -6.44316792e-01 4.47374344e-01
-2.36598969e-01 5.89258730e-01 -5.01825869e-01 3.37387234e-01
7.21213281e-01 -3.97698343e-01 1.79379538e-01 -2.16079935e-01
-4.77366112e-02 2.44758651e-01 -1.17787004e+00 1.48016870e+00
-3.52650017e-01 3.07793677e-01 1.95611089e-01 -1.52499461e+00
8.69703174e-01 2.91995317e-01 1.79787800e-01 -3.78088087e-01
1.17221445e-01 2.90936530e-01 -1.18132845e-01 -4.42148119e-01
2.07110539e-01 -2.63055176e-01 2.56243855e-01 5.09333909e-01
1.99076384e-01 -1.88152015e-01 -2.59623304e-02 3.66094798e-01
8.49776924e-01 9.24155414e-02 -2.82503315e-04 -4.76334989e-01
5.94495714e-01 -3.39714199e-01 6.90954685e-01 8.39515984e-01
-2.48759314e-01 -7.08642080e-02 2.83531994e-01 -1.30846620e-01
-4.85516131e-01 -1.32171988e+00 -2.63755083e-01 1.60763073e+00
2.73269117e-01 -2.07857385e-01 -7.52289593e-01 -1.23149133e+00
9.33943763e-02 5.30490518e-01 -6.28489375e-01 -3.71745348e-01
-5.72503507e-01 -3.59983802e-01 5.50770044e-01 5.56595504e-01
5.76086879e-01 -7.03492403e-01 -7.15523809e-02 -2.00392492e-02
1.46286651e-01 -8.67195845e-01 -5.00419736e-01 3.52472931e-01
-8.32715452e-01 -1.08283556e+00 -6.26634419e-01 -9.54258680e-01
8.43839765e-01 4.19139624e-01 1.03044224e+00 3.81808281e-01
1.24631017e-01 5.64699054e-01 -2.53270894e-01 -4.07806844e-01
-3.96670431e-01 5.44896349e-02 3.87116492e-01 -7.48790726e-02
5.56470752e-01 -9.24116433e-01 -4.02721405e-01 3.21462899e-01
-9.74748909e-01 4.90585491e-02 7.42991745e-01 8.90828073e-01
3.96394014e-01 2.38347962e-01 9.66278613e-01 -1.19558024e+00
4.24368799e-01 -3.91613543e-01 -7.63972223e-01 4.04896379e-01
-9.59726512e-01 2.63301581e-01 6.56870902e-01 -9.42278802e-01
-1.34356940e+00 -3.22579071e-02 -1.08422555e-01 -4.96413738e-01
-2.54203111e-01 5.45852542e-01 -1.03041627e-01 -2.76125282e-01
5.79722524e-01 1.64807290e-01 -1.02861792e-01 -6.62519991e-01
6.97712243e-01 4.45609182e-01 5.66870332e-01 -1.14134824e+00
1.08089375e+00 4.76436675e-01 -1.97256640e-01 -6.86933696e-01
-1.56441402e+00 -2.97030807e-01 -5.62215149e-01 -6.02906309e-02
6.11172795e-01 -1.03697002e+00 -9.03467834e-01 3.04244727e-01
-7.28873193e-01 -6.80918515e-01 -5.57419002e-01 6.54378593e-01
-1.97340325e-01 4.43442762e-01 -6.59684777e-01 -6.12762868e-01
5.41255251e-02 -1.23480248e+00 3.88961881e-01 5.58264673e-01
3.28583658e-01 -1.35521674e+00 -6.03502356e-02 4.32632804e-01
4.79036182e-01 -1.57175183e-01 1.10449564e+00 -9.07052159e-01
-5.26150107e-01 1.99636251e-01 -3.41011196e-01 6.88277066e-01
9.60074589e-02 -1.54836833e-01 -1.24078476e+00 -4.12166327e-01
8.84890035e-02 -8.66815388e-01 1.05670571e+00 2.02375233e-01
1.18934500e+00 -2.48829886e-01 -1.75712526e-01 5.29640317e-01
1.24453282e+00 -2.74571002e-01 2.60504484e-01 8.87573808e-02
9.97268498e-01 6.50559187e-01 3.08656275e-01 1.11830279e-01
5.36694407e-01 5.97816944e-01 4.53829706e-01 -3.46213579e-01
-2.60158032e-01 -4.76089209e-01 5.04040122e-01 1.28425217e+00
-1.02034524e-01 2.12587148e-01 -6.96460664e-01 5.10367095e-01
-1.74495447e+00 -5.69610894e-01 -1.61445096e-01 2.23341537e+00
1.50199759e+00 3.68264019e-02 -6.84415475e-02 -3.87138575e-02
5.66423118e-01 -1.19705483e-01 -5.11092663e-01 -2.64708608e-01
9.02255848e-02 4.06839013e-01 4.36250508e-01 5.47574520e-01
-1.03596234e+00 1.15555716e+00 5.74827671e+00 1.42817140e+00
-1.21857357e+00 2.44697317e-01 4.64220196e-01 2.81615794e-01
-3.73225302e-01 -2.69819722e-02 -1.14005208e+00 3.37372452e-01
7.46626556e-01 -2.17657223e-01 2.46436045e-01 1.10503256e+00
9.53110214e-03 -3.45900841e-02 -1.27753806e+00 6.37103856e-01
-9.19560269e-02 -9.74491477e-01 2.54333019e-04 1.94333583e-01
1.10848022e+00 -1.70493960e-01 3.93017530e-01 8.66536260e-01
8.17045689e-01 -1.04704368e+00 5.29874086e-01 1.71631157e-01
5.57666063e-01 -7.75132656e-01 5.78195572e-01 6.88602328e-01
-9.94448721e-01 6.82329535e-02 -5.74283898e-01 -1.11149505e-01
-2.47512132e-01 9.40082729e-01 -7.19104290e-01 5.17239928e-01
4.80717987e-01 6.95543587e-01 -5.67177653e-01 7.17906654e-01
-9.44226921e-01 1.18941438e+00 -1.81186438e-01 4.06849504e-01
4.01388973e-01 -2.76551306e-01 2.00791389e-01 1.24058461e+00
6.08469881e-02 2.83328295e-01 6.18251979e-01 9.92849231e-01
-3.51091802e-01 1.29947364e-02 -4.61306632e-01 1.66388378e-01
5.55460811e-01 1.37402797e+00 -3.84502083e-01 -5.02207518e-01
-5.16470790e-01 5.40990531e-01 7.52502322e-01 4.87881720e-01
-7.96870649e-01 -1.17360413e-01 5.74039280e-01 -8.73594731e-02
3.36247772e-01 1.30838424e-01 -9.98830423e-03 -1.22570014e+00
-3.22472900e-01 -8.09688091e-01 2.81432509e-01 -3.87693584e-01
-1.55678916e+00 2.32571915e-01 -1.02467358e-01 -1.02255464e+00
1.43932775e-01 -5.37003040e-01 -5.90278029e-01 7.97253013e-01
-2.07741165e+00 -1.07441115e+00 -2.97721773e-01 7.48634160e-01
9.87141728e-02 -5.44013008e-02 6.76089704e-01 4.34877932e-01
-5.88012874e-01 9.30541813e-01 8.37166682e-02 2.51735359e-01
9.33544338e-01 -1.53353286e+00 -4.28317875e-01 7.13007510e-01
2.06772491e-01 7.03947961e-01 5.16956687e-01 -3.91147703e-01
-1.04508007e+00 -1.13920617e+00 6.66190982e-01 -4.23663050e-01
8.00834477e-01 -1.14085831e-01 -1.37018967e+00 6.88244343e-01
1.53143108e-01 6.83347955e-02 7.98135877e-01 2.88694143e-01
-7.08535969e-01 -2.12113425e-01 -9.53691661e-01 4.15501535e-01
6.46398544e-01 -5.35292149e-01 -8.50840807e-01 3.30800205e-01
8.95865142e-01 -3.29871565e-01 -9.67999220e-01 5.25460839e-01
1.71754062e-01 -9.18389261e-01 9.06588078e-01 -5.65673232e-01
4.35664177e-01 -3.15598965e-01 1.25462890e-01 -1.46959817e+00
-6.59235641e-02 -2.54026860e-01 -2.52124459e-01 1.44880724e+00
1.46719337e-01 -5.75418293e-01 8.73715341e-01 5.39122283e-01
-3.65040988e-01 -9.04327929e-01 -5.97167373e-01 -1.05000317e+00
5.84401369e-01 -3.05978268e-01 3.98883283e-01 1.57416391e+00
-3.53380740e-02 3.37916404e-01 -2.90488422e-01 3.27477157e-01
7.58706868e-01 7.31343850e-02 6.76965415e-01 -1.42813540e+00
-3.61650050e-01 -4.93038416e-01 -4.11134437e-02 -1.63702750e+00
5.23361862e-01 -1.07857823e+00 7.60715827e-02 -1.06621492e+00
2.71453917e-01 -9.48503315e-01 -4.94553417e-01 7.30416000e-01
-4.72666979e-01 2.20729277e-01 -2.36593243e-02 1.46589503e-01
-5.76798499e-01 6.41825259e-01 1.58215153e+00 -5.38175777e-02
-1.95466131e-01 1.47489190e-01 -9.61624682e-01 1.07388985e+00
5.77554226e-01 -4.62776721e-01 -7.70754337e-01 -3.21457803e-01
2.16490924e-01 -4.01256800e-01 2.43309468e-01 -5.61479211e-01
3.05973828e-01 -2.63223916e-01 -1.63332894e-01 -1.64718449e-01
-9.08959471e-03 -8.29986393e-01 -5.41067958e-01 4.20992076e-01
-4.34686899e-01 -8.17369759e-01 6.71650320e-02 5.70599973e-01
-1.43575743e-01 -4.50757384e-01 1.00452125e+00 6.88001439e-02
-6.99502766e-01 2.39714593e-01 -1.10678367e-01 2.94715285e-01
8.53748679e-01 2.90748328e-02 -6.61644816e-01 -3.25915396e-01
-8.29188228e-01 3.82848829e-01 1.33819848e-01 7.31383043e-04
2.96991408e-01 -1.31379080e+00 -6.03608429e-01 9.00432467e-02
-1.11232542e-01 3.45201910e-01 1.18191764e-01 1.22566497e+00
6.45559505e-02 2.18792826e-01 2.67924964e-01 -6.94643199e-01
-1.12647688e+00 7.08200574e-01 1.14830822e-01 -3.38792473e-01
-6.08994186e-01 1.12423265e+00 8.29741895e-01 -6.69086158e-01
6.94123209e-01 -3.91462058e-01 -2.70638078e-01 2.40600910e-02
6.18303418e-01 2.35704109e-01 -1.41699240e-01 -3.62445891e-01
-5.20560741e-02 6.12364352e-01 -3.13463926e-01 4.28887695e-01
1.21377134e+00 -9.93499383e-02 6.02481365e-02 2.48634502e-01
8.06587398e-01 2.79090554e-01 -1.43247092e+00 -6.93496585e-01
7.44126216e-02 -1.95801377e-01 1.10894136e-01 -8.88684690e-01
-1.26774704e+00 1.04514420e+00 3.77758265e-01 2.33376503e-01
1.08417988e+00 2.19576389e-01 4.88676161e-01 4.61467147e-01
1.94865316e-01 -9.67609107e-01 3.86817575e-01 5.10385513e-01
3.38519245e-01 -1.35069072e+00 3.93141583e-02 -6.59401774e-01
-5.46635926e-01 8.19184363e-01 8.17372620e-01 -3.22401859e-02
7.19693422e-01 7.37908185e-02 1.82179958e-01 -6.96707144e-02
-6.10985100e-01 -4.21788305e-01 4.79715705e-01 4.16063666e-01
5.24494886e-01 -1.79906879e-02 -2.44774774e-01 8.68853629e-01
-8.73201042e-02 -9.04118270e-02 4.08386767e-01 5.95820963e-01
-8.60194266e-01 -1.13840270e+00 -1.31425560e-01 1.78580463e-01
-3.11058760e-01 -1.23767681e-01 -1.78259373e-01 7.21929729e-01
4.49170351e-01 9.94283020e-01 -3.66022825e-01 -2.36292318e-01
2.30442047e-01 2.35251593e-03 4.03690428e-01 -7.83153534e-01
-7.39253998e-01 4.58940789e-02 -3.18032682e-01 -3.40995729e-01
-6.11511052e-01 4.92734015e-02 -1.25584507e+00 -3.41766149e-01
-6.19457543e-01 5.59051812e-01 3.50670904e-01 1.08805442e+00
3.94423641e-02 6.13628507e-01 6.31600738e-01 -4.02829498e-01
-9.75760162e-01 -8.44919860e-01 -8.56186628e-01 4.01984811e-01
1.37079403e-01 -9.56514239e-01 -8.13177109e-01 1.88724875e-01] | [9.513175964355469, 3.334073066711426] |
259fe715-bcc8-4145-9540-44c956152cb8 | searching-for-alignment-in-face-recognition | 2102.05447 | null | https://arxiv.org/abs/2102.05447v2 | https://arxiv.org/pdf/2102.05447v2.pdf | Searching for Alignment in Face Recognition | A standard pipeline of current face recognition frameworks consists of four individual steps: locating a face with a rough bounding box and several fiducial landmarks, aligning the face image using a pre-defined template, extracting representations and comparing. Among them, face detection, landmark detection and representation learning have long been studied and a lot of works have been proposed. As an essential step with a significant impact on recognition performance, the alignment step has attracted little attention. In this paper, we first explore and highlight the effects of different alignment templates on face recognition. Then, for the first time, we try to search for the optimal template automatically. We construct a well-defined searching space by decomposing the template searching into the crop size and vertical shift, and propose an efficient method Face Alignment Policy Search (FAPS). Besides, a well-designed benchmark is proposed to evaluate the searched policy. Experiments on our proposed benchmark validate the effectiveness of our method to improve face recognition performance. | ['Zhen Lei', 'Feng Zhou', 'Chenxu Zhao', 'Jianzhu Guo', 'Yunxiao Qin', 'Qiang Meng', 'Xiaqing Xu'] | 2021-02-10 | null | null | null | null | ['face-alignment'] | ['computer-vision'] | [ 1.62752613e-01 -3.96509141e-01 -2.58336216e-01 -5.31919658e-01
-5.67050934e-01 -4.72843021e-01 6.91039979e-01 -4.05537963e-01
-1.33350506e-01 2.73063421e-01 -8.27226043e-02 1.43516943e-01
-2.23734275e-01 -5.22865057e-01 -3.77614319e-01 -7.09294975e-01
4.27113846e-02 2.73963630e-01 5.45449592e-02 2.62284994e-01
5.99454463e-01 1.03536141e+00 -1.69121814e+00 -2.30258733e-01
6.66284502e-01 1.30775952e+00 -3.03992108e-02 -6.66448101e-02
-1.01283021e-01 1.40527233e-01 -6.27803206e-01 -5.41290343e-01
4.29622650e-01 -4.17512298e-01 -6.35468781e-01 3.60447556e-01
5.21112084e-01 -1.42631009e-01 -1.36066629e-02 1.17913675e+00
6.85331047e-01 7.95875490e-02 5.51360905e-01 -1.17145908e+00
-2.10590750e-01 2.19438106e-01 -9.22227025e-01 1.26248688e-01
4.69630420e-01 -4.31494527e-02 4.28372592e-01 -1.18611252e+00
4.75978851e-01 1.32364666e+00 6.29578173e-01 6.19283199e-01
-8.79098177e-01 -9.16108429e-01 1.30482204e-02 2.59486824e-01
-1.64991820e+00 -8.91323626e-01 8.39964211e-01 -3.38452786e-01
4.86109018e-01 3.44288021e-01 3.95686954e-01 8.37744117e-01
-1.02803074e-01 3.75926852e-01 1.15032136e+00 -6.29363716e-01
1.38959279e-02 7.88504109e-02 -8.18944648e-02 9.48295176e-01
1.81277409e-01 6.68464676e-02 -2.88970262e-01 -9.33424011e-02
7.35971570e-01 1.08295016e-01 -4.04107690e-01 -3.41475844e-01
-7.54957080e-01 3.73787940e-01 2.52810717e-01 2.94806182e-01
-1.97590768e-01 -3.51348333e-02 2.34634250e-01 -8.91211107e-02
2.11613327e-01 9.72479507e-02 -2.34700665e-02 1.01789087e-01
-1.10407734e+00 -1.37572810e-01 7.03231215e-01 7.33974636e-01
8.11931849e-01 -1.50563344e-01 -4.47605997e-01 9.87935305e-01
4.47011143e-01 4.65231061e-01 4.47317868e-01 -5.97426891e-01
2.78325677e-01 7.32948601e-01 -2.97806226e-03 -1.38367879e+00
-2.33560488e-01 -1.71978891e-01 -7.26632774e-01 -4.41844314e-02
4.48717833e-01 -3.54416259e-02 -7.67130196e-01 1.37855268e+00
6.33337915e-01 7.74566233e-01 -2.80722797e-01 8.60229969e-01
6.21656358e-01 3.98310751e-01 -6.72865212e-02 -5.38699329e-01
1.48027408e+00 -9.15057123e-01 -6.99873149e-01 -8.18874761e-02
2.42127791e-01 -1.13302398e+00 5.93866944e-01 1.12852030e-01
-8.18949044e-01 -6.43988073e-01 -1.00604928e+00 2.82194018e-01
-2.11784214e-01 6.81142211e-01 4.92128968e-01 1.05039513e+00
-9.39018071e-01 5.89474976e-01 -7.12423682e-01 -4.44433928e-01
5.14841199e-01 5.53783774e-01 -5.45610845e-01 -8.25233236e-02
-7.83517003e-01 8.61230552e-01 2.17337668e-01 3.62976998e-01
-8.09046924e-01 -3.28076899e-01 -7.32508183e-01 4.72023599e-02
6.06039226e-01 -1.88686296e-01 8.69616270e-01 -8.10369551e-01
-1.72747207e+00 9.52928364e-01 -5.33896446e-01 -2.06485242e-01
4.66329932e-01 -6.86042756e-02 -5.86559057e-01 4.90980893e-02
-1.41670689e-01 3.16879094e-01 1.24229693e+00 -1.12684512e+00
-4.70158815e-01 -7.96778917e-01 -2.35583559e-01 -7.13470355e-02
-5.42270720e-01 5.21934748e-01 -9.45514202e-01 -5.34386158e-01
1.48499548e-01 -8.03929567e-01 7.92107955e-02 -1.16904236e-01
-2.19546378e-01 -4.30635571e-01 8.29255462e-01 -6.18643463e-01
1.48906422e+00 -2.21169114e+00 -1.10767737e-01 5.27700543e-01
-7.62759671e-02 5.27026176e-01 -3.19724753e-02 4.03138362e-02
-3.99511047e-02 1.29696131e-01 -8.68091881e-02 -4.50188279e-01
-3.80157866e-02 -1.90628365e-01 -2.23908275e-01 6.77919984e-01
3.16055954e-01 6.05036080e-01 -5.64097047e-01 -7.53221333e-01
2.10107729e-01 6.43253148e-01 -6.25601172e-01 2.09705591e-01
1.83045208e-01 4.01409775e-01 -4.88509655e-01 1.03815758e+00
1.06186235e+00 3.86582166e-02 1.94840059e-01 -4.63780373e-01
-2.04136521e-01 -1.18292227e-01 -1.52494502e+00 1.32548928e+00
-2.61709958e-01 1.71594173e-01 1.40252545e-01 -8.19300115e-01
1.30937076e+00 1.94862515e-01 6.62553906e-01 -4.64296222e-01
3.58467519e-01 2.98363745e-01 -5.22630550e-02 -4.14921522e-01
1.67962387e-01 2.80595005e-01 4.48207885e-01 4.94129658e-01
1.34447649e-01 4.02682960e-01 1.08713977e-01 -4.74711090e-01
7.34166443e-01 2.07075357e-01 4.32651609e-01 -2.40608349e-01
1.21424663e+00 -3.89959812e-01 6.14283204e-01 2.38494128e-01
-5.32451153e-01 5.30324340e-01 3.32348019e-01 -6.17746413e-01
-4.72571820e-01 -3.66253048e-01 -3.12440455e-01 7.84732103e-01
3.49487066e-01 -5.19276142e-01 -1.15577757e+00 -8.49545538e-01
-6.40326142e-02 1.39506325e-01 -7.63228118e-01 -3.80560644e-02
-8.23796034e-01 -6.19132996e-01 4.90254492e-01 2.45559424e-01
8.46465886e-01 -1.22200811e+00 -6.70022130e-01 -1.00732960e-01
2.44393453e-01 -9.78305101e-01 -8.13018024e-01 -2.41942704e-01
-5.72956860e-01 -1.21522284e+00 -7.54155636e-01 -9.23034370e-01
1.02623069e+00 3.62393439e-01 7.74742186e-01 4.65935051e-01
-4.27879721e-01 2.74531126e-01 -1.12530991e-01 -2.56959885e-01
1.23805359e-01 -5.12348022e-03 1.23657286e-01 6.13985419e-01
4.16375458e-01 -1.79607153e-01 -6.59370482e-01 8.62541676e-01
-4.93452042e-01 -3.86958987e-01 6.87568843e-01 7.21193910e-01
6.62781835e-01 -2.86149740e-01 3.57522756e-01 -5.92142522e-01
4.86783385e-01 -1.92315862e-01 -9.06074226e-01 7.11997986e-01
-4.49179679e-01 9.46136415e-02 3.42297524e-01 -2.90891975e-01
-9.51711237e-01 4.92708683e-01 -5.81040718e-02 -6.79409444e-01
-1.68230757e-01 1.59788027e-01 -5.20340741e-01 -5.41867971e-01
4.01105523e-01 2.40170330e-01 2.24827528e-01 -4.77923036e-01
2.24491000e-01 6.26492620e-01 4.33637381e-01 -7.20050752e-01
9.87756252e-01 5.26761830e-01 8.34436789e-02 -8.02143335e-01
-4.58487958e-01 -4.65698093e-01 -8.54914665e-01 -2.89948821e-01
5.20431101e-01 -2.95507550e-01 -1.14502656e+00 2.37554431e-01
-1.23042786e+00 3.55839878e-01 2.81998307e-01 3.51520538e-01
-2.56200701e-01 5.38839281e-01 4.26906124e-02 -9.93134379e-01
-4.99902427e-01 -1.57794225e+00 1.11472535e+00 6.20690823e-01
1.58461422e-01 -5.51040173e-01 -9.38963052e-03 1.66383237e-02
5.13506770e-01 2.76865885e-02 4.35965717e-01 -6.67909980e-01
-6.15722418e-01 -1.88776016e-01 -3.18554461e-01 -7.01377764e-02
3.28700423e-01 1.65311500e-01 -1.05551445e+00 -3.53769720e-01
1.43227026e-01 -1.18280891e-02 7.27775574e-01 1.68938562e-01
1.48886371e+00 -2.25475863e-01 -6.80923700e-01 9.97094929e-01
1.17629397e+00 5.53794563e-01 6.16550982e-01 3.22614521e-01
3.06269616e-01 6.24310911e-01 7.07577467e-01 4.39019173e-01
-1.34184971e-01 9.31913853e-01 3.08768868e-01 1.87958866e-01
-8.49220064e-03 -1.15227424e-01 3.19003969e-01 3.97625387e-01
-1.80993289e-01 2.32274070e-01 -8.32763135e-01 9.49390754e-02
-1.48165584e+00 -1.02683532e+00 6.67969584e-01 2.42542005e+00
6.84394419e-01 -2.18648717e-01 6.52961135e-02 1.71638414e-01
9.42309260e-01 1.46240667e-01 -3.58593464e-01 -4.24114503e-02
2.96197087e-01 4.29422736e-01 2.07092613e-01 4.57348794e-01
-1.29902506e+00 1.11922634e+00 5.82091141e+00 9.81792986e-01
-1.64366150e+00 -1.34785399e-01 7.10011721e-01 3.77896607e-01
2.34237090e-01 3.57296038e-03 -1.39102483e+00 4.73456413e-01
4.22376752e-01 -2.02769607e-01 5.71157098e-01 1.04736316e+00
-6.57706857e-02 1.89787760e-01 -1.10265052e+00 1.44681072e+00
4.49506253e-01 -1.12854505e+00 2.08360609e-02 -1.57791246e-02
4.47293222e-01 -5.84241033e-01 5.76350428e-02 2.15121642e-01
-3.83177608e-01 -1.30816031e+00 3.27312499e-01 5.13645113e-01
7.99013793e-01 -6.74639106e-01 5.48252583e-01 -1.84735190e-02
-1.56040955e+00 -3.51627730e-03 -5.02305508e-01 4.57584858e-01
-2.81175911e-01 2.68311262e-01 -8.94125819e-01 6.29748344e-01
4.69651282e-01 4.91175503e-01 -7.63033271e-01 1.42304146e+00
-1.40015662e-01 3.96215796e-01 -4.30443823e-01 -2.13699322e-02
-9.05956253e-02 -4.37624723e-01 2.94261187e-01 1.07415223e+00
5.66329479e-01 1.05361022e-01 2.74575621e-01 7.18300939e-01
-2.43206456e-01 5.36087275e-01 -5.25686085e-01 1.61936641e-01
7.88691521e-01 1.58231032e+00 -8.96582127e-01 -8.10051057e-03
-3.41339409e-01 7.60348260e-01 2.73324490e-01 1.18052639e-01
-8.75253558e-01 -3.11964631e-01 5.18401086e-01 6.27516657e-02
2.02169657e-01 -5.58870193e-03 -7.55019784e-02 -9.26308453e-01
2.30052575e-01 -1.04577720e+00 3.93519998e-01 -8.53042975e-02
-8.18318009e-01 8.86139333e-01 -1.42045408e-01 -1.26644993e+00
-1.24406613e-01 -6.16392553e-01 -7.33540416e-01 9.72336829e-01
-1.50488019e+00 -9.61365819e-01 -6.85976148e-01 5.92910469e-01
4.58547235e-01 -4.12877738e-01 6.88521266e-01 4.59427983e-01
-1.19290292e+00 1.07680655e+00 -2.10179821e-01 2.86665231e-01
8.05710256e-01 -6.51643157e-01 2.87520736e-01 8.55443060e-01
3.25685173e-01 9.98097003e-01 2.73533076e-01 -4.65213418e-01
-1.63148856e+00 -1.00818086e+00 5.41074932e-01 -2.04321653e-01
1.69283018e-01 -1.64410293e-01 -9.03351128e-01 5.30129433e-01
4.31852601e-02 4.14005280e-01 2.79724538e-01 -1.63394466e-01
-3.46652687e-01 -5.94800293e-01 -1.25040460e+00 6.37631476e-01
1.10974264e+00 -2.91408926e-01 -2.94074267e-01 2.53959507e-01
-1.40192792e-01 -4.80630517e-01 -5.49972117e-01 6.52415395e-01
7.36571193e-01 -8.95755827e-01 1.09294379e+00 -2.31881440e-01
-6.02510609e-02 -4.56898332e-01 1.56640157e-01 -1.04475307e+00
-1.11803792e-01 -9.15635765e-01 8.44640508e-02 1.64700365e+00
8.73830095e-02 -7.23092318e-01 8.61460388e-01 4.46572453e-01
1.77415088e-01 -1.02179360e+00 -9.66760695e-01 -6.61235213e-01
-5.19478142e-01 2.01845560e-02 9.27818060e-01 7.83896685e-01
-2.31882855e-01 1.64317954e-02 -3.28981608e-01 3.39569092e-01
5.64613581e-01 3.20993304e-01 8.70512187e-01 -1.29922903e+00
2.30461001e-01 -7.84544587e-01 -4.82889920e-01 -9.42252815e-01
1.91686630e-01 -8.34765136e-01 -7.54584745e-02 -1.04988801e+00
1.84311658e-01 -5.35399973e-01 -3.36647063e-01 5.68345606e-01
-2.66691059e-01 8.81899446e-02 1.18657313e-01 3.06623191e-01
-4.36509281e-01 5.34573734e-01 1.06220841e+00 -1.03582188e-01
-5.17020710e-02 6.87299147e-02 -4.33847547e-01 6.68489754e-01
7.69839823e-01 -3.54197949e-01 -1.15424126e-01 -2.20274240e-01
-5.32180488e-01 -1.96305454e-01 6.18597306e-02 -1.08691823e+00
4.30166513e-01 -1.80250123e-01 6.75788701e-01 -5.30552387e-01
2.79112339e-01 -1.00037086e+00 1.07566856e-01 4.57471102e-01
8.34719688e-02 3.12517196e-01 2.11476266e-01 2.27835983e-01
-2.55745918e-01 -5.23330092e-01 1.00883365e+00 1.48563877e-01
-7.75031269e-01 4.74674970e-01 4.62959498e-01 -2.95483708e-01
1.42305183e+00 -3.85592908e-01 -1.00035354e-01 9.81487632e-02
-3.31085920e-01 1.17901877e-01 3.27220529e-01 3.94295752e-01
6.25795305e-01 -1.28112650e+00 -5.16042233e-01 5.78566968e-01
-6.72725141e-02 -4.03179318e-01 -2.50929128e-02 8.71478498e-01
-5.11611819e-01 4.79407161e-01 -3.15667301e-01 -7.47246563e-01
-1.45900369e+00 5.42637825e-01 6.12916529e-01 -6.16404079e-02
-1.59779787e-01 8.36063325e-01 -4.04815227e-02 -1.30815625e-01
5.16285121e-01 -4.18027379e-02 -6.20969057e-01 1.89265624e-01
7.21347094e-01 3.21394175e-01 1.82671860e-01 -9.64973390e-01
-6.38239503e-01 1.22637773e+00 -9.88794044e-02 3.42532218e-01
1.01223338e+00 2.61460066e-01 -3.29675585e-01 -3.92744064e-01
1.10070765e+00 1.91794515e-01 -1.08413231e+00 -1.39185637e-01
4.40294385e-01 -9.49448586e-01 -2.03742042e-01 -4.74540174e-01
-1.32666016e+00 8.32478881e-01 9.02342021e-01 -6.33102804e-02
1.13443863e+00 -2.92287499e-01 3.92349780e-01 2.13530868e-01
5.25744438e-01 -7.99644053e-01 1.24010250e-01 3.96965772e-01
1.13478720e+00 -1.19199336e+00 4.63040397e-02 -7.95449853e-01
-1.85739949e-01 1.25432730e+00 8.43461633e-01 8.67911428e-02
7.73943245e-01 3.66580963e-01 5.03551476e-02 -1.36860266e-01
-1.71427235e-01 -2.70580083e-01 5.27673900e-01 3.07664096e-01
4.38467234e-01 -2.35142261e-01 -5.29509783e-01 5.88209569e-01
-3.61212268e-02 1.66222900e-02 -3.45809937e-01 7.36200929e-01
-3.89765918e-01 -1.30875134e+00 -7.15170860e-01 2.43231758e-01
-5.30298948e-01 2.61560023e-01 -3.87717664e-01 6.83718920e-01
2.49670818e-02 7.09115684e-01 -8.33452642e-02 -3.49251568e-01
3.48486930e-01 1.19420379e-01 5.21499872e-01 -4.77115393e-01
-5.02201796e-01 1.21818952e-01 -4.99636501e-01 -6.99327171e-01
-3.28993201e-01 -5.11530340e-01 -9.43719506e-01 2.13690624e-02
-5.05569637e-01 2.48089463e-01 7.01050103e-01 8.80695105e-01
5.46814442e-01 1.95705723e-02 1.00862777e+00 -9.80990350e-01
-7.85175025e-01 -9.45557058e-01 -1.86789140e-01 3.61987323e-01
-9.66596827e-02 -1.03789723e+00 -2.21123800e-01 -2.28552103e-01] | [13.267894744873047, 0.49896031618118286] |
ecc498e2-267b-4e3b-822c-8e01be0c3bdc | towards-continual-adaptation-in-industrial | null | null | https://dl.acm.org/doi/10.1145/3503161.3548232 | https://dl.acm.org/doi/pdf/10.1145/3503161.3548232?casa_token=bIrjY1SPABwAAAAA:JO9_iH9AVEo6zmYgk_axB5lgXWEpHSdy0FfrVYR0UAIhZ0JejPh2U0HcQffNC8ffDw04NG1z07034Q | Towards Continual Adaptation in Industrial Anomaly Detection | Anomaly detection (AD) has gained widespread attention due to its ability to identify defects in industrial scenarios using only normal samples. Although traditional AD methods achieved acceptable performance, they mainly focus on the current set of examples solely, leading to catastrophic forgetting of previously learned tasks when trained on a new one. Due to the limitation of flexibility and the requirements of realistic industrial scenarios, it is urgent to enhance the ability of continual adaptation of AD models. There- fore, this paper proposes a unified framework by incorporating continual learning (CL) to achieve our newly designed task of con- tinual anomaly detection (CAD). Note that, we observe that data augmentation strategy can make AD methods well adapted to su- pervised CL (SCL) via constructing anomaly samples. Based on this, we hence propose a novel method named Distribution of Nor- mal Embeddings (DNE), which utilizes the feature distribution of normal training samples from past tasks. It not only effectively alleviates catastrophic forgetting in CAD but also can be integrated with SCL methods to further improve their performance. Extensive experiments and visualization results on the popular benchmark dataset MVTec AD, have demonstrated advanced performance and the excellent continual adaption ability of our proposed method compared to other AD methods. To the best of our knowledge, we are the first to introduce and tackle the task of CAD. We believe that the proposed task and benchmark will be beneficial to the field of AD. Our code is available in the supplementary material. | ['Feng Zheng', 'Chengjie Wang', 'Jun Liu', 'Bin-Bin Gao', 'Bizhong Xia', 'Jinbao Wang', 'Jiawei Zhan', 'Wujin Li'] | 2022-10-10 | null | null | null | acmmm-2022-10 | ['continual-anomaly-detection'] | ['computer-vision'] | [ 2.36166418e-01 -2.03016013e-01 1.87295750e-01 -1.51960999e-01
-2.73317516e-01 -2.44882628e-01 4.80327934e-01 3.00699383e-01
-1.92402467e-01 5.94590724e-01 -3.68497521e-01 -2.74822205e-01
-1.81566402e-01 -7.10406899e-01 -6.09222949e-01 -5.77096403e-01
-4.44462337e-02 2.22483695e-01 3.70637327e-01 -1.39236897e-01
4.24138874e-01 6.25174999e-01 -1.85920990e+00 7.06225261e-02
1.28199530e+00 1.03812826e+00 2.55592257e-01 2.35382795e-01
-1.69710934e-01 4.84121531e-01 -6.48520589e-01 -3.83896053e-01
1.85138166e-01 -3.05826753e-01 -4.92956370e-01 2.76909053e-01
3.96509767e-01 -3.44359815e-01 -2.13883117e-01 9.66509700e-01
4.12299335e-01 3.22596103e-01 5.66803396e-01 -1.46973312e+00
-9.16924775e-01 1.81966558e-01 -5.36153436e-01 4.73229080e-01
1.79792181e-01 3.13717723e-01 7.22631037e-01 -1.23759305e+00
3.84883881e-01 9.03139770e-01 6.13626420e-01 6.66603565e-01
-9.62512136e-01 -5.74057043e-01 6.23609543e-01 5.18302023e-01
-1.16652215e+00 -2.05797926e-01 1.18290961e+00 -3.86767507e-01
7.85715342e-01 6.41799420e-02 7.41227984e-01 1.26187444e+00
4.10513371e-01 1.03485942e+00 7.21670985e-01 -4.26631898e-01
6.67142034e-01 1.10669337e-01 2.15300947e-01 4.52987671e-01
5.55330515e-01 -4.47551087e-02 -4.77174222e-01 -6.82404712e-02
5.34025967e-01 5.44792652e-01 -1.30233377e-01 -6.81009769e-01
-8.44824851e-01 7.24234462e-01 3.47252339e-01 3.36373657e-01
-6.34646416e-01 -1.74682543e-01 5.88832855e-01 5.05170345e-01
5.58641732e-01 5.65115571e-01 -5.03034353e-01 -1.97424456e-01
-5.16552389e-01 2.17776820e-01 3.63343060e-01 8.36173892e-01
6.18299067e-01 4.48453963e-01 7.23649785e-02 8.42038989e-01
1.73860474e-03 2.26406917e-01 7.31295824e-01 -5.86682856e-01
1.46684557e-01 8.12227190e-01 -1.49245083e-01 -8.63087714e-01
-9.66684595e-02 -6.01475298e-01 -8.47493470e-01 4.66883361e-01
2.08571091e-01 -7.70658627e-03 -1.00412476e+00 1.55692923e+00
3.04891467e-01 4.10241753e-01 4.27984856e-02 5.09460330e-01
7.87897035e-02 4.14505094e-01 2.62411162e-02 -2.44987771e-01
8.53968084e-01 -8.71133506e-01 -9.46563900e-01 -2.79107392e-01
7.00578213e-01 -7.48780310e-01 1.45021212e+00 8.45553637e-01
-7.61696100e-01 -7.14796841e-01 -1.37104523e+00 4.17914659e-01
-5.82547367e-01 7.04013035e-02 6.76125646e-01 4.89614964e-01
-6.48708105e-01 7.48681664e-01 -1.07879174e+00 -4.70803291e-01
6.44258618e-01 1.11954711e-01 -1.52372494e-01 -2.56356776e-01
-9.99505699e-01 6.30617976e-01 3.72353435e-01 2.47763634e-01
-8.54541779e-01 -8.06422472e-01 -8.41610610e-01 -3.01735014e-01
7.05656350e-01 -3.21024537e-01 1.11707425e+00 -6.78076506e-01
-1.20741630e+00 2.09109157e-01 -1.40879527e-02 -6.38712347e-01
5.25939524e-01 -8.90468955e-01 -8.17474663e-01 -1.36909172e-01
4.71524000e-02 1.87324956e-01 1.06495261e+00 -1.22951615e+00
-7.03979611e-01 -4.69906390e-01 -9.27657709e-02 -2.17282742e-01
-9.20288265e-01 -6.34411812e-01 -1.18853040e-01 -9.59902883e-01
1.54624090e-01 -8.90951574e-01 -1.48108587e-01 1.30190939e-01
-1.91831619e-01 -3.94860059e-01 1.43838537e+00 -5.05645931e-01
1.46500921e+00 -2.40324044e+00 -3.02441493e-02 -7.05812946e-02
2.99764294e-02 6.94637656e-01 -6.21407516e-02 3.10395151e-01
-2.54348457e-01 -2.60663092e-01 -5.94001293e-01 -2.64413238e-01
-1.63938612e-01 4.96199638e-01 -4.30506974e-01 3.78315538e-01
6.13332391e-01 6.61532938e-01 -9.26244557e-01 -2.41339713e-01
4.91653711e-01 1.91697031e-01 -6.17413700e-01 2.05028236e-01
-2.30807006e-01 3.04574311e-01 -4.00951207e-01 8.38141859e-01
7.88039029e-01 7.28945509e-02 -2.65542120e-01 -1.17094092e-01
3.01343929e-02 -3.14192086e-01 -1.25422478e+00 1.76034582e+00
-5.79990745e-01 1.59556091e-01 -3.10163558e-01 -1.11487246e+00
1.13649726e+00 2.13131249e-01 4.35983092e-01 -7.73750961e-01
-2.70183295e-01 4.63045418e-01 1.20474607e-01 -4.82748747e-01
3.93448144e-01 9.89705846e-02 2.74424683e-02 2.61393487e-01
8.78683403e-02 1.79008637e-02 2.96032131e-01 1.50575161e-01
1.21230137e+00 2.03931034e-01 2.04024583e-01 -6.54233992e-02
7.94592440e-01 -1.62625939e-01 6.58189893e-01 6.31478786e-01
-3.37324232e-01 4.28964823e-01 3.46748173e-01 -6.80028379e-01
-1.03535223e+00 -1.32940733e+00 -1.74251273e-01 6.89204097e-01
7.43628740e-02 -3.79744142e-01 -3.60728890e-01 -1.16208947e+00
2.17284203e-01 1.19950294e+00 -5.58499813e-01 -6.56509817e-01
-8.64319026e-01 -7.16085017e-01 3.94166373e-02 7.31349587e-01
5.21629333e-01 -1.12863481e+00 -3.52566034e-01 3.15696865e-01
3.22939634e-01 -8.57454777e-01 -1.39353707e-01 2.00368330e-01
-1.15713608e+00 -1.10985172e+00 -6.53527081e-01 -7.34922230e-01
7.53025413e-01 1.98469222e-01 9.15510774e-01 2.99024973e-02
-4.96981204e-01 5.64143598e-01 -5.01284242e-01 -6.25585973e-01
-4.51352209e-01 1.00980990e-01 4.10553902e-01 1.39459014e-01
5.31831086e-01 -7.46828854e-01 -6.67278707e-01 2.75787205e-01
-1.17198050e+00 -4.65214461e-01 9.60313737e-01 8.89567256e-01
5.80890656e-01 2.95477480e-01 1.08834875e+00 -1.03521669e+00
6.96820319e-01 -5.76976240e-01 -3.97886068e-01 3.49323414e-02
-1.15557742e+00 -2.32568520e-04 9.31346297e-01 -5.16806960e-01
-1.18558395e+00 -5.59621155e-02 -3.22944999e-01 -6.70024395e-01
-3.40617031e-01 2.65580148e-01 -1.40952513e-01 2.60495096e-01
5.83781838e-01 2.57769346e-01 1.68288723e-01 -8.43230724e-01
1.99132934e-01 4.99816984e-01 5.96987605e-01 -6.36498749e-01
8.41584086e-01 2.29847088e-01 5.84111735e-03 -7.60765374e-01
-7.93646693e-01 -3.45093131e-01 -8.40075076e-01 -1.39731452e-01
4.28320229e-01 -7.05115736e-01 -2.76480228e-01 6.65017784e-01
-9.35808241e-01 -2.34586447e-02 -7.38716841e-01 3.43810230e-01
-3.80736649e-01 6.85103714e-01 -4.38269228e-01 -9.80132759e-01
-1.46576777e-01 -7.94739664e-01 7.39908695e-01 -1.11081600e-01
-1.98396519e-01 -1.01873720e+00 -9.07559041e-03 -8.91662315e-02
4.40136194e-01 3.26272070e-01 1.02886987e+00 -9.58084524e-01
-3.71190131e-01 -4.81490791e-01 1.42587632e-01 9.61954117e-01
5.60505271e-01 -1.70809805e-01 -9.94325578e-01 -5.42537928e-01
1.98692739e-01 -1.50847673e-01 9.81874824e-01 2.22802646e-02
1.47324157e+00 -8.41461122e-02 -2.33813241e-01 6.02618270e-02
1.35054517e+00 4.65351224e-01 7.61110008e-01 4.95967388e-01
7.20508814e-01 3.26630145e-01 1.25179279e+00 6.09957516e-01
-2.10689038e-01 5.34129679e-01 6.63214445e-01 -1.02701917e-01
-2.58375078e-01 -1.58373937e-01 3.89499784e-01 9.94347274e-01
-1.61858983e-02 -1.57205641e-01 -7.38750696e-01 7.67400980e-01
-1.85688627e+00 -6.82951272e-01 -7.26882592e-02 2.22670078e+00
4.82865930e-01 5.15816927e-01 -9.89441425e-02 6.75487459e-01
5.98190129e-01 6.09927736e-02 -8.46219301e-01 -4.75894600e-01
1.20690644e-01 2.76690811e-01 -5.54271936e-02 6.64799213e-02
-1.14962113e+00 5.79904675e-01 5.29322100e+00 9.55395401e-01
-8.55322719e-01 1.86046567e-02 3.70403051e-01 5.38339168e-02
2.54578758e-02 -2.87501603e-01 -6.55683577e-01 7.07468510e-01
8.61455739e-01 -1.16091788e-01 1.10754766e-01 1.24845445e+00
-7.12499395e-02 1.20769935e-02 -1.33989954e+00 6.39643013e-01
1.19870208e-01 -7.84511805e-01 3.29100281e-01 -6.71021864e-02
6.87319398e-01 -4.95580852e-01 3.40098351e-01 5.60350895e-01
-7.45025948e-02 -5.42541385e-01 3.33718717e-01 4.50253487e-01
4.87111747e-01 -8.84234726e-01 1.07069159e+00 1.69415921e-01
-1.08604228e+00 -5.04597425e-01 -4.08655882e-01 -2.40829512e-02
1.51159152e-01 9.44123864e-01 -9.26897526e-01 7.89479613e-01
6.50210261e-01 9.48735774e-01 -9.02629077e-01 1.08841002e+00
-5.56615330e-02 5.99595249e-01 -1.16828978e-02 2.48235002e-01
1.08530363e-02 1.78260319e-02 6.73468888e-01 7.05726266e-01
8.07487011e-01 -5.84250450e-01 8.06381851e-02 7.83416688e-01
3.24168891e-01 2.19013784e-02 -9.18818712e-01 -1.15235962e-01
4.49077636e-01 9.57731545e-01 -4.95213568e-01 -1.64602414e-01
-5.54839849e-01 1.17164159e+00 1.59344137e-01 1.73466012e-01
-6.57274604e-01 -5.36685884e-01 7.46840119e-01 2.50914514e-01
5.01527786e-01 -2.83310592e-01 -3.43793780e-01 -8.85875523e-01
3.62904787e-01 -7.22782195e-01 6.17788851e-01 -4.07896906e-01
-1.74473679e+00 6.10653281e-01 -1.20520264e-01 -1.59862447e+00
-2.18767568e-01 -6.49921417e-01 -6.76590145e-01 3.10232699e-01
-1.41985309e+00 -1.01510942e+00 -2.64542818e-01 5.40572286e-01
1.02047431e+00 -5.56976855e-01 8.13879907e-01 5.23085237e-01
-7.95972824e-01 6.61522150e-01 1.55867845e-01 -3.40953350e-01
8.47394109e-01 -1.35784948e+00 4.34302926e-01 1.08575797e+00
-5.72026670e-02 5.62478364e-01 8.01658809e-01 -9.17096019e-01
-1.22931015e+00 -1.38356256e+00 3.59524012e-01 -4.56069648e-01
7.08999932e-01 -2.76501477e-01 -1.38838267e+00 6.51561737e-01
-8.71707499e-02 2.23941263e-02 5.00510693e-01 6.16562478e-02
2.26161368e-02 -4.65238541e-01 -1.14139605e+00 5.69712520e-01
9.96525109e-01 -9.87445638e-02 -7.24966168e-01 2.74261922e-01
9.11208570e-01 -8.41084644e-02 -7.85232186e-01 7.59812236e-01
9.56127271e-02 -9.05312777e-01 9.97461736e-01 -3.51377130e-01
4.11704689e-01 -3.13098282e-01 7.76998000e-03 -1.44166124e+00
-1.28461853e-01 -2.75162965e-01 -6.06792152e-01 1.49120617e+00
1.22538313e-01 -9.43407536e-01 6.06445074e-01 3.46379817e-01
-5.91006577e-01 -9.70381677e-01 -9.22365963e-01 -1.13015115e+00
4.25320230e-02 -5.79582870e-01 6.75974905e-01 7.95185506e-01
-3.70927185e-01 1.00731269e-01 -4.09938186e-01 1.78046316e-01
5.77075660e-01 -9.06809419e-02 7.09216118e-01 -1.47443688e+00
-9.39462408e-02 -1.36268428e-02 -6.12007678e-01 -7.87823200e-01
-7.08106682e-02 -5.95988393e-01 -1.25696421e-01 -1.13086987e+00
-2.72977293e-01 -5.62749863e-01 -7.60476410e-01 4.23711389e-01
-2.65424579e-01 2.07511827e-01 -1.55860722e-01 1.60699069e-01
-6.50667012e-01 9.56749201e-01 1.28764927e+00 -6.76136836e-02
-3.60441476e-01 1.81124434e-01 -4.99998987e-01 6.78901434e-01
9.87004757e-01 -3.41528654e-01 -7.01815307e-01 -1.00426167e-01
-1.26516715e-01 -6.18099570e-01 3.32864702e-01 -1.37060797e+00
6.42035007e-02 1.20806366e-01 4.99832451e-01 -6.44048214e-01
2.54635781e-01 -8.85856152e-01 -1.93206415e-01 5.98709166e-01
1.08196825e-01 3.49170595e-01 4.74619061e-01 1.03409123e+00
-3.46525431e-01 -2.98965603e-01 5.47497332e-01 -8.44249427e-02
-1.13161087e+00 2.42851362e-01 -3.78951281e-01 -1.76824018e-01
1.34745777e+00 -2.67991453e-01 -1.18753664e-01 -1.01457685e-02
-7.74597347e-01 1.37208223e-01 5.12157023e-01 7.21967340e-01
8.80381763e-01 -1.58277786e+00 -3.49377066e-01 6.22928619e-01
3.15027744e-01 1.35410637e-01 4.73048478e-01 8.29433799e-01
-2.84599602e-01 2.53824770e-01 -3.70112002e-01 -5.92172146e-01
-9.89618897e-01 1.00182641e+00 -1.19640708e-01 -1.52103469e-01
-9.39361572e-01 4.44888949e-01 2.48781219e-01 -2.60940254e-01
2.63658911e-01 -1.71930000e-01 -1.20119840e-01 -9.48625058e-02
3.97331417e-01 7.20783174e-01 4.66284007e-01 -7.60645419e-02
-3.56203467e-01 4.00559455e-01 -5.78114033e-01 3.91828358e-01
1.46781218e+00 -8.98184255e-02 8.05568844e-02 8.25834394e-01
7.91740417e-01 -3.45980488e-02 -1.25128949e+00 -1.72594845e-01
1.83911055e-01 -6.78293228e-01 -1.21231832e-01 -7.52415597e-01
-1.08534503e+00 9.54051614e-01 9.77522969e-01 2.05226645e-01
1.25168824e+00 -2.01778919e-01 1.06726420e+00 4.24924642e-01
3.43358606e-01 -1.17428398e+00 7.36558437e-01 2.76389241e-01
9.54587579e-01 -1.19047594e+00 3.98328044e-02 -3.47885698e-01
-6.25905514e-01 1.16640472e+00 1.07438695e+00 -2.16959193e-01
6.84559643e-01 2.38940343e-01 -9.72232968e-02 -2.39140153e-01
-6.27424777e-01 1.03343353e-01 1.28855705e-01 7.54449904e-01
2.31448561e-01 -3.79280031e-01 -3.13030124e-01 5.76063931e-01
2.12183192e-01 1.16374701e-01 4.67813164e-01 1.25416601e+00
-3.86187404e-01 -1.30559027e+00 -1.63742140e-01 6.12025678e-01
-2.86364585e-01 2.86294520e-01 -3.18608060e-02 8.99915040e-01
2.86276698e-01 8.20455968e-01 1.47125468e-01 -4.08894479e-01
6.08600378e-01 3.21130335e-01 2.15303555e-01 -5.65248847e-01
-3.47559899e-02 -2.50823766e-01 -2.28055015e-01 -5.09095490e-01
-5.40407933e-02 -7.76039779e-01 -9.91715372e-01 -1.36157125e-01
-3.79175216e-01 9.77647901e-02 4.34400171e-01 7.75569439e-01
5.52268803e-01 1.00320804e+00 5.75083137e-01 -5.92484951e-01
-7.23801196e-01 -1.00181782e+00 -7.85582125e-01 6.20778084e-01
2.81488031e-01 -1.10941529e+00 -6.03526115e-01 -4.68595251e-02] | [7.5176615715026855, 2.13736891746521] |
a02baec8-597b-4f7f-8ba8-96566312380a | cross-paced-representation-learning-with | 1803.01504 | null | http://arxiv.org/abs/1803.01504v1 | http://arxiv.org/pdf/1803.01504v1.pdf | Cross-Paced Representation Learning with Partial Curricula for Sketch-based Image Retrieval | In this paper we address the problem of learning robust cross-domain
representations for sketch-based image retrieval (SBIR). While most SBIR
approaches focus on extracting low- and mid-level descriptors for direct
feature matching, recent works have shown the benefit of learning coupled
feature representations to describe data from two related sources. However,
cross-domain representation learning methods are typically cast into non-convex
minimization problems that are difficult to optimize, leading to unsatisfactory
performance. Inspired by self-paced learning, a learning methodology designed
to overcome convergence issues related to local optima by exploiting the
samples in a meaningful order (i.e. easy to hard), we introduce the cross-paced
partial curriculum learning (CPPCL) framework. Compared with existing
self-paced learning methods which only consider a single modality and cannot
deal with prior knowledge, CPPCL is specifically designed to assess the
learning pace by jointly handling data from dual sources and modality-specific
prior information provided in the form of partial curricula. Additionally,
thanks to the learned dictionaries, we demonstrate that the proposed CPPCL
embeds robust coupled representations for SBIR. Our approach is extensively
evaluated on four publicly available datasets (i.e. CUFS, Flickr15K, QueenMary
SBIR and TU-Berlin Extension datasets), showing superior performance over
competing SBIR methods. | ['Nicu Sebe', 'Xavier Alameda-Pineda', 'Jingkuan Song', 'Elisa Ricci', 'Dan Xu'] | 2018-03-05 | null | null | null | null | ['sketch-based-image-retrieval'] | ['computer-vision'] | [ 9.07828733e-02 -5.40336311e-01 -6.20456755e-01 -3.46891433e-02
-1.35162342e+00 -6.94734216e-01 9.57456172e-01 2.94272035e-01
-4.35765356e-01 5.35034060e-01 2.78036028e-01 1.31274641e-01
-7.50150919e-01 -6.46496952e-01 -7.76240528e-01 -5.77065825e-01
1.25429645e-01 3.88120502e-01 -1.14214011e-01 -3.40999424e-01
3.40652645e-01 3.60497743e-01 -1.89511395e+00 2.70368218e-01
9.20406163e-01 1.01918972e+00 2.70038247e-01 3.88073355e-01
-3.42130154e-01 5.54840744e-01 -3.69530439e-01 -2.83050269e-01
2.76187867e-01 -2.10293621e-01 -7.68398523e-01 -1.11703254e-01
7.71482944e-01 -8.70349631e-02 -3.73487204e-01 8.49661887e-01
6.06607676e-01 3.20326477e-01 7.74368823e-01 -1.25407648e+00
-8.58443975e-01 1.73959017e-01 -5.44354498e-01 1.14076115e-01
5.01713634e-01 -3.54998887e-01 1.23878741e+00 -1.11955655e+00
9.70651448e-01 1.23388457e+00 6.20854437e-01 5.18306375e-01
-1.39294565e+00 -7.96014249e-01 4.16313618e-01 2.92222321e-01
-1.55530441e+00 -2.36926332e-01 1.14134800e+00 -3.70025456e-01
3.53107929e-01 1.74975783e-01 3.41087520e-01 1.44500196e+00
-2.75670141e-01 1.27184391e+00 1.37787497e+00 -5.00739634e-01
1.28246486e-01 2.44824931e-01 2.11914420e-01 4.99716222e-01
-4.12132442e-02 2.57509649e-01 -9.30540562e-01 -2.45520592e-01
8.54885042e-01 2.27356017e-01 -1.49147481e-01 -9.94741261e-01
-1.38934577e+00 8.96886587e-01 4.57504779e-01 5.19171476e-01
-2.45092437e-01 -5.87367602e-02 4.65487897e-01 5.85101247e-01
5.05486369e-01 3.43930155e-01 -3.12688053e-01 -2.08397284e-02
-1.08813727e+00 4.14599001e-01 5.45718551e-01 1.08339357e+00
8.37734163e-01 -2.17867434e-01 -3.10763180e-01 1.20687485e+00
1.92287281e-01 4.41570133e-01 5.58930218e-01 -5.62987268e-01
4.56641883e-01 5.02048552e-01 -8.88261423e-02 -9.98513520e-01
5.14758117e-02 -3.40692610e-01 -8.16845655e-01 5.11813760e-02
4.29920554e-01 3.73105556e-01 -8.35378766e-01 1.73647094e+00
1.97001666e-01 3.26940745e-01 1.58804059e-01 9.88043010e-01
1.18538427e+00 5.06734490e-01 2.33068556e-01 -2.88407726e-04
1.21814787e+00 -9.86339033e-01 -6.65589690e-01 1.13847874e-01
3.10983092e-01 -9.04910862e-01 1.12083328e+00 6.50556266e-01
-9.57740724e-01 -8.12844157e-01 -1.00447249e+00 -8.84538144e-02
-6.74858809e-01 1.70486197e-01 4.89902139e-01 4.03747976e-01
-9.37032640e-01 5.58427453e-01 -3.60153407e-01 -1.93217903e-01
5.08437276e-01 1.26388654e-01 -6.06803536e-01 -4.92705077e-01
-1.09415567e+00 7.18867302e-01 1.93235457e-01 -3.02979827e-01
-8.87559354e-01 -1.13305259e+00 -8.05019259e-01 -7.33417645e-02
4.43287522e-01 -4.56839025e-01 7.21092880e-01 -1.06779897e+00
-1.37520790e+00 9.02008176e-01 4.58123833e-02 -1.66131034e-01
6.15269244e-01 -2.17939094e-01 -2.83261836e-01 1.90384492e-01
5.43209203e-02 7.82064736e-01 1.09213793e+00 -1.35433412e+00
-4.23480779e-01 -3.45091522e-01 1.35260731e-01 3.27479362e-01
-8.07971895e-01 -1.12522401e-01 -8.04175377e-01 -9.77592945e-01
-1.35942511e-02 -8.67562711e-01 -8.58154893e-02 3.00087303e-01
2.25721240e-01 -4.44186807e-01 9.63412762e-01 -5.00822663e-01
1.14388454e+00 -2.00832915e+00 4.74770695e-01 1.23218484e-01
-7.30979722e-03 4.66730952e-01 -5.04777133e-01 4.86980528e-01
-9.40307304e-02 -1.52492017e-01 -2.60966755e-02 -5.07037342e-01
1.30820051e-01 2.94396937e-01 -3.10707152e-01 3.31802428e-01
2.99468786e-01 9.11457479e-01 -1.23605549e+00 -7.20661461e-01
3.63980591e-01 8.40220213e-01 -5.77893972e-01 3.05047423e-01
-5.77286398e-03 3.65163863e-01 -4.29196686e-01 8.13696384e-01
6.32094443e-01 -1.61372915e-01 1.14637874e-01 -2.60398835e-01
-9.02426615e-03 -2.41508707e-02 -1.45903528e+00 2.35819888e+00
-8.19199264e-01 4.44289893e-01 5.77427633e-03 -1.27351010e+00
1.07792985e+00 2.70258158e-01 6.23148322e-01 -1.05268204e+00
-3.16578895e-01 2.71667123e-01 -7.72037208e-01 -3.48316394e-02
6.37299478e-01 -6.76786229e-02 -1.99549541e-01 2.96065122e-01
3.95732403e-01 1.30314603e-02 1.59983292e-01 2.29457542e-01
8.11944902e-01 4.61053133e-01 2.15612948e-01 -3.51567149e-01
7.76701093e-01 -8.40528756e-02 3.32528234e-01 7.38250077e-01
-1.54189259e-01 7.95458019e-01 3.59710469e-03 -5.13102412e-01
-7.77808487e-01 -1.18926346e+00 -3.01865071e-01 1.30120671e+00
4.03975874e-01 -5.55922627e-01 -2.27299005e-01 -8.47996414e-01
1.79127991e-01 1.41433194e-01 -7.09568024e-01 -1.66302010e-01
-4.25888270e-01 -3.00993204e-01 2.37563640e-01 5.14333248e-01
3.31731230e-01 -9.41186190e-01 -3.14951777e-01 2.58256286e-01
-1.39378488e-01 -9.81653154e-01 -6.22042358e-01 5.43564074e-02
-6.72817230e-01 -9.66859818e-01 -1.21837258e+00 -8.70759606e-01
5.06293416e-01 6.11586213e-01 1.27684593e+00 4.74834144e-02
-6.71188772e-01 1.03467202e+00 -4.78108764e-01 -1.82941988e-01
-9.22084749e-02 1.02671608e-01 -3.13922316e-02 5.79812601e-02
2.96311885e-01 -4.09852862e-01 -5.91213346e-01 3.33215564e-01
-1.04816771e+00 -9.24692973e-02 6.48528218e-01 1.30704963e+00
8.84266734e-01 -2.14837372e-01 6.60125852e-01 -6.82384610e-01
6.09543145e-01 -5.12365282e-01 -5.42643964e-01 5.12974322e-01
-6.91592038e-01 3.05593908e-01 4.20906126e-01 -8.62032473e-01
-1.09502661e+00 1.49509564e-01 2.53212541e-01 -1.00242937e+00
-1.97187245e-01 5.01749396e-01 8.65002647e-02 -3.09933752e-01
5.62458038e-01 4.97863829e-01 5.86849898e-02 -4.99540001e-01
5.10611296e-01 3.63457382e-01 3.62387776e-01 -1.12050927e+00
9.76612866e-01 3.43674868e-01 -8.03113356e-02 -7.49120474e-01
-9.07062352e-01 -9.46879685e-01 -6.60886645e-01 -2.87435442e-01
6.23223364e-01 -1.18047965e+00 -4.43550736e-01 1.44545704e-01
-7.81078041e-01 8.03747848e-02 -2.97981799e-01 3.05919409e-01
-7.37763166e-01 5.42930841e-01 -2.25244895e-01 -7.82770336e-01
-3.56031001e-01 -1.06656837e+00 1.26588476e+00 1.63098350e-01
4.50534597e-02 -1.22450733e+00 4.23585266e-01 3.39626908e-01
4.97460485e-01 2.40252703e-01 5.91580689e-01 -4.98066932e-01
-5.14178276e-01 -9.61797684e-02 -3.11635941e-01 2.47842297e-01
1.48127139e-01 -3.73124480e-01 -1.02084148e+00 -5.78112662e-01
-5.32858908e-01 -7.82885134e-01 9.75332439e-01 -2.83615068e-02
1.26386452e+00 -1.82033598e-01 -2.44277418e-01 5.59385955e-01
1.72885048e+00 -2.92311132e-01 4.40281987e-01 4.19911504e-01
5.57675421e-01 6.52791023e-01 9.78752971e-01 4.19423193e-01
4.97829586e-01 1.02055418e+00 1.35623679e-01 -2.43730590e-01
-4.37018603e-01 -5.28380096e-01 2.34823436e-01 6.78722799e-01
-1.17616877e-02 2.83224553e-01 -7.40807891e-01 8.54754388e-01
-1.89643872e+00 -9.56223071e-01 2.30227411e-01 2.26102519e+00
1.00325382e+00 -2.96717197e-01 2.23012045e-01 2.25731283e-01
4.12473321e-01 2.16202945e-01 -2.35718444e-01 -1.25202715e-01
-8.23135674e-02 5.91498911e-01 3.00980419e-01 3.09657067e-01
-1.31466961e+00 9.73718226e-01 5.27890301e+00 1.18207526e+00
-9.88120735e-01 1.39747009e-01 2.08933860e-01 6.35410547e-02
-5.22549510e-01 -2.25170270e-01 -6.57849729e-01 2.95275718e-01
5.62540412e-01 -1.16449408e-01 4.96379703e-01 8.59645665e-01
-3.80488664e-01 2.75998205e-01 -1.15981364e+00 1.37710989e+00
2.41300046e-01 -1.35662484e+00 1.95438251e-01 -9.28580761e-02
9.66307044e-01 -2.59000748e-01 4.14562017e-01 6.32088542e-01
2.41705060e-01 -9.33736920e-01 6.14101768e-01 6.32553518e-01
9.15450156e-01 -8.44574869e-01 3.45566899e-01 1.80853024e-01
-1.49063516e+00 -1.11717209e-01 -4.05411094e-01 3.97224158e-01
-4.10272837e-01 1.89019009e-01 -4.08978254e-01 1.03933716e+00
8.28545451e-01 1.02589607e+00 -6.45588100e-01 9.67063308e-01
1.14038862e-01 1.74763769e-01 1.44559070e-02 1.73129708e-01
3.09828550e-01 -1.70660660e-01 4.19919997e-01 1.25161302e+00
2.13442311e-01 -1.72544345e-01 2.95691669e-01 8.49260986e-01
5.79964481e-02 2.38760754e-01 -7.26388395e-01 1.86601877e-02
2.89617121e-01 1.33884823e+00 -3.44581395e-01 -2.13569611e-01
-4.98065591e-01 1.01665163e+00 5.61483026e-01 3.18425894e-01
-4.77119714e-01 -1.28339410e-01 8.15960646e-01 -2.79542059e-02
5.29526114e-01 -2.40971759e-01 6.13079127e-03 -1.27217555e+00
-3.45186107e-02 -1.09979999e+00 7.05900967e-01 -3.77241939e-01
-1.74666703e+00 4.08535391e-01 1.42434478e-01 -1.60917139e+00
-1.12383306e-01 -5.86006582e-01 -2.87471533e-01 8.57869685e-01
-2.31050587e+00 -1.59879637e+00 -3.98550391e-01 1.12260008e+00
6.65018380e-01 -2.56967843e-01 1.03649127e+00 5.42177200e-01
-2.96034694e-01 9.80473876e-01 2.65749574e-01 -2.28383634e-02
1.13775134e+00 -1.30160165e+00 -1.81731701e-01 3.15089583e-01
5.57372630e-01 6.63667738e-01 3.54331166e-01 -2.79148281e-01
-1.80384302e+00 -8.76274586e-01 7.39407957e-01 -3.96619529e-01
5.97995937e-01 -3.82720172e-01 -1.01385605e+00 1.65135041e-01
-4.35086824e-02 5.55986166e-01 6.10178292e-01 2.85198569e-01
-8.56198847e-01 -3.02372843e-01 -1.07168496e+00 2.65326500e-01
1.11888802e+00 -7.66656697e-01 -6.08849227e-01 2.78490454e-01
2.26116538e-01 -3.64998579e-01 -1.10073292e+00 2.65307307e-01
7.28213370e-01 -6.63543522e-01 1.54493737e+00 -6.75674200e-01
4.38149184e-01 -1.71752736e-01 -2.90983856e-01 -1.13602793e+00
-2.00687781e-01 -1.55883878e-01 -1.07354522e-01 1.32626569e+00
-1.54242054e-01 -1.72557488e-01 7.82549500e-01 1.56293705e-01
1.73611149e-01 -5.87298214e-01 -9.54385459e-01 -1.01789820e+00
3.09831560e-01 -2.92707443e-01 4.12013531e-01 1.17896783e+00
-2.05378801e-01 8.78587514e-02 -5.96037209e-01 7.32480213e-02
8.74593139e-01 4.81224120e-01 8.97396326e-01 -1.53093565e+00
-1.36461571e-01 -6.00593150e-01 -3.74345541e-01 -9.33469176e-01
4.28789675e-01 -1.07360101e+00 -4.50983942e-02 -1.25659072e+00
3.76263678e-01 -9.21485007e-01 -7.99684823e-01 3.57804835e-01
-2.73591846e-01 3.55716497e-01 3.09527785e-01 1.40614659e-01
-9.43291187e-01 7.27252483e-01 1.17272794e+00 -4.11479205e-01
-2.32162829e-02 -2.30925053e-01 -5.41716397e-01 2.36906871e-01
3.43945533e-01 -2.97511637e-01 -5.66504359e-01 -2.90575594e-01
-7.67243886e-03 8.97291526e-02 5.48952520e-01 -8.44756067e-01
4.27598327e-01 -1.33887157e-01 3.94057602e-01 -5.92364430e-01
3.79759341e-01 -9.68061447e-01 -4.80119176e-02 1.38786733e-01
-5.86724579e-01 -2.04309314e-01 2.34346509e-01 8.65417838e-01
-5.45566142e-01 -1.13047726e-01 6.07310236e-01 -1.98827088e-01
-9.35577154e-01 4.66601461e-01 6.12413995e-02 8.72998461e-02
8.61123025e-01 -9.42606628e-02 -2.34973300e-02 -3.17833155e-01
-5.73727548e-01 2.57713646e-01 3.74773890e-01 8.83426845e-01
8.39918017e-01 -1.77066839e+00 -6.37263119e-01 2.23232150e-01
6.72018647e-01 -2.17253074e-01 4.08701152e-01 5.13888061e-01
1.03495970e-01 4.74251956e-01 -4.26885873e-01 -7.92054355e-01
-1.32422185e+00 7.95148730e-01 3.08595747e-02 -6.51350796e-01
-5.62274575e-01 8.16900313e-01 1.83625057e-01 -7.46239781e-01
5.07726192e-01 1.99730530e-01 -3.33676428e-01 4.34945762e-01
5.32852292e-01 1.82273194e-01 1.60962626e-01 -5.47090650e-01
-3.69573593e-01 9.10136819e-01 -2.35352814e-01 4.33673039e-02
1.42573607e+00 7.09909871e-02 2.99803346e-01 3.28196496e-01
1.41417325e+00 -1.68659180e-01 -1.21800625e+00 -8.32417071e-01
2.39174947e-01 -6.93376422e-01 2.92219579e-01 -7.79898524e-01
-1.03179014e+00 7.83634186e-01 8.47175360e-01 -2.85688132e-01
1.14862108e+00 -2.87511609e-02 5.66122591e-01 3.14278036e-01
5.83429396e-01 -1.18414557e+00 6.53413653e-01 2.87259638e-01
1.12719858e+00 -1.58367300e+00 5.54906577e-02 -7.85973147e-02
-6.29357576e-01 1.15721560e+00 5.09539068e-01 -1.47736967e-01
6.48725748e-01 -1.44267008e-01 -5.16031086e-02 -3.42272930e-02
-6.69591725e-01 -3.39445353e-01 8.04894686e-01 6.12462938e-01
3.59591126e-01 -1.24828435e-01 -2.67892033e-01 5.41595340e-01
1.79895356e-01 -4.46068384e-02 4.98395786e-02 1.03368890e+00
4.27815504e-02 -1.59571362e+00 -3.96280617e-01 1.47179425e-01
-1.64658487e-01 7.76624605e-02 -3.47136557e-01 8.53353143e-01
6.61460159e-04 5.97450197e-01 -7.15369582e-02 -1.22709289e-01
4.48672622e-01 -1.63464211e-02 6.91518664e-01 -3.99884909e-01
-6.77379727e-01 -2.73518674e-02 -3.05074602e-01 -6.48085237e-01
-6.83834434e-01 -6.34711921e-01 -6.95847213e-01 -1.15814363e-03
-1.94776043e-01 3.61808017e-02 6.67538047e-01 7.07445681e-01
3.11702520e-01 3.69960040e-01 6.57065511e-01 -8.93186927e-01
-6.05846286e-01 -5.94100237e-01 -3.97729278e-01 7.12290823e-01
4.42705989e-01 -1.08277404e+00 -1.27750486e-01 -1.69772342e-01] | [11.456860542297363, 0.8131424188613892] |
c8082f6b-3ffb-484b-94b3-f53acf0ed79f | client-driven-animated-gif-generation | null | null | https://link.springer.com/article/10.1007/s11042-020-10236-6#Abs1 | https://link.springer.com/article/10.1007/s11042-020-10236-6 | Client-driven Animated GIF Generation Framework Using an Acoustic Feature | This paper proposes a novel, lightweight method to generate animated graphical interchange format images (GIFs) using the computational resources of a client device. The method analyzes an acoustic feature from the climax section of an audio file to estimate the timestamp corresponding to the maximum pitch. Further, it processes a small video segment to generate the GIF instead of processing the entire video. This makes the proposed method computationally efficient, unlike baseline approaches that use entire videos to create GIFs. The proposed method retrieves and uses the audio file and video segment so that communication and storage efficiencies are improved in the GIF generation process. Experiments on a set of 16 videos show that the proposed approach is 3.76 times more computationally efficient than a baseline method on an Nvidia Jetson TX2. Additionally, in a qualitative evaluation, the GIFs generated using the proposed method received higher overall ratings compared to those generated by the baseline method. To the best of our knowledge, this is the first technique that uses an acoustic feature in the GIF generation process. | ['Eun-Seok Ryu', 'Jaehyoun Kim', 'Sangsoon Lee', 'Ghulam Mujtaba'] | 2021-02-12 | null | null | null | null | ['animated-gif-generation', 'music-genre-recognition'] | ['computer-vision', 'music'] | [ 2.19008625e-01 -3.27921748e-01 1.55963823e-01 -2.39808679e-01
-6.73759103e-01 -6.22957468e-01 3.53427142e-01 6.29506037e-02
-3.15379649e-01 2.97532946e-01 1.77561715e-02 -1.89562276e-01
2.32822672e-01 -6.34729624e-01 -6.34954810e-01 -4.02609289e-01
-9.67803448e-02 -1.31122708e-01 4.97828692e-01 1.97944865e-01
7.23349571e-01 3.90810281e-01 -2.12225580e+00 5.17199457e-01
7.46722400e-01 1.50157166e+00 5.98383904e-01 8.18984270e-01
6.76892102e-02 8.01925778e-01 -9.73040819e-01 -5.05194217e-02
1.92136675e-01 -3.30320299e-01 -5.44033289e-01 1.39649818e-02
6.68245494e-01 -9.60225224e-01 -7.24605769e-02 6.66235447e-01
6.79133713e-01 3.14439356e-01 2.06058800e-01 -1.37829566e+00
3.67298573e-02 3.83661985e-01 -4.48192298e-01 2.42181197e-01
9.07307744e-01 -1.13557667e-01 5.12456656e-01 -9.87753689e-01
7.77360797e-01 1.05601001e+00 4.14498031e-01 5.35389148e-02
-6.87946856e-01 -9.43411052e-01 -1.32258862e-01 2.57941335e-01
-1.42763710e+00 -4.10491019e-01 5.47776103e-01 -4.22550917e-01
9.12474930e-01 3.69255215e-01 8.72460961e-01 7.20050871e-01
1.58342078e-01 5.44690490e-01 1.02730179e+00 -4.27792192e-01
3.67286831e-01 -1.42220939e-02 -9.28211734e-02 5.14096797e-01
-1.07940875e-01 -2.44708627e-01 -8.12726498e-01 -3.47013593e-01
9.20501590e-01 -4.57380205e-01 -2.94293880e-01 3.78687605e-02
-1.23457134e+00 6.32483542e-01 -1.21590458e-01 -1.39872683e-02
-4.52948630e-01 1.27532527e-01 5.41395783e-01 2.21647590e-01
3.40338260e-01 2.41095781e-01 2.51095384e-01 -8.67210150e-01
-1.25391459e+00 2.62058735e-01 7.32100725e-01 8.92221093e-01
2.82034487e-01 3.32484752e-01 2.58905673e-03 5.21435559e-01
3.44842702e-01 5.29283524e-01 4.64221746e-01 -1.26052666e+00
4.86456782e-01 1.26122817e-01 2.77615935e-01 -1.15603495e+00
-3.21396515e-02 1.99480802e-01 -1.73106894e-01 5.26665747e-01
2.20106110e-01 -3.79589051e-01 -4.53943878e-01 1.17617333e+00
4.63092864e-01 3.46772730e-01 -2.55185097e-01 1.01603687e+00
9.82968807e-01 1.06625271e+00 -1.49566382e-01 -2.00595886e-01
1.23835361e+00 -9.19130921e-01 -7.51671195e-01 4.53014731e-01
1.70741424e-01 -1.28271079e+00 1.09785938e+00 9.06672060e-01
-1.25528252e+00 -9.01751161e-01 -1.42521310e+00 1.43047929e-01
1.89849779e-01 2.39528283e-01 5.65064728e-01 9.02900338e-01
-1.22327721e+00 5.67788422e-01 -7.49848247e-01 -3.03846180e-01
-1.46541417e-01 3.31302971e-01 1.37103638e-02 3.74440104e-01
-8.02797973e-01 1.97643206e-01 2.37861335e-01 -2.61745393e-01
-6.66279495e-01 -5.40352285e-01 -5.58144629e-01 6.44176155e-02
1.00829937e-01 -3.50756377e-01 1.56192875e+00 -1.21455896e+00
-2.06929278e+00 1.12674750e-01 -9.25084725e-02 -3.26439232e-01
3.31100106e-01 -5.17870903e-01 -3.29848439e-01 9.02500391e-01
-1.28950626e-01 8.68368447e-01 1.23433971e+00 -9.88994122e-01
-9.17452753e-01 1.74003989e-01 2.79315591e-01 5.11907876e-01
-3.96423340e-01 1.14341319e-01 -5.88232279e-01 -7.19868362e-01
-2.38541394e-01 -9.97865558e-01 2.30126649e-01 3.06516811e-02
-1.02001585e-01 -4.55208980e-02 1.09116662e+00 -6.07015312e-01
1.45889223e+00 -2.34754586e+00 -1.86452612e-01 2.86281526e-01
-9.88954082e-02 3.29055935e-01 1.03521034e-01 6.53224468e-01
1.41219914e-01 9.42346528e-02 6.02191746e-01 -2.27497503e-01
-2.91269153e-01 -2.30490625e-01 -5.75658083e-01 1.31223768e-01
-3.62066180e-01 2.46791802e-02 -7.05910563e-01 -5.48572183e-01
3.46582949e-01 8.85094345e-01 -7.65419662e-01 2.91383952e-01
1.56107947e-01 3.63637447e-01 -3.36399853e-01 4.09499824e-01
7.14570105e-01 1.23508228e-03 1.48165002e-01 -3.21035594e-01
-5.43808877e-01 2.98167676e-01 -1.37434256e+00 1.64294493e+00
-3.78104895e-01 8.54392648e-01 -1.23680271e-01 -3.87061924e-01
8.75185609e-01 5.84790468e-01 5.57185113e-01 -7.01341331e-01
2.96468586e-01 1.75547078e-01 -2.79336721e-01 -6.59277916e-01
9.54342842e-01 4.11829978e-01 1.79513574e-01 5.71181715e-01
-3.33554178e-01 -2.56583720e-01 3.62900287e-01 3.77084970e-01
8.58723164e-01 4.68287677e-01 5.25301509e-03 1.57424822e-01
2.99297839e-01 -1.18943833e-01 1.69551790e-01 6.02500498e-01
-2.78068334e-02 6.64834619e-01 2.40373477e-01 -2.75934517e-01
-1.05968463e+00 -1.01693857e+00 9.31652039e-02 8.73129368e-01
3.45528990e-01 -9.51387048e-01 -1.05099022e+00 -2.42343649e-01
-4.29409593e-01 4.80608970e-01 2.19920948e-02 3.63050520e-01
-4.69069421e-01 -7.31104985e-02 4.21772480e-01 3.94156009e-01
5.73000371e-01 -1.01882732e+00 -1.25905585e+00 1.97079927e-01
-3.66784036e-01 -1.11032987e+00 -6.12803757e-01 -6.47898495e-01
-1.01469648e+00 -7.60990977e-01 -4.82220888e-01 -7.52895594e-01
5.23131669e-01 5.29750407e-01 5.32563686e-01 8.25387165e-02
-2.30479941e-01 3.09783876e-01 -7.04818964e-01 -2.99551904e-01
-1.95213258e-01 -3.16814363e-01 6.96762875e-02 -8.37777182e-03
-8.27707276e-02 -3.71339440e-01 -6.83986604e-01 3.57131690e-01
-8.19921613e-01 4.88869369e-01 1.52867332e-01 3.93901855e-01
8.34595025e-01 2.81104088e-01 2.78882295e-01 -3.98811549e-01
6.82397187e-01 -4.60134476e-01 -7.67246723e-01 -2.54637241e-01
-3.02692980e-01 -2.45593309e-01 8.14460337e-01 -5.04273534e-01
-1.23848534e+00 1.38547897e-01 2.79088821e-02 -3.50849271e-01
-4.97695804e-02 3.57065111e-01 2.98673362e-01 -6.96907863e-02
2.35537976e-01 -1.34794861e-01 9.69360396e-02 -3.97595167e-01
-5.88508137e-03 1.09665251e+00 5.62585711e-01 -3.19435477e-01
4.81171548e-01 5.45341790e-01 -2.32844174e-01 -9.95669484e-01
-3.97097059e-02 -3.98330748e-01 -2.34701678e-01 -7.52313495e-01
6.41856611e-01 -1.02809644e+00 -8.80427241e-01 6.33714020e-01
-1.04049170e+00 -1.15275942e-01 1.78015769e-01 8.39736104e-01
-5.53043604e-01 4.04194146e-01 -6.25659406e-01 -8.25151026e-01
-5.13562739e-01 -1.24990284e+00 1.25085926e+00 3.55698228e-01
-4.36889082e-01 -3.23820144e-01 -1.71699356e-02 4.23811078e-01
4.26066130e-01 3.15619618e-01 2.81365216e-01 1.14513626e-02
-5.61254919e-01 -9.72081721e-02 -9.31008905e-02 7.38057867e-02
1.29345372e-01 6.06420398e-01 -8.30861509e-01 -3.87489259e-01
-6.43131658e-02 -1.00591928e-01 1.36094615e-01 2.70426303e-01
1.21822584e+00 -2.38197654e-01 -6.69206008e-02 5.10396302e-01
1.33707285e+00 8.68210852e-01 7.91984499e-01 3.21295351e-01
4.67655003e-01 2.25197524e-01 1.19489431e+00 9.83705103e-01
2.96018183e-01 9.69152868e-01 2.75505632e-01 1.20342746e-02
-2.01087400e-01 -2.01288134e-01 8.36276174e-01 1.20053124e+00
-3.84477645e-01 -4.52748060e-01 -4.90116298e-01 2.26633132e-01
-1.43245435e+00 -1.13246346e+00 -2.18616486e-01 2.37908363e+00
3.47445846e-01 -8.11186880e-02 1.97272435e-01 3.44561964e-01
7.95092881e-01 4.56386507e-02 -6.73954114e-02 -9.40319598e-01
3.51712674e-01 4.61906731e-01 1.50134772e-01 3.47069710e-01
-1.00693214e+00 4.45877433e-01 6.40194082e+00 8.67116451e-01
-1.48019552e+00 -1.09984212e-01 3.28633279e-01 -6.16290271e-01
1.01484112e-01 -6.22365065e-02 -5.43160200e-01 7.74213910e-01
1.37124193e+00 -2.29370058e-01 4.36022282e-01 8.96503925e-01
7.70009339e-01 -6.71357930e-01 -7.89314806e-01 1.20796335e+00
2.43950889e-01 -1.25906873e+00 6.52980134e-02 4.98669408e-02
4.89295632e-01 -3.14345419e-01 1.62042484e-01 -1.70692444e-01
-3.35782677e-01 -7.34829783e-01 1.18150020e+00 4.01125938e-01
8.79911423e-01 -1.10888767e+00 6.23706162e-01 -1.24775767e-01
-1.57238686e+00 1.76650211e-01 -7.78554901e-02 -2.71754742e-01
4.67893690e-01 2.99107343e-01 -1.04716206e+00 4.43042487e-01
1.01407874e+00 2.17716888e-01 -4.90340590e-01 1.17593539e+00
1.07652314e-01 8.33691776e-01 -4.49301898e-01 1.75041240e-02
4.47958224e-02 -7.84817263e-02 5.42338073e-01 1.04179680e+00
8.80714357e-01 2.18816593e-01 9.03095305e-02 2.57210940e-01
1.88879684e-01 4.85516071e-01 -4.95189697e-01 -2.32298374e-01
6.69324398e-01 1.25024879e+00 -1.03822899e+00 -5.98520696e-01
-3.47321898e-01 1.05701196e+00 -2.19270006e-01 1.47980481e-01
-1.23064435e+00 -9.19671535e-01 2.23445565e-01 3.21132094e-01
4.64029938e-01 -4.13776875e-01 8.22338909e-02 -7.02882051e-01
1.44966304e-01 -8.66269469e-01 1.70777306e-01 -1.21095717e+00
-5.40135622e-01 9.87830758e-01 1.61832571e-01 -1.79528952e+00
-6.39852881e-01 -8.32382515e-02 -4.75480407e-01 5.12278378e-01
-9.94718075e-01 -6.40811086e-01 -7.76142895e-01 4.91234720e-01
6.89747334e-01 1.33219928e-01 6.83970273e-01 5.34176230e-01
-2.23607436e-01 3.88705581e-01 -2.23985184e-02 -4.14077610e-01
7.77162850e-01 -8.16087723e-01 3.18210721e-01 7.53493547e-01
-1.51334321e-02 6.16685390e-01 6.06616437e-01 -8.65016580e-01
-1.52826667e+00 -6.43290102e-01 7.00332224e-01 2.02601612e-01
3.50301951e-01 -1.51581451e-01 -5.66308200e-01 2.69517869e-01
6.11361861e-01 -3.96890670e-01 8.34448934e-01 -6.30469322e-01
-2.83584781e-02 -7.63546005e-02 -1.11966240e+00 5.24304628e-01
6.00358188e-01 -4.44876194e-01 -2.31014088e-01 5.66359377e-03
4.43560898e-01 -7.99640238e-01 -9.04086351e-01 2.79183090e-02
1.00461328e+00 -1.18551624e+00 7.06186414e-01 4.30717319e-01
4.25516486e-01 -5.80406129e-01 -3.79743278e-02 -9.90140319e-01
2.36767411e-01 -9.82301891e-01 7.16220364e-02 1.13501298e+00
2.53099557e-02 -2.45823234e-01 4.26079690e-01 1.99069694e-01
-1.04883932e-01 -5.24041474e-01 -7.25684285e-01 -5.97371519e-01
-8.46723378e-01 -6.36705518e-01 5.50008178e-01 4.60906327e-01
2.40797326e-01 -6.04682490e-02 -5.43093324e-01 2.28335187e-01
4.96958166e-01 1.27497330e-01 9.08368409e-01 -9.51940715e-01
-4.76678222e-01 1.59871757e-01 -5.98039865e-01 -9.18697178e-01
-3.45193654e-01 -4.51259494e-01 -1.03555828e-01 -1.27162099e+00
-6.24862500e-02 -2.32369214e-01 9.21776965e-02 1.46859512e-01
1.22714147e-01 6.51715100e-01 6.24072969e-01 2.16818988e-01
-4.87615556e-01 2.06621930e-01 1.00667822e+00 2.64890790e-01
-3.23268771e-01 -1.78167060e-01 -2.98989087e-01 6.89659178e-01
7.06101477e-01 -3.03364366e-01 -7.81876743e-01 -3.99437875e-01
4.91822287e-02 3.30621183e-01 1.72673255e-01 -1.50635779e+00
2.02831060e-01 -2.79725585e-02 3.61954182e-01 -7.66861558e-01
5.72806358e-01 -7.19526887e-01 7.65043616e-01 4.57877457e-01
-1.03123672e-03 4.87671643e-01 4.14319038e-01 1.96024179e-01
-3.77113849e-01 -2.76202530e-01 4.40840870e-01 2.99359947e-01
-7.16093123e-01 -1.39813393e-01 -1.02522099e+00 -4.99578059e-01
1.18126810e+00 -5.47571599e-01 -6.12883531e-02 -7.93063462e-01
-3.99103194e-01 -3.01362723e-01 5.32410204e-01 5.85586607e-01
7.97019362e-01 -1.45362222e+00 -2.04303741e-01 2.87842065e-01
-4.91704911e-01 -3.26832324e-01 4.49662775e-01 7.14028358e-01
-1.18468177e+00 2.86150634e-01 -4.89919782e-01 -6.57509446e-01
-1.89138532e+00 1.70019999e-01 -3.31402898e-01 2.03550726e-01
-6.49723232e-01 5.45573473e-01 -1.35264341e-02 5.06349981e-01
2.63686806e-01 -3.47743124e-01 -3.82089138e-01 2.71341562e-01
1.02333879e+00 6.18808568e-01 1.63639918e-01 -7.62755036e-01
-2.27731466e-01 5.80271721e-01 9.43137407e-02 -6.28466487e-01
1.06518269e+00 -2.89900213e-01 1.49950713e-01 3.64516169e-01
1.12435913e+00 4.49569404e-01 -1.18829489e+00 5.00388145e-01
-5.00810206e-01 -9.26007688e-01 1.72889158e-02 -3.50461572e-01
-1.19197619e+00 6.20023131e-01 8.42886627e-01 8.22863542e-03
1.34482288e+00 -6.35690033e-01 1.05898511e+00 -9.52208415e-02
6.54760361e-01 -1.28147936e+00 2.76230156e-01 1.62387297e-01
7.01948822e-01 -5.38044035e-01 1.35987774e-02 -6.54179931e-01
-7.64685750e-01 1.46835148e+00 6.66631758e-01 -4.07451056e-02
3.94329250e-01 5.30801594e-01 1.04543127e-01 2.21281499e-03
-6.73750639e-01 3.83779287e-01 9.11753923e-02 3.84516597e-01
5.40074050e-01 -3.21463756e-02 -6.53981745e-01 1.56336695e-01
-6.02884233e-01 2.98227042e-01 9.64510143e-01 1.32643485e+00
-4.15274471e-01 -1.03505993e+00 -5.75918078e-01 2.68568039e-01
-6.68951452e-01 -4.60011624e-02 -1.40528485e-01 5.72910368e-01
-1.55877147e-04 1.40517259e+00 3.91812801e-01 -6.45866156e-01
1.95714653e-01 -1.33853272e-01 4.07617956e-01 -3.29214215e-01
-7.87130415e-01 3.95529747e-01 2.75545925e-01 -1.00707078e+00
-5.51763058e-01 -4.52055097e-01 -1.56916678e+00 -4.13609773e-01
-1.71977237e-01 1.96662977e-01 1.10798585e+00 4.49215293e-01
6.71217918e-01 4.03842300e-01 8.70669484e-01 -1.35532737e+00
6.61601126e-02 -5.84865510e-01 -3.56749535e-01 3.99934798e-01
-7.10135177e-02 -6.66232944e-01 -2.30487213e-01 3.23041022e-01] | [10.579748153686523, -1.0129733085632324] |
34ffe750-2af0-4871-aeb1-09cf03722c50 | partial-inference-in-structured-prediction | 2306.03949 | null | https://arxiv.org/abs/2306.03949v1 | https://arxiv.org/pdf/2306.03949v1.pdf | Partial Inference in Structured Prediction | In this paper, we examine the problem of partial inference in the context of structured prediction. Using a generative model approach, we consider the task of maximizing a score function with unary and pairwise potentials in the space of labels on graphs. Employing a two-stage convex optimization algorithm for label recovery, we analyze the conditions under which a majority of the labels can be recovered. We introduce a novel perspective on the Karush-Kuhn-Tucker (KKT) conditions and primal and dual construction, and provide statistical and topological requirements for partial recovery with provable guarantees. | ['Jean Honorio', 'Chuyang Ke'] | 2023-06-06 | null | null | null | null | ['structured-prediction'] | ['methodology'] | [ 6.09617710e-01 6.57594681e-01 -6.75357878e-01 -5.92808664e-01
-1.33514428e+00 -6.93959951e-01 2.10340485e-01 -6.35003522e-02
7.84749091e-02 1.02765453e+00 2.45003134e-01 -2.60457963e-01
-5.33482730e-01 -4.62214798e-01 -9.14775431e-01 -8.60658169e-01
-2.05130845e-01 8.03824604e-01 -3.70669365e-01 3.80606711e-01
1.64936975e-01 5.65926254e-01 -1.21020198e+00 1.68901488e-01
6.91988766e-01 7.27169454e-01 -1.04897998e-01 7.01728106e-01
3.63722950e-01 7.06034184e-01 -4.53914590e-02 -5.70441663e-01
4.72501874e-01 -4.67479020e-01 -1.27245677e+00 3.78440320e-01
7.21784174e-01 -2.64937162e-01 -3.38002145e-01 1.34636283e+00
3.25559437e-01 1.99715927e-01 9.83575463e-01 -1.46927667e+00
-6.09131515e-01 6.95840955e-01 -5.12133420e-01 -6.74946383e-02
6.22502148e-01 -4.40231085e-01 1.65373778e+00 -9.91577029e-01
1.05173051e+00 1.04814744e+00 7.00851440e-01 4.67709929e-01
-1.83891773e+00 -3.03907096e-01 1.08603977e-01 -1.17735796e-01
-1.48446441e+00 -5.66549838e-01 7.11982906e-01 -4.93766785e-01
3.82051408e-01 4.39460605e-01 1.00361899e-01 7.21172690e-01
-1.44167185e-01 9.89062071e-01 1.24706399e+00 -5.76134562e-01
2.79024780e-01 2.12722227e-01 4.10075784e-01 9.98373687e-01
2.34763250e-01 9.33770090e-03 -6.50714338e-01 -6.77017450e-01
5.20340025e-01 -1.37711257e-01 -2.52530277e-01 -8.13393593e-01
-9.04305398e-01 1.02824843e+00 2.03555942e-01 -8.50442350e-02
-1.18312456e-01 2.27627739e-01 -6.08450081e-03 1.03160322e-01
5.75215816e-01 1.23913482e-01 -2.50904709e-01 4.35196579e-01
-1.15809107e+00 1.33736044e-01 1.15712631e+00 1.40693676e+00
8.66027296e-01 -2.40709573e-01 -2.62093306e-01 4.60215956e-01
5.52045643e-01 6.76384091e-01 -5.86517751e-01 -1.22054470e+00
4.57287908e-01 -3.51156071e-02 2.63563961e-01 -8.64674389e-01
-2.61335403e-01 -4.54908252e-01 -6.90246642e-01 -1.18768759e-01
4.23857450e-01 1.19672023e-01 -7.75070071e-01 2.01036644e+00
4.71570909e-01 4.43369538e-01 -2.04579741e-01 6.84044659e-01
4.70924646e-01 5.70530534e-01 -1.01746276e-01 -8.59992802e-01
8.32710981e-01 -7.76446998e-01 -6.55469954e-01 -5.11059240e-02
8.07006598e-01 -4.45058316e-01 5.25846243e-01 2.02508286e-01
-1.29989719e+00 -3.40391174e-02 -8.76476467e-01 -2.54086435e-01
1.55763671e-01 1.04489446e-01 7.13747084e-01 6.80400491e-01
-1.28641963e+00 8.27677608e-01 -6.45874262e-01 -2.76054884e-03
1.77448139e-01 4.56961542e-01 -5.74491084e-01 -3.35179865e-01
-5.41345656e-01 6.79349065e-01 2.28704453e-01 1.39469177e-01
-1.15271318e+00 -4.16656047e-01 -7.06910491e-01 2.62689893e-03
5.60553670e-01 -4.40269411e-01 9.85385835e-01 -6.04795158e-01
-9.64094460e-01 1.20710325e+00 -5.23403466e-01 -1.48931965e-01
4.30577725e-01 1.56809270e-01 7.14480430e-02 2.38752767e-01
2.13834256e-01 3.43700349e-01 6.98830366e-01 -1.60894048e+00
-3.58658433e-01 -5.68341792e-01 -7.91028291e-02 1.50824159e-01
2.98007220e-01 -1.58231974e-01 -1.41119540e-01 -4.98903453e-01
5.36231458e-01 -1.20163143e+00 -3.12576264e-01 -1.73289701e-01
-8.06358814e-01 -9.25754234e-02 1.71334222e-01 -8.83107901e-01
9.16236281e-01 -1.98398221e+00 6.15872681e-01 7.36289799e-01
3.73812675e-01 -3.56397122e-01 -1.58800080e-01 4.37164843e-01
1.20763434e-02 1.32252216e-01 -5.22294164e-01 -8.06475520e-01
3.04728031e-01 3.62584651e-01 -4.97939497e-01 8.99924397e-01
-1.56708673e-01 7.80049920e-01 -9.08845544e-01 -5.33188462e-01
-1.31020606e-01 1.16853021e-01 -6.12349808e-01 1.58551320e-01
-2.38365695e-01 5.03671944e-01 -5.02325773e-01 9.03394103e-01
8.60069275e-01 -6.02621794e-01 6.72816634e-01 1.38898538e-02
3.37874413e-01 3.22408825e-01 -1.20370042e+00 1.53190506e+00
-6.57750964e-02 3.47024113e-01 7.07223237e-01 -1.16389275e+00
5.13994455e-01 1.85046226e-01 6.46917880e-01 -2.72729367e-01
-7.22983256e-02 1.60947055e-01 -7.44716942e-01 -7.75725916e-02
3.90630990e-01 -3.75027120e-01 -2.85815865e-01 7.88641334e-01
1.84667274e-01 1.73349828e-01 -1.06855948e-03 5.56851506e-01
1.05865741e+00 2.41688177e-01 3.13233614e-01 -4.35600042e-01
5.98957911e-02 -2.10861206e-01 6.69746995e-01 8.36732566e-01
-3.97412898e-03 4.02072549e-01 7.31104374e-01 -2.65925229e-02
-1.20608366e+00 -1.10653090e+00 -4.18671131e-01 1.25061774e+00
-6.07109740e-02 -3.14125746e-01 -7.73424327e-01 -7.90436864e-01
1.37321666e-01 7.60682225e-01 -7.80169547e-01 1.03411963e-02
-3.66061270e-01 -6.85054302e-01 3.53466630e-01 3.69210392e-01
-1.47185862e-01 -6.09279394e-01 1.46336839e-01 7.35115856e-02
-2.73202270e-01 -1.10589194e+00 -5.22905827e-01 3.53533953e-01
-1.07798815e+00 -1.06657672e+00 -3.85728925e-01 -8.24889004e-01
1.07200086e+00 1.40479639e-01 1.13152206e+00 -3.73354107e-02
-1.23304643e-01 5.18271148e-01 -7.14379847e-02 4.52325791e-01
-3.97771180e-01 5.53865358e-02 -8.51009786e-02 9.22288001e-03
-1.39830738e-01 -6.24134004e-01 -8.36034268e-02 2.00069755e-01
-5.71383774e-01 3.01595237e-02 3.50980192e-01 1.00027597e+00
1.13908899e+00 -2.01972514e-01 2.55099177e-01 -1.37168193e+00
3.72344583e-01 -5.90980709e-01 -7.85017848e-01 6.75448418e-01
-1.03993320e+00 5.21775305e-01 2.09134981e-01 -1.56492576e-01
-7.92167544e-01 4.90905404e-01 1.53159767e-01 -3.58281255e-01
1.26369148e-01 4.01052296e-01 -4.03586179e-01 -4.57314521e-01
3.94947499e-01 2.46960998e-01 -2.56048292e-01 -5.95935404e-01
6.83106780e-01 3.49627167e-01 3.67019296e-01 -9.62611914e-01
6.90490961e-01 6.64486885e-01 5.93811870e-01 -3.60557854e-01
-1.38759100e+00 -3.83095771e-01 -8.66980851e-01 -1.18007816e-01
6.14881396e-01 -6.21066093e-01 -6.96330369e-01 -8.43121633e-02
-8.26936007e-01 -3.32761258e-01 -4.72583622e-01 2.07716033e-01
-9.84028697e-01 6.93908691e-01 -6.25632644e-01 -1.01737916e+00
3.84448632e-03 -9.73035455e-01 1.16973782e+00 -2.68327296e-01
1.56609371e-01 -1.22092152e+00 3.76043767e-01 4.43836302e-01
-1.48910984e-01 3.73137295e-01 9.29817915e-01 -5.44221818e-01
-8.39725435e-01 -1.19803101e-01 -5.24143100e-01 1.01923950e-01
-4.25589263e-01 -3.51584464e-01 -8.69767010e-01 -5.46358943e-01
-5.90123702e-04 -3.33992392e-01 9.31077838e-01 5.35012186e-01
9.41418648e-01 -7.65157819e-01 -5.19261599e-01 8.04226100e-01
1.57699621e+00 -4.29446250e-01 2.90434659e-01 -2.83062905e-01
8.48948836e-01 5.41938365e-01 4.76845503e-01 5.02030730e-01
4.91753668e-01 4.87007856e-01 2.82035947e-01 3.69500965e-01
1.17962569e-01 -6.68291509e-01 1.37127414e-01 7.35651910e-01
1.44540995e-01 -2.30928317e-01 -5.71825445e-01 5.03702164e-01
-2.05617547e+00 -1.00126135e+00 -1.05981715e-01 2.39226437e+00
8.57015848e-01 -2.62636662e-01 9.92700160e-02 -1.06215499e-01
9.50594664e-01 1.70027420e-01 -4.95469958e-01 -1.33756056e-01
-1.20597757e-01 2.24119395e-01 1.07830167e+00 9.45726097e-01
-9.17696416e-01 6.61095500e-01 8.39051628e+00 9.47083056e-01
-2.43668020e-01 3.70625883e-01 6.83648586e-01 7.88257867e-02
-7.91883171e-01 4.78737682e-01 -8.78781438e-01 2.24769220e-01
9.24529612e-01 1.14956796e-01 9.44055438e-01 8.46080244e-01
-4.25466150e-01 1.04357295e-01 -1.41730964e+00 6.70128167e-01
1.24187820e-01 -1.08597457e+00 -3.80913079e-01 5.40042281e-01
1.13095355e+00 -2.03306243e-01 1.42826572e-01 -5.00549041e-02
7.04560637e-01 -1.20927405e+00 5.68021834e-01 6.42044723e-01
1.06647205e+00 -7.61226833e-01 2.38174930e-01 4.60858375e-01
-1.06084800e+00 1.08455062e-01 -3.83966118e-01 1.98333159e-01
3.61097157e-01 8.92265737e-01 -6.10755444e-01 5.26455462e-01
-1.72124743e-01 7.13435352e-01 -2.32727826e-02 8.61799479e-01
-3.72501612e-01 6.81579113e-01 -4.73743021e-01 3.36565584e-01
-8.51388276e-02 -5.14393330e-01 6.44179583e-01 1.12500870e+00
2.49931440e-01 3.14682245e-01 6.38294458e-01 8.64361882e-01
-4.93699938e-01 1.39677644e-01 -5.87654054e-01 1.83316432e-02
5.68640530e-01 1.06626868e+00 -7.63178766e-01 -2.42795601e-01
-2.75668316e-02 8.40452135e-01 6.51268065e-01 4.00637120e-01
-4.79546756e-01 2.00859010e-01 1.55979872e-01 -2.72446051e-02
2.62002379e-01 -8.14559162e-02 -5.78390896e-01 -1.35727906e+00
1.50474221e-01 -4.62101132e-01 5.72600543e-01 -4.62883681e-01
-1.36835217e+00 -5.28471284e-02 1.83647275e-01 -6.45223618e-01
-1.73436582e-01 -4.61705029e-01 -3.89076352e-01 7.67182589e-01
-1.12069654e+00 -1.29125583e+00 2.97772259e-01 5.68304598e-01
-1.15309559e-01 2.32188284e-01 8.12449872e-01 1.44706100e-01
-4.24972922e-01 5.88373184e-01 4.63731349e-01 -2.76433110e-01
2.70434231e-01 -1.63880038e+00 6.22726083e-02 8.87515962e-01
4.23776060e-01 3.49296093e-01 8.08328509e-01 -6.78798854e-01
-1.56549358e+00 -9.15649652e-01 9.77454662e-01 -4.03915077e-01
4.86858219e-01 -3.88982862e-01 -6.13970816e-01 1.24818945e+00
-1.44145772e-01 1.45768389e-01 8.19992185e-01 4.75356638e-01
-5.80493510e-01 1.91813707e-01 -1.49083030e+00 8.20258856e-02
1.29461896e+00 -8.36243391e-01 -1.39948681e-01 9.69066978e-01
3.49485159e-01 -2.45514035e-01 -9.34598923e-01 3.32895339e-01
3.79712999e-01 -5.80999255e-01 9.77906048e-01 -8.30016315e-01
7.17230439e-02 -1.14262002e-02 -6.74061000e-01 -9.41938102e-01
-6.68380976e-01 -8.11144710e-01 -4.00434792e-01 9.81245697e-01
5.63055575e-01 -5.83029449e-01 1.03683603e+00 8.94412756e-01
1.30826816e-01 -7.40960240e-01 -1.19785905e+00 -6.70729160e-01
2.84959376e-02 -2.46242210e-01 1.74644738e-01 9.53756988e-01
1.91149786e-01 3.41886729e-01 -9.97519255e-01 2.59708434e-01
1.23377144e+00 3.16272587e-01 3.32411289e-01 -1.33061779e+00
-6.10122561e-01 -6.58500493e-02 -2.49701247e-01 -1.03503287e+00
7.17235923e-01 -1.47574425e+00 1.95239097e-01 -1.43331778e+00
8.90018284e-01 -6.28124475e-01 -2.58118302e-01 3.65870357e-01
1.50209665e-01 6.29534200e-02 1.48965195e-01 5.19052923e-01
-9.63876784e-01 2.12867409e-01 9.58933771e-01 4.79787514e-02
2.67161727e-01 2.01877803e-01 -7.67394364e-01 2.88461804e-01
4.40653622e-01 -7.48241603e-01 -3.27284336e-01 -1.10113978e-01
7.21259773e-01 5.34727812e-01 4.05947298e-01 -2.37415180e-01
1.86702460e-01 -2.82825410e-01 3.20024453e-02 -6.36960566e-01
4.06148404e-01 -6.02283061e-01 4.37318414e-01 3.10963571e-01
-9.26663101e-01 -2.45834187e-01 -5.25044501e-01 1.00556982e+00
2.52610177e-01 -7.21446872e-01 7.78549194e-01 4.05928493e-02
-1.33388594e-01 4.33551550e-01 4.54417132e-02 3.15582961e-01
1.00263965e+00 1.78674132e-01 -2.04294652e-01 -4.02038187e-01
-1.17743182e+00 2.12723106e-01 6.29822969e-01 -5.15119851e-01
5.16984522e-01 -1.47780871e+00 -7.83959329e-01 -3.93013246e-02
-1.33692056e-01 -2.19697878e-01 1.74725249e-01 9.86591458e-01
-5.46831675e-02 3.05233717e-01 1.96879774e-01 -3.46869588e-01
-1.13137650e+00 7.40858197e-01 1.63250521e-01 -5.29011309e-01
-5.86682320e-01 8.06162834e-01 1.79535761e-01 -4.50944126e-01
1.70020267e-01 2.08326891e-01 4.15479451e-01 -9.09289345e-02
1.22786313e-01 6.47628605e-01 -1.34362683e-01 -8.23752642e-01
-2.62793362e-01 3.68587464e-01 -7.09370747e-02 -4.74559665e-01
1.21361101e+00 -2.54145205e-01 -3.71444434e-01 3.31625879e-01
1.34915316e+00 3.20819288e-01 -1.26946700e+00 -5.26544273e-01
2.49314699e-02 -5.39009333e-01 1.28016621e-01 -5.87324739e-01
-9.39897954e-01 6.39412582e-01 8.26868415e-02 5.00043631e-01
6.64681733e-01 4.55112159e-01 5.71395099e-01 3.29922855e-01
5.58544219e-01 -1.10778856e+00 -4.68209356e-01 1.92508832e-01
4.95666385e-01 -9.10692275e-01 1.90147430e-01 -6.92073882e-01
-5.46650469e-01 7.59707093e-01 -2.57325292e-01 -2.01872870e-01
7.55265713e-01 2.53424436e-01 -6.44971907e-01 -3.49840105e-01
-8.49677205e-01 -1.72229543e-01 5.45107245e-01 5.86134315e-01
1.58272207e-01 5.96161127e-01 -2.31120646e-01 1.80610299e-01
-9.21067894e-02 -3.93500507e-01 3.19105595e-01 7.40422130e-01
-6.04155898e-01 -1.11109495e+00 -1.21614687e-01 6.03356481e-01
-4.74507123e-01 2.28682607e-02 -5.58381736e-01 2.12656468e-01
-1.36766315e-01 9.15463984e-01 -4.86869872e-01 -1.86259687e-01
-1.71582341e-01 2.81309426e-01 7.57594705e-01 -6.21231318e-01
8.92414525e-02 2.10171118e-01 3.20179045e-01 -4.62259918e-01
-5.76259136e-01 -9.68122602e-01 -8.82549644e-01 -5.77511132e-01
-6.21288478e-01 2.65726864e-01 5.61235726e-01 1.06351888e+00
2.60898322e-01 -7.18326569e-02 8.98980439e-01 -5.51474154e-01
-9.21173513e-01 -7.13001728e-01 -1.00541210e+00 2.23805830e-01
2.16382965e-01 -5.45523405e-01 -6.37323439e-01 -1.65169667e-02] | [7.525025844573975, 4.229300498962402] |
3cdbe7f8-dcca-466a-8330-8a36a5803011 | multi-modal-facial-expression-recognition | 2303.08419 | null | https://arxiv.org/abs/2303.08419v2 | https://arxiv.org/pdf/2303.08419v2.pdf | Multi Modal Facial Expression Recognition with Transformer-Based Fusion Networks and Dynamic Sampling | Facial expression recognition is an essential task for various applications, including emotion detection, mental health analysis, and human-machine interactions. In this paper, we propose a multi-modal facial expression recognition method that exploits audio information along with facial images to provide a crucial clue to differentiate some ambiguous facial expressions. Specifically, we introduce a Modal Fusion Module (MFM) to fuse audio-visual information, where image and audio features are extracted from Swin Transformer. Additionally, we tackle the imbalance problem in the dataset by employing dynamic data resampling. Our model has been evaluated in the Affective Behavior in-the-wild (ABAW) challenge of CVPR 2023. | ['Chee Sun Won', 'NamHo Kim', 'Jun-Hwa Kim'] | 2023-03-15 | null | null | null | null | ['facial-expression-recognition'] | ['computer-vision'] | [ 3.59575987e-01 -3.05500329e-01 -3.19419242e-02 -7.16337740e-01
-1.01898146e+00 -2.60443270e-01 3.41619670e-01 2.62421686e-02
-2.74618357e-01 5.63503265e-01 3.95776421e-01 4.52281803e-01
-3.52547392e-02 -1.78933352e-01 -7.54720494e-02 -8.71099532e-01
8.28493759e-02 -1.33808702e-01 -4.71766323e-01 -3.93228382e-01
-2.10132733e-01 4.89510894e-01 -1.96844888e+00 7.71281600e-01
4.05961454e-01 1.57359505e+00 -6.21722102e-01 4.17679161e-01
5.96569211e-04 9.69670415e-01 -5.64771235e-01 -5.49130142e-01
-2.42509440e-01 -4.02542174e-01 -7.25558162e-01 2.63606578e-01
1.87457129e-01 -1.06797256e-01 3.45670909e-01 9.83976901e-01
8.55401695e-01 1.16278104e-01 4.84347016e-01 -1.78113186e+00
2.23130748e-01 1.35550052e-01 -1.08318734e+00 1.16596282e-01
7.10936248e-01 -1.19392566e-01 7.34744787e-01 -9.83784556e-01
5.79395592e-01 1.26756072e+00 6.38027072e-01 6.61432922e-01
-1.12124932e+00 -8.94297481e-01 -1.04013152e-01 6.30274534e-01
-1.69790316e+00 -9.54978347e-01 1.08217955e+00 -3.72993708e-01
4.89921123e-01 3.39529514e-01 6.44985080e-01 1.21279430e+00
-2.35468596e-01 9.43024457e-01 1.10354483e+00 -3.38135511e-01
-3.33731472e-02 8.42706207e-03 -1.70167282e-01 5.16758561e-01
-7.91839957e-01 -2.47270286e-01 -1.07352388e+00 -5.86427808e-01
-4.18536551e-02 -4.95356955e-02 -1.15715086e-01 3.08000684e-01
-5.23038268e-01 7.76909351e-01 -1.72692686e-02 2.59340286e-01
-9.17697012e-01 -6.29497916e-02 8.20454538e-01 3.31895322e-01
6.95328295e-01 -1.63666084e-01 -2.37242073e-01 -4.46172446e-01
-8.46697330e-01 2.84955531e-01 3.34593564e-01 1.27868563e-01
7.68757105e-01 2.18185154e-03 -3.20258021e-01 1.29150224e+00
3.56380969e-01 3.11489552e-01 6.12443805e-01 -1.17776954e+00
-3.01721711e-02 4.97166336e-01 -2.24408671e-01 -1.28843653e+00
-5.52734852e-01 2.23169059e-01 -1.00014162e+00 3.94671299e-02
-2.45973095e-01 -1.26247019e-01 -3.83945048e-01 1.94121206e+00
6.14762127e-01 3.53871673e-01 2.03107409e-02 8.66922200e-01
1.02315557e+00 3.51391494e-01 3.19412172e-01 -5.68650723e-01
1.68766677e+00 -4.66341823e-01 -9.99332011e-01 4.02858816e-02
4.74865496e-01 -6.24721408e-01 6.84429049e-01 7.60905981e-01
-9.23096359e-01 -4.17733550e-01 -7.22507179e-01 2.66158253e-01
-1.13547668e-01 4.12677974e-01 5.09409606e-01 7.57861912e-01
-9.10161674e-01 -1.01191588e-02 -4.63394850e-01 -1.35832474e-01
5.47830164e-01 4.52989578e-01 -9.40793872e-01 1.65160283e-01
-1.25264788e+00 4.70361739e-01 8.91031176e-02 3.80805522e-01
-4.75184739e-01 -5.05550623e-01 -1.03150022e+00 -2.38851726e-01
4.31487486e-02 -1.83914602e-01 1.24233103e+00 -1.66625977e+00
-1.65888560e+00 1.12405419e+00 -5.28224289e-01 -2.69007713e-01
1.20474510e-01 -1.04226977e-01 -8.41724396e-01 4.93579388e-01
-2.26948515e-01 5.94530761e-01 1.15907657e+00 -8.91398311e-01
-3.54023188e-01 -8.69393826e-01 -4.66418028e-01 1.11093245e-01
-2.64979273e-01 6.07583940e-01 -2.72060484e-01 -5.11133254e-01
-1.40521556e-01 -7.04280853e-01 1.47499025e-01 -1.36174798e-01
-2.17687786e-01 -4.14037496e-01 1.03426397e+00 -8.21186185e-01
1.19109344e+00 -2.68335819e+00 8.82210806e-02 5.36065161e-01
2.10478410e-01 1.43266499e-01 -2.21578851e-01 4.44941856e-02
-2.94550836e-01 -3.01112980e-01 1.35379657e-01 -8.39147389e-01
-5.06276824e-02 -2.19891369e-02 -2.71209925e-01 4.58825201e-01
5.12271821e-01 6.68063045e-01 -4.62433577e-01 -8.12705815e-01
2.15586834e-02 7.40252137e-01 -5.74715972e-01 2.68596977e-01
1.93117067e-01 5.88606715e-01 -1.83467016e-01 1.04602599e+00
6.55466020e-01 2.15747967e-01 4.61131558e-02 -4.38953608e-01
1.81913719e-01 -3.55730027e-01 -1.00339782e+00 1.76678944e+00
-3.85567904e-01 5.51604033e-01 5.02928019e-01 -9.54013824e-01
9.18230653e-01 6.94513440e-01 8.58720481e-01 -7.51781225e-01
5.62798381e-01 -7.99914896e-02 -3.49425644e-01 -7.16888666e-01
3.37765008e-01 -3.85916531e-01 -2.76398569e-01 2.44944513e-01
3.63488376e-01 1.35615962e-02 -4.15859558e-02 -9.58628878e-02
9.47845936e-01 -1.81637198e-01 3.45327646e-01 4.43986356e-01
8.26016784e-01 -5.69465637e-01 7.88720012e-01 -5.37830405e-02
-6.73459768e-01 4.69781339e-01 6.18782461e-01 -2.61580706e-01
-2.31245965e-01 -8.18681061e-01 -5.15700281e-02 1.43624163e+00
-4.05519605e-01 -6.52838469e-01 -7.52606690e-01 -4.83405441e-01
-2.73574740e-01 1.58447564e-01 -8.65189731e-01 -4.46544915e-01
2.73587438e-03 -8.42923760e-01 8.48294377e-01 3.49383950e-01
4.22865748e-01 -1.06350935e+00 -5.90034962e-01 2.19192579e-02
-7.10920155e-01 -1.17761660e+00 -1.91651270e-01 -1.24971487e-01
-1.28365889e-01 -9.44496870e-01 -3.73384953e-01 -3.05712938e-01
2.49945655e-01 -1.65618658e-01 8.28379989e-01 -2.44439498e-01
-7.06720650e-01 5.82312465e-01 -4.13797528e-01 -5.50679743e-01
-2.53641486e-01 -4.08533871e-01 1.77762657e-01 9.08191144e-01
6.11966133e-01 -6.40908480e-01 -4.34162736e-01 2.60344267e-01
-8.60460162e-01 -2.81755120e-01 2.43267536e-01 7.87479043e-01
6.31725132e-01 -2.10719600e-01 9.05045211e-01 -4.43756729e-01
7.52981365e-01 -5.63327074e-01 -8.78770575e-02 1.08455598e-01
9.10677761e-02 -4.28376794e-01 2.08599493e-01 -4.53736663e-01
-1.06377375e+00 3.74274313e-01 -5.69585979e-01 -7.54055023e-01
-2.57208377e-01 5.45655489e-01 -4.39524829e-01 -5.21523580e-02
4.11698073e-01 -2.49016900e-02 3.05306047e-01 -2.27534890e-01
9.36847106e-02 1.09706461e+00 6.51921570e-01 -4.40309763e-01
-7.96483606e-02 5.54727316e-01 1.95940565e-02 -1.07014167e+00
-7.04867244e-01 -6.12901747e-01 -4.02724415e-01 -7.35694349e-01
8.37124944e-01 -1.15113819e+00 -1.28573060e+00 6.86046779e-01
-1.07947683e+00 2.11607769e-01 1.80462562e-02 3.05245250e-01
-6.44300044e-01 1.96155608e-01 -4.44055617e-01 -1.18484318e+00
-4.96080399e-01 -9.83853459e-01 1.35838699e+00 2.21815333e-01
-7.00133741e-01 -3.68316263e-01 3.03394169e-01 5.61997473e-01
2.29911014e-01 5.64822376e-01 4.63857174e-01 -5.81048012e-01
4.06645566e-01 -2.33384222e-01 -1.19262792e-01 5.25807500e-01
1.78598806e-01 1.73317462e-01 -1.42130780e+00 9.52466205e-02
-2.10123494e-01 -1.08410394e+00 7.76184201e-01 1.14553154e-01
1.44754350e+00 -1.20093681e-01 1.27080485e-01 5.25852561e-01
7.50048518e-01 -4.00603423e-03 6.52087510e-01 -3.06262821e-01
3.32393050e-01 9.29905117e-01 7.60035098e-01 1.20103633e+00
2.97788799e-01 7.94361830e-01 3.70539874e-01 -6.90487474e-02
3.76409411e-01 1.18728183e-01 4.49572623e-01 5.08247674e-01
-1.27779573e-01 1.68323934e-01 -6.11020029e-01 3.86172444e-01
-1.71598959e+00 -1.26525903e+00 3.02701652e-01 1.61995447e+00
9.49326694e-01 -5.79936802e-01 4.17179763e-01 4.15970564e-01
5.90797484e-01 2.75398195e-01 -2.87740350e-01 -5.98810673e-01
-2.68222362e-01 3.94833148e-01 -4.03274417e-01 1.45092487e-01
-1.24821663e+00 5.09995759e-01 5.83007002e+00 1.02283871e+00
-1.35322106e+00 2.98976541e-01 7.15648770e-01 -3.06035340e-01
1.72626488e-02 -6.64825618e-01 -4.29972172e-01 4.07532096e-01
9.95819330e-01 4.41321693e-02 3.56799304e-01 6.16377771e-01
3.48806083e-01 -3.03453267e-01 -6.86718941e-01 1.57777715e+00
4.52278554e-01 -7.71162152e-01 -2.94153780e-01 -2.58589327e-01
1.74145594e-01 -1.54662549e-01 1.80000275e-01 3.04489583e-01
-2.71523476e-01 -1.01993394e+00 4.27412987e-01 6.31692290e-01
8.78864050e-01 -1.11294949e+00 5.42873204e-01 7.91342705e-02
-1.10257816e+00 -4.09781858e-02 3.56610060e-01 1.92986026e-01
1.05868787e-01 5.12823999e-01 -6.71071053e-01 5.36104858e-01
9.73309755e-01 6.43555164e-01 -3.36883336e-01 5.04523993e-01
9.12848786e-02 4.07699525e-01 -3.24738801e-01 2.58638293e-01
-2.95948446e-01 -7.43124112e-02 2.64449060e-01 1.20174158e+00
2.75610894e-01 1.84937537e-01 -1.52421191e-01 5.22770286e-01
-1.78140789e-01 4.10185307e-01 -4.89933163e-01 -1.63681597e-01
1.23443417e-01 1.55685055e+00 -1.92212984e-01 -1.58322584e-02
-3.63722295e-01 1.14652765e+00 1.09439790e-01 1.74640596e-01
-9.53691065e-01 -2.51690060e-01 1.13962555e+00 -4.11609977e-01
-6.52330071e-02 3.89570177e-01 4.28671420e-01 -1.01764548e+00
-5.03071621e-02 -9.87735569e-01 5.59367180e-01 -8.44529867e-01
-1.06917250e+00 6.81303263e-01 -2.34732345e-01 -1.17439866e+00
-7.03784943e-01 -3.38461310e-01 -5.01231313e-01 7.02483058e-01
-1.35715353e+00 -1.17995465e+00 -4.58280653e-01 1.06160700e+00
2.09732965e-01 -1.36668652e-01 1.12921953e+00 7.36712515e-01
-7.19743967e-01 7.83607244e-01 -4.27419692e-01 2.06301019e-01
1.00260746e+00 -5.13339937e-01 -5.07139444e-01 2.69583851e-01
8.95544887e-03 2.01975003e-01 6.47450566e-01 -1.22447230e-01
-1.26400805e+00 -1.13416922e+00 6.98587477e-01 3.83282825e-02
4.02780443e-01 -1.95872158e-01 -5.93200147e-01 3.61065328e-01
1.02315672e-01 3.47090513e-01 1.42884970e+00 1.32804051e-01
-5.14970899e-01 -3.26967895e-01 -1.31565726e+00 1.26203671e-01
4.17294443e-01 -9.09558296e-01 -1.16002157e-01 -2.37397656e-01
2.40582809e-01 -1.36104703e-01 -8.72651398e-01 7.54968166e-01
9.48941708e-01 -9.93120432e-01 8.55408311e-01 -7.55609810e-01
3.48463237e-01 -1.85359612e-01 -5.22854626e-01 -1.29641116e+00
2.59839237e-01 -6.33482873e-01 1.24791406e-01 1.62309206e+00
1.10156713e-02 -3.18280548e-01 5.72780192e-01 4.24482673e-01
3.90855163e-01 -6.38055325e-01 -1.21046591e+00 -1.54997543e-01
-8.09413314e-01 -7.24398792e-01 5.42776406e-01 1.06869268e+00
2.65131861e-01 3.15630198e-01 -6.79766536e-01 -2.01761544e-01
2.06225589e-01 4.84599583e-02 6.76640809e-01 -1.26072049e+00
-1.93601236e-01 -3.73759925e-01 -8.00668418e-01 -1.13729350e-01
7.57303596e-01 -5.82948148e-01 -2.53728181e-02 -6.37320578e-01
3.05017769e-01 2.72774816e-01 -4.34658349e-01 6.56420887e-01
4.03365456e-02 7.17224061e-01 1.57161802e-01 -4.04945880e-01
-8.30147326e-01 9.17032659e-01 4.12496775e-01 -2.61189759e-01
-5.03659770e-02 2.22159345e-02 -6.59030259e-01 8.92473757e-01
5.67124307e-01 -1.04120523e-01 -1.16343848e-01 1.85134515e-01
2.52169967e-01 3.40502828e-01 5.04455328e-01 -6.49414837e-01
-3.14844474e-02 -1.22233750e-02 4.60044056e-01 -6.06406748e-01
1.01389182e+00 -7.35855877e-01 1.20488644e-01 8.65402352e-03
-2.32959956e-01 1.69880703e-01 4.13043171e-01 2.96173215e-01
-7.81582117e-01 1.26498356e-01 7.95756221e-01 3.40422422e-01
-7.39051282e-01 3.04397523e-01 -5.21341920e-01 -1.02112472e-01
9.85897481e-01 2.03855485e-01 -1.24215811e-01 -6.37916923e-01
-1.08989692e+00 2.65123993e-01 -2.28996836e-02 4.00996953e-01
8.54082584e-01 -1.65981328e+00 -7.80457914e-01 5.29320419e-01
5.90871394e-01 -6.53576195e-01 7.98829198e-01 1.35225630e+00
2.07185939e-01 -2.99507409e-01 -3.24553579e-01 -6.31018281e-01
-1.99516666e+00 1.67991415e-01 2.85377771e-01 2.47612726e-02
2.74859726e-01 9.05125499e-01 -6.47985190e-02 -1.80521965e-01
3.17033470e-01 1.87335491e-01 -4.74385083e-01 8.97355020e-01
9.98202384e-01 1.54198587e-01 3.14184397e-01 -1.10650229e+00
-5.74111879e-01 4.79496211e-01 2.59228021e-01 -2.75019795e-01
1.30101514e+00 -2.58224994e-01 -4.40934360e-01 7.17882872e-01
1.43838096e+00 -3.95858549e-02 -6.27055168e-01 -2.64969885e-01
-1.70453250e-01 -3.05486441e-01 6.56406507e-02 -5.41178048e-01
-1.05554831e+00 1.05301881e+00 7.88474262e-01 6.01799451e-02
1.87412202e+00 5.02897501e-02 5.90555727e-01 1.40801117e-01
4.07495275e-02 -1.29828191e+00 1.60733819e-01 1.75440639e-01
1.01437962e+00 -1.32824838e+00 -3.03492457e-01 -6.62718043e-02
-1.05711114e+00 9.68152165e-01 4.15742815e-01 4.38713133e-01
9.04662549e-01 4.16669220e-01 2.67557144e-01 -1.82048321e-01
-9.47297394e-01 -3.29481632e-01 2.12373942e-01 4.72886264e-01
3.61023754e-01 -6.80421293e-02 8.02234635e-02 1.22616172e+00
6.19922914e-02 2.59391546e-01 -9.62730870e-02 7.19458103e-01
-1.81449071e-01 -8.91084731e-01 -4.44387585e-01 3.09560597e-01
-7.77057350e-01 1.79096073e-01 -7.14858532e-01 2.10493833e-01
1.89505652e-01 1.06895232e+00 1.56484414e-02 -6.81250811e-01
3.39046210e-01 5.73717892e-01 4.54713196e-01 -7.38902241e-02
-6.95921898e-01 2.36664429e-01 1.14534505e-01 -1.05695057e+00
-9.89190340e-01 -8.81488740e-01 -1.09318483e+00 -7.40925223e-02
8.41852203e-02 -1.15193399e-02 6.63381398e-01 9.52868521e-01
5.69078326e-01 3.14456910e-01 1.08948433e+00 -9.20892179e-01
-8.83110911e-02 -9.62175071e-01 -8.02873075e-01 5.72584271e-01
7.34094799e-01 -6.91429436e-01 -3.74534607e-01 1.24447942e-01] | [13.54776382446289, 2.127455472946167] |
c5a9657a-be9d-4454-8ae6-1212455ba033 | text-data-augmentation-made-simple-by | 1812.04718 | null | https://arxiv.org/abs/1812.04718v1 | https://arxiv.org/pdf/1812.04718v1.pdf | Text Data Augmentation Made Simple By Leveraging NLP Cloud APIs | In practice, it is common to find oneself with far too little text data to train a deep neural network. This "Big Data Wall" represents a challenge for minority language communities on the Internet, organizations, laboratories and companies that compete the GAFAM (Google, Amazon, Facebook, Apple, Microsoft). While most of the research effort in text data augmentation aims on the long-term goal of finding end-to-end learning solutions, which is equivalent to "using neural networks to feed neural networks", this engineering work focuses on the use of practical, robust, scalable and easy-to-implement data augmentation pre-processing techniques similar to those that are successful in computer vision. Several text augmentation techniques have been experimented. Some existing ones have been tested for comparison purposes such as noise injection or the use of regular expressions. Others are modified or improved techniques like lexical replacement. Finally more innovative ones, such as the generation of paraphrases using back-translation or by the transformation of syntactic trees, are based on robust, scalable, and easy-to-use NLP Cloud APIs. All the text augmentation techniques studied, with an amplification factor of only 5, increased the accuracy of the results in a range of 4.3% to 21.6%, with significant statistical fluctuations, on a standardized task of text polarity prediction. Some standard deep neural network architectures were tested: the multilayer perceptron (MLP), the long short-term memory recurrent network (LSTM) and the bidirectional LSTM (biLSTM). Classical XGBoost algorithm has been tested with up to 2.5% improvements. | ['Claude Coulombe'] | 2018-12-05 | null | null | null | null | ['text-augmentation'] | ['natural-language-processing'] | [ 3.68719071e-01 3.32054406e-01 -2.27765605e-01 -2.88803279e-01
-2.85329819e-01 -1.03002638e-01 8.71729136e-01 3.79392743e-01
-6.69109821e-01 9.19837773e-01 3.46904099e-01 -5.53682506e-01
1.45019457e-01 -7.90260434e-01 -5.42875707e-01 -5.76482475e-01
4.80136037e-01 6.89039767e-01 -1.45141318e-01 -6.88047290e-01
3.55943263e-01 2.99692780e-01 -1.54099715e+00 4.62612808e-01
8.69933307e-01 7.89700270e-01 -3.90544757e-02 4.69244480e-01
-8.02305996e-01 7.99560905e-01 -6.27334952e-01 -4.68761325e-01
2.08517835e-01 -1.76037520e-01 -9.38652277e-01 -4.33574617e-01
1.68989286e-01 1.50676951e-01 1.92826599e-01 9.56901729e-01
6.38318896e-01 -9.60665494e-02 3.06683332e-01 -9.56083834e-01
-6.88625336e-01 9.43418384e-01 -4.99955982e-01 2.52424747e-01
3.99032593e-01 -1.29575536e-01 5.58112442e-01 -6.77481592e-01
6.37750566e-01 9.41348195e-01 8.50754023e-01 3.59946728e-01
-1.06578887e+00 -5.55228412e-01 -1.41039044e-01 4.21700001e-01
-8.97623718e-01 -1.44174248e-01 8.14428031e-01 -3.26332986e-01
1.44654930e+00 3.36867273e-01 5.82901478e-01 1.35922301e+00
2.63488263e-01 3.70731354e-01 1.50485480e+00 -9.88986790e-01
3.96490768e-02 7.75283277e-01 4.41904008e-01 2.31090114e-01
2.19112754e-01 -1.71364620e-01 -4.65642989e-01 -2.80214883e-02
3.71335857e-02 -7.37338290e-02 1.06997721e-01 7.70543441e-02
-9.09433246e-01 9.91082728e-01 3.90992224e-01 1.03598845e+00
-6.42077923e-01 -1.87378511e-01 7.69206464e-01 5.74534714e-01
8.74492109e-01 5.58480561e-01 -7.42750406e-01 -2.03261048e-01
-8.30468118e-01 8.69597048e-02 8.40013623e-01 3.46131891e-01
5.86737096e-01 3.89626563e-01 -8.62014741e-02 8.84459496e-01
1.60768941e-01 2.39215538e-01 1.31169081e+00 -1.02485038e-01
6.46429300e-01 1.02869713e+00 -2.63274342e-01 -1.02245426e+00
-7.16261744e-01 -5.12059212e-01 -9.37739551e-01 1.88431099e-01
3.04795235e-01 -5.13446510e-01 -1.10979176e+00 1.48241973e+00
1.83775842e-01 -1.27405703e-01 9.57854092e-02 4.75472182e-01
6.86301172e-01 7.47482061e-01 1.80614471e-01 -2.18787044e-01
1.22928607e+00 -9.68638539e-01 -8.14170361e-01 -4.16464716e-01
8.62340510e-01 -9.58840489e-01 1.22450900e+00 4.26711529e-01
-9.33137774e-01 -5.70058584e-01 -1.05720496e+00 9.23840702e-03
-1.16307712e+00 8.88188742e-03 6.16835296e-01 1.08795238e+00
-1.08196187e+00 8.32477510e-01 -4.80444402e-01 -6.00076079e-01
1.25120804e-01 6.78385377e-01 -5.18513381e-01 3.35278958e-01
-1.23980582e+00 1.23460412e+00 7.53346264e-01 1.58530146e-01
-5.67892343e-02 -4.91542995e-01 -4.34330374e-01 -6.88397978e-03
1.26793474e-01 -6.25036299e-01 9.07312214e-01 -1.63456333e+00
-1.79973674e+00 9.96863425e-01 4.62688059e-02 -9.14302528e-01
3.40466619e-01 -4.00795668e-01 -4.30092812e-01 -4.09370273e-01
-1.68688565e-01 5.83711684e-01 6.03882015e-01 -6.95986807e-01
-4.42969143e-01 -5.43419302e-01 -2.55448222e-01 8.27146471e-02
-8.67755115e-01 3.37108374e-01 2.71033704e-01 -7.86391497e-01
4.42530848e-02 -9.81362045e-01 -2.13859558e-01 -7.72461236e-01
-4.02586490e-01 -5.10297194e-02 9.25152421e-01 -1.10911262e+00
1.01694345e+00 -1.72349954e+00 9.64111183e-03 2.97992855e-01
-2.73573488e-01 6.19638920e-01 2.04356804e-01 4.96698916e-01
-4.27773446e-01 3.35039645e-01 -7.52381235e-02 -3.94706696e-01
-1.95490435e-01 1.81633562e-01 -2.97312170e-01 1.65037334e-01
2.20170654e-02 7.72607863e-01 -3.41239125e-01 -1.46515459e-01
3.19770545e-01 5.67811370e-01 -3.49657625e-01 -2.06557110e-01
-1.60079837e-01 2.95076430e-01 -1.29696410e-02 3.77470583e-01
4.39782560e-01 8.32391679e-02 9.57590640e-02 5.68725169e-02
-3.00876409e-01 4.39830810e-01 -1.05043459e+00 1.36320877e+00
-6.01044357e-01 8.00270200e-01 -2.62969553e-01 -1.22485745e+00
1.19489288e+00 4.49431896e-01 2.97876865e-01 -7.99126208e-01
4.02346820e-01 5.52647889e-01 1.43179417e-01 -4.22037303e-01
5.74120402e-01 -1.89979762e-01 2.85771072e-01 2.85188109e-01
2.08288133e-01 2.18432501e-01 1.68398082e-01 -3.51332605e-01
7.44472921e-01 1.34441003e-01 2.99737781e-01 -9.32740346e-02
7.61717439e-01 8.41261297e-02 2.13968486e-01 4.77243692e-01
2.35034347e-01 4.16726768e-01 4.55867290e-01 -6.01523042e-01
-1.31187344e+00 -1.90520853e-01 6.74826577e-02 1.18847251e+00
-6.35246456e-01 -3.43600959e-01 -8.02268863e-01 -3.83424908e-01
-4.90286738e-01 7.95226574e-01 -4.74662930e-01 -1.36450514e-01
-6.18843019e-01 -1.05929697e+00 4.20562267e-01 3.43156964e-01
7.11010873e-01 -1.62972271e+00 -3.18667322e-01 3.35758537e-01
7.18598999e-03 -1.00028145e+00 5.07166147e-01 5.51892459e-01
-1.17185807e+00 -4.02015954e-01 -5.37339032e-01 -7.33839750e-01
3.16877961e-01 -2.93941289e-01 9.11930203e-01 -5.24119623e-02
-8.29784647e-02 -3.16327900e-01 -6.13344073e-01 -8.08166146e-01
-5.90736449e-01 6.12656236e-01 -8.98081139e-02 -1.08220559e-02
5.96152544e-01 -7.99721837e-01 -1.30692974e-01 -1.37430310e-01
-5.89358985e-01 3.59160230e-02 8.70773852e-01 9.13192511e-01
3.66754651e-01 -3.60179186e-01 7.74403274e-01 -1.24842072e+00
9.70346451e-01 -2.56552309e-01 -3.65619063e-01 -5.11567220e-02
-9.99326169e-01 6.05984554e-02 8.37134004e-01 -3.26014131e-01
-9.20488358e-01 2.34756060e-02 -6.98607862e-01 -4.32079285e-02
-2.70459622e-01 6.36229277e-01 6.12113960e-02 -1.64387509e-01
9.03511465e-01 1.01017639e-01 4.57790457e-02 -5.96944928e-01
3.51655632e-01 9.30022240e-01 1.46713406e-01 4.70329151e-02
5.51775455e-01 1.28259718e-01 -1.94971800e-01 -8.30188155e-01
-5.23014247e-01 -2.85894126e-01 -4.44560945e-01 7.93311223e-02
8.73484015e-01 -5.10966837e-01 -3.72308701e-01 6.12892628e-01
-1.09657311e+00 -1.79718807e-01 -4.13366705e-01 4.07397956e-01
-1.57601863e-01 1.99588478e-01 -6.49105072e-01 -6.30116224e-01
-9.91646409e-01 -8.98287058e-01 5.97665787e-01 4.10302490e-01
-3.31190020e-01 -9.28325713e-01 6.97683766e-02 6.50635123e-01
8.27360868e-01 1.71304315e-01 1.11867452e+00 -1.29471445e+00
9.38898474e-02 -4.94639963e-01 4.71273810e-02 7.24586725e-01
-1.02112986e-01 -4.79483824e-05 -1.23033381e+00 -4.49079871e-02
2.73338377e-01 -2.24193186e-01 6.19583666e-01 3.87650013e-01
9.18360710e-01 -3.67959797e-01 -2.43253469e-01 2.24048808e-01
1.25384760e+00 3.44798058e-01 7.84199119e-01 7.70130396e-01
5.44224143e-01 6.36121631e-01 2.79874057e-01 3.99292521e-02
-3.30004259e-03 6.51645184e-01 2.38638923e-01 -1.43582121e-01
9.28362831e-02 9.25708190e-03 3.09124887e-01 8.99943948e-01
-2.27512881e-01 5.63222840e-02 -1.10131681e+00 2.49431476e-01
-1.69464469e+00 -8.37727129e-01 -3.62714142e-01 2.31901860e+00
6.28539085e-01 7.14997172e-01 -1.19974660e-02 7.67070830e-01
6.15351617e-01 -2.63065435e-02 -9.25270095e-03 -1.17728317e+00
-2.47102350e-01 6.49398506e-01 5.09429455e-01 3.77692729e-01
-1.12449992e+00 9.91080821e-01 5.38436174e+00 7.35824049e-01
-1.72290075e+00 4.88388121e-01 7.33382225e-01 6.26854151e-02
6.74748644e-02 4.20805551e-02 -8.38035643e-01 5.77711463e-01
1.42571008e+00 1.09025449e-01 2.81238377e-01 9.92502868e-01
2.29625225e-01 6.71688244e-02 -4.24443990e-01 7.84552813e-01
2.31402576e-01 -1.44162238e+00 -7.71539882e-02 -1.07313972e-02
6.94561720e-01 3.65585327e-01 -1.18033558e-01 6.53041720e-01
-2.76082933e-01 -1.00940180e+00 2.64935851e-01 3.50666136e-01
1.56918570e-01 -7.68575609e-01 1.29859412e+00 3.30293745e-01
-5.17286360e-01 -4.39149112e-01 -3.03947806e-01 -4.48918343e-01
3.28777991e-02 5.14578998e-01 -1.12169206e+00 4.57599878e-01
7.19060481e-01 2.22427860e-01 -9.06443536e-01 8.49192858e-01
-1.76400632e-01 6.70785308e-01 -5.44507802e-01 -4.45531815e-01
3.70566219e-01 -2.95634806e-01 4.35315043e-01 1.10226595e+00
2.01592803e-01 -4.76889551e-01 -1.64739400e-01 4.83719289e-01
-1.39591351e-01 8.52344751e-01 -7.34222293e-01 -2.40364149e-01
4.21484448e-02 1.57945395e+00 -7.80339956e-01 -3.50418836e-01
-2.85941154e-01 7.71245599e-01 2.23582871e-02 -3.96703146e-02
-5.76027215e-01 -4.95721251e-01 6.41198084e-02 3.18744570e-01
-2.15548906e-03 1.73526168e-01 -6.17777765e-01 -7.51341462e-01
2.30964705e-01 -1.18479514e+00 1.67122766e-01 -8.16243470e-01
-1.06741703e+00 9.04400289e-01 -2.32186228e-01 -8.37949514e-01
-7.01780200e-01 -8.51692617e-01 -8.11922848e-01 9.87819076e-01
-1.14172840e+00 -1.42680657e+00 -2.75502950e-01 5.30653834e-01
5.16995132e-01 -5.73581517e-01 1.06862044e+00 6.02405727e-01
-5.30897856e-01 4.12701905e-01 1.58581123e-01 -4.39186506e-02
5.16389966e-01 -1.20665705e+00 3.98119926e-01 6.29917502e-01
1.82233989e-01 3.54113698e-01 8.24705899e-01 -6.10323310e-01
-1.01838911e+00 -7.93527305e-01 1.25206208e+00 -5.46073616e-02
6.67138517e-01 -5.71611524e-01 -1.07208979e+00 6.98739886e-01
5.94189525e-01 -2.96329498e-01 5.83309174e-01 3.41135114e-01
5.53688146e-02 -2.09959731e-01 -1.10577929e+00 6.55278742e-01
4.56129730e-01 -2.82896817e-01 -7.04475999e-01 5.70800424e-01
6.85910165e-01 -1.49747103e-01 -5.70053041e-01 3.81464064e-01
2.32155576e-01 -1.07636726e+00 7.51217723e-01 -7.34142601e-01
3.66692960e-01 1.24980472e-01 9.36238244e-02 -1.22204387e+00
2.51906037e-01 -5.92546999e-01 1.42610043e-01 1.46291077e+00
8.05436790e-01 -8.97607446e-01 1.13410091e+00 3.70250583e-01
-1.04013830e-01 -8.79199743e-01 -9.89515305e-01 -3.19286287e-01
1.15752004e-01 -4.73997027e-01 1.84865683e-01 1.03339767e+00
4.70101126e-02 7.52320170e-01 -4.35142577e-01 -4.88393992e-01
-1.20100506e-01 -1.89422280e-01 6.93864465e-01 -1.38155878e+00
-2.53181666e-01 -5.76572359e-01 -6.48086667e-01 -1.65873915e-01
2.12288812e-01 -8.77940059e-01 -5.95506608e-01 -1.50699127e+00
-2.62273431e-01 -2.05921099e-01 -1.34214103e-01 5.96767604e-01
1.12067737e-01 3.33660603e-01 2.53990859e-01 -1.93759978e-01
1.87383816e-01 3.09317857e-01 5.34390450e-01 -1.17809050e-01
-6.24526680e-01 1.67006254e-01 -7.06580937e-01 9.38912213e-01
1.13759315e+00 -5.37487507e-01 -1.70291558e-01 -4.60335501e-02
6.62089288e-01 -3.85513067e-01 -5.52565046e-03 -1.14277303e+00
1.20594665e-01 4.08468634e-01 4.44275737e-01 -5.21151125e-01
3.61658365e-01 -7.32998192e-01 6.82944432e-02 5.89674354e-01
-1.71620086e-01 3.93890768e-01 4.40778553e-01 -5.60669508e-03
-3.41064036e-01 -6.66157603e-01 6.07575893e-01 -1.59830898e-01
-5.09322822e-01 -4.33507502e-01 -5.62153459e-01 -3.23982656e-01
9.32597041e-01 -2.62612104e-01 -3.29133421e-01 -3.55803102e-01
-7.42600024e-01 -2.67277002e-01 -4.71947454e-02 5.57844460e-01
1.31555021e-01 -9.31495965e-01 -5.56518734e-01 1.19752422e-01
-2.28469104e-01 -3.59769434e-01 4.45014052e-02 1.02328122e+00
-6.02358639e-01 6.28014922e-01 -5.60461104e-01 -3.32206875e-01
-1.45371878e+00 5.16462088e-01 4.15586829e-01 -6.93154156e-01
-4.45553362e-01 5.89136362e-01 -7.26493001e-01 -6.79819703e-01
7.78392330e-02 -4.81549919e-01 -7.64529824e-01 3.44279557e-01
3.21500927e-01 3.27645630e-01 6.65976286e-01 -3.49349886e-01
-9.03901085e-02 4.24894571e-01 -1.70065016e-01 -1.53743654e-01
1.55438960e+00 3.27003211e-01 -3.63136172e-01 5.76234758e-01
9.74807203e-01 1.85142215e-02 3.73341190e-03 6.23093247e-02
3.82687807e-01 4.96955998e-02 2.00506270e-01 -1.11936522e+00
-8.44441831e-01 1.05218744e+00 1.04292285e+00 4.93895352e-01
1.10738492e+00 -5.33487141e-01 5.11705399e-01 6.68559730e-01
4.96049179e-03 -1.28509712e+00 -1.49034023e-01 6.07049406e-01
8.67869616e-01 -1.20326066e+00 -1.46591306e-01 8.66268054e-02
-5.64069748e-01 1.20362771e+00 6.07458353e-01 -6.73129633e-02
4.62202966e-01 3.48790616e-01 5.06056100e-02 1.58985872e-02
-5.69470882e-01 4.62122727e-03 1.36319140e-03 3.81409109e-01
7.84295321e-01 -1.45808324e-01 -8.55003536e-01 4.72595692e-01
-5.13021588e-01 3.11425477e-01 5.10269284e-01 8.03840876e-01
-3.13526630e-01 -1.47988832e+00 -4.60285246e-01 7.86887348e-01
-7.95136631e-01 -3.15785229e-01 -5.25704384e-01 9.75376070e-01
2.83952147e-01 7.67838776e-01 -5.86111620e-02 -3.81703883e-01
3.94191265e-01 6.98000371e-01 2.54292861e-02 -5.38210928e-01
-1.38431752e+00 -5.75530380e-02 3.49750489e-01 -5.99849597e-02
-4.72900152e-01 -5.11133492e-01 -1.08160233e+00 -4.01085526e-01
-3.76412988e-01 1.17055466e-02 1.44978476e+00 1.17184079e+00
2.71125615e-01 4.51774120e-01 2.99896866e-01 -7.78937101e-01
-3.39728713e-01 -1.66542768e+00 -1.48790911e-01 3.44259977e-01
-2.25156710e-01 -3.56745064e-01 -1.46542311e-01 -1.82659790e-01] | [10.967259407043457, 7.276968479156494] |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.